INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Synaptic plasticity has been hypothesized to underlie behavioral plasticity and memory in both vertebrates and invertebrates (Abel and Kandel 1998; Bailey et al. 2000a; Byrne and Kandel 1996; Huang et al. 1996; Lechner and Byrne 1998; Milner et al. 1998). Two well-studied mechanisms that are capable of modulating synaptic strength are post-tetanic potentiation (PTP) and long-term potentiation (LTP). Nearly all neuromuscular and many central synapses show PTP, the phenomenon in which synaptic potentials are potentiated for several minutes following a high-frequency (tetanic) burst of presynaptic action potentials. At most synapses studied, PTP is thought to be due to an increased level of residual Ca²⁺ in the presynaptic terminal (Thomson 2000; Zucker 1989, 1999). This increase in residual Ca²⁺ is believed to activate as yet unidentified processes that increase transmitter release. We will refer to this phenomenon as classic PTP. Another widespread form of frequency-dependent potentiation at central synapses is LTP. However, unlike PTP, induction of LTP is thought to require an increase in the Ca²⁺ concentration in the postsynaptic neuron that is, in many cases, mediated primarily by N-methyl-D-aspartate (NMDA) receptors (Bekkers and Stevens 1990; Kullmann and Siegelbaum 1995; Malenka and Nicoll 1999; Malinow et al. 2000).

Modulation of synaptic transmission by neuropeptide cotransmitters, at least superficially, shares several of the characteristics of classic PTP and LTP. The release of neuropeptide cotransmitters and the induction of both PTP and LTP often require high firing frequencies (Huang et al. 1996; Kupfermann 1991; Zucker 1999). Many neuropeptide cotransmitters increase synaptic transmission with a time course similar to PTP and LTP. Neuropeptides also can increase presynaptic or postsynaptic Ca²⁺ levels by modulating the conductance of voltage-gated Ca²⁺ channels or by releasing Ca²⁺ from internal stores (Kaczmarek 2000; Miller 1998; Tse and Tse 1999; Wickman and Clapham 1995; Wu and Saggau 1997). The similarities between physiological stimuli that induce neuropeptide release and PTP combined with the prevalent expression of neuropeptide cotransmitters suggests that a component of PTP in some neurons could be mediated by modulatory neuropeptides.

To investigate if neuropeptides were involved in PTP, we studied the neuromuscular synapses of anterior buccal muscle 3 (I3a) in Aplysia. This muscle and the motor neurons that innervate it generate rhythmic biting and swallowing movements during feeding (Church and Lloyd 1991, 1994; Cohen et al. 1978; Gardner 1971). There are several reasons to use this preparation to study the role of modulatory neuropeptides in PTP. First, the excitatory motor neurons that innervate I3a have been identified (B3 and B38) (Church and Lloyd 1991; Church et al. 1993). Because these neurons functionally innervate the same muscle fibers, PTP produced by both neurons can be assayed on the same postsynaptic target (Fox and Lloyd 1997). Second, the transmitters used by B3 and B38 are known. Both neurons use the same conventional fast acting transmitter, L-glutamate (Fox and Lloyd 1999), so differences in PTP between B3 and B38 are unlikely to be due to release of different conventional transmitters. However, B3 and B38 use different modulatory neuropeptide cotransmitters. B3 expresses and releases FMRFamide, whereas B38 expresses and releases the small cardioactive peptide (SCP). SCP dramatically increases EJPs evoked by B38, but not B3-evoked EJPs (Church et al. 1993; Fox and Lloyd 1997; Lotshaw and Lloyd 1990).

The mechanisms that underlie the effects of SCP have been investigated in I3a. Both the neuronal release and application of exogenous SCP dramatically increased cAMP levels in the muscle (Church et al. 1993; Fox and Lloyd 2000). Many of the short-term modulatory effects of SCP in this preparation are mediated by the cAMP pathway (Fox and Lloyd 2000; Lotshaw and Lloyd 1990). By contrast, FMRFamide only slightly increases EJPs evoked by either neuron. Thus the major difference between the effects of the neuropeptide cotransmitters is quantitative. At 1 µM, FMRFamide increases EJPs <50% while SCP increases B38-evoked EJPs >1,000% (Church et al. 1993; Fox and Lloyd 1997, 1998; Keating and Lloyd 1999; Lotshaw and Lloyd 1990). So, if neuropeptide cotransmitters participate in PTP, increases in B38-evoked EJPs could act as a sensitive assay for the effects of SCP released during a tetanus.

It was our hypothesis that SCP released from B38's terminals during tetanic stimulation would contribute to PTP for B38. We found that PTP for B38 was much larger than PTP for B3. Because no antagonist to SCP currently exists, we used several indirect approaches to implicate the release of SCP as the mediator of PTP for B38. First, we studied the effects of stimulation frequency of the tetanus and the effects of temperature on PTP. SCP release is known to be critically dependent on these parameters. Next, we selectively superfused the neuromuscular synapses with exogenous SCP to determine if this would occlude the effects of SCP released from B38 during a tetanus. We provide evidence that a large component of PTP for B38 is mediated by SCP release.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Aplysia californica (60-300 g) were obtained from Marinus (Long Beach, CA), maintained in circulating artificial sea water (ASW) at 16°C, and fed dried seaweed every 3 days.

