SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  • 1
    Although peptides are important modulators of synapses, their action on synapse-glia interactions remain unclear. The amphibian neuromuscular junction (NMJ) was used to examine the effects of substance P (SP) on perisynaptic Schwann cells (PSCs), glial cells at the frog NMJ, by monitoring changes in intracellular Ca2+.
  • 2
    SP induced Ca2+ responses that were mimicked by the neurokinin 1 receptor (NK-1) agonist septide and with a shorter delay by the SP fragment, SP(6–11). SP and SP(6–11) responses were blocked by NK-1 antagonists SR140333 and LY303870.
  • 3
    Ca2+ responses remained unchanged when extracellular Ca2+ was removed but were blocked after pertussis toxin (PTX) treatment, indicating that the receptors were linked to internal stores of Ca2+ via a PTX-sensitive G-protein.
  • 4
    The slowly hydrolysable NK-1 agonist [Sar9, Met(O2)11]-SP only induced Ca2+ responses when applied for a long period of time and not during brief, local applications, suggesting the involvement of SP hydrolysis. Acetylcholinesterase (AChE) may not be involved in SP degradation since Ca2+ responses evoked by SP were unchanged in the presence of the cholinesterase inhibitor neostigmine.
  • 5
    Ca2+ responses induced by muscarine and nerve stimulations were almost abolished when preceded by SP applications, while those induced by ATP were significantly reduced. The rundown of the nerve-evoked Ca2+ responses in PSCs was attenuated in the presence of SR140333.
  • 6
    These results indicate that endogenous SP is involved in the regulation of PSC activity and that SP is an important modulator of glial cell Ca2+ signalling and synapse-glia communication.

Neural control and modulatory signals are believed to occur between the pre- and postsynaptic elements of chemical synapses and to arise from the effects of neurotransmitters and neuromodulators. However, synapses are tightly surrounded by glial cells. These cells appear as potential targets for chemical signals released from neurons since they possess a variety of receptors for neurotransmitters and react to neurotransmitters released during synaptic activity (Dani et al. 1992; Chiu & Kriegler, 1994). The presence of functional receptors for transmitters on glial cells emphasizes the possibility that glial cells may, in turn, play modulatory roles in synaptic function. This is supported by the observations by Parpura et al. (1994) and Nedergaard (1994) that glial cells can modulate neuronal activity.

Peptides are widely distributed in central and peripheral nervous systems and are present with other neurotransmitters at synapses. Peptides are good candidates to mediate signalling between synapses and perisynaptic glial cells since they are responsible for numerous modulatory effects on neuronal and glial cells (Evans et al. 1990; De Koninck et al. 1994; Heath et al. 1994). The influence of neurochemicals on glial cells has been difficult to define at synapses in the CNS owing to the complexity of the connections. The frog neuromuscular junction (NMJ) provides a simple system where the relationship between synaptic components is kept intact and which allows in situ studies of synaptic-glial interactions. It has been demonstrated that the nerve-evoked release of neurotransmitters or local applications of neurotransmitters raised intracellular Ca2+ in perisynaptic Schwann cells (PSCs), glial cells at the frog NMJ, similar to astrocytes (Jahromi et al. 1992; Reist & Smith, 1992; Robitaille, 1995; Robitaille et al. 1997). Interestingly, Evans et al. (1986, 1990) observed that vasoactive intestinal peptide (VIP) and substance P (SP) hyperpolarized the membrane potential of Schwann cells at the squid giant axon.

The present study focuses on the role of SP in synapse-glia interactions since it is present in the nerve terminal of the frog NMJ (Matteoli et al. 1990) and since many of its physiological effects have been characterized (Pernow, 1983). SP is a tachykinin that acts mainly through the neurokinin-1 receptor (NK-1) which potentiates the phospholipase C pathway, elevates IP3 and releases intracellular Ca2+ (Guard & Watson, 1991).

Using a Ca2+ imaging technique, we report that SP induces the release of Ca2+ from internal stores by pertussis toxin (PTX)-sensitive G-proteins linked to an NK-1-like receptor. SP modulates Ca2+ responses induced by muscarinic (mACh) and purinergic (ATP) receptors and endogenous SP modulates the responsiveness of the cell to nerve-evoked transmitter release. It is proposed that glial cells may receive SP as a chemical messenger which would act as a modulator of glial cell Ca2+ signalling and synapse-glia interactions.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements

Experiments were performed on NMJs of cutaneous pectoris muscles of Rana pipiens frogs (5–7 cm body length; Wards, St-Catharines, Ontario, Canada and Connecticut Valley Biological Supply, Southampton, MA, USA). Animals were killed by double pithing as recommended and approved by the Animal Care Committee of the Université de Montréal. Nerve-muscle preparations were dissected and mounted in a recording chamber filled with Ringer solution (mm: 120 NaCl, 2 KCl, 1 NaHCO3, 1.8 CaCl2 and 5.0 or 15.0 Hepes; pH adjusted to 7.2 with NaOH). In some experiments, external Ca2+ was omitted and 5 mm MgCl2 added (Mg2+ Ringer).

Calcium imaging

The membrane-permeant form of the Ca2+ indicator fluo-3 (fluo-3 AM; Molecular Probes) was used for imaging Ca2+ changes in PSCs. The loading procedure has been described in detail previously (Jahromi et al. 1992; Georgiou et al. 1994; Robitaille, 1995; Robitaille et al. 1997). Briefly, nerve-muscle preparations were incubated in a solution containing 10 μm fluo-3 AM with a final concentration of 1.1 % dimethyl sulphoxide (DMSO; Sigma) and 0.02 % pluronic acid for 90–120 min at room temperature (21–23°C). Unless stated otherwise, preparations were perfused with cold Ringer solution (12–15°C) containing 20 μm tetrakis (2-pyridylmethyl) ethylenediamine (TPEN; Molecular Probes). TPEN was used for partial chelation of heavy metals that bind to fluo-3 and limit its response to Ca2+ (Jahromi et al. 1992).

