Redirecting valvular myofibroblasts into dormant fibroblasts through light-mediated reduction in substrate modulus

Abstract

Fibroblasts residing in connective tissues throughout the body are responsible for extracellular matrix (ECM) homeostasis and repair. In response to tissue damage, they activate to become myofibroblasts, which have organized contractile cytoskeletons and produce a myriad of proteins for ECM remodeling. However, persistence of myofibroblasts can lead to fibrosis with excessive collagen deposition and tissue stiffening. Thus, understanding which signals regulate de-activation of myofibroblasts during normal tissue repair is critical. Substrate modulus has recently been shown to regulate fibrogenic properties, proliferation and apoptosis of fibroblasts isolated from different organs. However, few studies track the cellular responses of fibroblasts to dynamic changes in the microenvironmental modulus. Here, we utilized a light-responsive hydrogel system to probe the fate of valvular myofibroblasts when the Young's modulus of the substrate was reduced from ~32 kPa, mimicking pre-calcified diseased tissue, to ~7 kPa, mimicking healthy cardiac valve fibrosa. After softening the substrata, valvular myofibroblasts de-activated with decreases in α-smooth muscle actin (α-SMA) stress fibers and proliferation, indicating a dormant fibroblast state. Gene signatures of myofibroblasts (including α-SMA and connective tissue growth factor (CTGF)) were significantly down-regulated to fibroblast levels within 6 hours of in situ substrate elasticity reduction while a general fibroblast gene vimentin was not changed. Additionally, the de-activated fibroblasts were in a reversible state and could be re-activated to enter cell cycle by growth stimulation and to express fibrogenic genes, such as CTGF, collagen 1A1 and fibronectin 1, in response to TGF-β1. Our data suggest that lowering substrate modulus can serve as a cue to down-regulate the valvular myofibroblast phenotype resulting in a predominantly quiescent fibroblast population. These results provide insight in designing hydrogel substrates with physiologically relevant stiffness to dynamically redirect cell fate in vitro.

Figure 1. Cell fate in response to substrate modulus reduction.

Valvular interstitial cells (VICs) were seeded on photodegradable poly(ethylene glycol) (PD-PEG) gels on day 0. At day 3, a portion of the stiff gels was softened with light (365 nm at 10 mW/cm²). The fate of VICs on continuously stiff, continuously soft and stiff-to-soft gels was subsequently examined on day 3 and/or day 5 based on immunocytochemistry, apoptosis staining, proliferative assay and mRNA expression.

Figure 2. Reduced myofibroblast activation in response to lowering substrate modulus.

VICs cultured on stiff, soft or stiff-to-soft gels as shown in Fig.1 were fixed on day 3 and day 5, and stained for α-smooth muscle actin (α-SMA). (A) Representative staining of the myofibroblast phenotype for VICs cultured on substrates with different stiffnesses on day 3 and day 5. Green: α-SMA. Blue: nuclei. Arrows: myofibroblasts characterized by organized α-SMA+ stress fibers. Star: a cell stained with diffusive α-SMA. Scale bar: 100 µm. (B) Quantification of the percent of myofibroblasts on the substrates based on staining in (A). The percentage of myofibroblasts on stiff-to-soft gels or soft gels was significantly lower than that on stiff gels. * indicates p<0.05.

Figure 3. Decreased number of myofibroblasts on stiff-to-soft gels was not due to apoptosis.

VICs cultured on different PD-PEG gels were stained with Annexin V linked with Alexa Fluor 594 and DAPI to detect apoptosis. (A) Scatter plot of Annexin V and DAPI staining for VICs cultured on stiff gels. Red box: apoptotic cells with high fluorescence in the Annexin V channel and low fluorescence in the DAPI channel. (B) A representative confocal image of a cell stained positively for Annexin V (red) overlaid with transmitted light DIC on stiff gels. Positively-stained cells were observed to exhibit a rounded morphology. Blue: Nucleus. Scale bar: 20 µm. (C) Quantification of apoptosis based on flow cytometry as shown in (A). Low levels of apoptosis were detected for VICs cultured on either gels or plastic plates. VICs treated with camptothecin, an apoptosis-inducing reagent, showed a much higher level of apoptosis than any gel-based culture condition or plastic plate. * indicates p<0.05.

Figure 4. VICs switch to a less activated and less proliferative fibroblast phenotype on softer substrates.

(A) 6 hours after switching stiff gels to soft on day 3, VICs were collected for mRNA quantification based on real-time PCR. Myofibroblast gene markers, α-SMA and connective tissue growth factor (CTGF), were significantly down-regulated in soft and stiff-to-soft conditions, compared with stiff. The fibroblast gene marker, vimentin, was expressed at a similar level on different substrates. (B) To measure proliferation, VICs cultured on stiff, soft or stiff-to-soft gels were chased with EdU for 3 hours on day 5. EdU incorporation into DNA was detected by labeling EdU with Alexa Fluor 488 and was quantified via flow cytometry. Relative proliferation on day 5 of VIC culture was calculated based on normalizing the percent of EdU+ cells in each condition to that of the stiff condition. VICs were less proliferative on soft and stiff-to-soft gels than on stiff gels. (C) Representative EdU staining of VICs cultured on PD-PEG gels. As expected, EdU staining (green) is co-localized with nuclei (blue). Scale bar: 100 µm. * indicates p<0.05. (D) VICs cultured on plastic plate, stiff, soft or stiff-to-soft gels were fixed on day 5 and stained for vimentin. De-activated fibroblasts on stiff-to-soft gels maintain the mesenchymal fibroblast fate. Green: vimentin. Blue: nuclei. Scale bar: 100 µm.

Figure 5. Deactivated VICs on stiff-to-soft gels enter the cell cycle with proliferative stimulus and activate myofibroblast the gene program in response to TGF-β1.

VICs cultured on stiff-to-soft gels were treated with either proliferative media with 15% FBS and fibroblast growth factor 2 (FGF-2) or fibrogenic chemokine (TGF-β1) on day 4 for 24 hours. (A) Proliferation was measured by EdU chase for 1 hour on day 5. VICs treated with growth stimulus had ∼4 fold more proliferating cells than those in control medium. (B) Myofibroblast gene markers, including CTGF, collagen 1a1 (Col1a1) and fibronectin 1 (FN1), were significantly up-regulated in deactivated VICs treated with TGF-β1. The mRNA level of α-SMA was not changed significantly by TGF-β1 treatment. (C) Immunocytochemistry of α-SMA showed similar levels of myofibroblasts on stiff-to-soft gels treated with or without TGF-β1. Green: α-SMA. Blue: nuclei. These results show that the de-activated fibroblasts have the potential to proliferate and to activate fibrogenic associated genes in response to chemical cues, but a stiffer substratum is likely required for α-SMA stress fiber formation. Scale bar: 100 µm. * indicates p<0.05.