Fibroblast growth factor 2 inhibits myofibroblastic activation of valvular interstitial cells.

Heart valve disease is a growing problem worldwide. Though very common in older adults, the mechanisms behind the development of the disease aren't well understood, and at present the only therapeutic option is valve replacement. Valvular interstitial cells (VICs) may hold the answer. These cells can undergo pathological differentiation into contractile myofibroblasts or osteoblasts, leading to thickening and calcification of the valve tissue. Our study aimed to characterise the effect of fibroblast growth factor 2 (FGF-2) on the differentiation potential of VICs. We isolated VICs from diseased human valves and treated these cells with FGF-2 and TGF-β to elucidate effect of these growth factors on several myofibroblastic outcomes, in particular immunocytochemistry and gene expression. We used TGF-β as a positive control for myofibroblastic differentiation. We found that FGF-2 promotes a 'quiescent-type' morphology and inhibits the formation of α-smooth muscle actin positive myofibroblasts. FGF-2 reduced the calcification potential of VICs, with a marked reduction in the number of calcific nodules. FGF-2 interrupted the 'canonical' TGF-β signalling pathway, reducing the nuclear translocation of the SMAD2/3 complex. The panel of genes assayed revealed that FGF-2 promoted a quiescent-type pattern of gene expression, with significant downregulations in typical myofibroblast markers α smooth muscle actin, extracellular matrix proteins, and scleraxis. We did not see evidence of osteoblast differentiation: neither matrix-type calcification nor changes in osteoblast associated gene expression were observed. Our findings show that FGF-2 can reverse the myofibroblastic phenotype of VICs isolated from diseased valves and inhibit the calcification potential of these cells.

Introduction

Heart valve disease is highly prevalent globally [1], and as populations age, the incidence of aortic stenosis and end-stage calcific aortic valve disease (CAVD) is set to drastically increase [2]. At present, the only therapeutic option for treating these diseases is replacement of the valve with a prosthesis [3]. Despite its prevalence, not much is known about the molecular mechanisms behind the development of heart valve disease. Gone are the days when age-related disease such as CAVD were described as simple ‘wear and tear’ [4-6]. Instead, recent research has pointed to dysregulation of resident cells as the likely culprit [7-9]. Valvular interstitial cells (VICs) are the fibroblasts that populate the interstitium of the valve. Aortic valves are composed of three distinct layers: the collagen-rich fibrosa on the aortic side, the elastin-rich ventricularis underneath, and the GAG-rich spongiosa sandwiched between [10]. VICs are a heterogenous cell population, and have distinct phenotypes in keeping with the different functions of the valve layers [11]. Surrounding the valve is a monolayer of valvular endothelial cells (VECs), also distinct from endothelial cells found elsewhere in the vasculature [12]. Gene expression profiles of VICs and VECs differ from valve to valve, from leaflet to leaflet, and even from side to side of the same leaflet [13-15]. The VICs and VECs work in tandem to maintain the integrity and extracellular matrix (ECM) homeostasis of the valve over the lifetime of the heart, some 3 billion beats [8, 16]. Previous research has highlighted the role of transforming growth factor β (TGF-β) stimulated ‘activation’ of VICs into myofibroblasts, termed ‘activated VICs’ or aVICs in the context of the heart valve [17, 18]. Myofibroblasts are a contractile phenotype of fibroblasts implicated in matrix production during wound healing. In health, these cells maintain tissue homeostasis, but excessive myofibroblastic activation is responsible for a host of fibro-contractile diseases [19]. In the aortic valve, this presents initially as thickening of the valve leaflet (aortic sclerosis) followed by progressive calcification leading to end-stage CAVD [20]. A thorough review of VIC physiology by Rutkovskiy et al. describes the bi-potent differentiation potential of these cells into aVICs or osteoblasts [16]. Both are responsible for valve calcification, albeit through two distinct mechanisms [8]. TGF-β is a powerful stimulator of myofibroblast activation through a number of biochemical pathways, in particular the canonical and non-canonical SMAD pathways [16, 21]. Recently, evidence has emerged that fibroblast growth factor 2 (FGF2) opposes this effect [22-24]. However, this evidence base is founded on either porcine cells [22, 23] or cells isolated from healthy human donors [24]. What remains unclear is if these anti-myofibroblast effects are able to halt or reverse pathological changes in already diseased valve cells. FGF-2 has also been shown to inhibit osteoblastic-type calcification of porcine VICs via the Notch1 pathway [25]. This study aimed to elucidate the effect of FGF-2 on key outcomes relating to 1) myofibroblastic activation, and 2) calcification of VICs from diseased patients. We aimed to compare the observed changes in FGF-2 stimulated cells to TGF-β stimulated VICs, which can be thought of as a positive control for myofibroblastic activation. In particular, our study aimed relate myofibroblastic changes to changes in gene expression. We assayed VICs isolated from patients with end-stage CAVD (a cell population displays a high degree of myofibroblastic activation from the outset), and any improvement in outcomes would constitute an in vitro therapeutic effect. We hypothesised that FGF-2 would oppose the myofibroblastic effect demonstrated in the TGF-β group in all areas and reverse the calcification potential of diseased VICs compared to control. We also hypothesised that osteoblastic differentiation would be inhibited by FGF-2 and promoted by TGF-β.

Materials and methods

hVIC isolation and expansion

Ethical approval was obtained for the use of human tissue from Auckland Health Research Ethics Committee (approval #AH1186). All participants provided written consent prior to tissue collection. 52 aortic valves were obtained from patients undergoing surgical valve replacement at Auckland City Hospital. All patients had end-stage CAVD. Explanted valve leaflets were agitated in phosphate buffered saline (PBS) containing 1000U/mL collagenase type II (Sigma) for 10 minutes at 37˚C to loosen the endothelium. The suspension resulting from this first digest was centrifuged and cultured separately to identify endothelial cells. Leaflets were then scraped with a scalpel blade to remove remaining endothelial cells, then rinsed in PBS. Calcified portions of the valve were removed, and the remaining tissue was finely minced. The tissue was further digested in 1000U/mL collagenase for 2 hours at 37˚C under agitation. The resulting suspension was passed through a 70μm cell strainer to remove debris, centrifuged, and resuspended in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Cells were passaged at 80% confluency and expanded to yield at least 1x106 cells per patient. All cultures were used between passage 2 and 5.

Experimental conditions

hVICs were cultured in 3 media formulations:

‘Control’ medium: DMEM supplemented with 10% FBS

‘FIB’ medium, previously described by Latif and colleagues [24]: DMEM supplemented with 2% FBS, 10ng/mL FGF-2 (Thermo Fisher) and 50ng/mL insulin (Sigma). Per Latif et al., the low serum content was used to reduce the inherent concentration of TGF-β present in FBS, and the insulin was added as mitotic agent

‘TGF-β’ medium: DMEM supplemented with 10% FBS and 20ng/mL TGF-β (Thermo Fisher). This media was our ‘positive control’ for aVIC differentiation

In all assays, hVICs from a single donor were initially seeded in control medium and allowed to adhere for 24 hours. Control medium was then aspirated and replaced with the one of the above formulations. Unless otherwise stated, endpoints were measured on day 2 and 4.

alamarBlue cell viability assay

7.5x103 hVICs were seeded in 950μL medium, in a 24-well plate with replicates of 4. Negative control wells contained each of the corresponding media without cells. At time = 0h, 24h, 48h, 72h, and 96h, 50μL of alamarBlue cell viability reagent (Thermo Fisher) was added to each well (final concentration = 5%) and incubated at 37˚C for 4 hours. Media was aspirated and added to a 96-well plate in triplicate. Fluorescence was measured at 530/590nm on a Synergy 2 microplate reader (Biotek). Absolute fluorescence values for 4 biological repeats were combined and are presented graphically normalised to each well on day 0. An ordinary two-way ANOVA was used to compare treatments to control at each time point. To investigate relative rates of proliferation, a simple linear regression model was used to determine the gradient of the line of best fit from 24hrs onwards. The 0h data point was excluded from the linear regression analysis because fluorescence values were not comparable in the FIB group; the change in serum content after the 0hrs caused inherent differences in fluorescence of the media. Gradients were normalised to control in each replicate, and a one-way ANOVA was used to compare the two treatment groups to control.

