Chapter 2 - Functional Circuit Development in the Auditory System · 2019-12-21 · 2.1...

This article was originally published in the Comprehensive Developmental Neuroscience: Neural Circuit Development and Function in the Brain published by Elsevier, and the attached copy is provided by Elsevier for the author's benefit and for the benefit of the author's institution, for non-commercial research and educational use including without limitation use in instruction at your institution, sending it to specific colleagues who you

know, and providing a copy to your institution’s administrator.

All other uses, reproduction and distribution, including without limitation commercial reprints, selling or licensing copies or access, or posting on open internet sites, your personal or institution’s website or repository, are prohibited. For exceptions, permission may be sought for such use through Elsevier's permissions site at:

http://www.elsevier.com/locate/permissionusematerial

Polley D.B., Seidl A.H., Wang Y. and Sanchez J.T. (2013) Functional Circuit Development in the Auditory

System. In: RUBENSTEIN J. L. R. and RAKIC P. (ed.) Comprehensive Developmental Neuroscience: Neural Circuit Development and Function in the Brain, volume 3, pp. 21-39 Amsterdam: Elsevier.

© 2013 Elsevier Inc. All rights reserved.

Neural Circuit Development and Function in the Brain: Comprehensiv

Neuroscience, Volume 3 http://dx.doi.org/10.1016/B978-0-12-397

Comprehensive Developmental Neuros

C H A P T E R

Author's personal copy

2

Functional Circuit Development in theAuditory System

D.B. Polley1, A.H. Seidl2, Y. Wang2, J.T. Sanchez21Eaton-Peabody Laboratory, Massachusetts Eye and Ear Infirmary & Dept. of Otology and Laryngology,

Harvard Medical School, Boston, MA, USA; 2Virginia Merrill Bloedel Hearing Research Center & Dept. of

Otolaryngology, Head and Neck Surgery, University of Washington, Seattle, WA, USA

O U T L I N E

2.1 Introduction to Auditory SystemDevelopment 222.1.1 A Neurobiological Approach to Studying

Auditory System Development 222.1.2 Basic Concepts of Cochlear Transduction 222.1.3 Scope of this Chapter 23

2.2 Development of Peripheral Circuits 232.2.1 Developing Networks Within the Cochlea 232.2.2 Development of the Place Code 242.2.3 Development of Afferent and

Efferent Circuits 25Phase 1: Well Before Hearing Onset 25

Phase 2: Shortly Before Hearing Onset 26

Phase 3: Shortly After Hearing Onset 26

2.2.4 Conclusions 27

2.3 Development of Brainstem Circuits 272.3.1 Functional Circuit Assembly in the

Brainstem 272.3.2 Development of Fine-Scale Connectivity

in the MSO 282.3.3 Development of Fine-Scale Connectivity

in the LSO 29

2.3.4 Afferent Regulation of Cochlear NucleusDevelopment 29

2.3.5 Afferent Regulation of 3rd-Order BrainstemNuclei 30

2.3.6 Influence of the Source and Pattern of AfferentActivity on Brainstem Circuits 31

2.3.7 Conclusions 31

2.4 Development of Auditory Midbrain andForebrain Circuits 322.4.1 Development of Thalamocortical Subplate

Circuitry 322.4.2 Postnatal Development of Local

Cortical Circuits 332.4.3 Afferent Regulation of Higher Auditory Circuit

Development 342.4.4 Developmental Regulation over Reinstating

Hearing in the Deaf 352.4.5 Experience-Dependent Influences on

Functional Circuit Development 362.4.6 Conclusions and Directions for

Future Research 37

References 38

e Develop

267-5.00

cience

21mental

136-9

: Neural Circuit Developme

# 2013 Elsevier Inc. All rights reserved.

nt and Function in the Brain, (2013), vol. 3, pp. 21-39

http://dx.doi.org/10.1016/B978-0-12-397267-5.00136-9

22 2. FUNCTIONAL CIRCUIT DEVELOPMENT IN THE AUDITORY SYSTEM


2.1 INTRODUCTION TO AUDITORYSYSTEM DEVELOPMENT

2.1.1 A Neurobiological Approach to StudyingAuditory System Development

The auditory system undergoes a series of profoundchanges from the time neural circuits begin forming inthe fetal brain to the day, years later, when a childfirst comprehends a complete sentence. The processesunfolding during this period are a fascinating mixtureof intrinsic molecular orchestration and activity-dependent refinement. In humans, auditory perceptualdevelopment is a protracted process that begins late inthe second trimester when the fetus first shows discrim-inative changes in heart rate to variations in specificsound features. Progressive improvements in the abilityto resolve variations in frequency, temporal patterns andspatial positions of sounds are observed throughout in-fancy and early childhood, and these changes often par-allel an increasing capacity to discriminate phonologicalunits of a child’s native language (Werker, 2005). While agreater understanding of the processes at work in hu-man auditory development is of paramount importance,these efforts are often complicated by an inability to iso-late the contributions of sensory system developmentfrom other cognitive and physical factors.

The use of model systems such as birds and rodentshas provided researchers with direct access to centraland peripheral auditory circuits, and has elucidatedmany of the basic mechanisms that underlie theirchanges during ontogeny. Songbirds and chickens havelong been popular model systems due to the greater easeof studying and manipulating the embryo in the egg (inovo) rather than in utero. As a result, we can now drawinformation from an extensive corpus of work detailingthe organization of brainstem and forebrain pathways,and the similarities by which songbirds acquire theirsong and humans acquire speech, particularly in theirdependence on auditory feedback during sensitive pe-riods of development. Interestingly, birds begin hearingapproximately eight days before hatching, around em-bryonic (E) day 11, while altricial mammals such asmice,rats and gerbils do not respond to airborne sound untilthe second postnatal (P) week of life. The relatively lateonset of hearing in rodents has aided research into thecellular and molecular changes that underlie abruptchanges in circuit development in the days before andafter the onset of hearing.

Recent studies suggest that molecular cues, whose ex-pression is genetically controlled, play an essential rolein the formation of topographically ordered connectionsin the auditory system. It is also evident that factorslinked to the flux of ions across the cell membrane dur-ing and following the action potential also support

I. CIRCUIT DEV

Comprehensive Developmental Neuroscience: Neural Circuit Deve

neuron survival and regulate the growth and topo-graphic specificity of axons and dendrites within audi-tory brain areas. In some brain areas, instructiveelectrical signals generated through the “closed-loop”spontaneous activity is sufficient to promote normalneural circuit development, whereas higher levels ofthe auditory pathway require structured activity pat-terns arising from acoustic signals to guide their finalstages of assembly. The influence of each activity-dependent and activity-independent factor waxes andwanes within defined windows of development, andpiecing together the chronology andmechanisms behindeach epoch of auditory system development representsone of the fundamental challenges for researchers in thisfield. The chapter describes the current state of knowl-edge concerning the interplay of these factors in the es-tablishment of functional circuits from the cochlea to thecerebral cortex.

2.1.2 Basic Concepts of Cochlear Transduction

The encoding of auditory information begins with ahighly specialized receptor organ known as the cochleain mammals and the basilar papilla in birds. In mam-mals, pressure waves are delivered into the fluid filledcochlea via the middle ear bones, setting the cochlearpartition into motion. The cochlear partition consists ofthe basilar membrane (BM), the organ of Corti, and thetectorial membrane (Figure 2.1(a)). As a result of thismo-tion and the mechanical properties of these structures, arelatively crude spectral analysis is performed by a vi-bratory pattern that occurs along the partition, a passiveprocess referred to as the traveling wave (Von Bekesy,1960). This traveling wave results in a vibratory patternsuch that high frequency sounds produce maximumdis-placement near the beginning (base) of the partition,while low frequency sounds vibrate maximally nearthe end (apex) of the partition.

The sensory receptors in mammals responsible fortransducing these hydro-mechanical vibrations are lo-cated in the organ of Corti and are known as innerand outer hair cells (IHC and OHC, respectively). Bothtypes of receptors are polarized epithelial cells that con-tain mechano-sensitive organelles located on their api-cal surface. These actin-filled hair-like processes aretermed stereocilia and contain mechanotransducerchannels near the tips of the ciliary bundles. The bun-dles are deflected as a result of partition movement,opening the channels and depolarizing the hair celldue to an influx of potassium (Kþ). This unconventionalform of depolarization activates voltage gated calcium(Ca2þ) channels, which triggers the release of the excit-atory neurotransmitter glutamate. Glutamate binds topostsynaptic receptors located on first-order spiral

ELOPMENT

lopment and Function in the Brain, (2013), vol. 3, pp. 21-39

Tectorial membrane

(a) (b)

High

Sound

frequencyLow

Basilar membrane

IHC

OHCOHC

OHC

FIGURE 2.1 Connections within the adult auditory system. Schematics illustrate the position of cell bodies and their interconnections inmammalian Organ of Corti (a) and initial components of the auditory pathway (b). (a) Two classes of mechanical sensory receptors, inner haircells (IHC) and outer hair cells (OHC) are innervated by dendrites that carry afferent signals to the brain (green) and efferent axons that conveysignals to the brain (gray). (B) Sounds of varying frequencies maximally excite particular regions along the basilar membrane (BM). Maximal sen-sitiveness aremapped systematically across the length of the BM, such that high frequency vibrations excite the base of the BM and low frequenciesexcite the apex. This tonotopic organization of frequency preference is preserved through topographically ordered projections to the central au-ditory system. Distinct populations of cells on both sides of the brainstem send efferent projections to the cochlea.

232.2 DEVELOPMENT OF PERIPHERAL CIRCUITS


ganglion neurons (green dendrites, Figure 2.1(a)), initi-ating saltatory conduction of action potentials to audi-tory nuclei in the brainstem.

