Thesis (M. Eng.)--Massachusetts Institute of Technology, Biological Engineering Division, 2007. / Includes bibliographical references (p. 53-57). / Alternative splicing is involved in numerous cellular functions and is often disrupted and involved in disease. Previous research has identified methods to distinguish alternative conserved exons (ACEs) in human and mouse. However, the cellular machinery, the spliceosome, does not use comparative genomics to decide when to include and when to exclude an exon. Human RefSeq exons obtained from the University of California Santa Cruz (UCSC) genome browser were analyzed for tissue-specific skipping. Expressed sequence tags (ESTs) were aligned to exons and their tissue of origin and histology were identified. ACEs were also identified as a subset of the skipped exons. About 18% of the exons were identified as tissue-specifically skipped in one of sixteen different tissues at four stringency levels. The different datasets were analyzed for both general features such as exon and intron length, splice site strength, base composition, conservation, modularity, and susceptibility to nonsense-mediated mRNA decay caused by skipping. Cis-element motifs that might bind protein factors that affect splicing were identified using overrepresentation analysis and conserved occurrence rate between human and mouse. / (cont.) Tissue-specific skipped exons were then classified with both a decision-tree based classifier (Random ForestsTM) and a support vector machine. Classification results were better for tissue-specific skipped exons vs. constitutive exons than for tissue-specific skipped exons vs. exons skipped in other tissues. / by Craig Jeremy Rothman. / M.Eng.

Thesis (M. Eng.)--Massachusetts Institute of Technology, Biological Engineering Division, 2005. / Includes bibliographical references (p. 77-80). / This study aims to investigate the relationships between growth parameters (agitation, glycerol concentration, salt concentration) and responses (biomass, growth rate, protein expression), by a 3-factor-3-level central composite factorial design. This experimental design involved running shake flask culture at 15 different experimental conditions with duplicates. Optical density (OD600), dry cell weight (DCW), and BCA Protein Assays were done on each experiment. Mathematical models in terms of these parameters' effects and their interactions were proposed for each of the responses. The significance of each effect and interaction, as well as the goodness-of-fit of mathematical models to data were examined by analysis of variance. It was found that biomass (with R²Adj=0.951) is a strong function of glycerol concentration (higher glycerol concentration leads to higher biomass), but it varies much less with agitation, and it is completely independent of salt concentration. Growth rate (R²Adj=0.901), however, varies strongly with agitation and salt concentration, but much more weakly with glycerol concentration. Protein production has a low R²Adj value of 0.746, implying that higher-order terms, e.g. x₁² and x₂², should be tested for significance in the model. / (cont.) Collected data were fitted to the proposed models by response surface regression, after which surface and contour plots of responses were generated to identify trends in them. High agitation (300 rpm in shaker) gave rise to both highest biomass and growth rate. In addition, biomass at high glycerol concentration (3% v/v) was almost twice as much as biomass at low glycerol concentration (1% v/v) at high agitation rate (19 g/L compared to 11 g/L). At the same agitation rate, growth rate shows the largest increase of 20.5% with increasing salt concentration from 0.7% to 2.1%. Protein production reached maximum of 7.3 mg/mL at medium agitation rate (250 rpm), high salt and glycerol concentrations. / by Joyce Chan. / M.Eng.