Neuron stimulation

Detailed experimental methods have been described previously (Fox and Lloyd 1997). Briefly, animals were immobilized with an injection of isotonic MgCl₂ and the dissection carried out in high-Ca²⁺ (33 mM; 3 times normal), high-Mg²⁺ (165 mM; 3 times normal) ASW (termed high Ca, Mg ASW). The buccal mass and buccal ganglia were removed and the mass bisected along the midline. Buccal nerve 2, which contains the peripheral axons of B3 and B38, was left intact (nerve designations, Gardner 1971; muscle nomenclature, Howells 1942; also see Lloyd 1988). The ganglia were desheathed and superfused with low-Ca²⁺ (0.5 mM; 0.05 times normal), high-Mg²⁺ (110 mM; 2 times normal) ASW (termed low Ca ASW). Neurons were normally impaled with two microelectrodes (2-4 M; filled with 3 M K acetate and 30 mM KCl), one to inject current and one to monitor membrane potential. B3 and B38 were identified by their position, size, and muscle innervation patterns (Church et al. 1993). Experiments were performed at room temperature (~24°C) unless otherwise stated. Individual spikes in motor neurons were driven by brief (10-20 ms) depolarizing current pulses. The frequency of action potentials within a burst was usually 16 Hz. Dose response curves were carried out by alternatively stimulating bursts in B3 and B38 at 50 s intervals (100 s intervals for each neuron) while increasing the concentration of SCP from 0.1 to 1 µM at 20 min intervals with no intervening washes. These long interburst intervals were used to minimize release of endogenous neuropeptide cotransmitters (Church et al. 1993; Lotshaw and Lloyd 1990; Whim and Lloyd 1990). Burst durations were adjusted (from 0.15 to 2 s) so that the compound EJPs evoked by B3 or B38 were similar in amplitude. PTP experiments were performed separately for each motor neuron. Bursts (usually 16 Hz) were stimulated in either B3 or B38 at 100 s intervals. After recording several "control" EJPs, PTP was induced by stimulating the neuron to fire a tetanus of 12 bursts (5-15 Hz) of 4 s duration with 7 s interburst intervals.

Measurement of I3a EJPs

EJPs were recorded with a perfusion electrode (Church et al. 1993; Fox and Lloyd 1997). The perfusion electrode consisted of a small chamber (100 µl) and aperture (~1.5 mm) that was positioned to press firmly down on a portion of the muscle (see Fig. 1 in Church et al. 1993). The inside of this electrode was perfused with ASW while the rest of the preparation was superfused with low-Ca ASW to suppress synaptic transmission and muscle contractions. This procedure permits the simultaneous recording from a population of muscle fibers thereby reducing sampling bias. It also confined the contractions to the small area of the muscle covered by the recording chamber and thus markedly reduced movement artifacts in the recordings. The earliest evoked muscle contractions occur after the sixth EJP so the early EJPs in a burst are recorded in the absence of any movement. EJPs were recorded by extracellular electrodes placed inside and just outside the wall of the perfusion apparatus. Signals were amplified using a Grass P15D AC amplifier. Neuropeptides were applied in ASW to the inner chamber of the perfusion electrode so the ganglia and the remainder of the muscle were not exposed to them. Typical application periods were 20 min. Burst durations were adjusted (from 0.15 to 2 s) so that the compound EJPs evoked by B3 or B38 were similar in amplitude. Temperature changes were also restricted to the inner chamber of the perfusion electrode. ASW temperature was controlled with a heat exchanger submerged in a water bath and was monitored with a miniature temperature probe (Yellow Springs Instruments No. 402) mounted in the perfusion electrode. Experiments at 16 and 24°C were performed on the same preparations, and the direction of the temperature change was alternated.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of tetanic stimulation of B38 resemble the effects of the bath application of SCP

Initially we compared the effects of PTP to those of the bath application of SCP on EJPs evoked by B3 and B38. If the release of SCP contributes significantly to PTP for B38, then the effects of the neuropeptide should resemble, at least in part, the effects of tetanic stimulation. First we determined the effects of exogenous SCP and FMRFamide on EJPs evoked by B3 and B38 in the same preparation. A perfusion electrode was used to record extracellular EJPs in a small portion of the I3a muscle that was perfused with ASW while the remainder of the muscle was superfused with low-Ca ASW to suppress synaptic transmission and muscle contractions (Church et al. 1993). This procedure permitted stable long-term recordings, rapid solution turnover, and simultaneous recordings from a population of fibers, thereby reducing sampling bias. Results from this technique are qualitatively similar to those obtained with intracellular electrodes (Fox and Lloyd 1997). Application of 1 µM SCP dramatically increased B38-evoked EJPs while having a much smaller effect on those evoked by B3 (Fig. 1). By contrast, 1 µM FMRFamide only slightly increased EJPs evoked by both neurons. The increases produced by both SCP and FMRFamide reversed within 1 h of washout. So if a significant component of PTP was due to SCP release, then we would expect tetanic stimulation of B38 to selectively potentiate its own EJPs and we would expect PTP to readily reverse after the end of the tetanic stimulation.

