Regulation of gene expression is critical to many aspects of life. From cell survival and proliferation to animal development and species propagation, improper gene regulation can have serious, often fatal, consequences. Therefore, understanding the processes that control gene expression can provide important biological insights. At the center of many of these regulatory processes are trans-acting proteins called transcription factors. Most transcription factors contain DNA-binding domains that recognize specific DNA sequences. These site-specific transcription factors target genes by recognizing binding sites in regulatory sequences called cis-regulatory modules (CRMs). However, many transcription factors recognize degenerate DNA-sequences that can be found frequently throughout the genome. Despite this potential for promiscuity, transcription factors control very specific in vivo functions. This "specificity paradox" is best understood in the context of one particular family of transcription factors: the Homeobox (Hox) proteins. Conserved in all bilaterians, Hox genes are best known for their roles in embryonic pattering and organogenesis. Characterized by a highly conserved DNA-binding domain called the homeodomain, all Hox proteins recognize similar `AT' rich sequences. One way Hox proteins achieve functional specificity is through cooperative DNA-binding with the cofactor Extradenticle (Exd) in invertebrates or Pbx in vertebrates. Using Drosophila melanogaster as a model system we conducted a structure-function analysis of three different Hox proteins, Sex combs reduced (Scr), Ultrabithorax(Ubx) and AbdominalA (AbdA) to understand how interactions with a shared cofactor can increase specificity. To identify amino acid sequence motifs that contribute to Exd-dependent functions, we generated and tested a series of mutant Hox proteins for cooperative DNA-binding ability in vitro, and for their ability to regulate target genes in vivo. The results of these studies demonstrate that while Scr uses a single conserved motif, more posteriorly expressed Hox proteins Ubx and AbdA use multiple, sometimes unique motifs to regulate Exd-dependent functions. This discrepancy between the quantity and quality of motifs endows AbdA with the ability to outcompete Scr for DNA-binding and regulation of an Exd-dependent target. In addition, by testing the ability for AbdA mutants to carry out a variety of in vivo functions, we observed that the different modes of interaction with Exd affect functional specificity. However, in the case of Ubx, we find that despite the contribution of Exd-interaction motifs to cooperative complex formation in vitro, none of these motifs are required individually or in combination for in vivo functions. Together, these data suggest that one technique Hox proteins use to differentiate themselves when interacting with a shared cofactor is through the utilization of different interaction motifs. Furthermore, having multiple modes of interaction can expand and alter their functional specificity. However, as illustrated by Ubx, the functional interactions between Hox proteins and cofactors can be more complex and may not require cooperative DNA-binding. In conclusion, the characterization of Hox-cofactor interactions helps us better understand how transcription factors select their targets and regulate gene expression in a highly specific manner.

Hedgehog (Hh) family proteins specify many cell fates in flies and mammals that depend on Hh concentration. Hh achieves differential expression of target genes by regulating the activity and stability of Cubitus interruptus (Ci) / Glioblastoma (Ci / Gli) transcription factors. Both modes of regulation require kinesin-like protein Costal2 (Cos2) (Kif7 in mammals) that acts in a complex with Fused (Fu) kinase. We used a group of fused (fu) alleles with truncations in the Fu-regulatory domain to confirm the importance of physical association between Cos2 and Fu for processing of Ci-155 to Ci-75 repressor in the absence of Hh. By specifically disrupting Cos2 interaction with Fu using Cos2deltaFu transgenes, we confirmed the importance of this interaction for stability of the Fu protein and for Ci processing when the transgene was expressed at natural levels (under control of genomic regulatory sequences). It is possible that Cos2/Fu interaction helps recruit processing-promoting kinases to Ci. We found no evidence for the importance of Suppressor of fused (Su(fu)), a protein known to limit Ci activity, in regulating Ci processing. Using a group of Cos2 variants under the control of genomic sequences, we found that expression of the wild type genomic Cos2 transgene (gCos2) or gCos2 deficient in Fu-dependent phosphorylation of Ser572 and Ser931 rescued animals to adulthood. Contrary to past observations of over-expressed phosphorylation variants, these gCos2 transgenes supported normal responses to Hh at the anterior-poster (AP) border of wing discs, a region of active Hh signaling. As shown previously, Ser572 had a role in stabilization of Ci-155 by activated Fu kinase. gCos2deltaFu greatly reduced the response to Hh and was also unable to rescue animal viability. The transgene allowed for induction of ptc-lacZ in response to activated Fu (GAP-Fu) but did not respond to activated Smo (SmoD123), suggesting that it fails to mediate Fu activation in response to activated Smo . Normal accumulation of Smo in the regions of active Hh signaling in cos2-null wing discs expressing Cos2deltaFu also suggested that Cos2deltaFu does not affect the accumulation component of Smo activation. We propose that Cos2deltaFu is defective in promoting Fu activation through cross-phosphorylation of Fu molecules that requires physical interaction with Cos2. Similarly, a gCos2-S182N variant with impaired binding to Ci failed to rescue animal viability and was defective in Ci processing, but gCos2-S182N supported almost normal responses to Hh at the AP border. Ectopic induction of ptc-lacZ in anterior cos2 clones with gCos2-S182N indicated a defect in limiting Ci activity possibly through physical interaction.

