The circadian visual system, 2005 · 2006-10-19 · Review The circadian visual system, 2005 L.P....

B R A I N R E S E A R C H R E V I E W S 5 1 ( 2 0 0 6 ) 1 – 6 0

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ loca te /b ra in res rev

Review

The circadian visual system, 2005

L.P. Morina,⁎, C.N. Allenb

aDepartment of Psychiatry and Graduate Program in Neuroscience, Stony Brook University, Stony Brook, NY 11794, USAbDepartment of Physiology and Pharmacology and Centre for Research on Occupational and Environmental Toxicology,Oregon Health and Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA

A R T I C L E I N F O

⁎ Corresponding author. Fax: +1 631 444 7534.E-mail address: lawrence.morin@stonybr

0165-0173/$ – see front matter © 2005 Elsevidoi:10.1016/j.brainresrev.2005.08.003

A B S T R A C T

Article history:Accepted 9 August 2005Available online 5 December 2005

The primary mammalian circadian clock resides in the suprachiasmatic nucleus (SCN), arecipient of dense retinohypothalamic innervation. In its most basic form, the circadianrhythm system is part of the greater visual system. A secondary component of the circadianvisual system is the retinorecipient intergeniculate leaflet (IGL) which has connections tomany parts of the brain, including efferents converging on targets of the SCN. The IGL alsoprovides a major input to the SCN, with a third major SCN afferent projection arriving fromthe median raphe nucleus. The last decade has seen a blossoming of research into theanatomy and function of the visual, geniculohypothalamic and midbrain serotonergicsystems modulating circadian rhythmicity in a variety of species. There has also been asubstantial and simultaneous elaboration of knowledge about the intrinsic structure of theSCN. Many of the developments have been driven by molecular biological investigation ofthe circadian clock and the molecular tools are enabling novel understanding of regionalfunction within the SCN. The present discussion is an extension of the material covered bythe 1994 review, “The Circadian Visual System.”

© 2005 Elsevier B.V. All rights reserved.

Keywords:CircadianSuprachiasmaticSCNIntergeniculate leafletIGL

ook.edu (L.P. Morin).

er B.V. All rights reserved.

mailto:lawrence. morin@stonybroo k.edu

http://dx.doi.org/10.1016/j.brainresre v.2005.08.003

2 B R A I N R E S E A R C H R E V I E W S 5 1 ( 2 0 0 6 ) 1 – 6 0

Abbreviations:5-HT, serotoninCALB, calbindinCALR, calretininCCK, cholecystokininENK, enkephalinGABA, gamma amino butyric acidGRP, gastrin releasing hormoneIGL, intergeniculate leafletNPY, neuropeptide YNT, neurotensinPACAP, pituitary adenylatecyclase-activating polypeptideSCN, suprachiasmatic nucleusSP, substance PSS, somatostatinVIP, vasoactive intestinalpolypeptideVP, vasopressin

Contents

1. Introduction and caveats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32. Psychophysics of the circadian rhythm system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43. Organization of the suprachiasmatic nucleus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

3.1. Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53.2. Distribution of cell phenotypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53.3. Intra-SCN connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

4. Retinal photoreceptors and projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84.1. Melanopsin ganglion cell characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84.2. Retinal ganglion cell projections and bifurcation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94.3. Classical and ganglion cell photoreceptor contribution to regulation of circadian rhythms, masking, the pupillary

light reflex and melatonin suppression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114.3.1. Circadian rhythms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114.3.2. Masking, pupillary light reflex and melatonin suppression. . . . . . . . . . . . . . . . . . . . . . . . . . 11

5. Neuromodulators of the retinohypothalamic tract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

5.1. Glutamate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125.2. N-Acetylaspartyl glutamate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125.3. Glutamate receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135.4. Nitric oxide and RHT signal transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135.5. PACAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145.6. Substance P. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

6. SCN structure–function relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176.1. Function of cellular neuromodulators in the SCN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

6.1.1. GABA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186.1.2. VIP and VIP/PACAP receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186.1.3. Calbindin and the central subnucleus of the hamster SCN . . . . . . . . . . . . . . . . . . . . . . . . 196.1.4. Bombesin and gastrin releasing peptide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216.1.5. Neuromedin U and neuromedin S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216.1.6. Vasopressin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216.1.7. Diurnality–nocturnality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

6.2. Acetylcholine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226.3. Neurophysiology of the SCN action potential rhythm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236.4. Inter-SCN differences in gene expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256.5. Regional expression of Fos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

6.5.1. Hamster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256.5.2. Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

6.6. Regional clock gene expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3B R A I N R E S E A R C H R E V I E W S 5 1 ( 2 0 0 6 ) 1 – 6 0

6.6.1. Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266.6.2. Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276.6.3. Hamster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

6.7. Stimulus induction of gene expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286.7.1. Photic stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286.7.2. Nonphotic stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286.7.3. Inconsistencies between phase response and gene expression . . . . . . . . . . . . . . . . . . . . . . 29

6.8. Polysialic acid and SCN function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307. SCN connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

7.1. Humoral factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

7.2. Efferent anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317.3. Afferent anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

8. Intergeniculate leaflet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348.1. Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348.2. Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348.3. Anatomy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

8.3.1. Geniculohypothalamic tract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348.3.2. Connectivity and relationship to sleep, visuomotor and vestibular systems . . . . . . . . . . . . . . . 34

8.4. Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358.4.1. Neurophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358.4.2. Nonphotic regulation of rhythmicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368.4.3. Photic regulation of rhythmicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

9. Serotonin and midbrain raphe contribution to circadian rhythm regulation . . . . . . . . . . . . . . . . . . . . . . . 389.1. Effects of serotonin-specific neurotoxins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389.2. Serotonin receptor function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389.3. Regulation and consequences of serotonin release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399.4. Serotonergic potentiation of light-induced phase shifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

10. Failure to re-entrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4111. Masking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4212. The future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

1. Introduction and caveats

There have been a large number of new developments withrespect to knowledge about the anatomy and physiology ofcircadian rhythm regulation since publication of our 1994review, “The Circadian Visual System” (Morin, 1994). As isnearly always the case with such projects, many of thedevelopments were well underway, and in some instancescompleted, prior to the actual publication of that review. Mostnotably, much of the presentation in the 1994 reviewconcerning the serotonergic system was incomplete orwrong and understanding of the issues changed substantiallysoon after publication of the review. Other issues emerged asunanticipatedmajor developments. One of these has been thediscovery of the photopigment, melanopsin, in retinal gangli-on cells and the subsequent establishment of those ganglioncells as a special class of photoreceptors.

The steady growth of molecular biological contributionsto the understanding of circadian clock mechanisms ininvertebrates has been translated into parallel evaluation ofclock mechanisms in mammals. This development hascreated new avenues for architectural analysis of thesuprachiasmatic nucleus (SCN) function according to theregional distributions of cell phenotypes comprising thecircadian clock.

The purpose of the current review is to continue thetheme of the original, but to do so with minimal redundan-cy. The focus of the current presentation will be onnecessary modifications to the perspective presented 10years ago, and to the substantial additions to the generalbody of knowledge relating to the circadian visual system.The title of the 1994 review, “The Circadian Visual System,”remains of some import. We believe it is worth emphasizingthat, taken as a whole, the research topic under consider-ation is “circadian,” is “visual” and certainly is a “system.”That is an aspect of the research topic of importance thatshould not be underestimated.

We do not intend to cover all facets of circadian rhythmresearch and we also expect to have inadvertently overlookedresearch worthy of inclusion here. To the extent that hashappened, we apologize. On a technical note, when specificcell types are mentioned as containing a particular neuromo-dulator, the comment refers to the fact that it has beendemonstrated by immunohistochemical means.

The work that is particularly ignored concerns the molec-ular biology of the SCN or other tissues. There will be no effortto review the burgeoning literature concerning the molecularmechanisms of the central circadian clock or of molecularoscillations in peripheral structures. Rather, we will touch onthis literaturewhen it can be used to amplify understanding of


the system properties governing circadian rhythm regulation.One issue at the forefront of the list of mysteries presentlyconfronting that understanding is the spatial distribution ofSCN neurons exhibiting clock-like behavior. The generalquestion is, “Are all SCN cells pacemakers?” While someexperimental work suggests that the answer is no, the topic isan area of intense research and is discussed in some detailbelow.

A second question is, “Do all pacemaker cells in the SCNoscillate with the same phase?” Evaluation of a variety ofmolecular oscillations in SCN cells has given the impressionthat the molecular behavior of a whole SCN reflects thebehavior of the individual cells. While this may be true onaverage, many individual SCN neurons do not expressrhythmicity in phase with the average of the whole (Quinteroet al., 2000). This development in the analytical approach topacemaker activity among groups of cells emphasizes theimportance of molecular activity within individual cells andpoints to the need to evaluate spatial relationships andconnectivity among oscillatory neurons within the SCN.

When microelectrode arrays are used for long-termrecordings, the action potential firing rhythm periods ofindividual of SCN neurons show a greater diversity indispersed cultures compared to organotypic cultures. Thesedata suggest that interneuronal communication is required tofine tune the period of the clock (Nakamura et al., 2002; Herzoget al., 1998; Liu et al., 1997a; Albus et al., 2002). The finalbehavioral period is an average of these individual neuronalperiods (Liu et al., 1997a). The demonstration of phasedifferences among SCN neurons is a variant on the largertheme indicating differences in phase between the two SCN ina single animal (De la Iglesia et al., 2000). Thus, from the pointof view of this review, utility of the molecular analysis ofcellular clocks serves to highlight the need to understand thesystem properties intrinsic to the SCN, as well as its regulationof efferent response to afferent information, whether thatinformation be visual or some other sensory modality.

2. Psychophysics of the circadian rhythm system

The Pickard et al. (1987) IGL lesion paper was one of the first toemploy nonsaturating light stimuli to elicit circadian rhythmphase responses. The method eliminated ceiling effectscaused by saturating light, providing a more accurate view ofpsychophysical sensitivity to light, rather than a simplestatement that animals respond (or not) to the stimulus.Subsequently, the psychophysics of circadian rhythm phaseresponse to light presented at CT19 has been explored ingreater detail. The hamster (Nelson and Takahashi, 1991, 1999;Zhang et al., 1996), mouse (van den Pol et al., 1998) and gerbil(Dkhissi-Benyahya et al., 2000) circadian systems estimate theimportance of an acute photic stimulus by effectively “count-ing” photons. It may be more appropriate to refer to thisactivity as “temporal integration”which is a broader term. Theprocess is demonstrated by the fact that there is a clearrelationship between the stimulus duration and irradiance(photons/cm2) with respect to the magnitude of the phaseshift. That is, dim, long duration stimuli have the same effectas brief but bright stimuli, if the irradiances (i.e., total photons

or Joules/cm2) are equivalent. Most impressively, the photicintensity information in a series of light pulses appears to besummed across the series (Nelson and Takahashi, 1991), aslong as the individual pulses are of sufficient duration (i.e.,minutes).

The integrative capacity of the circadian system appears tobe time-limited to the extent that, under test conditionsexplored to date, if a photic stimulus is sufficient to produce amaximal phase shift,more photons at the same time or up to 2h later have no additional effect on phase shift magnitude(Nelson and Takahashi, 1999). In such cases, the circadiansystem is deemed to be “saturated”with respect to its ability toshow a further phase response to light.

An unexpected result (van den Pol et al., 1998), demon-strated the important point that a form of “temporalintegration” may occur even when the light stimulus issubthreshold. For example, in the mouse or hamster, a brief(2–3 ms), bright (1013 photons/cm2/s) flash will not elicit aphase shift (Nelson and Takahashi, 1991; van den Pol et al.,1998), but a series of flashes totaling 120mswill yield a normalphase response (van den Pol et al., 1998). Thus, some form ofsummation has occurred to elicit the phase shifts. It is not yetclear, however, that there is a systematic relationship betweenthe number of flashes and the magnitude of the phase shift.Evidence from studies employing a series of 5 to 45 lightflashes, each only 10 μs long, suggests otherwise (Arvanito-giannis and Amir, 1999). Each such series effectively inducesFOS-IR in the SCN and yields 75 min phase shifts whendelivered to rats at CT13. Thus, the fact of a light-inducedphase shift may be separable from the process of“integration.”

The entire visual system has been intact in all studies todate that involve “temporal integration.” Therefore, there ispresently no information regarding the extent to which theintegration mechanism includes or excludes the retina, SCN,IGL, visual midbrain, the serotonergic or other systems.Unilateral enucleation eliminates 50% of the retinal photo-receptors and their associated RHT projections, but had noeffect on hamster re-entrainment rate (Stephan et al., 1982).On the other hand, the same publication showed that bothadvance and delay shifts of rats were retarded followingunilateral enucleation. D. Earnest (unpublished work cited byPickard and Turek, 1983) failed to find an effect of unilateralenucleation on hamster circadian period in LL. Similarly,unilateral enucleation appears to have no effect on light-induced FOS in rat (Beaulé et al., 2001a) or hamster SCN(Muscat and Morin, unpublished data), or in SCN of hamsterwith one eye occluded, except in response to high irradianceor long duration light (Muscat and Morin, in press). In each ofthese investigations, saturating stimuli were employed andthe experimental methods not sufficiently sensitive to deter-mine whether there is an effect of unilateral enucleation onrate of re-entrainment or any other measure of circadianrhythm regulation (but see Tang et al., 2002 for a broad set ofresults suggesting a large effect of unilateral enucleation).

Unilateral enucleation is also reported to shorten thecircadian period of rats in LL, while a unilateral SCN lesionhas no similar effect (Donaldson and Stephan, 1982). Giventhat the IGL regulates 50% of the hamster circadian periodresponse to LL (Pickard et al., 1987; Morin and Pace, 2002) (no


equivalent studies have been done in the rat), the effect ofenucleation could be attributable exclusively to loss of theretinal input to the IGL.

Light induction of FOS-IR in the gerbil SCN (Dkhissi-Benyahya et al., 2000) has been interpreted as obeying thetemporal integration rule found in the hamster for phaseshifts and melatonin suppression. However, over the sameirradiance range, the effect of light on FOS in ganglion cells ofthe gerbil retina appears to be all-or-none, rather thanadditive. Similar results have been found for light-inducedFOS in hamster IGL cells (Muscat andMorin, unpublished data)and imply that neither the ganglion cell nor IGL neuronproperties are altered by light in a manner consistent with“photon counting” responses of the circadian rhythm system.

3. Organization of the suprachiasmatic nucleus

3.1. Nomenclature

An accepted anatomical nomenclature affords the opportuni-ty to standardize the description of brain structures. This hasnot yet happened with respect to the SCN intrinsic anatomy.Historically, there have been two approaches to SCN organi-zation, one based on the rat model and the other based on thehamster. The rat SCN has been commonly divided intodorsomedial and ventrolateral divisions. Although these aregeographic designations, the conceptualization is based onthe fact that, in this species, the vasoactive intestinalpolypeptide (VIP) neurons plus most of the geniculohypotha-lamic tract (GHT) and retinal afferents occupy the latter, whilethe former is the location of most vasopressin (VP) neurons.The hamster SCN has a caudocentral region containingneurons identified by several different peptides and this hasbeen referred to as the “central subnucleus” which partiallyoverlaps with the region containing VIP neurons. In thisspecies, there has been no anatomical attempt to designate a“ventrolateral” SCN division, although some investigatorshave applied rat-derived descriptors to the hamster. Asindicated in Section 3.2, identification of a dorsomedial areabased upon the location of VP-immunoreactive neurons is ofdubious value, even in the rat.

The nomenclature has become further confused withreferences to a “core and shell” SCN organization. Forexample, in a 1996 review, one of us (LPM) used the terms asan analogy to facilitate understanding of the relationshipbetween the hamster central subnucleus and its surround(Miller et al., 1996). However, another contributor to thatarticle employed the same terms in a different review, alsopublished in 1996, as labels of regions more or less equivalentto the previously used dorsomedial and ventrolateral divi-sions in the rat SCN (Moore, 1996). Given the reasonablelikelihood that a common SCN organizational plan willemerge and be applicable to rodents, and that apparentstructural differences between rat and hamster SCN will bereconciled, we believe it is premature to adopt the “core andshell” terminology based upon their application to anatomicaldescription in one species or another. More important, and asdiscussed below, we are of the opinion that the “core andshell” terminology (A) is not supported by the data in any

species; (B) is not consistent with the apparent anatomicaldifferences between species; (C) has been employed withdifferent meanings in different species, as well as within thesame species; (D) unduly simplifies a structure which has astill-emerging level of organizational complexity and (E) ifaccepted in an unquestioning manner, will have a directnegative influence on experimental design, while augmentingthe likelihood of erroneous and incomplete interpretation ofnew and available data.

Another reason for caution in using the core/shellterminology is that they are based primarily on the localiza-tion of the neuronal cell body. SCN neurons have a soma thatis approximately 10–15 μm in diameter and generally contain1–3 dendrites. These dendrites extend for a distance of about250 μm. Given the small size of the SCN (300–400 μmwide and400–500 μm tall), a dendrite has the ability to project acrossmuch of the SCN (van den Pol, 1980, 1991; Silver and Brand,1979; Jiang et al., 1997). This implies, for example, that a cellbody in the dorsal part of the SCN could send a dendriteventrally into the region where the densest retinal terminalsare located. Indeed, individual dendrites from the dorsal SCNhave been shown to extend into the ventral SCN and furtherinto the optic chiasm (Silver and Brand, 1979). Rejection of“core and shell” nomenclature specifically does not meanthat there is no specialized organization within the SCN.However, we believe the research goal should be to determineanatomical reality with a view toward consistent andconstructive use of a nomenclature that is sufficiently fluidto permit revisions reflecting the evolution of knowledgeabout the SCN.

3.2. Distribution of cell phenotypes

Essential to the understanding of circadian rhythm systemfunction is knowledge about the intrinsic organization of theSCN, site of the predominant circadian clock in mammals. Ithas been commonly stated that all SCN neurons appear tocontain GABA (Moore and Speh, 1993; Morin and Blanchard,2001; Abrahamson and Moore, 2001), although an immuno-histochemical, electron microscopic study of the rat SCNsuggests that only 40–70% contain GABA (Castel and Morris,2000). There is a dense GABAergic plexus throughout thenucleus, but a clear spatial organization of the SCN isevident if other cellular characteristics are evaluated. Thetwo most obvious are the distribution of cell phenotypes andthe distribution of cells projecting to non-SCN targets. Theformer can be ascertained by immunohistochemical means,while the latter relies largely upon retrograde tracerinjections.

A cogent argument has been made supporting the viewthat the SCN consists of two major divisions, a dorsomedialarea related to the presence of VP neurons and a ventral arearelated to the presence of VIP cells, dense retinal afferents andneuropeptide Y-containing terminals of the GHT (Moore et al.,2002). However, the evidence seems to suggest the presence ofmore than two subdivisions within the SCN, with some moreexpressly related to cell phenotypes than others. For example,the rat SCN has VP cells distributed in a crescent extendingfrom a ventral location, medially, dorsally and laterally,almost encircling a region that contains the VIP neurons.

Fig. 1 – Schematic organization of the hamster SCN at a levelthrough the central subnucleus identified by SP, CALB, CALRand GRP immunoreactive cells. (A) Approximate locations ofcell types. VP, SS and CCK are found in a generallydorsomedial location, although this may vary according totime of day (indicated by shading). CCK cells are also evidentventrolaterally (black dots); (B) terminal fields of the threemajor afferent pathways. PACAP and glutamate from retinalprojections occupy the entire SCN, to a varying degree (seePart C of this figure). 5-HT terminals from the median raphenucleus fill essentially the entire nucleus, but only sparsely inthe central and dorsolateral regions. NT, NPY, ENK and GABAarrive from the IGL, although the GABA terminals from the IGLhave not been explicitly identified (see text) and (C)identifiable zones of retinal projections. Black = region ofdense retinal innervation, predominantly from thecontralateral retina. Cross-hatched = region of modestinnervation predominantly from the ipsilateral retina. Shadedgray = region of lesser, approximately equal, innervation fromeach retina. This diagram is a guide, not a rule. The lines donot represent absolute boundaries.


These two areas are distinct from, but appear to overlap with,a predominantly dorsolateral region containing calretinin(CALR) and enkephalin (ENK) neurons (Moore et al., 2002).Encompassed by the region of VIP cells is a smaller domain ofneurotensin neurons.

In the mouse, the regional situation resembles that of therat to the extent that there appear to be multiple SCN areascharacterized by one or more particular cell phenotypes. VPcells are scattered across the rostral SCN, are locatedpredominantly in the typical dorsomedial aspect of the centralSCN and are again scattered through the entire caudal SCN.Some cells are also present in the ventral and ventrolateralSCN. ENK cells are grouped as a collection in the dorsal SCN,while VIP neurons are generally localized to the ventral part ofthe nucleus. GRP cells occupy a more central region, slightlydorsal to the VIP neurons and ventral to the ENK cells(Abrahamson and Moore, 2001; Silver et al., 1999). A greenfluorescent protein reporter which minimizes interferencefrom GRP immunoreactivity in cell processes, indicates thatthese neurons occupy a large percentage of the central mouseSCN and that the GRP region extends dorsolaterally to theperimeter of the nucleus (Karatsoreos et al., 2004).

Interpretation of regional distinctiveness based on cellphenotype may be limited by the experimental methods. Forexample, recent data suggest that discrete VP and VIP regionsmay be more apparent than real. Immunohistochemistry forVP peptide generally reveals dense staining in the dorsome-dial hamster SCNwhere there is amix of cells and processes inwhich cells are often difficult to identify (e.g., Miller et al., 1996;Kalsbeek et al., 1993). In marked contrast, in situ hybridizationmethods detecting VP mRNA enable clear visualization ofindividual neurons. This procedure reveals that, at certaincircadian times, the gene is active over a much wider portionof the SCN (Silver et al., 1999; Hamada et al., 2001), includingareas that do not contain neurons clearly identifiable by VPpeptide content. In fact, at certain circadian times, virtually noVP mRNA at all is detected in the SCN (CT20), whereas at CT8,the VP gene appears to be active in the majority of anteriorSCN neurons (Hamada et al., 2004a). In situ hybridizationmethods also demonstrate a much larger distribution of VPmRNA in the rat SCN than expected from comparable resultsbased on immunohistochemistry (Dardente et al., 2002). Themethod-related differences are also apparent when evaluat-ing rat VIP. These results plus those for VP mRNA or peptideindicate clearly different, but extensively overlapping, dis-tributions of VP and VIP cells in the rat SCN. The data haveimplications for all allegations of discreet VP or VIP regionsand to any specific functions related to those cells or the areasin which they are purportedly found.

Despite the foregoing caution, analysis of the hamster SCNindicates that it shares with other rodents the apparentseparation of a VP region from a VIP region (Fig. 1) becausethere is no question that the most persistently evidentdorsomedial part of the VP region does not overlap with theVIP region (Hamada et al., 2004a; Card and Moore, 1984). Thehamster SCN also has a dorsolateral region generally devoid ofa presently known specific cell phenotype. The hamster SCNalso contains a feature initially thought to be unique to thisspecies, namely a small, fairly central region identifiable by itsdistinct sets of neurons. The regionwas originally identified as


the location of substance P (SP) cells (Morin et al., 1992)visualized with the use of colchicine. However, the sameregion contains numerous cells synthesizing calbindin (CALB)(Silver et al., 1996a) which ismore easily visualized. The regionalso contains neurons immunoreactive to gastrin releasingpeptide (GRP) (Miller et al., 1996) and calretinin (CalR; Silver etal., 1996a; Blanchard and Morin, unpublished data). Further-more, this caudocentral SCN subnucleus overlaps to somedegree with the VIP region (Aioun et al., 1998).

The complexity of the caudocentral region of the hamsterSCN is borne out by colocalization studies. About 8% (Silver etal., 1996a) or 65% (LeSauter et al., 2002), 37 and 33% of CalB cellshave colocalized SP, GRP and VIP, respectively. About 80–90%of SP cells and 43% of GRP cells also contain CalB and 60% ofVIP neurons located within the region of the CalB neuronsactually containing CalB (LeSauter et al., 2002). GRP has alsobeen observed colocalized with VIP (Aioun et al., 1998).Colocalization of CalB and cholecystokinin (CCK) does notoccur (LeSauter et al., 2002). CalR has not been examined forthe possibility of colocalization in the SCN, nor have mostother combinations of antigens.

The region containing VP neurons is matched reasonablywell by the locations of CCK cells. Nevertheless, to the limitedextent that the issue has been examined, there is nocolocalization of VP and CCK in the hamster (Blanchard andMorin, unpublished data). In themost ventrolateral part of thenucleus, CCK cells are evident at a location containing VPfibers (LeSauter et al., 2002).

The caudocentral SCN subnucleus may be special withrespect to circadian rhythm regulation. This issue is consid-ered separately below. From the comparative anatomicalperspective, it is useful to question the extent to which otherspecies have a clearly defined region of likely homology. Theanswer is not clear. The only other species known to date withSP cells at a similar location is the Djungarian hamster (Reussand Bürger, 1994). The grass rat has CALB-IR neurons clusteredin the SCN (Mahoney et al., 2000), apparently more centrallylocated than those in the hamster in which the subnucleus islocated in the caudal half of the SCN. In rats, CALB-IR neuronsare more prevalent in the ventral SCN, but are not as tightlyclustered as in the hamster (Mahoney et al., 2000; Arvanito-giannis et al., 2000). In mice, CALB neurons are present withinthe general central SCN area, the surrounding region andgenerally throughout the entire nucleus (Abrahamson andMoore, 2001). However, the mouse SCN does have a clearlydefined central region notable because of the absence of aspecific set of identifiable neurons (LeSauter et al., 2003). Thisregion correspondsmost closely to the location of neurotensinand GRP neuron clusters (Abrahamson and Moore, 2001). Aregion with reasonably clear homology to the hamster centralsubnucleus has been described in the goldenmantled squirrel,but is recognized by a distinct cluster of ENK, rather than SP,cells (Smale et al., 1991).

The rat SCN is more compressed along the dorsoventralaxis than it is in the mouse or hamster. A cell groupingobviously homologous to those in the hamster caudocentralsubnucleus has not been described in the rat. However, anargument can be made that cells containing GRP occupy asimilar position. The GRP cells, in particular, meet the criteriaof being tightly distributed, more or less centrally located

within the SCN, with a distribution not identical to anotherknown phenotype (Dardente et al., 2002 but see Moore et al.,2002 for a different perspective). The group of rat GRP cellsappears to be smaller than the group of VIP cells andpositioned within the distribution of those cells. A smallnumber of NT neurons has been described in the rat SCN in adistribution that overlaps with the lateral collection of VIPcells (Moore et al., 2002). Another candidate rat SCN homologof the caudocentral hamster SCN subnucleus consists of theregion occupied by CalR cells. These are present in a locationdistinctly different from the other cell phenotypes in thisspecies, overlapping the dorsomedial VP cells to some extent,but largely occupying a separate region in the dorsal part ofthe nucleus (Moore et al., 2002).

As has been noted previously (Morin, 1994), there are sub-stantialspeciesdifferencesinSCNstructure.However,thereareat least two organizational characteristicswhichmany speciesseemtoshare.ThesearetherespectivegenerallocationsoftheVPand VIP cell groups. The former tends to occupy a dorsomedialSCNposition, while the latter group sits centrally in the ventralpart of the nucleus (Fig. 1). This spatial characteristic has beenobservedintheSCNofrat(Moore,1983),hamster(CardandMoore,1984), grass rat (Smale and Boverhof, 1999), golden mantledgroundsquirrel(Smaleetal.,1991),Europeangroundsquirrel(Hutet al., 2002), Octodon degus (Goel et al., 1999), blind mole rat(Negroni et al., 1997),monkey (Moore, 1993) and human (Moore,1993).However,theminkandmuskshrewarestrikingexceptionsto this rule as these species have no VP neurons present at thelocation normally identified as the SCN (Martinet et al., 1995;Tokunagaetal.,1992).

Cassone et al. (1988) evaluated comparative SCN organi-zation by analyzing retinal projections to the SCN and therelative distribution of various peptidergic neurons. Theydistinguish “visual” and “nonvisual” regions of the SCN, withthe latter receiving a much reduced density of retinalprojections compared to the former. The “visual” regionusually contains VIP cells, although it is noteworthy thatthe specific location of this region varies. In the eutherianmammals examined (house mouse, guinea pig, cat, pig), the“visual” region resides in the ventral or ventrolateral SCN.Among marsupials, there are no VIP cells evident in the SCNof the kowari or stripe-faced dunnart, while in the bandicoot,VIP cells are codistributed medially and dorsomedially withthe VP neurons. The retinal projection of the Virginiaopossum and gray short-tail opossum distributes ventrallyand medially in the caudal half of the SCN (Cassone et al.,1988). In these two species, VIP and VP neurons arecodistributed through the retinorecipient region and mediallyin more rostral SCN. Thus, in general, VIP neurons tend to beassociated with the densely retinorecipient region, whereasthe VP neurons are also located there in some species, but notin others, including most of the eutherian mammals.Whether these differences in peptidergic cell distributionreflect the functional organization of the SCN remains to bedemonstrated. It is also worth noting that the laterodorsalSCN of most mammals is a “specialized” region remarkablebecause of the general absence of non-GABA cell phenotypes(Abrahamson and Moore, 2001; Moore et al., 2002; Silver et al.,1999; Card and Moore, 1984; Smale and Boverhof, 1999; Smaleet al., 1991; Goel et al., 1999; Cassone et al., 1988).


3.3. Intra-SCN connections

Studies of intra-SCN connections fall into two basic categories.One involves filling individual cells and mapping processeswithin the nucleus. The second involves the use of tracttracers applied iontophoretically or by small volume injection.The first procedure is limited by the fact that relatively fewcells can be examined, usually only one per SCN. A successfulapplication of the cell filling technique has demonstrated thatdendritic processes of CALB neurons in the hamster SCN havea predominantly dorsomedial orientation (Jobst et al., 2004).This suggests that CALB neurons receive input from dorsome-dially located neurons, many (but not all) of which contain VPor CCK (LeSauter et al., 2002).

The tract tracer application method has the advantage oflabeling more cells, but the disadvantage of usually spillinglabel into regions adjacent to the intended target. With thatcaveat in mind, small iontophoretic injections of biotinylateddextran amine (BDA) into the SCN have provided a good initialappraisal of connectivity within the nucleus. Confocal micro-scopic analysis of the injected brains shows that most, but notall, of the cell phenotypes are connected to CALB neurons (Fig.2), the focus of this particular study (LeSauter et al., 2002).Particularly noteworthy is the absence of efferent or afferentconnections between CALB and VP neurons. Interconnectionsamong other pairs of neuron phenotypes have not beenexamined.

Use of the Bartha pseudorabies virus as a tract tracer hasprovided a novel perspective on intrinsic SCN connectivity. Forexample, an injection into the dorsomedial hypothalamusretrogradely labels, by 48 h postinjection, many neurons in thedorsal third of the ipsilateral rat SCN (Leak et al., 1999). About20–24 h later, labeled cells are present throughoutmuch, if not

Fig. 2 – Connectivity of cell types within the hamster SCN.CCK = cholecystokinin; GRP = gastrin releasing peptide; VIP =vasoactive intestinal polypeptide; VP = vasopressin. Arrowsindicate the direction the cell types project. Numbers indicatethe percentage of cells with afferent contacts. Dashed arrowsindicate SCN-afferent projections containing serotonin (5-HT)and neuropeptide Y (NPY) from the median raphe nucleus(Mn) and intergeniculate leaflet (IGL), respectively. Datadrawn from Table 1 in LeSauter et al. (2002).

all, of the SCN. An injection of virus into the subparaven-tricular zone initially labels cells largely in the ventral twothirds of the rat SCN, but by 54 h, labeled cells denselypopulate the whole nucleus. The data suggest that cells in thedorsal SCN receive projections from cells in the ventral SCNand vice versa (Leak et al., 1999). It has not been determinedwhether all SCN cells become labeledwith virus after injectioninto a target of SCN efferents. Comparable studies do not existfor other species.

4. Retinal photoreceptors and projections

In nonmammalian species, the existence of photoreceptorsnot specialized for classical visual function has long beenrecognized (Gaston and Menaker, 1968; Menaker, 1968; Mena-ker and Keatts, 1968; Menaker and Underwood, 1976; Enaker etal., 1970; Underwood and Menaker, 1970). These photorecep-tors regulate photoperiodism and entrainment of circadianrhythms in nonmammals. For the same purposes, mammalsrequire photoreceptors in the eye (Foster et al., 2003; Yamazakiet al., 1999). A critical study by Foster et al. (1991) demonstratedthat rd/rdmice, sustaining thenear-total loss of rods and conesalong with their respective photopigments, retain substantial-ly normal phase responses to light. In addition, the rd/rd miceperform similarly to controls even after 767 days, althoughthere are no electroretinogram responses after 210 days of age(Provencio et al., 1994). The implication of these studies wasthat a novel retinal photoreceptor was sufficient for light-induced phase shifts in mammals. In 1998, Provencio et al.(1998a) identified the molecule, melanopsin, and described itspresence in Xenopus brain, eye and photosensitive skinmelanophores. The human and mouse melanopsin geneswere then cloned and demonstrated in cells of the mouse andmonkey retinal ganglion cell layer (Provencio et al., 2000).

4.1. Melanopsin ganglion cell characteristics

The results of Provencio and colleagues were rapidly adaptedto specifically identify photoreceptive retinal ganglion cells(Berson et al., 2002) with additional work demonstrating thatmelanopsin behaves as a photopigment capable of activating aG-protein (Newman et al., 2003; Melyan et al., 2005; Qiu et al.,2005; Panda et al., 2005; Isoldi et al., 2005). Although there isdebate concerning the shape of the action spectrum, the bestindications are that the melanopsin action spectrum peaks atabout 480 nm.

Several studies have been designed to evaluate lightsensitivity of photoreceptive cells contributing to the retino-hypothalamic tract (RHT). The results clearly demonstratedthat a special class of ganglion cells projecting to thesuprachiasmatic nucleus (SCN) are both photoreceptive inthe absence of known synaptic input (Berson et al., 2002) andcontain melanopsin (Hattar et al., 2002; Gooley et al., 2001). Inthe hamster retina, there are approximately 1600melanopsin-containing cells (Morin et al., 2003), comprising an estimated1.5% of all ganglion cells (Tiao and Blakemore, 1976; Hsiao etal., 1984; Rhoades et al., 1979). In the rat, the estimate is 2.5%(about 2455melanopsin cells) and, in themouse, 1% (about 730cells) (Hattar et al., 2002). These cells provide the retina with


what has been described as a “photoreceptive net” provided byoverlapping, large arbors of beaded dendrites (Berson et al.,2002; Provencio et al., 2002). Apparently, all portions of themelanopsin-containing cells are photoreceptive (Berson et al.,2002). Melanopsin of retinal origin has been described in theSCN of rats (Beaulé et al., 2003b), although it has not beenobserved in the SCN of mouse or hamster (Provencio,unpublished data; Blanchard and Morin, unpublished data).In the hamster, the melanopsin cells are spread homoge-neously throughout the retina, as they also are in the cat(Semo et al., 2005). The mouse may have a higher density ofmelanopsin in the superior retina (Hattar et al., 2002), a featurethat has been reported in the rat (Hattar et al., 2002; Hannibalet al., 2002) and the diurnal species, the Nile grass rat(Blanchard, Novak and Morin, unpublished data).

Nearly all the melanopsin-containing cells are present inthe ganglion cell layer of the retina. However, there is a smallpercentage of the cells found as “displaced” to the innernuclear layer (Berson et al., 2002). In either case, the dendriticarbor is most extensive along the border between the innernuclear and inner plexiform layers. The predominant locationof the dendrites is in the OFF sublayer of the inner plexiformlayer. The frequency histogram of soma diameters of mousemelanopsin ganglion shows amean of about 16.3 μm(Hattar etal., 2002). In the cat, SCN-projecting ganglion cell diameter issimilar to the rat (17.2 μm) with dendrites ramifying primarilyin the inner plexiform layer (Pu, 1999). The size distribution isnot symmetrical, but skewed toward larger diameters for bothspecies.

Photoreceptive ganglion cells were first described followingintra-SCN injection of a retrograde tracer. This permittedidentification of ganglion cells contributing to the RHT andcould be combined with recording from those cells in theabsence of input from other retinal neurons (Berson et al.,2002). With or without the presence of drugs blocking synapticinput, the cells are slow to respond to light stimuli withsaturating light pulses requiringhundreds ofmilliseconds, andnear threshold stimuli requiring about 60 s, to response onset.Latency to maximal depolarization is reduced as stimulusirradiance increases. Subsequently, the depolarization slowlydecays to a plateau that more or less endures for the durationof the stimulus, with the plateau proportional to the stimulusenergy. The cells are wavelength-sensitive with a peakresponse at about 484 nm, similar to the results from actionspectra for rodent circadian rhythm phase response (Takaha-shi et al., 1984; Provencio and Foster, 1995). For a fixedwavelength stimulus, peak depolarization increases linearlyuntil reaching a saturating irradiance. Spike frequency appearsrelated to the extent of depolarization. Subsequent to stimulusoffset, repolarization is very slow, takingminutes after stimuliof high irradiance. Spike discharges are maintained untilpolarity returns near baseline. The membrane depolarizationis produced by an inward current that is activated by the lightpulse (Warren et al., 2003). The current activates slowlyreaching a peak several seconds after the application of alight pulse. The time course of the membrane depolarizationand the inward current are similar. The signal transductionpathways couplingmelanopsin to the activation of the inwardcurrent are not known. The current was unaffected bysubstitution of extracellular Na+ and shows both inward and

outward rectification. These data are consistent with thecurrent beingmediated by channels of the trp family (transientreceptor potential) of channels (Warren et al., 2003). Propertiesof SCN-projecting ganglion cells (identified by retrogradetracer) recorded from a cat eyecup preparation in which therewas no elimination of rod/cone contribution indicate that theyexhibit sustained responses of the “on” or “on–off” centervariety; have a spectral sensitivity peak at about 500 nm andresponse is optimal with motionless or slow-moving stimuli(Pu, 2000).

4.2. Retinal ganglion cell projections and bifurcation

Retinal projections to the SCN were definitively described in1972 (Moore and Lenn, 1972; Hendrickson et al., 1972) andfocused scientific attention squarely on that nucleus as theprobable site of the neural circadian clock. Evaluation of theRHT has continued as methods have improved and morespecies are evaluated. The extent to which the retina projectsto each SCN is highly variable across species. The diurnalgrass rat has bilateral retinal innervation of the SCN, althougha preponderance arrives from the contralateral retina (Smaleand Boverhof, 1999). This differs from what is seen in thediurnal golden mantle ground squirrel in which RHT inner-vation is exclusively from the contralateral retina (Smale etal., 1991). A recent report shows modest ipsilateral SCNinnervation by the retina in the diurnal California groundsquirrel (Major et al., 2003). In contrast, following analysis ofdarkfield images, the hamster RHT was described as bilater-ality symmetrical (Johnson et al., 1988a). The issue of lateralityand pervasiveness of the retinal projection in the hamsterSCN has now been evaluated with confocal microscopy(Muscat et al., 2003). The data provide several novel observa-tions. First, nearly the entire SCN is innervated by the RHT,consistent with earlier HRP results (Pickard and Silverman,1981). Second, despite the extent of innervation, it is nothomogenous within the SCN. And third, the axons from oneretina project to most or all of the SCN bilaterally. The centraland dorsal SCN receive dense innervation predominantlyfrom the contralateral retina, while the ventromedial SCNreceives moderate innervation largely from the ipsilateralretina. Thus, innervation of the SCN by one retina is generallypervasive, but with clear regional specificity. Of particularinterest is the presence of the densest retinal innervation bothto the region of the CALB-containing central subnucleus of theSCN and to a region slightly more dorsal. The heavilyinnervated dorsal area corresponds to a zone in whichphosphorylated extracellular signal-regulated kinase (pERK)is found (Lee et al., 2003), a characteristic that is apparentlyimportant to pERK activity.

Other species also appear to have overlapping, but pre-ferred SCN regions of innervation by retinal projections. This isapparent in the rat, but not the mouse (Shivers and Muscat,2004). One apparent difference between the RHT terminalfields of these species and that of the hamster is thesubstantially absent innervation of the dorsomedial SCN thatlargely corresponds to the area in which most VP cells arepresent.

The retina projects to the SCN via the RHT, a major directprojection, and through a secondary, indirect pathway


through the IGL and the GHT. In addition, there arenumerous retinal projections to disparate brain regionsmany of which have potential, reciprocal connections tothe SCN through the IGL (Morin and Blanchard, 1998). Onesuch is a direct retinal projection to the dorsal raphenucleus, with an initial description in cats given in 1978(Foote et al., 1978). Interest was further heightened with thedemonstration, based on both anterograde and retrogradetracing data, of a similar pathway in rat (Shen and Semba,1994). This pathway has also been observed in gerbil (Fite etal., 1999, 2003) and O. degus (Fite and Janusonis, 2001). Thepathway has not been found in hamster, although it hasbeen sought several times (Morin and Blanchard, unpub-lished data; Fite, unpublished data).

Bifurcation of ganglion cell axons and subsequent projec-tion of these cells to the SCN and IGL (Pickard, 1985) appears tobe a common characteristic of the circadian visual system.The presence of ganglion cells that bifurcate to innervate boththe SCN and IGL has been recently confirmed, with theadditional information that similar bifurcation is present forindividual cells projecting to both the SCN and superiorcolliculus or OPT (Morin et al., 2003). Injection of tworetrograde tracers, one into each SCN, also demonstratesthat at least some ganglion cells bifurcate and send at leastone axon branch to each SCN. Further, at least some of thesecells contain melanopsin, as do some that project to the OPT.In contrast, individual ganglion cells apparently do notbifurcate and project bilaterally to the IGL, OPT or the SC.Other retinorecipient regions have not been simultaneouslyexamined for both bifurcation and melanopsin cell afferents.Nor has the question of individual ganglion cells projecting tomore than two visual targets been examined. The prospect islikely that individual ganglion cells have multiple axonalprocesses (beyond two) that would allow a rather smallnumber of photoreceptors to commonly influence a broadrange of visual functions.

A novel contribution to the analysis of retinal projectionsto the SCN comes from two studies by Abe and Rusak (1992).In the initial paper, hamsters were electrically stimulated inthe IGL with the result that FOS protein synthesis wasinduced in the centrodorsal SCN. The preliminary interpre-tation was that GHT activation was causal to FOS expression(Abe et al., 1992). However, a subsequent study (Treep et al.,1995) showed that FOS expression could not be stimulated inbilaterally enucleated animals. This result suggested analternate interpretation, namely, electrical stimulation mayhave elicited antidromically activated retinal ganglion cellswhich produced orthodromic activation of the SCN via thecollaterals from bifurcating axons in the RHT (Morin et al.,2003; Pickard, 1985). The resulting pattern of FOS proteinexpression would uniquely represent the distribution ofretinal projections that bifurcate and project to both theIGL and SCN. The extent to which such bifurcating projec-tions originate from ganglion cells that do or do not containmelanopsin is unknown.

Photoreceptive ganglion cells contributing to the RHTcontain melanopsin (Hattar et al., 2002; Gooley et al., 2001;Morin et al., 2003; Hannibal et al., 2002; Gooley and Saper,2003). To date, however, there has not been a comprehensivedescription of melanopsin ganglion cell projections to other

brain areas. The initial report of melanopsin cell targetsshowed X-gal staining for β-galactosidase activity in atransgenic mouse strain with tau-lacZ targeted into themelanopsin gene locus (Hattar et al., 2002). This allowedvisualization of β-galactosidase fused to a protein antero-gradely transportable to axon terminals. The blue reactionproduct emphatically revealed projections to the SCN, IGLand olivary pretectal nucleus (OPT). Subsequently, retrogradetracing studies have confirmed melanopsin-containing gan-glion cells projecting to these retinorecipient targets, as wellas to the ventral lateral preoptic area, and the subparaven-tricular zone of the hypothalamus (Morin et al., 2003; Gooleyand Saper, 2003; Sollars et al., 2003). Melanopsin cells in thehamster (Morin et al., 2003), but not the rat (Gooley andSaper, 2003), have been reported to project to the superiorcolliculus. The retrograde studies confirm the projectionsfrom melanopsin cells to initially defined major visualtargets, but suggest that not all the ganglion cells projectingto these areas contain melanopsin. Estimates are that 10–20% of ganglion cells projecting to the SCN do not containmelanopsin (Morin et al., 2003; Gooley and Saper, 2003;Sollars et al., 2003).

The suggestion that melanopsin cells and their projec-tions create a “nonimage forming” visual system has beenrecently put under pressure with the observation that atleast some of these cells project to the monkey dorsal lateralgeniculate nucleus (Peterson et al., 2003; Dacey et al., 2005).These cells appear to provide both irradiance and colorinformation to the dorsal lateral geniculate nucleus, a siteinvolving in the processing of feature-related information,i.e., images. Melanopsin cell projections to dorsal lateralgeniculate nucleus have not been found in the rat (Gooleyand Saper, 2003). As indicated above, the melanopsin cellsare known to project to several other retinorecipient nucleiand bifurcation may be a common feature of axonsprojecting through the RHT to the SCN. Therefore, it is likelythat the melanopsin cells projecting to the dorsal lateralgeniculate nucleus also bifurcate and innervate many, if notall, other subcortical visual regions.

An unexpected development has arisen from a thoroughevaluation of an engineered version of the Bartha strainpseudorabies virus (PRV152), as a retrograde transsynaptictracer (Pickard et al., 2002). Previously, viral tracing unequiv-ocally demonstrated a subset of retinal ganglion cells follo-wing injection into the contralateral eye. These cellspresumably gave rise to the RHT (Moore et al., 1995;Provencio et al., 1998b; Hannibal et al., 2001a). However, anew interpretation of these results is based on the fact thatPRV152 is exclusively a retrograde tracer that makes its wayto the SCN multisynaptically through the nucleus ofEdinger–Westphal in the midbrain oculomotor complex(Pickard et al., 2002). Cells in the SCN, IGL and OPT areinfected, as are several ganglion cell types in the contralat-eral retina. Only a subset of these contribute to the RHT.Viral transsynaptic tracing is a powerful method and has theadded advantage of revealing second order retinal neurons,bipolar and amacrine cells, that synapse with the primaryganglion cells (Belenky et al., 2003). This observation hassignificant implications for the manner in which lightregulates circadian rhythm phase.


4.3. Classical and ganglion cell photoreceptor contributionto regulation of circadian rhythms, masking, the pupillarylight reflex and melatonin suppression

4.3.1. Circadian rhythmsThree classes of photoreceptive cell types reside in themammalian retina. The current view is that while all effectsof light on the circadian visual system are accounted for bythe three types, no single photoreceptor is necessary forentrainment. The rd/rd mouse strain suffers postnataldegeneration of rods and cones. These individuals haveessentially normal phase response to light (Foster et al.,1991; Provencio et al., 1994; Freedman et al., 1999) and havean action spectrum for phase shifts that peaks at 480 nm(Yoshimura and Ebihara, 1996; Hattar et al., 2003) (but seeProvencio and Foster, 1995). Recently, entrainment stabilityat threshold levels of illumination was assessed (Mrosovsky,2003) and, with this simple procedure, only 10% of rd/rdmice were able to remain entrained to a 12:12 LDphotoperiod when the daytime light intensity was 15stops (b2 lx), compared to about 80% of wild-type mice.This result supports the view (Yoshimura et al., 1994) thatthere is a circadian system response to light that isexclusively rod and/or cone-sensitive. Rods may regulatehamster entrainment under very dim lighting conditions(Gorman et al., 2005). Neurophysiological methods alsosupport the view that both rods and cones (primarily 510nm green and secondarily 375 nm near-ultraviolet) providephotic information to the SCN (Aggelopoulos and Meissl,2000).

