The Role of Extracellular Matrix Stiffness in Regulating Cytoskeletal Remodeling via Vinculin in Synthetic Smooth Muscle Cells

Abstract

Vinculin is a key player in sensing and responding to external mechanical cues such as extracellular matrix stiffness. Increased matrix stiffness is often associated with certain pathological conditions including hypertension induced cellular cytoskeleton changes in vascular smooth muscle (VSM) cells. However, little is known on how stiffness affects cytoskeletal remodeling via vinculin in VSM cells. Thus, we utilized matrices with elastic moduli that simulate vascular stiffness in different stages of hypertension to investigate how matrix stiffness regulates cell cytoskeleton via vinculin in synthetic VSM cells. Through selecting a suitable reference gene, we found that an increase in physiologically relevant extracellular matrix stiffness (2–50 kPa) downregulates vinculin gene expression but upregulates vinculin protein expression. This discrepancy, which was not observed previously for non-muscle cells, suggests that the vinculin-mediated mecahnotransduction mechanism in synthetic VSM cells may be more complex than those proposed for non-muscle cells. Also adding to previous findings, we found that VSM cell growth may be impeded by substrates that are either too soft or too rigid.

1. Introduction

Hypertension has been identified as an important risk factor for cardiovascular mortality.[1, 2] Increased aortic stiffness predicts the development of hypertension and consequent cardiovascular diseases. A stiff aorta resists the deformation to pulse pressure, causes vascular remodeling, and leads to hypertension.[3–5] Prolonged hypertension further exacerbates aorta stiffening by accelerating aorta structural changes.[6, 7] Although both vascular smooth muscle (VSM) cells and extracellular matrix (ECM) were found to contribute to aortic stiffness, ECM is considered to be the major determinant of aorta’s rigidity and affects VSM cell functions.[8] VSM cells may possess either contractile or synthetic phenotypes, and synthetic phenotype is involved in cell proliferation and migration and contributes to vascular remodeling in hypertension.[9, 10]

The effect of ECM stiffness for non-muscle cell functions has been well documented. ECM stiffness regulates cell migrations, affects directly cancer metastasis,[11, 12] and plays an essential role in stem cell differentiation. [13] However, few studies have focused on how ECM stiffness alters VSM cell behaviors.[8] Among possible cell mechanotransduction pathways—ion channels, actin cytoskeleton adaption, and G-protein-coupled receptors (GPCR)[14, 15]—cell cytoskeleton remodeling through focal adhesions (FAs) has been considered to be a more important mechanism.[16–20]. FAs, which are the protein complex localized at cell-ECM adhesions, link cell cytoskeletons with integrin to relay forces bidirectionally. Among various FA proteins, vinculin, which binds and crosslinks filamentous actin (F-actin) when located to FAs, has been the focus of recent intensive research due to its indispensable role as a mechanotransducer.[21–23] Vinculin has a muscle specific splice isoform, metavinculin, whose expression is negligible in synthetic VSM cells.[24]

Vinculin is expressed ubiquitously and localizes at both cell-ECM and cell-cell contacts. It comprises an N-terminal head domain (Vh), a C-terminal tail domain (Vt), and an interconnecting hinge domain. Vinculin is indispensable in cytoskeleton reorganization through binding and crosslinking F-actin. Upon binding to F-actin, Vt exposes cryptic homodimerization sites and subsequently bundles actin filaments.[25–27] In non-muscle cells, vinculin gene levels correlate with vinculin protein expression, and increased matrix stiffness enhances the binding between Vt and F-actin, and thus affects cells’ mechanotransduction and alters cytoskeleton remodeling.[28] However, little is known on how ECM stiffness affects cell cytoskeletal remodeling in VSM cells via vinculin.[29] In this study, we utilized hydrogel-based matrices with elastic moduli, which simulate vascular stiffness in different stages of hypertension, to examine how ECM stiffness regulates vinculin behaviors as well as cytoskeleton organization in synthetic VSM cells.