Changes in fluorescence intensity were observed with a laser scanning confocal microscope (BioRad 600) equipped with an argon ion laser. The 488 nm line of the laser was attenuated to 1 % using neutral density filters. The emitted fluorescence was detected through a low pass filter (cut-off at 515 nm). Preparations were examined with a × 40 water immersion lens (Nikon, 0.55 NA or Olympus 0.75 NA, Tokyo, Japan). PSCs on surface NMJs were located using transmitted light microscopy, which allows reliable identification (Georgiou et al. 1994). Images were collected at an additional zoom factor (3 or 4) and fluorescence intensity measurements were made over the cell body area of the PSCs. Relative changes in fluorescence intensities (%ΔF/F), were expressed as:

F/F= (F - Frest)/Frest× 100,

  • image

where F is the fluorescence that reflects the intracellular level of Ca2+. The %ΔF/F of peak Ca2+ responses are presented. Baseline fluorescence intensities (Frest) for each cell studied were obtained before any drug applications. Images were collected in two ways. First, for time course measurements, images were acquired continuously at 645 ms intervals. Second, three or five successive frames of NMJs and PSCs loaded with fluo-3 were averaged at rest and at intervals of 5–10 s after local application of SP for up to 4 min. Fluorescence intensity was measured on a grey scale where black corresponds to lower, and white corresponds to higher levels of Ca2+. None of the Ca2+ responses saturated the range of the 256 grey levels of the system. Also, it is unlikely that the responses saturated the fluorescent Ca2+ indicator since larger Ca2+ responses than the ones obtained could be elicited in the same experimental conditions.

Statistical analysis

Each cell is considered as an independent observation and the results from all cells submitted to a given treatment are expressed as a mean ±s.e.m. Unless stated otherwise, the mean values represent the %ΔF/F of responding cells. For classification purposes, cells showing an increase in intracellular %ΔF/F of less than 30 % were considered as non-responding cells. Statistical analysis was performed using one-way analysis of variance (ANOVA). The P values were corrected by Bonferroni's method to allow multiple comparisons and were considered significant when P < 0.05. When data were compared between different groups, statistical significance was determined using the independent Student's unpaired or paired t tests, when appropriate, at a confidence level of 95 %.

Drug applications

Agonists were dissolved with the same solution used for perfusion. This ensured that the only difference between the two solutions was the presence of the agonist in the pipette. Local applications of agonists on PSCs were performed using a series of positive pressure pulses (1–10 pulses, 999 ms duration, 10 psi) applied to a glass micropipette (tip diameter, 2–3 μm) by a Picospritzer II (General Valve, Fairfield, NJ, USA). The area covered by the local application of agonist was about 50–100 μm as indicated by the displacement of the tissue. A deflection of the baseline was induced by the movement of the preparation caused by consecutive pulses of positive pressure necessary to apply the agonists on PSCs (see Figs 1, 2 and 3). When a cell did not respond to a first local application of an agonist, the pipette was repositioned closer and the agonist was reapplied. Local pressure applications of the perfusion solution on PSCs did not induce any Ca2+ responses.

image

Figure 1. . SP-induced Ca2+ responses in PSCs at frog NMJ

A, confocal images of Ca2+ responses to local applications of SP on PSCs. Relative fluo-3 fluorescence appears on a grey scale, where black corresponds to lower, and white corresponds to higher fluorescence. A1, three PSCs at rest; A2, during the Ca2+ response to local applications of 200 μm SP; A3, during the recovery period. The image of the responding PSCs does not necessarily show their peak responses since the onset was different for each cell. B, changes in cell body fluorescence (%ΔF/F) over time before, during and after the local drug applications (arrow) on the cell indicated by the arrow in A. Numbers beside the trace indicate the respective times at which images in A were taken. Note that a rebound in fluorescence intensity occurs during the recovery period. C, Ca2+ responses in PSCs evoked by a bath application (2 min) of SP (1 μm). One PSC is shown before (Rest) and at the peak of the Ca2+ response (SP). Scale bars, 10 μm.

Download figure to PowerPoint

image

Figure 2. . SP(6–11)- and septide-induced Ca2+ responses in PSCs at frog NMJ

A, Ca2+ responses in PSCs evoked by local applications of SP(6–11). Images show 4 cells before (A1) and after local applications of 20 μm SP(6–11) (A2), and during recovery following drug applications (A3). The images of responding PSCs do not necessarily show their peak responses since the onset was different for each cell. B, time course of the Ca2+ response in the cell indicated by the arrow in A. Numbers beside the trace indicate the times at which the images in A were taken. No oscillations were observed during the recovery period. C, delays for Ca2+ changes induced by local applications (arrow) of SP and SP(6–11) on the same PSC. Each trace illustrates the time course of Ca2+ responses for the first 45 s after the agonist applications, with an interval of 10 min between the two applications. 20 μm SP(6–11) (○) induced Ca2+ responses after 20.3 s and 200 μm SP (•) induced Ca2+ responses after 33.9 s. D, confocal images of Ca2+ responses to local applications of septide (20 μm), a SP agonist specific for NK-1 receptors. Images show two PSCs before (Control) and at the peak of the Ca2+ response evoked by local applications of septide (20 μm). Scale bars, 10 μm.

Download figure to PowerPoint

image

Figure 3. . Effects of an NK-1 blocker on Ca2+ responses induced by SP and SP(6–11)

A, time course of the Ca2+ changes for 3 PSCs in response to local applications of agonists (arrows) in the presence of the NK-1 receptor blocker, SR140333. Local applications of 20 μm SP(6–11) failed to induce Ca2+ responses in PSCs with SR140333 (4 μm) in the perfusate. After washout of the first agonist (first break in the x-axis), local applications of 200 μm SP also failed to induce Ca2+ responses in the presence of SR140333. However, 10 min later (second break in the x-axis), Ca2+ responses could be induced with 10 μm muscarine, indicating the specific action of SR140333 on NK-1 receptors. B, confocal images of the Ca2+ fluorescence in two PSCs in the presence of SR140333 (Rest), during application of SP (200 μm) in the presence of SR140333 (SP/SR140333) and after the washout of SR140333 (SP). Note that the effects of SR140333 are reversible. Scale bar, 10 μm.

Download figure to PowerPoint

To determine the effects of SP on Ca2+ responses induced by mACh and ATP receptors (Fig. 6), cells were exposed to local applications of SP (20 μm) 10 min before muscarine or ATP applications.

image

Figure 6. . Modulation by SP of muscarine and ATP-induced Ca2+ responses in PSCs

A, percentage of responding cells and the mean of Ca2+ response in PSCs induced by local applications of 10 μm muscarine (□) and induced by local applications of muscarine (10 μm) following SP applications (20 μm; 10 min intervals; ▪). Note that both the percentage of responding cells and the size of muscarine-induced Ca2+ responses were significantly reduced after SP applications. B, percentage of responding cells and the mean of Ca2+ responses in PSCs evoked by local application of ATP (□) and evoked by ATP following local applications of SP (▪). Following applications of 20 μm SP (10 min intervals) the percentage of cells responsive to 20 μm ATP was not affected, whereas the mean Ca2+ response was significantly different from the control. Numbers of cells tested are shown in parentheses. C, normalized increase of fluorescence expressed as a percentage of the Ca2+ response in PSCs evoked by the first application of ATP. On average, consecutive applications of 10 μm ATP (10 min intervals) induced Ca2+ responses of similar amplitude (6 cells). D, normalized increase of fluorescence expressed as a percentage of the Ca2+ response in PSCs evoked by the first application of ATP after SP (500 nM) was perfused for 10 min (bar), immediately after the first ATP application. Note that a rundown in the mean Ca2+ response induced by 10 μm ATP was observed (6 cells, different preparation from C). Significant differences from the control values are shown by asterisks: *P= 0.02, **P= 0.005, Student's t test.

