Assessment of the effect of potential antifibrotic compounds on total and αVβ6 integrin‐mediated TGF‐β activation

Abstract Transforming growth factor‐β (TGF‐β) plays an important role in the development of tissue fibrosis, and molecules inhibiting this pathway are attractive therapeutic targets for fibrotic diseases such as idiopathic pulmonary fibrosis (IPF). Activation of TGF‐β is the rate‐limiting step in TGF‐β bioavailability, and activation by the αVβ6 integrin is important in fibrosis of the lung, liver, and kidney. Activation of TGF‐β by αVβ6 requires direct cell–cell contact and measurable release of active TGF‐β in extracellular fluid compartments does not reflect tissue specific activation. The aim of this study was to determine the effect of antifibrotic compounds on both total, and specific αVβ6 integrin‐mediated TGF‐β activity. Using a transformed mink lung cell (TMLC) TGF‐β reporter, the effects of potential antifibrotic therapies including an activin‐like kinase (Alk5) inhibitor, Dexamethasone, Pirfenidone, N‐acetylcysteine (NAC), and BIBF1120 were assessed. Effects due to αVβ6 integrin‐mediated TGF‐β activity were measured using reporter cells cocultured with cells expressing αVβ6 integrins. These high‐throughput studies were validated using a phosphorylated Smad2 Enzyme‐Linked Immunosorbent Assay. Alk5 inhibitors are potent inhibitors of TGF‐β activity, whereas the novel antifibrotics, Pirfenidone, BIBF1120, and NAC are only moderate inhibitors, and Dexamethasone does not specifically affect TGF‐βactivity, but inhibits TGF‐β‐induced gene expression. None of the current small molecular inhibitors inhibit αVβ6‐mediated TGF‐β activity. These results demonstrate the potential of this high‐throughput assay of αVβ6‐specific TGF‐β activity and illustrate that currently available antifibrotics have limited effects on αVβ6 integrin‐mediated TGF‐β activity.


Introduction
Fibrotic diseases represent a significant, and escalating, burden globally, with fibrotic diseases of the lung, liver and kidney being responsible for 3.4% of the global mortality in 2010, having increased from 2.7% in 1990 (Lozano et al. 2012). Idiopathic pulmonary fibrosis (IPF) is an exemplar fibrotic disease. It is a chronic, progressive lung disease with a high mortality rate (Navaratnam et al. 2011), and at present, there is no effective therapy so patients progress to respiratory failure with a median survival of around 3 years (Gribbin et al. 2006). The incidence continues to rise with almost 5000 deaths per year in the United Kingdom.
There is, therefore, an urgent unmet need for novel therapeutic interventions.
The cause of IPF remains unknown, but the current hypothesis for its pathogenesis proposes that repeated epithelial injury and a lack of adequate repair of the alveolar-capillary basement membrane leads to infiltration of fibrogenic cells and subsequent excessive matrix production (Jenkins et al. 2012). Transforming growth factor-b (TGF-b) is a key cytokine that has been implicated in both epithelial repair and matrix deposition and there is considerable evidence that it plays a central role the pathogenesis of fibrotic diseases. Increased levels of TGF-b have been found in fibroblastic foci of IPF patients (Broekelmann et al. 1991), overexpression of active TGF-b induces fibrosis (Sime et al. 1997), whereas inhibition has prevented fibrosis (Wang et al. 1999) in animal models of disease. Despite the considerable evidence implicating TGF-b in pulmonary fibrosis, no specific inhibitors of TGF-b have emerged as therapies for IPF. TGF-b1 is a member of the highly conserved TGF-b superfamily, it is ubiquitously expressed and has pleiotropic effects on cells and organs throughout the body. Global inhibition of TGF-b activity has potent proinflammatory effects, and TGF-b inhibitors have been limited by their toxicity. It is likely that the pleiotropic effects of TGF-b are mediated through tight spatial and temporal regulation of its activation.
TGF-b1 is synthesized as a small latent complex in which the active TGF-b is noncovalently associated with the latency-associated peptide (LAP). This is then further associated with the latent TGF-b-binding proteins (LTBPs) to form the large latent complex, and through the LTBPs, is tethered to components of the extracellular matrix in its inactive form. Activation of TGF-b requires dissociation from the LTBPs. There are different mechanisms by which activation can occur including physical effects such as extremes of heat or pH (Brown et al. 1990) and by proteolytic cleavage involving plasmin (Lyons et al. 1988), tryptase (Tatler et al. 2008), thrombin (Taipale et al. 1992), MMP's (Mu et al. 2002) or thrombospondin (Crawford et al. 1998). However, in vivo, a major mechanism of TGF-b activation is via integrins (Munger et al. 1999).