The generic description above for sensory transduc-tion holds true for all hair cells in both birds and mam-mals. However, the separation into IHCs andOHCs, andtheir differences, are remarkable and unique to mam-mals. The IHCs are the true sensory receptor, receivingapproximately 95% of afferent innervation. In contrast,OHCs receive only around 5% of afferent contacts, sug-gesting a minor role in sensory transduction. Despitethis dichotomy, OHCs are thought to have highly-specialized “motor” functions. One such active functionis to enhance the bundle displacement caused by the vi-bration of the cochlear partition, acting as a positive feed-back on the bundle and the consequent amplification ofits displacement. A second active function involves anelongation and contraction of the OHC itself. As theOHC changes length, it feeds backmechanical force ontoits surrounding environment, a process termed electro-motility. This results in a highly significant and spatiallysegregated enhancement of the basilar membrane vibra-tory amplitude. Accordingly, the relatively crude sensi-tivity and frequency selectivity of the basilar membranethat arises through its passive biomechanical propertiesis substantially refined through hair cell active amplifi-cation. Not only do hair cells transmit signals to the cen-tral nervous system, they are also innervated by efferentaxons from the brain, which communicate with OHCs toextend the dynamic range of hair cell signaling and pro-tect the Organ of Corti from acoustic overexposure(Figure 2.1(b)). For detailed reviews of auditory trans-duction and the distinct mechanisms that create activecochlear amplification in mammals versus birds, readersare referred to (Hudspeth, 2008; Dallos, 2006)

I. CIRCUIT DEV


2.1.3 Scope of this Chapter

A functional circuit can be defined as the connectionbetween one or more cells or nuclei that transmit – andoften transform – a signal. As such, topics relating to theproliferation, delamination and migration of cells andthe development of their connecting projections in theauditory system are touched upon only briefly. Rather,this chapter is primarily concerned with describing therelative contributions of intrinsic molecular events,spontaneous action potentials and sound-evoked actionpotentials in the assembly of functional circuits withinthe peripheral and central auditory pathways.

This chapter is divided into three principal sections.The first section covers the ontogeny of local circuitswithin the cochlea and as well as the development ofafferent and efferent circuits connecting the cochlea tothe brain. The second section describes the establishmentof circuits within the developing auditory brainstem andthe influence of signaling from the auditory periphery.The final section addresses the formation of functional cir-cuits in the auditory midbrain and cortex. When appro-priate, we have cited notable seminal research papers,breakthrough findings and comprehensive review mate-rials so that the interested reader may avail themselves ofthese more focused sources of information.

2.2 DEVELOPMENT OF PERIPHERALCIRCUITS

2.2.1 Developing Networks Within the Cochlea

Cochlear hair cells initiate the process of hearing byconverting mechanical deflections of their stereociliabundles into electrochemical signals that are distributed

ELOPMENT


IHC

(a) (b)

IPC

IHCKo ISC

ATP

FIGURE 2.2 Transient microcircuits in the developing Organ of

Corti. (a) Schematic of the region surrounding the inner hair cell inthe pre-hearing rodent. Compared to the same region in the adultcochlea (Fig. 2.1(a)), the immature cochlea features a proliferation ofelongated inner supporting cells called Kolliker’s organ (Ko, in purple)and inner pillar cells (IPC). (b) Inner supporting cells in Ko (ISC, pur-ple) release ATP prior to hearing onset. ATP binds to purinergic recep-tors (black ellipses) to promote Ca2þ-dependent glutamate release fromribbon synapseswithin the inner hair cell (IHC, gray) and action poten-tials in auditory nerve fibers (green).



throughout the rest of the auditory system. Beforematureand normal transduction can occur, a number of criticaldevelopmental events take place between hair cells andnon-sensory cells within the cochlea. The specializedfunction of IHCs and OHCs depends in part upon devel-oping networks of non-sensory supporting cells withinthe organ of Corti and the lateral wall of the cochlea.

The precision of mechanoelectrical transduction canbe attributed, in part, to the unusual electrical potentialand ionic milieu in the endolymphatic space surround-ing the apical surface of the hair cell. Prior to and duringthe first week of hearing, an endocochlear potential isestablished between the endolymph and surroundingperilymph, which increases from 0 mV to þ80 mV. Theramping up of the electrical potential is complementedby the accumulation of high levels of Kþ in the endolym-phatic space, which further exaggerates the electricalgradient across the negative resting potential of the haircell membrane. The combination of high extracellular Kþ

and the positive endocochlear potential work synergisti-cally to effectively drive ionic currents through openmechanotransducer channels, creating the large andrapid receptor potential changes that mediate glutamaterelease at the synapse between the hair cell and the au-ditory nerve. The endocochlear potential is establishedthrough the development of tight cellular junctions be-tween local networks of epithelial cells, connective tissueand supporting cells that completely partition the endo-lymph from the surrounding perilymph. These tightlybound networks also efficiently recycle Kþ from the haircell back into the endolymphatic space where they canonce again be used in sensory transduction.

The spontaneous generation of action potentials fromsensory receptors is considered essential for normal neu-ral circuit development throughout the brain. In the de-veloping auditory system, the mechanisms responsiblefor spontaneous action potential activity are still unre-solved but recent reports suggest that this spontaneousactivity is generated by IHCs of the cochlea. The cartoonof the IHC region in the immature Organ of Corti repre-sents one proposed set of developmental changes that oc-cur in cochlear circuitry (Figure 2.1(a)). Compared toIHCs in mature animals, which are surrounded byone or two supporting cells (see Figure 2.1(a)), the pre-hearing Organ of Corti features a structure known asthe greater epithelial ridge, or Kollikers organ (Ko). Thisstructure consists of non-neuronal inner supporting cells(ISCs) that are present up to the onset of hearing. How-ever, by the time of hearing onset, Ko undergoes pro-grammed cell death and subsequent removal of themajority of ISCs. Despite this dramatic change in thestructure of the organ of Corti, recent studies have iden-tified a potential role for ISCs in the initiation of electricalsignaling within the auditory nerve (Tritsch, 2007). Oneto two weeks prior to the onset of hearing, the elongated

I. CIRCUIT DEV


ISCswithin Ko begin to spontaneously release adenosinetri-phosphate (ATP) into the extracellular space(Figure 2.2(b)). ATP activates purinergic receptors onneighboring IHCs, peripheral processes of the auditorynerve and on the ISCs themselves. Binding of ATP onthe IHC depolarizes the membrane potential, inducingCa2þ-dependent glutamate release and bursts of actionpotentials in auditory nerve fibers. ATP release is localand desynchronized along the length of the cochlea. Inthis manner, spatially and temporally independent vol-leys of electrical signals initiated by non-sensory neuronsentrain the firing patterns of SG and, ultimately, centralauditory neurons. This process is thought to play a rolein the strengthening of functional circuits prior to theonset of hearing.

Despite this role ofATP release fromKo, it remains un-certain how early action potential activity is patternedand whether ATP binding drives IHCmembrane voltageor providesweakermodulatory control.More recent datasuggests that during the first postnatalweek of life, devel-oping IHCs intrinsically generate the voltage changes thatelicit action potentials in SG neurons. The frequency andpattern of this spontaneous action potential activity variesbetween regions of the cochlea (i.e., high-frequency ver-sus low-frequency) and are modulated in multiple waysby the release of acetylcholine (ACh) and ATP near theIHCs (Johnson, 2011). It has been proposed that this pat-tern of action potential activity, along with ACh and ATPmodulation, could be important for guiding tonotopic or-ganization and the refinement of sensory informationalong the central auditory pathways before the occur-rence of experience-drive information becomes relevant.

2.2.2 Development of the Place Code

The basilar membrane acts as a spectral analyzer thattranslates vibration frequencies within the cochlear fluid

ELOPMENT


252.2 DEVELOPMENT OF PERIPHERAL CIRCUITS


pressure waves into positions of maximal displacementalong its length. In mature animals, the BM is relativelynarrow and taut at its base (violet region, Figure 2.1(b))compared to the apex, which is wider and more mobile(red region, Figure 2.1(b)). As previouslymentioned, thisstructural gradient confers a smooth shift of preferred vi-bration frequency along its length, with high frequenciesmaximally activating basal regions of the BM and lowfrequencies maximally activating apical areas. As withthe visual and somatosensory pathways, the spatial or-ganization of the receptor organ is maintained throughtopographic connections between the receptor epitheliaand successive levels of the central sensory pathways.In the auditory system, the one-dimensional tonotopicarrangement of preferred frequency along the lengthof the BM is preserved in tonotopic maps of preferredfrequency within the central auditory nuclei.

In nearly every respect, the development of basal(high frequency) regions of the cochlea occurs before api-cal (low frequency) regions; apical hair cells are the lastto differentiate and the last to be innervated by afferentand efferent nerve fibers that convey signals to and fromthe brain. Therefore, onewould predict that sensitivity tohigh frequency sounds would emerge before low fre-quencies in development. Interestingly, behavioral andneurophysiological hearing assessments in dozens ofavian and mammalian species show just the opposite:high frequency sensitivity is the last to mature. This de-velopmental mismatch implies that either tonotopic con-nections between the periphery and central auditorysystem are undergoing large-scale rewiring, or that de-velopmental changes within the cochlea cause a givenposition along the BM to vibrate at progressively higherfrequencies during the early period of hearing. Directneurophysiological recordings from first-order auditoryganglion neurons that innervate a fixed point within thebasal cochlea provide clear support for this latter sce-nario (Echteler, 1989). Basal cochlear regions were foundto be maximally sensitive to low frequencies at the onsetof the hearing and then become gradually responsive tohigher frequencies. This developmental shift in the co-chlear place code has been traced to a progressive mat-uration of OHC mechanics (Norton, 1991).

2.2.3 Development of Afferent and EfferentCircuits

The mature functional circuit linking the auditory pe-riphery to the brain has four essential processing sta-tions: 1) sensory hair cells in the auditory periphery, 2)first-order spiral ganglion cells (SG) that send a periph-eral dendrite to the IHCs andOHCs and a central axon tothe brain (i.e., the auditory nerve), 3) the cochlear nu-cleus (CN), a second-order auditory brain nucleus thatis heavily innervated by auditory nerve fibers, and 4)

I. CIRCUIT DEV


brainstem neurons whose efferent projections innervateIHCs, OHCs and neurons of the CN (Figure 2.3(a)). Thetwo types of sensory receptors in the mammaliancochlea (IHCs and OHCs) are each innervated by twotypes of SG neurons: Type I and Type II. Type I afferentscomprise approximately 95% of all SG neurons and eachcontact a single IHC. A single IHC, in turn, can be inner-vated by 10–20 Type I afferents, providing parallel andtopographically specific information transfer from a sin-gle IHC to the CN. By contrast, a single Type II SG neu-ron contacts 30–60 OHCs, providing a weaker, spatiallyintegrated signal to CN neurons. Although the OHCscontribute comparatively little afferent input to thebrain, they are the predominant targets of efferent axonsfrom the medial olivocochlear neurons (MOC) in thebrainstem. In mature animals, the central processes ofthe auditory nerve terminate in the CN. In some cases,one or two auditory nerve fibers contact a single CN neu-ron via amassive axosomatic synapse called the endbulbof Held. The process through which this circuit achievesits mature form reflects an interplay of molecular pro-cesses and intrinsic activity-dependent processes thatcan be broken down into three phases.

