Eukaryotic genomes contain many copies of ribosomal DNA (rDNA) encoding the RNA components of ribosomes, ranging from 150 tandem repeated copies in the budding yeast Saccharomyces cerevisiae to approximately 700 copies total in diploid human cells. It is not known what dictates rDNA copy number, but we hope to gain insight by identifying genetic regulators of rDNA copy number. Recent work in the Brewer/Raghuraman lab using S. cerevisiae reveals connections between rDNA copy number and cellular processes, such as genome replication and replicative aging. These studies suggest that too few rDNA copies cannot satisfy ribosome demand and too many copies can interfere with DNA replication. Thus, certain genetic pathways must function to maintain such balance. Our preliminary survey finds that some strains with specific single gene deletions have varying rDNA copy numbers, suggesting genetic control of the flexibility of rDNA copy number. Our goal is to identify genetic regulators of rDNA copy number in a systematic genome-wide screen. I have developed a high throughput screen to look for strains with variation in rDNA copy number by measuring the size of chromosome XII, the chromosome that carries the rDNA locus. I will use this method to screen a yeast library of approximately 5000 non-essential single gene deletions. After finding and analyzing mutant strains with chromosome XII size changes, we will have a more comprehensive understanding of the cellular mechanism that controls rDNA repeat number. As rDNA repeat number is linked to replicative lifespan, we hope to uncover the common genetic pathways that regulate rDNA copy number and longevity.

 A significant proportion of the 23 known human-specific duplicated genes are implicated in neurological processes. The Rho GTPase activating protein 11 (ARHGAP11) is one such gene that is also found in an area associated with schizophrenia, intellectual disability, and autism. We determined that the human-specific duplicated copy (ARHGAP11B) duplicated 5.3 million years ago (mya) shortly after the human-chimpanzee divergence (~ 6.5 mya). The human-specific duplicated copy is not found in all humans, and represents an incomplete version of ARHGAP11A, suggesting a potentially alternative function. The goal of my project was to identify sequence differences between A (ancestral) and B (duplicated) paralogs (i.e., genes related by duplication within the same species) and use these differences to assess gene variation and expression. By sequencing 64 different transcripts (i.e. protein coding regions) we found nine “paralog-specific variants” (or PSVs) allowing us to distinguish between A and B. We assessed the level of gene expression (how much a gene is “used”) for each paralog in human fetal brain samples and found that A was expressed approximately two times more than B. Interestingly, we found similar results from other tissue types including blood. Further, we have used these PSVs to identify potentially pathogenic mutations of ARHGAP11A from targeted large-scale sequencing data of a collection of children with intellectual disability. We are using these data to understand the function of different members of the ARHGAP11 gene family both on neurological development and potentially on the evolution of traits that make us uniquely human.

DNA replication is vital for cell division as it produces exact copies of all chromosomes for the two daughter cells.  During S-phase of the cell cycle, DNA replication begins at specific, defined locations within a chromosome known as origins of replication.  Interestingly, some origins activate in early S-phase while others activate late.  The biological significance of this temporal control of replication is largely unknown.  In multiple species, centromeric DNA, the site of recruitment for proteins required for separation of chromosomes during mitosis, has been found to be early replicating.  The Brewer/Raghuraman lab has recently shown that centromeres in the budding yeast Saccharomyces cerevisiae promote their own early replication by advancing the activation time of their neighboring origins.  The conservation of early centromere replication coupled with the Brewer/Raghuraman lab’s recent finding suggests that early centromere replication may be important for chromosome maintenance.  We set out to determine the consequences of a late replicating centromere by artificially delaying its replication time in S. cerevisae, through replacing its nearby origins with drug resistant markers.  To enhance sensitivity of the experiment, I impaired the activity of the spindle checkpoint that monitors correct chromosome segregation.  I also constructed a control chromosome with an early replicating centromere and the same markers.  I will then compare the degree of chromosomal instability, as measured by survival on drug plates, of the control cells to those that contain a late replicating centromere.  We expect the experimental group to show an increase in the number of cells that have lost the chromosome compared to the control cells.  Even if the effect is small we will perform long-term growth experiments to analyze how the cells cope with reduced genetic stability.  The results of this study will provide a better understanding of the biological importance of the temporal replication program.