Immunocytochemistry

1x104 hVICs were seeded in 1mL medium in 24-well plates and cultured for 2 and 4 days. Media were changed on day 2. At the designated endpoints, wells were washed twice in PBS and fixed in 4% paraformaldehyde at room temperature for 20 minutes. hVICs were permeabilised in 0.1% v/v Triton X100 in PBS for 15 minutes, before rinsing three times in PBS. Cells were then blocked in PBS supplemented with 5% serum from the species in which the secondary antibody was raised. Five immunocytochemistry (ICC) groups were used, detailed in Table 1. The vimentin/α-smooth muscle actin (α-SMA) protocol was used to measure cell morphology data (area, major and minor axes, aspect ratio, and circularity), as well as α-SMA staining intensity. The von Willebrand factor protocol was used to ensure the cultures were free from endothelial cells. In this protocol, endothelial cell culture obtained from the first collagenase digest was used as a positive control. Paxillin staining was used to resolve and count the number of focal adhesions per cell. SMAD2/3 staining was used to identify nuclear localisation of the phosphorylated SMAD2/SMAD3 complex. Scleraxis staining was used to identify the intracellular location of scleraxis, a transcription factor associated with tendon development and cardiac myofibroblasts.

Table 1

Immunocytochemistry (ICC) protocols.

ICC group	Primary antibody	Protocol	Secondary antibody	Protocol	Counterstain	Protocol
Vimentin and α-SMA	Vimentin, rabbit anti-human (abcam, 92547)	1/800 dilution, incubated for 12 hours at 4˚C	Donkey anti-rabbit IgG Alexa Fluor 488 conjugated, (Invitrogen, A32790)	1/400 dilution, incubated for 1 hour at room temperature	DAPI	2μg/mL, incubated for 5 minutes at room temperature
	Α-SMA, Cy3 conjugated, mouse anti-human (Sigma C6198)	1/200 dilution, incubated for 12 hours at 4˚C	n/a (directly conjugated)
Von Willebrand factor (vWF)	vWF rabbit anti-human (abcam, 6994)	1/200 dilution, incubated for 12 hours at 4˚C	Donkey anti-rabbit IgG Alexa Fluor 488 conjugated, (Invitrogen, A32790)	1/400 dilution, incubated for 1 hour at room temperature	DAPI	2μg/mL, incubated for 5 minutes at room temperature
					Phalloidin	1/50 dilution, incubated for 1 hour at room temperature
Paxillin	Paxillin, mouse anti-human (Invitrogen, MA513356)	1/50 dilution, incubated for 12 hours at 4˚C	Goat anti-mouse IgG Alexa Fluor 488 conjugated, (abcam, ab150113)	1/400 dilution, incubated for 1 hour at room temperature	none
SMAD2/3	Activated SMAD2/3 complex, rabbit anti-human (Invitrogen, PA5110155)	1/100 dilution, incubated for 12 hours at 4˚C	Donkey anti-rabbit IgG Alexa Fluor 488 conjugated, (Invitrogen, A32790)	1/400 dilution, incubated for 1 hour at room temperature	Phalloidin	1/50 dilution, incubated for 1 hour at room temperature
Scleraxis	Scleraxis, mouse anti-human (Santa Cruz Biotech, 518082)	1/100 dilution, incubated for 12 hours at 4˚C	Goat anti-mouse IgG Alexa Fluor 488 conjugated, (abcam, ab150113)	1/400 dilution, incubated for 1 hour at room temperature	Phalloidin	1/50 dilution, incubated for 1 hour at room temperature

ICC images were captured at 40x magnification on an Olympus CKX41 microscope fitted with an Olympus DP74 camera and processed using ImageJ. For the vimentin/α-SMA protocol, images were taken from n = 6 biological replicates. For each timepoint, 6 images were taken per replicate for a total of 36 images per experimental group. Images were thresholded in the green channel (vimentin) and cell morphology data was obtained from cell outlines. Cells were excluded from analysis if they overlapped other cells, or were undergoing mitosis (i.e., 2 visible nuclei), otherwise all visible cells were counted. For the paxillin stained cells, images were taken from n = 4 biological replicates. For each time point, 10 images were taken per replicate for a total of 40 images per experimental group. Images were thresholded and outlined, before being ‘despeckled’ to remove artefacts and non-specific staining. Mature focal adhesions were then counted manually. For SMAD2/3 staining, n = 3 biological repeats were imaged on day 2. For scleraxis staining, n = 3 biological repeats were stained on day 2 and 4.

Alizarin red calcification stain

hVICs (n = 4) were seeded, cultured, and fixed as described in the ICC section. 2% w/v alizarin red in diH2O was passed through a 0.2μm filter and adjusted to pH 4.1 with 12M HCl. 2 wells per replicate were incubated in the alizarin red for 15 minutes at room temperature before washing 5 times with PBS. Calcific nodules were counted manually, and representative images were captured at 40x magnification.

Gene expression

5x104 hVICs were seeded in 6-well plates in 2.5mL medium and cultured for 2 and 4 days. Cells were harvested by adding 600μL RLT lysis buffer (Qiagen) supplemented with 1% v/v β-mercaptoethanol and dissociated from the plastic using a cell-scraper. Lysates were homogenised by passing through a 22-gauge needle 5 times. RNA was extracted using the RNeasy mini kit (Qiagen) according to the manufacturer’s protocol. Genomic DNA was depleted on-column using DNase I (Qiagen). RNA concentration and purity was assessed using Nanodrop. RNA integrity was assessed using RNA screen tape (Agrilent). 500ng of RNA was reverse transcribed using the SuperScript III system (Invitrogen), modified with 100mM RNaseOUT RNase inhibitor (Invitrogen). Real-time quantitative polymerase chain reaction (RT-qPCR) was performed on the samples using Taqman probes detailed in S1 Table. Briefly, cDNA, probes, and mastermix were added to a 384-well plate in triplicate in accordance with the manufacturer’s protocol and cycled on a Quantstudio 5 real-time PCR machine. 18S was used as the housekeeper gene. Genes were considered ‘not detectable’ if mean Ct values exceeded 37. Gene expression changes were calculated using the 2-ΔΔCt (Livak) method [26], where both FIB and TGF-β groups were compared to the control group at day 2. Gene expression results are reported as ‘fold change’ and are presented graphically as log2(fold change). RT-qPCR experiments were performed twice to ensure reproducibility; presented data is from the first replicate.

