Portal de Periódicos da CAPES

Presynaptic target of Ca 2+ action on neuropeptide and acetylcholine release in Aplysia californica

2001; Wiley; Volume: 535; Issue: 3 Linguagem: Inglês

10.1111/j.1469-7793.2001.00647.x

1469-7793

Kiyoshi Ohnuma, Matthew D. Whim, Richard D. Fetter, Leonard K. Kaczmarek, Robert S. Zucker,

Olfactory and Sensory Function Studies

The Journal of PhysiologyVolume 535, Issue 3 p. 647-662 Free Access Presynaptic target of Ca2+ action on neuropeptide and acetylcholine release in Aplysia californica Kiyoshi Ohnuma, Kiyoshi Ohnuma Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720, USASearch for more papers by this authorMatthew D. Whim, Matthew D. Whim Department of Pharmacology, University College London, London WC1E 6BT, UKSearch for more papers by this authorRichard D. Fetter, Richard D. Fetter Howard Hughes Medical Institute, University of California, Berkeley, CA 94720, USASearch for more papers by this authorLeonard K. Kaczmarek, Leonard K. Kaczmarek Department of Pharmacology, Yale University School of Medicine, New Haven, CT 06520, USASearch for more papers by this authorRobert S. Zucker, Robert S. Zucker Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720, USASearch for more papers by this author Kiyoshi Ohnuma, Kiyoshi Ohnuma Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720, USASearch for more papers by this authorMatthew D. Whim, Matthew D. Whim Department of Pharmacology, University College London, London WC1E 6BT, UKSearch for more papers by this authorRichard D. Fetter, Richard D. Fetter Howard Hughes Medical Institute, University of California, Berkeley, CA 94720, USASearch for more papers by this authorLeonard K. Kaczmarek, Leonard K. Kaczmarek Department of Pharmacology, Yale University School of Medicine, New Haven, CT 06520, USASearch for more papers by this authorRobert S. Zucker, Robert S. Zucker Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720, USASearch for more papers by this author First published: 01 September 2001 https://doi.org/10.1111/j.1469-7793.2001.00647.xCitations: 19 Corresponding author R. S. Zucker: Molecular and Cell Biology Department, 111 Life Sciences Addition, University of California, Berkeley, CA 94720-3200, USA., Email: zucker@socrates.berkeley.edu AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract 1 When buccal neuron B2 of Aplysia californica is co-cultured with sensory neurons (SNs), slow peptidergic synapses are formed. When B2 is co-cultured with neurons B3 or B6, fast cholinergic synapses are formed. 2 Patch pipettes were used to voltage clamp pre- and postsynaptic neurons and to load the caged Ca2+ chelator o-nitrophenyl EGTA (NPE) and the Ca2+ indicator BTC into presynaptic neurons. The relationships between presynaptic [Ca2+]i and postsynaptic responses were compared between peptidergic and cholinergic synapses formed by cell B2. 3 Using variable intensity flashes, Ca2+ stoichiometries of peptide and acetylcholine (ACh) release were approximately 2 and 3, respectively. The difference did not reach statistical significance. 4 ACh quanta summate linearly postsynaptically. We also found a linear dose-response curve for peptide action, indicating a linear relationship between submaximal peptide concentration and response of the SN. 5 The minimum intracellular calcium concentrations ([Ca2+]i) for triggering peptidergic and cholinergic transmission were estimated to be about 5 and 10 μm, respectively. 6 By comparing normal postsynaptic responses to those evoked by photolysis of NPE, we estimate [Ca2+]i at the release trigger site elicited by a single action potential (AP) to be at least 10 μm for peptidergic synapses and probably higher for cholinergic synapses. 7 Cholinergic release is brief (half-width ≈200 ms), even in response to a prolonged rise in [Ca2+]i, while some peptidergic release appears to persist for as long as [Ca2+]i remains elevated (for up to 10 s). This may reflect differences in sizes of reserve pools, or in replenishment rates of immediately releasable pools of vesicles. 8 Electron microscopy revealed that most synaptic contacts had at least one morphologically docked dense core vesicle that presumably contained peptide; these were often located within conventional active zones. 