Impaired integrin α5 /β1 -mediated hepatocyte growth factor release by stellate cells of the aged liver.

Abstract Hepatic blood flow and sinusoidal endothelial fenestration decrease during aging. Consequently, fluid mechanical forces are reduced in the space of Disse where hepatic stellate cells (HSC) have their niche. We provide evidence that integrin α5/β1 is an important mechanosensor in HSC involved in shear stress‐induced release of hepatocyte growth factor (HGF), an essential inductor of liver regeneration which is impaired during aging. The expression of the integrin subunits α5 and β1 decreases in liver and HSC from aged rats. CRISPR/Cas9‐mediated integrin α5 and β1 knockouts in isolated HSC lead to lowered HGF release and impaired cellular adhesion. Fluid mechanical forces increase integrin α5 and laminin gene expression whereas integrin β1 remains unaffected. In the aged liver, laminin β2 and γ1 protein chains as components of laminin‐521 are lowered. The integrin α5 knockout in HSC reduces laminin expression via mechanosensory mechanisms. Culture of HSC on nanostructured surfaces functionalized with laminin‐521 enhances Hgf expression in HSC, demonstrating that these ECM proteins are critically involved in HSC function. During aging, HSC acquire a senescence‐associated secretory phenotype and lower their growth factor expression essential for tissue repair. Our findings suggest that impaired mechanosensing via integrin α5/β1 in HSC contributes to age‐related reduction of ECM and HGF release that could affect liver regeneration.


Supplemental data
Supplemental Table S1: Differential expression of markers associated with ECM in whole liver tissue from aging rats. The gene expression in liver tissue from young (2 months) and old (22 months) rat livers was investigated by Affymetrix micorarrays (n = 3 for each group). Negative fold change implicates lower expression in hepatic tissue from old rats (p < 0.05). All ECM-associated genes, that were significantly altered in the liver tissue from 22months-old rats, showed a reduction. Among integrins, only integrin α5 (Itga5) was markedly regulated.

