Thesis (Ph. D.)--University of Hong Kong, 2005. / Title proper from title frame. Also available in printed format.

No description available.

Abstract The complete exon-intron organization of the gene coding for the mouse α1(XIII) collagen chain, Col13a1, was characterized from genomic clones and multiple transcription initiation points were determined. Detailed comparison of the human and mouse genes showed that the exon-intron structures are completely conserved between the species, and both genes have their 5' untranslated region preceded by a highly conserved putative promoter region. The chromosomal location of the mouse gene was determined to be at chromosome 10, band B4, between markers D10Mit5 – (2.3 ± 1.6 cM) – Col13a1 – (3.4 ± 1.9 cM) – D10Mit15. The location of the genes for both the catalytically important α-subunit of prolyl 4-hydroxylase (P4HA) and human type XIII collagen (COL13A1) were previously mapped to 10q21.3-23.1. Prolyl-4-hydroxylase catalyzes the formation of 4-hydroxyproline in collagens by the hydroxylation of peptide-bound proline and plays a crucial role in the synthesis of these proteins. The order and transcriptional orientation of the COL13A1 and P4HA was determined. These two genes were found to lie at tail to tail orientation on chromosome 10 and the distance between these genes was determined to be about 550 kbp. To study the function of type XIII collagen we used gene targeting in ES cells to generate a mouse line that carries a mutated type XIII collagen gene. Instead of normal protein, mutant mice express type XIII collagen with an altered amino-terminus in which the cytosolic and the transmembrane domains have been replaced with an unrelated sequence. The homozygous mice are fertile and viable but they show alterations in skeletal muscles, mainly wavy sarcolemma and increased variation in muscle fiber diameter. Ultrastructural studies revealed additional abnormalities such as streaming of z-disks, accumulation and enlargement of mitochondria, and disorganized myofilaments. The basement membranes of the muscle cells showed areas of detachment from the plasma membrane and the fibrillar matrix of the cells was less compact than in control animals. Fibroblasts cultured from mutant mice had normal levels of type XIII collagen but exhibited decreased adhesion to substratum which might be explained by a reduced anchoring strength of the altered protein.

Abstract This study was initiated to elucidate the complete genomic structures of type XV collagen in man and mouse and the functional properties of their promoters, as well as to obtain knowledge of the biological role of type XV collagen during development and maturity using immunofluorescence and transgenic techniques. The cloning and characterization of genomic clones revealed that the human COL15A1 gene is 145-kb in size and consists of 42 exons, and the mouse Col15a1 gene is 110-kb with 40 exons. The genomic organization of the two genes was found to be highly conserved, except for two regions of divergence. The nuclease S1 protection analysis revealed multiple transcription initiation sites in both genes, which is in accordance with the overall genomic structures of their 5'-flanking sequences. Transient cell transfection experiments with varying lengths of 5'-deletion constructs identified the fragments necessary for basic promoter activity in both genes and those implicated in the positive and negative regulation of the mouse Col15a1 gene. Furthermore, the involvement of transcription factor Sp1 in the gene regulation of the human COL15A1 gene was demonstrated. A mouse specific polyclonal antibody against type XV collagen was generated and utilized in the localization of type XV collagen protein in developing and mature mouse tissues. Type XV collagen was deposited early in the development and was particularly prominent in capillaries. Spatio-temporal differences in the expression of type XV collagen in various capillary types was demonstrated. Early expression was also detected in the skeletal muscle and peripheral nerves, while expression in the heart, lung, and kidney appeared to be developmentally regulated. Transgenic mice lines expressing truncated type XV collagen driven by either short or long endogenous type XV collagen promoters were generated. The two promoters conferred different tissue-specificities and expression levels, the longer one resulting in more endogenous-like expression. Despite some expression at both mRNA and protein levels, the truncated type XV collagen did not cause any obvious phenotypic or histological changes in any of the lines driven by the shorter promoter fragment. In heterozygote matings of one of the lines driven by the longer promoter fragment, however, a portion of the transgene positive mice appeared to be lost prenatally. Furthermore, pregnancy terminations in this line indicated a high number of abortions beginning at about 11 days of development. Further studies are needed before detailed conclusions on the consequences of the generated mutation can be drawn. The elucidation of the genomic structure of the human COL15A1 gene provides the necessary database for screening mutations in patient samples for candidate diseases caused by this collagen. The genomic clones and the mouse-specific antibody against type XV collagen are valuable tools also in future projects. The knowledge of the developmental dynamics of type XV collagen is of great value, as it helps to understand the physiological consequences that the as yet unidentified mutations in type XV collagen may cause in humans.

