The role of the hippocampus in trace conditioning: temporal discontinuity or task difficulty?

It is well established that the hippocampal formation is critically involved in the acquisition of trace memories, a paradigm in which the conditioned (CS) and unconditioned stimuli (US) are separated by a temporal gap (Solomon et al., 1986). The structure is reportedly not critical for the acquisition of delay memories, where the CS and the US overlap in time (Berger & Orr, 1983; Schmaltz & Theios, 1972). Based on these results, it is often stated that the hippocampus is involved in "filling the gap" or otherwise associating the two stimuli in time. However, in addition to the presence of a temporal gap, there are other differences between trace and delay conditioning. The most apparent difference is that animals require many more trials to learn the trace task, and thus it is inherently more difficult than the delay task. Here, we tested whether the hippocampus was critically involved in delay conditioning, if it was rendered more difficult such that the rate of acquisition was shifted to be analogous to trace conditioning. Groups of rats received excitotoxic lesions to the hippocampus, sham lesions or were left intact. Using the same interstimulus intervals (ISI), control animals required more trials to acquire the trace than the delay task. As predicted, animals with hippocampal lesions were impaired during trace conditioning but not delay conditioning. However, when the delay task was rendered more difficult by extending the ISI (a long delay task), animals with hippocampal lesions were impaired. In addition, once the lesioned animal learned the association between the CS and the US during delay conditioning, it could learn and perform the trace CR. Thus, the role of the hippocampus in classical conditioning is not limited to learning about discontiguous events in time and space; rather the structure can become engaged simply as a function of task difficulty.

Neurophysiological substrates of context conditioning in Hermissenda suggest a temporally invariant form of activity-dependent neuronal facilitation.

The neurophysiological basis for context conditioning is conceptually problematic because neurophysiological descriptions of activity-dependent (associative) forms of neuronal plasticity uniformly assume that a specific temporal relationship between signals is necessary for memory induction. In the present experiments, this problem is addressed empirically by presenting, as a temporally diffuse contextual signal, a stimulus that results in known neural modifications following punctate (temporally contiguous) pairings with an aversive unconditioned stimulus. Hermissenda were trained to discriminate between adjoining contexts that were distinguished only in that one was lit and one was dark. Thirty unsignaled rotations were presented during each of three 15-min sessions in one of the two (lit or dark) contexts. Prior to training, animals displayed a slight preference for the lit context. After exposure to unsignaled rotation, animal's preferences shifted strongly to the dark context if unsignaled rotations were presented in the light, and tended (nonsignificantly) to the lit context if unsignaled rotations were presented in the dark. The B photoreceptors of the Hermissenda eye undergo several forms of activity-dependent facilitation (e.g., an increase in neuronal input resistance and evoked spike frequency) following pairings of punctate light (CS) and presynaptic vestibular stimulation (US). Similar facilitation in the B photoreceptor was observed following in vitro training that mimicked context conditioning in which presynaptic vestibular stimulation was presented repetitively during a continuous 7.5-min light. Subsequently, Ca(2+)-imaging experiments were conducted with Fura-2AM. It was determined that intracellular Ca(2+), the CS-induced second messenger critical for the induction of activity-dependent facilitation, was elevated in the B photoreceptor throughout the 7.5-min light presentation. These results indicate that activity-dependent facilitation within similar neural structures can underlie learning about both temporally diffuse contextual stimuli and temporally punctate CS-US pairings. These results suggest that a common mechanism may underlie learning about diffuse contextual stimuli as well as punctate-conditioned stimuli, provided that the stimuli are processed similarly in each type of conditioning arrangement. Consequently, the expression of different responses to contextual and discrete stimuli are likely to reflect a higher property of the neural network, and do not necessarily arise from unique underlying mechanisms.

Interactive contributions of intracellular calcium and protein phosphatases to massed-trials learning deficits in Hermissenda.

