B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

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

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

Research Report

Sustained depolarization-induced propagation of [Ca2+]ioscillations in cultured DRG neurons: The involvement ofextracellular ATP and P2Y receptor activation

Yan Zenga, Xiao-hua Lvb, Shao-qun Zengb,⁎, Shun-lian Tiana, Man Lia, Jing Shia,⁎aDepartment of Neurobiology, Tongji Medical School, Huazhong University of Science and Technology, HUST, 13 Hangkong Road, Wuhan430030, PR ChinabThe Key Laboratory of Biomedical Photonics of Ministry of Education, Huazhong University of Science and Technology, Wuhan 430074,PR China

A R T I C L E I N F O

⁎ Corresponding authors. Fax: +86 27 8369 260E-mail address: sj@mails.tjmu.edu.cn (J. SAbbreviations: DRGdorsal root ganglia;

tetraacetate; DMSOdimethyl sulfoxide; [Ca2+]receptors; PLCphospholipase C; ERthe endoEagle's medium; HEPESN-2-hydroxyethylpiptetraacetic acid; PPADSPyridoxal-5-phosphat

0006-8993/$ – see front matter © 2008 Elsevidoi:10.1016/j.brainres.2008.08.085

A B S T R A C T

Article history:Accepted 24 August 2008Available online 10 September 2008

Recently emerging evidence implicates a number of neuroactive substances and theirreceptors in mediating complex cell-to-cell communications in the ganglia. In the presentstudy, we characterized the nonsynaptic chemical coupling mediated by extracellular ATPin dorsal root ganglia (DRG) neuron cultures by using the real time imaging of ATP, whole-cell patch clamping, in conjunction with confocal calcium imaging. Sustaineddepolarization by electrical stimulation evoked intracellular Ca2+ concentrations ([Ca2+]i)oscillations in individual DRG neurons, and subsequent ATP-dependent propagation [Ca2+]ioscillations to surrounding non-stimulated neighbors. [Ca2+]i oscillations were suppressedby inositol-1,4,5-trisphosphate (IP3) receptor antagonist 2-APB, but not ryanodine. Thepropagation of [Ca2+]i oscillations was prevented by the presence of the ATP-degradingenzyme, apyrase, and completely abolished by the blockase of G protein-coupled purinergicreceptors-PLC-IP3 pathway with suramin, U73122 or 2-APB. In parallel, sustaineddepolarization elicited robust ATP release and diffusion from the stimulation site.Moreover, exogenous application of ATP to DRG cultures in large concentration elicits the[Ca2+]i oscillations in most neurons. Taken together, this data demonstrates that sustainedmembrane depolarization elicited ATP release, acting through a highly sensitive P2Yreceptors/IP3-mediated signaling pathway to mediate the propagation of intercellular Ca2+

signaling, which suggest a novel signaling pathway for neuronal communication in DRG.© 2008 Elsevier B.V. All rights reserved.

Keywords:Dorsal root gangliaElectrical stimulationATP release[Ca2+]i oscillationIntercellular communication

8.hi).ATPadenosine 5′-triphosphate disodium salt; EGTAethyleneglycol bis(aminoethylether)iintracellular Ca2+ concentration; IP3Rsinositol-1,4,5-trisphosphate receptors; RyRsryanodineplasmic reticulum; TGthapsigargin; BSSbalanced salt solution; DMEMDulbecco's modifiederazine-N-2-ethanesulfonic acid; BAPTA-AM1, 2-bis (2-aminophenoxy) ethane-N,N,N',N'-e-6-azophenyl-2′4′-disulfonic acid

er B.V. All rights reserved.

13B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

1. Introduction

Spike activity was elicited by an intracellular voltage step inprimary afferent neurons in dorsal root ganglia (DRG) in vivo(Devor and Wall 1990) and in vitro (Utzschneider et al. 1992)that excited adjacent passive neurons that share the sameDRG. It appears that such cell-to-cell interactions within DRGresult from neither synaptic nor electrical junctions, becauseadjacent neurons are isolated in individual satellite cellsheaths (Lieberman 1976). It has been inferred that chemicallymediated cross-depolarization and the net increase in theinput resistance of neurons contributed to the phenomenon(Amir and Devor 1996, 2000), but exact mechanisms remainelusive. During the last decade, the general assumption hasbeen that the cell body (soma) of a neuron, which contains thenucleus, is mainly responsible for synthesis of macromole-

Fig. 1 – Sustained depolarization-evoked [Ca2+]i oscillations in DRresponse to 500 ms depolarization pulses in ROI of DRG culturescorresponding holding current recordings. (C) 500 ms, 1 Hz electneuron under current clamp conditions. (D) corresponding actionrepresentative [Ca2+]i wave propagation within a DRG neuron inobserved first in the cell body, then the [Ca2+]i wave propagated

cules and has a limited role in cell-to-cell communication.Huang and Neher (1996) first reported that depolarization insmall DRG neurons (15–25 μm in diameter, C-type) couldtrigger capacitance increase, indicating somatic exocytosis.Among others, some groups have found quantal release ofpreloaded transmitters or dyes, and pain-related peptidessuch as CGRP and substance P from the DRG neuron somata inresponse to action potentials (Zhang et al. 1995; Bao et al. 2003;Zhang et al. 2004; Ouyang et al. 2005). The somatic secretionmost likely provides a way for an excited DRG neuron in anautocrine or paracrine fashion via extrasynaptic mechanismsto modulate the activity of adjacent ganglion neurons(including itself), and thereby affect the transmission ofsensing and nociceptive inputs to the spinal cord and highercentral nervous system (CNS). As for reception of the chemicalsignals, DRG neurons are known to contain a wealth ofreceptors for various neurotransmitters such as purinergic

G neuron cultures. (A) a typical trace of [Ca2+]i oscillations in. (B) The fluorimetric Ca2+ signals are accompanied by therical currents induced a typical [Ca2+]i oscillations in a DRGpotentials evoked by current stimulus trains. (E)

response to sustained depolarizations. Signal initiation wasalong the neuron axons (white arrows). Bar, 20 μm.

14 B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

receptors, glutamate receptors and tachykinin receptors(Lovinger and Weight 1988; Sato et al. 1993; Shinder andDevor 1994; Brechenmacher et al. 1998; Marvizon et al. 2002;Yang et al. 2003; Wu et al. 2004). Therefore, conditions withinDRG may permit extracellular messenger-mediated nonsy-naptic functional coupling.