Process
Gene Symbol Fold Change ANOVA p-value ECM-associated (see separate Excel file "Supplemental Table S2")   Two separate tables are given: one table containing all identified proteins and one table  containing all quantified proteins. The tables include data from protein identification as was as quantitative data from label-free quantification.
Supplemental Table S3: Differential expression of markers associated with ECM and integrins in HSC from aging rats. The gene expression of HSC isolated from young (2 months) and old (22 months) rat livers was investigated by Affymetrix micorarrays (n = 3 for each group). Negative fold change implicates lower expression while positive values indicate higher expression in HSC from old rats (p < 0.05). The expression of Mmp13 increased markedly on mRNA level in array analysis, whereas matrix proteins such as Lama2, Lamc1, Nid1, Nid2, collagens, and Fn1 decreased in HSC from old rats. Furthermore, several integrins were significantly downregulated such as Itga5. (see separate Excel file "Supplemental Table S5") Supplemental Table S5: Differential expression of markers associated with quiescence or activation in HSC from aging rats. The gene expression of HSC isolated from young (2 months) and old (22 months) rat livers was investigated by Affymetrix microarrays (n = 3 for each group). Negative fold change implicates lower expression while positive values indicate higher expression in HSC from old rats. The quiescent-associated markers Sparcl1, Pparγ, and reelin as well as activation-associated markers fibronectin, periostin, nestin, and desmin were significantly downregulated in HSC from old rats (p < 0.05), while Gfap and α-Sma remained unchanged on mRNA level in array analysis.  Table S6: Differential expression of markers associated with SASP in HSC from aging rats. The gene expression of HSC isolated from young (2 months) and old (22 months) rat livers was investigated by Affymetrix microarrays (n = 3 for each group). Negative fold change implicates lower expression while positive values indicate higher expression in HSC from old rats (p < 0.05). Many genes associated with SASP exhibited altered expression in HSC from old rats, when compared to HSC from young rats. The expression of many growth factors declined, but inflammation-and cell migration-associated genes exhibited elevated expression in array analysis.
Supplemental Figure S1: Experimental setup. Isolated HSC and liver tissue from young (2 months) and old (22 months) rats were compared to unravel possible alterations of stellate cells and their niche during aging. (A) The liver tissues of young and old rats were also analyzed by microarrays (see above; n = 3 for each age group). The array results were evaluated by qPCR analysis and completed by immunofluorescence as well as Western blot to identify age-related alterations in the liver (n = 3-10). (B) To enable quantitative assessment of the matrisome, liver tissues from both age groups were decellularized for proteome analysis (n = 3 for both age groups). (C) The blood serum was collected from young and old rats and analyzed regarding cytokines indicating a SASP. (D) HSC of both age groups were isolated and enriched by density gradient centrifugation and further purified by FACS using their characteristic vitamin A (retinoid) fluorescence. After one day of culture, the HSC showed typical cell morphology and vitamin A fluorescence (blue). The HSC from young rats contained fewer retinoids compared to those obtained from old rats. The RNA of HSC was harvested and analyzed by microarrays (Affymetrix GeneChip Rat Gene 2.0 ST Array; n = 3 for each age group). The medium of the HSC was collected for protein arrays and ELISA to investigate their secretome with respect to a SASP.
Supplemental Figure S2: Quality control of microarray raw data. (A, B) To evaluate the reliability of the data sets obtained by microarray analysis for liver tissue and isolated HSC (cultured for 1 day) from young (2 months) and old (22 months) rats, the signal intensities of samples on genes chips were compared. (C, D) Pearson correlation analysis of the microarray data for liver tissue and isolated HSC from both age groups was performed and revealed that the gene expression of liver samples seemed to be equal. In contrast, significant differences between HSC from young and old rats were observed when Pearson correlation was applied. (E) Gene expression data of whole liver tissue from both groups were compared by principal component analysis (PCA) and showed that the samples of young rats were highly similar and different from old rats. (F) PCA of gene expression arrays in isolated HSC from 2 (blue) and 22 (red) months old rats. HSC from young rats exhibited highly similar gene expression and were different from samples obtained from old rats, which clustered together, but exhibited a higher variation. Each dot represents a data set from a single animal (n = 3 for each age group).
Supplemental Figure S3: Gene ontology (GO) term analysis of differentially expressed genes in HSC from old compared to young rats with respect to biological processes. GO term analysis of differentially expressed genes obtained by Affymetrix microarrays was performed with the software GOrilla (October 2018; fold enrichment > 1.5; p < 0.05).
Supplemental Figure S4: Senescence-and inflammation-associated factors in blood serum and HSC. (A) The rat cytokine array revealed no obvious differences of cytokines in the serum of young and old rats, with the exception of CXCL7, which was significantly lower in the serum samples from 22-months-old animals. Mean pixel density of samples from 2months-old rats was set to 100% and data are presented as means ± SEM (n = 4, *p < 0.05). Cytokines released into culture medium by freshly isolated HSC from young and old rats was also analyzed. Only TIMP1 and CXCL1 were detected by the rat cytokine array in all culture supernatants. (B) TIMP1 concentration remained unchanged, (C) whereas CXCL1 increased by 3-fold in culture supernatants of HSC from old rats (n = 4 for each age group). Data from young rats were set to 100%. (D) Analysis of CXCL3 by ELISA revealed a 10-fold increase in CXCL3 concentration in HSC culture media from old rats. (E) IL6 concentration in culture supernatants of HSC from old rats increased by 4-fold compared to young rats as analyzed by ELISA. Cytokine array and ELISA data were normalized to cell number. The data are presented as means ± SEM (cytokine arrays: n = 3; ELISA: n = 5 for 2-months and n = 3 for 22-months-old rats; *p < 0.05).