Human immunodeficiency virus type 1 (HIV-1) is a member of the lentivirus subfamily of retroviruses. HIV-1 expresses multiple genes from a single provirus by alternative splicing. Early in viral expression, fully spliced 2-kb viral RNA is exported from the nucleus and encodes the viral regulatory protein, Rev, which is essential for nuclear transport of partially spliced and unspliced genomic-length RNA. Rev binds to an RNA structural element called the Rev response element (RRE) and mediates nuclear export through the leucine-rich nuclear export signal (NES) pathway. The human Rev Interacting Protein (hRIP) interacts specifically with the Rev NES. Rev NES mutants that are unable to export Rev-dependent RNAs are also unable to bind to hRIP. The hRIP cDNA encodes a 562 amino acid protein containing an N-terminal zinc finger with homology to Arf GAP domains, a central serine and threonine rich region, and C-terminal phenylalanine-glycine (FG) repeats characteristic of nucleoporins. To identify an hRIP ortholog in a genetically tractable organism, we performed database searches using the N-terminal zinc finger of hRIP. Using this approach, we identified a novel gene in Schizosaccharomyces pombe. Alignment of the entire reading frame of the putative ortholog with hRIP indicates similarity with the serine/threonine rich region and with the FG repeats, suggesting that S. pombecould be a good model system to study the cellular function of hRIP. We find that the S. pombe ORF is an essential gene, which encodes a 483 amino acid protein that is also able to interact with the NES of HIV-1 Rev. Based on being an essential gene, and the presence of a putative Arf GAP domain, the ORF was named an Arf GAP essential for viability, agv1+. We show that Agv1 is not directly involved in the nuclear export of poly(A+) RNA or 5S rRNA, nuclear export of leucine-rich NES-containing proteins, or nuclear import of nuclear localization signal (NLS)-containing proteins. However, Agv1 does appear to play a role in the cytoplasmic localization of 5S rRNA. We demonstrate that loss of Agv1 alters the localization of endoplasmic reticulum (ER) membrane and Golgi membrane resident proteins, accumulates intracellular membrane, and blocks processing of carboxypeptidase Y. Furthermore, the S. cerevisiae ADP-ribosylation factor (Arf) GTPase activating protein (GAP) Glo3, but not a catalytically inactive Glo3 mutant [R59K], is able to partially compensate for the loss of Agv1 function in temperature sensitive strains, indicating that Agv1 is an S. pombe Arf GAP with some functional features similar to S. cerevisiae Glo3.

All living creatures change their gene expression program in response to nutrient availability and metabolic demands. Nutrients and metabolites can directly control transcription and activate second-­‐messenger systems. In bacteria, metabolites also affect post-­‐transcriptional regulatory mechanisms, but there are only a few isolated examples of this regulation in eukaryotes. Here, I present evidence that RNA-­‐binding by the stem cell translation regulator Musashi-­‐1 (MSI1) is allosterically inhibited by 18-­‐22 carbon ω-­‐9 monounsaturated fatty acids. The fatty acid binds to the N-­‐terminal RNA Recognition Motif (RRM) and induces a conformational change that prevents RNA association. Musashi proteins are critical for development of the brain, blood, and epithelium. I identify stearoyl-­‐CoA desaturase-­‐1 as a MSI1 target, revealing a feedback loop between ω-­‐9 fatty acid biosynthesis and MSI1 activity. To my knowledge, this is the first example of an RNA-­‐binding protein directly regulated by fatty acid. This finding may represent one of the first examples of a potentially broad network connecting metabolism with post-­‐transcriptional regulation.

Synaptic plasticity, in its broadest sense, can be defined as the ability of synapses to be modified structurally and functionally in response to various internal and external factors. Growing evidence has established that at the very core of these modifications are alterations in the cytoskeletal architecture. This discovery has led to the unearthing of a number of signaling pathways that might be involved in cytoskeletal regulation and also in the regulation of other aspects of synapse development and plasticity. In this regard, polarity proteins and secreted morphogens such as the Wnt proteins, typically involved in embryonic development, are emerging as critical determinants of synaptic growth and plasticity. However, their mechanism of action at synapses needs further investigation. Additionally, not much is known about how these morphogens are secreted or transported across synapses. Using the Drosophila larval NMJ as a model system, I have addressed aspects related to the issues mentioned above in the subsequent body of work. In the first half of my thesis, I have uncovered a role for the aPKC/Baz/Par-6 polarity protein complex in the regulation of the postsynaptic actin cytoskeleton in conjunction with the lipid and protein phosphatase PTEN. In the second half of my thesis, I have contributed to the elucidation of mechanisms underlying the secretion of Wg, the Drosophila Wnt homolog. Our findings suggest that Wnts might be secreted via a previously unidentified mechanism involving the release of exosome like vesicles from the presynapse and this process requires Evi/Wntless (Evi), a protein dedicated to Wnt secretion. Alterations in signaling pathways and aberrant cytoskeletal regulation lead to a variety of neurological disorders. The body of work in this thesis will provide a deeper understanding of the mechanisms involved in synaptic plasticity and provide a basis for uncovering similar pathways in the context of vertebrate synapses.