A flurry of investigations followed the discovery ofmelanopsin photoreceptive retinal ganglion cells. Melanop-sin gene knockout (KO) mice were made by two differentmethods, yet the effects on light-induced rhythm responseswere very similar (Panda et al., 2002; Ruby et al., 2002a). Theanimals entrain and free-run properly in normal light–darkphotoperiod and constant dark conditions, respectively.However, phase response to light is attenuated in the KOanimals. With a 480 nm stimulus, phase shift magnitudewas only about half that of wild-type mice at each of three,nonsaturating irradiance levels (Panda et al., 2002). Asaturating, white light pulse also yielded a diminishedphase shift from the KO animals, although it was abouttwo thirds normal (Ruby et al., 2002a). A second defect inresponse to light by the melanopsin KO animals was thefailure of constant light (LL) to elicit a properly lengthenedcircadian period. In both strains, LL (white light) effectivelyincreased the period, but in melanopsin KO animals, thechange was only about 55–65% of controls (Panda et al., 2002;Ruby et al., 2002a). It is worth noting that similar effects onlight-induced circadian period lengthening can be achievedby IGL lesions (Morin and Pace, 2002), possibly because suchlesions would destroy the functional effectiveness of mela-nopsin cell projections to the IGL (Hattar et al., 2002). Lightadministered at CT16 was able to induce c-fos mRNA in SCNcells of melanopsin KO animals, although there has been notest of the normality of this expression (Ruby et al., 2002a).An important outcome of the melanopsin KO studies is theimplication that both classical photoreceptors and intrinsi-cally photoreceptive, melanopsin-containing retinal ganglion

cells contribute, approximately equally, to circadian rhythmregulation by light.

Subsequently, two sets of investigators, again usingsomewhat different methods, addressed the issue of photo-receptor types and circadian rhythm regulation using animalsdefective in both melanopsin production and classical photo-receptor function. Again, the results are substantially thesame: animals lacking melanopsin but bearing the rd/rdgenotype (Panda et al., 2003) or undergoing rod and conedegeneration plus having defective phototransduction incones (Hattar et al., 2003) free-run under normal light–darkphotoperiod conditions. Lack of entrainment was not alteredby light intensity nor was circadian period under LL differentfrom the circadian period under DD. In addition, light wasunable to inhibit nocturnal melatonin synthesis (Panda et al.,2003). It should be noted that the defects in rhythm responseto light are not the result of loss of ganglion cells that normallysynthesizemelanopsin or their projections (Hattar et al., 2003).

4.3.2. Masking, pupillary light reflex and melatonin suppressionOther accessory visual functions are also modified by theabsence of melanopsin. Although light-induced negativemasking is reduced, particularly at lower light intensities, inrd/rd mice, the melanopsin KO alone yields masking that isabout 30% greater than in the rd/rd mice (Panda et al., 2003).When the melanopsin KO and rd/rd traits are combined, themice do not exhibit negative masking to light (Hattar et al.,2003; Panda et al., 2003). Melanopsin does not seem to beinvolved in the positive masking that occurs in response todim light exposure (Mrosovsky and Hattar, 2003), but appar-ently contributes to negative masking in the presence of light≥10 lx (∼6.7 μW/cm2).

Pupillary constriction in response to photic stimuli is muchreduced in rd/rdmice and the response has a longer latency. Athigh irradiances, pupillary response equals that in wild-typemice, but the response of rd/rdmice is markedly attenuated atintermediate and lower irradiances (Hattar et al., 2003; Pandaet al., 2003; Lucas et al., 2001). Unlike masking and phaseresponse to light, the pupillary light reflex is relativelyunmodified in the absence of melanopsin. Only at irradiancesN10 × 1012 photons/cm2/s is there a modest sensitivity deficit.The full effect of light on the pupillary light reflex can beaccounted for by the additivity of rod/cone and melanopsinphotoreceptor contributions (Panda et al., 2003; Lucas et al.,2003), an observation supported by the absence of anypupillary light reflex in melanopsin KO animals lacking rodand cone function (Hattar et al., 2003; Panda et al., 2003). Asingle opsin/vitamin A-based photopigment with a peaksensitivity around 480 nm appears to drive the pupillarylight reflex in rodless, coneless mice (Hattar et al., 2003; Lucaset al., 2001).

A contribution of the pupil to circadian rhythm regulationhas not been studied and is seldom considered as a factorregulating circadian rhythm response to light. However, it isuseful to note (a) that both classical and melanopsin photo-receptors contribute to both responses (Hattar et al., 2003;Panda et al., 2003; (b) that, within limits, phase responsemagnitude is proportional to the total number of photonsassessed by the circadian visual system across the duration oflight exposure (Nelson and Takahashi, 1991; (c) that the pupil


area can vary by a factor of 16 (Hattar et al., 2003; Lucas et al.,2003) and (d) that the pupil acts to gate exposure of thecircadian visual system to the very stimulus type to which itoptimally detects and responds.

Mice which cannot synthesize melanopsin or which bearthe rd/rd mutation show, to the extent tested, normal light-induced suppression of pineal melatonin (Lucas and Foster,1999; Lucas et al., 1999). The two photoreceptor classes appearredundant to the extent that loss of both eliminatesmelatoninsuppression by light (Panda et al., 2003).

5. Neuromodulators of the retinohypothalamictract

5.1. Glutamate

It is generally believed, based on abundant indirect evidence,that glutamate is the primary neurotransmitter of theretinohypothalamic tract. This topic (up to 1996) has beenextensively reviewed (Ebling, 1996; Hannibal, 2002a) and willnot be considered in depth here. However, despite thenumerous studies of function involving use of glutamatereceptor agonists and antagonists, there ismuch less certaintyregarding its actual presence in the RHT. Most convincing inthis regard has been the electron microscopic demonstrationthat glutamate immunoreactivity occurs in cholera-HRPidentified terminals of retinal projections to the mouse andrat SCN (De Vries et al., 1993; Castel et al., 1993). Electricalstimulation of the optic nerve induces SCN release of [3H]-glutamate or [3H]-aspartate (but not [3H]-GABA) loaded ontorat slice preparations (Liou et al., 1986). However, directstimulation of the SCN yields even greater release of the twoamino acids, suggesting their presence in nonretinal term-inals and SCN cells, in addition to RHT terminals. Pharmaco-logical studies suggest that aspartate may be the most potentexcitatory neurotransmitter in the rat RHT (Shibata et al.,1986). Two laboratories have evaluated the uptake andretrograde transport of [3H]-aspartate injected into the ratSCN. While the two investigations agreed with respect tononretinal projections to the SCN, one identified retinalganglion cells contributing to the RHT (Devries et al., 1995)and the other did not (Moga and Moore, 1996). More recently,glutamate has been immunohistochemically identified in theRHT colocalized with PACAP which may identify most, if notall, retinal projections to the SCN (Hannibal et al., 2000).

Vesicular glutamate transporters, VGluT1 and VGluT2,account for nearly all the glutamate transporters in the brainand are generally abundant in retinorecipient nuclei(Fujiyama et al., 2003; Land et al., 2004), for example, in thedorsal lateral and ventral lateral geniculate nuclei. However,they are not abundant in either the IGL (where their presenceis weak, at best) or the SCN. The two transporters are weaklyevident in the SCN, but not at the identical locations (Fujiyamaet al., 2003). Unilateral enucleation has little, if any, effect ontransporter visibility in the SCN, although bilateral enucle-ation may do so.

One emergent difficulty resulting from the studies ofglutamate function has been the general inability to obtain alight-type phase response curve (PRC) to direct SCN applica-

tion of excitatory amino acid. The initial study usingglutamate generated a PRC more of the NPY-type (Meijer etal., 1988), while aspartate had no effect (De Vries and Meijer,1991). Subsequently, NMDA was shown to elicit phase delayswhen administered at CT13.5 and advances at CT19, consis-tent with a light-type PRC (Mintz and Albers, 1997) and a full,light-type PRC (although the subjective night is longer andmaximal delays equal maximal advances) to direct SCNapplication of NMDA, have been demonstrated (Mintz et al.,1999). The effects of NMDA on circadian rhythm phase wereblocked by NMDA receptor antagonists. In vivo glutamateyields a PRC that differs from that for NMDA for reasons thatare not clear.

Injection of NMDA onto the SCN during the early subjectivenight elicits light-like phase delays which cannot be blockedby simultaneous tetrodotoxin (TTX) application (Gamble et al.,2003). NMDA treatment during the subjective day has no directeffect on rhythm phase (Mintz et al., 1999), but it can inhibitphase advances induced by the GABAA agonist, muscimol(Gamble et al., 2003; Mintz et al., 2002). In this respect, NMDAduring the subjective day mimics the effect of light (Mintz etal., 2002). However, the action of NMDA during the day isblocked by TTX treatment suggesting that sodium-dependentaction potentials are necessary for effects of light on rhythmphase during the subjective day, but not during the subjectivenight.

A corollary to the issue of glutamate in the RHT concernsthe biochemical pathway that is likely necessary for proces-sing the neurotransmitter in the SCN. This pathway involvesthe de novo synthesis of glutamate in neurons and glia, plusthe recycling of extracellular glutamate through glia back toneurons (Hertz, 2004). Extracellular glutamate concentrationin the peri-SCN region is greatest during the hours of darkness(Rea et al., 1993a; Glass et al., 1993). Glial fibrillary acidic acidconcentration in SCN astrocytes is lowest after lights off (Glassand Chen, 1999).

5.2. N-Acetylaspartyl glutamate

A significant question that has not been resolved is whetherglutamate is the sole amino acid transmitter in the RHTcontributing to circadian rhythm regulation. N-Acetylaspartylglutamate (NAAG) is also present in the RHT (Moffett et al.,1990, 1991; Tieman et al., 1991), although at the synaptic cleft,NAAG is converted to its constituent amino acids and each isavailable for postsynaptic activity (Baslow, 2000; Neale et al.,2000; Coyle, 1997). However, unlike NAAG, neither aspartatenor glutamate is released from retinal ganglion cell terminalsin a calcium-dependent manner (the RHT has not beenspecifically tested; see Tieman and Tieman, 1996 for areview). Whether this constitutes the source of photicallyinduced glutamate release or the transmitter is independent-ly released by RHT terminals remains to be discovered. Themost recently available information (Moffett, 2003) suggeststhe presence of NAAG in a portion, but not all, of the rat RHT(ventrolateral part of the SCN, i.e., not the whole retinoreci-pient region). Analysis of prospective NAAG function iscomplicated by the presence of numerous NAAG-IR neuronsin the SCN with particularly dense expression in dorsomedialcells (Moffett, 2003).


5.3. Glutamate receptors

One presumption, given that the RHT is likely to use glutamateas a transmitter and that glutamate agonists or antagonistsalter circadian clock function, is that there are glutamatereceptors in the SCN. The most comprehensive study ofglutamate receptor localization in the SCN has been per-formed with hamsters (Stamp et al., 1997) and demonstratesthat the GluR1 receptor is most widely distributed in the SCNwith greatest density in a region corresponding approximatelyto that which receives dense innervation from the contralat-eral retina (Muscat et al., 2003). Both the GluR2/3 and GluR4receptors are distributed primarily in the dorsomedial SCN,although there are also substantial GluR4 receptors in thecentral and ventral part of the nucleus.

One important consideration regarding glutamate recep-tors is the fact that they can also be abundant on astrocytes(Porter and McCarthy, 1997), and may indirectly modulateneuronal activity by altering glial response to neurotransmit-ters (Haak et al., 1997). Astrocytes may also be a source ofglutamate that can activate neuronal glutamate receptors (Liuet al., 2004). Astrocytes may be part of a route yieldingglutamate receptor-activated release of astrocytic endothelialnitric oxide synthase that can act in parallel with glutamate-activated neuronal nitric oxide synthase through which lightcan induce NO release in the SCN thereby altering circadianrhythm phase response (Caillol et al., 2000).

5.4. Nitric oxide and RHT signal transduction

Early articulation of the prospect that nitric oxide mightcontribute to circadian rhythm regulation was provided byAmir (1992) who tested whether the SCN mediated theincreased heart rate response to acute photic stimulation.Infusion of L-NG-nitro-arginine-methyl ester (L-NAME), acompetitive inhibitor of nitric oxide synthase, over the SCNgreatly attenuated the heart rate increase. This attenuationcould be overcome by simultaneous infusion of excess L-arginine, the natural substrate for the enzyme. Similar effectson the heart rate response were also obtained with aglutamate NMDA-type receptor antagonist and it was sug-gested (Amir, 1992) that nitric oxide mediates the effects oflight on the SCN.

Subsequently, Ding et al. (1994) published an extensiveseries of experiments probing the contribution of nitric oxideto photic signal transduction in the SCN. As a precursor to thework, the investigators evaluated phase response of the SCNneuronal firing rhythm to acute perfusion with glutamateapplied to a rat brain slice preparation. This yielded a very tidyglutamate-induced PRC (Ding et al., 1994) that closely mim-icked the normal rat light PRC (Summer et al., 1984), withglutamate yielding concentration-dependent phase shiftsacross a range of 10−6 to 10−2 M glutamate. The phasic effectsof glutamateweremimicked by in vitro and in vivo applicationof NMDA (Ding et al., 1994; Colwell and Menaker, 1992; Colwellet al., 1990, 1991; Takeuchi et al., 1991), a specific agonist forone type of glutamate receptor.

Using in vitro activation of the NMDA glutamate receptorsequelae as the indicator of a functional photic signalingpathway, the photic effects were mimicked by administering

nitric oxide generators to the slice (Ding et al., 1994). Sodiumnitroprusside, S-nitroso-N-acetyl-penicillamine and hydrox-ylamine application yielded equivalent, glutamate-like phaseadvances and delays during the subjective night, but no phasechange during the subjective day. Further, the glutamate-induced phase shifts could be blocked by any of several typesof competitive substrates for nitric oxide synthase. Theprocedures were extended to the in vivo situation in whichL-NAMEwas able to completely block hamster phase advancesinduced by light. This blockadewas reversed by increasing theavailability of L-arginine substrate (Ding et al., 1994). L-NAMEwas also able to attenuate hamster phase delays (Watanabe etal., 1995). The data suggest that light acting through the RHTreleases glutamate in the SCN, activating the NMDA receptorson SCN neurons. This increases intracellular calcium concen-tration, followed by nitric oxide synthase stimulation andnitric oxide production (Ding et al., 1994). Nitric oxide may actas a short distance neuromodulator in the photic signaltransduction pathway. The effects of melatonin and serotoninon rat circadian rhythm phase may also require nitric oxidesynthesis activation in the SCN (Starkey, 1996).

The foregoing basic principles have been linked to light-induced phosphorylation of the cAMP response elementbinding protein (CREB) in SCN neurons (Ginty et al., 1993).The appearance of phosphorylated CREB (pCREB) is theearliest known postsynaptic biochemical marker of photicsignal transduction in the SCN. In vivo light-induced pCREBimmunoreactivity in rat SCN cells attained maximal expres-sionwithin 10min following a 10min 150 lx light pulse at CT19and this could bemimicked by glutamate application to a slicepreparation (Ding et al., 1997). The effects of glutamate onpCREB immunoreactivity were blocked by an NMDA receptorblocker (APV) or a nitric oxide synthase inhibitor (L-NAME). Asyet, the steps between postsynaptic glutamate action andinduction of CREB phosphorylation or nitric oxide synthesis/release have not been determined. It is important to note thatlight- or glutamate-induced pCREB and nitric oxide activity aredependent on circadian clock phase indicating that theresponsiveness of postsynaptic targets of the RHT is limitedat certain times by feedback from the circadian clock.Whetherthis means the postsynaptic cells are intrinsically “clock” cellsor they receive rhythmic inhibitory feedback from elsewhereis not known.

Nitric oxide synthase activity in hamsters is rhythmicallyrelated to the presence of an alternating light–dark cycle. Lightadministered at any time of day elevated enzyme activity andthe day–night enzyme rhythm did not persist in DD (Ferreyraet al., 1998; Agostino et al., 2004; Chen et al., 1997). This isconsistent with the view that the consequences of nitric oxiderelease, but not the release itself, are regulated by thecircadian clock.

There is no complete consistency in the literatureconcerning nitric oxide function in circadian rhythm regula-tion. Light-induced FOS protein, thought to be an eventdownstream of CREB phosphorylation (Kornhauser et al.,1990), has been blocked in rat SCN, but not IGL, by peripheralinjection of L-NAME (Amir and Edelstein, 1997). In contrast, thesame procedure in hamsters has blocked light-inducedactivity rhythm shifts, but not light-induced FOS in the SCN(Weber et al., 1995). This may indicate a species difference.


However, in the hamster, it is further recognition of the factthat FOS protein expression is not a necessary component ofphase response to light (Colwell et al., 1993a,b).

The anatomical evidence in support of the SCN as a site fornitric oxide modulation of circadian rhythmicity is notparticularly encouraging. The earliest descriptions of potentialnitric oxide synthesizing substrate based on NADPH-diapho-rase staining in the circadian rhythm system revealed no cellsin the rat SCN (Vincent and Kimura, 1992) and sparse cells inthe Siberian hamster SCN (Decker and Reuss, 1994). Occasion-al neurons immunoreactive for the neuronal form of nitricoxide synthase have also been observed in rat SCN, and someof these contain VIP (Reuss et al., 1995). More recently, a denseplexus of processes immunoreactive for the neuronal form ofnitric oxide synthase has been identified surrounding a fewimmunoreactive cells in ventral rat SCN (Chen et al., 1997), butthis has not been seen by other investigators (Caillol et al.,2000; Wang and Morris, 1996). The most convincing anatom-ical study indicates no cells in the SCN of rat or mouseidentifiable by NADPH-diaphorase staining, but in bothspecies, immunoreactivity for neuronal nitric oxide synthaserevealed distinct cell populations within the SCN. In themouse, the labeled neurons are evenly distributed through theSCN with about one tenth of the SCN neurons labeled. Incontrast, all labeled rat neurons are grouped laterally in theventral SCN among immunoreactive neuronal processes.Neither labeled cells nor processes were evident in thedorsomedial SCN (Wang and Morris, 1996). However, underspecial slice conditions, no nitric oxide synthase immunore-activity was observed in the rat SCN although the ventrallateral part of the nucleus is densely labeled with cyclic GMPimmunoreactivity. NADPH-diaphorase is not always a usefulindicator of nitric oxide synthase (see Wang and Morris, 1996for discussion). A subsequent report using similar methods,but different antisera, identified no cells containing neuronalnitric oxide synthase in the hamster SCN and a few in theventral rat SCN. Nonneuronal cells (probably astrocytes) andprocesses immunoreactive for endothelial nitric oxidesynthase were observed in the SCN of both species (Caillol etal., 2000). Despite the biochemical manipulations indicatingthat nitric oxide synthase activity is necessary for phaseresponse to light or glutamate, mutant mice lacking the genecoding for endothelial (Kriegsfeld et al., 2001) or neuronal(Kriegsfeld et al., 1999) nitric oxide synthase apparentlyentrain normally to the light–dark cycle and light inducesnormal FOS expression in the SCN. As suggested by Caillol etal. (2000), it is possible that there are parallel neuronal andglial pathways by which photic information gains access tothe circadian clock. If this is the case, then the only effectiveknockout might require deleting the genes coding for bothendothelial and neuronal nitric oxide synthase.

5.5. PACAP

Pituitary adenylate cyclase activating polypeptide (PACAP) didnot become a molecule of interest to the circadian rhythmfield until 1997 when Hannibal et al. (1997) described itspresence in retinal ganglion cells plus a projection of thesecells to retinorecipient regions of the SCN and IGL. The samepaper demonstrated that, in a concentration-dependent

fashion, PACAP administered at CT6 onto a rat SCN slicepreparation could induce 3.5 h phase advances (tests at CT14or 19 yielded no phase change). VIP, which can act onreceptors similar to those for PACAP, also yielded phaseadvances, but the concentration sensitivity and resultingphase response were much lower (Hannibal et al., 1997). Theargument was offered that PACAP constitutes a neuromodu-lator in the RHTwhich, along with glutamate, normally acts tomodify circadian rhythm phase in response to retinalphotoreception. Subsequently, large, concentration-depen-dent, PACAP-induced (CT6) phase advances were observed inhamster SCN slices (Harrington et al., 1999). However,substantial phase delays were also observed if 1 nM PACAPwas administered at CT14, but lesser or greater concentrationshad progressively diminished effects. In addition, small phasedelays or advances were obtained in vivo following infusion of1 nM PACAP onto the hamster SCN at CT14 and CT18,respectively (Harrington et al., 1999). Thus, PACAP in vivomimicked the effect of light in terms of the direction of theelicited phase shift (tests during the subjective day were notperformed).

Replication of the foregoing PACAP effects has provenelusive. Additional in vivo hamster tests have usually, butnot always, yielded phase delays following administration ofvarious PACAP concentrations at CT14. Only phase delayswere obtained at CT18, and small phase delays (noadvances) were observed with one PACAP concentrationgiven at CT6 (Piggins et al., 2001a). Recently, a thirdinvestigation of hamster in vivo effects obtained smallphase delays following a moderately high concentration ofPACAP administered at CT14 (Bergstrom et al., 2003). As isobvious, there are numerous inconsistencies in the availabledata which require reconciliation before a function of PACAPcan be accepted.

The Hannibal lab has developed a monoclonal PACAPantibody (Hannibal et al., 1995) that has enabled production ofa large amount of information regarding the anatomy ofstructures containing this peptide (see Hannibal, 2002a for areview). PACAP is present in the RHT (Hannibal and Fahrenk-rug, 2004; Hannibal et al., 1997, 2002; Bergstrom et al., 2003)where it colocalizes with glutamate (Hannibal et al., 2000).PACAP-ir neurons in the rat SCN have also been reported(Piggins et al., 1996), although there may have been cross-reactivity with VIP. In addition, PACAP appears to be found inaxons and terminals in most of the retinorecipient regions ofthe rat brain (Hannibal and Fahrenkrug, 2004) including theIGL, ventral lateral preoptic area and subparaventricularhypothalamus. Two exceptions may be the commissuralpretectal area and the dorsal lateral geniculate nucleus,although that information is difficult to discern. PACAP iswidespread in brain (Hannibal, 2002b) and it is not alwaysclear whether the PACAP observed is actually present inretinal projections or is from other sources.

PACAP immunoreactivity, in addition to identifying projec-tions to the SCN, identifies a small population of retinalganglion cells in the rat most of which are found in thesuperior retina (Hannibal et al., 2002). Most, if not all, of thesecells also contain melanopsin and glutamate. Hannibal et al.(2002) has suggested that all ganglion cells projecting to theSCN contain both PACAP and melanopsin. However, the issue


has not been directly evaluated using retrograde tracersinjected into the SCN. The retrograde tracer procedure hasshown that 10–20% of RHT neurons do not contain melanop-sin (Morin et al., 2003; Gooley and Saper, 2003; Sollars et al.,2003). The apparently close relationship between PACAP andmelanopsin raises the possibility that PACAP released bymelanopsin cell electrical activity may generate prolongedpostsynaptic activity in the SCN that endures beyond the rapidrelease effects of glutamate.

The suggestion that all melanopsin-containing ganglioncells contain PACAP also implies a major role for that peptidein normal photic regulation of circadian rhythms, assuming itacts as a neuromodulator. In the context of light andglutamate effects on circadian rhythm phase, in vivo applica-tion of PACAP antibody or receptor antagonists to the SCNattenuates the phase delaying or advancing effects of light(Bergstrom et al., 2003; Chen et al., 1999). Ordinarily, glutamateapplied to an SCN slice preparation mimics the effect of lighton the phase of the neuron firing rate rhythm with phasedelays and advances induced during the early and latesubjective night, respectively (Ding et al., 1994; Chen et al.,1999).When glutamate and PACAP are coinfused onto the slicepreparation, phase delays are enhanced and phase advancesare abolished. In contrast, the PAC1 receptor antagonist,PACAP6-38, enhanced glutamate-induced phase advanceswhile blocking phase delays (also blocked by PACAP antibodyinfusion). The implication is that glutamate cannot inducephase delays without cofunctioning PACAP, whereas gluta-mate is better able to induce phase advances in the absence offunctional PACAP (Chen et al., 1999).

The role of PACAP in circadian rhythm regulation has beenrecently addressed through the use of PACAP KOmice (Colwellet al., 2004; Kawaguchi et al., 2003). The results aremixed, bothwithin and between studies. Absence of the gene coding forPACAP does not modify rate to re-entrain to an advanced ordelayed LD12:12 photoperiod, phase response to dark pulses atCT6, the circadian period in LL or the negative masking effectof light on wheel running (Colwell et al., 2004). Nevertheless,the same study showed that, in the PACAP KO mice, bothadvance and delay phase shifts to normally saturating ornonsaturating light stimuli administered at CT16 or 23 aregreatly reduced, the circadian period is shorter in DD and thephase angle of entrainment is earlier. The second study(Kawaguchi et al., 2003) failed to find a difference in circadianperiod during DD or an effect of the KO on phase delays to lightat CT15, although the phase advances to light at CT21 wereabout half of the expectation. To make matters moreconfusing, wild-type mice had greater light-induced FOSprotein in the SCN at CT15 than did PACAP KO mice, but notat the time at which light yielded reduced phase advances(CT21). At the present time, it is difficult to reconcile theobservation (Colwell et al., 2004) of (a) a large effect of the geneKO on phase response to light pulses with the absence of aneffect on rate of re-entrainment; (b) the presence of reducedFOS at a time at which the same stimulus yields a normalrhythmphase shift; (c) the presence of a reduced phase shift attime at which the same stimulus yields normal FOS expres-sion and (d) a failure of the two studies (Colwell et al., 2004;Kawaguchi et al., 2003) to find equally deficient phase delaysin PACAP KO mice.

5.6. Substance P

Confusion also abounds with respect to the presence of SP inthe RHT. In contrast to PACAP, there is more positive evidenceregarding function of SP than there is for its anatomicalpresence in the RHT. SP has been suggested as a peptide nativeto retinal ganglion cells projecting to the SCN (Takatsuji et al.,1991; Mikkelsen and Larsen, 1993) and the distribution of SPfibers and terminals in the rat SCN is similar to that for retinalinput (Piggins et al., 2001b). Other attempts to confirm a SPprojection have not succeeded (Otori et al., 1993; Hartwich etal., 1994), and the presence of a SP terminal plexus in the SCNvaries substantially across species (Piggins et al., 2001b). Afairly comprehensive analysis of the rat RHT has demonstrat-ed PACAP and anterograde tracer in the SCN terminal plexusof the RHT, but apparently there is no colocalization with SP(Hannibal and Fahrenkrug, 2002). Similarly, SP does notcolocalize with PACAP in the rat retina. According to onereport, most retinal SP cells are present in the inner nuclearlayer and outer part of the inner plexiform layer. The few SPcells in the rat ganglion cell layer differ in size from the PACAPneurons and are considered to be “displaced amacrine cells”(Hannibal and Fahrenkrug, 2002). This contradicts an earlierreport demonstrating SP-containing ganglion cells identifiedby retrograde label as projecting to the SCN (Takatsuji et al.,1991). However, the results of Hannibal and Fahrenkrug (2002)(Negroni et al., 1997) are consistent with the presence of SP inthe SCN surviving bilateral enucleation (Otori et al., 1993). Asimilar conclusion has been reached by other investigatorsfollowing bilateral enucleation of either rat or hamster(Hartwich et al., 1994).

The SP receptor, neurokinin 1 (NK1), is generally absentfrom the ventral SCN of both rat and hamster (Mick et al., 1994,1995), with low levels of NK1 receptor present in neuronswhich are found along the dorsolateral margin of the rat SCN.Another report (Takatsuji et al., 1995) also indicates a generalabsence of the receptor from the ventral rat SCN, butdemonstrates its presence on a few cells in the most ventralpart of the nucleus and in the underlying optic tract. Most evi-dent receptor is on neurons and dendrites scattered throughthe dorsal and lateral part of the nucleus. Dendrites of thelabeled neurons are described as extending ventrally and, inenucleated animals, NK1 is found in dendrites contactingdegenerating retinal projections via asymmetric axo-dendriticsynapses (Takatsuji et al., 1995). This suggests that SP-contain-ing retinal fibers terminate on cellular elements in the SCNwith NK1 as a receptor. A detailed evaluation of NK1 receptordistribution in the rat, mouse, golden and Siberian hamstersgenerally confirms the above observations as holding acrossspecies (Piggins et al., 2001b). The major point of difference isthe absence of receptor from the ventral SCN in all speciesexcept the rat. In all species, the SCN is surrounded by a clearmargin of dense NK1 receptor and there is a distinct cluster ofcells identifiable by NK1 immunoreactivity just inside extend-ing to just outside the dorsolateral border of the SCN.

SP is present in fibers and terminals of the rat, mouse,golden and Siberian hamster IGL (Morin et al., 1992; Piggins etal., 2001b), although the origin of this input is not known. TheNK1 receptor is densely distributed throughout the entire IGLof the same species (Piggins et al., 2001b).


With respect to function, SP yielded increased 2-deoxyglu-cose uptake in the rat SCN during the subjective night, but notduring the day. It also produced a light-like PRC in the actionpotential rhythm (Shibata et al., 1992; Hamada et al., 1999; Kimet al., 2001). Both effects were blocked by the SP receptorantagonists. Cycloheximide also blocked SP-induced phaseshifts in the slice preparation (Shibata et al., 1994). About 43%of SCN neurons were excited by application of SP to rat brainslices (11% inhibited), but there was no night/day difference(Shirakawa and Moore, 1994). Similar results have been foundin the hamster SCN (Piggins et al., 1995a). In both species, SPmodulated neuronal response to glutamate (Shirakawa andMoore, 1994; Piggins et al., 1995a). SP also induced glutamaterelease in rat SCN slices (Hamada et al., 1999). In optic nervestimulated slices, an SP antagonist reduced excitatory post-synaptic currents in rat SCN neurons. The effect on NMDA andnon-NMDA receptor-mediated responses was similar (Kim etal., 1999). However, SP-induced phase delays could be blockedby NMDA or non-NMDA receptor antagonists, whereas SP

antagonists were unable to block glutamate-induced phasedelays (Kim et al., 2001). The latter results are interpreted asindicating that SP works upstream of glutamate.

Intracranial SP application to the hamster SCN in the earlysubjective night elicited small phase delays. The effects werenot dose-dependent in the pico/nanomolar range andwerenotseenat other circadian times (Piggins andRusak, 1997). Theuseof SP receptor agonists or antagonists has yielded moreconsistent and sometimes larger effects. Light-induced FOSwas significantly reduced by pretreatment with the nonspe-cific SP antagonist, spantide. The reduction was concentra-tion-specific and varied according to SCN region. Greaterreduction occurred in the dorsocaudal SCN than in theventrocaudal SCN and in the rostral SCN than in the caudalSCN (Abe et al., 1996). The authors compare the effects ofspantide to those of NMDA and non-NMDA antagonists withrespect to the distribution of light-induced FOS and concludedthat the glutamate receptor antagonists were involved in lightresponsiveness of the ventrocaudal SCN, whereas SP


contributed to light responsiveness of dorsocaudal SCNneurons. Intraperitoneal injection of an NK1 receptor antago-nist had no effect on hamster rhythms in DD, but in LL, phaseadvances were obtained following treatment at CT8 (Challet etal., 1998a, 2001). These were not associated with increasedwheel running at the time of injection. NK1 antagonists had noeffect on light-induced phase delays at any of three intensities(Challet et al., 1998a, 2001). However, at CT19, they attenuatedlight-induced phase advances by about 30–45%, if adminis-tered prior to the light pulse. The antagonists were unable toblock the effect of light on phase response if administered afterthe light pulse (Challet et al., 2001). The effect of spantide onrhythm regulation in hamsters has apparently not beeninvestigated.

6. SCN structure–function relationships

The structure–function relationship that appears to be emerg-ing is one of both time and space. That is to say, the phases ofintrinsic oscillatory events vary according to SCN region.Attempts to refine the designation, “region,” according to cellphenotype are progressing, butmust be considered as fluid. Asis discussed below, the actual number of “regions” in the SCNis highly debatable (2, 3 or even 4 for the rat; Figs. 3A–E), evenwithin a single species, as is the correspondence of a nominalregion to a particular cell type or organizational attribute.

Perhaps the dominant question pertaining to function ofSCN neurons concerns the extent to which each is anoscillator. The prevailing view, until fairly recently, has beenthat the SCN is composed of many intrinsically oscillatoryneurons which interactively synchronize to yield a mono-phasic rhythm (Herzog et al., 1998; Liu et al., 1997a). Gillette

Fig. 3 – Suggested organization of the SCN varies according tolaboratory, criteria for region designation and the evolution of id“shell” nomenclature (Leak andMoore, 2001). Note fairly substanSCN. (B) Geographical structure using “dorsomedial” (DM) and “vNagano et al. (2003). The area VL′ set off by the dashed line has bdifferently from either the DM or VL, but was not acknowledged bthe variation in density of FOS-IR cells (dots). Drawn from the histFOS-IR cellswithin themiddle of the 3 divisions of (C2) results fromto demarcate a lateral region that did not appear to vary with resperiventricular, dorsomedial central, ventrolateral) derived fromhas been added to set off a region that variesminimally with respdivisions (dorsomedial, ventrolateral main, ventrolateral medial)receives sparse or no retinal innervation (Kawamoto et al., 2003)F1—ZT23, F2—ZT14-light induced) and Per1 (dots; F3—ZT23, F4—(2000). The circular central area was drawn to encompassmost oflight-induced Per1 tends to be absent from the region of rhythmicthe area showing rhythmic Per1 expression (F3—ZT23). (G) Regionfrom histology in (King et al., 2003). The central area in panel G1 wCT0) and retained in panels G2 which shows the widespread expindicated by the location of VP-IR cells and fibers or GRP cells asKaratsoreos et al. (2004). (I) Solid lines diagram themouse SCN “co2004). The dashed line indicates the border of the “core” as preseHAMSTER: (J) SCN and CALB-identifiable central subnucleus (dra(dots) in the rostral SCN at (K1) CT20 and (K2) CT4 or in the caudHamada et al., 2004b), outlines included, with the enclosed ventrin this study).

et al. (1993) provided evidence, using extracellular recordingmethods, that separated dorsomedial and ventrolateral partsof the rat SCN oscillated more or less equally in vitro,although amplitude of the rhythm from the ventrolateralregion was reduced. This early attempt to distinguishregions of functional significance in which one SCN areadiffers from another with respect to oscillatory spikedischarge provided a negative result, but presaged a topicthat has evolved along with the availability of molecularbiological tools. More recently, more refined methods haveallowed greater detail to be gleaned from individual SCNslices (Schaap et al., 2001, 2003). Subtle differences in phaseand waveform are apparent both between the bilateral SCNand across regions within a single rat SCN.

Whereas most attempts to evaluate rhythmicity of SCNcells have been done with reference to location or phenotype,Saeb-Parsy and Dyball (2003) evaluated cells based on theirtargets of projection. Cells projecting to the arcuate and/or thesupraoptic nucleus had two peaks in firing rate, one in theearly night and the other in the late night, in contrast withcells projecting elsewhere.

As described below (Section 6.1.3), there is accumulatingevidence that at least some cells may not oscillate. However,attempts to identify individual SCN pacemaker cells, whilesuccessful in doing so, have not been able to provideinformation regarding the percentage of SCN cells that areintrinsically oscillatory or their particular cell characteristics(Herzog et al., 1997, 1998; Liu and Reppert, 2000, Liu et al.,1997a;Welsh et al., 1995; Aujard et al., 2001). Three generalitiesarise from these studies (Liu and Reppert, 2000). One is that thebehavior of oscillatory cells evinces a wider range of intrinsiccircadian periods if the cells are isolated from one anotherthan if they are coupled. The second is that GABA is able to

species, rostrocaudal location within the SCN, time of day,eas. RAT: (A1 is rostral to A2) Structure using the “core” andtial differences in position that depend on the level within theentrolateral” (VL) designations. Drawn from histology ineen added to indicate a portion of the SCN that behavedy the authors. (C1, C2) Divisional structure as suggested fromology in (Beaulé et al., 2001b). Note that the smaller number oflight exposure. The dashed line in panels C1 has been addedpect to FOS induction. (D) Three divisions (dorsomedialthe expression of Per1 (Yan and Okamura, 2002). Dashed lineect to Per1 expression. Compare to panels C1 and C2. (E) Threederived from the expression of VIP-IR and Per1. The VLmed. MOUSE: (F) Expression of Cry1-IR (dots;ZT14-light induced). Drawn from histology in Field et al.the Cry1-IR cells and held constant in panels F2–F4. Note thatCry1 expression (F1—ZT23), but has substantial overlap withs indicated by antiphase rhythmic expression of Per2. Drawnas drawn to encompass most of the Per2-IR cell nuclei (dots;ression of Per2 in nuclei (dots) at CT12. (H) Compartmentsindicated by a green fluorescent protein signal. Drawn fromre” and “shell” according to one formulation (Yan and Silver,nted in another formulation (Abrahamson and Moore, 2001).wn from Hamada et al., 2001). (K) Cells expressing VP mRNAal SCN at (K3) CT20 and (K4) CT4 (drawn from histology ofolateral area designating the region containing CALB neurons


shift the phase of individual, noninteracting SCN neurons in aphase-dependent manner. Third, the phases of oscillationsamong a collection of independent cells are synchronized byrepeated bath application of GABA at the same time every day.There is no information concerning the cell phenotypewhether from hamsters (Herzog et al., 1998), rats (Welsh etal., 1995) or mice (Liu and Reppert, 2000; Liu et al., 1997a).

The ability to evaluate the simultaneous rhythmic behav-ior of multiple individual cells has provided a new perspec-tive on the oscillatory organization of the SCN. In mouse SCNwith a green fluorescent protein (GFP) reporter of Per1 geneactivity (Kuhlman et al., 2000), most neurons oscillate insynchronized fashion, but with numerous distinct phases(Quintero et al., 2003). An estimated 11% (LD) and 26% (DD) ofthe cells evaluated were judged as nonrhythmic. Unexpect-edly, cells tend to be grouped, having one of three preferredphases that differ if the animals have been entrained or free-running. In SCN obtained from mice in LD, 78% of neuronshad peak GFP expression at ZT5, ZT8 or ZT11 (most cellspeaked at ZT8). With a history of DD, most cells peak at CT12and 92% had peaks in the CT12, CT15 or CT18 phase groups.The half-maximum phase plot indicates CT12 for the DD-housed animals and ZT8 for the LD housed animals. Severalvaluable observations are evident in this study (Quintero etal., 2003): (1) not all SCN neurons are rhythmic; (2) differentrhythmic neurons have different phases; (3) there are threepreferred phases of rhythmic SCN neurons, a characteristicpreviously observed using multiunit in vivo electrophysio-logical methods (Meijer et al., 1997); (4) the preferred phase ofrhythmic neurons varies according to photoperiod history.The results support the conclusion that, assuming clockgenes actually determine the pacemaker properties of SCNneurons, other factors control the phase relationship be-tween the locomotor rhythm and circadian clock phase. Thisis demonstrated by the fact that the preferred phase of geneexpression is about 4 h earlier in LD housed mice comparedto DD housed mice when assessed relative to activity onset(Quintero et al., 2003).

An exception to the above evidence concerning the timingof oscillations in SCN neuronal activity has been observed inthe hamster. Whereas there is a neuronal activity rhythmpeak recorded approximately once every 24 h if the SCN tissueis cut in the coronal plane, there are two distinct peaks if theSCN is cut in the horizontal plane (Jagota et al., 2000). Thephases of the two peaks aremodulated by photoperiod history(Mrugala et al., 2000). This phenomenon appears to apply onlyto hamsters, and has not been seen in mouse or rat (Burgoonet al., 2004). These results support the view that there arefunctionally distinct regions within the SCN; different regionsoscillate differently; and the “standard” perspective (here, thecoronal view) is not always the fully correct and adequateperspective.

6.1. Function of cellular neuromodulators in the SCN

6.1.1. GABAAs indicated above, GABA is the predominant neurotransmit-ter in cells of the rat, mouse and hamster SCN. It is, therefore,not a surprise that manipulation of GABA neurotransmissionin the SCN alters circadian rhythm function. One potential

complication to the interpretation of the functional analysesof GABA in the SCN is the presence of a GABAergic afferentprojection from the IGL (Morin and Blanchard, 2001).

The basic observation is that peri-SCN administration ofthe GABAA agonist, muscimol, substantially reduces themagnitude of light-induced phase shifts, while similar appli-cation of the GABAA antagonist, bicuculline, substantiallyincreases phase shift magnitude (Gillespie et al., 1996). Theseresults are inconsistent with prior data obtained usingsystemic drug injections (Ralph and Menaker, 1987, 1989).Nevertheless, the effects of intracranial (i.e., SCN) drugapplication appear to be repeatable and consistent, althoughthe effect of bicuculline may depend on circadian time(Gamble et al., 2003; Mintz et al., 2002; Gillespie et al., 1997).Similarly, bicuculline augments and muscimol suppresseslight-induced Fos protein in the hamster SCN (Gillespie et al.,1999).

The GABAB receptor may also be involved in modulation oflight-induced phase shifts. Peri-SCN injections of a GABAB

receptor agonist, baclofen, block light-induced phase delaysand advances, but the antagonist, CGP-35348, has no effect(Gillespie et al., 1997). Baclofen also blocks glutamate-inducedphase delays and this action is not affected by TTX (Mintz etal., 2002). The GABAB agonist also effectively blocks light-induced Fos at CT13.5 and CT19 (Gillespie et al., 1999),although an antagonist has no effect. The effect of GABAB

receptor activation is the result of an inhibition of glutamaterelease from the RHT terminals (Jiang et al., 1995).

The GABAA agonist, muscimol, is also able to induce phaseadvances when administered to DD-housed hamsters duringthemiddle subjective day (Novak et al., 2004; Smith et al., 1989)and this effect is inhibited by a concomitant light pulse (Mintzet al., 2002).

6.1.2. VIP and VIP/PACAP receptorsVIP neurons are clustered in the ventral SCN of mostmammalian species and use VIP as a neuromodulator tocommunicate with other neurons in the SCN through activityat the PAC1 or VPAC2 receptor, both of which are found in theSCN (Cagampang et al., 1998a,b). A complicating factor in theanalysis of receptor function in the context of the circadiansystem is the fact that both receptors bind VIP and PACAPwithapproximately equal affinity. PACAP-containing terminals arefound in projections from the retina, as well as from other,unidentified brain regions (Hannibal and Fahrenkrug, 2004;Hannibal et al., 1997). VIP peptide has been implicated in theregulation of circadian rhythm phase and period, althoughconflicting observations have been reported (Gillespie et al.,1996; Piggins et al., 1995b; Albers et al., 1991). More recent workhas focused on function of the PAC1 and VPAC2 receptors withrespect to circadian rhythm regulation and the results havebeen quite provocative.

Calcium imaging methods suggest that the effects ofPACAP on SCN neurons are elicited through the PAC1 receptorbecause of the absence of response to equimolar VIPapplication (Kopp et al., 1999). The PAC1 knockout mousetends to have a shorter circadian period in DD than wild-typeand light-induced phase shifts are less after stimulation atcertain circadian times compared to thewild type (Hannibal etal., 2001b). Fos protein and Per1/2 gene induction were also


much less in the knockout animals following a light pulseduring the early subjective night, although they were notdifferent during the late subjective night.

The effects of PAC1 receptor knockout are small comparedto the consequences of losing a functional VPAC2 receptor.The latter appears to be an essential component of thecircadian rhythm system, at least in mice. Its absence in theVipr2 KOmouse is associated with nearly complete loss of thecircadian locomotor rhythm assessed in constant dim redlight (Cutler et al., 2003). Similarly, rhythms in SCN geneexpression (Per1, Per2, Cry1, VP were lost, although a vastlyattenuated, but statistically significant, rhythm in BMAL1expressionwas evident (Harmar et al., 2002; Cutler et al., 2002).In the Vipr2 KO mice, there was also very little light-inducedFOS protein compared to what is normal in the wild type. Inthe SCN slice preparation, the ensemble circadian rhythm inneuronal firing rate is lost in the Vipr2 KOmouse (Cutler et al.,2003). An additional feature of this study was the demonstra-tion that a specific VPAC2 receptor antagonist had the sameeffect on the ensemble neuronal firing rhythm of wild-typeanimals as the gene defect does on the Vipr2 KO animal. Incontrast to the gene knockout model, a transgenic mouseoverexpressing the VPAC2 receptor has a much longercircadian period and appears to be more sensitive to light asindicated by faster re-entrainment to an advance of the light–dark photoperiod (Shen et al., 2000).

At the level of individual SCN cells, there is also an absenceof rhythmic gene expression in Vipr2 KO mice (Harmar et al.,2002). An explanation for the function of the VPAC2 receptorwithin the context of circadian clock structure and intercel-lular pacemaker regulation has not been determined, al-though the suggestion has been made that rhythmic VIPserves to maintain coupling between oscillating neurons inthe SCN (Harmar et al., 2002). This explanation does notaccount for the fact that about 15% of the congenitallyanophthalmic ZRDCT-Anmouse strain are arrhythmic despitethe presence of VIP neurons in the SCN (Laemle andOttenweller, 2001). Thus far, there is a demonstrable require-ment for the presence of the VPAC2 receptor in order for miceto display circadian rhythmicity, but no demonstration thatVIP itself is explicitly necessary or greatly involved in rhythmgeneration or regulation.

Nearly half themouse SCNneurons immunoreactive for VPalso have the VPAC2 receptor as do about 31% of the VIP andCalR neurons (Kallo et al., 2004). In general, there is a highdegree of regional correspondence between the distribution ofmouse neurophysin (marker for VP) and for PHI-containing(marker for VIP) neurons with the distribution of cellscontaining VPAC2 receptor (King et al., 2003), although thedegree of colocalization was not established in this paper. Theregional correspondence is not perfect as a central zone in themouse SCN is devoid of VIP or VP neurons, but contains a fewcells with VPAC2 receptors.

The specific importance of VIP for circadian rhythms hasbeen emphasized in mice (Colwell et al., 2003; Aton et al.,2005). VIP neurons project widely from the SCN to varioustargets (Kalsbeek et al., 1993), but also appear to have intra-SCN connections (LeSauter et al., 2002; Daikoku et al., 1992;van den Pol and Gorcs, 1986). Two thirds of cells in themouse SCN have spontaneous inhibitory postsynaptic

potentials that are enhanced by VIP. These cells aredistributed throughout the SCN with no discernible region-related pattern (Itri and Colwell, 2003). VIP infused into thehamster peri-SCN region or onto the rat SCN slice during theearly subjective night induces phase delays, while infusionsduring the late subjective night yield phase advances, with aPRC similar to that for light (Piggins et al., 1995b; Albers etal., 1991; Reed et al., 2002) (although there is controversyconcerning whether or not PHI and GRP are necessary forobtaining phase shifts). In mice lacking genes coding for PHIand VIP, a variety of rhythm-related measures were altered.In particular, the circadian period was shorter by about anhour, phase response to light was greatly reduced oreliminated, there was more variability in activity onsetunder LD and cohesive rhythms were generally not main-tained under DD (Colwell et al., 2003). Mice lacking only thegene coding for VIP had similar characteristics. Most notably,about two thirds of the animals had running rhythms in DDthat expressed multiple noncircadian components. Impor-tantly, the behavior of VIP KO mice was strikingly similar tomice lacking the gene (Vipr2) coding for the VPAC2 receptor(Aton et al., 2005). The importance of VIP to normalrhythmicity in the VIP KO mice has been established byreinstating SCN rhythmicity and synchrony among SCNneurons by periodic application of VIP agonist to a tissueculture preparation, but this has no effect on VPAC2 KO mice(Aton et al., 2005).

6.1.3. Calbindin and the central subnucleus of the hamsterSCNThere has been emphasis on the central subnucleus of theSCN and the role its constituent CALB cells may play incircadian rhythm regulation. A subset of SCN neuronsexpresses the Ca2+ binding protein CALB (Arvanitogiannis etal., 2000; LeSauter and Silver, 1999). CALB contains six high-affinity Ca2+ binding domains and contributes to intracellularCa2+ signaling and buffering (Leathers et al., 1990; Roberts,1994). As described above, the CALB neurons are clusteredwithin the caudocentral SCN. This region contains about 250CALB neurons and is approximately 300 μm long, 180 μm highand 130 μm wide (Silver et al., 1996a).