9 Both cholinergic and peptidergic vesicles are docked within active zones, but cholinergic vesicles may be located closer to Ca2+ channels than are peptidergic vesicles. Many neurons contain both classical transmitters, such as ACh and glutamate, as well as peptide co-transmitters (Kupfermann, 1991). There are some differences in the secretion of these two types of transmitters. For instance, release of classical transmitters can occur with a single AP, whereas peptide secretion often requires high frequency bursts of activity (Dutton & Dyball, 1979; Jan & Jan, 1982; Whim & Lloyd, 1989; Cropper et al. 1990; Peng & Horn, 1991; but see Whim et al. 1997). Thus, the relationship between transmitter release and [Ca2+]i may differ between the two transmitter types. With classical transmitters, the amount of transmitter release depends on the third to fourth power of extra- or intracellular [Ca2+] (Dodge & Rahamimoff, 1967; Katz & Miledi, 1970; Dudel, 1981; Augustine et al. 1985; Heidelberger et al. 1994; Landò & Zucker, 1994). This highly non-linear relationship was attributed to three or four successive Ca2+ ions binding to an as yet unspecified target to trigger transmitter release. In contrast, synaptic peptide release appears to depend linearly on [Ca2+]i (Sakaguchi et al. 1991; Peng & Zucker, 1993). The calcium stoichiometry of release from non-neural peptidergic cells has been reported to be higher (Thomas et al. 1993; Proks et al. 1996). In addition, although the [Ca2+]i needed to release classical transmitters is estimated to approach or even exceed 100 μm (Roberts et al. 1990; Adler et al. 1991; Llinás et al. 1992; Landò & Zucker, 1994), the [Ca2+]i required to release peptide transmitters has been suggested to be about 1 μm (Cazalis et al. 1987; Lindau et al. 1992; Peng & Zucker, 1993). Small clear vesicles containing classical transmitters cluster and attach to the presynaptic membrane and form active zones. However, peptide secretion is believed not to occur at active zones (De Camilli & Jahn, 1990; Golding, 1994; Leenders et al. 1999). Comparing peptidergic and classical synapses is difficult, because differences may arise from the different cells releasing the two transmitter types, or from essential differences in the transmitter releasing systems. Whim et al. (1997) produced both peptidergic and cholinergic synapses using the same buccal ganglion neuron, B2, from Aplysia. This neuron synthesizes and releases ACh (Lloyd et al. 1985) and the small cardioactive peptides A and B (SCPA and SCPB; SCPs) (Lloyd et al. 1986), which are localized to dense-core vesicles (DCVs) (Kreiner et al. 1986; Reed et al. 1988). When B2 (or B1) is co-cultured with a SN, a slow peptidergic synapse forms. However, when B2 is co-cultured with neuron B3 (or B6), a cholinergic synapse forms (Whim et al. 1997). Although this peptidergic synapse behaves similarly to other peptidergic synapses, it has a unique property: a single AP is sufficient to induce a postsynaptic response when making a soma-soma synapse (Whim et al. 1997). Using photolabile Ca2+ chelators to manipulate [Ca2+]i, we have compared the Ca2+ sensitivities of the presynaptic targets mediating peptide and classical transmitter release as detected by the choice of postsynaptic partner. Although clear differences were seen between them, the differences were smaller than expected. Electron microscopy of peptidergic synapses revealed that they also have classical synapses containing active zone structures. The identification of peptidergic active zones is novel, and may explain the ability of a single AP to evoke a postsynaptic response. METHODS Cell culture Cells were isolated and maintained by standard techniques (Schacher & Proshansky, 1983; Whim et al. 1997). Aplysia californica weighing ≈1 g and ≈100 g were anaesthetized by injection of isotonic MgCl2 and used for buccal and pedal-pleural ganglia, respectively. Ganglia were removed and incubated in 1 % protease (Sigma type IX) in sterile normal artificial seawater (NASW) at 34 °C for 1.5 or 2.5 h, respectively. Neurons were removed using finely pulled glass probes and maintained