Printer Friendly

Phylogenetic analysis of learning-related neuromodulation in molluscan mechanosensory neurons.

The past two decades of research on the cellular bases of learning and memory in taxonomically diverse model systems have revealed a number of mechanistic commonalities (Byrne et al. 1993; Carew and Sahley 1986; Dudai 1989; Hawkins et al. 1993; Martinez and Kesner 1991; Quinn 1984; Squire and Butters 1992). However, no study has directly addressed the question of how specific cellular mechanisms of learning and memory change phylogenetically. Studies in other biological systems suggest that such an evolutionary analysis not only determines which mechanisms have been diversified or conserved during evolution, but can also correlate putative causal mechanisms with specific phenotypic change, thus giving unique insights into these mechanisms. For example, in the study of ion channel structure and function, comparative studies have successfully elucidated which protein motifs tend to be altered by evolution to produce ion channel diversity, and which motifs have been conserved to maintain a constrained function (e.g., Hille 1991). Furthermore, studies on the role of homeobox genes in pattern formation and morphogenesis have been able to directly relate phylogenetic changes in specific gene expression to specific changes in patterns of development (Akam et al. 1994; Patel 1994), thereby generating evidence for particular mechanistic hypotheses. In a similar way, the present study addresses the question of how cellular mechanisms of learning have evolved. We document phylogenetic variation in two neuromodulatory phenomena. This allows us to deduce the evolutionary history of learning-related neuromodulation and also presents unique opportunities to test cellular models of learning and memory.

The molluscan clade we investigated provides a useful starting point for a comparative analysis because many features of the cellular mechanisms of learning in one member of the clade, the sea hare Aplysia californica Cooper, 1863, are well understood (Byrne et al. 1993; Carew 1987; Kandel et al. 1987; full species name is used at first appearance, thereafter only the genus name is used to indicate each particular species). Although learning and memory have been studied in a variety of behaviors in Aplysia (Carew 1987), by far the greatest effort has addressed learning mechanisms in defensive withdrawal reflexes, in which the modulatory transmitter serotonin (5-hydroxytryptamine or 5-HT) is an important mediator of several forms of learning (Mackey et al. 1989; Glanzman et al. 1989; Mercer et al. 1991). One of the primary sites of action of 5-HT is the mechanosensory neurons that initiate the withdrawal reflexes. Two particularly consistent changes effected by 5-HT in sensory neurons are spike broadening (a longer duration action potential; Baxter and Byrne 1990; Mercer et al. 1991; Sugita et al. 1994b) and increased excitability (more action potentials for a given intracellular depolarizing current; Klein et al. 1986; Baxter and Byrne 1990; Mercer et al. 1991). These changes to sensory neuron firing properties contribute to increased reflex responses produced by various forms of learning, including dishabituation, sensitization (both long-term and short-term forms) and classical conditioning (see reviews by Byrne et al. 1993; Carew 1987; Kandel et al. 1987).

The species examined here belong to a monophyletic group of opisthobranch gastropods (Phylum Mollusca). This group first appeared during the Carboniferous, some 340 million years ago (mya; Moore et al. 1952), and diversified significantly during the Jurassic (192 mya; Knight et al. 1960). By testing for 5-HT-induced spike broadening and increased excitability in homologous sensory neurons of these species and mapping these neuromodulatory traits onto the cladogram of their phylogenetic relationships, we deduce how these traits most likely evolved. More specifically, we were interested in the following questions: Which of these traits is phylogenetically more ancient? Can either trait be lost once it has originated? How tightly coupled are these traits evolutionarily?

MATERIAL AND METHODS

Phylogenetic Methods

The phylogenetic relationships of the study species (see below) were derived from previous work done by a variety of molluscan systematists. We extracted from the literature as many phylogenetically informative characters as possible. We narrowed this list to characters that are most widely recognized by opisthobranch systematists (T. Gosliner, Calif. Acad. Sci., pers. comm., 1996) and generated a character matrix. We chose Bulla gouldiana over other possible out-group species because it has a relatively large number of anatomical and morphological traits considered plesiomorphic to (i.e., shared by) the entire opisthobranch clade (Gosliner 1981), and because its nervous system is sufficiently large for intracellular recording. We used the Phylogenetic Algorithm Using Parsimony (PAUP) program (Our thanks to Terrence Gosliner for entering the matrix on PAUP) to generate a most parsimonious cladogram. We mapped the presence or absence of 5-HT-induced increases in spike duration and excitability onto this cladogram and generated an hypothesis to account for the evolution of these two traits using the principle of parsimony (Hennig 1966; Maddison et al. 1984). We evaluated competing hypotheses by direct observation rather than computer analysis as the number of possible evolutionary hypotheses was quite small.

Study Species

Adult individuals of A. californica, Dolabella auricularia, Bursatella leachii, Dolabrifera dolabrifera, Akera bullata, and B. gouldiana, were obtained from the wild by suppliers in California (Aplysia, Alacrity Marine; Bulla, Marinus), Florida (Bursatella, University of Miami Aplysia facility), Hawaii (Dolabella and Dolabrifera, thanks to M. Switzer-Dunlap, M. Hadfield, and D. Gentile), and England (Akera, thanks to D. Seaward, P. Benjamin, and Blade Biological). All animals were held in artificial sea water with a recirculating biofilter exchange system; the temperate species (Aplysia, Bulla, and Akera) were held at 14 [degrees] C to 16 [degrees] C, the tropical species (Dolabella, Bursatella, and Dolabrifera) were held at 20 [degrees] C to 22 [degrees] C. All animals were used within 90 d of their arrival in Colorado.

Physiology

Animals were anesthetized prior to dissection by injection of isotonic Mg[Cl.sub.2] approximately equal to their body volume. The pleural/pedal ganglia were removed and pinned out in a Sylgard-lined dish. The sheath of each ganglion was surgically removed. The dish was perfused with an artificial seawater solution (ASW: 460 mM NaCl, 55 mM Mg[Cl.sub.2], 11 mM Ca[Cl.sub.2], 10 mM KCl, 10 mM Trizma, pH 7.6, temperature 20 [degrees] C to 22 [degrees] C) for the remainder of the experiment. In receptive field experiments, the posterior body wall (Walters et al. 1983a,b) was dissected and left connected to the central nervous system via peripheral pedal nerves.

Single-electrode intracellular recording technique was used for all experiments. Neurons were impaled with single borosilicate glass microelectrodes (8-15 M[ohms]) filled with a 3M KCl solution. Voltage was amplified and intracellular current injected through a Getting (Model 5A) microelectrode amplifier, with a bridge balance to allow injection of current and simultaneous measurement of voltage. The signal was monitored on an oscilloscope (Tektronix Model 511A) and recorded either directly into a computer (Macintosh IIci) using commercial data collection hardware and software (Super-Scope; G.W. Instruments) or into a VCR-adapter recording system (Vetter, Inc.) for subsequent playback and analysis.

To examine the distal axon anatomy of putative sensory neuron homologs and to confirm a sensory-neuron-like response to nerve stimulation, action potentials were elicited with an electrical current (5-10 msec, 10-50 V, capacity coupled, using a stimulus isolation unit) passed through a suction electrode of a diameter appropriate to make a tight seal on the largest posterior-projecting peripheral nerve of the pedal ganglion (most likely homologous to nerve P9 in Aplysia).

To test the receptive field of putative sensory-neuron homologs, the tail and central nervous system was dissected from the animal and pinned to the bottom of a Sylgard-lined dish (see Walters et al. 1983a). Upon impaling a putative tail-sensory homolog, we tested whether stimulating the dissected tail with a glass probe elicited action potentials in the soma of the putative homolog.