Partial lesions of the SCN that completely destroyed theCALB region yielded animals with arrhythmic locomotoractivity. In contrast, if the lesions spared a portion of one orboth CALB regions, behavioral rhythmicity was maintained(LeSauter et al., 1996). The circadian locomotor rhythm isrestored after implantation of fetal SCN tissue into the thirdventricle of previously SCN-lesioned adult hamsters (Ralph etal., 1990). Restoration of rhythmicity requires the presence ofCALB neurons within the transplanted grafts and the strengthof the restored locomotor rhythm is correlated with thenumber of surviving CALB neurons (LeSauter and Silver,1999). Small lesions of the SCN that damage the CALB regionbilaterally eliminate a number of behavioral and physiologicalrhythms. If the CALB region on either side was partially orcompletely spared, the rhythms remained intact (Kriegsfeld etal., 2004). These data suggest that the region of the SCNcontaining these CALB positive neurons is important in thegeneration of circadian rhythms. An important question to beanswered is whether it is the CALB neurons (or other neurons


in this region) or whether it is the network connectionsbetween the cells in this region that are critical for thegeneration of circadian rhythms.

The central subnucleus receives dense input from theretina via the retinohypothalamic tract and the CALB neuronsare directly contacted by retinal projections (Muscat et al.,2003; Bryant et al., 2000). A light pulse during the subjectivenight increased FOS expression in the majority of CALBneurons of the hamster SCN. These data led to the develop-ment of a model in which the CALB neurons serve as a “gate”for photic input to the SCN (Antle et al., 2003a). This modelmust be reconciled with anatomical data demonstrating thatthe RHT projects to all regions of the hamster SCN and the factthat the densest retinal terminal field extends well beyond theCALB subnucleus (Muscat et al., 2003). In contrast to results inthe hamster, CALB and FOS colocalization are rare in the grassrat (Mahoney et al., 2000). Destruction of CALB neurons in therat SCN does not alter light-induced FOS expression, free-running rhythms or photic entrainment in that species(Beaulé and Amir, 2002).

Neurons in the hamster central subnucleus do not show arhythmic expression of messenger RNAs for Per1 and Per2(Hamada et al., 2001, 2004a). However, light can induce theexpression of both these clock genes in the central subnu-cleus. This has led to the conclusion that the centralsubnucleus is a nonrhythmic region of the SCN, while thedorsomedial region corresponding to the general location ofVP neurons is rhythmic because the cells in this area expressPer1 and Per2 in a rhythmic manner. However, it may not bepossible to determine the rhythmicity of individual neuronswith in situ hybridization techniques if only a portion of theneurons in a region are rhythmic. Further, electrophysiolog-ical evidence suggests that at least some neurons in thecentral subnucleus fire action potentials in circadian manner.The mean spontaneous firing rate (SFR) of non-CALB neuronsin the central subnucleus has a peak at ZT6 and a nadir at ZT19(Jobst and Allen, 2002). Whether the rhythmicity of these non-CALB neurons is driven by endogenous clocks located inneurons of the central subnucleus or whether it is due anetwork phenomena of interconnections between rhythmicand nonrhythmic neurons is an important area for futureresearch. In contrast, CALB neurons had a mean SFR of lessthan 1 Hz and showed no circadian pattern of action potentialfiring (Jobst and Allen, 2002). It is noteworthy, but of unknownsignificance, that some CALB neurons contact a neighboringCALB or non-CALB cell via gap junctions (Jobst et al., 2004). Inthe mouse, increased firing frequency is correlated withincreased expression of Per1 (Kuhlman et al., 2003). Thissuggests that CALB neurons are a functionally distinctsubpopulation within the SCN that do not contain a rhythmiccircadian clock. The data raise the intriguing possibility thatthe interactions between rhythmic and nonrhythmic neuronsare required to generate circadian locomotor rhythms. Thereis also the untested possibility that the GRP, SP and CALRneurons that occupy the central subnucleus along with theCALB cells are likewise nonrhythmic.

CALB expression is dynamic depending on the genotype,developmental age and the lighting conditions. Tau hamstershave significantly more CALB expressing neurons than wild-type hamsters (LeSauter et al., 1999). In mice, there is

significantly more CALB expression in young animals thatgradually decrease as the animals approach adulthood (Ikedaet al., 2003).

CALB expression is inversely correlated with the durationof the light cycle (Menet et al., 2003). Hamsters maintained onshort duration photoperiods had significantly more CALBexpression than those maintained on long photoperiods(Menet et al., 2003). Hamsters reared in DD or short photope-riod conditions had more CALB than controls (Tournier et al.,2003). Similarly, rats maintained in LL conditions had fewerCALB-containing neurons then in rats maintained in DD(Arvanitogiannis et al., 2000). Neonatal monosodium gluta-mate (MSG) exposure increased the number of CALB-expres-sing SCN neurons and prevented the disruption of circadianrhythms produced by LL in rats (Arvanitogiannis et al., 2000).The parallel effects of MSG on CALB and LL changes onrhythms suggest that CALB may be important in the regula-tion of response to LL.

CALB appears to play a key role in the photic responses inhamsters. CALB is present in the cell nucleus and cytoplasmof central subnucleus SCN neurons during the day, but isabsent from the cell nucleus during the night (Hamada et al.,2003). Reduction of CALB expression by antisense oligonu-cleotides blocked light-induced phase shifts during thenight, but light could produce shifts of both behavioral andPer rhythms during the day (Hamada et al., 2003). Interest-ingly, these effects occurred despite the fact that the CALBmRNA was only reduced 38% and the CALB protein onlyreduced 28% (Hamada et al., 2003). Also interesting was therapidity of the effect, which occurred only 4 h after theantisense injections. Therefore, while the CALB neurons donot have rhythmic clock gene expression or show rhythmsin action potentials, CALB expression in the nucleus may becontrolled in a circadian manner. Thus, these cells appear tobe rhythmic according to at least one criterion, but not byothers. The mechanisms controlling this cellular regulationare unknown. They may involve rhythmic feedback from thedorsomedial “rhythmic” SCN or may result from theinteraction between rhythmic and nonrhythmic neurons inthe central subnucleus.

The central subnucleus is heterogeneous and CALB cellsconstitute a minority of the neurons (LeSauter et al., 2002;Jobst and Allen, 2002). A better understanding of mechanismsunderlying intercellular communication between non-CALBand CALB neurons and how they respond to synaptic inputwill be important to understand how the SCN generates afunctional circadian output. Double-label methods havedemonstrated that 94% of CALB neurons also containedGABA (Jobst et al., 2004). Of note, most of the cellular profilesin the central subnucleus contain GABA, althoughmany GABAneurons in the central subnucleus did not contain CALB. Inaddition to CALB and GABA colocalization in the centralsubnucleus, glutamate decarboxylase 67 (GAD67), one of tworelated isoforms of the synthetic enzyme for GABA, wasdetected in CALB cell bodies. Immunoreactivity for the GABAA

receptor α2 and β2/3 subunits was prominent in the SCN (Jobstet al., 2004). Confocal microscopic analysis demonstrated thatβ2/3 and α2 did not colocalize with CALB. Labeling for the β2/3subunits was diffusely distributed around CALB neurons,whereas labeling for the α2 subunit appeared as discrete


puncta. The data indicate that α2 and β2/3 subunits contributeto the composition of GABAA receptors in the CALB region. Alimitation of this study is that the GABAA receptor subunitswere not localized to either postsynaptic or presynapticregions. GAD65, an isoform of GAD exclusively located withinsynaptic terminals (Erlander et al., 1991), is expressed in apunctate pattern throughout the SCN, including the centralsubnucleus (Jobst et al., 2004).

6.1.4. Bombesin and gastrin releasing peptideBombesin contains 14 amino acids, GRP has 27 and they sharea 7 amino acid sequence homology. There is a very smallliterature concerning the effects of bombesin on SCN function.Immunoreactive cells are abundant in the Djungarian ham-ster (Reuss, 1991), although cross-reactivity with GRP is apotential problem (Panula et al., 1984). Bombesin increased thefiring rate of about 50–60% of golden hamster neurons studied,while inhibiting the rate in half of the remainder (Piggins andRusak, 1993; Piggins et al., 1994). Neuromedin B, another in theclass of bombesin-like peptides, activated about 40% of SCNneurons and inhibited very few (Piggins et al., 1994). Theresponses of individual cells to application of several differentbombesin-like peptides were similar in more than 80% of thetests. Responses to the peptides were attenuated by pretreat-ment with antagonists to the GRP-preferring receptor which ispresent in the SCN (Ladenheim et al., 1992, 1993). More cellswere activated by GRP and neuromedin B at ZT12-16 than atZT4-8 (Piggins et al., 1994). Bombesin has not been similarlytested.

GRP was originally reported to induce light-like phaseshifts in hamster only if infused onto the SCN along with VIPand PHI (Albers et al., 1991), but may be equivalentlyfunctional if applied alone (Piggins et al., 1995b). The actionpotential rhythm in rat or hamster slice preparationsexhibited large, light-like phase responses to GRP andthese effects could be blocked by bombesin 2 receptorantagonists. Bombesin 2 and bombesin 1 receptors wereidentified in SCN, although there is no evidence thatbombesin 1 contributes to rhythm regulation (McArthur etal., 2000). GRP injected into the hamster third ventricle at CT13 induces sizable phase delays in association with induc-tion of Per1, Per2 and FOS (Antle et al., 2005). Interestingly,the gene expression occurs almost exclusively in the denselyretinorecipient region in the dorsal SCN, just above thecentral subnucleus identified by CALB cell presence (Muscatet al., 2003). This is the same area in which pERK isexpressed, dependent upon an intact RHT, during thesubjective night (Lee et al., 2003). And, phosphorylation ofERK is apparently necessary for phase response to GRPbecause if phosphorylation is pharmacologically blocked, thephase shifts are also blocked (Antle et al., 2005). Althoughspontaneously rhythmic ERK phosphorylation does notoccur in the absence of the RHT, the effect of enucleationon phase response to GRP has not been examined.

In the mouse, GRP infused into the peri-SCN region atCT16 elicited small phase delays and induced Per and Fosgene expression (Aida et al., 2002). GRP receptor was alsofound throughout most of the SCN, although density wasclearly greatest in the dorsocaudal part of the nucleus. Micelacking the GRP receptor had normal phase delays in

response to a 15 min 30 lx light pulse at CT16, but unlikewild-type animals, the delays were not greater if the light was300 lx. Small (approximately 20%) decrements were also seenin Per2 (but not Per1) and Fos response to the 300 lx light (Aidaet al., 2002).

In ovariectomized, estrogen-primed rats, pituitary prolac-tin is released in a daily surge timed by the SCN (Mai and Pan,1993, 1995; Freeman et al., 2000). Properly timed bombesininfusion into the peri-SCN region blocks the prolactin surge(Mai and Pan, 1995). The implication is that bombesin neuronsparticipate in the efferent control of prolactin release timing.This raises the general issue of whether individual SCN cellphenotypes are concerned with the control of single ormultiple efferent systems.

6.1.5. Neuromedin U and neuromedin SNeuromedin U (NMU) is a 24 amino acid peptide that shares a 7amino acid sequence homology with the somewhat largerneuromedin S (NMS) (Mori et al., 2005). NMU and its receptors,NMU-R1 and NMU-R2, are both present in the rat and mouseSCN (Mori et al., 2005; Graham et al., 2003, 2005; Nakahara etal., 2004). NMU and NMU-R1 have high amplitude rhythmicgene expression that persists in DD, whereas rhythmicexpression of the NMU-R2 gene is lower amplitude (Grahamet al., 2003; Nakahara et al., 2004). NMU is present in some, butnot all, VP neurons and some, but not all, NMU-containingneurons also have VP; it apparently does not colocalize in VIPneurons (Graham et al., 2003). The distribution of NMU in rat ismost similar to the distribution of VIP neurons, but is presentthroughout much of the SCN. NMU-R1 and R2 receptordistributions are similar to that for the peptide. Phaseadvances of about an hour followed ventricular injection ofNMU at CT0 into rats; injections at CT6 yielded phase delays ofabout 0.6 h; and from CT8 to CT23, there was no phaseresponse. FOS protein expression was widely induced byintraventricular NMU (Graham et al., 2003). Mice lacking thegene coding for NMU have rhythmic feeding behavior, but thephase angle of entrainment may be altered (Hanada et al.,2004).

NMS is distributed in the rat SCN in a pattern similar to VIPand there was rhythmic expression peaking at ZT11 (Mori etal., 2005). The rhythm was lost in DD and a light pulse acutelyreduced expression. Intraventricular injection of NMS at CT23induced 1 h phase delays and, at CT6, about 2 h phaseadvances with a dead zone between CT9 and CT22. FOSexpression in dorsal SCN cells was also induced by NMS. ThePRC for NMS was similar to that for NPY (Albers and Ferris,1984; Albers et al., 1984), whereas that for NMU appeared to besomewhat different. The general shape of the PRC for NMSwas very much light-like, but was shifted such that phasedelays began at CT24 with a cross-over point at CT3 and thephase advance region ended at approximately CT8-9 (Naka-hara et al., 2004).

6.1.6. VasopressinThe Brattleboro rat is deficient in VP, yet has circadianrhythms (Ingram et al., 1998). It has been concluded that VPis not necessary for circadian rhythm expression. In fact,transplanted SCN from Brattleboro rats will restore normalcircadian rhythmicity in arrhythmic hosts (Boer et al., 1999).


Although VP is not a necessity for rhythm expression inBrattleboro rats, transplanted SCN from normal animals canrestore rhythmicity in paraventricular hypothalamic neuronsand this rhythmicity is likely to be the result of VP stimulation(Tousson and Meissl, 2004). The issue is rendered morecomplicated by the demonstration that rhythmicity of VPcells evaluated in vitro shows advances and delays accordingto circadian time of VIP stimulation. Interestingly, the samestudy failed to observe a phase shifting effect of glutamatewhich would be expected to act either directly on the VP cellsor on the VIP cells projecting to the VP cells (Watanabe et al.,2000).

6.1.7. Diurnality–nocturnalityIn the grass rat, a diurnal rodent, the above effect of GABAagonists administered during the subjective day may occur atphases opposite to those effective in the nocturnal hamster(Novak and Albers, 2004a). In particular, muscimol injected inthe peri-SCN region at this time induces phase delays ratherthan phase advances and these effects are not blocked by TTX.Opposite to the effect in nocturnal species, phase shiftscaused by peri-SCN muscimol are not inhibited by simulta-neous light exposure (Novak and Albers, 2004b). However,nocturnal and diurnal species appear to respond similarlywith respect to the ability of GABAA and GABAB agonists toinhibit phase response to light (Novak and Albers, 2004b;Novak et al., 2004). Assuming that the action of injected GABAagonists is occurringwithin the SCN, then the antiphase GABAactivity in diurnal vs. nocturnal animals is the first indicationthat the SCN contributes to this species-specific behavioraltrait.

In a subparaventricular region contiguous to the dorsalSCN, the lower subparaventricular zone, the grass rat hashigher expression of FOS during the night than during theday, an effect not observed in rats (Smale et al., 2001;Schwartz et al., 2004; Nunez et al., 1999). Moreover, the day–night pattern of expression is roughly opposite that whichoccurs in the SCN of either the Norway lab rat or grass rat.The suggestion has been made that the neuron firing raterhythm can become inverted in the peri-SCN region fornocturnal (Inouye and Kawamura, 1979), but not for diurnal(Sato and Kawamura, 1984), species and may contribute tothe manifestation of nocturnal or diurnal activity traits(Smale et al., 2003).

6.2. Acetylcholine

Cholinergic innervation of the SCN has been of interest formany years and there are numerous reports of cholinergicreceptors on SCN cells and modification of circadian rhythmfunction by cholinergic agonists or antagonists (see Morin,1994 for a review). Several anatomical reports have describedcholine acetyltransferase (ChAT) immunoreactive processesin the SCN, although the histology has not been robust and thedemonstrable fibers have been located in somewhat differentplaces, depending upon the particular study (van den Pol andTsujimoto, 1985; Bina et al., 1993; Ichikawa and Hirata, 1986;Kiss and Halasz, 1996). Cholinergic fibers have also beenshown to make contacts directly on SCN neuron (Kiss andHalasz, 1996). Perhaps the most convincing evidence of

cholinergic input to the SCN is the demonstration that,following an intra-SCN injection of retrograde tracer, labeledcells containing ChAT are found in several brain nucleiincluding the lateral dorsal and pedunculopontine tegmen-tum, areas concerned with sleep regulation (Bina et al., 1993).Retrogradely labeled cholinergic cells were also found in thebasal forebrain. Excitotoxic lesions of the nucleus basalismagnocellularis destroyed about 75% of cholinergic cells inthis location yielding a decrease in cholinergic processes inthe SCN (Madeira et al., 2004). Bilateral nucleus basalis lesionsalso diminished the numbers of countable VP and VIP cells inthe SCN by about 30% and reduced volume andmRNA levels ofVP and VIP cells by about 34%.

The immunotoxin, 192 IgG-saporin, destroys cells bearingthe low affinity nerve growth factor receptor, p75NTR, whichare generally cholinergic (Leanza et al., 1995). Intraventricu-lar infusion of the immunotoxin abolished p75NTR in the ratSCN without affecting VIP neurons (Moga, 1998). However,intrahypothalamic immunotoxin injection into the rat great-ly reduced numbers of CALB neurons (Beaulé and Amir,2002). The circadian period length was related to the extentof cell loss. Typically, the period was much shorter inlesioned rats and about a third of them became arrhythmic.Animals sustaining immunotoxin lesions of cholinergic cellsalso had about a 50% reduction in light-induced FOS cellcounts. In another study (Erhardt et al., 2004), 192 IgG-saporin was injected into the ventricle or directly into theSCN, but neither procedure yielded deficits in CALB or VIPneurons and changes in circadian period or general rhyth-micity were not observed. Nevertheless, the immunotoxingenerally reduced the ability of light to produce phase delaysor advances. The rhythm-related effect of 192 IgG-saporintreatment may be related to disruption of effective retinalprojections to the SCN. Unilateral enucleation reducesoptical density of p75NTR in the contralateral SCN by about50% (Bina et al., 1997).

As reviewed (Morin, 1994; O'Hara et al., 1998), the cholin-ergic agonist, carbachol, has generally yielded light-like phaseresponses. However, the topic is not without controversy. Forexample, peri-SCN carbachol induced robust phase delayswhen injected at CT14 and advances at CT22, but alsoproduced large phase advances in about half the animalsreceiving the drug at CT6 (Bina and Rusak, 1996). The phaseshifts during the subjective night are consistent with a light-like action, but the shifts during the day are not. The responseswere blocked by pretreatment with atropine, but not bymecamylamine, suggesting that a muscarinic receptor med-iates the cholinergic effects. The functional presence ofmuscarinic receptors in the SCN region is supported byample additional data, particularly from SCN slice preparationstudies (Liu and Gillette, 1996; Liu et al., 1997b; Artinian et al.,2001) and there is evidence supporting a rhythm functionspecifically for the M1 receptor (Gillette et al., 2001).

Despite the foregoing, the nicotinic receptor may alsomediate the effects of acetylcholine on circadian rhythmicity.The nicotinic antagonist, mecamylamine, blocked the phaseshifting effects of light at CT19 and greatly attenuated light-induced FOS in the dorsomedial hamster SCN with littleinhibition more ventrally (Zhang et al., 1993). In the rat slicepreparation, the PRC to nicotine was phase-independent with


advances occurring throughout the circadian day and thephase shifts could be blocked by mecamylamine (Trachsel etal., 1995). In vivo, nicotine administered subcutaneously atCT16 elicited significant phase delays and induced FOS in theSCN and mecamylamine greatly attenuated FOS expression(Ferguson et al., 1999). A preliminary description of ratresponse to nicotine indicated that phase shifts were unreli-able following systemic administration, but were light-likefollowing nicotine intraventricular infusion (O'Hara et al.,1998).

There are also reports that a strain of rats bred forcholinergic supersensitivity has a shorter circadian periodthan controls, an altered phase angle of entrainment andreduced FOS induction in SCN neurons (Shiromani et al., 1988;Ferguson and Kennaway, 1999).

6.3. Neurophysiology of the SCN action potential rhythm

Action potential firing by SCN neurons is regulated as acircadian rhythm. Higher frequencies are observed during thesubjective day than during the subjective night with themeanpopulation firing frequency peaking near CT6 (Gillette, 1991).During the subjective day, action potential firing frequenciesrange from less than 1.0 Hz to 14 Hz, compared to thesubjective night when action potentials fire with frequenciesof 0.5 to 5.0 Hz. The presence of fast firing neurons during theday is responsible for the increased mean population firingfrequency observed in SCN slices (Gillette, 1991). This circadi-an rhythm of action potential firing is a property of individualneurons because the firing difference ismaintainedwhen SCNneurons are grown in dispersed cell cultures (Welsh et al.,1995; Herzog et al., 1997; Abe et al., 2000; Honma et al., 1998a;Shirakawa et al., 2000).

A current hypothesis proposes that themolecular circadianclock sends signals to regulate the activity of ion channelsresponsible for setting the membrane potential and actionpotential firing rate. This hypothesis predicts that disruptionof clock gene function would alter the action potential firing ofSCN neurons. Mutations of clock genes alter the period ofaction potential firing rhythms. Microelectrode array record-ings of slice cultures from Clock homozygous mice demon-strated neurons with a period near 27 h that was significantlylonger than neurons of wild-type or Clock heterozygous mice(periodnear 24h) (Nakamura et al., 2002). In a similar study, theperiod of Clock heterozygous mice was longer than that of thewild type mice (24.5 vs. 23.3 h), while the Clock homozygousmice were arrhythmic (Herzog et al., 1998). In the tau mutanthamster,whichhas a shortened circadianperiod, the circadianrhythm of action potential firing is also shortened to approx-imately 20 h (Liu et al., 1997a). SCN neurons from mice withmutations of the cryptochrome 1 or cryptochrome 2 genes arearrhythmic (Albus et al., 2002). Also consistent with thismodelare the observations of McMahon and colleagues who used atransgenic mouse expressing a green fluorescent proteindriven by the Per1 promoter to demonstrate that the frequencyof action potential firing was linearly correlated with Per1transcription (Quintero et al., 2003).

SCN neurons express a wide repertoire of ion channels andregulation of the activity of these ion channels by signals fromthe molecular clock would lead to regulation of action

potential firing. The presence of a diffusible messenger issupported by the observation that spontaneous firing of SCNneurons “washes out” during the dialysis of the intracellularmilieu that occurs during whole-cell recording (Schaap et al.,1999). Currently, the signaling pathways connecting themolecular clock in the nucleus to changes in membranepotential are unknown. However, significant progress hasbeenmade toward describing the ion channels responsible forthe regulation of spontaneous firing.

Manymembrane properties of SCN neurons show a diurnalor circadian variation. SCN neurons have input resistances inthe range of 1–3 GΩwhen measured with whole cell recordingtechniques. The input resistance of SCN neurons is higherduring the day compared to the night in slices maintained ineither LD or DD (De Jeu et al., 1998; Kuhlman and McMahon,2004). The resting membrane potential of SCN neurons hasbeen reported to be more depolarized during the day thanduring the night (De Jeu et al., 1998; Kuhlman and McMahon,2004; Pennartz et al., 2002). These differences in membranepotential are maintained in rats maintained in DD consistentwith their regulation by the circadian clock (Kuhlman andMcMahon, 2004). However, a significant day–night differencein the membrane potential has not always been observed(Schaap et al., 1999; Kim and Dudek, 1993; Teshima et al.,2003). Difficulties in estimating the “resting” membranepotential of spontaneously firing neurons and dialysis ofintracellular contents may contribute to these differentresults. The resting membrane potential has been proposedto be set, in part, by a Ba2+-sensitive, fast-activating, outwardlyrectifying K+n current (De Jeu et al., 2002). Interestingly, thiscurrent did not show a circadian rhythm in activity, althoughgiven the high input resistance of SCN neurons, a smallchange in this current could significantly change the restingmembrane potential. The membrane potential of some SCNneurons has a low frequency (2–7 Hz) oscillation that ismediated by L-type Ca2+ channels and is observedwhen actionpotentials are blocked by TTX (Jiang et al., 1997; De Jeu et al.,1998; Pennartz et al., 2002; Jackson et al., 2004).

Action potentials in SCN neurons are characterized by arapid upstroke of membrane depolarization (∼+25 mV) fol-lowed by membrane potential repolarization. An afterhyper-polarization (AHP; ∼−80 mV), a membrane potential thathyperpolarizes below the resting membrane potential andwhich is dependent on extracellular Ca2+, follows the actionpotential upstroke (Cloues and Sather, 2003; Thomson, 1984;Thomson and West, 1990). The rapid upstroke of the actionpotential is driven by current flowing through voltage-gatedNa+ channels and is completely blocked by TTX (Thomson andWest, 1990; Huang, 1993; Wheal and Thomson, 1984).

During the interval between individual action potentials(interspike interval), the membrane potential shows a steadydepolarization that reaches spike threshold (Kononenko et al.,2004; Pennartz et al., 1997). This interspike interval can bemodulated by synaptic input, particularly inhibitory postsyn-aptic potentials generated by GABAergic input (Pennartz et al.,1998; Kononenko and Dudek, 2004). These “depolarizingramps” are sensitive to the Na+ channel blocker, TTX, and toriluzole, a compound used to treat amyotrophic lateralsclerosis, that blocks a persistent Na+ current. Voltage-gatedNa+ channels rapidly activate and inactivate during a


membrane depolarization. However, many neurons have asmall Na+ current that shows very slow or no inactivation (forreview see Crill, 1996). This “slowly-inactivating” or “persis-tent” Na+ current (INa,S or INa,P) is TTX-sensitive and mayrepresent current flowing through a noninactivating class ofvoltage-gated Na+ channels or a modal variation of thenormally transient Na+ current. In SCN neurons, a slowlyinactivating Na+ current (in I = 50–250ms at − 45mV) is blockedby TTX, and riluzole similar to the INa,P. The INa,Smay representa third modal state of voltage-gated Na+ channels (Jackson etal., 2004).

It has been hypothesized that the INa,S is primarilyresponsible for driving the membrane depolarization duringthe interspike interval (Kononenko and Dudek, 2004; De Jeuand Pennartz, 2002). Jackson et al. (2004) recorded from acutelyisolated dorsomedial SCN neurons and used action potentialwaveforms as voltage signals under voltage clamp to deter-mine the currents flowing during the interspike interval. Theionic currents flowing during the action potential waveformare the same as those that flow during the action potentialfiring. The majority of the current during the interspikeinterval was carried by TTX-sensitive Na+ current. A nimodi-pine-sensitive Ca2+ current was also present during the duringthe interspike interval. However, the amplitude and chargecarried by the Ca2+ current are only 26% of that carried by theNa+ current (Jackson et al., 2004). Further, the Ca2+ currentturns on later in the interspike interval than the Na+ currentand is not significant until the membrane potentialapproaches threshold. The authors conclude that the slowlyinactivatingNa+ current is primarily responsible for setting theaction potential frequency while contribution of voltage-gatedCa2+ currents to the action potential firing frequency is small(Jackson et al., 2004).

Other candidate ion currents have been proposed tocontribute to setting the action potential frequency (Kim andDudek, 1993; Pennartz et al., 1997). The H-current (IH), anonselective cation channel activated by membrane depolar-ization has been hypothesized as contributing a depolarizingcurrent, but a direct test of this hypothesis did not yield anyevidence for an IH contribution (Akasu et al., 1993; De Jeu andPennartz, 1997). The A-current (IA), a fast-activating, rapidlyinactivating K+ current present in all SCN neurons, maycontribute to setting the duration of the interspike interval(Huang, 1993; Bouskila and Dudek, 1995). The low-threshold T-typeCa2+ currenthas also beenproposedas a contributor to thesetting of action potential frequency (Akasu et al., 1993). Nodirect evidence for the IH or IT has been presented.

The AHP that follows membrane potential repolarizationalso contributes to setting action potential firing frequency.The AHP duration correlates with the spontaneous firing rate,the slower theAHPdecay, the slower the firing rate (Cloues andSather, 2003). If theAHP is blocked, then the rate of firingof SCNneurons increases. In SCN neurons, the AHP is dependent onCa2+ entering the cell via predominately L- and R-type voltage-gated Ca2+ channels that activate both large-conductance(BKCa) and small-conductance (SKCa) Ca2+-activated-K+ chan-nels. Blockade of apamin-sensitive SKCa channels (only SK2subunits in SCN) has been reported to produce an increase orno effect on spontaneous action potential frequency. Thedifference probably reflects a diversity of SKCa expression in

SCN neurons. Block of iberiotoxin-sensitive BKCa did not alterfiring frequency. An apamin- and iberiotoxin-insensitive KCa

current was diurnally modulated and contributes to the spikefrequency (Cloues and Sather, 2003). The AHP is dependent onboth small conductance (SKCa) and large conductance (BKCa)Ca2+-activated-K+ channels.

Abundant experimental evidence demonstrates that Ca2+

plays an important role in the activity of SCN neurons. First,removal of extracellular Ca2+ abolishes the circadian pattern ofaction potential firing (Shibata et al., 1984). Second, some SCNneurons have a regular pattern of action potential firingcharacterized by small deviations in the interspike interval(Thomson, 1984). Reduction of the extracellular Ca2+ concen-trationdisrupts this regular firing pattern,whereas an increasepotentiates theactionpotential spike amplitudeand the rate ofmembrane repolarization (Thomson andWest, 1990). Removalof Ca2+ or block of Ca2+ channels with high concentrations ofextracellular Mg2+ reduces the spike amplitude, rate ofmembrane potential repolarization and the amplitude of theAHP (Thomson and West, 1990). Third, there is a circadianrhythm of intracellular Ca2+ levels with higher levels duringthe subjective day thanduring the subjectivenight (Ikeda et al.,2003; Colwell, 2000). The Ca2+ rhythm is not blocked by TTXalthough action potentials are completely eliminated. Ryano-dine, which depletes intracellular Ca2+ stores, significantlyblocked both the intracellular Ca2+ rhythm and disrupted theaction potential firing (Ikeda et al., 2003).

GABA is the most prominent neurotransmitter in the SCN(Moore and Speh, 1993; Morin and Blanchard, 2001) and thereare abundant inhibitory postsynaptic potentials that can berecorded in SCN neurons (Jiang et al., 1997; Kim and Dudek,1992). Activation of GABAA receptors is traditionally thought toproduce cellular inhibition due to activation of a chlorideconductance and membrane hyperpolarization or the shunt-ing of excitatory currents. However, in the SCN, descriptions ofthe actions of GABA on the SCN neuronal activity have beencomplicated and controversial. Themajority of recordings hadshown GABA to be an inhibitory neurotransmitter during theday (Bos and Mirmiran, 1993; Gribkoff et al., 1999; Liou andAlbers, 1991; Liou et al., 1990). However, other investigatorsdescribed the activation of GABAA receptors to be primarilyexcitatory, producing an increase in firing activity or mem-brane depolarization during the day reducing action potentialfiring and producing a membrane hyperpolarization duringthe night (Wagner et al., 1997, 2001). This diurnal rhythmicitywas proposed to be caused by a daily change in theintracellular chloride concentration (Wagner et al., 1997,2001). A high daytime intracellular chloride concentrationwas proposed to account for the observations. An excitatoryGABA response would be observed when the equilibriumpotential of Cl− (ECl) is less negative than the restingmembrane potential. Contrary to those results, a differentgroup of investigators reported that GABAA receptor activationproduces only inhibitory effects during the day and excitatoryeffects in a subpopulation of SCN neurons during the night (DeJeu and Pennartz, 2002). The reversal potential during thenight was more depolarized than during the day consistentwith an increased intracellular chloride concentration. Thespecific reasons for these dramatically different observationsremain unknown. Technical issues such as style of whole cell


recording, duration of GABA application and the GABAconcentrationmay all contribute to the different observations.

6.4. Inter-SCN differences in gene expression

One of the more intriguing developments during the last 10years has been the observation that the two SCN canoscillate out of phase (De la Iglesia et al., 2000). In malehamsters with split rhythms caused by LL exposure, theputative circadian clock gene, Per1, is maximally expressedin one SCN while another clock gene, BMAL1, is simulta-neously expressed maximally in the other SCN (normally,Per1 and BMAL1 are expressed about 180° out of phase andmaximal expression cannot be seen simultaneously; Honmaet al., 1998b; Shearman et al., 2000). High Per1 expressionnormally occurs during the middle subjective day, acharacteristic that is also true in SCN from hamsters withsplit rhythms. The genes coding for VP and FOS proteins arealso expressed rhythmically in normal animals at the samephase as Per1. In the animals with split rhythms, activity ofthese genes remains temporally synchronized with that forPer1. Despite the continuous presence of light, photicallyinduced Per or FOS expression does not appear to occurduring the subjective night (De la Iglesia et al., 2000). Thisobservation is curious because it implies that a photic actionnormally occurring in DD-housed animals in is beingblocked, at least in the case of animals with split rhythms.Apparently, LL-housed animals with unsplit rhythms havenot been tested to determine if there is light-induced geneexpression during the subjective night. It would also seempossible, although perhaps unlikely, that the rhythmic geneexpression observed in SCN of animals with split rhythms isinduced by a temporally gated action of light rather thanbeing an expression of intrinsic cellular rhythmicity.

The observation that the two SCN may oscillate inantiphase has raised the possibility that each SCN mayhave specific control over the timing of individual physio-logical events. This has been examined by evaluating thecontrol of the gonadotropin, luteinizing hormone (LH), inanimals with split running rhythms. Ordinarily, estrogen-treated, ovariectomized female hamsters have a single dailyLH peak timed by a circadian clock (Alleva et al., 1971;Bittman and Goldman, 1979), but in animals with splitrhythms, there are two such LH peaks in antiphase (Swannand Turek, 1985). The circadian clock action occurs throughregulation of pituitary LH release by septal-preoptic neuronssynthesizing gonadotropin releasing hormone (GNRH) (De laIglesia et al., 1995). Activation of GNRH neurons, as indicatedby FOS expression, is maximal at the time of LH release(Urbanski et al., 1991; Doan and Urbanski, 1994). Whenestrogen-treated, ovariectomized hamsters have splitrhythms in LL, there is daily FOS expression in GNRH cellspredominantly (about 5:1) localized to the side of theforebrain in which SCN FOS expression is also dominant(about 5.5:1) (De la Iglesia et al., 2003). The implications ofthis form of lateralized circadian organization for regulatoryphysiology are far-reaching and there are preliminary datasupporting the view that each SCN controls the ipsilateralexpression of rhythmicity in peripheral organs (Yamazaki etal., 2000).

6.5. Regional expression of Fos

6.5.1. HamsterAn early observation (Rea, 1992) was that light did nothomogeneously induce FOS expression throughout the ham-ster SCN. Rather, the protein was heavily induced in theventral half of the nucleus both at CT13 and CT18, but therewas also large induction of FOS in the dorsal half at CT18. Thisstudy lacked no-light controls, but the result has beencorroborated (Chambille et al., 1993; Guido et al., 1999). FOSinduction by light is similar in the dorsal and middle thirds ofthe hamster SCN at both ZT15 and ZT22 (no response andlarge response, respectively) and in the ventral third (verylarge responses at both times) (Guido et al., 1999). Also, the FOSresponse to light is greater in the ventral third of the SCN thanin the other two thirds at both ZT15 and ZT22. Much of theventral third of the SCN, in which dense, light-induced FOSoccurs, corresponds to the central subnucleus that can beidentified by the presence of CALB neurons and dense retinalinnervation. One point of the Guido et al. (1999) results is thatFOS induction by light occurs throughout the SCN, not just inthe central subnucleus.

The effects of GABA agonists on facilitation or inhibition oflight-induced FOS are proportionally equivalent in rostral,middle and caudal thirds of the SCN (Gillespie et al., 1999).When FOS counts weremade through an entire SCN section atthe midcaudal level, there was no response at all to a 1 minlight exposure at ZT15, whereas the same stimulus at ZT18, 21or 24 yielded a near maximal result. A longer exposure to lightat ZT15 greatly increased the FOS response, but never to thelevel seen at the other times. A complementary analysis(Chambille et al., 1993) has additionally demonstrated thatwhile light at CT14 induces a small increase in FOS in therostral 3/4 of the SCN, there is substantial FOS induction in thisarea when light is administered at CT19. The above-citedpapers demonstrate a region by time interaction with respect togene induction by light stimulation, with region being definedby rostrocaudal or dorsoventral level. Further, the Guido et al.(1999) work indicates that the sensitivity of SCN neurons tophotic input varies according to time during the subjectivenight. Implicit, but not studied, is the likelihood of region bytime interaction with respect to sensitivity to light. Theobservations of region by time interaction with respect tolight-induced FOS expression have not had an impact onmorerecent developments focused on photic regulation of putativeclock genes.

6.5.2. RatFos gene expression spontaneously oscillates in the SCN withmost of the amplitude accounted for by the expression in thedorsomedial region (Schwartz et al., 1994; Koibuchi et al.,1992). However, a low amplitude oscillation is present in theventrolateral region (Sumova et al., 1998). Regional FOSoscillation persists in vitro (Prosser et al., 1994; Geusz et al.,1997). Multiple Fos-family genes are light-induced in theventrolateral region of the rat SCN. These include c-fos, fra-2and fosB. It has been suggested that each gene is a candidatemediator of light effects on SCN neurons and regulation ofcircadian rhythm phase that may enable reasonably normalrhythm function in the absence of normal gene activity


(Schwartz et al., 2000; Honrado et al., 1996). A double labelstudy has shown that only a small percentage of VP cellscontain spontaneously expressed FOS and that a similarlysmall percentage of cells expressing FOS also contain VP(Sumova et al., 2000). The daily profile of spontaneouslyexpressed FOS is altered according to the length of theentraining photoperiod, a phenomenon with implications forphotoperiodism (see Sumova et al., 2004 for a review).

Analysis of light-induced FOS expression in various SCNregions has been rendered more complex by the observationthat light can inhibit it (Beaulé et al., 2001b). In the rat, lightinduces large increases in FOS expression in neuronsthroughout the ventral SCN. As indicated above, the dorsome-dial region does not show light-induced gene expression, butdoes have constitutive FOS that varies according to circadiantime. A particularly novel result in this study is the demon-stration of an actual drop in the number of cells expressingFOS in an apparently specialized sector of the dorsomedialSCN (Fig. 3C1, 2) following light administered at CT2 or CT14,or in entrained animals, at ZT1 or ZT2 (Beaulé et al., 2001b).

6.6. Regional clock gene expression

6.6.1. MouseThe advent of digoxigenin procedures for in situ hybridiza-tion has permitted analysis of light-induced clock genes atthe cellular level. As indicated by these procedures, lightapplied at CT16 induced Per1 mRNA in the ventral, but notthe dorsomedial, mouse SCN (Shigeyoshi et al., 1997). Incontrast, the dorsomedial region exhibited a circadianoscillation of Per1 expression with a peak near CT4 andsimilar rhythmicity in Per1 expression was present through-out the mouse SCN. A two component model of SCNoperation was proposed in which the SCN is organized intodorsomedial “type A” cells and ventral “type B” cells(Shigeyoshi et al., 1997). The latter are strongly influencedby light, while the former are light-independent. Similarresults have been obtained in the rat (Yan et al., 1999).

In the Per1-GFP mouse, the extent of regionality appears todepend upon the lighting history. In LD housed animals,approximately equal numbers of cells rhythmic for Per1 werefound in the medial and lateral halves of the SCN (Quintero etal., 2003). In contrast, in DD housed animals, about 66% of themedial neurons were rhythmic compared to 34% in the lateralSCN. There were also differences in the dorsoventral distribu-tions of rhythmic cells, with twice as many found in theventral SCN of LD housed animals compared to similarpercentages in the two halves of DD housed mice. Asdescribed above, the Per1-GFP mice have rhythmic neuronswith peak activity at one of three preferred phases. In DDhoused animals, the vast majority of cells peaking at CT18 arepresent in the medial SCN, whereas the medial–lateraldifference becomes progressively less at CT15 and CT12,respectively. A similar relationship between peak phase andlaterality is evident in LD housed mice, but is not aspronounced (Quintero et al., 2003). In organotypic slicecultures of SCN from mice carrying the Per1promoter-drivenluciferase reporter gene, most cells identified by the reportershow oscillatory gene expression and these cells are widelydistributed throughout the SCN (Yamaguchi et al., 2003).

In the organotypic slice preparation, luminescence rhythmpeaks of Per1 expression in cells of the dorsalmouse SCN havea broad distribution of phases, but most are maximal inadvance of those in middle and ventral thirds of the nucleus(Quintero et al., 2003). The cells rhythmically expressing Per1 inthe dorsal third are unable to collectively sustain a cohesiverhythm, unlike those more ventrally located. It is noteworthy,however, that each dorsal cell continues a seemingly normaloscillation uncoupled from the others in the same dorsalregion. This suggests that themore ventral oscillatory cells areimparting a signal to those in the dorsal SCN, therebycontrolling phase of rhythmic cells in that region. In addition,the investigators note that results from the horizontal slicepreparation indicate the dorsal neurons are not necessary tomaintain either rhythmicity of the ventral cells or theiroscillatory synchrony (Yamaguchi et al., 2003). The presenceof preferred phases by regionally specific collections ofoscillatory neurons (Quintero et al., 2003; Yamaguchi et al.,2003) is consistent with the view described for hamsters(Hamada et al., 2004a) (discussed below), although theinterpretation may be substantially different. On the onehand, discussion of mouse SCN tends to emphasize thepresence of many oscillatory neurons collectively synchro-nized, but with different phases of peak Per1 gene expression.On the other, the hamster work describes the “slow spread ofgene expression” from one area to another. The appearance of“spread” may be the consequence of what has been describedin mice (Yamaguchi et al., 2003), namely, the presence ofoscillatory neurons with grouped or a gradient of phases. Themouse data have an implication for the direction of causation.The dorsal oscillatory neurons cannot maintain synchronywhen separated from those more ventrally located. At least inthe mouse, this strongly suggests that the direction ofcausality with respect to phase angle and synchrony amongoscillators is from themore ventral rhythmic cells to the dorsalrhythmic cells.

Other studies of the mouse SCN note a central region (Fig.3F3) with greatly reduced rhythmic Per1 and Per2 expression(Karatsoreos et al., 2004; LeSauter et al., 2003; King et al., 2003;Reddy et al., 2002). This area corresponds closely to thedistribution of GRP neurons (note: the GRP cells were identifiedin a CALB-GFP reporter mouse in which, for some unknownreason, about 90% of the GFP cells in the SCN also containedGRP, and for an equally unknown reason, no cells in the SCNknown to contain CALB were identified by GFP (Karatsoreos etal., 2004). However, the most central part of the regioncontaining putative GRP cells has distinct expression of bothPer1andPer2at ZT2orCT2, timeswhen rhythmic expression isminimal elsewhere in the SCN (King et al., 2003). Previously, asimilar sized, fairly discrete collection of neurons in thecentrodorsal (not dorsomedial) mouse SCN was described asexpressing Per1 at ZT0 or CT0 (Hastings et al., 1999).

With respect to light-induced clock gene expression in themouse, a light pulse at CT13-14 has been reported to causePer1 expression throughout the central and dorsolateral SCNin a region complementary to the medial area in whichrhythmic Per1 is expressed (Karatsoreos et al., 2004). There is ahigh degree of correspondence between light-induced Per1and the GRP cell distribution. The same appears true for FOSexpression. The region of light-induced Per1 and GRP cells is


completely filled with retinal projections which tend to bemuch sparser in themedial part of the SCN in which rhythmicPer1 or Per2 expression predominate. However, another reportindicates that light-induced Per1 is found in the ventral thirdof the SCN corresponding largely to the area containing PHIneurons (King et al., 2003). This is consistentwith observationsthat VIP neurons containing Per1 jump from 31% to 59%following a light pulse at ZT21 (Kuhlman et al., 2003). Incontrast, Per1 is expressed in about 35% of VP neurons at thistime, but is not elevated in response to light. VIP and VP cellsexpressing Per1 comprise about 51% of the total neuronsspontaneously expressing that gene at ZT10, indicating thatthere is no specific correspondence between cell phenotypeand rhythmic clock gene activity. The information alsoindicates that rhythmic gene expression is not solely theprovince of dorsomedial neurons, as the VIP cells areconsistently present in the ventral part of the SCN.

Light-induced Per2 protein in the SCN is described (Kinget al., 2003) as being expressed with breadth and magnitudeequivalent to that which occurs at its maximal rhythmicexpression. One specific characteristic of the rhythmicexpression is its relative absence from the GRP cell region(Karatsoreos et al., 2004; LeSauter et al., 2003; King et al.,2003) (Figs. 3G1, 2; H). Cells identifiable as containing theVPAC2 receptor had both rhythmic and light-inducedexpression of Per2 (King et al., 2003). There appear to besome differences in the results of two comparable mousestudies (Karatsoreos et al., 2004; King et al., 2003), but it isdifficult to determine whether or not they are real or simplythe result of substantially different methods.

6.6.2. RatA comprehensive analysis of Per1 and Per2 gene expression inthe entire rat SCN has produced a modification of thedorsomedial/ventrolateral SCN divisional model to includethree functionally distinct regions (Yan and Okamura, 2002).The single dorsomedial region has been divided into periven-tricular and central parts (Fig. 3D). The former overlaps a fairlythin, dorsomedial portion of the SCN identifiable by VPneurons. The central portion overlaps the ventrolateral partof the VP cell group as well as the dorsal part of the VIP celldistribution. The third, ventrolateral, portion of the SCNgenerally corresponds to the bottom 80% of the regioncontaining VIP cells. The three regions are reported to bedistinct with respect to their expression of Per1. For example,the peak expression in the periventricular division is at CT4,but at CT8 in the central division. Overall, the profiles of Per1expression are different in the three regions. A seconddistinguishing feature is the Per1 response to light which, forthe periventricular and central divisions, is nonexistent atCT16, while extremely robust for the ventrolateral division. Incontrast to Per1, rhythmic expression of Per2 is similar in theperiventricular and the central divisions, thereby indicating adifferent SCN divisional structure.

An analysis of Per2 protein has shown rhythmic expressionin dorsomedial and ventrolateral rat SCN, but no induction ineither location after a light pulse at CT13 (Beaule et al., 2003a).

It has been conceptually convenient to consider the rat SCNas having two parts, the dorsomedial and ventrolateral“divisions.” Moreover, it has also been convenient to view

the ventrolateral division as the sole SCN region with retinalinnervation and VIP neurons, while the dorsomedial divisionhas no retinal innervation and contains VP neurons. However,convenience must give way to a more complex reality assuggested by the presence of two functionally distinct parts ofthe dorsomedial rat SCN (Yan and Okamura, 2002). Similarly,when the ventrolateral region containing VIP neurons hasbeen examined closely, it also can be subdivided intodifferently functioning units. VIP cells identify an ellipticalarea occupying the ventral half of the rat SCN. However, cellsin the medial quarter of this region do not express light-induced Per1 mRNA nor, as demonstrated in adjacentsections, does this area receive much, if any, retinal innerva-tion (Kawamoto et al., 2003). Thus, the nominal ventrolateralrat SCN containing VIP cells is divisible into a lateral part inwhich Per1 is induced by light and a medial part in which thecells do not have this response (Fig. 3E).

The foregoing results are of significance because theydemonstrate, at least for the rat, that (a) cells throughout theentire SCNoscillate, although themagnitude is greater in someregions than in others; (b) there are regional distinctionsbetween the rhythmic expression of Per1 and Per2; (c) there areat least four functionally distinct regions within the SCN; (d)the SCN appears to have a compartment in which cells rapidlyexpress Per genes in response to light, but this doesnot occur inthe other three compartments and (e) there is a large, butclearly imperfect, anatomical correspondence between knowncell phenotypes and the three regions. In addition, it isnoteworthy that the ventrolateral division, unlike the others,does not extend the entire length of the SCN (Yan andOkamura, 2002).