View larger version (19K): [in this window] [in a new window]   Fig. 1. Effects of small cardioactive peptide (SCP) or FMRFamide on B3- and B38-evoked excitatory junction potentials (EJPs). Compound EJPs were evoked by alternately stimulating bursts of action potentials (16 Hz) in B3 and B38 at 50 s intervals (100 s intervals for each neuron). In this and the following figures, burst durations were initially adjusted so that the amplitude of the final EJPs in a burst were similar for both neurons. A: application of 1 µM SCP dramatically increased B38-evoked EJPs and had a much smaller effect on those evoked by B3. Note that the increase in EJPs produced by SCP reversed fully within 1 h of washout. B: application of 1 µM FMRFamide only slightly increased EJPs evoked by B3 or B38. Recordings are from the same preparation.

Next we examined the heterosynaptic effects of tetanic stimulation of B3 or B38. It was important to determine that stimulation of the motor neurons did not release other substances that modulated the efficacy of synaptic transmission. For example, it is possible that glutamate released from the motor neurons could induce a lasting plasticity similar to LTP. We found that tetanic stimulation of one of the motor neurons had very little effect on subsequent EJPs evoked by the other neuron (Figs. 2 and 3). Tetanic stimulation of B38 only slightly potentiated B3-evoked EJPs [by 22 ± 8% (mean ± SE) at 8 s after the tetanus; n = 5], which is similar to the effects observed for the application of exogenous SCP (Fig. 6). Tetanic stimulation of B3 had an even smaller effect on B38-evoked EJPs (potentiated by 13 ± 11%, n = 5; Figs. 2 and 3). The effects of released neuropeptides were expected to be quite small as exogenous SCP only moderately increased B3-evoked EJPs (Church et al. 1993; Fox and Lloyd 1997) and exogenous FMRFamide has only a small effect on B38-evoked EJPs. FMRFamide, at 1 µM, increased B38-evoked EJPs by 40 ± 16% (n = 8). Therefore the effects of heterosynaptic stimulation could be explained by the release of neuropeptides expressed in B3 and B38.

View larger version (29K): [in this window] [in a new window]   Fig. 2. Post-tetanic potentiation (PTP) of B3- and B38-evoked EJPs by homosynaptic and heterosynaptic tetanic stimulation. Compound EJPs were evoked by stimulating bursts of action potentials (16 Hz) in either B3 or B38 at 100 s intervals. Several control EJPs (C) were recorded before PTP was induced by stimulating one of the motor neurons to fire a tetanus (12 bursts of 4 s duration at 15 Hz with 7 s interburst intervals; - - -) in the interval between the motor neuron bursts. Homosynaptic stimulation of both neurons was more effective at producing PTP than heterosynaptic stimulation. Note that the PTP was much larger for B38 than for B3.

View larger version (17K): [in this window] [in a new window]   Fig. 3. Time course of PTP caused by tetanic stimulation of B3 and B38. Homosynaptic stimulation of B38 caused PTP that was much larger than PTP for B3 (n = 6). Heterosynaptic stimulation only slightly potentiated EJPs (n = 5). Typically, PTP fully reversed within 20 min after the tetanus. Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts.

Homosynaptic stimulation of B38 produced PTP that resembled the effects of exogenous SCP application. Both neurons exhibited PTP, but the B38-evoked EJPs were potentiated much more than those evoked by B3 (Figs. 2 and 3). When PTP was measured at 108 s after the end of the tetanic stimulation, a period that allows more rapid processes such as facilitation and augmentation to decay (Zucker 1989), tetanic stimulation of B38 potentiated B38-evoked EJPs by 342 ± 88% (n = 6) and tetanic stimulation of B3 potentiated B3-evoked EJPs by 52 ± 18% (n = 6). PTP for both neurons typically reversed within 15 min. These results are consistent with the hypothesis that a major component of PTP for B38 is caused by SCP release. Unfortunately, no antagonist for SCP exists, so we tested this hypothesis indirectly. Initially we varied two parameters known to strongly influence neuropeptide release, the frequency of stimulation during the tetanus and temperature.

PTP is highly dependent on the frequency of tetanic stimulation

Neuropeptide release from Aplysia motor neurons is highly dependent on firing frequency (e.g., Vilim et al. 1996, 2000; Whim and Lloyd 1989, 1990, 1994). We already estimated the frequency dependence of SCP release from B38 in I3a using several indirect approaches (Church et al. 1993). SCP release occurred at frequencies <10 Hz and increased with higher frequencies. We examined whether the amplitude of PTP was dependent on frequency for B3 and B38 (Fig. 4). We found that for all frequencies tested, PTP was larger for B38 than for B3. PTP for B3 was dependent on tetanic firing frequency; however, PTP for B38 showed a stronger dependence on frequency (Fig. 4). The fact that PTP for B38 was larger at all stimulation frequencies supports our hypothesis that the release of SCP from B38 mediates a major component of PTP for B38. However, it is possible that mechanisms that underlie classic PTP also contributed to PTP for B38. It is interesting to note that the frequencies that elicited PTP for B38 were also shown by previous indirect measurements to release SCP (Church et al. 1993). For example, B38 normally fires at 10 Hz during feeding-like motor programs and this frequency both releases SCP and produces substantial PTP.

