Stefania Ceruti a,1, Marta Fumagalli a,1, Giovanni Villa a,Claudia Verderio b,1, Maria P. Abbracchio a,∗,1
a Laboratory of Molecular and Cellular Pharmacology of Purinergic Transmission, Department of Pharmacological Sciences,University of Milan, via Balzaretti 9, 20133 Milan, Italy
b CNR Institute of Neuroscience and Department of Pharmacology, University of Milan, via Vanvitelli 32, 20129 Milan, Italy
Received 18 April 2007; received in revised form 23 August 2007; accepted 4 October 2007Available online 26 November 2007
bstract
Receptors for extracellular nucleotides (the P2X-calcium channels and the phospholipase C-coupled P2Y receptors) play key roles in painignaling, but little is known on their function in trigeminal ganglia, whose hyperactivation leads to the development of migraine pain. Heree characterize calcium signaling via P2X3 and P2Y receptors in primary mouse neuron-glia trigeminal cultures. Comparison with intactanglion showed that, in dissociated cultures, sensory neurons retain, at least in part, their physical relationships with satellite glia. RT-PCRndicated expression of P2X2/P2X3 (confirmed by immunocytochemistry) and of all cloned P2Y receptors. Single-cell calcium imagingith subtype-selective P2-agonists/antagonists revealed presence of functional neuronal P2X3, as well as of ADP-sensitive P2Y1,12,13 andTP-activated P2Y2/P2Y4 receptors on both neurons and glia. Calcium responses were much higher in glia, that also responded to UDP,
uggesting functional P2Y6 receptors. To study whether trigeminal ganglia P2 receptors are modulated upon treatment with pro-inflammatory
gents, cultures were acutely (up to 3 min) or chronically (24 h) exposed to bradykinin. This resulted in potentiation of algogenic P2X3
eceptor-mediated calcium responses followed by their down-regulation at 24 h. At this exposure time, P2Y receptors responses in satellitelia were instead upregulated, suggesting a complex modulation of P2 receptors in pain signaling.
Signaling via extracellular adenine (ATP, ADP) and uracilucleotides (UTP, UDP and UDP-glucose) has long beenssociated with sensory systems, where these moleculesct as co-transmitters and/or neuromodulators [1] throughctivation of ionotropic P2X and G-protein-coupled P2Yeceptors [2]. Stimulation of P2X receptors induces intra-ellular calcium responses by promoting calcium entry from
he extracellular space [3]. Seven distinct P2X receptors haveeen characterized (the P2X1–7 receptors) [3], and functionaltudies have also highlighted the existence of heteromeric
omplexes (e.g., P2X2/3) [4,5]. Concerning P2Y receptors,ight distinct subtypes are currently recognized, which areubdivided into two distinct subgroups. The first encom-asses the P2Y1, P2Y2, P2Y4, P2Y6 and P2Y11 receptors,ainly coupled to Gq proteins, activation of phospholipase-(PLC) and release of calcium from intracellular stores,
hereas the other one includes the P2Y12, P2Y13 and P2Y14eceptors coupled to Gi and, in the case of P2Y12 and P2Y13,o both inhibition of intracellular cAMP and PLC stimu-ation [6]. While P2X receptors only respond to adenineucleotides, some P2Y receptors can also or exclusivelyespond to uracil nucleotides, such as UTP, UDP and UDP-
lucose [3,6].
Among other actions, a specific role of extracellularucleotides as main algogenic transmitters has been recog-ized [7–9]. Application of ATP to the skin strongly evokes
ain sensations [10], and lowers the threshold temperature forctivation of capsaicin-sensitive TRPV1 receptors, suggest-ng that, in the presence of ATP, these receptors can triggerhe pain sensation at the normal body temperature [10].xtracellular ATP concentrations are dramatically increasedfter tissue trauma, inflammation, visceral distension andigraine, and the resulting P2 receptor activation is likely to
ontribute to and to modulate the consequent pain sensations11]. On this basis, the therapeutic application of a pharma-ological manipulation of purinergic system in various painonditions has been anticipated [1,7].
It is now widely accepted that the P2X3 and the P2X2/3hannels, which are expressed on peripheral sensitive nervess well as on nociceptive neurons in the CNS and in sensoryanglia, play a fundamental role in transducing ATP-drivenainful signals [1,11,12]. However, evidence is emerging thatther calcium-linked P2 receptors (e.g., metabotropic PLC-oupled P2Y receptors) may also participate to pain [10].hus, there is currently a growing interest in the characteriza-
ion of calcium signaling via specific P2Y-receptor subtypesnder physiological conditions and in the establishment andaintenance of pathological neuropathic pain states (e.g.,
llodynia and migraine).Most of the currently available information on P2 recep-
ors in pain transmission is based on results obtained inorsal root ganglia (DRG) [13,14], or in the spinal cord [15].ery little is known on purinergic transmission in trigemi-al ganglion (TG), an important sensory station integratingociceptive stimuli from most of the head and facial dis-ricts and involved in migraine, a form of neuropathic pain.unctional P2X3 and P2X2/3 receptors have been found onsubpopulation of small and medium diameter neurons in
oth DRG and TG [16–19], while P2X3 immunoreactivityn larger peptidergic neurons occurs exclusively in the latter20]. Discordant data have been reported for P2X1: posi-ive immunostaining has been recently demonstrated in TGeurons [21], at variance from previous work showing neg-tive results, despite the presence of a low level of mRNA17,22]. Concerning G protein-coupled receptors, the pres-nce of P2Y1,2,4,6 has been demonstrated both in DRG andG neurons [14,23,24]. However, a definitive proof of their
unctionality in TG is still lacking, while (with the only excep-ion of P2Y6 [25]) these receptors have been found to beunctional in DRG [10,26,27]. Moreover, very few data arevailable regarding the presence and function of P2 receptorsn TG glial cells (see also below).
As mentioned above, TG is involved in migraine, a com-on, life-long disease affecting more than 10% of population
28]. Anatomically, the core of TG is represented by sensoryeurons that are surrounded by satellite glial cells (SCGs), aype of glia typical of sensory ganglia, which wraps aroundeurons forming a distinct and maybe functional morpholog-
cal unit [29]. While it is generally recognized that migraineain is triggered by the activation of sensitive fibers ofG neurons innervating meningeal blood vessels, emerg-
ng data suggest that SGCs may also contribute [29,30].
mt((
43 (2008) 576–590 577
ctivation of the trigeminovascular system indeed deter-ines the release of vasoactive neuropeptides, e.g., calcitonin
ene-related peptide (CGRP), substance P, as well as ofro-inflammatory agents and transmitters (e.g., serotonin,radykinin, prostaglandins and ATP) that contribute to neu-ogenic inflammation [28] acting on both neurons and SGCsn TG. To date, most of the studies on P2 receptors in TG haveocused on neurons, and only limited information is availableor SGCs. Calcium imaging studies in intact TG suggestedhat only P2Y1,2,4 (but not P2X1,3 [31]) were functional inGCs, and no information is currently available on other P2Yeceptors in glial cells from sensory ganglia. Moreover, someata are available on the effect of nociceptive agents (e.g.,GRP) on the functionality of algogenic P2X3 on trigeminalanglia neurons [32], but very little is known on the possi-le alterations of neuronal P2Y receptors after exposure tooxious stimuli.
On this basis, here we have evaluated the presence andunction of P2 receptors in mouse TG primary cultures, aodel system allowing the investigation of the structural
nd functional properties of TG neurons and SGCs. Weave utilized single-cell calcium imaging to analyze func-ional responses to subtype-selective P2 ligands. Based onaucity of data on the role of glial P2 receptors in pain signal-ng, specific attention has been given to the characterizationf P2Y receptors in SGCs. We have also investigated thehanges of P2 receptors-mediated calcium transients in TGultures after either acute (3 min) or long-term (24 h) treat-ent with bradykinin, a pro-inflammatory algogen believed
o act through CGRP and PGE2 release [33].