Statistics

Data were processed using Prism (GraphPad). For semiquantitative analysis of immunocytochemistry data (cell size and shape parameters), an ordinary 2-way ANOVA with Šidák correction was used to determine inter-group and inter-time point differences. The two experimental groups were compared to the control group within each time point. For gene expression data, a mixed effects 2-way ANOVA was applied to the ΔΔCt values. Each experimental group was compared to the control group within the same time point. Across all statistical models, differences between experimental group versus control were considered significant if p<0.05, and are marked with *, differences where p<0.01 are marked with #.

Results

Cell culture and gene expression: quality control

Of the 52 aortic valves obtained, 40 resulted in successful cell populations. Valves that were heavily calcified, or from individuals >80 years of age typically did not yield enough viable cells to form a culture. We used vWF immunostaining to ensure absence of valvular endothelial cells (hVECs) in culture (Fig 1). As a positive control, cells obtained from the first collagenase digest were plated and stained alongside the VIC cultures. The hVEC positive control culture showed strong presence of vWF, while the ‘VIC’ only cultures were negative for vWF. Measures of nucleic acid purity and integrity are available in S2 Table.

Fig 1

vWF immunostaining.

hVICs and hVECs stained for vWF (green), phalloidin (red), and DAPI (blue). Scale bars represent 100μm.

FIB media is pro-proliferative

hVICs treated with FIB medium containing FGF-2 caused a significant increase in cell viability, indicative of an increase in cell number (Fig 2A). Marked increase in fluorescence from time 0h to 24h can be explained by inherent differences in the fluorescence of media containing different serum concentrations, and so time = 0h was excluded from linear regression analysis. Lines of best fit were added to fluorescence values of each replicate independently, with a mean r2 value of 0.895 (σ2 = 0.082). Absolute gradient values varied substantially between repeats, and so each repeat was normalised to the gradient of the control line. Gradients in the FIB group were significantly greater than in control (p = 0.0057), indicating that hVICs grown in this medium proliferated at a higher rate. TGF-β did not influence viability. A limitation of the linear regression method is it assumes cells proliferate in a linear fashion, which in practice they don’t [27]. However, we believe assessing the rate of change over three 24-hour periods is sufficiently short to apply a linear fit.

Fig 2

FIB media is pro-proliferative.

(A) Fluorescence data from alamarBlue proliferation assay (n = 4), normalised to day 0. The two treatment groups were compared to control at each time point using an ordinary 2-way ANOVA and (B) the gradients of the line of best fit derived from a simple linear regression model of 4 repeats, normalized to control. Both treatments were compared to control using an ordinary one-way ANOVA. * p<0.05, #p<0.01.

FGF-2 promotes quiescent-like morphology

hVICs displayed variable morphologies depending on the media formulation. Cells seeded in control and FIB media tended to grow separately or on top of each other, while cells in TGF-β supplemented media formed extensive cell-cell junctions, and in areas of high density formed a monolayer sheet (Fig 3A). Cells in the control group showed the most variability in their shape and growth pattern, while cells in the FIB and TGF-β groups tended to display similar characteristics across repeats: namely, FIB cells were slender, spindle-shaped, with small bodies and long processes, and TGF-β cells were uniformly very large and round. The variable that differed most markedly between groups was cell area. FIB-media hVICs over both time points measured, on average, 2134 μm2 compared to 3238 μm2 in the control group, representing a 35% reduction (Fig 3B). TGF-β treated hVICs on the other hand were more than double the size of control cells, with a mean area of 7943 μm2. FIB-treated cells had a significantly higher aspect ratio (Fig 3C, p = 0.0014) and significantly lower circularity (Fig 3D, p<0.0001), indicating the cells were more spindle-like in shape. In both these variables, TGF-β treated cells did not differ from control. Finally, the length of the major axis was somewhat reduced in the FIB group (Fig 3E, p = 0.0218), but markedly increased in the TGF-β group (p<0.0001). Interestingly, 2-way ANOVA did not find any significance in the ‘time factor’, indicating no significant change between time points in any of the morphology variables measured.

Fig 3

FGF-2 and TGF-β differentially influence cell morphology.

(A) depicts representative images of hVICs on day 2 and 4, immunostained for α-SMA (red), vimentin (green), and DAPI (blue). Scale bars represent 100μm. The following morphological characteristics are derived from pooling days 2 and 4, as the 2-way ANOVA showed that the ‘time’ factor did not significantly affect any parameter. Comparisons are ordinary one-way ANOVAs between control and the experimental groups. For image analysis, 6 biological replicates were used (n = 6); 6 representative images were taken from each biological replicate per treatment per timepoint, representing >75 cells per experimental group. These are: (B) cell area in μm2, (C) aspect ratio, the long axis divided by the short axis of each cell, (D) the ‘circularity’ of each cell. This measure is defined by 4π(area/perimeter2), and E) the length of the major axis. * p<0.05, #p<0.01, error bars represent 95% CI.

FGF-2 inhibits α-SMA expression

α-SMA is a well-characterised marker of aVICs [17, 22, 24, 28]. Immuno-staining (Fig 3A) revealed heterogenous expression of α-SMA in the control group on day 2, with approximately half the cells showing some degree of expression, indicating a high degree of inherent activation. By day 4, thick bands of α-SMA were present in the majority of control cells. In the FIB group, over both time points, α-SMA staining was exceedingly low. The TGF-β treated hVICs, however, expressed large amounts of α-SMA over both time points, organised into thick, mature bands that extended across cell boundaries, linked by focal adhesions. The apparent down- and up-regulation of α-SMA in FIB and TGF-β groups respectively is supported by the RT-qPCR results (Fig 4). FIB-treated hVICs showed 7.3-fold reduction in ACTA2 expression on day 2 (95% CI: 5.2–12.3), and a 7.9-fold reduction on day 4 (95% CI: 3.5–112) when compared to day 2 control. Conversely, TGF-β treated hVICs demonstrated a 3.5-fold upregulation on day 2 (95% CI: 2.1–5.0) and 2.5-fold upregulation on day 4 (95% CI: 0.2–4.7) compared to control. Vimentin is a cytoskeletal protein present in all VIC phenotypes, and perhaps unsurprisingly, staining was uniform across treatment groups and time points (Fig 3A). The FGF-2 treated cells did show a small but significant upregulation in VIM expression at day 2 and 4, while TGF-β had no effect (Fig 4). Our findings indicate that α-SMA and its corresponding gene ACTA2 remain useful markers of TGF-β stimulated activation of hVICs. FGF-2 treatment reduced visible immune-staining of α-SMA and expression of ACTA2.

Fig 4

FGF-2 downregulates ACTA2 expression.

RT-qPCR results of ACTA2 (α-SMA) and VIM (vimentin) expression. In a mixed effects ANOVA, experimental groups were compared to control at each time point. * p<0.05, #p<0.01, dots represent biological replicates (n = 7 on day 2, n = 5 on day 4), error bars represent 95% CI.