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2008. / "September 2008." / Includes bibliographical references and index. / Glioblastoma multiforme (GBM) is the most aggressive adult brain tumor and remains incurable despite multimodal intensive treatment regimens. EGFRvIII is a truncated extracellular mutant of the EGF receptor (EGFR) that is commonly found in GBMs and confers tumorigenic behavior. Although much work has been done over the past decade to elucidate pathways involved in EGFRvIII receptor signaling, the global map of the signaling networks that it activates remains incomplete, making it difficult to assess downstream components involved in EGFRvill-mediated transformation. To gain a molecular understanding of the mechanisms by which EGFRvIII acts, we have employed a mass spectrometry-based phosphoproteomic approach to quantitatively map cellular signaling events activated by this receptor. Using this approach, we have determined the major downstream pathways activated as a function of titrated EGFRvIII receptor levels. This analysis highlighted several aspects of EGFRvIII tumor biology, including crosstalk between EGFRvIII and other receptor tyrosine kinases. Specifically, we have identified the c-Met receptor as a co-target in the treatment of EGFRvIII positive GBM cells, and have shown that an EGFR and c-Met combination inhibitor strategy may be applicable in overcoming the poor efficacy of EGFR kinase inhibitor monotherapy in GBM patients. We then went on to investigate the mechanisms by which signaling networks are regulated in response to site-specific tyrosine mutations on EGFRvIII. This analysis has revealed a receptor compensation mechanism that is capable of restoring network architecture, upon the loss of a major tyrosine phosphorylation site on EGFRvIII. / (cont.) This is, to our knowledge, the first demonstration of signal compensation at the level of receptor phosphorylation and highlights an unexpected level of complexity within the signaling network. Our data also indicates that EGFRvIII fine-tunes the activity of the Erk pathway; some Erk activity is required for growth but excessive pathway activation results in cell death. We believe that the sensitivity to modulation of the Erk pathway may be exploited as a potential means of therapy for EGFRvIII positive tumors. Taken together, our study highlights the utility of quantitative phosphoproteomic analysis as a tool to gain molecular insights in cancer biology and a means for drug target discovery. / by Hua Ming Paul Huang. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2006. / Vita. / Includes bibliographical references. / The continued success of protein therapeutics has put a strain on industry's ability to meet the large demand. Creating a more productive expression host for the manufacture of these proteins is a potential solution. Although heterologous proteins are frequently made in organisms as disparate as E. coli and bovines, the single-celled organism S. cerevisiae has emerged as a well-qualified candidate due to its approachable genetic and fermentation attributes as well as its ability to stably fold disulfide bonded and multi domain proteins. Because S. cerevisiae screens for enhanced protein secretion have traditionally utilized low-throughput and often plate-based methods, a high-throughput, liquid phase assay could offer a real advantage in secretory selection. In this thesis, yeast surface display is investigated as a potential proxy for heterologous protein secretion. Although ultimately unsuitable as a screening proxy, the surface display experiments did show a novel method of improving protein secretion by co-expressing a more stably folded protein with the protein of interest. In these studies the secretion of an scFv-Aga2p fusion was stimulated 1 0-fold by the concomitant surface expression of BPTI. / (cont.) BPTI surface expression also stimulated the secretion of secreted scFv three-fold suggesting a niche for protein coexpression as well as secretion by way of Aga2p fusions. A new screening method was developed that involves the capture of secreted protein on the surface of the cell where it can be labeled and sorted by FACS. This new method was verified to achieve thirty-five fold enrichment per pass for a three-fold enhanced protein secretor making it easily suitable for screening. The new screening methodology, the Cell Surface Secretion Assay (CeSSA), was also modeled and verified with time course data that enabled optimization of sort parameters and predicted sort outcomes based on user-derived selection parameters. The CeSSA was used to screen a library of mutant yeast alpha mating factor leader sequences for improved secretion of the scFv 4m5.3. The improved leaders imparted up to a twenty-fold improvement in scFv secretion per cell and up to thirty-fold improvement after expression tuning. These engineered leader sequences also conferred improved secretion on other scFv's and proteins including whole IgG. Moreover, the leader sequence mutants give indications of where the important residues in secretory leaders lie and the aberrations in protein traffic that result in reduced secretion. / by James A. Rakestraw. / Ph.D.

