Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2005. / Includes bibliographical references. / Chondroitin sulfate is a critical component of articular cartilage due to its contribution to the tissue's resistance to compressive deformation. Alterations in the biosynthesis of this molecule over time could impact the ability of the tissue to perform its necessary functions. Several factors have been shown to alter the biosynthesis of chondroitin sulfate in cartilage; among them are age, disease, depth of tissue, and mechanical compression. Specifically, mechanical compression has been shown to have a significant effect on the sulfation pattern and chain length and number in cartilage explant studies. The mechanisms that govern these alterations, however, have not been determined. The purpose of this study is to examine the effects of mechanical compression on chondroitin sulfate biosynthesis and analyze the roles of two possible mechanisms; enzyme transcription and organelle deformation. The effects of mechanical compression on the transcription rates of enzymes associated with the biosynthesis of chondroitin sulfate have not been previously studied. To perform this study in a bovine model, portions of the bovine genome had to be sequenced, PCR primers designed, and bulk expression levels determined. Static compression resulted in the significant up-regulation of two genes of interest: chondroitin sulfate and GalNAc 4S,6-sulfotransferase. / (cont.) Dynamic compression resulted in the significant up-regulation of the three sulfotransferases responsible for the bulk of sulfation in cartilage tissue. These results indicate a transient mechanotransduction reaction that differs based on the load regime. The effect of mechanical loading on the biosynthesis of chondroitin sulfate has been studied previously, however, this study seeks to examine more comprehensive loading regimes. Static compression and release resulted in an increase in 6-sulfation and a decrease in 4-sulfation that lasted to 48 hours after release of compression. Dynamic compression and release had the opposite effect on sulfation ratio, with an increase in 4-sulfation compared to 6-sulfation. The transcription changes seen in this study do not indicate the changes that occur in the end products of synthesis. Other factors may play a larger role, such as precursor availability or transport through the Golgi apparatus. Intracellular organelles react to static compression of the surrounding tissue in one of two manners. The majority of organelles deform much as the nucleus, proportionally in volume and shape to the cell. The Golgi apparatus appears to retain a significant portion of its volume relative to the cell and other organelles. In addition, it reforms structurally into a highly ordered stacked appearance. / (cont.) Osmotic forces within the Golgi may allow it to balance the osmotic load in the cytoplasm and resist compression and altered trafficking of the Golgi may in turn produce the altered appearance. Recent microscopy experiments on the Golgi apparatus utilizing two-photon microscopy have allowed us to examine the reaction of live tissue to static compression. These results illustrate the significant, but differing, effects of static and dynamic compression on the biosynthesis of chondroitin sulfate. The effects of these compression types on the transcription of enzymes responsible for this biosynthesis cannot fully explain the changes seen in newly synthesized chondroitin sulfate. Organelle reorganization has been shown to occur in response to static load and it is possible that altered organelle trafficking plays a role in this altered biosynthesis. Further studies are necessary to determine the final effect of the altered transcription and organelle structure on the manufacture of this important cartilage molecule. / by Jon D. Szafranski. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2005. / Includes bibliographical references. / Acute traumatic joint injury in young adults leads to an increased risk for the development of osteoarthritis (OA) later in life irrespective of surgical intervention to stabilize the injured joint. Although the mechanism by which injury leads to joint degeneration remains to be elucidated, several injury-related factors may contribute to the development of OA. These factors include but are not limited to altered mechanical loading and initiation of a cellular response in cartilage or other joint tissues at the time of the injury. Three in vitro models of joint injury were investigated to separately evaluate the effects on cartilage of mechanical overloading and damage to joint capsule and synovial lining. Models of injury included (1) mechanical injury lto cartilage explants alone, (2) co-culture of normal cartilage explants with an excised specimen of joint capsule tissue, and (3) co-culture of mechanically injured cartilage explants with excised joint capsule tissue. These models have been shown previously to result in matrix damage and decreased biosynthesis by the chondrocytes. We measured gene expression levels of matrix molecules and matrix proteases and found them to be expressed in control cartilage at levels ranging over five orders of magnitude, and to be differentially regulated in these three models of joint injury. / (cont.) Expression of matrix molecules including collagen II and aggrecan were unaffected by injurious compression or co-culture with joint capsule tissue during the first 24 hours; however, the combination of injurious compression followed by co-culture resulted in a -50% decrease in expression by 24 hours. Matrix proteases aggrecanase-2 (ADAMTS-5) and stromelysin (MMP-3) showed increased expression of 40-250-fold by 12 hours following injurious compression and 6-12-fold during 24 hours of co-culture with joint capsule tissue. Aggrecanase-1 (ADAMTS-4) and collagenase-3 (MMP-13) showed larger magnitude increases in expression during co-culture (6-8-fold; 6-24 hours) compared to injurious compression (2-4-fold; 6-24 hours). Expression of transcription factors, c-fos and c-jun, was rapidly increased by injurious compression (40-100-fold within one hour) but was less affected by co-culture with joint capsule tissue (increased 3-5-fold; 1-24 hours). Expression level results displayed a general trend toward matrix degradation in the models of joint injury with specific differences apparent between the models. Analysis of matrix protein fragments in the same injury models showed cleavage of aggrecan at the aggrecanase site in the interglobular domain by 16 days following injurious compression and during co-culture with joint capsule tissue. / (cont.) Equilibrium and dynamic stiffness of cartilage explants were decreased by 30-35% immediately after injurious compression but were unaffected through 16 days of co-culture with joint capsule tissue. Specific changes in gene expression and activity of matrix proteases observed in these injury models may be indicative of some of the molecules responsible in the initial phase of cartilage degradation observed clinically following joint injury. / Jennifer H. Lee. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2008. / Includes bibliographical references (p. 82-92). / The practical viability of biomolecule-nanostructure hybrids depends critically on the functional and structural stability of biomolecules in application environments. Noncovalent interactions of biochemical functional groups with nanostructure surfaces can significantly disrupt biomolecular structure and function. We report a systematic study of the effect of DNA sequence on the binding interaction between gold nanoparticles and thiolated DNA (AuNp-DNA). Base specific noncovalent nucleotide adsorption on gold surfaces can affect nucleotide bioavailability in solution. Systematic investigation of DNA oligonucleotide sequence, the location of specific sequence motifs, and the effect of nanoparticle size was performed. Sequence effects on DNA coverage and oligonucleotide adsorption affinities.were studied by Langmuir isotherm analysis. The nanoparticle coverage at saturating concentrations of thiolated DNA varied with oligonucleotide sequence. Saturation coverages correlated well with complement hybridization efficiency. From this we concluded that noncovalent interactions between nucleotides and the particle surface effect both hybridization and DNA coverage and adsorption. This hypothesis was confirmed by chemical treatment of the particle surface to eliminate noncovalent interactions. Upon treatment the effect of sequence on hybridization efficiency was removed. The effect of sequence is not consistent across nanoparticle sizes. Different bases show the highest saturation coverages and hybridization efficiencies on different AuNp sizes. These results allow for sequence selection guidelines based on AuNp size for sizes ranging from 4-11nm. For smaller particles (<5nm) adenine rich sequences show the highest saturation coverage and hybridization efficiency. / (cont.) For mid-sized particles (~7.5nm), guanine sequences show the highest saturation coverage and hybridization efficiency. Larger particles (>10nm) show little sequence dependent behavior and are likely the best choice for uses where sequence choice is limited. Sequence selection based on these guidelines will provide AuNp-DNA conjugates with the highest possible oligonucleotide bioavailability, maximizing their utility in biotechnology applications. / by Katherine A. Brown. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2006. / Includes bibliographical references. / Ex vivo expansion of hematopoietic stem cells (HSCs) is a long-standing challenge faced by both researchers and clinicians. To date, no robust, efficient method for the pure, ex vivo expansion of human HSCs has been demonstrated. Previous methods primarily induced the expansion of committed hematopoietic progenitor cells (HPCs), yielding even less pure populations of HSCs. This research was based on the hypothesis that, like for other adult stem cells (ASCs), the major barrier to expanding HSCs ex vivo is in preferentially regulating the asymmetric self-renewal of HSCs without loss in their ability to produce differentiated committed HPCs. This laboratory has shown that a p53-dependent pathway specifically controls the self-renewal pattern of several types of ASCs and thereby provides an effective means for expansion of ASCs in culture. The method, which involves the use of purine metabolites to achieve suppression of asymmetric cell kinetics, is referred to as SACK. The utility of the p53-dependent pathway was investigated for directing expansion of human HSCs. In order to support this investigation, the proliferation of HPCs in in vitro cultures was repressed by culturing cells without hematopoietic growth factors and cytokines. / (cont.) This allowed the in vitro detection of SACK-effects on a small sub-population of cells, predicted to include HSCs. In order to determine the self-renewal capacity and multilineage potential of SACK- cultured cells, they were transplanted into non-obese diabetic/severe combined immunodeficient (NOD/SCID) mice. In vivo transplantation investigations exhibited 1.9- fold to 4.5-fold increased engraftment efficiency with SACK-agents compared to SACK-free controls, suitable for clinical applications. This result suggests that SACK-culture expands a population of SCID-repopulating cells (SRCs) that yields self-renewal and multilineage engraftment in NOD/SCID mice. Accordingly, increased engraftment efficiency for successful clinical applications may be achieved after additional optimization of HSC expansion. To obtain the full therapeutic potential of expanded HSCs, development of methods for independently marking putative ASCs for future analyses and gene therapy was explored. This early success with human HSCs supports the basic hypothesis that the SACK approach may be applicable to expansion of many types of ASCs. / by Rouzbeh R. Taghizadeh. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2004. / Includes bibliographical references. / Biological cells such as endothelial or muscle cells respond to mechanical stimulation with activation of specific intracellular and extracellular signaling pathways and cytoskeletal remodeling, a process termed mechanotransduction. Intracellular mechanosensors are thought to be activated by conformational changes induced by local cellular deformations. Since these mechanosensors have been speculated to be located in several cellular domains including the cell membrane, the cytoskeleton, and the nucleus, it is necessary to achieve a detailed understanding of subcellular mechanics. In this work, we present novel methods to independently quantify cytoskeletal displacements, mechanical coupling between the cytoskeleton and the extracellular matrix, and nuclear mechanics based on high resolution tracking of cellular structures and receptor bound magnetic beads in response to applied strain or microscopic forces. These methods were applied to study the effects of several human disease associated mutations on subcellular mechanics and to examine the interaction between known protein function and specific changes in cellular mechanical properties and mechanotransduction