Computational Inference of Genome-Wide Protein-DNA Interactions Using High-Throughput Genomic Data

Transcriptional regulation has been studied intensively in recent decades. One important aspect of this regulation is the interaction between regulatory proteins, such as transcription factors (TF) and nucleosomes, and the genome. Different high-throughput techniques have been invented to map these interactions genome-wide, including ChIP-based methods (ChIP-chip, ChIP-seq, etc.), nuclease digestion methods (DNase-seq, MNase-seq, etc.), and others. However, a single experimental technique often only provides partial and noisy information about the whole picture of protein-DNA interactions. Therefore, the overarching goal of this dissertation is to provide computational developments for jointly modeling different experimental datasets to achieve a holistic inference on the protein-DNA interaction landscape.

We first present a computational framework that can incorporate the protein binding information in MNase-seq data into a thermodynamic model of protein-DNA interaction. We use a correlation-based objective function to model the MNase-seq data and a Markov chain Monte Carlo method to maximize the function. Our results show that the inferred protein-DNA interaction landscape is concordant with the MNase-seq data and provides a mechanistic explanation for the experimentally collected MNase-seq fragments. Our framework is flexible and can easily incorporate other data sources. To demonstrate this flexibility, we use prior distributions to integrate experimentally measured protein concentrations.

We also study the ability of DNase-seq data to position nucleosomes. Traditionally, DNase-seq has only been widely used to identify DNase hypersensitive sites, which tend to be open chromatin regulatory regions devoid of nucleosomes. We reveal for the first time that DNase-seq datasets also contain substantial information about nucleosome translational positioning, and that existing DNase-seq data can be used to infer nucleosome positions with high accuracy. We develop a Bayes-factor-based nucleosome scoring method to position nucleosomes using DNase-seq data. Our approach utilizes several effective strategies to extract nucleosome positioning signals from the noisy DNase-seq data, including jointly modeling data points across the nucleosome body and explicitly modeling the quadratic and oscillatory DNase I digestion pattern on nucleosomes. We show that our DNase-seq-based nucleosome map is highly consistent with previous high-resolution maps. We also show that the oscillatory DNase I digestion pattern is useful in revealing the nucleosome rotational context around TF binding sites.

Finally, we present a state-space model (SSM) for jointly modeling different kinds of genomic data to provide an accurate view of the protein-DNA interaction landscape. We also provide an efficient expectation-maximization algorithm to learn model parameters from data. We first show in simulation studies that the SSM can effectively recover underlying true protein binding configurations. We then apply the SSM to model real genomic data (both DNase-seq and MNase-seq data). Through incrementally increasing the types of genomic data in the SSM, we show that different data types can contribute complementary information for the inference of protein binding landscape and that the most accurate inference comes from modeling all available datasets.

This dissertation provides a foundation for future research by taking a step toward the genome-wide inference of protein-DNA interaction landscape through data integration.

Characterization of a Full-Length TTP Family Member Association with RNA Sequence Elements

