Contribution of the α8 Integrin Chain to the Expression of Extracellular Matrix Components

Abstract

In the kidney, the α8 integrin chain (itga8) is expressed in mesenchymal cells and is upregulated in fibrotic disease. We hypothesized that itga8 mediates a profibrotic phenotype of renal cells by promoting extracellular matrix and cytokine expression. Genetic itga8 deficiency caused complex changes in matrix expression patterns in mesangial and smooth-muscle cells, with the only concordant effect in both cell types being a reduction of collagen III expression. Silencing of itga8 with siRNA led to a decline of matrix turnover with repression of matrix metalloproteinases and reduction of matrix production. In contrast, de novo expression of itga8 in tubular epithelial cells resulted in reduced collagen synthesis. Overexpression of itga8 in fibroblasts did not change the expression of matrix molecules or regulators of matrix turnover. Thus, the influence of itga8 on the expression of matrix components was not uniform and celltype dependent. Itga8 seems unlikely to exert overall profibrotic effects in renal cells.

Keywords::

• integrins
• fibrosis
• extracellular matrix
• mesenchymal cells
• α8 integrin deficiency
• renal cells

INTRODUCTION

The α8 integrin chain (itga8) is expressed on mesenchymal cells, like vascular smooth-muscle cells, mesangial cells of the kidney, some fibroblasts and undifferentiated epithelial cells (Benoit et al., 2009; Hartner et al., 1999; Levine et al., 2000; Schnapp et al., 1995a). After dimerization with the β1 integrin chain, it serves as a receptor for fibronectin, vitronectin, osteopontin, tenascin C, nephronectin, and presumably for fibrillin-1 (Bouzeghrane et al., 2005; Brandenberger et al., 2001; Denda et al., 1998; Schnapp et al., 1995b). α8 integrins contribute to the regulation of various cell functions like adhesion, migration, proliferation, and apoptosis (Bieritz et al., 2003; Farias et al., 2005; Hartner et al., 2008).

Integrins are also known to play a role in the regulation of the expression of extracellular matrix molecules and fibrosis. For example, an overexpression of αvβ5 integrin in cultivated fibroblasts enhanced the expression of the α2(I) collagen gene (Asano et al., 2004), while silencing of the β3 integrin chain in fibroblasts led to reduced expression levels of collagen I (Suarez et al., 2013). Several in vivo studies revealed profibrotic effects of integrins: Underexpression of the αvβ6 integrin in an animal model of renal fibrosis attenuated the expression and deposition of collagen I and collagen III (Ma et al., 2003). In a rat model of glomerular scarring α1β1 integrin is overexpressed and treatment with an inhibitory antibody to α1β1 integrin reduced fibrosis (Cook et al., 2002). Similarly, an inhibition of αvβ3 integrin using specific RGD peptides resulted in a reduction of glomerulosclerosis during experimental glomerular disease (Amann et al., 2012). Moreover, induction of glomerulosclerosis by adriamycin in α2 integrin-deficient mice also resulted in a reduction of glomerular damage in comparison to wildtype mice (Borza & Pozzi, 2012). Thus, several integrin chains might serve as a promising therapeutic target to delay the progression of fibrotic diseases.

The itga8 is expressed in fibroblasts in lung and liver fibrosis and in the scarring myocardium (Bouzeghrane et al., 2004; Levine et al., 2000). Moreover, the itga8 is upregulated in glomeruli undergoing fibrotic remodeling (Hartner et al., 1999). Profibrotic agents like TGFβ-1 and angiotensin II are able to induce the expression of the itga8 in different cell types, supporting the notion that α8β1 integrin might exert profibrotic effects (Bouzeghrane et al., 2004; Hartner et al., 1999). Mesangial cells deficient for the itga8 downregulate the expression of α-smooth muscle actin (a marker of myofibroblast activation) and the profibrotic cytokine CTGF (Marek et al., 2010), further arguing for a contribution of α8β1 integrin to the induction of extracellular matrix expression in fibrotic areas. On the other hand, underexpression of the itga8 did not protect from fibrotic changes in the heart or the kidney (Hartner et al., 2002, 2009). To clarify if α8β1 integrin has a profibrotic potential, we studied the effects of under- or overexpression of the itga8 in mesenchymal and epithelial cells on the expression of extracellular matrix molecules and profibrotic cytokines.

