Tropomyosin 2.1 collaborates with fibronectin to promote TGF-β1-induced contraction of human lung fibroblasts

Many lung diseases are characterized by fibrosis, leading to impaired tissue patency and reduced lung function. Development of fibrotic tissue depends on two-way interaction between the cells and the extra-cellular matrix (ECM). Concentration-dependent increased stiffening of the ECM is sensed by the cells, which in turn increases intracellular contraction and pulling on the matrix causing matrix reorganization and further stiffening. It is generally accepted that the inflammatory cytokine growth factor β1 (TGF-β1) is a major driver of lung fibrosis through the stimulation of ECM production. However, TGF-β1 also regulates the expression of members of the tropomyosin (Tm) family of actin associating proteins that mediate ECM reorganization through intracellular-generated forces. Thus, TGF-β1 may mediate the bi-directional signaling between cells and the ECM that promotes tissue fibrosis. Using combinations of cytokine stimulation, mRNA, protein profiling and cellular contractility assays with human lung fibroblasts, we show that concomitant induction of key Tm isoforms and ECM by TGF-β1, significantly accelerates fibrotic phenotypes. Knocking down Tpm2.1 reduces fibroblast-mediated collagen gel contraction. Collectively, the data suggest combined ECM secretion and actin cytoskeleton contractility primes the tissue for enhanced fibrosis. Our study suggests that Tms are at the nexus of inflammation and tissue stiffening. Small molecules targeting specific Tm isoforms have recently been designed; thus targeting Tpm2.1 may represent a novel therapeutic target in lung fibrosis.


Introduction
In the western world, fibrosis is a major etiological factor in an estimated 45% of deaths [1] and interstitial lung diseases typified by fibrosis, including idiopathic pulmonary fibrosis (IPF), are significant contributors to global disease burden [2]. Increased extracellular matrix (ECM) secretion and deposition by tissue-resident fibroblasts (and other cells) increases tissue stiffness that characterizes fibrosis in IPF [3]. Moreover, fibroblasts pull on the ECM via forces exerted through the cellular actin cytoskeleton and this too can lead to increased tissue stiffening [4]. In previous studies, the inflammatory mediator TGF-β 1 has been shown to induce ECM deposition [5,6]. In separate studies, TGF-β 1 has been shown to increase expression of the actin-associating tropomyosin (Tms) that are required for cells to exert force on the external environment [7]. We sought to determine whether TGF-β 1 induces concomitant ECM deposition and Tms in human lung fibroblasts, thereby generating a feed-back loop that promotes tissue fibrosis.
Tissue desmoplasia (growth of fibrous or connective tissue) is initiated through large-scale ECM structural deregulation causing increased ligand density, fibril attachment and crosslinking, sustained by unregulated production of growth factors, cytokines and secretion of atypical matrix components. Damaged epithelial, endothelial and smooth muscle cells (and other structural cells) release cascades of inflammatory mediators. These mediators promote increased platelet formation, vasodilation, the stimulation of collagen-secreting fibroblasts and the secretion of matrix-metalloproteinases that degrade and reorganize the ECM. Cells sense the external tissue mechanical features by exerting intracellular-derived forces on the matrix generated by myosin-motor mediated contraction of bundles F-actin (stress fibres) [8]. The stress fibers are coupled via adaptor proteins to the cytoplasmic tail of ECM-bound integrins and 'pull' on ECM fibers on the external cell surface through the integrins [9]. The magnitude of the force required to deform the ECM by the cell is therefore related to the ECM stiffness and, in this manner, cells probe and respond to the mechanical forces in local tissue. A key aspect of the fibrotic circuit is the ability of the cells to sense forces in the external environment and to then contribute to the stiffening of the external tissue by relaying corresponding force onto the ECM [4]. Tms decorate the actin filaments and the specific isoform regulates the contractile function of the associated actin filament [10]. The Tm isoform Tpm2.1 is an example of a Tm that stabilizes actin stress fibres and thereby increases the contractile force derived through the stress fibres and exerted onto the surrounding tissue. Previous studies have shown that TGF-β 1 induces Tpm2.1 [11], but this has not previously been reported in human lung fibroblasts. In this study, we focus on Tpm2.1 (an isoform coded from the TM1 gene) along with other key Tm isoforms associated with actin stabilization coded from the TM1 (Tpm1.6, Tpm1.7) and TM3 (Tpm3.1) gene.
IPF is characterized by excessive ECM deposition causing the lung parenchyma to stiffen, resulting in the progressive loss of lung function. While anti-fibrotic therapeutics, such as nintedanib and pirfenidone, have shown promise [12], a greater understanding of the underlying molecular mechanisms that underpin fibrotic progression may yield novel therapeutic targets. Despite a recognition that fibroblasts play a key role in pathogenesis and tissue dysfunction in pulmonary fibrosis, most studies to date have focused on the molecules and pathways that regulate ECM production, and in particular those induced by the pro-fibrotic inflammatory mediator, transforming growth factor β 1 (TGF-β 1 ). TGF-β 1 is clearly a therapeutic target in IPF and other fibrotic conditions. However, treatment approaches targeting TGFβ 1 signaling has proved challenging due to its integral role in health. These obstacles may prove to be overcome by innovative dosing regimens or other interventions [13].
Notably, a recent study revealed that in response to TGF-β 1 , lung fibroblasts showed greater cytoskeletal reorganization and importantly lung fibroblasts from IPF patients were stiffer when compared to control [14], suggesting that increased cellular stiffness may contribute to IPF pathology. Similar studies by Sarna et al. demonstrated change in stiffness of cells derived from people with asthma [15], underscoring the importance of fibrosis in respiratory diseases more broadly. Thus, in this study we test whether, in addition to its well-established ability to induce expression of the ECM fibronectin [5,6], TGFβ 1 , also increases tropomyosin expression in human lung fibroblasts. We profile tropomyosin isoform and ECM production following TGF-β 1 treatment and measure the functional outcomes by collagen gel contraction.