. Methods
.1. Cell cultures and pharmacological treatments
Primary cultures from TG of P11-12 C57-Black/6 miceCharles River Lab, Calco, Italy) were prepared, as previouslyescribed [34]. Briefly, after decapitation TG were rapidlyxcised and dissociated in 0.25 mg/ml trypsin, 1 mg/ml col-agenase and 0.2 mg/ml DNAse (Sigma–Aldrich, Milan,taly) in F12 medium (Invitrogen, Gibco, Italy) at 37 ◦C.nzymes were then inactivated by adding 10% fetal calferum and 0.125 mg/ml trypsin inhibitor (Sigma–Aldrich).ells were centrifuged at 1000 rpm for 5 min, resus-ended in F12 medium + 10% fetal calf serum and platednto poly-l-lysine-coated 24-mm diameter glass cover-lips. Cells were maintained in culture for 48 h. In somepecific experiments, after 24 h of culture, cultures werereated with 100 nM bradykinin (BK; Sigma–Aldrich), and
aintained at 37 ◦C for additional 24 h before applica-ion of various P2 agonists and antagonists. For calcium
easurements, the following pharmacological agents wereested: ��-methylene-ATP (��-meATP; 100 �M), ADP10 and 100 �M), UTP (10 and 100 �M), UDP-glucose100 �M), N6-methyl-2-deoxyadenosine 3′,5′-bisphosphate
5 Calcium
(2oa
2
pptcitiAIUBpr4stcorigoPtw(uJHtusom
2
r6(l7sn([�
CpaSb5b
2
est
2
Fb1Htsiwcauimra
2
(RrodwsaGTp
2
78 S. Ceruti et al. / Cell
MRS2179; 100 �M), Cangrelor (10 �M), Reactive Blue-(RB-2, 100 �M), Suramin (100 �M). All reagents were
btained from Sigma–Aldrich, except for Cangrelor that waskind gift of The Medicines Company, Parsippany, NJ, USA.
.2. Immunocytochemistry
Primary trigeminal cultures were fixed at room tem-erature for 25 min with 4% paraformaldehyde in 0.1 Mhosphate-buffered saline (PBS; Euroclone, Celbio) con-aining 0.12 M sucrose. To reduce non-specific staining,ells were incubated for 45 min in 0.25% albumin dilutedn PBS with 0.1% Triton X-100, before incubation withhe primary antibody. The following primary antibod-es were utilized: rabbit anti-P2X3 (1:250 o/n at 4 ◦C;lomone Labs, Jerusalem, Israel), mouse anti-�-Tubulin-
II (�-TubIII; 1:100 o/n at 4 ◦C; Chemicon, Temecula, CA,SA), rabbit anti-�-TubIII (1:1000 o/n at 4 ◦C; Covance,erkeley, CA, USA), mouse anti-2′,3′-cyclic-nucleotide 3′-hosphodiesterase (CNPase, 1:100, 2.5 h RT; Chemicon),abbit anti-glial fibrillary acidic protein (GFAP, 1:300 o/n at◦C; Dako, Glostrup, Denmark), and rabbit anti-glutamine
ynthetase (GS, 1:100, 2.5 h RT; Novus Biologicals, Little-on, CO, USA). Bandeiraea simplicifolia isolectin B4 directlyonjugated to fluorescein isothiocyanate (FITC-IB4, 1:100/n at 4 ◦C; Sigma–Aldrich) was also utilized. Cells wereinsed three times with PBS with 0.1% Triton X-100, and thenncubated for 1 h at RT with the secondary goat anti-rabbit andoat anti-mouse antibodies, conjugated to AlexaFluor®488r AlexaFluor®594 (1:600; Molecular Probes, Invitrogen).rimary and secondary antibodies were diluted in PBS con-
aining 0.25% albumin and 0.1% Triton X-100. Coverslipsere finally mounted with Permafluor mounting medium
Shandon Italscientifica, Genova, Italy), and analyzed bysing an inverted fluorescence microscope (200M; Zeiss,ena, Germany) equipped with a CCD camera (AxioCamRm; Zeiss), connected to a PC computer equipped with
he software Axiovision (Zeiss). This software also enableds to measure neuron diameters. Control experiments of non-pecific staining were performed without primary antibodyr, for P2X3 immunoreactivity, by preadsorption of the pri-ary antibody with the immunizing peptide.
.3. Immunohistochemistry
The excised trigeminal ganglia were fixed in Carnoyeagent (absolute ethanol:chloroform:glacial acetic acid,:3:1; Merck, Milan, Italy), embedded in ParaplastSigma–Aldrich), and cut in the horizontal plane along theong axis of the ganglion on a microtome at a thickness of�m. After dewaxing and rehydration, paraffin-embedded
ections were incubated for 45 min in PBS containing 10%
ormal goat serum (Sigma–Aldrich) and 0.1% Triton X-100Sigma–Aldrich) to prevent non-specific antibody binding35]. Sections were then double-stained with rabbit anti--tubIII (1:1000), and mouse anti-GS antibody (1:250;
Wala
43 (2008) 576–590
hemicon) or mouse anti-CNPase (1:100) o/n at 4 ◦C. Toerform co-immunostaining with CNPase, rabbit anti-GSntibody (1:100, 2.5 h at RT), Novus Biological) was used.ections were then rinsed three times with PBS, and incu-ated for 1 h RT with the AlexaFluor® 488- or AlexaFluor®
94-conjugated secondary antibodies (see above). All anti-odies were diluted in PBS containing 0.1% Triton X-100.
.4. Confocal microscopy
Labelled primary cultures or trigeminal sections werexamined with a Bio-Rad Radiance 2100 confocal micro-cope (Bio-Rad, Milan, Italy); images were acquired usinghe LaserSharp 2000 software [35].
.5. Intracellular calcium measurements
Cultures were loaded for 45 min at 37 ◦C with 2 �Mura-2 pentacetoxy methylester in Krebs-Ringer solutionuffered with HEPES (KRH; 125 mM NaCl, 5 mM KCl,.2 mM MgSO4, 2 mM CaC12, 10 mM glucose, and 25 mMEPES/NaOH, pH 7.4), as previously described [36], and
ransferred to the recording chamber of an inverted micro-cope (Axiovert 100TV; Zeiss) equipped with a calciummaging unit. Polychrome IV (TILL Photonics, Germany)as used as light source. Fura-2 fluorescence images were
ollected with a CCD camera Imago-QE (Till Photonics),nd analyzed with the Tillvision 4.0.1 software. Images weresually acquired at 1–2 ratio/s. The rate of acquisition wasncreased up to 8 ratios/s upon cell stimulation with ��-
eATP. Neurons were distinguished from glial cells by theiresponsiveness to 50 mM KCl. The total number of cellsnalyzed for any given condition is indicated as n.
.6. Total RNA isolation and PCR analysis
Total RNA was extracted using the TRIZOL® reagentInvitrogen) according to the manufacturer’s instructions.etrotranscription to cDNA and PCR reactions were car-
ied out as previously described [36]. For a complete listf the P2X and P2Y-receptors primers utilized, see [36]. Toetect the expression of mouse P2Y4 the following primersere used: sense 5′CTTTGGCTTTCCCTTCTTGA3′, anti-
ense 5′GTCCGCCCACCTGCTGATGC3′. c-DNA was alsomplified with primers for �-Tub-III: sense 5′CTTCCAGCT-ACACACTCAC3′, antisense 5′AGACACAAGGTGG-TGAGGT3′. Primers were designed using the Oligo 4.0rimer analysis software.