MATERIALS AND METHODS

Cultivation of mesangial cells and vascular smooth muscle cells

Mesangial cells were isolated from kidneys of wildtype or α8 integrin-deficient mice (obtained from U. Müller, Basel) by the sieving method (Hartner et al., 1999) using 63, 75, and 38 μm grid sieves. Cultured wildtype and α8 integrin-deficient mesangial cells were characterized as described by Bieritz et al. (2003). Rat mesangial cells from Sprague-Dawley rats were isolated by the sieving method using 106, 180, 75 μm grid sieves. Mesangial cells were grown in culture flasks coated with 10 μg/ml fibronectin, in Dulbecco's modified Eagle's Medium (DMEM high glucose with L-Glutamin; PAA Laboratories GmbH, Pasching, Austria) containing 10% FCS, 1% penicillin–streptomycin, 0.1% insulin (Sigma, Deisenhofen, Germany) in a 95% air and in 5% CO₂-humidified atmosphere at 37°C. Mesangial cells were used for experiments in passages 5–18.

Vascular smooth muscle cells were isolated from mouse aorta similar as described by Strehlow et al. (2003) for rat vascular smooth muscle cells. Briefly, the aortas were excised, washed in phosphate-buffered saline with 1% penicillin–streptomycin, and fat was removed with a fine forceps. The aortas were then incubated in DMEM containing 1mg/ml collagenase type I (Sigma), 0.3mg/ml elastase (Serva, Heidelberg, Germany), and 0.3mg/ml trypsin inhibitor type II (Sigma) for 15–20 min at 37°C. The aorta was washed and the adventitia was stripped with fine forceps. The vessels were incised longitudinally and the endothelial cells were gently scraped off. The aortas were then minced with scissors and transferred to reaction tubes containing the same enzymatic solution as described above, incubated in 37°C for 60–90 min until 90% of the cells were dispersed under the microscope. The cells were centrifuged at 5000rpm for 2 min, then resuspended in 3 ml DMEM with 20% fetal calf serum (FCS), 2% penicillin–streptomycin, and cultured in plates or flasks coated with 10 μg/ml fibronectin, in a 95% air and in 5% CO₂-humidified atmosphere at 37°C for experiments. Cultured cells were verified to be vascular smooth-muscle cells using immunostaining with anti-smooth-muscle actin antibody (Sigma). Vascular smooth-muscle cells were used for experiments in passages 0–1.

Cultivation of NIH3T3 fibroblasts

NIH3T3 cells (ATCC) were grown in culture flasks coated with 10 μg/ml fibronectin, in DMEM high glucose with L-Glutamin; PAA Laboratories GmbH containing 10% FCS, 1% penicillin–streptomycin in a 95% air and in 5% CO₂-humidified atmosphere at 37°C.

Cultivation of primary tubular epithelial cells

Human primary tubular epithelial cells (a gift from Sven Kroening, Department of Nephrology and Hypertension, University Hospital of Erlangen) were isolated as described before (Kroening & Goppelt-Struebe, 2010). Cell isolation was performed from tumor-free areas of tumor nephrectomies. The use of human material was approved by the local ethics committee (Ethik Kommission der Medizinischen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg, reference number 3755). Cells were grown in culture flasks coated with 10 μg/ml fibronectin, in DMEM/Ham's with 2,5% FCS, 1% glutamin, 1% penicillin-streptomycin, 1% ITS supplement, 0,04% T3, 0,005% EGF, 0,0072% hydrocortison in a 95% air – 5% CO₂-humidified atmosphere at 37°C and were used for experiments freshly isolated.