Phase 1: Well Before Hearing Onset

In rodents, the peripheral processes of SG neurons in-nervate basal regions of the cochlea approximately fivedays before birth, or approximately 17 days prior tohearing onset (Figure 2.3(b)). Within a day of growinginto the peripheral epithelia, afferents can be sorted intoType I or Type IImorphologies. It is generally agreed thatthere are no gross errors orwidely exuberant connectionsbetween SG dendrites and sensory hair cells. However,the exact precision of longitudinal (basal to apical) and ra-dial (IHCtoOHC) innervationpatternsbyType Iafferentsis not entirely understood. Genetic fate mapping studieshave described highly precise and rapid initial targeting(Koundakjian, 2007), while ultrastructural and histo-chemical studies find that Type I afferents initially inner-vatemultiple IHCsandOHCs(Echteler,1992).Theweightof evidence favors this latter characterization. ACh-releasing efferent fibers grow into the sensory epitheliaat the same time or slightly before afferent fibers. Com-pared to afferent innervation, however, efferent fibersshowacleardevelopmental shift in their spatial targeting,as olivocochlear efferent fiber innervation is initiallybiased towards IHCs rather than OHCs (Pujol, 1978).

The central projections of SG neurons reach their tar-gets in the CN approximately one day before their pe-ripheral dendrites innervate the sensory epithelia.Despite the fact that the cochlear nucleus is still formingat this stage, projections from the SG are remarkably pre-cise and demonstrate a clear spatial organization longbefore intrinsic or sensory-evoked action potential sig-naling begins. Although little is known about the

ELOPMENT


MatureWeeks beforehearing onset

Shortly afterhearing onset

Endbulb of Held

CN

SG

IHC

OH

C

Apex Base(a) (b)

(c) (d)

None

None

None

High

Low

None

High

Low

MO

C

Shortly beforehearing onset

FIGURE 2.3 Development of afferent and efferent circuits linking the periphery and the brain. (a) In mature mammals, Type I (ellipse) andType II (rectangle) spiral ganglion neurons (SG) extend a peripheral processes to a single row of inner hair cells (IHC) or three rows of outer haircells (OHC), respectively. The central projection of SG neurons form elaborate axosomatic synapses called endbulbs of Held on specific types ofneurons within the cochlear nucleus (CN, triangle). Efferent axons from medial olivocochlear (MOC) neurons innervate OHCs. Recording from aSG neuron innervating to the base of the cochlea would reveal occasional action potentials in the absence of sound or in response to low frequencysounds but elevated firing rates in response to high frequency sound (sound onset and offset represented by the trapezoidal stimulus). (b) Weeksbefore hearing onset, patterns of peripheral, central and efferent connections have been established, but lack the precise organization seen in ma-ture animals. (c) In the days leading up to hearing onset, Type I/ II andMOC connections are sorted into their correct hair cell targets and rhythmictrains of Ca2þ spikes initiated in IHCs entrain the firing patterns of Type I SG neurons. Endbulb synapses are beginning to form. (d) In the daysfollowing hearing onset, peripheral innervation patterns are largely mature and central projection are approaching the topographic specificity andsynaptic structure observed in mature animals. SG neurons innervating basal regions of the cochlea are responsive to low frequency tones, ratherthan high, due to immature basilar membrane and stereocilia mechanics.



molecules or cellular interactions participating in the es-tablishment of auditory nerve fiber topography, severalgrowth factors and receptors are expressed at approxi-mately the time that connections are being established.In particular, differential distribution of the Eph/ephrinfamily of receptor tyrosine kinases in the auditory nerveand CN suggests an axon guidance mechanisms thatshapes the formation of topographic connectivity withinthe auditory brainstem (Cramer, 2005).

Phase 2: Shortly Before Hearing Onset

Following the initial period of exuberant connectivity,the dendrites fromType I ganglion cells begin to coalescearound individual IHCs (Figure 2.3(c)). Conversely, ef-ferent MOC fibers are in a transitional state duringwhich immature connections to IHCs are maintainedalongside newly formed connections with OHCs. IHC

I. CIRCUIT DEV


mediated spontaneous spiking is robust during this pe-riod and electrophysiological recordings reveal tempo-rally patterned IHC Ca2þ spikes that are mimicked bythe firing patterns of SG neurons (Tritsch, 2010). In addi-tion, functional synapses between SG and CN neuronsare established, but the shape and overall size of end-bulbs in the CN are still quite immature.

Phase 3: Shortly After Hearing Onset

At this stage, peripheral innervation of the receptorepithelia is essentially mature with Type I and TypeII afferent endings making appropriate contacts withIHCs and OHCs, respectively. MOC efferents drop theirIHC connection and make nearly exclusive contact withOHCs. Rhythmic spontaneous bursting gives way to sto-chastic spontaneous action potentials intermingled withtemporall and topographically structured sound-evoked

ELOPMENT


272.3 DEVELOPMENT OF BRAINSTEM CIRCUITS


responses. However, recordings from SG neurons inner-vating basal IHCs reveal an immature preference for lowfrequency sounds, as described in the section on placecode development above. The topography from theSG neurons to the CN is initially very precise. However,more recent findings in mammals show that subtletopographic refinement continues around the onsetof hearing and continues for several months, suggest-ing a combination of intrinsic and extrinsic activity-dependent mechanisms (Leake, 2002). Additionally,endbulb synapses continue to undergo clear structuralmodifications. Though they do not yet reach the state ofa fully developed calyx that engulfs most of the CN cellbody in mature animals, the axon terminals increasein diameter by an order of magnitude relative to thepre-hearing period (Jhaveri, 1982; Ryugo, 1982).

2.2.4 Conclusions

In a matter of weeks, what began as an outpocketingof epithelial cells near the embryonic hindbrain becomesan exquisite functional circuit, capable of encoding me-chanical vibrations spanning three orders of magnitudein frequency, a 120dB dynamic range for amplitudeencoding (a million-million-fold change in signal en-ergy), and sensitivity to sub-atomic stereocilia displace-ments with microsecond mechanical response times.The physical attributes of sound are captured by atonotopically-organized array of sensory hair cells,which form topographic connections with neurons inthe CN to initiate the psychological experience of hear-ing. This complex circuitry arises through the interactionof sensory and non-sensory supporting cells within thedeveloping cochlea in addition to afferent and efferentconnections that link the cochlea to the brain. Other thana modest topographic refinement in the spatial distribu-tion of central projections to the CN and a fairly dramatictransformation of the endbulb synaptic terminal shapeand overall size, these circuits reach a mature form inde-pendent of sensory input. Instead,maturation appears toreflect the dominant influence of genetic programmingand molecular guidance cues with a subordinate rolefor cell-cell interactions thatmay include - but are not lim-ited to – internally generated action potential patterns.

In humans, the development of peripheral circuitry isparalleled by enormous changes in infant phonologicalperception. The earliest stages of perceptual refinementcan be attributed, at least in part, to the physical matura-tion of the outer ear, middle ear and cochlea. However,the scope of perceptual processes that continue to comeonline after peripheral circuits have matured as well astheir dependence upon normally patterned sensory in-put both point toward the essential role of central audi-tory circuits in the development of hearing.

I. CIRCUIT DEV


2.3 DEVELOPMENT OF BRAINSTEMCIRCUITS

2.3.1 Functional Circuit Assembly in theBrainstem

The assembly of functional circuits within the audi-tory system has been the subject of intense study overthe past 30 years. Auditory brainstem nuclei are derivedfrom progenitor cells within the hindbrain that migrateto their appropriate positions shortly before hair cellsin basal regions of the cochlea are born. Like the centralprojection of SG neurons to the CN, guidance of embry-onic brainstem neurons and assembly of their intercon-nections are thought to be mediated by the Eph/ephrinsignaling, although the detailed signaling pathwayshave yet to be defined (Cramer, 2005). Axonal connec-tions into second- and third-order nuclei are fullyformed several days before hearing onset in birds androdents. Direct electrical stimulation of the afferentaxons in pre-hearing animals reveal that these con-nections form functional synapses shortly after theyinnervate their target nuclei (Jackson, 1982). The tonoto-pically organized connection between the auditory spi-ral ganglion neurons and the CN is preserved inprojections to higher-order nuclei. That this topographyis initially present from the time connections are formedfurther supports the influence of activity-dependentmechanisms on circuit formation in the auditorysystem.

Functional circuits in the auditory brainstem of mam-mals (Figure 2.4(a)) and birds (Figure 2.4(b)) can be sepa-rated into three functional divisions: 1) an excitatoryprojection from the second-order auditory nucleus(green), 2) inhibitory inputs (red), and 3) third-ordernuclei that establish binaural sensitivity by integratingthese excitatory and inhibitory inputs (white).

In mammals, CN neurons receive afferent input fromthe SG and project ipsilaterally to the lateral superior ol-ive (LSO) and contralaterally to themedial nucleus of thetrapezoid body (MNTB). The medial superior olive(MSO) gets binaural excitatory inputs from both the ip-silateral and contralateral CN. The MNTB extends a gly-cineric (inhibitory) projection ipsilaterally to the MSOand LSO.

In birds, cochlear ganglion neurons extend a centralprocess into two second-order brainstem nuclei, nucleusmagnocellularis (NM) and nucleus angularis (NA). NMneurons make bilateral projections to nucleus laminaris(NL), a third-order nucleus analogous to the MSO inmammals. Whereas the mammalian MSO receives apowerful direct inhibitory input from MNTB, the inhib-itory brainstem nucleus in birds, SON, modulates theconvergent excitatory strengths of bilateral inputs toNA, NM, and NL.