Every fall, Northeastern America monarch butterflies (Danaus plexippus) undergo an extraordinary migration to their overwintering site in Central Mexico. During their long migration, monarch migrants use sun compass to navigate. To maintain a southward flying direction, monarch migrants compensate for the continuously changing position of the sun by providing timing information to the compass using their circadian clock. Animal circadian clocks depend primarily on a negative transcriptional feedback loop to track time. I started my work to re-construct the monarch butterfly circadian clock negative feedback loop in cell culture, focusing on homologs of Drosophila clock genes. It turned out that in addition to a Drosophila-like cryptochrome (cry1) gene, a second mammalian-like cry2 gene exists in monarch butterflies and many other insects, except in Drosophila. The two CRYs showed distinct functions in our initial assays in cultured Drosophila Schneider 2 (S2) cells. CRY2 functions as a potent transcriptional repressor, while CRY1 is light sensitive but shows no obvious transcriptional activity. The existence of two cry genes in insects changed the Drosophila-centric view of insect circadian clock. During the course of my study, our lab obtained a monarch cell line called DpN1 cells. These cells possess a light-driven clock and contributed tremendously to the research on monarch circadian clock. Using this cell line, I provided strong evidence supporting monarch CRY2’s role as a major circadian clock repressor and identified a protein-protein protective interaction cascade underlying the CRY1-mediated resetting of the molecular oscillator in DpN1 cells. I continued my work trying to understand how insect CRY2 inhibits transcription. I provided evidence suggesting the involvement of monarch PER in promoting CRY2 nuclear entry in both S2 cells and DpN1 cells. Finally, I mapped CRY2’s transcriptional inhibitory activity onto its N-terminal domain. Collectively, my research helped to change our view of insect clocks from a Drosophila-centric standpoint to a much more diverse picture. My studies also advanced the understanding of monarch circadian clock mechanism, and provides a foundation for further studies.

The pancreas is located below the liver and adjacent to the small intestine where it connects to the duodenum. It consists of exocrine and endocrine components. The exocrine portion makes enzymes which are deposited in the duodenum to digest fats, proteins, and carbohydrates. Exocrine tissue also makes bicarbonates that neutralize stomach acids. The endocrine portion produces hormones such as insulin and glucagon which are released into the blood stream. These hormones regulate glucose transport into the body's cells and are crucial for energy production. The pancreas is associated with diseases such as cancer, diabetes, Annular pancreas and Nesidioblastosis. Annular pancreas and Nesidioblastosis are congenital malformations associated with excess endocrine tissue of the pancreas and its structures. Understanding the development of the pancreas might lead to insight of these diseases. The pancreas arises from the endoderm. In zebrafish, Nodal signaling activates mix-type and gata genes that then function together to regulate sox32 expression which is necessary and sufficient to induce endoderm formation. Interestingly, sox32 is exclusive to zebrafish and works synergistically with pou5f1 to regulate its own expression and turn on sox17 expression. sox17is evolutionarily conserved from zebrafish to mouse and is necessary for endoderm formation. Signals from within the endoderm and the surrounding mesoderm specify regions in the endoderm to develop into the pancreas and other endodermal organs. Sonic hedgehog (shh) expression in the foregut establishes the anterior boundary of the pancreas primordium while cdx4 expression establishes the posterior boundary, but what regulates these factors is unclear. We determined that two Three Amino Acid Loop Extension (TALE) homeodomain transcription cofactors, Meis3 and Pbx4, regulate shh expression in the anterior endoderm. Disrupting either meis3 or pbx4 reduces shh expression in the anterior endoderm. As a result, anterior ectopic insulin expression occurs outside the normal pancreatic domain. Therefore, we discovered upstream regulatory factors of shhexpression in the anterior endoderm, which is necessary for patterning the endoderm and pancreas primordium. We performed an ENU (N-ethyl-N-nitrosurea) haploid screen to look for endocrine pancreas mutants and to find other factors involved in pancreas development and patterning. From the screen, we characterized two mutants. We identified an aldh1a2 mutant, aldh1a2um22, which blocks the production of Retinoic Acid (RA) from vitamin A. While RA is known to be necessary for differentiation of the pancreas and liver, we also found it to be necessary for intestine differentiation. Two other aldh family genes exist in the zebrafish genome, but our data suggests that aldh1a2is the only Aldh that functions in endoderm differentiation and it is maternally deposited. From the screen, we discovered a second mutant, 835.4, that spontaneously arose within the background. pou5f1 expression is normal in mutant embryos, but sox32 expression is reduced and sox17 expression is lost. Downstream endoderm genes of sox17 are also lost and as a result no endodermal organs develop. Rescue experiments indicate that the mutation is located between sox32 and sox17 in the endoderm pathway. We currently have not been successful at mapping this mutation and therefore are unable to rule out the possibility that it lies in the sox17 gene. However, our data suggest that the mutation occurs in a new gene that is necessary for sox17 expression, potentially working with sox32 and/or pou5f1.