ATP, a ubiquitous neurotransmitter/neuromodulator in theCNS and the peripheral nervous system, for decades has beenknown to operate as a very important intercellular transmitterby the activation of purinergic receptors to mediate inter-cellular communication in various cells (Guidotti 1996; Piazzaet al. 2007; Wu et al. 2006). It is considered the dominantextracellular messenger for astrocyte-to-astrocyte or astro-cyte-to-neuron intercellular communication in the brain(Guthrie et al. 1999). It also is a transmitter especiallyimportant for signaling injurious nociceptive informationand sensory regulation. Nonsynaptic release of ATP fromelectrically stimulated DRG somata has been demonstrated toinhibit schwann cell proliferation (Stevens et al. 2004), toarrest schwann cell development at the premyelinating stage(Fields, 2006), and especially to play a crucial role in bidirec-tional communication between neurons and surroundingsatellite glial cells (Zhang, et al., 2007). It is of interest thatATP released from the keratinocytes in response to mechan-ical stimulus induces an increase of [Ca2+]i in the DRG neuronsafter a time lag (Koizumi et al. 2004). The evidence above mayimplicate that the possibility that ATP also acts in directneuron-to-neuron communication in the DRG. Therefore, wepropose that activation-dependent ATP release from somamay act as amediator of neuron-to-neuron signals in the DRG.

In the present study, using real time ATP imaging, and acombined whole-cell patch clamp and confocal calciumimaging, we showed an obvious communication mechanismin DRG neurons: nonsynaptic chemical communication weremediated by the extracellular ATP-P2Y coupled to phospho-lipase C receptors system.We characterized the release of ATPfrom cell somata with a sustained depolarization, togetherwith the oscillatory increase of free [Ca2+]i at the soma,subsequently ATP diffusion from the release site to a relativelydistant receptive membrane, and then P2Y receptor-mediateddelayed [Ca2+]i oscillations in surrounding DRG neurons.Therefore, ATP acting as a vital intercellular signals, plays amajor role in mediating paracrine communications betweenneurons-neurons in DRG cultures.

Fig. 2 – Depolarization-evoked [Ca2+]i oscillations were IP3

receptors dependent. Ca2+ signals evoked by the applicationof depolarizing pulses from −80 mV to −20 mV in avoltage-clamped neurons bathed in various solution.[Ca2+]i oscillations still were evoked in Ca2+-free solution(Ca2+-free external solution plus 2 mM EGTA), the solutionplus 20 μMNi2+, or plus ryanodine (10 μM, 20 min).Whereas,[Ca2+]i oscillations were suppressed in the solution plus2-APB at 100 μM and plus TG (at 20 μM for 60 min), or byremoval of internal Ca2+ (with 20 μM BAPTA-AM), althoughsome Ca2+ influx were shown under these three conditions.Low panel shows the change of [Ca2+]i peaks as time lapse ineach group. The data were expressed as F/F0. The datarepresent the means±S.E.

2. Results

2.1. Sustained depolarization-induced [Ca2+]i oscillationsin stimulated DRG neurons

Time-lapse confocal Ca2+ imaging were used to study thespatial and temporal profiles of depolarization-inducedchanges in [Ca2+]i in rat DRG neuron cultures. Under whole-cell patch-clamp conditions, holding at-80 mV, using stepstimulations from –80 mV to-20 mV of durations 500 ms in5 mV increments, could evoke repetitive [Ca2+]i elevations(Figs. 1A and B, n=56) approximately 100–150 ms afterdepolarization onsets, which characterized with a meanfrequency of 9–10 times/min. When the stimulation ceased,

the [Ca2+]i oscillations continued for total of 10–14 min. Anoscillatory [Ca2+]i response was defined when [Ca2+]i peakswere repeated more than three times within a single neuron.We decided to find out if calcium oscillations could begenerated by more physiological stimulations. We applied500 ms, 1 Hz current stimulus trains into cultured DRGneurons, the characteristic [Ca2+]i oscillations still appeared ina lot of stimulated neurons with different amplitudes (Figs. 1Cand D, n=34). Interestingly, in some stimulated neurons (Fig.1E, n=15), the [Ca2+]i oscillations exhibited a nonuniformpropagation pattern, briefly, the first increase in [Ca2+]i

15B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

occurred at the soma of the neuron, then a Ca2+ wave spreadto distal sections only along the large caliber axis, which wasthen repeated to create an oscillatory pattern. It has recently

Fig. 3 – Depolarization-evoked propagation of [Ca2+]i oscillationsrise in [Ca2+]i occurred first in the stimulated cell 3 (black trace) an2 (green trace), and then propagated cell 4 (orange trace) and cell[Ca2+]i wave in different conditions, which shows a significant inneurons after applying apyrase at 80 U/ml, suramin at 300 μM, Uextracellular Ca2+ (with 2 mM EGTA) failed to inhibit the propagavelocity, as approximated by the slope of the left graph, was notinitiation and slowedwith increasing distance. (right) themagnitwas highest at the place of origin and decreased with growing d

been shown in other neurons that release of Ca2+ frominternal stores can also spread within a neuron as regen-erative waves (Nakamura et al. 1999; Kapur et al. 2001; Simon

in DRG neuron cultures. (A) a typical experiment showing thed then propagatedmultidirectional to cell 1 (red trace) and cell5 (blue trace). (B) the correspondingly colored traces forhibition of the propagation of [Ca2+]i oscillations in DRG73122 at 10 μM, and 2-APB at 100 μM. whereas removal oftion. (C) (left) the latency to rise in [Ca2+]i, wave propagationlinear with distance. [Ca2+]i spread faster near the site ofude wasmeasured, similarly, the amplitude of the Ca2+ signalistance. Bar, 20 μm.

16 B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

et al. 2005; Choi et al. 2006). The ER network may act as afunctional tunnel for rapid Ca2+ transport generating Ca2+

signals in local domains in neurons (Choi et al. 2006).