Download figure to PowerPoint

Bath applications of SP and its agonists were performed in two ways. First, in a closed bath, the perfusion was stopped and the control solution was removed from the bath and replaced by the agonist solution for 2 min. The agonist solution was then washed out with the Ringer perfusate. Second, the perfusate was exchanged with a solution containing the agonist for 10 min and then perfused with the control physiological solution. Similar results were obtained with the two methods.

Stock solutions of the NK-1 receptor antagonists (S)1-{2-[3,4-dichlorophenyl)-1-(3-isopropoxyphenyl-acetyl)piperidin-3-yl]ethyl}-4-phenyl-1-azoniabicyclo[2.2.2]octane chloride (SR140333) (6.5 and 100 mm) and (R)-1-[N-(2-methoxybenzyl)acetylamino]-3-(1H-indol-3-yl)-2-[N-(2-(4-(piperidin-1-yl)piperidin-1-yl)acetylamino]propane (LY303870) were prepared in DMSO, divided into aliquots and kept at −80°C. Both antagonists were used in Ringer solution, at a final concentration of 1–4 μm (0.001–0.004 % DMSO, respectively). Neostigmine (3 μg ml−1) was dissolved in Ringer-TPEN solution. The antagonist and neostigmine solutions were perfused at least 20 min prior to drug applications. Neither SR140333, LY303870 nor neostigmine affected the baseline fluorescence of cells during the perfusion (not shown). Stock solutions of NK-1 agonists [Sar9, Met(O2)11]-SP (1 mm) and septide (10 mm) were prepared in distilled water and DMSO, respectively. In some experiments, ω-conotoxin GVIA (ω-CgTx; 1 μm) was used to block Ca2+ channels responsible for the release of neurotransmitters (Robitaille et al. 1996). The toxin was applied for 30 min prior to any agonist applications.

Pertussis toxin experiments

Preparations were incubated overnight (12–14 h) at 20–22°C in normal Ringer solution containing 2 μg ml−1 of PTX. The PTX solution was replaced with fresh toxin solution twice during the incubation period. Muscles were then processed for Ca2+ imaging as indicated above.

Nerve-evoked Ca2+ responses

Nerve-muscle preparations were loaded with fluo-3 AM as described above and Ca2+ responses in PSCs induced by synaptic transmission were examined. Motor axons were stimulated at 50 Hz for 30 s using a suction electrode coupled to an S-88 Grass stimulator. Intervals of 25 min were allowed between trials when repetitive trains of stimuli were performed on the same preparation. To prevent muscle contractions evoked by transmitter release, preparations were incubated with unlabelled α-bungarotoxin (α-BuTx; 10 μg ml−1 for 10 min; Molecular Probes) which blocks muscle ACh receptors but has no effect on PSCs (Jahromi et al. 1992; Robitaille et al. 1997).

Drug supplies

Peptides were purchased from Bachem California (SP, SP(6–11), SP(1–7), [Sar9, Met(O2)11]-SP) or Peptides International Inc (SP; Louisville, KY, USA). ATP and muscarine were obtained from Sigma, neostigmine bromide, ω-CgTx and PTX were from RBI (Natick, MA, USA) and SR140333 and LY303870 were from Dr J. C. Brelière of Sanofi (Montpellier, France) through Dr R. Couture (Département de physiologie, Université de Montréal, Canada).

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements

SP-mediated Ca2+ responses in PSCs

In 93 % of PSCs studied (26 of 28 cells; 8 muscles), local applications of 200 μm SP induced an increase in intracellular Ca2+ (261 ± 29 %). On average, the Ca2+ response began 30.9 ± 2.4 s after the application of SP and was sustained for several seconds. Figure 1A and B shows three PSCs loaded with fluo-3 at rest (A1), 40 s after local applications of SP (200 μm, 10 pulses; A2) and during the recovery period (A3). In 67 % of cells, oscillations were observed after the initial Ca2+ response decayed to baseline. Similar Ca2+ responses could be elicited by consecutive applications of SP at intervals of 10 min (data not shown), indicating that the Ca2+ response induced by SP did not desensitize or that recovery from any desensitization that might have occurred had already taken place. Similar Ca2+ responses were observed when presynaptic nerve terminal Ca2+ channels were blocked with ω-CgTx (1 μm) and when postsynaptic cholinergic receptors were blocked with α-BuTx (1.3 μm). These observations suggest that the responses induced by SP were not caused by the indirect contribution of pre- and postsynaptic elements.

Because SP undergoes a large dilution by the perfusate after its extrusion from the pipette, a more precise evaluation of the concentration of SP required to induce Ca2+ responses was determined by direct bath application. As shown in Fig. 1C, 500 nM to 1 μm SP was sufficient to induce similar Ca2+ responses to that observed with 200 μm SP in the pipette (9 of 11 cells responded). This is within the range of the SP concentration used by Evans et al. (1986) on Schwann cells at the giant axon of the squid.

Characterization of neurokinin receptors on PSCs

Three tachykinin receptor types have been characterized pharmacologically (NK-1, NK-2 and NK-3). Each type will bind any of the natural mammalian tachykinins but SP has preferential affinities for the NK-1 type (Nakanishi, 1991; Maggi et al. 1993).

The carboxy fragment of SP (SP(6–11)) was used as a first attempt to characterize the receptor involved in induction of Ca2+ responses by SP, since SP(6–11) is known to act preferentially on NK-1 receptors with a higher potency than SP itself (Bury & Mashford, 1976; Piercey et al. 1982). Figure 2A shows four PSCs before (A1) and shortly after local applications of SP(6–11) (A2), and during recovery (A3). When applied locally, 20 μm SP(6–11) (3 pulses) evoked Ca2+ responses of 275 ± 17 %, in fifty-eight of sixty-five cells tested (89 %; 19 muscles). The mean amplitude of Ca2+ responses evoked by the SP C-terminal fragment was identical to that found with SP (P > 0.05, Student's t test). A striking difference, however, was that SP(6–11) induced Ca2+ responses with shorter delays (Fig. 2B), since the onset was only 12.5 ± 1.6 s, which was statistically different from the delay obtained with SP (P < 00001, Student's t test). When applied on the same cell (Fig. 2C), SP(6–11) (20 μm) induced a Ca2+ response with a delay of 20.3 s in comparison to 33.9 s for SP (200 μm). No Ca2+ response was observed using the amino terminal fragment of SP, SP(1–7) (20 μm, 3 pulses; data not shown).