.7. Western blotting analysis
Whole-cell lysates were prepared and analyzed by
estern blotting as previously described [36]. Briefly,
pproximately 20 �g aliquots from each protein sample wereoaded on 11% sodium-dodecylsulphate polyacrilamide gels,nd blotted onto nitrocellulose membranes (Bio-Rad Labora-
Calcium
td1tt(aC3(tfioprbMwo
2
igSM*
3
3m
cfi(cp1s�(awc51am(
wp
tScapiwitd
asdC(
wIaDsdaCe
wSc(po
3c
tvaticitKrSurc
S. Ceruti et al. / Cell
ories, Milan, Italy). Filters were saturated with 10% non-fatry milk in Tris-buffered saline (TBS) (1 mM Tris–HCl,5 mM NaCl, final concentrations, pH 8) for 1 h at roomemperature, and then incubated overnight at 4 ◦C withhe following primary antibodies: mouse anti-GS antibody1:250 in 5% non-fat dry milk in TBS; Chemicon), rabbitnti-CNPase (1:200 in 5% non-fat dry milk in TBS; Santaruz Biotechnologies), rabbit anti-�-actin (�-act; 1:750 in% non-fat dry milk in TBS; Sigma), mouse anti-�-TubIII1:4000 in 5% non-fat dry milk in TBS; Promega). Fil-ers were then washed in TBS-T (TBS plus 0.1% Tween,nal concentration), incubated for 1 h with goat anti-rabbitr anti-mouse secondary antibody conjugated to horseradisheroxidase (1:4000 or 1:2000 in 5% non-fat dry milk in TBS,espectively; Sigma). Detection of proteins was performedy enhanced chemiluminescence (Amersham Biosciences,ilan, Italy) and autoradiography. Specificity of GS signalas tested by performing immunoreactions in the presencef the secondary antibody alone.
.8. Statistical analysis
All results are expressed as mean ±S.E.M. of at least threendependent experiments. Statistical significance betweenroups was derived from one-way ANOVA followed bycheffe’s analysis performed with the StatView software foracIntosh. Two degrees of significance were considered:
P < 0.05 or **P < 0.01.
. Results
.1. Primary mouse trigeminal cultures as in vitroodel of trigeminal ganglia
Fig. 1A shows the typical aspect of mouse trigeminalultures 48 h after plating, consisting of neurons, identi-ed by staining with the neuronal marker �-tubulin-III�-tubIII; Fig. 1B), surrounded by glial-like cells. Severalultured neurons displayed the typical pseudo-unipolar mor-hology of sensory neurons (Fig. 1C), and accounted for the1.1 ± 0.97% of total cell population (n = 8340; 12 cover-lips from eight independent experiments). 37.3 ± 1.5% of-tubIII-positive neurons were also positive for isolectin B4
IB4), a widely accepted marker of sensory neurons (Fig. 1Dnd E; see also below). As shown in Fig. 1F and in lineith previous data [23,34], 41.5 ± 5.8% of total neuronal
ells were small neurons with a diameter lower than 15 �m,3.4 ± 4.1% were medium neurons with a diameter between5 and 25 �m and only 5.1 ± 0.7% were large neurons withdiameter greater than 25 �m. About 40% of small- andedium-size but no large-size neurons were IB4-positive
Fig. 1F).As mentioned above, several cultured neurons sat on and
ere wrapped by glial cells, likely belonging to the SGCsopulation (see Section 1). A widely accepted marker for
0
mT
43 (2008) 576–590 579
hese cells is glutamine synthetase (GS) [29]. However, sinceGCs also share common characteristics with oligodendro-ytes, to identify SGCs, cultures were double-stained withn anti-GS antibody and an anti-2′,3′-cyclic nucleotide 3′-hosphodiesterase (CNPase) antibody. Fig. 1G–I shows thatn our experimental model CNPase/GS double-positive cellsere found wrapping around �-tubIII-positive neurons (as
dentified in Fig. 1J–L). These data demonstrate for the firstime that CNPase can be utilized to identify not only oligo-endrocytes but also at least a subpopulation of SGCs.
Immunostaining with antibodies directed against thestroglial marker, glial fibrillary acidic protein (GFAP)howed that very few positive astrocytic-like cells wereetected in TG cultures (Fig. 1M–O). Of interest, GFAP- andNPase-immunoreactivity never colocalized in TG cultures
Fig. 1O; see also below).In sections from intact TG ganglia, GS-positive SGCs
ere localized around neurons (Fig. 2A), as expected [29].nterestingly, also CNPase immunoreactivity was distributedround �-tubIII-positive cells, as shown in Fig. 2B. Panels–F show a clear colocalization of these two markers in a
ubpopulation of SGCs, thus confirming the data obtained inissociated cultures (see above). Moreover, Western blottingnalysis demonstrated that significant expression of GS andNPase was preserved in dissociated cultures compared toxpression levels in intact TG ganglia (Fig. 2C).
The comparable expression of these markers, togetherith the preservation of clusters of neurons surrounded byGCs in culture (Fig. 1G–L) resembling the morphologi-al and probably functional unit observed in intact gangliaFig. 2A and B and D and F), validate the use of dissociatedrimary cultures as an adequate model to study the functionf neurons and SGCs in TG.
.2. Both neurons and satellite cells in mouse trigeminalultures bear P2 receptors
Previous studies have highlighted a role for ATP in cell-o-cell signaling in intact TG and indicated the presence ofarious P2 receptors on both neurons and glia [11,31]. Tossess if functional responses to ATP are indeed preserved inrigeminal cultures, we tested the ability of this nucleotide tonduce [Ca2+]i increases after loading cells with the fluores-ent calcium dye Fura-2 (Fig. 3A–D). ATP induced transientncreases of [Ca2+]i both in glial cells and neurons (Fig. 3C),he latter identified by a subsequent challenge with 50 mMCl (Fig. 3D and E). While 23 ± 6.6% (n = 119) of neurons
esponded to ATP with [Ca2+]i increases, the vast majority ofGCs (89 ± 4.7%, n = 146) showed clear calcium transientspon nucleotide application (Fig. 3F). The mean calciumesponse to ATP, evaluated as increase in 340/380 fluores-ence ratio (�F340/380), was 0.34 ± 0.04 in neurons and
.71 ± 0.04 in SGCs (Fig. 3G).
Among P2 receptors, P2X3 undoubtedly represents theost characterized subtype in TG sensory neurons [34,37].o assess the presence and function of this receptor in our
580 S. Ceruti et al. / Cell Calcium 43 (2008) 576–590
Fig. 1. Immunocytochemical characterization of mouse trigeminal cultures: (A) representative brightfield showing a cluster of neurons on a layer of SGCs;(B) same field as in (A) after staining with the neuronal specific marker �-tubIII; (C) a typical pseudounipolar �-tubIII-positive sensory neuron; (D) confocalmicroscopy image of trigeminal sensory neurons labeled with fluorescein-conjugated isolectin B4 (FITC–IB4); (E) same field as in (D) after double stainingwith FITC–IB4 and an anti-�-tubIII antibody, demonstrating IB4 staining and colocalization with �-tubIII only in medium/small size neurons; (F) histogramsshowing the distribution of �-tubIII-positive neurons according to the size (diameter) of their soma (indicated as tot, black columns). The size distribution ofIB4-positive neurons was also calculated and expressed as percentage of the total neuronal population (grey columns). The classification proposed by Simonettiet al. [34] was utilized to rank cells according to their diameter (small neurons: diameter <15 �m; medium neurons: diameter between 15 and 25 �m; largeneurons: diameter 25 �m); (G) representative brightfield showing a neuron surrounded by a CNPase-positive SGC that also expresses GS (H and I); (J) exampleof a brightfield showing a neuron sat on and wrapped by a group of SGCs labeled with anti-CNPase antibody; (K) same as in (J) showing �-tubIII staining;(L) same as in (J) and (K) showing merging of CNPase and �-tubIII immunoreactivity; (M) example of a brightfield showing an astrocyte (see white arrow);as confirmed by immunostaining for the typical astrocytic marker GFAP (N); (O) a representative field showing that different populations of glial cells expressCNPase and GFAP. In all micrographs, scale bar: 15 �m.