in sterile culture medium at room temperature (20-22 °C). The medium consisted of 26 %Aplysia haemolymph; 4 % fetal bovine serum; 70 % NASW, supplemented with penicillin (50 u ml−1), streptomycin (50 μg ml−1), vitamin mixture (0.5 × minimal essential medium (MEM)), and non-essential (0.2 × MEM) and essential amino acids without l-glutamine (0.2 × MEM). Although fetal bovine serum was added to suppress electrical coupling (Carrow & Levitan, 1989), strong electro-coupling sometimes occurred. Such cell pairs were not used for experiments. Neurons generally could be identified by visual criteria alone (Church et al. 1993; Whim & Lloyd, 1994), and any anomalous ganglia were rejected. The NASW contained (mm): 460 NaCl, 10.4 KCl, 55 MgCl2, 11 CaCl2 and 15 Na-Hepes (pH 7.5: ≈1070 mosmol kg−1). All of the experiments used the cultured soma-soma synapse configuration (Haydon, 1988; Whim et al. 1997). Buccal B2, B3, and B6 neurons, and the pleural SNs were isolated and separately maintained in droplets of culture medium on dishes. After more than 1 day the primary neurite had been reabsorbed, and the pre- and postsynaptic cells were joined. The neurons adhered to each other and were used 1-4 days later. Cholinergic pairs consisted of presynaptic B2 and postsynaptic B3 or B6 (to simplify, we call this B2-B3) and peptidergic pairs consisted of presynaptic B2 and postsynaptic SN (B2-SN). Although there was no difference between B3 and B6 as ACh detectors (Whim et al. 1997; Kehoe & McIntosh, 1998), B3 neurons were mainly used. B3 are bigger than B2 neurons and can be more easily distinguished from B2 after combination. For recording, neuron pairs were transferred to a glass-bottomed experimental chamber. Although we did not observe any differences in the calcium dependency of transmission between B2-B3 and B2-B6 or between B1-SN and B2-SN, we cannot rule out the possibility that the postsynaptic target cell could modify the characteristics of transmitter release. Electrical recordings Whole-cell voltage-clamp recordings of both the presynaptic and the postsynaptic neurons were performed with an EPC-9 double patch-clamp amplifier (EPC-9/2) with PULSE software (HEKA, Lambrecht/Pfalz, Germany). Two types of patch pipettes and filling solutions were used. One type of pipette (outer diameter of the tip ≈2 μm) had resistances of ≈1 MΩ measured in NASW, and was used with the normal pipette solution containing (mm): 435 potassium asparate, 70 KCl, 11 glucose, 10 potassium glutathione, 5 Na2-ATP, 0.1 GTP, 1.2 MgCl2, 5 K-Hepes (pH 7.3: ≈1050 mosmol kg−1). Gigohm seals were obtained using this pipette and solution. A second type of pipette had a resistance of ≈0.3 MΩ (outer diameter of the tip ≈5 μm), and was used for loading BTC and NPE into the presynaptic neuron. The filling solution contained (mm): 416 potassium asparate, 70 KCl, 11 glucose, 10 potassium glutathione, 5 Na2-ATP, 0.1 GTP, 1.2 MgCl2, 0.25 BTC, 5 NPE, 3.75 CaCl2 (75 % of NPE), 50 K-Hepes (pH 7.3; ≈1050 mosmol kg−1). Using this pipette, we obtained seal resistances up to 200 MΩ. The holding potential of presynaptic B2 was -46 mV. The holding potential of postsynaptic SN was -31 mV and that of postsynaptic B3 was -36 mV because it was sometimes unstable at -31 mV. The liquid junction potential of the extracellular solution against the internal solution was measured to be +7 mV, and all clamp potentials were corrected accordingly. Experiments were performed at room temperature. [ca2+]i measurement and elevation After mounting the experimental chamber onto the stage of an upright IX 70 microscope (Olympus, Tokyo, Japan), the chamber was perfused with NASW. [Ca2+]i was measured by the dual-wavelength ratiometric method using BTC (Iatridou et al. 1994), which was excited with light alternated between 400 and 480 nm using a Polychrome II monochrometer (TILL Photonics, Martinsried, Germany), and the resulting fluorescence signal at wavelengths greater than 510 nm was measured using an R928 photo-multiplier tube (PMT; Hamamatsu, Shizuoka, Japan). We used a UApo/340 objective (× 40, n.a. = 1.35, oil immersion, Olympus), a 505 DC XR-UV dichroic mirror, and HQ 535/50 emission filter (Chroma Technology