pathways. Initial experiments were targeted to the role of membrane adhesion receptors. Experiments with cells expressing a mutant form of the integrin-associated molecule tetraspanin CD151 revealed that CD151 plays a key role in selectively strengthening α6βl integrin-mediated adhesion to laminin-1. We then studied cytoplasmic behavior using cells from mice with an αB-Crystallin mutation (R120G) that causes desmin-related myopathy. These studies showed impaired passive cytoskeletal mechanics in adult mouse cardiac myocytes. Finally, we studied cells deficient in the nuclear envelope / (cont.) protein lamin A/C and showed that lamin A/C deficient cells have increased nuclear deformation, defective mechanotransduction, and impaired viability under mechanical strain, suggesting that the tissue specific effects observed in laminopathies such as Emery-Dreifuss muscular dystrophy or Hutchinson-Gilford progeria may arise from varying degrees of impaired nuclear mechanics and transcriptional regulation. In conclusion, our methods provide new and valuable tools to examine the role of subcellular biomechanics on mechanotransduction in normal and mutant cells, leading to improved understanding of disease mechanisms associated with altered cell mechanics. / by Jan Lammerding. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2006. / Includes bibliographical references. / Endogenous and exogenous oxidants react with cellular macromolecules to generate a variety of electrophiles that react with DNA produce cytotoxic and mutagenic adducts. One source of such electrophiles is deoxyribose in DNA itself. Oxidation of each position in deoxyribose generates a unique spectrum of products, many of which are highly reactive with DNA bases and lead to formation of adducts. The objective of this thesis was to clarify the chemistry of deoxyribose oxidation, with a focus on C4'-oxidation that gives rise to 3'- phosphoglycolate residues on the DNA backbone and releases base propenal or malondialdehyde, and to investigate the role of base propenals in the formation of an important endogenous DNA adduct, MI dG. First, an index of total deoxyribose oxidation was developed, one that provides a means to compare different oxidizing agents. This method exploits the reaction of aldehyde- and ketone-containing deoxyribose oxidation products with 14C-methoxyamine to form stable oxime derivatives that are quantified by accelerator mass spectrometry. Sensitive GC/MS methods were developed to quantify 3'-phosphoglycolate residues from deoxyribose C4'-oxidation and HPLC/post-column derivatization methods were developed to quantify the corresponding base propenal or malondialdehyde. / (cont.) Combined with the quantification of total deoxyribose oxidation and the alternative product of C4'-oxidation, the 4'-ketoaldehyde abasic site, under the same conditions, these results offered direct insights into the partitioning of C4'-oxidation and the chemical mechanisms of deoxyribose oxidation in DNA. With a foundation of deoxyribose oxidation chemistry and analytical methods, the in vitro DNA oxidative damage induced by y-irradiation, Fe2+/EDTA, bleomycin and peroxynitrite was explored. The results revealed that malondialdehyde was neither sufficient nor necessary for the formation of MldG, while base propenal was effective in generating MldG. These observations were extended to an E. coli cell model in which the membrane content of polyunsaturated fatty acids was controlled. The results revealed that lipid peroxidation caused by y-irradiation was insufficient to produce MldG in cells and the level of MldG adducts was inversely correlated with the quantity of membrane polyunsaturated fatty acids when cells were treated with peroxynitrite. Finally, M1dG showed a moderate (-50%) increase in tissues from a mouse model of inflammation, while etheno-adducts induced by lipid peroxidation increased -3- fold. These results are again consistent with lipid peroxidation as a minor source of MldG. / by Xinfeng Zhou. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Biological Engineering Division, 2005. / Includes bibliographical references (p. 252-285). / Glycosaminoglycans (GAGs) are complex polysaccharides that exist both on the cell surface and free within the extracellular matrix. The intrinsic sequence variety stemming from the large number of building blocks that compose this biopolymer leads to substantial information density as well as to the ability to regulate a wide variety of important biological processes. With the recent and progressive emergence of biochemical and analytical tools to probe GAG structure and function, efforts can be taken to understand the role of GAGs in cell biology and in disease in the various physiological locations where GAGs can exist. As a first step to probe the functions of GAGs, the heparin/heparan sulfate-GAG (HSGAG)-fibroblast growth factor (FGF) system was examined. Understanding the role of HSGAGs in inducing FGF2 dimerization led to the development of a novel engineered protein that was found to be effective at promoting functional recovery in stroke. Subsequently, methods to isolate HSGAGs from the cell surface were optimized and the ability of HSGAGs to support FGF signaling was investigated. Cell surface HSGAGs can define the responsiveness of a given cell to FGF1 and FGF2 through multiple receptor isoforms. Stromal cell derived HSGAGs were also identified as critical regulators of tumor cell growth and metastasis, effecting not only FGF2., but also 1-integrin signaling. / (cont.) Other GAGs, including dermatan sulfates, were characterized as modulators of FGFs and vascular endothelial growth factors. Finally, FGFs and HSGAGs were found to have important roles in maintaining epithelial monolayer integrity, with syndecan-l serving as a critical factor in inflammatory bowel disease. In addition to understanding HSGAGs in their normal physiological settings, techniques to internalize them were developed. Poly(3-amino ester)s were found to condense heparin and enable its endocytosis into cells. Internalized heparin is preferentially taken up by cancer cells, which often have a faster endocytic rate than non-transformed cells, and promotes apoptotic cell death. Internalized heparin can also be used as a tool to probe cell function. In Burkitt's lymphoma, poly(3-amino ester)-heparin conjugates served to identify cell surface HSGAGs as an important modulator of cell growth that can be harnessed to inhibit growth. Finally, studies that sought to broaden the scope of GAG biology were undertaken. Cell surface HSGA(:is were identified as mediators of vascular permeability. Furthermore a novel technique to immobilize GAGs was employed. The interactions between GAG and substrate were via hydrogen bonding. Immobilization of GAGs alters their properties, such that they can affect cells in ways distinct from GAGs free in the ECM. / (cont.) Furthermore, immobilized GAGs can regulate cancer cell adhesion, growth and progression, and may offer a new way to regulate the activity of cancer cells. In addition to directly providing new potential therapeutics and drug targets, these studies represent a foundation to enable additional studies of GAG function. Future work harnessing the techniques presented may open new avenues of research and facilitate the development of novel GAG-based therapeutics. / by David Berry. / Ph.D.

Thesis (S.M.)--Massachusetts Institute of Technology, Biological Engineering Division, 2006. / Includes bibliographical references (leaves 44-45). / There is a great need in the fields of biology, medicine, and pharmaceuticals to create high-throughput devices for the detection of specific cell states in a heterogeneous mixture of cells. The desire is to differentiate among diseased and healthy cells, cell age, and cell type with the minimum amount of sample pretreatment. This project addresses this need by developing microfluidic devices that exploit the adhesion differences between cell states and cell types to rapidly count cells of different types without the need for labels. There are two avenues in which to explore cell adhesion differences with these devices, the first is a net electrostatic change at the surface of the cell wall and the second is the presence of specific cell-membrane adhesion proteins. It is hypothesized that the forced interaction of the cell wall with the microfabricated microcapillary walls would result in a differential velocity based on cell type that could be detected simply using a microscope and video camera or an interferometer. The eventual integration of cell velocity detection would result in a portable all-inclusive lab-on-a-chip system that could be used in the field for detecting the presence of diseases, such as malaria and cancer as well as in a lab setting for drug discovery. / by Kristen M. Naegle. / S.M.