S. Ceruti et al. / Cell Calcium 43 (2008) 576–590 581
Fig. 2. In situ labeling of neuronal and SGCs in intact mouse trigeminal ganglia: (A) and (B) confocal microscopy images of trigeminal ganglion sectionsstained for the neuronal marker �-tubIII, and GS (A) or CNPase (B). GS- and CNPase-positive cells are found surrounding neurons; (C) Western blottinganalysis of GS and CNPase expression in cultured cells and in trigeminal ganglia. For GS, in the tissue sample, the specifity of the band at the expected molecularweight of 45 kDa was evaluated by incubation with the secondary antibody alone (right panel). Expression of �-tubulin is also shown; �-actin is used as ani f a triga icrograa
eeMnfstWPac�lr50gi
eTtiP
Pbt
3r
nviracot0tUR
nternal loading control; (D–F), GS (D) and CNPase (E) immunostaining ond CNPase in SGCs (arrows) surrounding neurons (indicated with “n”). Mll micrographs, scale bar: 15 �m.
xperimental setting, we performed double immunostainingxperiments with an anti-�-tubIII and an anti-P2X3 antibody.ost neurons (i.e., 73.6 ± 2.7% of �-tubIII-positive cells,= 440) expressed P2X3 (Fig. 4A); this receptor was indeed
ound in almost all (95.3 ± 0.9%, n = 324) IB4-positive sen-ory neurons (Fig. 4B). The presence of P2X3 in mouserigeminal cultures was confirmed by RT-PCR (Fig. 4C), and
estern blot analysis (data not shown). Expression of the2X2 receptor (that can form heterodimers with P2X3) waslso found whereas P2X1 was not expressed (Fig. 4C). Cal-ium transients induced by the P2X1/P2X3 receptor agonist�-meATP (here acting only on the P2X3 subtype, due to the
ack of expression of the P2X1 receptor) indicated that theeceptor was functional in trigeminal neurons (Fig. 4D), with8.0 ± 4.7% of responding neurons (mean calcium increase:.12 ± 0.01, �F340/380, n = 115). P2X3 was never found inlial cells, as assessed by both immunostaining and calciummaging (data not shown, n = 105).
Trigeminal cells have been demonstrated to express sev-ral P2Y receptors [10,23] but their systematic analysis in
G cultures has never been performed. We thus analyzed
he presence of all P2Y receptors cloned from rodent tissuesn both intact ganglia and mouse trigeminal cultures. RT-CR analysis showed that all known P2Y receptors (i.e., the
sriP
eminal ganglion section. Merging (F) clearly shows co-localization of GSphs have been taken at a higher magnification with respect to (A and B). In
2Y1,2,4,6,12,13,14 receptors) were expressed with no apprecia-le differences in their expression profile between the intactissue and the dissociated cultures (Fig. 5).
.3. Characterization of functional calcium-linked P2Yeceptors in neuronal trigeminal cells
While the presence of a functional P2X3 receptor in TGeurons has been previously well documented [34,37,38],ery few data [21] are available concerning the functional-ty of P2Y receptors in these cells. We thus analyzed theesponse of TG neurons to the most commonly utilized P2Ygonists by single-cell calcium imaging. A significant per-entage of neurons responded to either ADP (8.5 ± 3.9%f total, n = 142; Fig. 6A–C) or UTP (13.0 ± 4.1% ofotal, n = 142; Fig. 6C), with a mean calcium response of.44 ± 0.06 �F340/380 and 0.34 ± 0.09 �F340/380, respec-ively (Fig. 6D and E). No cells showed responses to eitherDP or UDP-glucose (Fig. 6C, n = 83 and 43, respectively).esponse to ADP is likely mediated by P2Y1, as demon-
trated by the almost complete blockade exerted by the P2Y1eceptor antagonist MRS2179 (Fig. 6F and G). In some exper-ments, ADP response was also partially antagonized by the2Y12/P2Y13 antagonist Cangrelor, suggesting that some
582 S. Ceruti et al. / Cell Calcium 43 (2008) 576–590
Fig. 3. ATP induces calcium increases in primary cultures from mouse trigeminal ganglia: (A–D) series of pseudocolor images of the same Fura-2-loaded cells(shown in (A)) at the 380 nm fluorescence) under basal conditions (F340/380, (B)) and after application of 100 �M ATP (at peak [Ca2+]i response, (C)), followedby 50 mM KCl (D), here utilized as a depolarizing agent to identify neurons. Graded colors from blue to green and yellow indicate increasing F340/380; scalebar: 15 �m; (E) representative temporal plot of [Ca2+]i changes recorded from two cells identified in (A–D) (blue circle: neuron, blue trace in (E) red circle:SGC, red trace in (E)), and stimulated with 100 �M ATP followed by 50 mM KCl; (F) Histograms showing the percentage of ATP-responding cells (from at ed calco to color
nw
3r
Pr(a
aniiUlU
otal of 119 neurons and 146 satellite cells); (G) quantification of ATP-inducbtained in 10 independent experiments.(For interpretation of the references
eurons may also express these receptor subtypes togetherith P2Y1 (data not shown).
.4. Characterization of functional calcium-linked P2Yeceptors in trigeminal SGCs
In a similar way, we characterized the functionality of
2Y receptors in TG SGCs. Application of P2Y agonistsesulted in [Ca2+]i increases in most SGCs with 78.0 ± 5.2%n = 320) and 83.0 ± 4% (n = 331) of cells responding to ADPnd UTP, respectively (mean calcium increases: 0.73 ± 0.03
ngi(
ium increases in neurons and SGCs, expressed as mean ±S.E.M. of resultsin this figure legend, the reader is referred to the web version of the article.)
nd 0.8 ± 0.3). Only a minor percentage of cells (20 ± 6%,= 237) instead responded to UDP, with a mean calcium
ncrease of 0.4 ± 0.06 (Fig. 7A and B). Examples of agonist-nduced calcium transients are reported in Fig. 7C; 100 �MTP-evoked calcium response was more sustained and pro-
onged with respect to the same concentration of ADP andDP, this difference being less evident at 10 �M UTP (data
ot shown). Experiments performed with MRS2179 and Can-relor showed that, in glial cells, response to ADP mainlynvolves P2Y1 with a smaller contribution by P2Y12/P2Y13Fig. 7D). Suramin and RB2 inhibited UTP-induced [Ca2+]i
S. Ceruti et al. / Cell Calcium 43 (2008) 576–590 583
Fig. 4. Expression of P2X3 receptors and coupling to intracellular calcium increases in TG neurons: (A) double immunostaining of cultures with anti-�-tubIII andanti-P2X3 receptor antibodies, showing that most �-tubIII-positive cells also express P2X3 (see text for details); (B) representative confocal image showing doublestaining of cultures with an anti-P2X3 receptor antibody and FITC–IB4; most IB4-positive sensory neurons express P2X3 (see text for details). Scale bars: 15 �m;(C) RT-PCR analysis showing expression of P2X2 and P2X3 receptors in trigeminal cultures. No signal was detected for P2X1 receptor. No amplification productswere detected in RNA samples that were not subjected to retrotranscription (indicated as –, RT). �-tubIII was used as an internal control for RT-PCR amplification;( minal nf figure
ti
3t
a[teie1Tras
u0iw(
tranr(mB
D) representative temporal plot of [Ca2+]i increases recorded from two trigeollowed by 50 mM KCl.(For interpretation of the references to color in this
ransients by 76.6% and 76.7%, respectively, supporting thenvolvement of P2Y2/P2Y4 receptors (Fig. 7E).