Co., Brattleboro, VT, USA). The excitation light exposed a 50 μm square region, which included the entire presynaptic neuron. The fluorescence intensity was measured from a 40 μm square region centred within the excitation region. [Ca2+]i was determined from the ratio (R) of the fluorescence signals at both wavelengths, according to (Grynkiewicz et al. 1985): (1) where KdApp is the apparent dissociation constant, Rmin is the minimum ratio in zero Ca2+, and Rmax is the maximum ratio at saturated Ca2+. These calibration constants were obtained from in-cuvette calibrations. To obtain instantaneous increases in [Ca2+]i, 75 % Ca2+-loaded NPE (Ellis-Davies & Kaplan, 1994) was perfused into the presynaptic neuron from the patch pipette, and 200 J discharge flashes of UV light were applied to the whole presynaptic area (100 μm square) by fibre-optic coupling through a dual port condenser that used a sapphire mirror to combine photolysis light at 92 % transmittance and fluorescence excitation light at 8 % reflectance. The flash-lamp was a modified Chadwick-Helmuth system (El Monte, CA, USA) described previously (Landò & Zucker, 1989; Zucker, 1994). A single 200 J discharge of UV flash with 0.3 neutral density filter photolysed 50 % of DM-nitrophen half-loaded with Ca2+, measured in micro-cuvettes on the microscope stage as described previously (Zucker, 1993). Flash intensity was sometimes reduced with neutral-density (ND) filters transmitting between 12 and 76 %. Calibrations of BTC with NPE In-cuvette calibrations under high ionic strength (500-600 mm) were performed in the presence of NPE, using the Ca2+ buffer 1,3-diaminopropan-2-ol-tetraacetic acid (DPTA) to control moderate [Ca2+] (Neher & Zucker, 1993). First, because aspartate binds Ca2+, the pK of aspartate was measured. [Ca2+] in a solution containing 350 mm aspartate and 175 mm CaCl2 (pH = 7.3) was measured at 59 mm with a Ca2+-sensing electrode (MI-600, Microelectrodes, Inc., Bedford, NH, USA). Second, 111 % concentrated solutions containing the top eight chemicals listed in Table 1 (excluding BTC and NPE) were made. Using a Ca2+-sensing electrode, [Ca2+] of those solutions was measured and from this and the known buffering effect of aspartate, the dissociation constant (pK) of DPTA was calculated (Table 1). Third, 10 times concentrated solutions containing BTC and NPE were made and mixed in a ratio of 1:9 with the 111 % concentrated solutions to make small volumes of final calibration solutions A-E (Table 1). In all but the zero-Ca2+ solution, NPE was complexed with Ca2+ to prevent distortion of the [Ca2+] level set by DPTA. [Ca2+] of those solutions were calculated using our measured pK values of aspartate and DPTA, and published values for NPE (Ellis-Davies & Kaplan, 1994). Finally, rectangular glass cuvettes (50 μm path length, VitroCom, Mountain Lakes, NJ, USA) filled with the calibration solutions A-E were put in glass-bottomed chambers and covered with light mineral oil. Fluorescence ratios were measured on the microscope and fitted to eqn (1) to calculate calibration constants. Table 1. Solutions used to calibrate BTC with NPE under high ionic strength Concentrations (mm) A B C D E K-Asparate* [1.23] 300 300 305 310 360 KCl 65 65 65 65 10 K-Hepes (pH 7.3) 150 150 150 150 150 Glucose 11 11 11 11 11 K2-Glutathione 10 10 10 10 10 CaCl2 0 7.5 25 35 45 K3-DPTA* [4.5] 0 50 50 50 0 K2-EGTA* [∼7] 50 0 0 0 0 BTC* 0.25 0.25 0.25 0.25 0.25 NPE* [∼7] (CaCl2) 5(0) 5(5) 5(5) 5(5) 5(5) Measured [Ca2+] (μm) A B C D E 111 % solution < 0.01 5.88 33.1 74.6 6440 Pre-flash < 0.01 5.95 33.1 74.5 6320 Post-flash < 0.01 7.52 37.9 87.4 6610 Top part of table shows composition of calibration solutions A–E. *Ca2+-binding compound with pK values shown in square brackets. Bottom part of table shows measured [Ca2+] in the 111 % solution containing the first eight constituents and used to make final solutions, or calculated using pK values in final solutions before (Pre-flash) and after (Post-flash) single full strength (no neutral density filter) photolysis of NPE. The results are shown in Fig. 1. Before UV photolysis, the dissociation constant (Kd) was 19.4 μm, KdApp was 158 μm, and Rmax/Rmin was 24.7. After a