Understanding the factors that contribute to the origin and maintenance of species, and elucidating the mechanisms that influence species distribution across the landscape are two goals that are fundamental to evolutionary biology and ecology. I combined field and laboratory experiments, a robust phylogeny, and species distribution data from herbaria to test a series of hypotheses that address variation in the distribution of species, and, the evolution and maintenance of reproductive isolating barriers, key components for understanding speciation. I used species in the genus Collinisa to test the following main hypotheses: 1) An extension of Bakers Law: Among similar aged sister-taxa pairs, species more proficient at autonomous self-fertilization should be better colonizers and thus should have larger range sizes than their sister-taxa that are less proficient at autonomous selfing, 2) Bateson-Dobzhansky-Muller (BDM) model of reproductive isolation: Intrinsic postzygotic isolation barriers increase as divergence time increases among species, and 3) Reinforcement of reproductive isolation: early selfing can evolve in response to heterospecific pollen receipt, and may thus act as a prezygotic reproductive isolating barrier that is reinforced in sympatry. I found that species most proficient at selfing had significantly larger range sizes than their sister-taxa that were less proficient at selfing. Thus, mating system did explain differences in the ranges sizes of similar aged sister-taxa. To address the second hypothesis, I first tested for allopatric speciation in this genus, and found strong support for allopatric speciation in the California clade of Collinsia. But I did not find strong support overall, likely because of large range-shifts in the northeastern clade, which obscured the expected pattern of increasing range overlap with increasing divergence time. In support of the BDM model, I found that post-mating pre- and postzygotic isolation increased with increasing divergence time. And finally, in support of the final prediction, I found that when C. rattanii was sympatric with C. linearis it self-pollinated at a significantly earlier stage, suggesting that earlier selfing may be acting to reinforce of reproductive isolation in sympatry.

Bacteriophage tails, despite differences in their morphology, all play a key role in host recognition and DNA injection. It is widely believed that the tail, especially the baseplate / tail tip, has to undergo conformational change and protein rearrangement during infection. This change has been observed in both the long, contractile tail of T4 and the short tail of T7. In contrast, little is known about this aspect of the long, non-contractile tail of bacteriophage Ü. Four proteins are involved in the Ü tail tip assembly. I present evidence that gpJ, gpI, gpL and gpK are components of the tail tip complex. My results also suggest that there may be about three copies of gpI and three copies of gpL involved in the Ü tail assembly. In addition, I have successfully purified gpL, which contains eight cysteines. My results show that when the conserved cysteine at position 173, 182 or 205 is mutated to serine, the mutant protein is defective in tail assembly. However, the C212S mutant accumulates a small amount of tail. Further analysis of this mutant indicates that C212 may have roles in both tail assembly and DNA injection. gpK is required for Ü tail assembly, but is not detected in the mature virion. Two different amber mutations were introduced into gene K. Neither of these mutants is able to complement in vivo. However, the short amber fragment is unable to assemble Ü tail whereas phage-like particles with little infectivity accumulate in the long amber fragment lysate. The results indicate that the function of gene K can be bypassed to some extent in the Kam768 mutant, but not in the Kam6 mutant.