Thesis (Ph. D. in Molecular and Systems Bacterial Pathogenesis)--Massachusetts Institute of Technology, Biological Engineering Division, 2005. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Includes bibliographical references. / Helicobacter hepaticus infection of A/JCr mice is a model of liver cancer resulting from chronic active inflammation. We monitored hepatic global gene expression profiles and correlated them to histological liver lesions in H. hepaticus infected and control male A/JCr mice at 3 months, 6 months, and 1 year of age. We used an Affymetrix-based oligonucleotide microarray platform on the premise that a specific genetic expression signature at isolated time points would be indicative of disease status. Model based expression index comparisons generated by dChip yielded consistent profiles of differential gene expression for H. hepaticus infected male mice with progressive liver disease versus uninfected control mice within each age group. Linear discriminant analysis and principal component analysis allowed segregation of mice based on combined age and lesion status, or age alone. Up-regulated genes present throughout the 12 month study involved inflammation, tissue repair, and host immune function. Upregulation of putative tumor and proliferation markers correlated with advancing hepatocellular dysplasia. Transcriptionally down-regulated genes in mice with liver lesions included those related to peroxisome proliferator, cholesterol, and steroid metabolism pathways. Transcriptional profiling of hepatic genes documented gene expression signatures in the livers of H. hepaticus infected male A/JCr mice with chronic progressive hepatitis and preneoplastic liver lesions, complemented the histopathological diagnosis, and suggested molecular targets for the monitoring and intervention of disease progression prior to the onset of hepatocellular neoplasia. Our laboratory, in collaboration with Professors Suerbaum and Schauer, recently identified a / (cont.) 70kb genomic island in Helicobacter hepaticus strain ATCC 51488 as a putative pathogenicity island (HhPAI) (Suerbaum et al, PNAS, 2003). This region within H. hepaticus contains genes HH0233-HH0302, a differential GC content, several long tandem repeats but no flanking repeats, and three components of a type IV secretion system (T4SS). A/JCr mice were experimentally infected with three naturally occurring strains of H. hepaticus including the type strain H. hepaticus ATCC 51488 strain (Hh 3B1) isolated from A/JCr mice, MIT 96-1809 (Hh NET) isolated from mice shipped from the Netherlands, and MIT-96-284 (HhG) isolated from mice acquired from Germany.4 HhNET (missing most of the HhPAI) infected male A/JCR mice exhibited a significantly lower prevalence (p<.05) of hepatic lesions at 6 months post infection than Hh 3B1 with an intact HhPAI. Hh G also has a large segment of the genomic island deleted, but not as many genes are deleted as compared to Hh NET. Hh G also demonstrated a lower prevalence of hepatic lesions. This variable pathological effect was evident in male mice only. The severity of chronic active inflammation in the liver of the H. hepaticus infected A/JCr mice depended on H. hepaticus liver colonization levels. The in vivo results support the presence of the HhPAI as a legitimate virulence determinant and predictor of severity of liver lesions in H. hepaticus infected A/JCr male mice. To further determine the differences in virulence of the H. hepaticus strains Hh 3B1, Hh NET, Hh G and an isogenic mutant H. ... / by Samuel R. Boutin. / Ph.D.in Molecular and Systems Bacterial Pathogenesis

Thesis (M. Eng.)--Massachusetts Institute of Technology, Biological Engineering Division, 2004. / Includes bibliographical references (leaves 84-93). / Non-parenchymal cells (NPC's) are integral to recreate the native hepatic microenvironment and necessary to maintain in vivo liver function. A variety of in vitro culture systems have been developed to address different aspects of liver physiology and architecture in order to recreate the microenvironment. These in vitro co-culture strategies have been limited by their inability to systematically characterize the addition of non-parenchymal cells. In this dissertation, I use gene expression levels quantified by real-time RT-PCR to determine tissue composition. The identified genes demonstrate significant cell-type specificity, magnitude, and stability of expression in vivo and throughout each step of cell isolation process. In the course of this development, we establish protocols to accurately isolate and count an enriched fraction of primary NPC's. Experiments on the perfusion and isolation process prove that there exists an inverse correlation between perfusion flow rate and NPC yield and viability. Further, we have characterized the tissue composition of each step in the cell isolation process and the resulting NPC population to confirm that a significant number of each NPC type is delivered to in vitro co-culture. System output analysis of spheroids co-cultured at physiological ratios and seeded into the milliF bioreactor shows the presence of stellate cells, but the absence of endothelial (EC) and kupffer cells (KC). The same analysis of 2D collagen gel sandwiches shows the presence of all NPC cell types. This indicates that our process is currently limited by the ability of EC's and KC's to incorporate into spheroid aggregates. Future work that validates the temporal expression stability of the identified genes in different in vitro culture systems / (cont.) and environments will enable determination of relative levels of NPC incorporation and will allow correlations to be made between operational features of in vitro systems, the resulting culture microenvironment, and observed tissue function. / by Brent M. Schreiber. / M.Eng.