.5. Effect of bradykinin on calcium signaling viarigeminal neuronal P2X3 and glial P2Y receptors
We then studied the effect of the pro-inflammatory medi-tor bradykinin (BK), which is involved in activation of TG33], on the calcium responses mediated by P2X3 recep-ors in neurons and P2Y receptors in glial cells. Acutexposure to 100 nM BK evoked changes in the [Ca2+]in the 35.11 ± 4.72% of neurons (n = 188; 11 independentxperiments), with respect to the 3.76 ± 2.01% (n = 319;1 independent experiments) of SGCs, suggesting that, in
G, BK receptors are almost exclusively functional in neu-
onal cells. In BK-responding neurons, the mean calciummplitude induced by the P2X3 agonist ��-meATP wasignificantly increased to 144.63 ± 0.16% if compared to
ctna
eurons upon stimulation with the P2X1/P2X3 agonist ��-meATP (100 �M),legend, the reader is referred to the web version of the article.)
ntreated controls set to 100% (mean calcium increase:.24 ± 0.03, n = 25, with respect to 0.17 ± 0.01 �F340/380n vehicle-treated cells, n = 48; Fig. 8A and B). No changesere observed in P2Y receptor-mediated calcium increases
data not shown).On the contrary, a 24-h exposure of trigeminal cultures
o 100 nM BK did not modify the amplitude of calciumesponses induced by ��-meATP (Fig. 8C), but resulted inhigh significant reduction of the percentage of responsiveeurons, which was decreased to 28.5 ± 5.7% (n = 125) withespect to 62.1 ± 5.8% in control untreated cultures (n = 118)Fig. 8E). Interestingly, significant changes in P2Y receptor-ediated responses were recorded in SGCs. In particular,K-treated cultures showed a significant increase of the mean
alcium amplitude (expressed as �F340/380 and normalizedo BK-untreated control cells exposed to the same P2 ago-ist set to 1.00) of 1.36 ± 0.11 (n = 60), 1.39 ± 0.07 (n = 119)nd 1.44 ± 0.11 (n = 75) after exposure to ATP, ADP, and
584 S. Ceruti et al. / Cell Calcium 43 (2008) 576–590
Fig. 5. Trigeminal ganglia and cultures express all cloned P2Y receptors. Total RNA isolated from either intact trigeminal ganglia (indicated as “tissue”)o ulture”)s etectedt receptor
Uadarcc
4
tPgnapmrtstsbluSBi
4
Tt
tnrwt
opcf
1
2
3
r from primary trigeminal cultures 48 h after preparation (indicated as “cpecific for each indicated P2Y receptor (see Section 2). No products were dhe housekeeping gene �-actin is shown. No significant differences in P2Y-epresentative experiment out of three is shown.
TP, respectively (Fig. 8D), with no changes in the percent-ge of responding cells (Fig. 8F). A trend to increase, whichid not reach the statistical significance, was also observedfter exposure to UDP (Fig. 8D), whereas the percentage ofesponding SGCs was dramatically increased (Fig. 8F, per-entage of UDP-responding cells with respect to untreatedontrol cells: 253.9 ± 5.9%, n = 265).
. Discussion
In this paper, we have characterized the presence, func-ionality, and cellular localization of the P2X3 channels and2Y metabotropic purinoceptors in primary mixed neuron-lia cultures from TG ganglia. Results show that TG sensoryeurons express functional P2X3 and ADP- and UTP-ctivated metabotropic receptors. Based on RT-PCR andharmacological data, we suggest that the latter responses areediated by P2Y1, P2Y2/4 and, to a lesser extent, P2Y12/13
eceptors. As expected, SGCs do not express the P2X3 recep-or subtype, but are highly responsive to ADP, UTP, and UDP,uggesting the presence of functional P2Y1,2,4,6,12,13 recep-ors. Moreover, exposure to the pro-inflammatory agent BKignificantly modified neuronal P2X3 receptor function in aiphasic way, with sensitisation after an acute exposure fol-owed by a reduction of P2X3-mediated calcium responsespon chronic application. Functionality of P2Y receptors inGCs was also significantly enhanced by chronic exposure toK, suggesting a complex modulation of P2 receptors upon
nflammatory conditions.
.1. Characterization of TG cultures
To characterize calcium signaling via P2 receptors inG, we have taken advantage of in vitro trigeminal cultures
hat have been recently reported as a reliable experimen-
was subjected to RT and the resulting cDNA was amplified with primersin the absence of retrotranscription (indicated as –). Parallel expression of
r expression were detected between freshly isolated tissue and cultures. A
al model to study the purinergic regulation of trigeminaleurons [32,34]. However, these studies only focused on neu-onal P2X3, and no evidence was provided that such a modelould be adequate to also assess the contribution of glial cells
o purinergic signaling in TG.As a first step, we have characterized TG cultures in terms
f presence and structural organization of the different cellopulations. As demonstrated by Figs. 1 and 2, TG primaryultures are a reliable in vitro model of sensory ganglia. Inact:
. Neurons survive 48 h after TG dissociation, display a goodviability, as indicated by their responsiveness to depolar-ization and by MTT assay (data not shown), despite thefact that cells were grown in the absence of exogenouslyadded NGF. Since this well-known pro-survival factor canalso alter pain transmission by acting as a pro-algogenicsubstance and increase the excitability of sensory neu-rons [39], it was not added to cultures. Moreover, most ofthe neurons maintain the typical morphological featuresof TG neurons with a single axon that forms a T-shapebifurcation (Fig. 1C).
. Analysis of frequency distribution of neurons according totheir somatic size (Fig. 1F) shows that small and mediumnociceptive neurons represent almost the totality of theneuronal population, in full agreement with data obtainedin cell cultures from sensory ganglia [40,41], includingTG [34]. Consistent with previous literature data [22], apercentage of these neurons is stained by IB4, a knownmarker of a subpopulation of unmyelinated primary affer-ents nociceptive neurons (Fig. 1D–F) [42].
. Finally, our cell cultures maintain to some extent the
anatomical organization of TG in vivo. In sensory gan-glia, neurons are tightly surrounded and wrapped by SGCs(Fig. 2), which play a fundamental role in controllingneuronal environment and firing [29]. After dissection
S. Ceruti et al. / Cell Calcium 43 (2008) 576–590 585
Fig. 6. Identification of functional P2Y receptors in trigeminal neurons: (A) series of pseudocolor images of Fura-2 loaded cells before (F340/380, basal) andat peak [Ca2+]i responses after the application of 100 �M ADP, followed by 50 mM KCl; (B) identification of neurons was confirmed by immunopositivity to�-tubIII. Scale bars: 15 �m; (C) histograms showing the percentage of neurons responding to the indicated P2Y agonists; (D) histograms showing the mean[Ca2+]i increase, evaluated as changes in the 340/380 fluorescence ratio (�F340/380), induced in neuronal cells by the P2Y agonists ADP and UTP; (E)representative temporal plots of [Ca2+]i increases recorded from neurons upon stimulation with 100 �M either ADP or UTP, as indicated, followed by 50 mMKCl. In the case of ADP, the neuron is the cell identified by a white circle in (A and B). (F) Representative temporal plot of [Ca2+]i changes recorded fromt the P2Y( nses ±SS legend
fi(iesSaa
p(
4s
tiui
wo neurons stimulated with 100 �M ADP before and after application ofG) quantification of data reported in (F) expressed as mean calcium respocheffe’s analysis.(For interpretation of the references to color in this figure
and culturing, neurons still retain, at least in part, theirphysical relationships with SGCs (Fig. 1G–L), suggest-ing that also their functional and modulatory activitiesare preserved.
Glutamine synthase (GS) expression is taken as a use-ul marker to identify SGCs [29], as also confirmedn our experimental model in both dissociated culturesFig. 1H), and intact ganglia (Fig. 2A and D). Interest-ngly, a significant percentage of GS-positive cells alsoxpress CNPase (Figs. 1G–I and 2B and D–F), therefore
uggesting that, besides representing a known marker ofchwann cells, CNPase can be also utilized to identifyt least a subpopulation of SGCs in both immunohisto-nd immunocytochemical studies. Finally, very few GFAP-
Cmni
1 antagonist MRS2179 (100 �M), followed by exposure to 50 mM KCl;.E.M. *P < 0.05 with respect to ADP alone, by one ANOVA followed by
, the reader is referred to the web version of the article.)
ositive cells (i.e., astrocytes) were found in TG culturesFig. 1M–O).
.2. Identification of functional P2 receptors byingle-cell calcium imaging in TG cultures
We have chosen to assess the functionality of the P2 recep-ors by single-cell calcium imaging based on previous studiesn other cellular systems showing their coupling to mod-lation of [Ca2+]i [36,43,44]. Moreover, a number of datandicates that upregulation of Ca2+ currents and enhanced
a2+ entry into sensory cells could contribute to the develop-ent of pathological pain [45]. Interestingly, several kinds of
europathic pain, such as familial hemiplegic migraine, arendeed associated to mutations of P/Q calcium channels [46].