New DNA sequencing technologies have increased the throughput of sequencing by several orders of magnitude, reaching upwards of 50 GB of data per day on a single instrument. However, advancing our understanding of how genetic variation impacts human health and disease depends on studies that survey very large numbers of individuals, and as such may only be economical by analyzing specific subsets of the genome. This is complicated by the fact that traditional methods of preparing DNA samples for sequencing do not allow selection for such candidate genes. One solution is the use of “capture by circularization” by molecular inversion probes (MIPs). MIPs allow dozens of genes to be targeted and amplified simultaneously; however, individual MIPs suffer from notoriously variable performance relative to one another, which results in uneven coverage of target regions. Prior work highlighted the potential for statistical analysis to improve MIP design and selection, inviting further testing. After performing a large scale sequencing run using an unbiased set of 12,000 MIPs, logistic regression was used to build a model capable of scoring MIPs prior to synthesis. New design software incorporating this model into the selection process was developed and tested by designing MIPs to over 60 new genes. Analysis of the resulting sequence data demonstrated the successful in silico identification of challenging genomic regions and a reduction in the number of gaps in coverage, which should support the use of MIPs in future large scale studies.

Tubes are the foundation for many organs in all animals. Although tube formation is critical for development, relatively little is known about how cells coordinate their behaviors to form tubes. To investigate this process, we are studying the development of the dorsal appendages (DAs), two respiratory structures that are formed in Drosophila ovaries and protrude from the fly egg. We use a genetic approach to disrupt gene function, then study cell behaviors during tube formation. Previously, our lab showed that trimeric G proteins play a role in DA formation. Trimeric G proteins act as signal transducers and Gα, one subunit of the trimeric G-protein complex, is responsible for signal specificity. Although in situ hybridization showed that six of seven Drosophila Gαs are expressed in the ovary, functional studies demonstrated that Gα12 was most important. RNAi against Gα12 in a subset of DA-forming cells resulted in 90% of egg chambers exhibiting abnormal DAs. We hypothesize that concertina and its partial duplicate CG40005, the genes coding for Gα12, work together in DA-forming cells to coordinate cell behaviors. To test this hypothesis, we generated null mutations in cta and CG40005. We induced random deletions in the DNA via irradiation and screened for loss of a visible marker carried on a transposon near cta. Through complementation tests, we showed that 22 deletions extended into cta, or the neighboring essential gene, light. Although the genomic region contains many repeated sequences, I developed PCR tests and found 18 deletions disrupting cta, CG40005, and light, and one deleting only CG40005. Finally, rescue studies using large BAC clones generated by the P[acman] project will let us examine cell behaviors during DA formation in the absence of cta and CG40005. Since DA formation resembles neural-tube formation in vertebrates, our studies will give insight into tube-forming processes in other animals.

The nuclear transport pathway performs the fundamental function of moving cargo between the cytoplasm and nucleus in all eukaryotes. Nuclear transport is carried out through a highly conserved mechanism across all eukaryotes. Yet, in Drosophila, several components of the nuclear transport apparatus, including importins, evolve rapidly under positive selection. Genetic conflict with selfish elements has been suggested as a possible cause for this pattern of rapid evolution. Here, we present a comprehensive phylogenomic analysis of importin gene evolution in Drosophila. Importins are adapter molecules that directly mediate the transport of cargo into the nucleus. Our analysis reveals a recurrent pattern of gain and loss of importin paralogs across independent lineages of Drosophila. Interestingly, we discovered that almost all new copies of importins have acquired a testes-specific expression pattern since their birth through gene duplication. This pattern of repeated gains of testes-specific copies of importins suggests a function in suppressing genetic conflicts such as segregation distortion in the male germline. Segregation distorters such as SD in Drosophila melanogaster act by impairing nuclear transport in the testes. We are currently performing functional tests for the hypothesis that an increased dosage of these testes-specific importins may serve a role in suppressing segregation distortion in the male germline by restoring nuclear transport during spermatogenesis.