Using Hermissenda as subjects, massed-trials training deficits were examined. Associative pairings of light and rotation induced a progressively greater conditioned foot contraction in response to light as the intertrial interval (ITI) was extended (up to 8 min). In contrast, a short ITI (30 s) produced no evidence of learning. In a corresponding in vitro conditioning experiment that mimicked training of the intact animal, facilitation of neuronal excitability in the animal's B photoreceptors paralleled the results obtained in vivo. Imaging of intracellular Ca2+ using Fura-2 indicated that Ca2+ levels remained elevated during short ITIs. This Ca2+ accumulation appears to induce activation of protein phosphatases because normal facilitation of the B photoreceptors was induced with a short ITI if training occurred in the presence of a phosphatase inhibitor. These results suggest that intracellular Ca2+ and protein phosphatases contribute interactively to the kinetics of memory formation and provide evidence that an accumulation of intracellular Ca2+ across training trials may impede memory formation.

Ubiquitous molecular substrates for associative learning and activity-dependent neuronal facilitation.

Recent evidence suggests that many of the molecular cascades and substrates that contribute to learning-related forms of neuronal plasticity may be conserved across ostensibly disparate model systems. Notably, the facilitation of neuronal excitability and synaptic transmission that contribute to associative learning in Aplysia and Hermissenda, as well as associative LTP in hippocampal CA1 cells, all require (or are enhanced by) the convergence of a transient elevation in intracellular Ca2+ with transmitter binding to metabotropic cell-surface receptors. This temporal convergence of Ca2+ and G-protein-stimulated second-messenger cascades synergistically stimulates several classes of serine/threonine protein kinases, which in turn modulate receptor function or cell excitability through the phosphorylation of ion channels. We present a summary of the biophysical and molecular constituents of neuronal and synaptic facilitation in each of these three model systems. Although specific components of the underlying molecular cascades differ across these three systems, fundamental aspects of these cascades are widely conserved, leading to the conclusion that the conceptual semblance of these superficially disparate systems is far greater than is generally acknowledged. We suggest that the elucidation of mechanistic similarities between different systems will ultimately fulfill the goal of the model systems approach, that is, the description of critical and ubiquitous features of neuronal and synaptic events that contribute to memory induction.

Intracellular Ca2+ and adaptation of voltage responses to light in Hermissenda photoreceptors.

Fluorescent imaging of Ca2+ and intracellular recordings were used to assess Ca2+ increases and voltage responses during light presentations in Hermissenda B photoreceptors. Ca2+ levels increased and were sustained during a relatively long exposure to light. Repeated presentations of a brief light induced an elevation of intracellular Ca2+ that persisted throughout short interlight intervals, but which dissipated during long interlight intervals. In all instances, the magnitude of the intracellular Ca2+ signal was inversely related to the amplitude of the light-induced generator potential. Blocking of voltage-dependent Ca2+ channels did not significantly affect the magnitude of the Ca2+ signal, suggesting that the intracellular Ca2+ response arises primarily from release from intracellular stores. These results indicate that Ca2+ plays an important role in the modulation of the voltage responses to light, acting to suppress the response during repetitive or prolonged stimulation.

Incremental redistribution of protein kinase C underlies the acquisition curve during in vitro associative conditioning in Hermissenda.

An incremental increase in the excitability (i.e., input resistance, evoked spike frequency) of B photoreceptors in Hermissenda accompanied successive pairings of light and presynaptic stimulation of vestibular hair cells (simulating light-rotation pairings in an intact animal). Analysis of protein kinase C (PKC) in the Hermissenda's photoreceptors indicated a training-induced incremental reduction of PKC in cytosolic compartments, a tendency toward an increase in membrane compartments, and a small decrease in total enzyme activity (possibly owing to a downregulation or conversion of PKC to a calcium-independent state). Neither the biophysical or biochemical effects were observed in Hermissenda exposed to unpaired light and rotation or in those trained in the presence of the selective PKC inhibitor NPC-15437 (which had no effect on synaptic interactions or light-induced generator potentials). These results suggest that the intracellular redistribution of a protein kinase contributes critically to the kinetics of new learning.

