Chronic stress exposure can produce sensitization of norepinephrine release in the terminal fields of locus coeruleus (LC) neurons. The present studies explore the potential localization and mechanism underlying the sensitized response of LC neurons in rats following chronic exposure to cold (2 weeks; 5*C). Single unit recordings of LC neurons in halothane-anesthetized rats were used to compare the effect of intraventricular administration of corticotropin releasing hormone (CRH; 0.3-3.0µg) in control and previously cold-exposed rats. The CRH-evoked increase in LC neuron activity was enhanced following chronic cold exposure, without alteration in basal activity. The enhanced activation was only apparent at higher doses of CRH, resulting in an increased slope of the dose-response relationship for CRH in previously cold-exposed rats. It is concluded that the sensitization of CRH-evoked norepinephrine release in cold-exposed rats is accompanied by sensitization of LC neuron activity. We hypothesized that the response of LC neurons to multiple excitatory inputs is enhanced. Using in vitro intracellular recordings, we subsequently examined whether CRH exerts a direct effect on LC neurons, and which ionic currents and second messenger systems are likely affected by CRH. It was demonstrated that CRH dose-dependently increases the firing rate of LC neurons through a direct (TTX-insensitive) mechanism by decreasing a potassium conductance via adenylate cyclase and protein kinase A. The CRH-evoked activation of LC neurons is, at least in part, mediated by CRH1 receptors. In subsequent in vitro experiments using intracellular recordings, the electrophysiological properties of LC neurons were compared between control and cold-exposed rats. We observed that the excitability and input resistance of LC neurons was enhanced in slices from cold-exposed rats. In addition, the accommodation of spike firing was reduced and there was a strong trend toward a reduction of the post-activation inhibitory period. These data demonstrate that the stress-induced sensitization of LC neurons is, at least in part, maintained in vitro and suggest that alterations in electrophysiological properties of LC neurons contribute to the chronic stress-induced sensitization of central noradrenergic function observed in vivo. Furthermore, these data suggest that an alteration in auto-inhibitory control of LC activity is involved in chronic stress-induced alterations.

How does the brain keep track of relevant spatial locations when the eyes move? In extrastriate, parietal and frontal cortex, and in the superior colliculus, neurons update stimulus representations in conjunction with eye movements. This updating reflects a transfer of visual information, from neurons that encode a salient location before the saccade, to neurons that encode the location after the saccade. Copies of the oculomotor command corollary discharge signals likely initiate this transfer. Spatial updating, or remapping, is thought to contribute to the maintenance of stable spatial representations as the eyes move. We investigated the circuitry that supports spatial updating in the primate brain. Our central hypothesis was that the forebrain commissures provide the primary route for remapping spatial locations across visual hemifields, which entails the interhemispheric transfer of visual information. Further, we hypothesized that these commissures provide the primary route for corollary discharge signals, generated in one hemisphere, to initiate spatial updating in the opposite hemisphere. We tested these hypotheses by measuring spatial behavior and neural activity in two split-brain macaques. In behavioral experiments, we observed striking initial impairments in the monkeys ability to update stimuli across visual hemifields. Surprisingly, however, we found that both animals were ultimately capable of performing these across-hemifield sequences. Both monkeys readily performed the same spatial task when updating required an interhemispheric transfer of corollary discharge signals, suggesting that these signals are transferred via subcortical pathways in the normal monkey. In physiological experiments, we found that neurons in lateral intraparietal cortex of the split-brain monkey can remap stimuli across visual hemifields, albeit with a reduction in the strength of remapping activity. These neurons were robustly active when within-hemifield updating was initiated by a saccade into the opposite hemifield. Our findings suggest that both visual and corollary discharge signals from opposite hemispheres can converge to update spatial representations in the absence of the forebrain commissures. These investigations provide new evidence that a unified and stable representation of visual space is supported by a redundant circuit, comprised of cortical as well as subcortical pathways, with a remarkable capacity for reorganization.