2. Materials and Methods

2.1. Cell culture

Rat thoracic aorta A10 cells (ATCC) were used as the model for synthetic VSM cells.[30] Following ATCC’s protocol, A10 cells were seeded on collagen I coated hydrogel matrices (2, 8, and 50 kPa) and the polystyrene surface ( > GPa) (Matrigen). The hydrogel-based matrices were selected to simulate the increased ECM stiffness of vascular tissues over hypertension development.[31, 32] Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (Corning). Bright field images of cells cultured on matrices were acquired for analyzing cell density and areas with ImageJ (NIH). Immunofluorescence images were also taken to monitor changes in cytoskeleton structure and vinculin localization in FAs. For western blotting and total mRNA preparation, cells on all matrices were harvested when cell confluency reached ~90% on the polystyrene surface. Cells were washed 2X with warm serum-free DMEM and lysed. At least three independent samples were prepared for both reverse transcription quantitative PCR reactions and western blotting.

2.2. Reverse transcription quantitative PCR (RT-qPCR)

RT-qPCR was employed to quantify vinculin gene (Vcl) expression levels in A10 muscle cells cultured on matrices with different stiffness (2, 8, 50 kPa, and polystyrene (> GPa)). Total cellular RNA samples were prepared using Qiagen RNeasy Plus Mini Kit (Qiagen). cDNA samples were synthesized using the High Capacity RNA-to-cDNA kit (Applied Biosystems) and were subsequently used as templates for PCR reactions. The transcriptional levels of Vcl were quantified using an ABI 7500 RT-qPCR system with TaqMan Gene Expression Assays for vinculin (Vcl, assay ID: Rn01755894_m1; Applied Biosystems) and reference gene HPRT1 (assay ID: Rn01527840_m1). The relative Vcl expression levels were calculated using the ΔΔCt method. Gene expression levels of other housekeeping genes, including GAPDH (assay ID: Rn01775763_g1), B2M (assay ID: Rn03928990_g1), and α-actin (ACTA1, assay ID: Rn01426628_g1), were also measured.

2.3. Western blotting analysis of cell lysates

To quantify protein expression levels of vinculin isoforms, cells were lysed using radioimmunoprecipitation assay lysis and extraction buffer (RIPA, ThermoFisher) supplemented with protease and phosphatase inhibitor cocktail tablet (Pierce). The lysates were then centrifuged at 15,000 × g and 4°C for 15 min. The supernatant that contained cytosolic and membrane-bound vinculin protein was transferred to a clean tube, while the pellet was washed with RIPA buffer to remove residual supernatant. To extract vinculin proteins bound to cytoskeleton, the pellet was re-suspended in RIPA buffer supplemented with SDS (1.2%, final concentration) and incubated at room temperature for 30 min. The suspension was centrifuged at 15,000 × g and 4°C for 15 min.

The samples were resolved on 12% gels and transferred to PVDF membranes for western blotting. Anti-vinculin antibody (Sigma-Aldrich) that reacts with both vinculin and metavinculin was used to detect the presence of metavinculin in cell lysates, and GAPDH was used as the loading control for western blot experiments. Recombinant vinculin protein (Abcam) and chicken gizzard muscle extract sample[33] were used as the positive controls for vinculin and metavinculin, respectively. HRP-conjugated secondary antibodies were purchased from ThermoFisher. Images were acquired with a Bio-Rad ChemiDoc XRS+ Imaging station.

2.4. Immunofluorescence imaging

Cells were cultured overnight on 6-well hydrogel bound glass coverslips coated with collagen I (Matrigen). Cells were then fixed, permeabilized, and stained with rabbit anti-vinculin antibody (Cell Signaling Technology) overnight at 4°C. Next, cells were washed and incubated at room temperature for 1 h with Alexa Fluor 488 goat anti-rabbit IgG and Alexa Fluor 568 phalloidin (Molecular Probes) to visualize vinculin in FAs and actin cytoskeleton, respectively. Immunofluorescence images were acquired on a Zeiss LSM 800 microscope equipped with Airyscan. Images were processed using the software Zen 2.3.

