Cellular stress, TGF-β, and EMT: what are the true origins of fibrosis?

Back to top Introduction

Fibrosis results from excessive extracellular matrix deposition by myofibroblasts in the course of chronic inflammation and wound healing, and is a key pathogenic process in many organs, including kidneys, lung, and liver (see figure Fibrosis). Indeed, it is estimated that 45% of deaths in the developed world result from fibrosis (1). Epithelial-mesenchymal transition, a reversible process in which epithelial cells transdifferentiate into cells with mesenchymal characteristics, is thought to be a step in the progression to cancer metastasis and is also widely considered to be a crucial process in fibrosis alongside inflammation and apoptosis (2). In vitro studies show that growth factors like TGF-β and FGF-2 can trigger several changes in epithelial cells, including loss of epithelial markers, acquisition of mesenchymal markers, and increased levels of ECM component and matrix metalloproteinase production (3). However, a review published in 2011 argues against the involvement of EMT in renal fibrosis, citing insufficient specificity of commonly-used phenotypic markers and a lack of strong evidence for this transition in cell fate tracing and structural studies (4). Moreover, in vivo evidence regarding EMT in fibrosis is mixed. This review summarizes recent findings in renal and pulmonary fibrosis research, and their potential impact on this debate.

Back to top EMT regulators show relevance in renal fibrosis: HIPK2 and Alk3

Jin et al. recently identified one of the molecular drivers of kidney fibrosis, showing that homeodomain-interacting protein kinase 2 (HIPK2) links disease-related processes, like DNA damage and oxidative stress, with EMT leading to fibrosis (5). The work initially focused on HIV-associated nephropathy (HIVAN), using the Tg26 murine model and infected human renal tubular epithelial cells (hRTEC), but the findings were quickly extended to other experimental models of kidney disease, including unilateral ureteral obstruction (UUO), and folic-acid-induced renal fibrosis. Moreover, HIPK2 expression was associated with fibrosis in patient samples of diabetic nephropathy, severe IgA nephropathy, and focal segmental glomerulosclerosis in addition to HIVAN samples, suggesting that HIPK2 may regulate fibrosis across a broad range of kidney diseases.

After identifying HIPK2 as an important kinase in HIVAN using a combination of computational and experimental tests, the team interrogated its related signaling pathways using hRTECs or the proximal tubular epithelial cell line HK2. Overall, their results suggest that reactive oxygen species or DNA damage resulting from infection lead to inhibition of the ubiquitin ligase SIAH1, releasing its post-translational repression of HIPK2 and therefore opening up the floodgates of EMT-promoting pathways like TGF-β, Wnt, and Notch, apoptosis-promoting p53, and inflammatory NFκB signaling, all of which combine to produce fibrosis (see figure HIPK2 in fibrosis). Therefore, treatments targeting HIPK2 could potentially stop this cascade in its tracks, preventing fibrosis and halting the progression of kidney disease.
On the flip side, a new study from Sugimoto et al shows therapeutic promise of an agonist for a bone morphogenic protein 7 (BMP7) receptor, activin-like kinase 3 (Alk3), in reversing fibrosis and promoting kidney regeneration (6). Like HIPK2, Alk3 also functions across a broad range of fibrosis-related kidney disease models, including diabetic nephropathy, nephrotoxic serum-induced kidney fibrosis (NTN), the UUO model, and a murine model of Alport syndrome. BMP7 is an antagonist of TGF-β signaling that has been previously shown to reverse EMT (7) and is downregulated in kidney injury, thus suggesting the benefit of targeting Alk3. The authors showed that deletion of Alk3 in renal tubular epithelium worsened fibrosis during NTN and enhanced EMT and macrophage influx, suggesting that Alk3 might be an effective therapeutic target.