The most common structural platform in biology, the βα-repeat classes of proteins, are represented by the (βα)8TIM barrel topology and the α/β/α sandwich, CheY-like topology. Previous studies on the folding mechanisms of several members of these proteins have suggested that the initial event during refolding involves the formation of a kinetically trapped species that at least partially unfolds before the native conformation can be accessed. The simple topologies of these proteins are thought to permit access to locally folded regions that may coalesce in non-native ways to form stable interactions leading to misfolded intermediates. In a pair of TIM barrel proteins, αTS and sIGPS, it has been shown that the core of the off-pathway folding intermediates is comprised of locally connected clusters of isoleucine, leucine and valine (ILV) residues. These clusters of Branched Aliphatic Side Chains (BASiC) have the unique ability to very effectively prevent the penetration of water to the underlying hydrogen bond networks. This property retards hydrogen exchange with solvent, strengthening main chain hydrogen bonds and linking tertiary and secondary structure in a cooperative network of interactions. This property would also promote the rapid formation of collapsed species during refolding. From this viewpoint, the locally connected topology and the appropriate distribution of ILV residues in the sequence can modulate the energy landscapes of TIM barrel proteins. Another sequence determinant of protein stability that can significantly alter the structure and stability of TIM barrels is the long-range main chain-side chain hydrogen bond. Three of these interactions have been shown to form the molecular underpinnings for the cooperative access to the native state in αTS. Global analysis results presented in Chapter II and Chapter III, suggest that the off-pathway mechanism is common to three proteins of the CheY-like topology, namely CheY, NT-NtrC and Spo0F. These results are corroborated by Gō-simulations that are able to identify the minimal structure of kinetically trapped species during the refolding of CheY and Spo0F. The extent of transient, premature structure appears to correlate with the number of ILV side chains involved in a large sequence-local cluster that is formed between the central β-sheet and helices α2, α3 and α4. The failure of Gō-simulations to detect off-pathway species during the refolding of NT-NtrC may reflect the smaller number of ILV side chains in its corresponding hydrophobic cluster. In Chapter IV, comparison of the location of large ILV clusters with the hydrogen exchange protected regions in 19 proteins, suggest that clusters of BASiC residues are the primarily determinants of the stability cores of globular proteins. Although the location of the ILV clusters is sufficient to determine a majority of the protected amides in a protein structure, the extent of protection is over predicted by the ILV cluster method. The survey of 71 TIM barrel proteins presented in Chapter V, suggests that a specific type of long-range main chain-side chain hydrogen bond, termed “βα hairpin clamp” is a common feature in the βα-repeat proteins. The location and sequence patterns observed demonstrate an evolutionary signature of the βαβ modules that are the building blocks of several βα-repeat protein families. In summary, the work presented in this thesis recognizes the role of sequence in modulating the folding free energy landscapes of proteins. The formation of off-pathway folding intermediates in three CheY-like proteins and the differences in the proposed extent of structure formed in off-pathway intermediates of these three proteins, suggest that both topology and sequence play important and concerted roles in the folding of proteins. Locally connected ILV can clusters lead to off-pathway traps, whereas the formation of the productive folding path requires the development of long-range nativelike topological features to form the native state. The ability of ILV clusters to link secondary and tertiary structure formation enables them to be at the core of this cooperative folding process. Very good correlations between the locations of ILV clusters and both strong protection against exchange and the positions of folding nuclei for a variety of proteins reported in the literature support the generality of the BASiC hypothesis. Finally, the discovery of a novel pattern of H-bond interactions in the TIM barrel architecture, between the amide hydrogen of a core ILV residue with a polar side chain, bracketing βαβ modules, suggests a means for establishing cooperativity between different types of side chain interactions towards formation of the native structure. See Additional Files for copies of the source code for the global analysis program and the cluster analysis program.

