Integrins, myofibroblasts, and organ fibrosis


  • Potential conflict of interest: Nothing to report.

Henderson NC, Arnold TD, Katamura Y, Giacomini MM, Rodriguez JD, McCarty JH, et al. Targeting of αV integrin identifies a core molecular pathway that regulates fibrosis in several organs. Nat Med 2013;19:1617-1624. (Reprinted with permission.)


Myofibroblasts are the major source of extracellular matrix components that accumulate during tissue fibrosis, and hepatic stellate cells (HSCs) are believed to be the major source of myofibroblasts in the liver. To date, robust systems to genetically manipulate these cells have not been developed. We report that Cre under control of the promoter of Pdgfrb (Pdgfrb-Cre) inactivates loxP-flanked genes in mouse HSCs with high efficiency. We used this system to delete the gene encoding αV integrin subunit because various αV-containing integrins have been suggested as central mediators of fibrosis in multiple organs. Such depletion protected mice from carbon tetrachloride-induced hepatic fibrosis, whereas global loss of β3, β5 or β6 integrin or conditional loss of β8 integrins in HSCs did not. We also found that Pdgfrb-Cre effectively targeted myofibroblasts in multiple organs, and depletion of the αV integrin subunit using this system was protective in other models of organ fibrosis, including pulmonary and renal fibrosis. Pharmacological blockade of αV-containing integrins by a small molecule (CWHM 12) attenuated both liver and lung fibrosis, including in a therapeutic manner. These data identify a core pathway that regulates fibrosis and suggest that pharmacological targeting of all αV integrins may have clinical utility in the treatment of patients with a broad range of fibrotic diseases.


Integrins are transmembrane receptors involved in cell-cell and cell-matrix signaling pathways. Each integrin is typically formed by the noncovalent pairing of one α subunit, of which 18 types are known to exist, and one β subunit, of which 8 types are known to exist. Together, 24 distinct heterodimers have been identified to date.[1] This observed high variability in dimeric integrin assembly translates into diversity in ligand-binding specificity (most integrin ligands being matrix components) and functional nonredundancy.[2] It is well established that, in multiple organs, integrins can control the release and activation of the important profibrogenic cytokine transforming growth factor beta (TGF-β).[3-5] Importantly, this regulatory effect of integrins on TGF-β activity appears to primarily involve αV-containing integrins. Tissue myofibroblasts, which are the dominant effector cells during fibrosis in multiple organs, express multiple αV-containing integrins.[6] Taken together, these findings suggest that αV integrin itself may be essential for the development of fibrosis, regardless of its associated β subunit.

The goal of the present study by Henderson et al. was to examine the role of αV integrin in liver fibrosis, and specifically to determine whether targeted deletion of this integrin in myofibroblasts prevents fibrosis. For that purpose, the authors used Itgavflox/flox; Pdgfrb-Cre mice to perform experimental models of fibrosis in liver (carbon tetrachloride intoxication [CCl4] for 6 weeks), lungs (bleomycin instillation for 14 and 28 days), and kidneys (unilateral ureteric obstruction [UUO] for 14 days). The approach used here was to induce targeted deletion of αV integrin gene specifically in cells displaying Pdgfrb promoter activity—that is, myofibroblasts in the fibrosing organ. Without wanting to overstate our case, we suggest that this superbly clever methodology (along with the experimental designs applied) makes this work required reading in the liver fibrosis field for 2014.

The authors initially validated Pdgfrb-Cre mice as a model to study and track tissue pericytes or myofibroblasts by crossing them with dual mTmG-GFP[7] and/or single Ail4[8] reporter mice. Using the well-characterized CCl4-induced liver fibrosis model, they showed that Pdgfrb is a reliable marker for liver pericytes and hepatic stellate cells (HSCs), as most reporter-labeled cells expressed Pdgfrb and desmin in normal and fibrotic livers and α-SMA de novo in fibrotic livers. A similar expression pattern for these markers was observed in pericytes from lungs and kidneys, using the bleomycin and UUO fibrosis models, respectively. In aggregate, these findings suggest that the Pdgfrb promoter is an effective way to target myofibroblasts in a variety of organs. The constitutive expression of Pdgfrb in quiescent HCSs was confirmed in Pdgfrb-BAC-eGFP reporter mice. This is an interesting observation, since Pdgfrb had been previously characterized only as an HSC activation marker.[9] Here, use of other myofibroblast markers such as Lrat[10] or Cytoglobin[11] would have helped to better characterize liver Pdgfrb-expressing cells. Indeed, liver myofibroblasts are increasingly recognized as heterogeneous groupings of cells as liver fibrosis progresses.[12, 13] In particular, alpha-smooth muscle actin (α-SMA)-expressing myofibroblasts and collagen-producing myofibroblasts are distinct, but overlapping, populations.[14] Besides myofibroblasts, “vascular wall” cells, likely vascular smooth muscle cells, also express Pdgfrb, desmin, and α-SMA, meaning that vascular wall cells could also contribute functionally to any observed phenotype of Pdgfrb-Cre mice. It is also worth noting that there is no scientific consensus as to the ideal markers or set of markers defining hepatic stellate cells or liver myofibroblasts. That said, in all three fibrosis models, the authors showed increased expression of profibrogenic collagen and α-SMA genes among others in primary tdTomato+-sorted Ail4; Pdgfrb-Cre pericytes isolated from fibrotic tissues, when compared to controls.