single 100 % flash photolysis (no ND filter), Kd was 20.1 μm, KdApp was 146 μm, and Rmax/Rmin was 22.6. Our measurement of Kd is within the range of published values (Iatridou et al. 1994; Zhao et al. 1996). Although there was a small caged chelator-indicator interaction (Zucker, 1992; Neher & Zucker, 1993; Kasai et al. 1996), the effect was negligible, especially under 40 μm. We used the post-flash value to calculate [Ca2+]i. Figure 1Open in figure viewerPowerPoint BTC calibration curves with NPE under high ionic strength (about 550 mm) before and after UV flash photolysis Fluorescence ratios were normalized by Rmin before photolysis and plotted on a semi-log scale. The filled circles represent in-cuvette pre-photolysis measurements and the filled triangles represent post-photolysis fluorescence ratios (no ND filter; see text). The letters (A-E) represent solution names (see Table 1); the leftmost point is in nominally zero-Ca2+ medium. Lines are best-fit curves to in-cuvette values based on eqn (1) (dashed line: pre-flash, continuous line: post-flash). Inset shows a high magnification view at low concentration with linear scales. Most of the experiments were done in this range. Because the point around 10 μm (solution B) fell above the curve, we did not include it in calculating calibration values. Although these two curves are slightly different because of caged chelator-indicator interactions (Zucker, 1992), the difference is negligible at low Ca2+ concentrations. We used the post-flash value (Kd= 20 μm) to convert fluorescence ratios to [Ca2+]i resulting from NPE photolysis. In situ ratios are represented as the open circles (pre-flash) and as the open triangles (post-flash). We also performed in situ calibrations. Calibration solutions were like the solutions listed in Table 1 but with Cs+ substituted for K+. Calibration solutions were loaded into the SN using a patch pipette. In situ calibration results are close to in-cuvette results at some points (Fig. 1), but these were not used to calculate calibration constants. There were several problems. Because the neuron has high volume (about 50 μm in diameter and spherical), gradual filling from the patch pipette causes the fluorescence ratio to reach a plateau very slowly (more than 30 min). The ratio was often lower than expected, because [Ca2+]i never reached the level of the pipette solution due to cellular Ca2+ uptake and extrusion processes. We tried using lithium-substituted ASW to suppress the Na+-Ca2+ exchanger, but it had little effect. Furthermore, especially when high [Ca2+]i solutions were introduced, most neurons lost membrane integrity as indicated by a sudden increase in fluorescence ratio before that ratio and [Ca2+]i approached steady levels during perfusion from the patch pipette. In such cases, the sudden ratio increase probably reflected membrane breakdown and Ca2+ inflow. Due to this problem it was impossible to obtain any data at all at the highest [Ca2+]i levels, and the in situ data we did obtain did not show a clear sign of saturation as [Ca2+]i increases, preventing estimation of KdApp or Rmax from this data. The fact that fluorescence ratios were still rising when cells were lost indicates that fluorescence ratios for in situ calibrations underestimate the true values, especially at higher [Ca2+]i levels. This is exactly what a comparison of the in situ and cuvette data shows, and so we think the in situ results are consistent with (and essentially validate) the more accurate cuvette measurements. When in-cuvette calibration Rmin was used, the resting (pre-flash) [Ca2+]i of presynaptic neurons was occasionally estimated at a negative value, and at other times as high as 2 μm. These cells were perfused with a Ca2+-loaded Ca2+ buffer (NPE) which should have buffered [Ca2+]i to 0.4 μm, which is the level of [Ca2+]i that we both calculated and measured in the presynaptic filling solution (using Ca2+-sensitive electrodes). At low values of R, negative estimates of [Ca2+]i can only arise from errors in Rmin, and the largest deviations we observe from 0.4 μm would arise from errors of about 20-25 % in Rmin. Variation in Rmin probably arises from variability in the