Thesis (Ph. D. in Applied Biosciences)--Massachusetts Institute of Technology, Biological Engineering Division, 2004. / Includes bibliographical references (leaves 119-130). / Tissue inhibitors of metalloproteinases (TIMPs) regulate tumor growth, progression and angiogenesis in a variety of experimental cancer models and in human malignancies. However, numerous studies have revealed important differences between TIMP family members in their ability to inhibit angiogenic processes in vitro and angiogenesis in vivo despite their universal ability to inhibit matrix metalloproteinase (MMP) activity. To address these differences, structure-function studies were conducted to identify and characterize the anti-angiogenic domains of TIMP-2, the endogenous MMP inhibitor that uniquely inhibits capillary endothelial cell (EC) proliferation and angiogenesis in vivo. Only the carboxy-terminal domain of TIMP-2 (T2C) and not the MMP-inhibitory N-terminal domain (T2N), inhibited capillary EC proliferation. Although both T2N and T2C inhibited embryonic angiogenesis, only T2C potently inhibited mitogen-stimulated angiogenesis. These findings demonstrate that TIMP-2 possesses two distinct types of anti-angiogenic activities which can be uncoupled from each other. The anti-proliferative activity of T2C was further mapped to the 24-amino acid peptide, Loop 6, which proved to be a potent inhibitor of both embryonic and nitogen-stimulated angiogenesis in vivo. Initial studies into the mechanism(s) by which Loop 6 inhibits angiogenesis revealed that the anti-proliferative effects of Loop 6 are due, at least in part, to the inhibition of cell cycle progression and not to the induction of apoptosis. This inhibition was associated with increased levels of cell cycle inhibitor p27. Although Loop 6 did not compete with bFGF for binding to its receptor, / (cont.) five potential cell surface complexes were observed in crosslinking studies of capillary EC treated with ¹²⁵I-labeled T2C or Loop 6. Finally, given the high degree of homology between TIMP-2 and TIMP-4, we hypothesized that TIMP-4 might share anti-proliferative and MMP inhibition- independent anti-angiogenic activities with TIMP-2. Our results demonstrate that although TIMP-4 inhibits capillary EC migration, it does not inhibit capillary EC proliferation. Furthermore, TIMP-4 did not result in significant inhibition of embryonic angiogenesis in the CAM. These results suggest that TIMP-2 is unique among TIMP family members in its ability to inhibit angiogenesis via two distinct pathways. One of these activities, housed within Loop 6, results in the potent inhibition of angiogenesis in vivo. / by Cecilia A. Fernández. / Ph.D.in Applied Biosciences

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, February 2008. / Vita. / Includes bibliographical references. / Oxidatively damaged DNA has been strongly associated with cancer, chronic degenerative diseases and aging. Guanine is the most frequently oxidized base in the DNA, and generation of a guanine radical cation (G'") as an intermediate in the oxidation reaction leads to migration of a resulting cationic hole through the DNA n-stack until it is trapped at the lowest-energy sites. These sites reside at runs of guanines, such as 5'-GG-3' sequences, and are characterized by the lowest sequence-specific ionization potentials (IPs). The charge transfer mechanism suggests that hotspots of oxidative DNA damage induced by electron transfer reagents can be predicted based on the primary DNA sequence. However, preliminary data indicated that nitrosoperoxycarbonate (ONOOCO2"), a mediator of chronic inflammation and a one-electron oxidant, displayed unusual guanine oxidation properties that were the focus of present work. As a first step in our study, we determined relative levels of guanine oxidation, induced by ONOOCO2 in all possible three-base sequence contexts (XGY) within double-stranded oligonucleotides. These levels were compared to the relative oxidation induced within the same guanines by photoactivated riboflavin, a one-electron reagent. We found that, in agreement with previous studies, photoactivated riboflavin was selective for guanines of lowest IPs located within 5'-GG-3' sequences. In contrast, ONOOCO2" preferentially reacted with guanines located within 5'-GC-3' sequences characterized by the highest IPs. This demonstrated that that sequence-specific IP was not a determinant of guanine reactivity with ONOOCO2". Sequence selectivities for both reagents were double-strand specific. Selectivity of ONOOCO2 for 5'-GC-3' sites was also observed in human genomic DNA after ligation-mediated PCR analysis. / (cont.) Relative yields of different guanine lesions produced by both ONOOCO2" and riboflavin varied 4- to 5-fold across all sequence contexts. To assess the role of solvent exposure in mediating guanine oxidation by ONOOCO2", relative reactivities of mismatched guanines with ONOOCO2" were measured. The majority of the mismatches displayed an increased reactivity with ONOOCO2 as compared to the fully matched G-C base-pairs. The extent of reactivity enhancement was sequence context-dependent, and the greatest levels of enhancement were observed for the conformationally flexible guanine- guanine (G-G) mismatches and for guanines located across from a synthetic abasic site. To test the hypothesis that the negative charge of an oxidant influences its reactivity with guanines in DNA, sequence-selective guanine oxidation by a negatively charged reagent, Fe+2-EDTA, was assessed and compared to guanine oxidation produced by a neutral oxidant, y-radiation. Because both of these agents cause high levels of deoxyribose oxidation, a general method to quantify sequence-specific nucleobase oxidation in the presence of direct strand breaks was developed. This method exploited activity of exonuclease III (Exo III), a 3' to 5' exonuclease, and utilized phosphorothioate-modified synthetic oligonucleotides that were resistant to Exo III activity. This method was employed to determine sequence-selective guanine oxidation by Fe+2-EDTA complex and y-radiation and to show that both agents produced identical guanine oxidation pattems and were equally reactive with all guanines, irrespective of their sequence-specific IPs or sequence context. / (cont.) This showed that negative charge was not a determinant of Fe+2-EDTA-mediated guanine oxidation. Finally, the role of oxidant binding on nucleobase damage was assessed by studying sequence-selective oxidation produced by DNA-bound Fe+2 ions in the presence of H202. We found that the major oxidation targets were thymines located within 5'-TGG-3' motifs, demonstrating that while guanines were a required element for coordination of Fe+2 to DNA, they were not oxidized. Our results suggest that factors other than sequence-specific IPs can act as major determinants of sequence-selective guanine oxidation, and that current models of guanine oxidation and charge transfer in DNA cannot be used to adequately predict the location and identity of mutagenic lesions in the genome. / by Yelena Margolin. / Ph.D.

Thesis (Sc. D.)--Massachusetts Institute of Technology, Biological Engineering Division, February 2007. / Includes bibliographical references (leaves 108-124). / Our understanding of why biological systems are designed in a particular way would benefit from biophysically-realistic models that can make accurate predictions on the time-evolution of molecular events given arbitrary arrangements of genetic components. This thesis is focused on constructing such models for gene expression during bacteriophage T7 infection. T7 gene expression is a particularly well suited model system because knowledge of how the phage functions is thought to be relatively complete. My work focuses on two questions in particular. First, can we address deficiencies in past simulations and measurements of bacteriophage T7 to improve models of gene expression? Second, can we design and build refactored surrogates of T7 that are easier to understand and model? To address deficiencies in past simulations and measurements, I developed a new single-molecule, base-pair-resolved gene expression simulator named Tabasco that can faithfully represent mechanisms thought to govern phage gene expression. I used Tabasco to construct a model of T7 gene expression that encodes our mechanistic understanding. The model displayed significant discrepancies from new system-wide measurements of absolute T7 mRNA levels during infection. / (cont.) I fit transcript-specific degradation rates to match the measured RNA levels and as a result corrected discrepancies in protein synthesis rates that confounded previous models. I also developed and used a fitting procedure to the data that let us evaluate assumptions related to promoter strengths, mRNA degradation, and polymerase interactions. To construct surrogates of T7 that are easier to understand and model, I began the process of refactoring the T7 genome to construct an organism that is a more direct representation of the models that we build. In other words, instead of making our models evermore detailed to explain wild-type T7, we started to construct new phage that are more direct representations of our models. The goal of our original design, T7. 1, was to physically define, separate, and enable unique manipulation of primary genetic elements. To test our initial design, we replaced the left 11,515 bp of the wild-type genome with 12,179 bp of engineered DNA. The resulting chimeric genome encodes a viable bacteriophage that appears to maintain key features of the original while being simpler to model and easier to manipulate. I also present a second generation design, T7.2, that extends the original goals of T7.1 by constructing a more direct physical representation of the T7 model. / by Sriram Kosuri. / Sc.