Lysogeny is a defining feature of temperate bacteriophages. Temperate bacteriophages are able to establish lysogeny by integrating into the host chromosome or extrachromosomally, as a plasmid-like prophage in the host cytoplasm. In either case, host factors are involved in both the establishment and maintenance of lysogeny. Mycobacteriophage L5 forms an integrated prophage through the action of a phage-encoded tyrosine recombinase. L5 integrase (Int) binds to the phage attachment site, attP, and the bacterial attachment site, attB. Int binds attP bivalently by binding to the core, where strand exchange occurs, and to arm-type binding sites that flank the core. A host-encoded DNA binding protein, mIHF, is required for recombination and binds between the core and arm-type binding sites of attP. mIHF is thought to catalyze the bending of attP required for bivalent binding between the core and arm-type binding sites. attP core consists of a seven base pair overlap region flanked on either side by imperfect inverted repeats that make up the recombinase binding elements or RBEs. The RBEs were shown to be essential for core binding by creating attPs with mutations in the RBEs. When both of the RBEs are mutated Int can neither perform recombination, nor create specific complexes that involve core binding. When the right side RBE is wild type with mutant left side RBE, recombination can occur, and complexes involving core binding are observed. However, Int is unable to catalyze recombination when the left side RBE is wild type and the right is mutated, indicating that contact must be made with attP on the right side of core for a stable Int/attP core complex to be formed. The middle domain connects the two outer domains, but its function in recombination is not well understood. To determine the function of the middle domain, mutations were made at conserved residues in the middle domain of Int, and these mutants were characterized. These mutants are able to catalyze recombination, but do not form the recombinagenic complexes observed by wild type Int. Increasing the mIHF concentration in these reactions augments the recombination efficiency and stabilizes recombinagenic complexes that involve Int/core binding.

DNA methylation in mammals is involved in several essential processes including X chromosome inactivation, genomic imprinting, and host defense against mobile genetic elements. How methylation is targeted to specific sequences in the germ line and how specific methylation patterns are maintained during development is not fully understood. Genomic methylation is established in the gamete, decreases dramatically during preimplantation development, and is re-established after implantation of the blastocyst. However, the methylation present at imprinted loci is specifically maintained during preimplantation development. Imprinted genes are located in clusters, and within each gene cluster parent-of-origin specific expression is governed by an imprinting center (IC). The ICs of the maternally imprinted murine Snrpn, Kcnq1, and Igf2r gene clusters coincide with their differentially methylated domains (DMDs). We have shown that specific DMD sequences are required to establish differential methylation at an imprinted locus. Hybrid transgenes were generated using a non-imprinted derivative of the imprinted RSVIgmyc mouse transgene, Ig/myc, and sequences from endogenous imprinted gene DMDs. Addition of specific DMD sequences to the Ig/myc transgene restored its imprinting. Only the tandem repeats found within the Snrpn, Kcnq1, and Igf2r DMDs were capable of establishing maternal-specific transgene methylation. These experiments have also shown that the methylation on imprinted gene DMD sequences is specifically maintained during the early stages of preimplantation development. These results clearly demonstrate the importance of tandem repeats in the process of genomic imprinting. DNA methylation is also critical for silencing intracisternal A particle (IAP) transposition within the genome. It is thought that maintenance of IAP element methylation during preimplantation is critical to repress IAP element transcription and transposition. The methylation of IAP element long terminal repeat (LTR) sequences was analyzed at the blastocyst stage of preimplantation development using the bisulfite genomic sequencing technique. These experiments have shown that methylation is maintained on the majority of IAP elements at the blastocyst stage of preimplantation development. However, the methylation on specific IAP elements is completely lost at this time, including the methylation of single IAP element LTRs.