3. Results

We employed recombinant vinculin protein (Abcam) and chicken gizzard muscle extract as positive controls for vinculin and metavinculin, respectively. No metavinculin band (~150 kDa) was observed in any of A10 cell lysis supernatant or cytoskeleton samples, which confirmed previous observation that metavinculin is absent in synthetic VSM cells.

3.1. The role of ECM stiffness in vinculin gene (Vcl) expression

We measured Vcl levels in VSM cells cultured on substrates with four elastic moduli. To accurately evaluate Vcl expression levels, we first examined several housekeeping genes (HPRT1, B2M, GAPDH, ACTA1, and 18S RNA) to select a suitable reference gene to normalize Vcl levels from different matrices.[34, 35] Among these housekeeping genes, HPRT1 was selected due to its expression stability over different ECM stiffness. As matrix stiffness changed, significantly different C_t (threshold cycle) values were observed for all housekeeping genes except B2M and HPRT1. Further examination suggested HPRT1 as a better reference gene than B2M since the amplification curves of HPRT1 were almost identical for samples with different stiffness. Therefore, we employed HPRT1 as the reference gene for the subsequent vinculin gene expression experiments.

Vcl levels of A10 cells cultured on different matrices are summarized in Figure 1. It is evident that on a soft matrix (2 kPa), VSM cells had a relatively high Vcl level. When matrix stiffness increased to 8 kPa, Vcl level decreased steadily. When ECM stiffness further increased to 50 kPa, Vcl level decreased by 2-fold as compared with that at 2 kPa. When ECM became extremely rigid (polystyrene, > GPa), however, Vcl level increased to the level comparable to that at 2 kPa.

Figure 1.

Gene expression levels of vinculin (Vcl) relative to that of the reference gene HPRT1 in A10 cells cultured on matrices with different stiffness. Data are presented as mean ± SD. * Denotes P value < 0.05 compared to the 2 kPa.

3.2. The role of ECM stiffness in vinculin protein expression

We characterized protein expression of vinculin in VSM cells cultured on ECMs with different stiffness since metavinculin was absent in cultured A10 cells. To avoid possible disturbance to protein expression during cell lysis process, we washed cells with pre-warmed and serum-free DMEM medium immediately before adding RIPA lysis buffer. GAPDH was chosen to normalize vinculin quantity in the supernatant fraction (cytosol and membrane) across different samples because of its relatively high abundance and stable protein expression levels in tissues with different stiffness levels.[36] α-actin wasn’t used because of possible large variations in its protein expression levels over different stiffness.[36] Despite high Vcl levels at both 2 kPa and >GPa (polystyrene), vinculin protein expression level at 2 kPa in cytosol and membrane bound is significantly less than that from polystyrene (Figure 2). Another interesting observation is that as stiffness increased up to 50 kPa, Vcl level decreased but vinculin protein expression increased. On an extreme rigid surface (> GPa), however, Vcl level agreed well with protein expression level. Compared with cytosol and membrane fraction, the amount of vinculin bound to cytoskeleton under all ECM stiffness was negligible.

Figure 2.

Vinculin protein expression in A10 cells cultured on matrices with different stiffness: 2, 8, 50 kPa, and polystrene (> GPa). A) western blot of A10 samples: supernatant fractions (Lanes 1–4), cytoskeleton fractions (Lanes 5–8), recombinant vinculin (Lane 9), and chicken gizzard muscle extract (Lane 10). Vinculin bands (~ 130 kDa) and metavinculin band (~150 kDa) were labeled with arrows. 10 μg of total protein sample was loaded on each lane. B) The amount of vinculin in supernatant fractions relative to GAPDH. Data are presented as mean ± SD. * Denotes p value < 0.05 compared to the 2 kPa.