View larger version (18K): [in this window] [in a new window]   Fig. 4. PTP was dependent on the frequency of stimulation in the tetanus. PTP was measured at 108 s after the end of the tetanic stimulation, a period that allows more rapid processes such as facilitation and augmentation to decay (Zucker 1989). PTP increased with tetanus frequency for both neurons; however, the increase was larger for B38. Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts (n = 6).

PTP is highly dependent on temperature

Neuropeptide release at Aplysia neuromuscular synapses is extremely temperature dependent, increasing dramatically at lower temperatures (Vilim et al. 1996; Whim and Lloyd 1990). For example, in another buccal neuromuscular preparation from Aplysia (termed the ARC), release of SCP was directly measured using a radioimmunoassay (RIA) during the stimulation of a SCP-expressing motor neuron B15. Decreasing the temperature from 19 to 16°C caused a approximately fourfold increase in SCP release from B15's terminals (Vilim et al. 1996). Neuronal release of SCP also potently increases cAMP levels in the ARC. Thus changes in SCP release due to temperature can be monitored indirectly using cAMP levels in the muscle (Whim and Lloyd 1990). Stimulation of B15 with a paradigm that had no effect on cAMP levels at 22°C caused a 14 ± 5-fold increase at 16°C (Whim and Lloyd 1990). This increase in cAMP levels is most likely due to increased SCP release and not a change in cAMP metabolism because the same decrease in temperature had no effect on the amplitude of the increase in cAMP caused by exogenous SCP. Pilot studies indicated that there is similar relationship between temperature and SCP release from B38's terminals in the I3a muscle (Church et al. 1993). Therefore if SCP mediated a major component of PTP and if more SCP is released at lower temperatures, we would also expect PTP for B38 to increase at lower temperatures. However, before we tested the effects of tetanic stimulation on PTP, we first determined the effects of temperature on basal synaptic transmission.

Changing temperature can have dramatic and unpredictable effects on synaptic transmission. Depending on the temperature range and the preparation tested, reducing the temperature can either increase or decrease the rest potential, input resistance, amplitude of synaptic potentials, or excitability of the neuron (Hardingham and Larkman 1998; Stephens 1985; Stephens and Atwood 1982; Volgushev et al. 2000a,b; Weight and Erulkar 1976). We tested the effects of reducing the temperature 8°C, from 24 to 16°C, on EJPs evoked by both B3 and B38. Both of these temperatures are within the normal physiological range for Aplysia because they normally live and feed at these temperatures (Kupfermann and Carew 1974). We found that reducing the temperature increased the delay between the action potentials evoked in the motor neurons cell bodies and the onset of the EJPs for both neurons. The onset of the EJPs was delayed by 26 ± 4% for B3 and 28 ± 2% for B38 (n = 6). This increase could be due to a reduction in the action potential conduction velocity or an increase in synaptic delay or both. These mechanisms have been shown to function in other systems (Baldo et al. 1983; Charlton and Atwood 1979; Katz and Miledi 1965; Lester 1970). We also found that at lower temperatures the EJP amplitude was reduced to a similar degree for both neurons (Fig. 5). In other systems, reductions in the amplitude of evoked synaptic potentials have been attributed to a decrease in transmitter release that is due to conduction block of action potentials, a reduction in the probability of evoked transmitter release, or changes in the electrical properties of the postsynaptic membrane (Barrett et al. 1978; Barton and Cohen 1982; Volgushev et al. 2000a,b; Weight and Erulkar 1976). Even though the EJP amplitude was reduced, there was little effect on facilitation within a burst or the effectiveness of exogenous SCP to enhance EJPs. Facilitation within a burst was measured as the ratio of the amplitude of the third EJP to that of the first EJP. For B3, the third EJP was increased by 163 ± 8% over the first EJP at 24°C and by 181 ± 24% at 16°C (means ± SE; n = 6). For B38, the third EJP was increased by 174 ± 29% at 24°C and 168 ± 21% at 16°C (n = 6). So, for both neurons, lowering temperature did not effect facilitation within the burst. Exogenous SCP also increased EJPs to a similar degree at both temperatures. The effects of bath application of two concentrations of SCP (0.1 and 1 µM) were compared at 24 and 16°C. Experiments at both temperatures were performed on the same preparations and the direction of the temperature shift was varied systematically. EJPs evoked by both neurons at 16°C were smaller than those evoked at 24°C, however, the increase produced by SCP was essentially identical at both temperatures (Fig. 6). So, reducing the temperature had a large effect on basal synaptic transmission, but no effect on the facilitation within a burst or the increase in EJP amplitude produced by exogenous SCP.

View larger version (15K): [in this window] [in a new window]   Fig. 5. Amplitude of B3- and B38-evoked EJPs were decreased at lower temperatures. A: compound EJPs were evoked by alternately stimulating bursts of action potentials (16 Hz) in B3 and B38 at 50-s intervals (100-s intervals for each neuron). Reducing the temperature by 8°C (from 24 to 16°C) reduced the amplitude of EJPs evoked by both B3 and B38 to a similar degree. B: summary of the effects of temperature on EJPs. Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts (n = 6).

