Pavlovian Conditioning in Hermissenda: A Circuit Analysis
http://www.100md.com
《生物学报告》
Department of Neurobiology and Anatomy, University of Texas Medical School, Houston, Texas 77030
Abstract
An understanding of associative learning requires (1) an adequate description of the experimental conditions under which learning is produced, (2) a knowledge of what is learned or the determination of the content of learning, and (3) an explanation of how learning generates changes in behavior (Rescorla, 1980). These basic issues are being addressed at both the behavioral and cellular/molecular levels by the analysis of associative learning in animals with relatively uncomplex nervous systems. Use of Pavlovian conditioning of invertebrates as a model for associative learning has led to the identification of cellular and synaptic mechanisms underlying the formation of basic associations. However, an understanding of the associative processes that form the basis for Pavlovian conditioning requires an explanation not only of the mechanisms of temporal contiguity or predictability between the conditioned stimulus (CS) and the unconditioned stimulus (US), but also of how changes produced in the nervous system by conditioning are expressed in behavior. Studies with invertebrates have provided the opportunity to examine how associative learning is expressed in the neural circuitry that supports the generation of learned behavior.
Abbreviations: CR, conditioned response ? CS, conditioned stimulus ? EPSP, excitatory postsynaptic potential ? IPSP, inhibitory postsynaptic potential ? UR, unconditioned response ? US, unconditioned stimulus
Introduction
The nudibranch mollusc Hermissenda crassicornis, the subject of this review, is one preparation that has contributed to an understanding of Pavlovian conditioning at the cellular, molecular, and systems level. The Hermissenda central nervous system is relatively simple, which makes it possible to study identifiable neurons in the neural circuitry that supports conditioning. Identified neurons in the pathway of a conditioned stimulus (CS) have been studied in detail using biochemical, biophysical, and molecular techniques. The two sensory structures mediating the CS and the US (unconditioned stimulus) are centrally located, and thus their synaptic projections remain intact after surgical isolation of the nervous system or in experiments with semi-intact preparations. Mechanisms of CS-US contiguity have been identified and have been the focus of biophysical, biochemical, and molecular analyses. Recent studies have led to the identification of neurons that contribute to the neural circuitry supporting the unconditioned response (UR) and conditioned response (CR). Since conditioning can be studied in semi-intact preparations, an explanation of how conditioning is expressed in the generation of behavior is now feasible.
Pavlovian Conditioning
Pavlovian conditioning in Hermissenda involves stimulation of visual and graviceptive sensory pathways. The conditioning procedure consists of pairing light, the CS, with rotation or orbital shaking, the US. Rotation of the statocyst has been shown to be an adequate stimulus for evoking depolarizing generator potentials and an increase in spike activity in Hermissenda hair cells (Alkon, 1975). Two URs are elicited by rotation—a reduced rate of forward locomotion and foot-shortening (Alkon, 1974; Crow and Alkon, 1978; Lederhendler et al., 1986; Matzel et al., 1990a). Pavlovian conditioning in Hermissenda results in the acquisition of two different CRs. Conditioning produces both light-elicited inhibition of normal positive phototaxis (Crow and Alkon, 1978, 1980; Crow and Offenbach, 1983; Crow, 1985a) and CS-elicited foot contraction (Lederhendler et al., 1986). Inhibition of phototaxis produced by conditioning is expressed by a light-dependent inhibition in the initiation of locomotion (Crow and Offenbach, 1983) and a reduced rate of forward locomotion in light (Farley and Alkon, 1982; Matzel et al., 1990a). The two CRs are theorized to develop independently (Matzel et al., 1990a), which is consistent with recent cellular studies showing that the URs involve different components of the neural circuitry responsible for foot contraction and ciliary locomotion (Crow and Tian, 2003a, b, 2004). This review focuses on a discussion of how conditioning-dependent modifications of synaptic function and intrinsic cellular excitability in identified components of the neural circuit that supports ciliary locomotion result in the generation of phototactic inhibition.
Conditioning in the two different behavioral response systems that support the two CRs is sensitive to both CS-US contiguity and manipulation of the forward interstimulus-interval (Matzel et al., 1990b). Moreover, both the foot contraction and the inhibition of phototaxis produced by conditioning involve the development or emergence of a new response to the CS rather than the potentiation, through US presentations, of an already existing response to the CS that is referred to as reflex potentiation, or alpha conditioning (e.g., Schreurs, 1989; Sahley and Crow, 1998). In both CRs there is a transfer of functional aspects of the response-evoking properties of the US to the CS (Crow and Alkon, 1978; Lederhendler et al., 1986; Matzel et al., 1990a). This feature probably accounts for the increased complexity of the circuit supporting the CS and US, the multiple sites of CS-US pathway convergence in the nervous system, and the multiple synaptic interactions within the neural network supporting behavior.
Primary Sensory Neurons of the of the Conditioned and Unconditioned Stimuli Pathways
The two sensory structures that are stimulated by the CS and US have been described in detail by Alkon and colleagues (Alkon and Fuortes, 1972; Alkon, 1973a, b; Alkon and Bak, 1973; Detwiler and Alkon, 1973). In addition, the convergence sites providing for synaptic interactions between the CS and US pathways have been identified (Alkon, 1973a,b; Alkon et al., 1978, 1993; Akaike and Alkon, 1980; Crow and Tian, 2000, 2002a, b, 2003a, 2004).
Photoreceptors
Each eye of Hermissenda contains five photoreceptors: three classified as type B and two as type A. The photoreceptors can be further classified according to their location within the eye. There are medial and lateral A and B photoreceptors and one central B photoreceptor. The synaptic connections between the type B photoreceptors and between type B and type A photoreceptors are in the neuropil of the cerebropleural ganglion, and the synaptic interactions between B photoreceptors are mutually inhibitory (Alkon and Fuortes, 1972; Alkon, 1973a; Crow et al., 1979; Senft et al., 1982; Frysztak and Crow, 1993). Light produces a depolarizing generator potential and an increase in spike activity in both type A and B photoreceptors (Dennis, 1967; Alkon and Fuortes, 1972).
Hair cells
The sensory structures stimulated by the US are the two central gravity-detecting statocysts (Alkon and Bak, 1973; Detwiler and Alkon, 1973; Detwiler and Fuortes, 1973; Alkon, 1975). Each statocyst contains 13 hair cells whose cell bodies are located around the perimeter of the statocyst. Hair cells that are opposite one another in the statocyst are mutually inhibitory (Detwiler and Alkon, 1973). Statocyst hair cells contact calcium carbonate particles, called statoconia, by interacting with motile cilia that project into the lumen of the statocyst from the apical region of the somas (Alkon, 1975). Rotation or gravity causes the statoconia to press against the motile cilia of hair cells in front of the centrifugal or gravitational force vector, resulting in a depolarizing generator potential and an increase in spike activity (Alkon, 1975). Hair cells in back of the centrifugal force vector hyperpolarize in response to rotation.