We next used septide, a specific NK-1 agonist (Torrens et al. 1995; Hastrup & Schwartz, 1996) to confirm that SP acts through this type of receptor. As shown in Fig. 2D, local applications of septide (20 μm) induced a mean Ca2+ response of 204 ± 17 % in nineteen of the twenty-one cells tested. These values are not significantly different from those observed with SP (P > 0.05, ANOVA).

To test further whether SP acts on an NK-1 receptor, a potent and selective non-peptide NK-1 receptor antagonist, SR140333, was used (Emonds-Alt et al. 1993). Figure 3A shows that in PSCs previously confirmed to be responsive to SP, Ca2+ responses were no longer observed in the presence of SR140333 (1–4 μm) in the perfusate, following local applications of 200 μm SP (9 ± 3 % in 22 cells; 3 muscles; P < 0.05, ANOVA) or 20 μm SP(6–11) (12 ± 7 % in 14 cells; 2 muscles) (Fig. 3A). However, the ability of muscarine (10 μm) to induce Ca2+ responses was not affected by the presence of this NK-1 antagonist. The blockade of NK-1 receptors was reversible since Ca2+ responses were induced by SP upon washout of SR140333 (Fig. 3B). It is unlikely that the blockade of the response was due to an indirect modulation of the synapse, since SR 140333 had no effect on transmitter release per se (data not shown).

Similar results were obtained in the presence of another NK-1 antagonist, LY303870 (4 μm), where the mean response induced by local application of SP (200 μm) was only 26 ± 10 % in all cells tested (3 muscles, 14 cells; P < 0.05, ANOVA). Hence, the results obtained with the various agonists and antagonists suggest that SP acts on an NK-1 receptor.

SP-mediated release of Ca2+ from intracellular stores via a PTX-sensitive G-protein

NK-1 receptors are usually coupled to PTX-insensitive GTP binding proteins (G-proteins). To test this possibility, muscles were incubated in PTX which ADP-ribosylates the α-subunit of certain G-proteins and prevents their activation. In all cells exposed to PTX treatment (Fig. 4A), no changes in Ca2+ levels were detected in response to local applications of 200 μm SP (14 ± 5 % in 22 cells; 3 muscles; P < 0.05, ANOVA). This blockade was not due to the inability of cells to generate any responses after such treatment since local applications of muscarine (Fig. 4A, muscarine), known to induce the release of Ca2+ through a PTX-insensitive mechanism (Robitaille et al. 1997), induced Ca2+ responses in the same cells. In addition, in control muscles incubated overnight in normal Ringer solution without PTX, SP was still able to induce Ca2+ responses similar to control preparations (303 ± 33 %; 15 cells; 3 muscles; P > 0.05, ANOVA). Hence, these results suggest that, unlike the classical NK-1 receptor, the receptors activated by SP at PSCs are sensitive to PTX.

image

Figure 4. . Cellular mechanisms of SP-evoked Ca2+ responses in PSCs

Confocal images of Ca2+ responses in PSCs to local applications of SP. A, images of 2 PSCs incubated for 14 h in a PTX solution (2 μg ml−1) before (Rest) and 25 s after local applications of SP (200 μm). Note the absence of Ca2+ responses in these conditions. However, Ca2+ responses could be induced by muscarine (10 μm) in the same cells. B, images showing 2 PSCs, from a different preparation to A, at resting level (Rest) and 33 s after local applications of SP following 20 min of perfusion with Mg2+ Ringer solution. Ca2+ responses were still induced by SP when no Ca2+ was added to the saline. Scale bars, 10 μm.

Download figure to PowerPoint

To determine the origin of the Ca2+ mobilized by SP, Mg2+ Ringer solution was perfused 20 min prior to SP applications. As shown in Fig. 4B, Ca2+ responses induced by SP when no external Ca2+ was added were on average 258 ± 26 % (23 cells; 4 muscles), which were not significantly different from the value obtained with 200 μm SP in normal Ringer solution (P > 0.05, ANOVA). Taken together, these data suggest that SP acts through a receptor that activates a PTX sensitive G-protein and induces the release of Ca2+ from internal stores.

Is SP active through a degradation product?

The shorter delay of Ca2+ responses induced by the fragment SP(6–11) suggests that a hydrolysis product of SP may be the active substance. If this is the case, a slowly hydrolysable agonist of SP should induce Ca2+ responses when applied for a long period of time (bath-applied) but may not induce a Ca2+ response when applied briefly (local applications) since it may not undergo sufficient degradation before it is washed away from the NMJ. We used the SP analogue, [Sar9,Met(O2)11]-SP, an NK-1 agonist that is more resistant to endopeptidase activity and more stable in vitro (Regoli et al. 1994). As shown in Fig. 5A, prolonged application (2 min) of [Sar9, Met(O2)11]-SP induced Ca2+ responses in 92 % of cells tested (12/13 cells) with a mean response of 207 ± 20 % in the responding cells. This value was not significantly different from the size of the responses induced by SP (P > 0.05, ANOVA). However, brief applications of 200 μm[Sar9,Met(O211]-SP failed to induce Ca2+ responses (10 ± 10 %; 5 cells; 2 muscles) in the same cells in which a comparable concentration of SP was effective in inducing Ca2+ responses (P= 0.003, Student's paired t test) (Fig. 5B).

image

Figure 5. . SP is active through a degradation product

A, confocal images of Ca2+ responses to a bath application (2 min) of [Sar9, Met(O2)11]-SP, a SP analogue more resistant to enzymatic degradation. Images show two cells before (Control) and at the peak of the Ca2+ response induced by [Sar9, Met(O2)11]-SP (200 μm) which occurred after 120 s, B, Ca2+ responses to local applications of [Sar9, Met(O2)11]-SP. Images show one cell before (Rest), 20 s after local applications of [Sar9, Met(O2)11]-SP (200 μm), and at the peak of the Ca2+ response evoked by local applications of SP (200 μm). C, Ca2+ responses obtained by local applications of SP, 20 min after addition of 3 μg ml−1 neostigmine. Images show 2 cells before (Rest) and 35 s after local applications of SP (200 μm). Note that Ca2+ responses were still evoked by SP in the presence of neostigmine. Scale bars, 10 μm.

Download figure to PowerPoint

Acetylcholinesterase (AChE) is abundant and highly functional at the frog NMJ and has been shown to be capable of hydrolysing SP (Chubb et al. 1980). If AChE were responsible for the degradation of SP, the AChE inhibitor neostigmine should be effective in protecting SP from hydrolysis and, hence, in preventing the induction of Ca2+ responses by SP. As shown in Fig. 5C, the presence of neostigmine (11 μm) did not block the effect of SP on evoked Ca2+ responses (245 ± 55 %; 7 cells; 2 muscles; P > 0.05, ANOVA). The effectiveness of neostigmine was confirmed by the increase in the amplitude and duration of postsynaptic responses (data not shown). Thus, if degradation of SP is important for its action on PSCs, that degradation is not prevented by neostigmine.