2.2. An internal Ca2+ release is responsible for thesustained-induced [Ca2+]i oscillations

It is interesting that the oscillatory [Ca2+]i during depolariza-tion were not modified by perfusion with low-calcium, orcalcium-free solutions with 2 mM EGTA (Fig. 2. n=24), and theblocker of Ca2+ channels Ni2+(Nickel chloride) (20 μM) (Fig. 2.n=24), which demonstrated that Ca2+ entry was not requiredfor [Ca2+]i oscillations, and suggested that Ca2+ release frominternal storesmay be required to generate [Ca2+]i oscillations.We went on to identify the internal source of the [Ca2+]ioscillations. DRG neurons could not develop [Ca2+]i oscilla-tions after application of BAPTA-AM at 20 μM for 60 min toload the fast Ca2+ chelator BAPTA into all cells in the dishes(Fig. 2, n=12), demonstrating an increase in cytosolic Ca2+ wasrequired to generate [Ca2+]i oscillations. If Ca2+ release fromintracellular Ca2+ stores was indeed involved in stimuli-induced [Ca2+]i oscillations, the effect of stimuli effects shouldalso have been abolished when intracellular Ca2+ stores weredepleted, thus DRG neurons were preincubated with thapsi-gargin (TG) at 20 μM for 60 min to deplete intracellular Ca2+

stores. Stimuli did not produce any intracellular oscillation in[Ca2+]i of neurons tested (Fig. 2, n=12), suggesting that the ERCa2+ store was the main source of [Ca2+]i oscillations. Themain ER Ca2+ store was mediated either by inositol trispho-sphate receptors (IP3Rs) or ryanodine receptors (RyRs)(Albrieux et al. 1997, Ouyang et al. 2005). In order to determinethe contribution of IP3Rs and RyRs channel to [Ca2+]i oscilla-tions, DRG neuron were pretreated with ryanodine, RyRs-specific agents at 10 μM for 20 min, this concentration isknown to inhibit the RyRs of DRG neurons (Albrieux et al.1997, Ouyang et al. 2005), and had no effect on depolarization-induced [Ca2+]i oscillations, suggesting that Ca2+ release didnot require the activation of RyRs (Fig. 2, n=15). In contrast,when cells pretreated with 2-aminoethyl diphenylborateat100 μM (2-APB), a plasma membrane permeable antagonistof the IP3Rs, the depolarization-induced Ca2+ response wascompletely eliminated (Fig. 2, n=15). These results indicatethat depolarization-induced [Ca2+]i oscillations are the resultof the actions of Ca2+ mobilization from IP3-sensitive Ca2+

stores. These observations suggest that a mobilization ofintracellular Ca2+ stores may be mainly involved in the Ca2+

oscillations for sustained depolarization, although Ca2+ influxfrom extracellular buffer also contributed to [Ca2+]i transients(Fig. 2).

2.3. [Ca2+]i oscillations propagated to surroundingneurons without stimulations

This research also showed that, in addition to the stimulatedneurons, [Ca2+]i oscillation s were followed by the induction ofa propagating [Ca2+]i wave with incidence of activating in5±2.7 surrounding passive neurons (Figs. 3A and B, n=39),which appeared a oscillatory pattern in each case after 11–17 sfrom the stimulated one according to the distance, namelythe propagating of [Ca2+]i oscillations. By plotting the latency

between the onsets of the signal in the neurons as a functionof increasing distance, it could be seen that the propagationvelocity was not uniform. The [Ca2+]i oscillations spread fasternear the site of initiation and slowed as it traveled out (Fig. 3C,n=19). There also was a gradual decrease in amplitude as thewave spread (Fig. 3D, n=19). [Ca2+]i oscillations near theinitiation site were always the greatest in the amplitude andthe rate of rise compared with the oscillations measuredfurther away. The mean propagation velocity of the [Ca2+]ioscillations was 14.2±5.8 μm/s. All this data suggest thesecretion of a diffusible messenger.

We found that the propagation of [Ca2+]i oscillations wasabolished by 80 U/ml ATP-degrading enzyme apyrase (Figs. 3Aand B, n=20). Hence, ATP may be responsible for [Ca2+]ioscillation of a population of DRG neurons without stimula-tions by paracrine communication. Suramin is a sulfatedpolyanion that can inhibit the activity of a variety of proteinsand was found to inhibit the ATP induced rise in [Ca2+]i inglobular substance vesicles (Suzuki et al., 1997). To determinewhether the depolarization-associated propagation of [Ca2+]ioscillations was similarly sensitive to suramin, DRG cultureswere treated with 300 μM suramin. After prolonged (N10 min)exposure to suramin, the propagations were inhibited, sug-gesting that this agent blocked the ability of depolarization toinduce the propagation of [Ca2+]i oscillations.

Next, pretreatment of neurons with 10 μM phospholipaseC (PLC) antagonist U73122 (Figs. 3A and B, n=12) or 100 μM 2-APB (Figs. 3A and B, n=12), respectively, after inhibited thepropagation, suggesting ATP all the way through to theactivation of P2Y G protein-coupled receptors and theirsignaling cascades PLC-IP3 system-acting on internal Ca2+

restore mediated the process. In addition, removal of externalCa2+ from the bath (with 2 mM EGTA) did not change thepropagation of [Ca2+]i oscillations (Figs. 3A and B, n=12),suggesting that the depolarization-induced propagation ofCa2+ oscillations is independent of external Ca2+. So weconclude that the present model is based on purinergictransmission, where the underlying mechanism may be thediffusion of ATP in the extracellular space, which mobilizedIP3 and released Ca2+ from intracellular stores.

2.4. Depolarization-induced release and diffusion of ATPfrom DRG neurons

This study sought to directly demonstrate directly thedepolarization-evoked release of ATP from DRG neurons,therefore, we modified the luciferin–luciferase chemilumi-nescence bioassay for detecting ATP levels from living cells.Neurons were bathed in a solution containing the luciferin–luciferase reagents and images were captured with 1 s apart.Fig. 4A shows the DIC image and bioluminescence imagesbefore and after depolarization respectively. The standardcalibration curve obtained under this condition showed a highcorrelation between the bioluminescence intensity and theATP concentration with a correlation coefficient of 0.986 overa concentration range of 100 nM–10 mM (Fig. 4B). The restinglevel of the bioluminescence signal was very low; then, it thenwas increased to a level sufficient to evoke increases in [Ca2+]iin neurons (285±159 μM) in response to depolarization (Fig.4C, n=28). To visualize the spatiotemporal dynamics of the

Fig. 4 – Depolarization-evoked ATP release and diffusion from DRG neurons. The fluorescence intensity is indicated by pseudocolors that are used to enhance the changes in brightness associated with the changes in ATP. Blue indicates a low fluorescentlevel. As ATP rises, the fluorescence intensity increased and the color changed gradually to green and then to red. (A) the imagesshowphoton intensity (yellowdots) in a field of DRG neurons bathed in luciferin–luciferase reagent before (−30 s) and after (30 s)depolarization. The position of pipette (white arrow) is shown in a DIC image of DRG neurons, and the spatial distribution ofphoton intensity following electrical stimulation shown as a two-dimensional pseudo color surface plot before and serialimages after depolarization. (B) a typical bioluminescence intensity–ATP concentration relationship was shown after 30 s afterstimulation under these conditions. Various concentrations of ATP standard solution were injected in the presence of theluciferin–luciferase reagent, and fluorescence were then accumulated for 30 s. (C) normalized fluorescence value in theappropriate point away from stimulation point from different treatments, which show that 100 μM TG for 20 min–60 min or10 μM U73122 inhibited depolarization-induced ATP release. Bar, 20 μm.