Neuroanatomy

To test their proximal neuroanatomy, we impaled the somas of putative sensory homologs with low resistance (3-5 M[ohms]), thin-walled (0.78 mm inner diameter; 1.0 mm outer diameter) electrodes and filled the cells with Lucifer Yellow (5%-7%) via pressure injection (80 psi in multiple pulses of 0.1-60 sec duration using a Picospritzer II, General Valve Corp.). Injected neurons were allowed to incubate at 4 [degrees] C for 0.5-2 h, then fixed in 4% paraformaldehyde (pH 7.4), and finally dehydrated with three to five washes of 95% ethanol. Dehydrated ganglia were immersed in methyl salicylate for more than 20 min, mounted, and viewed with a Leitz compound microscope equipped for epifiuorescent illumination (H2 filter block; 455 nm long-pass excitation filter). Three-dimensional images were captured using a laser confocal microscopy system (see below).

To stain nervous systems for sensorin A, a peptide specific to Aplysia sensory neurons (Brunet et al. 1991), we obtained a polyclonal antibody to this peptide (courtesy of Robert Hawkins, Columbia University) and performed an immunocytochemical procedure modified from Wright et al. (1995). All treatments were performed at 4 [degrees] C. Washes were 2, 5, 10, and 15 min in succession. The ganglia were dissected out of the animals and fixed for 4 h in 4% paraformaldehyde in ASW and washed in phosphate buffered saline (PBS; 50 mM [Na.sub.2]HP[O.sub.4] and 140 mM NaCl in distilled water adjusted to pH 7.4). Then the ganglia were rinsed in PBS and soaked in a 4% (in PBS) solution of Triton X-100 (Pierce) for 1 h, after which they were blocked for 1 h in 10% goat serum and 0.4% Triton in PBS. After washing in PBS, the ganglia were incubated (1:750 dilution) in polyclonal rabbit anti-sensorin antibody with 2% goat serum and 0.5% Triton X-100 in PBS for 72 h. The ganglia were blocked again (same solution as above) for 1 h, and then incubated for 3-4 h in secondary antibody (goat antirabbit) with either an FITC (Cappel, Inc.), or a CY3 (Pierce) fluorescent tag diluted 1:50 in 2% goat serum and 0.5% Triton X-100 in PBS. After rinsing in PBS, the ganglia were cleared in 10% n-propyl gallate in glycerol, mounted, and viewed under a Leitz compound microscope with epifluorescent illumination. Optical images were then captured with a confocal laser microscope (see below).

Confocal Image Capture

After viewing ganglia and orienting them using conventional epifluorescent microscopy, we captured optical sections on computer using a laser confocal microscopy system (CLSM MultiProbe 2001, Molecular Dynamics). The 25 mW argon laser used for excitation had primary emission lines of 457, 488, and 514 nm. Lucifer Yellow-labeled neurons were viewed using an excitation wavelength of 457 nm with a long-pass filter of 510 nm. To view CY3-labeled neurons, we used an excitation wavelength of 514 nm and a long-pass filter of 600 nm. We collected 30-40 optical sections (2-8 [Mu] in thickness), which were then assembled into a three-dimensional image of the preparation using standard software (ImageSpace, Molecular Dynamics). No image enhancement algorithms (e.g., smoothing, sharpening, etc.) were used on these images, although the look up tables were sometimes altered to optimize contrast.

Experimental Protocol

In each experiment, neurons were accepted if resting potentials were stable for more than 5 min, and if spike height was more than 40 mV. Upon achieving a stable intracellular recording of a mechanosensory homolog, two kinds of intracellular current injections were alternated with a 30-sec interval to assay spike duration and excitability. Action potentials for measurement of spike duration were induced with short (2-4 msec) current injections. This short current pulse insured that all of the falling phase and part of the rising phase of the action potential could be recorded in the absence of current input. Spike duration was measured (Super-Scope software; G.W. Instruments) as the time from when membrane voltage reached the peak of the action potential to when it returned to 33% of that peak on the falling phase of the action potential. Action potentials for measurement of excitability were induced with a longer (500 msec) pulse of depolarizing current. The current pulse amplitude was adjusted at the beginning of the experiment to produce one (rarely two) action potentials. Membrane potential was allowed to vary naturally throughout the course of the experiment.

Experiments were performed in a continuously perfused 3-5 ml chamber. At least four baseline measures of spike duration and excitability were recorded in ASW before applying 5-HT through the perfusion system. The perfusion rate was approximately 8 ml [min.sup.-1]. To allow complete perfusion of the dish, we tested spike duration and number of spikes of the fifth through eighth stimuli after application of 5-HT (i.e., approximately 5-8 min after 5-HT onset). The mean of these tests was divided by the average of the four stimuli just prior to 5-HT application to give a proportion change due to 5-HT. After application of 5-HT, ASW was again applied to the preparation, and the spike width and number of spikes of the fifth through eighth stimuli after return to ASW were recorded. To reduce the chances that increases in spike duration or excitability were due to nonspecific effects, we excluded experiments in which any observed increases were not at least partly reversible. Thus, any experiment that showed an increase in spike duration and/or number of spikes in 5-HT, which was not reduced by the wash in ASW, were excluded (Lack of reversibility invalidated no experiments in Aplysia, Dolabella, Bursatella, or Bulla, and only 6 of 33 experiments in Dolabrifera, probably due to the negligible effect of 5-HT in this species). Only experiments in which both spike duration and spike number were recorded are included in this analysis.

A pairwise t-test on the proportion increase was used to evaluate whether 5-HT caused any significant change to either spike duration or number of spikes. Two-sample t-tests were used to compare the proportion increase observed between Aplysia and each of the other species. An analysis of variance (ANOVA) and a post hoc protected LSD (Fisher 1949) was used to compare firing property effects between all species pairs. Because the spike number data tended to be somewhat skewed, we log-transformed the proportion increase in spike number prior to this analysis. We report means [+ or -] standard error of means. All statistical test probabilities are two-tailed.

RESULTS

Phylogenetics

We identified 20 phylogenetically useful character traits from the literature (Appendix). We narrowed these to 12 characters which are more generally recognized among molluscan systematists (Table 1; T. Gosliner, pers. comm., 1996). Using this character matrix, and assuming Bulla to be the outgroup, PAUP generated a single shortest tree [ILLUSTRATION FOR FIGURE 1 OMITTED], which had a consistency index of 1.00. The most closely related species [TABULAR DATA FOR TABLE 1 OMITTED] to A. californica was D. auricularia. Bursatella leachii and D. dolabrifera comprise a sister clade to Dolabella and Aplysia. Akera bullata is a species from a sister family to the aplysiids (Morton 1972; Ghiselin 1966). We assume this well-supported cladogram to be fixed, allowing us to map our two neuromodulatory characters onto it.