The most exciting aspect of the study is the observation that αV integrin is important in the survival and/or function of liver myofibroblasts. The authors demonstrated that αV integrin expression, which gradually increases in primary liver Pdgfrb-labeled HSCs upon culture activation, can be efficiently suppressed in primary liver Itgavflox/flox; Pdgfrb-Cre HSCs in similar conditions. Here, the authors did not describe whether (or to what extent) αV integrin expression is reduced in quiescent Itgavflox/flox; Pdgfrb-Cre HSCs, which harbor an active Pdgfrb. This is a critical point to address, because it is unknown whether this could induce potential phenotypic changes in HSCs, even prior to myofibroblastic differentiation. Next, the authors presented the main finding of the article: targeted deletion of αV integrin in myofibroblasts is protective against organ fibrosis in vivo and in vitro. Histologic analysis of fibrotic liver, lung, and kidney tissues from Itgavflox/flox; Pdgfrb-Cre mice showed significant (although partial) reduction in matrix deposition, myofibroblastic differentiation, and phosphorylation of the TGF-β downstream target Smad3 as compared to controls. Of note, while the chosen time course for bleomycin model is appropriate,[15] the one chosen for the CCl4 liver model appears to be brief.[16] Still, the findings of fibrosis reduction are highly noteworthy.

As noted above, integrins function only as obligate heterodimers, with α and β subunits necessary for function; interestingly, this heterodimer arrangement is preserved even in primitive organisms.[17] Thus, the authors investigated whether β integrins known to be associated with αV affected experimental liver fibrosis. Surprisingly, the answer was (more or less), no. Genetic deletion of the αV-associated integrins β3, 5, 6, or 8 in mice did not confer protection against CCl4-induced liver fibrosis (although deletion of β1 integrin could not be tested in vivo due to a lack of mouse viability). It may be inferred that an array of β; integrins may associate with αV in a way necessary for liver myofibroblast function.

How does αV integrin regulate liver myofibroblast functions? The authors provide in vitro evidence that the answer may be by way of regulation of the TGF-β signal transduction cascade. Specifically, αV integrin deletion in culture-activated Itgavflox/flox; Pdgfrb-Cre HSCs or antibody-mediated αV integrin inhibition in culture-activated tdTomato+-sorted Ail4; Pdgfrb-Cre HSCs led to decreased expression of collagen and α-SMA, but not TGF-β. However, a TGF-β reporter assay demonstrated that suppression of αV integrin induced defective TGF-β/p-Smad3 signaling in Itgavflox/flox; Pdgfrb-Cre HSCs. This suggests that the effect observed did not result from reduced cytokine expression, but rather from diminished integrin-mediated capacity to activate TGF-β. This hypothesis is intriguing and needs to be explored in greater depth.

Is it possible that the findings of this work may be applied in a translational fashion? To address this, the authors tested the effectiveness of a small molecule inhibitor of αV integrin, CWHM12, in liver and lung fibrosis models. In both models, CWHM12 treatment decreased organ fibrosis and impaired TGF-β/p-Smad3 signaling. Thus, it is indeed plausible that an anti-αV integrin approach will be rational to investigate as a means to prevent or treat liver (and kidney and lung) fibrosis. However, caveats are important to consider. The experience of Crohn's disease treatment with the humanized α4 integrin monoclonal blocking antibody natalizumab provides a cautionary tale.[18] Natalizumab presumably works by blocking α4 integrin function(s) in white blood cells and is effective in Crohn's disease treatment.[19] However, natalizumab treatment is associated with a risk of progressive multifocal leukoencephalopathy, as these same white blood cells are known to cross the blood-brain barrier (which is why this drug is effective for treatment of multiple sclerosis).[20-22] Still, we are optimistic that αV-mediated therapy is highly worth pursuing, and we look forward to laboratory refinements to better understand the mechanisms of action.