The aVb6 integrin is an epithelial cell-restricted integrin that is expressed at low levels in normal lung but dramatically increased in fibrotic disease, and there is considerable evidence implicating this integrin in the pathogenesis of IPF (Munger et al. 1999;Horan et al. 2008;Puthawala et al. 2008;Xu et al. 2009), renal fibrosis (Ma et al. 2003;Hahm et al. 2007), and hepatic fibrosis (Wang et al. 2007;Popov et al. 2008;Sullivan et al. 2010). Therefore, inhibition of aVb6 integrin-mediated TGF-b activation is an appealing strategy for the development of novel therapies in fibrotic disease, through targeting inhibition of TGF-b activity to the site of disease thus avoiding significant "off-target" effects.
This study used a high-throughput reporter cell assay, and a phosphorylated Smad2 (pSmad2) Enzyme-Linked Immunosorbent Assay (ELISA) to determine the total, and avb6 integrin-specific, TGF-b activity of four potential antifibrotic compounds, Pirfenidone, NAC, dexamethasone, and BIBF1120. Pirfenidone, NAC, and BIBF1120 were weak inhibitors of total TGF-b activation, and dexamethasone had no intrinsic effect of TGF-b activation but had moderate effects on TGF-b signalling. These studies demonstrate an urgent need to develop novel anti-aVb6 integrin inhibitors.

Flow cytometry
Cells were harvested by trypsinization. Nonspecific interactions were blocked with goat serum for 20 min at 4°C. Cells were then washed in PBS containing Ca 2+ and Mg 2+ before incubating with the anti-aVb6 mouse monoclonal antibody, 10D5, at 10 lg mL À1 for 1 h at 4°C. After washing again, cells were incubated with goat anti-mouse secondary antibody labelled with Phycoerythrin at a 1:200 dilution for 20 min at 4°C. Fluorescence was measured using a FACSCanto II flow cytometer (BD, Franklin Lakes, NJ), and analysed using FlowJo software (Treestar, OR).

TMLC active TGF-b assay
To measure the inhibition of aVb6 mediated TGF-b activity, a coculture of MEF cells stably transfected with the aVb6 integrin was used. MEF-b6 cells were seeded into 96-well plates at a density of 2.5 9 10 5 cells mL À1 in DMEM with 10% serum. The cells were incubated overnight to allow adherence. TMLC cells were harvested and resuspended at a concentration of 5 9 10 5 cells mL À1 in DMEM + FCS, and any inhibitors or antibodies added to the correct concentration. Medium was removed from the MEF-b6 cells and replaced with 100 ll of appropriate TMLC cell suspension. After 16 h incubation at 37°C, the cells were washed with PBS and then lysed with 50 lL of reporter lysis buffer (Promega, Hampshire, UK). The cell lysate was then transferred to a luminometry plate and 100 lL of luciferase assay reagent added (Promega). Luciferase activity was measured immediately using a FLUOstar omega multi-mode microplate reader (BMG Labtech, Ortenburg, Germany).
To measure total active TGF-b, TMLC cells were plated alone and serum starved overnight. A quantity of 2 ng mL À1 TGF-b was added to each well with the appropriate inhibitor concentration and incubated for 16 h at 37°C. Harvest and measurement was identical to coculture experiments.

Phosphorylated Smad2 ELISA
MEF-b6 cells were plated in 10-cm petri dishes and incubated at 37°C until 80% confluence. Inhibitors were added along with fresh medium, containing 10% FCS and 2 ng mL À1 TGF-b. After 4 h of incubation at 37°C, nuclear protein was prepared using Nuclear Extract kit (Active Motif, La Hulpe, Belgium). Protein concentration was determined using BCA Protein Assay kit (Pierce, Rockland, IL). A quantity of 10 lg of nuclear protein was loaded into a Pathscan pSmad2 ELISA assay (Cell Signalling, Danvers, MA). Absorbance was read at 450 nm using a FLUOstar omega multi-mode microplate reader (BMG Labtech).