One obvious burden imposed on investigators by theincreased number of SCN subdivisions is the need tominimizeconfusion by using a common nomenclature characterized byfully understandable definitions. It is difficult to comparestudies when, in one study, a region may be referred to as the“VP” area, a second study refers to the same area as the“dorsomedial region,” while a third study emphasizes adifference between the “medial dorsomedial” and the “lateraldorsomedial.” The difficulty is compounded by the fact that,because the information is so new, the laboratories makingmost of the anatomical observations have yet to achieve astandard nomenclature across publications.

6.6.3. HamsterA functional analysis of SCN divisions has also been made inthe hamster, although without the same level of detail. A two-compartment description has been demonstrated, with oneunresponsive to light and the other highly responsive to light(Hamada et al., 2001). The shape of the hamster SCN lendsitself to a descriptive terminology more closely related topeptide content of cells than to the geographic descriptors ofdorsomedial vs. ventrolateral. Thus, there is more emphasison the central subnucleus and the fact that cells in this regionbehave differently than the more dorsomedial cells. As in therat, there is a pronounced endogenous rhythm in Per1 andPer2 gene activity among neurons closely associated with thedistribution of VP cells in the hamster SCN. As a generality, thedistribution of these cells wraps around the medial half of thecollection of CALB neurons in the central subnucleus. Cells in


the CALB-containing region express the Per1 and Per2 genesrobustly in response to light. It is also the case that neurons inthe CALB cell region do not have endogenous oscillations inPer1, Per2 or Fos gene expression unlike their more medialcounterparts (Hamada et al., 2001). It is not known to whatextent the Per1 or Per2 gene expression occurs rhythmically inVP neurons or is light-induced in CALB neurons.

Further analysis of the hamster SCN has demonstratedthat while themost robust Per gene response to light occurs incells in and near the CALB cell region, a similar phenomenonoccurs among cells in the dorsomedial SCN, the region of VPneurons (Hamada et al., 2004a). The time course of light-induced Per1 gene expression in cells of this region issignificantly longer than in the CALB region, suggesting analtogether different pathway for induction. Nevertheless, theresult is consistent with the smaller number of retinalprojections to the dorsomedial region than to the centralCALB cell region in the hamster (Muscat et al., 2003). It is alsothe case that the VP gene itself not only has an intrinsicoscillatory expression in SCN neurons, it is also light-induciblein a slow onsetmanner similar to that for the Per1 gene in cellsof the VP region (Hamada et al., 2004a).

Rhythmic expression of VP mRNA in the hamster SCNappears to have at least two characteristics: there is alwaysexpression in the dorsomedial region (minimal at CT20), and isgreatly increased in that region at CT8, while also appearingthroughout a much larger part of the entire SCN (Hamada etal., 2004a). Per gene expression has similar rhythmic expres-sion characteristics, but Per1 and Per2 mRNAs are virtuallynonexistent at CT20. The investigators suggest that rhythmicgene expression is initiated among a small number ofdorsomedial cells and spreads slowly through more caudaland ventrolateral parts of the SCN.

There is clear intra-SCN organization to the temporal andspatial rhythmic expression of the Per and VP genes. However,in an oscillatory system where there is no beginning and noend, it is not clear that an origin of the gene expression rhythmcan be defined. More important is the issue of whether cells inone part of the SCN transmit neural or humoral rhythmicinformation to other cells that is requisite for the expressionor phasing of rhythmicity in those cells. Related to this issue isthe question of cellular “gradients” in gene expression withinthe SCN (Hamada et al., 2004a; Yan and Okamura, 2002). A“gradient” implies gradual spatial change, a view contrary toobservations of rhythmic gene expression occurring withinspecific sectors of the SCN (see Fig. 3 in Yan and Okamura,2002). The observations of temporal and spatial rhythmic or-ganization within the SCN are likely to be of great importancewith respect to understanding the movement of informationwithin the SCN itself as well as to the understanding offunction attributable to the various sectors of the SCN.

To our knowledge, there has not been an investigation oflight-induced clock gene expression comparable to the com-prehensive study of FOS by Guido et al. (1999). However,possibly unlike the situation in the rat (Yan and Okamura,2002), a CT19 lightpulse first induces Per1 expression inanareaof the hamster SCN encompassing the central subnucleus, butthis is followed, with time, by induction of gene expression incells throughout most of the dorsomedial region as indicatedby the location of VP neurons (Hamada et al., 2004a). Whether

there is a “periventricular” part of the hamster SCN, as there isfor the rat (Yan andOkamura, 2002), remains to be determined.

6.7. Stimulus induction of gene expression

6.7.1. Photic stimuliRea (1989) showed photic induction of FOS protein in SCNneurons in 1989. During the past decade, there have beenmany other investigations directed at the expression of avariety of genes thought to be components of the circadianclock. Substantial information exists which shows rapidinduction of Per1 and Per2, but not Cry1, Cry2 or Per3, geneexpression following a light pulse (Shigeyoshi et al., 1997; Bestet al., 1999; Field et al., 2000; Zylka et al., 1998; Shearman et al.,1997). Brief (15 min) light pulses at ZT14 yielded Per1 and Per2gene expression increases that were 26 and 30%, respectively,greater than controls when measurements were made 1h later. However, the changes in mRNA levels occurredwithout a noticeable increase in protein which is not evidentuntil many hours later (Field et al., 2000). These resultsemphasize the investigators' point that there is not a reliablerelationship between the phase shift magnitude and the levelof light-induced Per gene expression.

In a novel experiment, Paul et al. (2003) tested whetherglutamate receptors mediate light-induced suppression ofnocturnal pineal melatonin as well as light-induced expres-sion of Per1 mRNA. As expected, both nocturnal environmen-tal light and the glutamate receptor agonist, NMDA, applied tothe SCN suppressed pineal melatonin and elevated Per1mRNA. Joint application of AMPA or NMDA receptor antago-nists with the NMDA blocked the effect of light on Per1 mRNAexpression, but in the same animals, the antagonists wereunable to prevent the light-induced reduction in melatonin.This study demonstrates the existence of two different path-ways in the SCN. Both are light-responsive, with one presum-ably related to the regulation of circadian rhythm phaseresponse and the other contributing to photic inhibition ofnocturnal pinealmelatonin. The study also raises an issue thatshould be of concern to all rhythm investigators. That is,whether the changes observed in the SCN are directly relatedto the phenomenon under investigation or simply correlates.This issue has been extended to evaluation of photic inductionof the clock genes, Per1 and Per2 (Paul et al., 2004). The resultsindicate that both light and NMDA induced the expected geneexpression in the SCN and also suppressed pineal melatonin.Blockade of sodium-dependent action potentials with TTXblocked both light-induced gene expression and melatoninsuppression consequent to either light or NMDA. In contrast,induction of Per1 and Per2 by NMDA was not blocked by TTX(Paul et al., 2004). The results of the above studies areconsistent with the view that there are two distinct mechan-isms regulating light-induced gene expression and light-suppression of pineal melatonin.

6.7.2. Nonphotic stimuliWhereas light during the subjective night induces Per geneactivity in the SCN and causes rhythm phase shifts, it haslittle, if any, effect on these two measures during thesubjective day. In contrast, phasically active nonphoticstimuli administered during the subjective day elicit rhythm


shifts when applied during the subjective day which isassociated with decreased Per gene activity in the SCN.Several classes of nonphotic stimuli have this effect,including benzodiazepine treatment (Yokota et al., 2000),novel wheel activity (Maywood and Mrosovsky, 2001),injection-related arousal (Maywood et al., 1999), peri-SCNinfusion of NPY (Maywood et al., 2002; Fukuhara et al., 2001)and serotonin receptor agonists (Horikawa et al., 2000).Antisense oligonucleotide to the Per1 gene administereddirectly into the peri-SCN region during the subjective daysuppresses Per1 mRNA expression by about 60% (Hamada etal., 2004b). This treatment at CT6 induces a phase advancethat averages about 60 min, a result consistent with the viewthat enhanced Per1 expression, in this case caused by anonphotic stimulus, induces phase shifts.

A proposed unifying hypothesis suggests that elevatedPer gene expression and Per protein synthesis is causal tophase shifts induced either by light or by nonphotic stimuli(Maywood and Mrosovsky, 2001). Moreover, the hypothesismakes an unconventional specific prediction that lightshould be able to modify circadian clock function evenduring the “dead zone” of the PRC. The rationale is that,during the normal dead zone, light cannot alter phasebecause Per expression is already high and further expres-sion cannot be elicited by light during the subjectivedaytime. However, light can be shown to alter SCN activityduring the subjective day because light exposure at that timeprevents the drop in Per gene activity normally induced bynonphotic stimuli (Maywood and Mrosovsky, 2001) or directNPY infusion into the peri-SCN region (Maywood et al., 2002).The converse of this procedure is also demonstrable withnonphotic stimuli presented during the subjective night ableto attenuate light-induced phase shifts (Mistlberger andAntle, 1998; Ralph and Mrosovsky, 1992).

According to the Maywood and Mrosovsky (2001) unifyinghypothesis, nonphotic stimulus should achieve its attenuatingeffects on light during the subjective night by reducing theextent of light-induced Per gene activity. However, direct testsof this proposition have yielded mixed results. In the firstinstance, benzodiazepine treatment produced an approxi-mately 50% reduction in light-induced (5 lx, 15 min, CT20)phase advances, but had no effect on phase delays (Yokota etal., 2000). The effect on phase advanceswas associatedwith anapproximately 40–50% drop in Per1 or Per2 mRNA (thecorresponding test for phase delays was not performed). Incontrast, access to a novel running wheel reduced light-induced phase delays without altering light-induced phaseadvances, but the novel wheel activity had no effect on light-induced Per1 mRNA (Christian and Harrington, 2002). In asimilar investigation, access to a novel wheel for 1 h after alight pulse resulted in virtual abolition of phase advances inresponse to less intense illumination and substantial attenu-ation of response tomore intense illumination. However, as inthe above study, therewasno reduction in either Per1mRNAorFOS protein associated with the reduced phase response tolight (Edelstein et al., 2003a). As noted by the authors, Per1appears to be insufficient for phase shifts in locomotorrhythms and that, “The relationship between light andnonphotic stimuli is more complex than has been previouslysuggested (Maywood and Mrosovsky, 2001)” (see Edelstein et

al., 2003a for a discussion that considers a variety of issuesrelated to the above inconsistencies).

6.7.3. Inconsistencies between phase response and geneexpressionLight-induced Fos gene activity is closely correlated with theability of light to entrain circadian rhythms, but there is ageneral consensus that FOS is not an essential part of eitherthe clock or the input pathway mediating entrainment(Colwell et al., 1993a,b,c). Moreover, FOS protein induction inresponse to light is usually, but not always, predictive ofphase response. Quipazine injection at CT8 is significantlyreduced FOS expression in hamster SCN, but there was noeffect on phase shifts at this time and in mice, this drug hadno effect on phase response, but increased FOS expressionduring the subjective day (Antle et al., 2003b). In miceentrained to a skeleton photoperiod of 1 h light pulsespresented to create a 1:14:1:8 LDLD photoperiod, someanimals ran during the long, and others during the short,interval of dark. The light manipulations revealed that onlythe light pulse occurring near activity onset (ZT12) appearedto influence locomotor rhythm phase. Irrespective of theeffect on entrainment, among animals running during thelong dark interval, light pulses at the beginning or the end ofthat interval equally induced Fos mRNA expression. More-over, if animals ran during the short dark interval, Fos mRNAexpression was induced by both pulses, but was much greaterafter the pulse at ZT21 (Schwartz et al., 1996). Thus, in thisparadigm, there is a clear disparity between the entrainingcapability of light and the magnitude of Fos gene expressionin response to light. Molecular absence of Fos does notprevent entrainment of mice (Honrado et al., 1996). In vivoblockade of Fos availability is associated with an inability oflight to induce phase shifts in rats (Wollnik et al., 1995).However, JunB gene activity was simultaneously blocked inthis study and the negative effects on light-regulated phasecontrol may have been the consequence of the suppression ofthe activity of both genes, rather than only Fos (see Schwartzet al., 2000 for a discussion).

Edelstein et al. (2003a) suggest that their data fromnonphotic stimulus conditions are consistent with the viewthat several differentmechanisms contribute to the regulationof circadian phase depending on the nature of the stimulusinvolved. Most importantly, they conclude that phase shiftscanoccurwithoutnecessarilymodulating Per geneexpression.This is contrary to the general view that phase shifts aredependent upon Per gene activity, although not at allinconsistent with the available literature. It is also the casethat Per1 and Per2 expression is not the same in the SCN ofanimals showing rhythm splitting in LL as it is in animalsshowing a form of induced splitting in DD referred to as“behavioral decoupling” (De la Iglesia et al., 2000; Edelstein etal., 2003b). In the latter case, “splitting” consisting of a dailybout of induced locomotion is followed by a resting intervalduring which Per gene expression is increased throughoutmost of the SCN. However, when the elevated gene expressionoccurred during the rest phase following bouts of spontaneouslocomotion, Per gene expressionwas also elevated, but only inthe dorsomedial part of the nucleus (Edelstein et al., 2003b).Also, despite having several characteristics similar to


spontaneous splitting in LL conditions, the DD-housed behav-iorally decoupled animals had no indications of bilateralasymmetry in Per gene expression in the two SCN, unlike thesituation in animals showing LL-induced splitting (cf., De laIglesia et al., 2000).

There are several forms of dissociation between theregulation of gene expression and the phasic effects oflight on circadian rhythms. Perhaps the simplest is thetemporary loss of synchrony between gene expressionrhythms during clock resynchronization to a shifted photo-period. The re-entrainment rate of certain gene rhythms inmouse is faster than that of others and depends upon thedirection of the phase shift. (Reddy et al., 2002). The rhythmof mPer1/2 expression in the mouse phase advances rapidlyin response to light, while re-entrainment lags for mCry1. Incontrast, the rhythms of mCry1 and mPer1 or mPer2 re-entrain at the same rate during a phase delay. These geneexpression results match corresponding behavioral datashowing slower advances and faster delays (Reddy et al.,2002). A variation on the theme of temporary desynchroni-zation among gene expression rhythms is the demonstrationthat they (Per1, Per2, Cry1) shift rapidly in the ventrolateralrat SCN, but more slowly in the dorsomedial region (Naganoet al., 2003).

A second form of dissociation has been demonstrated inmice lacking the PAC1 receptor (Hannibal et al., 2001b). A lightpulse that induces a 100 min phase delay in wild-type miceelicits a larger phase delay in the mutant mice. However, themutant mice have a vastly attenuated Fos or Per geneinduction response to light. A substantially different experi-ment has yielded the observation that imposition of a 6h photoperiod advance rapidly induces a large phase advancein the Per1-luc bioluminescence rhythm relative to unshiftedcontrols. This phase differential persists for at least 6 days inDD. In contrast, such advances did not occur in in vivo SCNelectrophysiological or behavioral rhythms (Vansteensel et al.,2003a).

A third form of dissociation addresses the question ofwhether particular genes, or change in their expression, arealways necessary for rhythm expression. One example is thegene, Clock. Mice homozygous for a mutant form of Clockrapidly which when absent yields mice that rapidly becomearrhythmic in DD (Vitaterna et al., 1994). More recently,however, a similar test procedure has been applied to Clockmutant mice, but with an LL background to the rhythmexpression. In this circumstance, many of the mice exhibitclear circadian rhythms despite the absence of the Clockgene (Spoelstra et al., 2002). Other investigators havesuggested that running may be an inadequate variable toassess in order to determine whether or not the Clockmouse in DD is rhythmic, and present data supporting theview that such mice are indeed fundamentally rhythmic andentrainable, but with a much longer period than normal(Kennaway et al., 2003).

In quite a different situation with themixed in vivo/in vitropreparation, novel wheel access for 3 hduring the subjectiveday rather quickly reduced Per1 expression in the SCN asmight be expected during the midsubjective daytime. Howev-er, under the conditions of the experiment, there was noassociated shift in the phase of the SCN firing rate rhythm in

animals so treated (Yannielli et al., 2002). On the other hand, ifan additional 2 h of wheel access was permitted (with orwithout additional running), a the firing rate rhythmadvancedabout 2.4 h. Brains taken 5 h, rather than 3 h, after the initialaccess to wheels had reduced Per2 expression, but Per1 hadreturned to normal. One conclusion is that clock resetting inresponse to activity-related stimulation may require a 3–5h poststimulus latency (Yannielli et al., 2002; Antle andMistlberger, 2000), whereas other nonphotic stimuli or lightmay not require the same amount of time (Mead et al., 1992;Sumova and Illnerova, 1996).

6.8. Polysialic acid and SCN function

There are a variety of possible neurobiological roles that maybe played by the molecule, polysialic acid (PSA) and itscarrier, the neural cell adhesion molecule (NCAM), whichform PSA-NCAM (particularly during development and in thecontext of neural plasticity; see Rutishauser and Land-messer, 1996). PSA-NCAM is distributed along plasmamembranes of appositions between neurons, at synapsesand between glia (Shen et al., 1997, 1999). It is of interest thatthe SCN of Siberian (Glass et al., 1994) and golden hamsters(Fedorkova et al., 2002), mice (Shen et al., 1997) and rats(Prosser et al., 2003) contains concentrations of PSA thatfluctuates rhythmically. In LD, PSA concentration dropsprecipitously and by more than 90% during the dark hours.Although the shape of the curve describing PSA content inthe SCN is substantially different during DD, clear rhythmic-ity persists (Glass et al., 2003a). Moreover, light greatly booststransient PSA expression in the SCN (but not in hippocam-pus). NCAM rhythmicity is similar in the SCN, although itsrhythm in DD is much less robust than that for PSA (Glass etal., 2003a).

Mice lacking the NCAM-180 isoform that carries PSA havea shorter than normal circadian period, a longer activityphase and become arrhythmic within 3 weeks of DD (Shen etal., 1997). The mutant mice also show light-induced phasedelays smaller than control mice (Glass et al., 2000a). Therole of PSA itself has been investigated using an enzyme,endoneuraminidase (endoN), to remove PSA. Animals trea-ted with endoN have shorter circadian periods in DD,although they do not appear to become arrhythmic (Shenet al., 1997). Similar to the mutant animals, the endoN-treated mice have diminished phase delays in response tolight; they also have reduced light-induced FOS expression(Glass et al., 2000a).

In contrast to the diminished effects of light in micesustaining endoN removal of PSA, hamsters so treated hadaugmented phase response to nonphotic stimuli (61%increase to 3 h arousal; 120% increase to 8-OH-DPATinjection) applied at ZT6. In the PSA-depleted rat SCNslice preparation, direct application of 8-OH-DPAT elicited a41% greater phase advance (Fedorkova et al., 2002). Also inthe slice preparation (Prosser et al., 2003), PSA levelsremain rhythmic. Glutamate treatment of the slice induceda 100% increase of PSA in the SCN, but this was blocked byendoN pretreatment. Treatment with endoN also blockedthe phase shifting effects of glutamate or electricalstimulation of the slice. The strong implication of this


research is that retinal input to the SCN requires PSA to befunctional.

A further analysis of the molecular requirements formnormal mouse rhythmicity was performed using three strainsdeficient, to varying degrees, in PSA and NCAM. Animalslacking both PSA and all forms of NCAMbecome arrhythmic inDD and tended to run so much when lights are on thatentrained rhythms are often difficult to discern (Shen et al.,2001). A similar strain, but not lacking NCAM isoform 120 hasnormal entrainment under LD, but is arrhythmic in DD,suggesting that NCAM-180 and/or NCAM-140 are required toavoid arrhythmicity in PSA deficient mice. The conclusionfrom these studies is that some combination of PSA andNCAMis necessary for normal rhythmicity and entrainment. Themechanisms through which this action is achieved arepresently unknown.

7. SCN connectivity

7.1. Humoral factors

One of the more surprising results during the last decade ofresearch on SCN function has been the demonstration thatrhythmic events can be controlled, in all likelihood, bynonneural linkage to the circadian clock. The availability ofperinatal SCN transplant methods enabled this conclusion.The procedure developed with hamsters by Ralph et al. (1990)used embryonic SCN to restore locomotor rhythms in lesioned,arrhythmic hosts, but SCN efferent connections were restoredto an uncertain degree (Canbeyli et al., 1991; Lehman et al.,1987). Subsequently, transplantation of viable, membraneencapsulated SCN restored locomotor rhythmicity withoutany apparent neural connections to adjacent hypothalamus(Silver et al., 1996b). In contrast, rat to hamster or mouse tohamster SCN transplants restore rhythmicity in associationwith substantial growth of donor SCN projections into hosthypothalamus, preoptic area and septum (Sollars and Pickard,1993; Sollars et al., 1995; Lehman et al., 1998). It appears to bethe case that behavioral rhythms, but not endocrine rhythms,are restored by transplantation of embryonic SCN (Meyer-Bernstein et al., 1999; LeSauter and Silver, 1998).

Despite declarations that the SCN contains the timingapparatus governing transplanted circadian periodicity, thereare other possible explanations of the phenomenon and theremay be species differences in the extent to which period is theexclusive property of the SCN (Sollars et al., 1995). It is also thecase that the specific location of the transplantwithin the hostbrain is relatively unimportant as long as the site is in or nearthe third ventricle (such as adjacent lateral ventricle) (LeSau-ter et al., 1996; Lehman et al., 1987; Aguilar-Roblero et al., 1994).In addition, transplants consisting of dissociated SCN cells inthe anterior hypothalamic parenchyma or medial hypothala-mus also restore rhythmicity (Silver et al., 1990). Precision ofactivity onset reinstated is related to the transplant distancecaudal or dorsal (but not rostral) to the SCN. Based on thesedata, the suggestion has been made that the target of aneurohumor conveying period information from the SCN islocated near or slightly rostral to the SCN (LeSauter et al.,1997).

7.2. Efferent anatomy

Watts et al. (1987) identified six general regions to which therat SCN projects. Subsequent studies using the anterogradetracer, Phaseolus vulgaris leucoagglutinin, yielded generallysimilar pathways in the hamster brain for which there is (1)an anterior projection to the ventral lateral septum, bednucleus of the stria terminalis and anterior paraventricularthalamus; (2) a periventricular hypothalamic projection tothe medial preoptic region, subparaventricular zone, theparaventricular, ventromedial and dorsomedial nuclei andthe premammillary area and (3) a posterior projection to theposterior paraventricular thalamus, precommissural nucleusand olivary pretectal nucleus (Kalsbeek et al., 1993; Morin etal., 1994). In the hamster, it is likely that the cells projectingto the olivary pretectal nucleus are actually located in theperi-SCN, as is indicated by retrograde tracer injections intothe pretectum (Morin and Blanchard, 1998). Such cells arealso seen following retrograde tracer injections into the IGL,ventrolateral geniculate, posterior limitans nucleus, com-missural pretectal nucleus and medial pretectal nucleus(Morin and Blanchard, 1998; Morin et al., 1992; Vidal andMorin, 2005). In particular, cells projecting to the IGL fromthe SCN region do not reside within the SCN proper, butrather around its dorsal perimeter with a substantialprobability that their dendrites extend ventrally into thenucleus. Watts identified an IGL-efferent projection in theSCN of rats and the cells of origin can be seen within theSCN in this species (Watts and Swanson, 1987). The hamsteralso differs from the rat in that it has a distinct caudalprojection that ascends to innervate posterior paraventricu-lar thalamus (a separate projection innervates the anteriorparaventricular thalamus in both rats and hamsters) (Morinet al., 1994). Given that retrograde tracer studies identifycells around, but not in, the SCN for some of the putativeSCN targets (IGL, VLG, pretectum), it is distinctly possiblethat others (such as the posterior amygdala or posteriorparaventricular thalamus) might also receive afferents fromperi-SCN cells.

Lesion methods have been used in conjunction withimmunohistochemistry to identify SCN projections havingspecific peptide content (Kalsbeek et al., 1993). In thehamster, VP cells project to the medial preoptic, parvicellularparaventricular and dorsomedial hypothalamic nuclei andthe anterior paraventricular thalamus. Cells containing VIPproject to the anterior paraventricular thalamus, parvicellu-lar paraventricular hypothalamus, subparaventricular areaand the dorsomedial hypothalamus. SCN efferent projectionscontaining GRP were identified in the subparaventriculararea, dorsomedial hypothalamus and supraoptic nucleus.The topography of projection neurons in mouse SCN hasbeen inferred (Abrahamson and Moore, 2001) from thepresence of VP or VIP immunoreactive fibers in many ofthe areas identified as receiving SCN projections in otherspecies.

Analysis of SCN projection neurons by retrograde tracermethods provides at least a preliminary indication of astructure–function relationship. Successful retrograde proce-dures indicate both targets of SCN efferent projections and thelocation of neurons providing those efferents. Retrograde


tracing methods have provided much of the recent informa-tion regarding the putative “core” and “shell” organization ofthe SCN (Moore et al., 2002). In this regard, there are tworeferences that need to be compared. The 1987 study of Wattsand Swanson (1987) reached conclusions different from thoseof Leak and Moore (2001), although the two investigationswere equally broad and used basically similar methods. As aresult, it is not possible to determine which interpretation iscorrect. On the one hand, Watts and Swanson (1987) empha-sized the result that, “...cells in dorsal as well as ventral partsof the SCh project to each of the terminal fields examined.” (p.230, abstract). Three cases showing labeled cells followingretrograde tracer injection into the paraventricular hypothal-amus were illustrated. In Case E46, the labeled cells werepredominant in the dorsomedial SCN with few in the ventralpart of the nucleus. In Case E47, there were many more cellslabeled and the great preponderance was found in the ventralhalf of the SCN. In Case PVH33, fewer labeled cells werepresent, and the greatest density of such cells was in theventrolateral quadrant of the SCN. On the other hand, apredominance of retrogradely labeled neurons in the dor-somedial SCN compared to the ventral part following para-ventricular nucleus tracer injection has been reported (Leakand Moore, 2001; Vrang et al., 1995a). The observation ofinjection-related variability by Watts and Swanson is consis-tent with an equivalent observation by Leak and Moore whonoted opposite patterns of dorsomedial:ventrolateral SCN celllabeling that depended upon the precise location of tracerinjection in the subparaventricular zone (medial injec-tion = 5.6:1; lateral injection 1:5.6). In the hamster, tracerinjection into the paraventricular hypothalamus yields a“pattern of retrogradely labeled cells in the SCN […] uniformacross cases […] scattered throughout the subdivisions of theSCN, and no tendency for backfilled cells to be located in thedorsomedial SCNwas observed” (Munch et al., 2002; p. 53), i.e.,there is an absence of topographic organization in the speciesin which it might be most expected. The Watts and Swansonstudy remains themost substantial examination of the extentto which individual SCN cells containing a particular neuro-peptide project to various targets. In general, tracer injectioninto a target nucleus retrogradely labels SCN neurons contain-ing VP and VIP, although the great majority do not containeither peptide.

In the mouse, SCN cells projecting to the subparaventri-cular area have been shown to express FOS protein inresponse to night time light exposure (De la Iglesia andSchwartz, 2002). In the hamster, SCN neurons projecting to theparaventricular hypothalamus have also been shown toexpress FOS in response to light. The percentage of the cellsexpressing FOS is much greater at ZT19 than at ZT14 (Munchet al., 2002). It was also determined that some of the same cellscontained VP, VIP or GRP. Of these, cells with the combinedtraits of projecting to the paraventricular nucleus, expressingFOS in response to light and containing VIP were mostcommon and the combination in which the cells containedVP was rare. Direct SCN projections to gonadotrophin releas-ing hormone cells of the basal forebrain (De la Iglesia et al.,1995; Van der Beek et al., 1997) and to the rat paraventricularhypothalamic nucleus and corticotropin releasing hormonecells, in particular, have been reported (Vrang et al., 1995a,b).

SCN neurons have also been reported to project to, and makelikely contact with, both hypocretin (orexin) and melaninconcentrating hormone neurons in the rat lateral hypothala-mus (Abrahamson et al., 2001).

Transneuronal viral tracers have been utilized to deter-mine polysynaptic routes from the SCN. Following virusinjection into the rat locus coeruleus, cells are retrogradelylabeled in the dorsomedial hypothalamic nucleus, a knowntarget of SCN neurons. The transneuronal tracer also labelsSCN cells, but this does not occur following lesions of thedorsomedial hypothalamus. Furthermore, the circadianrhythm in firing rate of locus coeruleus cells is abolishedafter by the same lesions, suggesting that the SCN control ofrhythmic neuronal activity in the locus coeruleus is via atwo synapse circuit with a first order synapse in thedorsomedial hypothalamic nucleus (Aston-Jones et al.,2001). Injection of virus into the hamster medial vestibularnucleus also labels a few neurons in the SCN. This is apolysynaptic pathway that probably includes the locuscoeruleus which projects to the medial vestibular nucleus(Horowitz et al., 2005).

Viral tracers injected into fat pads (brown or whiteadipose tissue) identified neurons in Siberian hamster andrat SCN (Bamshad et al., 1998, 1999). The actual number ofretrogradely labeled SCN neurons is small, and an estimated2% of these also contain VP (Shi and Bartness, 2001).Injections placed in the adrenal medulla (Ueyama et al.,1999), pancreas (Buijs et al., 2001) and other structures (seeUeyama et al., 1999; Bartness et al., 2001 for reviews) labelcells in the SCN. The general theme of these studies is thatthe SCN regulates sympathetic outflow, a feature evident inthe viral tract tracing of pineal innervation (Larsen, 1999;Larsen et al., 1998; Teclemariam-Mesbah et al., 1999). How-ever, it appears that there is also SCN control of parasym-pathetic output (Ueyama et al., 1999; Kalsbeek et al., 2000).The combination of the two types of peripheral innervationis consistent with the presence of a circuitous polysynapticroute from the SCN through the midbrain to the eyeidentified by virus injection into that organ (Pickard et al.,2002; Smeraski et al., 2004).

There is a large amount of convergence by SCN-efferentprojections onto brain regions receiving IGL input (Fig. 4). Inaddition, many of the same areas are also innervated byretinal and olfactory system projections (Johnson et al., 1988a;Morin and Blanchard, 1999; Morin et al., 1994; Cooper et al.,1994; Ling et al., 1998).

7.3. Afferent anatomy

The SCN-afferent projections in the hamster were initiallydescribed in a fairly comprehensive fashion by Pickard (1982)in the early 1980s. There have been two additional suchinvestigations, and in another species, the rat (Moga andMoore, 1997). In this study, a combination of retro- andanterograde tracing studies demonstrated additional projec-tions to the SCN arising from infralimbic cortex, lateralseptum, substantia innominata, ventral subiculum, subforni-cal organ, ventromedial hypothalamus, arcuate nucleus,posterior hypothalamus, pretectal nuclei, median raphe andIGL, but not in several of the brainstem regions identified by

Fig. 4 – Convergence of SCN (solid lines) and IGL (dashed lines) projections onto diencephalic and basal forebrain regions(black areas) in the hamster. Gray areas receive innervation only from the IGL. Only the premammillary area (white) receivesinput solely from the SCN. Thickness of the lines indicates approximate density of the projection. Note that the hamsterSCN does not project to the IGL, although peri-SCN cells do so. See text regarding further convergence of visual and olfactoryinput to many of these areas. Drawn after Morin et al. (1994) and data from Morin and Blanchard (1999).


Bina et al. (1993), as containing cells projecting to the SCN. Theobservations of Moga and Moore (1997) in the rat aresubstantially similar, but broader than, those of the older,more limited hamster study (Pickard, 1982). As part of a largeranalysis, Krout et al. (2002) evaluated monosynaptic inputs tothe rat SCN identified by CTB labeling. These investigatorsobserved about 40 regions sending direct projections to theSCN, largely in agreement with the results of Moga and Moore(1997). However, Krout et al. (2002) also observed a morewidespread distribution of labeled cells in areas such as thelocus coeruleus and divisions of the parabrachial complex,among others, but did not see such cells in the substantiainnominata and only a few or none in the olivary pretectalnucleus. In all studies involving tracer injection into the SCN,the most likely cause of label disparity across injections is theextent of label spread outside the nucleus, uptake byprocesses in close proximity to the nucleus or uptake by cellprocesses extending significant distances to enter the nucleus(Morin and Blanchard, 1998; Morin et al., 1992, 1994; Gerfenand Sawchenko, 1984).

The above-mentioned report (Krout et al., 2002) describedabout 40 brain regions with cells monosynaptically project-ing to the SCN. Injection of an exclusively retrograde(Pickard et al., 2002), transneuronal virus label into the SCNexpanded the number of brain nuclei with cells projecting tothe rat SCN to about 100–140 regions (variation may relate tolabel spread outside the SCN) (Krout et al., 2002). The regionswith likely polysynaptic projections to the SCN includenumerous additional olfactory structures, divisions of the

cortex, hippocampus, nuclei of the amygdala, thalamus,hypothalamus, as well as nuclei in the midbrain andmedulla from which there are very few reported direct SCNprojections. While this study is important, it should be notedthat any uptake by terminals in the peri-SCN region is likelyto identify regions not connected to the SCN. A similarthorough examination of polysynaptic projections to theSCN has not been systematically completed in other species.However, some information is available for the hamster andshows both similarities and significant dissimilarities withthe rat data. In particular, there appears to be little ornothing in the way of a polysynaptic projection from thehamster tectum and pretectum to the SCN, yet moreoculomotor midbrain structures contain labeled cells thanin the rat SCN (Horowitz et al., 2004). A polysynapticprojection from the medial vestibular nucleus to the SCNhas been reported in the hamster.

Focused studies have documented SCN-afferent projec-tions from the pretectal complex, with cells identified in theolivary, medial and posterior pretectal nuclei (rat) (Mikkelsenand Vrang, 1994) and from the median raphe nucleus (Meyer-Bernstein and Morin, 1996; Hay-Schmidt et al., 2003). Cholin-ergic cells in the substantia innominata, preoptic nucleus,diagonal band of Broca, medial septum, pedunculopontine,laterodorsal tegmental and parabigeminal nuclei have alsobeen reported to project to the SCN in the rat (Bina et al., 1993).As noted above, there is disagreement concerning which, ifany, these nuclei send projections to the SCN (Moga andMoore, 1997; Krout et al., 2002).


8. Intergeniculate leaflet

8.1. Nomenclature

The IGL has been, and continues to be, confused with theventral lateral geniculate nucleus. Focus on the term, “leaflet,”as a descriptor of the IGL has drawn attention away from theactual boundaries of the nucleus and, until fairly recently, hasnot had a clearly identifiable homolog in nonrodent species. Itis now established that the medial division of the ventrallateral geniculate nucleus is the cat homolog of the IGL(Nakamura and Itoh, 2004; Van der Gucht et al., 2003; Pu andPickard, 1996). In the monkey, the homologous structure isknown as the pregeniculate nucleus (Van der Gucht et al.,2003; Moore, 1989). However, there is uncertainty regardingthe neuromodulator content of the GHT across primatespecies (Chevassus and Cooper, 1998).

In the hamster, the definition of the IGL has graduallyevolved to include all regions of the lateral geniculate complexin which NPY neurons and cells projecting to the SCN may befound. As indicated from the developmental data describedbelow, NPY cells found in medial areas near the acousticradiation and ventral to the leaflet portion of the IGL have notcompleted full migration (Botchkina and Morin, 1995). Theregion of the laggard migratory NPY cells also containsneurons projecting to the SCN which are not found elsewherein the lateral geniculate complex outside of the IGL (Morin andBlanchard, 1995; Morin et al., 1992). There is also acorresponding group of enkephalin-containing cells, especial-ly in the medioventral part of the IGL (Morin and Blanchard,1995). This description of the hamster IGL is generallyconsistent with that for the rat (Moore and Card, 1994),although there are slight species differences with respect toits cross-sectional appearance at various rostrocaudal levels.

8.2. Development

IGL development has been described for the hamster (Botch-kina and Morin, 1995). The IGL is probably best considered aspart of the so-called ventral thalamus with a developmentalorigin different from the dorsal lateral geniculate nucleus(Altman and Bayer, 1979a,b). The hamster IGL develops out ofthe reticular neuroepithelium with the first NPY cells evidenton embryonic day 9–10. These are evident against a back-ground of radial glial cells extending dorsolaterally. NPYneurons apparently migrate along a specialized set of radialglial processes reaching their final destination in what willbecome the IGL on E14. The radial glial cell bodies dislocatefrom the ventricular germinal zone on E15, and transform intothe astrocytes that become evident in the IGL across postnataldays 5–10. Some NPY neurons do not achieve completemigration and remain relatively displaced in what is nowconsidered ventromedial IGL. Fiber outgrowth from NPY-containing IGL neurons is evident by P0, and by P3, theypenetrate the SCN. The adult-like NPY terminal plexus in theSCN is achieved by approximately P10.

The P3 arrival in the SCN of the NPY component of the GHTapproximately coincides with the end of SCN synaptogenesis(Speh and Moore, 1990), arrival of serotonergic innervation

(Botchkina and Morin, 1993) and precedes the initial retinalinnervation of the SCN by about 2 days (Speh andMoore, 1993).The arrival of retinal projections in the lateral geniculateregion occurs on E13, but there is a delay in terminalarborization until about E15.5 (Jhaveri et al., 1991). The timingof retinal innervation of the IGL is not known.

8.3. Anatomy

8.3.1. Geniculohypothalamic tractThe NPY neurons were the first IGL cell phenotype shown toproject to the SCN (Card and Moore, 1982; Moore et al., 1984).The initial description of IGL projections to the contralateralnucleus in the rat indicated that many of the cells containedenkephalin, but none of this phenotype projected to the SCN(Card and Moore, 1989). There are substantial species differ-ences with respect to IGL organization and in the hamster,only a very small percentage of NPY or enkephalin neuronsproject to the contralateral IGL (Morin and Blanchard, 1995). Inboth species, GABA neurons are present in the IGL and projectto the SCN (Morin and Blanchard, 2001; Moore and Card, 1994).In the rat, there is a fourth, unidentified, cell type (Moore andCard, 1994). In the hamster, a fourth class consists ofneurotensin neurons (Morin and Blanchard, 2001). It is usefulto recognize, as is further indicated below, that while amplenumbers of the IGL neurons also project to the pretectum, onlya small percentage contain any of the various neuromodula-tors thus far evaluated, implying existence of at least oneother cell type in hamster.

NPY cells can be used to identify the IGL as they occurnowhere else in the lateral geniculate complex (Morin andBlanchard, 2001; Moore and Card, 1994). In the hamster,neurotensin, which is almost completely colocalized withNPY, can likewise identify the nucleus, albeit with somewhatgreater difficulty because it is present in only about half theNPY neurons. There is colocalization ranging from 7% (percentof GABA cells containing enkephalin) to 98% (percent ofneurotensin cells containing NPY) of all identified peptideswith each other in hamster IGL neurons (Morin and Blanchard,2001).

8.3.2. Connectivity and relationship to sleep, visuomotor andvestibular systemsHarrington (1997) has reviewed the substantial literatureconcerning the ventral part of the lateral geniculate com-plex. This review was titled, with a good deal of foresight,“The ventral lateral geniculate nucleus and intergeniculateleaflet: interrelated structures in the visual and circadiansystems.” Most of the literature cited in that review wasunaware that an “intergeniculate leaflet” existed, let alonethe possibility that it might have functions different fromthose of the ventral lateral geniculate. At the time of thereview, it was established that the IGL contributes to thephotic regulation of circadian period and entrainment, andthat IGL lesions could block the ability of certain nonphoticstimuli to induce phase responses normally consistent withan NPY-type PRC (Janik and Mrosovsky, 1994; Wickland andTurek, 1994; see Morin, 1994; Harrington, 1997 for reviews).The prediction in Harrington's review, evident in the title,that the IGL has functions beyond those of circadian rhythm


regulation has not yet been directly established, but thegroundwork is now in place for doing so.

The most important anatomical feature of the IGL appearsto be the enormity of its connections with the rest of the brain(cf., Fig. 4, which only includes a portion of the ascendingprojections). Not only are they widespread, but they aregenerally bilateral and often reciprocal. The magnitude ofIGL connectivity was initially suggested by an investigation ofhypothalamic-IGL connectivity in the gerbil (Mikkelsen, 1990).This study, and others in hamster (Morin and Blanchard, 1999)and rat (Horvath, 1998; Moore et al., 2000) revealed a majorefferent target extending from the lateral optic chiasmmedially through the lateroanterior hypothalamus to themidline, including the SCN, and dorsally to include thesubparaventricular zone and adjacent perifornical anteriorhypothalamus. In addition, IGL projections extend to severaldivisions of the bed nucleus of the stria terminalis and varietyof basal forebrain areas extending laterally as far as thepiriform cortex (Morin and Blanchard, 1999). Projections tonuclei of the midline thalamus are also fairly extensive (Morinand Blanchard, 1999; Moore et al., 2000). Many of the sameregions also send projections back to the IGL (Morin andBlanchard, 1999; Vrang et al., 2003).

Comparison of regions receiving retinal projectionsreveals convergence with IGL efferents. Four distinct routesproviding retinal innervation of rostral diencephalon havebeen described (Johnson et al., 1988a; Morin and Blanchard,1999): (a) to the SCN and anterior hypothalamus; (b) to theventral lateral hypothalamic area; (c) rostrally to the ventralpreoptic area and (d) to the bed nucleus of the striaterminalis (Cooper et al., 1994) (earlier identified as project-ing to the anterodorsal thalamic nucleus; Johnson et al.,1988a). With very few exceptions, IGL projections innervatethe retinorecipient regions, although the extent of IGLprojections is substantially greater than those from theretina (Morin and Blanchard, 1999). In a similar vein, areasreceiving IGL input are also very likely to receive projectionsfrom the SCN (Morin et al., 1994) and from the olfactorysystem (Cooper et al., 1994).

IGL connections with the midbrain and brainstem are justas extensive as those for the forebrain. The concept of a“subcortical visual shell” has been invoked to facilitatediscussion of connectivity of the IGL (Morin and Blanchard,1998). The subcortical visual shell consists of 12 contiguousretinorecipient nuclei of the thalamus and midbrain. All butthose projecting to cortex (lateral posterior and dorsal lateralgeniculate nuclei) receive both ipsi- and contralateral projec-tions from the IGL. In addition, most of the regions also supplyreciprocal input to the IGL bilaterally (Morin and Blanchard,1998), although ipsilateral connections are more robust. Thefunctional importance of IGL connections with nuclei of thesubcortical visual shell has not yet been tested. However, thepresence of bilaterally symmetrical connectivity stronglysuggests involvement of the IGL in the regulation of visuo-motor function (for a review, see Morin and Blanchard, 1998;Horowitz et al., 2004).

The further suggestion has been made that the IGL maycontribute to two discrete classes of events, one concerningvisuomotor, and the other circadian rhythm, regulation(Horowitz et al., 2004, 2005; Morin and Blanchard, 2005). The

IGL has at least two classes of neurons identified by theirprojections to the SCN, one being so connected and the othernot. Those IGL cells that contribute GHT input to the SCNappear not to receive any input from the pretectumor superiorcolliculus despite abundant IGL-afferent projections fromthose areas (Morin and Blanchard, 1998; Horowitz et al.,2004). This further suggests that photic input to the pretectumand tectum does not influence circadian rhythmicity bypassing through the IGL and GHT. This inference is supportedby lesion studies indicating that loss of the IGL, but not visualmidbrain areas afferent to the IGL, modifies circadian rhythmresponse to light (Morin and Pace, 2002).

There have been recent reports describing brainstemorigins of IGL-afferent connections (Horowitz et al., 2004;Vrang et al., 2003). Two of these, a serotonergic input fromthe dorsal raphe and a noradrenergic projection from thelocus coeruleus, have been previously documented (Meyer-Bernstein and Morin, 1996; Kromer and Moore, 1980). Otherregions sending projections to the IGL include severalnuclei of the oculomotor complex, the caudal cuneiform,pedunculopontine and laterodorsal tegmental nuclei (Hor-owitz et al., 2004; Vrang et al., 2003). In addition, projectionsfrom the pararubral nucleus (which also is retinorecipient)and vestibular nuclei to the IGL have been described(Horowitz et al., 2004). Several of these regions, includingthe medial vestibular nucleus, locus coeruleus, divisions ofthe dorsal tegmental and pontine tegmental nuclei and thedorsal raphe, are also recipients of projections from thehypocretin/orexin neurons in the lateral hypothalamic area(Horowitz et al., 2005; Mintz et al., 2001). These regions arealso thought to be involved in the modulation of sleep andarousal processes (Saper et al., 2001). Most also receiveprojections directly from the IGL (however, the locus coer-uleus and vestibular nuclei do not) (Morin and Blanchard,2005). Thus, it seems likely that the IGL contributes to theregulation of sleep and arousal in addition to its role incircadian rhythm regulation and its hypothesized role as amediator of visuomotor function.

8.4. Function

8.4.1. NeurophysiologyA novel observation has been made concerning the firing rateof IGL neurons. In a series of paper, Lewandowski andcolleagues have described a slow oscillation in the rat IGLthat is not evident in adjacent thalamic structures. Maximalfiring rate is approximately 20Hz occurring as a burst that lastsabout 70 s with the overall oscillation of firing having a periodof about 106 s (Lewandowski et al., 2000). Unlike the firingpattern observed in the dorsal lateral geniculate, there is nocorrelation between IGL cell firing and the pattern of firing inthe visual cortex. A lesion of one IGL does not disruptrhythmicity in the other (Lewandowski et al., 2002). The originof the oscillatory activity is unknown, but is not exclusivelyintrinsic to the IGL as it fails to persist in a slice The function ofthis burst oscillation is not known, nor is it even known to berelated to circadian rhythm regulation. However, it is presentonly in the light with bursting suppressed by dark (Bina et al.,1993). The bursting pattern can be obliterated and rendered atonic discharge pattern by bicuculline or picrotoxin applied


intraventricularly (Blasiak and Lewandowski, 2004a). In addi-tion, the IGL burst discharge frequency is increased about 100%by lesions of the dorsal raphe nucleus (Blasiak and Lewan-dowski, 2003), although serotonin agonists injected intraper-itoneally unexpectedly induce an increase in dischargefrequency as well. The implications of the slow oscillatorybursting activity are not clear and may be related more tosuspected visuomotor functions of the IGL.

Intriguingly, the ultradian rhythm in IGL neuron burstdischarge is very similar to what has been described as a burstpattern in SCN neurons (Miller and Fuller, 1992). About 20% ofthe cells recorded in and near the SCN burst at a rate of about10 Hz with an average period of about 143 s. It may be that theIGL burst rhythm is derived from the SCN region because theIGL rhythm is lost in a slice preparation (Blasiak andLewandowski, 2004b).

8.4.2. Nonphotic regulation of rhythmicityOne major question concerning the influence of nonphoticstimuli on circadian rhythm regulation is whether it ismediated by activity (i.e., muscle/joint movement) or acorrelate of activity. Dark pulses, triazolam and novel wheellocomotion each elicit large phase advances when adminis-tered at CT6 (see Morin, 1994; Mrosovsky, 1996) and each ofthese is usually associated with increased wheel running. Atleast two experiments have now demonstrated that differentparts of the brain mediate the phase shifts elicited bydifferent stimuli. In the clearest case (Marchant and Morin,1999), the benzodiazepine, triazolam, failed to elicit phaseshifts at any circadian time in hamsters sustaining neurotoxiclesions of the deep superior colliculus. Lesions of thepretectum were at least partially effective in eliminatingsuch responses and, because of the lesion method, it has notbeen possible to fully distinguish between collicular andpretectal effects. However, lesions confined to the superficiallayers of the superior colliculus had no effect on phaseresponse to triazolam.

In contrast to the ability of midbrain lesions to block phaseshift responses to triazolam, the same lesions had no effect onphase shifts induced by prolonged running in a novel wheel.Thus, the neural substrates for triazolam and novel wheel-induced phase shifts are different, although each appears torequire an intact GHT as a final common path conveyinginformation to the SCN (Janik and Mrosovsky, 1994; Wicklandand Turek, 1994; Johnson et al., 1988b; Marchant et al., 1997).These results support previous observations that chlordiaz-epoxide (a benzodiazepine) and fentanyl (an opiate mu opioidreceptor agonist), neither of which enhances locomotion inhamsters, nevertheless induce equivalent phase shifts (Bielloand Mrosovsky, 1993; Maywood et al., 1997; Meijer et al., 2000)to those induced by triazolam.