Vesicular trafficking, the movement of vesicles between organelles and the plasma membrane for secretion, consists of multiple highly regulated processes. Many protein families function as specificity and regulatory determinants to ensure correct vesicle targeting and timing of trafficking events. The SNARE proteins dock and fuse vesicles to their target membranes. Sec1/Munc18 (SM) proteins regulate membrane fusion through interactions with the SNAREs—SM proteins have been shown to act as both inhibitors and stimulators of SNARE assembly and membrane fusion. However, the details of these SM protein functions are not understood. Constructing a model of SM protein function has been challenging due to the various modes of interactions reported between SM proteins and their SNAREs. SM proteins interact with their cognate SNAREs and SNARE complexes through several distinct modes. The most conserved mode is an interaction with the syntaxin N-peptide; other modes of binding, such as the syntaxin closed conformation, are hypothesized to be specific for specialized cell types. In order to elucidate the general function of SM proteins, I investigated the function of the endosomal SM protein Vps45p by analyzing its interactions with its cognate syntaxin Tlg2p and its role in SNARE assembly. I had two main hypotheses: that the Tlg2p N-peptide does not solely mediate the interaction between Vps45p and Tlg2p; and that Vps45p functions to stimulate SNARE complex assembly. I systematically mapped the interaction between Vps45p and Tlg2p using various Tlg2p truncations containing the different domains of Tlg2p and discovered a second binding site on Tlg2p that corresponds to the closed conformation. The neuronal SM-syntaxin pair interacts in a similar manner, indicating that this interaction mode is conserved. To characterize the closed conformation binding mode further, and determine its relationship to the N-peptide binding mode, I developed a quantitative fluorescent electrophoretic mobility shift assay. Results indicate that these two sites do not bind simultaneously and that the N-peptide binding modulates the closed conformation affinity. Furthermore, I monitored the effect of Vps45p on SNARE complex assembly using size exclusion chromatography. Under the conditions tested, Vps45p did not appear to stimulate SNARE complex assembly. The work presented here addresses several puzzling issues in the field and significantly contributes to the construction of a new mechanistic model for SM protein function. In this new model, the SM protein is recruited to the membrane by its interaction with the syntaxin N-peptide. The SM protein then binds the syntaxin closed conformation thus inhibiting SNARE complex assembly. Upon dissociation of the SM protein from the closed conformation, an event perhaps regulated by the SM protein, syntaxin opens and interacts with the other SNAREs to form a SNARE complex. Fusion ensues, stimulated by the SM protein.

It has become increasingly clear that transcription factors (TFs) play crucial roles in the development and day-to-day homeostasis that all biological systems experience. TFs target particular genes in a genome, at the appropriate place and time, to regulate their expression so as to elicit the most appropriate biological response from a cell or multicellular organism. TFs can often be grouped into families based on the presence of similar DNA binding domains, and these families are believed to have expanded and diverged throughout evolution by several rounds of gene duplication and mutation. The extent to which TFs within a family have functionally diverged, however, has remained unclear. We propose that systematic analysis of multiple aspects, or parameters, of TF functionality for entire families of TFs could provide clues as to how divergent paralogous TFs really are. We present here a multiparameter integrated network of the activity of the basic helix-loop-helix (bHLH) TFs from the nematode Caenorhabditis elegans. Our data, and the resulting network, indicate that several parameters of bHLH function contribute to their divergence and that many bHLH TFs and their associated parameters exhibit a wide range of connectivity in the network, some being uniquely associated to one another, whereas others are highly connected to multiple parameter associations. We find that 34 bHLH proteins dimerize to form 30 bHLH dimers, which are expressed in a wide range of tissues and cell types, particularly during the development of the nematode. These dimers bind to E-Box DNA sequences and E-Box-like sequences with specificity for nucleotides central to and flanking those E-Boxes and related sequences. Our integrated network is the first such network for a multicellular organism, describing the dimerization specificity, spatiotemporal expression patterns, and DNA binding specificities of an entire family of TFs. The network elucidates the state of bHLH TF divergence in C. elegans with respect to multiple functional parameters and suggests that each bHLH TF, despite many molecular similarities, is distinct from its family members. This functional distinction may indeed explain how TFs from a single family can acquire different biological functions despite descending from common genetic ancestry.