The reversible phosphorylation of proteins mediates a wide range of biological processes that range from signal transduction cascades to regulation of protein abundance. However, little is known about the mechanisms and evolution of phosphorylation networks. Despite the extraordinary advances in genome sequencing of many yeast species, evolutionary studies on the phosphoproteome of yeast species have been limited to the experimental analysis of phosphorylation in one species and theoretical analysis of the conservation of phospho-acceptor residues with other species. To properly study the evolutionary conservation phosphorylation in yeast, we are utilizing mass spectrometry to study and compare the phosphoproteome of 21 species of yeast, including representative species from each clade of the yeast phylogenetic tree. Most of these species have not been studied before by proteomics. This high-throughput phosphoproteomic study on the yeast species will contribute to the construction of phosphoproteome datasets, which can be exploited for comparative analysis of phosphorylation between the species. One of main aims is to identify phosphorylation sites that show a high degree of conservation across the 21 species of yeast. These highly conserved phosphorylation sites likely regulate the same functions across the different species, and these functions can be further studied and correlated to our understanding of the mechanistics of phosphorylation in key biological processes, as well as the diseases that occur when the signaling pathways that involve these sites are defective. The second aim of this project is to study the correlation between the conservation of the phosphoproteome with the conservation of the genome and transcriptome, which will expand our understanding of the evolution of phenotypic diversity. Finally, the large-scale dataset of the proteomes and phosphoproteomes holds a great potential for other types of computational analysis, and also serves as a foundational reference for future experiments.

In an effort to study evolution and mechanisms of adaptation, seven populations of S. cerevisiae were cultivated in chemostat under nutrient limitation. Whole genome sequencing and Single Nucleotide Polymorphisms (SNPs) analysis were performed on clones isolated at 200 generations. I sought to test the effect of each mutation on the cellular fitness thus to characterize beneficial and passengers mutations. By combining PCR and Sanger sequencing, I confirmed five mutations (4 SNPs, one indel) detected by whole genome sequencing on one clone. Using backcross-mating techniques with the ancestral strains, I was able to isolate each mutation and test their individual fitness. I found that 4 of the mutations are passenger mutations that do not impact the fitness of the cells. One of the mutations, a non-synonymous change in SGF73 leading to the truncation of the protein, increased the competitive fitness by 24%. Mutations leading to the truncation of SGF73 were also found in three other independent clones. The presence of similar mutations in four independent populations suggests that the loss of the protein is an important process of adaptation to sulfate limitation. I am now studying the dynamic and the spread of this mutation in the population over 200 generations. SGF73 is the human ortholog of ataxin-7 of the SAGA complex associated with spinocerebellar ataxia diseases and our findings may be important to study the role of this protein in the development of human disease.

Our goal is to use recently available large-scale genome sequencing data to build a comprehensive database of all known mutations in the Fanconi anemia (FA) genes, and then use these data to identify subsets of mutations that may be most susceptible to mutation-specific therapies. Fanconi anemia (FA) is a recessive disease caused by mutations in any one of 15 genes (FANC genes) that are part of a coordinated pathway that senses and repairs DNA damage. These mutations will serve as the focus for experimental work to determine how much FANC protein function is required to suppress the FA cellular phenotype, and which FANC mutations retain residual activity and may be better-suited for drug or small molecule therapeutics. This work was stimulated by the recent success of mutation-specific small molecule therapeutics in other recessive diseases such as cystic fibrosis. In complementary research we are using FA patient-derived cell lines to search for existing drugs and small molecules that suppress the FA cellular phenotype in response to aldehyde damage. Aldehyde damage is an unavoidable consequence of normal human physiology, and aldehydes are a common type of environmental mutagen and carcinogen. In contrast to our other research, this work is not focused on individual disease-causing mutations. We will present the screening assay being used for this part of the project, and preliminary results. We hope to identify FA disease-associated mutations that are the best candidates for future drug or small molecule research by virtue of retaining residual protein function, and thus the potential for rescue of function. We also hope to provide one example of a small molecule that can suppress the FA cellular phenotype regardless of the underlying mutation. Both results should aid the search for patient- and mutation-specific or more general therapies to prevent or treat Fanconi anemia.