3.3. The impact of ECM stiffness on the amount of vinculin in FAs and actin cytoskeleton morphology

We also examined how the amount of vinculin in FAs and actin cytoskeleton morphology changed over ECM stiffness. The immunofluorescent images of F-actin (Figure 3) show that the amount of vinculin protein localized at FAs increased with stiffness as reflected by the increased number and size of FAs (Figure 3). This observation coincides with the elevated vinculin protein expressions in stiffer ECMs (Figure 2) despite the biphasic relationship between Vcl levels and matrix stiffness. As the amount of vinculin in FAs increased with stiffness, actin cytoskeleton changed dramatically. When VSM cells grew on a soft matrix (2 kPa), the actin bundles were small, thin, and irregularly arranged. In contrast, the cells cultured on more rigid surfaces (50 kPa hydrogel and extreme rigid surface (> GPa)), which also had significantly higher vinculin expression levels, possessed well organized actin skeleton and thick actin bundles. Thus, actin cytoskeleton in cultured VSM cells may be remodeled in the presence of elevated amount of vinculin in FAs to cope with stiffer external mechanical environment.

Figure 3.

Immunofluorescent images of A10 cells cultured on collagen coated matrices with selected stiffness. Actin was stained with phalloidin (red) while vinculin was visualized in green. scale bar: 10 μm.

3.4. The impact of ECM stiffness on cell growth and morphology

Previous studies of non-muscle cell lines revealed that cell growth might be significantly enhanced on stiff substrates.[37, 38] In this study, morphologies and cell densities of A10 cells cultured on three elastic moduli (2, 8, 50 kPa) were examined using differential interference contrast (DIC) microscopy (Zeiss, Figure 4A). A10 cells growing on soft matrices (2, 8 kPa) mostly assumed a spherical shape, suggesting a weak attachment of cells toward the matrix. As the stiffness increased to 50 kPa, cells spread and became polarized, and the cell area increased ~4.6-fold (Figure 4 B). Such an increase in cell areas agrees well with the trend of vinculin protein expression level and cytoskeleton remodeling when ECM became stiffer (Figures 2 and 3). However, cell densities on different ECM showed a different trend. Cell density first increased ~2.2-fold from 2 kPa to 8 kPa but dropped to ~1.1-fold at 50 kPa. Since the same number of A10 cells (1:4 passage ratio) were seeded on all matrices, the difference in cell density suggested matrices that are too soft (2 kPa) or too rigid (50 kPa) may impede VSM cell growth.

Figure 4.

A10 cell growth and morphology were affected by ECM stiffness. A) representative differential interference contrast (DIC) images of A10 cells cultured on collagen-coated ECM with different stiffness (scale bar: 50 μm); B) summary of A10 relative cell areas (n=30), and C) summary of cell densities (n=30) cultured on physiologically relevant 2–50 kPa hydrogel-based matrices. Data are presented as mean ± SD. ** Denotes p value < 0.01 compared to controls (2 kPa).

4. Discussion

Cell sensing and responding to mechanical cues such as ECM stiffness are critical for accomplishing various cellular functions. VSM cell behaviors change in response to increased ECM stiffness in pathological conditions, including hypertension and cardiovascular diseases.[39, 40] In this study, we investigated how ECM stiffness regulates vinculin behaviors (gene and protein expressions) and subsequent cell cytoskeletal changes in synthetic phenotype VSM cells, because vinculin expression levels have been suggested to affect cytoskeletal remodeling in non- muscle cells.[20, 41, 42] We observed a novel phenomenon that ECM stiffness has a differential effect on vinculin gene and protein expression levels in synthetic VSM cells.

