Blood-brain barrier disruption in Parkinsons disease and Parkinsons disease models that involve dopaminergic neurodegeneration has been minimally evaluated despite mounting evidence for its involvement. Oxidative stress and neuroinflammation are both involved in Parkinsons disease pathology and also both contribute to blood-brain barrier dysfunction, creating the likelihood that blood-brain barrier disruption is also a pathological feature of the disease. Disruption of the blood-brain barrier can lead to an increased susceptibility to neuronal injury and potentially neurodegeneration due to the invasion of peripheral factors such as immunoglobulins and environmental toxins into the brain. An understanding of mechanisms by which blood-brain barrier disruption occurs may lead to the development of new approaches for the treatment of neurologic diseases such as Parkinsons disease.

The regulation of transmitter release at the neuromuscular junction is tightly regulated by the influx of calcium in the presynaptic nerve terminal. Interestingly, the probability that release sites at the neuromuscular junction will liberate transmitter during each action potential is very low. The reasons for this low probability of release are not well understood. To test the hypothesis that individual N-type calcium channels open with a low probability, single channel recordings of N-type voltage-gated calcium channels were performed. Using this approach I determined the conductance of these channels, their probability of gating during an action potential waveform, and the magnitude of calcium flux during a single channel opening. I conclude from these studies that N-type voltage-gated calcium channels have a very low probability of opening (< 5%) during an action potential and the characteristics of calcium entry during single channel openings can help to explain the low probability of transmitter release at release sites in the neuromuscular junction. To understand how calcium current is activated physiologically, the activation and resulting current from N-type calcium channels elicited by different action potential waveforms were studied. This work was carried out at both room temperature and 37°C to provide a physiological context. Using the whole-cell patch clamp techniques, I studied the activation of current during various action potential shapes and conditions, and the kinetics of N- and L-type current activation. Using this approach I determined that N-type channels activate more slowly than L-type. Furthermore, depending on the action potential shape used and the temperature, action potentials can activate varying proportions (I/Imax) of N-type calcium current (ranging from 10-100%). Under physiological conditions using a frog motoneuron action potential waveform I determined that there was a very low proportion of calcium current activated by a natural action potential (~32%). Adenosine 5´-triphosphate (ATP) is co-released with acetylcholine (ACh) at the neuromuscular junction, and has been found to inhibit transmission. I used the cutaneous pectoris muscle of the Rana pipiens to study ATP-mediated modulation of ACh release. Intracellular postsynaptic recordings were used as a measure of ACh release, and agents that perturb the ATP signaling were examined.

Spike timing dependent plasticity (STDP) is a form of synaptic plasticity that depends on the relative time of activation of a presynaptic neuron and its postsynaptic neuron. STDP in the synapses made by Schaffer collateral afferents onto hippocampal CA1 pyramidal neurons (CA3- CA1 synapses) is NMDA receptor dependent. The objective of the current study was to develop and test a technique of glutamate iontophoresis that could replace the role of presynaptic neurotransmitter release at the CA3-CA1 synapse, so that the postsynaptic mechanisms involved in the induction of STDP could be isolated for study. Therefore, this document describes: (1) fabrication of electrodes that could be used for millisecond-level microiontophoresis in acute slice preparations of the juvenile rat hippocampus; (2) characterization of the properties and limitations of microiontophoresis in slice tissue, specifically for activation of postsynaptic ionotropic glutamate receptors at the CA3-CA1 synapse; (3) induction of STDP by pairing microiontophoresis with postsynaptic depolarization; (4) characterization of the properties and limitations of microiontophoretically induced STDP. It was determined that microiontophoresis is a viable technique to study the postsynaptic mechanisms of STDP at the CA3-CA1 synapse. My results also show that microiontophoretically induced STDP exhibits many of the same general properties as STDP induced either synaptically or by exogenously applied agonists. Microiontophoretically induced STDP also exhibits other features that will need to be considered during the design and interpretation of further experiments.