Skeletal muscle differentiation requires synergy between tissue-specific transcription factors, chromatin remodeling enzymes and the general transcription machinery. Here we demonstrate that two distinct protein arginine methyltransferases are required to complete the differentiation program. Prmt5 is a type II methyltransferase, symmetrically dimethylates histones H3 and H4 and has been shown to play a role in transcriptional repression. An additional member of the Prmt family, Carm1 is a type I methyltransferase, and asymmetrically methylates histone H3 and its substrate proteins. MyoD regulates the activation of the early class of skeletal muscle genes, which includes myogenin. Prmt5 was bound to and dimethylates H3R8 at the myogenin promoter in a differentiation-dependent fashion. When proteins levels of Prmt5 were reduced by antisense, disappearance of H3R8 dimethylation and Prmt5 binding was observed. Furthermore, binding of Brg1 to regulatory sequences of the myogenin promoter was abolished. All subsequent events relying on Brg1 function, such as chromatin remodeling and stable binding by muscle specific transcription factors such as MyoD, were eliminated. Robust association of Prmt5 and dimethylation of H3R8 at myogenin promoter sequences was observed in mouse satellite cells, the precursors of mature myofibers. Prmt5 binding and histone modification were observed to a lesser degree in mature myofibers. Therefore, these results indicate that Prmt5 is required for dimethylating histone at the myogenin locus during skeletal muscle differentiation in order to facilitate the binding of Brg1, the ATPase subunit of the chromatin remodeling complex SWI/SNF. Further exploration of the role of Prmt5 during the activation of the late class of muscle genes revealed that though Prmt5 is associated with and dimethylates histones at the regulatory elements of late muscle genes in tissue and in culture, it was dispensable for late gene activation. Previous reports had indicated that Carm1 was involved during late gene activation. We observed that Carm1 was bound to and responsible for dimethylating histones at late muscle gene promoters in tissue and in culture. In contrast to Prmt5, a complete knockout of Carm1 resulted in abrogation of late muscle gene activation. Furthermore, loss of Carm1 binding and dimethylated histones resulted in a disappearance of Brg1 binding and chromatin remodeling at late muscle gene loci. Time course chromatin immunoprecipitations revealed that Carm1 binding and histone dimethylation occurred concurrently with the onset of late gene activation. In vitro binding assays revealed that an interaction between Carm1, myogenin and Mef2D exists. These results demonstrate that Carm1 is recruited to the regulatory sequences of late muscle genes via its interaction with either myogenin or Mef2D and is responsible for dimethylates histones in order to facilitate the binding of Brg1. Therefore, these results indicate that during skeletal muscle differentiation, distinct roles exist for these Prmts such that Prmt5 is required for activation of early genes while Carm1 is essential for late gene induction.

There are two copies of each core histone in a nucleosome, however, it is unclear whether post-translational modifications on each molecule function redundantly or if symmetrical modifications are required to properly regulate gene expression. We tried to address this question by breaking nucleosomal symmetry and measuring its impact on gene expression. Our strategy includes re-engineering specific residues at the H3-H3 interface, generating pairs of mutant proteins, which were predicted by computational methods to form obligate heterodimers. Using S. cerevisiae as a model system, we tested the viability of strains with mutant histones, and analyzed the interaction between by co-immunoprecipitation from mononucleosome preparations. We also measured the changes of gene expression in the strains bearing single-tailed or tailless H3 heterodimers. The data suggested that the best computationally-derived H3 pair was frequently, but not exclusively heterodimeric in vivo. In order to obtain a more stringent H3 heterodimer, random mutagenesis was performed on four codons in the original computational design, and then genetic screening of the mutant libraries was performed.