Overexpression of the itga8 in NIH3T3 cells, primary tubular epithelial cells, and mesangial cells

Transfection was performed with 1 × 10⁵ cells and 3 μg plasmid DNA per well in a 6-well plate. Briefly, the α8 integrin vector described in Bieritz et al. (2003) was recloned to generate an α8 integrin expression vector under control of the CMV promoter (CMVitga8), with an additional mRNA stabilizing intron from the rabbit β-globin gene inserted ahead of the start codon. As control transfection, pcDNA 3.1 (Invitrogen, Darmstadt, Germany) was used. 3μg of plasmid DNA was mixed with 9 μl of HiPerFect (Qiagen, Hilden, Germany) and 96 μl of DMEM to restore the transfection mix. After 10 min incubation at room temperature, the transfection mix was added drop-by-drop on freshly washed cells in fresh culture medium. For NIH3T3 cells and primary tubular epithelial cells, mRNA was isolated 24 and 48 h after transfection without additional stimulation, respectively, 24 and 48 h after stimulation with 10 μg/ml of TGFβ-1 after preceding transfection. For α8 integrin- deficient and wildtype mouse, mesangial cells mRNA was isolated 24, 48, 72, and 96 h after first transfection, with cells being retransfected in the same way 48 h after first transfection for the isolation time points 72 and 96 h.

Silencing of the itga8 in rat mesangial cells

Gene silencing of the itga8 was achieved by transfection of mesangial cells with α8 integrin siRNA (si itga8) according to the fast protocol from the HiPerFect Transfection Reagent Handbook (Qiagen). Mesangial cells were transfected in the cell suspension with a final concentration of 5nM si itga8 (sense: GAC CUC CUC AGG AUG AAA UdT dT, antisense: AUU UCA UCC UGA GGA GGU CdT dT, Qiagen) before seeding to allow silencing for 72 h. A nonsilencing siRNA (sense: UUC UCC GAA CGU GUC ACG UdT dT, antisense: ACG UGA CAC GUU CGG AGA AdTdT, Qiagen) was used as control.

Isolation of mRNA and real-time PCR

To evaluate mRNA expression levels, total RNA was obtained from harvested cells by extraction with RNeasy^® Mini columns (Qiagen). First-strand cDNA was synthesized with TaqMan reverse transcription reagents (Applied Biosystems, Weiterstadt, Germany) using random hexamers as primers. Final RNA concentration in the reaction mixture was adjusted to 100 ng/μl. Reactions without Multiscribe reverse transcriptase were used as negative controls for genomic DNA contamination. PCR was performed with an ABI PRISM 7000 Sequence Detector System and SYBR Green (Applied Biosystems) according to the manufacturer's instructions. The relative amount of the specific mRNA was normalized with respect to 18S rRNA. Primers used for amplification are listed in Table S1 (see Supplementary Table 1 available online at http://informahealthcare.com/doi/abs/10.3109/15419061.2013.876012). All samples were run in triplicates.

Western blot analysis

Protein was isolated from harvested cells with cell lysis buffer (RIPA-buffer) by incubating on ice for 30 min with intermittent vortexing, followed by centrifugation for 10 min at 13 000 rpm to get protein containing supernatant. Protein concentration of cell lysates was determined using a protein assay kit (Pierce, Rockford, IL, USA). Protein samples containing 70 μg of total protein were denatured by boiling for 5 min and separated on a 6% denaturing SDS-PAGE gel. After electrophoresis, the gels were electroblotted onto PVDF membranes (Pall Filtron, Karlstein, Germany), blocked with Roti-Block (Roth, Karlsruhe, Germany) for 1 h and incubated with the primary antibody overnight. Immunoreactivity was visualized with a secondary horseradish peroxidase-conjugated anti-goat IgG antibody 1:40000 in TBS/T, or a horseradish peroxidase-conjugated anti-rabbit IgG antibody (both Santa Cruz Biotechnology, Heidelberg, Germany) 1:10000 in TBS/T using the ECL system according to the manufacturer's instructions (Amersham, Braunschweig, Germany).