Dendrites are classically regarded as the brains listeners, while neuronal output is thought to be the exclusive privilege of the axon. Although we now appreciate the complexity of dendritic integration, the role of dendrites as output structures has received less attention. This is becoming an increasingly important topic, as the list of cell types with release competent dendrites continues to grow. One boon of coupling dendritic activity to dendritic release is that outputs from a single neuron typically thought to occur from fixed sites with stereotyped dynamics may occur for signals of varying spatial extent, timecourse, and release efficacy. In essence, dendritic output may inherit the same diversity characteristic of events in excitable dendrites. Here I studied dendritic transmitter output and its modulation in cells of the accessory olfactory bulb a CNS structure critical for processing species-specific chemical signals called pheromones. Because of the stereotypy of its inputs, the prevalence of dendritic transmitter release from its cells, and its well-defined outputs, the AOB offers a superb model system for studying the integrative and output properties of dendrites. I first characterized basic excitable properties of the apical dendrites of mitral cells (the principal AOB neurons), and observed that they conduct non-decremental action potentials (APs). In addition to APs, these dendrites were also found to support compartmentalized, synaptically-evoked calcium spikes. Both APs and local spikes were triggers of dendritic glutamate release and feedback inhibition, suggesting that neuronal output can be flexibly routed to particular populations of postysynaptic cells. I next asked whether the relative efficacy of particular dendritic events as triggers of transmitter release can be altered, as this could provide an additional level of control over single neuron output. I found that metabotropic glutamate receptors (mGluRs) play a key role in controlling dendritic output from AOB mitral cells and an obligatory role in concomitant feedback inhibition. This work culminates with the demonstration of a new principle of neuronal signaling: the ability of mGluRs to gate a transition between phasic and tonic dendritic transmitter release. Taken in total, these results extend our understanding of how the outputs from single neurons are controlled.

Alterations in the inhibitory circuitry of the dorsolateral prefrontal cortex (DLPFC) appear to contribute to the impairments in working memory observed in individuals with schizophrenia. Consistent with this idea, a microarray study indicated that the mRNA levels of GABAA receptor &alpha;4 and &delta; subunits were lower in the DLPFC of subjects with schizophrenia. However, although &alpha;4 and &delta; subunits co-assemble to form functional receptors, the differences in &alpha;4 and &delta; mRNA expression in schizophrenia were not correlated. We assessed the mRNA levels of &alpha;4 and &delta; in the DLPFC of 23 subjects with schizophrenia matched to control subjects by in situ hybridization. The level of &alpha;4 mRNA was lower only in subjects with schizophrenia receiving medications at the time of death, whereas the level of &delta; mRNA was significantly lower in schizophrenia, regardless of the medications used at the time of death. We also found that across postnatal development of monkey DLPFC the level of &alpha;4 mRNA decreased with age, whereas that of &delta; mRNA increased in a manner similar to that previously observed for the &alpha;1 subunit. Given that &alpha;1 mRNA levels are lower in schizophrenia and &alpha;1 subunits can co-assemble with &delta; subunits, lower &delta; mRNA in schizophrenia could represent lower GABAA &alpha;1&beta;x&delta; rather than &alpha;4&beta;x&delta; receptors. Studies suggest that reduced signaling through excitatory synapses, as hypothesized to be present in schizophrenia, give rise to decreased expression of &delta; subunit mRNA. To test this hypothesis, we measured the levels of &delta; subunit mRNA in the prefrontal cortex of four rodentmodels of reduced cortical excitatory drive: 1) NMDAR NR1 hypomorphic mice, 2) rats with adult mediodorsal thalamic nuclei lesions, 3) rats with neonatal ventral hippocampal lesions and 4) TrkB hypomorphic mice reported to have decreased dendritic arborization. However, the mRNA levels of &delta; subunit were unchanged in the PFC of any of the animal models analyzed. Thus, although reduced signaling through excitatory synapses might be a pathogenetic mechanism for other abnormalities in schizophrenia, the convergence of the findings from this study do not support the hypothesis that it accounts for the lower expression of GABAA receptor &delta; subunit mRNA.