The site-specific recombination systems of bacteriophages and other mobile elements fall into two categories that are named for the integrase protein that catalyzes integration and excision of the phage genome. These integrases utilize either a tyrosine or a serine residue to carry out the nucleophilic attack of the DNA. In many cases the directionality of the reaction, that is whether the enzyme catalyzes integration or excision, is determined by an additional phage encoded protein referred to as RDF, or recombination directionality factor. Two prophage-like elements, phiRv1 and phiRv2, were found through sequencing Mycobacterium tuberculosis strain H37Rv. These are absent from the vaccine bacillus M. bovis BCG, and are often found in virulent strains of M. bovis and M. tuberculosis. Through the work presented here, one of these elements, phiRv1, was found to encode an active recombination system, with a serine integrase and RDF. The phiRv1 element is found within a degenerate repeated element, REP13E12, which is present in seven non-identical copies in M. tuberculosis and M. bovis BCG. In vivo studies have revealed that four of the seven 13E12 elements can serve as integration sites (attBs) for a plasmid carrying a reconstructed attP and integrase, and that multiple integrations can occur. The fast growing saprophyte, M. smegmatis, also supports integration, although inefficiently. In M. smegmatis, the phiRv1 plasmid integrates into at least two 13E12 repeats that are quite different from those found in M. bovis BCG and M. tuberculosis. Inefficiency is overcome by providing M. smegmatis with an attB site from BCG. These integrated plasmids are stable, and excision occurs in the presence of the phiRv1 RDF encoded by Rv1584c. In vitro assays were developed for both integration and excision. All the substrate site requirements are relatively small. Integration occurs slowly but efficiently in the presence of excess attB or on an intramolecular attP-attB plasmid. In the presence of RDF integration is inhibited and excision is stimulated. The RDF binds to a specific sequence in both attB and attL, although the way in which it functions may be through protein-protein interactions with integrase.

This thesis explores the theory that gap junctions play a role in adrenocortical homeostasis and function. To test this hypothesis, gap junction mediated communication was investigated in the intact adrenal gland. In addition, to determine the impact of altered adrenal trophic state on gap junction distribution and expression, adrenal glands from hypophysectomized mice as well as normal and neoplastic human adrenal tissues were evaluated. Gap junction mediated intercellular communication was done in primary mouse adrenal glands using a modified Lucifer Yellow dye communication assay. Dye communication showed that the adrenal cortex was communication competent and that functional cell-cell communication was directly related to the abundance of connexin 43 expression. There was an absence of dye communication in the outer cortical zona glomerulosa (ZG) that was contrasted by a high level of communication in the inner zones of the zonae fasciculata/reticularis (ZF/ZR). The removal of the endogenous pituitary derived ACTH stimulus by hypophysectomy resulted in a significant loss of connexin 43 expression in the ZF/ZR that was associated with a diminished trophic state. The loss of connexin 43 expression was concurrent with features associated with a diminished functional state of the gland including loss of lipid droplets and reduced expression of the ZG specific cytochrome P450 aldosterone synthase (P450 aldo). A comparison of adrenal glands from connexin 43 deficient mice (Cx 43 -/-) demonstrated several features that were similar to those observed in hypophysectomized mice, including glandular atrophy and reduced P450 aldo expression in the ZG. Connexin 43 expression and distribution in the normal human adrenal gland was identical to that previously described in the adrenal glands from other mammals. Connexin 43 expression was abundant in the ZF/ZR and nearly absent from the parenchymal cells of the ZG. Immunocytochemial analysis of neoplastic tissues showed that benign (adenoma) and malignant (carcinoma) neoplasms lost 30% and 90% respectively, of their connexin 43 expression in the ZF/ZR when compared to normal tissues. The data presented here demonstrate that connexin 43 expression is related to physiological changes that affect adrenal trophic state and suggest that gap junction mediated cell-cell communication is involved in the function of the adrenal cortex. Furthermore, because connexin 43 deficient animals also exhibit glandular changes consistent with a diminished state of the adrenal gland, it supports a role for proper gap junction expression in adrenal cortex development and function.

In vertebrates, alternative splicing regulates gene expression in a cell- and developmental stage-specific manner, producing functional diversity of the corresponding proteins. Although its importance has been underscored, control mechanisms of alternative splicing are poorly understood. The N-methyl-D-aspartate receptor NR1 subunit (NMDAR1, GRIN1) contains three cassette exons (NI, CI and CII), and which are specifically expressed in mammalian brain. In this study, a minigene splicing reporter system of the CI cassette exon was utilized to study its splicing silencing mechanism in mammalian cell lines. This work focuses primarily on the identification of exonic UAGG motifs and a 5 splice site region G cluster. Individual motifs play a silencing role but the combination of all three is required for strong silencing. Whereas insertion of an extra UAGG motif in the exon shows almost complete silencing, removal of all three silencers results in loss of silencing. Therefore, the UAGG and G cluster motifs confer flexibility of control of CI cassette exon splicing and the number of the motifs determines the silencing strength in various tissues. The UAGG motif interacts with hnRNP A1 and the strong silencing of hnRNP A1 requires the G cluster. The splicing enhancing role of hnRNP F or H/H is identified, and this effect largely requires the G cluster. Using bioinformatics searches, new groups of skipped exons containing the UAGG and GGGG (as a G cluster) motifs are identified from the human and mouse genomes. This study suggests that hnRNP H1 and H3 (HNRPH1 and HNRPH3) may be auto-regulated at the level of splicing. Overall, this work provides evidence for a splicing silencing mechanism that is important for the tissue-specificity of the CI cassette exon. This work also shows that the motif pattern can be used computationally to identify additional skipped exons that contain combinations of UAGG and GGGG motifs.