Interneurons in the Pathway of the Unconditioned Response
Statocyst hair cells project to photoreceptors and three identified types of interneurons in the cerebropleural ganglia (Akaike and Alkon, 1980; Tabata and Alkon, 1982; Goh and Alkon, 1984; Crow and Tian, 2004). Hair cells form monosynaptic connections with photoreceptors and with type Ie and Ii interneurons; they also project polysynaptically to type Ib interneurons, which have recently been identified. Type Ie and Ii interneurons project polysynaptically to type IIIi inhibitory interneurons, and type IIIi interneurons inhibit ciliary-activating motor neurons (Crow and Tian, 2003a). In summary, rotation or orbital shaking—the US—depolarizes statocyst hair cells, resulting in excitation of type Ie interneurons; the interneurons excite type IIIi inhibitory interneurons, which then inhibit or decrease the spike activity of ciliary motor neurons. Therefore, an increase in the spike activity of type IIIi interneurons results in inhibition of ciliary locomotion. Activation of the identified components of the circuit supporting the US explains the effect of rotation on ciliary locomotion. However, activation of the circuitry must also provide an explanation for the elicitation of the negative geotactic response expressed in Hermissenda. Recently identified type Ib interneurons were shown to form monosynaptic excitatory connections with contractile motor neurons and ciliary motor neurons (Crow and Tian, 2004). Therefore, hair cell activation of Ib interneurons could contribute to the generation of ciliary activity underlying a geotactic response.
Convergence of the Pathways of the Conditioned and Unconditioned Stimuli
Recently, progress has been made on identifying multiple sites of synaptic interactions between cells in the CS and US pathways (for review, see Crow, 2004). The initial site of convergence between the CS and US is at the primary sensory neurons of the visual and graviceptive pathways. The synaptic projections from statocyst hair cells to the photoreceptors are both monosynaptic and polysynaptic. Hair cells and photoreceptors form reciprocal monosynaptic inhibitory connections (Alkon, 1973b). Caudal hair cells inhibit photoreceptors, and cephalic hair cells are inhibited by type B photoreceptors. The second site of convergence between the CS and US pathways involves aggregates of identified interneurons in the cerebropleural ganglia. The synaptic organization of the interneurons that make up the convergence site between the visual and graviceptive pathways have now been characterized and described in considerable detail (Alkon et al., 1978; Akaike and Alkon, 1980; Crow and Tian, 2000, 2002a, 2003a, 2004). Photoreceptors and hair cells project to aggregates of "on" and "off" cells designated as type Ie and Ii cerebropleural interneurons. An additional site of convergence between the CS and US pathways is the recently identified type Ib interneurons that form monosynaptic connections with contractile motor neurons and ciliary-activating motor neurons (Crow and Tian, 2004).
Circuitry Supporting the Generation of the Conditioned Response
How does illumination modulate the spike activity of ciliary motor neurons? The results of recent work have shown that illumination of photoreceptors decreases the frequency of IPSPs recorded from identified VP1 ciliary motor neurons (Crow and Tian, 2003a). Interestingly, extrinsic current depolarization of type Ie and Ii interneurons during illumination increases the frequency of IPSPs, measured during the period of current stimulation in ciliary motor neurons. These results indicate that light regulates the activity of ciliary motor neurons by hyperpolarizing type Ii interneurons ("off" cells) and depolarizing type Ie interneurons ("on" cells). The integration of synaptic input from type I interneurons results in a net decrease in the spike activity of type IIIi inhibitory interneurons during light. This decrease in spike activity reduces the inhibition of ciliary motor neurons by type IIIi interneurons, resulting in an increase in their spike activity and an increase in ciliary activity on the foot. In an untrained animal, the frequency of IPSPs in ciliary motor neurons is decreased more effectively by the light-induced hyperpolarization of type Ii interneurons than by the concomitant excitation of type IIIi interneurons provided by light-induced excitatory input from depolarized type Ie interneurons. Therefore, the effect of light on the activity of ciliary motor neurons is an increase in spikes and increased ciliary movement.
We have examined how the neural circuitry underlying light-elicited ciliary locomotion contributes to the inhibition of phototaxis produced by conditioning. Previous research has shown that Pavlovian conditioning produces changes in both intrinsic excitability and synaptic function in identified photoreceptors. Cellular correlates of conditioning have now been examined in photoreceptors (for a review, see Crow, 2004), type I interneurons (Crow and Tian, 2002b), and ciliary motor neurons (Crow and Tian, 2003b). As shown in Figure 3A–C, conditioning increases the spike activity that the CS elicits in type Ie interneurons significantly compared to the type Ie spike activity of controls that received independent random presentations of the CS and US (with the restriction that the two stimuli could not overlap in time—i.e., were pseudorandom [Figs. 3B–C]). An examination of the complex EPSP in type Ie interneurons revealed that the CS evoked a larger amplitude depolarization with a greater frequency of smaller EPSPs in conditioned preparations (Figs. 3D–F) than in the unconditioned pseudorandom controls (Figs. 3E–G). The mean peak amplitude of the type Ie complex EPSP evoked by the CS was significantly larger in conditioned animals (Fig. 3H). In addition, the monosynaptic EPSP elicited in type Ie interneurons (Fig. 3J) by a single spike in a lateral B photoreceptor (Fig. 3I) was facilitated in conditioned animals as compared to pseudorandom controls (Fig. 3M). Synaptic changes for type Ii interneurons were similar to those for Ie interneurons in comparisons between conditioned animals and pseudorandom controls. The complex IPSP (Figs. 4A–C) elicited in type Ii interneurons by the CS was of significantly larger amplitude in conditioned animals than in pseudorandom controls (Fig. 4B-C). The greater inhibition the CS evoked in type Ii interneurons produced a significant decrease in the spike activity of these interneurons relative to pseudorandom controls (Fig. 4D). The effect of light on the spike activity of type Ii interneurons would be expected to depress excitation of type IIIi interneurons more strongly in conditioned animals. In addition, the amplitude of monosynaptic IPSPs recorded in type Ii interneurons (Fig. 4F) and elicited by a single spike in a lateral type B photoreceptor (Fig. 4E) was facilitated in conditioned animals relative to controls (Fig. 4I). Taken collectively, the conditioning correlates detected in type I interneurons can be explained by well-documented intrinsic changes in the type B photoreceptors (Crow and Alkon, 1980; West et al., 1982; Alkon et al., 1982, 1985; Farley and Alkon, 1982, 1987; Crow, 1985b, 1988; Crow and Forester, 1991; Frysztak and Crow, 1993, 1994, 1997; Gandhi and Matzel, 2000).