FGF-2 inhibits nuclear translocation of SMAD2/3

The canonical TGF-β signalling pathway involves the phosphorylation and complex formation of SMAD2 and SMAD3, followed by the migration of this complex to the nucleus where it affects translation of TGF-β associated genes [29]. Control group hVICs showed a moderate degree of nuclear translocation of activated SMAD2/3 complex, visible as bright nuclei on immunostaining (Fig 5A). This effect was entirely absent in FIB-media cells, although staining did show presence of the complex in the cytoplasm. The TGF-β treated cells had the greatest degree of nuclear translocation, with very bright nuclei present in all cells. FGF-2 treatment did not meaningfully alter the expression of SMAD2 mRNA (Fig 5B and 5C), only showing a marginal increase in SMAD3 expression on day 4. Only TGF-β significantly influenced mRNA expression of these genes, with a 2.4-fold decrease in SMAD3 expression on day 2 (95% CI: 2.2–2.8, p<0.0001) and a 2.8-fold decrease on day 4 (95% CI: 1.7–8.9, p<0.0001). The data reaffirm the well-established canonical TGF-β signalling pathway in hVICs. FGF-2 does not, however, influence the expression of the SMAD genes.

Fig 5

FGF-2 reduces SMAD immunostaining, no effect on expression of SMAD genes.

(A) shows hVICs on day 2 stained for activated SMAD2/3 complex (green), counterstained with phalloidin (red), scale bars represent 100μm. Representative images of day 4 did not appreciably differ from day 2, and are available in the supplementary materials (S3). (B) shows RT-qPCR results of expression of SMAD2 and SMAD3 respectively at day 2 and 4. * p<0.05, #p<0.01, dots represent biological replicates (n = 7 on day 2, n = 5 on day 4), error bars represent 95% CI.

FGF-2 inhibits formation of focal adhesions

Focal adhesions are the protein complexes that join actin filaments of adjacent cells together to form a functional unit [30]. As these adhesions transmit tension forces from cell to cell, their presence is associated with the contractile nature of myofibroblasts. We used paxillin as a target for immunostaining to resolve focal adhesions in our hVIC cultures (Fig 6A). In the control group, focal adhesions appeared as brightly stained nodules along the cell margin, typically between 2 and 5μm in length. In the FGF-2 treated cells, the number of these foci was starkly reduced, and instead, paxillin staining appeared diffusely around the cytoplasm. By contrast, the TGF-β treated cells showed many mature focal adhesions formed along cell-cell junctions. We thresholded the images taken and counted the number of focal adhesions per cell (Fig 6B). There was no significant inter-time point difference. FGF-2 caused a significant decrease in the number of paxillin-positive focal adhesions, with a 2-fold decrease at day 2 and 4. Conversely, the TGF-β group showed a 2-fold increase at both time points. FGF-2 appears to inhibit formation of focal adhesions. TGF-β appears to promote their formation.

Fig 6

FGF-2 inhibits focal adhesion formation.

(A) shows paxillin staining of hVICs on day 2, scale bars represent 100μm. Representative images of day 4 did not appreciably differ from day 2, and are available in the supplementary materials (S3). (B) is the number of mature focal adhesions per cell. n = 4 biological replicates were used, 10 images per replicate per group. Experimental groups were compared to control using a 2-way ANOVA. * p<0.05, #p<0.01, error bars represent 95% CI.

FGF-2 inhibits nodular-type calcification, but did not affect osteoblastic differentiation of hVICs

Alizarin red staining was used to resolve calcification formed by hVICs on day 2 and 4 (Fig 7A). Specifically, we looked to distinguish the nodular-type calcification formed around apoptotic bodies from the diffuse calcification produced by osteoblast-like phenotype described in the literature [16, 20, 28]. All groups showed a significant increase in the number of calcific foci over the time course (Fig 7B, p<0.0001), though they did not significantly differ from each other on day 2. By day 4 the control group cells formed numerous calcific nodules. These nodules were small, measuring between 5 and 20 μm across, typical of apoptotic bodies. The FIB-media hVICs showed significantly fewer nodules compared to control (p<0.0001), with each well containing less than 20 resolvable calcifications. TGF-β treated cells on the other hand, formed considerably more nodules (p = 0.0001).

Fig 7

FGF-2 inhibits nodular calcification.

(A) shows alizarin red stained hVICs on day 4. Scale bars represent 100μm. (B) shows the number of calcific nodules per well on day 2 and day 4, counted manually from n = 4 biological repeats, with 2 wells per group. A 2-way ANOVA compared the experimental groups with control at each time point. * p<0.05, #p<0.01, error bars represent 95% CI.

Interestingly, no evidence of diffuse osteoblast-like calcification was seen. Nor was there any change in genes associated with osteoblastic differentiation (Fig 8). We performed RT-qPCR on RUNX2, the ‘master transcription factor’ of the osteoblast lineage and found no difference between groups (p = 0.9122). Levels of OPN and OPG were highly variable across replicates, and did not differ between groups (p = 0.2583 and 0.5175 for OPN and OPG respectively). Expression of BGLAP and SP7 was below detectable levels. CDH11, a cadherin isoform associated with both osteoblasts and fibroblast more generally [31, 32], did not show any changes in expression under FGF-2 or TGF-β stimulation (p = 0.3112). Only biglycan (BGN), a bone-associated glycoprotein, demonstrated a significant upregulation in the TGF-β group (p = 0.0074 on day 2, p = 0.0006 on day 4), however this increase was small in magnitude. On the whole, expression of osteoblast-associated genes appeared unaffected by the treatments.

Fig 8

FGF-2 has no effect on expression of osteoblast-associated genes.

Results of RT-qPCR on a panel of 7 osteoblast associated genes. Note that expression of BGLAP and SP7 was below the detectable level. * p<0.05, #p<0.01, dots represent biological replicates (n = 7 on day 2, n = 5 on day 4), error bars represent 95% CI.

FGF-2 inhibits expression of matrix synthesis, in opposition to TGF-β

Six matrix associated genes were assayed in this study at day 2 and day 4 (Fig 9). None of these genes showed a significant time-dependent change in expression. The overall effect of FGF-2 tended towards downregulation of structural matrix proteins and upregulation of proteinases, indicating a net matrix ‘breakdown’ effect. COL1A1 encodes the major precursor to collagen 1, and FGF-2 caused a marginal but statistically significant downregulation of this gene on day 2. A striking difference was noted in the expression of elastin, where FGF-2 caused a 5.0-fold decrease in ELN mRNA (95% CI: 4.2–6.2) on day 2. MMP1, which codes for matrix metalloproteinase 1, an enzyme that digests collagen, was significantly upregulated at both timepoints, with a 9.8-fold and 7.0-fold increase on days 2 and 4 respectively. Changes to MMP2 (type IV collagenase) and TIMP1 (inhibitor of MMPs) did show some significant changes under FGF-2 stimulation, but these were marginal in magnitude. TGF-β treated hVICs in most instances demonstrated opposite effects on expression of matrix associated genes, i.e. the there was a tendency towards matrix synthesis. COL1A1 and COL3A1 expression was significantly increased on both days 2 and 4. Notably, levels of ELN mRNA were increased 3.7-fold on day 2 (95% CI: 2.5–3.7) and 5.0-fold on day 4 (95% CI: 3.3–6.7). One result that departed from the general theme of TGF-β stimulated matrix synthesis was the 2-fold increase in MMP2 in the TGF-β group at both day 2 and 4. Overall, the RT-qPCR results show that FGF-2 stimulates a movement towards a matrix breakdown gene expression profile, while TGF-β stimulates a movement towards an increase in matrix associate genes.