Post-transcriptional regulation of cytoplasmic mRNAs is an efficient mechanism of regulating the amounts of active protein within a eukaryotic cell. RNA sequence elements located in the untranslated regions of mRNAs can influence transcript degradation or translation through associations with RNA-binding proteins. Tristetraprolin (TTP) is the best known member of a family of CCCH zinc finger proteins that targets adenosine-uridine rich element (ARE) binding sites in the 3’ untranslated regions (UTRs) of mRNAs, promoting transcript deadenylation through the recruitment of deadenylases. More specifically, TTP has been shown to bind AREs located in the 3’-UTRs of transcripts with known roles in the inflammatory response. The mRNA-binding region of the protein is the highly conserved CCCH tandem zinc finger (TZF) domain. The synthetic TTP TZF domain has been shown to bind with high affinity to the 13-mer sequence of UUUUAUUUAUUUU. However, the binding affinities of full-length TTP family members to the same sequence and its variants are unknown. Furthermore, the distance needed between two overlapping or neighboring UUAUUUAUU 9-mers for tandem binding events of a full-length TTP family member to a target transcript has not been explored. To address these questions, we recombinantly expressed and purified the full-length C. albicans TTP family member Zfs1. Using full-length Zfs1, tagged at the N-terminus with maltose binding protein (MBP), we determined the binding affinities of the protein to the optimal TTP binding sequence, UUAUUUAUU. Fluorescence anisotropy experiments determined that the binding affinities of MBP-Zfs1 to non-canonical AREs were influenced by ionic buffer strength, suggesting that transcript selectivity may be affected by intracellular conditions. Furthermore, electrophoretic mobility shift assays (EMSAs) revealed that separation of two core AUUUA sequences by two uridines is sufficient for tandem binding of MBP-Zfs1. Finally, we found evidence for tandem binding of MBP-Zfs1 to a 27-base RNA oligonucleotide containing only a single ARE-binding site, and showed that this was concentration and RNA length dependent; this phenomenon had not been seen previously. These data suggest that the association of the TTP TZF domain and the TZF domains of other species, to ARE-binding sites is highly conserved. Domains outside of the TZF domain may mediate transcript selectivity in changing cellular conditions, and promote protein-RNA interactions not associated with the ARE-binding TZF domain.

In summary, the evidence presented here suggests that Zfs1-mediated decay of mRNA targets may require additional interactions, in addition to ARE-TZF domain associations, to promote transcript destabilization and degradation. These studies further our understanding of post-transcriptional steps in gene regulation.

RNA Localization and Translational Regulation on the Endoplasmic Reticulum

mRNA localization is emerging as a critical cellular mechanism for the spatiotemporal regulation of protein expression and serves important roles in oogenesis, embryogenesis, cell fate specification, and synapse formation. Signal sequence-encoding mRNAs are localized to the endoplasmic reticulum (ER) membrane by either of two mechanisms, a canonical mechanism of translation on ER-bound ribosomes (signal recognition particle pathway), or a poorly understood direct ER anchoring mechanism. In this study, we identify that the ER integral membrane proteins function as RNA-binding proteins and play important roles in the direct mRNA anchoring to the ER. We report that one of the ER integral membrane RNA-binding protein, AEG-1 (astrocyte elevated gene-1), functions in the direct ER anchoring and translational regulation of mRNAs encoding endomembrane transmembrane proteins. HITS-CLIP and PAR-CLIP analyses of the AEG-1 mRNA interactome of human hepatocellular carcinoma cells revealed a high enrichment for mRNAs encoding endomembrane organelle proteins, most notably encoding transmembrane proteins. AEG-1 binding sites were highly enriched in the coding sequence and displayed a signature cluster enrichment downstream of encoded transmembrane domains. In overexpression and knockdown models, AEG-1 expression markedly regulates translational efficiency and protein functions of two of its bound transcripts, MDR1 and NPC1. This study reveals a molecular mechanism for the selective localization of mRNAs to the ER and identifies a novel post-transcriptional gene regulation function for AEG-1 in membrane protein expression.

Yeast screens identify the RNA polymerase II CTD and SPT5 as relevant targets of BRCA1 interaction.