View larger version (16K): [in this window] [in a new window]   Fig. 6. SCP increases EJPs at 16 and 24°C to a similar degree. Pooled data from experiments in which compound EJPs were evoked by alternately stimulating bursts of action potentials (16 Hz) in B3 and B38 at 50 s intervals (100 s intervals for each neuron). Application of exogenous SCP (0.1 and 1 µM) increased EJPs much more for B38 than B3. The effects of SCP were similar at both temperatures. Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts (n = 6).

Next we tested the effects of this temperature shift on PTP. Lowering the temperature has little effect on the amplitude of classic PTP for several preparations from vertebrates and invertebrates (Fisher et al. 1997; Schlapfer et al. 1975; Zengel et al. 1980). As discussed in the preceding text, lowering temperature dramatically increases the release of SCP. Thus reducing the temperature could potentially distinguish between classic PTP and PTP mediated by SCP release. Experiments at both temperatures were performed on the same preparations and the direction of the temperature shift was varied systematically. Because PTP appeared to plateau at higher stimulation frequencies (at 24°C; Fig. 4), we concentrated our experiments on lower frequencies (5-10 Hz; Fig. 7). For B3, no significant PTP was induced with 5 Hz tetanus. When a higher frequency tetanus was used, PTP for B3 tended to be reduced at 16°C. By contrast, PTP produced by B38 was larger at 16°C for all of the frequencies tested. At tetanus frequencies of 5 and 7.5 Hz, PTP at 16°C was twice the amplitude of PTP at 24°C. Also at 16°C, PTP appeared to reach a plateau in the range of 7.5 to 10 Hz tetanic stimuli. Thus lowering the temperature selectively increased PTP for B38 and appeared to shift the plateau to a lower frequency. These results suggest that different mechanisms underlie PTP for B3 and B38 and also supply indirect support for our hypothesis that the release of SCP contributes to PTP for B38.

View larger version (16K): [in this window] [in a new window]   Fig. 7. Lower temperatures selectively increased PTP for B38. Pooled data from the 108 s time points were used. Lowering the temperature had little effect on PTP for B3 at 5 Hz but tended to reduce PTP for B3 at higher frequencies. The reduction in PTP for B3 was not significant (P > 0.05 at all frequencies; t-test). PTP for B38 was larger at 16°C for all 3 frequencies of tetanic stimulation. The increase in PTP for B38 was significant at all frequencies tested (P < 0.01 at 5 and 7.5 Hz; P < 0.05 at 10 Hz). Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts (n = 6 for B3 and n = 9 for B38).

Interpreting the results of the experiments in which the temperature was changed was complicated because PTP was measured at 108 s after the end of the tetanic stimulation. This time point was chosen to allow for the decay of shorter term synaptic plasticities such as facilitation and augmentation (Zucker 1989). However, PTP also decayed during this period, so PTP measured at 108 s was a function of two processes; the initial amplitude of PTP and its decay in the interval after the end of the tetanus. Thus it was possible that the observed increase in PTP at lower temperatures was not due to a change in initial PTP amplitude but instead was caused by a reduction in the rate of decay of PTP. Indeed, the time constant for the decay of PTP has been shown to be highly sensitive to temperature in both vertebrate and invertebrate preparations (Fisher et al. 1997; Schlapfer et al. 1975; Zengel et al. 1980). For example, in central ganglia of Aplysia, the maximum amplitude of PTP was not affected by reducing the temperature 12°C (from 20 to 8°C), whereas the time constant for the decay of the PTP was dramatically increased (Schlapfer et al. 1975). PTP for B38 at 5 Hz and B3 at all frequencies was too small in most experiments to accurately calculate the time constant for decay or the initial PTP amplitude. However, we were able to plot the decay of PTP for B38 at frequencies >5 Hz and extrapolate the initial PTP amplitude at the end of the tetanus. The decay of PTP from 108 to 1,008 s was well fit by a single exponential at both temperatures. The 8 s time point was always higher than the extrapolated initial PTP (Fig. 8), presumably reflecting the contribution of other processes such as augmentation and consequently this time point was not used. The initial PTP amplitude (y intercept) and the time constants for decay were determined. We found that the initial PTP amplitude was increased at 16°C for B38, 2.0 ± 0.2-fold at 7.5 Hz and 1.8 ± 0.4-fold at 10 Hz (n = 8). Surprisingly, there was no effect of temperature on the decay of PTP for B38 at either frequency (Fig. 8). So, the increase in PTP measured at 108 s was due to an increase in the initial PTP amplitude and not a change in the time constant for its decay.

View larger version (18K): [in this window] [in a new window]   Fig. 8. Time course of PTP caused by tetanic stimulation of B38 at 16 and 24°C. Homosynaptic stimulation of B38 at 7.5 Hz caused PTP that was larger at 16 than at 24°C. Note that the time course of decay of PTP was essentially identical at both temperatures. Lines are exponential fits of pooled data from 108 to 1,008 s. The 8 s time points were not used for these fits. Increases in EJP amplitudes were plotted on a log scale. Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts (n = 9).