ELOPMENT


MSO

MNTB

SON

BP

LSO

NM2

4

1 3

2

NL

CG

NA

CN

4

34

2

(a)

(b)

1 SG

FIGURE 2.4 Organization of mammalian and avian brainstem circuitry. Schematic coronal sections of the auditory brainstem nuclei in ro-dents (A) and chicken (B). Brainstem circuits feature central projections from the primary sensory ganglion neurons (1) to 2nd order auditory nucleiin the brainstem (2). 3rd order brainstem nuclei (4) integrate inputs from 2nd order nuclei and local inhibitory nuclei (3). Excitatory projections areshown in green, inhibitory in red, nuclei in grey. For mammals, SG¼spiral ganglion, CN¼cochlear nucleus, LSO¼ lateral superior olive,MSO¼medial superior olive,MNTB¼medial nucleus of the trapezoid body. For birds, BP¼basilar papilla, CG¼cochlear ganglion, NA¼nucleusangularis, NM¼nucleus magnocellularis, NL¼nucleus laminaris, SON¼superior olivary nucleus. For both schematics, the dorsal surface of thebrainstem is facing up.



2.3.2 Development of Fine-Scale Connectivity inthe MSO

The spatial position of a visual or tactile stimulus canbe encoded according to where, along the two-dimensional layout of the retina or skin, activity is great-est. In the auditory system, the functional layout of thecochlea is reduced to a single dimension mapped tosound frequency. The horizontal position of a soundsource in spacemust be computed centrally, by neuronssensitive to differences in the loudness or timing ofsounds arriving to each ear. Neurons in the MSOof mammals (NL in birds) are specialized to extract

I. CIRCUIT DEV


microsecond differences in the timing of excitatory in-puts from the CN (or NM) associated with each ear. Inmammals with low frequency hearing, such as gerbils,this precise tuning to interaural time differences (ITDs)appears to be enhanced by developmental changesin the subcellular positioning of inhibitory glycinergicinputs from the MNTB (Werthat, 2008). Glycinergicterminals from MNTB are evenly distributed acrossthe dendrites and cell bodies of MSO neurons roundthe time of hearing onset. In the days and weeks that fol-low, the less effective inhibitory synapses on the distalends of the dendritic tree are eliminated, sparing the

ELOPMENT




proximal axosomatic inhibitory synapses (Figure 2.5).Alterations at the subcellular level are paralleled bya progressive refinement in the overall topographicbreadth of presynaptic axon innervation across the topo-graphic map in MSO.. These anatomical changes areassociated with increased temporal precision and effi-cacy of synaptic inhibition onto MSO neurons as wellas sharpening of neural ITD tuning functions in theauditory brainstem, suggesting a direct link betweenphysiological maturation and its structural underpin-nings. Taken together, the developmental shifts in thetopographic and subcellular distribution of inhibitoryinputs to the MSO work synergistically to sharpen thespatial and temporal sensitivity of MSO neurons tothe excitatory CN inputs arriving from each side of thebrain.

Brainstem responseMSO cell

Dev

elop

men

t

Aroundhearing onsetWell afterhearing onset White

noise

Whitenoise

Immature,coarseITD tuning

Developed,properITD tuning

Undeveloped,coarseITD tuning

Unchanged,properITD tuning

FIGURE 2.5 Correlation between development of inhibitory in-

puts and ITD tuning. Before hearing onset, inhibitory glycinergic syn-apses (red dots) and excitatory glutamatergic synapse (green dots) onMSO neurons are distributed over soma and dendrites. After hearingonset, the glycinergic synaptic inputs get refined to the cell body. Thisrefinement depends on meaningful acoustic experience and can beinterrupted by the exposure of omnidirectional white noise duringhearing onset. White noise exposure at an adult stage has no effecton glycine receptor distribution. ITD tuning responses in the auditorybrainstem correlate with the development of inhibitory inputs to MSO.When animals were exposed to white noise between P10 and P25, ITDtuning responses recorded were found to have the same characteristicsas right after hearing onset, when neuronal responses to ITDs are im-mature and coarse. Under control conditions, i.e. with normal exposureto sound, ITD sensitivity developed normally. White noise exposure inadult animals has no effect on ITD tuning.

I. CIRCUIT DEV


2.3.3 Development of Fine-Scale Connectivity inthe LSO

LSO neurons are sensitive to interaural sound leveldifferences via tuning to the relative strength of excit-atory inputs from the ipsilateral ear and inhibitoryinputs from the contralateral ear. Unlike the MSO, con-tralateral inputs are only expressed indirectly, via a localinhibitory connection from the MNTB. The sharp binau-ral tuning of LSO neurons is thought to arise from theintegration of tonotopically matched excitatory inputsfrom the ipsilateral CN and MNTB. By contrast toMSO synapses, which are predominantly remodeled af-ter hearing onset, topographic plasticity in the LSO pro-ceeds in two phases: a functional silencing of MNTBinhibitory inputs prior to hearing onset followed by astructural pruning of MNTB axon terminals after the on-set of hearing.

The pre-hearing functional refinement phase takesplace contemporaneously with the period of ATP-induced Ca2þ spikes in IHCs. Accordingly, recordingsfrom MNTB neurons at this age reveal spontaneous ac-tion potential bursts with the same rhythmic patternsobserved in SG neurons (Tritsch, 2010). These activitypatterns also depend upon an intact connection withIHCs, whichmay indicate that intrinsic spontaneous “testpatterns” generated in the Organ of Corti prior to hearingonset may play a role in both peripheral and central cir-cuit formation. During this 3–4 day period, the spatialspread of MNTB-derived inhibition shrinks by approxi-mately 75%, considerably sharpening the breadth of inhi-bition along the tonotopic axis (Kim, 2003).

The onset of hearing ushers in a phase of structural re-finement. In the week following hearing onset, dendriticarbors of postsynaptic LSO neurons increase in branch-ing complexity yet are culled to more confined spacewithin the topographic map. These postsynaptic modifi-cations are accompanied by the physical elimination oftonotopically misaligned presynaptic axon terminalsfrom the MNTB. For a more detailed description of in-hibitory circuit development the reader is referred toChapter 131 of this volume by K. Kandler et. al.

2.3.4 Afferent Regulation of Cochlear NucleusDevelopment

These functional and structural changes in the brain-stem synaptic networks during the weeks surroundinghearing onset provide correlational evidence that spon-taneous and sensory-evoked events are needed toachieve the elegant organization of neurons and connec-tivity observed in mature animals. The strong test of thishypothesis has been carried out by dozens of studiesover the last sixty years that examine age-dependentdeviations from normative development following

ELOPMENT




disruption or complete elimination of afferent inputfrom the cochlea. The seminal study by Rita Levi-Montalcini (Levi-Montalcini, 1949) removed the otocyst(the precursor of hair cells and ganglion neurons) ofchicks at an early stage of embryonic development,thereby depriving auditory brainstem nuclei of cochlearinput. She noted that neurons in NA and NM developednormally until E11 (the approximate time of hearingonset in the chick), at which point the size and overallnumber of surviving neurons rapidly declined. This ob-servation suggested a dependence of CN neuron sur-vival upon an intact connection with the peripheryand has inspired a number of researchers to delve deeperinto the role of afferent signaling in the formation andmaintenance of brainstem circuitry.

Subsequent studies in the chick and rodents havefurther characterized the nature and timing of CN de-generation following deafferentation. Unilateral otocystremoval at E3 or cochlea removal prior to sexual matu-rity in chicks or cochlear destruction shortly after birthin rodents produces a variety of changes, including:1) the CN on the ablated side has significantly fewerneurons than the intact side and the surviving neuronshave smaller soma and neuropil area; 2) an ectopic –yet tonotopically aligned – projection from the normallyinnervated CN grows into the CN on the deafferentedside; and 3) in mammals, the endbulb of Held develop-ment described in Figure 2.3 never reaches the fully ma-ture state (reviewed in Harris, 2006).

In most animals, these plasticity effects are strictlylimited to a developmental critical period. As notedabove, the effects of otocyst removal in the E3 chickare not apparent for another eight days, when activeconnections with the periphery are first established.The critical period for CN neuron survival ends asabruptly as it begins, as cochlear destruction in gerbilsduring the first postnatal week causes 45-90% of CNcells to die, yet has no effect on the number of survivingneurons when the same manipulation is performed atP9, just prior to the onset of hearing (Tierney, 1997).The remarkably rapid changes in susceptibility ofCN neurons to deprivation appears to reflect a differen-tial weighting of factors that promote versus inhibitcell death.

Collectively, studies of CN development demonstratethat the proliferation, migration and formation of appro-priate topographic connections are complete before ac-tion potentials begin to appear in auditory nervefibers. The arrival of normal spontaneous and sound-evoked afferent action potentials has little effect on cel-lular morphology or topographic specificity. However,pathological deviations from the normal developmentaltrajectory (e.g., cochlear removal) that occur within adefined critical period window radically alter CN orga-nization, leading to pronounced cell death and atrophyof surviving cells. The cellular mechanisms that close

I. CIRCUIT DEV


the critical period of CN vulnerability are not yet fullyestablished, but gene array analyses suggest that glialproliferation and up-regulation of immunity-relatedgenes may play an important role.

In addition to deafferentation-induced cell death andatrophy, hearing loss also induces a long-term enhance-ment of neuronal excitability. In vitro measurements ofneurons in acute slices of the CN demonstrate a shift to-wards enhanced excitation and diminished inhibition.The loss of balanced excitation and inhibition arises froman abnormal sorting of membrane-bound ligand-gatedneurotransmitter receptors and voltage-gated ion chan-nels. Ongoing research is exploring the hypothesis thatthis pathological over-excitability in the brainstem andother stations of the central auditory pathways may bethe source of tinnitus, the perception of phantom soundsthat can accompany hearing loss.

2.3.5 Afferent Regulation of 3rd-OrderBrainstem Nuclei

All auditory-evoked signals in the brain are initiallyrouted through the CN. Therefore, alterations in CNmorphology resulting from cochlear ablation could alsoimpact the organization of downstream nuclei. Indeed,the post-hearing structural refinement of axon terminalsfrom the MNTB to the MSO and LSO described previ-ously is substantially diminished without a connectionto an intact cochlea. Moreover, unilateral cochlea re-moval before hearing onset enhanced an otherwise weakphysical connection between the normally innervatedCN and the opposing side of the MSO.