SP-modulated Ca2+ responses induced by mACh and ATP receptors in PSCs

It was shown that SP modulates cholinergic and purinergic mechanisms in various cells (Krnjevic & Morris, 1974; Silva et al. 1988; Simmons et al. 1990). Since mACh and ATP receptors are known to activate PSCs (Jahromi et al. 1992; Robitaille, 1995; Robitaille et al. 1997), it was of interest to determine the action of SP on these receptors. First, the effects of SP on the activation of mACh receptors was tested. As shown in Fig. 6A (□), 10 μm muscarine induced Ca2+ responses of 350 ± 35 % in 94 % of the cells tested (17/18 cells; 5 muscles). However, only 22 % of cells tested (6/27 cells; 8 muscles) were responsive to muscarine when preceded (10 min before) by local applications of 20 μm SP (▪), a concentration that did not induce Ca2+ responses. Furthermore, the size of Ca2+ responses in responding cells was reduced to 154 ± 42 %, which was significantly different from Ca2+ responses evoked by muscarine not preceded by SP (P= 0.005, Student's t test). Neither the normalized shape nor the time course of the responses was affected (data not shown). The effect of SP on muscarine-induced ACh receptors seems to occur quickly, since the reduction of the Ca2+ responses is maximal and complete within a few minutes of the application of SP. In addition, it was unlikely that the loss of responsiveness to muscarine was due to a depletion of internal stores of Ca2+ since, in these experiments, SP was used at a concentration which did not induce any Ca2+ responses.

Second, we investigated whether exposure to SP affected Ca2+ responses induced by ATP. As shown in Fig. 6B (□), ATP (20 μm) induced mean Ca2+ responses of 292 ± 22 % in responding cells (40 cells; 6 muscles). Ten minutes following local applications of 20 μm SP, the percentage of cells responding to local applications of 20 μm ATP (▪) was not affected (22/23 responding cells; 5 muscles). However, the mean Ca2+ responses (209 ± 27 %) were significantly different from control (P= 0.02, Student's t test). Neither the normalized shape nor the time course of the responses was affected (data not shown), which is similar to the effects of SP on muscarine-induced responses. As with muscarine-induced responses, the reduction of Ca2+ responses by ATP was probably not due to depletion of internal stores of Ca2+, since 20 μm SP did not induce Ca2+ responses in these experiments.

Modulation by SP of ATP-induced responses occurred over a long period of time. Figure 6C shows that, on average, consecutive applications of 10 μm ATP (10 min intervals) induced Ca2+ responses of similar amplitude for up to 80 min. However, in a separate set of experiments, bath application of SP (500 nM) prior to application of ATP reduced the ATP-induced Ca2+responses (Fig. 6D). This effect was maximal in about 30–40 min and persisted for more than 120 min. Our results support a role for SP as an inhibitory neuromodulator of PSCs.

Is endogenous SP responsible for the rundown of nerve-evoked Ca2+ responses?

A rundown of nerve-evoked Ca2+ responses in PSCs was observed with a repetitive train of stimuli (Jahromi et al. 1992; Robitaille, 1995). Since SP downregulates muscarine and ATP-induced Ca2+ responses, and since both neurotransmitters contribute to the nerve-evoked responses in PSCs, we tested the hypothesis that the rundown of nerve-evoked Ca2+ responses may be caused by the action of SP.

The rundown of nerve-evoked Ca2+ responses in PSCs induced by repetitive nerve stimulations was studied in the absence and presence of an NK-1 receptor blocker perfused throughout the experiment to prevent the activation of the receptor by endogenous SP. As previously reported, successive trains of stimuli to the motor nerve induced smaller Ca2+ responses (Fig. 7A, continuous trace; data from second and third trains not shown; Fig. 7B, □). Often, no Ca2+ response could be evoked by a fourth train. However, in the presence of the NK-1 antagonist SR140333 (1 μm) (Fig. 7A, dotted trace, data from second and third train not shown; Fig. 7B, ▪), the rundown of nerve-evoked Ca2+ responses was less pronounced and large responses could still be evoked by a fourth train of stimuli.

image

Figure 7. . Endogenous SP modulates nerve-evoked Ca2+ signals in PSCs

A, Ca2+ changes in PSCs over time before, during (bar), and after release of transmitter induced by repetitive nerve stimulation (50 Hz, 30 s). In control (continuous trace), nerve-evoked Ca2+ responses were characterized by progressive rundown during successive stimulations at intervals of 25 min (data from second and third trains not shown). Often, no Ca2+ response could be evoked by a fourth train. However, in the presence of the NK-1 blocker SR140333 (1 μm) (dotted trace) the rundown of nerve-evoked Ca2+ responses was less pronounced (data from second and third trains not shown). In some cases, oscillations were also observed. B, mean Ca2+ responses evoked by repetitive trains of stimuli to the motor nerve (50 Hz, 30 s; at intervals of 25 min) for controls (□) and cells in the presence of SR140333 (▪). Note the persistence of the Ca2+ response evoked by the third and fourth train of stimuli in the presence of SR140333. Significant differences from the control values are indicated by asterisks: *P= 0.02, **P= 0.005, Student's t test. C, changes in fluo-3 fluorescence over time in 3 PSCs (arrowheads) induced by repetitive nerve stimulation (50 Hz, 30 s) to evoke the release of neurotransmitters. Ca2+ responses are associated with their respective PSCs by an arrow. SP (20 μm) was locally applied to the top cell. Scale bar, 10 μm.

Download figure to PowerPoint

The release of peptides from dense core vesicles is slower than the release of classical neurotransmitters (Huang & Neher, 1996). Hence, one would predict that the downregulation by SP of nerve-evoked Ca2+ responses in PSCs should not occur on the first train of stimuli since SP would be released after ATP and ACh, the neurotransmitters responsible for the production of the Ca2+ responses in PSCs. Moreover, the effects of ACh and ATP should take place before SP since ACh and ATP have much shorter action latencies (∼1 s) compared with SP (∼30 s). Consistent with the delayed release of peptides, the first nerve-evoked Ca2+ responses were unaffected by the presence of the NK-1 antagonist (Fig. 7A and B). SR140333 had no effect on the amount of transmitter release, thus ruling out the possibility that the changes in nerve-evoked Ca2+ responses by SR140333 were due to a change in transmitter release (data not shown). These results suggest that endogenous SP or another related peptide activated NK-1-like receptors, which caused a large portion of the rundown observed.