Exogenous SCP occludes a large component of PTP for B38

Currently no SCP antagonist exists, so we determined if exogenous SCP would occlude the effects of released SCP and thereby selectively reduce PTP observed after tetanic stimulation of B38. We used two concentrations of exogenous SCP, 0.1 and 1.0 µM, and these experiments were done at 24°C. One potential problem with this approach is that SCP itself potently increases EJPs. This raises the possibility that PTP for B38 might be reduced because the EJPs are already increased close to their maximal amplitude in the presence of SCP. To ensure that this was not the case, we initially used the lower SCP concentration. At 0.1 µM, SCP increases EJPs 263 ± 119% compared with 831 ± 210% for 1 µM SCP tested on the same preparations (n = 6; Fig. 6). We examined the effects of 0.1 µM SCP on both neurons and found that PTP was reduced dramatically for B38, whereas PTP for B3 was actually increased slightly (Figs. 9 and 10). Because exogenous SCP increased EJPs, we wanted to verify that B38-evoked EJPs were not maximally potentiated after the tetanus. We tested for this by measuring facilitation within bursts. The ratio of the amplitude of the third and first EJPs in a burst was quantified and we found that it was not changed by 0.1 µM SCP or by tetanic stimulation (Fig. 11). Indeed, it was only slightly reduced by the combination of a tetanus and 0.1 µM SCP. Thus facilitation within a burst, that presumably is too rapid to be mediated by SCP release, was not significantly changed by exogenous SCP and the effects of SCP were quite specific to PTP for B38.

View larger version (22K): [in this window] [in a new window]   Fig. 9. Exogenous SCP selectively reduced PTP for B38. Compound EJPs were evoked by stimulating bursts of action potentials (16 Hz) in either B3 or B38 at 100 s intervals. Several control EJPs (C) were recorded before PTP was induced by stimulating 1 of the motor neurons to fire a tetanus (12 bursts of 4 s duration at 15 Hz with 7 s interburst intervals; - - -) in the interval between the motor neuron bursts. Application of exogenous SCP (0.1 µM) increased EJPs and reduced PTP for B38 while only slightly changing PTP for B3.

View larger version (19K): [in this window] [in a new window]   Fig. 10. Time course of PTP in the absence or the presence of exogenous 0.1 µM SCP. PTP was induced by stimulating B3 or B38 to fire a tetanus (12 bursts of 4 s duration at 15 Hz with 7 s interburst intervals). Application of 0.1 µM SCP reduced PTP for B38 while having little effect on PTP for B3. Graphed values are the means ± SE of the 3rd EJP of the motor neuron bursts (n = 6).

View larger version (19K): [in this window] [in a new window]   Fig. 11. Facilitation within bursts evoked by B38 was not changed by 0.1 µM SCP or PTP. Pooled data from the 108 s time points were used. Facilitation within a burst was not significantly changed by application of SCP or the tetanic stimulation alone (P > 0.05; t-test). However, SCP and tetanic stimulation in combination did slightly reduce the facilitation within a burst (P < 0.05). Graphed values are the means ± SE of the ratio of the 3rd EJP to the 1st EJP in a burst (n = 6).

Next the effect of a higher concentration of SCP (1 µM) was examined on PTP evoked by both neurons. At this concentration, the effects of SCP are nearly maximal. For example, application of a 10-fold higher concentration of exogenous SCP (10 µM) increases EJPs only ~20% more than 1 µM. Thus one might expect that the effects of SCP released from B38 might be fully occluded by 1 µM exogenous SCP. EJPs in 1 µM SCP were very large, so we restricted our analyses to the amplitude of the first EJP. However, the EJPs measured at 108 s after the tetanus were not near their maximal amplitude as they were only 23 ± 7% (n = 6) of the largest EJPs observed within the 4 s tetanic bursts. We found that 1 µM SCP dramatically reduced PTP for B38 while it only slightly reduced PTP for B3 (Fig. 12). It is also quite striking that in 1 µM SCP, PTP for B38 appeared similar to that observed in B3 especially at later time points (Fig. 12). So most of the PTP for B38 was selectively occluded by this high concentration of exogenous SCP thus supporting our hypothesis that it is mediated by SCP released from B38's terminals during tetanic stimulation. However, a component of PTP for B38 was not occluded suggesting that another mechanism, possibly classic PTP, also contributes. If this is the case, the second mechanism would only contribute a small component of the total PTP for B38.

View larger version (19K): [in this window] [in a new window]   Fig. 12. Time course of PTP in the absence or presence of exogenous 1 µM SCP. PTP was induced by stimulating B3 or B38 to fire a tetanus (12 bursts of 4 s duration at 15 Hz with 7 s interburst intervals). A: application of 1 µM SCP markedly reduced PTP for B38 while causing a smaller reduction in PTP for B3. B: PTP for B3 and B38 were similar in the presence of 1 µM SCP. Graphed values are the means ± SE of either individual EJPs or the 1st EJP of the motor neuron bursts (n = 5).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It was our hypothesis that SCP released from B38's terminals during tetanic stimulation mediated a major component of PTP for B38. The results, albeit indirect, support this hypothesis. Increasing the frequency of stimulation in the tetanus increased PTP for both neurons, but the effect was more pronounced for B38. Thus changing the tetanus frequency did not expose differences in the mechanisms that underlie PTP for B3 or B38. This was probably due to the fact that the amplitude of classic PTP and neuropeptide release both increase with the frequency of stimulation in the tetanus.