Antibodies for Western blot analysis

A goat polyclonal antibody to the itga8 (R&D Systems, Wiesbaden, Germany) was used in dilution of 1:1000 in 1xRoti-Block. As a loading control, rabbit polyclonal antibody to β-tubulin (Abcam, Cambridge, UK) was used in dilution of 1:5000 in 1% FCS/PBS.

Adhesion assay

48 h after transfection, 10000 mesangial cells per chamber slide were seeded on the α8 integrin ligand fibronectin (10 μg/ml). Cells were allowed to attach for 1 h. After washing, fixation with methanol and staining with hematoxylin, adhered cells were counted in six medium power views (× 200) per chamber slide.

Statistical analyses

A t-test was used to test significance of differences between groups. A P value less than 0.05 was considered significant. Values are displayed as means ± standard deviation (SD).

RESULTS

Effects of the underexpression of the itga8 in mesangial and vascular smooth muscle cells

First, we studied the influence of a genetic knock out of α8 integrin on the expression of extracellular matrix molecules and profibrotic cytokines in mesangial und vascular smooth muscle cells. In mesangial cells, a deficiency in the itga8 led to inconsistent changes in the expression patterns of extracellular matrix molecules and regulators of matrix turnover. On the one hand, we detected a decline of matrix metalloproteinase expression and the expression of some extracellular matrix components like collagens I and III and CTGF (Table 1). On the other hand, an induction of the profibrotic growth factor TGFβ-1 and other extracellular matrix components like collagen IV, osteopontin, and fibrillin-1 was detected in α8 integrin-deficient mouse mesangial cells compared to wildtype mesangial cells (Table 1).

Table 1. Expression analysis of matrix components and regulators of matrix turnover in α8 integrin chain underexpressing cells.

Expression of the α8 integrin chain (itga8), collagen I (Col I), collagen III (Col III), collagen IV (Col IV), fibrillin-1 (FBN), fibronectin (FN), osteopontin (OPN), matrix metalloproteases, MMP-2 and MMP-9, tissue inhibitors of metalloproteases, TIMP-1 and TIMP-2, connective tissue growth factor (CTGF), and transforming growth factor-β 1 TGFβ-1 in α8 integrin-deficient mouse mesangial cells (α8-/-MC), in rat mesangial cells 72 h after performing transient knock down of the α8 integrin chain by siRNA (si itga8) and α8 integrin-deficient vascular smooth-muscle cells (α8-/-VSMC). Data are means± standard deviation in relation to wildtype (wt) or control siRNA-treated mesangial cells (co). Results are representative for at least three similar experiments. Red: significant downregulation, p < 0.05. Green: significant upregulation p < 0.05. n.d. = not detected.

Aortic vascular smooth muscle cells were isolated from wildtype and α8 integrin-deficient mice. Similar to the effects of itga8 deficiency in mesangial cells, itga8 deficiency in vascular smooth-muscle cells led to an ambiguous change of matrix components, with a significant induction of TIMP-1 and CTGF (Table 1). Otherwise, collagen III, fibrillin-1, and osteopontin were significantly downregulated in α8 integrin-deficient vascular smooth-muscle cells (Table 1). Except the common reduction of collagen III expression, the pattern of expressional changes in matrix molecules was different in α8 integrin-deficient mesangial cells and vascular smooth-muscle cells.

Silencing of the itga8 in rat mesangial cells was performed to detect changes in the expression patterns of extracellular matrix components and profibrotic molecules due to an acute and transient knock down of the itga8. 72 h after silencing with siRNA for the itga8, the strongest knock down effect of α8 integrin expression could be seen (Figure 1A and B). At this time point, a significant reduction of several matrix components like collagen III, collagen IV, fibrillin-1, and matrix proteinases was found, whereas an induction of CTGF could be detected using a real-time PCR compared to that of control siRNA-treated mesangial cells (Table 1).Thus, knock out and silencing of the itga8 in mesangial and vascular smooth-muscle cells led to incoherent changes in the expression of matrix components and profibrotic cytokines which seemed to be depending on cell type and permanent or transient knock down of the itga8.