17B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

18 B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

depolarization-evoked release of ATP from DRG neurons, theextracellular ATP levels were plotted as pseudo color imagesusing the Excel 2-D surface plot program. As shown in Fig. 4A,the levels of extracellular ATP after electrical stimulationwere highest at the site of stimulation, subsequently diffusedto a relatively distant receptive membrane and decreasedconcentrically, which was consistent with spreading rate of[Ca2+]i oscillations with a correlation coefficient of 0.99. Theseresults show that the depolarization-evoked propagation of[Ca2+]i oscillations corresponded to the release of ATP fromDRG neurons. In fact, pretreated with 20 μM TG for 60 min or10 μMU73122 inhibited depolarization-induced ATP release ina Ca2+-containing bath (Fig. 4C, n=24). This data suggests thatATP release is associated with an increase in cytosolic Ca2+

Fig. 5 – ATP-induced [Ca2+]i oscillations and transients by activatilive fluorescence intensity were taken by confocal microscope frsolution, 250μMATP in Ca2+-free solution, 100μMATP in Ca2+-freplusing 100 μM 2-APB, and 250 μM ATP plusing 10 μM U73122.intensity of each group. (C) the change of [Ca2+]i peaks as time laprepresent the means±S.E.

and the increase in [Ca2+]i must be present to cause thisincrease in ATP levels.

2.5. Characterization of ATP evoked [Ca2+]i increases inDRG neurons

In support of our proposition that the depolarization-inducedpropagation of [Ca2+]i oscillations in DRG neurons was linkeddirectly to the effect of ATP released from DRG neurons, weexplored ATP-induced [Ca2+]i dynamics in DRG neuron cul-tures. ATP at 10–150 μM elicited [Ca2+]i transient responses,sustained [Ca2+]i rise and a periodic repetitive [Ca2+]i transientsin 76±5% of DRG neurons (Figs. 5A and B, n=29). ATP at 150–250 μM elicited repetitive [Ca2+]i oscillations in 56±7% of DRG

on of P2Y receptors in DRG neurons. (A) the representatives ofom DRG neurons exposed to 250 μM ATP in Ca2+-containede solution, 250μMATP plusing 300μMsuramin, 250μMATP(B) corresponding to the time curves based on fluorescencese in each group. The data were expressed as F/F0. The data

19B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

neurons with a frequency of approximately 100 ms (-1) inextracellular Ca2+ contained solution (Fig. 5A and B, n=32).This data suggests that the quantity of DRG neurons in thepreparations responding to ATP with periodical [Ca2+]i oscilla-tions depended on the concentration of ATP. ATP is capable oftriggering [Ca2+]i transients in various cells via either metabo-tropic P2Y or ionotropic P2X purinoreceptors. To determinethe effects of ATP on Ca2+ release from ER, we used Ca2+-freebuffers to obviate the potentially confounding effects of store-operated Ca2+ entry and Ca2+ influx via P2X receptors. If suchoscillations in [Ca2+]i are mediated by the P2X receptors theycan be negated in the absence of extracellular Ca2+. Surpris-ingly, when the application of ATP was repeated in Ca2+-freesolution, the [Ca2+]i oscillations were virtually unchanged(Figs. 5A and B, n=22), and P2X receptor antagonist pyridoxal-phosphate-6-azophenyl-2′, 4′-disulfonic acid (PPADS) at 10 μMcould not eliminated [Ca2+]i oscillations (data not shown,n=12), indicating that the Ca2+ oscillations triggered by ATPmainly resulted from Ca2+ movements between intracellularcompartments rather than Ca2+ influx through P2X receptors.

3. Discussion

The present findings demonstrated and delineate a commu-nication mechanism of sustained depolarization-inducedcalcium signaling, manifesting as intercellular waves of[Ca2+]i oscillations, that is, in DRG neurons, the result ofcalcium-induced secretion of ATP and activation of puriner-gic receptors. Our study shows that; sustained membranedepolarization or action potentials trains applied to DRGneurons, elicits IP3-mediated [Ca2+]i oscillations, facilitatesthe release of ATP that activated its neighboring neurons viametabolic ATP receptors, then, mobilizes internal Ca2+ occursin these neurons, that appears to be a propagation of [Ca2+]ioscillations. Therefore, it can be assumed that activity-dependent regulation of ATP contributes to sustaineddepolarization-induced dynamic interaction in DRG neurons.

Several special electrical stimulation patterns in someexperiments have been used to elicit [Ca2+]i transients orsparks which couple to secretion activities in single DRGneurons (Zhang et al. 2004; Ouyang et al. 2005). The character-ized [Ca2+]i oscillations in DRG neurons elicited by sustainedstep depolarizations have not been reported before, althoughoscillations of cytoplasmic Ca2+ levels have been observed inmany kinds of excitable and nonexcitable cells (Tsien andTsien 1990; Fewtrell, 1993; Stefan et al. 2000;Yoshida et al.2003). As an initial step toward understanding the propagationof [Ca2+]i oscillations during this membrane depolarizationprocess, the mechanisms of [Ca2+]i oscillations in single DRGneurons, evoked by the depolarization itself, were investi-gated. Since sustained depolarizations could induce [Ca2+]ioscillations in the absence of extracellular Ca2+, we think thatthemechanisms of ultradian oscillations appear to depend oncycles of intracellular Ca2+ fluxes from Ca2+-sequesteringorganelles into the cytoplasm, followed by a subsequent Ca2+

reduction. The receptors for the second messenger IP3,constitute a family of Ca2+ channels responsible for themobilization of intracellular Ca2+ stores. IP3 mediated-oscilla-tory patterns in response to various factors have been

determined (Keizer et al. 1995; Fanchaouy et al. 2005; Kawanoet al. 2006). The increase in the levels of IP3 results in theopening of Ca2+ channels present in the ER and the subse-quent release of Ca2+ into the cytosol (Taylor and Richardson1991). We characterized the IP3 pathway in depolarization-mediated [Ca2+]i oscillations. The [Ca2+]i oscillations wereattenuated by applications of either a PLC inhibitor, orinhibitors for IP3 receptors. Thus our depolarization patternevoked repetitive intracellular Ca2+ mobilization that mightalso explain the selective existence of oscillatory [Ca2+]i in DRGneurons. Our results support the notion that membranedepolarization may directly trigger Ca2+ release from internalCa2+ stores in some excitable cells (Bers et al. 1991) and thexenopus oocyte (Charpentier and Kado 1999). However we stillquestioned whether the extracellular Ca2+ influx is necessaryfor the IP3-dependent Ca2+ release because that Ca2+-inducedCa2+ release in sensory neurons has been suggested followingdifferent electrical stimulation protocols (Ouyang et al. 2005).The effect of stimuli on Ca2+ dynamics in Ca2+-free bathsolution with EGTA was examined. Stimuli still inducedcalcium oscillations in most of DRG neurons (Fig 2). Therefore,sustained depolarization could elicit [Ca2+]i oscillations with-out Ca2+ influx.