The second experiment differentiating response to tria-zolam from response to novel wheel involved serotonin-specific neurotoxic lesions of the median or dorsal raphenuclei. Loss of serotonin neurons from the median (but notthe dorsal) raphe eliminated phase shifts in response totriazolam treatment (Meyer-Bernstein and Morin, 1998).Novel wheel access elicited equivalent phase shifts regard-less of dorsal raphe, median raphe or control lesions. Thus,it is clear that while a final common path from the IGL to

SCN mediates the phase shift response to two distinctstimuli, there are different anatomical substrates subservingthose stimuli. Whether the ability of DD to elicit phase shiftresponses corresponding to the NPY-type PRC depends uponlocomotion has not been determined.

NPY, the first neuromodulator identified in the IGL (Card etal., 1983), has also received the greatest amount of attentionwith respect to IGL and GHT function. Stimulation of the IGLhas been thought to elicit phase shifts because of NPY releasefrom GHT terminals in the SCN (see Morin, 1994). This concepthas been reaffirmed by antagonism of NPY activity in vivo byinfusion of NPY antiserum onto the SCN which blocked phaseshifts in response to novel wheel running (Biello et al., 1994).

As indicated above, cells contributing to the hamster GHTalso contain GABA, neurotensin and enkephalin. Neurotensinappears not to have been evaluated with respect to its abilityto modify circadian rhythmicity in vivo. However, a CT6application of neurotensin to the rat SCN in a slice preparationgenerally reduced neuron firing (Coogan et al., 2001), whileinducing large phase advances of the firing rhythm (Meyer-Spasche et al., 2002).

A small literature exists concerning the function ofenkephalin in the hamster GHT. The delta opioid receptorvisualized by immunohistochemistry is densely distributedon cells in the SCN, particularly in the ventral andmedial partsof the nucleus (Byku et al., 2000). A few cells in the IGL also areidentifiable from the delta opioid receptor labeling. Theobserved distribution of receptor label in the SCN is consistentwith the distribution of enkephalin terminals (Morin et al.,1992). Delta opioid agonists administered at CT8 or CT10 (butnot at other times) will induce moderate phase advances thatare not related to the amount of activity also induced by thedrugs (Byku and Gannon, 2000a,b). Agonists of kappa or muopioid receptors apparently do not elicit phase shifts, althoughother reports indicate that fentanyl, a mu opioid agonist, cangenerate an NPY-type PRC (Meijer et al., 2000; Vansteensel etal., 2003b). The delta (but not mu or kappa) opioid agonists arealso able to prevent light-induced phase shifts when admin-istered in advance of a light pulse at CT19 (Tierno et al., 2002).This result is consistent with a presynaptic inhibitory actionon RHT terminals, as suggested by the receptor localization inthe SCN (Byku et al., 2000). However, electrophysiologicalevidence suggests that the primary site of action of theeffective delta opioid agonists might be outside the SCN(Cutler et al., 1999).

The ability of delta opioid agonists to block light-inducedphase shifts is similar to the effect of NPY. In vivo infusion ofNPY onto the SCN greatly attenuates light-induced phaseadvances and delays (Lall and Biello, 2002, 2003a,b; Weber andRea, 1997). In contrast, infusion of NPY antiserum onto theSCN augments phase advances to light during the latesubjective night (Biello, 1995). Wheel running, which isthought to release NPY into the SCN through the GHT (Bielloet al., 1994), appears to attenuate light-induced phaseadvances, but not delays (Mistlberger and Antle, 1998; Ralphand Mrosovsky, 1992). Inhibition of phase response to light isobservable in the slice preparation if the light pulse isadministered in vivo followed by removal of the brain andacute treatment of the SCN with NPY (Yannielli and Harring-ton, 2000). Similarly, NMDA stimulation of the SCN slice yields


large phase shifts that are blocked by coapplication of NPY(Yannielli and Harrington, 2001; Yannielli et al., 2004) and theNPY-blocked phase response to NMDA can be restored bysimultaneous treatment with an NPY Y5 (but not Y1) receptorantagonist. NPY induces phase advances when applied byitself to the SCN slice during the subjective day, but noobservable phase shift during the subjective night (Harringtonand Schak, 2000; Biello et al., 1997b). In vivo studies usingintracranial injections showed that a Y5 receptor antagonistenhances light-induced phase shifts. If the locomotor rhythmphase response to light is blocked by intracranial NPY, theshifts can be restored to normal by coinjection of NPY and theY5 receptor antagonist (Lall and Biello, 2003a; Yannielli et al.,2004). Additional in vivo studies also demonstrated probableinvolvement of the Y5 receptor in the NPY suppression oflight-induced phase shifts (Lall and Biello, 2003a) and lendcredibility to the view that light and NPY (consequent tolocomotor activity) normally act together to regulate phaseshift magnitude (Lall and Biello, 2002).

The Mistlberger lab has added a new twist to the topic ofnonphotic control of circadian rhythm phase. Initially, therewere two significant observations. One was that 24 h forcedtreadmill activity blocked light-induced phase delays, where-as simple treadmill activity simultaneous with the light pulsedid not alter phase. The second observation yielded theconverse result, namely, a shorter interval on the treadmill(6 h) blocked phase response to light (Mistlberger et al., 1997).

Studies were then performed, using the Aschoff type IIprocedure, to elaborate on the ability of “sleep deprivation” toinduce rhythm phase shifts. In these experiments, sleepdeprivation was accomplished using gentle handling (notconfounded by simultaneous locomotion) beginning at CT6. A3 h procedure elicited a significant phase advance shift and a 1h procedure yielded an even greater (about 150 min) shift(Antle and Mistlberger, 2000). The shifts appear to occur inabout two thirds of the animals and it was observed that thefewer the number of interventions needed to keep an animalawake, the larger the phase shift to the sleep deprivationprocedure. Additionally, exposure to light (10 lx) during a 3h gentle handling interval effectively eliminated phaseadvances. This observation suggests that the phase shiftsconsequent to “gentle handling” are not caused by the absenceof sleep, but rather by a correlate. Such a correlate could be“stress” generated by the sleep deprivation procedure.

In a comprehensive series of experiments (Mistlberger etal., 2003), several of the stress-related issues consequent tosleep deprivation were clarified. In particular, stress asreflected by elevated cortisol levels, is not associated withphase advance shifts, nor are shifts elicited by restraint whichalso elevates cortisol (Mistlberger et al., 2003). Nor is sleepdeprivation caused by the “pedestal over water” methodassociated with phase shifts. In contrast, mere access to a 1m2 open field did cause phase shifts similar in magnitude tothose resulting from gentle handling. Phase shift magnitudewas related to the amount of forward locomotion and thoseanimals classified as “shifters” engaged in considerably morelocomotion than the nonshifters. The conclusion of thesestudies is that some level of locomotion is indeed necessary toachieve phase advance shifts (Mistlberger et al., 2003).Observations that such shifts occur following the simple

intraperitoneal injection of saline (Mead et al., 1992; Hastingset al., 1992) may reflect an action of this stimulus complexthrough an alternate route to the circadian clock not requiringlocomotion for phase shifts (cf., benzodiazepine; Marchantand Morin, 1999; Biello and Mrosovsky, 1993), or the responseto saline may in fact include a locomotor component notmeasured in those studies.

One variation on the theme of nonphotic stimuli capable ofinducing phase advances has involved use of “dark pulses”administered at CT6. The dark pulse is associated withincreased locomotion and consequent phase shifts in ham-sters housed in LL (Boulos and Morin, 1985; Boulos and Rusak,1982). Phase advance responses can be greatly increased (up to7 h) if the dark pulses coincide with the “gentle handling”paradigm for sleep deprivation (Mistlberger et al., 2002).Although the reference to dark pulses as a nonphotic stimulusimplies that it is of the same class as benzodiazepines, novelwheel, gentle handling and saline injection, themechanism ofaction of dark pulses is not necessarily the same. In particular,IGL lesions block the ability of triazolam (Johnson et al., 1988b),novel wheel (Janik and Mrosovsky, 1994; Wickland and Turek,1994) and saline injection (Maywood et al., 1997) (gentlehandling has not been tested) to elicit phase advances, butthey apparently do not block such responses to dark pulses(Harrington and Rusak, 1986).

The effect of restraint has also become more ambiguousthan expected when considered in the context of dark pulses.Restraint during the midsubjective day has no effect onrhythm phase (Van Reeth et al., 1991; Rosenwasser andDwyer, 2002). However, during the late subjective day, briefor 6 h restraint will induce phase delays. At both circadiantimes, restraint blocks the ability of a dark pulse to induce aphase advance (Rosenwasser and Dwyer, 2002; Van Reeth andTurek, 1989; Reebs et al., 1989). Thus, on the one hand, phaseadvances during the midsubjective day are blocked byunknown mechanism invoked by restraint, but one that isapparently not directly related to the circadian clock; on theother hand, blockade of the dark-induced phase advance atdusk may be the simple consequence of a restraint-inducedphase shift of equal magnitude but opposite direction (Rosen-wasser and Dwyer, 2002; Dwyer and Rosenwasser, 2000).

8.4.3. Photic regulation of rhythmicityOne major function of the IGL is modulation of the circadianperiod during LL conditions (Pickard et al., 1987; Harringtonand Rusak, 1986). However, the visual system projects to avariety of midbrain regions that connect with the IGL leavingthe possibility that photic information effecting the circadianclock acts indirectly through, rather than directly in, the IGL.This perspectivehasbeenenhancedby thedemonstration thatvisual midbrain structures are involved in circadian rhythmresponse to a benzodiazepine and that phase shift magnitudeto a novel wheel stimulus is modulated by light (Marchant andMorin, 1999). IGL lesions reduce the circadian period length-ening effect of LL by about 50%, but do not alter the normalperiod in DD (Morin and Pace, 2002). This result suggests theIGLmediates the tonic effects of light. In contrast, large lesionsof the pretectum and tectum that eliminate rhythm responseto benzodiazepine treatment (Marchant andMorin, 1999) haveno effect on circadian period in LL (Morin and Pace, 2002).


In mouse, IGL lesions lengthen the circadian period of DD-housed mice, but do not modify the period lengthening effectof LL (Pickard, 1994; Lewandowski and Usarek, 2002). In the rat,IGL lesions do not alter the circadian period in DD nor do theyprevent LL-induced circadian rhythm disruption (Edelsteinand Amir, 1999a). Period in LL before rhythm disruption wasnot assessed. Under skeleton photoperiod conditions (1 h oflight from ZT0-1 and ZT11-12, rats with IGL lesions generallyfail to entrain indicating the necessity of the IGL under theseconditions; Edelstein and Amir, 1999b). Equivalent studieshave not been done in other species.

Just as NPY is able to attenuate light-induced phase shifts,effects of NPY on phase are also inhibited by light. Novelwheel-induced phase advances are greatly attenuated by asubsequent light pulse. Similarly, NPY-induced phase advan-ces are curtailed by a light pulse (Biello and Mrosovsky, 1995).A glutamate agonist (NMDA), injected onto the SCN to mimicthe effect of photic input, also greatly reduces the phaseadvance effects of NPY administered in the midsubjective day(Gamble et al., 2004). At CT20, NMDA injection onto the SCNinduces phase advances (Mintz et al., 1999), but when givensimultaneously with NPY, small phase delays occur (Gambleet al., 2004). This result, in conjunction with the observationthat the effect occurs in the presence of TTX, suggests thatNPY is acting postsynaptically to induce phase shifts, rather asa presynaptic inhibitor of RHT function.

Glutamate receptor antagonists, known tomodify SCN andrhythm response to light (Abe et al., 1992; Rea et al., 1993b;Vindlacheruvu et al., 1992), fail to alter light-induced FOSexpression in the rat IGL (Edelstein and Amir, 1998). This isconsistent with the likelihood that there appears to be little orno glutamatergic innervation of the IGL (Fujiyama et al., 2003).

9. Serotonin and midbrain raphe contributionto circadian rhythm regulation

The 1994 review (Morin, 1994) fell victim to the typical problemof being partially out of date by the time it was published. Notopic was more affected by this than serotonin and its effecton circadian rhythmicity. In the year of publication, researchon the topic began to blossom and there have been severalmajor reviews published (Morin, 1999; Rea and Pickard, 2000a;Mistlberger et al., 2000; Yannielli and Harrington, 2004).

One of the more important changes in perspectiveconcerning the role of serotonin has been a change of focusaway from the dorsal raphe nucleus, previously thought to bethe source of SCN innervation by serotonin fibers (Smale et al.,1990; Azmitia and Segal, 1978). Anatomical studies, using threeseparate methods (anterograde and retrograde tract tracing ofraphe projections, and lesion studies designed to destroyraphe projections), have unequivocally demonstrated that themedian raphe sends mixed serotonergic and nonserotonergicprojections to the SCN, and the dorsal raphe sends mixedserotonergic and nonserotonergic projections to the IGL (Mogaand Moore, 1997; Meyer-Bernstein and Morin, 1996; Hay-Schmidt et al., 2003). This simple anatomy is complicated bythe presence of reciprocal serotonergic and nonserotonergicconnections betweendorsal andmedian raphenuclei (Tischlerand Morin, 2003).

9.1. Effects of serotonin-specific neurotoxins

The sense of the literature is that, in hamsters, serotonin hasan inhibitory effect on rhythm response to light which, inanimals lacking median raphe serotonin neurons, is evidentas a larger amplitude PRC, greater prevalence of, and shorterlatency to, light-induced rhythm splitting and longer circadianperiods under LL (Meyer-Bernstein andMorin, 1996; Morin andBlanchard, 1991; Bradbury et al., 1997). No effects on rhyth-micity are seen following destruction of only the dorsal rapheserotonergic cells. Most recently, evaluation of irradianceresponse curves in animals with neurotoxic lesions of medianraphe serotonin neurons has shown similar sensitivities oflesioned and unlesioned controls to light, but a greater phaseresponse by lesioned animals to high intensity (probablysaturating) light (Muscat et al., 2005).

The drug of abuse, MDMA (‘ecstasy’), is rapidly toxic toserotonin terminals (Battaglia et al., 1991). Treatment of ratswith this drug for several days reduces phase advances of therhythm in SCN firing rate by about 50% in response to 8-OH-DPAT applied to a slice at ZT6 (Biello and Dafters, 2001). Theeffects of MDMA on rhythmicity have been extended to theevaluation of its effects on hamsters in vivo. After druginjections of MDMA, phase advances to 8-OH-DPAT adminis-tered at CT8 were blocked. Light-induced phase delays wereconvincingly reduced about 50% (Colbron et al., 2002),although the effect is not consistent with the literatureshowing that serotonin loss increases phase response tolight. Damage to the serotonin terminals varies according tobrain region, with some areas being completely unaffected bythe drug (Battaglia et al., 1991). In the above MDMA studies, itis unclear as to which part of, or to what extent, the serotoninsystem has been damaged.

9.2. Serotonin receptor function

A major development in the study of serotonergic effects oncircadian rhythm regulation has been the increased focus onparticular receptor types and their function. One criticaldevelopment has been the recognition that 8-OH-DPAT, longthought to a 5HT1A-specific agonist and frequently usedbecause of that specificity (seeMorin, 1999), is actually specificto the 5HT1A, 5HT5A and 5HT7 receptors (Lovenberg et al.,1993; Sprouse et al., 2004a; and reviewed by Gannon, 2001).There are now reports indicating that the SCN contains the5HT1A, 5HT1B, 5HT2A, 5HT2C, 5HT5A and 5HT7 receptor types(Sprouse et al., 2004a,b; Belenky and Pickard, 2001; Duncan etal., 1999, 2000; Prosser et al., 1993; Oliver et al., 2000; Moyer andKennaway, 1999). A useful review of the evidence favoring thepresence and function of 5HT7 receptors in the SCN waspublished in 2001 (Gannon, 2001), but needs to be interpretedin the context of subsequent information.

The 5HT5A receptor has most recently been appreciated asa serotonin receptor type contributing to circadian rhythmregulation. It is distributed in the dorsal and median raphe, aswell as in the SCN and IGL (Duncan et al., 2000). Thus, the5HT1A, 5HT5A and 5HT7 receptors are found in similar placesand respond to similar agonists/antagonists thereby increas-ing the difficulty of discerning receptor-specific function. Onesuggestion is that 8-OH-DPAT injection into the dorsal raphe


induces phase shifts via 5HT7 (but not 5HT5A) receptoractivation (Duncan et al., 2004), whereas systemic injectionof the drug induces phase shifts by activating the 5HT1Areceptor (Tominaga et al., 1992). The phase response ofhamsters to dorsal raphe application of serotonin 5HT1A/5A/7 agonists diminishes with age (Duncan et al., 2004; Penev etal., 1995), and thismay be the result of specific loss of available5HT7 receptors (Duncan et al., 1999).

In the rat, the 5HT2C receptor may contribute to theregulation of circadian rhythm phase. The receptor type ispresent in the SCN (Moyer and Kennaway, 1999). Agonists ofthe 5HT2C receptor are able to induce phase delays in rats andinduction of FOS in SCN neurons at CT18, but not at CT6(Kennaway and Moyer, 1998), and generally mimic the effectsof light. The suggestion has beenmade (Kennaway andMoyer,1998) that previous observations of quipazine-induced phaseshifts in rat (Prosser et al., 1993; Kennaway and Moyer, 1998)may have been the result of that drug's affinity for the 5HT2class of receptors. Kennaway andMoyer (1998) suggest that theliterature is ambiguous with respect to the certainty that therat SCN contains 5HT1A and 5HT7 receptors. Further, theysuggest the possibility that light may act through the knownretina-dorsal raphe projection in the rat (not evident in thehamster; see Section 4.2) and it is themodulation of the 5HT2Creceptors on this pathway that results in light-like phaseresponse and FOS induction (Kennaway and Moyer, 1998).Consistent with this view is the observation that reduction ofserotonin in rats by pCPA injection attenuates light-inducedFOS (Moyer and Kennaway, 2000), suggesting that normalserotonin innervation is necessary for light-induced FOS.Agonist activation of 5HT2A/C receptors induces substantially(but not completely) light-like FOS expression in the rat SCN(Kennaway et al., 2001). The investigators suggest that, in therat, but probably not in the mouse or hamster, there are twopathways by which light can modulate SCN rhythmicity. Oneof these is the RHT, while the other involves the retinalprojection to the dorsal raphe which, it is hypothesized, sendsa direct serotonergic projection to the SCN that effectsglutamate release at that location (Kennaway et al., 2001). Adorsal raphe projection to the SCN is probably absent in the rat(Moga and Moore, 1997), as it is in hamster, but there isabundant dorsal raphe innervation of adjacent hypothalamus(Moga and Moore, 1997; Vertes, 1991) and peri-SCN neuronsmodulated by serotonin from the dorsal raphe may provideglutamate signaling to the SCN. Involvement of the 5HT2Creceptor in rat rhythm regulation is further supported by theobservation that agonists specific for this type of receptorinduce expression of FOS, Per1 and Per2 in the SCN. This effectwas more robust at ZT16 than ZT22 (Varcoe et al., 2003).

The issue of species differences has been addressed bydirect comparison of mouse and hamster responses to avariety of 5HT1A, 1B, 2 or 7 agonists. The results are complexwith differing phase and spontaneous Fos-inductionresponses in each species, plus differing effects on light-induced phase shifts and Fos-induction. In addition, responsehas also been shown to be related to whether the drug wasinjected centrally (SCN) or administered peripherally (Antle etal., 2003b). One aspect of the complexity of serotonergicactivity relative to the effects of light is the demonstrationthat phase response to agonists can vary enormously depend-

ing upon prior light exposure (Knoch et al., 2004). Hamstersexposed to LL for as little as 1 day can have very large phaseshifts in response to intracranial NPY given at CT6, a sleepdeprivation procedure at ZT18 or 8-OH-DPAT at ZT18 or ZT21(Aschoff type 2 test). LL exposure also eliminates the earlynight time peak in peri-SCN serotonin and reduces the overalllevel to approximately the typical late afternoon levels seenunder LD conditions (Knoch et al., 2004).

The 5HT1B receptor has received substantial attentionwithrespect to its role in rhythm regulation. The receptor isidentifiable in the SCN, specifically on terminals of the RHT(Belenky and Pickard, 2001; Pickard et al., 1999) and there is anapproximately 35% reduction in 5HT1B binding sites in theSCN after bilateral enucleation (Pickard et al., 1996; Manriqueet al., 1999). 5HT1B receptor agonistsmarkedly attenuate light-induced phase shifts and the induction of FOS in the SCN,although cells in a central region that appears to include theCALB neurons continue to express FOS (Pickard et al., 1996,1999; Pickard and Rea, 1997; Shimazoe et al., 2004). Light-induced suppression of nocturnal melatonin is also alleviatedby treatment with 5HT1B agonists (Rea and Pickard, 2000b).Infusion of receptor agonist directly onto the SCN eliminatedphase response to light (Pickard et al., 1996).

Use of the 5HT1B receptor knockout mouse has yieldedsubstantial additional data. Phase shifts elicited by light in thisstrain are not modified by 5HT1B receptor agonists. Electricalstimulation of the optic nerve or glutamate application to thebath elicit excitatory postsynaptic currents (EPSCs) in SCNneurons recorded in a slice preparation. Agonists for the5HT1B receptor greatly reduce EPSCs in the wild-type mouse,but have no effect in the 5HT1B receptor KO mouse (Smith etal., 2001; Pickard et al., 1999). Although the circadian period ofwild-typemice does not differ from that of the 5HT1B KOmice,the period is longer in LL consistent with the view that thisstrain lacks inhibitory control of retinal input to the circadianvisual system (Sollars et al., 2002). Most recently, bicuculline-sensitive miniature inhibitory currents were recorded on SCN(which are predominantly GABAergic; Moore and Speh, 1993;Morin and Blanchard, 2001). Frequency of the inhibitory cur-rents is reduced by a 5HT1B receptor agonist, but the agonisthadno effect on the currents in SCNneurons of 5HT1BKOmice(Bramley et al., 2005). The authors suggest that abundant5HT1B receptors are present presynaptically on GABA term-inals, and thereby modulate GABA release within the SCN.

9.3. Regulation and consequences of serotonin release

Extracellular serotonin increases abruptly from daytime levelsto higher night time levels in both the peri-SCN and peri-IGLregion (Dudley et al., 1998; Grossman et al., 2004). Similarly,extracellular serotonin increases in both the SCN and IGL inresponse to certain nonphotic stimuli (Dudley et al., 1998;Grossman et al., 2000, 2004). Serotonin availability in thesenuclei is also augmented by electrical stimulation of the dorsalraphe nucleus (Grossman et al., 2004; Dudley et al., 1999; Glasset al., 2000b, 2003b).

Despite these effects on serotonin release, it is not yetknown to what extent the stimulation-induced transmitterincrease in either the SCN or IGL contributes to circadianrhythm phase control. Studies involving electrical stimulation

Fig. 5 – Diagrammatic conceptualization of the functionalserotonergic circuitry governing circadian rhythm regulation.Based on an unpublished update provided courtesy of Dr. J.D.Glass, modified from the previous version (Glass et al.,2003b). The presumptive intra-IGL GABA connection is basedupon unpublished work from the Glass laboratory. There aredata for (Glass et al., 2003b) and against (Meyer-Bernstein andMorin, 1998) the likelihood that behavioral input inducesphase shifts by acting through the serotonin neurons of thedorsal raphe nucleus.


of the dorsal ormedian raphe inwhich serotonin neuronshavebeen destroyed have not been performed. Arguing against anormal effect of serotonin on rhythm phase control is theobservation that treatment of hamsters with a 5HT1A receptorantagonist elevates serotonin release in the peri-SCN region tolevels comparable to those seen during novel wheel access orraphe electrical stimulation without augmenting phase shiftsin response to novel wheel activity (Antle et al., 2000). Nor dosystemic or SCN injections of 5HT1/2, 5HT1A and 5HT2/7receptor antagonists alter novel wheel-induced phase shifts(Antle et al., 1998). Consistent with those results is theobservation that the serotonergic median raphe projectionsappear to be essential for triazolam-induced phase shifts, butnot for those induced by novel wheel activity (Meyer-Bernsteinand Morin, 1998).

It is possible that the electrical stimulation activatesnonserotonergic raphe cells projecting to the SCN or IGL(Meyer-Bernstein and Morin, 1996). The most directly relateddata on the topic come from the mouse in which a serotonin-specific neurotoxin was injected intracranially into the regionof the SCN (Marchant et al., 1997; Edgar et al., 1997) and, incombination with IGL lesions, block activity-related phasecontrol in the mouse. The suggestion is that the serotonergicprojections to the SCN destroyed by the neurotoxin arenecessary for suchphasecontrol.However, in theexperiments,serotonin depletion occurred widely within the diencephalonand localization to the SCN cannot be assured. In the hamster,small injections of neurotoxin that destroyed nearly allserotonergic innervation to the SCN without damaging mostof the surrounding hypothalamus had very little effect oncircadian rhythmicity compared to the consequences of loss ofallmedianrapheprojections (Meyer-BernsteinandMorin,1996;Meyer-Bernstein et al., 1997; Bobrzynska et al., 1996). There aretwo implications of these studies. One is that median rapheprojections to non-SCN targets (also destroyed by intrarapheneurotoxin injection) may contribute to the regulation ofhamster circadian rhythmicity. The other is that the denseserotonergic projection to the SCN (destroyed by the intra-SCNneurotoxin) may beminimally involved in such regulation.

One perplexing issue has been the fact that electricalstimulation of the dorsal or the median raphe nuclei yieldssimilar release of serotonin in the SCN (Dudley et al., 1999) andstimulation of either nucleus during the subjective day elicitsequivalent magnitude phase advances (Meyer-Bernstein andMorin, 1999). A model has been developed (Glass et al., 2003b)that provides a reasonable reconciliation of behavioral,pharmacological and anatomical data (Fig. 5; see Glass et al.,2003b for an extensive explanation and presentation of thedata implicating GABA). The similar effectiveness of stimula-tion occurring in either dorsal or median raphe may bepossible because of the serotonergic–nonserotonergic inter-connection between the two nuclei (Tischler and Morin, 2003).

9.4. Serotonergic potentiation of light-induced phase shifts

It is generally the case that a manipulation during thesubjective night that modifies phase advances will alsomodify phase delays. Such is not true for the drug BMY 7378.Administered at CT19, this drug potentiates the phase shiftingeffect of light given shortly thereafter. However, no such effect

occurs when the drug/light combination occurs at CT14. Thereason for this is not clear, although it may result fromcancellation of light-induced phase delays by drug-inducedphase advances when each is administered at CT14. At CT19,BMY 7378 alone induces small (1 h) phase advances, but theCT19 combination of light and drug yields shifts more than300% larger than those induced by light alone (light-inducedshifts are about 1.5 h; potentiated shifts are about 5 h; Gannon,2003). This phenomenon is similar to what has been obtainedwith other drugs with different effects on phase response inthe absence of light (Gannon, 2003; Rea et al., 1995; Moriya etal., 1998). For example, NAN-190 administered alone at CT14 orCT19-20 has essentially no effect on rhythm phase (Rea et al.,1995; Moriya et al., 1996), but potentiates both phase advancesand delays when administered at those times in combinationwith light (Rea et al., 1995). One interpretation (Gannon, 2003)is that the drugs, which have mixed agonist/antagonistproperties, facilitate light-induced phase shifts by acting asagonists on presynaptic 5HT1A receptors of median rapheserotonin cell dendrites to inhibit release of serotonin in theSCN. Elsewhere in the brain, the drugs act on postsynaptic5HT1A receptors which are, presumably, antagonized by thedrugs. WAY 100635, a 5HT1A antagonist, has no effect onlight-induced phase advances (Smart and Biello, 2001), and itis thought that the particular combination of raphe cellautoreceptor activation plus the antagonism of SCN postsyn-aptic receptors is necessary for the large light-induced phaseadvances potentiated by the mixed agonist/antagonists


(Gannon, 2003). If this hypothesis is correct, then it may bepossible to mimic the effect of the mixed agonist/antagonistdrugs by concomitant injection of a 5HT1A antagonist and a5HT1A agonist known to act on raphe autoreceptors.

The complex control of serotonergic involvement inrhythm regulation is further suggested by the observationthat depletion of brain serotonin (88% lower in the peri-SCNregion) by systemic pCPA treatment greatly augments phaseshifts in response to SCN application of serotonin or the5HT1A/7 receptor agonist, 8-OH-DPAT (shifts are about 100minwith pCPA treatment vs. 30 min without) (Ehlen et al., 2001).The results suggest that a necessary condition for normalphase response to agonists is lack of activity at the postsyn-aptic receptors. Short-term changes in postsynaptic receptornumber or sensitivity to endogenous or exogenous agonistsmay be critical to their effect on rhythmphase (see Ehlen et al.,2001 for a discussion).

In animals pretreated with pCPA, phase advances resultingfrom8-OH-DPAT infusiononto the SCNatCT6were reducedbyabout 68% in animals treated with 5HT7 receptor antagonists(Ehlen et al., 2001). On the one hand, the result supports theview that 5HT7 receptors mediate the effect of serotoninagonists applied to the SCN. On the other, there is a significantresidual phase shift effect of 8-OH-DPAT that may beattributable to its affinity for the 5HT1A receptor. It is alsointeresting that light stimulation simultaneous with 8-OH-DPAT (i.e., CT6) in pCPA treated animals completely blocks thephase shifting effect of the 5HT1A/7 agonist (Ehlen et al., 2001).The in vivo result is consistent with observations thatglutamate application to the SCN slice preparation, whichpresumably mimics the excitatory effect of light exposure,blocks the phase shifting effect of NPY or serotonin (Biello etal., 1997a; Prosser, 1998). The presumption is that the action ofserotonin takes place in the SCN, although there is evidence forIGL involvement in the response to 8-OH-DPATaswell (Challetet al., 1998b).

Fig. 6 – Failure to re-entrain to a shifted photoperiod bySiberian hamsters. The initial photoperiodwas LD16:8 (whiteand black bars at the top of each record; lights off = ZT12)which, on the day of the arrowhead, was changed. On thatday at ZT17, the lights were turned on for 2 h. The next day,the LD16:8 photoperiod was delayed by 3 h. Shading areasindicate the times of lights off under the modifiedphotoperiod. This combination of light manipulationsprevents re-entrainment in most animals. The expectedre-entrainment (A) sometimes follows the phase shift, but isless likely to occur than the lack of re-entrainment (B, C). Onecase (C) demonstrates neither re-entrainment nor a clearfree-running circadian rhythm, and a significant percentageof animals become arrhythmic. Unpublished data courtesy ofDr. N.F. Ruby.

10. Failure to re-entrain

It is axiomatic that an entrained animal will re-entrain to ashifted light–dark cycle. In this era of transgenic animalsspecifically designed to show abnormal behavioral traits thatwill allow novel probing of the way things usually work, it isironic that one of the more provocative developments duringthe last decade has been the discovery that a standardlaboratory rodent, the Siberian hamster, is a model forfailure to re-entrain. The initial observation occurred in bothyoung and old animals entrained initially to an LD 16:8photoperiod, then subjected to a 5 h delay of the photope-riod. About 90% of animals had free-running locomotor andbody temperature rhythms for many weeks after the shift orbecame arrhythmic (Ruby et al., 1996, 1997) (Fig. 6).Melatonin treatment, known to modify the phase angle ofentrainment in Siberian hamsters (Margraf and Lynch, 1993),reduced the percentage of animals failing to re-entrain to ashifted photoperiod (Ruby et al., 1997). Analysis of melatoninlevels has shown that animals subjected to a 5 h phase delayhave very low circulating melatonin at a time that it isexpected to be high (Ruby et al., 2000), suggesting that the


absence of melatonin may contribute to the inability toachieve a stable phase angle of entrainment after the shift.

Siberian hamsters re-entrain properly to 1 or 3 h delays ofphotoperiodphase, and to 1hadvances, butmuch lesswell to 3h advances. Generally, these animals do not re-entrain prop-erly following 5 h advance or delay shifts in the photoperiod,although the effect ismodifiable by the exactmanner inwhichthe photoperiod is shifted (Ruby et al., 1998, 2004). If Siberianhamsters are exposed to a 5 h phase delay in the photoperiod,then immediately placed in DD for 14 days, all animals will re-entrain when the shifted photoperiod is resumed. If DDexposure is delayed for several days, re-entrainment probabil-ity is reduced to less than 50% (Ruby et al., 2002b).

The Siberian hamster is noted for the existence ofindividuals in the general population that fail to undergo go-nadal regressionwhen exposed to a short day photoperiod andbreeding has selected this trait in a strain of animals identifiedas ‘nonresponders.’ Photoperiodic time measurement for re-production control relies on the use of a circadian clock (Elliottet al., 1972; Rusak and Morin, 1976; Stetson and Watson-Whitmyre, 1976). Comparison of the responder and nonre-sponder strains with the phase shift paradigm has shown thatthey also have the traits of re-entrainment and nonre-entrainment, respectively (Ruby et al., 2004).

Most recently, Siberian hamsters that fail to re-entrain to a 5h photoperiod shift have been exposed to light to induce SCNgene activity. In animals that re-entrain, Fos and Per1 geneactivity is induced, as expected, in SCN neurons. In contrast,gene induction by light apparently does not occur in animalsthat fail to re-entrain (Barakat et al., 2004). This is conceivablyrelated to the absence of light-induced gene expression ingoldenhamsters showing split rhythms in LL (De la Iglesia et al.,2000). The collective results of these studies indicate that thelighting conditions dictate subsequent response of the Siberianhamster circadian system to light. The most extreme con-clusion is that a consequence of the 5 h phase shift renders thecircadian rhythm system unresponsive to the effects of light.

11. Masking

Mrosovsky (1999) is synonymous with masking and hasprovided a substantial review of the topic. Masking isimportant to the study of circadian rhythms for severalreasons. One is purely technical to the extent that, under avariety of circumstances in which there are questionsconcerning whether the evident rhythm is imposed by theenvironment or is from endogenous sources, tests must beconducted to obtain an explicit answer. For example, micelacking the genes coding for cryptochrome1 (Cry1) andcryptochrome2 (Cry2) show nocturnal locomotor activityunder LD conditions, but are arrhythmic under DD (Albuset al., 2002; Van der Horst et al., 1999). The transfer into DD isthe test of persistent rhythmicity with the same initial phaseas seen under LD conditions. In the case of the Cry1/2 KOmice, the result of the transfer demonstrated that theabsence of wheel running during the light portion of theprevious daily photoperiod occurred because the runningwas strongly suppressed by the negative masking effects oflight.

A second major reason for studying masking concerns theanatomy of the visual system. Light-induced masking is avisual function, but the visual system anatomy contributing toit has not been satisfactorily elucidated.

Normal mice generate a characteristic irradiance responsecurvewhen tested for light-inducedmasking of runningwheelrevolutions. Through a middle range of intensities, thestronger the stimulus, the greater the extent of negativemasking. With very dim stimuli, activity level actuallyincreases slightly (positive masking; see Mrosovsky, 1999;Redlin and Mrosovsky, 1999a). At the level of the retina, bothclassical and melanopsin photoreceptors have been shown tocontribute to masking (Hattar et al., 2003; Panda et al., 2003).Melanopsin cells influence masking responses to the higherlight intensities (Mrosovsky and Hattar, 2003). Loss of rodfunction in mice eliminates the positive masking seen at dimillumination levels, but for some strains, there may also be asignificant reduction in the masking response to higherintensities (Mrosovsky et al., 1999, 2000).

The analysis of visual projections contributing to negativemasking remains incomplete. Mrosovsky and colleagues haveevaluated contributions of the visual cortex, superior collicu-lus, lateral geniculate and SCN without obtaining datademonstrating that oneof these areas is necessary for negativemasking. In fact, themost common effect of lesions is actuallyincreased masking. This is true for lesions of the visual cortex,superior colliculus, dorsal lateral geniculate and IGL, althoughit has not yet been possible to distinguish the effects ofdorsolateral geniculate lesions from lesions aimed at the IGL(Edelstein et al., 2001; Redlin et al., 1999, 2003). Lesions of theSCN render animals arrhythmic and masking experimentsmust, of necessity, be conducted differently from those inwhich animals have normal circadian locomotor rhythms.When SCN lesioned animals are given access to runningwheels in a 3.5 h light:3.5 h dark schedule, about 90% of theirwheel revolutions occur during the dark intervals and thesame animals show a strong preference for darkened vs. alighted chamber (Redlin and Mrosovsky, 1999b). These resultsindicate that negativemasking is not abolished by SCN lesions.By inference, it would also appear that retinal projectionspassing through the SCN to the subparaventricular zone anddorsomedial hypothalamus are not necessary for maskingbecause they would also be destroyed by the lesions. However,these conclusions must be tempered by a recent reportindicating loss of negative masking in SCN-lesioned hamsters(Li et al., 2005). This study did not address possible maskingfunctions of retinal projections destroyed while passingthrough and around the SCN to other hypothalamic sites.

Despite the absence of a demonstration that a particularstructure is necessary for masking, the Mrosovsky workcollectively supports the view that two visual processesgovern the behavior. One, involving thalamus, midbrain andcortical visual structures, provides a normal inhibitory effecton the magnitude of the masking response to light. The other,presumably in one of the two dozen or so other retinorecipientbrain regions (Morin and Blanchard, 1999; Horowitz et al.,2004) not yet tested, or in the SCN (Li et al., 2005), actuallycontrols the negative masking response to light.

As indicated above, the rhythmicity of Cry1 andCry2 doubleKO mice under LD conditions has been explained as a


phenomenon of negative masking because these mice are ar-rhythmicunderDDconditions (Albus et al., 2002; VanderHorstet al., 1999). This issue has also been tested in the standardfashion (1 h light pulse at various levels of illumination). Thedouble KO animals show masking which is no different fromthe wild-type controls, i.e., they are not more sensitive to theactivity suppressing effects of light than thewild-type animals(Mrosovsky, 2001). These mice show bouts of “predark”locomotion the magnitude of which may be related to theprevailing light intensity. These animals may have an alteredphase angle of entrainment that requires specific plottingmethods to discern (see Fig. 3 in Mrosovsky, 2001).

12. The future

Circadian rhythmicity is a pervasive feature of mammalianphysiology and behavior. Thus, it should not be surprising thatdiscoveries concerning regulation of the circadian visualsystem have importance with respect to understanding theworkings of the brain and body. In this context, the discoveryof melanopsin as a photoreceptive molecule in a subset ofganglion cells is one of the great strides forward in the study ofthe circadian visual system during the last decade. However,as is already being demonstrated (Dacey et al., 2005), it isunlikely that the importance of these ganglion cells liesexclusively with photic regulation of circadian rhythmicity.The melanopsin cells of the retina are likely to project widelyin the visual system and influence most, if not all, retinor-ecipient regions.

It is clear that the visual regulation of circadian rhythmsis influenced by the classical visual photoreceptors as wellas the intrinsically photoreceptive ganglion cells. There is anopportunity for the circadian rhythm field to contribute tounderstanding of retinal function by establishing the ana-tomical and functional relationships between rods, conesand melanopsin cells with respect to rhythm response tolight.

The most obvious influence of circadian rhythm researchon the understanding of normal bodily function may bederived from the large amount of research (most of whichhas not been considered by this review) concerned with themolecular regulation of clock-like activity in the SCN. Thesestudies have discovered a new window into the exploration ofmammalian physiology through the demonstration of oscil-lating clock genes in peripheral tissues. While not necessarilycircadian oscillations themselves (but see Yoo et al., 2004 tothe contrary), under normal conditions, the genes do oscillatewith many of the same characteristics as those in the SCN(Yagita et al., 2001). The widespread presence of oscillatingpresumptive clock genes undoubtedly has enormous ramifi-cations for the normal functioning of organs as well as indisease states.

Within the SCN, systematic exploration of the relationshipbetweencell phenotype and functionhasbeen initiated. Part ofthis effort has focused on cell level analysis of gene expression.Another has sought to describe the rhythmic activity ofparticular cell types. Both have been directed, at least partially,toward the intrinsic organization of thewhole SCN. The overallresearch effort has barely addressed a small fraction of the cell

types and already there is typical scientific confusionwith onemethod showing an unexpected absence of rhythmicity incertain cells (Hamada et al., 2001), while another methoddemonstrates an unexpected form of rhythmicity in the sametype of cell (Hamada et al., 2003). This topic is rich and likely tocontribute substantially to such issues as the bimodal rhyth-micity observed in horizontal slices through the hamster SCN,function-specific rhythm regulation and the general problemof specifying which cells are pacemakers.

Research focused on the IGL component of the circadianvisual system has refined the view that it contributes to bothphotic and nonphotic regulation of rhythmicity. However, theresearch on IGL connectivity has pointed toward a function forthis nucleus that may extend well beyond its role in circadianrhythm regulation. In particular, the IGL appears to havewidespread, bilateral and often reciprocal connections withbrain regions involved in the regulation of sleep, visuomotorand vestibular system activities. The nature of the functionalrelationships among these systems remains to be seen, as isthe circadian system role of the IGL with respect to theirregulation.

Theserotonergicsysteminfluenceonrhythmregulationhasbeen substantially elucidated, but large parts of that influenceremain a mystery. One characteristic, in particular, is compel-ling and not yet understood. Certain manipulations of theserotonergic systemcontribute to very large, rapid phase shiftsto photic stimuli. How the circadian clock can overcome theinertia of a stable, entrained system is not understood, but thepresence of such large shifts gives hope that there may beforthcoming manipulations that can reliably overcome thedetrimental effects of jet lag or modify the severity of timingdisorders such as advanced sleep phase syndrome.

The last decade of research on the circadian visual systemhas been extremely exciting and productive. The currentdecade will see this trend continue unabated with a muchenhanced recognition of the contributions made by investi-gators in the field of circadian visual system research to theunderstanding of general brain function and the normalphysiology of bodily systems.

Acknowledgments

Supported by NIH grants NS22168 and MH64471 and NationalSpace Biomedical Research Institute grant HPF0027 viaCooperative Agreement NCC9-58 with the National Aeronau-tics and Space Administration (to LPM) and NIH grantsNS36607 and NS40782 (to CNA).

R E F E R E N C E S

Abe, H., Rusak, B., 1992. Stimulation of the hamster ventral lateralgeniculate nucleus induces Fos-like immunoreactivity insuprachiasmatic nucleus cells. Neurosci. Lett. 148, 185–189.

Abe, H., Rusak, B., Robertson, H.A., 1992. NMDA and non-NMDAreceptor antagonists inhibit photic induction of Fos protein inthe hamster suprachiasmatic nucleus. Brain Res. Bull. 28,831–835.

Abe, H., Honma, S., Shinohara, K., Honma, K., 1996. Substance Preceptor regulates the photic induction of Fos-like protein in


the suprachiasmatic nucleus of Syrian hamsters. Brain Res.708, 135–142.

Abe, M., Herzog, E.D., Block, G.D., 2000. Lithium lengthens thecircadian period of individual suprachiasmatic nucleusneurons. NeuroReport 11, 3261–3264.

Abrahamson, E.E., Moore, R.Y., 2001. Suprachiasmatic nucleus inthe mouse: retinal innervation, intrinsic organization andefferent projections. Brain Res. 916, 172–191.

Abrahamson, E.E., Leak, R.K., Moore, R.Y., 2001. Thesuprachiasmatic nucleus projects to posterior hypothalamicarousal systems. NeuroReport 12, 435–440.

Aggelopoulos, N.C., Meissl, H., 2000. Responses of neurones ofthe rat suprachiasmatic nucleus to retinal illuminationunder photopic and scotopic conditions. J. Physiol. 523,211–222.

Agostino, P.V., Ferreyra, G.A., Murad, A.D., Watanabe, Y.,Golombek, D.A., 2004. Diurnal, circadian and photic regulationof calcium/calmodulin-dependent kinase II and neuronal nitricoxide synthase in the hamster suprachiasmatic nuclei.Neurochem. Int. 44, 617–625.

Aguilar-Roblero, R., Morin, L.P., Moore, R.Y., 1994. Morphologicalcorrelates of circadian rhythm restoration induced bytransplantation of the suprachiasmatic nucleus in hamsters.Exp. Neurol. 130, 250–260.

Aida, R., Moriya, T., Araki, M., Akiyama, M., Wada, K., Wada, E.,Shibata, S., 2002. Gastrin-releasing peptide mediates photicentrainable signals to dorsal subsets of suprachiasmaticnucleus via induction of Period gene in mice. Mol. Pharmacol.61, 26–34.

Aioun, J., Chambille, I., Peytevin, J., Martinet, L., 1998. Neuronscontaining gastrin-releasing peptide and vasoactive intestinalpolypeptide are involved in the reception of the photic signal inthe suprachiasmatic nucleus of the Syrian hamster: animmunocytochemical ultrastructural study. Cell Tissue Res.291, 239–253.

Akasu, T., Shoji, S., Hasuo, H., 1993. Inward rectifier andlow-threshold calcium currents contribute to the spontaneousfiring mechanism in neurons of the rat suprachiasmaticnucleus. Pflugers Arch. 425, 109–116.

Albers, H.E., Ferris, C.F., 1984. Neuropeptide Y: role in light–darkentrainment of hamster circadian rhythms. Neurosci. Lett. 50,163–168.

Albers, H.E., Ferris, C.F., Leeman, S.E., Goldman, B.D., 1984. Avianpancreatic polypeptide phase shifts hamster circadianrhythms when microinjected into the suprachiasmatic region.Science 223, 833–835.

Albers, H.E., Liou, S.-Y., Stopa, E.G., Zoeller, R.T., 1991. Interactionof colocalized neuropeptides: functional significance in thecircadian timing system. J. Neurosci. 11, 846–851.

Albus, H., Bonnefont, X., Chaves, I., Yasui, A., Doczy, J., Van derHorst, G.T.J., Meijer, J.H., 2002. Cryptochrome-deficient micelack circadian electrical activity in the suprachiasmatic nuclei.Curr. Biol. 12, 1130–1133.

Alleva, J.J., Waleski, M.V., Alleva, F.R., 1971. A biological clockcontrolling the estrous cycles of the hamster. Endocrinology88, 1368–1379.

Altman, J., Bayer, S.A., 1979a. Development of the diencephalon inthe rat: V. Thymidine-radiographic observations oninternuclear and intranuclear gradients in the thalamus.J. Comp. Neurol. 188, 473–500.

Altman, J., Bayer, S.A., 1979b. Development of the diencephalon inthe rat: IV. Quantitative study of the time of origin of neuronsand the internuclear chronological gradients in the thalamus.J. Comp. Neurol. 188, 455–472.

Amir, S., 1992. Blocking NMDA receptors or nitric oxide productiondisrupts light transmission to the suprachiasmatic nucleus.Brain Res. 586, 336–339.

Amir, S., Edelstein, K., 1997. A blocker of nitric oxide synthase,NG-nitro-L-arginine methyl ester, attenuates light-induced Fos

protein expression in rat suprachiasmatic nucleus. Neurosci.Lett. 224, 29–32.

Antle, M.C., Mistlberger, R.E., 2000. Circadian clock resettingby sleep deprivation without exercise in the Syrianhamster. J. Neurosci. 20, 9326–9332.

Antle, M.C., Marchant, E.G., Niel, L., Mistlberger, R.E., 1998.Serotonin antagonists do not attenuate activity-induced phaseshifts of circadian rhythms in the Syrian hamster. Brain Res.813, 139–149.

Antle, M.C., Glass, J.D., Mistlberger, R.E., 2000. 5-HT1A autoreceptorantagonist-induced 5-HT release in the hamstersuprachiasmatic nuclei: effects on circadian clock resetting.Neurosci. Lett. 282, 97–100.

Antle, M.C., Foley, D.K., Foley, N.C., Silver, R., 2003a. Gates andoscillators: a network model of the brain clock. J. Biol. Rhythms18, 339–350.

Antle, M.C., Ogilvie, M.D., Pickard, G.E., Mistlberger, R.E., 2003b.Response of the mouse circadian system to serotonin 1A/2/7agonists in vivo: surprisingly little. J. Biol. Rhythms 18, 145–158.

Antle, M.C., Kriegsfeld, L.J., Silver, R., 2005. Signaling within themaster clock of the brain: localized activation ofmitogen-activated protein kinase by gastrin-releasing peptide.J. Neurosci. 25, 2447–2454.