The increase in the amplitude of the complex PSPs in type I interneurons may be due to both the light-evoked enhanced excitability of the photoreceptors produced by conditioning (for review, see Crow, 2004) and the facilitation of the monosynaptic PSPs between lateral B photoreceptors and type I interneurons (Crow and Tian, 2002b). However, postsynaptic changes intrinsic to type I interneurons could also contribute to the modifications of type I interneurons detected in conditioned animals. We initially addressed this issue by investigating cellular excitability in type Ie interneurons from conditioned and pseudorandom controls. We found that the intrinsic excitability of type Ie interneurons was enhanced in conditioned animals compared to pseudorandom controls (Crow and Tian, 2003b). For each current level tested (Fig. 5), the current pulse evoked more spikes in Ie interneurons of conditioned animals than in in pseudorandom controls. In the group summary data (Fig. 5C) this difference is statistically significant. In conditioned animals, the intrinsic enhanced excitatory of type Ie interneurons should increase the frequency of IPSPs produced in type IIIi interneurons by light-evoked depolarization of the type Ie interneurons. In contrast, the light-evoked hyperpolarization of type Ii interneurons observed before conditioning decreases the spike activity of type IIIi interneurons and decreases the frequency of IPSPs in VP1 ciliary motor neurons. We have examined the effects of enhanced Ie excitability on IPSP frequency in VP1 ciliary motor neurons. Extrinsic current stimulation of Ie interneurons resulted in more IPSPs recorded in VP1 ciliary motor neurons in conditioned animals than in pseudorandom controls (Fig. 6). Analysis of the group data (Fig. 6E) revealed that the difference was statistically significant. The increase in the number of IPSPs in VP1 ciliary motor neurons of conditioned animals could be accounted for by the increase in the number of spikes produced in type IIIi interneurons by the increased excitation of type Ie interneurons.
Consistent with these data are the results of an examination of light-elicited spike activity recorded in VP1 ciliary motor neurons of conditioned animals and pseudorandom controls (Fig. 7). The presentation of the CS in conditioned animals resulted in a decrease in the spike activity in VP1 ciliary motor neurons and an inhibition of firing during the light (Fig. 7A). In contrast, pseudorandom controls exhibited an increase in spike activity in VP1 ciliary motor neurons during the light step (Fig. 7B). The group data (Fig. 7C) indicated that the CS produced a statistically significant inhibition of VP1 spike activity compared to pre-CS baseline activity and pseudorandom controls.
In summary, a combination of intrinsic excitability changes and modifications in synaptic function at specific loci in the circuit responsible for light-dependent ciliary locomotion can explain the generation or expression of the CR elicited by the CS. Before conditioning, the light-dependent inhibition of type Ii interneurons is very effective in regulating the IPSP frequency of ciliary motor neurons and their subsequent firing. Indeed, the net effect of light-dependent excitation of type Ie interneurons and light-dependent inhibition of type Ii interneurons is a decrease in the IPSP frequency of ciliary motor neurons, an increase in their spike activity, and increased ciliary movement. The induction of intrinsic enhanced excitability of type Ie interneurons produced by conditioning reconfigures the neural circuit such that light, the CS, increases IPSP frequency in ciliary motor neurons. As a result, the spike activity of the ciliary motor neurons is decreased during illumination and phototaxis is inhibited.
Conclusions and Discussion
Progress in determining how Pavlovian conditioning is expressed in the generation of phototactic behavior in Hermissenda is encouraging, and supported by recent work in identifying the neural circuit that controls ciliary locomotion and how it is affected by light (CS) and graviceptive input (US). The analysis of Pavlovian conditioning in the neural circuit that generates ciliary locomotion shows that both enhanced cellular excitability and synaptic facilitation are expressed in identified circuit components at different loci within the network. The distributed nature of the cellular and synaptic plasticity associated with this example of Pavlovian conditioning suggests that an adequate explanation of conditioned behavior requires both an analysis of neural circuits and the identification of mechanisms of CS-US contiguity at convergence sites between the CS and US pathways. Consistent with the view that learning may initially involve changes in pre-existing synaptic connections, the inhibition of phototactic behavior produced by conditioning involves modifications of existing synaptic connections between photoreceptors and identified type I interneurons. However, the possibility that, with conditioning, new connections form between neurons in the neural circuit that modulates ciliary locomotor behavior cannot be dismissed.
An earlier analysis of visual control of locomotion suggested that the enhanced inhibition of the medial A photoreceptor by the B photoreceptor and the subsequent decrease in spike activity of interneurons and a motor neuron with conditioning may contribute to decreased phototaxis (Goh and Alkon, 1984). However, the type A photoreceptors are typically not active in the dark, so their increased inhibition by B photoreceptors in the light cannot account for the decrease in the spike activity of pedal motor neurons to below pre-light baseline levels (Richards and Farley, 1987; Hodgson and Crow, 1992) or for the decrease, elicited by the CS in conditioned animals, in the spike activity of ciliary motor neurons to below baseline (Crow and Tian, 2003b). In addition, the putative motor neuron examined in the earlier study was proposed to contribute to the turning of animals, not to their ciliary locomotion (Goh and Alkon, 1984). Ciliary motor neurons have only recently been identified in Hermissenda (Crow and Tian, 2003a).
The analysis of conditioning correlates has revealed that the first site of intrinsic cellular and synaptic plasticity is at the initial site of convergence between the CS and US pathways—that is, in the primary sensory neurons of the CS pathway. The mechanisms of temporal contiguity between the CS and US involve enhancements in both cellular excitability and synaptic strength. The changes in photoreceptor excitability produced by conditioning involve reductions in several well-characterized K+ conductances in type B photoreceptors. The second site of enhanced intrinsic excitability is the type Ie interneurons. The membrane conductances underlying enhanced excitability intrinsic to type Ie interneurons have not yet been analyzed. The cellular and synaptic changes identified following conditioning are distributed at several loci within the network and therefore are not localized to a single synaptic site or neuron. The distributed nature of learning-dependent changes may account for the complexity of Pavlovian conditioning in Hermissenda—specifically for the emergence of a new response to the CS following conditioning.
Acknowledgments
This research was supported by National Institutes of Mental Health Grant MH-58698 to T. Crow. We thank Diana Parker for assistance with this manuscript.
Footnotes
Received 26 October 2005; accepted 24 February 2006.
Literature Cited
Akaike, T., and D. L. Alkon. 1980. Sensory convergence on central visual neurons in Hermissenda. J. Neurophysiol. 44:501–513.
Alkon, D. L. 1973a. Neural organization of a molluscan visual system. J. Gen. Physiol.61:444–461.
Alkon, D. L. 1973b. Intersensory interactions in Hermissenda. J. Gen. Physiol. 62:185–202.
Alkon, D. L. 1974. Associative training of Hermissenda. J. Gen. Physiol. 64:70–84.
Alkon, D. L. 1975. Responses of hair cells to statocyst rotation. J. Gen. Physiol.66:507–530.
Alkon, D. L., and A. Bak. 1973. Hair cell generator potentials. J. Gen. Physiol.61:619.
Alkon, D. L., and M. G. Fuortes. 1972. Responses of photoreceptors in Hermissenda. J. Gen. Physiol. 60:631–649.
Alkon, D. L., T. Akaike, and J. Harrigan. 1978. Interaction of chemosensory, visual, and statocyst pathways in Hermissenda crassicornis. J. Gen. Physiol. 71:177–194.
Alkon, D. L., I. Lederhendler, and J. J. Shoukimas. 1982. Primary changes of membrane currents during retention of associative learning. Science215:693–695.