The team therefore developed a BMP7 signaling agonist, THR-123, and demonstrated that it had similar properties to BMP7 in inhibiting induction of the EMT program, proinflammatory cytokine production, apoptosis, and fibrosis in several disease models. Moreover, THR-123 actually led to reversal of fibrosis when given at 6 weeks after NTN induction. However, THR-123 conferred no benefit in Alk3-deficient mice subjected to NTN or ischemia reperfusion injury compared to wild-type mice, confirming that THR-123 acts through Alk3. The team's data also suggests that combination therapy of angiotensin-converting enzyme inhibitor(ACEi) drugs and BMP7 agonists may be an effective anti-fibrotic treatment strategy, as adding THR-123 to ACEi treatment led to greater preservation of renal function and inhibition of mesangial matrix expansion, tubular atrophy, and interstitial volume expansion in mice with diabetic nephropathy.

Back to top Who is responsible for pulmonary fibrosis: epithelial cells or resident stromal cells?

Pulmonary fibrosis is initiated by several potential triggers, including infection, chemotherapy, and environmental toxins, as well as undefined causes such as in IPF. EMT is thought to be partially responsible for the generation of myofibroblasts in this setting, in addition to other sources such as fibrocytes and existing mesenchymal cells (8). Two recent in vitro studies suggest how and why EMT might be triggered in the lung, describing how mechanical and endoplasmic reticulum (ER) stress induce this process in alveolar type II epithelial cells (AECs) via distinct pathways. However, an in vivo study in mice shows evidence that the cells responsible for fibrosis may not be AECs after all.

Back to top Mechanical stress triggers EMT in alveolar epithelial cells in vitro

 Mechanical stress on AECs can result from mechanical ventilation or abnormal amounts of stretch due to scarring. Heise et al demonstrated that mechanical stretch applied to AECs led to diminished levels of the epithelial marker E-cadherin and higher levels of EMT markers like vimentin and α-smooth muscle actin (α-SMA) (9). They also noted an increase in has3 expression levels and in the proportion of low molecular mass hyaluronan (sHA) in supernatant of stretched cells. HA is known to be involved in EMT in other settings (10), and experiments with sHA alone, an HA antagonist, and cells deficient for the HA receptor all suggested that hyaluronan was indeed the responsible agent in the induction of EMT via stretch.

HA can act as a "danger-associated molecular pattern" for the innate immune system, triggering signaling through Toll-like receptors 2 and 4 in response to tissue injury. Accordingly, knocking out the TLR adaptor MyD88 abrogated AEC's induction of EMT in response to stretch, suggesting that it is HA's interaction with TLRs that drives this process. The Wnt/β-catenin pathway was also shown to be involved, as inhibiting the Wnt-induced gene WISP-1 blocked EMT in AECs.

Back to top ER stress triggers EMT in alveolar epithelial cells in vitro

Mechanical stress is not the only stressor that can send AECs down the path of EMT in vitro. Signs of ER stress had been noted in 3 forms of IPF, one of which included a mutation in the surfactant protein C gene, SFTPC. Tanjore et al delved into this phenomenon in order to discover the link between ER stress and fibrosis, and revealed that ER stress in AECs triggers EMT (11). When ER stress was induced by stable expression of the mutated surfactant protein C or tunicamycin treatment, morphological changes occurred that were consistent with fibroblasts. Additionally, these cells showed upregulation of several key mesenchymal markers such as vimentin and N-cadherin, and lower levels of the epithelial marker E-cadherin, strongly suggesting that EMT was taking place (see figure Epithelial-mesenchymal transition).

Potentially explaining these findings, the team found that induction of ER stress led to activation of several EMT-related pathways, including TGF-β, Wnt/β-catenin, and Src, as measured by increased phosphorylation of Smad2/3 and Src and increased localization of β-catenin in the nucleus, where it colocalized with phospho-Smad2/3. Inhibiting Smad2 and Src partially blocked the transition induced by tunicamycin, but neutralizing TGF-β or blocking Wnt signaling had no effect, indicating that extracellular signaling was likely not involved. Additionally, ER-stress-induced EMT was transient and reversible when the stressor was removed.