Figure 1. Silencing of the α8 integrin chain by transfection of α8 integrin siRNA (si itga8) in mesangial cells. A, Expression of the α8 integrin chain normalized to 18S rRNA detected using real-time PCR, depicted as fold induction versus nonsilencing siRNA (si co). n = 6, data are means ± standard deviation. * p < 0.05. B, Western blot analysis of the α8 integrin chain 72 h after silencing of the α8 integrin chain in rat mesangial cells (RMC) by siRNA (si itga8). si co = transfection with nonsilencing siRNA. β-tubulin served as loading control.

Effects of overexpression of the itga8 in mesangial cells

24, 48, 72, and 96 h after transfection of wildtype mouse mesangial cells with the expression vector CMVitga8 and with the control plasmid pcDNA3.1, extracellular matrix components and profibrotic cytokines were measured using the real-time PCR. Overexpression of α8 integrin was detectable at all time points (Figure 2). While no relevant changes of matrix component expression could be detected after 24 h, a significant downregulation of MMP-2 and TIMP-2 after 48 and 96 h was measured compared to transfection control (Table 2). Regarding collagen expression, a significant change could only be detected after 72 h with significant downregulation of collagen III. After 96 h, CTGF was significantly downregulated in itga8 overexpressing mouse mesangial cells.

Figure 2. Overexpression of the α8 integrin chain in mesangial cells. Real-time PCR analysis of the expression of the α8 integrin chain in wildtype mouse mesangial cells normalized to 18S rRNA referred to control transfection (pcDNA). Results are representative for at least three similar experiments. Data are means ± standard deviation. *p < 0.05.

Table 2. Expression analysis of matrix components and regulators of matrix turnover in mesangial cells overexpressing the α8 integrin chain.

Transfection of wildtype mouse mesangial cells with an α8 integrin-coding plasmid (CMVitga8). Expression of the α8 integrin chain (itga8), collagen I (Col I), collagen III (Col III), collagen IV (Col IV), fibrillin-1 (FBN), fibronectin (FN), osteopontin (OPN), MMP-2, MMP-9, TIMP-1, TIMP-2, CTGF, TGFβ-1 normalized to 18S rRNA detected using a real-time PCR, depicted as fold induction versus control transfection with the empty pcDNA vector (pcDNA) 24, 48, 72, and 96 h after transfection. Results are representative for at least three similar experiments. Red: significant downregulation, p < 0.05. Green: significant upregulation p < 0.05.

24, 48, 72 and 96 h after transfection of α8 integrin-deficient mouse mesangial cells with the expression plasmid CMVitga8, extracellular matrix components were measured using the real-time PCR. A reexpression of the itga8 was detected at all time points investigated. The effects of the reexpression of the itga8 in α8 integrin-deficient mesangial cells on the expression of matrix components and profibrotic cytokines were similar to the effects of the overexpression of the itga8 in wildtype mouse mesangial cells (data not shown).

To confirm the functionality of the overexpressed itga8, we performed adhesion assays with wildtype and α8 integrin-deficient mouse mesangial cells. 48 h after transfection, cells were seeded on the α8 integrin ligand fibronectin and after 1 h, adhered cells were counted. Both wildtype and α8 integrin-deficient mouse mesangial cells adhered better on fibronectin after transfection with the expression plasmid CMVitga8 (Figure 3).

Figure 3. Adhesion of α8 integrin-overexpressing wildtype (wt MC, A) and α8 integrin-deficient (α8-/- MC, B) mouse mesangial cells. Forty-eight hours after transfection of cells with the expression plasmid CMVitga8 or the control vector pcDNA, cells were seeded on fibronectin and adhered cells were counted. C: exemplary photomicrographs of α8 integrin-deficient mesangial cells. Results are representative for three independent experiments. Data are means ± standard deviation. * p < 0.05.

Effects of α8 integrin overexpression in fibroblasts and tubular epithelial cells

To evaluate the effect of de novo expression of the itga8 in cells not expressing the itga8 under physiological conditions, we performed overexpression of the itga8 in fibroblasts and tubular epithelial cells.