BRCA1 has been implicated in numerous DNA repair pathways that maintain genome integrity, however the function responsible for its tumor suppressor activity in breast cancer remains obscure. To identify the most highly conserved of the many BRCA1 functions, we screened the evolutionarily distant eukaryote Saccharomyces cerevisiae for mutants that suppressed the G1 checkpoint arrest and lethality induced following heterologous BRCA1 expression. A genome-wide screen in the diploid deletion collection combined with a screen of ionizing radiation sensitive gene deletions identified mutants that permit growth in the presence of BRCA1. These genes delineate a metabolic mRNA pathway that temporally links transcription elongation (SPT4, SPT5, CTK1, DEF1) to nucleopore-mediated mRNA export (ASM4, MLP1, MLP2, NUP2, NUP53, NUP120, NUP133, NUP170, NUP188, POM34) and cytoplasmic mRNA decay at P-bodies (CCR4, DHH1). Strikingly, BRCA1 interacted with the phosphorylated RNA polymerase II (RNAPII) carboxy terminal domain (P-CTD), phosphorylated in the pattern specified by the CTDK-I kinase, to induce DEF1-dependent cleavage and accumulation of a RNAPII fragment containing the P-CTD. Significantly, breast cancer associated BRCT domain defects in BRCA1 that suppressed P-CTD cleavage and lethality in yeast also suppressed the physical interaction of BRCA1 with human SPT5 in breast epithelial cells, thus confirming SPT5 as a relevant target of BRCA1 interaction. Furthermore, enhanced P-CTD cleavage was observed in both yeast and human breast cells following UV-irradiation indicating a conserved eukaryotic damage response. Moreover, P-CTD cleavage in breast epithelial cells was BRCA1-dependent since damage-induced P-CTD cleavage was only observed in the mutant BRCA1 cell line HCC1937 following ectopic expression of wild type BRCA1. Finally, BRCA1, SPT5 and hyperphosphorylated RPB1 form a complex that was rapidly degraded following MMS treatment in wild type but not BRCA1 mutant breast cells. These results extend the mechanistic links between BRCA1 and transcriptional consequences in response to DNA damage and suggest an important role for RNAPII P-CTD cleavage in BRCA1-mediated cancer suppression.

Interaction of Cryptococcus neoformans Rim101 and protein kinase A regulates capsule.

Cryptococcus neoformans is a prevalent human fungal pathogen that must survive within various tissues in order to establish a human infection. We have identified the C. neoformans Rim101 transcription factor, a highly conserved pH-response regulator in many fungal species. The rim101 multiply sign in circle mutant strain displays growth defects similar to other fungal species in the presence of alkaline pH, increased salt concentrations, and iron limitation. However, the rim101 multiply sign in circle strain is also characterized by a striking defect in capsule, an important virulence-associated phenotype. This capsular defect is likely due to alterations in polysaccharide attachment to the cell surface, not in polysaccharide biosynthesis. In contrast to many other C. neoformans capsule-defective strains, the rim101 multiply sign in circle mutant is hypervirulent in animal models of cryptococcosis. Whereas Rim101 activation in other fungal species occurs through the conserved Rim pathway, we demonstrate that C. neoformans Rim101 is also activated by the cAMP/PKA pathway. We report here that C. neoformans uses PKA and the Rim pathway to regulate the localization, activation, and processing of the Rim101 transcription factor. We also demonstrate specific host-relevant activating conditions for Rim101 cleavage, showing that C. neoformans has co-opted conserved signaling pathways to respond to the specific niche within the infected host. These results establish a novel mechanism for Rim101 activation and the integration of two conserved signaling cascades in response to host environmental conditions.

Smoothened signal transduction is promoted by G protein-coupled receptor kinase 2.

Deregulation of the Sonic hedgehog pathway has been implicated in an increasing number of human cancers. In this pathway, the seven-transmembrane (7TM) signaling protein Smoothened regulates cellular proliferation and differentiation through activation of the transcription factor Gli. The activity of mammalian Smoothened is controlled by three different hedgehog proteins, Indian, Desert, and Sonic hedgehog, through their interaction with the Smoothened inhibitor Patched. However, the mechanisms of signal transduction from Smoothened are poorly understood. We show that a kinase which regulates signaling by many "conventional" 7TM G-protein-coupled receptors, G protein-coupled receptor kinase 2 (GRK2), participates in Smoothened signaling. Expression of GRK2, but not catalytically inactive GRK2, synergizes with active Smoothened to mediate Gli-dependent transcription. Moreover, knockdown of endogenous GRK2 by short hairpin RNA (shRNA) significantly reduces signaling in response to the Smoothened agonist SAG and also inhibits signaling induced by an oncogenic Smoothened mutant, Smo M2. We find that GRK2 promotes the association between active Smoothened and beta-arrestin 2. Indeed, Gli-dependent signaling, mediated by coexpression of Smoothened and GRK2, is diminished by beta-arrestin 2 knockdown with shRNA. Together, these data suggest that GRK2 plays a positive role in Smoothened signaling, at least in part, through the promotion of an association between beta-arrestin 2 and Smoothened.