Statistical analysis
All values are expressed as mean AE SEM. The minimum number of replicates was three. Statistical analysis was performed using GraphPad Prism 6 software. Comparisons with control were performed using a one-way analysis of variance (ANOVA) test. A P < 0.05 was considered significant.

Currently available antifibrotic compounds are weak inhibitors of TGF-b receptor signalling
To determine the effect of potential antifibrotic compounds on TGF-b activity and signalling TMLCs were used. TMLCs were stimulated with exogenous TGF-b in the presence of a small molecular Alk5 inhibitor (SB525334A), Dexamethasone, Pirfenidone, BIBF1120, and NAC. SB525334A inhibited TGF-b activity with an IC 50 of~0.5 lmol/L (Fig. 1A). Dexamethasone was a weak inhibitor of TGF-b activity with maximal inhibition of TGF-b induced reporter activity of between 20% and 50%. Dexamethasone inhibited TGF-b induced reporter activity in the nanomolar range ( Fig. 1B), Pirfenidone and NAC had no effect on TGF-b reporter activity ( Fig. 1C and D). None of the concentrations used were toxic to TMLC cells ( Fig. 2A-D). BIBF1120 did appear to inhibit TGF-b-induced reporter cell activity but this was due to toxic effects on the reporter cells (Fig. 2E).
Total cellular TGF-b activation can be discriminated from aVb6 integrin-specific activation using an aVb6 expressing and TMLC reporter coculture assay To determine whether antifibrotic compounds were able to inhibit total cellular TGF-b or aVb6 integrin-specific TGF-b activation, a coculture assay using an avb6 integrin-expressing cell line and the TMLC reporter cell line were used. High levels of the aVb6 integrin were confirmed on the experimental cells and the absence of this integrin was confirmed on TMLCs by flow cytometry (Fig. 3A and B). To confirm the coculture assay could selectively distinguish between aVb6 mediated TGF-b and total cellular TGF-b activation, the aVb6 specific blocking antibody 10D5 and the pan-TGF-b blocking antibody 1D11 were used. 10D5 lead to concentration-dependent inhibition of the coculture with an IC 50 of 0.25 lg mL À1 , without effecting TMLC reporter cells at any concentration tested (Fig. 3C) . Similarly, an IgG2a control had no effect on either the coculture or the reporter cells alone (Fig. 3D). Surprisingly, 1D11 had no effect on either the coculture or the reporter cells at concentrations below 2.5 lg mL À1 (Fig. 3E), but at a concentration of 25 lg mL À1 a 50% inhibition was observed in the coculture. The IgG1 control had no effect on either the coculture or the reporter cells (Fig. 3F). None of the antibodies used had toxic effects on either the reporter cells or the coculture (Fig. 4).

Currently available antifibrotic compounds are poor inhibitors of aVb6-mediated TGF-b activation
Having confirmed that the coculture assay could distinguish between total and aVb6 integrin-specific TGF-b activity, the potential antifibrotic compounds SB525334A, dexamethasone, Pirfenidone, and NAC were assessed. SB525334 and Dexamethasone inhibited both reporter cells and coculture in a concentration-dependent manner with an IC 50 of 5 nmol/L and 1 nmol/L respectively ( Fig. 5A and B) suggesting inhibition of total cellular, rather than avb6-specific TGF-b activity. Pirfenidone and NAC had no effect on aVb6-dependent or total cellular TGF-b activation. (Fig. 5C and D).

Currently available antifibrotic compounds are moderate inhibitors of Smad2 phosphorylation
The most proximal step following TGF-b receptor ligation is phosphorylation of the receptor Smad2. To determine the effect of the potential antifibrotic compounds on TGF-b activation and receptor inhibition an ELISA measuring levels of pSmad2 was used. SB525334A inhibited pSmad2 in a concentration-dependent manner with an IC 50 of 100 nM (Fig. 6A) similar to previous data using the coculture assay (Figs. 1A and 5A). In contrast, dexamethasone had no effect on pSmad2 levels ( Fig. 6B), demonstrating that all effects on reporter cell assays ( Fig. 1B and 5B) were due to direct effects on the PAI1 promotor, not through inhibition of TGF-b activation or receptor ligation. Pirfenidone, BIBF1120, and NAC had moderate but variable effects on pSmad2 levels ( Fig. 6C and D) supporting observations in coculture assays.