Invasive species pose an enormous threat to native species and imposes substantial costs on the US economy. Although the threat of exotic species is well recognized, the general ecological mechanisms that underlie these invasions remain elusive. Predictions both for invasion success and the success of biological control remain poor. In this dissertation, I harness plant competition theory to predict the success of invasions and biocontrol, using a model system composed of invasive purple loosestrife (Lythrum salicaria), native broad-leaved cattail (Typha latifolia), and Galerucella calmariensis, a leaf-feeding beetle widely released to control loosestrife. In chapter 1, I introduce the problem and summarize my results. In chapter 2, I extend resource competition theory to competition for light: species coexistence is possible if one species is sufficiently taller and with less dense foliage than its competitor. This integrated model of competition for light is easily parameterized through measurements of light availability in monoculture and thus can easily be tested in the field. In chapter 3, I test the ability pf three models of plant competition (response to resource availability, plant size, and resource reduction) to predict competitive outcomes between loosestrife and cattail. My experimental design included monoculture mesocosms in which to measure plant traits, and mixture mesocosms in which to determine competitive outcomes. Surprisingly, while loosestrife was, on average, negatively affected by the presence of cattail, cattail was not, on average, negatively affected by loosestrife. Indeed, at high fertility, cattail was strongly negatively affected by loosestrife in the absence of insect herbivores of both species, yet was strongly facilitated when herbivores of both species were present. The facilitation of cattail by loosestrife was likely due to density dependent predation by cattails natural enemies. Cattail abundance in mixture was not predicted by any of the three models, which is not surprising considering the lack of an overall competitive effect of loosestrife on cattail. In contrast, loosestrife abundance in mixture was well predicted by species height, as predicted by the plant size model. These results suggest that competitive traits may predict invasion success and biocontrol, but only when species interact only through competition for resources.

The L1 Ta subfamily of Long INterspersed Elements (LINEs) consists exclusively of human-specific L1 elements with a copy number of ~520 in the human genome. Four hundred sixty-eight L1 Ta elements were extracted from the draft human genomic sequence and screened by polymerase chain reaction (PCR) assays to determine their phylogenetic origin and to determine their contribution to human genomic diversity. PCR analysis indicated that 45% of the L1 Ta elements screened are polymorphic for insertion presence or absence. Sequence analysis of the L1 Ta elements produced definitive evidence of 3transduction, gene conversion involving an older pre-existing L1 element, as well as several potential retrotransposition competent elements. The average age of the L1 Ta subfamily was estimated at 1.99 million years, indicating the subfamilys expansion subsequent to the divergence of humans from African apes. PCR based screening in non-human primate genomes of the orthologous sites for 249 human L1 Ta elements resulted in the recovery of various types of sequence variants for ~12% of these loci. Sequence analysis was employed to capture the nature of the observed variation. Half of the orthologous loci differed from the predicted sizes due to localized sequence variants that occurred as a result of common mutational processes in ancestral sequences, often including regions containing simple sequence repeats. Additional sequence variation included genomic deletions that occurred upon L1 insertion, as well as successive mobile element insertions that accumulated within a single locus over evolutionary time. We estimate the overall frequency of parallel independent insertion events at L1 insertion sites in seven different primate species to be very low (0.52%). In addition, no cases of insertion site homoplasy involved the integration of a second L1 element at any of the loci, but rather largely involved secondary insertions of Alu elements. No independent mobile element insertion events were found at orthologous loci in the human and chimpanzee genomes. Therefore, L1 insertion polymorphisms appear to be essentially homoplasy free characters well-suited for the study of population genetics and phylogenetic relationships within closely related species.