Phospholipases and arachidonic acid contribute independently to sensory transduction and associative neuronal facilitation in Hermissenda type B photoreceptors.

During contiguous pairings of light and rotation, B photoreceptors in the Hermissenda eye undergo an increase in excitability that contributes to a modification of several light-elicited behaviors. This excitability increase requires a light-induced rise in intracellular Ca2+ in the photoreceptor concomitant with transmitter binding to G protein-coupled receptors as a result of presynaptic vestibular hair cell stimulation. Phospholipases and arachidonic acid (ArA) are here reported to be involved in independent signal transduction pathways that underlie both receptor function and activity-dependent facilitation of the B photoreceptor. 4-Bromophenacyl bromide (BPB), an inhibitor of phospholipases A2 (PLA2) and C (PLC), blocked the generation of light-induced depolarizing generator potentials, but had no affect on the inhibitory postsynaptic potential (IPSP) in the B cell that results from hair cell stimulation. Quinacrine, which predominantly blocks the activity of PLA2 in neurons, had no affect on either the light response or the IPSP, but did block increases in excitability (i.e. increased input resistance and elicited spike rate) of the B cell that results from pairings of light and presynaptic vestibular stimulation (i.e., in vitro associative conditioning). Neither nordihydroquararetic acid (NDGA), which inhibits metabolism of ArA by cyclooxygenase, nor indomethacin, which inhibits lipoxygenase metabolism of ArA, affected the light response or IPSP, but both blocked the increases in excitability in the B cell that accompanied in vitro conditioning. In combination with earlier results, these data suggest that ArA activates PKC in a synergistic fashion with Ca2+ and diacylglycerol in the B cell, and suggest that PLA2-induced ArA release, though not necessary for transduction of light or the hair cell-induced IPSP in the B cell, is a critical component of the convergence of signals that precipitates associative facilitation in this system.

Variations in learning reflect individual differences in sensory function and synaptic integration.

With the invertebrate Hermissenda as subjects, variability in acquisition of a learned association between light and rotation was correlated with the magnitude of the unconditioned responses elicited by these stimuli. Moreover, learning was facilitated by increasing stimulus intensity. In the isolated nervous system, pairings of light and mechanical stimulation of the animal's vestibular hair cells resulted in an increase in the excitability of B photoreceptors (an in vitro index of learning) that was strongly correlated with the strength of the synaptic interaction between the hair cells and the photoreceptors and weakly correlated with the magnitude of the light response in the photoreceptors. Because these in vitro results are not attributable to motor or motivational variables, they suggest that the efficacy of synaptic integration between sensory systems and sensory transduction is the primary determinant of the variability in learning.

Bidirectional regulation of neuronal potassium currents by the G-protein activator aluminum fluoride as a function of intracellular calcium concentration.

Hydrolysis-resistant activation of G-proteins by extracellular perfusion of fluoride ions was examined in Type B cells isolated from the cerebral ganglion of the marine mollusc Hermissenda. Under single-electrode voltage-clamp, modulation by aluminum fluoride ions of several classes of outward K+ currents as well as an inward Ca2+ current was observed. Following injection of the Ca2+ chelator EGTA, aluminum fluoride ions selectively increased a slow, voltage-dependent K+ current (IK) within 5 min of application, while in the absence of EGTA, aluminum fluoride ions induced a small, transient reduction of IK. Neither the magnitude nor steady-state inactivation of a fast, voltage-dependent K+ current (IA), nor a slow, Ca2+-dependent K+ current (IK-Ca), were affected by aluminum fluoride ions. In contrast, when perfusion of aluminum fluoride ions was accompanied by a repetitive depolarization and a concomitant increase in intracellular Ca2+, both IA and the combined late currents (IK and IK-Ca) were markedly reduced, a reduction which was not observed following depolarization alone or if the pairing of aluminum fluoride ions and depolarization was preceded by an injection of EGTA. The reduction of membrane conductance by the pairing of aluminum fluoride ions with depolarization could not be accounted for by an increased Ca2+ conductance, as aluminum fluoride ions produced only a small decrease in the voltage-dependent Ca2+ current. In total, these results indicate that regulatory G-proteins may bidirectionally modulate neuronal K+ currents, the direction of which is dependent on intracellular Ca2+ concentration. Such a dual regulatory mechanism may contribute to the modulation of membrane excitability observed when presynaptic activity is paired with postsynaptic depolarization, and thus may contribute to some forms of activity-dependent plasticity involving metabatropic receptors.