Heat shock proteins of 70kDa (Hsp70s) and their J domain-containing Hsp40 cofactors are conserved chaperone pairs that facilitate diverse cellular processes. One essential Hsp70 in the endoplasmic reticulum (ER) lumen, BiP (Kar2p in yeast), participates in polypeptide translocation into the ER, protein folding, and ER-associated degradation (ERAD). Like other Hsp70s, BiP contains an N-terminal ATPase domain, followed by a substrate binding domain and a C-terminal lid domain. To better define how substrate affinity and Hsp40 interaction affect BiP function, I constructed and characterized a mutation, R217A, in the putative J domain-interacting surface of yeast BiP. The mutation compromises ATPase stimulation by Sec63p, an Hsp40 required for translocation, but stimulation by Jem1p, an Hsp40 required for ERAD, is robust. In accordance with these data, yeast expressing R217A BiP exhibit translocation defects, but no ERAD defects, and a genetic interaction study using this mutant yielded data consistent with defects in translocation. In contrast, mutations in the substrate binding domain that either disrupt an ionic contact with the lid or remove this domain are deficient for peptide-stimulated ATPase activity. Expression of these mutants in yeast results in varying translocation and ERAD defects. Taken together, these data indicate that BiP can distinguish between its ER-resident cochaperones, and that optimal substrate binding is a key determinant of BiP function. Next, I tested the hypothesis that the functional specificity of Hsp70s is regulated by cognate Hsp40s. If this is true, one might expect divergent Hsp70-Hsp40 pairs to be unable to function in vivo. However, I discovered that a mammalian ER-lumenal Hsp40, ERdj3, when directed to the yeast cytosol, was able to rescue the temperature-sensitive growth phenotype of yeast containing mutant alleles in two cytosolic Hsp40s, HLJ1 and YDJ1. Moreover, ERdj3 activated the ATPase activity of Ssa1p, the yeast cytosolic Hsp70 that partners with Hlj1p and Ydj1p. Intriguingly, ERdj3 mutants that were compromised for substrate binding were unable to rescue the hlj1ydj1 growth defect, even though they stimulated Ssa1p ATPase activity. These data suggest that the substrate binding properties of certain Hsp40snot simply the formation of unique Hsp70-Hsp40 pairsis critical to specify in vivo function.

Alternative pre-mRNA splicing adjusts the transcriptional output of the genome by generating related mRNAs from a single primary transcript, thereby expanding protein diversity. A fundamental unanswered question is how splicing factors achieve specificity in the selection of target substrates despite the recognition of information-poor sequence motifs. The CUGBP2 splicing regulator plays a key role in the brain region-specific silencing of the NI exon of the NMDA R1 receptor. However, the sequence motifs utilized by this factor for specific target exon selection and the mechanism of splicing silencing are not understood. Here, I use chemical modification footprinting to map the contact sites of CUGBP2 to GU-rich motifs closely positioned at the boundaries of the branch sites of the NI exon, and demonstrate a mechanistic role for this specific arrangement of motifs for the regulation of branchpoint formation. General support for a branch site-perimeter-binding model is indicated by the identification of a group of novel target exons with a similar configuration of motifs that are silenced by CUGBP2. These results reveal an autoregulatory role for CUGBP2 as indicated by its direct interaction with functionally significant RNA motifs surrounding the branch sites upstream of exon 6 of the CUGBP2 transcript, itself. The perimeter-binding model explains how CUGBP2 can effectively embrace the branch site region to achieve the specificity needed for the selection of exon targets and the fine-tuning of alternative splicing patterns.

Escherichia coli bacteriophage HK97 assembles its mature capsid from 415 copies of the major capsid protein gp5. After assembly with maturation protease, the precursor capsid called prohead undergoes proteolysis and several stages of maturation including expansion and crosslinking to form a mature capsid. During capsid assembly and maturation, gp5 proteins assemble with each other and undergo refolding and conformational changes. Thus, some of the gp5 residues must play essential roles in the local network of intra- and inter- protein interactions, prohead assembly, and the different stages of capsid maturation. Here I establish a genetic suppressor approach to investigate the functional features of gp5 residues in vivo. By correlating the results from biochemical, structural, and genetic studies, specific protein interactions during HK97 capsid assembly and maturation have been determined. First I report that gp5 residue V163 facilitates HK97 head protein crosslinking, possibly by providing a local hydrophobic environment to promote crosslinking. I demonstrate that mutant gp5 V163D exhibits a specific defect in the head protein crosslinking reaction. Genetic results show that Val, Leu, Thr, Ile and Cys are tolerated at gp5 residue 163, suggesting a certain limited size and its hydrophobicity are important. Second, I propose that gp5 residues 231 and 178 play central and essential roles in HK97 capsid assembly by mediating assembly of proheads from capsomers. Single substitutions at residue 231 or 178, e.g. D231L, block the assembly of capsomers into proheads. Interestingly, the deficiency of the mutant substitution D231L can be suppressed by substitutions K178V, K178I, or K178N. Instead of the wild type pair of residues D231 and K178, alternative pairs, D231L/K178V, D231L/K178I, and D231L/K178N, yield viable phage HK97. Structural analysis confirms that this interaction between residues 231 and 178 is involved at the early HK97 prohead assembly from capsomers. Both cases underline that genetic suppressor studies offer useful in vivo data to determine the functional importance of the HK97 gp5 residues.