586 S. Ceruti et al. / Cell Calcium 43 (2008) 576–590
Fig. 7. Identification of functional P2Y receptors in trigeminal SGCs: (A and B) histograms showing the percentage of responding SGCs (A) or the mean[Ca2+]i increase (B) induced by the indicated P2Y-receptor agonists; (C) examples of [Ca2+]i changes recorded from SGCs upon stimulation with various P2Yagonists; (D) left: representative temporal plots of [Ca2+]i changes recorded from four SGCs stimulated with 100 �M ADP before and after application of theP2Y1-selective antagonist MRS2179 (100 �M), as indicated (see text for details). Right: quantification of mean calcium increase ±S.E.M. after stimulationwith 100 �M ADP before and after exposure to 100 �M MRS2179 (n = 25 cells) or 10 �M Cangrelor (n = 20 cells). Values were normalized to ADP alone set to1.00; (E) left: representative temporal plots of [Ca2+]i changes recorded from three satellite cells stimulated with 100 �M UTP before and after application ofthe relatively P2Y2/P2Y4-selective antagonist Reactive Blue-2 (RB-2; 100 �M), as indicated. Right: quantification of mean calcium increase ±S.E.M. inducedby 100 �M UTP before and after exposure to 100 �M Suramin (n = 24 cells) or 100 �M RB-2 (n = 30 cells). Values were normalized to UTP alone, set to 1.00.*P < 0.05 and **P < 0.01 with respect to corresponding agonist alone, by one way ANOVA followed by Scheffe’s analysis.
S. Ceruti et al. / Cell Calcium 43 (2008) 576–590 587
Fig. 8. Acute and chronic exposure to bradykinin (BK) differentially affect the function of neuronal P2X3 and glial P2Y receptors: (A) representative temporalplots of [Ca2+]i changes induced by ��-meATP in vehicle- or acutely BK-treated (100 nM) neurons. The quantification of results, normalized to vehicle-treatedcontrol set to 1.00, is shown in panel (B); (C–F) results obtained after a chronic (24 h) exposure of trigeminal cultures to either vehicle (–BK) or 100 nM BK;( ponsesc meATP( h respe
savcsap
eiea
C) mean calcium responses to ��-meATP in neurons; (D) mean calcium resells) recorded in SGCs; (E) mean percentage of neurons responding to ��-with vehicle-treated control cells set to 100%). *P < 0.05 and **P < 0.01 wit
Analysis of single-cell calcium changes induced by expo-ure to ATP revealed increase in the [Ca2+]i in both neuronsnd glial cells after 48 h in culture (Fig. 3). Since ATP acti-ates various receptor subtypes, we then characterized the
ontribution of different P2 receptor subtypes by applyingelective agonists. As a first step, since very few data arevailable in TG, we focused on P2X3, the most importanturinoceptor involved in pain transmission. Expression lev-
sT[s
to various nucleotides (normalized to corresponding vehicle-treated control; (F) mean percentage of glial cells responding to ATP, ADP, UTP and UDPct to vehicle-treated cells, one-way ANOVA followed by Scheffe’s analysis.
ls of P2X3 and of the heterodimer P2X2/P2X3 are increasedn different inflammatory [47], and neuropathic [48] mod-ls of pain, and pharmacological inhibition of P2X2/P2X3ctivation reduces or abolishes pain sensation [49]. Recent
tudies reported that P2X3 receptors are also increased inG primary cultures after exposure to pro-algogenic CGRP
32,34], suggesting the possible importance of this receptorubtype also in migraine pain. According to literature data
5 Calcium
[7s(maavdpPtTmest
PRrntbsRtaistatmcbtIeUpgaw1((Utasc
tTa
Ptsfrsftc
er
4d
ictbnrr[erowtsB(oamatdBrfiitsibP
tFi
88 S. Ceruti et al. / Cell
23,34], P2X3 was found to be expressed in approximately0% of neurons (Fig. 4A), and, relevant to pain transmis-ion, in almost the totality of IB4-positive sensory neuronsFig. 4B), as also demonstrated in DRG [50]. In approxi-ately 60% of neurons, P2X3 is coupled to [Ca2+]i increases
s demonstrated by application of the relatively selectivegonist ��-meATP (Fig. 4D). In line with previous obser-ation [31], no glial expression of the P2X3 receptor wasetected, by either immunocytochemistry or functional cou-ling to [Ca2+]i. Primary trigeminal cultures also express2X2, which likely dimerizes with P2X3 contributing to pain
ransmission, but not the P2X1 receptor subtype (Fig. 4C).aken together, these results suggest that TG sensory neuronsaintain in culture their physiological responses to ATP, and
xpress functional algogenic P2X3 receptors, confirming theuitability of this in vitro model to study modulation of painransmission by the purinergic system.
As mentioned above, we also aimed at characterizing2Y receptors on both neurons and SGCs in TG cultures.egarding the presence of these receptors in sensory neu-
ons, immunohistochemical analysis on intact DRG, TG, andodose ganglia has demonstrated that P2Y1 and P2Y4 recep-ors are expressed by two distinct neuronal populations [23],ut, while some functional data are available in DRG [13,14],o far no studies have been performed in TG. Interestingly,T-PCR analysis showed that all cloned murine P2Y recep-
ors (i.e., P2Y1,2,4,6,12,13,14) are expressed in both intact TGnd in trigeminal cultures 48 h after plating (Fig. 5). Coex-stence of different P2 receptors has been demonstrated ineveral cell types, such as astrocytes [44], microglia [36], cen-ral nervous system neurons [51], and cardiomyocytes [43],nd seems indeed to represent more a rule than an excep-ion. Different receptors may thus be inserted into the plasma
embrane and/or recruited by cells, depending upon spe-ific functional states ([36]; see also below). Results obtainedy single-cell calcium recordings from TG neurons indicatehat only some P2Y receptors are functional in these cells.n particular we show, for the first time, that TG neuronsxpress functional metabotropic ADP-sensitive P2Y1 andTP-responding P2Y2/P2Y4 receptors (Fig. 6). In a smallercentage of neurons, P2Y12/P2Y13 were also found, sug-esting that, under some circumstances, their activation maylso contribute to ADP-mediated effects. Increase of [Ca2+]iere much more sustained over time after application of00 �M UTP with respect to the same concentration of ADPFig. 6), this difference being less evident at 10 �M UTPnot shown). This finding suggests that a higher number ofTP-sensitive receptors might be present in neurons and/or
hat, in the case of UTP, additional means to increase [Ca2+]ire recruited after mobilization of calcium from intracellulartores (e.g., calcium-dependent calcium entry via membranealcium channels).
The present results also add novel information regardinghe presence and function of calcium-linked P2Y receptors onG SGCs, for which very limited data have been so far avail-ble. In one single study on intact TG ganglion, P2Y1, and
oaPi
43 (2008) 576–590
2Y2/P2Y4 receptors on SGCs were demonstrated to coupleo increase of intracellular calcium concentrations [31]. Ashown in Fig. 7, in dissociated TG cultures, SGCs expressunctional metabotropic P2Y1 and P2Y12/P2Y13 receptorsesponsive to ADP, as well as P2Y2/P2Y4 receptors respon-ive to UTP. Only a minor [Ca2+]i response to UDP wasound, indicating that P2Y6 receptors marginally contributeo P2-mediated calcium increases in these cells, at least underontrol conditions (see also below).
Taken together, our results represent the first systematicvaluation of P2Y-receptor expression and function in neu-onal and non-neuronal TG cells.