Alkon, D. L., M. Sakakibara, R. Forman, J. Harrigan, I. Lederhendler, and J. Farley. 1985. Reduction of two voltage-dependent K+ currents mediates retention of a learned association. Behav. Neural Biol.44:278–300.
Alkon, D. L., M. J. Anderson, A. J. Kuzirian, D. F. Rogers, D. M. Fass, C. Collin, T. J. Nelson, I. M. Kapetanovic, and L. D. Matzel. 1993. GABA-mediated synaptic interaction between the visual and vestibular pathways of Hermissenda. J. Neurochem. 61:556–566.
Crow, T. 1985a. Conditioned modification of phototactic behavior in Hermissenda. I. Analysis of light intensity. J. Neurosci.5:209–214.
Crow T. 1985b. Conditioned modification of phototactic behavior in Hermissenda. II. Differential adaptation of B-photoreceptors. J. Neurosci.5:215–223.
Crow, T. 1988. Cellular and molecular analysis of associative learning and memory in Hermissenda. Trends Neurosci. 11:136–147.
Crow, T. 2004. Pavlovian conditioning of Hermissenda: current cellular, molecular and circuit perspectives. Learn. Mem.11:229–238.
Crow, T., and D. L. Alkon. 1978. Retention of an associative behavioral change in Hermissenda. Science 201:1239–1241.
Crow, T., and D. L. Alkon. 1980. Associative behavioral modification in Hermissenda: cellular correlates. Science209:412–414.
Crow, T., and J. Forrester. 1991. Light paired with serotonin in vivo produces both short- and long-term enhancement of generator potentials of identified B-photoreceptors in Hermissenda. J. Neurosci. 11:608–617.
Crow, T., and N. Offenbach. 1983. Modification of the initiation of locomotion in Hermissenda: behavioral analysis. Brain Res.271:301–310.
Crow, T., and L.-M. Tian. 2000. Monosynaptic connections between identified A and B photoreceptors and interneurons in Hermissenda: evidence for labeled-lines. J. Neurophysiol.84:367–375.
Crow, T., and L.-M. Tian. 2002a. Morphological characteristics and central projections of two types of interneurons in the visual pathway of Hermissenda. J. Neurophysiol. 87:322–332.
Crow, T., and L.-M. Tian. 2002b. Facilitation of monosynaptic and complex PSPs in type I interneurons of conditioned Hermissenda. J. Neurosci. 22:7818–7824.
Crow, T., and L.-M. Tian. 2003a. Interneuronal projections to identified cilia-activating pedal neurons in Hermissenda. J. Neurophysiol. 89:2420–2429.
Crow, T., and L.-M. Tian. 2003b. Neural correlates of Pavlovian conditioning in components of the neural network supporting ciliary locomotion in Hermissenda. Learn. Mem. 10:209–216.
Crow, T., and L.-M. Tian. 2004. Statocyst hair cell activation of identified interneurons and foot contraction motor neurons in Hermissenda. J. Neurophysiol. 92:2874–2883.
Crow, T., E. Heldman, V. Hacopian, R. Enos, and D. L. Alkon. 1979. Ultrastructure of photoreceptors in the eye of Hermissenda labelled with intracellular injections of horseradish peroxidase. J. Neurocytol.8:181–195.
Dennis, M. J. 1967. Electrophysiology of the visual system in a nudibranch mollusc. J. Neurophysiol.30:1439–1465.
Detwiler, P. B., and D. L. Alkon. 1973. Hair cell interactions in the statocyst of Hermissenda. J. Gen. Physiol. 62:618–642.
Detwiler, P. B., and M. G. Fuortes. 1973. Responses of hair cells in the statocyst of Hermissenda. J. Physiol. 251:107–129.
Farley, J., and D. L. Alkon. 1982. Associative neural and behavioral change in Hermissenda: consequences of nervous system orientation for light- and pairing-specificity. J. Neurophysiol.48:785–807.
Farley, J., and D. L. Alkon. 1987. In vitro associative conditioning of Hermissenda: cumulative depolarization of type B photoreceptors and short-term associative behavioral changes. J. Neurophysiol.57:1639–1668.
Frysztak, R. J., and T. Crow. 1993. Differential expression of correlates of classical conditioning in identified medial and lateral type A photoreceptors of Hermissenda. J. Neurosci. 13:2889–2897.
Frysztak, R. J., and T. Crow. 1994. Enhancement of type B and A photoreceptor inhibitory synaptic connections in conditioned Hermissenda. J. Neurosci. 14:1245–1250.
Frysztak, R. J., and T. Crow. 1997. Synaptic enhancement and enhanced excitability in presynaptic and postsynaptic neurons in the conditioned stimulus pathway of Hermissenda. J. Neurosci. 17:4426–4433.
Gandhi, C. C., and L. D. Matzel. 2000. Modulation of presynaptic action potential kinetics underlies synaptic facilitation of type B photoreceptors after associative conditioning in Hermissenda. J. Neurosci. 20:2022–2035.
Goh, Y., and D. L. Alkon. 1984. Sensory, interneuronal, and motor interactions within Hermissenda visual pathway. J. Neurophysiol.52:156–169.
Hodgson, T. M., and T. Crow. 1992. Cellular correlates of classical conditioning in identified light responsive pedal neurons of Hermissenda crassicornis. Brain Res. 570:267–271.
Lederhendler, I. I., S. Gart, and D. L. Alkon. 1986. Classical conditioning of Hermissenda: origin of a new response. J. Neurosci.6:1325–1331.
Matzel, L.D., B. G. Schreurs, and D. L. Alkon. 1990a. Pavlovian conditioning of distinct components of Hermissenda’s responses to rotation. Behav. Neural Biol.54:131–145.
Matzel, L. D., B. G. Schreurs, I. Lederhendler, and D. L. Alkon. 1990b. Acquisition of conditioned associations in Hermissenda: additive effects of contiguity and the forward interstimulus interval. Behav. Neurosci.104:597–606.
Rescorla, R. A. 1980. Pavlovian Second-Order Conditioning: Studies in Associative Learning. Lawrence Erlbaum, Hillsdale, NJ.
Richards, W. G., and J. Farley. 1987. Motor correlates of phototaxis and associative learning. Brain Res. Bull.19:174–189.
Sahley, C. L., and T. Crow. 1998. Invertebrate learning: current perspectives. Pp. 197–209 in Learning and Memory, J. L. Martinez, Jr. and R. P. Kesner, eds. Academic Press, New York.
Schreurs, B. G. 1989. Classical conditioning of model systems: a behavioral review. Psychobiology17:145–155.
Senft, S. L., R. D. Allen, T. Crow, and D. L. Alkon. 1982. Optical sectioning of HRP-stained molluscan neurons. J. Neurosci. Meth.5:153–159.
Tabata, M., and D. L. Alkon. 1982. Positive synaptic feedback in visual system of nudibranch mollusk Hermissenda crassicornis. J. Neurophysiol. 48:174–191.
