The organization of retinal ganglion cell (RGC) axons within the optic nerve (ON), chiasm (OCh), and tract of the frog, Rana pipiens, was examined using both light and electron microscopy. In one group of experiments, anterograde and retrograde transport of horseradish peroxidase (HRP) revealed the spatial reorganizations of RGC axons between the retina and central retinorecipient targets. Optic axons form a concentric representation or map of the retinal surface across the distal ON cross-section. Within this map, the distal ON contains two representations of the temporal and nasal retinal quadrants. Fiber reorganizations within the proximal ON result in axons from central RGCs becoming positioned within the dorsal region of the nerve and in axons of more peripheral RGC becoming localized to the ventral, anterior, and posterior margins of the nerve. As a result of this proximal ON fiber reorganization, age-related RGC axons become grouped in laminae that are aligned perpendicular to the posterodorsal-anteroventral axis of the optic projection. Axons occupying complementary positions in different retinal representations converge in a dorsal-to-ventral sequence as the projection passes through the perichiasm. Within the prechiasm, axons separate from the laminated optic projection to form four fascicles; the projection to corpus geniculatum, the projection to the neuropil of Bellonci, the basal optic root, and the marginal optic tract (MOT). In addition, electron microscopic examination of the optic fiber spectrum revealed the morphological organization of the retinofugal projection within the ON and the lateral division of the marginal optic tract (lMOT). Within the ON, myelinated and unmyelinated axons of various sizes are homogeneously distributed. Optic axons are reorganized within the MOT such that within the lMOT large myelinated optic axons become concentrated within ventromedial regions and small myelinated axons become concentrated within the dorsolateral regions. In rostral lMOT larger unmyelinated axons are also concentrated along the ventromedial regions. The morphological organization of the optic axons within lMOT, established by the reorganization of the retinofugal projection within MOT, anticipates the innervation pattern of the optic axons within the laminated, superficial layers of the optic tectum.

Recent anatomical and neurochemical studies suggest that the dorsal raphe nucleus (DRN) of the brainstem is composed of several subdivisions that have different afferent and efferent connections and are different neurochemically. In the present study, DRN subdivisions projecting to visual system structures were studied in a highly visual rodent, the Mongolian gerbil (Meriones unguiculatus). In one group of gerbils, the fluorescent neural tract tracer Fluoro-Gold was injected into the superior colliculus (SC), and brain sections also were double-labeled for serotonin (5-hydroxytryptamine, 5-HT). In another group of gerbils, the neural tract tracer cholera toxin subunit B (CTB) was injected into the SC, lateral geniculate nucleus (LGN), or primary visual cortex (V1). The location of retrogradely labeled cells in the DRN was analyzed at five rostro-caudal levels to determine DRN subdivisions projecting to these visual structures. At the three most rostral coronal levels, virtually all cells projecting to the SC occurred in the lateral DRN; most of these cells were serotonergic. At these rostral levels, virtually all cells projecting to the LGN also were observed in the lateral DRN. At the two most caudal levels, retrogradely labeled cells were observed in the ventromedial and interfascicular DRN following tracer injections into the SC or LGN. Following CTB injections into the primary visual cortex, virtually all retrogradely labeled cells occurred in the ventromedial DRN at all rostro-caudal levels. Therefore, the ventral caudal region of the DRN appeared to be the only DRN region projecting to all three visual areas (SC, LGN, and V1). In order to determine whether functional differences might exist among DRN subdivisions, the expression of c-fos, an immediate-early gene, was analyzed in all DRN subdivisions throughout the light-dark cycle. In all subdivisions, the highest c-Fos expression was observed one hour after the onset of the dark period. However, ANOVA analysis revealed that the diurnal variation in c-Fos expression was statistically significant only in the caudal ventral region of the DRN, and gradually became less prominent in the caudal-to-rostral and ventral-to-dorsal directions. A portion of this Dissertation has been previously published in the Journal of Comparative Neurology (Janušonis et al., 1999).