Discussion
TGF-b activation and signalling pathways are central to the pathogenesis of fibrosis. However, the determination of biologically relevant TGF-b in experimental systems is challenging due to the tight spacial and temporal regulation of TGF-b activation in tissues. Activation of TGF-b by the aVb6 integrin is likely to be a central driver of lung fibrosis, but measurement of aVb6 integrin-mediated TGF-b activation is especially challenging because it does not lead to free release of measurable amounts of active TGF-b. Assays that rely entirely on measuring TGFb in the supernatants of biological fluids will give unreliable assessments of TGFb activity because they do not take into account cell-associated activation of TGFb. Furthermore, because total TGFb is synthesized in considerable excess of biologically active TGFb (Xu et al. 2009), and activation of TGFb is the rate-limiting step in TGFb bioavailability (Annes et al. 2003) assays that measure synthesis of TGFb at the message level will give little indication of its bioavailability. Therefore, we sought to systematically assess the ability of currently clinically available antifibrotic compounds to inhibit cell-free, total cellular, and avb6 integrin-mediated TGFb activity in cell-based assays.
We use an in vitro model utilizing MEFs stably transfected with wild-type b6 integrins. Although the avb6 integrin is restricted to epithelial cells in vivo, this system has a number of advantages for measuring avb6mediated TGFb activation in vitro. Although primary epithelial cells express high levels of avb6 integrins, expression is lost in epithelial cell lines, therefore, MEF-b6 cells offer a high level of uniform avb6 expression in rapidly proliferating cells. Furthermore, avb6 integrinmediated TGFb activation relies on cellular contraction (Shi et al. 2011), and fibroblasts are more contractile than epithelial cells, thus the measurable signal is amplified. Our data demonstrate a rational, and practical, strategy to determine the relative contribution of total and avb6 integrin-specific TGF-b activation using in vitro assays that can be used to assess the preclinical potential of antifibrotic molecules. SB525334A is an Alk5/type 1 TGF-b receptor kinase inhibitor that inhibits Alk5 phosphorylation of Smad3 with an IC 50 of 14 nmol/L in kinase assays (Grygielko et al. 2005). Furthermore, it inhibits TGF-b induced PAI1 mRNA and procollagen gene synthesis with an IC 50 in the nanomolar range (Grygielko et al. 2005). It is reassuring, if unsurprising, that inhibition of physiological concentrations of cell-associated TGF-b occurred with SB525334A in the nanomolar range, consistent with these previous observations. Although SB525334A inhibits experimental models of pulmonary artery hypertension (Thomas et al. 2009), renal fibrosis (Grygielko et al. 2005), and mesenchymal tumours, it has antiapoptotic and mitogenic effects in epithelial cells raising concerns regarding potential toxicity that may limit its therapeutic utility (Laping et al. 2007).
Somewhat surprisingly, the pan-TGF-b-isoform neutralizing antibody 1D11 did not completely inhibit total cellular, or aVb6 integrin-mediated TGF-b activity at concentrations below 2.5 lg mL À1 . Original descriptions of 1D11 demonstrate that it is able to inhibit the growth inhibitory activity of both TGF-b1 and TGF-b2 at concentrations of 10 lg mL À1 (Dasch et al. 1989). Although our data would confirm that this concentration of 1D11 . Solid line = 10 mg mL À1 10D5 antibody. Dotted line = secondary antibody only. The integrin aVb6-neutralizing antibody 10D5 lead to a concentration-dependent reduction in luciferase activity in coculture, without effecting levels of luciferase expression from TMLC reporter cells alone (C). An IgG2a isotype control antibody had no effect in either system (D). The pan-isoform TGF-b neutralizing antibody 1D11 lead to partial inhibition of luciferase activity from both coculture and TMLC reporter cells (E). IgG1 Isotype control antibody had no effect on either system (F). Solid line = coculture. Dotted line = TMLC cells only. All experiments were performed in triplicate and repeated three times. Data presented are the mean of three independent experiments and expressed as a percentage of untreated controls. Data expressed as mean AE standard error. *P < 0.05, **P < 0.01, ***P < 0.001. is sufficient to inhibit TGF-b activity, a comparison of its TGF-b inhibitory effect in relation to other inhibitors has not previously been performed, and it is apparent that it has fairly low-level TGF-b inhibitory activity. This is probably due to the mechanism of action of TGF-b activity in vitro and in vivo, which in many circumstances, including activation by aVb6 integrins, requires cytoplasmic forces to alter the structure of the latent TGF-b complex that is tethered to the extracellular matrix (Fontana et al. 2005;Wipff et al. 2007). The active TGF-b is not released into the extracellular fluid in large quantities because cells have to be in direct cell-cell contact to respond to the TGF-b (Munger et al. 1999). Thus, it is likely that the steric hindrance provided by the matrixlatent TGF-b complex is sufficient to inhibit the effect of 1D11 in physiological conditions, and this may explain the disappointing effects of humanized TGF-b-neutralizing antibodies in clinical trials of fibrotic disease (Denton et al. 2007).