Effect of itga8 expression on the expression of extracellular matrix components, regulators of matrix turnover and profibrotic cytokines in NIH3T3 cells: 24 and 48 h after transfection of NIH3T3 cells with the expression plasmid CMVitga8, the expression of the itga8 and extracellular matrix components was evaluated using the real-time PCR and compared to NIH3T3 cells transfected with the control plasmid pcDNA3.1. Even though significant overexpression of the itga8 was achieved, there was no significant induction or repression of the gene expression of matrix components or regulators of matrix turnover at both time points investigated (Table 3).

Table 3. Expression analysis of matrix components and regulators of matrix turnover in α8 integrin chain overexpressing fibroblasts.

Transfection of NIH3T3 fibroblasts with an α8 integrin-coding plasmid (CMVitga8). Expression of the α8 integrin chain (itga8), collagen I (Col I), collagen III (Col III), collagen IV (Col IV), fibrillin-1 (FBN), fibronectin (FN), osteopontin (OPN), MMP-2, MMP-9, TIMP-1, and TIMP-2, normalized to 18S rRNA detected using a real-time PCR, depicted as fold induction versus control transfection with the empty pcDNA vector (pcDNA) 24 and 48 h after transfection. Results are representative for at least three similar experiments. Green: significant upregulation p < 0.05.

Effect of itga8 expression on the expression of extracellular matrix components, regulators of matrix turnover and profibrotic cytokines in human tubular epithelial cells: 24 and 48 h after transfection of human tubular epithelial cells with the expression plasmid CMVitga8, an overexpression of the itga8 was detected (Figure 4A). Additional TGFβ-1 stimulation together with the transfection with CMVitga8 in tubular epithelial cells, resulted in itga8 expression levels similar to those achieved with CMVitga8 transfection only (not shown). Tubular epithelial cells changed their appearance after transfection with CMVitga8 to a more mesenchymal-like phenotype with cells becoming more spread or spindle-like (Figure 4B). Moreover, expression of the epithelial marker E-cadherin was reduced in these cells (Figure 4C). In tubular epithelial cells overexpressing α8 integrin, expression of extracellular matrix components and profibrotic cytokines was measured and compared to human tubular epithelial cells transfected with the control plasmid pcDNA3.1. While no significant changes could be seen 24 h after transfection with CMVitga8, a significant reduction of MMP-2, -9, collagen I, and IV expression compared to control transfection was detected using the real-time PCR after 48 h (Figure 5). This effect was partly reversed by additional stimulation with TGFβ-1. In addition, 48 h with or without additional stimulation with TGFβ-1 and preceding transfection with CMVitga8, a significant upregulation of fibronectin was detected (Figure 5). Thus, overexpression or de novo expression of α8 integrin in different cell types did not lead to uniform changes in the expression patterns of extracellular matrix components and profibrotic cytokines.

Figure 4. Overexpression of the α8 integrin chain in tubular epithelial cells. A, Overexpression of the α8 integrin chain after transfection of human tubular epithelial cells with an α8 integrin coding plasmid 24 and 48 h after transfection. Expression of the α8 integrin chain normalized to 18S rRNA detected using the real-time PCR, depicted as fold induction versus control transfection with the empty pcDNA vector (pcDNA). Results are representative for at least three similar experiments. B, representative photographs of CMVitga8 transfected (CMVitga8) and pcDNA control transfected (pcDNA) human tubular epithelial cells 72 h after transfection. α8 integrin chain overexpressing human tubular epithelial cells displays a more mesenchymal phenotype. C, E-cadherin expression in CMVitga8 transfected (CMVitga8) and pcDNA control transfected (pcDNA) human tubular epithelial cells. Data are means ± standard deviation. * p < 0.05.