To seek further evidence for this possibility, the ability of PSCs to respond to nerve-evoked release of neurotransmitters was tested after local applications of SP. SP (20 μm) was locally applied to one PSC in a field and the changes in fluorescence induced by nerve stimulation were compared with the changes observed in neighbouring, non-treated PSCs. To limit the spill-over of SP onto neighbouring cells, five positive pressure pulses (200 ms, 20 psi) were applied at intervals of 5 s. As shown in Fig. 7C, previous application of SP significantly reduced the size of the Ca2+ responses induced by nerve stimulations. Indeed, from three preparations, the mean response was 312 ± 24 % in untreated PSCs (6 cells) but was only 65 ± 27 % in PSCs to which SP was applied prior to nerve stimulation (4 cells). These results indicate that SP downregulates the nerve-evoked Ca2+ responses in PSCs and further support the possibility that endogenous SP is involved in PSC modulation.

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements

We present evidence that SP activates NK-1-like receptors at PSCs of the frog NMJ and induces the release of Ca2+ from internal stores via a PTX-sensitive G-protein. SP actions reduce the ability of mACh and ATP receptors to increase the level of intracellular Ca2+ and the release of an endogenous SP-like substance contributes to the rundown of the nerve-evoked Ca2+ responses.

SP receptors at PSCs of the frog NMJ

SP induces an elevation of intracellular Ca2+ in several glial cell types (Delumeau et al. 1991; Heath et al. 1994). In cultured rodent astrocytes and human astrocytoma cell lines, activation of NK-1 receptors increases IP3 synthesis, leading to the release of Ca2+ from intracellular stores (Marriott et al. 1991; Beaujouan et al. 1991; Johnson & Johnson, 1992). SP also hyperpolarizes the plasma membrane of Schwann cells at the squid giant axon (Evans & Villegas, 1988; Evans et al. 1986, 1990). At PSCs of the frog NMJ, SP induces the release of Ca2+ from internal stores as indicated by the lack of effect of removing extracellular Ca2+. The blockade of SP-induced Ca2+ responses by SR140333 and LY303870, specific NK-1 receptor antagonists which do not affect either NK-2 or NK-3 receptors (Emonds-Alt et al. 1993; Cellier et al. 1996; Iyengar et al. 1997), supports the possibility that SP acts on NK-1 receptors in PSCs. Moreover, the fragment SP(6–11), known to preferentially activate NK-1 receptors, and the NK-1 agonists septide and [Sar9, Met(O2)11]-SP induce Ca2+ responses in PSCs. Hence, these results indicate that, in PSCs, SP mediates its action by activating an NK-1 receptor, causing the release of Ca2+ from internal stores.

However, part of the characterization of the SP receptors in PSCs differs from the classical view of the NK-1 type of receptor. Indeed, in most studies, it has been reported that SP actions lack sensitivity to PTX (Delumeau et al. 1991) whereas we observed that PTX prevented all Ca2+ responses to SP. It is unlikely that the actions of PTX on PSCs are due to non-specific effects since SP induced similar responses in control muscles left overnight in frog Ringer solution (no PTX) and in acutely prepared muscles. Moreover, muscarine induces Ca2+ responses after PTX treatment, which is consistent with its insensitivity to PTX (Robitaille et al. 1997). In mast cells, a direct action of SP, not mediated by any receptors, induced phospholipase C activation and released Ca2+ from internal stores via a PTX-sensitive G-protein (Bueb et al. 1990). Such a mechanism is unlikely in PSCs since SP effects are blocked by the NK-1 antagonists SR140333 and LY303870. This difference in the sensitivity to PTX of SP receptors, combined with the potent effects of NK-1 receptor agonists and antagonists on SP-induced Ca2+ responses, suggests that the receptors on PSCs are members of a subclass of NK-1 receptors. The differences observed with different preparations may be related to the cell types or the conditions of the preparations (e.g. culture versusin situ). This may be an indication of the specific functions of this type of receptor in perisynaptic glial cells, since the involvement of different G-proteins implies different actions exerted by the receptor.

The molecular machinery associated with the NK-1 receptor of PSCs differs from the muscarinic and purinergic receptors found on these cells, as indicated by the induction of oscillations by SP which are not usually observed in Ca2+ responses induced by muscarine or ATP (Robitaille, 1995; Robitaille et al. 1997). This may suggest that, although a single type of internal store appears to be present in PSCs, different Ca2+ mechanisms may be involved and different second messenger cascades may be linked to the NK-1 receptors and the muscarinic and purinergic receptors.

Hydrolysis of SP

The evidence obtained from PSCs indicates that hydrolysis of SP may be a necessary part of its action. This possibility is supported by two observations. First, there was a long delay (∼30 s) between local applications of SP and the onset of the response, consistent with SP degradation producing an active product, and the carboxyl fragment of SP induced Ca2+ responses with a shorter delay. Second, the slowly hydrolysable agonist, [Sar9, Met(O2)11]-SP induced Ca2+ responses only when bath-applied for a long period of time (>2 min) and not by local, brief applications. The final proof of SP hydrolysis at the frog NMJ will require identification of the proteolytic enzyme involved in this process. An obvious candidate would have been AChE which is present at the frog NMJ and is known to hydrolyse SP (Chubb et al. 1980). However, our results indicate that the interaction between AChE and neostigmine had no effect on SP-induced Ca2+ responses. Although SP(6–11) is an excellent candidate for the active component, we cannot rule out the possibility that other fragments are also active. However, the amino terminal fragment SP(1–7) had no effect on PSCs.

SP modulates muscarine- and ATP-induced Ca2+ responses

We found that SP reduced the ability of muscarine and ATP to induce Ca2+ responses. It has been reported that SP inhibits ACh-evoked nicotinic excitation of Renshaw cells (Krnjevic & Morris, 1974) and decreases the response to M1 but increases the response to M2 receptors in rat vas deferens (Silva et al. 1988). A desensitization of the inhibition of the M-current in sympathetic neurons by SP has also been reported (Simmons et al. 1990).

The mechanisms by which SP could modulate the muscarine-induced changes in Ca2+ are unclear because the type of mACh receptor on PSCs does not correspond to any of the five known types (Robitaille et al. 1997). It is unlikely that SP modulation occurs directly at the level of the intracellular Ca2+ stores since there is only one type of internal store associated with an IP3 receptor (R. Robitaille, unpublished observation). Hence, preventing the activation of IP3 receptors should lead to a similar decline in Ca2+ responses mediated by all receptors. This is not the case since responses were still induced by ATP (although at a reduced level) and by SP. However, preliminary data indicate that SP actions are mediated by the activation of protein kinase C since specific blockers of this enzyme prevented the action of SP (K. Olofsdotter, M. J. Bourque and R. Robitaille, unpublished observations).