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2008. / Includes bibliographical references (leaves 66-75). / Eight distinct nuclear shapes, or morphologies, have been discovered in human proto-organs and tumors, including bell-shaped nuclei with stem-like properties. These bell-shaped, or "metakaryotic," nuclei are abundant in fetal tissues and neoplasias, but rare in normal adult somatic tissues. Metakaryotic nuclei employ an unusual process for division in which DNA synthesis, partial genomic condensation, and separation of the two nuclei in a cup-from-cup fashion occur concurrently, as shown by Feulgen densitometry and single-stranded DNA assays by Dr. Elena Gostjeva. This is clearly different from the sequential steps of S-phase DNA synthesis, chromatin condensation, chromosomal separation, and genomic segregation that occur in mitotic eukaryotic cells. In order to discover how a genome apparently devoid of chromosomes might be organized, this thesis focused on recognizable DNA sequences common to all chromosomes: centromeres and telomeres. Fluorescence In Situ Hybridization (FISH) with pan-centromeric and pan-telomeric probes was applied to samples of human tissue. (A collaborating lab used centromeric and telomeric antibodies to confirm results.) An optimized FISH protocol was developed specifically for metakaryotic nuclei and tested in both human cell lines and eukaryotic cells as experimental controls. Staining of metakaryotic nuclei resulted in approximately 23 centromeric regions in each, unlike the expected number of 46 regions seen in eukaryotic nuclei. Many of these staining regions contained paired centromere signals, or doublets. This suggested a genomic organization of homologous chromosomes paired at their centromere regions. If this were the case, one would expect 46 telomeric signals per nuclei, if telomeres were also homologously paired. / (cont.) Unexpectedly, an average of 23 telomeric regions were found in many, if not all, bell-shaped metakaryotic nuclei. This, along with the observation of a condensed double ring around the mouth of the bell-shaped nuclei, suggested the possibility of a genome organized as paired, continuous genomic circles. Studies of telomere joining in metakaryotic nuclei by Dr. Per Olaf Ekstrom have provided further evidence for the paired genomic circle model. The results in this thesis are an original contribution to the field of stem cell physiology, a starting point for further investigation of DNA organization, synthesis, and repair in these metakaryotic cells, and hopefully will lead to a greater understanding of human development, growth, and cancer. / by Amanda Natalie Gruhl. / Ph.D.

Thesis (S.M.)--Massachusetts Institute of Technology, Biological Engineering Division, 2005. / Includes bibliographical references (leaves 40-43). / Adult stem cells are potentially useful in many biomedical applications that can save lives and increase the quality of a patient's life, such as tissue engineering, cell replacement, and gene therapy. However, these applications are limited because of the difficulty in isolating and expanding pure populations of adult stem cells (ASCs). A major barrier to ASC expansion in vitro is their property of asymmetric cell kinetics. Our lab has developed a method, Suppression of Asymmetric Cell Kinetics (SACK), to expand ASCs in vitro by shifting their cell kinetics program from asymmetric to symmetric. We have found that guanine nucleotide precursors can be used to convert the kinetics of adult stem cells from asymmetric to symmetric, which promotes their exponential expansion. Previously, we have used the SACK method to derive hepatic and cholangiocyte stem cell strains from adult rat livers in vitro. These cell strains provide an assay to evaluate whether growth factors and cytokines previously implicated in proliferation of progenitor cells act by converting the kinetics of the stem cells in the population from asymmetric to symmetric, and thus identify new SACK agents. We are evaluating three agents, Wnt, IGF- 1, and Sonic hedgehog (Shh). / (cont.) Wnt has been found to cause self-renewal and proliferation of hematopoietic stem cells (HSCs) in vitro. IGF- 1 also plays a role in hematopoietic progenitor self-renewal in vivo as well as in tissue maturation. Shh has been implicated in the proliferation of primitive neural cells as well as in cellular proliferation during invertebrate development. Thus far, we have found that Wnt peptide shifts the cell kinetics from asymmetric to symmetric and may reduce the generation time, whereas IGF-1 appears only to affect generation time. Studies involving Shh are currently underway. We are also currently investigating whether Wnt acts additively or synergistically with guanine nucleotide precursors to shift cell kinetic symmetry. Discovering new SACK agents will allow us to obtain purer populations of ASCs that can be used to study properties unique to stem cells. Furthermore, the observation that Wnt shifts the kinetics of adult rat hepatic stem cells from asymmetric to symmetric implicates the involvement of similar cell kinetics symmetry mechanisms in the proliferation effect of Wnt on murine and human HSCs. / by Michal Ganz. / S.M.