Reducing the temperature, however, proved to be a useful procedure to distinguish between the mechanisms that underlie PTP for B3 and B38. Decreasing the temperature tended to reduce PTP for B3, while increasing PTP for B38. Thus different mechanisms appear to be responsible for PTP in the two motor neurons. We suspect that the differences in PTP are due to the release of SCP for the following reasons. In Aplysia, decreasing the temperature has been shown to increase SCP release in another buccal muscle (the ARC). SCP release was detected directly using a RIA or indirectly by measuring its effects on cAMP levels in the ARC (Vilim et al. 1996; Whim and Lloyd 1990). A similar relationship between temperature and the effects of SCP released from B38 on cAMP levels was also observed for the I3a muscle (Church et al. 1993). We also found that the effectiveness of SCP at increasing EJPs was identical for both temperatures tested. That is, application of exogenous SCP increased B38-evoked EJPs the same amount at both 16 and 24°C. It may seem surprising that the increase produced by SCP was essentially identical at both temperatures; however, it is mediated by cAMP and previous studies in the ARC have shown that exogenous SCP increases the cAMP levels in the muscle fibers to the same degree at both 16 and 22°C (Whim and Lloyd 1990). Taken together, these results suggest that the mechanism(s) that underlie PTP for B38 were different from those for B3 and that it is possible that PTP for B3 and a small component of PTP for B38 is a form of classic PTP, whereas most of the PTP for B38 is mediated by SCP released from its terminals. Surprisingly, we found that the decay rates of PTP for B38 were essentially identical at 16 and 24°C. This suggests that the rate-limiting mechanism underlying the decay of PTP is relatively unaffected by temperature over this range. One possible explanation for this observation is that the decay of PTP is limited by the diffusion of SCP away from the synapse, a process that would be essentially independent of temperature over this small range. Indeed, it is likely that diffusion is the primary mechanism used to reduce SCP concentrations since the SCP receptors do not desensitize in this system and buccal muscles show no uptake or proteolysis of SCP (Whim and Lloyd 1989).

Application of exogenous SCP also proved to be a useful procedure to distinguish between the mechanisms that underlie PTP and supported the hypothesis that most of the PTP for B38 was due to SCP release. Exogenous SCP occluded much of the PTP for B38 and presumably the effects of SCP released from B38, but there was a small component of PTP for B38 that was not occluded and was similar to PTP for B3. Indeed, in the presence of 1 µM SCP, PTP for B38 was similar to that for B3. EJPs evoked by both neurons were increased by exogenous SCP, so it was possible that the increase in PTP was smaller because the control EJPs (before the tetanus) in SCP were larger. This clearly was not the case for the lower SCP concentration as facilitation within bursts still occurred in 0.1 µM SCP. Indeed, even though EJP amplitude increased in SCP or after the tetanic stimulation, facilitation within bursts was essentially unchanged. EJP amplitude also was not maximal in the experiments using 1 µM SCP because much larger EJPs were observed during the long tetanic bursts. Although SCP appears to be acting by occluding the effects of endogenously released SCP, we cannot rule out the possibility that it is acting through other unrelated mechanisms. It is interesting to note that PTP was elicited using a stimulation paradigm that was designed to emulate active feeding (Church and Lloyd 1994) and that at least the lower concentration of SCP (0.1 µM) used to occlude PTP may occur in vivo. Thus physiological activity in B38 may elicit PTP that is predominantly dependent on the release of SCP from its own terminals.

Even though many other mechanisms can modulate the strength of central and peripheral synapses in Aplysia, we do not believe that they significantly contributed to PTP in this system. Both activity-dependent and -independent heterosynaptic enhancement of synaptic transmission have been described for central and peripheral synapses in Aplysia. However, heterosynaptic mechanisms did not contribute to the PTP produced by tetanic stimulation because neuronal activity in the I3a preparation was precisely controlled by removing all of the ganglia except for the buccal ganglia, suppressing ongoing activity in the buccal ganglion with low-Ca ASW, and controlling the activity of the motor neurons with intracellular electrodes. In the absence of motor neuron stimulation, little activity was detected with the perfusion electrode. Instead, it appears that PTP, at least for B38, was produced by two activity-dependent homosynaptic mechanisms; classic PTP and potentiation of EJPs by the release of SCP from B38's terminals. Other forms of activity-dependent homosynaptic plasticity like long-term potentiation (LTP) or activation of presynaptic autoreceptors for glutamate, the fast-acting conventional transmitter, do not appear to significantly contribute to the PTP at these synapses. A phenomenon very similar to LTP has been described for Aplysia central synapses, however, it is unlikely to contribute to PTP at I3a because the tetanic stimulation used to evoke PTP did not evoke significant LTP. In addition, PTP for B3 or B38 decayed with a time constant of ~8 min, whereas conventional LTP lasts for more than an hour for Aplysia sensory neurons (Lin and Glanzman 1994; Murphy and Glanzman 1997). The occlusion by SCP of most PTP for B38 also suggests that autoreceptors for glutamate do not contribute to this component of PTP. However, it is possible that activation of glutamate receptors underlies a small component of PTP. We could not determine whether the effects of PTP were pre- or postsynaptic. However, this system warrants further study because excitatory presynaptic autoreceptors appear to be quite rare, especially for neuropeptide cotransmitters (Juaneda et al. 2000; Malcangio and Bowery 1999; Miller 1998; Wu and Saggau 1997), and this system may allow the examination of an understudied mechanism of neuromodulation.