Fig 9

FGF-2 promotes matrix breakdown expression pattern.

RT-qPCR results on a panel of matrix associated genes. * p<0.05, #p<0.01, dots represent biological replicates (n = 7 on day 2, n = 5 on day 4), error bars represent 95% CI.

FGF-2 promotes expression of bone morphogenic proteins

Bone morphogenic proteins (BMPs) are members of the TGF-β family, influencing bone development. FGF-2 promoted an upregulation of BMP2, with a 5.0-fold increase on day 2 (95% CI: 2.1–8.0) and a 4.4-fold increase on day 4 (95% CI: 3.3–5.6) (Fig 10). There was a small but statistically significant increase in BMP4 expression by day 4. TGF-β had the opposite effect: BMP2 was down regulated 2.7-fold on day 4 (95% CI: 1.5–13.1); and BMP4 showed significant downregulation, with a 2.8-fold decrease on day 2 (95% CI: 2.0–4.6) and a 3.7-fold decrease on day 4 (95% CI: 2.5–6.8). Expression of BMP6 was low across all groups, and demonstrated a high degree of variability between repeats. Despite this, TGF-β did induce a 2.1-fold upregulation of BMP6 on day 2 (95% CI: 1.7–2.4).

Fig 10

FGF-2 upregulates BMP2 expression.

RT-qPCR results of BMPs 2, 4, and 6. In a mixed effects model, experimental groups were compared to control at each time point. * p<0.05, #p<0.01, dots represent biological replicates (n = 7 on day 2, n = 5 on day 4), error bars represent 95% CI.

Scleraxis: A potential marker of hVIC myofibroblasts?

Scleraxis is a transcription factor typically associated with tendon development [33] and cardiac myofibroblasts [34]. We assayed the expression of SCX as part of our gene expression panel and found a marked degree of change in the experimental groups (Fig 11B). FGF-2 caused a 2.8- and 2.3-fold decrease on days 2 and 4 respectively. Conversely, TGF-β caused a 6.9-fold and 8.9-fold upregulation on day 2 and 4 respectively. Following this result, we used immuno-staining to identify presence of scleraxis in the hVICs. All experimental groups showed presence of scleraxis, localised to the nucleus (Fig 11A). However, there were no observed differences between the groups.

Fig 11

Scleraxis, potential myofibroblast marker, is upregulated by TGF-β.

(A) shows representative images of hVICs stained for scleraxis (green) and phalloidin (red) on day 4. Representative images of day 2 did not appreciably differ from day 4, and are available in the supplementary materials (S3). Scale bars represent 100μm. (B) shows RT-qPCR results of SCX (scleraxis) gene expression. In a mixed effects model, experimental groups were compared to control at each time point. * p<0.05, #p<0.01, dots represent biological replicates (n = 7 on day 2, n = 5 on day 4), error bars represent 95% CI.

Discussion

Our study found that FGF-2 had an anti-myofibroblast effect on hVICs isolated from patients with end-stage CAVD. Myofibroblasts are central to wound healing and growth of tissues [19]. However, excessive activation of VICs to myofibroblasts is associated with valve sclerosis and eventual calcification [16, 21]. In a study on healthy hVICs, Latif and colleagues described the use of FGF-2 to inhibit the myofibroblastic activation of VICs using a ‘FIB’ media formulation that contained 10ng/mL FGF-2 and a reduced serum content to lower endogenous TGF-β [24]. This study aimed to build on this work by demonstrating the inhibitory effect of FGF-2 on the myofibroblast activation of hVICs and formation of calcific nodules, with a focus on changes in gene expression. Latif et al’s work showed that FGF-2 could retain hVICs in quiescence, but is not known if FGF-2 could reverse myofibroblastic activation in cells that were already activated. Unique to this study was the use of hVICs isolated from patients with end-stage CAVD; we wanted to determine if the therapeutic effect seen in healthy hVICs would remain intact in a population of diseased cells with a greater degree of myofibroblastic activation.

To examine the pro- or anti-myofibroblast effect of FGF-2, it was first necessary to define exactly what constitutes a ‘myofibroblast’. Previous work in this area has used presence of α-SMA as a marker [17, 22, 24, 28], and while this is a well-validated approach, α-SMA is also present to a lesser degree in healthy valve tissue, particularly during growth and development [11]. Other researchers have used ‘increased contractility’ [22, 24], collagen deposition by hydroxyproline assay [22], desmin [17], or the activation of known TGF-β pathways as endpoints [22, 35]. In our study, we used TGF-β stimulation as a positive control to build up a comprehensive picture of what a hVIC-derived myofibroblast looks like. We found that TGF-β treated VICs take on a distinct morphology: when cultured on plastic cells become very large and rounded, and lay down mature bands of α-SMA that terminate in focal adhesions. Focal adhesion formation was also increased by TGF-β. The gene expression profile of TGF-β treated hVICs was pro-matrix forming, with significant upregulations in genes coding collagen 1, collagen 3, and elastin. This effect has been demonstrated extensively in the literature, where TGF-β stimulation consistently upregulates matrix gene expression and deposition of VICs [35, 36]. MMP2, a matrix protease that breakdown collagen IV-rich basement membranes, was also upregulated. This has been shown in other cell types [37], and indeed is a mechanism behind cell invasion through basement membranes in tumour metastasis [38, 39].

Across most of our outcomes, FGF-2 induced an opposite effect to that caused by TGF-β stimulation. FGF-2 treated cells were more typical of quiescent fibroblasts in shape: they were of significantly smaller area and demonstrated a ‘spindle-like’ morphology. These findings were in line with Latif and colleagues [24]. A primary endpoint for aVICs, α-SMA expression, was reduced in the FGF-2 group, and enhanced in the TGF-β group. The gene expression profile showed reduced levels of matrix-forming proteins, and a large increase in MMP1, indicating a net degradative effect. This effect is seen in fibroblasts elsewhere: Kashpur and colleagues found similar reductions in matrix associated mRNAs coupled with upregulation of MMP1 in dermal fibroblasts treated with FGF-2 [40]. How far these conclusions can translate to an in vivo setting remains unclear. In a gene expression analysis of normal vs stenotic human valves, Bossé and colleagues found that stenotic valves (i.e. valves with a high degree of myofibroblasts) showed a general upregulation in MMPs; MMP1 and MMP12 in particular [7].