Cell culture
Human lung fibroblasts were isolated from the parenchyma of lungs obtained from patients undergoing lung transplantation or surgical resection for thoracic malignancies, in accordance with procedures approved by the Sydney South West Area Health Service. Tables 1 and  2 show patient characteristics. Human lung fibroblasts were dissected and purified as previously described [16,17]. Primary human fibroblasts were cultured in Dulbecco's Modified Eagles Medium (DMEM) supplemented with 5% heat-inactivated fetal bovine serum, 0.5 mM L-glutamine, 20 mM HEPES and 1000 units/mL of penicillin, 1000 µg/mL of streptomycin, and 2.5 µg/ mL of amphotericin B. All cells were seeded at 1.3 × 10 4 cells/cm 2 , cultured for 48 h and underwent quiescence (DMEM supplemented with 0.1% bovine serum albumin, 0.5 mM L-glutamine, 20 mM HEPES and 1,000 units/mL of penicillin, 1,000 µg/mL of streptomycin, and 2.5 µg/ mL of amphotericin B) for 24 h prior to experimentation. All cell cultures tested negative for mycoplasma prior to experimentation and only cell cultures at less than 6 passages were used.

Collagen gel contraction assay
Collagen gel contraction assays were performed by adapting previous publications [19,20]. Briefly, 1 × 10 5 human lung fibroblasts in media (as above) were seeded within 1.3 mg/mL collagen (rat tail Type I) gels. Cells were incubated for 48 h at 37 °C. The cells then underwent quiescence for 24 h at 37 °C and the gels released from the wells by running a sterile yellow tip around the side of the well. Gels were left untreated (control) or stimulated with TGF-β 1 (2 ng/mL) and images taken at 0, 24, 48, 72 and 96 h using a Kodak Imager. Triplicate technical replicates were performed and collagen gel area at each time point measured by ImageJ [18].