The ghost shrimp, Lepidophthalmus louisianensis, burrows up to meters deep in oxygen-limited marine sediments along the Gulf coast. During low tides these animals are subjected to extended periods of anoxia. The main objective of this study was to assess survival under anoxia and evaluate the physiological mechanisms that underlie the anoxia tolerance of this species. I observed large specimens of L. louisianensis (>2g) having an LT₅₀ of 64 h under anoxia at 25º C. Smaller specimens (<1g) have a significantly higher LT₅₀ of 113 h under identical conditions (P<0.0001). I measured whole body lactate levels in shrimp exposed to anoxia for up to 72 h, and recorded significant accumulation of this anaerobic end product (ANOVA, P<0.001). I also measured adenylates and arginine phosphate in shrimp exposed to anoxia for up to 48 h, and after a 24-h recovery period. Adenylates were not significantly altered during the anoxia regime, and reductions in arginine phosphate occurred after 12 and 24 h, but returned to normoxic values during recovery (ANOVA, P<0.001). While reserves of arginine phosphate are used to some extent to buffer losses in ATP, substantial contribution to the maintenance of energetic status comes from the high rate of anaerobic glycolysis. Energized mitochondria isolated from ghost shrimp hepatopancreas possess a pronounced ability to take up exogenous Ca²⁺ (compared to mitochondria-free controls) as measured by following the external free Ca²⁺ concentration with the fluorogenic dye Fluo-5N. Importantly, Ca²⁺ was not released from the mitochondrial matrix at any level of exogenous Ca²⁺ tested (up to 1.0 mM, in the presence of 5 mM phosphate). Thus, Ca²⁺ does not stimulate opening of the mitochondrial permeability transition pore, which potentially could help prevent apoptotic and necrotic cell death during extended periods of anoxia. (Supported by NIH grant 1-RO1-GM0-71345-01 and by SIGMA XI GIAR).

Reinforcement occurs when natural selection strengthens behavioral discrimination to prevent costly interpopulation matings, such as when matings produce sterile hybrids. This evolutionary process can complete speciation, thereby providing a direct link between Darwins theory of natural selection and the origin of new species. My dissertation presents the first study on the genetics of reinforcement. This study is framed in a conceptual body that explains how genomic architecture, selection and recombination, interact to facilitate divergence in the presence of gene flow. In addition, in my dissertation I produced a dense recombination map for D. pseudoobscura, which together with the genome sequence opens many possibilities for classic population genetic and genomic analyses in this system. I examine a case of speciation by reinforcement in Drosophila. I present the first high-resolution genetic study of variation within species for female mating discrimination that is enhanced by natural selection. I show that reinforced mating discrimination is inherited as a dominant trait, exhibits variability within species, and may be influenced by a known set of candidate genes involved in olfaction. My results show that the genetics of reinforced mating discrimination is different from the genetics of mating discrimination between species, suggesting that overall mating discrimination might be a composite phenomenon, which in Drosophila could involve both auditory and olfactory cues. Examining the genetics of reinforcement provides a unique opportunity for both understanding the origin of new species in the face of gene flow and identifying the genetic basis of adaptive female species preferences, two major gaps in our understanding of speciation.

Studies of speciation in the marine environment have historically compared broad-scale distributions and presumed larval dispersal to infer the geographic barriers responsible for allopatric speciation. However, many marine clades show high species diversity in geographically restricted areas where barriers are not obvious and larval dispersal should bring sister taxa into contact. Genetic differentiation at spatial scales <1000 km could facilitate speciation by mechanisms other than the gradual accumulation of reproductive isolation during extended allopatry, such as ecological adaptation to local environmental conditions or the rapid evolution of genes tied to mate recognition. The role of each of these possibilities has not been simultaneously explored for any species-rich marine taxon. The most species-rich genus of Neotropical reef fishes is Elacatinus (Gobiidae), with 27 species. I examine potential mechanisms underlying this richness through analyses of three genetic markers to investigate genetic and ecological differentiation between closely related taxa and among island populations. Phylogenetic analyses indicate that sister taxa of Elacatinus occur within the same oceans but occupy geographically separate ranges. Sister taxa usually differ by coloration, and distantly related sympatric species frequently differ by habitat. These differences suggest that some combination of coloration and ecological differences may facilitate assortative mating in sympatry or at range boundaries. The ranges of several Elacatinus taxa adjoin at Mona Passage and in the central Bahamas, both in the Caribbean Sea. These boundaries separate island populations by as few as 23 km, yet these populations are genetically distinct. Populations not separated by these breaks also show strong genetic structuring. Coalescent analyses suggest these populations have been demographically closed for up to 800,000 years. Such strong genetic structuring suggests that pelagic larvae are retained at natal populations, despite a three week larval duration (determined from otolith growth rings). My results suggest that local retention of pelagic larvae, coupled with biogeographic breaks, has generated or maintained strong genetic population structure which may facilitate adaptation to local ecological conditions. Such adaptations may explain observed divergence along ecological and coloration gradients. Repeated radiations among allopatrically distributed sister taxa may explain much of the high diversity in Elacatinus.