Unlike the inhibitory projections from the MNTB tothe LSO and MSO, excitatory projections from the CNundergo considerably less developmental refinementand are largely insensitive to cochlear removal. TheCN, for example, forms a giant excitatory synapse ontoMNTB neurons called the calyx of Held. This calyx isthe largest synapse in the mammalian brain and featuresglutamatergic CN terminals that almost completelyenvelop the MNTB neuron, ensuring high-fidelity trans-mission necessary for sharp interaural time- and level-dependent tuning observed in the MSO and LSO,respectively. In mature brains, one calyx innervates asingle MNTB neuron. Although the calyx undergoessubstantial changes in shape during the first weeks ofpostnatal development, the one-to-one connectivity ispresent from the time CN projections initially arrive inthe MNTB and is established with or without spontane-ous or evoked action potential from the auditory nerve(Hoffpauir, 2006).

The effects of deafferentation have been extensivelystudied in NL of the chicken. The intrinsic organizationand connectivity patterns within NL lend themselves to

ELOPMENT




studies of afferent regulation of individual dendrites.First, NL exhibits a clear gradient of dendritic geometrythat varies from small, short and stubby to large, longand elaborate across the high-to-low tonotopicmap. Thisgradient appears to begin to form around the time ofhearing onset. Second, NL neurons have two sets of sym-metrical dendrites, one set oriented dorsally to contactglutamatergic inputs from the ipsilateral NM, the otherventrally to contact glutamatergic inputs from the con-tralateral NM. This organization permits researchers todirectly compare the effects of unilateral manipulations(which would affect the dorsal dendrites on the ipsilat-eral NL and the ventral dendrites on the contralateralNL) to transection of the NM axons at midline (whichwould affect the ventral dendrites on both sides of thebrain).

Using both of these approaches, researchers have dis-covered an interesting dichotomy in the afferent regula-tion of intrinsic features versus functional circuitproperties. Removal of synaptic input to one side ofNL induces a progressive retraction of dendrites onthe corresponding side on a surprisingly short timescale(Figure 2.6). Tracking these changes over time revealsthat retraction begins within 1 hour, can last over 2weeks, and can amount to as much as a 60% reductionin length, demonstrating that NL neurons can rapidlyregulate significant amounts of membrane surface de-voted to specific excitatory inputs (Deitch, 1984). By con-trast to the rapid calibration of dendritic length allocatedto the dorsal versus ventral NM inputs, the short-to-longgradient of relative dendrite length across the tonotopicaxis of NL is largely unaffected by cochlear removal.

Normalbalanced activity

Ventral inputremoved

Dorsal inputstimulated

FIGURE 2.6 Compartment-specific regulation of afferent activity

on dendritic structure in the chick nucleus laminaris (NL). NL neu-rons have bipolar dorsal and ventral dendrites (black), receiving highlysegregated excitatory inputs by an axon from either the ipsilateral orcontralateral ear via the nucleus magnocellularis (NM). (Left), Undernormal physiological conditions, the two dendritic domains receivebalanced inputs (green), each elicited by one ear, and their dendritesare of similar length. (Middle), Following the deafferentation of the in-puts to one domain, e.g. when axons from one side of the brain get cut,deprived dendritic branches rapidly retract, while the length of theother dendrite remains unchanged. (Right) Physiological stimulationof the axon inputs to one set of dendrites (red lightning bolt), but notthe other, leads to a growth of the stimulated dendritic branches anda retraction of unstimulated dendritic domain.

I. CIRCUIT DEV


2.3.6 Influence of the Source and Patternof Afferent Activity on Brainstem Circuits

The effects of cochlear removal on brainstem circuitspoint towards a panoply of developmental events thatdepend upon afferent signals from the periphery.Cochlear removal eliminates both spontaneous andsound-evoked action potentials in addition to the phys-ical degeneration of SG neurons. Because the depen-dence on afferent activity is most commonly observedafter hearing onset, one might assume that sound-evoked activity provides important signals for the fine-tuning of brainstem circuits. An alternative explanationholds that any afferent action potential signaling, be itspontaneous or evoked, could be sufficient for normalassembly of brainstem circuits.

To isolate the relative contributions of spontaneousaction potentials versus sound-evoked activity, re-searchers have compared the effect of pharmacologicallysilencing all afferent action potentials versus simplyblocking the transduction of acoustic signals. The effectsof tetrodotoxin infusion into the inner ear at the timeelectrical signaling first begins were indistinguishablefrom the effects of cochlear removal: approximately40% of the neurons died with widespread neuronal atro-phy in the survivors (Born, 1988). On the other hand,simply blocking sound-evoked activity (by disruptingthe sound transmission mechanisms of the outer or mid-dle ear) without eliminating spontaneous activity in au-ditory nerve did not cause atrophy or cell death in theCN (Tucci, 1985). These results suggest that action poten-tials, or more probably the voltage-gated changes in glu-tamate and calcium signaling, provide a necessarysource of trophic support during the critical period ofCN development. Although the particular patterningof sound-evoked action potentials was not necessaryfor the normal cellular maturation of this second-ordernucleus, it had a significant impact on stimulus selectiv-ity and circuit formation in downstream nuclei. Forinstance, rearing animals in omnidirectional noise inter-feres with low frequency signals necessary to calculateinteraural time differences. Absent these instructiveenvironmental signals, the MSO fails to develop at a nor-mal pace and features widely branching axodendriticMNTB synapses rather than the topographically focusedaxosomatic synapses found in normally reared animals(Seidl, 2005); Figure 2.5). Similarly, interaural leveldifference tuning in brainstem neurons is significantlyaltered when owls are reared with an earplug thatdeprives them of normally calibrated binaural cues(Mogdans, 1994).

2.3.7 Conclusions

In summary, functional circuit development and re-finement in the auditory brainstem is thought to reflect

ELOPMENT


Actx

5 IC

MG

Rt

6

7

FIGURE 2.7 Organization of higher auditory circuits. Schematicof a horizontal section through a rodent brain. Convergent bilateral in-put from the brainstem (white arrows) projects to the inferior colliculus(IC), an auditorymidbrain structure (5). The IC connects to the auditorythalamus (6) and cortex (7) through a complex chain of feedforward,feedback and intrinsic excitatory (green) and inhibitory (red) connec-tions. MG¼medial geniculate body, Rt¼ reticular nucleus of the thal-amus, Actx¼auditory cortex. Top of the figure is rostral (i.e., anterior),bottom is caudal (i.e., posterior).



an interplay betweenmolecular cues coupledwith spon-taneous and evoked action potential activity. Factorssuch as connectional topography in the LSO and spatialgradients of dendritic length in NL appear to be gov-erned by intrinsic, activity-independent mechanismsand form prior to the onset of hearing. Cochlear actionpotentials before and after the onset of hearing predom-inantly regulate the size and subcellular positioning ofsynaptic contacts. The development of precise connec-tivity is less categorical: the proper development of axo-nal projections from the MNTB to MSO and LSO isdependent upon afferent signaling after hearing onset,the topographic specificity of SG connections into theCN are subtly modified by afferent signaling, while ex-citatory projections from the CN to their ipsilateral andcontralateral brainstem targets develop independentlyof cochlear signaling (although they make aberrant con-nections should one cochlea be removed).

Compared to peripheral circuits, which havematuredinmany respects by hearing onset, brainstem circuits un-dergo additional refinement after sound-evoked activityis introduced to the system. These phenomena are likelyto contribute to the progressive improvement in infantauditory perceptual acuity in humans and animals. Forexample, human infants become increasingly capableof resolving the fine positioning of sounds along a hori-zontal plane, which may reflect the developmental cali-bration of synaptic properties within the MSO and LSO.Other aspects of phonological development, includingthe neural specializations that underlie the acquisitionof language, can only be understood by examining thedevelopment of functional circuits in higher levels ofthe central auditory system.

2.4 DEVELOPMENT OF AUDITORYMIDBRAIN AND FOREBRAIN CIRCUITS

Higher auditory circuits are assembled from a diverseset of excitatory and inhibitory connections arising fromthree critical sources: 1) the inferior colliculus (IC), amidbrain auditory brain structure, 2) the medial genicu-late (MG) and reticular (Rt) divisions of the thalamus,and 3) the auditory cortex (Actx) (Figure 2.7). These nu-clei are interconnected through a complex array of feed-forward, feedback and intrinsic connections. Inputsfrom the two ears are heavily intermixed, given the feed-forward binaural inputs from the MSO and LSO as wellas interhemispheric connections between each IC andActx. Both the IC and MG are obligatory relays for audi-tory signals reaching the Actx and, consistent with brain-stem circuits, a tonotopic organization within theseconnections is evident 5-7 days before birth in rodents,long before the initiation of afferent signaling from theperiphery.

I. CIRCUIT DEV


2.4.1 Development of Thalamocortical SubplateCircuitry

Compared to the wealth of studies in NL, LSO andMSO, functional circuit developments in higher auditorybrain areas have yet to be characterized in detail at thecellular level. One emerging exception is the observationof a transient microcircuit linking the neonatal thalamusand cortex. Cortical neurons originate from the ventric-ular zone and reach their final positions within the six-layered cortical plate via local radial migration or, inthe case of several classes of inhibitory interneurons,via long-distance horizontal migration along the rostralstream (additional information on patterning and neuralmigration within the cerebral cortex can be found inRubenstein and Rakic, 2013). Subplate neurons (SPN) re-side in the white matter beneath the cortical plate, phys-ically interposed between the ascending MG axons andthe incipient circuits forming in the more superficiallayers of the cerebral cortex (for a comprehensive reviewof subplate neurons, see (Kanold, 2010)).

SPN development occurs at an accelerated pace com-pared to the other cortical neurons; SPNs are among the

ELOPMENT


332.4 DEVELOPMENT OF AUDITORY MIDBRAIN AND FOREBRAIN CIRCUITS


first neurons to appear in the cerebral cortex, the firstcortical neurons to fire action potentials, and are almostcompletely eliminated around the time of hearing onset.SPNs have elaborate dendritic trees that integrate excit-atory inputs from the MG and local inhibitory inputs.SPN axon terminals ramify extensively in the samelayers of the cerebral cortex that will receive the bulkof direct axonal input from the MG in subsequent stagesof development. SPNs have high input resistances, rela-tively depolarized resting membrane potentials and areelectrically coupled to one another via gap junctions,making them a sensitive and potent source of excitatoryinput to Actx prior to hearing onset.