The aVb6-neutralizing antibody, 10D5, did not completely inhibit the reporter system. However, it is able to inhibit TGF-b activity in the coculture at concentrations as low as 0.25 lg mL À1 consistent with previous reports (Weinreb et al. 2004), and has no effect on reporter cells that don't express aVb6 integrins demonstrating its specificity for aVb6 mediated pathways. The lack of complete inhibition may reflect steric hindrance, alternative pathways of TGF-b activation, and possibly the presence of active TGF-b within the cell culture medium. We have observed similar effects in serum-free conditions, and although other aVb6-neutralizing antibodies are considerably more potent in the same system (Weinreb et al. 2004), they were also unable to completely neutralize TGF-b activity. This suggests that alternative TGF-b activation pathways, possibly via TGF-b2 (Neurohr et al. 2006), are playing a role in homeostatic TGF-b activation.
Pirfenidone is a promising antifibrotic molecule licensed for the treatment of IPF (Jenkins 2013). The precise mechanism of Pirfenidone's action remains unknown but it has been described as an inhibitor of TGFb signalling and activity (Cho and Kopp 2010) and is thought to exert its clinical effects through inhibition of TGFb production (Azuma 2012). It can inhibit TGF-b gene expression in the lungs of bleomycin-treated hamsters (Iyer et al. 1999), and kidneys of cyclosporine-treated rats (Shihab et al. 2002), furthermore, it can inhibit TGF-b2 protein and mRNA expression in malignant human cells (Burghardt et al. 2007). However, no studies have systematically assessed the effect of Pirfenidone on TGFb activation. Our studies have used concentrations of Pirfenidone that are biologically relevant, similar to levels observed in plasma from patients (Shi et al. 2007;Rubino et al. 2009) and would suggest that Pirfenidone has no TGF-b1,3 inhibitory activity in the range of 30-1000 lmol/L (equivalent to 5.8-185 lg mL À1 ). Pirfenidone has been previously shown to inhibit TGF-b1-induced responses in human lung cells at between 100 and 500 lg mL À1 (Nakayama et al. 2008;Hisatomi et al. 2012) and it has been shown to specifically inhibit TGF-b2-induced Smad3 nuclear translocation in lens epithelial cells at 500 lg mL À1 (Yang et al. 2013). Our studies would suggest that physiological concentrations of Pirfenidone have no effect on aVb6 integrin-mediated TGF-b activation, which is the predominant pathway for TGF-b1 activation in the alveoli. We cannot exclude the possibility that Pirfenidone preferentially inhibits TGF-b2 activation pathways that are independent of the avb6 integrin (see Fig. 7). Therefore, Pirfenidone in combination with avb6neutralizing strategies may be beneficial as combination therapies in fibrotic disease targeting different aspects of TGFb signalling pathways.