The limbic and central noradrenergic systems are sensitive to stress and demonstrate pathophysiology in individuals with mood and anxiety disorders. Neurons from the central nucleus of the amygdala (CeA) form synapses onto the dendrites of noradrenergic neurons of the locus coeruleus (LC) in the rostrolateral peri-coerulear area. The CeA-LC pathway is thought to contain GABA and corticotropin-releasing hormone (CRH), both of which have opposing effects on LC activity. The current study further characterized the CeA-LC pathway in vivo by examining extracellular electrophysiological LC activity during electrical and pharmacological manipulation of the CeA in halothane-anesthetized rats. The majority of LC neurons exhibited an excitatory response to electrical CeA stimulation, with a small group of neurons responding with pure inhibition or antidromic activation. Pharmacological activation of the CeA confirmed excitatory responses to electrical stimulation, whereas pharmacological inactivation of the CeA had no effect on LC activity. Excitatory responses to electrical CeA stimulation were partially attenuated following ventricular infusions of the CRH antagonist, D-Phe-CRH, but were completely attenuated following infusions of the excitatory amino acid antagonist, kynurenic acid. Ventricular administration of kynurenic acid during electrical CeA stimulation also revealed an inhibitory period that simultaneously occurs with excitatory responses. This study confirms previous findings suggesting that the CeA partially mediates excitatory responses via CRH. These findings further suggest that there is a glutamatergic component to excitatory responses following activation of the CeA. In addition, these data suggest that upon activation, the CeA provides a simultaneous inhibitory drive to LC neurons. Collectively, these data provide additional support for an excitatory limbic input to neurons of the LC. The present data also suggest that the CeA may selectively modulate LC activity across its projection sites via the simultaneous inhibitory drive from the CeA. Activation of the CeA-LC pathway, particularly during stress responses, may be critical for noradrenergic modulation of cortical and limbic areas involved in attention, emotional learning, and responses to stressful stimuli.

Previous transneuronal tracing studies in the rat and ferret have identified regions throughout the spinal cord, medulla, and pons that are synaptically linked to the diaphragm muscle; however, the extended circuits that innervate the diaphragm of the cat have not been well defined. The N2C strain of rabies virus has been shown to be an effective transneuronal retrograde tracer of the polysynaptic circuits innervating a single muscle. Rabies was injected throughout the left costal region of the diaphragm in the cat to identify brain regions throughout the neuraxis that influence diaphragm function. Infected neurons were localized throughout the cervical and thoracic spinal cord with a concentration of labeling in the vicinity of the phrenic nucleus where diaphragm motoneurons are known to reside. Infection was also found throughout the medulla and pons particularly around the regions of the dorsal and ventral respiratory groups and the medial and lateral reticular formations but also in several other areas including the caudal raphe nuclei, parabrachial nuclear complex, vestibular nuclei, ventral paratrigeminal area, lateral reticular nucleus, and retrotrapezoid nucleus. Infection was also localized in fastigial and dentate deep cerebellar nuclei. Additionally infected neurons were observed in the midbrain, particularly in the periaqueductal grey matter and mesencephalic reticular nucleus but also in other regions including the pedunculopontine nucleus, cuneiform nucleus, Edinger-Westphal nucleus, nucleus locus coeruleus, and the red nucleus. Diencephalic labeling was most concentrated within the regions of the perifornical area and the medial parvocellular division of the paraventricular nucleus of the hypothalamus but was also found in several other areas. Infected neurons were also observed in the cruciate sulcus of the cerebral cortex as well as in prefrontal and insular cortices. This data suggests that the central circuits innervating the diaphragm muscle in the cat are highly complex and integrated with the central circuits controlling a variety of other non-respiratory behaviors.

Motivated behaviors are mediated by a neuronal circuit that includes the amygdala, prefrontal cortex and nucleus accumbens. The amygdala signals the affective significance of cues in the environment, and the prefrontal cortex and accumbens contribute to planning and selecting actions. Thus, the physiological interactions within this circuit may be neural correlates for the interface between motivation and voluntary movement. Using in vivo electrophysiological recordings, we measured the effects of amygdala activation on excitatory transmission in the prefrontal cortex-to-nucleus accumbens pathway. In the prefrontal cortex, a subset of neurons that projected to the accumbens was excited by amygdala stimulation. Prefrontal-to-accumbens projecting neurons were also excited by Pavlovian conditioned odors, suggesting a role for this pathway in the expression of conditioned fear. In neurons of the accumbens, amygdala input facilitated and shifted the latency of spikes elicited by prefrontal cortex stimulation. Facilitation was greatest when amygdala and prefrontal cortex were stimulated nearly synchronously at subthreshold intensities, suggesting that accumbens neurons integrate both the timing and strength of their afferent inputs. In additional accumbens recordings, high intensity amygdala activation induced an activity-dependent depression of prefrontally elicited spiking; thus, the amygdala is able to modulate prefrontal-to-accumbens transmission by several mechanisms. Taken together, these findings show that the amygdala can influence neuronal transmission at two critical motor-related sites in the brain. In this way, affective information may constrain and gate action, resulting in optimal behavioral responses to salient motivational stimuli.