Their precocious morphological and biophysical de-velopment combined with their physical position be-tween MG and Actx make SPNs ideal interlocutors inthe postnatal assembly of thalamocortical circuits. In-deed, as represented in Figure 2.8, MG axons innervatethe subplate days before birth and wait there for daysor weeks (depending on the species) before innervatingthe middle cortical layers. During this waiting period,SPNs make excitatory glutamatergic projections to excit-atory and inhibitory neuron subtypes within Actx,potentially providing a source of activity-dependentsynaptic refinement before connections are establishedwith the MG.

2.4.2 Postnatal Development of LocalCortical Circuits

Cortical circuits undergo a subsequent wave of refine-ment during the second postnatal week, after the

Shortly after birth

Act

xS

ubp

late

MG

Days before

FIGURE 2.8 Subplate neurons form a transient microcircuit linking

mediate excitatory signaling from the medial geniculate nucleus of the thcells in the auditory cortex (Actx, triangle). Shortly after birth (left), SPNIn the days hearing onset (middle), SPNs mediate direct excitatory projectellipses¼glutamate receptor, red ellipses¼GABA receptor, black ellipses

I. CIRCUIT DEV


majority of SPNs have been eliminated. At the onsetof hearing, sound-evoked responses in Actx are re-stricted to frequencies at the center of the hearing rangepresented at high sound levels. Over the following 2–3days, response thresholds decrease and sensitivity to abroader range of sound frequencies begins to emerge.Following this brief period of change, which almost cer-tainly reflects peripheral maturation, recordings fromActx neurons reveal a gradual improvement in theirability to synchronize action potential timing to rapidmodulations of an incoming sound source. This im-provement in cortical temporal processing is a hallmarkfeature of central auditory development and thought tobe essential for the accurate encoding of important envi-ronmental sounds such as speech.

Through targeted recordings of excitatory and inhib-itory neuron subtypes in Actx using an in vitro brainslice preparation, researchers have discovered that theprogressive elimination of temporal “sluggishness” to-wards the end of the second postnatal week is associatedwith a confluence of synaptic and intrinsic changes inActx neurons. In addition to afferent input from MG,acetylcholine (ACh)-positive axon terminals from neu-rons located in the basal forebrain begin innervatingthe Actx in postnatal week 1 and reach adult levels bypostnatal week 2. The maturation of cholinergic termi-nals coincides with significant changes in the levels orcomposition of nicotinic ACh receptors and NMDA re-ceptors (nAChRs and NMDARs, respectively) in Actx(Metherate, 2003). Although ACh does not excite Actxneurons directly, it can enhance excitatory transmissionindirectly by binding to nAChRs, which in turn modu-late glutamatergic NMDARs. The decline of nAChRs

hearing onset Hearing onset

the auditory thalamus and cortex. Subplate neurons (SPNs, squares)alamus (MG, diamond) to excitatory (triangle) and inhibitory (circle)s are believed to be the exclusive source of subcortical input to Actx.ions to Actx before they are eliminated, shortly thereafter (right). Green¼electrical synapse.

ELOPMENT




in Actx by postnatal week 3 works synergistically withchanges in AMPA and NMDA glutamate receptor sub-unit composition to reduce the duration and increasethe amplitude of excitatory postsynaptic currents inActx (Figure 2.7(a)). Thus, the transient appearance ofnAChRs during postnatal week 2 represents a critical pe-riod during which ACh can prolong the time course ofNMDA-dependent synaptic excitation around the pe-riod of hearing onset.

The sharpening of synaptic excitation in Actx aroundthe time of hearing onset is mirrored by a progressive en-hancement of synaptic inhibition. Presynaptically, thevoltage-gated Kþ channel subtypes present in GABAer-gic interneurons change significantly over the first threeweeks of postnatal development to adjust the restingmembrane potential and shorten the refractory periodfollowing an action potential. These biophysical changesare complemented by the progressive loss of GABAB re-ceptors on presynaptic interneurons, enabling highersustained firing rates without fatiguing. Postsynapti-cally, the GABAA receptor subunit composition in excit-atory Actx neurons changes over the first weeks ofdevelopment to eliminate the a3 subunit in favor ofthe a1 and b2/3, both of which are associated with fasterinhibitory synaptic current rise times (Figure 2.9(a)).Collectively, these intrinsic and synaptic changes inthe time course of excitatory and inhibitory synaptic

Pre-hearing

Exc

itatio

nIn

hib

ition

1.5 months 3 months A

Early hearing Adult

D

(a)

(d) (e)V

RC

FIGURE 2.9 Afferent regulation of higher auditory circuits. (a-c) Blaaptic currents recorded from brain slices containing auditory cortex (a, c) oand stronger synaptic currents over development based on changes in intricochlear ablation, wherein neurons are hyperexcitable and display immatu(d) Lateral surface of the cat brain. Gray denotes total area of auditory coiological signals are measured in panels e and f. D¼dorsal, R¼ rostral, V¼plotted upwards, and outward positive IPSCs are plotted downwards, bevoked from activation of a stimulating electrode implanted proximal to(f) cats. Current traces in (b) and (c) are adapted from data presented in

I. CIRCUIT DEV


transmission endows Actx neurons with an improvedability to track rapid temporal fluctuations in sound sig-nals with high fidelity (for review see Sanes, 2009).

2.4.3 Afferent Regulation of Higher AuditoryCircuit Development

Detailed characterizations of auditory brainstem cir-cuit development in the absence of cochlear signalinghave revealed a combination of activity-dependent andactivity-independent processes. To assess the generalityof these principles throughout the central auditorypathways, researchers have also examined the role ofan intact periphery on the maturation of IC and Actxcircuits. The potentially confounding influence of CNdegeneration has been avoided in this type of experi-ment by bilaterally removing the cochleae after the crit-ical period for CN cell death, but before the onset ofhearing, such that higher auditory circuits have the po-tential to be shaped by spontaneous – but not sound-evoked – action potentials.

Recordings made in the acute brain slice preparationweeks after cochlear ablation reveal dramatic alterationsin the strength and time course of excitatory and inhib-itory synaptic currents in the IC (Figure 2.9(b)) andActx (Figure 2.9(c)). Compared to control slices taken

dult 1.5 months 3 months Adult

Control 400 ms

50 ms 20 ms

50 pA

Max

Min

(b) (c)

(f)

5 pA

20 ms

-20 pA -5 pA

Coch. Ablation

Cor

tical

res

pon

se

ck lines represent excitatory (upward) and inhibitory (downward) syn-r inferior colliculus (b) neurons. Auditory cortex neurons display fasternsic and synaptic properties. This process is arrested following bilateralre temporal dynamics. Current amplitude is plotted on the vertical axis.rtex in normal cats. Black square denotes area from which neurophys-ventral C¼caudal. Horizontal bar¼1 cm. Inward negative EPSCs are

y convention. (e-f) Areal extent of neural responses in auditory cortexthe auditory nerve fibers in acutely deafened (e) or congenitally deafSanes and Bao, 2009.

ELOPMENT




from normally hearing animals, neurons in cochlear-ablated slices areas show weaker, prolonged inhibitionthat is qualitatively similar to activity found in normalanimal prior to hearing onset. Synaptic excitation is alsoprolonged in a similar fashion to pre-hearing animals,yet the amplitudes are far greater than those observedin the course of normal development, suggesting thatdeafferentation tips the homeostatic balance between ex-citation and inhibition towards greatly enhanced excit-ability, in a similar fashion to the CN (Kotak, 2005).

Recent studies have identified a combination of intrin-sic and synaptic factors that may explain the failureof synaptic inhibition to mature normally. In terms ofintrinsic biophysical mechanisms, the first postnatalweeks are marked by a substitution of voltage-gated Kþ

channels that mediate faster membrane kinetics as wellas the appearance of a Kþ-dependent intracellular chlo-ride transporter, KCC2. During the first week of postnataldevelopment, when KCC2 expression levels are low,intracellular chloride concentrations are higher thanthe electrochemical equilibrium potential and GABA re-lease from inhibitory neurons induces a depolarizationof the membrane potential in postsynaptic neurons,rather than the expected hyperpolarization. As levelsof membrane-bound KCC2 increase in the second weekof postnatal development, greater amounts of intracel-lular chloride are extruded into the extracellular space,thereby lowering the equilibrium potential and estab-lishing the normal hyperpolarizing influence of GABA.This developmental process is arrested in the IC ofcochlea-ablated animals, where the inhibitory reversalpotential can be elevated by as much as 24 mV abovenormally hearing age-matched controls, reducing thehyperpolarizing effect of GABA binding. Althoughoverall expression levels of KCC2 are not affected,pharmacological experiments reveal that cochlear abla-tion arrests the normal age-dependent maturation ofKCC2 function, contributing to reduced levels of inhib-itory signaling (Vale, 2003).

A synaptic piece of the puzzle was discoveredthrough a comparison of GABAA receptor subunit com-position in the Actx of normal and deafferented animals.Recall from the description above that GABAA receptorscomposed of the a1 and b2/3 subunits appear during thesecond postnatal week and are partially responsible forthe transition to shorter, larger-amplitude inhibitorysynaptic currents. Following bilateral cochlear ablation,juvenile Actx neurons fail to express the mature form ofthe membrane-bound GABAA receptor, thereby pre-venting the sharpening of the inhibitory postsynapticcurrents observed in age-matched controls (Kotak,2008). Taken together, these results demonstrate a broadspectrum of intrinsic and synaptic events that fail to de-velop in the absence of sound evoked-activity. Thesum total of these events render IC and Actx neurons

I. CIRCUIT DEV


incapable of tracking rapid fluctuations in sound proper-ties, an essential characteristic of normal hearing.

2.4.4 Developmental Regulation overReinstating Hearing in the Deaf

Synaptic transmission studies in the brain slice prep-aration shed some light on the molecular targets of affer-ent signaling and help to identify the complications andpossibilities associated with reinstating hearing in deafindividuals. Unlike birds, and other non-mammalianvertebrates, which can regrow hair cells throughout life,mammals are born with all the cochlear hair cells theywill ever have. Nevertheless, hearing is a possibilityfor profoundly deaf individuals through the use of thecochlear implant, a neural prosthetic device that by-passes the dysfunctional transduction machinery withinthe cochlea and reinstates afferent signals through directelectrical stimulation of auditory nerve fibers. Approxi-mately 200,000 individuals have been fitted with co-chlear implants over the past 40 years and it wasdiscovered early on that the age of surgical implantationplays a crucial role in the quality of hearing experiencedby cochlear implant users. While post-lingually deaf in-dividuals often recover acceptable hearing and speechrecognition whether they are implanted as children oradults, congenitally deaf individuals stand the bestchance of experiencing the full benefit of the cochlear im-plant if they undergo the implantation procedure at anearly age, typically by the time they are 7 years old(Dorman, 2007).