Figure 5. Expression analysis of matrix components and regulators of matrix turnover in α8 integrin chain overexpressing tubular epithelial cells. Transfection of human tubular epithelial cells with an α8 integrin-coding plasmid. Expression of collagen I (Col I), collagen III (Col III), collagen IV (Col IV), fibronectin (FN), CTGF, TGFβ-1, osteopontin (OPN), fibrillin-1 (FBN), MMP-2, MMP-9, TIMP-1, and TIMP-2 was measured 48 h after transfection without additional TGFβ-1 (CMVitga8) and with additional TGFβ-1 stimulation (CMVitga8 + TGFβ-1). Expression is presented normalized to 18S rRNA detected using the real-time PCR, depicted as fold induction versus control transfection with the empty pcDNA vector (pcDNA). Results are representative for at least three similar experiments. Data are means ± standard deviation. *p < 0.05 CMVitga8 versus pcDNA. ^#p < 0.05 CMVitga8 versus CMVitga8 + TGFβ-1.

DISCUSSION

The data from our in vitro study in different cell types reveal heterogeneous effects of α8 integrin expression in different cell types regarding the regulation of extracellular matrix components. Moreover, there were differences in the regulation of matrix components and regulators of matrix turnover depending on whether the expression of the itga8 was permanently or transiently reduced. Overexpression of the itga8 did not lead to increases in the expression of matrix components (except for fibronectin in tubular epithelial cells), but more likely to a downregulation of several matrix molecules. Thus, our results do not argue for an unequivocal profibrotic effect of itga8 expression, neither in mesenchymal nor in epithelial cells, although data from previous experiments (Marek et al., 2010) suggested that the expression of the itga8 might promote a phenotypic change to a more activated mesenchymal phenotype, which is responsible for increased matrix production of cells affected.

In vivo studies in animal models revealed a coincidence of the induction of the itga8 with the development of fibrosis in lung, liver, heart, and kidney (Bouzeghrane et al., 2004; Hartner et al., 1999; Levine et al., 2000), arguing for profibrotic features of the itga8. Nevertheless, itga8 deficiency did not protect from the development of fibrosis in the kidney (Hartner et al., 2002). Recently, de novo expression of the itga8 was detected in tubulointerstitial fibrosis after unilateral ureteral obstruction (UUO) in rats and mice. Here, tubular interstitial epithelial cells and fibroblasts expressed the itga8 and its ligands in response to UUO, suggesting that α8β1 integrin might act to convey profibrotic signals. However, mice with itga8 deficiency presented with even more renal fibrosis after UUO than wildtype mice (Hartner et al., 2012). To determine whether the expression of the itga8 has direct effects on matrix accumulation, we analyzed the influence of itga8 over- or underexpression on the expression patterns of several matrix components and profibrotic cytokines in epithelial and mesenchymal cells. Underexpression of other integrin chains frequently resulted in a reduction of the expression of matrix molecules: Fibroblasts deficient for the β3 integrin chain expressed less fibronectin and reexpression of the β3 integrin chain in these cells upregulated fibronectin expression (Balasubramanian et al., 2012). SiRNA experiments attenuating the expression of β3 integrin or α2 integrin chains in fibroblasts led to a downregulation of collagen I expression (Honda et al., 2013; Suarez et al., 2013). Along with the loss of the itga8 in mesangial cells or vascular smooth muscle cells due to genetic or transient knock down, we detected a significant downregulation of some matrix components such as collagen III in mouse and rat mesangial cells as well as in mouse vascular smooth-muscle cells, suggesting a profibrotic role of the itga8. On the other hand, some profibrotic mediators were upregulated when the itga8 was knocked out, which is in keeping with findings in keratinocytes where downregulation of the α3 integrin chain resulted in increased expression levels of fibronectin and laminin (Wen et al., 2010). In our study, osteopontin was upregulated in α8 integrin-deficient mouse mesangial and vascular smooth-muscle cells as well as TGFβ-1 in α8 integrin-deficient mouse mesangial cells and CTGF in α8 integrin-deficient mouse vascular smooth muscle cells and rat mesangial cells with a transient knock down of the itga8. In contrast, CTGF was downregulated in α8 integrin-deficient mouse mesangial cells. Altogether, loss of the itga8 did not lead to uniform effects in all experiments, but changed the expression of extracellular matrix and its regulators depending on the cell type and on the kind of knock down of the itga8. These findings could at least partly be explained by differences in compensatory mechanisms after itga8 knock down: For example, in α8 integrin-deficient mesangial cells, not in α8 integrin-deficient vascular smooth-muscle cells or mesangial cells with a transient knock down of the itga8, an increased expression of the α2 integrin chain is observed, which might account for at least some of the expressional differences detected in these cells (Marek et al., 2010). However, whether the increase in the expression of the α2 integrin chain in α8 integrin-deficient mesangial cells could result in TGFβ-1 upregulation, remains unclear, because α2 integrin-deficient mice were shown to have increased TGFβ-1 expression levels (Girgert et al., 2010). Moreover, reexpression of the itga8 in α8 integrin-deficient mesangial cells did not restore the characteristics of wildtype mesangial cells regarding the expression patterns of matrix molecules, regulators of matrix turnover of profibrotic cytokines. This also argues for the existence of compensatory changes in α8 integrin-deficient mesangial cells.