The extracellular signal-regulated kinase (ERK) cascade can transduce cell-surface signals to the nucleus in post-synaptic neurons during hippocampus-dependent learning and hippocampus-dependent synaptic plasticity, yet, whether the cascade can convey information about stimulus frequency or synaptic modification direction to the nucleus during plasticity events has not been addressed. The objective of the current study was to investigate whether ERK regulation differs as a function of stimulus frequency and in accordance with synaptic modification direction by comparing ERK regulation during LTP in area CA1 of the hippocampus in vivo to previous findings for ERK regulation during LTD in area CA1 in vivo (Thiels et al., 2002). The ultimate goal was to determine whether ERK functions as a general or as a specific plasticity kinase during synaptic plasticity events in the hippocampus. Using a combination of in vivo electrophysiology, pharmacology and Western blot analysis, I demonstrate that: (1) LTP induced by high-frequency stimulation applied to commissural fiber inputs to area CA1 pyramidal cells in the adult hippocampus in vivo is accompanied by a rapid yet transient increase in ERK2 activation; (2) blockade of NMDA receptors by MK-801 blocks both LTP induction and the associated increase in ERK2 activation; (3) HFS delivered in the presence of the ERK kinase inhibitor SL327 fails to produce a persistent potentiation; (5) phosphorylation of the transcriptional regulator cAMP response element-binding protein (CREB) is increased after HFS; and (6) inhibition of ERK2 activation by SL327 blocks this observed increase in pCREB. The similarity of the current findings with previous findings for ERK2 activation and regulation during LTD in area CA1 in vivo, suggests that the ERK cascade conveys a general as opposed to a specific plasticity signal during these two forms of synaptic plasticity in area CA1 in vivo. Differences in the coupling of ERK2 activation to CREB phosphorylation between LTP and LTD (Thiels et al., 2002), suggest that other signaling cascades are most likely operative in determining the direction of synaptic modification during bidirectional synaptic plasticity in the hippocampus.

The inferotemporal cortex in primates is thought to be the primary region that subserves object recognition. The studies presented here help to elucidate the role of IT in higher visual processing by addressing three specific outstanding issues. In the first study, we sought to determine whether IT neurons respond similarly to patterns that are perceptually confused. We considered a behavioral phenomenon whereby lateral mirror images are confused more frequently than vertical mirror images. By presenting mirror images to the monkey while simultaneously recording from IT neurons, we found that neurons differentiate less effectively between lateral mirror images than between vertical mirror images. This phenomenon may underlie the perceptual confusion documented in behavioral studies. In the second study, we sought to determine whether activity in IT reflects experience-based changes in perception. We tested this by first training monkeys to discriminate shape orientation. We then recorded from IT neurons while monkeys performed an orientation discrimination task with trained orientations, and passively viewed orientations of trained and untrained shapes. We found that training to discriminate between orientations of a shape significantly increases the ability of IT neurons to discriminate between those same orientations. This neuronal selectivity correlated with the monkeys ability to discriminate orientation. These data suggest that training-induced changes in perception are supported by processes in IT. Some IT neurons respond to the onset of a visual stimulus by firing a series of bursts at a frequency of around 5 Hz. One explanation for this phenomenon is that stimuli in the visual scene compete, with alternating success, for processing resources in IT. In the third study, we tested this by examining the oscillatory activity of IT neurons in response to the presentation of multiple stimuli, a central preferred image and a peripheral non-preferred image. We observed that the onset of a central pattern in the presence of the peripheral stimulus elicited strong oscillations phase-locked to pattern-onset. Onset of the peripheral stimulus in the presence of the central pattern elicited a succession of inhibitory troughs phase-locked to stimulus-onset. These results are congruent with a model of mutual inhibition of competing neuronal populations.

N-methyl-D-aspartate (NMDA) receptors are broadly involved in the CNS physiological and pathological processes. The voltage-dependent block by external Mg2+ is a signature characteristic of the NMDA receptors and is partly responsible for the many important roles NMDA receptors play. The work included in this Dissertation was designed to advance our understanding of the mechanism of Mg2+ block of NMDA receptors by exploring the permeant ion and subunit dependence of this process. Whole-cell and outside-out patch recordings from primary cultures of rat cortical neurons or heterologous mammalian cell lines were performed in combination with kinetic modeling. I report that Mg2+ inhibition of whole-cell NMDA currents in cortical neurons, which express NMDA receptors with NR2A or NR2B NR2 subunits, is very sensitive to ionic conditions. This phenomenon can be explained by a kinetic model which incorporates external permeant ion binding sites within the pore. Permeant ions binding to these sites prevents Mg2+ blocking or unblocking the channel. The general mechanisms of Mg2+ channel block of NR1/2D receptors is fundamentally similar to that of cortical receptors. However, Mg2+ block of NR1/2D receptors is much weaker than cortical receptors, mostly due to faster Mg2+ unblocking. Permeant ions also greatly affect Mg2+ block of NR1/2D receptors. The results can be explained by a kinetic model that incorporates two external and one internal permeant ion binding sites in the channel of NMDA receptors. When these sites are occupied by permeant ions, Mg2+ blocking or unblocking is affected. Thus, the research included in this Dissertation has deepened our understanding of the mechanism of Mg2+ block . The work also provides insights into NMDA receptor structure and gating.