Many mechanisms converge to enhance the strength of the I3a neuromuscular synapses including facilitation, classic PTP, modulation by neuropeptide cotransmitters, and heterosynaptic facilitation by modulatory serotonergic neurons (Church et al. 1993; Fox and Lloyd 1997, 1998; Lotshaw and Lloyd 1990). Here we have described the convergence of two forms of synaptic enhancement, classic PTP and the homosynaptic modulation by a neuropeptide cotransmitter. However, the convergence of modulatory mechanisms is not unique to neuromuscular synapses in Aplysia. Several activity-dependent and -independent mechanisms interact to modulate the strength of sensory neuron synapses in Aplysia. For example, three different Ca²⁺-dependent mechanisms operate on similar time scales to enhance the strength of these synapses. One mechanism, classic PTP, acts presynaptically, whereas the other two mechanisms elevate the Ca²⁺ concentration in the postsynaptic neuron by pathways that are both dependent and independent of N-methyl-D-aspartate receptor activation (Bao et al. 1997, 1998; Lin and Glanzman 1994; Schaffhausen et al. 2001; Walters and Byrne 1984). Persistent forms of synaptic enhancement also converge to modulate sensory neuron synapses (Bailey et al. 2000a,b; Sutton and Carew 2000). It is interesting to note that we have previously identified two forms of persistent modulation that can prime the I3a neuromuscular system and increase the effectiveness of subsequent modulation. First 5HT causes a persistent increase in the amplitude of EJPs and contractions that lasts many hours (Fox and Lloyd 1997, 1998). Although these effects are most prominent at high-serotonin concentrations (~1 µM), they were also observed with lower concentrations (~0.1 µM) or after stimulation of the serotonergic neurons in patterns and frequencies observed during feeding-like motor programs (Fox and Lloyd 1998; Kupfermann and Weiss 1982). Second, SCP application persistently increases cAMP levels and also increases the responsiveness of the neuromuscular preparation to subsequent activation. This persistent increase in cAMP levels is prominent at higher concentrations of SCP (~1 µM) but was also observed at lower concentrations (~0.1 µM). We have shown that stimulation of B38 in patterns and frequencies observed during feeding-like motor programs, releases SCP at concentrations that can increase EJPs, mediates the predominant component of PTP, and may persistently increase cAMP levels. Although the pattern and frequencies of tetanic stimulation used in this study resembled those observed during feeding-like motor programs, the overall duration of the tetanus was much shorter than the duration of typical feeding episodes. PTP was induced with a tetanus that lasted for just over a minute, and a feeding episode can last for over an hour (Kupfermann and Carew 1974). It is likely that extending the duration of tetanic stimulation would also increase neuropeptide release. In fact, this has been investigated in the ARC neuromuscular system. It was found that neuropeptide release does not peak until the motor neuron is stimulated for 40 min. Remarkably, the peak amount of SCP released was ~100-fold higher than the amount released during the first few minutes of stimulation (Karhunen et al. 2001). So, if SCP release from B38 follows a similar time course, longer-duration tetanic stimulation should release more SCP, cause more PTP, and persistently elevate cAMP levels thereby increasing the responsiveness of the system for many hours after a feeding episode.

In conclusion, we wanted to determine whether SCP released from B38's terminals during tetanic stimulation contributed to PTP for B38. Our results support this hypothesis and are consistent with known properties of SCP release in this system. Increasing the tetanus frequency or decreasing the temperature have been shown to increase SCP release and enhance PTP for B38. Finally, exogenous SCP dramatically reduced PTP for B38, but had little effect on PTP for B3. Thus physiological stimulation of B38 may elicit PTP that is predominantly dependent on the release of SCP from its own terminals. Finally, both the amplitude of PTP and the mechanisms underlying it can be very different for two motor neurons innervating the same target muscle. It is possible that the release of neuropeptides during tetanic stimulation could contribute to PTP in other systems as well. For example, both vertebrate and insect motor neurons express neuropeptides that have been shown to potentiate EJPs at their respective neuromuscular synapses (Laufer and Changeux 1989; Ohhashi and Jacobowitz 1988; Uchida et al. 1990; Zhong 1995; Zhong and Pena 1995). In addition, similar mechanisms may be prevalent in the CNS because so many neurons express peptide cotransmitters.