Abstract A data base search using the amino acid sequence of Saccharomyces cerevisiae Etr1p, the last enzyme of mitochondrial fatty acid synthesis type II (FAS II), revealed a highly similar human protein, NRBF-1. Expression of NRBF-1 in a yeast etr1Δ strain rescued its respiratory deficiency. NRBF-1 resides in mitochondria in cultured HeLa cells. The recombinant NRBF-1 is enzymatically active, reducing 2E-enoyl-CoAs to acyl-CoAs in an NADPH-dependent manner. Altogether, our data showed that NRBF-1 is a mitochondrial 2-enoyl-CoA reductase/2-enoyl thioester reductase (MECR/ETR1), the human functional counterpart of yeast Etr1p. In addition, MECR was also isolated from bovine heart. It turns out that mammals contain a mitochondrial FAS II pathway, in addition to cytoplasmic FAS I. To investigate the functional mechanism of MECR/ETR1 at the molecular level, the protein was crystallized and the crystal structure determined. The apo-structure of MECR/ETR1 contains two sulfates in the nucleotide binding site and the domain arrangement resembles the NADPH-containing holo-structure of yeast Etr1p. The predicted mode of NADPH-binding and kinetic data suggest that Tyr94 and Trp311 play critical roles in catalysis. A pocket was found in the structure extending away from the catalytic site that can accommodate fatty acyl chains up to 16 carbons. An acyl carrier protein (ACP) binding site was also suggested. To study the physiological function of mouse Mecr, two lines of transgenic mice overexpressing Mecr were generated. The Mecr transgenic mice developed cardiac and mitochondrial abnormalities. The phenotyping was carried out using echocardiography, heart perfusion, histology, and endurance testing. Our results suggest Mecr plays a role in mitochondrial and heart function. Therefore, inappropriate expression of the genes of FAS II may result in the development of cardiomyopathy.

Abstract Lysyl hydroxylase (LH, EC 1.14.11.4) catalyzes the post-translational hydroxylation of lysyl residues in collagens and other proteins with collagenous domains. The hydroxylysyl residues participate in the formation of collagen cross-links, and some of the hydroxylysyl residues are further glycosylated. Three lysyl hydroxylase isoforms LH1, LH2 and LH3, encoded by three individual genes have been characterized and one isoform, LH3 is a multifunctional enzyme containing lysyl hydroxylase, collagen galactosyltransferase (GT, E.C. 2.4.1.50) and glucosyltransferase (GGT, E.C. 2.4.1.66) activities in vitro. In this thesis the genes for the mouse lysyl hydroxylases were each mapped to a different chromosome. In addition, the roles of the lysyl hydroxylase isoforms were characterized in mice by studying their expression during development and the distribution of LH2 and LH3 in adult mice. The results revealed a widespread expression of the mouse lysyl hydroxylases during embryonic development whereas LH2 and LH3 showed tissue- or cell-specific expression patterns in the adult. Alternative splicing of the gene for LH2 also showed developmental and tissue-specific regulation. The different functions of LH3 were studied in vivo by generating three different LH3 manipulated mouse lines. Analysis of the mouse lines revealed that LH3 has lysyl hydroxylase and glucosyltransferase activities in vivo, and that, in particular, the glucosyltransferase activity of LH3 is essential for normal development. The loss of glucosyltransferase activity caused disruption of basement membranes leading to embryonic lethality while the absence of lysyl hydroxylase activity led to ultrastructural alterations in muscle and basement membranes and disorganization of collagen fibrils. The disruption of basement membrane was due to an intracellular accumulation of unglycosylated type IV collagen, whereas the ultrastructural alterations were related to the abnormal aggregation and distribution of underglycosylated type VI collagen. The results demonstrate that hydroxylysine-linked glycosylations are critical for the secretion of type IV collagen and its assembly into basement membranes, and for the assembly and distribution of type VI collagen.

No description available.