In rodents, there exists a strong interaction between the reproductive and circadian systems. For this thesis the female hamster was used as a model for the study of this interaction. Studies described in chapter II investigated whether the circadian regulation of reproductive processes may be through direct input of the suprachiasmatic nucleus (SCN) to neurons containing estrogen receptor (ER) and/or to neurons containing luteinizing hormone releasing hormone (LHRH). The anterograde tracer Phaseolus vulgaris leucoagglutinin (PHA-L) was applied to the SCN and double label immunocytochemistry for PHA-L and either ER or LHRH was carried out. Both ER- and LHRH-immunoreactive cells show appositions with SCN efferents or with efferents of the subparaventricular nucleus and the retrochiasmatic area. Results suggest that the circadian system can regulate reproductive processes via input to LHRH- and/or ER-containing neurons. Studies described in chapter 111 investigated whether effects of estrogen on circadian rhythms may be exerted through estrogen-binding systems afferent to the SCN. Immunocytochemistry for ER and the retrograde tracer cholera toxin B subunit, after its application to the SCN, demonstrated that some areas contain relatively high percentages of SCN afferent neurons which show ER immunoreactivity. Retrograde tracing results were compared with results of anterograde tracing from some of the sites containing SCN afferents. Furthermore, using a combined retrograde and anterograde tracing technique, SCN input to some SCN afferent neurons was demonstrated. However, no evidence of reciprocity between single ER-immunoreactive cells and the SCN was found. Results indicate the existence of estrogen binding systems afferent to the SCN which might mediate the effects of gonadal steroid hormones on circadian rhythms. Studies in chapter IV analyze the effects of blockade of SCN axonal output by local unilateral application of tetrodotoxin (TTX) on the LH surge. Injections of TTX on either the morning or the afternoon of proestrus were unable to block the LH surge. Results favor the interpretation that the SCN output signal responsible of the circadian gating of the LH surge occurs before the onset of the light period on the day of proestrus.

The frog neuromuscular junction is a unique model that allowed us to selectively remove cellular components from the neuromuscular junction and create preparations with varying degrees of nerve terminal stability. We found further evidence that frog terminal Schwann cells communicate with their cellular partners, as terminal Schwann cells responded with changes in number or morphology as a result of changes in synaptic integrity. Terminal Schwann cells divided at synaptic sites in response to a regenerating nerve terminal. Terminal Schwann cells also had morphological changes in response to changes in status of their cellular partners; they extended processes in response to removal of the nerve terminal. Orientation and length of these processes was profoundly affected by the presence or absence of muscle fiber and nerve terminal. Similar to observations at the mammalian neuromuscular junction, terminal Schwann cells appear to play a role in reinnervation, as frequently regenerating nerve terminals were within the confines of terminal Schwann cells and their processes. I also investigated the organization of actin within preparations with varying amounts of nerve terminal stability, including developing nerve terminals and regenerating adult nerve terminals that were forming either stable or unstable connections. Previously, F-actin stained target-deprived nerve terminals in a ladder-like pattern and was concentrated in the nonrelease domains (Dunaevsky and Connor 2000). I found that β-actin was similarly distributed and localized to the nonrelease domains of nerve terminals at intact neuromuscular junctions. Further, association of actin with these particular domains appeared to be important for nerve terminal stability. As nerve terminals acquired increasing stability during development, they acquired this domain specific distribution of F-actin. Additionally, although synaptic sites with stable regenerating nerve terminal acquired this ladder-like pattern of F-actin, it was very rare for unstable regenerating nerve terminals to do so. I also tested the dynamic nature of F-actin with pharmacological perturbation. F-actin at nonrelease domains was found to be very stable. This stability of the F-actin based cytoskeleton further suggests that F-actin at the nonrelease domains of nerve terminals may play a role in the stability of motor nerve terminals.