The modulation by SP of ATP-induced Ca2+ responses appears to occur by different mechanisms, as suggested by the fact that SP only reduces the size of the Ca2+ responses, not the number of responding cells. In addition, the full effects on ATP-evoked responses occur slowly (> 30 min) whereas the effects on muscarine-induced responses occur within minutes. Since Ca2+ responses induced by ATP are mediated by three types of receptors (P2x, P2Y and A1; Robitaille, 1995), the effect of SP could be related to any of these receptors or a combination. For instance, SP was shown to potentiate the inhibitory action of adenosine in dorsal horn neurons (De Koninck et al. 1994) and reduce the L-type Ca2+ current in isolated colonic myocytes (Lee et al. 1995). This is an attractive possibility since L-type Ca2+ channels and A1 adenosine receptors are present in PSCs (Robitaille, 1995; Robitaille et al. 1996). Since mACh and ATP receptors regulate different cellular processes (Georgiou et al. 1994), the differential effects of SP on these receptors suggest that it modulates different mechanisms of cellular machinery in PSCs.

Modulation of PSCs by endogenous peptides

SP and calcitonin gene-related peptide are present in the dense-core vesicles of the presynaptic terminal of the frog NMJ (Matteoli et al. 1990). In the present study, evidence for the release of endogenous SP, or a related peptide, is supported by the observations that the rundown of the nerve-evoked Ca2+ responses following successive trains of stimuli was greatly reduced when the activation of NK-1 receptors was prevented by the antagonist SR140333 and that SP selectively reduced the nerve-evoked Ca2+ responses in PSCs to which it was locally applied. Moreover, the lack of effect of the NK-1 antagonist on the first nerve-evoked Ca2+ responses further supports the concept that the release of peptides is slow and delayed in comparison to the release of the classical neurotransmitters. However, it is not known which receptor types (mACh, ATP or both) are modulated by the activity of the NK-1 receptor during transmitter release.

SP is known to modulate transmitter release at the frog NMJ. Hence, the effects of SP on the PSCs may have been indirectly mediated by the action of SP on the presynaptic terminal which would then affect the PSCs. This is unlikely since SP had the same effects on PSCs when transmitter release and postsynaptic activity were blocked using blockers of presynaptic Ca2+ channels and postsynaptic ACh receptors.

The involvement of other endogenous substances besides SP with NK-1 receptors cannot be ruled out since these receptors can be activated by various neurokinins, such as NKA and NKB. Also, the remaining rundown observed after the first train may be related to the desensitization of Ca2+ responses produced by the mACh receptor (Robitaille et al. 1997) or may be caused by other endogenous substances such as calcitonin gene-related peptide that could act on their own receptor subtypes. These data allow us to conclude that the activation of NK-1-like receptors by an endogenous SP-like substance modulates the responsiveness of PSCs to other neurotransmitters.

Synapse-glia interactions

Synaptic modulation of perisynaptic glial cells occurs with the release of neurotransmitters such as glutamate, ACh or ATP. Our results suggest that the release of endogenous peptides is also involved, acting as modulators. Hence, we conclude that synapse-glia interactions are modulated by endogenous peptides similar to many other functions in the nervous system. Interestingly, it has been shown that SP application at the frog NMJ has pre- and postsynaptic effects (Akasu, 1986). This raises the possibility that the action of SP could be a consequence of its modulation of PSCs which would then modulate the activity of the synapse.