Unfolded protein response genes regulated by CED-1 are required for Caenorhabditis elegans innate immunity.

The endoplasmic reticulum stress response, also known as the unfolded protein response (UPR), has been implicated in the normal physiology of immune defense and in several disorders, including diabetes, cancer, and neurodegenerative disease. Here, we show that the apoptotic receptor CED-1 and a network of PQN/ABU proteins involved in a noncanonical UPR response are required for proper defense to pathogen infection in Caenorhabditis elegans. A full-genome microarray analysis indicates that CED-1 functions to activate the expression of pqn/abu genes. We also show that ced-1 and pqn/abu genes are required for the survival of C. elegans exposed to live Salmonella enterica, and that overexpression of pqn/abu genes confers protection against pathogen-mediated killing. The results indicate that unfolded protein response genes, regulated in a CED-1-dependent manner, are involved in the C. elegans immune response to live bacteria.

Mass spectrometry-based thermal shift assay for protein-ligand binding analysis.

Described here is a mass spectrometry-based screening assay for the detection of protein-ligand binding interactions in multicomponent protein mixtures. The assay utilizes an oxidation labeling protocol that involves using hydrogen peroxide to selectively oxidize methionine residues in proteins in order to probe the solvent accessibility of these residues as a function of temperature. The extent to which methionine residues in a protein are oxidized after specified reaction times at a range of temperatures is determined in a MALDI analysis of the intact proteins and/or an LC-MS analysis of tryptic peptide fragments generated after the oxidation reaction is quenched. Ultimately, the mass spectral data is used to construct thermal denaturation curves for the detected proteins. In this proof-of-principle work, the protocol is applied to a four-protein model mixture comprised of ubiquitin, ribonuclease A (RNaseA), cyclophilin A (CypA), and bovine carbonic anhydrase II (BCAII). The new protocol's ability to detect protein-ligand binding interactions by comparing thermal denaturation data obtained in the absence and in the presence of ligand is demonstrated using cyclosporin A (CsA) as a test ligand. The known binding interaction between CsA and CypA was detected using both the MALDI- and LC-MS-based readouts described here.

Osmolyte-induced folding of an intrinsically disordered protein: folding mechanism in the absence of ligand.

Understanding the interconversion between thermodynamically distinguishable states present in a protein folding pathway provides not only the kinetics and energetics of protein folding but also insights into the functional roles of these states in biological systems. The protein component of the bacterial RNase P holoenzyme from Bacillus subtilis (P protein) was previously shown to be unfolded in the absence of its cognate RNA or other anionic ligands. P protein was used in this study as a model system to explore general features of intrinsically disordered protein (IDP) folding mechanisms. The use of trimethylamine N-oxide (TMAO), an osmolyte that stabilizes the unliganded folded form of the protein, enabled us to study the folding process of P protein in the absence of ligand. Transient stopped-flow kinetic traces at various final TMAO concentrations exhibited multiphasic kinetics. Equilibrium "cotitration" experiments were performed using both TMAO and urea during the titration to produce a urea-TMAO titration surface of P protein. Both kinetic and equilibrium studies show evidence of a previously undetected intermediate state in the P protein folding process. The intermediate state is significantly populated, and the folding rate constants are relatively slow compared to those of intrinsically folded proteins similar in size and topology. The experiments and analysis described serve as a useful example for mechanistic folding studies of other IDPs.

Thermodynamic analysis of ligand-induced changes in protein thermal unfolding applied to high-throughput determination of ligand affinities with extrinsic fluorescent dyes.