Statistical analysis
Statistical analysis was performed with GraphPad Prism 8 using Student's unpaired t test and two-way ANOVA then Bonferroni's multiple comparison test or multiple linear regression (least squares). P values < 0.05 were sufficient to reject the null hypothesis for all analyses. A 3D scatter plot was generated in R.

TGF-β 1 increases fibronectin mRNA expression and protein expression
While it is well established that the cytokine TGF-β 1 increases fibronectin expression in human lung fibroblasts [5,6], our first aim was to determine the temporal kinetics of TGF-β 1 -induced fibronectin mRNA expression in human lung fibroblasts. These are important foundational experiments to establish the conditions for further exploration. Cells were either untreated (vehicle control) or stimulated with TGF-β 1 (2 ng/mL) for 0, 1, 4, 8, 24, and 48 h and fibronectin mRNA expression was quantified by real-time RT-PCR. No significant increase in fibronectin mRNA was detected within the first 8 h of TGF-β 1 stimulation (Fig. 1). However, following TGF-β 1 stimulation for 24 h a significant increase in fibronectin mRNA expression was observed (3.8 ± 0.8-fold) and sustained at 48 h (5.1 ± 1.6-fold) (P < 0.05) (Fig. 1). To determine if TGF-β 1 also induced fibronectin protein expression, Western blot analysis was conducted following 24 and 48 h of TGF-β 1 stimulation. Untreated cells show a small, but not significant, increase in fibronectin protein over time in untreated cells (Fig. 2). In comparison, TGF-β 1 induced an increase in fibronectin protein at both 24 and 48 h post-stimulation (Fig. 2a) and densitometric analysis confirmed that TGF-β 1 significantly increased fibronectin protein upregulation at 48 h (4.3 ± 1.1-fold) (P < 0.05) (Fig. 2b).

TGF-β 1 induces human lung fibroblast-mediated collagen gel contraction
The ability of human lung fibroblasts to contract threedimensional collagen gels is an in vitro model of ECM remodelling, a functional characteristic of lung fibrosis in vivo [22,23]. To investigate the effects of TGF-β 1 on human lung fibroblast-mediated collagen gel contraction, primary fibroblasts from n = 9 donors with a range of diagnoses (see Table 1) were used. Over the 96 h time course, fibroblasts stimulated with TGF-β 1 displayed significantly increased contraction of the collagen gels from 48 h when compared to the vehicle control ( Fig. 5: P < 0.05).

TGF-β 1 -induced upregulation of fibronectin, Tpm1.6/1.7 and Tpm2.1 protein: inter-donor variability
While TGF-β 1 -induced fibronectin is a known contributor to the fibrotic lung fibroblast phenotype [23], the role TGF-β 1 -induced Tm expression plays in regulating collagen gel contraction was next assessed. Thus, parallel  Western blot experiments were conducted using the same primary human lung fibroblasts used for the collagen gel contraction assay (Fig. 5). Fibroblasts were either left untreated (vehicle control) or stimulated with TGF-β 1 for 0, 24 and 48 h. Protein levels of fibronectin (Fig. 6a), Tpm2.1 (Fig. 6b), Tpm1.6/1.7 (Fig. 6c) were quantitated and analysed by densitometry. The results are expressed as protein fold change compared to control t = 0 and are expressed for each donor (as well as the mean for the n = 9 donors at each time point). Although there was a significant impact of TGF-β 1 on resultant protein levels (P < 0.05), not surprisingly, there was inter-donor variability between the TGF-β 1 -induced responses with TGF-β 1 stimulating the fibroblast production of both fibronectin and Tms to different extents (Fig. 6). The biological variability is unsurprising given the range of disease diagnoses in the group of cell donors (Table 1). It is also important to note that the goal of the study was not to link disease to the resultant responses (and the study is not powered to do so). Rather, our aim was to examine cooperation between Tms and fibronectin on contractility of human lung fibroblasts. Accordingly, we next examined the possibility that TGF-β 1 -induced collagen gel contraction may be greater if TGF-β 1 stimulation upregulates protein production of both fibronectin and TM1 isoforms.