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2008. / Includes bibliographical references (leaves 244-249). / This thesis utilizes modeling approaches rooted in statistical physics and physical chemistry to investigate several aspects of cellular signal transduction at both the modular and global levels. Design principles of biological networks and cell signaling processes pertinent to disease progression emerge from these studies. It is my hope that knowledge of these principles may provide new mechanistic insights and conceptual frameworks for thinking about therapeutic intervention into diseases such as cancer and diabetes that arise from aberrant signaling. Areas of interest have emphasized the role of scaffold proteins in protein kinase cascades, modeling relevant biophysical processes related to T cell activation, design principles of signal transduction focusing on multisite phosphorylation, quantifying the notion of signal duration and the time scale dependence of signal detection, and entropy based models of network architecture inferred from proteomics data. These problems are detailed below. The assembly of multiple signaling proteins into a complex by a scaffold protein guides many cellular decisions. Despite recent advances, the overarching principles that govern scaffold function are not well understood. We carried out a computational study using kinetic Monte Carlo simulations to understand how spatial localization of kinases on a scaffold may regulate signaling under different physiological condition. Our studies identify regulatory properties of scaffold proteins that allow them to both amplify and attenuate incoming signals in different biological contexts. In a further, supplementary study, simulations also indicate that a major effect that scaffolds exert on the dynamics of cell signaling is to control how the activation of protein kinases is distributed over time[2]. / (cont.) Scaffolds can influence the timing of kinase activation by allowing for kinases to become activated over a broad range of times, thus allowing for signaling across a broad spectrum of time scales. T cells orchestrate the adaptive immune response and are central players in maintenance of functioning immune system. Recent studies have reported that T cells can integrate signals between interrupted encounters with Antigen Presenting Cells (APCs) in such a way that the process of signal integration exhibits a form of memory. We carried out a computational study using a simple mathematical model of T cell activation to investigate the ramifications of interrupted T cell-APC contacts on signal integration. We considered several mechanisms of how signal integration at these time scales may be achieved. In another study, we investigated the role of spatially localizing signaling components of the T cell signaling pathway into a structure known as the immunological synapse. We constructed a minimal mathematical model that offers a mechanism for how antigen quality can regulate signaling dynamics in the immunological synapse These studies involving the analysis of signaling dynamics led us to investigate how differences in signal duration might be detected. Signal duration (e.g. the time scales over which an active signaling intermediate persists) is a key regulator of biological decisions in myriad contexts such as cell growth, proliferation, and developmental lineage commitments. Accompanying differences in signal duration are numerous downstream biological processes that require multiple steps of biochemical regulation. We present an analysis that investigates how simple biochemical motifs that involve multiple stages of regulation can be constructed to differentially process signals that persist at different time scales[3]. / (cont.) Topological features of these networks that allow for different frequency dependent signal processing properties are identified. One role of multisite phosphorylation in cell signaling is also investigated. The utilization of multiple phosphorylation sites in regulating a biological response is ubiquitous in cell signaling. If each site contributes an additional, equivalent binding site, then one consequence of an increase in the number of phosphorylations may be to increase the probability that, upon disassociation, a ligand immediately rebinds to its receptor. How such effects may influence cell signaling systems is not well understood. A self-consistent integral equation formalism for ligand rebinding, in conjunction with Monte Carlo simulations, was employed to further investigate the effects of multiple, equivalent binding sites on shaping biological responses. Finally, this thesis also seeks to investigate cell signaling at a global scale. Advances in Mass Spectrometry based phosphoproteomics have allowed for the real-time quantitative monitoring of entire proteomes as signals propagate through complex networks in response to external signals. The trajectories of as many as 222 phosphorylated tyrosine sites can be simultaneously and reproducibly monitored at multiple time points. We develop and apply a method using the principle of maximum entropy to infer a model of network connectivity of these phosphorylation sites. The model predicts a core structure of signaling nodes, affinity dependent topological features of the network, and connectivity of signaling nodes that were hitherto unassociated with the canonical growth factor signaling network. Our combined results illustrate many complexities in the broad array of control properties that emerge from the physical effects that constrain signal propagation on complex biological networks. / (cont.) It is the hope of this work that these studies bring coherence to seemingly paradoxical observations and suggest that cells have evolved design rules that enable biochemical motifs to regulate widely disparate cellular functions. / by Jason W. Locasale. / Ph.D.