An important characteristic of multicellular organisms is the control that the tissue architecture exerts on the fate of individual cells. Epithelial cells sense their location through interactions with the extracellular matrix (ECM) and remove themselves by programmed cell death (anoikis) when those interactions are disturbed. Importantly, anoikis is a line of defense that has to be circumvented by cancerous epithelial cells to be able to leave their home environment and establish long distance metastases. Here, by combining a genome-wide RNAi screen and a novel algorithm to study copy number alterations (ISAR-DEL), we identify the BridgingIntegrator3 (BIN3) as a novel 8p21 tumor suppressor gene whose inactivation promotes escape from anoikis in epithelial cancers. Mechanistically, we link the tumor suppression function of BIN3 to its ability to relocate to the cell membrane after cell detachment and to induce a proapoptotic cascade. This death signaling is mediated by CDC42 activation of the P38α stress pathway and the consequent accumulation of the apoptotic facilitator BimEL. Our results identify BIN3 as a novel epithelial tumor suppressor gene, provide novel insights on the mechanisms of attachment tumor suppressor checkpoint and highlight the importance of anoikis escape in driving cell transformation and metastasis in human cancer.

The Drosophila eye periphery undergoes peripheral patterning in response to a graded Wingless signal emanating from the surrounding head capsule. High levels of Wg signaling lead to the formation of the Pigment Rim. The pigment rim is a thick band of pigment cells that serves to optically insulate the eye from extraneous light rays. It is composed mainly of the pigment cells that surrounded the outermost row of ommatidia in the developing pupal eye. These peripheral ommatidia undergo timed developmental apoptosis, leaving the remaining pigment cells to coalesce and form the pigment rim. Earlier work showed that high levels of Wingless signaling induced the expression of Escargot, Wingless and Notum in a subset of the cells of the peripheral ommatidia, namely the cone cells. But the mechanism of apoptosis of the entire ommatidia remained unclear. My work focuses on the mechanism by which Wingless leads to the apoptosis of the different cell types of the ommatidia in a concerted manner. In this thesis, I show that the peripheral apoptosis follows a precisely timed sequence of events. I also show that ectopic expression of Wingless at high levels causes the entire eye to respond in a manner similar to the peripheral ommatidia. In order to elucidate the mechanism of Wingless induced apoptosis, I analyzed the effects of manipulations of the Wingless signaling pathway in the subsets of the cells of the ommatidia. I found that the expression of Escargot in the cone cells is required for their collapse, while the Wingless expression appears to be a booster signal for the apoptosis of the remaining cells of the ommatidia. I also show that the activation of Wingless signaling in the cone cells alone is insufficient for apoptosis of the ommatidia, thereby suggesting a combinatorial response of all the cell types. Lastly, I present a logical conundrum in the response of the photoreceptors to manipulations in Wingless signaling. In conclusion, I present a possible model of a concerted response of the different cell types of the ommatidia to lead to their apoptosis.

Both parathyroid hormone (PTH) and the sympathetic tone promote Rankl expression in osteoblasts and osteoclast differentiation by enhancing cAMP production, through an unidentified transcription factor for PTH and ATF4 for the sympathetic tone. How two extracellular cues using the same second messenger in the same cell elicit different transcriptional events is unknown. Here we show that PTH favors Rankl expression by triggering the ubiquitination of HDAC4, a class II histone deacetylase, partly via Smurf2. HDAC4 degradation releases MEF2c that transactivates the Rankl promoter. On the other hand, sympathetic signaling in osteoblasts favors the accumulation of HDAC4 and its association with ATF4. In this setting, HDAC4 increases Rankl expression. Through this interaction with ATF4, HDAC4 also influences Osteocalcin expression, and its endocrine and cognitive functions. This study shows that through its ability to differently connect distinct extracellular cues to their genome, HDAC4 is a global regulator of osteoblast functions.

Both parathyroid hormone (PTH) and the sympathetic tone promote Rankl expression in osteoblasts and osteoclast differentiation by enhancing cAMP production, through an unidentified transcription factor for PTH and ATF4 for the sympathetic tone. How two extracellular cues using the same second messenger in the same cell elicit different transcriptional events is unknown. Here we show that PTH favors Rankl expression by triggering the ubiquitination of HDAC4, a class II histone deacetylase, partly via Smurf2. HDAC4 degradation releases MEF2c that transactivates the Rankl promoter. On the other hand, sympathetic signaling in osteoblasts favors the accumulation of HDAC4 and its association with ATF4. In this setting, HDAC4 increases Rankl expression. Through this interaction with ATF4, HDAC4 also influences Osteocalcin expression, and its endocrine and cognitive functions. This study shows that through its ability to differently connect distinct extracellular cues to their genome, HDAC4 is a global regulator of osteoblast functions.