Although every cell in the human body has nearly the same DNA sequence, there are many different cell types performing different functions; for example, lymphoblast cells which differentiate into white blood cells or endothelial cells which provide a lining throughout the human vascular system. But cells do not always function as they should, for example, overproduction of white blood cells can be caused by lymphoblastic leukemia, a type of cancer. I intend to use data on the sensitivity of parts of the genome to deoxyribonuclease (DNase) digestion to explain the differences in cell behavior. Regions of the genome structure that are hypersensitive to DNase have already been shown to affect gene expression. I want to determine if portions of the genome which are mildly sensitive, or mesosensitive, to DNase play a role in cell differentiation. My hypothesis is that DNase can identify large scale regions, on the order of thousands of base pairs, that harbor cell type specific genes. I used Segway, software which uses a dynamic Bayesian network, to segment the genome based on DNase sensitivity data into regions of low sensitivity, mesosensitivity, and hypersensitivity on six different cell types. From my segmentations I removed the hypersensitive regions from considerations, and compared the low and mesosensitive regions to gene expression data. I found that regions Segway labels mesosensitive had higher levels of expression than low sensitivity regions. I expect to find that many of the regions responsible for cell type specific behavior are mesosensitive. If this is the case, I hope the results will be robust enough such that anyone with access to DNase data can determine what genes are active in a population of cells, enabling them to better predict the behavior of the cells. In furthering our understanding of gene regulation, we may better understand diseases such as leukemia.

Protein phosphorylation is a common regulatory mechanism carried out by the cell, as adding a phosphate group to a particular protein can switch its activity on or off.  Signaling pathways employ this control method to instigate series of phosphorylation events that elicit a particular cell response.  Our research involves extracting and characterizing the phosphoproteome, or the cell’s entire set of expressed phosphorylated proteins, in both yeast and mammalian cell models.  To generate lists of phosphoproteins present in these cell models, we use mass spectrometry, an analytic technique that produces characteristic spectra for individual phosphopeptides that are then matched to a library of known protein spectra through the use of computer algorithms.  In our preliminary results, we have generated large-scale data sets identifying thousands of phosphorylated proteins present in each species and cell type.  The goal of the first project, the yeast study, is to determine the degree of evolutionary conservation in phosphorylation events across 19 yeast species to learn about the basic principles of phosphoregulation.  The conservation of phosphorylation can also give insight into phosphorylation events in other species evolutionarily related to yeast, such as humans.  The second project, the mammalian cell study, examines the phosphoproteome of different breast cancer cell types in order to identify common phosphorylated proteins, from which we can deduce a molecular signature for breast cancer.  This signature can be used in future studies to design a universal protein inhibitor drug as an alternative anti-cancer therapy.  Additionally, by identifying the phosphoproteins exclusive to each cell type, we can uniquely characterize specific types of breast cancer.  Knowledge of the proteins active in one type of breast cancer may facilitate development of inhibitor drugs as a form of personalized medicine.

Chemostats are a continuous culture system in which cells are grown in a tightly controlled, chemically constant environment where culture density is constrained by limiting specific nutrients. Unlike batch cultures, the selective pressure imposed in a chemostat is constant. This has made them a desirable tool for evolution and competition experiments. Traditionally, chemostat experiments with S. cerevisiae have been constrained due to the limited number of cultures that can be run at once. In order to solve this problem our team is working with a new culturing system that allows 64 chemostats to be run simultaneously. In our experiment we evolved 96 biological replicates of haploid MATa yeast cells under either constant sulfur, phosphate, or glucose limiting environments for 300 generations. We are now using a variety of techniques including array-comparative genomic hybridization and whole genome sequencing of evolved populations to characterize mutations that have been selected for during the course of these evolution experiments. These techniques have revealed that amplification of genes that code for nutrient transporters are a common mutation that occur in the evolved populations over the course of the experiments. Therefore, we will also be experimenting with strains that have the most common amplified genes knocked-out in the hopes of identifying alternative pathways for nutrient-limited fitness increases. Changes in the genome can be correlated to changes in relative fitness, which we assayed through competition with a green fluorescent protein (GFP) expressing ancestral strain every 50 generations. Ultimately it is our hope that higher replicate number will better reveal the spectrum of genes important for adaptive growth in specific nutrient-limited environments.