colour of excitation light, in which case Rmax should covary with Rmin. Therefore, Rmin was adjusted to set the resting [Ca2+]i equal to 0.4 μm, and Rmax was adjusted to keep the ratio Rmax/Rmin constant. This procedure is similar to that used in earlier studies employing low-affinity dyes (e.g. Heinemann et al. 1994). We also tried using dimethoxynitrophenyl-EGTA-4 (DMNPE-4) to try to produce higher [Ca2+] rises than are possible with NPE. The efficiency of uncaging DMNPE-4 should be five times higher than for NPE (Ellis-Davies & Kaplan, 1994; DelPrincipe et al. 1999). In addition, because it has higher affinity for Ca2+ than NPE before photolysis, this chelator can be loaded with more Ca2+ than NPE at resting free [Ca2+]i levels. We performed in-cuvette calibrations of BTC with DMNPE-4. The pre-flash calibration constants were almost the same as those with BTC alone. However, the post-flash values changed dramatically: Kd decreased to 65 %, KdApp decreased to 60 %, Rmin decreased to 87 %, and Rmax decreased to 59 %. Even with corrected values for calibration constants, the [Ca2+]i did not rise as much as expected on photolysis of DMNPE-4. Thus, the light sensitivity of DMNPE-4 was less than expected, and there was a large DMNPE-4-indicator interaction, which made it very difficult to estimate [Ca2+] changes accurately after photolysis (Zucker, 1992; Neher & Zucker, 1993). It is possible that these problems are particular to the samples of DMNPE-4 and BTC available to us, but because of them we decided not to pursue the use of DMNPE-4 in our preparation. puffing scpb onto SNs Using a U-shaped tube with gravity feed and electrical valve control (Krishtal & Pidoplichko, 1980; Fenwick et al. 1982), SCPB was puffed onto a SN. The speed of concentration change was measured by puffing diluted (85 %) NASW onto the SN and measuring changes in liquid junction potential in an adjacent electrode. The 10-90 % rise time of solution change was 0.18 ± 0.03 s (mean ±s.e.m.), which is faster than the initiation of a peptidergic response in SN (Fig. 5). Figure 5Open in figure viewerPowerPoint The SN is a linear, but delayed, detector of peptide Aa, 60 s application of the peptide SCPB at 3, 30 and 300 nm, and 3 and 30 μm (indicated by black bar) induced prolonged inward current in the SN. Desensitization was not observed. Ba, 1 s application of SCPB (3 nm to 30 μm) mimics flash-evoked inward current in the same SN. Ab and Bb show the same responses (middle traces) and their derivatives (lower traces) on a faster time scale; upper traces track the time course of local concentration change (see Methods). C, dose-response curve for SCPB concentration and the maximum rate of inward current change. The open circles represent 60 s application and filled squares represent 1 s applications of SCPB. Lines are best-fit curves based on the Hill equation. Hill coefficients were 0.85 ± 0.06 (60 s application, n= 4) and 1.09 ± 0.09 (1 s application, n= 6). Electron microscopy B2-SN cell pairs were cultured on Aclar film (Ted Pella Inc., Redding, CA, USA), a tissue culture substrate that can be sectioned for electron microscopy. The cell pairs remained tightly adhered to the Aclar film during fixation, processing and embedding, thus maintaining the orientation of the cells and allowing their identification in tissue blocks and thin sections. Cell pairs were fixed with 2 % glutaraldehyde or 2 % glutaraldehyde-1 % formaldehyde in 0.1 m sodium cacodylate buffer, pH 7.4 for 2 h at room temperature. Following this primary fixation, the specimens were post-fixed with 1 % OsO4 in 0.1 m sodium cacodylate buffer, stained en bloc with 1 % uranyl acetate in distilled water, dehydrated in a graded ethanol series and embedded in Eponate 12 (Ted Pella, Inc.). Serial 80-85 nm sections were cut with a Reichert-Jung Ultracut E ultramicrotome (Leica, Deerfield, IL, USA) and mounted on formvar-coated slot grids. The plane of sectioning was oriented perpendicular to the plane of contact between the cell pairs, and block faces were trimmed asymmetrically so the SN and B2 cells could be easily and uniquely identified in the electron microscope. Following sequential staining with uranyl acetate