The sleeping rat cycles between two well characterized physiological states, slow-wave sleep (SWS) and rapid eye-movement sleep (REM), often identified by the presence of large irregular activity (LIA) and theta activity, respectively, in the hippocampal EEG. Inspection of the activity of ensembles of hippocampal CA1 complex-spike cells along with the EEG reveals the presence of a third physiological state, distinctly different from both REM and SWS in both hippocampal EEG and population activity. The EEG during this state abruptly flattens for a few seconds, appearing very similar to the small-amplitude irregular activity (SIA) hippocampal EEG state reported in the literature to occur when rats are startled out of sleep. The flattening of the EEG is accompanied by a striking pattern of spike activity in the population of hippocampal pyramidal cells, wherein a small subset of cells becomes very active while the rest become quiet; the same subset of cells is usually active across long sequences of SIA. This dissertation shows (1) that these active cells are place cells whose place fields are in the location in which the rat is sleeping; (2) that the spontaneous SIA observed during sleep corresponds to the SIA state of increased alertness that has been reported in the literature to occur when rats are startled out of sleep; (3) that SIA is accompanied by a desynchronized neocortical EEG and low amplitude EMG; (4) that the cells active in SIA reflect a memory for the location in which the rat fell asleep, rather than an assessment of its location based on current sensory information; and (5) that the generation of SIA is likely to involve an increase of serotonin levels in the medial septal nucleus. It is proposed that SIA serves as a neural substrate for maintaining context memory during sleep, and that it reflects a partial arousal in response to internal or external stimuli that allows the animal to assess whether full awakening is warranted, without disrupting the sleep cycle.

Classical Neurotransmitter (NT) release is dependent upon the influx of calcium (Ca2+) through voltage gated Ca2+ channels. Once the local concentration of intracellular calcium increases to approximately 100 ÝM the calcium sensor detects the calcium ions and the process of vesicle fusion goes to completion. As such, transmitter release is very sensitive to alterations in the influx of Ca2+ into the presynaptic active zone. The objective of these experiments is to study how a known interaction between core complex proteins and presynaptic calcium channels might be important in the modulation of vesicle fusion. Syntaxin 1A (stx1A) is part of the minimal protein machinery necessary for vesicle fusion. Stx1A has also been shown to interact with Ca2+ channels in vitro. The functional significance of this interaction however is unclear. Work in expression systems supports the idea that stx1A functionally interacts with the N-type Ca2+ channel causing it to become inactivated. It has also been proposed that the Ca2+ channel- syntaxin interaction functions to co-localize the release machinery in the vicinity of the Ca2+ channel. Essentially, this would put the trigger for release next to the source of Ca2+ influx. To help discern whether this stx1A-N-type calcium channel interaction alters channel gating or simply serves to co-localize the two proteins, I employed a combination of molecular manipulations of syntaxin and the calcium channel. Electrophysiological recordings assaying NT release were then used to determine how these manipulations altered vesicle fusion. Data from two sets of experiments yielded the following information. 1) When the portion of the N-type Ca2+ channel that binds with syntaxin 1A was injected into Xenopus embryos, the strength of NT release onto the postsynaptic cell appeared to decrease. This was suggested by increases in both paired pulse facilitation and tetanic potentiation after injection of the competitive peptide. 2) Recordings of paired pulse facilitation and tetanic potentiation with injection of a mutant form of stx1A, which couples with the calcium channel but which has been shown not to modulate the channel, suggest an increase in synaptic strength as assayed through a tendency for both paired pulse and tetanic potentiation to decrease following injections of the mutant. Taken together, these results suggest that the interaction between stx1A and N-type calcium channels influences the level of NT release. They do not however, definitively distinguish between the interaction strictly being structural or modulatory. Instead, the data suggests that both may be occurring in vivo and a balance between these modulatory influences determines that what may be physiologically important.