Calcium influx and release from intracellular stores contribute differentially to activity-dependent neuronal facilitation in Hermissenda photoreceptors.

A series of experiments is described that elucidates the sources of Ca2+ that contribute to activity-dependent neuronal facilitation in Hermissenda B photoreceptors during associative conditioning. In an in vitro preparation, pairings of a 4-s light with a 3-s mechanical stimulation of presynaptic hair cells increased the input resistance and elicited spike rate (i.e., excitability) of the B photoreceptors in the Hermissenda eye, indicative of a Ca(2+)-dependent process that is analogous to associative conditioning in the intact animal. This increase in excitability was reduced but not eliminated when hyperpolarizing current was applied to the B cell during the pairings, suggesting that voltage-dependent influx of Ca2+ contributed only a portion of the total calcium signal necessary for facilitation. Moreover, no increase in excitability was observed when a comparable current-induced depolarization of the photoreceptor was substituted for light-induced depolarization. In other experiments, Ca(2+)-dependent inactivation of a light-induced Na+ current was used as an index of intracellular Ca2+ concentration. It was determined that light caused a large increase in intracellular Ca2+ concentration regardless of whether the photoreceptor was allowed to freely depolarize in response to light or was voltage clamped at its resting membrane potential. Current-induced depolarization produced a smaller increase, while presynaptic stimulation had no measurable effect. Intracellular injections of either heparin, an antagonist of intracellular Ca2+ release, or EGTA, a general Ca2+ chelator, induced comparable reductions of light-induced Ca2+ accumulation. Finally, intracellular injections of heparin blocked the pairing-induced increases in B cell excitability as effectively as injections of EGTA. Taken as a whole, these data suggest that Ca2+ release from intracellular stores may be sufficient for the induction of facilitation in this preparation, while Ca2+ influx through voltage-dependent channels may have an additive effect and provide further evidence for the ubiquitous role of Ca2+ in learning-related forms of neuronal plasticity.

Trial-spacing effects in Hermissenda suggest contributions of associative and nonassociative cellular mechanisms.

In behaving Hermissenda, a preparatory conditioned response developed across repeated pairings of light (conditioned stimulus; CS) and rotation (unconditioned stimulus; US) with intertrial intervals (ITIs) of 60 and 120 s, but not 30 s. Likewise, contiguous in vitro stimulation of the visual and vestibular receptors, an analog of behavioral conditioning, resulted in an increase in the input resistance (i.e., excitability, a correlate of conditioning) of the B photoreceptors of the Hermissenda's eye, but only with ITIs greater than 60 s. Calcium signaling in the B cell, critical to the induction of this neuronal plasticity, was attenuated with shorter ITIs owing to (a) a reduction of the light-induced generator potential and hence voltage-dependent Ca2+ influx during the light CS, (b) a depression of the Ca2+ current that persisted throughout shorter ITIs, and (c) a steady-state inactivation of the Ca2+ current as a result of a sustained depolarization persisting from the previous trial. These results are consistent with a 2-process theory of associative learning in which a primary process (Ca2+ influx) may be opposed by a secondary process (depression of the Ca2+ current) during short ITIs.