Medizinische Fakultät - Digitale Hochschulschriften der LMU - Teil 12/19

Introduction: Human mesenchymal stem cells (hMSC) are easily obtainable from bone mar-row and possess the ability to differentiate into osteoblasts. Therefore, they have been sug-gested as a suitable source for bone regeneration. HMSC are equipped with a variety of in-tegrins that mediate essential cell-matrix interactions. Collagen I represent approximately 90% of the bone protein content. Cell attachment to collagen I is mediated by three members of the integrin receptor family named a1b1, a2b1 and a11b1 integrins. The main aim of this doctoral thesis was to investigate the basal expression of those integrins in hMSC and to func-tionally analyze the knockdown effect of a single collagen I-binding integrin on hMSC behav-ior in vitro. Materials and methods: HMSC were cultured on collagen I-coated surface. A lentiviral trans-fer of a1-, a2- and a11-specific shRNA was applied for downregulation of the corresponding integrin mRNA. Quantitative PCR and western blot analysis were used to assess the basal ex-pression, knockdown efficiency and integrin compensation. Colorimetric adhesion assay was used for estimation of the extent of cells attachment. HMSC spreading and migration was ob-served by time lapse experiments. JC-1 staining was used for investigation of the initiation of apoptosis. Results: Quantitative PCR were used to assess the basal expression of collagen I-binding integrins in three hMSC donors. We found that these integrins are differently expressed as integrin a11 had the highest and integrin a2 the lowest expression. Next, we applied lentiviral delivery of target-specific short hairpin RNA (shRNA) in order to knockdown each of the collagen I-binding integrins and compared them to the hMSC transduced with a sequence against a non-human gene abbreviated as shRNA control. We achieved significant downregulation (> 80%) of the collagen I-binding integrin mRNA and protein. Subsequently to the transduction, we did not noticed pronounce morphological cell changes, however, a clear decrease of a2- and a11-knockdown hMSC numbers was observed during cultivation. Using a quantitative adhesion assay, we estimated that 120 min after plating only 30% of integrin a11-deficent cells were able to attach to collagen I. In contrast, at the same time point, 70% of integrin a2-knockdown hMSC were attached while integrin a1- and shRNA control hMSC have already reached 100% cell adhesion. Furthermore, a time lapse-based investigation showed that integrin a1- and shRNA control hMSC need approximately 35 min to fully spread on collagen I. In contrast, integrin a2- and a11-knockdown hMSC took approximately double more time for spreading in comparison to shRNA control hMSC. Additionally, we analyzed the migration capability of the four different hMSC lines. The average path which integrin a1- and shRNA control hMSC passed was approximately 170 µm with mean speed of 11.5 µm/h. In parallel integrin a2 and a11-deficient hMSC migrated to a distance of approximately 70 µm with a velocity of 5 µm/h. Since it was observed a lost of a2- and a11-deficient hMSC, next we performed JC-1 staining that visualizes mitochondrial leakage, a hallmark of apoptosis. The majority of integrin a2- and a11-knockdown hMSC exhibited mitochondrial leakage whereas integrin a1- and shRNA control hMSC showed intact mitochondria. Finally, we used quantitative PCR to investigate whether there were compensatory effects between the three integrin receptors. We detect that knockdown of integrin a1 led to upregulation of a2 and a11. Similarly, when integrin a2 was downregulated, integrin a1 and a11 expression increased. Interestingly, knockdown of integrin a11 caused only a slight increase in integrin a1 but not in a2 expression. We also observed that upon osteogenic stimulation, integrin a2 and a11-deficient hMSC further reduced in number and did not mineralize the matrix even on a single cell level. Moreover, our preliminary investigation in hMSC-derived from osteoporosis suf