Reactive oxygen species (ROS) are thought to play a significant role in fibrogenesis (Liu and Gaston Pravia 2010) and NAC is used for its antioxidant properties as a therapy in IPF (Demedts et al. 2005). ROS is thought to specifically activate latent TGF-b1 (Jobling et al. 2006), and NAC has been shown to inhibit Smad3 phosphorylation in lung epithelial cells (Felton et al. 2009). Although there are no data to suggest oxidant-mediated injury plays a role in avb6 integrin-mediated TGFb activation, the role of NAC in aVb6 integrin-mediated TGF-b activation has not been studied. The concentrations of NAC used in this study range from the physiological, 0.3 mmol/L, (Borgstrom et al. 1986) to previously used in vitro concentrations, 5 mmol/L (Felton et al. 2009) and 10 mmol/ L (Lavrentiadou et al. 2001). In contrast with the studies by Felton et al. (2009), our data do not suggest that NAC has consistent TGF-b inhibitory activity. This may be because the previous studies used a prolonged culture protocol to assess effects of NAC on EMT. Our data do suggest a trend towards inhibition of pSmad2 at high concentrations of NAC, but these concentrations are of . aVb6-dependent TGF-b activity was determined by coculturing MEF-b6 cells with TMLC reporter cells in the presence of inhibitors, effects on total cellular TGF-b activity were determined by culturing TMLC reporter cells in the presence of inhibitors alone. SB525334 lead to a concentration-dependent decrease in both total cellular, and aVb6 integrin-dependent, luciferase activity (A). Dexamethasone had a similar but less potent effect on both total, and aVb6 integrin-dependent luciferase activity (B). Pirfenidone had no effect on total cellular TGF-b activity and no effect on aVb6 integrin-dependent luciferase activity (C). NAC at low concentrations had marginal effects on total cellular TGF-b, which decreased at higher concentrations and no effect on aVb6 integrin-mediated TGF-b (D). Solid line = coculture. Dotted line = TMLC cells only. All experiments were performed in triplicate and repeated three times. Data presented are the mean of three independent experiments and expressed as a percentage of untreated controls. Data expressed as mean AE standard error. *P < 0.05, **P < 0.01, ***P < 0.001. uncertain biological relevance and do not support a role for NAC in avb6 integrin-mediated TGFb activation. Dexamethasone inhibited TGF-b-stimulated, cell-associated, and avb6 integrin-mediated reporter activities; however, it had no effect on Smad2/3 phosphorylation. These data confirm the known effect of ligated glucocorticoid receptor, namely that it inhibits transcriptional activation of the PAI-1 promoter by Smad3 (Song et al. 1999), but demonstrate that dexamethasone has no effect on TGF-b activation pathways (see Fig. 7).
BIBF 1120 is highly selective for angiogenic tyrosine kinase receptors with an IC 50 in the nanomolar range for vascular endothelial growth factor, platelet derived growth factor, and fibroblast growth factor in kinase assays and can inhibit ligand induced mitogen-activated protein kinase and Akt phosphorylation at 0.3 lm L À1 (Hilberg et al. 2008). Alk5 is a serine/threonine kinase receptor and thus it was not anticipated that BIBF1120 would have any effect on TGFb signalling. However, at high concentrations BIBF1120 did show a trend towards inhibiting TGFb activity by coculture and pSmad2 assays. Although BIBF 1120 had toxic effects on TMLC reporter cells at moderate concentrations, it was tolerated by experimental cells. Furthermore, the pSmad2 ELISA is adjusted for protein content, limiting the confounding effects of cellular toxicity, suggesting that the effects of BIBF 1120 on pSmad2 are not due to generalized cytopathic effects. However, the concentrations required to effect TGFb Figure 6. To confirm the effects on reporter cells MEF-b6 cells were cultured in the presence of inhibitor for 4 h and pSmad2 was measured by ELISA. SB525334 lead to a concentration-dependent inhibition of pSmad2 (A), Dexamethasone had no effect on pSmad2 (B). Pirfenidone lead to moderate concentration-dependent inhibition of pSmad2 (C). NAC had variable and nonsignificant inhibition of pSmad2 (D). BIBF1120 lead to concentration-dependent inhibition of pSmad2 at high concentration (E). All experiments were performed in triplicate and repeated three times. Data presented are the mean of three independent experiments and expressed as a percentage of untreated controls. Data expressed as mean AE standard error. *P < 0.05, **P < 0.01, ***P < 0.001. pathways are much higher than therapeutic concentrations (Mross et al. 2010), suggesting that BIBF1120 is unlikely to have any clinical effect via TGF-b signalling.
These data demonstrate that current antifibrotic therapies have, at best, marginal effects on TGF-b activation at therapeutic concentrations. Although these data are not entirely unexpected, they are important because TGFb activation and signalling pathways are central in the pathogenesis of IPF. Furthermore, these data highlight practical, and robust preclinical assays for determining total and aVb6 integrin-mediated TGF-b activation. Given the likely importance of avb6 integrin-mediated TGF-b activation in the development of fibrogenesis there is a clear unmet need for further development of molecules that target this pathway. These assays will be useful preclinical assessment tools in the development of such compounds.