Our findings regarding hVIC calcification are twofold: 1) FGF-2 inhibits the nodular-type calcification, and 2) no evidence of diffuse, osteoblast-like calcification occurred in our cultures. The TGF-β stimulated formation of nodules in porcine cell culture has been extensively described in the literature [22, 28, 41, 42], and mirrors what we saw with TGF-β treatment of our hVICs. Both Cushing et al. and Gao et al. describe a similar reduction in calcification under FGF-2 stimulation [22, 25]. However, an interesting departure from the literature is the apparent absence of change in cadherin 11 gene expression in our study. Cadherin-11 is described by Hutcheson and colleagues as necessary in the formation of ‘dystrophic’ nodules in porcine cells [28]–an effect absent here. It is widely accepted that TGF-β treatment does induce osteoblastic differentiation of VICs, including the upregulation of osteoblast associated genes [16, 43, 44]. Our study did not reveal marked changes in osteoblast related genes upon FGF-2 or TGF-β stimulation, nor did we find evidence of ossification typical of osteoblast-like cells. Gao and colleagues did find FGF-2 mediated downregulations of osteogenic genes RUNX2, OPN, and SP7, though they used osteogenic medium [25]. The only finding suggestive of osteoblast differentiation was TGF-β stimulated upregulation of biglycan, an effect described by Song and colleagues as a crucial part of ‘pre-osteogenic programming’ in hVICs [45]. It is well established that hVICs are a heterogenous cell population, with various authors recognising up to five distinct phenotypes [16, 21]. It is unclear from this literature if all VICs undergo osteoblast differentiation, or if only certain subtypes are susceptible. It is unclear from our study if our culture conditions could induce osteogenesis in hVICs, or if the cells we isolated from diseased valves were somehow less responsive to osteoblastic differentiation than porcine or healthy human cells used by other researchers. The osteoblastic VICs appear to arise from quiescent VICs independent of the myofibroblastic lineage [46], and it is possible that the cells used here were committed myofibroblasts. FGF-2 stimulation upregulated the expression of BMPs 2 and 4, while TGF-β treatment caused a downregulation of the same. This effect has been described [44], and may be ascribed to the negative feedback TGF-β exerts on expression of TGF-β family of signallers, of which the BMPs are members [47]. Our finding is in opposition to those of Gao and colleagues, who describe the down-regulation of BMP2 in response to FGF-2 stimulation via the Notch1 pathway, though the FGF-2 in this case was added to osteogenic medium [25]. The disparate findings may represent the effect of FGF-2 on the two distinct modes of VIC calcification (nodular and osteoblastic), and Gao et al.’s downregulation of osteoblast genes and calcification occurred via Notch1-mediated obVIC inhibition—an effect not seen in our study.

We found FGF-2 to slightly improve viability on alamarBlue assay. Both the absolute absorbance values, and the rate of increase of said values were raised in the FIB group, and did not differ from control in the TGF-β group. Other researchers have found similar [25] or the opposing effects [24]. Our study has used insulin as a mitotic agent in the FIB medium, and this addition may explain our findings here. Lam and colleagues, who used the same insulin concentration in their FGF-2 supplemented media, also found increases in proliferation [48].

The canonical TGF-β signalling pathway was intact in our hVIC culture, evidenced by the localisation of the activated SMAD2/3 complex to the nucleus upon immunostaining. We found that FIB media inhibited this localisation, as much less nuclear staining was seen compared to control, in line with previous studies on porcine cells [22, 23]. It is unclear if the low SMAD2/3 localisation was due to the FGF-2 alone or the low endogenous TGF-β in the FIB media, though Cushing et al.’s findings in porcine VICs support FGF-2’s inhibitory action [22].

Scleraxis emerged as a potential marker of myofibroblastic VICs. Scleraxis is known to be a critical factor in conversion of cardiac fibroblasts to myofibroblasts via the canonical TGF-β signalling pathway [34, 49], but it is unclear from this study if this upregulation represents a true in vivo effect in valve cells specifically. Scleraxis is traditionally thought of as a tendon marker [33], but embryological studies in mice show scleraxis as a key transcription factor for heart valve development [50], and its upregulation has been ascribed to canonical TGF-β2 signalling [51].

In all, FGF-2 promoted a quiescent-like phenotype in hVICs isolated from calcific valves, while TGF-β promoted a myofibroblastic ‘aVIC’ phenotype. Our findings are summarised in Fig 12.

Fig 12

Summary of findings.

Conclusion

Our findings suggest that FGF-2 opposes the formation of aVICs in an in vitro setting. We assayed cells from patients with end stage valve disease, i.e. cells with a high degree of inherent activation, and found that outcomes associated with myofibroblastic activation were returned to ‘quiescent-type’ levels by FGF-2. We used the well-validated aVIC marker, α-SMA, as well as a panel of gene expression analyses to build a comprehensive picture of ‘myofibroblastic activation’ and noticed that FGF-2 pushed cells in the opposite direction. Pro-quiescent observations include the downregulation of ECM production and inhibition of calcific nodule formation. FGF-2 showed an in vitro therapeutic effect, returning cells from an apparent diseased state to a quiescent state.

List of probes assayed in RT-qPCR.

(DOCX)

Click here for additional data file.

Measures of RNA purity and integrity.

(DOCX)

Click here for additional data file.

Immunocytochemistry from alternate timepoints (day 4 SMAD2/3, day 4 paxillin, and day 2 scleraxis) A) shows hVICs on day 4 stained for activated SMAD2/3 complex (green), counterstained with phalloidin (red), B) shows hVICs on day 4 stained for activated paxillin (green), C) shows hVICs on day 2 stained for activated scleraxis (green), counterstained with phalloidin (red). Scale bars represent 100μm.

(TIF)

Click here for additional data file.

21 Mar 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Katherine Yutzey, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_auSimilarity Check/iThenticate Results (28%)thors_affiliations.pdf.

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

3. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

4. Please ensure that you refer to Figure 10 in your text as, if accepted, production will need this reference to link the reader to the figure.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This manuscript compares VIC phenotype upon treatment with FGF2 or TGFb1, with a specific focus on myofibroblastic and osteogenic markers. This is possibly the most clearly written paper that this reviewer has seen in a long time; it was a pleasant read. That said, this reviewer does have several concerns:

1) Some of the findings in this paper have been previously published by others – e.g., VIC morphology measures with FGF2 or TGFb1 treatment, aSMA upon FGF2 or TGFb1 treatment. Several of these previous works are indeed referenced by the authors, but many of the authors’ findings – which mirror those previous studies – do seem a bit redundant in the context of already-published work. Is the unique contribution of the current study that the VICs were from diseased valves? If so, that aspect should be better emphasized to differentiate this study from earlier ones. If not, it would be useful to perhaps place some of the ‘repeat’ data in the supplement and focus the main paper on measures that have not been previously reported.

2) The authors state that “FGF-2 interferes with the nuclear translocation of the [Smad 2/3] complex” because translocation was lower in the FIB condition than in the Control. However, as noted by the authors, the Control condition (with 10% FBS) already has more TGFb1 in it than the FIB condition. So, it seems likely that Smad 2/3 translocation may be lower in FIB media simply because there is less TGFb1 present than in the Control media? In other words, less Smad 2/3 translocation can just indicate less TGFb1 to stimulate it, not an active inhibitory role of FGF2.

3) The authors used hVICs from 40 different donors. Were those cell populations pooled together to perform the experiments? Or were different experiments performed with different donors? Additional information is needed on how many donors are represented in each experiment.

4) There are many places where it sounds like the authors have treated with FGF2 and then treated those cultures with TGFb1 (which, as an aside, would be an interesting future experiment). For example, the abstract indicates treatment with “FGF2 and TGFb1”, and there are multiple variations of the conclusion that “FGF-2 inhibits the nodular-type calcification that results from TGF-β stimulation”. Other statements such as “FGF-2 reduced the calcification potential of VICs” (in the abstract) also made this reviewer think that this paper was going to be about an active calcification stimulus being added after FGF2.

5) Another VIC/FGF2 study just came out last month (PMID: 35074856), likely after the authors had already finalized this manuscript. This work is distinct, but should be cited/discussed as appropriate in the current manuscript.

6) A table containing patient demographics (age, sex, other relevant info if available) should be included.