The purpose of this study was to determine if administering two painful stimuli used to assess motor responses in patients with brain injury, nail bed compression (NBC) and supraorbital pressure (SOP), had an effect on intracranial pressure (ICP), mean arterial pressure (MAP), cerebral perfusion pressure (CPP), heart rate (HR), and motor responses, in unconscious patients. Thirteen unconscious adult male and female subjects with brain injury were enrolled in the study. All subjects had normal ICP and CPP, and were hemodynamically stable. After collection of baseline values of the dependent variables, NBC and SOP were delivered on both sides of the body while ICP, MAP, CPP and HR were recorded from the bedside monitor. Subjects were recorded on video tape for motor responses to NBC and SOP and later given a motor score on the Glasgow Coma Scale. Pressures used to administer NBC and SOP were measured in psi by devices constructed for purposes of this study. Data were analyzed using MANOVA statistical procedures to detect differences within subjects on all physiological values from baseline. Findings from this study were that NBC and SOP administered on both sides of the body resulted in statistically significant increases in ICP, MAP, CPP, and HR from baseline for a brief, yet unsustained time period (p = $<$0.05). The ICP had returned to baseline in 30 seconds, the MAP and CPP within four minutes, and HR by the second minute after administration of NBC and SOP. ANOVA statistical procedure was used to detect differences in motor scores when NBC and SOP were given. There were no statistical differences between motor scores on the GCS with NBC or SOP. The mean pressures that were measured on NBC and SOP were 77.11 $\pm$ 30.98 and 86.1 $\pm$ 8.54 psi, respectively. These data suggest that NBC and SOP do have an effect on physiological indices of cerebral perfusion by an increasing ICP, MAP, CPP, and HR for a brief period of time but return to baseline quickly. Therefore, administering painful stimuli to evaluate motor responses in unconscious patients who have normal ICP and CPP is probably safe. Also, the data suggest that NBC and SOP produce similar motor responses and could both be used to assess unconscious patients.

Sexual behavior in female guinea pigs and rats is dependent upon circulating ovarian steroid hormones. The actions of these hormones on behavior are mediated by intracellular receptors located within interconnected brain regions. This complex neural network integrates somatosensory and hormonal information relevant to sexual behavior. An understanding of the anatomy of this network is important for understanding how behavior is generated. This dissertation investigated the neural projections from a region important for the induction of sexual behavior, the ventrolateral hypothalamus. Based on the steroid-sensitive neural network model described in rats, several experiments were designed to test predictions of this neural model in guinea pigs. In the first experiment, estrogen receptor- and estradiol-induced progestin receptor-containing cells were localized in the midbrain, a major projection site of the steroid receptor-rich region within the ventrolateral hypothalamus. In the second experiment, the anterograde tracer Phaseolus vulgaris-Leucoagglutinin, an anterograde tract-tracer, was deposited within the estrogen receptor-containing region of the ventrolateral hypothalamus. Projections from this area were found in most other estrogen receptor-containing sites, including the midbrain central gray, often closely associated with estrogen receptor-containing cells. The third experiment examined the distributions of substance P, a peptide found in ovarian steroid hormone receptor-containing cells in the ventrolateral hypothalamus and estrogen receptor-containing cells in the midbrain central gray. Substance P-immunoreactive boutons were found closely associated with some estrogen receptor-containing cells suggesting substance P modulation of ovarian steroid hormone receptor-containing cells. These connections, taken together with the hypothalamic projections closely associated with estrogen receptor-containing cells in the midbrain, support the idea that some estrogen receptor-containing cells may be directly linked. These studies provide information on the neural connections between estrogen receptor-containing regions and cells which may be important for regulating functions of the steroid hormone sensitive neural network.

Mice with a deletion of the hypothalamic basic helix-loop-helix transcription factor Nhlh2 (N2KO) display adult onset obesity, implicating Nhlh2 in the neuronal circuits regulating energy availability. Nhlh2 co-localizes with the hypothalamic thyrotropin-releasing hormone (TRH) neurons in the paraventricular nucleus (PVN) and proopiomelanocortin (POMC) neurons in the arcuate nucleus. N2KO mice become obese due to reduced physical activity in the absence of hyperphagia making them a unique mouse model for the study weight gain, obesity and energy expenditure. Signals that regulate Nhlh2 and the effects of Nhlh2 on peripheral tissues remain largely unknown. The research presented here utilized numerous techniques to investigate the effects of changes in energy availability on Nhlh2 expression. We show that Nhlh2 expression decreases significantly with food deprivation and cold exposure. Nhlh2 expression is stimulated with food return or leptin injection following food deprivation or return to room temperature following cold exposure. These data suggest that Nhlh2 gene expression responds positively to increased energy availability and negatively to reduced energy availability. These findings combined with the phenotype of N2KO mice led us to propose that Nhlh2 integrates energy availability inputs in various hypothalamic nuclei to drive expression of genes required for body weight maintenance. Investigation into peripheral tissues in N2KO mice revealed that responses of genes in the hypothalamus-pituitary-thyroid axis, muscle, and brown and white adipose tissue to changes in energy availability require Nhlh2 expression. The responses of serum total T4 levels and UCP1 mRNA and UCP3 mRNA to energy availability signals are altered in N2KO mice. In addition, N2KO mice maintain body temperature with cold exposure but are unable to maintain body weight. In summary, my work provides new insight into the role of Nhlh2 in coordinating energy availability signals to downstream genes required for body weight maintenance and thermoregulation.