.3. Acute and chronic bradykinin exposureifferentially affect P2-receptor function in TG cultures
Prolonged exposure of nociceptors to algogenic and pro-nflammatory mediators may lead to neuronal [52], and glialhanges [53] contributing to pathological pain chronicisa-ion [9,54,55]. Among other pro-inflammatory substances,radykinin (BK) is indeed a known activator of sensoryeurons [56] and has been demonstrated to both increase neu-onal firing rate in TG nucleus caudalis and to enhance theelease of CGRP from cultured dorsal horn and TG neurons33]. Moreover, literature data suggest that a brief, acute (50 s)xposure to BK can rapidly sensitise the function of P2X3-eceptor subtype in an expression model (Xenopus Laevisocytes) [57]. Since SGCs are also sensitive to BK [58],e deemed it interesting to study the effect of an acute (up
o 3 min) or a prolonged (24 h) BK exposure on purinergicignaling in both neurons and SGCs. Acute application ofK-induced responses in a significant percentage of neurons
see also below) and significantly enhanced the amplitudef neuronal P2X3-mediated peak calcium response (Fig. 8And B), thus confirming the data obtained in an expressionodel [57]. Chronic BK treatment instead induced a strong
nd highly significant reduction of neuronal P2X3 recep-or function (Fig. 8C and E). These data represent the firstemonstration of modulation of P2X3 receptor functions byK in a native system. Thus, a biphasic response of P2X3
eceptor to painful situations, with increased or decreasedring depending on the length of the exposure to the nox-
ous and pro-inflammatory stimulus, might exist, suggestinghat different pharmacological manipulations of this receptorubtype should be utilized in acute or chronic pain. Interest-ngly, in TG cultures, a 1-h exposure to CGRP, which cane released by BK [33], also led to upregulation of neuronal2X3 receptors [32].
Upon chronic exposure to BK, metabotropic P2Y recep-ors on SGCs were concomitantly increased (Fig. 8D and). Of interest, very few SGCs (about 3%) responded with
ncrease in [Ca2+]i upon BK exposure compared to the 35%
f responding neuronal cells (Fig. 8A; see Section 3). Thisllows us to hypothesize that the observed changes in SGC2Y receptors functionality might be due to a yet-to-be
dentified factor released by neurons upon BK stimulation,
Calcium
tiPnIhinariUtaforbs
aiFbpatapac
A
hSsP(O
R
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
S. Ceruti et al. / Cell
hus contributing to cell-to-cell communication in TG dur-ng pain transmission. Concerning the significance of these2Y-receptor changes, data on the role of P2Y-mediated sig-aling in modulation of pain are at present contradictory.n fact, both algogenic and analgesic P2Y-mediated effectsave been described. For example, the P2Y1 subtype cannhibit P2X3 receptor and voltage-activated calcium chan-els functions in DRG neurons, therefore, suggesting annti-algogenic role for ADP [26]. On the other hand, P2Y2eceptor activation on DRG neurons facilitates capsaicin-nduced membrane currents through TRPV1 [10], suggestingTP to be an algogenic stimulus. Surprisingly, administra-
ion of UTP and UDP intratechally in vivo was found to benalgesic [59], probably due to inhibition of cytokine releaserom glial cells. It might be also hypothesized that changesf these receptors after chronic exposure to pain stimuli mayepresent an initial attempt to decrease pain transmission,ut may eventually contribute to further worsening painfulensations and to perpetuate pathological pain.
Globally, these data suggest that TG cultures representn appropriate experimental model to study calcium signal-ng via purinergic receptors in this important sensory station.unction of P2 receptors on both neurons and glial cells cane easily monitored by single-cell calcium imaging underseudo-physiological conditions as well as upon exposure tolgogenic and pro-inflammatory substances. This model mayhus prove useful to study alterations of calcium signalingnd dysregulation of receptor function under experimentalaradigms reproducing in vitro the conditions of sustainedlgogen release and activity known to occur in vivo duringhronic pain.
cknowledgements
Authors are deeply grateful to Dr. Paolo Gelosa forelp with immunohistochemical analysis, and to Mr. Giulioimonutti for technical assistance at the confocal micro-cope. Cangrelor was a kind gift of The Medicines Company,arsippany, NJ, USA. This work was supported by a grantGGP04037) from the Italian Comitato Telethon Fondazionenlus to MPA.
eferences
[1] G. Burnstock, Purinergic P2 receptors as targets for novel analgesics,Pharmacol. Ther. 110 (2006) 433–454.
[2] M.P. Abbracchio, G. Burnstock, Purinoceptors: are there families ofP2X and P2Y purinoceptors? Pharmacol. Ther. 64 (1994) 445–475.
[4] A. Nicke, H.G. Baumert, J. Rettinger, et al., P2X1 and P2X3 recep-
tors form stable trimers: a novel structural motif of ligand-gated ionchannels, EMBO J. 17 (1998) 3016–3028.
[5] G.E. Torres, T.M. Egan, M.M. Voigt, Hetero-oligomeric assembly ofP2X receptor subunits. Specificities exist with regard to possible part-ners, J. Biol. Chem. 274 (1999) 6653–6659.
[
43 (2008) 576–590 589
[6] M.P. Abbracchio, G. Burnstock, J.M. Boeynaems, et al., InternationalUnion of Pharmacology LVIII: update on the P2Y G protein-couplednucleotide receptors: from molecular mechanisms and pathophysiologyto therapy, Pharmacol. Rev. 58 (2006) 281–341.
[7] G. Burnstock, A unifying purinergic hypothesis for the initiation ofpain, Lancet 347 (1996) 1604–1605.
[8] G. Burnstock, J.N. Wood, Purinergic receptors: their role in nocicep-tion and primary afferent neurotransmission, Curr. Opin. Neurobiol. 6(1996) 526–532.
[9] S.B. McMahon, W.B. Cafferty, F. Marchand, Immune and glial cellfactors as pain mediators and modulators, Exp. Neurol. 192 (2005)444–462.
10] Z. Gerevich, P. Illes, P2Y receptors and pain transmission, PurinergicSignal. 1 (2004) 3–10.
11] B.A. Chizh, P. Illes, P2X receptors and nociception, Pharmacol. Rev.53 (2001) 553–568.
12] R.A. North, P2X3 receptors and peripheral pain mechanisms, J. Physiol.554 (2004) 301–308.
13] Z. Gerevich, S.J. Borvendeg, W. Schroder, et al., Inhibition of N-typevoltage-activated calcium channels in rat dorsal root ganglion neuronsby P2Y receptors is a possible mechanism of ADP-induced analgesia,J. Neurosci. 24 (2004) 797–807.
14] M. Sanada, H. Yasuda, M. Omatsu-Kanbe, et al., Increase in intra-cellular Ca(2+) and calcitonin gene-related peptide release throughmetabotropic P2Y receptors in rat dorsal root ganglion neurons, Neu-roscience 111 (2002) 413–422.
15] G. Collo, R.A. North, E. Kawashima, et al., Cloning of P2X5 and P2X6
receptors and the distribution and properties of an extended family ofATP-gated ion channels, J. Neurosci. 16 (1996) 2495–2507.
16] C.C. Chen, A.N. Akopian, L. Sivilotti, D. Colquhoun, G. Burnstock,J.N. Wood, A P2X purinoceptor expressed by a subset of sensoryneurons, Nature 377 (1995) 428–431.
17] S.P. Cook, L. Vulchanova, K.M. Hargreaves, R. Elde, E.W. McCleskey,Distinct ATP receptors on pain-sensing and stretch-sensing neurons,Nature 387 (1997) 505–508.
18] C. Lewis, S. Neidhart, C. Holy, R.A. North, G. Buell, A. Sur-prenant, Coexpression of P2X2 and P2X3 receptor subunits canaccount for ATP-gated currents in sensory neurons, Nature 377 (1995)432–435.
19] S.J. Robertson, M.G. Rae, E.G. Rowan, C. Kennedy, Characteriza-tion of a P2X-purinoceptor in cultured neurones of the rat dorsal rootganglia, Br. J. Pharmacol. 118 (1996) 951–956.
20] R. Ambalavanar, M. Moritani, D. Dessem, Trigeminal P2X3 receptorexpression differs from dorsal root ganglion and is modulated by deeptissue inflammation, Pain 117 (2005) 280–291.
21] J. Luo, G.F. Yin, Y.Z. Gu, et al., Characterization of three types of ATP-activated current in relation to P2X subunits in rat trigeminal ganglionneurons, Brain Res. 1115 (2006) 9–15.
22] Z. Xiang, X. Bo, G. Burnstock, Localization of ATP-gated P2X receptorimmunoreactivity in rat sensory and sympathetic ganglia, Neurosci.Lett. 256 (1998) 105–108.