Neuronal Homology

We used up to seven characteristics, each relatively unique to tail-sensory neurons in Aplysia, to identify tail-sensory homologs in the other study species (Table 2). Although some of these characteristics may be more powerful indicators of homology than others, in that they are relatively more unique to tail-sensory neurons (location, physiology, proximal neuroanatomy, and immunostaining), none of these characters is sufficient, by itself, to indicate homology; rather they each, relatively independently, strengthen the proposition of homology. The characteristics we used included the following: I location [ILLUSTRATION FOR FIGURE 2 OMITTED]: pleural ganglion, ventral, caudal, near the pleural-pedal connective (Walters et al. 1983a). II Size: smallest neurons in the pleural ganglion (Walters et al. 1983a). III Resting physiology: no spontaneous action potentials or post-synaptic potentials (Walters et al. 1983a). IV Expression of a peptide specific to mechanosensory neurons (Brunet et al. 1991). V Proximal neuroanatomy [ILLUSTRATION FOR FIGURE 2 OMITTED]: single axon running from the soma in the pleural ganglion to the adjacent pedal ganglion (Walters et al. 1983a). VI Distal neuroanatomy [ILLUSTRATION FOR FIGURE 2 OMITTED]: antidromic stimulation of distal peripheral nerve from the pedal ganglion produces short latency high-fidelity action potential rising from stable baseline (Walters et al. 1983a). VII Receptive field: mechanical stimulation of posterior body wall produces a burst of action potentials in the sensory neuron soma (Walters et al. 1983a).

For example, putative tail-sensory neurons of Dolabrifera satisfied each of these criteria as follows [ILLUSTRATION FOR FIGURE 3 OMITTED]. Putative homologs were identified as the smallest neurons in the ventral-caudal part of the pleural ganglion (characteristics I and II; stained cells in Fig. 3A2 demonstrate this location). Cells in this location were impaled and further verified as sensory neurons by their quiet resting potential (characteristic III; see baseline of [ILLUSTRATION FOR FIGURE 3D OMITTED]). Immunocytochemical experiments indicated that a high proportion of the cells in this location were immunopositive to sensorin, a peptide that in Aplysia is expressed exclusively in sensory neurons (characteristic IV; [ILLUSTRATION FOR FIGURE 3A2 OMITTED]; [ILLUSTRATION FOR FIGURE 3A1 OMITTED] shows Aplysia for comparison). Similar results were obtained in three additional preparations.

Both the proximal and distal neuroanatomy of putative sensory-neuron homologs in Dolabrifera closely resembled that of Aplysia tail-sensory neurons. Based on intracellular Lucifer Yellow fills (five preparations), the proximal neuroanatomy (characteristic V) of these cells [ILLUSTRATION FOR FIGURE 3B2 OMITTED], resembled that of Aplysia [ILLUSTRATION FOR FIGURE 3B1 OMITTED]: a single axon exits the soma and runs without major branches into the pedal ganglion. We investigated how unique this neuroanatomy is in Aplysia by filling 25 neurons whose position and physiology indicated that they were not sensory neurons. The neuroanatomy of each of these neurons (data not shown) differed markedly from that of tail-sensory neurons [ILLUSTRATION FOR FIGURE 3B OMITTED]. In particular, each of the nonsensory neurons showed multiple axons, major branches in a single axon, and/or axons that ran to the cerebral or visceral ganglia instead of the pedal ganglion. Thus, the proximal anatomy of tail-sensory neurons in Aplysia appears to be relatively unique, which increases the utility of this trait for testing homology.

The distal neuroanatomy (characteristic VI) of putative tail-sensory homologs in Dolabrifera was also similar to that of tail-sensory neurons in Aplysia. A high fidelity one-for-one antidromic spike in response to a high frequency (50 Hz) electrical stimulation of a pedal nerve was observed [ILLUSTRATION FOR FIGURE 3C OMITTED]. These antidromic spikes indicate that the recorded neuron has an axon in the pedal nerve, and this pattern of response (constant latency, flat baseline) is strikingly similar to that observed in tail-sensory neurons of Aplysia given similar stimulation (Walters et al. 1983a, their [ILLUSTRATION FOR FIGURE 4B OMITTED]). We repeated this experiment three additional times with similar results.

Finally, the receptive field of these neurons was the same as that of tail-sensory neurons of Aplysia (characteristic VII). In reduced preparations consisting of the tail connected by its nerves to the central nervous system, stimulation of the tail with a glass probe activated a barrage of action potentials with no prepotentials ([ILLUSTRATION FOR FIGURE 3D OMITTED]; similar results observed in two additional preparations). Again, this response is quite similar to the response of tail-sensory neurons of Aplysia to tactile stimulation of the tail (Walters et al. 1983a, their [ILLUSTRATION FOR FIGURE 4C OMITTED]). These results in Dolabrifera strongly suggest that the neurons studied are homologous to the tail-sensory neurons of Aplysia.

Although not all seven of the above tests for homology were used in all of the species studied, each species satisfied the tests performed (Table 2). Location, relative size, and resting physiology criteria were satisfied for all species in the course of doing the physiology experiments (see below). To increase our confidence in homology, we tested for one to four additional characteristics in the ingroup species, including all four characteristics in Dolabrifera, the species whose physiology deviated most dramatically from that of Aplysia (see below). In principle, this means our confidence in the homology for each species is not equal (Dolabrifera [greater than] Dolabella [greater than] Bursatella [greater than] Akera/Bulla). However, in light of the reliable correspondence between these traits in Aplysia and Dolabrifera, we think it quite likely that the neurons studied were homologous.

Effect of 5-HT on Sensory Neurons in Aplysiay and Dolabella

5-HT broadened action potentials and increased excitability in both Aplysia and Dolabella. For example, in Aplysia tail-sensory neurons [ILLUSTRATION FOR FIGURE 4A OMITTED], the narrow (ca. 1 msec) action potential typical of sensory neurons in Aplysia was observed in control conditions (PRE, [ILLUSTRATION FOR FIGURE 4A OMITTED]). Upon application of 10 [[micro]molar] 5-HT (5-HT, [ILLUSTRATION FOR FIGURE 4A OMITTED]), the action potential in this example broadened some 60%, and returned to near control levels after washout of the 5-HT with ASW. 5-HT also had a powerful effect on excitability in these sensory neurons ([V.sub.m], middle three traces of [ILLUSTRATION FOR FIGURE 4A OMITTED]). Before 5-HT, the 500 msec current pulse elicited a single spike. Upon application of 5-HT, excitability was greatly increased, resulting in nine spikes. This increase in excitability was readily reversed during washout with ASW solution. Both of these responses in Aplysia are well documented (reviewed by Byrne et al. 1993; Kandel et al. 1987).

The effect of 5-HT on tail-sensory neuron homologs in the closely related species, Dolabella, was very similar to that in Aplysia [ILLUSTRATION FOR FIGURE 4B OMITTED]. In one example, 5-HT increased spike duration (top panel, [ILLUSTRATION FOR FIGURE 4B OMITTED]) by approximately 50%. 5-HT also enhanced excitability. The same depolarizing current pulse that elicited one spike in ASW produced six spikes in the presence of 5-HT ([V.sub.m], middle three traces, [ILLUSTRATION FOR FIGURE 4B OMITTED]).

Effect of 5-HT on Sensory Neurons in Bursatella and Dolabrifera

We next tested for 5-HT-induced effects in a clade somewhat more distantly related to Aplysia. The effect of 5-HT on the firing properties of tail-sensory-neuron homologs in one species from this clade, Bursatella, was similar to its [TABULAR DATA FOR TABLE 2 OMITTED] effects in Aplysia and Dolabella [ILLUSTRATION FOR FIGURE 4C OMITTED]. In this example, 5-HT caused approximately 50% broadening of the action potential (top panel, [ILLUSTRATION FOR FIGURE 4C OMITTED]), and an increase in excitability from one to five action potentials ([V.sub.m], middle three traces, [ILLUSTRATION FOR FIGURE 4C OMITTED]). Unlike its strong effects on sensory-neuron firing properties in the first three species, serotonin had little or no effect in Dolabrifera [ILLUSTRATION FOR FIGURE 4D OMITTED]. In this example, when 10 [[micro]molar] 5-HT was applied to the nervous system, the action potential broadened only slightly ([ILLUSTRATION FOR FIGURE 4D OMITTED], top traces) and 5-HT had no effect on excitability ([ILLUSTRATION FOR FIGURE 4D OMITTED], [V.sub.m], bottom three traces).