Three Paradigms of Emotional Learning Differentially Affect Brainstem, Hypothalamic, and Limbic Circuits in the Rat Elizabeth A. Myers, M.S. University of Pittsburgh, 2004 Noradrenergic (NA) signaling in limbic forebrain regions, such as the central nucleus of the amygdala (CeA), shapes the encoding and expression of emotional learning, and modulates responses to stress and anxiety. The present study examined whether categorically different emotional stress models, [cholecystokinin (CCK), trimethylthiazoline (TMT), and yohimbine (YO)] support behaviorally aversive conditioning and differentially activate CRH-positive neurons in the hypothalamus, medullary and pontine NA neurons, and ascending inputs to the CeA. A conditioned flavor avoidance (CFA) paradigm using a flavor preference test was implemented as a measure of aversive conditioning for each stressor. In a terminal experiment, rats received either an injection of CCK (10 µg/kg, i.p.), YO (5 mg/kg. i.p.), or 15 min exposure to an aversive odor, TMT, and were perfused 60-120 min later. In a subset of rats, retrograde neural tracer was microinjected into the CeA prior to stressor treatment and perfusion. Brainstem and forebrain sections were processed for immunocytochemical localization of cFos and either dopamine beta hydroxylase (DbH) to identify NA neurons, corticotropin-releasing hormone (CRH), or neural tracer to identify hindbrain CeA-projecting neurons. All stressors activated hypothalamic CRH neurons, produced a relatively strong CFA in a 2-bottle choice test, and recruited similar proportions of CeA-projecting neurons arising from the parabrachial nucleus, a projection path critical for this behavioral paradigm. All stressors recruited NA neurons within the medullary A2 cell group to a similar extent, whereas those in the medullary A1 cell group and pontine A6 cell group were recruited more selectively by TMT and YO compared to CCK. Afferent inputs to the CeA arising in these hindbrain cell groups were activated in a parallel manner, with TMT and YO recruiting a much greater proportion of CeA-projecting neurons in the A1 and A6 cell groups. These findings lend support to the working hypothesis that different emotional stimuli may potentially influence emotional learning via stressor-specific ascending NA projection pathways to the CeA. In general, elucidating stressor-specific neural circuitry may provide new insight into how to design effective therapeutic measures for a wide range of human disorders and conditions involving the NA system.

Glutamate-induced oxidative toxicity in HT22 cells and primary immature cortical cultures provides an excellent model system for studying oxidative stress-dependent neurodegeneration. Glutamate treatment leads to cysteine and subsequent glutathione depletion, followed by the steady accumulation of reactive oxygen species (ROS). This form of cell death depends upon the persistent activation, via phosphorylation, of extracellular signal regulated kinase (ERK) kinase-1/2 (ERK-1/2) that occurs during oxidative stress. However the mechanisms responsible for this chronic activation of ERK during oxidative stress have not been well characterized. In this thesis, I demonstrate that ERK activation is dependent upon the tonic activity of the phosphatidylinositol-3-kinase (PI3K)-Akt pathway and the subsequent activation of MEK. Furthermore, the persistent ERK activation that leads to cellular toxicity can be driven by the oxidative-dependent inactivation of ERK-phosphatases. Thus the balance of activating kinase activity and inactivating phosphatase activity dynamically regulates ERK-dependent signaling and is a major determinant of neuronal cell responses to oxidative stress. The overexpression of a negative regulator of the ERK MAPK pathway, the ERK-specific phosphatase MKP3, led to protection of both HT22 cells and primary immature cortical cultures from oxidative toxicity. Furthermore, a catalytically inactive form of MKP3 (MKP3 C293S) was shown to physically restrict activate ERK to the cytoplasm. Because overexpression of MKP3 C293S was also shown to be neuroprotective, translocation of active ERK to the nucleus, but not ERK activation alone, must be required for glutamate-induced oxidative toxicity. Collectively, these results clearly place ERK activation as a necessary event that leads to neuronal cell death during oxidative stress and have revealed some unique mechanisms by which ROS accumulation drives ERK activation.

This main purpose of this dissertation is to describe a set of three experiments that were designed to explore neural systems supporting navigation and spatial cognition. Additionally, the theoretical significance of the experiments and their relationship to the current literature will be discussed in detail. This introductory section provides some motivation for the study of navigation and spatial cognition, followed by an overview of the three experiments, a discussion of previous work leading to each experiment, and a brief description of the methodology used for each study. Experiment 1 utilizes high-density neurophysiological recording techniques to examine hippocampal place cell activity in cue-rich environments where the rats are exposed to cue conflict situations, and concludes that the hippocampal representation of spatial location is concordant. Experiment 2 takes advantage of the transynaptic retrograde transport of pseudorabies virus to define the synaptology of a pathway through which cells whose activity code for head direction could receive vestibular sensory information. Experiment 3 also uses pseudorabies virus to define polysynaptic pathways from the hippocampal formation to the hypothalamus via the lateral septum, which could provide the pathways for spatial information to be transported to areas involved in motivated behavior. The last chapter of this dissertation provides an extensive discussion of some issues surrounding the data presented, and brings together the experimental conclusions to construct generalizations regarding neural substrates of navigation and spatial cognition.