Since both cell morphology and metabolic activities are significantly affected by matrix rigidity, it is vital to select proper reference gene(s) for evaluating Vcl levels and internal control(s) for comparing vinculin protein expression over matrices with various stiffness.[34, 43] Among the housekeeping genes surveyed for suitability, HPRT1 showed the most stable gene expression levels as reflected by steady C_t (threshold cycle) values across a wide range of stiffness (2 kPa ~ GPa) and was thus selected as the reference gene. Expression levels of ACTA1 varied significantly as ECM stiffness changed, and this agrees with the morphological differences among actin fibers in VSM cell cytoskeleton on different matrices (Figure 3).

Vcl was downregulated when ECM rigidity increased within physiological relevant ranges (2–50 kPa). When ECM stiffness increased from 2 to 8 kPa—simulating arterial stiffness of adults, Vcl level reduced by ~1.3-fold. When ECM became more rigid (50 kPa)—representing increased stiffness in hypertension, Vcl level decreased to half of that at 2 kPa. It is known that extracellular domain of integrins, including α1β1 and α2β1, specifically bind collagen.[44] On a soft matrix, β1 integrin is downregulated through lysosomal degradation pathway.[45] Because vinculin in FAs engages cytoskeleton to ECM through integrin and is involved in cell spread and growth, a higher Vcl level on the soft matrix could be attributed to genetic compensation mechanism.[46] In fact, a recent study shows that the transcriptional program in endothelial cells induced by stiffness is regulated by GATA Binding Protein 2 (GATA2) transcription factor, and that genes involved in cell-ECM adhesion are upregulated on soft ECM (0.2 kPa).[47] It is intriguing that Vcl on the extreme stiff matrix (polystyrene, > GPa) was upregulated to the level comparable with that on the soft (2 kPa) matrix. Such an observation could be attributed to the upregulation of integrin in VSM cells on collagen-coated polystyrene surface, which increases the number of focal adhesions and results in upregulation of Vcl.

While Vcl levels gradually decreased as physiologically relevant ECM stiffness increased from 2 to 50 kPa, vinculin protein expression levels increased. This finding is different from that observed in non-muscle cells, where vinculin protein expression increases with gene expression. One possible reason for the discrepancy between gene and protein expression could be the lack of polypyrimidine tract binding protein (PTB), an RNA-binding protein, in cytoplasm. PTB is abundant in smooth muscle tissues and regulates mRNA translation.[48] Previous research suggested that vinculin protein expression in non-muscle cells is regulated by PTB, which typically locates within nucleus but transiently relocates to cytoplasm.[49] Compared with that on more rigid ECMs, only a small fraction of PTB may relocate to cytoplasm when cultured on soft ECMs such as 2 and 8 kPa, and thus results in low amount of vinculin protein on soft matrices. It should be noted that the amount of vinculin in FAs also increased with ECM stiffness and resulted in cytoskeletal reorganization. In addition, cell density on matrices with different stiffness showed a different trend and suggested VSM cell growth may be impeded by matrices that are either too soft or too rigid. This finding is different from and complements to previous studies, which showed that non-muscle cell growth might be significantly enhanced on stiff substrates.[37, 38]

In conclusion, we observed a novel phenomenon that ECM stiffness has a differential effect on vinculin gene and protein expression levels in synthetic VSM cells. Within physiological stiffness range, increased ECM stiffness leads to downregulation of vinculin gene expression but upregulation of vinculin protein expression. Beyond vinculin-ligand binding as suggested by previous studies, our results show that vinculin-mediated mechanotransduction in synthetic VSM cells may also involve the regulation of both vinculin gene and protein expression levels, and that only vinculin protein levels correlate with cell cytoskeleton remodeling under different ECM stiffness. Also adding to previous findings, our results show that VSM cell growth may be impeded by substrates that are either too soft or too rigid. Overall, our work sheds light on possible mechanotransduction mechanisms through which vinculin behaviors in synthetic VSM cells are regulated in response to matrix stiffness and may inform potential therapeutics for regulating cellular response to pathological ECM stiffening in hypertension development.