and Sato's lead (Sato, 1968), the sections were examined and photographed on a 1200EX/II electron microscope (JEOL, Tokyo, Japan) operating at 80 kV. The distribution of vesicles as a function of distance from active zone membranes in the presynaptic bouton-like contacts was examined in 34 serial section micrographs of four B2-SN cell pairs. We measured the minimal distance between all synaptic vesicles and the electron dense membrane specialization of the active zone out to a distance of 1000 nm from the active zone membrane (Hess et al. 1993; Reist et al. 1998). The presynaptic bouton-like contacts used in these vesicle counts were slightly oblong in shape, measuring 2.3-4.4 μm in length and 2.0-3.3 μm in width. Drugs The caged Ca2+ chelators NPE and DMNPE-4 were gifts from Dr G. C. R. Ellis-Davies (Medical College of Pennsylvania Hahnemann University). BTC was obtained from Molecular Probes, Inc. (Eugene, OR, USA). α-Conotoxin IMI was obtained from BACHEM (Torrance, CA). SCPB was obtained from American Peptide Company (Sunnyvale, CA, USA). DPTA, protease type IX (bacterial), tetrodotoxin, and penicillin-streptomycin (lyophilized) were obtained from Sigma (St. Louis, MO, USA). MEM amino acids solution without l-glutamine solution (50 ×), MEM non-essential amino acid solution (100 ×), MEM vitamin mixture (lyophilized), and fetal bovine serum were obtained from GIBCO BRL (Grand Island, NY, USA). Presentation of data Results are expressed as means ±s.e.m. RESULTS Cholinergic and peptidergic release evoked by a depolarising pulse Figure 2 shows examples of cholinergic (B2-B3) and peptidergic (B2-SN) postsynaptic currents (PSCs) evoked by a single 5 ms depolarisation to 19 mV. The postsynaptic responses show a sharp threshold to gradual increases in pulse potential, are unaffected by prolongation of the pulse, and are blocked by 100 μm tetrodotoxin, and are therefore mediated by APs in unclamped presynaptic processes. Depolarizations of 5 ms evoke single APs in Aplysia neurons, where typical AP duration is ≈5-10 ms and refractory period is ≈20 ms (unpublished observations). Figure 2Open in figure viewerPowerPoint Cholinergic and peptidergic PSCs evoked by an AP Upper trace, presynaptic [Ca2+]i; middle trace, postsynaptic current (Ipost); lower trace, rate of Ipost change evoked by a single presynaptic 5 ms depolarisation (arrow) from -46 mV to +19 mV in cholinergic (A, B2-B3) and peptidergic (B, B2-SN) cell pair. The volume averaged presynaptic (soma) [Ca2+]i change was undetectable during a 5 ms depolarising pulse evoking a single AP. This suggests that both cholinergic and peptidergic release are evoked by local regions of submembrane [Ca2+]i. Transmission at these synapses has been studied by Whim et al. (1997). Cholinergic transmission is inhibitory and involves opening of Cl− channels (Gardner & Kandel, 1977), while peptidergic transmission is excitatory and involves second messenger-mediated closing of K+ channels, which are normally open (Ocorr & Byrne, 1985; Baxter & Byrne, 1989). Using postsynaptic perfusion of a solution with 72 mm Cl− and 550 mm K+, and NASW as a bathing solution with 600 mm Cl− and 10 mm K+, the cholinergic and peptidergic transmission have reversal potentials of -53 and -100 mV, respectively. Recording postsynaptic responses under voltage clamp around -35 mV, cholinergic opening of Cl− channels produces a Cl− influx or outward current while peptidergic closing of K+ channels reduces K+ efflux and produces an inward current (Ocorr & Byrne, 1985). The responses are also distinguishable kinetically. Cholinergic inhibitory postsynaptic currents (IPSCs) start after a synaptic delay of a few milliseconds following presynaptic depolarisation and last only about 0.1 s. This is at least five times faster than the desensitisation rate of the desensitising component of cholinergic responses (Kehoe & McIntosh, 1998), and so apparently reflects the duration of the release of transmitter. Excitatory synaptic currents (EPSCs), which are at least partially, and perhaps wholly, due to release of SCPs (Whim et al. 1997), start about 1 s after presynaptic depolarisation and last for more than 10 s. The slow decay of peptidergic resp