Artinian, L.R., Ding, J.M., Gillette, M.U., 2001. Carbonmonoxide andnitric oxide: interacting messengers in muscarinic signaling tothe brain's circadian clock. Exp. Neurol. 171, 293–300.

Arvanitogiannis, A., Amir, S., 1999. Resetting the rat circadianclock by ultra-short light flashes. Neurosci. Lett. 261, 159–162.

Arvanitogiannis, A., Robinson, B., Beaulé, C., Amir, S., 2000.Calbindin-D28k immunoreactivity in the suprachiasmaticnucleus and the circadian response to constant light in the rat.Neuroscience 99, 397–401.

Aston-Jones, G., Chen, S., Zhu, Y., Oshinsky, M.L., 2001. A neuralcircuit for circadian regulation of arousal. Nat. Neurosci. 4,732–738.

Aton, S.J., Colwell, C.S., Harmar, A.J., Waschek, J., Herzog, E.D.,2005. Vasoactive intestinal polypeptide mediates circadianrhythmicity and synchrony in mammalian clock neurons. Nat.Neurosci. 8, 476–483.

Aujard, F., Herzog, E.D., Block, G.D., 2001. Circadian rhythms infiring rate of individual suprachiasmatic nucleus neuronsfrom adult and middle-aged mice. Neuroscience 106,255–261.

Azmitia, E.C., Segal, M., 1978. An autoradiographic analysis of thedifferential ascending projections of the dorsal and medianraphe nuclei in the rat. J. Comp. Neurol. 179, 641–668.

Bamshad, M., Aoki, V.T., Adkison, M.G., Warren, W.S., Bartness,T.J., 1998. Central nervous system origins of the sympatheticoutflow to white adipose tissue. Am. J. Physiol. 275, R291–R299.

Bamshad, M., Song, C.K., Bartness, T.J., 1999. CNS origins of thesympathetic nervous system outflow to brown adipose tissue.Am. J. Physiol. 276, R1569–R1578.

Barakat, M.T., O'Hara, B.F., Cao, V.H., Larkin, J.E., Heller, H.C., Ruby,N.F., 2004. Light pulses do not induce c-fos or per1 in the SCN ofhamsters that fail to reentrain to the photocycle. J. Biol.Rhythms 19, 287–297.

Bartness, T.J., Song, C.K., Demas, G.E., 2001. SCN efferents toperipheral tissues: implications for biological rhythms. J. Biol.Rhythms 16, 196–204.

Baslow, M.H., 2000. Functions of N-acetyl-L-aspartate andN-acetyl-L-aspartylglutamate in the vertebrate brain: role inglial cell-specific signaling. J. Neurochem. 75, 453–459.

Battaglia, G., Sharkey, J., Kuhar, M.J., de Souza, E.B., 1991.Neuroanatomic specificity and time course of alterations in ratbrain serotonergic pathways induced by MDMA(3,4-methylenedioxymethamphetamine): assessment usingquantitative autoradiography. Synapse 8, 249–260.

Beaulé, C., Amir, S., 2002. Effect of 192 IgG-saporin on circadianactivity rhythms, expression of p75 neurotrophin receptors,


calbindin-D28K, and light-induced fos in the suprachiasmaticnucleus in rats. Exp. Neurol. 176, 377–389.

Beaulé, C., Arvanitogiannis, A., Amir, S., 2001a. Entrainment to adiscrete light pulse at dawn is associated with acutesuppression of FOS in the shell region of the suprachiasmaticnucleus. Soc. Neurosci. (abstract 182.6).

Beaulé, C., Arvanitogiannis, A., Amir, S., 2001b. Light suppressesfos expression in the shell region of the suprachiasmaticnucleus at dusk and dawn: implications for photic entrainmentof circadian rhythms. Neuroscience 106, 249–254.

Beaule, C., Houle, L.M., Amir, S., 2003a. Expression profiles ofPER2 immunoreactivity within the shell and core regions ofthe rat suprachiasmatic nucleus: lack of effect of photicentrainment and disruption by constant light. J. Mol.Neurosci. 21, 133–147.

Beaulé, C., Robinson, B., Lamont, E.W., Amir, S., 2003b. Melanopsinin the circadian timing system. J. Mol. Neurosci. 21, 73–89.

Belenky, M.A., Pickard, G.E., 2001. Subcellular distribution of 5-HT(1B) and 5-HT(7) receptors in the mouse suprachiasmaticnucleus. J. Comp. Neurol. 432, 371–388.

Belenky, M.A., Smeraski, C.A., Provencio, I., Sollars, P.J.,Pickard, G.E., 2003. Melanopsin retinal ganglion cells receivebipolar and amacrine cell synapses. J. Comp. Neurol. 460,380–393.

Bergstrom, A.L., Hannibal, J., Hindersson, P., Fahrenkrug, J., 2003.Light-induced phase shift in the Syrian hamster (Mesocricetusauratus) is attenuated by the PACAP receptor antagonistPACAP6-38 or PACAP immunoneutralization. Eur. J. Neurosci.18, 2552–2562.

Berson, D.M., Dunn, F.A., Takao, M., 2002. Phototransduction byretinal ganglion cells that set the circadian clock. Science 295,1070–1073.

Best, J.D., Maywood, E.S., Smith, K.L., Hastings, M.H., 1999. Rapidresetting of the mammalian circadian clock. J. Neurosci. 19,828–835.

Biello, S.M., 1995. Enhanced photic phase shifting after treatmentwith antiserum to neuropeptide Y. Brain Res. 673, 25–29.

Biello, S.M., Dafters, R.I., 2001. MDMA and fenfluramine alter theresponse of the circadian clock to a serotonin agonist in vitro.Brain Res. 920, 202–209.

Biello, S.M., Mrosovsky, N., 1993. Circadian phase-shifts inducedby chlordiazepoxide without increased locomotor activity.Brain Res. 622, 58–62.

Biello, S.M., Mrosovsky, N., 1995. Blocking the phase-shifting effectof neuropeptide Y with light. Proc. R. Soc. Lond., B Biol. Sci. 259,179–187.

Biello, S.M., Janik, D., Mrosovsky, N., 1994. Neuropeptide Y andbehaviorally induced phase shifts. Neuroscience 62, 273–279.

Biello, S.M., Golombek, D.A., Harrington, M.E., 1997a. NeuropeptideY and glutamate block each other's phase shifts in thesuprachiasmatic nucleus in vitro. Neuroscience 77, 1049–1057.

Biello, S.M., Golombek, D.A., Schak, K.M., Harrington, M.E., 1997b.Circadian phase shifts to neuropeptide Y in vitro: cellularcommunication and signal transduction. J. Neurosci. 17,8468–8475.

Bina, K.G., Rusak, B., 1996. Muscarinic receptors mediatecarbachol-induced phase shifts of circadian activity rhythmsin Syrian hamsters. Brain Res. 743, 202–211.

Bina, K.G., Rusak, B., Semba, K., 1993. Localization of cholinergicneurons in the forebrain and brainstem that project to thesuprachiasmatic nucleus of the hypothalamus in rat. J. Comp.Neurol. 335, 295–307.

Bina, K.G., Rusak, B., Semba, K., 1997. Sources of p75-nerve growthfactor receptor-like immunoreactivity in the ratsuprachiasmatic nucleus. Neuroscience 77, 461–472.

Bittman, E.L., Goldman, B.D., 1979. Serum levels of gonadotrophinsin hamsters exposed to short photoperiods: effects ofadrenalectomy and ovariectomy. J. Endocrinol. 83, 113–118.

Blasiak, T., Lewandowski, M.H., 2003. Dorsal raphe nucleus

modulates neuronal activity in rat intergeniculate leaflet.Behav. Brain Res. 138, 179–185.

Blasiak, T., Lewandowski, M.H., 2004a. Blockade of GABAAreceptors disrupts isoperiodic neuronal oscillations in theintergeniculate leaflet of the rat. Brain Res. 1009, 82–87.

Blasiak, A., Lewandowski, M.H., 2004b. In vitro extracellularrecording of spontaneous activity of the intergeniculate leafletneurons. Brain Res. 1015, 82–86.

Bobrzynska, K.J., Vrang, N., Mrosovsky, N., 1996. Persistence ofnonphotic phase shifts in hamsters after serotonin depletionin the suprachiasmatic nucleus. Brain Res. 741, 205–214.

Boer, G.J., van Esseveldt, K.E., van der Geest, B.A., Duindam, H.,Rietveld, W.J., 1999. Vasopressin-deficient suprachiasmaticnucleus grafts re-instate circadian rhythmicity insuprachiasmatic nucleus-lesioned arrhythmic rats.Neuroscience 89, 375–385.

Bos, N.P.A., Mirmiran, M., 1993. Effects of excitatory and inhibitoryamino acids on neuronal discharges in the culturedsuprachiasmatic nucleus. Brain Res. Bull. 31, 67–72.

Botchkina, G.I., Morin, L.P., 1993. Development of the hamsterserotoninergic system: cell groups and diencephalicprojections. J. Comp. Neurol. 338, 405–431.

Botchkina, G.I., Morin, L.P., 1995. Specialized neuronal and glialcontributions to development of the hamster lateral geniculatecomplex and circadian visual system. J. Neurosci. 15, 190–201.

Boulos, Z., Morin, L.P., 1985. Entrainment of split circadian activityrhythms in hamsters. J. Biol. Rhythms 1, 1–15.

Boulos, Z., Rusak, B., 1982. Circadian phase response curves fordark pulses in the hamster. J. Comp. Physiol. 146, 411–417.

Bouskila, Y., Dudek, F.E., 1995. A rapidly activating type of outwardrectifier K+ current and A-current in rat suprachiasmaticnucleus neurones. J. Physiol. 488 (Pt. 2), 339–350.

Bradbury, M.J., Dement, W.C., Edgar, D.M., 1997.Serotonin-containing fibers in the suprachiasmatichypothalamus attenuate light-induced phase delays in mice.Brain Res. 768, 125–134.

Bramley, J.R., Sollars, P.J., Pickard, G.E., Dudek, F.E., 2005. 5-HT1Breceptor-mediated presynaptic inhibition of GABA release inthe suprachiasmatic nucleus. J. Neurophysiol. 93, 3157–3164.

Bryant, D.N., LeSauter, J., Silver, R., Romero, M.T., 2000. Retinalinnervation of calbindin-D28K cells in the hamstersuprachiasmatic nucleus: ultrastructural characterization.J. Biol. Rhythms 15, 103–111.

Buijs, R.M., Chun, S.J., Niijima, A., Romijn, H.J., Nagai, K., 2001.Parasympathetic and sympathetic control of the pancreas: arole for the suprachiasmatic nucleus and other hypothalamiccenters that are involved in the regulation of food intake.J. Comp. Neurol. 431, 405–423.

Burgoon, P.W., Lindberg, P.T., Gillette, M.U., 2004. Differentpatterns of circadian oscillation in the suprachiasmaticnucleus of hamster, mouse, and rat. J. Comp. Physiol., A 190,167–171.

Byku, M., Gannon, R.L., 2000a. SNC 80, a delta-opioid agonist,elicits phase advances in hamster circadian activity rhythms.NeuroReport 11, 1449–1452.

Byku, M., Gannon, R.L., 2000b. Opioid induced non-photic phaseshifts of hamster circadian activity rhythms. Brain Res. 873,189–196.

Byku, M., Legutko, R., Gannon, R.L., 2000. Distribution of δ opioidreceptor immunoreactivity in the hamster suprachiasmaticnucleus and intergeniculate leaflet. Brain Res. 857, 1–7.

Cagampang, F.R., Piggins, H.D., Sheward, W.J., Harmar, A.J., Coen,C.W., 1998a. Circadian changes in PACAP type 1 (PAC1) receptormRNA in the rat suprachiasmatic and supraoptic nuclei. BrainRes. 813, 218–222.

Cagampang, F.R., Sheward, W.J., Harmar, A.J., Piggins, H.D., Coen,C.W., 1998b. Circadian changes in the expression of vasoactiveintestinal peptide 2 receptor mRNA in the rat suprachiasmaticnuclei. Brain Res. Mol. Brain Res. 54, 108–112.


Caillol, M., Devinoy, E., Lacroix, M.C., Schirar, A., 2000. Endothelialand neuronal nitric oxide synthases are present in thesuprachiasmatic nuclei of Syrian hamsters and rats. Eur. J.Neurosci. 12, 649–661.

Canbeyli, R.S., Lehman, M.N., Silver, R., 1991. Tracing SCN graftefferents with Dil. Brain Res. 554, 15–21.

Card, J.P., Moore, R.Y., 1982. Ventral lateral geniculate nucleusefferents to the rat suprachiasmatic nucleus exhibit avianpancreatic polypeptide-like immunoreactivity. J. Comp.Neurol.206, 390–396.

Card, J.P., Moore, R.Y., 1984. The suprachiasmatic nucleus of thegolden hamster: immunohistochemical analysis of cell andfiber distribution. Neuroscience 13, 390–396.

Card, J.P., Moore, R.Y., 1989. Organization of lateralgeniculate–hypothalamic connections in the rat. J. Comp.Neurol. 284, 135–147.

Card, J.P., Brecha, N., Moore, R.Y., 1983. Immunohistochemicallocalization of avian pancreatic polypeptide-likeimmunoreactivity in the rat hypothalamus. J. Comp. Neurol.217, 123–136.

Cassone, V.M., Speh, J.C., Card, J.P., 1988. Comparative anatomy ofthemammalianhypothalamic suprachiasmatic nucleus. J. Biol.Rhythms 3, 71–91.

Castel, M., Morris, J.F., 2000. Morphological heterogeneity of theGABAergic network in the suprachiasmatic nucleus, the brain'scircadian pacemaker. J. Anat. 196, 1–13.

Castel, M., Belenky, M., Cohen, S., Ottersen, O.P., Storm-Mathisen,J., 1993. Glutamate-like immunoreactivity in retinal terminalsof the mouse suprachiasmatic nucleus. Eur. J. Neurosci. 5,368–381.

Challet, E., Naylor, E., Metzger, J.M., MacIntyre, D.E., Turek, F.W.,1998a. An NK1 receptor antagonist affects the circadianregulation of locomotor activity in golden hamsters. Brain Res.800, 32–39.

Challet, E., Scarbrough, K., Penev, P.D., Turek, F., 1998b. Roles ofsuprachiasmatic nuclei and intergeniculate leaflets inmediating the phase-shifting effects of a serotonergic agonistand their photic modulation during subjective day. J. Biol.Rhythms 12, 410–421.

Challet, E., Dugovic, C., Turek, F.W., Olivier, V.R., 2001. Theselective neurokinin 1 receptor antagonist R116301 modulatesphotic responses of the hamster circadian system.Neuropharmacology 40, 408–415.

Chambille, I., Doyle, S., Servière, J., 1993. Photic induction andcircadian expression of Fos-like protein.Immunohistochemical study in the retina andsuprachiasmatic nuclei of hamster. Brain Res. 612, 138–150.

Chen, D., Hurst, W.J., Ding, J.M., Faiman, L.E., Mayer, B., Gillette,M.U., 1997. Localization and characterization of nitric oxidesynthase in the rat suprachiasmatic nucleus: evidence for anitrergic plexus in the biological clock. J. Neurochem. 68,855–861.

Chen, D., Buchanan, G.F., Ding, J.M., Hannibal, J., Gillette, M.U.,1999. Pituitary adenylyl cyclase-activating peptide: a pivotalmodulator of glutamatergic regulation of thesuprachiasmatic circadian clock. Proc. Natl. Acad. Sci.U. S. A. 96, 13468–13473.

Chevassus, A.L., Cooper, H.M., 1998. Is there ageniculohypothalamic tract in primates? A comparativeimmunohistochemical study in the circadian system ofstrepsirhine and haplorhine species. Brain Res. 805, 213–219.

Christian, C.A., Harrington, M.E., 2002. Three days of novelwheel access diminishes light-induced phase delays in vivowith no effect on per1 induction by light. Chronobiol. Int.19, 671–682.

Cloues, R.K., Sather, W.A., 2003. Afterhyperpolarization regulatesfiring rate in neurons of the suprachiasmatic nucleus.J. Neurosci. 23, 1593–1604.

Colbron, S., Jones, M., Biello, S.M., 2002. MDMA alters the response

of the circadian clock to a photic and non-photic stimulus.Brain Res. 956, 45–52.

Colwell, C.S., 2000. Circadian modulation of calcium levels in cellsin the suprachiasmatic nucleus. Eur. J. Neurosci. 12, 571–576.

Colwell, C.S., Menaker, M., 1992. NMDA as well as non-NMDAreceptor antagonists can prevent the phase-shifting effects oflight on the circadian system of the golden hamster. J. Biol.Rhythms 7, 125–136.

Colwell, C.S., Ralph, M.R., Menaker, M., 1990. Do NMDA receptorsmediate the effects of light on circadian behavior. Brain Res.523, 117–120.

Colwell, C.S., Foster, R.G., Menaker, M., 1991. NMDA receptorantagonists block the effects of light on circadian behavior inthe mouse. Brain Res. 554, 105–110.

Colwell, C.S., Kaufman, C.M., Menaker, M., 1993a. Photic inductionof Fos in the hamster suprachiasmatic nucleus is inhibited bybaclofen but not by diazepam or bicucullin. Neurosci. Lett. 163,177–181.

Colwell, C.S., Kaufman, C.M., Menaker, M., Ralph, M.R., 1993b.Light-induced phase shifts and Fos expression in the hamstercircadian system: the effects of anesthetics. J. Biol. Rhythms 8,179–188.

Colwell, C.S., Kaufman, C.M., Menaker, M., 1993c. Phase-shiftingmechanisms in themammalian circadian system: new light onthe carbachol paradox. J. Neurosci. 13, 1454–1459.

Colwell, C.S., Michel, S., Itri, J., Rodriguez, W., Tam, J., Lelievre, V.,Hu, Z., Liu, X., Waschek, J.A., 2003. Disrupted circadian rhythmsin VIP- and PHI-deficient mice. Am. J. Physiol.: Regul., Integr.Comp. Physiol. 285, R939–R949.

Colwell, C.S., Michel, S., Itri, J., Rodriguez, W., Tam, J., Lelievre, V.,Hu, Z., Waschek, J.A., 2004. Selective deficits in the circadianlight response in mice lacking PACAP. Am. J. Physiol.: Regul.,Integr. Comp. Physiol. 287, R1194–R1201.

Coogan, A.N., Rawlings, N., Luckman, S.M., Piggins, H.D., 2001.Effects of neurotensin ondischarge rates of rat suprachiasmaticnucleus neurons in vitro. Neuroscience 103, 663–672.

Cooper, H.M., Parvopassu, F., Herbin, M., Magnin, M., 1994.Neuroanatomical pathways linking vision and olfaction inmammals. Psychoneuroendocrinology 19, 623–639.

Coyle, J.T., 1997. The nagging question of the function ofN-acetylaspartylglutamate. Neurobiol. Dis. 4, 231–238.

Crill, W.E., 1996. Persistent sodium current in mammalian centralneurons. Annu. Rev. Physiol. 58, 349–362.

Cutler, D.J., Munekawa, K., Mason, R., 1999. Electrophysiologicaleffects of opioid receptor activation on Syrian hamstersuprachiasmatic nucleus neurones in vitro. Brain Res. Bull. 50,119–125.

Cutler, D.J., Hardin, H., Rees, H., Maywood, E., Hastings, M.H.,Shena, H.M., Martens, J.M., Harrer, C., Piggins, H.D., 2002. Themouse VPAC2 receptor confers SCN cellular rhythmicity andresponsiveness to vasoactive intestinal polypeptide (VIP) invitro. Soc. Res. Biol. Rhythms 8, 37.

Cutler, D.J., Haraura, M., Reed, H.E., Shen, S., Sheward, W.J.,Morrison, C.F., Marston, H.M., Harmar, A.J., Piggins, H.D., 2003.The mouse VPAC2 receptor confers suprachiasmatic nucleicellular rhythmicity and responsiveness to vasoactiveintestinal polypeptide in vitro. Eur. J. Neurosci. 17, 197–204.

Dacey, D.M., Liao, H.W., Peterson, B.B., Robinson, F.R., Smith, V.C.,Pokorny, J., Yau, K.W., Gamlin, P.D., 2005.Melanopsin-expressing ganglion cells in primate retina signalcolour and irradiance and project to the LGN. Nature 433,749–754.

Daikoku, S., Hisano, S., Kagotani, Y., 1992. Neuronal associations inthe rat suprachiasmatic nucleus demonstrated byimmunoelectron microscopy. J. Comp. Neurol. 325, 559–571.

Dardente, H., Poirel, V.J., Klosen, P., Pevet, P., Masson-Pevet, M.,2002. Per and neuropeptide expression in the ratsuprachiasmatic nuclei: compartmentalization anddifferential cellular induction by light. Brain Res. 958, 261–271.


Decker, K., Reuss, S., 1994. Nitric oxide-synthesizing neurons inthe hamster suprachiasmatic nucleus: a combined NOS- andN. Brain Res. 666, 284–288.

De Jeu, M., Pennartz, C.M.A., 1997. Functional characterization ofthe H-current in SCN neurons in subjective day and night: awhole-cell patch-clamp study in acutely prepared brain slices.Brain Res. 767, 72–80.

De Jeu, M., Pennartz, C.M.A., 2002. Circadian modulation of GABAfunction in the rat suprachiasmatic nucleus: excitatory effectsduring the night phase. J. Neurophysiol. 87, 834–844.

De Jeu, M., Hermes, N., Pennartz, C.M.A., 1998. Circadianmodulation of membrane properties in slices of ratsuprachiasmatic nucleus. NeuroReport 9, 3725–3729.

De Jeu, M., Geurtsen, A., Pennartz, C.M.A., 2002. A Ba(2+)-sensitiveK(+) current contributes to the resting membrane potential ofneurons in rat suprachiasmatic nucleus. J. Neurophysiol. 88,349–878.

De la Iglesia, H.O., Schwartz, W.J., 2002. A subpopulation ofefferent neurons in the mouse suprachiasmatic nucleus is alsolight responsive. NeuroReport 13, 857–860.

De la Iglesia, H.O., Blaustein, J.D., Bittman, E.L., 1995. Thesuprachiasmatic area in the female hamster projects toneurons containing estrogen receptors and GnRH.NeuroReport 6, 1715–1722.

De la Iglesia, H.O., Meyer, J., Carpino Jr., A., Schwartz, W.J., 2000.Antiphase oscillation of the left and right suprachiasmaticnuclei. Science 290, 799–801.

De la Iglesia, H.O., Meyer, J., Schwartz, W.J., 2003. Lateralization ofcircadian pacemaker output: activation of left- and right-sidedluteinizing hormone-releasing hormone neurons involves aneural rather than a humoral pathway. J. Neurosci. 23,7412–7414.

De Vries, M.J., Meijer, J.H., 1991. Aspartate injections into thesuprachiasmatic region of the Syrian hamster do not mimicthe effects of light on the circadian activity rhythm. Neurosci.Lett. 127, 215–218.

De Vries, M.J., Nunes Cardozo, B., Van der Want, J., De Wolf, A.,Meijer, J.H., 1993. Glutamate immunoreactivity in terminals ofthe retinohypothalamic tract of the brown Norwegian rat.Brain Res. 612, 231–237.

Devries, M.J., Lakke, E.J.F., 1995. Retrograde labeling of retinalganglion-cells and brain neuronal subsets by [H-3]D-aspartateinjection in the Syrian-hamster hypothalamus. Brain Res. Bull.38, 349–354.

Ding, J.M., Chen, D., Weber, E.T., Faiman, L.E., Rea, M.A., Gillette,M.U., 1994. Resetting the biological clock:mediation of nocturnalcircadian shifts by glutamate and NO. Science 266, 1713–1717.

Ding, J.M., Faiman, L.E., Hurst,W.J., Kuriashkina, L.R., Gillette, M.U.,1997. Resetting the biological clock: mediation of nocturnalCREB phosphorylation via light, glutamate, and nitric oxide.J. Neurosci. 17, 667–675.

Dkhissi-Benyahya, O., Sicard, B., Cooper, H.M., 2000. Effects ofirradiance and stimulus duration on early gene expression(Fos) in the suprachiasmatic nucleus: temporal summationand reciprocity. J. Neurosci. 20, 7790–7797.

Doan, A., Urbanski, H.F., 1994. Diurnal expression of Fos inluteinizing hormone-releasing hormone neurons of Syrianhamsters. Biol. Reprod. 50, 301–308.

Donaldson, J.A., Stephan, F.K., 1982. Entrainment of circadianrhythms: retinofugal pathways and unilateral suprachiasmaticnucleus lesions. Physiol. Behav. 29, 1161–1169.

Dudley, T.E., DiNardo, L.A., Glass, J.D., 1998. Endogenousregulation of serotonin release in the hamster suprachiasmaticnucleus. J. Neurosci. 18, 5045–5052.

Dudley, T.E., DiNardo, L.A., Glass, J.D., 1999. In vivo assessment ofthe midbrain raphe nuclear regulation of serotonin release inthe hamster suprachiasmatic nucleus. J. Neurophysiol. 81,1468–1477.

Duncan, M.J., Short, J., Wheeler, D.L., 1999. Comparison of the

effects of aging on 5-HT7 and 5-HT1A receptors in discreteregions of the circadian timing system in hamsters. Brain Res.829, 39–45.

Duncan, M.J., Jennes, L., Jefferson, J.B., Brownfield, M.S., 2000.Localization of serotonin5A receptors in discrete regions of thecircadian timing system in the Syrian hamster. Brain Res. 869,178–185.

Duncan, M.J., Grear, K.E., Hoskins, M.A., 2004. Aging andSB-269970-A, a selective 5-HT7 receptor antagonist, attenuatecircadian phase advances induced by microinjections ofserotonergic drugs in the hamster dorsal raphe nucleus. BrainRes. 1008, 40–48.

Dwyer, S.M., Rosenwasser, A.M., 2000. Effects of light intensity andrestraint ondark-pulse-inducedcircadianphaseshiftingduringsubjectivenight inSyrianhamsters. J. Biol. Rhythms15,491–500.

Ebling, F.J.P., 1996. The role of glutamate in the photic regulation ofthe suprachiasmatic nucleus. Prog. Neurobiol. 50, 109–123.

Edelstein, K., Amir, S., 1998. Glutamatergic antagonists do notattenuate light-induced Fos protein in rat intergeniculateleaflet. Brain Res. 810, 264–268.

Edelstein, K., Amir, S., 1999a. The intergeniculate leaflet does notmediate the disruptive effects of constant light on circadianrhythms in the rat. Neuroscience 90, 1093–1101.

Edelstein, K., Amir, S., 1999b. The role of the intergeniculate leafletin entrainment of circadian rhythms to a skeletonphotoperiod. J. Neurosci. 19, 372–380.

Edelstein, K., Mrosovsky, N., 2001. Behavioral responses to light inmice with dorsal lateral geniculate lesions. Brain Res. 918,107–112.

Edelstein, K., De la Iglesia, H.O., Schwartz, W.J., Mrosovsky, N.,2003a. Behavioral arousal blocks light-induced phase advancesin locomotor rhythmicity but not light-induced Per1 and fosexpression in the hamster suprachiasmatic nucleus.Neuroscience 118, 253–261.

Edelstein, K., De la Iglesia, H.O., Mrosovsky, N., 2003b. Periodgene expression in the suprachiasmatic nucleus ofbehaviorally decoupled hamsters. Brain Res. Mol. Brain Res.114, 40–45.

Edgar, D.M., Reid, M.S., Dement, W.C., 1997. Serotonergic afferentsmediate activity-dependent entrainment of the mousecircadian clock. Am. J. Physiol. 273, R265–R269.

Ehlen, J.C., Grossman, G.H., Glass, J.D., 2001. In vivo resetting of thehamster circadian clock by 5-HT7 receptors in thesuprachiasmatic nucleus. J. Neurosci. 21, 5351–5357.

Elliott, J.A., Stetson, M.H., Menaker, M., 1972. Regulation of testisfunction in golden hamsters: a circadian clock measuresphotoperiodic time. Science 178, 771–773.

Enaker, M., Roberts, R., Elliott, J., Underwood, H., 1970. Extraretinallight perception in the sparrow: 3. The eyes do not participatein photoperiodic photoreception. Proc. Natl. Acad. Sci. U. S. A.67, 320–325.

Erhardt, C., Galani, R., Jeltsch, H., Cassel, J.C., Klosen, P., Menet, J.S.,Pevet, P., Challet, E., 2004. Modulation of photic resetting in ratsby lesions of projections to the suprachiasmatic nucleiexpressing p75 neurotrophin receptor. Eur. J. Neurosci. 19,1773–1788.

Erlander, M.G., Tillakaratne, N.J., Feldblum, S., Patel, N., Tobin, A.J.,1991. Two genes encode distinct glutamate decarboxylases.Neuron 7, 91–100.

Fedorkova, L., Rutishauser, U., Prosser, R., Shen, H., Glass, J.D.,2002. Removal of polysialic acid from the SCN potentiatesnonphotic circadian phase resetting. Physiol. Behav. 77,361–369.

Ferguson, S.A., Kennaway, D.J., 1999. Emergence of alteredcircadian timing in a cholinergically supersensitive rat line.Am. J. Physiol. 277, R1171–R1178.

Ferguson, S.A., Kennaway, D.J., Moyer, R.W., 1999. Nicotine phaseshifts the 6-sulphatoxymelatonin rhythm and induces c-Fos inthe SCN of rats. Brain Res. Bull. 48, 527–538.


Ferreyra, G.A., Cammarota, M.P., Golombek, D.A., 1998. Photiccontrol of nitric oxide synthase activity in the hamstersuprachiasmatic nuclei. Brain Res. 797, 190–196.

Field,M.D.,Maywood, E.S., O'Brien, J.A.,Weaver, D.R., Reppert, S.M.,Hastings, M.H., 2000. Analysis of clock proteins in mouse SCNdemonstrates phylogenetic divergence of the circadianclockwork and resetting mechanisms. Neuron 25, 437–447.

Fite, K.V., Janusonis, S., 2001. Retinal projection to the dorsal raphenucleus in the Chilean degus (Octodon degus). Brain Res. 895,139–145.

Fite, K.V., Janusonis, S., Foote, W., Bengston, L., 1999. Retinalafferents to the dorsal raphe nucleus in rats and Mongoliangerbils. J. Comp. Neurol. 414, 469–484.

Fite, K.V., Birkett, M.A., Smith, A., Janusonis, S.F., McLaughlin, S.,2003. Retinal ganglion cells projecting to the dorsal raphe andlateral geniculate complex in Mongolian gerbils. Brain Res. 973,146–150.

Foote, W.E., Taber-Pierce, E., Edwards, L., 1978. Evidence for aretinal projection to the midbrain raphe of the cat. Brain Res.156, 135–140.

Foster, R.G., Provencio, I., Hudson, D., Fiske, S., De Grip, W.,Menaker, M., 1991. Circadian photoreception in the retinallydegenerate mouse (rd/rd). J. Comp. Physiol., A 169, 39–50.

Foster, R.G., Provencio, I., Bovee-Geurts, P.H., DeGrip, W.J., 2003.The photoreceptive capacity of the developing pineal glandand eye of the golden hamster (Mesocricetus auratus).J. Neuroendocrinol. 15, 355–363.

Freedman, M.S., Lucas, R.J., Soni, B., Von Schantz, M., Munoz, M.,David-Gray, Z., Foster, R.G., 1999. Regulation of mammaliancircadian behavior by non-rod, non-cone, ocularphotoreceptors. Science 284, 502–504.

Freeman, M.E., Kanyicska, B., Lerant, A., Nagy, G., 2000. Prolactin:structure, function, and regulation of secretion. Physiol. Rev.80, 1523–1631.

Fujiyama, F., Hioki, H., Tomioka, R., Taki, K., Tamamaki, N.,Nomura, S., Okamoto, K., Kaneko, T., 2003. Changes ofimmunocytochemical localization of vesicular glutamatetransporters in the rat visual system after the retinofugaldenervation. J. Comp. Neurol. 465, 234–249.

Fukuhara, C., Brewer, J.M., Dirden, J.C., Bittman, E.L., Tosini, G.,Harrington, M.E., 2001. Neuropeptide Y rapidly reduces Period 1and Period 2 mRNA levels in the hamster suprachiasmaticnucleus. Neurosci. Lett. 314, 119–122.

Gamble, K.L., Novak, C.M., Paul, K.N., Albers, H.E., 2003.Tetrodotoxin blocks the circadian effects of NMDA during theday but not at night. NeuroReport 14, 641–644.

Gamble, K.L., Novak, C.M., Albers, H.E., 2004. Neuropeptide Y andN-methyl-D-aspartic acid interact within the suprachiasmaticnuclei to alter circadian phase. Neuroscience 126, 559–565.

Gannon, R.L., 2001. 5HT7 receptors in the rodent suprachiasmaticnucleus. J. Biol. Rhythms 16, 19–24.

Gannon, R.L., 2003. Serotonergic serotonin1a mixed agonists/antagonists elicit large-magnitude phase shifts in hamstercircadian wheel-running rhythms. Neuroscience 119, 567–576(Ref Type: Generic).

Gaston, S., Menaker, M., 1968. Pineal function: the biological clockin the sparrow? Science 160, 1125–1127.

Gerfen, C.R., Sawchenko, P.E., 1984. An anterogradeneuroanatomical tracing method that shows the detailedmorphology of neurons, their axons and terminals:immunohistochemical localization of an axonallytransported plant lectin, Phaseolus vulgaris leucoagglutinin(PHA-L). Brain Res. 290, 219–238.

Geusz, M.E., Fletcher, C., Block, G.D., Straume, M., Copeland, N.G.,Jenkins, N.A., Kay, S.A., Day, R.N., 1997. Long-term monitoringof circadian rhythms in c-fos gene expression fromsuprachiasmatic nucleus cultures. Curr. Biol. 7, 758–766.

Gillespie, C.F., Huhman, K.L., Babagbemi, T.O., Albers, H.E., 1996.Bicuculline increases and muscimol reduces the

phase-delaying effects of light and VIP/PHI/GRP in thesuprachiasmatic region. J. Biol. Rhythms 11, 137–144.

Gillespie, C.F., Mintz, E.M., Marvel, C.L., Huhman, K.L., Albers, H.E.,1997. GABA(A) and GABA(B) agonists and antagonists alter thephase-shifting effects of light when microinjected into thesuprachiasmatic region. Brain Res. 759, 181–189.

Gillespie, C.F., Van der Beek, E.M., Mintz, E.M., Mickley, N.C.,Jasnow, A.M., Huhman, K.L., Albers, H.E., 1999. GABAergicregulation of light-induced c-Fos immunoreactivity withinthe suprachiasmatic nucleus. J. Comp. Neurol. 411,683–692.

Gillette, M.U., 1991. SCN electrophysiology in vitro: rhythmicactivity and endogenous clock properties. In: Klein, D.C.,Moore, R.Y., Reppert, S.M. (Eds.), Suprachiasmatic Nucleus: TheMind's Clock. Oxford Univ. Press, New York, pp. 125–143.

Gillette, M.U., DeMarco, S.J., Ding, J.M., Gallman, E.A., Faiman, L.E.,Liu, C., McArthur, A.J., Medanic, M., Richard, D., Tcheng, T.K.,Weber, E.T., 1993. The organization of the suprachiasmaticcircadian pacemaker of the rat and its regulation byneurotransmitters andmodulators. J. Biol. Rhythms 8, S53–S58.

Gillette, M.U., Buchanan, G.F., Artinian, L., Hamilton, S.E.,Nathanson, N.M., Liu, C., 2001. Role of the M1 receptor inregulating circadian rhythms. Life Sci. 68, 2467–2472.

Ginty, D.D., Kornhauser, J.M., Thompson, M.A., Bading, H., Mayo,K.E., Takahashi, J.S., Greenberg, M.E., 1993. Regulation of CREBphosphorylation in the suprachiasmatic nucleus by light anda circadian clock. Science 260, 238–241.

Glass, J.D., Chen, L., 1999. Serotonergic modulation of astrocyticactivity in the hamster suprachiasmatic nucleus. Neuroscience94, 1253–1259.

Glass, J.D., Hauser, U.E., Blank, J.L., Selim, M., Rea, M.A., 1993.Differential timing of amino acid and 5-HIAA rhythms insuprachiasmatic hypothalamus. Am. J. Physiol. 265,R504–R511.

Glass, J.D., Lee, W., Shen, H., Watanabe, M., 1994. Expression ofimmunoreactive polysialylated neural cell adhesion moleculein the suprachiasmatic nucleus. Neuroendocrinology 60,87–95.

Glass, J.D., Shen, H., Fedorkova, L., Chen, L., Tomasiewicz, H.,Watanabe, M., 2000a. Polysialylated neural cell adhesionmolecule modulates photic signaling in the mousesuprachiasmatic nucleus. Neurosci. Lett. 280, 207–210.

Glass, J.D., DiNardo, L.A., Ehlen, J.C., 2000b. Dorsal raphe nuclearstimulation of SCN serotonin release and circadianphase-resetting. Brain Res. 859, 224–232.

Glass, J.D., Watanabe, M., Fedorkova, L., Shen, H., Ungers, G.,Rutishauser, U., 2003a. Dynamic regulation of polysialylatedneural cell adhesion molecule in the suprachiasmatic nucleus.Neuroscience 117, 203–211.

Glass, J.D., Grossman, G.H., Farnbauch, L., DiNardo, L., 2003b.Midbrain raphe modulation of nonphotic circadian clockresetting and 5-HT release in themammalian suprachiasmaticnucleus. J. Neurosci. 23, 7451–7460.

Goel, N., Lee, T.M., Smale, L., 1999. Suprachiasmatic nucleus andintergeniculate leaflet in the diurnal rodent Octodon degus:retinal projections and immunocytochemical characterization.Neuroscience 92, 1491–1509.

Gooley, J.J., Saper, C.B., 2003. A broad role for melanopsin innon-visual photoreception based on neuroanatomicalevidence in rats. J. Neurosci. 23, 7093–7106.

Gooley, J.J., Lu, J., Chou, T.C., Scammell, T.E., Saper, C.B., 2001.Melanopsin in cells of origin of the retinohypothalamic tract.Nat. Neurosci. 4, 1165.

Gorman, M.R., Kendall, M., Elliott, J.A., 2005. Scotopic illuminationenhances entrainment of circadian rhythms to lengtheninglight:dark cycles. J. Biol. Rhythms 20, 38–48.

Graham, E.S., Turnbull, Y., Fotheringham, P., Nilaweera, K.,Mercer, J.G., Morgan, P.J., Barrett, P., 2003. Neuromedin U andNeuromedin U receptor-2 expression in the mouse and rat


hypothalamus: effects of nutritional status. J. Neurochem. 87,1165–1173.

Graham, E.S., Littlewood, P., Turnbull, Y., Mercer, J.G., Morgan, P.J.,Barrett, P., 2005. Neuromedin-U is regulated by the circadianclock in the SCN of the mouse. Eur. J. Neurosci. 21, 814–819.

Gribkoff, V.K., Pieschl, R.L., Wisialowski, T.A., Park, W.K., Strecker,G.J., De Jeu, M.T.G., Pennartz, C.M.A., Dudek, F.E., 1999. Areexamination of the role of GABA in the mammaliansuprachiasmatic nucleus. J. Biol. Rhythms 14, 126–130.

Grossman, G.H., Mistlberger, R.E., Antle, M.C., Ehlen, J.C., Glass, J.D.,2000. Sleep deprivation stimulates serotonin release in thesuprachiasmatic nucleus. NeuroReport 11, 1929–1932.

Grossman, G.H., Farnbauch, L., Glass, J.D., 2004. Regulation ofserotonin release in the Syrian hamster intergeniculate leafletregion. NeuroReport 15, 103–106.

Guido, M.E., Goguen, D., De Guido, L., Robertson, H.A., Rusak, B.,1999. Circadian and photic regulation of immediate-early geneexpression in the hamster suprachiasmatic nucleus.Neuroscience 90, 555–571.

Haak, L.L., Heller, H.C., van den Pol, A.N., 1997. Metabrotropicglutamate receptor activationmodulates kainate and serotonincalcium response in astrocytes. J. Neurosci. 17, 1825–1837.

Hamada, T., Yamanouchi, S., Watanabe, A., Shibata, S., Watanabe,S., 1999. Involvement of glutamate release in substanceP-induced phase delays of suprachiasmatic neuron activityrhythm in vitro. Brain Res. 836, 190–193.

Hamada, T., LeSauter, J., Venuti, J.M., Silver, R., 2001. Expression ofPeriod genes: rhythmic and non-rhythmic compartments ofthe suprachiasmatic nucleus pacemaker. J. Neurosci. 21,7742–7750.

Hamada, T., LeSauter, J., Lokshin, M., Romero, M.T., Yan, L.L.,Venuti, J.M., Silver, R., 2003. Calbindin influences response tophotic input in suprachiasmatic nucleus. J. Neurosci. 23,8820–8826.

Hamada, T., Antle, M.C., Silver, R., 2004a. Temporal and spatialexpression patterns of canonical clock genes andclock-controlled genes in the suprachiasmatic nucleus. Eur. J.Neurosci. 19, 1741–1748.

Hamada, T., Antle, M.C., Silver, R., 2004b. The role of Period1 innon-photic resetting of the hamster circadian pacemaker inthe suprachiasmatic nucleus. Neurosci. Lett. 362, 87–90.

Hanada, R., Teranishi, H., Pearson, J.T., Kurokawa, M., Hosoda, H.,Fukushima, N., Fukue, Y., Serino, R., Fujihara, H., Ueta, Y.,Ikawa, M., Okabe, M., Murakami, N., Shirai, M., Yoshimatsu, H.,Kangawa, K., Kojima, M., 2004. Neuromedin U has a novelanorexigenic effect independent of the leptin signalingpathway. Nat. Med. 10, 1067–1073.

Hannibal, J., 2002a. Neurotransmitters of the retino-hypothalamictract. Cell Tissue Res. 309, 73–88.

Hannibal, J., 2002b. Pituitary adenylate cyclase-activatingpeptide in the rat central nervous system: animmunohistochemical and in situ hybridization study.J. Comp. Neurol. 453, 389–417.

Hannibal, J., Fahrenkrug, J., 2002. Immunoreactive substance P isnot part of the retinohypothalamic tract in the rat. Cell TissueRes. 309, 293–299.

Hannibal, J., Fahrenkrug, J., 2004. Target areas innervated byPACAP-immunoreactive retinal ganglion cells. Cell Tissue Res.316, 99–113.

Hannibal, J., Mikkelsen, J.D., Clausen, H., Holst, J.J., Wulff, B.S.,Fahrenkrug, J., 1995. Gene expression of pituitary adenylatecyclase activating polypeptide (PACAP) in the rathypothalamus. Regul. Pept. 55, 133–148.

Hannibal, J., Ding, J.M., Chen, D., Fahrenkrug, J., Larsen, P.J.,Gillette, M.U., Mikkelsen, J.D., 1997. Pituitary adenylate cyclase-activating peptide (PACAP) in the retinohypothalamic tract: apotential daytime regulator of the biological clock. J. Neurosci.17, 2637–2644.

Hannibal, J., Moller, M., Ottersen, O.P., Fahrenkrug, J., 2000. PACAP

and glutamate are co-stored in the retinohypothalamic tract. J.Comp. Neurol. 418, 147–155.

Hannibal, J., Vrang, N., Card, J.P., Fahrenkrug, J., 2001a. Light-dependent induction of cFos during subjective day and night inPACAP-containing ganglion cells of the retinohypothalamictract. J. Biol. Rhythms 16, 457–470.

Hannibal, J., Jamen, F., Nielsen, H.S., Journot, L., Brabet, P.,Fahrenkrug, J., 2001b. Dissociation between light-inducedphase shift of the circadian rhythm and clock gene expressionin mice lacking the pituitary adenylate cyclase activatingpolypeptide type 1 receptor. J. Neurosci. 21, 4883–4890.

Hannibal, J., Hindersson, P., Knudsen, S.M., Georg, B., Fahrenkrug,J., 2002. The photopigment melanopsin is exclusively presentin pituitary adenylated cyclase-activatingpolypeptide-containing retinal ganglion cells of theretinohypothalamic tract. J. Neurosci. 22, RC191 (1-7).

Harmar, A.J., Marston, H.M., Shen, S., Spratt, C., West, K.M.,Sheward, W.J., Morrison, C.F., Dorin, J.R., Piggins, H.D., Reubi,J.-C., Kelly, J.S., Maywood, E., Hastings, M.H., 2002. The VPAC2receptor is essential for circadian function in the mousesuprachiasmatic nuclei. Cell 109, 497–508.

Harrington, M.E., 1997. The ventral lateral geniculate nucleus andthe intergeniculate leaflet: interrelated structures in the visualand circadian systems. Neurosci. Biobehav. Rev. 21, 705–727.

Harrington, M.E., Rusak, B., 1986. Lesions of the thalamicintergeniculate leaflet alter hamster circadian rhythms.J. Biol. Rhythms 1, 309–325.

Harrington, M.E., Schak, K.M., 2000. Neuropeptide Y phaseadvances the in vitro hamster circadian clock during thesubjective day with no effect on phase during the subjectivenight. Can. J. Physiol. Pharm. 78, 87–92.

Harrington, M.E., Hoque, S., Hall, A., Golombek, D.A., Biello, S.,1999. Pituitary adenylate cyclase activating peptide phaseshifts circadian rhythms in a manner similar to light.J. Neurosci. 19, 6637–6642.

Hartwich, M., Kalsbeek, A., Pevet, P., Nurnberger, F., 1994. Effects ofillumination and enucleation on substance-P-immunoreactivestructures in subcortical visual centers of golden-hamster andWistar rat. Cell Tissue Res. 277, 351–361.

Hastings, M.H., Mead, S.M., Vindlacheruvu, R.R., Ebling, F.J.P.,Maywood, E.S., Grosse, J., 1992. Non-photic phase shifting ofthe circadian activity rhythm of Syrian hamsters: the relativepotency of arousal and melatonin. Brain Res. 591, 20–26.

Hastings, M.H., Field, M.D., Maywood, E.S., Weaver, D.R., Reppert,S.M., 1999. Differential regulation of mPER1 andmTIM proteinsin the mouse suprachiasmatic nuclei: new insights into a coreclock mechanism. J. Neurosci. 19, RC11.

Hattar, S., Liao, H.W., Takao, M., Berson, D.M., Yau, K.W., 2002.Melanopsin-containing retinal ganglion cells: architecture,projections, and intrinsic photosensitivity. Science 295,1065–1070.

Hattar, S., Lucas, R.J., Mrosovsky, N., Thompson, S., Douglas, R.H.,Hankins, M.W., Lem, J., Biel, M., Hofmann, F., Foster, R.G., Yau,K.W., 2003. Melanopsin and rod-cone photoreceptive systemsaccount for all major accessory visual functions in mice.Nature 424, 76–81.

Hay-Schmidt, A., Vrang, N., Larsen, P.J., Mikkelsen, J.D., 2003.Projections from the raphe nuclei to the suprachiasmaticnucleus of the rat. J. Chem. Neuroanat. 25, 293–310.

Hendrickson, A.E., Wagoner, N., Cowan, W.M., 1972. Anautoradiographic and electron microscopic study ofretinohypothalamic connections. Z. Zellforsch. 135, 1–26.

Hertz, L., 2004. Intercellular metabolic compartmentation in thebrain: past, present and future. Neurochem. Int. 45, 285–296.

Herzog, E.D., Geusz, M.E., Khalsa, S.B., Straume, M., Block, G.D.,1997. Circadian rhythms in mouse suprachiasmatic nucleusexplants on multimicroelectrode plates. Brain Res. 757,285–290.

Herzog, E.D., Takahashi, J.S., Block, G.D., 1998. Clock controls


circadian period in isolated suprachiasmatic nucleus neurons.Nat. Neurosci. 1, 708–713.

Honma, S., Shirakawa, T., Katsuno, Y., Namihira, M., Honma, K.,1998a. Circadian periods of single suprachiasmatic neurons inrats. Neurosci. Lett. 250, 157–160.