Statistics from seven experiments confirm that 5-HT had very little effect on the tail-sensory homologs of Dolabrifera compared to Aplysia [ILLUSTRATION FOR FIGURE 5 OMITTED]. Not only was the modest broadening induced by 5-HT in the sensory neurons of Dolabrifera indistinguishable from no change (t = 1.73; P = 0.13; n = 7), but this negligible broadening was also significantly less than that observed in the tail-sensory neurons of Aplysia (t = 2.48; P = 0.033; [ILLUSTRATION FOR FIGURE 5A OMITTED]). In addition, 5-HT had no significant effect on excitability in Dolabrifera [ILLUSTRATION FOR FIGURE 5B OMITTED]. The slight 5-HT-induced increase in number of spikes was not significantly different from pre-5-HT baseline levels (t = 1.84; P = 0.11; n = 7). In addition, this negligible increase was significantly smaller than that seen in the tail-sensory neurons of Aplysia (t = 11.13; P [less than or equal to] 0.001).

Because its tail-sensory homologs showed neither 5HT-dependent spike broadening nor increased excitability, we focused additional experiments on Dolabrifera to explore the limits of this lack of response. To test the hypothesis that a different concentration of 5-HT might be required to enhance spike duration and excitability comparable to that observed in Aplysia, we repeated these experiments with 5-HT concentrations ranging over three orders of magnitude (0.1-100 [[micro]molar]). The protocol was the same as before, except that in each preparation we began with 0.1 [[micro]molar] 5-HT and repeated the experiment four times, incrementing the applied 5-HT concentration by a factor of 10 each time. A different sensory neuron was used at each 5-HT concentration. In spite of some variation in the proportion change in spike duration [ILLUSTRATION FOR FIGURE 6A OMITTED] and excitability [ILLUSTRATION FOR FIGURE 6B OMITTED], no significant spike broadening or increased excitability was observed with any of the applied 5-HT concentrations.

All of the above experiments used a single depolarizing current level to test for excitability. To test whether excitability using different depolarizing current levels might be more sensitive to 5-HT in Dolabrifera, we repeated the excitability experiments with a range of depolarizing current levels, 0.1, 0.2, and 0.3 nA, before and after applying [10.sup.-5] 5-HT [ILLUSTRATION FOR FIGURE 7 OMITTED]. Current amplitude had no significant impact on 5-HT-induced increases in excitability (repeated measures two-way ANOVA, Current X 5-HT interaction, [F.sub.2,12] = 1.201; NS). No significant increase in excitability was observed at any of the three current amplitudes (paired t-tests, t = 1.64, 1.10, and 1.36; NS). Thus, these broader tests demonstrated that the firing properties of tail-sensory homologs of Dolabrifera are extremely resistant, if not impervious, to applied 5-HT.

Effect of 5-HT on Sensory Neurons in Akera and Bulla

To gain a broader understanding of the evolution of 5-HT-induced increases in spike duration and excitability, we performed additional experiments in two outgroup species, A. bullata and B. gouldiana [ILLUSTRATION FOR FIGURE 1 OMITTED]. Due to limited availability, we were able to perform only one experiment in Akera [ILLUSTRATION FOR FIGURE 4E OMITTED]. In that experiment, 5-HT caused no change in spike duration ([ILLUSTRATION FOR FIGURE 4E OMITTED], top panel), but did increase the number of spikes to six from one in control conditions ([ILLUSTRATION FOR FIGURE 4E OMITTED], [V.sub.m], middle thee panels). Similar results were obtained for Bulla [ILLUSTRATION FOR FIGURE 4F OMITTED]. 5-HT had no effect on spike duration ([ILLUSTRATION FOR FIGURE 4F OMITTED], top panel), but increased excitability from one spike in control conditions to six spikes in 5-HT ([ILLUSTRATION FOR FIGURE 4F OMITTED], [V.sub.m], middle three panels).

Summary Statistics

We performed 5-10 experiments like the ones described above in five of the six study species ([ILLUSTRATION FOR FIGURE 8 OMITTED]; the one experiment in Akera is also plotted for comparison). Although a fair amount of variation between species exists, the primary result is that significant 5-HT-induced increases in spike duration were observed in the tail-sensory homologs of Aplysia, Dolabella, and Bursatella, whereas no such changes were observed in the homologs of Dolabrifera, Akera, or Bulla. All species showed statistically significant increases in excitability except Dolabrifera.

Spike duration in control conditions appeared to vary substantially among these species (top panels [ILLUSTRATION FOR FIGURE 4A-F OMITTED]). A statistical comparison of the grouped data confirm this suggestion (Table 3). Specifically, an ANOVA of all the data was highly significant. A post hoc comparison indicated that the relative order beginning with the briefest action potential was Bulla = Aplysia [less than] Bursatella [less than] Dolabella [less than] Dolabrifera.

DISCUSSION

Our data indicate that two neuromodulatory processes thought to contribute to several forms of learning in Aplysia vary significantly among opisthobranch gastropods. They also suggest how these processes have evolved.

To gain a more specific understanding of their evolution, 5-HT-induced modulation of spike duration and excitability were mapped onto the cladogram of Fig. 1 using parsimony methods [ILLUSTRATION FOR FIGURE 9 OMITTED]. In the case of excitability, because all species except Dolabrifera show 5-HT-induced increases, this neuromodulatory trait was most likely present (bold striped line) in the common ancestor of the aplysiids, and was subsequently lost (lightly shaded striped line) in the lineage leading to Dolabrifera, sometime after the split of that lineage from the one leading to Bursatella.

The phylogenetic interpretation of spike broadening is somewhat more complex. In particular, although the common opisthobranch ancestor probably lacked spike broadening (both Akera and Bulla lack the trait), the lineage leading to Dolabrifera may have either lost spike broadening, or never had it. Because Aplysia, Dolabella, and Bursatella all show spike broadening, the trait was probably present (bold solid line, [ILLUSTRATION FOR FIGURE 9 OMITTED]) in the common ancestor of all four species and subsequently lost (light solid line, [ILLUSTRATION FOR FIGURE 9 OMITTED]) in the lineage leading to Dolabrifera. However, because spike broadening is not present in Bulla or Akera, an equally parsimonious hypothesis (not shown in [ILLUSTRATION FOR FIGURE 9 OMITTED]) is that spike broadening was not present in the common ancestor of the aplysiids, but appeared independently twice, once just prior to the split between Aplysia and Dolabella and once in the lineage leading to Bursatella after its split with the lineage leading to Dolabrifera.

Further experiments with other aplysiid species are needed to resolve this issue. For example, other genera are thought to be more closely related to Dolabrifera (Petalifera and Phyllaplysia) on the one hand and to Bursatella (Steilocheilus and Notarchus) on the other (Eales 1944; Marcus 1972). If sensory homologs of the species closely related to Dolabrifera all showed spike broadening, then the interpretation of Figure 9 would be favored. If such genera did not show spike broadening, the issue would remain unresolved. An alternative course would be to compare the spike broadening of Bursatella versus Aplysia in more detail for clues about common ancestry. For example, spike broadening in Aplysia can be observed in response to activators of protein kinase C (Sugita et al. 1992). If these activators were ineffective on the sensory neurons of Bursatella, this would support the idea that spike broadening was independently evolved in Bursatella. Alternatively, a voltage-clamp study of the tail-sensory homologs might uncover significant deviations from such studies already done in Aplysia (e.g., Baxter and Byrne 1989, 1990; Jarrard et al. 1993).