West, A., E. Barnes, and D. L. Alkon. 1982. Primary changes of voltage responses during retention of associative learning. J. Neurophysiol.48:1243–1255.(Terry Crow and Lian-Ming )
Abstract
An understanding of associative learning requires (1) an adequate description of the experimental conditions under which learning is produced, (2) a knowledge of what is learned or the determination of the content of learning, and (3) an explanation of how learning generates changes in behavior (Rescorla, 1980). These basic issues are being addressed at both the behavioral and cellular/molecular levels by the analysis of associative learning in animals with relatively uncomplex nervous systems. Use of Pavlovian conditioning of invertebrates as a model for associative learning has led to the identification of cellular and synaptic mechanisms underlying the formation of basic associations. However, an understanding of the associative processes that form the basis for Pavlovian conditioning requires an explanation not only of the mechanisms of temporal contiguity or predictability between the conditioned stimulus (CS) and the unconditioned stimulus (US), but also of how changes produced in the nervous system by conditioning are expressed in behavior. Studies with invertebrates have provided the opportunity to examine how associative learning is expressed in the neural circuitry that supports the generation of learned behavior.
Abbreviations: CR, conditioned response ? CS, conditioned stimulus ? EPSP, excitatory postsynaptic potential ? IPSP, inhibitory postsynaptic potential ? UR, unconditioned response ? US, unconditioned stimulus
Introduction
The nudibranch mollusc Hermissenda crassicornis, the subject of this review, is one preparation that has contributed to an understanding of Pavlovian conditioning at the cellular, molecular, and systems level. The Hermissenda central nervous system is relatively simple, which makes it possible to study identifiable neurons in the neural circuitry that supports conditioning. Identified neurons in the pathway of a conditioned stimulus (CS) have been studied in detail using biochemical, biophysical, and molecular techniques. The two sensory structures mediating the CS and the US (unconditioned stimulus) are centrally located, and thus their synaptic projections remain intact after surgical isolation of the nervous system or in experiments with semi-intact preparations. Mechanisms of CS-US contiguity have been identified and have been the focus of biophysical, biochemical, and molecular analyses. Recent studies have led to the identification of neurons that contribute to the neural circuitry supporting the unconditioned response (UR) and conditioned response (CR). Since conditioning can be studied in semi-intact preparations, an explanation of how conditioning is expressed in the generation of behavior is now feasible.
Pavlovian Conditioning
Pavlovian conditioning in Hermissenda involves stimulation of visual and graviceptive sensory pathways. The conditioning procedure consists of pairing light, the CS, with rotation or orbital shaking, the US. Rotation of the statocyst has been shown to be an adequate stimulus for evoking depolarizing generator potentials and an increase in spike activity in Hermissenda hair cells (Alkon, 1975). Two URs are elicited by rotation—a reduced rate of forward locomotion and foot-shortening (Alkon, 1974; Crow and Alkon, 1978; Lederhendler et al., 1986; Matzel et al., 1990a). Pavlovian conditioning in Hermissenda results in the acquisition of two different CRs. Conditioning produces both light-elicited inhibition of normal positive phototaxis (Crow and Alkon, 1978, 1980; Crow and Offenbach, 1983; Crow, 1985a) and CS-elicited foot contraction (Lederhendler et al., 1986). Inhibition of phototaxis produced by conditioning is expressed by a light-dependent inhibition in the initiation of locomotion (Crow and Offenbach, 1983) and a reduced rate of forward locomotion in light (Farley and Alkon, 1982; Matzel et al., 1990a). The two CRs are theorized to develop independently (Matzel et al., 1990a), which is consistent with recent cellular studies showing that the URs involve different components of the neural circuitry responsible for foot contraction and ciliary locomotion (Crow and Tian, 2003a, b, 2004). This review focuses on a discussion of how conditioning-dependent modifications of synaptic function and intrinsic cellular excitability in identified components of the neural circuit that supports ciliary locomotion result in the generation of phototactic inhibition.
Conditioning in the two different behavioral response systems that support the two CRs is sensitive to both CS-US contiguity and manipulation of the forward interstimulus-interval (Matzel et al., 1990b). Moreover, both the foot contraction and the inhibition of phototaxis produced by conditioning involve the development or emergence of a new response to the CS rather than the potentiation, through US presentations, of an already existing response to the CS that is referred to as reflex potentiation, or alpha conditioning (e.g., Schreurs, 1989; Sahley and Crow, 1998). In both CRs there is a transfer of functional aspects of the response-evoking properties of the US to the CS (Crow and Alkon, 1978; Lederhendler et al., 1986; Matzel et al., 1990a). This feature probably accounts for the increased complexity of the circuit supporting the CS and US, the multiple sites of CS-US pathway convergence in the nervous system, and the multiple synaptic interactions within the neural network supporting behavior.
Primary Sensory Neurons of the of the Conditioned and Unconditioned Stimuli Pathways
The two sensory structures that are stimulated by the CS and US have been described in detail by Alkon and colleagues (Alkon and Fuortes, 1972; Alkon, 1973a, b; Alkon and Bak, 1973; Detwiler and Alkon, 1973). In addition, the convergence sites providing for synaptic interactions between the CS and US pathways have been identified (Alkon, 1973a,b; Alkon et al., 1978, 1993; Akaike and Alkon, 1980; Crow and Tian, 2000, 2002a, b, 2003a, 2004).
Photoreceptors
Each eye of Hermissenda contains five photoreceptors: three classified as type B and two as type A. The photoreceptors can be further classified according to their location within the eye. There are medial and lateral A and B photoreceptors and one central B photoreceptor. The synaptic connections between the type B photoreceptors and between type B and type A photoreceptors are in the neuropil of the cerebropleural ganglion, and the synaptic interactions between B photoreceptors are mutually inhibitory (Alkon and Fuortes, 1972; Alkon, 1973a; Crow et al., 1979; Senft et al., 1982; Frysztak and Crow, 1993). Light produces a depolarizing generator potential and an increase in spike activity in both type A and B photoreceptors (Dennis, 1967; Alkon and Fuortes, 1972).
Hair cells
The sensory structures stimulated by the US are the two central gravity-detecting statocysts (Alkon and Bak, 1973; Detwiler and Alkon, 1973; Detwiler and Fuortes, 1973; Alkon, 1975). Each statocyst contains 13 hair cells whose cell bodies are located around the perimeter of the statocyst. Hair cells that are opposite one another in the statocyst are mutually inhibitory (Detwiler and Alkon, 1973). Statocyst hair cells contact calcium carbonate particles, called statoconia, by interacting with motile cilia that project into the lumen of the statocyst from the apical region of the somas (Alkon, 1975). Rotation or gravity causes the statoconia to press against the motile cilia of hair cells in front of the centrifugal or gravitational force vector, resulting in a depolarizing generator potential and an increase in spike activity (Alkon, 1975). Hair cells in back of the centrifugal force vector hyperpolarize in response to rotation.