Honma, S., Ikeda, M., Abe, H., Tanahashi, Y., Namihira, M., Honma,K., Nomura,M., 1998b. Circadian oscillation of BMAL1, a partnerof a mammalian clock gene Clock, in rat suprachiasmaticnucleus. Biochem. Biophys. Res. Commun. 250, 83–87.

Honrado, G.I., Johnson, R.S., Golombek, D.A., Spiegelman, B.M.,Papaioannou, V.E., Ralph, M.R., 1996. The circadian system ofc-fos deficient mice. J. Comp. Physiol., A 178, 563–570.

Horikawa, K., Yokota, S., Fuji, K., Akiyama, M., Moriya, T.,Okamura, H., Shibata, S., 2000. Nonphotic entrainment by5-HT1A/7 receptor agonists accompanied by reduced Per1 andPer2 mRNA levels in the suprachiasmatic nuclei. J. Neurosci.20, 5867–5873.

Horowitz, S.S., Blanchard, J.H., Morin, L.P., 2004. Intergeniculateleaflet and ventral lateral geniculate nucleus afferentconnections: an anatomical substrate for functional input fromthevestibulo-visuomotor system. J. Comp.Neurol. 474, 227–245.

Horowitz, S.S., Blanchard, J.H., Morin, L.P., 2005. Medial vestibularconnections with the hypocretin (orexin) system. J. Comp.Neurol. 487, 127–146.

Horvath, T.L., 1998. An alternate pathway for visual signalintegration into the hypothalamo-pituitary axis:retinorecipient intergeniculate neurons project to variousregions of the hypothalamus and innervate neuroendocrinecells including those producing dopamine. J. Neurosci. 18,1546–1558.

Hsiao, K., Sachs, G.M., Schneider, G.E., 1984. A minute fraction ofSyrian golden hamster retinal ganglion cells project bilaterally.J. Neurosci. 4, 359–367.

Huang, R.C., 1993. Sodium and calcium currents in acutelydissociated neurons from rat suprachiasmatic nucleus.J. Neurophysiol. 70, 1692–1703.

Hut, R.A., van der Zee, E.A., Jansen, K., Gerkema, M.P., Daan, S.,2002. Gradual reappearance of post-hibernation circadianrhythmicity correlates with numbers ofvasopressin-containing neurons in the suprachiasmaticnuclei of European ground squirrels. J. Comp. Physiol., B 172,59–70.

Ichikawa, T., Hirata, Y., 1986. Organization of cholineacetyltransferase-containing structure in the forebrain of therat. J. Neurosci. 6, 281–292 (Ref Type: Abstract).

Ikeda, M., Sugiyama, T., Wallace, C.S., Gompf, H.S., Yoshioka, T.,Miyawaki, A., Allen, C.N., 2003. Circadian dynamics of cytosolicand nuclear Ca2+ in single suprachiasmatic nucleus neurons.Neuron 38, 253–263.

Ingram, C.D., Ciobanu, R., Coculescu, I.L., Tanasescu, R., Coculescu,M., Mihai, R., 1998. Vasopressin neurotransmission and thecontrol of circadian rhythms in the suprachiasmatic nucleus.Prog. Brain Res. 119, 351–364.

Inouye, S.T., Kawamura, H., 1979. Persistence of circadianrhythmicity in a mammalian hypothalamic “island”containing the suprachiasmatic nucleus. Proc. Natl. Acad. Sci.U. S. A. 76, 5962–5966.

Isoldi, M.C., Rollag, M.D., de Lauro Castrucci, A.M., Provencio, I.,2005. Rhabdomeric phototransduction initiated by thevertebrate photopigment melanopsin. Proc. Natl. Acad. Sci.U. S. A. 102, 1217–1221.

Itri, J., Colwell, C.S., 2003. Regulation of inhibitory synaptictransmission by vasoactive intestinal peptide (VIP) in themouse suprachiasmatic nucleus. J. Neurophysiol. 90,1589–1597.

Jackson,A.C., Yao,G.L., Bean, B.P., 2004.Mechanismof spontaneousfiring in dorsomedial suprachiasmatic nucleus neurons.J. Neurosci. 24, 7985–7998.

Jagota, A., De la Iglesia, H.O., Schwartz, W.J., 2000. Morning and

evening circadian oscillations in the suprachiasmatic nucleusin vitro. Nat. Neurosci. 3, 372–376.

Janik, D., Mrosovsky, N., 1994. Intergeniculate leaflet lesions andbehaviorally-induced shifts of circadian rhythms. Brain Res.651, 174–182.

Jhaveri, S., Edwards, M.A., Schneider, G.E., 1991. Initial stages ofretinofugal axon development in the hamster: evidence for twodistinct modes of growth. Exp. Brain Res. 87, 371–382.

Jiang, Z.-G., Allen, C.N., North, R.A., 1995. Presynaptic inhibition bybaclofen of retinohypothalamic excitatory synaptictransmission in rat suprachiasmatic nucleus. Neuroscience 64,813–819.

Jiang, Z.-G., Yang, Y.-Q., Liu, Z.P., Allen, C.N., 1997. Membraneproperties and synaptic inputs of suprachiasmatic nucleusneurons in rat brain slices. J. Physiol. 499, 141–159.

Jobst, E.E., Allen, C.N., 2002. Calbindin neurons in the hamstersuprachiasmatic nucleus do not exhibit a circadian variation inspontaneous firing rate. Eur. J. Neurosci. 16, 2469–2474.

Jobst, E.E., Robinson, D.W., Allen, C.N., 2004. Potential pathwaysfor intercellular communication within the calbindinsubnucleus of the hamster suprachiasmatic nucleus.Neuroscience 123, 87–99.

Johnson, R.F., Morin, L.P., Moore, R.Y., 1988a. Retinohypothalamicprojections in the hamster and rat demonstrated using choleratoxin. Brain Res. 462, 301–312.

Johnson, R.F., Smale, L., Moore, R.Y., Morin, L.P., 1988b. Lateralgeniculate lesions block circadian phase shift responses to abenzodiazepine. Proc. Natl. Acad. Sci. U. S. A. 85, 5301–5304.

Kallo, I., Kalamatianos, T., Wiltshire, N., Shen, S., Sheward, W.J.,Harmar, A.J., Coen, C.W., 2004. Transgenic approach revealsexpression of the VPAC2 receptor in phenotypically definedneurons in the mouse suprachiasmatic nucleus and in itsefferent target sites. Eur. J. Neurosci. 19, 2201–2211.

Kalsbeek, A., Teclemariam-Mesbah, R., Pevet, P., 1993. Efferentprojections of the suprachiasmatic nucleus in the goldenhamster (Mesocricetus auratus). J. Comp. Neurol. 332, 293–314.

Kalsbeek, A., Fliers, E., Franke, A.N., Wortel, J., Buijs, R.M., 2000.Functional connections between the suprachiasmatic nucleusand the thyroid gland as revealed by lesioning and viral tracingtechniques in the rat. Endocrinology 141, 3832–3841.

Karatsoreos, I.N., Yan, L., LeSauter, J., Silver, R., 2004. Phenotypematters: identification of light-responsive cells in the mousesuprachiasmatic nucleus. J. Neurosci. 24, 68–75.

Kawaguchi, C., Tanaka, K., Isojima, Y., Shintani, N., Hashimoto, H.,Baba, A., Nagai, K., 2003. Changes in light-induced phase shiftof circadian rhythm in mice lacking PACAP. Biochem. Biophys.Res. Commun. 310, 169–175.

Kawamoto, K., Nagano, M., Kanda, F., Chihara, K., Shigeyoshi, Y.,Okamura, H., 2003. Two types of VIP neuronal components inrat suprachiasmatic nucleus. J. Neurosci. Res. 74, 852–857.

Kennaway, D.J., Moyer, R.W., 1998. Serotonin 5-HT2c agonistsmimic the effect of light pulses on circadian rhythms. BrainRes. 806, 257–270.

Kennaway, D.J., Moyer, R.W., Voultsios, A., Varcoe, T.J., 2001.Serotonin, excitatory amino acids and the photic control ofmelatonin rhythms and SCN c-FOS in the rat. Brain Res. 897,36–43.

Kennaway, D.J., Voultsios, A., Varcoe, T.J., Moyer, R.W., 2003.Melatonin and activity rhythm responses to light pulses inmice with the Clock mutation. Am. J. Physiol.: Regul., Integr.Comp. Physiol. 284, R1231–R1240.

Kim, Y.I., Dudek, F.E., 1992. Intracellular electrophysiologicalstudy of suprachiasmatic nucleus neurons in rodents:inhibitory synaptic mechanisms. J. Physiol. (London) 458,247–260.

Kim, Y.I., Dudek, F.E., 1993. Membrane properties of ratsuprachiasmatic nucleus neurons receiving optic nerve input.J. Physiol. (London) 464, 229–243.

Kim, Y.I., Kim, S.H., Kim, D.Y., Lee, H.W., Shin, H.C., Chung, J.M.,


Han, H.C., Na, H.S., Hong, S.K., 1999. Electrophysiologicalevidence for the role of substance P in retinohypothalamictransmission in the rat. Neurosci. Lett. 274, 99–102.

Kim, D.Y., Kang, H.C., Shin, H.C., Lee, K.J., Yoon, Y.W., Han, H.C.,Na, H.S., Hong, S.K., Kim, Y.I., 2001. Substance P plays a criticalrole in photic resetting of the circadian pacemaker in the rathypothalamus. J. Neurosci. 21, 4026–4031.

King, V.M., Chahad-Ehlers, S., Shen, S., Harmar, A.J., Maywood, E.S.,Hastings, M.H., 2003. A hVIPR transgene as a novel tool for theanalysis of circadian function in the mouse suprachiasmaticnucleus. Eur. J. Neurosci. 17, 822–832.

Kiss, J., Halasz, B., 1996. Synaptic contacts between cholinergicafferents and suprachiasmatic neurones of the rat.NeuroReport 7, 1961–1964.

Knoch, M.E., Gobes, S.M., Pavlovska, I., Su, C., Mistlberger, R.E.,Glass, J.D., 2004. Short-term exposure to constant lightpromotes strong circadian phase-resetting responses tononphotic stimuli in Syrian hamsters. Eur. J. Neurosci. 19,2779–2790.

Koibuchi, N., Sakai, M.,Watanabe, K., Yamaoka, S., 1992. Change inFos-like immunoreactivity in the suprachiasmatic nucleus inthe adult male rat. NeuroReport 3, 501–504.

Kononenko, N.I., Dudek, F.E., 2004. Mechanism of irregular firing ofsuprachiasmatic nucleus neurons in rat hypothalamic slices.J. Neurophysiol. 91, 267–273.

Kononenko, N.I., Shao, L.R., Dudek, F.E., 2004. Riluzole-sensitiveslowly inactivating sodium current in rat suprachiasmaticnucleus neurons. J. Neurophysiol. 91, 710–718.

Kopp, M.D., Schomerus, C., Dehghani, F., Korf, H.W., Meissl, H.,1999. Pituitary adenylate cyclase-activating polypeptide andmelatonin in the suprachiasmatic nucleus: effects on thecalcium signal transduction cascade. J. Neurosci. 19, 206–219.

Kornhauser, J.M., Nelson, D.E., Mayo, K.E., Takahashi, J.S., 1990.Photic and circadian regulation of c-fos gene expression in thehamster suprachiasmatic nucleus. Neuron 5, 127–134.

Kriegsfeld, L.J., Demas, G.E., Lee Jr., S.E., Dawson, T.M., Dawson, V.L., Nelson, R.J., 1999. Circadian locomotor analysis ofmalemicelacking the gene for neuronal nitric oxide synthase (nNOS−/−).J. Biol. Rhythms 14, 20–27.

Kriegsfeld, L.J., Drazen, D.L., Nelson, R.J., 2001. Circadian organi-zation in male mice lacking the gene for endothelial nitricoxide synthase (eNOS−/−). J. Biol. Rhythms 16, 142–148.

Kriegsfeld, L.J., LeSauter, J., Silver, R., 2004. Targeted microlesionsreveal novel organization of the hamster suprachiasmaticnucleus. J. Neurosci. 24, 2449–2457.

Kromer, L.F., Moore, R.Y., 1980. A study of the organization of thelocus coeruleus projections to the lateral geniculate nucleus inthe albino rat. Neuroscience 5, 255–271.

Krout, K.E., Kawano, J., Mettenleiter, T.C., Loewy, A.D., 2002. CNSinputs to the suprachiasmatic nucleus of the rat. Neuroscience110, 73–92.

Kuhlman, S.J., McMahon, D.G., 2004. Rhythmic regulation ofmembrane potential and potassium current persists in SCNneurons in the absence of environmental input. Eur. J.Neurosci. 20, 1113–1117.

Kuhlman, S.J., Quintero, J.E., McMahon, D.G., 2000. GFPfluorescence reports Period 1 circadian gene regulation in themammalian biological clock. NeuroReport 11, 1479–1482.

Kuhlman, S.J., Silver, R., LeSauter, J., Bult-Ito, A., McMahon, D.G.,2003. Phase resetting light pulses induce Per1 and persistentspike activity in a subpopulation of biological clock neurons.J. Neurosci. 23, 1441–1450.

Ladenheim, E.E., Jensen, R.T., Mantey, S.A., Moran, T.H., 1992.Distinct distributions of two bombesin receptor subtypes in therat central nervous system. Brain Res. 593, 168–178.

Ladenheim, E.E., Jensen, R.T., Mantey, S.A., Taylor, J.E., Coy, D.H.,Moran, T.H., 1993. Bombesin receptor antagonists differentiatereceptor subtypes in rat brain. Eur. J. Pharmacol. 235, 121–125.

Laemle, L.K., Ottenweller, J.E., 2001. The relationship between

circadian rhythmicity and vasoactive intestinal polypeptide inthe suprachiasmatic nucleus of congenitally anophthalmicmice. Brain Res. 917, 105–111.

Lall, G.S., Biello, S.M., 2002. Attenuation of phase shifts to light byactivity or neuropeptide Y: a time course study. Brain Res. 957,109–116.

Lall, G.S., Biello, S.M., 2003a. Attenuation of circadian light inducedphase advances and delays by neuropeptide Y and aneuropeptide Y Y1/Y5 receptor agonist. Neuroscience 119,611–618.

Lall, G.S., Biello, S.M., 2003b. Neuropeptide Y, GABA and circadianphase shifts to photic stimuli. Neuroscience 120, 915–921.

Land, P.W., Kyonka, E., Shamalla-Hannah, L., 2004. Vesicularglutamate transporters in the lateral geniculate nucleus:expression of VGLUT2 by retinal terminals. Brain Res. 996,251–254.

Larsen, P.J., 1999. Tracing autonomic innervation of the rat pinealgland using viral transneuronal tracing. Microsc. Res. Tech. 46,296–304.

Larsen, P.J., Enquist, L.W., Card, J.P., 1998. Characterization ofthe multisynaptic neuronal control of the rat pineal glandusing viral transneuronal tracing. Eur. J. Neurosci. 10,128–145.

Leak, R.K., Moore, R.Y., 2001. Topographic organization ofsuprachiasmatic nucleus projection neurons. J. Comp. Neurol.433, 312–334.

Leak, R.K., Card, J.P., Moore, R.Y., 1999. Suprachiasmaticpacemaker organization analyzed by viral transynaptictransport. Brain Res. 819, 23–32.

Leanza, G., Nilsson, O.G., Wiley, R.G., Bjorklund, A., 1995.Selective lesioning of the basal forebrain cholinergic systemby intraventricular 192 IgG-saporin: behavioural, biochemicaland stereological studies in the rat. Eur. J. Neurosci. 7,329–343.

Leathers, V.L., Linse, S., Forsen, S., Norman, A.W., 1990.Calbindin-D28K, a 1 alpha,25-dihydroxyvitamin D3-inducedcalcium-binding protein, binds five or six Ca2+ ions with highaffinity. J. Biol. Chem. 265, 9838–9841.

Lee, H.S., Nelms, J.L., Nguyen, M., Silver, R., Lehman, M.N., 2003.The eye is necessary for a circadian rhythm in thesuprachiasmatic nucleus. Nat. Neurosci. 6, 111–112.

Lehman, M.N., Silver, R., Gladstone, W.R., Kahn, R.M., Gibson, M.,Bittman, E.L., 1987. Circadian rhythmicity restored by neuraltransplant. Immunocytochemical characterization of the graftand its integration with the host brain. J. Neurosci. 7,1626–1638.

Lehman, M.N., LeSauter, J., Silver, R., 1998. Fiber outgrowth fromanterior hypothalamic and cortical xenografts in the thirdventricle. J. Comp. Neurol. 391, 133–145.

LeSauter, J., Silver, R., 1998. Output signals of the SCN. Chronobiol.Int. 15, 535–550.

LeSauter, J., Silver, R., 1999. Localization of a suprachiasmaticnucleus subregion regulating locomotor rhythmicity.J. Neurosci. 19, 5574–5585.

LeSauter, J., Lehman, M.N., Silver, R., 1996. Restoration of circadianrhythmicity by transplants of SCN “micropunches”. J. Biol.Rhythms 11, 163–171.

LeSauter, J., Romero, P., Cascio, M., Silver, R., 1997. Attachment siteof grafted SCN influences precision of restored circadianrhythm. J. Biol. Rhythms 12, 327–338.

LeSauter, J., Stevens, P., Jansen, H., Lehman, M.N., Silver, R., 1999.Calbindin expression in the hamster SCN is influenced bycircadian genotype and by photic conditions. NeuroReport 10,3159–3163.

LeSauter, J., Kriegsfeld, L.J., Hon, J., Silver, R., 2002. Calbindin-D28Kcells selectively contact intra-SCN neurons. Neuroscience 111,575–585.

LeSauter, J., Yan, L., Vishnubhotla, B., Quintero, J.E., Kuhlman, S.J.,McMahon, D.G., Silver, R., 2003. A short half-life GFP mouse


model for analysis of suprachiasmatic nucleus organization.Brain Res. 964, 279–287.

Lewandowski, M.H., Usarek, A., 2002. Effects of intergeniculateleaflet lesions on circadian rhythms in themouse. Behav. BrainRes. 128, 13–17.

Lewandowski, M.H., Blasiak, T., Domoslawski, J., Wolkowska, A.,2000. Ultradian rhythmic neuronal oscillation in theintergeniculate leaflet. NeuroReport 11, 317–321.

Lewandowski, M.H., Blasiak, T., Blasiak, A., 2002. Are ultra-slowisoperiodic oscillations in rat intergeniculate leaflet neuronsdependent on reciprocal connection with its contralaterallylocated counterpart? Neurosci. Lett. 330, 243–246.

Li, X., Gilbert, J., Davis, F.C., 2005. Disruption of masking byhypothalamic lesions in Syrian hamsters. J. Comp. Physiol., ANeuroethol. Sens. Neural Behav. Physiol. 191, 23–30.

Ling, C., Schneider, G.E., Jhaveri, S., 1998. Target-specificmorphology of retinal axon arbors in the adult hamster. Vis.Neurosci. 15, 559–579.

Liou, S.Y., Albers, H.E., 1991. Single unit response of neuronswithin the hamster suprachiasmatic nucleus to neuropeptideY. Brain Res. Bull. 27, 825–828.

Liou, S.-Y., Shibata, S., Iwasaki, K., Ueki, S., 1986. Optic-nervestimulation-induced increase of release of H-3 glutamate andH-3 aspartate but not H-3 GABA from the suprachiasmaticnucleus in slices of rat hypothalamus. Brain Res. Bull. 16,527–531.

Liou, S.-Y., Shibata, S., Albers, H.E., Ueki, S., 1990. Effects of GABAand anxiolytics on the single unit discharge ofsuprachiasmatic neurons in rat hypothalamic slices. BrainRes. Bull. 25, 103–107.

Liu, C., Gillette, M.U., 1996. Cholinergic regulation of thesuprachiasmatic nucleus circadian rhythm via a muscarinicmechanism at night. J. Neurosci. 16, 744–751.

Liu, C., Reppert, S.M., 2000. GABA synchronizes clock cells withinthe suprachiasmatic circadian clock. Neuron 25, 123–128.

Liu, C., Weaver, D.R., Strogatz, S.H., Reppert, S.M., 1997a. Cellularconstruction of a circadian clock: period determination in thesuprachiasmatic nuclei. Cell 91, 855–860.

Liu, C., Ding, J.M., Faiman, L.E., Gillette, M.U., 1997b. Coupling ofmuscarinic cholinergic receptors and cGMP in nocturnalregulation of the suprachiasmatic circadian clock. J. Neurosci.17, 659–666.

Liu, Q.S., Xu, Q., Arcuino, G., Kang, J., Nedergaard, M., 2004.Astrocyte-mediated activation of neuronal kainate receptors.Proc. Natl. Acad. Sci. U. S. A. 101, 3172–3177.

Lovenberg, T.W., Baron, B.M., De Lecea, L., Miller, J.D., Prosser, R.A.,Rea, M.A., Foye, P.E., Racke, M., Slone, A.L., Siegel, B.W.,Danielson, P.E., Sutcliffe, J.G., Erlander, M.G., 1993. A noveladenylyl cyclase-activating serotonin receptor (5-HT7)implicated in the regulation of mammalian circadian rhythms.Neuron 11, 449–458.

Lucas, R.J., Foster, R.G., 1999. Neither functional rodphotoreceptors nor rod or cone outer segments are requiredfor the photic inhibition of pineal melatonin. Endocrinology140, 1520–1524.

Lucas, R.J., Freedman, M.S., Munoz, M., Garcia-Fernández, J.-M.,Foster, R.G., 1999. Regulation of the mammalian pineal bynon-rod, non-cone, ocular photoreceptors. Science 284,505–507.

Lucas, R.J., Douglas, R.H., Foster, R.G., 2001. Characterization of anocular photopigment capable of driving pupillary constrictionin mice. Nat. Neurosci. 4, 621–626.

Lucas, R.J., Hattar, S., Takao, M., Berson, D.M., Foster, R.G., Yau, K.W., 2003. Diminished pupillary light reflex at high irradiancesin melanopsin-knockout mice. Science 299, 245–247.

Madeira, M.D., Pereira, P.A., Silva, S.M., Cadete-Leite, A.,Paula-Barbosa, M.M., 2004. Basal forebrain neurons modulatethe synthesis and expression of neuropeptides in the ratsuprachiasmatic nucleus. Neuroscience 125, 889–901.

Mahoney, M.M., Nunez, A.A., Smale, L., 2000. Calbindin and Foswithin the suprachiasmatic nucleus and the adjacenthypothalamus of Arvicanthis niloticus and Rattus norvegicus.Neuroscience 99, 565–575.

Mai, L.M., Pan, J.T., 1993. Central administration of bombesinblocks the estrogen-induced afternoon prolactin surge.Neuroendocrinology 57, 40–44.

Mai, L.M., Pan, J.T., 1995. Bombesin acts in the suprachiasmaticnucleus to affect circadian changes in tuberoinfundibulardopaminergic neuron activity and prolactin secretion.Endocrinology 136, 4163–4167.

Major, D.E., Rodman, H.R., Libedinsky, C., Karten, H.J., 2003. Patternof retinal projections in the California ground squirrel(Spermophilus beecheyi): anterograde tracing study using choleratoxin. J. Comp. Neurol. 463, 317–340.

Manrique, C., Héry, F., Faudon, M., François-Bellan, A.M., 1999.Indirect evidence for an association of 5-HT1B binding siteswith retinal and geniculate axon terminals in the ratsuprachiasmatic nucleus. Synapse 33, 314–323.

Marchant, E.G., Morin, L.P., 1999. The hamster circadian rhythmsystem includes nuclei of the subcortical visual shell.J. Neurosci. 19, 10482–10493.

Marchant, E.G., Watson, N.V., Mistlberger, R.E., 1997. Bothneuropeptide Y and serotonin are necessary for entrainment ofcircadian rhythms in mice by daily treadmill runningschedules. J. Neurosci. 17, 7974–7987.

Margraf, R.R., Lynch, G.R., 1993. Melatonin injections affectcircadian behavior and SCN neurophysiology in Djungarianhamsters. Am. J. Physiol. 264, R615–R621.

Martinet, L., Bonnefond, C., Peytevin, J., Monnerie, R., Marcilloux, J.C.,1995. Vasoactive intestinal polypeptide in the suprachiasmaticnucleus of themink (Mustela vison) could play a key role in photicinduction. J. Neuroendocrinol. 7, 69–79.

Maywood, E.S., Mrosovsky, N., 2001. A molecular explanation ofinteractions between photic and non-photic circadianclock-resetting stimuli. Gene Expr. Patterns 1, 27–31.

Maywood, E.S., Smith, E., Hall, S.J., Hastings, M.H., 1997. A thalamiccontribution to arousal-induced, non-photic entrainment ofthe circadian clock of the Syrian hamster. Eur. J. Neurosci. 9,1739–1747.

Maywood, E.S., Mrosovsky, N., Field, M.D., Hastings, M.H., 1999.Rapid down-regulation of mammalian Period genes duringbehavioral resetting of the circadian clock. Proc. Natl. Acad.Sci. U. S. A. 96, 15211–15216.

Maywood, E.S., Okamura, H., Hastings, M.H., 2002. Opposingactions of neuropeptide Y and light on the expression ofcircadian clock genes in the mouse suprachiasmatic nuclei.Eur. J. Neurosci. 15, 216–220.

McArthur, A.J., Coogan, A.N., Ajpru, S., Sugden, D., Biello, S.M.,Piggins, H.D., 2000. Gastrin-releasing peptide phase-shiftssuprachiasmatic nuclei neuronal rhythms in vitro. J. Neurosci.20, 5496–5502.

Mead, S., Ebling, F.J.P., Maywood, E.S., Humby, T., Herbert, J.,Hastings, M.H., 1992. A nonphotic stimulus causesinstantaneous phase advances of the light-entrainablecircadian oscillator of the Syrian hamster but does not inducethe expression of c-fos in the suprachiasmatic nuclei.J. Neurosci. 12, 2516–2522.

Meijer, J.H., van der Zee, E.A., Dietz, M., 1988. Glutamate phaseshifts circadian activity rhythms in hamsters. Neurosci. Lett.86, 177–183.

Meijer, J.H., Schaap, J., Watanabe, K., Albus, H., 1997. Multiunitactivity recordings in the suprachiasmatic nuclei: in vivo vs. invitro models. Brain Res. 753, 322–327.

Meijer, J.H., Ruijs, A.C.J., Albus, H., Van de Geest, B., Duindam, H.,Zwinderman, A.H., Dahan, A., 2000. Fentanyl, a μ-opioidreceptor agonist, phase shifts the hamster circadianpacemaker. Brain Res. 868, 135–140.

Melyan, Z., Tarttelin, E.E., Bellingham, J., Lucas, R.J., Hankins, M.W.,


2005. Addition of humanmelanopsin rendersmammalian cellsphotoresponsive. Nature 433, 741–745.

Menaker, M., 1968. Extraretinal light perception in the sparrow: I.Entrainment of the biological clock. Proc. Natl. Acad. Sci.U. S. A. 59, 414–421.

Menaker, M., Keatts, H., 1968. Extraretinal light perception in thesparrow: II. Photoperiodic stimulation of testis growth. Proc.Natl. Acad. Sci. U. S. A. 60, 146–151.

Menaker, M., Underwood, H., 1976. Extraretinal photoreception inbirds. Photophysiology 23, 299–306.

Menet, J.S., Vuillez, P., Pevet, P., 2003. Calbindin expression in thehamster suprachiasmatic nucleus depends on day-length.Neuroscience 122, 591–598.

Meyer-Bernstein, E.L., Morin, L.P., 1996. Differential serotonergicinnervation of the suprachiasmatic nucleus and theintergeniculate leaflet and its role in circadian rhythmmodulation. J. Neurosci. 16, 2097–2111.

Meyer-Bernstein, E.L., Morin, L.P., 1998. Destruction ofserotonergic neurons in the median raphe nucleus blockscircadian rhythm phase shifts to triazolam, but not to novelwheel access. J. Biol. Rhythms 13, 494–505.

Meyer-Bernstein, E.L., Morin, L.P., 1999. Electrical stimulation ofthe median or dorsal raphe nuclei reduces light-induced FOSprotein in the suprachiasmatic nucleus and causes circadianactivity rhythm phase shifts. Neuroscience 92, 267–279.

Meyer-Bernstein, E.L., Blanchard, J.H., Morin, L.P., 1997. Theserotonergic projection from the median raphe nucleus to thesuprachiasmatic nucleus modulates activity phase onset, butnot other circadian rhythm parameters. Brain Res. 755,112–120.

Meyer-Bernstein, E.L., Jetton, A.E., Matsumoto, S.I., Markuns, J.F.,Lehman, M.N., Bittman, E.L., 1999. Effects of suprachiasmatictransplants on circadian rhythms of neuroendocrine functionin golden hamsters. Endocrinology 140, 207–218.

Meyer-Spasche, A., Reed, H.E., Piggins, H.D., 2002. Neurotensinphase-shifts the firing rate rhythm of neurons in the ratsuprachiasmatic nuclei in vitro. Eur. J. Neurosci. 16, 339–344.

Mick, G., Maeno, H., Kiyama, H., Tohyama, M., 1994. Marginaltopography of neurons expressing the substance P receptorin the rat suprachiasmatic nucleus. Mol. Brain Res. 21,157–161.

Mick, G., Shigemoto, R., Kitahama, K., 1995. Localization ofsubstance P receptors in central neural structures controllingdaily rhythms in nocturnal rodents. C. R. Acad. Sci. III 318,209–217.

Mikkelsen, J.D., 1990. Projections from the lateral geniculatenucleus to the hypothalamus of the Mongolian gerbil (Merionesunguiculatus): an anterograde and retrograde tracing study.J. Comp. Neurol. 299, 493–508.

Mikkelsen, J.D., Larsen, P.J., 1993. Substance P in thesuprachiasmatic nucleus of the rat: an immunohistochemicaland in situ hybridization study. Histochemistry 100, 3–16.

Mikkelsen, J.D., Vrang, N., 1994. A direct pretectosuprachiasmaticprojection in the rat. Neuroscience 62, 497–505.

Miller, J.D., Fuller, C.A., 1992. Isoperiodic neuronal activity insuprachiasmatic nucleus of the rat. Am. J. Physiol. 263,R51–R58.

Miller, J.D., Morin, L.P., Schwartz, W.J., Moore, R.Y., 1996. Newinsights into themammalian circadian clock [Review] [230 refs]Sleep 19, 641–667.

Mintz, E.M., Albers, H.E., 1997. Microinjection of NMDA into theSCN region mimics the phase shifting effect of light inhamsters. Brain Res. 758, 245–249.

Mintz, E.M., Marvel, C.L., Gillespie, C.F., Price, K.M., Albers, H.E.,1999. Activation of NMDA receptors in the suprachiasmaticnucleus produces light-like phase shifts of the circadian clockin vivo. J. Neurosci. 19, 5124–5130.

Mintz, E.M., van den Pol, A.N., Casano, A.A., Albers, H.E., 2001.Distribution of hypocretin-(orexin) immunoreactivity in the

central nervous system of Syrian hamsters (Mesocricetusauratus). J. Chem. Neuroanat. 21, 225–238.

Mintz, E.M., Jasnow, A.M., Gillespie, C.F., Huhman, K.L., Albers, H.E., 2002. GABA interacts with photic signaling in thesuprachiasmatic nucleus to regulate circadian phase shifts.Neuroscience 109, 773–778.

Mistlberger, R.E., Antle, M.C., 1998. Behavioral inhibition oflight-induced circadian phase resettting is phase andserotonin dependent. Brain Res. 786, 31–38.

Mistlberger, R.E., Landry, G.J., Marchant, E.G., 1997. Sleepdeprivation can attenuate light-induced phase shifts ofcircadian rhythms in hamsters. Neurosci. Lett. 238, 5–8.

Mistlberger, R.E., Antle, M.C., Glass, J.D., Miller, J.D., 2000.Behavioral and serotonergic regulation of circadian rhythms.Biol. Rhythm Res. 31, 240–283.

Mistlberger, R.E., Belcourt, J., Antle, M.C., 2002. Circadian clockresetting by sleep deprivation without exercise in Syrianhamsters: dark pulses revisited. J. Biol. Rhythms 17,227–237.

Mistlberger, R.E., Antle, M.C., Webb, I.C., Jones, M., Weinberg, J.,Pollock, M.S., 2003. Circadian clock resetting by arousal inSyrian hamsters: the role of stress and activity. Am. J. Physiol.:Regul., Integr. Comp. Physiol. 285, R917–R925.

Moffett, J.R., 2003. Reductions in N-acetylaspartylglutamate andthe 67 kDa form of gluamic acid decarboxylaseimmunoreactivities in the visual system of albino andpigmented rats after optic nerve transections. J. Comp. Neurol.458, 221–239.

Moffett, J.R., Williamson, L., Palkovits, M., Namboodiri, M.A.A.,1990. N-acetylaspartylglutamate: a transmitter candidate forthe retinohypothalamic tract. Proc. Natl. Acad. Sci. U. S. A. 87,8065–8069.

Moffett, J.R., Williamson, L.C., Neale, J.H., Palkovits, M.,Namboodin, M.A.A., 1991. Effect of optic nerve transection onN-acetylaspartylglutamate immunoreactivity in the primaryand accessory optic projection systems in the rat. Brain Res.538, 86–94.

Moga, M.M., 1998. 192 IgG-saporin abolishes p75 neurotrophinreceptor immunoreactivity in rat SCN. NeuroReport 9,3197–3200.

Moga, M.M., Moore, R.Y., 1996. Putative excitatory amino acidprojections to the suprachiasmatic nucleus in the rat. BrainRes. 743, 171–177.

Moga, M.M., Moore, R.Y., 1997. Organization of neural inputs to thesuprachiasmatic nucleus in the rat. J. Comp. Neurol. 389,508–534.

Moore, R.Y., 1983. Organization and function of a central nervoussystem oscillator: the suprachiasmatic hypothalamic nucleus.Fed. Proc. 42, 2783–2789.

Moore, R.Y., 1989. The geniculohypothalamic tract in monkey andman. Brain Res. 486, 190–194.

Moore, R.Y., 1993. Organization of the primate circadian system.J. Biol. Rhythms 8, S3–S9.

Moore, R.Y., 1996. Entrainment pathways and the functionalorganization of the circadian timing system. In: Buijs, R.,Kalsbeek, A., Romijn, H.J., Pennartz, C.M.A., Mirmiran, M. (Eds.),Hypothalamic Integration of Circadian Rhythms. Elsevier,Amsterdam, pp. 101–117.

Moore, R.Y., Card, J.P., 1994. Intergeniculate leaflet: ananatomically and functionally distinct subdivision of thelateral geniculate complex. J. Comp. Neurol. 344, 403–430.

Moore, R.Y., Lenn, N.J., 1972. A retinohypothalamic projection inthe rat. J. Comp. Neurol. 146, 1–14.

Moore, R.Y., Speh, J.C., 1993. GABA is the principalneurotransmitter of the circadian system. Neurosci. Lett. 150,112–116.

Moore, R.Y., Gustafson, E.L., Card, J.P., 1984. Identicalimmunoreactivity of afferents to the rat suprachiasmaticnucleus with antisera against avian pancreatic polypeptide,


molluscan cardiaoexcitatory peptide and neuropeptide Y.Cell Tissue Res. 236, 41–46.

Moore, R.Y., Speh, J.C., Card, J.P., 1995. The retinohypothalamictract originates from a distinct subset of retinal ganglion cells.J. Comp. Neurol. 352, 351–366.

Moore, R.Y., Weis, R., Moga, M.M., 2000. Efferent projections of theintergeniculate leaflet and the ventral lateral geniculatenucleus in the rat. J. Comp. Neurol. 420, 398–418.

Moore, R.Y., Speh, J.C., Leak, R.K., 2002. Suprachiasmatic nucleusorganization. Cell Tissue Res. 309, 89–98.

Mori, K., Miyazato, M., Ida, T., Murakami, N., Serino, R., Ueta, Y.,Kojima, M., Kangawa, K., 2005. Identification of neuromedin Sand its possible role in the mammalian circadian oscillatorsystem. EMBO J. 24, 325–335.

Morin, L.P., 1994. The circadian visual system. Brain Res. Rev. 67,102–127.

Morin, L.P., 1999. Serotonin and the regulation of mammaliancircadian rhythmicity. Ann. Med. 31, 12–33.

Morin, L.P., Blanchard, J.H., 1991. Depletion of brain serotonin by5,7-DHT modifies hamster circadian rhythm response to light.Brain Res. 566, 173–185.

Morin, L.P., Blanchard, J.H., 1995. Organization of the hamsterintergeniculate leaflet: NPY and ENK projections to thesuprachiasmatic nucleus, intergeniculate leaflet and posteriorlimitans nucleus. Vis. Neurosci. 12, 57–67.

Morin, L.P., Blanchard, J.H., 1998. Interconnections among nucleiof the subcortical visual shell: the intergeniculate leaflet is amajor constituent of the hamster subcortical visual system.J. Comp. Neurol. 396, 288–309.

Morin, L.P., Blanchard, J.H., 1999. Forebrain connections of thehamster intergeniculate leaflet: comparison with those ofventral lateral geniculate nucleus and retina. Vis. Neurosci. 16,1037–1054.

Morin, L.P., Blanchard, J.H., 2001. Neuromodulator content ofhamster intergeniculate leaflet neurons and their projection tothe suprachiasmatic nucleus or visual midbrain. J. Comp.Neurol. 437, 79–90.

Morin, L.P., Blanchard, J.H., 2005. Descending projections of thehamster intergeniculate leaflet: relationship to the sleep/arousal and visuomotor systems. J. Comp. Neurol. 487, 204–216.

Morin, L.P., Pace, L., 2002. The intergeniculate leaflet, but not thevisual midbrain, mediates hamster circadian rhythm responseto constant light. J. Biol. Rhythms 17, 217–226.

Morin, L.P., Blanchard, J.H., Moore, R.Y., 1992. Intergeniculateleaflet and suprachiasmatic nucleus organization andconnections in the hamster. Vis. Neurosci. 8, 219–230.

Morin, L.P., Goodless-Sanchez, N., Smale, L., Moore, R.Y., 1994.Projections of the suprachiasmatic nuclei, subparaventricularzone and retrochiasmatic area in the golden hamster.Neuroscience 61, 391–410.

Morin, L.P., Blanchard, J.H., Provencio, I., 2003. Retinal ganglioncell projections to the hamster suprachiasmatic nucleus,intergeniculate leaflet and visual midbrain: bifurcation andmelanopsin immunoreactivity. J. Comp. Neurol. 465,401–416.

Moriya, T., Yamanouchi, S., Fukushima, T., Shimazoe, T., Shibata,S., Watanabe, S., 1996. Involvement of 5-HT1A receptormechanisms in the inhibitory effects of methamphetamine onphotic responses in the rodent suprachiasmatic nucleus. BrainRes. 740, 261–267.

Moriya, T., Yoshinobu, Y., Ikeda, M., Yokota, S., Akiyama, M.,Shibata, S., 1998. Potentiating action of MKC-242, a selective5-HT1A receptor agonist, on the photic entrainment of thecircadian activity rhythm in hamsters. Br. J. Pharmacol. 125,1281–1287.

Moyer, R.W., Kennaway, D.J., 1999. Immunohistochemicallocalization of serotonin receptors in the rat suprachiasmaticnucleus. Neurosci. Lett. 271, 147–150.

Moyer, R.W., Kennaway, D.J., 2000. Serotonin depletion decreases

light induced c-fos in the rat suprachiasmatic nucleus.NeuroReport 11, 1021–1024.

Mrosovsky, N., 1996. Locomotor activity and non-photicinfluences on circadian clocks. Biol. Rev. Camb. Philos. Soc. 71,343–372.

Mrosovsky, N., 1999. Masking: history, definitions, andmeasurement. Chronobiol. Int. 16, 415–429.

Mrosovsky, N., 2001. Further characterization of the phenotype ofmCry1/mCry2-deficient mice. Chronobiol. Int. 18, 613–625.

Mrosovsky, N., 2003. Contribution of classic photoreceptors toentrainment. J. Comp. Physiol., A 189, 69–73.

Mrosovsky, N., Hattar, S., 2003. Impaired masking responses tolight in melanopsin-knockout mice. Chronobiol. Int. 20,989–999.

Mrosovsky, N., Foster, R.G., Salmon, P.A., 1999. Thresholds formasking responses to light in three strains of retinallydegenerate mice. J. Comp. Physiol., A 184, 423–428.

Mrosovsky, N., Salmon, P.A., Foster, R.G., McCall, M.A., 2000.Responses to light after retinal degeneration. Vision Res. 40,575–578.

Mrugala, M., Zlomanczuk, P., Jagota, A., Schwartz, W.J., 2000.Rhythmic multiunit neural activity in slices of hamstersuprachiasmatic nucleus reflect prior photoperiod. Am. J.Physiol.: Regul., Integr. Comp. Physiol. 278, R987–R994.

Munch, I.C., Moller, M., Larsen, P.J., Vrang, N., 2002. Light-inducedc-fos expression in suprachiasmatic nuclei neurons targetingthe paraventricular nucleus of the hamster hypothalamus:phase dependence and immunochemical identification.J. Comp. Neurol. 442, 48–62.

Muscat, L., Morin, L.P., in press. Bionocular contributions to theresponsiveness and integrative capacity of the circadianrhythm system to light. J. Biol. Rhythms.

Muscat, L., Huberman, A.D., Jordan, C.L., Morin, L.P., 2003. Crossedand uncrossed retinal projections to the hamster circadiansystem. J. Comp. Neurol. 466, 513–524.

Muscat, L., Tischler, R.C., Morin, L.P., 2005. Functional analysis ofthe role of the median raphe as a regulator of hamstercircadian system sensitivity to light. Brain Res. 1044 (1), 59–66.

Nagano, M., Adachi, A., Nakahama, K., Nakamura, T., Tamada, M.,Meyer-Bernstein, E., Sehgal, A., Shigeyoshi, Y., 2003. An abruptshift in the day/night cycle causes desynchrony in themammalian circadian center. J. Neurosci. 23, 6141–6151.

Nakahara, K., Hanada, R., Murakami, N., Teranishi, H., Ohgusu, H.,Fukushima, N., Moriyama, M., Ida, T., Kangawa, K., Kojima, M.,2004. The gut-brain peptide neuromedin U is involved in themammalian circadian oscillator system. Biochem. Biophys.Res. Commun. 318, 156–161.

Nakamura, H., Itoh, K., 2004. Cytoarchitectonic and connectionalorganization of the ventral lateral geniculate nucleus in the cat.J. Comp. Neurol. 473, 439–462.

Nakamura, W., Honma, S., Shirakawa, T., Honma, K., 2002. Clockmutation lengthens the circadian period without dampingrhythms in individual SCN neurons. Nat. Neurosci. 5, 399–400.

Neale, J.H., Bzdega, T., Wroblewska, B., 2000.N-Acetylaspartylglutamate: the most abundant peptideneurotransmitter in the mammalian central nervous system.J. Neurochem. 75, 443–452.

Negroni, J., Nevo, E., Cooper, H.M., 1997. Neuropeptidergicorganization of the suprachiasmatic nucleus in the blind molerat (Spalax ehrenbergi). Brain Res. Bull. 44, 633–639.

Nelson, D.E., Takahashi, J.S., 1991. Sensitivity and integration in avisual pathway for circadian entrainment in the hamster(Mesocricetus auratus). J. Physiol. (London) 439, 115–145.

Nelson, D.E., Takahashi, J.S., 1999. Integration and saturationwithin the circadian photic entrainment pathway of hamsters.Am. J. Physiol.: Regul., Integr. Comp. Physiol. 277, R1351–R1361.

Newman, L.A., Walker, M.T., Brown, R.L., Cronin, T.W., Robinson,P.R., 2003. Melanopsin forms a functional short-wavelengthphotopigment. Biochemistry 42, 12734–12738.


Novak, C.M., Albers, H.E., 2004a. Novel phase-shifting effects ofGABAA receptor activation in the suprachiasmatic nucleus of adiurnal rodent. Am. J. Physiol.: Regul., Integr. Comp. Physiol.286, R820–R825.

Novak, C.M., Albers, H.E., 2004b. Circadian phase alteration byGABA and light differs in diurnal and nocturnal rodents duringthe day. Behav. Neurosci. 118, 498–504.

Novak, C.M., Ehlen, J.C., Huhman, K.L., Albers, H.E., 2004. GABA(B)receptor activation in the suprachiasmatic nucleus of diurnaland nocturnal rodents. Brain Res. Bull. 63, 531–535.

Nunez, A.A., Bult, A., McElhinny, T.L., Smale, L., 1999. Dailyrhythms of Fos expression in hypothalamic targets of thesuprachiasmatic nucleus in diurnal and nocturnal rodents.J. Biol. Rhythms 14, 300–306.

O'Hara, B.F., Edgar, D.M., Cao, V.H.,Wiler, S.W., Heller, H.C., Kilduff,T.S., Miller, J.D., 1998. Nicotine and nicotinic receptors in thecircadian system. Psychoneuroendocrinology 23, 161–173.

Oliver, K.R., Kinsey, A.M., Wainwright, A., Sirinathsinghji, D.J.,2000. Localization of 5-ht(5A) receptor-like immunoreactivityin the rat brain. Brain Res. 867, 131–142.

Otori, Y., Tominaga, K., Fukuhara, C., Yang, J., Yamazaki, S.,Cagampang, F.R.A., Okamura, H., Inouye, S.-I.T., 1993.Substance P-like immunoreactivity in the suprachiasmaticnucleus of the rat. Brain Res. 619, 271–277.

Panda, S., Sato, T.K., Castrucci, A.M., Rollag, M.D., DeGrip, W.J.,Hogenesch, J.B., Provencio, I., Kay, S.A., 2002. Melanopsin(Opn4) requirement for normal light-induced circadian phaseshifting. Science 298, 2213–2216.

Panda, S., Provencio, I., Tu, D.C., Pires, S.S., Rollag, M.D., Castrucci,A.M., Pletcher, M.T., Sato, T.K., Wiltshire, T., Andahazy, M., Kay,S.A., Van Gelder, R.N., Hogenesch, J.B., 2003. Melanopsin isrequired for non-image-forming photic responses in blindmice. Science 301, 525–527.

Panda, S., Nayak, S.K., Campo, B., Walker, J.R., Hogenesch, J.B.,Jegla, T., 2005. Illumination of the melanopsin signalingpathway. Science 307, 600–604.

Panula, P., Yang, H.Y., Costa, E., 1984. Comparative distribution ofbombesin/GRP- and substance-P-like immunoreactivities inrat hypothalamus. J. Comp. Neurol. 224, 606–617.

Paul, K.N., Fukuhara, C., Tosini, G., Albers, H.E., 2003.Transduction of light in the suprachiasmatic nucleus:evidence for two different neurochemical cascadesregulating the levels of Per1 mRNA and pineal melatonin.Neuroscience 119, 137–144.

Paul, K.N., Gamble, K.L., Fukuhara, C., Novak, C.M., Tosini, G.,Albers, H.E., 2004. Tetrodotoxin administration in thesuprachiasmatic nucleus prevents NMDA-induced reductionsin pineal melatonin without influencing Per1 and Per2 mRNAlevels. Eur. J. Neurosci. 19, 2808–2814.

Penev, P.D., Zee, P.C., Wallen, E.P., Turek, F.W., 1995. Aging altersthe phase-resetting properties of a serotonin agonist onhamster circadian rhythmicity. Am. J. Physiol. 268, R293–R298.

Pennartz, C.M., Bierlaagh, M.A., Geurtsen, A.M., 1997. Cellularmechanisms underlying spontaneous firing in ratsuprachiasmatic nucleus: involvement of a slowly inactivatingcomponent of sodium current. J. Neurophysiol. 78, 1811–1825.

Pennartz, C.M., Bos, N.P., Jeu, M.T., Geurtsen, A.M., Mirmiran, M.,Sluiter, A.A., Buijs, R.M., 1998. Membrane properties andmorphology of vasopressin neurons in slices of ratsuprachiasmatic nucleus. J. Neurophysiol. 80, 2710–2717.

Pennartz, C.M., de Jeu, M.T., Bos, N.P., Schaap, J., Geurtsen, A.M.,2002. Diurnal modulation of pacemaker potentials and calciumcurrent in the mammalian circadian clock. Nature 416,286–290.