During animal development, cells are born with the potential to adopt different cell fates. Many of the mechanisms that control cell fate choice are highly conserved. Here I study how the fates of two important cell types, the Anchor Cell (AC) and the Distal Tip Cell (DTC), are specified during C. elegans gonadogenesis. The AC is an important signaling hub that directs uterine and vulval development. Its specification and function require the E protein HLH-2/Da/E2A, an essential basic Helix-Loop-Helix transcription factor. In the developing gonad, two cells Z1.ppp and Z4.aaa have naturally variable fates. They require hlh-2 for the potential to adopt AC fate, for the AC/VU decision to specify one AC and one ventral uterine precursor cell (VU), and for execution of AC fate. hlh-2 has been shown to be post-transcriptionally regulated during the AC/VU decision: HLH-2 protein is observed in the AC and not in the VU, though hlh-2 is transcribed in both cells. Until this work, little was known about how hlh-2 is post-transcriptionally regulated in C. elegans. Understanding how E proteins are regulated is also important in other contexts, because increased E protein activity has been associated with developmental defects and lymphoma. Here, a novel dimerization-dependent mechanism for HLH-2 down-regulation is identified in the VU. I provide evidence that HLH-2 homodimers promote AC competence, the AC/VU decision, and AC function, and that HLH-2 homodimers are recognized in the VU and targeted for degradation. The human ortholog E2A is found to be regulated similarly, raising the possibility that the mechanism for negative regulation of HLH-2 and E2A is conserved. A simple model could explain the difference in stability of HLH-2 homodimers: lin-12/Notch activity is low in the AC but high in the VU, where it promotes the turnover of HLH-2 homodimers. The C. elegans gonad also contains two distal tip cells (DTC), which direct the shape of the developing gonad and promote germline proliferation. In addition to AC fate, HLH-2 is required for the specification and function of the DTCs. However, it was not known what dimerizes with HLH-2 to promote DTC specification. Here, LIN-32/Atonal and HLH-12 are identified as two functionally redundant partners for HLH-2 in promoting DTC fate and function. Loss of both lin-32 and hlh-12 causes a complete failure of DTC migration, which likely reflects a highly penetrant failure of DTC specification. lin-32 and hlh-12 are both expressed in the DTC around the time of specification, consistent with a cell-autonomous role. These results suggest that LIN-32 and HLH-12 can heterodimerize with HLH-2 in the DTC to specify fate and promote migration. This work advances our understanding of how HLH-2 is regulated and how it functions with different dimerization partners to specify different cell fates during gonad development.