Tpm2.1 collaborates with fibronectin to promote TGF-β 1 -induced contraction of human lung fibroblasts
To address the possibility that fibrosis severity may be linked to the donor's fibroblast response to TGF-β 1 stimulation, we performed multiple linear regression to test the hypothesis that fibronectin may collaborate with one or more TM1 isoforms to regulate TGF-β 1 -induced (compared to α-tubulin as a loading control); c Tpm1.6/1.7 (compared to α-tubulin as a loading control)). Results shown are densitometric analysis (normalised to loading controls and expressed as fold increase compared to control at 0 h). Data are protein levels detected in human lung fibroblasts from each of the nine donors with mean represented as black bars. Statistical analysis was performed using two-way ANOVA then Bonferroni's multiple comparisons test (where * denotes a significant effect of TGF-β 1 (P < 0.05)) collagen gel contraction. To do this, we represented the collagen gel contraction (from Fig. 5) as TGFβ 1 -induced collagen gel contraction (fold change over control at same time point) and measured area under the curve (AUC). With TGF-β 1 -induced collagen gel contraction (AUC) as the dependent variable, we conducted multiple linear regression to examine whether there was a statistically significant relationship between the dependent variable (contraction) and the explanatory variables (i.e. fibronectin and Tm isoforms) induced by TGF-β 1. The model is represented as: Y = β 0 + β 1 *B + β 2 *C; where Y is TGF-β 1 -induced collagen gel contraction (AUC); β 0 is the intercept (constant term) ; β 1 is the coefficient relating to B-TGF-β 1 -induced fibronectin protein upregulation at 48 h; β 2 is the coefficient relating to C-TGF-β 1 -induced Tpm1.6/1.7 or Tpm2.1 protein upregulation at 48 h). There was no significant relationship (P = 0.5578) between TGF-β 1induced fibronectin and Tpm1.6/1.7 upregulation and resulting contraction (Y = 32.52 + 2.26*B -0.38*C). In contrast, the combination of fibronectin and Tpm2.1 significantly predicted the extent of TGF-β 1 -induced contraction (Y = 56.99 + 9.174*B -27.02*C: P = 0.0221). Thus, fibronectin together with Tpm2.1, but not Tpm1.6/1.7, promotes TGF-β 1 -induced contraction of human lung fibroblasts. This cooperative relationship is represented graphically in a 3D scatter plot (Fig. 7).

Knocking down Tpm2.1 reduces human lung fibroblast-mediated collagen gel contraction
To demonstrate the importance of Tpm2.1 in human lung fibroblast-mediated collagen gel contraction, primary fibroblasts from n = 6 donors with a range of diagnoses (see Table 2