Parkinsons disease (PD) neurodegeneration is characterized by loss of the dopaminergic cells of the substantia nigra, and has been linked to oxidative stress and mitochondrial dysfunction. The reactive neurotransmitter dopamine (DA) may play a role in neuronal vulnerability. DA oxidation has been shown to elicit dopaminergic toxicity in animal models, covalently modify proteins, and affect mitochondrial function. However, mitochondrial protein targets of DA modification are unknown. In this study, I utilized proteomic techniques to identify and characterize mitochondrial proteins altered following in vitro exposure to DA oxidation. Using two-dimensional difference in-gel electrophoresis and mass spectrometry analyses, I identified a subset of mitochondrial proteins that exhibited decreased abundance following exposure of isolated rat brain mitochondria to DA quinone (DAQ). Losses of two of these proteins, mitochondrial creatine kinase (MtCK) and mitofilin were further confirmed by Western blot analyses. Western blot also confirmed significant decreases of these two proteins in differentiated PC12 cells exposed to DA. I next utilized two-dimensional gel electrophoresis with autoradiography to identify proteins covalently modified by DAQ. I identified a subset of proteins covalently modified by ¹⁴C-DA from rat brain mitochondria exposed to ¹⁴C-DAQ and from differentiated SH-SY5Y cells exposed to ¹⁴C-DA. Proteins including mortalin/GRP75/mtHSP70, subunits of Complex I, MtCK, and mitofilin, amongst other proteins, were found to be covalently modified. We chose to further examine mitofilin, a protein implicated in maintaining mitochondrial structure. To characterize the effect of altered mitofilin levels on cell viability, I utilized overexpression and knockdown techniques to modulate mitofilin expression in dopaminergic cell lines, differentiated PC12 and SH-SY5Y cells, and examined their response to dopaminergic toxins, DA and rotenone. I found that increased mitofilin expression was protective against both DA- and rotenone-induced toxicity in both cell lines, and decreased mitofilin enhanced DA-induced toxicity in differentiated SH-SY5Y cells. Therefore, in this thesis, I identified a subset of mitochondrial and cellular proteins that are potential targets of DA-induced modification, and may have roles in PD pathogenesis. Modulating the expression level of one of these proteins, mitofilin, affected the cellular response to toxins, and may play a role in dopaminergic cell vulnerability.

Lateral inhibition is a circuit motif found throughout the nervous system that often generates contrast enhancement and center-surround receptive fields. While widespread lateral inhibition of mitral cells (the principal output neurons of the bulb) is mediated by granule cells (inhibitory interneurons), due to the distributed representation of odorant-evoked activity it is currently unclear how this inhibition is specified or what functions it mediates. Given the reciprocal nature of the connection between mitral and granule cells, mitral cells that are the targets of lateral inhibition (postsynaptic) are able to modulate the activity of the granule cells mediating this lateral inhibition. This led us to hypothesize that lateral inhibition could be modulated by the activity of both pre- as well as postsynaptic mitral cells. I first characterize the lateral interactions between mitral cells and show that the effectiveness of lateral inhibition between them is dependent on the activity of the postsynaptic neuron such that lateral inhibition is only able to reduce the firing rate within a specific range of postsynaptic firing rates. I call this activity-dependent lateral inhibition. I then investigated this novel form of inhibition further and provide evidence indicating that it results from cooperative activation of the inhibitory granule cells. I investigated the mechanism and functional implications of these physiological results further using computational techniques. First, I show that integration of activity between pre- and postsynaptic mitral cells within a granule cell is sufficient to result in activity-dependent inhibition. I then used this model to show that activity-dependent inhibition is able to enhance the contrast and decorrelate initially similar input patterns to the network in a spatially independent manner. These results provide evidence for a novel form of neuronal interaction dependent on the activity of neurons within the network which could have important functional implications in olfaction as well as other brain areas.