To compare the effects of the de novo expression of the itga8 in epithelial cells with overexpression of the itga8 in mesenchymal cells, we additionally performed transfection experiments in mesangial cells, in fibroblasts and in tubular epithelial cells. Again, the effects of an overexpression of the itga8 turned out to be dependent on the cell type used for experiments. For example, overexpression of the itga8 in mesangial cells and in tubular epithelial cells showed a time dependent significant downregulation of matrix metalloproteases, while in fibroblasts overexpressing the itga8, the expression of matrix metalloproteases was the same as in control cells. Overexpression of other integrins even induced the expression of matrix metalloproteases in some cell types, for example overexpression of αvβ6 integrin in a keratinocyte cell line upregulated the expression of matrix metalloprotease-9 (Thomas et al., 2001). Synergistic action of integrins and TGFβ-1 on the expression of matrix components and regulators has been described frequently (Munger and Sheppard, 2011). For example, integrin α3β1 potentiates the TGFβ-1 mediated induction of matrix metalloprotease-9 (Lamar et al., 2008). Moreover, TGFβ-1-induced collagen expression was enhanced in cells overexpressing αvβ3 integrin (Hayashida et al., 2010). Therefore, we stimulated tubular epithelial cells overexpressing the itga8 with TGFβ-1 to test if α8β1 integrin might exert a profibrotic action by interacting with profibrotic cytokines. Additional TGFβ-1 stimulation of tubular epithelial cells seemed to somewhat antagonize the effects of the overexpression of the itga8, arguing against any synergistic cooperation of α8β1 integrin and TGFβ-1.

CONCLUSIONS

Taken together, the expression of the itga8 does not have uniform effects on the regulation of extracellular matrix. The α8 integrin-mediated regulation of the expression of matrix molecules, regulators of matrix turnover and profibrotic cytokines is complex and strongly depends on the cell type used and on the type of manipulation of the expression of the itga8. More likely, the expression of the itga8 seems to promote mesenchymal characteristics of cells, without inducing profibrotic properties by itself. Thus, using α8β1 integrin as a target for antifibrotic therapy might not be a useful approach.

Acknowledgments

We thank Dr. Ulrich Muller, San Diego, for kindly providing us with the α8 integrin-deficient mouse strain, Dr. Sven Kroening, Erlangen, for providing us with primary tubular epithelial cells, Dr. Kerstin Strehlow, Homburg, for assistance in establishing the technique of mouse vascular smooth muscle cell isolation and Dr. Margarete Goppelt-Struebe for critical discussion.

Declaration of interest: The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper.

This study was supported by a grant from the Deutsche Forschungsgemeinschaft, Bonn; Sonderforschungsbereich 423, TP A2 to AH and an educational grant from the Interdisciplinary Center for Clinical Research, University Hospital of Erlangen to IM. The funding bodies played no role in the design, in the collection, analysis or interpretation of the data, in the writing of the manuscript and in the decision to submit the manuscript for publication.