  • Akasu, T. (1986). The effects of substance P on neuromuscular transmission in the frog. Neuroscience Research 3, 275284.DOI: 10.1016/0168-0102(86)90020-9
  • Beaujouan, J. C., Teutsch, B., Saffroy, M., Petitet, F., Torrens, Y. & Glowinski, J. (1991). NK-1 receptors are the only class of tachykinin receptors found on mouse cortical astrocytes. Peptides 12, 813820.
  • Bueb, J. L., Mousli, M., Bronner, C., Rouot, B. & Landry, Y. (1990). Activation of Gi-like proteins, a receptor-independent effect of kinins in mast cells. Molecular Pharmacology 38, 816822.
  • Bury, R. W. & Mashford, M. L. (1976). Biological activity of C-terminal partial sequences of substance P. Journal of Medical Chemistry 19, 854856.
  • Cellier, E., Fayolle, C., Hipskind, P. A., Iyengar, S. & Couture, R. (1996). Peripheral effects of three novel non-peptide tachykinin NK1 receptor antagonists in the anaesthetized rat. European Journal of Pharmacology 318, 377385.
  • Chiu, S. Y. & Kriegler, S. (1994). Neurotransmitter-mediated signaling between axons and glial cells. Glia 11, 191200.
  • Chubb, I. W., Hodgson, A. J. & White, G. H. (1980). Acetylcholinesterase hydrolyzes substance P. Neuroscience 5, 20652072.
  • Dani, J. W., Chernjavsky, A. & Smith, S. J. (1992). Neuronal activity triggers calcium waves in hippocampal astrocyte networks. Neuron 8, 429440.DOI: 10.1016/0896-6273(92)90271-E
  • De Koninck, Y., Salter, M. W. & Henry, J. L. (1994). Substance P released endogenously by high-intensity sensory stimulation potentiates purinergic inhibition of nociceptive dorsal horn neurons induced by peripheral vibration. Neuroscience Letters 176, 128132.DOI: 10.1016/0304-3940(94)90887-7
  • Delumeau, J. C., Petitet, F., Cordier, J., Glowinski, J. & Premont, J. (1991). Synergistic regulation of cytosolic Ca2+ concentration in mouse astrocytes by NK1 tachykinin and adenosine agonists. Journal of Neurochemistry 57, 20262035.
  • Emonds-Alt, X., Doutremepuich, J. D., Heaulme, M., Neliat, G., Santucci, V., Steinberg, R., Vilain, P., Bichon, D., Ducoux, J. P., Proietto, V., Van Broeck, D., Soubrié, P., Le Fur, G. & Brelière, J. C. (1993). In vitro and in vivo biological activities of SR140333, a novel potent non-peptide tachykinin NK1 receptor antagonist. European Journal of Pharmacology 250, 403413.DOI: 10.1016/0014-2999(93)90027-F
  • Evans, P. D., Reale, V., Merzon, R. M. & Villegas, J. (1990). Substance P modulation of the membrane potential of the Schwann cell of the squid giant nerve fibre. Glia 3, 393404.
  • Evans, P. D., Reale, V. & Villegas, J. (1986). Peptidergic modulation of the membrane potential of the Schwann cell of the squid giant nerve fibre. Journal of Physiolology 379, 6182.
  • Evans, P. D. & Villegas, J. (1988). The action of vasoactive intestinal peptide antagonists on peptidergic modulation of the squid Schwann cell. Journal of Experimental Biolology 138, 259269.
  • Georgiou, J., Robitaille, R., Trimble, W. S. & Charlton, M. P. (1994). Synaptic regulation of glial protein expression in vivo. Neuron 12, 443455.DOI: 10.1016/0896-6273(94)90284-4
  • Guard, S. & Watson, S. P. (1991). Tachykinin receptor types: classification and membrane signaling mechanisms. Neurochemistry International 18, 149165.DOI: 10.1016/0197-0186(91)90180-L
  • Hastrup, H. & Schwartz, T. W. (1996). Septide and neurokinin A are high-affinity ligands on the NK-1 receptor: evidence from homologous versus heterologous binding analysis. FEBS Letters 399, 264266.DOI: 10.1016/S0014-5793(96)01337-3
  • Heath, M. J., Womack, M. D. & MacDermott, A. B. (1994). Substance P elevates intracellular calcium in both neurons and glial cells from the dorsal horn of the spinal cord. Journal of Neurophysiology 72, 11921198.
  • Huang, L.-Y. & Neher, E. (1996). Ca2+-dependent exocytosis in the somata of dorsal root ganglion neurons. Neuron 17, 135145.DOI: 10.1016/S0896-6273(00)80287-1
  • Iyengar, S., Hipskind, P. A., Gehlert, D. R., Schober, D., Lobb, K. L., Nixon, J. A., Helton, D. R., Kallman, M. J., Boucher, S., Couture, R., Li, D. L. & Simmons, R. M. A. (1997). LY303870, a centrally active neurokinin-1 antagonist with a long duration of action. Journal of Pharmacological and Experimental Therapeutics 280, 774785.
  • Jahromi, B. S., Robitaille, R. & Charlton, M. P. (1992). Transmitter release increases intracellular calcium in perisynaptic Schwann cells in situ. Neuron 8, 10691077.DOI: 10.1016/0896-6273(92)90128-Z
  • Johnson, C. L. & Johnson, C. G. (1992). Characterization of receptors for substance P in human astrocytoma cells: Radioligand binding and inositol phosphate formation. Journal of Neurochemistry 58, 471477.
  • Krnjevic, K. & Morris, M. E. (1974). An excitatory action of substance P on cuneate neurones. Canadian Journal of Physiology and Pharmacology 52, 736744.
  • Lee, H. K., Shuttleworth, C. W. & Sanders, K. M. (1995). Tachykinins activate nonselective cation currents in canine colonic myocytes. American Journal of Physiology 269, C13941401.
  • Maggi, C. A., Patacchini, R., Rovero, P. & Giachetti, A. (1993). Tackykinin receptors and tachykinin receptor antagonists. Journal of Autonomic Pharmacology 13, 2393.
  • Marriott, D. R., Wilkin, G. P. & Wood, J. N. (1991). Substance P-induced release of prostaglandins from astrocytes: regional specialisation and correlation with phosphoinositol metabolism. Journal of Neurochemistry 56, 259265.
  • Matteoli, M., Haimann, C. & De Camilli, P. (1990). Substance P-like immunoreactivity at the frog neuromuscular junction. Neuroscience 37, 271275.DOI: 10.1016/0306-4522(90)90213-N
  • Nakanishi, S. (1991). Mammalian tachykinin receptors. Annual Review of Neuroscience 14, 123136.DOI: 10.1146/annurev.ne.14.030191.001011
  • Nedergaard, M. (1994). Direct signaling from astrocytes to neurons in cultures of mammalian brain cells. Science 263, 17681771.
  • Parpura, V., Basarsky, T. A., Liu, F., Jeftinija, K., Jeftinija, S. & Haydon, P. G. (1994). Glutamate-mediated astrocyte-neuron signalling. Nature 369, 744747.DOI: 10.1038/369744a0
  • Pernow, B. (1983). Substance P. Pharmacological Reviews 35, 85141.
  • Piercey, M. F., Dobry, P. J., Einspahr, F. J., Schroeder, L. A. & Masiques, N. (1982). Use of substance P fragments to differentiate substance P receptors of different tissues. Regulatory Peptides 3, 337349.DOI: 10.1016/0167-0115(82)90057-X
  • Regoli, D., Boudon, A. & Fauchére, J. L. (1994). Receptors and antagonists for substance P and related peptides. Pharmacological Reviews 46, 551599.
  • Reist, N. E. & Smith, S. J. (1992). Neurally evoked calcium transients in terminal Schwann cells at the neuromuscular junction. Proceedings of the National Academy of Sciences of the USA 89, 76257629.
  • Robitaille, R. (1995). Purinergic receptors and their activation by endogenous purines at perisynaptic glial cells of the frog neuromuscular junction. Journal of Neuroscience 15, 71217131.
  • Robitaille, R., Bourque, M. J. & Vandaele, S. (1996). Localization of L-type Ca2+ channels at perisynaptic glial cells of the frog neuromuscular junction. Journal of Neuroscience 16, 148158.
  • Robitaille, R., Jahromi, B. S. & Charlton, M. P. (1997). Muscarinic Ca2+ responses resistant to muscarinic antagonists at perisynaptic Schwann cells of the frog neuromuscular junction. The Journal of Physiology 504, 337347.
  • Silva, W. I., Miranda, H. F., Wolstenholme, W. W., Cuevas, N. & Ucros, A. (1988). Neuromodulatory actions of substance P on the muscarinic receptors of the vas deferens of the rat. Neuropeptides 11, 153157.DOI: 10.1016/0143-4179(88)90068-6
  • Simmons, M. A., Becker, J. B. & Mather, R. J. (1990). Desensitization of the inhibition of the M-current in sympathetic neurons: effects of ATP analogs, polyanions, and multiple agonist applications. Neuron 4, 557562.DOI: 10.1016/0896-6273(90)90113-T
  • Torrens, Y., Beaujouan, J. C., Saffroy, M. & Glowinski, J. (1995). Involvement of septide-sensitive tachykinin receptors in inositol phospholipid hydrolysis in the rat urinary bladder. Peptides 16, 587594.DOI: 10.1016/0196-9781(95)00016-D

Acknowledgements

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements

This work was supported by an MRC grant (no. MT12057) and an FCAR team grant to R. R. R. R. was an MRC and a FRSQ Scholar and an Alfred P. Sloan fellow. M. J. B. was supported by a studentship from the FCAR groupe de recherche sur le système nerveux central. We thank Vincent Castellucci, John Georgiou, Réjean Couture and Allan Smith for reading the manuscript. We also thank Réjean Couture for helpful discussion and for kindly providing SR140333 and LY303870.