Human Gene Regulation

In the following thesis I examine the topic of human gene regulation in two distinct sets of experiments. The first set of experiments examines how human diversity contributes to variation in gene regulation and expression. The second set of experiments examines how miRNA expression distinguishes basal cell carcinoma subtypes and contributes to overall phenotypes.;RNA Pol II and NFκB binding was quantitated via chromatin immunoprecipitation followed by sequencing in ten β-lymphoblast cell lines from Han Chinese, Tokyo Japanese, CEPH Europeans, and Nigerian Yorubans. A total of 15,552 NFκB and 19,061 RNA Pol II binding sites were identified. 7.5% of NFκB and 25% of RNA Pol II binding sites varied significantly between at least two individuals. When a structural variant was present in the binding region, in 90% of the cases for NFκB and 80% of the cases for RNA Pol II binding followed the correct trend (i.e. a duplication of the binding region increased binding whereas a deletion reduced it). RNA expression showed a 0.475 correlation with NFκB binding and 0.461 with RNA Pol II binding. The genes near variable binding sites showed an enrichment for involvement in immunity and defense (p=0.045).;Two genes, Toll-like Receptor 1 (TLR1) and ATPase subunit beta-1 (ATP1B1), identified as being proximal to variable RNA Pol II bind sites were further studied. These genes were. TLR1 codes a lipoprotein and peptidoglycan receptor that recognizes gram-positive bacteria, and ATP1B1 codes a subunit of the Na+/K+ pump. In addition to validating expression variability at these genes, I attempted to link their expression to the surrounding genotype. The genotypes surrounding both genes were both analyzed in the standard Kidd Lab collection, which consists of 2739 individuals in 58 populations. In addition, HGDP, HapMap, and 1000Genomes datasets were also used for select analysis.;TLR1 showed no significant variance in expression in either luciferase assays on common regulatory-region haplotypes or in expression assays in β-lymphoblast cell lines under basal and PAM3CSK 4 stimulated conditions. Luciferase constructs identified two novel polymorphisms on a primarily East Asian haplotype. The haplotypes formed by rs5743618 and rs5743551 were also studied. The derived allele of rs5743618 causes reduced TLR1 surface trafficking, whereas the derived allele of rs5743551 has been linked to reduced risk of sepsis. The linkage disequilibrium between these two SNPs only exists primarily in Europe, suggesting that rs5743551 may only be a useful marker in these populations.;ATP1B1 showed significant differences in gene expression between thirty nine β-lymphoblast cell lines. In addition, a thorough analysis of a 1MB region surrounding ATP1B1 was performed on a panel of 140 HGDP SNPs. A region of positive selection extending approximately 350kb and including ATP1B1, NME7, BLZF1, C1orf114, and SLC19A2 was identified using nHS and REHH testing. This positive selection occurred on two distinct haplotypes – one common in East, Central, and South Central Asian populations, and one occurring in Southwest Asian and European populations. A missense polymorphism, rs1082180, in BLZF1 was in LD with the East Asian haplotype. There was no significant difference in expression between any of the haplotypes with signatures of selection and the ancestral haplotype.;ADH1B was studied due to strong evidence of positive selection on the derived allele of an upstream variant, rs3811801. The derived allele of rs3811801 is found only on an East Asian haplotype with the hypermorphic derived allele of rs1229984 (48His). Both states of rs3811801 were isolated successfully between two luciferase constructs. Luciferase assays did not show any significant difference in expression between the ancestral or the derived states of rs3811801.;Basal cell carcinomas are the most common form of cancer in the US. BCCs differentiate into five different subtypes. Global miRNA expression was quantified in tumors from the infiltrative and nodular subtypes. Seven of eight infiltrative tumors formed a tight cluster in a principle components analysis. Pathological review of the remaining tumor identified a mix of nodular and infiltrative elements. Nodular tumors did not cluster tightly, likely reflecting broader histopathologic diversity in this class, but trended towards forming a separate group. qPCR assays of six miRNAs that were significantly different between the classes validated five of them. One of these was miR-183, a miRNA that inhibits invasion and metastasis in several types of malignancies. Mir-183 was consistently lower in infiltrative versus nodular and may be one element underlying its invasive behavior.