Immunocytochemistry was used to study the distribution of gamma-aminobutyric acid (GABA) throughout the central visual nuclei and retina in Rana pipiens. In the diencephalon, intensely-labeled GABA immunoreactive neurons and nerve fibers were observed within the neuropil of Bellonci (nB) and corpus geniculatum (CG), while only immunoreactive puncta were found in the rostral visual nucleus (RVN). In the pretectal region, the posterior thalamic nucleus (nPT) contained the most intensely-labeled GABA immunoreactive perikarya and nerve fibers in the entire brain. Lightly immunoreactive perikarya were also found in the large-celled nucleus lentiformis mesencephali (nLM), as well as in the pretectal gray which contains neurons postsynaptic to the retinal terminal zones within nLM. In the optic tectum (OT), both immunoreactive perikarya and fibers were found within superficial layers 8 and 9; whereas only densely-packed immunoreactive perikarya were evident in the deep tectal layers (i.e. 2, 4, 6). The nucleus of the basal optic root (nBOR) contained a small number of lightly-labeled GABA immunoreactive perikarya mostly located in the dorsal half of the nucleus. A large number of perikarya within the nucleus isthmi (NI) were also lightly immunostained. In the retina, GABA immunoreactivity (both somata and fibers) was observed in all layers except the outer nuclear layer (ONL). Besides GABAergic putative horizontal and amacrine cells in the inner nuclear layer (INL), about 30% of total neurons within ganglion cell layer (GCL) expressed GABA immunoreactivity. Double-labeling studies indicated that about half of the GABA-containing perikarya in the GCL were retinal ganglion cells (RGCs). In addition, three GABAergic projection pathways existing in the visual system of Rana pipiens were demonstrated: (1) from RGCs to the contralateral OT; (2) from nBOR to the pretectal nLM; and (3) bilaterally from the NI to the OT. These results indicate that GABA is an important neurotransmitter in the frog visual system.

The neuroanatomical distribution and the physiological role of the biogenic amine octopamine was investigated in the prosomal central nervous system (CNS) of the horseshoe crab, Limulus polyphemus. Perfusion of octopamine onto the isolated CNS elicited the rhythmic feeding motor pattern in the entocoxal motor nerves which innervate the feeding musculature. The specific effects of octopamine on the CNS and the pharmacology of its action are described. The stimulation of the feeding motor program at concentrations comparable to that of octopamine by the octopamine agonists NC-5, NC-7, and NC-13, the diterpene adenylate cyclase activator forskolin and the cyclic AMP analogue, 8-bromo cyclic AMP indicated that octopamine's effects may be mediated by a second messenger. The putative octopamine antagonists tested were ineffective or weakly blocked the effects of octopamine. The distribution and localization of octopamine in the prosomal CNS was determined by light and electron microscopic immunocytochemistry. Sixteen discrete clusters of octopamine-like immunoreactive (Oct-LIR) neurons were found in the circumesophageal ring (CER) of fused thoracic ganglia. The distribution of these immunoreactive cells are described. The immunoreactive somata are 40 to 100 $\mu$m in size and are found in clusters of 12-24 cells. There is extensive distribution of Oct-LIR nerve fibers in Limulus; the wide distribution of Oct-LIR provides an anatomical basis for the several effects of octopamine in Limulus. The subcellular localization of octopamine in Oct-LIR terminals in the CER was determined by postembedding immunoelectron microscopy with a 5 nm immunogold label. Labelled terminals are morphologically unique; they contain large, dense-core granules of a distinct shape, typically cylindrical with an indentation or depression in one end. These large granules are typically 100-150 nm in diameter and 150-400 nm in length. The dense labelling of these unusual granules with the gold particles indicates that octopamine is sequestered in or associated with these granules.