Regardless of the precise details of the above evolutionary hypotheses, a conservative interpretation of these data is that spike broadening was more recently evolved than was increased excitability. Clearly, a wider survey of opisthobranchs would strengthen this interpretation. The pleural ganglion is well conserved among other opisthobranchs, as well as pulmonates (the likely sister taxon of the opisthobranchs; Gosliner 1981), and other gastropods (Bullock and Horridge 1965). Thus, the prospects of locating tail-sensory homologs in these related species are good, and further tests of this evolutionary interpretation are therefore feasible.

Work on the development of neuromodulation in Aplysia (Marcus and Carew 1989) showed that 5-HT-induced spike broadening emerges much later in development than does increased excitability. This developmental sequence is consistent with the evolutionary patterns detected in the present study. Phylogenetically more ancestral traits are typically expressed earlier in development than are more recently derived traits (e.g., Raff and Kaufman 1983).

Are these two traits evolving independently, or do they appear to change as a functional unit? Although the data are not sufficient to give a statistical answer to this question, some independent evolution appears likely, because increased excitability exists in the two outgroup species in the absence of spike broadening [ILLUSTRATION FOR FIGURE 9 OMITTED]. However, spike broadening might still be mechanistically linked to increased excitability, since both traits are absent in Dolabrifera. Thus, spike broadening may have evolved as an elaboration of the mechanism that leads to increased excitability.

Finally, the data strongly support the possibility that learning mechanisms can be lost evolutionarily. The capacity of 5-HT to increase excitability was quite likely lost in the lineage leading to Dolabrifera [ILLUSTRATION FOR FIGURE 9 OMITTED]. Although the data are more equivocal, spike broadening may also have been lost in that same lineage.

Because 5-HT-induced increases in excitability in tail-sensory homologs of Bursatella and Dolabella were somewhat lower than those in the tail-sensory neurons of Aplysia [ILLUSTRATION FOR FIGURE 9 OMITTED] differences may exist in the effect of 5-HT on the tail-sensory neurons of these two species, or alternatively in the 5-HT-independent firing properties of these neurons. The latter possibility seems likely because the spike duration (in control conditions) of tail-sensory homologs of both Bursatella and Dolabella was greater than the spike duration of Aplysia tail-sensory neurons (Table 3). In particular, because the tail-sensory-neuron action potential in Aplysia was narrower, this quality alone may be sufficient to explain the difference in the number of spikes fired during a 500 msec current pulse in the presence of 5-HT: both cells may be firing at a maximum rate that is limited not by the effects of 5-HT, but simply by the time it takes the neuron to fire and recover. Thus, the neuromodulatory differences observed between Dolabella and Bursatella versus Aplysia [ILLUSTRATION FOR FIGURE 8B OMITTED] may reflect background biophysical differences in these sensory neurons rather than a difference in their neuromodulatory response per se. Similar arguments may be made for the single experiment in Akera, although the lack of a statistical sample size precludes firm discussion. On the other hand, because spike duration in the tail-sensory homologs of Bulla was actually briefer than that of tail-sensory neurons in Aplysia, the reduced effect of 5-HT on excitability in Bulla may reflect real differences in the subcellular response of these tail-sensory homologs to 5-HT, relative to that of tail-sensory neurons in Aplysia. Future experiments on the pharmacology and biophysics of 5-HT-induced modulation in tail-sensory homologs of Bulla will be required to help resolve the evolutionary relationships of excitability changes in these two species.

Evolution of Intracellular Signaling

Our results indicate that 5-HT-induced increases in spike duration and excitability are either severely reduced or missing in Dolabrifera, and raise the question of which components of the intracellular communication system that produce these firing property changes in Aplysia might be missing or inactivated in Dolabrifera. To answer this question, we need to consider the intracellular events that produce these firing property changes in Aplysia. Although spike broadening and increased excitability in Aplysia are both initiated by receptor binding of 5-HT (reviewed by Byrne et al. 1993), these two processes may be activated by two different receptors (Mercer et al. 1991) or by the same receptor with different sensitivities (Jarrard et al. 1993). These receptors access second messenger cascades via membrane bound G-proteins. Increases in excitability are thought to involve a cAMP cascade that ultimately results in closure of a potassium channel (the "S" channel; reviewed by Kandel et al. 1987; Byrne et al. 1993). Increased spike duration is also caused to some extent by the closure of the S-channel, but recent work suggests that spike broadening is primarily due to a slowing of the voltage dependent potassium channel ([I.sub.kv]; Baxter and Byrne 1989; Goldsmith and Abrams 1992). Furthermore, these changes to [I.sub.kv] may be mediated by a cAMP-independent pathway, perhaps through protein kinase C (Baxter and Byrne 1990; Sugita et al. 1994a), although debate persists concerning this point (Goldsmith and Abrams 1992; Jarrard et al. 1993).
TABLE 3. Action potential duration of mechanosensory neurons in six
species of opisthobranch gastropods. SE, Standard error of the mean.
One-way ANOVA: [F.sub.4,34] = 22.74; P [less than or equal to]
0.0001; data from Akera not included due to low sample size. Post
hoc statistical comparisons (Fisher's protected LSD) at the P
[less than or equal to] 0.05 level; [Aplysia = Bulla] [less than]
Bursatella [less than] Dolabella [less than] Dolabrifera.

                            Mean (ms)         SE         N

Bulla gouldiana               1.10           0.14        8
Akera bullata                 2.88             -         1
Aplysia californica           0.97           0.06       10
Dolabella auricularia         2.25           0.23        5
Bursatella leachii            1.61           0.12        9
Dolabrifera dolabrifera       2.96           0.32        7


Which of these intracellular events might be missing or inactivated in the sensory neurons of Dolabrifera? Because both spike broadening and increased excitability are absent in Dolabrifera, the simplest hypothesis is that something in common to both processes is defunct. Because a single serotonin receptor may be the only element (if any) in common to spike broadening and increased excitability, it is a reasonable candidate locus. If this hypothesis is correct, then activators of the two second messenger systems should readily increase excitability and spike duration, exactly as they do in Aplysia (Baxter and Byrne 1985, 1990; Byrne et al. 1993; Castellucci et al. 1980; Sugita et al. 1992). Thus, for example, membrane permeable cAMP and phorbol diacetate (a potent activator of protein kinase C) should increase excitability and spike duration (respectively) as readily in Dolabrifera as in Aplysia (Baxter and Byrne 1990; Sugita et al. 1992). If so, other modulatory transmitters, such as small cardiac peptide, that are known to increase cAMP levels and sensory neuron excitability (Abrams et al. 1984; Billy and Walters 1989), might still have a strong effect. On the other hand, if differing titers of cAMP can account for both spike broadening and increased excitability, as suggested by Abrams and colleagues (Abrams and Goldsmith 1992; Goldsmith and Abrams 1992; Jarrard et al. 1993), other intracellular processes would be common to both excitability and spike duration changes, and could account for the absence of these traits in Dolabrifera. Alternatively, two different evolutionary changes to these cellular signaling pathways might have independently lesioned these two neuromodulatory traits.