23] H.Z. Ruan, G. Burnstock, Localisation of P2Y1 and P2Y4 receptors indorsal root, nodose and trigeminal ganglia of the rat, Histochem. CellBiol. 120 (2003) 415–426.
24] K. Kobayashi, T. Fukuoka, H. Yamanaka, et al., Neurons and glialcells differentially express P2Y receptor mRNAs in the rat dorsal rootganglion and spinal cord, J. Comp. Neurol. 498 (2006) 443–454.
25] N. Svichar, A. Shmigol, A. Verkhratsky, P. Kostyuk, ATP induces Ca2+release from IP3-sensitive Ca2+ stores exclusively in large DRG neu-rones, Neuroreport 8 (1997) 1555–1559.
26] Z. Gerevich, C. Muller, P. Illes, Metabotropic P2Y1 receptors inhibitP2X3 receptor-channels in rat dorsal root ganglion neurons, Eur. J.
Pharmacol. 521 (2005) 34–38.
27] D.C. Molliver, S.P. Cook, J.A. Carlsten, D.E. Wright, E.W. McCleskey,ATP and UTP excite sensory neurons and induce CREB phosphoryla-tion through the metabotropic receptor, P2Y2, Eur. J. Neurosci. 16(2002) 1850–1860.
5 Calcium
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[[
[
[
[
[
[
[
[
[
[
[
[
90 S. Ceruti et al. / Cell
28] D. Pietrobon, Migraine: new molecular mechanisms, Neuroscientist 11(2005) 373–386.
29] M. Hanani, Satellite glial cells in sensory ganglia: from form to func-tion, Brain Res. Brain Res. Rev. 48 (2005) 457–476.
30] P.S. Cherkas, T.Y. Huang, T. Pannicke, M. Tal, A. Reichenbach, M.Hanani, The effects of axotomy on neurons and satellite glial cells inmouse trigeminal ganglion, Pain 110 (2004) 290–298.
31] M. Weick, P.S. Cherkas, W. Hartig, et al., P2 receptors in satelliteglial cells in trigeminal ganglia of mice, Neuroscience 120 (2003)969–977.
32] E. Fabbretti, M. D’Arco, A. Fabbro, M. Simonetti, A. Nistri, R.Giniatullin, Delayed upregulation of ATP P2X3 receptors of trigem-inal sensory neurons by calcitonin gene-related peptide, J. Neurosci.26 (2006) 6163–6171.
33] D.W. Jenkins, L.A. Sellers, W. Feniuk, P.P. Humphrey, Characteriza-tion of bradykinin-induced prostaglandin E2 release from cultured rattrigeminal ganglion neurones, Eur. J. Pharmacol. 469 (2003) 29–36.
34] M. Simonetti, A. Fabbro, M. D’Arco, et al., Comparison of P2X andTRPV1 receptors in ganglia or primary culture of trigeminal neuronsand their modulation by NGF or serotonin, Mol. Pain 2 (2006) 11.
35] P. Ciana, M. Fumagalli, M.L. Trincavelli, et al., The orphanreceptor GPR17 identified as a new dual uracil nucleotides/cysteinyl-leukotrienes receptor, EMBO J. 25 (2006) 4615–4627.
36] F. Bianco, M. Fumagalli, E. Pravettoni, et al., Pathophysiological rolesof extracellular nucleotides in glial cells: differential expression ofpurinergic receptors in resting and activated microglia, Brain Res. BrainRes. Rev. 48 (2005) 144–156.
37] P.M. Dunn, Y. Zhong, G. Burnstock, P2X receptors in peripheral neu-rons, Prog. Neurobiol. 65 (2001) 107–134.
38] M. Shinoda, N. Ozaki, H. Asai, K. Nagamine, Y. Sugiura, Changesin P2X3 receptor expression in the trigeminal ganglion followingmonoarthritis of the temporomandibular joint in rats, Pain 116 (2005)42–51.
39] D. Julius, A.I. Basbaum, Molecular mechanisms of nociception, Nature413 (2001) 203–210.
40] Y.Z. Gu, G.F. Yin, B.C. Guan, Z.W. Li, Characteristics of P2Xpurinoceptors in the membrane of rat trigeminal ganglion neurons,Sheng Li Xue Bao 58 (2006) 164–170.
41] K. Nagamine, N. Ozaki, M. Shinoda, et al., Mechanical allodynia andthermal hyperalgesia induced by experimental squamous cell carci-noma of the lower gingiva in rats, J. Pain 7 (2006) 659–670.
42] H. Wang, C. Rivero-Melian, B. Robertson, G. Grant, Transganglionictransport and binding of the isolectin B4 from Griffonia simplicifolia Iin rat primary sensory neurons, Neuroscience 62 (1994) 539–551.
43] C. Banfi, S. Ferrario, O. De Vincenti, et al., P2 receptors in humanheart: upregulation of P2X6 in patients undergoing heart transplanta-
[
43 (2008) 576–590
tion, interaction with TNFalpha and potential role in myocardial celldeath, J. Mol. Cell Cardiol. 39 (2005) 929–939.
44] M. Fumagalli, R. Brambilla, N. D’Ambrosi, et al., Nucleotide-mediatedcalcium signaling in rat cortical astrocytes: role of P2X and P2Y recep-tors, Glia 43 (2003), 218-203.
45] A.M. van den Maagdenberg, D. Pietrobon, T. Pizzorusso, et al., ACacna1a knockin migraine mouse model with increased susceptibilityto cortical spreading depression, Neuron 41 (2004) 701–710.
mediated responses in rat dorsal root ganglion neurons, J. Neurosci. 22(2002) 93–102.
48] K. Kage, W. Niforatos, C.Z. Zhu, K.J. Lynch, P. Honore, M.F. Jarvis,Alteration of dorsal root ganglion P2X3 receptor expression and func-tion following spinal nerve ligation in the rat, Exp. Brain Res. 147(2002) 511–519.
50] E.J. Bradbury, G. Burnstock, S.B. McMahon, The expression of P2X3
purinoreceptors in sensory neurons: effects of axotomy and glial-derived neurotrophic factor, Mol. Cell Neurosci. 12 (1998) 256–268.
51] S. Amadio, N. D’Ambrosi, F. Cavaliere, et al., P2 receptor modulationand cytotoxic function in cultured CNS neurons, Neuropharmacology42 (2002) 489–501.
52] J.-G. Wang, J.A. Strong, W. Xie, J.-M. Zhang, Local inflammation inrat dorsal root ganglion alters excitability and ion currents in small-diameter sensory neurons, Anesthesiology 107 (2007) 322–332.
53] P. Dublin, M. Hanani, Satellite glial cells in sensory ganglia: their possi-ble contribution to inflammatory pain, Brain Behav. Immun. 21 (2007)592–598.
54] L.R. Watkins, S.F. Maier, Glia: a novel drug discovery target for clinicalpain, Nat. Rev. Drug Discov. 2 (2003) 973–985.
55] J. Wieseler-Frank, S.F. Maier, L.R. Watkins, Glial activation and patho-logical pain, Neurochem. Int. 45 (2004) 389–395.
56] J.B. Calixto, D.A. Cabrini, J. Ferreira, M.M. Campos, Kinins in painand inflammation, Pain 87 (2000) 1–5.
57] M. Paukert, R. Osteroth, H.S. Geisler, et al., Inflammatory mediatorspotentiate ATP-gated channels through the P2X(3) subunit, J. Biol.Chem. 276 (2001) 21077–21082.
58] F. Heblich, S. England, R.J. Docherty, Indirect actions of bradykininon neonatal rat dorsal root ganglion neurones: a role for non-neuronal
cells as nociceptors, J. Physiol. 536 (2001) 111–121.
59] M. Okada, T. Nakagawa, M. Minami, M. Satoh, Analgesic effects ofintrathecal administration of P2Y nucleotide receptor agonists UTP andUDP in normal and neuropathic pain model rats, J. Pharmacol. Exp.Ther. 303 (2002) 66–73.