The quantification of protein-ligand interactions is essential for systems biology, drug discovery, and bioengineering. Ligand-induced changes in protein thermal stability provide a general, quantifiable signature of binding and may be monitored with dyes such as Sypro Orange (SO), which increase their fluorescence emission intensities upon interaction with the unfolded protein. This method is an experimentally straightforward, economical, and high-throughput approach for observing thermal melts using commonly available real-time polymerase chain reaction instrumentation. However, quantitative analysis requires careful consideration of the dye-mediated reporting mechanism and the underlying thermodynamic model. We determine affinity constants by analysis of ligand-mediated shifts in melting-temperature midpoint values. Ligand affinity is determined in a ligand titration series from shifts in free energies of stability at a common reference temperature. Thermodynamic parameters are obtained by fitting the inverse first derivative of the experimental signal reporting on thermal denaturation with equations that incorporate linear or nonlinear baseline models. We apply these methods to fit protein melts monitored with SO that exhibit prominent nonlinear post-transition baselines. SO can perturb the equilibria on which it is reporting. We analyze cases in which the ligand binds to both the native and denatured state or to the native state only and cases in which protein:ligand stoichiometry needs to treated explicitly.

Thermodynamic analysis of a molecular chaperone binding to unfolded protein substrates.

Molecular chaperones are a highly diverse group of proteins that recognize and bind unfolded proteins to facilitate protein folding and prevent nonspecific protein aggregation. The mechanisms by which chaperones bind their protein substrates have been studied for decades. However, there are few reports about the affinity of molecular chaperones for their unfolded protein substrates. Thus, little is known about the relative binding affinities of different chaperones and about the relative binding affinities of chaperones for different unfolded protein substrates. Here we describe the application of SUPREX (stability of unpurified proteins from rates of H-D exchange), an H-D exchange and MALDI-based technique, in studying the binding interaction between the molecular chaperone Hsp33 and four different unfolded protein substrates, including citrate synthase, lactate dehydrogenase, malate dehydrogenase, and aldolase. The results of our studies suggest that the cooperativity of the Hsp33 folding-unfolding reaction increases upon binding with denatured protein substrates. This is consistent with the burial of significant hydrophobic surface area in Hsp33 when it interacts with its substrate proteins. The SUPREX-derived K(d) values for Hsp33 complexes with four different substrates were all found to be within the range of 3-300 nM.

SplicerAV: a tool for mining microarray expression data for changes in RNA processing.

BACKGROUND: Over the past two decades more than fifty thousand unique clinical and biological samples have been assayed using the Affymetrix HG-U133 and HG-U95 GeneChip microarray platforms. This substantial repository has been used extensively to characterize changes in gene expression between biological samples, but has not been previously mined en masse for changes in mRNA processing. We explored the possibility of using HG-U133 microarray data to identify changes in alternative mRNA processing in several available archival datasets. RESULTS: Data from these and other gene expression microarrays can now be mined for changes in transcript isoform abundance using a program described here, SplicerAV. Using in vivo and in vitro breast cancer microarray datasets, SplicerAV was able to perform both gene and isoform specific expression profiling within the same microarray dataset. Our reanalysis of Affymetrix U133 plus 2.0 data generated by in vitro over-expression of HRAS, E2F3, beta-catenin (CTNNB1), SRC, and MYC identified several hundred oncogene-induced mRNA isoform changes, one of which recognized a previously unknown mechanism of EGFR family activation. Using clinical data, SplicerAV predicted 241 isoform changes between low and high grade breast tumors; with changes enriched among genes coding for guanyl-nucleotide exchange factors, metalloprotease inhibitors, and mRNA processing factors. Isoform changes in 15 genes were associated with aggressive cancer across the three breast cancer datasets. CONCLUSIONS: Using SplicerAV, we identified several hundred previously uncharacterized isoform changes induced by in vitro oncogene over-expression and revealed a previously unknown mechanism of EGFR activation in human mammary epithelial cells. We analyzed Affymetrix GeneChip data from over 400 human breast tumors in three independent studies, making this the largest clinical dataset analyzed for en masse changes in alternative mRNA processing. The capacity to detect RNA isoform changes in archival microarray data using SplicerAV allowed us to carry out the first analysis of isoform specific mRNA changes directly associated with cancer survival.

Trafficking of G protein-coupled receptors.