Reviewer #2: In this study human aortic valve interstitial cells isolated from diseased valves were used for in vitro studies on how FGF-2 affects cell progression toward disease. The authors found that FGF-2 added to culture medium inhibited the activation of interstitial cells to myofibroblasts and decreased the formation of calcified nodules. Several comments/questions on the manuscript are below:

1) How were the 2 and 4 days time points chosen? It is not clear why there are two different time points, and in some figures it is not clear if the image analysis is for day 2 or day 4 images (i.e. Fig 2, Fig 3). Why are there no day 4 images in Fig 5 and Fig 6?

2) In Fig 11 the control image seems to be at a different magnification.

3) Figure legends can be improved. For all legends there is no mention of the number of replicates (n = XX). In image analysis graphs there is no mention of the number of images measured or cells/image measured. Please add this information. In the calcific nodule counting methods there is no mention of the number of replicates or number of wells analyzed.

4) Statistical analysis and annotation on figures is not clear. I believe a two way anova was performed, so in Fig 2, for example, are annotations denoting statistical significance relative to control or other treatments? If there is statistical significance between different treatments that should be indicated on the figures. I have similar questions about the other figures as well, statistical significance between treatments should be tested.

5) Page 16: This, coupled with the apparent absence of diffuse calcification suggests that FGF-2 inhibits calcification of hVICs with a mechanism independent of osteoblast differentiation.

I find this statement confusing and a bit broad. These cells were isolated from calcified valves that had already moved past osteoblastic differentiation and into large calcified nodules. It seems that pro-osteoblastic RUNX2 gene expression would normally happen earlier in the disease process. End stage valve disease cells may be less "plastic" and responsive to signaling. Could this statement be softened with some qualifiers about the initial state of the cells that were isolated, or can the authors add this point to the discussion?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Gretchen Mahler

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

26 May 2022

Response to reviewers

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_auSimilarity Check/iThenticate Results (28%)thors_affiliations.pdf.

We are eager to adhere to the style guide, and we think we have done this. If there are any areas where we have failed, please let us know

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

Our cover letter now includes:

“We were granted ethical approval to collect human tissue from patients at Auckland City Hospital. Māori, the indigenous population of New Zealand, view data as the ‘shadow’ of their tissue, and as such, data availability and data governance are very tightly controlled. Our ethical approval was granted by the Auckland Health Research Ethics Committee (AHREC), who have stipulated that all raw data must be stored securely on the premises of the University of Auckland, under the guardianship of either Marcus Ground or Professor Jillian Cornish. The key restriction that AHREC imposes on us is that the data must be housed in Auckland under our care. We are able to, however, share this data with other researchers upon any reasonable request. Data requests can be made to Marcus Ground (groma788@student.otago.ac.nz), Professor Jillian Cornish (j.cornish@auckland.ac.nz) or to AHREC directly at humanethics@auckland.ac.nz”

We hope you understand that the treatment of data as both personal and sacred by Maori in New Zealand means that we have to be ‘guardians’ of the data itself. This means we must keep it securely on our password-protected servers or locked in cabinets at the University of Auckland, per our AHREC approval. We are able to share data upon reasonable request.

3. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

We have removed the reference to data not presented here. We have instead added both time points to figure 5

4. Please ensure that you refer to Figure 10 in your text as, if accepted, production will need this reference to link the reader to the figure.

We have now included reference to figure 10

Reviewer #1: This manuscript compares VIC phenotype upon treatment with FGF2 or TGFb1, with a specific focus on myofibroblastic and osteogenic markers. This is possibly the most clearly written paper that this reviewer has seen in a long time; it was a pleasant read. That said, this reviewer does have several concerns:

1) Some of the findings in this paper have been previously published by others – e.g., VIC morphology measures with FGF2 or TGFb1 treatment, aSMA upon FGF2 or TGFb1 treatment. Several of these previous works are indeed referenced by the authors, but many of the authors’ findings – which mirror those previous studies – do seem a bit redundant in the context of already-published work. Is the unique contribution of the current study that the VICs were from diseased valves? If so, that aspect should be better emphasized to differentiate this study from earlier ones. If not, it would be useful to perhaps place some of the ‘repeat’ data in the supplement and focus the main paper on measures that have not been previously reported.

We acknowledge that we have used many of the same methodologies as previous researchers. However, to our knowledge, this is the first use of FGF-2 on cells explicitly from diseased valves, and is also the first that demonstrates ‘reversal’ of already activated cells. The cell population used by Latif et al. were from healthy donors, and so were mostly quiescent. Whereas the cell population used here is already highly myofibroblastic (We refer you the number of red-stained cells in the day 2 control group of Figure 3A). In any case, we have made the following changes to address your concern:

• The introduction has been amended to highlight the fact that this study is the first to expand previous FGF-2 findings onto diseased cells

• The first paragraph of the discussion has been amended to highlight what makes this study different

2) The authors state that “FGF-2 interferes with the nuclear translocation of the [Smad 2/3] complex” because translocation was lower in the FIB condition than in the Control. However, as noted by the authors, the Control condition (with 10% FBS) already has more TGFb1 in it than the FIB condition. So, it seems likely that Smad 2/3 translocation may be lower in FIB media simply because there is less TGFb1 present than in the Control media? In other words, less Smad 2/3 translocation can just indicate less TGFb1 to stimulate it, not an active inhibitory role of FGF2.

We agree we have overreached with this conclusion. The SMAD2/3 section of the results has been amended to remove suggestion that FGF-2 treatment alone caused this change, and have added a small sentence in the SMAD2/3 paragraph of the discussion. We direct you to Cushing et al. 2008 who showed thar FGF-2 does inhibit SMAD2/3 signalling in healthy pig cells: (doi: 10.1096/fj.07-087627)

3) The authors used hVICs from 40 different donors. Were those cell populations pooled together to perform the experiments? Or were different experiments performed with different donors? Additional information is needed on how many donors are represented in each experiment.

We did not pool donors. The ‘experimental conditions’ section of the methods has been amended to reflect this.

4) There are many places where it sounds like the authors have treated with FGF2 and then treated those cultures with TGFb1 (which, as an aside, would be an interesting future experiment). For example, the abstract indicates treatment with “FGF2 and TGFb1”, and there are multiple variations of the conclusion that “FGF-2 inhibits the nodular-type calcification that results from TGF-β stimulation”. Other statements such as “FGF-2 reduced the calcification potential of VICs” (in the abstract) also made this reviewer think that this paper was going to be about an active calcification stimulus being added after FGF2.

We agree that in several instances we have incorrectly made it sounds like FGF-2 and TGF-β treatments are conflated. We have amended the following points:

• The final sentence of the ‘FGF-2 inhibits formation of focal adhesions’ section of the results has been tweaked to remove the notion that somehow the FGF-2 worked against TGF-β in the same experiment

• The final sentence of the ‘FGF-2 inhibits expression of matrix synthesis, in opposition to TGF-β’ section of the results has been tweaked to remove the notion that somehow the FGF-2 worked against TGF-β in the same experiment

• The sentence in the conclusion that you mentioned was misleading has been amended to reflect the FGF-2 mediated changes only. And the TGF-β related changes have been expanded in the following sentence

• The sentence referring to the up/down regulation of BMPs has been amended to separate out the FGF-2 and TGF-β arms of the study

• The sentence about FGF-2 and SMAD2/3 staining: the word ‘interfered’ was changed to ‘inhibited’ to remove the idea that somehow the two treatments are happening at once

• The conclusion has been amended to remove the notion that FGF-2 inhibited TGF-β effects on cells

That being said, our claims about FGF-2’s effects (inhibition of calcification etc.) are made when we compared FGF-2 to control (no treatment). I.e. if we removed the TGF-β group from the analysis entirely, then we would still see the reduction of calcification on alizarin red staining, etc. FGF-2 did in fact cause a >two thirds reduction in nodular calcifications when compared to control on day 4.