One of the most interesting findings of our study is thatrelease of ATP in response to electrical nerval stimulationplays a pivotal role in neuron-neuron communication inDRGs. Our study showed that the inhibition of the P2Y-PLC-IP3 pathway completely block nerve-stimulated induced[Ca2+]i oscillations propagation in DRG neurons withoutaffecting [Ca2+]i transients in stimulated neurons (Fig. 4B).In this pathway, ATP was secreted from excited DRGneurons, and it diffused to neighboring cells and stimulatedthe P2Y receptors, resulted in the activation of PLC toproduce IP3, which diffused in the cytoplasm to release Ca2+

from intracellular stores in the neurons without stimula-tion. We could abolish the propagation of [Ca2+]i oscillationsby blocking any site through this pathway. These results,together with the finding that P2Y receptors are expressedstrongly in DRGs (Fig. 5), suggest that somatic ATP releasedfrom DRG neurons directly affects the activity of passiveneurons by activation of P2Y receptors. Our findings aredifferent from the Zarei report that sensory neurons havethe machinery to form direct purinergic synapses on eachother (Zarei et al. 2004). The oscillatory [Ca2+]i increaseobserved in passive neurons (Fig. 4A) in our studying is alsoconsistent with the slow P2Y receptor-mediated [Ca2+]ioscillation by adding exogenous ATP into buffer (Fig. 5).Our results also suggest that neuronal soma could commu-nicate with each other in the ganglia by propagatinginformation through neuronal-neurons. Modified luciferin–luciferase chemiluminescence bioassay allows us to identifyATP released from the somata of DRG neurons (Fig. 3).

Koizumi and colleagues showed that Ca2+ waves aretransmitted to the sensory neurons in DRG through activationof P2Y2 receptors mediated by ATP and by using sametechniques that have been used in the current article (Koizumiet al. 2004). This conclusion is further confirmed by resultsshowing release events disappeared in the presence ofapyrase (Fig. 4. 1).We also determined the mode of ATPsecretion. In the traditional view, the only role of action

20 B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

potential−induced membrane depolarization is to open thevoltage-gated Ca2+ channels, resulting in an elevation of the[Ca2+]i, which in turn triggers vesicular exocytosis and secre-tion of vesicular contents (Katz 1969).

In the experiment outline in this paper, we found that ATPrelease occurred in the absence of extracellular calcium butdepended on intracellular calcium; the depletion of intracel-lular calcium stores with TG blocked ATP release. A largenumber of observations have described that locally nonuni-form Ca2+ levels during IP3-dependent Ca2+ oscillations maycontribute to exocytosis and fluid secretion (Kasai andAugustine 1990; Rooney et al. 1990; Nathanson et al. 1992;Toescu et al. 1992; Hille 1994; Tse et al. 1994). Both thesustained [Ca2+]i increase and [Ca2+]i oscillations can stimulatea larger increase in exocytosis. Even the possibility thatvoltage, and not calcium, can regulate the release of trans-mitter has also been claimed in fresh DRG neurons (Zhang andZhou 2002). Some retinal neurons also appear to releasetransmitters by a mechanism that does not require calciuminflux (Schwartz 1987). In addition, ATP release, due tomechanical or chemical stimulation, was unaffected by theblocking of Ca2+ with TG or BAPTA/AM chelators in somereports (White and Leslie 1982; Wang et al. 2000; Andersonet al. 2004).

These observations challenge the conventional view.Therefore, we propose that IP3-dependent Ca2+ release fromthe ER may provide the Ca2+ required for ATP exocytosis inthis mode of ATP secretion. DRG neurons used the changes intheir intracellular calcium levels to trigger release of ATP intothe extracellular space, then to activate receptors on othernearby neurons and to trigger a propagated calcium wavethat can travel several hundred micrometers over a timescaleof s.

After the release from DRG neurons, ATP effects on cellularfunction are generally transduced by two classes of purinergicreceptors, the G protein-coupled P2Y receptors and the ATP-gated, calcium-permeable P2X receptor channels. DRG neu-rons express a complex array of purinergic receptors, mainlycomposed of P2X2 and P2X3 in small-and middle-sized DRGneurons, and P2Y1, P2Y2, and P2Y4 subunits in all-sizedneurons (Burnstock et al. 2000; Tominaga et al. 2001; Mor-iyama et al. 2003; Wynn et al. 2004; Cockayne et al. 2005; Chenet al. 2005; Koizumi et al. 2004; Luo et al. 2006; Seino et al. 2006).Depolarization-evoked [Ca2+]i oscillations propagation toneighboring DRG neuronal largely depended on intracellularCa2+ via P2Y receptors rather than attributes to Ca2+ influx viaP2X receptors in this study. We excluded the involvement ofP2X receptors in Ca2+ entry by demonstrating the depolariza-tion-induced Ca2+ oscillations in the presence of Ca2+-free bathor P2X receptor antagonist PPADS. Moreover, Gallagher andSalter (2003) have obtained velocities of 8 μm/s and 16 μm/s forpropagation of the Ca2+ wave through the release of ATP atastrocyte junctionswith P2Y1 and P2Y2 receptors, respectively.The velocities 14 μm/s detected in cultured DRG neurons arewithin the range of those that observed in cultured astrocytes,which also implying a P2Y receptors mediated calcium wave.Although our data roughly characterized that P2Y receptorswere responsible for depolarization-induced propagation of[Ca2+]i oscillations, it is likely the P2Y1, P2Y2 or P2Y4 receptorsmay be involved in [Ca2+]i oscillations. Therefore, the compre-

hensive characterization of all P2Y subtypes of DRG neurons isunder way in our laboratory.