Discussion
Fibrotic tissue remodeling has a significant impact on morbidity and mortality worldwide. Treatment options are limited, and the fundamental mechanisms of disease progression are poorly understood. Thus, there is an urgent need to investigate fibrosis at the cellular and molecular level. The expression pattern of tropomyosin isoforms is critical, as the pattern of tropomyosin isoform decoration of actin filaments is a key determinant of the dynamics of the associated actin filament. We have shown that TGF-β 1 induces the expression of Tpm1.6/1.7 and 2.1 and that via concomitant induction of Tpm2.1 and fibronectin protein expression, TGF-β 1 induces human lung fibroblast-mediated collagen gel contraction. During the fibrotic process, the actin cytoskeletons of fibroblasts in the tissue play a twofold role. Firstly, inflammatory cytokines induce enlarged actin fibers and focal adhesions. Secondly, the cells sense the increasing rigidity of the ECM by exerting traction forces through their focal adhesions (mechanosensing) and then convert this rigidity signal into increased actin fiber formation through the activation of biochemical signaling pathway (mechanotransduction) [24]. Increased actin fiber formation enables the cells to exert greater forces through the focal adhesions, which contracts the ECM and thereby disrupts the tissue structure [25]. Collectively, these changes to the fibroblast cellular structure therefore underpin the development of fibrosis. This interface of inflammation and mechanobiology is a key determinant of cellular architecture and tissue functioning. Currently, we do not fully understand why cells tip over into aberrant actin filament and focal adhesion formation to create the stiff, immobile fibroblast cells that are characteristic of fibrotic tissue. Intriguingly, a recent study by Jaffar et al. [14] demonstrated greater cellular stiffness in fibroblasts from people with IPF and greater cytoskeletal reorganization in response to TGF-β 1 . We propose that tropomyosins are key to actin stabilisation that characterize fibroblasts in fibrotic tissue.
Tropomyosins are a multi-isoform family of proteins that assemble into polymers and lay in the major groove of polymerized actin filaments [10]. The association of the tropomyosin polymer with actin filaments in turn determines the association of molecules that control actin filament turnover [26][27][28]. Ultimately, the pattern of tropomyosin decoration of actin filaments defines the function of the associated filament [29,30]. The tropomyosin isoforms we have examined in this study have been associated with actin stabilisation, a key characteristic of cellular stiffening that occurs in fibrosis. Tpm1.6 stabilises stress fibres [31] and Tpm1.6/1.7 was recently shown to be recruited to actin filaments to increase actomyosin contractility [32]. Notably, the actin stabilizing role of isoform Tpm2.1 (previously known as Tm1) has previously led some investigators to describe it as a pro-fibrotic response gene [33]. Tpm3.1 (previously known as Tm5NM1) has been shown to stimulate actin stress fibre formation due to actin stabilisation [26,34] in the context of cancer cell migration. Tm isoforms can exert distinct effects that depend on the cell's physical environment (such as ECM) and thus it is important to note that cell-and disease-type differences can exist.
Tropomyosins are key players in actin dynamics and despite the key role of the actin cytoskeleton influencing fibrosis, surprisingly few studies have investigated the role of the pro-fibrotic cytokines TGF-β 1 on tropomyosin expression, and none in lung fibroblasts from human subjects. Most studies to date have focused on the role and regulation of tropomyosins in TGF-β 1 -induced epithelial-mesenchymal transition in transformed epithelial cell lines (such as A549 [35], NMuMG [33,36]) or primary lens epithelium [37,38]. Schevzov et al. have shown that expression of Tm isoforms in human and mouse lung fibroblasts [7] and we are the first to extend these studies in human lung fibroblasts to show the inflammatory cytokine TGF-β 1 also induce Tpm1.6/1.7, Tpm2.1, but not Tpm3.1.
We examined the hypothesis that cooperation exists between tropomyosin isoforms upregulated by TGF-β 1 and fibronectin on the resultant collagen gel contraction, a model of lung stiffening [23]. We observed variability in patient specific expression of fibronectin, Tpm1.6/1.7 and Tpm2.1 induced by TGF-β 1 (at 48 h) from n = 9 donors. Interestingly, multiple linear regression demonstrated a statistically significant relationship between the extent of Tpm2.1 and fibronectin induced by TGF-β 1 and resultant collagen gel contraction. This would align with the known function of Tpm2.1 to stabilise actin and promote fibrosis [33], in addition to known ability of TGFβ 1 to synthesise and deposit fibronectin into surrounding remodelled environment. Taken together, our results link Tpm2.1 to lung fibroblast stiffening that characterises fibrosis in IPF. However, while Fig. 7 offers a correlation between fibronectin, Tpm2.1 and TGF-β 1 -induced contractility, correlation need not mean causation. We conducted further experimentation (Fig. 8) to show that knocking down Tpm2.1 reduces human lung fibroblastmediated collagen gel contraction. collagen gel size (percentage compared to 0 h). Statistical analysis was performed using (b) Student's unpaired t test and (c) two-way ANOVA then Bonferroni's multiple comparisons test (where * denotes that significantly knocking down Tpm2.1 protein represses human lung fibroblast-mediated collagen gel contraction (P < 0.05)). Data are mean ± SEM values from n = 6 human lung fibroblast primary cell cultures