Schizophrenia is a debilitating neuropsychiatric disorder with a worldwide prevalence of about 1%. Although the heritability of schizophrenia is high, there are many sporadic cases; that is, affected individuals with no family history of mental illness. We are sequencing persons with schizophrenia in the absence of family history of mental illness (probands), their unaffected parents, and whenever possible, an unaffected older sibling, in order to identify de novo mutations that may be responsible for the disorder. Our hypothesis is that de novo events will be observed in many different genes, and that the functions of the impacted genes will differ between probands and their unaffected siblings. From exome sequence, we filter variants to include only those predicted to be both de novo and damaging to the protein product. Next, we validate all candidate variants passing these filters by Sanger sequencing, using diagnostic PCR primers to amplify genomic DNA of probands, siblings, and parents. Results to date suggest that approximately 50% of probands and siblings carry at least one damaging de novo mutation. The analysis of pathways in which the genes converge is in progress. Next, we will target these candidate genes and determine their complete variant profiles in 5000 cases and 5000 controls. These results can help us identify critical pathways that are affected in schizophrenia.

Breast cancer tumors are classified within different subtypes based on the expression of three types of cell surface receptor proteins.  These proteins connect the extracellular environment with cellular activities via signaling pathways.  To elicit cell responses, signaling pathways employ protein phosphorylation as a regulatory mechanism, as adding a phosphate group to a particular protein can switch its activity on or off.  However, normal signaling can go awry if mutations in signaling proteins lead the cell to believe its components are always phosphorylated and active.  Consequently, the entire signaling pathway functions abnormally, which can lead to uncontrolled cell growth characteristic of cancer.  Given its potential role as a molecular mechanism of disease, our research focuses on obtaining a global view of the phosphoproteome, or the cell’s entire set of expressed, phosphorylated proteins, of 10 breast cancer cell lines.  To generate lists of phosphoproteins present in these cell lines, we are concurrently extracting and analyzing cell line phosphopeptide samples via mass spectrometry, an analytic technique that produces characteristic spectra for individual phosphopeptides that are then matched to a library of known protein spectra through the use of computer algorithms.  In our preliminary results, we have generated large-scale data sets identifying thousands of phosphorylated proteins present in two cell lines and generated side-by-side comparisons of the proteins expressed in each line.   By identifying common phosphorylated proteins across all of the cell lines, our goal is to deduce a molecular signature for breast cancer.  This signature can be used in future studies to design a universal protein inhibitor drug as an alternative anti-cancer therapy.   By identifying the phosphoproteins exclusive to each cell type, we can uniquely characterize specific proteins active in one type of breast cancer which may facilitate development of inhibitor drugs as a form of personalized medicine.

Despite recent sweeping advances in DNA sequencing technology, whole genome sequencing remains impractical as a routine medical service. Nevertheless, recent work suggests that rare genetic mutations may play large roles in modulating disease progression. More affordable targeted DNA sequencing is currently necessary to study variants of low frequency in the human population and attain sufficient statistical power to link these variants with specific disease phenotypes. Previous studies have demonstrated the effectiveness of a capture by circularization approach using molecular inversion probes (MIPs) whereby thousands of individual sites in the genome can be captured simultaneously for DNA sequencing. Yet, non-uniformity across large numbers of targeted sites persists as a major barrier to more widespread adoption. In this work, we show that statistical analysis of past MIP sequencing runs readily informs modifications to the process of probe design and selection. Additionally, logistic regression on carefully selected MIP features enhances in silico identification of ineffective probes that leave gaps in sequence coverage. Richer data sets from MIP sequencing runs representing a broader range of MIP characteristics and interactions may offer further guidance in perfecting the design of MIPs, achieving greater uniformity across targeted sites and ultimately lowering the cost of sequencing.