Peterson, B.B., Liao, H.W., Dacey, D.M., Yau, K.-W., Gamlin, P.D.R.,Robinson, F.R., Marshak, D.W., 2003. Functional architecture ofthe photoreceptive ganglion cell in primate retina:morphology, mosaic organization and central targets ofmelanopsin immunostained cells. ARVO Abst. 5182.

Pickard, G.E., 1982. The afferent connections of thesuprachiasmatic nucleus of the golden hamster with emphasison the retinohypothalamic projection. J. Comp. Neurol. 211,65–83.

Pickard, G.E., 1985. Bifurcating axons of retinal ganglion cellsterminate in the hypothalamic suprachiasmatic nucleus andthe intergeniculate leaflet of the thalamus. Neurosci. Lett. 55,211–217.

Pickard, G.E., 1994. Intergeniculate leaflet ablation alters circadianrhythms in the mouse. NeuroReport 5, 2186–2188.

Pickard, G.E., Rea, M.A., 1997. TFMPP, a 5HT1B receptor agonist,inhibits light-induced phase shifts of the circadian activityrhythm and c-Fos expression in the mouse suprachiasmaticnucleus. Neurosci. Lett. 231, 95–98.

Pickard, G.E., Silverman, A.J., 1981. Direct retinal projections tohypothalamus, piriform cortex and accessory optic nuclei inthe golden hamster as demonstrated by a sensitiveanterograde horseradish peroxidase technique. J. Comp.Neurol. 196, 155–172.

Pickard, G.E., Turek, F.W., 1983. The suprachiasmatic nuclei: twocircadian clocks? Brain Res. 268, 201–210.

Pickard, G.E., Ralph, M.R., Menaker, M., 1987. The intergeniculateleaflet partially mediates effects of light on circadian rhythms.J. Biol. Rhythms 2, 35–56.

Pickard, G.E., Weber, E.T., Scott, P.A., Riberdy, A.F., Rea, M.A., 1996.5HT1B receptor agonists inhibit light-induced phase shifts ofbehavioral circadian rhythms and expression of theimmediate-early gene c-fos in the suprachiasmatic nucleus.J. Neurosci. 16, 8208–8220.

Pickard, G.E., Smith, B.N., Belenky, M., Rea, M.A., Dudek, F.E.,Sollars, P.J., 1999. 5-HT1B receptor-mediated presynapticinhibition of retinal input to the suprachiasmatic nucleus.J. Neurosci. 19, 4034–4045.

Pickard, G.E., Smeraski, C.A., Tomlinson, C.C., Banfield, B.W.,Kaufman, J., Wilcox, C.L., Enquist, L.W., Sollars, P.J., 2002.Intravitreal injection of the attenuated pseudorabies virus PRVBartha results in infection of the hamster suprachiasmaticnucleus only by retrograde transsynaptic transport viaautonomic circuits. J. Neurosci. 22, 2701–2710.

Piggins, H.D., Rusak, B., 1993. Electrophysiological effects ofpressure-ejected bombesin-like peptides on hamstersuprachiasmatic nucleus neurons in vitro. J. Neuroendocrinol.5, 575–581.

Piggins, H.D., Rusak, B., 1997. Effects of microinjections ofsubstance P into the suprachiasmatic nucleus region onhamster wheel-running rhythms. Brain Res. Bull. 42,451–455.

Piggins, H.D., Cutler, D.J., Rusak, B., 1994. Effects ofionophoretically applied bombesin-like peptides on hamstersuprachiasmatic nucleus neurons in vitro. Eur. J. Pharmacol.271, 413–419.

Piggins, H.D., Cutler, D.J., Rusak, B., 1995a. Ionophoretically appliedsubstance P activates hamster suprachiasmatic nucleusneurons. Brain Res. Bull. 37, 475–479.

Piggins, H.D., Antle, M.C., Rusak, B., 1995b. Neuropeptides phaseshift the mammalian circadian pacemaker. J. Neurosci. 15,5612–5622.

Piggins, H.D., Stamp, J.A., Burns, J., Rusak, B., Semba, K., 1996.Distribution of pituitary adenylate cyclase activatingpolypeptide (PACAP) immunoreactivity in the hypothalamusand extended amygdala of the rat. J. Comp. Neurol. 376,278–294.

Piggins, H.D., Marchant, E.G., Goguen, D., Rusak, B., 2001a.Phase-shifting effects of pituitary adenylate cyclase activatingpolypeptide on hamster wheel-running rhythms. Neurosci.Lett. 305, 25–28.

Piggins, H.D., Samuels, R.E., Coogan, A.N., Cutler, D.J., 2001b.Distribution of substance P and neurokinin-1 receptorimmunoreactivity in the suprachiasmatic nuclei and


intergeniculate leaflet of hamster, mouse, and rat. J. Comp.Neurol. 438, 50–65.

Porter, J.T., McCarthy, K.D., 1997. Astrocytic neurotransmitterreceptors in situ and in vivo. Prog. Neurobiol. 51, 439–455.

Prosser, R.A., 1998. Neuropeptide Y blocks serotonergic phaseshifts of the suprachiasmatic circadian clock in vitro. Brain Res.808, 31–41.

Prosser, R.A., Dean, R.R., Edgar, D.M., Heller, H.C., Miller, J.D., 1993.Serotonin and the mammalian circadian system: I. In vitrophase shifts by serotonergic agonists and antagonists. J. Biol.Rhythms 8, 1–16.

Prosser, R.A., Macdonald, E.S., Heller, H.C., 1994. c-fos mRNA inthe suprachiasmatic nuclei in vitro shows a circadian rhythmand responds to a serotonergic agonist. Mol. Brain Res. 25,151–156.

Prosser, R.A., Rutishauser, U., Ungers, G., Fedorkova, L., Glass, J.D.,2003. Intrinsic role of polysialylated neural cell adhesionmolecule in photic phase resetting of the mammaliancircadian clock. J. Neurosci. 23, 652–658.

Provencio, I., Foster, R.G., 1995. Circadian rhythms in mice can beregulated by photoreceptors with cone-like characteristics.Brain Res. 694, 183–190.

Provencio, I., Wong, S., Lederman, A.B., Argamaso, S.M., Foster,R.G., 1994. Visual and circadian responses to light in agedretinally degenerate mice. Vision Res. 34, 1799–1806.

Provencio, I., Jiang, G., De Grip, W.J., Hayes, W.P., Rollag, M.D.,1998a. Melanopsin: an opsin in melanophores, brain, and eye.Proc. Natl. Acad. Sci. U. S. A. 95, 340–345.

Provencio, I., Cooper, H.M., Foster, R.G., 1998b. Retinalprojections in mice with inherited retinal degeneration:implications for circadian photoentrainment. J. Comp.Neurol. 395, 417–439.

Provencio, I., Rodriguez, I.R., Jiang, G., Hayes, W.P., Moreira, E.F.,Rollag, M.D., 2000. A novel human opsin in the inner retina.J. Neurosci. 20, 600–605.

Provencio, I., Rollag, M.D., Castrucci, A.M., 2002. Photoreceptive netin the mammalian retina. Nature 415, 493.

Pu, M., 1999. Dendritic morphology of cat retinal ganglion cellsprojecting to suprachiasmatic nucleus. J. Comp. Neurol. 414,267–274.

Pu, M., 2000. Physiological response properties of cat retinalganglion cells projecting to suprachiasmatic nucleus. J. Biol.Rhythms 15, 31–36.

Pu, M.L., Pickard, G.E., 1996. Ventral lateral geniculate nucleusafferents to the suprachiasmatic nucleus in the cat. Brain Res.725, 247–251.

Qiu, X., Kumbalasiri, T., Carlson, S.M., Wong, K.Y., Krishna, V.,Provencio, I., Berson, D.M., 2005. Induction of photosensitivityby heterologous expression of melanopsin. Nature 433,745–749.

Quintero, J.E., Kuhlman, S.J., McMahon, D.G., 2000. MousePeriod1-driven GFP gene expression cycles in SCN slices and inindividual SCN cells. Soc. Res. Biol. Rhythms 7, 87.

Quintero, J.E., Kuhlman, S.J., McMahon, D.G., 2003. The biologicalclock nucleus: a multiphasic oscillator network regulated bylight. J. Neurosci. 23, 8070–8076.

Ralph, M.R., Menaker, M., 1987. Bicuculline blocks circadian phasedelays but not advances. Neurochem. Int. 11, 55–62.

Ralph, M.R., Menaker, M., 1989. GABA regulation of circadianresponses to light: I. Involvement of GABAA-benzodiazepineand GABAB receptors. J. Neurosci. 9, 2858–2865.

Ralph, M.R., Mrosovsky, N., 1992. Behavioral inhibition of circadianresponses to light. J. Biol. Rhythms 7, 353–359.

Ralph, M.R., Foster, R.G., Davis, F.C., Menaker, M., 1990.Transplanted suprachiasmatic nucleus determines circadianperiod. Science 247, 975–978.

Rea, M.A., 1989. Light increases Fos-related proteinimmunoreactivity in the rat suprachiasmatic nuclei. Brain Res.Bull. 23, 577–581.

Rea, M.A., 1992. Different populations of cells in thesuprachiasmatic nuclei express c-fos in association withlight-induced phase delays and advances of the free-runningactivity rhythm in hamsters. Brain Res. 579, 107–112.

Rea, M.A., Pickard, G.E., 2000a. Serotonergic modulation of photicentrainment in the Syrian hamster. Biol. Rhythm Res. 31,284–314.

Rea, M.A., Pickard, G.E., 2000b. A 5-HT1B receptor agonist inhibitslight-induced suppression of pineal melatonin production.Brain Res. 858, 424–428.

Rea, M.A., Ferriera, S., Randolph, W., Glass, J.D., 1993a. Daily profileof the extracellular concentration of glutamate in thesuprachiasmatic region of the Siberian hamster. Proc. Soc. Exp.Biol. Med. 204, 104–109.

Rea, M.A., Buckley, B., Lutton, L.M., 1993b. Local administration ofEAA antagonists blocks light-induced phase shifts and c-fosexpression in hamster SCN. Am. J. Physiol. 265, R1191–R1198.

Rea, M.A., Barrera, J., Glass, J.D., Gannon, R.L., 1995. Serotonergicpotentiation of photic phase shifts of the circadian activityrhythm. NeuroReport 6, 1289–1292.

Reddy, A.B., Field, M.D., Maywood, E.S., Hastings, M.H., 2002.Differential resynchronisation of circadian clock geneexpression within the suprachiasmatic nuclei of micesubjected to experimental jet lag. J. Neurosci. 22,7326–7330.

Redlin, U., Mrosovsky, N., 1999a. Masking of locomotor activity inhamsters. J. Comp. Physiol., A 184, 429–437.

Redlin, U., Mrosovsky, N., 1999b. Masking by light in hamsters withSCN lesions. J. Comp. Physiol., A 184, 439–448.

Redlin, U., Vrang, N., Mrosovsky, N., 1999. Enhanced maskingresponse to light in hamsters with IGL lesions. J. Comp.Physiol., A 184, 449–456.

Redlin, U., Cooper, H.M., Mrosovsky, N., 2003. Increased maskingresponse to light after ablation of the visual cortex in mice.Brain Res. 965, 1–8.

Reebs, S.G., Lavery, R.J., Mrosovsky, N., 1989. Running activitymediates the phase-advancing effects of dark pulses onhamster circadian rhythms. J. Comp. Physiol., A 165,811–818.

Reed, H.E., Cutler, D.J., Brown, T.M., Brown, J., Coen, C.W., Piggins,H.D., 2002. Effects of vasoactive intestinal polypeptide onneurones of the rat suprachiasmatic nuclei in vitro.J. Neuroendocrinol. 14, 639–646.

Reuss, S., 1991. Photoperiod effects on bombesin- andcholecystokinin-like immunoreactivity in the suprachiasmaticnuclei of the Djungarian hamster (Phodopus sungorus).Neurosci. Lett. 128, 13–16.

Reuss, S., Bürger, K., 1994. Substance P-like immunoreactivity inthe hypothalamic suprachiasmatic nucleus of Phodopussungorus—Relation to daytime, photoperiod, sex and age.Brain Res. 638, 189–195.

Reuss, S., Decker, K., Rosseler, L., Layes, E., Schollmayer, A.,Spessert, R., 1995. Nitric oxide synthase in the hypothalamicsuprachiasmatic nucleus of rat: evidence from histochemistry,immunohistochemistry and Western blot; and colocalizationwith VIP. Brain Res. 695, 257–262.

Rhoades, R.W., Hsu, L., Parfett, G., 1979. An electronmicroscopicanalysis of the optic nerve in the golden hamster. J. Comp.Neurol. 186, 491–504.

Roberts, W.M., 1994. Localization of calcium signals by a mobilecalcium buffer in frog saccular hair cells. J. Neurosci. 14,3246–3262.

Rosenwasser, A.M., Dwyer, S.M., 2002. Phase shifting the hamstercircadian clock by 15-minute dark pulses. J. Biol. Rhythms 17,238–247.

Ruby, N.F., Saran, A., Kang, T., Franken, P., Heller, H.C., 1996.Siberian hamsters free run or become arrhythmic after a phasedelay of the photocycle. Am. J. Physiol. 271, R881–R890.

Ruby, N.F., Kang, T., Heller, H.C., 1997. Melatonin attenuates photic


disruption of circadian rhythms in Siberian hamsters. Am. J.Physiol. 273, R1540–R1549.

Ruby, N.F., Joshi, N., Heller, H.C., 1998. Phase shift magnitude anddirection determine whether Siberian hamsters reentrain tothe photocycle. J. Biol. Rhythms 13, 506–517.

Ruby, N.F., Dubocovich, M.L., Heller, H.C., 2000. Siberian hamstersthat fail to reentrain to the photocycle have suppressedmelatonin levels. Am. J. Physiol.: Regul., Integr. Comp. Physiol.278, R757–R762.

Ruby, N.F., Brennan, T.J., Xie, X., Cao, V., Franken, P., Heller, H.C.,2002a. Role of melanopsin in circadian responses to light.Science 298, 2211–2213.

Ruby, N.F., Joshi, N., Heller, H.C., 2002b. Constant darkness restoresentrainment to phase-delayed Siberian hamsters. Am. J.Physiol.: Regul., Integr. Comp. Physiol. 283, R1314–R1320.

Ruby, N.F., Barakat, M.T., Heller, H.C., 2004. Phenotypic differencesin reentrainment behavior and sensitivity to nighttime lightpulses in Siberian hamsters. J. Biol. Rhythms 19, 530–541.

Rusak, B., Morin, L.P., 1976. Testicular responses to photoperiodare blocked by lesions of the suprachiasmatic nuclei in goldenhamsters. Biol. Reprod. 15, 366–374.

Rutishauser, U., Landmesser, L., 1996. Polysialic acid in thevertebrate nervous system: a promoter of plasticity in cell–cellinteractions. Trends Neurosci. 19, 422–427.

Saeb-Parsy, K., Dyball, R.E., 2003. Defined cell groups in the ratsuprachiasmatic nucleus have different day/night rhythms ofsingle-unit activity in vivo. J. Biol. Rhythms 18, 26–42.

Saper, C.B., Chou, T.C., Scammell, T.E., 2001. The sleep switch:hypothalamic control of sleep and wakefulness. TrendsNeurosci. 24, 726–731.

Sato, T., Kawamura, H., 1984. Circadian rhythms in multiple unitactivity inside and outside the suprachiasmatic nucleus in thediurnal chipmunk (Eutamias sibiricus). Neurosci. Res. 1, 45–52.

Schaap, J., Bos, N.P., de Jeu, M.T., Geurtsen, A.M., Meijer, J.H.,Pennartz, C.M., 1999. Neurons of the rat suprachiasmaticnucleus show a circadian rhythm inmembrane properties thatis lost during prolonged whole-cell recording. Brain Res. 815,154–166.

Schaap, J., Albus, H., Eilers, P.H., Detari, L., Meijer, J.H., 2001. Phasedifferences in electrical discharge rhythms between neuronalpopulations of the left and right suprachiasmatic nuclei.Neuroscience 108, 359–363.

Schaap, J., Albus, H., VanderLeest, H.T., Eilers, P.H., Detari, L.,Meijer, J.H., 2003. Heterogeneity of rhythmic suprachiasmaticnucleus neurons: implications for circadian waveform andphotoperiodic encoding. Proc. Natl. Acad. Sci. U. S. A. 100,15994–15999.

Schwartz, W.J., Takeuchi, J., Shannon, W., Davis, E.M., Aronin, N.,1994. Temporal regulation of light-induced Fos and Fos-likeprotein expression in the ventrolateral subdivision of the ratsuprachiasmatic nucleus. Neuroscience 58, 573–583.

Schwartz, W.J., Peters, R.V., Aronin, N., Bennett, M.R., 1996.Unexpected c-fos gene expression in the suprachiasmaticnucleus of mice entrained to a skeleton photoperiod. J. Biol.Rhythms 11, 35–44.

Schwartz, W.J., Carpino Jr., A., De la Iglesia, H.O., Baler, R., Klein,D.C., Nakabeppu, Y., Aronin, N., 2000. Differential regulationof fos family genes in the ventrolateral and dorsomedialsubdivisions of the rat suprachiasmatic nucleus.Neuroscience 98, 535–547.

Schwartz, M.D., Nunez, A.A., Smale, L., 2004. Differences inthe suprachiasmatic nucleus and lower subparaventricularzone of diurnal and nocturnal rodents. Neuroscience 127,13–23.

Semo, M., Munoz, L.M., Foster, R.G., Jeffery, G., 2005. Melanopsin(Opn4) positive cells in the cat retina are randomly distributedacross the ganglion cell layer. Vis. Neurosci. 22, 111–116.

Shearman, L.P., Zylka, M.J., Weaver, D.R., Kolakowski Jr., L.F.,Reppert, S.M., 1997. Two period homologs: circadian

expression and photic regulation in the suprachiasmaticnuclei. Neuron 19, 1261–1269.

Shearman, L.P., Sriram, S., Weaver, D.R., Maywood, E.S., Chaves,I., Zheng, B., Kume, K., Lee, C.C., van der Horst, G.T., Hastings,M.H., Reppert, S.M., 2000. Interacting molecular loops in themammalian circadian clock. Science 288, 1013–1019.

Shen, H., Semba, K., 1994. A direct retinal projection to the dorsalraphe nucleus in the rat. Brain Res. 635, 159–168.

Shen, H., Watanabe, M., Tomasiewicz, H., Rutishauser, U.,Magnuson, T., Glass, J.D., 1997. Role of neural cell adhesionmolecule and polysialic acid inmouse circadian clock function.J. Neurosci. 17, 5221–5229.

Shen, H., Glass, J.D., Seki, T., Watanabe, M., 1999. Ultrastructuralanalysis of polysialylated neural cell adhesion molecule in thesuprachiasmatic nuclei of the adult mouse. Anat. Rec. 256,448–457.

Shen, S., Spratt, C., Sheward, W.J., Kallo, I., West, K., Morrison, C.F.,Coen, C.W., Marston, H.M., Harmar, A.J., 2000. Overexpressionof the human VPAC2 receptor in the suprachiasmatic nucleusalters the circadian phenotype of mice. Proc. Natl. Acad. Sci.U. S. A. 97, 11575–11580.

Shen, H., Watanabe, M., Tomasiewicz, H., Glass, J.D., 2001. Geneticdeletions of NCAM and PSA impair circadian function in themouse. Physiol. Behav. 73, 185–193.

Shi, H., Bartness, T.J., 2001. Neurochemical phenotype ofsympathetic nervous system outflow from brain to white fat.Brain Res. Bull. 54, 375–385.

Shibata, S., Shiratsuchi, A., Liou, S.Y., Ueki, S., 1984. The role ofcalcium ions in circadian rhythm of suprachiasmatic nucleusneuron activity in rat hypothalamic slices. Neurosci. Lett. 52,181–184.

Shibata, S., Liou, S.Y., Ueki, S., 1986. Influence of excitatory aminoacid receptor antagonists and of baclofen on synaptictransmission in the optic nerve to the suprachiasmatic nucleusin slices of rat hypothalamus. Neuropharmacology 25, 403–409.

Shibata, S., Tsuneyoshi, A., Hamada, T., Tominaga, K., Watanabe,S., 1992. Effect of substance P on circadian rhythms of firingactivity and the 2-deoxyglucose uptake in the ratsuprachiasmatic nucleus in vitro. Brain Res. 597, 257–263.

Shibata, S., Watanabe, A., Hamada, T., Watanabe, S., 1994.Protein-synthesis inhibitor blocks(R,S)-α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid(AMPA)- or substance P-induced phase shift of the circadianrhythm of neuronal activity in the rat suprachiasmatic nucleusin vitro. Neurosci. Lett. 168, 159–162.

Shigeyoshi, Y., Taguchi, K., Yamamoto, S., Takekida, S., Yan, L.,Tei, H., Moriya, T., Shibata, S., Loros, J.J., Dunlap, J.C., Okamura,H., 1997. Light-induced resetting of a mammalian circadianclock is associated with rapid induction of the mPer1transcript. Cell 91, 1043–1053.

Shimazoe, T., Nakamura, S., Kobayashi, K., Watanabe, S.,Miyasaka, K., Kono, A., Funakoshi, A., 2004. Role of 5-HT1Breceptors in entrainment disorder of Otsuka Long EvansTokushima fatty (OLETF) rats. Neuroscience 123, 201–205.

Shirakawa, T., Moore, R.Y., 1994. Responses of rat suprachiasmaticnucleus neurons to substance P and glutamate in vitro. BrainRes. 642, 213–220.

Shirakawa, T., Honma, S., Katsuno, Y., Oguchi, H., Honma, K.,2000. Synchronization of circadian firing rhythms incultured rat suprachiasmatic neurons. Eur. J. Neurosci. 12,2833–2838.

Shiromani, P.J., Overstreet, D., Levy, D., Goodrich, C.A., Campbell,S.S., Gillin, J.C., 1988. Increased REM sleep in rats selectivelybred for cholinergic hyperactivity. Neuropsychopharmacology1, 127–133.

Shivers, K., Muscat, L., 2004. Relative distribution of crossed anduncrossed retinohypothalamic tract (RHT) projections to thesuprachiasmatic nucleus (SCN) in rat and mouse. Soc.Neurosci. 765.14.


Silver, J., Brand, S., 1979. A route for direct retinal input to thepreoptic hypothalamus: dendritic projections into the opticchiasm. Am. J Anat. 155, 391–401.

Silver, R., Lehman, M.N., Gibson, M., Gladstone, W.R., Bittman, E.L.,1990. Dispersed cell suspensions of fetal SCN restore circadianrhythmicity in SCN-lesioned adult hamsters. Brain Res. 525,45–58.

Silver, R., Romero, M.T., Besmer, H.R., Leak, R.K., Nunez, J.M.,LeSauter, J., 1996a. Calbindin-D28K cells in the hamster SCNexpress light-induced Fos. NeuroReport 7, 1224–1228.

Silver, R., LeSauter, J., Tresco, P.A., Lehman, M.N., 1996b. Adiffusible coupling signal from the transplantedsuprachiasmatic nucleus controlling circadian locomotorrhythms. Nature 382, 810–813.

Silver, R., Sookhoo, A.I., LeSauter, J., Stevens, P., Jansen, H.T.,Lehman, M.N., 1999. Multiple regulatory elements result inregional specificity in circadian rhythms of neuropeptideexpression in mouse SCN. NeuroReport 10, 3165–3174.

Smale, L., Boverhof, J., 1999. The suprachiasmatic nucleus andintergeniculate leaflet of Arvicanthis niloticus, a diurnal muridrodent from East Africa. J. Comp. Neurol. 403, 190–208.

Smale, L., Michels, K.M., Moore, R.Y., Morin, L.P., 1990. Destructionof the hamster serotonergic system by 5,7-DHT: effects oncircadian rhythm phase, entrainment and response totriazolam. Brain Res. 515, 9–19.

Smale, L., Blanchard, J.H., Moore, R.Y., Morin, L.P., 1991.Immunocytochemical characterization of the suprachiasmaticnucleus and the intergeniculate leaflet in the diurnal groundsquirrel, Spermophilus lateralis. Brain Res. 563, 77–86.

Smale, L., Castleberry, C., Nunez, A.A., 2001. Fos rhythms in thehypothalamus of Rattus and Arvicanthis that exhibit nocturnaland diurnal patterns of rhythmicity. Brain Res. 899, 101–105.

Smale, L., Lee, T., Nunez, A.A., 2003. Mammalian diurnality: somefacts and gaps. J. Biol. Rhythms 18, 356–366.

Smart, C.M., Biello, S.M., 2001. WAY-100635, a specific 5-HT1Aantagonist, can increase the responsiveness of themammalian circadian pacemaker to photic stimuli. Neurosci.Lett. 305, 33–36.

Smeraski, C.A., Sollars, P.J., Ogilvie, M.D., Enquist, L.W., Pickard,G.E., 2004. Suprachiasmatic nucleus input to autonomiccircuits identified by retrograde transsynaptic transport ofpseudorabies virus from the eye. J. Comp. Neurol. 471, 298–313.

Smith, R.D., Inouye, S.-I.T., Turek, F.W., 1989. Centraladministration of muscimol phase-shifts the mammaliancircadian clock. J. Comp. Physiol. 164, 805–814.

Smith, B.N., Sollars, P.J., Dudek, F.E., Pickard, G.E., 2001.Serotonergic modulation of retinal input to the mousesuprachiasmatic nucleus mediated by 5-HT1B and 5-HT7receptors. J. Biol. Rhythms 16, 25–38.

Sollars, P.J., Pickard, G.E., 1993. Time course of fiber outgrowthfrom fetal anterior hypothalamic heterografts. Brain Res. 614,212–219.

Sollars, P.J., Kimble, D.P., Pickard, G.E., 1995. Restoration ofcircadian behavior by anterior hypothalamic heterografts.J. Neurosci. 15, 2109–2122.

Sollars, P.J., Ogilvie, M.D., Rea, M.A., Pickard, G.E., 2002. 5-HT1Breceptor knockout mice exhibit an enhanced response toconstant light. J. Biol. Rhythms 17, 428–437.

Sollars, P.J., Smeraski, C.A., Kaufman, J.D., Ogilvie, M.D., Provencio,I., Pickard, G.E., 2003. Melanopsin and non-melanopsinexpressing retinal ganglion cells innervate the hypothalamicsuprachiasmatic nucleus. Vis. Neurosci. 20, 601–610.

Speh, J.C., Moore, R.Y., 1990. Synaptogenesis andretinohypothalamic tract (RHT) development in the hamstersuprachiasmatic nucleus (SCN). Society for Neuroscience 16,602 (Ref Type: Abstract).

Speh, J.C., Moore, R.Y., 1993. Retinohypothalamic tractdevelopment in the hamster and rat. Dev. Brain Res. 76,171–181.

Spoelstra, K., Oklejewicz, M., Daan, S., 2002. Restoration ofself-sustained circadian rhythmicity by themutant clock allelein mice in constant illumination. J. Biol. Rhythms 17, 520–525.

Sprouse, J., Reynolds, L., Braselton, J., Schmidt, A., 2004a.Serotonin-induced phase advances of SCN neuronal firing invitro: a possible role for 5-HT5A receptors? Synapse 54,111–118.

Sprouse, J., Reynolds, L., Li, X., Braselton, J., Schmidt, A., 2004b.8-OH-DPAT as a 5-HT7 agonist: phase shifts of the circadianbiological clock through increases in cAMP production.Neuropharmacology 46, 52–62.

Stamp, J.A., Piggins, H.D., Rusak, B., Semba, K., 1997. Distribution ofionotropic glutamate receptor subunit immunoreactivity inthe suprachiasmatic nucleus and intergeniculate leaflet of thehamster. Brain Res. 756, 215–224.

Starkey, S.J., 1996. Melatonin and 5-hydroxytryptaminephase-advance the rat circadian clock by activation of nitricoxide synthesis. Neurosci. Lett. 211, 199–202.

Stephan, F.K., Donaldson, J.A., Gellert, J., 1982.Retinohypothalamic tract symmetry and phase shifts ofcircadian rhythms in rats and hamsters. Physiol. Behav. 29,1153–1159.

Stetson, M.H., Watson-Whitmyre, M., 1976. Nucleussuprachiasmaticus: the biological clock in the hamster?Science 191, 197–199.

Summer, T.L., Ferraro, J.S., McCormack, C.E., 1984. Phase-responseand Aschoff illuminance curves for locomotor activity rhythmof the rat. Am. J. Physiol. 246, R299–R304.

Sumova, A., Illnerova, H., 1996. Melatonin instantaneously resetsintrinsic circadian rhythmicity in the rat suprachiasmaticnucleus. Neurosci. Lett. 218, 181–184.

Sumova, A., Travnickova, Z., Mikkelsen, J.D., Illnerova, H., 1998.Spontaneous rhythm in c-Fos immunoreactivity in thedorsomedial part of the rat suprachiasmatic nucleus. BrainRes. 801, 254–258.

Sumova, A., Travnickova, Z., Illnerova, H., 2000. Spontaneous c-Fosrhythm in the rat suprachiasmatic nucleus: location and effectof photoperiod. Am. J. Physiol.: Regul., Integr. Comp. Physiol.279, R2262–R2269.

Sumova, A., Bendova, Z., Sladek, M., Kovacikova, Z., Illnerova, H.,2004. Seasonal molecular timekeeping within the rat circadianclock. Physiol. Res. 53 (Suppl. 1), S167–S176.

Swann, J.M., Turek, F.W., 1985. Multiple circadian oscillatorsregulate the timing of behavioral and endocrine rhythms infemale golden hamsters. Science 228, 898–900.

Takahashi, J.S., DeCoursey, P.J., Bauman, L., Menaker, M., 1984.Spectral sensitivity of a novel photoreceptive systemmediating entrainment of mammalian circadian rhythms.Nature 308, 186–188.

Takatsuji, K., Miguel-Hidalgo, J.-J., Tohyama, M., 1991. SubstanceP-immunoreactive innervation from the retina to thesuprachiasmatic nucleus in the rat. Brain Res. 568, 223–229.

Takatsuji, K., Senba, E., Mantyh, P.W., Tohyama, M., 1995. Arelationship between substance P receptor and retinal fibers inthe rat suprachiasmatic nucleus. Brain Res. 698, 53–61.

Takeuchi, Y., Takahashima, M., Katoh, Y., Nishikawa, T.,Takahashi, K., 1991. N-Methyl-D-aspartate, quisqualate andkainate receptors are all involved in transmission of photicstimulation in the suprachiasmatic nucleus in rats. Brain Res.563, 127–131.

Tang, I.H., Murakami, D.M., Fuller, C.A., 2002. Unilateral opticnerve transection alters light response of suprachiasmaticnucleus and intergeniculate leaflet. Am. J. Physiol. 282,R569–R577.

Teclemariam-Mesbah, R., ter Horst, G.J., Postema, F., Wortel, J.,Buijs, R.M., 1999. Anatomical demonstration of thesuprachiasmatic nucleus-pineal pathway. J. Comp. Neurol.406, 171–182.

Teshima, K., Kim, S.H., Allen, C.N., 2003. Characterization of an


apamin-sensitive potassium current in suprachiasmaticnucleus neurons. Neuroscience 120, 65–73.

Thomson, A.M., 1984. Slow, regular discharge in suprachiasmaticneurons is calcium dependent, in slices of rat brain.Neuroscience 13, 761–767.

Thomson, A.M., West, D.C., 1990. Factors affecting slow regularfiring in the suprachiasmatic nucleus in vitro. J. Biol. Rhythms5, 59–75.

Tiao, Y.-C., Blakemore, C., 1976. Regional specialization in thegolden hamster's retina. J. Comp. Neurol. 168, 439–458.

Tieman, S.B., Tieman, D.G., 1996. N-Acetylaspartylglutamateimmunoreactivity in human retina. Vision Res. 36, 941–947.

Tieman, S.B., Moffett, J.R., Irtenkauf, S.M., 1991. Effect of eyeremoval on N-acetylaspartylglutamate immunoreactivity inretinal targets of the cat. Brain Res. 562, 318–322.

Tierno, A., Fiore, P., Gannon, R.L., 2002. Delta opioid inhibition oflight-induced phase advances in hamster circadian activityrhythms. Brain Res. 937, 66–73.

Tischler, R.C., Morin, L.P., 2003. Reciprocal serotonergicconnections between the hamster median and dorsal raphenuclei. Brain Res. 981, 126–132.

Tokunaga, A., Ono, K., Kondo, S., Tanaka, H., Kurose, K., Nagai, H.,1992. Retinal projections in the house musk shrew, Suncusmurinus, as determined by anterograde transport of WGA-HRP.Brain Behav. Evol. 40, 321–329.

Tominaga, K., Shibata, S., Ueki, S., Watanabe, S., 1992. Effects of5-HT1A receptor agonists on the circadian rhythm ofwheel-running activity in hamsters. Eur. J. Pharmacol. 214,79–84.

Tournier, B.B., Menet, J.S., Dardente, H., Poirel, V.J., Malan, A.,Masson-Pevet, M., Pevet, P., Vuillez, P., 2003. Photoperioddifferentially regulates clock genes' expression in thesuprachiasmatic nucleus of Syrian hamster. Neuroscience 118,317–322.

Tousson, E., Meissl, H., 2004. Suprachiasmatic nuclei grafts restorethe circadian rhythm in the paraventricular nucleus of thehypothalamus. J. Neurosci. 24, 2983–2988.

Trachsel, L., Heller, H.C., Miller, J.D., 1995. Nicotinephase-advances the circadian neuronal activity rhythm in ratsuprachiasmatic nuclei explants. Neuroscience 65, 797–803.

Treep, J.A., Abe, H., Rusak, B., Goguen, D.M., 1995. Two distinctretinal projections to the hamster suprachiasmatic nucleus.J. Biol. Rhythms 10, 299–307.

Ueyama, T., Krout, K.E., Van Nguyen, X., Karpitskiy, V., Kollert, A.,Mettenleiter, T.C., Loewy, A.D., 1999. Suprachiasmatic nucleus:a central autonomic clock. Nat. Neurosci. 2, 1051–1053.

Underwood, H., Menaker, M., 1970. Extraretinal light perception:entrainment of the biological clock controlling lizardlocomotor activity. Science 170, 190–193.

Urbanski, H.F., Doan, A., Pierce, M., 1991. Immunocytochemicalinvestigation of luteinizing hormone- releasing hormoneneurons in Syrian hamsters maintained under long or shortdays. Biol. Reprod. 44, 687–692.

van den Pol, A.N., 1980. The hypothalamic suprachiasmaticnucleus of rat: intrinsic anatomy. J. Comp. Neurol. 191,661–702.

van den Pol, A.N., 1991. The suprachiasmatic nucleus:morphological and cytochemical substrates for cellularinteraction. In: Klein, D.C., Moore, R.Y., Reppert, S.M. (Eds.),Suprachiasmatic Nucleus: The Mind's Clock. Oxford Univ.Press, New York, pp. 17–50.

van den Pol, A.N., Gorcs, T., 1986. Synaptic relationships betweenneurons containing vasopressin, gastrin-releasing peptide,vasoactive intestinal polypeptide, and glutamatedecarboxylase immunoreactivity in the suprachiasmaticnucleus: dual ultrastructural immunocytochemistry withgold-substituted silver peroxidase. J. Comp. Neurol. 252,507–521.

van den Pol, A.N., Tsujimoto, K.L., 1985. Neurotransmitters of

the hypothalamic suprachiasmatic nucleus:immunocytochemical analysis of 25 neuronal antigens.Neuroscience 15, 1049–1086.

van den Pol, A.N., Cao, V., Heller, H.G., 1998. Circadian system ofmice integrates brief light stimuli. Am. J. Physiol.: Regul., Integr.Comp. Physiol. 275, R654–R657.

Van der Beek, E.M., Horvath, T.L., Wiegant, V.M., Buijs, R.M.,1997. Evidence for a direct neuronal pathway from thesuprachiasmatic nucleus to the gonadotropin-releasinghormone system: combined tracing and light and electronmicroscopic immunocytochemical studies. J. Comp. Neurol.384, 569–579.

Van der Gucht, E., Hof, P.R., Arckens, L., 2003. Neurochemicalorganization, architectonic subdivision and three-dimensionalreconstruction of cat ventral lateral geniculate nucleus andmonkey pregeniculate nucleus. Soc. Neurosci. 33, 68.7.

Van der Horst, G.T.J., Muijtjens, M., Kobayashi, K., Takano, R.,Kanno, S., Takao, M., De Wit, J., Verkerk, A., Eker, A.P.M., VanLeenen, D., Buijs, R., Bootsma, D., Hoeijmakers, J.H.J., Yasui, A.,1999. Mammalian Cry1 and Cry2 are essential for maintenanceof circadian rhythms. Nature 398, 627–630.

Van Reeth, O., Turek, F.W., 1989. Stimulated activity mediatesphase shifts in the hamster circadian clock induced by darkpulses or benzodiazepines. Nature 339, 49–51.

Van Reeth, O., Hinch, D., Tecco, J.M., Turek, F.W., 1991. The effectsof short periods of immobilization on the hamster circadianclock. Brain Res. 545, 208–214.

Vansteensel, M.J., Yamazaki, S., Albus, H., Deboer, T., Block, G.D.,Meijer, J.H., 2003a. Dissociation between circadian Per1 andneuronal and behavioral rhythms following a shiftedenvironmental cycle. Curr. Biol. 13, 1538–1542.

Vansteensel, M.J., Deboer, T., Dahan, A., Meijer, J.H., 2003b. Onsetand offset following GABA-ergic and opioid receptor activation.J. Biol. Rhythms 18, 297–306.

Varcoe, T.J., Kennaway, D.J., Voultsios, A., 2003. Activation of5-HT2C receptors acutely induces Per gene expression in therat suprachiasmatic nucleus at night. Brain Res. Mol. Brain Res.119, 192–200.

Vertes, R.P., 1991. A PHA-L analysis of ascending projections ofthe dorsal raphe nucleus in the rat. J. Comp. Neurol. 313,643–668.

Vidal, L., Morin, L.P., 2005. Hypothalamic and zona incerta neuronsexpressing hypocretin, but not melanin concentratinghormone, project to the hamster intergeniculate leaflet.Neuroscience 134, 1081–1090.

Vincent, S.R., Kimura, H., 1992. Histochemical mapping of nitricoxide synthase in the rat brain. Neuroscience 46, 755–784.

Vindlacheruvu, R.R., Ebling, F.J.P., Maywood, E.S., Hastings, M.H.,1992. Blockade of glutamatergic neurotransmission in thesuprachiasmatic nucleus prevents cellular and behaviouralresponses of the circadian system to light. Eur. J. Neurosci. 4,673–679.

Vitaterna, M.H., King, D.P., Chang, A.-M., Kornhauser, J.M., Lowrey,P.L, McDonald, J.D., Dove, W.F., Pinto, L.H., Turek, F.W.,Takahashi, J.S., 1994. Mutagenesis and mapping of a mousegene, Clock, essential for circadian behavior. Science 264,719–725.

Vrang, N., Larsen, P.J., Moller, M., Mikkelsen, J.D., 1995a.Topographical organization of the ratsuprachiasmatic-paraventricular projection. J. Comp. Neurol.353, 585–603.

Vrang, N., Larsen, P.J., Mikkelsen, J.D., 1995b. Direct projectionfrom the suprachiasmatic nucleus to hypophysiotrophiccorticotropin-releasing factor immunoreactive cells in theparaventricular nucleus of the hypothalamus demonstrated bymeans of Phaseolus vulgaris-leucoagglutinin tract tracing. BrainRes. 684, 61–69.

Vrang, N., Mrosovsky, N., Mikkelsen, J.D., 2003. Afferentprojections to thehamster intergeniculate leaflet demonstrated


by retrograde and anterograde tracing. Brain Res. Bull. 59,267–288.

Wagner, S., Castel, M., Gainer, H., Yarom, Y., 1997. GABA in themammalian suprachiasmatic nucleus and its role in diurnalrhythmicity. Nature 387, 598–603.

Wagner, S., Sagiv, N., Yarom, Y., 2001. GABA-induced current andcircadian regulation of chloride in neurones of the ratsuprachiasmatic nucleus. J. Physiol. 537, 853–869.

Wang, H., Morris, J.F., 1996. Presence of neuronal nitric oxidesynthase in the suprachiasmatic nuclei of mouse and rat.Neuroscience 74, 1059–1068.

Warren, E.J., Allen, C.N., Brown, R.L., Robinson, D.W., 2003.Intrinsic light responses of retinal ganglion cells projecting tothe circadian system. Eur. J. Neurosci. 17, 1727–1735.

Watanabe, A., Ono, M., Shibata, S., Watanabe, S., 1995. Effect of anitric oxide synthase inhibitor, N-nitro-L-arginine methylester,on light-induced phase delay of circadian rhythm ofwheel-running activity in golden hamsters. Neurosci. Lett. 192,25–28.

Watanabe, K., Vanecek, J., Yamaoka, S., 2000. In vitro entrainmentof the circadian rhythm of vasopressin-releasing cells insuprachiasmatic nucleus by vasoactive intestinal polypeptide.Brain Res. 877, 361–366.

Watts, A.G., Swanson, L.W., 1987. Efferent projections of thesuprachiasmatic nucleus: II. Studies using retrograde transportof fluorescent dyes and simultaneous peptideimmunohistochemistry in the rat. J. Comp. Neurol. 258,230–252.

Watts, A.G., Swanson, L.W., Sanchez-Watts, G., 1987. Efferentprojections of the suprachiasmatic nucleus: I. Studies usinganterograde transport of Phaseolus vulgaris leucoagglutinin inthe rat. J. Comp. Neurol. 258, 204–229.

Weber, E.T., Rea, M.A., 1997. Neruopeptide Y blocks light-inducedphase advances but not delays of the circadian activity rhythmin hamsters. Neurosci. Lett. 231, 159–162.

Weber, E.T., Gannon, R.L., Michel, A.M., Gillette, M.U., Rea, M.A.,1995. Nitric oxide synthase inhibitor blocks light-inducedphase shifts of the circadian activity rhythm, but not c-fosexpression in the suprachiasmatic nucleus of the Syrianhamster. Brain Res. 692, 137–142.

Welsh, D.K., Logothetis, D.E., Meister, M., Reppert, S.M., 1995.Individual neurons dissociated from rat suprachiasmaticnucleus express independently phased circadian firingrhythms. Neuron 14, 697–706.

Wheal, H.V., Thomson, A.M., 1984. The electrical properties ofneurones of the rat suprachiasmatic nucleus recordedintracellularly in vitro. Neuroscience 13, 97–104.

Wickland, C., Turek, F.W., 1994. Lesions of the thalamicintergeniculate leaflet block activity-induced phase shifts inthe circadian activity rhythm of the golden hamster. Brain Res.660, 293–300.

Wollnik, F., Brysch, W., Uhlmann, E., Gillardon, F., Bravo, R.,Zimmermann, M., Schlingensiepen, K.H., Herdegen, T., 1995.Block of c-Fos and JunB expression by antisenseoligonucleotides inhibits light-induced phase shifts of themammalian circadian clock. Eur. J. Neurosci. 7, 388–393.

Yagita, K., Tamanini, F., van der Horst, G.T., Okamura, H., 2001.Molecular mechanisms of the biological clock in culturedfibroblasts. Science 292, 278–281.

Yamaguchi, S., Isejima, H., Matsuo, T., Okura, R., Yagita, K.,Kobayashi, M., Okamura, H., 2003. Synchronization of cellularclocks in the suprachiasmatic nucleus. Science 302, 1408–1412.

Yamazaki, S., Goto, M., Menaker, M., 1999. No evidence forextraocular photoreceptors in the circadian system of theSyrian hamster. J. Biol. Rhythms 14, 197–201.

Yamazaki, S., Numano, R., Abe, M., Hida, A., Takahashi, R., Ueda,M., Block, G.D., Sakaki, Y., Menaker, M., Tei, H., 2000. Resettingcentral and peripheral circadian oscillators in transgenic rats.Science 288, 682–685.

Yan, L., Okamura, H., 2002. Gradients in the circadian expressionof Per1 and Per2 genes in the rat suprachiasmatic nucleus. Eur.J. Neurosci. 15, 1153–1162.

Yan, L., Silver, R., 2004. Resetting the brain clock: time course andlocalization of mPER1 and mPER2 protein expression insuprachiasmatic nuclei during phase shifts. Eur. J. Neurosci. 19,1105–1109.

Yan, L., Takekida, S., Shigeyoshi, Y., Okamura, H., 1999. Per1 andPer2 gene expression in the rat suprachiasmatic nucleus:circadian profile and the compartment-specific response tolight. Neuroscience 94, 141–150.

Yannielli, P.C., Harrington, M.E., 2000. Neuropeptide Y applied invitro can block the phase shifts induced by light in vivo.NeuroReport 11, 1587–1591.

Yannielli, P.C., Harrington, M.E., 2001. The neuropeptide YY5receptor mediates the blockade of “photic-like”NMDA-induced phase shifts in the golden hamster.J. Neurosci. 21, 5367–5373.

Yannielli, P., Harrington, M.E., 2004. Let there be “more” light:enhancement of light actions on the circadian system throughnon-photic pathways. Prog. Neurobiol. 74, 59–76.

Yannielli, P.C., McKinley, B.J., Harrington, M.E., 2002. Is novelwheel inhibition of per1 and per2 expression linked to phaseshift occurrence? Neuroscience 112, 677–685.

Yannielli, P.C., Brewer, J.M., Harrington, M.E., 2004. Blockade of theNPY Y5 receptor potentiates circadian responses to light:complementary in vivo and in vitro studies. Eur. J. Neurosci. 19,891–897.

Yokota, S.I., Horikawa, K., Akiyama, M., Moriya, T., Ebihara, S.,Komuro, G., Ohta, T., Shibata, S., 2000. Inhibitory action ofbrotizolam on circadian and light-induced per1 and per2expression in the hamster suprachiasmatic nucleus.Br. J. Pharmacol. 131, 1739–1747.

Yoo, S.H., Yamazaki, S., Lowrey, P.L., Shimomura, K., Ko, C.H.,Buhr, E.D., Siepka, S.M., Hong, H.K., Oh,W.J., Yoo, O.J., Menaker,M., Takahashi, J.S., 2004. PERIOD2∷LUCIFERASE real-timereporting of circadian dynamics reveals persistent circadianoscillations in mouse peripheral tissues. Proc. Natl. Acad. Sci.U. S. A. 101, 5339–5346.

Yoshimura, T., Ebihara, S., 1996. Spectral sensitivity ofphotoreceptors mediating phase shifts of circadian rhythms inretinally degenerate CBA/J (rd/rd) and normal CBA/N(+/+) mice.J. Comp. Physiol., A 178, 797–802.

Yoshimura, T., Nishio, M., Goto, M., Ebihara, S., 1994. Differences incircadian photosensitivity between retinally degenerate CBA/Jmice (rd/rd) and normal CBA/N mice (+/+). J. Biol. Rhythms 9,51–60.

Zhang, Y., Zee, P.C., Kirby, J.D., Takahashi, J.S., Turek, F.W., 1993. Acholinergic antagonist, mecamylamine, blocks light-inducedFos immunoreactivity in specific regions of the hamstersuprachiasmatic nucleus. Brain Res. 615, 107–112.

Zhang, Y., Kornhauser, J.M., Zee, P.C., Mayo, K.E., Takahashi, J.S.,Turek, F.W., 1996. Effects of aging on light-inducedphase-shifting of circadian behavioral rhythms, Fos expressionand CREB phosphorylation in the hamster suprachiasmaticnucleus. Neuroscience 70, 951–961.

Zylka, M.J., Shearman, L.P., Weaver, D.R., Reppert, S.M., 1998.Three period homologs in mammals: differential lightresponses in the suprachiasmatic circadian clock andoscillating transcripts outside of brain. Neuron 20,1103–1110.

The circadian visual system, 2005 · 2006-10-19 · Review The circadian visual system, 2005 L.P....

Documents

Transcript of The circadian visual system, 2005 · 2006-10-19 · Review The circadian visual system, 2005 L.P....