Interneurons in the Pathway of the Unconditioned Response
Statocyst hair cells project to photoreceptors and three identified types of interneurons in the cerebropleural ganglia (Akaike and Alkon, 1980; Tabata and Alkon, 1982; Goh and Alkon, 1984; Crow and Tian, 2004). Hair cells form monosynaptic connections with photoreceptors and with type Ie and Ii interneurons; they also project polysynaptically to type Ib interneurons, which have recently been identified. Type Ie and Ii interneurons project polysynaptically to type IIIi inhibitory interneurons, and type IIIi interneurons inhibit ciliary-activating motor neurons (Crow and Tian, 2003a). In summary, rotation or orbital shaking—the US—depolarizes statocyst hair cells, resulting in excitation of type Ie interneurons; the interneurons excite type IIIi inhibitory interneurons, which then inhibit or decrease the spike activity of ciliary motor neurons. Therefore, an increase in the spike activity of type IIIi interneurons results in inhibition of ciliary locomotion. Activation of the identified components of the circuit supporting the US explains the effect of rotation on ciliary locomotion. However, activation of the circuitry must also provide an explanation for the elicitation of the negative geotactic response expressed in Hermissenda. Recently identified type Ib interneurons were shown to form monosynaptic excitatory connections with contractile motor neurons and ciliary motor neurons (Crow and Tian, 2004). Therefore, hair cell activation of Ib interneurons could contribute to the generation of ciliary activity underlying a geotactic response.
Convergence of the Pathways of the Conditioned and Unconditioned Stimuli
Recently, progress has been made on identifying multiple sites of synaptic interactions between cells in the CS and US pathways (for review, see Crow, 2004). The initial site of convergence between the CS and US is at the primary sensory neurons of the visual and graviceptive pathways. The synaptic projections from statocyst hair cells to the photoreceptors are both monosynaptic and polysynaptic. Hair cells and photoreceptors form reciprocal monosynaptic inhibitory connections (Alkon, 1973b). Caudal hair cells inhibit photoreceptors, and cephalic hair cells are inhibited by type B photoreceptors. The second site of convergence between the CS and US pathways involves aggregates of identified interneurons in the cerebropleural ganglia. The synaptic organization of the interneurons that make up the convergence site between the visual and graviceptive pathways have now been characterized and described in considerable detail (Alkon et al., 1978; Akaike and Alkon, 1980; Crow and Tian, 2000, 2002a, 2003a, 2004). Photoreceptors and hair cells project to aggregates of "on" and "off" cells designated as type Ie and Ii cerebropleural interneurons. An additional site of convergence between the CS and US pathways is the recently identified type Ib interneurons that form monosynaptic connections with contractile motor neurons and ciliary-activating motor neurons (Crow and Tian, 2004).
Circuitry Supporting the Generation of the Conditioned Response
How does illumination modulate the spike activity of ciliary motor neurons? The results of recent work have shown that illumination of photoreceptors decreases the frequency of IPSPs recorded from identified VP1 ciliary motor neurons (Crow and Tian, 2003a). Interestingly, extrinsic current depolarization of type Ie and Ii interneurons during illumination increases the frequency of IPSPs, measured during the period of current stimulation in ciliary motor neurons. These results indicate that light regulates the activity of ciliary motor neurons by hyperpolarizing type Ii interneurons ("off" cells) and depolarizing type Ie interneurons ("on" cells). The integration of synaptic input from type I interneurons results in a net decrease in the spike activity of type IIIi inhibitory interneurons during light. This decrease in spike activity reduces the inhibition of ciliary motor neurons by type IIIi interneurons, resulting in an increase in their spike activity and an increase in ciliary activity on the foot. In an untrained animal, the frequency of IPSPs in ciliary motor neurons is decreased more effectively by the light-induced hyperpolarization of type Ii interneurons than by the concomitant excitation of type IIIi interneurons provided by light-induced excitatory input from depolarized type Ie interneurons. Therefore, the effect of light on the activity of ciliary motor neurons is an increase in spikes and increased ciliary movement.
We have examined how the neural circuitry underlying light-elicited ciliary locomotion contributes to the inhibition of phototaxis produced by conditioning. Previous research has shown that Pavlovian conditioning produces changes in both intrinsic excitability and synaptic function in identified photoreceptors. Cellular correlates of conditioning have now been examined in photoreceptors (for a review, see Crow, 2004), type I interneurons (Crow and Tian, 2002b), and ciliary motor neurons (Crow and Tian, 2003b). As shown in Figure 3A–C, conditioning increases the spike activity that the CS elicits in type Ie interneurons significantly compared to the type Ie spike activity of controls that received independent random presentations of the CS and US (with the restriction that the two stimuli could not overlap in time—i.e., were pseudorandom [Figs. 3B–C]). An examination of the complex EPSP in type Ie interneurons revealed that the CS evoked a larger amplitude depolarization with a greater frequency of smaller EPSPs in conditioned preparations (Figs. 3D–F) than in the unconditioned pseudorandom controls (Figs. 3E–G). The mean peak amplitude of the type Ie complex EPSP evoked by the CS was significantly larger in conditioned animals (Fig. 3H). In addition, the monosynaptic EPSP elicited in type Ie interneurons (Fig. 3J) by a single spike in a lateral B photoreceptor (Fig. 3I) was facilitated in conditioned animals as compared to pseudorandom controls (Fig. 3M). Synaptic changes for type Ii interneurons were similar to those for Ie interneurons in comparisons between conditioned animals and pseudorandom controls. The complex IPSP (Figs. 4A–C) elicited in type Ii interneurons by the CS was of significantly larger amplitude in conditioned animals than in pseudorandom controls (Fig. 4B-C). The greater inhibition the CS evoked in type Ii interneurons produced a significant decrease in the spike activity of these interneurons relative to pseudorandom controls (Fig. 4D). The effect of light on the spike activity of type Ii interneurons would be expected to depress excitation of type IIIi interneurons more strongly in conditioned animals. In addition, the amplitude of monosynaptic IPSPs recorded in type Ii interneurons (Fig. 4F) and elicited by a single spike in a lateral type B photoreceptor (Fig. 4E) was facilitated in conditioned animals relative to controls (Fig. 4I). Taken collectively, the conditioning correlates detected in type I interneurons can be explained by well-documented intrinsic changes in the type B photoreceptors (Crow and Alkon, 1980; West et al., 1982; Alkon et al., 1982, 1985; Farley and Alkon, 1982, 1987; Crow, 1985b, 1988; Crow and Forester, 1991; Frysztak and Crow, 1993, 1994, 1997; Gandhi and Matzel, 2000).