Back to top Pulmonary fibrosis occurs via stromal cells in vivo, independent of AEC EMT

The real test of a phenomenon's disease relevance, however, is whether it can be observed in vivo. Contrary to these in vitro studies, a new study in vivo suggests that IPF may not involve EMT by AECs at all. Rock et al used confocal microscopy and immunofluorescence to clarify the role of AEC2 cells in a murine bleomycin-induced pulmonary fibrosis model (12). To track the changes in this population during fibrosis, they lineage-labeled Sftpc-positive AEC2 cells and examined their proliferative ability and changes in surface markers and morphology. While a sizeable proportion of AEC2 cells gave rise to AEC1 cells upon fibrosis induction, they did not express the typical mesenchymal markers vimentin, αSMA, or others, as measured by immunofluorescence and confirmed by qPCR.
These findings suggest that epithelial-mesenchymal transition is not the origin of myofibroblasts in this fibrosis model. While further experiments in the paper also seem to rule out NG2-positive pericyte-like cells as the source for αSMA-positive myofibroblasts, the authors observed NG2-negative cells proliferating within fibrotic regions. They propose that other pericyte-like cells may give rise to the αSMA-positive myofibroblast population

Back to top Conclusions

Evidence exists on both sides of the question over EMT's role in fibrosis, and further observations will be needed to determine whether a full transition to myofibroblasts truly occurs in vivo. Whether EMT occurs may depend on the stimulus or the organ in which fibrosis is being initiated, and differences may exist between in vivo murine models and human disease. If epithelial cells are ultimately found not to be a source of myofibroblasts in fibrosis, the question remains: do all ECM component-secreting myofibroblasts arise from resident fibroblasts and fibrocytes? Are there other sources that remain to be clarified? And if epithelial cells are not making a full transition, what is the significance of the observations that led to this hypothesis? Further studies will need to address these questions in order to advance our understanding and ability to treat fibrosis.

Back to top References

1. Wynn, T.A. (2008) Cellular and molecular mechanisms of fibrosis. J. Pathol. 214, 199. Kalluri, R., and Weinberg, R.A. (2009) The basics of epithelial-mesenchymal transition. J. Clin. Invest. 119, 1420.
2. Strutz, F. et al. (2002) Role of basic fibroblast growth factor-2 in epithelial-mesenchymal transformation. Kidney Int. 61, 1714.
3. Kriz, W., Kaissling, B., and Le Hir, M. (2011) Epithelial-mesenchymal transition in kidney fibrosis: fact or fantasy? J. Clin. Invest. 121, 468.
4. Jin, Y. et al. (2012) A systems approach identifies HIPK2 as a key regulator of kidney fibrosis. Nat. Med. 18, 580.
5. Sugimoto, H. et al. (2012) Activin-like kinase 3 is important for kidney regeneration and reversal of fibrosis. Nat. Med. 18, 396.
6. Zeisberg, et al. (2003) BMP-7 counteracts TGF-β1-induced epithelial-to-mesenchymal transition and reverses chronic renal injury. Nat. Med. 9, 964.
7. Wynn, T.A. (2011) Integrating mechanisms of pulmonary fibrosis. J. Exp. Med. 208, 1339.
8. Heise, R.L., Stober, V., Cheluvaraju, C., Hollingsworth, J.W., and Garatziotis, S. (2011) Mechanical stretch induces epithelial-mesenchymal transition in alveolar epithelia via hyaluronan activation of innate immunity. J. Biol. Chem. 286, 17435.
9. Toole, B.P., Zoltan-Jones, A., Misra, S., and Ghatak, S. (2005) Hyaluronan: a critical component of epithelial-mesenchymal and epithelial-carcinoma transitions. Cells Tissues Organs. 179, 66.
10. Tanjore, H., et al. (2011) Alveolar epithelial cells undergo epithelial-to-mesenchymal
11. transition in response to endoplasmic reticulum stress. J. Biol. Chem. 286, 30972.
12. Rock, J.R. et al. (2011) Multiple stromal populations contribute to pulmonary fibrosis without evidence for epithelial to mesenchymal transition. Proc. Nat. Acad. Sci. 108, E1475.
13. Lopez-Novoa, J.M., and Nieto, M.A. (2009) Inflammation and EMT: an alliance towards organ fibrosis and cancer progression. EMBO Mol. Med. 1, 303.