Protein degradation in C. elegans muscle cells is very tightly controlled by opposing signaling cascades. Excessive FGFR, low IGFR or TGF-B signaling, starvation or denervation all lead to muscle protein breakdown. Here we show that under conditions of activated FGFR, low IGFR or low TGF-B activity or mutationally hyperactived MAP kinase, soluble protein in muscle cytosol is degraded through autophagy. Degradation under these conditions is prevented by a reduction of function mutation in unc-51 (encodes Atg1 homologue), by RNAi knockdown of BEC-1 (beclin), or ATG-7 (Atg7 homologue) or by treatment with N6,N6-dimethyladenosine, a presumed inhibitor of type III PtdIns-3-kinase. We infer that the MAP kinase cascade positively regulates autophagy when the FGFR signal is too high or the IGFR signal too low . However, we have yet to identify how a MAPK signal plays into TGF-B regulation of autophagy. Protein degradation following starvation or disruption of cholinergic signaling (ACh-deficient cha-1 mutant) is not prevented by interference with autophagy, consistent with the known sensitivity to proteasome inhibitors. We have also implicated the signaling protein RSK as necessary for autophagic signaling either in parallel with or downstream of MAPK The metabolic reserve of protein in muscle can be mobilized by catabolism in response to failure of any one of a variety of signals of 'healthy' conditions. To achieve versatility and flexibility of response, muscle cells evidently use multiple proteolytic systems and regulate them independently by integrating more than one regulatory input to each degradation system.

Many complex mechanisms regulate cytoskeletal-dependant changes in cell morphology and behavior embryonic and adult life. The actin-binding protein Shroom 3 (Shrm3) appears to interact with Rho kinase (Rok) to direct the assembly of a contractile actomyosin network in neuroepithelial cells, causing apical constriction during neural tube formation. An ortholog of Shrm3 was identified, Drosophila Shroom (dShrm), that contains homology to the Shrm3 domain responsible for constriction. When properly targeted, the invertebrate domain also exhibits the ability to cause apical constriction, suggesting that the pathway is conserved in invertebrates. Drosophila melanogaster will likely provide a powerful model system to study the localization and function of Shrm during tissue morphogenesis. The project has three goals: to classify the endogenous expression of dShrm, to investigate the effects of over expressing dShrm in various tissues, and to determine a potential interaction between dShrm and dRok. There are two isoforms of dShrm we have primarily studied, dShrmA and dShrmB. Using in situ hybridization and immunohistochemistry, we have shown that endogenous dShrm proteins localizes to adherens junctions (dShrmA) and the apical plasma membranes (dShrmB) of cells in the ectoderm, trachea system, and other cell populations during embryogenesis. Over expression of dShrm in the ectoderm, as well as the eye and wing imaginal discs, causes dramatic defects in tissue architecture. Over expression of dShrmA with dRok has shown that there is an interaction between these two proteins. Further work will focus on defining the mechanism by which dShrm functions and which tissues require its activity during embyogenesis.

Disorazole C1 (DZ) is a synthesized polyene macrodiolide, originally identified as a minor metabolite from the myxobacterium fermentation broth. I examined the cellular responses to DZ in noncancer and cancer cells, and compared the results to vinblastine and taxol. In noncancer cells, DZ induced a prolonged mitotic arrest, followed by mitotic slippage, which was associated with cyclin B degradation, but did not require p53. Four apoptosis detection assays, including examination of poly(ADP-ribose) polymerase cleavage, cytochrome C release, mitochondrial depolarization, and annexin V staining, were conducted and demonstrated little induction of apoptosis in noncancer cells treated with DZ. However, I observed an activated apoptotic pathway in cancer cells, suggesting that normal and malignant cells respond differently to DZ, and indicating a potential therapeutic application for this compound. GpIba is a transmembrane subunit of the von Willebrand factor receptor on platelet surfaces, functioning in platelet adhesion and activation. Recent research has revealed that GpIba also plays roles in transformation and genomic instability as an oncoprotein. In my studies, GpIba colocalized with F-actin and filamin A, an actin-binding protein, at the cleavage furrow in anaphase cells, suggesting a novel role for GpIba in cytokinesis. However, when GpIba was overexpressed in noncancer cells with p53 knockdown, GpIba, F-actin, and filamin A became mislocalized. RhoA, a cytokinesis regulator, was localized asymmetrically, although Aurora B, an important cytokinesis kinase, retained its correct localization. Additionally, anaphase bridges were observed in these GpIba-overexpressing cells, along with elevated percentages of cytokinesis failure and binucleation, indicating that GpIba overexpression led to cytokinesis defects/failure and tetraploidization. Conversely, in cancer cells that endogenously overexpress GpIba, I observed decreases in the percentages of binucleation and mitotic defects upon GpIba knockdown. Furthermore, down-regulation of Aurora B was demonstrated to mediate the mechanism through which GpIba overexpression led to cytokinesis failure, and the filamin A-binding domain and the signal peptide of GpIba were shown to be indispensible for this mechanism. These results add a new component of the existing cancer cell genetic instability pathways, by suggesting that overexpression of GpIba in cancer cells disrupts normal cytokinesis and causes tetraploidization through down-regulation of Aurora B.