The increase in the amplitude of the complex PSPs in type I interneurons may be due to both the light-evoked enhanced excitability of the photoreceptors produced by conditioning (for review, see Crow, 2004) and the facilitation of the monosynaptic PSPs between lateral B photoreceptors and type I interneurons (Crow and Tian, 2002b). However, postsynaptic changes intrinsic to type I interneurons could also contribute to the modifications of type I interneurons detected in conditioned animals. We initially addressed this issue by investigating cellular excitability in type Ie interneurons from conditioned and pseudorandom controls. We found that the intrinsic excitability of type Ie interneurons was enhanced in conditioned animals compared to pseudorandom controls (Crow and Tian, 2003b). For each current level tested (Fig. 5), the current pulse evoked more spikes in Ie interneurons of conditioned animals than in in pseudorandom controls. In the group summary data (Fig. 5C) this difference is statistically significant. In conditioned animals, the intrinsic enhanced excitatory of type Ie interneurons should increase the frequency of IPSPs produced in type IIIi interneurons by light-evoked depolarization of the type Ie interneurons. In contrast, the light-evoked hyperpolarization of type Ii interneurons observed before conditioning decreases the spike activity of type IIIi interneurons and decreases the frequency of IPSPs in VP1 ciliary motor neurons. We have examined the effects of enhanced Ie excitability on IPSP frequency in VP1 ciliary motor neurons. Extrinsic current stimulation of Ie interneurons resulted in more IPSPs recorded in VP1 ciliary motor neurons in conditioned animals than in pseudorandom controls (Fig. 6). Analysis of the group data (Fig. 6E) revealed that the difference was statistically significant. The increase in the number of IPSPs in VP1 ciliary motor neurons of conditioned animals could be accounted for by the increase in the number of spikes produced in type IIIi interneurons by the increased excitation of type Ie interneurons.
Consistent with these data are the results of an examination of light-elicited spike activity recorded in VP1 ciliary motor neurons of conditioned animals and pseudorandom controls (Fig. 7). The presentation of the CS in conditioned animals resulted in a decrease in the spike activity in VP1 ciliary motor neurons and an inhibition of firing during the light (Fig. 7A). In contrast, pseudorandom controls exhibited an increase in spike activity in VP1 ciliary motor neurons during the light step (Fig. 7B). The group data (Fig. 7C) indicated that the CS produced a statistically significant inhibition of VP1 spike activity compared to pre-CS baseline activity and pseudorandom controls.
In summary, a combination of intrinsic excitability changes and modifications in synaptic function at specific loci in the circuit responsible for light-dependent ciliary locomotion can explain the generation or expression of the CR elicited by the CS. Before conditioning, the light-dependent inhibition of type Ii interneurons is very effective in regulating the IPSP frequency of ciliary motor neurons and their subsequent firing. Indeed, the net effect of light-dependent excitation of type Ie interneurons and light-dependent inhibition of type Ii interneurons is a decrease in the IPSP frequency of ciliary motor neurons, an increase in their spike activity, and increased ciliary movement. The induction of intrinsic enhanced excitability of type Ie interneurons produced by conditioning reconfigures the neural circuit such that light, the CS, increases IPSP frequency in ciliary motor neurons. As a result, the spike activity of the ciliary motor neurons is decreased during illumination and phototaxis is inhibited.
Conclusions and Discussion
Progress in determining how Pavlovian conditioning is expressed in the generation of phototactic behavior in Hermissenda is encouraging, and supported by recent work in identifying the neural circuit that controls ciliary locomotion and how it is affected by light (CS) and graviceptive input (US). The analysis of Pavlovian conditioning in the neural circuit that generates ciliary locomotion shows that both enhanced cellular excitability and synaptic facilitation are expressed in identified circuit components at different loci within the network. The distributed nature of the cellular and synaptic plasticity associated with this example of Pavlovian conditioning suggests that an adequate explanation of conditioned behavior requires both an analysis of neural circuits and the identification of mechanisms of CS-US contiguity at convergence sites between the CS and US pathways. Consistent with the view that learning may initially involve changes in pre-existing synaptic connections, the inhibition of phototactic behavior produced by conditioning involves modifications of existing synaptic connections between photoreceptors and identified type I interneurons. However, the possibility that, with conditioning, new connections form between neurons in the neural circuit that modulates ciliary locomotor behavior cannot be dismissed.
An earlier analysis of visual control of locomotion suggested that the enhanced inhibition of the medial A photoreceptor by the B photoreceptor and the subsequent decrease in spike activity of interneurons and a motor neuron with conditioning may contribute to decreased phototaxis (Goh and Alkon, 1984). However, the type A photoreceptors are typically not active in the dark, so their increased inhibition by B photoreceptors in the light cannot account for the decrease in the spike activity of pedal motor neurons to below pre-light baseline levels (Richards and Farley, 1987; Hodgson and Crow, 1992) or for the decrease, elicited by the CS in conditioned animals, in the spike activity of ciliary motor neurons to below baseline (Crow and Tian, 2003b). In addition, the putative motor neuron examined in the earlier study was proposed to contribute to the turning of animals, not to their ciliary locomotion (Goh and Alkon, 1984). Ciliary motor neurons have only recently been identified in Hermissenda (Crow and Tian, 2003a).
The analysis of conditioning correlates has revealed that the first site of intrinsic cellular and synaptic plasticity is at the initial site of convergence between the CS and US pathways—that is, in the primary sensory neurons of the CS pathway. The mechanisms of temporal contiguity between the CS and US involve enhancements in both cellular excitability and synaptic strength. The changes in photoreceptor excitability produced by conditioning involve reductions in several well-characterized K+ conductances in type B photoreceptors. The second site of enhanced intrinsic excitability is the type Ie interneurons. The membrane conductances underlying enhanced excitability intrinsic to type Ie interneurons have not yet been analyzed. The cellular and synaptic changes identified following conditioning are distributed at several loci within the network and therefore are not localized to a single synaptic site or neuron. The distributed nature of learning-dependent changes may account for the complexity of Pavlovian conditioning in Hermissenda—specifically for the emergence of a new response to the CS following conditioning.
Acknowledgments
This research was supported by National Institutes of Mental Health Grant MH-58698 to T. Crow. We thank Diana Parker for assistance with this manuscript.
Footnotes
Received 26 October 2005; accepted 24 February 2006.
Literature Cited
Akaike, T., and D. L. Alkon. 1980. Sensory convergence on central visual neurons in Hermissenda. J. Neurophysiol. 44:501–513.
Alkon, D. L. 1973a. Neural organization of a molluscan visual system. J. Gen. Physiol.61:444–461.
Alkon, D. L. 1973b. Intersensory interactions in Hermissenda. J. Gen. Physiol. 62:185–202.
Alkon, D. L. 1974. Associative training of Hermissenda. J. Gen. Physiol. 64:70–84.
Alkon, D. L. 1975. Responses of hair cells to statocyst rotation. J. Gen. Physiol.66:507–530.
Alkon, D. L., and A. Bak. 1973. Hair cell generator potentials. J. Gen. Physiol.61:619.
Alkon, D. L., and M. G. Fuortes. 1972. Responses of photoreceptors in Hermissenda. J. Gen. Physiol. 60:631–649.
Alkon, D. L., T. Akaike, and J. Harrigan. 1978. Interaction of chemosensory, visual, and statocyst pathways in Hermissenda crassicornis. J. Gen. Physiol. 71:177–194.
Alkon, D. L., I. Lederhendler, and J. J. Shoukimas. 1982. Primary changes of membrane currents during retention of associative learning. Science215:693–695.