The genetic underpinnings of human hair growth are complex, relying upon several mechanisms to regulate gene expression. As such, inherited skin and hair disorders can arise from a variety of mutational events in the genome, from changes in single nucleotides to structural rearrangements of chromosomes. Importantly, inherited conditions affecting hair growth can be used as models to interrogate the molecular basis of the disease, and obtain novel insight into mechanisms and pathways required for normal hair growth. The approaches used to identify pathogenic mutations in a given disorder will depend on the inheritance pattern and disease prevalence in the population. In rare, Mendelian disorders of the skin and hair, the genetic architecture is often composed of rare variants with large effects, detected by linkage or whole-exome/genome sequencing. In contrast, polygenic disorders are composed of common and rare variants that contribute small to moderate effects and can be detected using Genome-Wide Association Studies (GWAS). The primary goal of my thesis research is to identify and characterize genetic mechanisms controlling human hair growth. To accomplish this, I have studied three inherited conditions affecting human hair growth as genetic models in which I performed detailed functional and molecular analyses of the causal genetic lesions and their downstream effects on gene expression in the hair follicle. My thesis work revolved around the use of three major approaches to identify and characterize genetic mechanisms underlying human hair growth: 1. Identifying genomic effects on human hair growth in a rare, sporadic Mendelian disorder (Chapter II), 2. Characterizing single-gene effects on human hair growth in a rare, familial condition (Chapter III), and 3. Functional analysis of rare, non-coding variants in a complex polygenic autoimmune disease, alopecia areata (Chapter IV). In the first part of my thesis work, I investigated the genetic mechanism associated with X-linked hypertrichosis (XLH), a very rare condition of excessive hair overgrowth. We identified a large interchromosomal insertion at chrXq27.1 that completely cosegregated with the phenotype, and was consistent with findings of interchromosomal insertions in two previously reported XLH families. Remarkably, the insertions in all three families occur at the exact same palindromic sequence, and because the sequences contained within each insertion are distinct, we hypothesized that the presence of the insertion (rather than its content) may be responsible for the excessive hair overgrowth phenotype. I then tested the impact of the insertion on the expression of the surrounding genes and found that FGF13 levels were selectively and dramatically reduced in patient hair follicles, suggesting a position effect as a result of the interchromosomal insertion. We postulate that the presence of this insertion disrupts key inter- and intrachromosomal interactions required for normal hair growth. In the second part of my thesis, I identified single-gene mutations that affect human hair growth by investigating the genetic basis of autosomal recessive congenital generalized hypertrichosis terminalis (CGHT) in a consanguineous family. We performed whole-exome sequencing and identified a novel, rare splice variant in ABCA5 that cosegregates with the phenotype in a homozygous recessive manner. I found that ABCA5 is highly expressed in human skin and hair follicles, and its expression pattern is conserved in mouse tissues as well. The ABCA5 mutation in CGHT leads to a complete loss-of-function in patient hair follicles, as well as reduced lysosome function and cholesterol transport, a finding consistent with defects in Abca5-/- mice. Moreover, we identified a deletion spanning ABCA5 in an unrelated sporadic CGHT case and found that ABCA5 levels were dramatically reduced in patient hair follicles. Collectively, our findings point to a novel role for ABCA5 in regulating hair growth. In the third part of my thesis, I characterized the genetic architecture of a complex, polygenic disease affecting hair growth by studying rare, non-coding variants in alopecia areata. The Christiano lab previously performed the first GWAS, identifying the ULBP3/6 locus on chr.6q25.1 encoding NKG2D T cell receptor ligands as the most significant association outside the HLA region. A strong upregulation of these ligands was observed on both human and mouse hair follicles, and we recently showed that T cells bearing the NKG2D receptor are both necessary and sufficient to induce disease in the mouse model. To identify the susceptibility variants at ULBP3/6 in human AA, we performed targeted deep resequencing and functional genomics studies and identified three rare, novel variants that reside within ULBP6 regulatory elements and CTCF binding sites. I found that these variants disrupt CTCF binding and regulatory activity in vitro, and CTCF binding is enriched at ULBP6 in vivo. Future studies examining long-range reporter activity and CTCF-mediated interactions will define the role of CTCF in the repression of the ULBP3/6 genes in the human hair follicle. Collectively, I used a variety of genetic approaches to identify novel genes and mechanisms controlling human hair growth.

Phosphate deficiency mimics Vitamin D deficiency and generates a rachitic state that impairs the ability of bone to mineralize and delays fracture healing. Previous studies have shown that hypophosphatemia alters many biological functions, and oxidative phosphorylation was the top biological process that was predicted affected by this perturbation. The goals of this study were to characterize the temporal mRNA expression for genes associated with oxidative phosphorylation and determine if phosphate deficiency more broadly altered the expression of genes associated with intermediate metabolism during fracture healing. Three in bred strains of skeletally mature male mice had stabilized fractures made in the right femur. The experimental group (Pi) was fed a low phosphate diet starting two days before fracture and continued for 16 days after which, a normal diet was re-introduced. The control group (Ctrl) was fed a normal diet throughout the study. RNA was extracted from fracture calluses and the RNA was quantified by microarray analysis. Analysis of covariance (ANCOVA) identified genes that were differentially expressed between the Pi and Ctrl groups (q-value ≤ 0.005), and analysis was performed on a subset of 577 genes related to the central elements of intermediate metabolism that were identified using the Kyoto Encyclopedia of Genes and Genomes (KEGG). To categorize genes that have similar temporal responses to the phosphate perturbation, the ratio of Pi/Ctrl expressions for each strain was independently clustered using normal mixtures method. These clusters were compared across the three strains to ascertain which groups of genes responded together in all three strains, and which biological functions bifurcated due to genetic variation between the strains. Oxidative phosphorylation and the tricarboxylic acid pathways showed the tightest temporal control in response to dietary phosphate restriction across all strains of mice. Analysis of the temporal profile showed the response to phosphate was strain dependent with the B6 strain showing the strongest response to Pi restriction. The re-introduction of phosphate also restored and/or abolished the delay in oxidative phosphorylation expression in a strain specific manner. AJ and C3 strains showed more similar rebounding profiles compared to the B6 stain. Additionally, it was found that Complex IV was affected by the phosphate deficient diet in manner different from the other components of ETC. Some intermediate metabolic pathways that also appeared to be affected in similar manners across multiples strains included glycolysis, and arginine and proline metabolism. Overlaying these various functions, temporal coordination was observed and suggests the functions may be mediated by a common upstream regulator. Further investigation of this potential upstream regulator can help to better describe the mechanisms that cause a delayed fracture healing in a phosphate deficient state.

No description available.