To summarizing, there are considerable notions in theliterature concerning the question of whether transmissionintra DRG is chemical, involving the release of chemicalsubstances (Amir and Devor 1996), because synaptic contactsare extremely rare in DRG neurons (Shinder et al. 1998) andgap junctions are not present in DRG neuronal cells (Huang etal. 2005). We observed here the release of ATP through Ca2+-dependent pattern from cell soma and their diffusion from therelease site to a relatively distant receptive membrane, in theabsence of a closely apposed postsynapticmembrane, and thegradual propagation of [Ca2+]i oscillations mediated by neuro-nal P2Y receptors linked to the IP3-Ca2+ intracellular signaltransduction cascade. This spread process of [Ca2+]i oscilla-tions provides one pathway by which DRG neurons, byreleasing signaling molecules that including ATP, may alsoplay active roles in adjacent neuronal behavior. Thesefindings, as well as the approaches presented here, mayprovide important insight into the mechanisms by whichneurons are depolarized by neighboring neurons in the sameganglion with repetitive stimulation, and store-mediated Ca2+

signaling in modulating synaptic transmission and neuronalexcitability.

4. Experimental procedures

4.1. Preparation of DRG neuronal cultures

Primary cultures of DRG neurons were prepared from post-natal Sprague Dawley rats in 4–6 d (Tongji Medical CollegeLaboratory Animal Center, Wuhan, China). The pups weredeeply anesthetized with an overdose of ketalar and decapi-tated. DRG were dissected in D-Hank's buffered salt solution;incubated in 0.25% trypsin in Dulbecco's modified Eagle'smedium (DMEM) at 37 °C for 15 min; washed three times witha 1:1 mixture of DMEM and Nutrient F12 supplemented with20% fetal calf serum, 2.2 mM L-glutamine, 15 mM HEPES,50 μg/mL gentamicin, and 30 mM D-glucose at roomtemperature; and triturated 40–60 times using a fire-polishedPasteur pipette. Neurons were seeded at a density of 2×10 4

cells/ml onto 25 μg/ml poly-L-lysine-coated glass coverslips.After 24 h, this medium was completely replaced with amaintenance medium containing 96% Neurobasal, 2% B27,and 1% L-glutamine. This medium was subsequently givenhalf-changes every 3 days. With the use of this protocol, 95–97% of the cultured cells were DRG neurons. The neuronswere cultured for 8–9 days at 37 °C in a humidified atmo-sphere with 5% CO2 before the experimental were performed.

4.2. Electrophysiology

Whole-cell patch-clamp recordings were performed by usingthe PC2-C and pCLAMP software (INBIO, Wuhan, China).Under whole-cell voltage clamp conditions, membrane vol-tage was held at −80 mV with the patch pipette (resistance of2–4 M). Step stimulation from −80 mV to −20 mV of durations500 ms in 5 mV increments were applied every 20 s. Under

21B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

whole-cell current clamp conditions, 500 ms, 1 Hz currenttrains were applied. Standard external solution contained (inmM): NaCl 150, KCl 5.0, CaCl2 2.0, MgCl2 1.0, HEPES 10, TEA Cl 10and glucose 10 (pH 7.4 adjusted with NaOH). For a Ca2+-freebath, we removed CaCl2 from the external solution and added2 mM EGTA in all cases. The standard internal solutioncontained (in mM) CsCl 153, NaCl 10, HEPES 10，TEA Cl 10 andATP 4 (pH 7.2). Whole cell membrane currents changes werefiltered at 5 kHz and at a sampling rate of 20 kHz via a Digidata1200 series interface. A single neuron of 25–45 μm in diameterwas selected. Only one neuron per coverslip was used forelectric stimulation. We carried out all experiments at roomtemperature (22–25 °C).

4.3. Ca2+ imaging

To examine the ability of depolarization to evoke Ca2+

dynamics, cells were loaded with the Ca2+ indicator dyeFluo-3 (5 μM Fluo-3 acetoxymethyl ester (AM), 0.005% PluronicF-127, and 0.1% DMSO for 30 min). Coverslips were thensubjected to whole-cell patch clamping in conjunction withconfocal calcium imaging (FV1000, Olympus, Tokyo, Japan),and imaged using the 20× (N.A.=0.75) or 40× (N.A.=0.9)objectives. The fluorescent dye was excited with light at488 nm from an argon laser. Emitted light was filtered with a515±35-nm filter. Relative changes were assessed in intracel-lular Ca2+ concentration ([Ca2+]i) in the regions of interest (ROI)selected in the soma and the neurites. The fluorescencesignals over time were normalized with respect to the restingfluorescence intensity (F0) and expressed as F/F0.

4.4. Imaging of ATP release

ATP release from DRG neurons was detected with a real timeluciferin–luciferase bioluminescence assay (Koizumi et al.2004). In brief, after an initial 30 min superfusion period,superfusion was stopped and the cell chamber was filled withBSS containing a luciferase reagent (ATP bioluminescenceassay kit CLS II; Roche Diagnostics, Mannheim, Germany). ATPbioluminescence was detected and visualized by using a high-sensitivity single photon-counting camera (C2400-35; Hama-matsu Photonics) to correlate photon counts with increases inextracellular ATP. The absolute ATP concentration wasestimated using a standard ATP solution through followingformula (ATP bioluminescence assay kit CLS II).

ATP Mð Þ = S Rluð Þ−B Rluð ÞI Rluð Þ � Amount of ATP standard added Mð Þ

4.5. Drugs

The following chemicals were used: ATP, poly-L-lysine, L-glutamine, HEPES, D-glucose, ryanodine, suramin hexasodiumsalt, PPADS, EGTA, TG, 2-aminoethyl 2-APB, BAPTA-AM, U-73122, and DMSO were bought from Sigma (St. Louis, MO);Fluo-3 acetoxymethyl ester (AM) and Pluronic F-127 fromMolecular Probes (Eugene, OR); DMEM, Dulbecco's mediumNutrient F12, Neurobasal medium, 20% fetal calf serum, andB27 from Invitrogen (San Diego, CA); ATP bioluminescence

assay kit CLS II from Roche Diagnostics (Mannheim, Ger-many). For 2-APB, BAPTA-AM, and U73122 dissolved in DMSO,the control cultures included 0.1%DMSO in themedium. Otherreagents were purchased from Shanghai Life TechnologyCompany. All chemicals were of analytical grade.

4.6. Statistics

Means±SEM of n are shown. n indicates the number of cells.The significance of difference between means was deter-mined, when appropriate, using Student's t test, and pairedt test. Pb0.05 was considered statistically significant.

Acknowledgments

We are grateful to Prof. Shao-qun Zeng for the invaluableassistance in the experiment of whole-cell patch-clamprecordings and calcium imaging. We also thank Dr. Guo Wenfor the help in cell culture, as well as Drs. Xiao-quan Yang andNing Sun for their helpful insights on data analysis on themanuscript. This work was supported by the National NaturalScience Foundation of China.

R E F E R E N C E S

Amir, R., Devor, M., 1996. Chemically mediated cross-excitation inrat dorsal root ganglia. J. Neurosci. 16, 4733–4741.