5) Another VIC/FGF2 study just came out last month (PMID: 35074856), likely after the authors had already finalized this manuscript. This work is distinct, but should be cited/discussed as appropriate in the current manuscript.

We thank you for directing us to this paper - indeed we had mostly finished our manuscript. We absolutely agree that it should be discussed. To achieve this, we have:

- Made brief mention of it in the introduction

- Made brief mention of the anti-calcific effects in the discussion

- Added Gao’s BMP2 findings to the BMP2 section of the discussion. We have highlighted Gao’s use of osteogenic media, and how the disparate findings may be due to this

- Added the proliferation findings to the alamarBlue section of the discussion

6) A table containing patient demographics (age, sex, other relevant info if available) should be included.

Due to the nature of our ethics approval we aren’t able to share demographic information about our tissue donors. Our consenting nurses did record this information, but it is not available to researchers. All we know is that the donors were undergoing valve replacement surgery.

Reviewer #2: In this study human aortic valve interstitial cells isolated from diseased valves were used for in vitro studies on how FGF-2 affects cell progression toward disease. The authors found that FGF-2 added to culture medium inhibited the activation of interstitial cells to myofibroblasts and decreased the formation of calcified nodules. Several comments/questions on the manuscript are below:

1) How were the 2 and 4 days time points chosen? It is not clear why there are two different time points, and in some figures it is not clear if the image analysis is for day 2 or day 4 images (i.e. Fig 2, Fig 3). Why are there no day 4 images in Fig 5 and Fig 6?

With respect to our myofibroblast endpoints (α-SMA, cell morphology, etc.), we wanted to use two timepoints to determine if there was some directionality in the changes we saw (i.e. was the FGF-2 effect transient? Was is static, but sustained? Did it increase over the time course?).

We chose to use day 2 and day 4 as timepoints as this, in our experience, is the time course in which hVICs undergo visible changes in their phenotype, but importantly, occurs prior to the cells become confluent (which generally occurs by day 6). Using low seeding densities (<3000 cells/cm2) to extend the growth phase, in our experience, can result in non-proliferating cultures.

We also drew on the literature. Other researchers have used VICs over a similar time course:

• Cushing et al. (doi: 10.1096/fj.07-087627) examined FGF-2 stimulation of porcine VICs over 2 days

• Monzack et al. (PMID: 21863660) examined osteoblast potential of VICs over 5 days

• Walker et al. (PMID: 15217906) treated VICs with TGF-β over the course of 2 days

• Gao et al. (PMID 35074856) used VICs over the course of 2 days

We agree there is inconsistency in the presentation of time points. To address your concerns:

• Figure 3B-E (the morphology analysis) is the result of pooling the day 2 and day 4 timepoints. We have updated the caption to reflect this. We pooled these data because in all morphology metrics, the 2-way ANOVA showed no significant differences between the time points, and we didn’t want to overload the graph with information.

• Figure 5 (Smads). We have now included both rtPCR time points on the graph. We did not see appreciable differences in staining so we only wanted to present one set of images. Day 4 staining is now included in supplementary figure 3.

• Figure 6 (paxillin). We did not see appreciable differences in the day 2 and day 4 paxillin staining and so we wanted to only present one set of images. We have now included day 4 staining in the supplementary figure 3

• Figure 11 (scx). We chose to stain for scleraxis because we saw gene expression changes at both time points. We did not see appreciable differences in the day 2 and day 4 scleraxis staining and so we wanted to only present one set of images. We have now included day 2 staining in the supplementary figure 3

2) In Fig 11 the control image seems to be at a different magnification.

We agree it looks fishy, but we promise the magnifications are the same. In our experience, there is a great degree of pleomorphism with the hVICs - particularly those in the control or TGF-β groups. For reference, have a look at these two images taken at the same magnification:

3) Figure legends can be improved. For all legends there is no mention of the number of replicates (n = XX). In image analysis graphs there is no mention of the number of images measured or cells/image measured. Please add this information. In the calcific nodule counting methods there is no mention of the number of replicates or number of wells analyzed.

We agree with your concerns about the lack of ‘N’s. To make this more transparent, we’ve done the following:

• Sample sizes for the various image analyses has been added to the ‘immunocytochemistry’ and the ‘alizarin red’ sections of the methods

• Captions from figures 2-11 have been amended to include replicates and well numbers where relevant

4) Statistical analysis and annotation on figures is not clear. I believe a two way anova was performed, so in Fig 2, for example, are annotations denoting statistical significance relative to control or other treatments? If there is statistical significance between different treatments that should be indicated on the figures. I have similar questions about the other figures as well, statistical significance between treatments should be tested.

We agree that our reporting of stats is lacking. To make our statistics more transparent, we have made the following changes:

• More explicit reference to the 2-way ANOVA has been added to the ‘statistics’ section. Specifically we have described explicitly what ‘image analysis’ is, and what values went into the gene expression 2-way ANOVAs

• We did not seek to measure differences between the two experimental groups - only the difference between each experimental group and the control group. We have explicitly stated this in the ‘statistics’ section

• The ‘alamarBlue’ section of the methods has been altered to make the comparisons more explicit.

• Captions have been amended where relevant

5) Page 16: This, coupled with the apparent absence of diffuse calcification suggests that FGF-2 inhibits calcification of hVICs with a mechanism independent of osteoblast differentiation.

I find this statement confusing and a bit broad. These cells were isolated from calcified valves that had already moved past osteoblastic differentiation and into large calcified nodules. It seems that pro-osteoblastic RUNX2 gene expression would normally happen earlier in the disease process. End stage valve disease cells may be less "plastic" and responsive to signaling. Could this statement be softened with some qualifiers about the initial state of the cells that were isolated, or can the authors add this point to the discussion?

We agree that this statement was over-reaching. It has been removed from the ‘results’ section. To provide further clarification, we have amended the fourth paragraph of our discussion to clarify that our diseased cells may have reduced ability to undergo osteoblastic differentiation. We have included reference to Monzack 2011 (PMID 21863660), who describes how obVICs may arise from qVICs independent of the myofibroblast lineage

Submitted filename: Response to reviewers.docx

Click here for additional data file.

7 Jun 2022

Fibroblast growth factor 2 inhibits myofibroblastic activation of valvular interstitial cells

PONE-D-22-04010R1

Dear Dr. Ground,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Katherine Yutzey, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: My comments have been addressed, but it may still be helpful to better annotate statistical significance on the figures (i.e. add a bracket between statistically significant treatments)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Gretchen Mahler

**********

9 Jun 2022

PONE-D-22-04010R1

Fibroblast growth factor 2 inhibits myofibroblastic activation of valvular interstitial cells

Dear Dr. Ground:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr Katherine Yutzey

Academic Editor

PLOS ONE