Understanding the mechanisms that lead to axial elongation in the mouse has direct relevance to elucidating the etiology of vertebral defects in humans. Through the characterization of a spontaneous mouse mutant, kinked tail, and the analysis of Tbx6 protein modulation in vivo, I uncovered two distinct mechanisms affecting axial elongation in the mouse. The kinked tail mutation is a spontaneous mutation, inherited dominantly, that results in a kinky tail phenotype in heterozygotes and early embryonic lethality in homozygotes. Defective axial elongation in kinked tail heterozygotes is displayed as shortened tails and multiple tail kinks resulting from wedge, hemi- and fused vertebrae, similar to those observed in scoliosis patients. These vertebral defects are likely due to a primary notochord defect that is thickened and branched. Kinked tail homozygotes fail to undergo gastrulation due to defective distal visceral endoderm cell migration, ultimately resulting in lethality by embryonic day 8.5. The defective cell migration is further compounded by basement membrane defects and gross dysmorphology of the mutant embryo. Tbx6, a T-box transcription factor, is essential for posterior somite formation, patterning and viability of the mouse embryo. I sought to understand Tbx6 protein regulation and the phenotypic consequences of modulating Tbx6 protein levels in vivo. In vitro analyses revealed that Tbx6 is a relatively stable protein that appears to be regulated in part by the proteasome in addition to other mechanisms. In vivo, less than heterozygous levels of Tbx6 protein results in rib and vertebral defects, enlarged tailbuds and axial shortening while greater than wildtype levels of Tbx6 protein results in small embryonic tailbuds, axial shortening, and lethality. I further examined the consequences of Tbx6 misexpression using a 3-component transgenic system. The primitive streak and presomitic mesoderm are affected in those embryos that misexpress Tbx6. Altogether, my analysis of the spontaneous mutation, kinked tail, demonstrated a possible role of the notochord in proper axial elongation, and the analysis of Tbx6 protein modulation further clarified the importance of maintaining proper levels of Tbx6 for normal axis elongation and embryonic development.

Recently, the biotic context for sexual and mating system evolution in plants has received special attention, yet the significance of interactions with antagonists has only begun to be revealed. We investigated the effect of florivory on reproduction and inbreeding depression by simulating damage on selfed and outcrossed progeny of hermaphrodites of Fragaria virginiana and recording the response of reproduction, as well as measuring tolerance to florivory. While both florivory and inbreeding affected reproduction, their effects were independent with respect to sexual traits but not an asexual trait; inbreeding depression was florivory- and family-dependent, specifically, for plantlet production. Plants were intolerant to florivory in terms of flowers, moderately tolerant in terms of fruit, and most tolerant in terms of plantlets. However, only under severe damage was intolerance statistically significant. Inbreeding did little to change these patterns. Our findings suggest that florivory does not consistently influence inbreeding depression, but its indirect effects on plantlet production could lead to increased geitonogamous selfing. This, combined with previous work demonstrating increased autogamous selfing following weevil damage and knowledge of the mechanism of sex determination in this system, suggests that damage by weevils could contribute to the maintenance of hermaphrodites in gynodioecious F. virginiana populations.