G protein-coupled receptors (GPCRs) play an integral role in the signal transduction of an enormous array of biological phenomena, thereby serving to modulate at a molecular level almost all components of human biology. This role is nowhere more evident than in cardiovascular biology, where GPCRs regulate such core measures of cardiovascular function as heart rate, contractility, and vascular tone. GPCR/ligand interaction initiates signal transduction cascades, and requires the presence of the receptor at the plasma membrane. Plasma membrane localization is in turn a function of the delivery of a receptor to and removal from the cell surface, a concept defined most broadly as receptor trafficking. This review illuminates our current view of GPCR trafficking, particularly within the cardiovascular system, as well as highlights the recent and provocative finding that components of the GPCR trafficking machinery can facilitate GPCR signaling independent of G protein activation.

Hybrid transgenic mice reveal in vivo specificity of G protein-coupled receptor kinases in the heart.

G protein-coupled receptor kinases (GRKs) phosphorylate activated G protein-coupled receptors, including alpha(1B)-adrenergic receptors (ARs), resulting in desensitization. In vivo analysis of GRK substrate selectivity has been limited. Therefore, we generated hybrid transgenic mice with myocardium-targeted overexpression of 1 of 3 GRKs expressed in the heart (GRK2 [commonly known as the beta-AR kinase 1], GRK3, or GRK5) with concomitant cardiac expression of a constitutively activated mutant (CAM) or wild-type alpha(1B)AR. Transgenic mice with cardiac CAMalpha(1B)AR overexpression had enhanced myocardial alpha(1)AR signaling and elevated heart-to-body weight ratios with ventricular atrial natriuretic factor expression denoting myocardial hypertrophy. Transgenic mouse hearts overexpressing only GRK2, GRK3, or GRK5 had no hypertrophy. In hybrid transgenic mice, enhanced in vivo signaling through CAMalpha(1B)ARs, as measured by myocardial diacylglycerol content, was attenuated by concomitant overexpression of GRK3 but not GRK2 or GRK5. CAMalpha(1B)AR-induced hypertrophy and ventricular atrial natriuretic factor expression were significantly attenuated with either concurrent GRK3 or GRK5 overexpression. Similar GRK selectivity was seen in hybrid transgenic mice with wild-type alpha(1B)AR overexpression concurrently with a GRK. GRK2 overexpression was without effect on any in vivo CAM or wild-type alpha(1B)AR cardiac phenotype, which is in contrast to previously reported in vitro findings. Furthermore, endogenous myocardial alpha(1)AR mitogen-activated protein kinase signaling in single-GRK transgenic mice also exhibited selectivity, as GRK3 and GRK5 desensitized in vivo alpha(1)AR mitogen-activated protein kinase responses that were unaffected by GRK2 overexpression. Thus, these results demonstrate that GRKs differentially interact with alpha(1B)ARs in vivo such that GRK3 desensitizes all alpha(1B)AR signaling, whereas GRK5 has partial effects and, most interestingly, GRK2 has no effect on in vivo alpha(1B)AR signaling in the heart.

Independent beta-arrestin 2 and G protein-mediated pathways for angiotensin II activation of extracellular signal-regulated kinases 1 and 2.

Stimulation of a mutant angiotensin type 1A receptor (DRY/AAY) with angiotensin II (Ang II) or of a wild-type receptor with an Ang II analog ([sarcosine1,Ile4,Ile8]Ang II) fails to activate classical heterotrimeric G protein signaling but does lead to recruitment of beta-arrestin 2-GFP and activation of extracellular signal-regulated kinases 1 and 2 (ERK1/2) (maximum stimulation approximately 50% of wild type). This G protein-independent activation of mitogen-activated protein kinase is abolished by depletion of cellular beta-arrestin 2 but is unaffected by the PKC inhibitor Ro-31-8425. In parallel, stimulation of the wild-type angiotensin type 1A receptor with Ang II robustly stimulates ERK1/2 activation with approximately 60% of the response blocked by the PKC inhibitor (G protein dependent) and the rest of the response blocked by depletion of cellular beta-arrestin 2 by small interfering RNA (beta-arrestin dependent). These findings imply the existence of independent G protein- and beta-arrestin 2-mediated pathways leading to ERK1/2 activation and the existence of distinct "active" conformations of a seven-membrane-spanning receptor coupled to each.