The absence of 5-HT-induced increases in spike duration and excitability in Dolabrifera tail-sensory homologs may not represent the only neuromodulatory differences between this species and the rest of the aplysiids. Indeed, several different 5-HT-induced effects known in Aplysia may also be reduced or absent in Dolabrifera. For example, does the tail sensorimotor synapse in Dolabrifera show spike-broadening-independent facilitation such as is known to occur in Aplysia (Gingrich and Byrne 1985; Hochner et al. 1986; Braha et al. 1990; Sugita et al. 1992; Klein 1994, 1995)? Investigations of such effects will not only help to sort out the nature of neuromodulation in Dolabrifera but may also provide evidence pertaining to the causal interconnections of these different processes in Aplysia.

Phylogenetic Lesions: The Use of Phylogenetic Variation to Test Mechanistic Hypotheses

Over the past two decades, work in Aplysia has made a substantial contribution to our understanding of likely cellular and subcellular mechanisms of learning and memory (reviews by Kandel et al. 1987; Byrne et al. 1993). For example, this work clearly established that neuronal modulation plays a key role in mediating different forms of short-term learning. Furthermore, it established the likelihood that different forms of learning - such as sensitization, a nonassociative generalized increase in reflex responding after a noxious stimulus, and classical conditioning, an associative increase in responding to an innocuous stimulus after it has been repeatedly paired with a noxious stimulus - share many mechanistic features in common (Hawkins et al. 1983; Walters and Byrne, 1983; Byrne et al. 1993). Thus, the models of learning and memory in Aplysia are rich in detail with generalizations that are useful to other systems (e.g., Dudai 1989).

However, many of the specific features of these models are based on correlational data. Tests of sufficiency or necessity of particular features are rare (although they do exist; Mercer et al. 1991; Glanzman et al. 1989). For example, serotonin-induced increases in spike duration and excitability of sensory neurons are thought to contribute to both sensitization as well as classical conditioning in Aplysia. However, it is technically impossible to test the necessity of spike broadening and increased excitability for the enhanced withdrawal reflex that occurs during these forms of learning. That is, one cannot selectively eliminate spike broadening in sensory neurons and test for reductions in learning capability.

Species like Dolabrifera, Bulla, and perhaps Akera, which are missing one or both 5-HT-induced effects on firing properties, could well be useful in the future as "phylogenetic lesions" to test the necessity of these effects on learning. For example, one would predict that noxious stimuli should have little or no effect on the synaptic strength of the sensorimotor connection in the tail withdrawal reflex of Dolabrifera. Taken to the limit, behavioral sensitization as well as classical conditioning should also be reduced in Dolabrifera. Recent work indicates that sensitization is indeed severely reduced, if not entirely absent in this species (Wright and Maynard 1994). Similar behavioral experiments are underway to test other forms of facilitatory learning such as dishabituation (non-associative facilitation of a habituated reflex) and classical conditioning, in the tail withdrawal reflex of Dolabrifera.

Akera and Bulla may be useful in the future to investigate the relative roles of sensory-neuron spike broadening and increased excitability in different forms of learning, because spike broadening but not increased excitability is missing in this species. For example, recent work in Aplysia suggests that the mechanisms underlying long-term memory for sensitization may be partly independent of those underlying short-term memory (Emptage and Carew 1993). Selective pharmacological elimination of 5-HT-induced spike broadening, which left excitability increases intact, abolished short-term synaptic facilitation without affecting long-term synaptic facilitation. Therefore species (such as Bulla or Akera), which show 5-HT induced increases in excitability, but not in spike duration, may express little or no short-term sensitization, yet simultaneously show long-term sensitization comparable to that of Aplysia.

Phylogenetic lesions may also be useful in testing specific mechanistic hypotheses of associative learning processes. For example, the most widely cited model for the mechanism of associative conditioning of the gill-withdrawal reflex in Aplysia involves an activity dependent amplification of 5-HT-induced effects in sensory neurons (Walters and Byrne 1983; Carew et al. 1983; Elliot et al. 1994), resulting in enhanced spike broadening and synaptic facilitation. This model predicts that Dolabrifera behavior should have a reduced capacity for associative conditioning, because there are virtually no 5-HT effects to be amplified. An alternative 5-HT-independent model, based on recent work on cultured Aplysia neurons (Lin and Glanzman 1994a,b), involves postsynaptic activity of the motor neuron and is partially dependent on NMDA receptors. This "Hebbian" mechanism resembles models of associative synaptic plasticity invoked for many mammalian systems (see e.g., Baudry and Davis 1991). If this mechanism is intact in Dolabrifera, then we would be in the unique position of being able to test its relative role (synaptic as well as behavioral) in the absence of 5-HT-related effects.

Adaptive Interpretation

Our results indicate that one of the aplysiid species, D. dolabrifera, is missing two mechanisms that, in the tail-withdrawal reflex of Aplysia, contribute to facilitatory learning phenomena such as dishabituation, sensitization, and classical conditioning. It is important to realize that this lack of expression is observed only in two particular mechanisms of learning. The presumed consequent reduction of behavioral expression of different forms of learning has not yet been rigorously tested. However, preliminary behavioral experiments (Wright and Maynard 1994) indicate that sensitization in Dolabrifera is absent, or at least severely reduced.

These results raise the following question: What is the adaptive advantage of not expressing a form of learning that appears to be widely expressed in related species? This question is hampered by our lack of understanding of the adaptive role of facilitatory learning in natural populations of opisthobranchs. The usual interpretation is that if an individual's environment becomes more noxious, perhaps because of the presence of a fish or invertebrate predator (Carefoot 1987; Pennings 1990, 1994), the individual sea slug will be more successful if it more vigorously withdraws its body parts from potential harm. However, no study has demonstrated that, for example, a sensitized group of Aplysia is more successful at avoiding physical damage (e.g., from predation) than is a nonsensitized control group. Thus, this interpretation must remain conjecture. If correct, however, why has the lineage leading to Dolabrifera lost this capacity?

Dolabrifera has a rather unique habitat on the undersides of rocks in the shallows inshore of fringing reefs (Kay 1979; Wright, pers. obs.). Unlike Aplysia and Bursatella, Dolabrifera is exclusively nocturnal (Kay 1979). Thus, visual predation on this species may be quite rare, reducing the adaptive advantage of sensitization. Unfortunately, a similar nocturnal habit is also found in Dolabella, a species that shows robust learning related neuromodulation (although behavioral sensitization has not been tested). Clearly, adaptive interpretation of the evolution of mechanisms of sensitization will be severely hampered until more examples of clades that have lost this reportedly ubiquitous form of behavioral plasticity have been correlated with the natural histories of those clades. However, these results raise the clear possibility that even simple forms of learning may have ecological costs that outweigh their benefits.

In conclusion, because this system is so well defined at both evolutionary and mechanistic levels, it may in the future contribute to our understanding, not only of the evolution of specific mechanisms of learning, but of cellular signaling processes in general.

ACKNOWLEDGMENTS

We are grateful to E. Marcus and D. Bird for extensive discussions. We also thank D. McLennan, R. Espinoza, K. Nishikawa, and T. H. Bullock, T. Gosliner, A. R. Palmer, and an anonymous reviewer for critically reading the manuscript. We thank R. Hawkins for providing us with sensorin antibody, and R. Bourret for advice. We thank M. Ghiselin and T. Gosliner for assistance with the opisthobranch phylogeny. We thank the Department of Anatomy and Neurobiology at Colorado State University for use of their confocal microscope facility. This work was supported in part by the National Science Foundation IBN-9511215, D.K. was supported by a Hughes Undergraduate Research Fellowship. B.M. was supported by the Program in Neuronal Growth and Development at Colorado State University.