Through careful study of cochlear implants in a spe-cial breed of congenitally deaf cats, researchers have be-gun to understand how auditory brain areas representsignals delivered through the cochlear implant and themanner by which these representations are shapedthrough development and experience. As an experimen-tal control, normally hearing cats are acutely deafenedwith an ototoxic drug and immediately fit with a co-chlear implant. Neural recordings are made from theActx of acutely or congenitally deaf cats at various agesin response to brief electrical pulses delivered the audi-tory nerve (Figure 2.9(d)). A comparison of activationpatterns across development in acutely deafened cats re-veals an exuberant spatial spread of neural activityacross the Actx in young kittens that is culled to a topo-graphically restricted activation area by 3 months of age(Figure 2.9(e)). By contrast, activating the implant incongenitally deaf kittens at 1.5 months evokes a weakcortical response (Figure 2.9(f)). At 3 months postnatal,deaf cats show the exuberant activation patterns compa-rable to normally hearing kittens at 1.5 months and theseactivation areas are not consolidated until early adult-hood (Kral, 2005). The hyper-excitability of the Actx incongenitally deaf cats at three months may stem from

ELOPMENT




the diminished synaptic inhibition and augmented syn-aptic excitation observed in brain slices of rodents thatundergo cochlear ablation in infancy (Figure 2.9(c)).

Although the mechanisms governing the recovery ofhearing in cochlear implant users remain unclear, onepossible clue has been found through analyzing the end-bulb of Held synapse in deaf cats that began hearingthrough chronic use of the cochlear implant at an youngage. As described previously, the endbulb synapse failsto develop normally in the absence of cochlear signaling.Strikingly, the endbulb synapse from cats using the co-chlear implant for 3 months beginning at an early agewas largely indistinguishable from normally hearingcats (Ryugo, 2005). Therefore, reinstating afferent activ-ity to the central auditory system in early life can rescuethe progressive synaptic degradation observed at sev-eral levels of the central auditory system.

2.4.5 Experience-Dependent Influenceson Functional Circuit Development

For the most part, the influence of afferent signalingon neural circuit development has been described inthe context of all-or-nothing manipulations; compari-sons to normally hearing animals are made through co-chlea removal, genetic deafness, or pharmacologicalsilencing of the auditory nerve. However, many facetsof auditory perceptual development depend upon thespecific patterns of auditory experience rather than itspresence or absence. For example, before a child uttersa first word in its native language, it will have been ex-posed to hundreds of thousands of words that bear thephonemic structure specific to that language. One schoolof developmental cognitive psychology has argued thatthis repeated auditory exposure to speech sounds, whenappropriately timed during childhood development,“primes” the auditory and vocal motor areas of the brainto specialize in the nuances of a given language. A halfcentury of inspired neurobiology and neuroethology re-search has shown that the processes through whichsongbirds learn their vocal repertoire offers a strikingparallel to human language acquisition. Song and speechlearning both involve a complex interplay between in-nate predispositions and experience, both forms of learn-ing are shaped by developmental critical periods, bothrequire skilled control over the motor vocal apparatus,and both depend upon precisely calibrated auditoryfeedback (for a review of song learning, the reader is re-ferred to (Brainard, 2002).

Although the circuitry underlying song acquisition ismore closely linked to sensorimotor integration than au-ditory processing per se, auditory experience has alsobeen found to exert a profound influence over the devel-opment of dedicated auditory circuitry within IC and

I. CIRCUIT DEV


Actx. In rodents, a tonotopic map of preferred sound fre-quency tuning can be readily delineated from the pri-mary field of the auditory cortex (AI), which is visiblealong the lateral surface of the brain (Figure 2.10(a)). Inmature animals, the tonotopic map features an orderlylow-to-high gradient of frequency tuning across theposterior-to-anterior extent of AI, with similar tuningorganized into orthogonal iso-frequency contours(Figure 2.10(b)). Most recording sites feature closelymatched frequency tuning for tones presented to eachear, albeit tuning evoked from contralateral stimuli ismore robust than the weaker, high-threshold tuningfor tones presented to the ipsilateral ear. By contrastto the adult map, AI recordings made shortly afterthe onset of hearing (P11-14) reveal narrowly tuned,high threshold responses restricted to the middle fre-quency regions of the tonotopic map (Figure 2.10(c)).

A series of studies over the past decade demonstratethat experimentally modifying the ambient acoustic en-vironment experienced by young animals during criticalperiods of higher auditory circuit development candramatically alter the normal developmental sequencedescribed in Figures 2.10(b) and 2.10(c). For example,rearing litters of rodents in a sound environment domi-nated by the repeated presentation of a single tone at afixed frequency more than doubles the area ofthe tonotopic map tuned to that exposure frequency(Figure 2.10(d)). As illustrated in Figure 2.10(d), expo-sure to a repetitive middle frequency tone for just 96hours beginning at the onset of hearing is associatedwith an expansion of the corresponding area withinthe tonotopic map at the expense of flanking frequencyrepresentations. The tonotopic map retains this distortedorganization into adulthood despite the fact that animalshear normal acoustic stimuli from P16 onwards. Con-versely, rearing animals older than P16 to the repeatedsingle frequency has no effect on the map regardless ofthe length of exposure time (de Villers-Sidani, 2007). Sim-ilar effects of tone rearing have been reported in the IC.

Whereas rearing animals in fixed frequency toneenvironments induces a region-specific reallocation ofterritory within the cortical map, rearing animals inbroad-spectrum noise is associated with a degenerationof Actx tonotopy and an abundance of recording siteswith abnormally broad tuning (Figure 2.10(e)). Similarto the extended visual cortex critical periods observedin animals reared in total darkness, the critical periodfor the effects of single tone exposure is abnormallyprolonged in animals reared in white noise. Thus, depri-vation of patterned acoustic input postpones the flipof a molecular switch that normally limits the criticalperiod, and the effects of singe tone exposure can distortthe map many weeks after the critical period wouldnormally have closed (Chang, 2003). That a critical pe-riod normally observed in infancy can be delayed

ELOPMENT


Low

V

R

D

C

Mid High

Max Loud

LowTone frequency

High Contralateralear

Contralateralear

Ipsilateralear

Ipsilateralear

SoftMin

Tone

-evo

ked

firin

g ra

te

Preferred

(a) (b) (c)

(f)(e)(d)

frequency tuning

FIGURE 2.10 Experience-dependent map reorganization in the developing auditory cortex. (a) Photograph of the lateral surface of the rat-brain. Color gradient denotes the position and orientation of the tonotopic mapwithin the primary auditory cortex (AI). Horizontal line¼1 cm. (b-f) Schematics of tonotopic maps reconstructed from microelectrode recordings of tone-evoked action potentials at approximately 100 differentpoints within AI. At each recording site, the frequency response area (grayscale surface plots) is determined by presenting tones at various fre-quencies and levels and identifying the frequency to which the neuron is tuned at its threshold intensity (indicated by the color of each dot). (B)Normal adult animals display a smooth and orderly tonotopic gradient in which neurons at the caudal and rostral boundaries of AI prefer low andhigh frequency tones, respectively. Neurons are tuned to the same frequency for sounds presented to each ear, yet responses to the contralateral earare more robust than the ipsilateral ear. (c) Tonotopy at the onset of hearing is narrow, high-threshold and restricted to low frequency areas of themap. Open circles denote recordings sites with no evoked response. (d) Rearing rodents in acoustic environments dominated by presentation of asinglemiddle frequency tone (represented by green colors and vertical arrows on surface plots) show an over-representation of that tone frequencywithin the tonotopic map such that recording sites at distant points within the map have similar tuning. (e) Rearing rodents in continous whitenoise is associated with a disrupted tonotopic map featuring numerous recording sites with abnormally broad frequency tuning (represented bygray dots). (f) Reversibly closing the contralateral ear during critical periods of development is associated with an over-representation of low fre-quency tones, the loss of contralateral tuning at several points (open circles) and an enhancement of tuning strength to tones delivered to thedevelopmentally unobstructed ipsilateral ear.



into adulthood supports the possibility that progressivedevelopment of auditory feature representation in Actxmay be more closely linked to an experiential timeline,rather than a strict chronological timeline.

A parallel can also be drawn between the reversiblelid suture method used extensively in studies of visualcortex development and the effects of reversible ear ca-nal ligation on Actx and IC development. Ligating theear canal temporarily interferes with the transmissionof acoustic signals to the middle ear, and ultimatelythe brain, particularly at high frequencies. By reopeningthe ear canal prior to recording from the contralateral AI,researchers have observed a degraded tonotopic mappopulated by weak high-threshold responses to contra-lateral tones and the occasional absence of contralateraltuning in higher frequency regions of the map (Popescu,2010). By contrast to the contralateral bias observed innormal animals (Figure 2.(10)b), the quality of tuningfor stimuli delivered to the ipsilateral ear is often superiorto the normally dominant contralateral ear (Figure 2.10(f)). Interestingly, ipsilateral inputs were only enhanced

I. CIRCUIT DEV


when the ear canal was ligated in infant and juvenileanimals, but not in adulthood. Collectively, these resultsindicate that the allocation of representational resourceswithin theActx and IC is quite dynamic and can be adap-tively reassigned to the inputs that are most prominentduring critical periods of early postnatal development.

2.4.6 Conclusions and Directions for FutureResearch

Compared to the periphery and brainstem, the assem-bly of circuits in higher auditory brain areas appears tobe particularly sensitive to sound-evoked patterns of af-ferent action potentials. Eliminating hearing or even per-turbing the normal balance of signals between the ears oracross various frequency channels can dramatically alterauditory signal processing. This plasticity has beendocumented at multiple levels of analysis, ranging fromthe synaptic transmission between two neurons up to thecoordinated arrangement of frequency tuning across

ELOPMENT




hundreds of thousands of neurons. Although circuit re-organization at higher levels of the auditory system isstriking, additional information on the normative courseof development in the IC, MG and Actx as well as a dee-per mechanistic understanding about the contributionsof transient microcircuits such as the SPN to circuit as-sembly prior to hearing onset will greatly aid efforts tounderstand how the stable networks that underlie nor-mal hearing are constructed. Absent data that bridgethe gap between synapses and neural networks in nor-mally developing animals, it is difficult to knowwhetheractivity- or experience-dependent modifications derailnormal development or introduce an abnormal outcomefollowing the conclusion of normal maturation.