Alzheimer's disease (AD) is the most prevalent form of age-associated dementia. Cholesterol dysregulation is linked with AD onset. Besides age, the most important risk factor associated with AD is the inheritance of the epsilon-4 allele of apolipoprotein E, a cholesterol transporter. In addition, while hypercholesterolemia has been shown to be an independent risk factor for AD, the nature of the cholesterol-AD link is still not clear. This gap in our understanding is partly due to a lack of knowledge about cholesterol metabolism in the central nervous system (CNS). The low-density lipoprotein receptor (LDLR) is the main receptor of apoE and a central regulator of serum cholesterol levels. Therefore, we sought to characterize the potential participation of LDLR in AD pathogenesis and/or progression. Previous reports with similar aims came to contradictory conclusions. Such studies assessed potential changes in AD in the absence of LDLR by utilizing the LDLR-/- mouse model and crossing it to AD mouse models. Initially we evaluated LDLR-/- mice as a suitable model to study AD. We found that LDLR-/- mice overexpressed a functional splice-variant of LDLR, LDLRDelta4. Moreover, its protein localized in similar regions as the LDLR did in control mice. Finally, we determined that LDLRDelta4 bound apoE, which underscores the impact of the isoform's function in the CNS. We then focused on characterizing changes to LDLR in AD models. We found that APP overexpression in cells increased LDLR mRNA and protein. APP overexpression and Abeta treatment shifted LDLR localization. An AD mouse model showed increased LDLR in hippocampus. Conversely, LDLR levels were decreased in APP-/- mice. Finally, we found that microtubules were affected in cells overexpressing APP. In conclusion, the data presented argue for the importance of LDLR-mediated regulation of cholesterol during AD progression. Also, LDLR may participate in the initial pathogenic insults leading to amyloid deposition, which make it a potential therapeutic target to treat AD. Finally, we propose that APP/Abeta overexpression disrupts microtubule formation; this alteration affects protein trafficking. One of the proteins affected is LDLR, the repercussions of which may ultimately result in cholesterol dysregulation.

Cardiac remodeling involves changes in heart shape, size, structure, and function after injury to the myocardium. The proinflammatory adaptor protein myeloid differentiation protein 88 (MyD88) contributes to cardiac remodeling. To investigate whether excessive MyD88 levels initiate spontaneous cardiac remodeling at the whole-organism level, we generated a transgenic MyD88 mouse model with a cardiac-specific promoter. MyD88 mice (male, 20–30 g, n=~80) were born at the expected Mendelian ratio and demonstrated similar morphology of the heart and cardiomyocytes with that of wild-type controls. Although heart weight was unaffected, cardiac contractility of MyD88 hearts was mildly reduced, as shown by echocardiographic examination, compared with wild-type controls. Moreover, the cardiac dysfunction phenotype was associated with elevation of ANF and BNP expression. Collectively, our data provide novel evidence of the critical role of balanced MyD88 signaling in maintaining physiological function in the adult heart.

While the cardiotoxicity of doxorubicin (DOX) is known to be partly mediated through the generation of reactive oxygen species (ROS), the biochemical mechanisms by which ROS damage cardiomyocytes remain to be determined. This study investigates whether S-glutathionylation of mitochondrial proteins plays a role in DOX-induced myocardial injury using a line of transgenic mice expressing the human mitochondrial glutaredoxin 2 (Glrx2), a thiotransferase catalyzing the reduction as well as formation of protein-glutathione mixed disulfides, in cardiomyocytes. The total glutaredoxin (Glrx) activity was increased by 76% and 53 fold in homogenates of whole heart and isolated heart mitochondria of Glrx2 transgenic mice, respectively, compared to those of nontransgenic mice. The expression of other antioxidant enzymes, with the exception of glutaredoxin 1, was unaltered. Overexpression of Glrx2 completely prevents DOX-induced decreases in NAD- and FAD-linked state 3 respiration and respiratory control ratio (RCR) in heart mitochondria at days 1 and 5 of treatment. The extent of DOX-induced decline in left ventricular function and release of creatine kinase into circulation at day 5 of treatment was also greatly attenuated in Glrx2 transgenic mice. Further studies revealed that heart mitochondria overexpressing Glrx2 released less cytochrome c than did controls in response to treatment with tBid or a peptide encompassing the BH3 domain of Bid. Development of tolerance to DOX toxicity in transgenic mice is also associated with an increase in protein S-glutathionylation in heart mitochondria. Taken together, these results imply that S-glutathionylation of heart mitochondrial proteins plays a role in preventing DOX-induced cardiac injury.