Search

ILD represents a heterogeneous spectrum of diseases caused by excessive deposition of extracellular matrix (ECM) into the interstitial space leading to distortion of the lung architecture and impairment of gas exchange. In this thesis the design of an in vitro model is described to test the pathways involved in the fibrotic response. The model was first tested on fibroblast and epithelial cell lines, two cell types that are implicated as key players in pulmonary fibrosis, treating cells with the pro-inflammatory cytokine Transforming Growth Factor-β (TGF-β), ensuring that this model was reliable and reproducible. Once a model had been established, it was utilised by treating cells with plasma from patients with IPF and Systemic Sclerosis Associated ILD (SSc-ILD). This was to assess whether treatment with plasma from patients with fibrotic lung disease would induce any changes in these cell lines. When cells were treated with plasma there were no changes in gene expression of the MRC5 fibroblasts but there was an increase in proliferation of the fibroblast cells when treated with IPF patient plasma. When the A549 epithelial cells were treated with plasma from patients with IPF, there was a significant reduction in COL3A1 and FN-EDA. This was associated with a significant downregulation of N-cadherin. These changes are not those expected in EMT. Conversely treatment of the same cell line with SSc-ILD plasma led to a significant increase in gene expression of COL3A1 and FN-EDA. Finally, this model was adapted to imitate the constant movement of the lungs by stretching the cells in vitro. Stretch alone caused no significant differences in any of the genes associated with fibrosis in either cell type. However, when used alongside TGF-β treatment there was an attenuation of the fibrotic response in COL1A1, FN-EDA and CCN2 in the fibroblasts and COL1A1 and Slug in the epithelial cells. In conclusion, this thesis has produced a novel way to explore some of the key pathways involved in fibrosis and the changes that plasma treatment induces on cell lines.

Osteoarthritis is a complex multifactorial degenerative disease that is characterised by progressive loss of articular cartilage and inflammation of the synovium resulting pain and loss of mobility. HuR is a ubiquitously expressed RNA binding protein involved in the post-transcriptional regulation of target mRNAs and can reduce their rate of turnover. Previously published data suggests that HuR is spatially regulated within developing cartilage and is involved in the regulation of chondrocyte genes and osteoarthritis. Furthermore, knockout of HuR in mice at pre-implantation stage leads a skeletal dysplasia, among other phenotypes. This project aimed to determine whether HuR activity is important for maintaining cartilage and skeletal phenotype appropriately. An inducible, cartilage-specific HuR knockout mouse model and a non-inducible limb bud and mesenchyme-specific HuR knockout mouse model were generated utilising the Cre/LoxP system, were Cre expression is driven by an aggrecan promotor and a Prx1 enhancer, respectively. Using these novel transgenic mouse models of HuR knockout, the role of HuR in embryonic development, during osteoarthritis and in chondrocytes was characterised. Induced aggrecan-driven knockout of HuR leads to severe dysregulation of skeletal development during embryogenesis, resulting in a lack of mineralisation of axial components including the ribs and spine, and craniofacial structures. Similarly, Prx1-driven knockout of HuR during embryonic development leads to altered mineralisation of the skeleton, most strikingly in the developing limbs. This is similar in many respects to the phenotype observed following a broader knockout of HuR at an earlier developmental stage and suggests a role for HuR during endochondral ossification. In vitro assays using murine primary chondrocytes suggest that HuR may be required for the appropriate regulation of a number of known chondrocyte marker genes including Sox9, Runx2 and Mmp13, of which the latter two are known to be critically involved in the regulation of chondrocyte hypertrophy and endochondral ossification. The roles of RNA binding protein-mediated post transcriptional gene control is still not well understood in orthopaedic tissues. This study indicates that spatial and temporal mis-control of the protein HuR can lead to impaired skeletal mineralisation, suggesting that it performs an important function in chondrogenesis and/or osteogenesis.

Laminin N-terminus α31 (LaNt α31), a member of the relatively new LaNt protein family, is a product of alternative splicing from the laminin (LM) α3 encoding gene (LAMA3). Prior to these studies, very little was known about LaNt α31 biology; it had been shown to co-localise with LM α3β3γ2 (LM332) in the basement membrane (BM) of the skin and to play a role in adhesion and migration of epidermal cells. Still, no direct insight into the mechanism behind this effect has been described. In this study, we performed the first analysis of LaNt α31 in the eye. Immunohistochemistry reveals that the protein is differentially distributed across the regions of the epithelium in the anterior surface of the eye. Specifically, LaNt α31 localises intracellularly through all layers of the corneal epithelium, but is restricted to the basal layers of the limbus and conjunctiva. Hence, we sought to investigate about LaNt α31 functional roles in corneal epithelial cells (hTCEpi) and its interplay with LM during corneal epithelial matrix assembly. Our functional studies demonstrate that knockdown of intracellular LaNt α31 has no discernible effect on hTCEpi, in which LaNt α31 is not normally matrix-associated. However, when overexpressed by adenoviral delivery, LaNt α31 is deposited into the extracellular matrix (ECM) and causes a number of phenotypic effects including; impaired cell adhesion to LM111, changes to LM organisation, increased cell spreading, early recruitment of Collagen (Col) type XVII to β4 integrin in hemidesmosome (HD) like-complexes and mislocalisation of focal adhesions (FAs). Increased LaNt α31 expression also leads to significantly decreased motility at times of new matrix synthesis, with this migration defect rescued when hTCEpi are provided with a cell-derived preformed ECM. Live imaging experiments indicate that GFP-tagged LaNt α31 clusters at the periphery of cells, where it co-distributes and is deposited along with LM β3. As cells move across a preformed matrix, clusters of LaNt α31 rapidly form and dissociate intracellularly, alongside LM deposits. To study the effects of overexpression in-vivo, we generated a human keratin 14 (hK14) promoter-driven LaNt α31 overexpressing mouse model by embryo microinjection. Using this construct, we obtained far fewer offspring than would normally be expected and were not able to detect expression of the protein product from the transgene in the animals carrying the construct. Based on these data, we hypothesise that, in this model system, LaNt α31 overexpression has induced lethality. Together, these data demonstrate that LaNt α31 influences the assembly of LM matrix in hTCEpi affecting fundamental cell functions, and implicate the LaNt proteins as regulators of other LM dependent processes including wound repair, differentiation and tissue homeostasis.