The past fifty years of auditory developmental re-search have revealed a great deal about the interplay be-tween intrinsic molecular events and dynamic electricalsignaling in functional circuit maturation. The scope ofresearch in the coming years may widen to include otherbiological factors that help to shape functional circuits.Although neurons are the central players in functionalcircuits, they do not operate in isolation. Networks ofdeveloping non-sensory glial cells, cells that form the de-veloping vasculature and the molecules that form theextracellular matrix play may key roles in modulatingneural signaling and providing a physical substrate forgrowth. As we have seen in the cochlea, the non-sensorycells that make up Kolliker’s organ may act as the cata-lyst for electrical signaling throughout the pre-hearingauditory system and it is probable that exciting new con-nections between neural development and the non-neural cells and molecules that support them remainto be discovered. On the other end of the spectrum, it willbe important to link the observations made at a neurobi-ological level to their potential behavioral consequencesmeasured at the level of the whole organism. Demon-strations of abnormal circuit connectivity or map organi-zation are worthwhile and interesting in their own right,but take on additional importance when they can be as-sociated with changes in perceptual abilities. Studiesthat bridge the gaps between these various levels of anal-ysis and stations of processing along the auditory path-way promise to reveal more about the etiology andtherapeutic possibilities for treating hearing impair-ment, and may teach us a great deal about the funda-mental principles of neural development.

References

Born, D.E., Rubel, E.W., 1988. Afferent influences on brain stem audi-tory nuclei of the chicken: Presynaptic action potentials regulateprotein synthesis in nucleus magnocellularis neurons. The Journalof Neuroscience 8, 901–919.

Brainard, M.S., Doupe, A.J., 2002.What songbirds teach us about learn-ing. Nature 417, 351–358.

I. CIRCUIT DEV


Chang, E.F., Merzenich, M.M., 2003. Environmental noise retards audi-tory cortical development. Science 300, 498–502.

Cramer, K.S., 2005. Eph proteins and the assembly of auditory circuits.Hearing Research 206, 42–51.

Dallos, P., Zheng, J., Cheatham, M.A., 2006. Prestin and the cochlearamplifier. The Journal of Physiology 576, 37–42.

de Villers-Sidani, E., Chang, E.F., Bao, S., Merzenich, M.M., 2007. Criticalperiod window for spectral tuning defined in the primary auditorycortex (A1) in the rat. The Journal of Neuroscience 27, 180–189.

Deitch, J.S., Rubel, E.W., 1984. Afferent influences on brain stem audi-tory nuclei of the chicken: Time course and specificity of dendriticatrophy following deafferentation. The Journal of ComparativeNeurology 229, 66–79.

Dorman, M.F., Sharma, A., Gilley, P., Martin, K., Roland, P., 2007.Central auditory development: Evidence from CAEP measure-ments in children fit with cochlear implants. Journal of Communi-cation Disorders 40, 284–294.

Echteler, S.M., 1992. Developmental segregation in the afferentprojections to mammalian auditory hair cells. Proceedings of theNational Academy of Sciences of the United States of America 89,6324–6327.

Echteler, S.M., Arjmand, E., Dallos, P., 1989. Developmental alterations inthe frequency map of the mammalian cochlea. Nature 341, 147–149.

Harris, J.A., Rubel, E.W., 2006. Afferent regulation of neuron number inthe cochlear nucleus: Cellular and molecular analyses of a criticalperiod. Hearing Research 216–217, 127–137.

Hoffpauir, B.K., Grimes, J.L., Mathers, P.H., Spirou, G.A., 2006.Synaptogenesis of the calyx of Held: Rapid onset of function andone-to-onemorphological innervation. The Journal of Neuroscience26, 5511–5523.

Hudspeth, A.J., 2008. Making an effort to listen: Mechanical amplifica-tion in the ear. Neuron 59, 530–545.

Jackson, H., Hackett, J.T., Rubel, E.W., 1982. Organization and develop-ment of brain stem auditory nuclei in the chick: Ontogeny of post-synaptic responses. The Journal of Comparative Neurology 210,80–86.

Jhaveri, S., Morest, D.K., 1982. Sequential alterations of neuronal archi-tecture in nucleus magnocellularis of the developing chicken: aGolgi study. Neuroscience 7, 837–853.

Johnson, S.L., Eckrich, T., Kuhn, S., et al., 2011. Position-dependentpatterning of spontaneous action potentials in immature cochlearinner hair cells. Nature Neuroscience 14, 711–717.

Kanold, P.O., Luhmann, H.J., 2010. The subplate and early corticalcircuits. Annual Review of Neuroscience 33, 23–48.

Kim,G., Kandler, K., 2003. Elimination and strengtheningof glycinergic/GABAergic connections during tonotopic map formation. NatureNeuroscience 6, 282–290.

Kotak, V.C., Takesian, A.E., Sanes, D.H., 2008. Hearing loss preventsthe maturation of GABAergic transmission in the auditory cortex.Cerebral Cortex 18, 2098–2108.

Kotak, V.C., Fujisawa, S., Lee, F.A., Karthikeyan, O., Aoki, C.,Sanes, D.H., 2005. Hearing loss raises excitability in the auditorycortex. The Journal of Neuroscience 25, 3908–3918.

Koundakjian, E.J., Appler, J.L., Goodrich, L.V., 2007. Auditory neuronsmake stereotyped wiring decisions before maturation of theirtargets. The Journal of Neuroscience 27, 14078–14088.

Kral, A., Tillein, J., Heid, S., Hartmann, R., Klinke, R., 2005. Postnatalcortical development in congenital auditory deprivation. CerebralCortex 15, 552–562.

Leake, P.A., Snyder, R.L., Hradek, G.T., 2002. Postnatal refinement ofauditory nerve projections to the cochlear nucleus in cats. The Jour-nal of Comparative Neurology 448, 6–27.

Levi-Montalcini, R., 1949. Development of the acousticovestibularcenters in the chick embryo in the absence of the afferent root fibersand of descending fiber tracks. The Journal of Comparative Neurol-ogy 91, 209–242.

ELOPMENT




Metherate, R., Hsieh, C.Y., 2003. Regulation of glutamate synapses bynicotinic acetylcholine receptors in auditory cortex. Neurobiologyof Learning and Memory 80, 285–290.

Mogdans, J., Knudsen, E.I., 1994. Site of auditory plasticity in the brainstem (VLVp) of the owl revealed by early monaural occlusion. Jour-nal of Neurophysiology 72, 2875–2891.

Norton, S.J., Bargones, J.Y., Rubel, E.W., 1991. Development of otoa-coustic emissions in gerbil: Evidence for micromechanical changesunderlying development of the place code. Hearing Research 51,73–91.

Popescu, M.V., Polley, D.B., 2010. Monaural deprivation disruptsdevelopment of binaural selectivity in auditory midbrain and cor-tex. Neuron 65, 718–731.

Pujol, R., Carlier, E., Devigne, C., 1978. Different patterns of cochlearinnervation during the development of the kitten. The Journal ofComparative Neurology 177, 529–536.

Rubenstein, J.L.R., Rakic, P., 2013. Patterning and Cell Types Specifica-tion in the Developing CNS and PNS.

Ryugo, D.K., Fekete, D.M., 1982. Morphology of primary axosomaticendings in the anteroventral cochlear nucleus of the cat: A studyof the endbulbs of Held. The Journal of Comparative Neurology210, 239–257.

Ryugo, D.K., Kretzmer, E.A., Niparko, J.K., 2005. Restoration of audi-tory nerve synapses in cats by cochlear implants. Science 310,1490–1492.

Sanes, D.H., Bao, S., 2009. Tuning up the developing auditory CNS.Current Opinion in Neurobiology 19, 188–199.

I. CIRCUIT DEV


Seidl, A.H., Grothe, B., 2005. Development of sound localization mech-anisms in the mongolian gerbil is shaped by early acoustic experi-ence. Journal of Neurophysiology 94, 1028–1036.

Tierney, T.S., Russell, F.A.,Moore, D.R., 1997. Susceptibility of develop-ing cochlear nucleus neurons to deafferentation-induced deathabruptly ends just before the onset of hearing. The Journal ofComparative Neurology 378, 295–306.

Tritsch, N.X., Yi, E., Gale, J.E., Glowatzki, E., Bergles, D.E., 2007. Theorigin of spontaneous activity in the developing auditory system.Nature 450, 50–55.

Tritsch, N.X., Rodriguez-Contreras, A., Crins, T.T., Wang, H.C.,Borst, J.G., Bergles, D.E., 2010. Calcium action potentials in hair cellspattern auditory neuron activity before hearing onset. NatureNeuroscience 13, 1050–1052.

Tucci, D.L., Rubel, E.W., 1985. Afferent influences on brain stem audi-tory nuclei of the chicken: Effects of conductive and sensorineuralhearing loss on n. magnocellularis. The Journal of ComparativeNeurology 238, 371–381.

Vale, C., Schoorlemmer, J., Sanes, D.H., 2003. Deafness disrupts chlo-ride transporter function and inhibitory synaptic transmission.The Journal of Neuroscience 23, 7516–7524.

Von Bekesy, G., 1960. Experiments inHearing.McGraw-Hill, NewYork.Werker, J.F., Yeung, H.H., 2005. Infant speech perception bootstraps

word learning. Trends in Cognitive Science 9, 519–527.Werthat, F., Alexandrova, O., Grothe, B., Koch, U., 2008. Experience-

dependent refinement of the inhibitory axons projecting to themedial superior olive. Developmental Neurobiology 68, 1454–1462.

ELOPMENT


Chapter 2 - Functional Circuit Development in the Auditory System · 2019-12-21 · 2.1...

Documents

Transcript of Chapter 2 - Functional Circuit Development in the Auditory System · 2019-12-21 · 2.1...