Amir, R., Devor, M., 2000. Functional cross-excitation betweenafferent A-and C-neurons in dorsal root ganglia. Neuroscience95, 189–195.

Anderson, C.M., Bergher, J.P., Swanson, R.A., 2004. ATP-inducedATP release from astrocytes. J. Neurochem. 88, 246–256.

Albrieux, M., Sardet, C., Villaz, M., 1997. The two intracellular Ca2+release channels, ryanodine receptor and inositol1,4,5-trisphosphate receptor, play different roles duringfertilization in ascidians. Dev. Biol. 189, 174–185.

Bao, L., Jin, S.X., Zhang, C., et al., 2003. Activation of delta opioidreceptors induces receptor insertion and neuropeptidesecretion. Neuron 37, 121–133.

Bers, D.M., 1991. Excitation-Contraction Coupling and CardiacContractile Force. Kluwer Academic Publishers, Dordrecht.

Brechenmacher, C., Larmet, Y., Feltz, P., et al., 1998. Cultured ratsensory neurones express functional tachykinin receptorsubtypes 1, 2 and 3. Neurosci. Lett. 241, 159–162.

Burnstock, G., 2000. P2X receptors in sensory neurones. Br. J.Anaesth. 84, 476–488.

Charpentier, G., Kado, R.T., 1999. Induction of Na+ channel voltagesensitivity in Xenopus oocytes depends on Ca2+ mobilization.J. Cell. Physiol. 178, 258–266.

Chen, Y., Li, G.W., Wang, C., et al., 2005. Mechanisms underlyingenhanced P2X receptor-mediated responses in theneuropathic pain state. Pain 119, 38–48.

Choi, Y.M., Kim, S.H., Chung, S., et al., 2006. Regional interaction ofendoplasmic reticulum Ca2+ signals between soma anddendrites through rapid luminal Ca2+ diffusion. J. Neurosci. 26,12127–12136.

Cockayne, D.A., Dunn, P.M., Zhong, Y., et al., 2005. P2X2 knockoutmice and P2X2/P2X3 double knockout mice reveal a rolefor the P2X2 receptor subunit in mediating multiple sensoryeffects of ATP. J. Physiol. 567, 621–639.

Devor, M., Wall, P.D., 1990. Cross-excitation in dorsal root gangliaof nerve-injured and intact rats. J. Neurophysiol. 64,1733–1746.

22 B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

Fanchaouy, M., Serir, K., Meister, J.J., et al., 2005. Intercellularcommunication: role of gap junctions in establishing thepattern of ATP-elicited Ca2+ oscillations and Ca2+-dependentcurrents in freshly isolated aortic smooth muscle cells.Cell Calcium 37, 25–34.

Fewtrell, C., 1993. Ca2+ oscillations in non-excitable cells. Annu.Rev. Physiol. 55, 427–454.

Fields, R.D., 2006. Nerve impulses regulate myelinationthrough purinergic signalling. Novartis Found. Symp. 276,148–158.

Gallagher, C.J., Salter, M.W., 2003. Differential properties ofastrocyte calciumwavesmediated by P2Y1 and P2Y2 receptors.J. Neurosci. 23, 6728–6739.

Guidotti, G., 1996. ATP transport and ABC proteins. Chem. Biol. 3,703–706 Review.

Guthrie, P.B., Knappenberger, J., Segal, M., et al., 1999. ATP releasedfrom astrocytes mediates glial calcium waves. J. Neurosci. 19,520–528.

Hille, B., 1994. Modulation of ion-channel function byG-protein-coupled receptors. Trends Neurosci. 17, 531–536.Review.

Huang, L.Y., Neher, E., 1996. Ca2+-dependent exocytosis inthe somata of dorsal root ganglion neurons. Neuron 17,135–145.

Huang, T.Y., Cherkas, P.S., Rosenthal, D.W., et al., 2005. Dyecoupling among satellite glial cells in mammalian dorsal rootganglia. Brain Res. 1036, 42–49.

Kapur, A., Yeckel, M., Johnston, D., 2001. Hippocampal mossy fiberactivity evokes Ca2+ release in CA3 pyramidal neurons via ametabotropic glutamate receptor pathway. Neuroscience 107,59–69.

Kasai, H., Augustine, G.J., 1990. Cytosolic Ca2+ gradients triggeringunidirectional fluid secretion from exocrine pancreas. Nature348, 735–738.

Katz, B., 1969. The release of neural transmitter substances.Thomas, Springfield, Illinois.

Kawano, S., Otsu, K., Kuruma, A., et al., 2006. ATP autocrine/paracrine signaling induces calcium oscillations and NFATactivation in humanmesenchymal stem cells. Cell Calcium 39,313–324.

Keizer, J., Li, Y.X., Stojilkovic, S., et al., 1995. InsP3-induced Ca2+

excitability of the endoplasmic reticulum. Mol. Biol. Cell 6,945–951.

Koizumi, S., Fujishita, K., Inoue, K., et al., 2004. Ca2+ waves inkeratinocytes are transmitted to sensory neurons: theinvolvement of extracellular ATP and P2Y2 receptor activation.Biochem. J. 380, 329–338.

Lieberman, A.R., 1976. Sensory ganglia. In: Landon, DN (Ed.),The peripheral nerve. Chapman and Hall, London, p. 188.

Lovinger, D.M., Weight, F.F., 1988. Glutamate induces adepolarization of adult rat dorsal root ganglion neuronsthat is mediated predominantly by NMDA receptors. Neurosci.Lett. 94, 314–320.

Luo, J., Yin, G.F., Gu, Y.Z., et al., 2006. Characterization ofthree types of ATP-activated current in relation to P2Xsubunits in rat trigeminal ganglion neurons. Brain Res. 1115,9–15.

Marvizon, J.C., McRoberts, J.A., Ennes, H.S., et al., 2002.Two N-methyl-D-aspartate receptors in rat dorsal root gangliawith different subunit composition and localization. J. Comp.Neurol. 446, 325–341.

Moriyama, T., Iida, T., Kobayashi, K., et al., 2003. Possibleinvolvement of P2Y2 metabotropic receptors in ATP-inducedtransient receptor potential vanilloid receptor 1-mediatedthermal hypersensitivity. J. Neurosci. 23, 6058–6062.

Nakamura, T., Barbara, J.G., Nakamura, K., et al., 1999. Synergisticrelease of Ca2+ from IP3-sensitive stores evoked by synapticactivation of mGluRs paired with backpropagating actionpotentials. Neuron. 24, 727–737.

Nathanson, M.H., Padfield, P.J., O'Sullivan, A.J., et al., 1992.Mechanism of Ca2+ wave propagation in pancreatic acinarcells. J. Biol. Chem. 267, 18118–18121.