LITERATURE CITED

ABRAMS, T. W., AND B. A. GOLDSMITH. 1992. cAMP modulation of multiple [K.sup.+] currents contributes to both action potential broadening and increased excitability in Aplysia sensory neurons. Soc. Neurosci. Abstr. 18:16.

ABRAMS, T. W., V. F. CASTELLUCCI, J. S. CAMARDO, E. R. KANDEL, AND P. E. LLOYD. 1984. Two endogenous neuropeptides modulate the gill and siphon withdrawal reflex in Aplysia by pre-synaptic facilitation involving cAMP-dependent closure of a serotonin-sensitive potassium channel. Proc. Nat. Acad. Sci., USA 81:7956-7960.

AKAM, M., P. HOLLAND, P. INGHAM, AND G. WRAY, EDS. 1994. The evolution of developmental mechanisms. Company of Biologists, London.

BAXTER, D. A., AND J. H. BYRNE. 1985. Forskolin-modulated membrane currents in Aplysia tail-sensory neurons. Soc. Neurosci. Abstr. 11:789.

-----. 1989. Serotonergic modulation of two potassium currents in the pleural sensory neurons of Aplysia. J. Neurophysiol. 64: 665-679.

-----. 1990. Differential effects of cAMP and serotonin on membrane current, action potential duration, and excitability in somata of pleural sensory neurons of Aplysia. J. Neurophysiol. 64: 978-990.

BAUDRY, M., AND J. L. DAVIS, EDS. 1991. Long-term potentiation: A debate of current issues. MIT Press, Cambridge, MI.

BILLY, A. J., AND E. T. WALTERS. 1989. Modulation of mechano-sensory threshold in Aplysia by serotonin, small cardioctive peptideB, FMRFamide, acetylcholine, and dopamine. Neurosci. Lett. 105:200-204.

BRAHA, O., N. DALE, B. HOCHNER, M. KLEIN, T. W. ABRAMS, AND E. R. KANDEL. 1990. Second messengers involved in the two processes of presynaptic facilitation that contribute to sensitization and dishabituation in Aplysia sensory neurons. Proc. Nat. Acad. Sci. USA 87:2040-2044.

BRUNET, J.-F., E. SHAPIRO, S. A. FOSTER, E. R. KANDEL, AND Y. LINO. 1991. Identification of a peptide specific for Aplysia sensory neurons by PCR-based differential screening. Science 252: 856-859.

BULLOCK, T. H., AND G.A. HORRIDGE. 1965. Structure and function in the nervous systems of invertebrates. Freeman, San Francisco, CA.

BYRNE, J. H. 1987. Cellular analysis of associative learning. Physiol. Rev. 67:329-439.

BYRNE, J. H., R. ZWARTJES, R. HOMAYOUNI, S. D. CRITZ, AND A. ESKIN. 1993. Roles of second messenger pathways in neuronal plasticity and in learning and memory. Pp. xxxx-xxxx in S. Senolikar and A. C. Nairn, eds. Advances in second messenger and phosphoprotein research. Raven Press, New York.

CAREFOOT, T. H. 1987. Aplysia: Its biology and ecology. Annu. Rev. Oceanogr. Mar. Biol. 25:167-284.

CAREW, T. J. 1987. Cellular and molecular advances in the study of learning in Aplysia. Pp. 177-204 in J. P. Changeaux and M. Konishi, eds. The neural and molecular basis of learning. Wiley, New York.

CAREW, T. J., AND C. L. SAHLEY. 1986. Invertebrate learning and memory: From behavior to molecules. Annu. Rev. Neurosci. 9: 435-487.

CAREW, T. J., R. D. HAWKINS, AND E. R. KANDEL. 1983. Differential classical conditioning of a defensive withdrawal reflex in Aplysia californica. Science 219:397-400.

CASTELLUCCI, V. F., E. R. KANDEL, J. H. SCHWARTZ, F. D. WILSON, A. C. NAIRN, AND P. GREENGARD. 1980. Intracellular injection of the catalytic subunit of cyclic AMP-dependent protein kinase simulates facilitation of transmitter release underlying behavioral sensitization in Aplysia. Proc. Nat. Acad. Sci. USA 77: 7492-7496.

DUDAI, Y. 1989. The neurobiology of memory. Oxford Univ. Press, New York.

EALES, N. B. 1944. Aplysiids from the Indian ocean. Proc. Malacol. Soc. Lond. 2:1-22.

ELLIOT, L. S., R. D. HAWKINS, E. R. KANDEL, AND S. SCHACHER. 1994. Pairing-specific, activity-dependend presynaptic facilitation at Aplysia sensory-motor neuron synapses in isolated cell culture. J. Neurosci. 14:368-383.

EMPTAGE, N. J., AND T. J. CAREW. 1993. Long-term synaptic facilitation in the absence of short-term facilitation in Aplysia neurons. Science 262:253-256.

FISHER, R. A. 1949. The design of experiments. Oliver and Boyd, Edinburgh, Scotland.

FRANZEN, A. 1955. Comparative morphological investigations into the spermiogenesis among Mollusca. Zool. Bid. Upps. 30:399-456.

GHISELIN, M. T. 1966. Reproductive function and phylogeny of opisthobranch gastropods. Malacologia 3:327-378.

GINGRICH, K. J., AND J. H. BYRNE. 1985. Simulation of synaptic depression, posttetanic potentiation, and presynaptic facilitation of synaptic potentials from sensory neurons mediating gill-withdrawal reflex in Aplysia. J. Neurophysiol. 53:652-669.

GLANZMAN, D. L., S. L. MACKEY, R. D. HAWKINS, A. M. DYKE, P. E. LLOYD, AND E. R. KANDEL. 1989. Depletion of serotonin in the nervous system of Aplysia reduces the behavioral enhancement of gill withdrawal as well as the heterosynaptic facilitation produced by tail shock. J. Neurosci. 9:4200-4213.

GOLDSMITH, B. A., AND T. W. ABRAMS. 1992. cAMP modulates multiple [K.sup.+] currents increasing spike duration and excitability in Aplysia sensory neurons. Proc. Nat. Acad. Sci. USA 89: 11481-11485.

GOSLINER, T. M. 1981. Origins and relationships of primitive members of the Opisthobranchia. Biol. J. Linn. Soc. 16:197-225.

-----. 1991. Morphological parallelism in opisthobranch gastropods. Malacologia 32:313-327.

GUIART, J. 1901, Contribution a l' etude des Gasteropodes Opisthobranches et en pariculier des Cephalaspides. Mem. Soc. zool. Fr. 14:5-219.

HAWKINS, R. D., T. W. ABRAMS, T. J. CAREW, AND E. R. KANDEL. 1983. A cellular mechanism of classical conditioning in Aplysia: Activity-dependent amplification of presynaptic facilitation. Science 219:400-405.

HAWKINS, R. D., E. R. KANDEL, AND S. A. SIEGELBAUM. 1993. Learning to modulate transmitter release: Themes and variations in synaptic plasticity. Annu. Rev. Neurosci. 1993:625-665.

HENNIG, W. 1966. Phylogenetic systematics. Univ. of Illinois Press, Urbana.

HILLE, B. 1991. Ionic channels of excitable membranes Sinauer, Sunderland, MA.

HOCHNER, B., M. KLEIN, S. SCHACHER, AND E. R. KANDEL. 1986. Additional component in the cellular mechanism of presynaptic facilitation contributes to behavioral dishabituation in Aplysia. Proc. Nat. Acad. Sci. USA 83:8794-8798.