Alkon, D. L., M. Sakakibara, R. Forman, J. Harrigan, I. Lederhendler, and J. Farley. 1985. Reduction of two voltage-dependent K+ currents mediates retention of a learned association. Behav. Neural Biol.44:278–300.
Alkon, D. L., M. J. Anderson, A. J. Kuzirian, D. F. Rogers, D. M. Fass, C. Collin, T. J. Nelson, I. M. Kapetanovic, and L. D. Matzel. 1993. GABA-mediated synaptic interaction between the visual and vestibular pathways of Hermissenda. J. Neurochem. 61:556–566.
Crow, T. 1985a. Conditioned modification of phototactic behavior in Hermissenda. I. Analysis of light intensity. J. Neurosci.5:209–214.
Crow T. 1985b. Conditioned modification of phototactic behavior in Hermissenda. II. Differential adaptation of B-photoreceptors. J. Neurosci.5:215–223.
Crow, T. 1988. Cellular and molecular analysis of associative learning and memory in Hermissenda. Trends Neurosci. 11:136–147.
Crow, T. 2004. Pavlovian conditioning of Hermissenda: current cellular, molecular and circuit perspectives. Learn. Mem.11:229–238.
Crow, T., and D. L. Alkon. 1978. Retention of an associative behavioral change in Hermissenda. Science 201:1239–1241.
Crow, T., and D. L. Alkon. 1980. Associative behavioral modification in Hermissenda: cellular correlates. Science209:412–414.
Crow, T., and J. Forrester. 1991. Light paired with serotonin in vivo produces both short- and long-term enhancement of generator potentials of identified B-photoreceptors in Hermissenda. J. Neurosci. 11:608–617.
Crow, T., and N. Offenbach. 1983. Modification of the initiation of locomotion in Hermissenda: behavioral analysis. Brain Res.271:301–310.
Crow, T., and L.-M. Tian. 2000. Monosynaptic connections between identified A and B photoreceptors and interneurons in Hermissenda: evidence for labeled-lines. J. Neurophysiol.84:367–375.
Crow, T., and L.-M. Tian. 2002a. Morphological characteristics and central projections of two types of interneurons in the visual pathway of Hermissenda. J. Neurophysiol. 87:322–332.
Crow, T., and L.-M. Tian. 2002b. Facilitation of monosynaptic and complex PSPs in type I interneurons of conditioned Hermissenda. J. Neurosci. 22:7818–7824.
Crow, T., and L.-M. Tian. 2003a. Interneuronal projections to identified cilia-activating pedal neurons in Hermissenda. J. Neurophysiol. 89:2420–2429.
Crow, T., and L.-M. Tian. 2003b. Neural correlates of Pavlovian conditioning in components of the neural network supporting ciliary locomotion in Hermissenda. Learn. Mem. 10:209–216.
Crow, T., and L.-M. Tian. 2004. Statocyst hair cell activation of identified interneurons and foot contraction motor neurons in Hermissenda. J. Neurophysiol. 92:2874–2883.
Crow, T., E. Heldman, V. Hacopian, R. Enos, and D. L. Alkon. 1979. Ultrastructure of photoreceptors in the eye of Hermissenda labelled with intracellular injections of horseradish peroxidase. J. Neurocytol.8:181–195.
Dennis, M. J. 1967. Electrophysiology of the visual system in a nudibranch mollusc. J. Neurophysiol.30:1439–1465.
Detwiler, P. B., and D. L. Alkon. 1973. Hair cell interactions in the statocyst of Hermissenda. J. Gen. Physiol. 62:618–642.
Detwiler, P. B., and M. G. Fuortes. 1973. Responses of hair cells in the statocyst of Hermissenda. J. Physiol. 251:107–129.
Farley, J., and D. L. Alkon. 1982. Associative neural and behavioral change in Hermissenda: consequences of nervous system orientation for light- and pairing-specificity. J. Neurophysiol.48:785–807.
Farley, J., and D. L. Alkon. 1987. In vitro associative conditioning of Hermissenda: cumulative depolarization of type B photoreceptors and short-term associative behavioral changes. J. Neurophysiol.57:1639–1668.
Frysztak, R. J., and T. Crow. 1993. Differential expression of correlates of classical conditioning in identified medial and lateral type A photoreceptors of Hermissenda. J. Neurosci. 13:2889–2897.
Frysztak, R. J., and T. Crow. 1994. Enhancement of type B and A photoreceptor inhibitory synaptic connections in conditioned Hermissenda. J. Neurosci. 14:1245–1250.
Frysztak, R. J., and T. Crow. 1997. Synaptic enhancement and enhanced excitability in presynaptic and postsynaptic neurons in the conditioned stimulus pathway of Hermissenda. J. Neurosci. 17:4426–4433.
Gandhi, C. C., and L. D. Matzel. 2000. Modulation of presynaptic action potential kinetics underlies synaptic facilitation of type B photoreceptors after associative conditioning in Hermissenda. J. Neurosci. 20:2022–2035.
Goh, Y., and D. L. Alkon. 1984. Sensory, interneuronal, and motor interactions within Hermissenda visual pathway. J. Neurophysiol.52:156–169.
Hodgson, T. M., and T. Crow. 1992. Cellular correlates of classical conditioning in identified light responsive pedal neurons of Hermissenda crassicornis. Brain Res. 570:267–271.
Lederhendler, I. I., S. Gart, and D. L. Alkon. 1986. Classical conditioning of Hermissenda: origin of a new response. J. Neurosci.6:1325–1331.
Matzel, L.D., B. G. Schreurs, and D. L. Alkon. 1990a. Pavlovian conditioning of distinct components of Hermissenda’s responses to rotation. Behav. Neural Biol.54:131–145.
Matzel, L. D., B. G. Schreurs, I. Lederhendler, and D. L. Alkon. 1990b. Acquisition of conditioned associations in Hermissenda: additive effects of contiguity and the forward interstimulus interval. Behav. Neurosci.104:597–606.
Rescorla, R. A. 1980. Pavlovian Second-Order Conditioning: Studies in Associative Learning. Lawrence Erlbaum, Hillsdale, NJ.
Richards, W. G., and J. Farley. 1987. Motor correlates of phototaxis and associative learning. Brain Res. Bull.19:174–189.
Sahley, C. L., and T. Crow. 1998. Invertebrate learning: current perspectives. Pp. 197–209 in Learning and Memory, J. L. Martinez, Jr. and R. P. Kesner, eds. Academic Press, New York.
Schreurs, B. G. 1989. Classical conditioning of model systems: a behavioral review. Psychobiology17:145–155.
Senft, S. L., R. D. Allen, T. Crow, and D. L. Alkon. 1982. Optical sectioning of HRP-stained molluscan neurons. J. Neurosci. Meth.5:153–159.
Tabata, M., and D. L. Alkon. 1982. Positive synaptic feedback in visual system of nudibranch mollusk Hermissenda crassicornis. J. Neurophysiol. 48:174–191.
West, A., E. Barnes, and D. L. Alkon. 1982. Primary changes of voltage responses during retention of associative learning. J. Neurophysiol.48:1243–1255.(Terry Crow and Lian-Ming )