Ouyang, K., Zheng, H., Qin, X., et al., 2005. Ca2+ sparks andsecretion in dorsal root ganglion neurons. Proc. Natl. Acad. Sci.U. S. A. 102, 12259–12264.

Ouyang, K., Wu, C., Cheng, H., 2005. Ca2+-induced Ca2+ release insensory neurons: low gain amplification confers intrinsicstability. J. Biol. Chem. 280, 15898–15902.

Piazza, V., Ciubotaru, C.D., Gale, J.E., et al., 2007. Purinergicsignalling and intercellular Ca2+ wave propagation in the organof Corti. Cell. Calcium. 41, 77–86.

Rooney, T.A., Sass, E.J., Thomas, A.P., 1990. Agonist-inducedcytosolic calcium oscillations originate from a specific locus insingle hepatocytes. J. Biol. Chem. 265, 10792–10796.

Sato, K., Kiyama, H., Park, H.T., et al., 1993. AMPA, KA and NMDAreceptors are expressed in the rat DRG neurones. Neuroreport4, 1263–1265.

Schwartz, E.A., 1987. Depolarization without calcium can releasegamma-aminobutyric acid from a retinal neuron. Science 238,350–355.

Seino, D., Tokunaga, A., Tachibana, T., et al., 2006. The role of ERKsignaling and the P2X receptor on mechanical pain evoked bymovement of inflamed knee joint. Pain 123, 193–203.

Shinder, V., Devor, M., 1994. Structural basis of neuron-to-neuroncross-excitation in dorsal root ganglia. J. Neurocytol. 23,515–531.

Shinder, V., Amir, R., Devor, M., 1998. Cross-excitation in dorsalroot ganglia does not depend on close cell-to-cell apposition.Neuroreport 9, 3997–4000.

Simon, N.J., Chi, U.C., Per, U., et al., 2005. Signaling microdomainsregulate inositol 1,4,5-trisphosphate-mediated intracellularcalcium transients in cultured neurons. J. Neurosci. 25,2853–2864.

Stevens, B., Ishibashi, T., Chen, J.F., et al., 2004. an activity-dependent axonal signal regulating MAP kinase andproliferation in developing Schwann cells. Neuron Glia Biol. 1,23–34.

Stefan, I.M., Zoltán, C., Martin, F.S., 2000. Origin sites of calciumrelease and calcium oscillations in frog sympathetic neurons.J. Neurosci. 20, 9059–9070.

Suzuki, Y., Yoshitomo, N.K., Maruyama, K., et al., 1997.Construction and characterization of a full length-enrichedand a 5′-end-enriched cDNA library. Gene 200, 149–156.

Taylor, C.W., Richardson, A., 1991. Structure and function ofinositol trisphosphate receptors. Pharmacol. Ther. 51,97–137.

Tominaga, M., Wada, M., Masu, M., 2001. Potentiation of capsaicinreceptor activity by metabotropic ATP receptors as a possiblemechanism for ATP-evoked pain and hyperalgesia. Proc. Natl.Acad. Sci. U. S. A. 98, 6951–6956.

Toescu, E.C., O'Neill, S.C., Petersen, O.H., et al., 1992. Caffeineinhibits the agonist-evoked cytosolic Ca2+ signal in mousepancreatic acinar cells by blocking inositol trisphosphateproduction. J. Biol. Chem. 267, 23467–23470.

Tse, F.W., Tse, A., Hille, B., 1994. Cyclic Ca2+ changesin intracellular stores of gonadotropes duringgonadotropin-releasing hormone-stimulated Ca2+oscillations. Proc. Natl. Acad. Sci. U. S. A. 91, 9750–9754.

Tsien, R.W., Tsien, R.Y., 1990. Calcium channels, stores, andoscillations. Annu. Rev. Cell Biol. 6, 715–760.

Utzschneider, D., Kocsis, J., Devor, M., 1992. Mutual excitationamong dorsal root ganglion neurons in the rat. Neurosci. Lett.146, 53–56.

Wang, Z., Haydon, P.G., Yeung, E.S., 2000. Direct observation ofcalcium-independent intercellular ATP signaling in astrocytes.Anal. Chem. 72, 2001–2007.

White, T.D., Leslie, R.A., 1982. Depolarization-induced release ofadenosine 5′-triphosphate from isolated varicosities derived

23B R A I N R E S E A R C H 1 2 3 9 ( 2 0 0 8 ) 1 2 – 2 3

from the myenteric plexus of the guinea pig small intestine.J. Neurosci. 2, 206–215.

Wu, Z.Z., Guan, B.C., Li, Z.W., et al., 2004. Sustained potentiationby substance P of NMDA-activated current in rat primarysensory neurons. Brain Res. 1010, 117–126.

Wu, D.M., Minami, M., Kawamura, H., et al., 2006. Electrotonictransmission within pericyte-containing retinal microvessels.Microcirculation. 2006 13, 353–363.

Wynn, G., Ma, B., Ruan, H.Z., et al., 2004. Purinergic componentof mechanosensory transduction is increased in a ratmodel of colitis. Am. J. Physiol. Gastrointest. Liver Physiol. 287,647–657.

Yang, Y.L., Yao, K.H., Gu, Y.Z., et al., 2003. Three kinds of current inresponse to substance P in bullfrog DRG neurons. Brain Res.981, 70–77.

Yoshida, H., Marunaka, Y., Nakahari, T., 2003. [Ca2+]i oscillationsinduced by high [K+]o in acetylcholine-stimulated rat

submandibular acinar cells: regulation by depolarization,cAMP and pertussis toxin. Exp. Physiol. 88, 369–379.

Zarei, M.M., Toro, B., McCleskey, E.W., 2004. Purinergic synapsesformed between rat sensory neurons in primary culture.Neuroscience 126, 195–201.

Zhang, C., Zhou, Z., 2002. Ca2+-independent butvoltage-dependent secretion in mammalian dorsal rootganglion neurons. Nat. Neurosci. 5, 425–430.

Zhang, X., Aman, K., Hokfelt, T., 1995. Secretory pathways ofneuropeptides in rat lumbar dorsal root ganglion neuronsand effects of peripheral axotomY. J. Comp. Neurol. 352, 481–500.

Zhang, C., Xiong, W., Zheng, H., et al., 2004. Calcium-anddynamin-independent endocytosis in dorsal root ganglionneurons. Neuron 42, 225–236.

Zhang, X., Chen, Y., Wang, C., Huang, L.-Y.M., 2007. Neuronalsomatic ATP release triggers neuron–satellite glial cellcommunication in dorsal root ganglia. PNAS 104, 9864–9869.