JARRARD, H. E., B. A. GOLDSMITH, AND T. W. ABRAMS. 1993. In Aplysia sensory neurons, the neuropeptide SCPb and serotonin differ in efficacy both in modulating cellular properties and in activating adenylyl cyclase: Implications for mechanisms underlying presynaptic facilitation. Brain Res. 616:188-199.

KANDEL, E. R., B. HOCHNER, M. SHYSTER, S. A. SIEGELBAUM, R. D. HAWKINS, V. F. CASTELLUCCI, AND T. W. ABRAMS. 1987. Synaptic modulation and learning: new insights into synaptic transmission from the study of behavior. Pp. 471-518 in G. M. Edelman, W. E. Gall, and W. M. Cowan, eds. Synaptic function. Wiley, New York.

KAY, A. 1979. Hawaiian marine shells. Bernice Pauahi Bishop Museum, Honolulu, HI.

KLEIN, M. 1994. Synaptic augmentation by 5HT at rested Aplysia sensorimotor synapses: Independence of action potential prolongation. Neuron 13:159-166.

-----. 1995. Modulation of ion currents and regulation of transmitter release in short-term synaptic plasticity: The rise and fall of the action potential. Invertebr. Neurosci. 1:15-24.

KLEIN, M., B. HOCHNER, AND E. R. KANDEL. 1986. Facilitatory transmitters and cAMP can modulate accommodation as well as transmitter release in Aplysia sensory neurons: Evidence for parallel processing in a single cell. Proc. Nat. Acad. Sci. USA 83: 7994-7998.

KNIGHT, J. B., L. R. Cox, A. M. KEEN, A. G. SMITH, R. L. BATTEN, E. L. YOCHELSON, N. H. LUDBROOK, R. ROBERTSON, C. M. YONGE, AND R. C. MOORE. 1960. Mollusca I. Pp. 12-1351 in R. C. Moore and C. W. Pitrat, eds. Treatise on invertebrate paleontology. Geol. Soc. Am. Univ. of Kansas Press.

LIN, X. Y., AND D. L. GLANZMAN. 1994a. Hebbian induction of long-term potentiation of Aplysia sensorimotor synapses: Partial requirement for activation of an NMDA-related receptor. Proc. R. Soc. Lond. B Biol. Sci. 255:215-221.

-----. 1994b. Long-term potentiation of Aplysia sensorimotor synapses in cell culture: Regulation by postsynaptic voltage. Proc. R. Soc. Lond. B Biol. Sci. 255:113-118.

MACKEY, S. L., E. R. KANDEL, AND R. D. HAWKINS. 1989. Identified serotonergic neurons LCB1 and RCB1 in the cerebral ganglia of Aplysia produce presynaptic facilitation of siphon sensory neurons. J. Neurosci. 9:4227-4235.

MADDISON, W. P., M. J. DONOGHUE, AND D. R. MADDISON. 1984. Outgroup analysis and parsimony. Syst. Zool. 33:83-103.

MARCUS, E. A., AND T. J. CAREW. 1989. Development of neuro-modulation in the pleural sensory neurons of Aplysia. Soc. Neurosci. Abstr. 15:1265.

MARCUS, E. D. B.-R. 1972. On the Anaspidea (Gastropoda: Opisthobranchia) of the warm waters of the western Atlantic. Bull. Mar. Sci. 22:841-874.

MARTINEZ, J. L. J., AND R. P. KESNER, EDS. 1991. Learning and memory. A biological view. Academic Press, New York.

MERCER, A. R., N. J. EMPTAGE, AND T. J. CAREW. 1991. Pharmacological dissociation of modulatory effects of serotonin in Aplysia sensory neurons. Science 254:1811-1813.

MOORE, R., C. LALIKER, AND A. FISHER. 1952. Invertebrate fossils. McGraw-Hill, New York.

MORTON, J. E. 1972. The form and functioning of the pallial organs in the Opisthobranch Akera bullata with a discussion on the nature of the gill in Notaspidea and other Tectibranchs. Veliger 14:337-349.

PATEL, N.H. 1994. Developmental evolution: Insights from studies of insect segmentation. Science 266:581-589.

PENNINGS, S. C. 1990. Multiple factors promoting narrow host range in the sea hare, Aplysia californica. Oecologia 82:192-200.

-----. 1994. Interspecific variation in chemical defenses in the sea hares (Opisthobranchia: Anaspidea). J. Exp. Mar. Biol. Ecol. 180:203-219.

QUINN, W. G. 1984. Work in invertebrates on the mechanisms underlying learning. Pp. xxxx-xxxx in P. Marler and H. S. Terrace, eds. The biology of learning. Springer, Berlin.

RAFF, R. A., AND T. C. KAUFMAN. 1983. Embryos, genes, and evolution. The developmental-genetic basis of evolutionary change. Macmillan, New York.

SQUIRE, L. R., AND N. BUTTERS, EDS. 1992. Neuropsychology of memory. Guilford Press, New York.

SUGITA, S., J. R. GOLDSMITH, D. A. BAXTER, AND J. H. BYRNE. 1992. Involvement of protein kinase C in serotonin-induced spike broadening and synaptic facilitation in sensorimotor connections of Aplysia. J. Neurophysiol. 68:643-651.

SUGITA, S., D. A. BAXTER, AND J. H. BYRNE. 1994a. Activation of protein kinase C mimics serotonin-induced modulation of a voltage-dependent potassium current in pleural sensory neurons of Aplysia. J. Neurophysiol. 72:1240-1249.

-----. 1994b. cAMP independent effects of 8-(4-parachlorophenylthio)-cyclic AMP on spike duration and membrane currents in pleural sensory neurons of Aplysia. J. Neurophysiol. 72: 1250-1259.

THOMPSON, T. E. 1973. Euthyneuran and other molluscan spermatozoa. Malacologia 14:167-206.

WALTERS, E. T., AND J. H. BYRNE. 1983. Associative conditioning of single sensory neurons suggests a cellular mechanism for learning. Science 219:405-408.

WALTERS, E. T., J. H. BYRNE, T. J. CAREW, AND E. R. KANDEL. 1983a. Mechanoafferent neurons innervating the tail of Aplysia. I. Response properties and synaptic connections. J. Neurophysiol. 50:1522-1542.

WALTERS, E. T., J. H. BYRNE, T. J. CAREW, AND E. R. KANDEL. 1983b. Mechanoafferent neurons innervating the tail of Aplysia. II. Modulation by sensitizing stimuli. J. Neurophysiol. 50:1543-1559.

WILLIAMS, G. C. 1975. Phlogenetic implications of the degree of concentration within the opisthobranch nervous system. Master's thesis, San Francisco State Univ., San Francisco, CA.

WRIGHT, W. G., AND B. MAYNARD. 1994. Evolution of learning-related neuromodulation in opisthobranch molluscs. Soc. Neurosci. Abrstr. 20:230.

WRIGHT, W. G., K. JONES, P. SHARP, AND B. MAYNARD. 1995. Widespread anatomical projections of the serotonergic modulatory neuron, CB1, in Aplysia. Invertebr. Neurosci. 1:173-183.

[TABULAR DATA FOR APPENDIX OMITTED]
COPYRIGHT 1996 Society for the Study of Evolution
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wright, William G.; Kirschman, David; Rozen, Danny; Maynard, Barbara
Publication:Evolution
Date:Dec 1, 1996
Words:9326
Previous Article:Suboptimal virulence of an insect-parasitic nematode.
Next Article:The geography of coevolution: comparative population structures for a snail and its trematode parasite.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters