One Molecule, Two Opposite Biological Effects: The Many Faces of Matrix Metalloproteases in the Pathogenesis of Idiopathic Pulmonary Fibrosis

Idiopathic pulmonary fibrosis (IPF) is a chronic lung disease of unknown etiology and limited therapeutic options that remains a leading cause of death among those with interstitial lung diseases. Thus, it is characterized by the unrelenting accumulation of scar tissue, resulting in the destruction of lung architecture and the progressive decline of lung function (1).

The pathogenesis is uncertain, but strong evidence indicates that the aberrant activation of airways and alveolar epithelial cells initiates the development of the disease through the secretion of numerous mediators, including several MMPs (matrix metalloproteinases) (1–3).

MMPs are a family of zinc-dependent matrixins that participate in extracellular matrix degradation but also process and cleave diverse bioactive mediators, such as growth factors, cytokines, and chemokines, playing a critical role in a wide variety of biological and pathological processes (4). From these, a growing body of evidence has demonstrated that MMP-9 is elevated in IPF lungs being expressed by different types of lung cells (4, 5). Outstandingly, this enzyme has a bidirectional relationship with TGF-β1, likely the strongest profibrotic mediator. Thus, Thy-1⁻ fibroblasts, which are usually located in the fibroblast/myofibroblast foci, stimulated by lung epithelium-produced TGF-β1 synthesize MMP-9, and MMP-9 activates latent TGF-β1, contributing to the increase in the pool of active TGF-β1 (4–7).

In this issue of the Journal, Espindola and colleagues (pp. 458–470) evaluated the expression of MMP-9 in IPF airway basal-like cells and the effects of MMP-9 inhibition on fibrotic mechanisms with the hypothesis that targeting this enzyme would attenuate the fibrotic response (8).

First, the investigators aimed to identify the cells expressing MMP-9 in IPF and normal lungs and found a marked increase in the percentage of MMP-9⁺ cells in airway basal-like (ABC-like) cells’ EpCAM⁺ CD45⁻, which also express CCR10. Of note, these cells obtained from IPF lungs displayed a significant increase of upstream transcriptional regulators, including TGF-β1, and of several activated canonical pathways, whereas exogenous treatment with active TGF-β1 markedly increased the expression of MMP-9.

Then, they determined the effect of andecaliximab on the activation of the TGF-β1 pathway (SMAD2 phosphorylation). Andecaliximab is a potent humanized monoclonal antibody that binds MMP-9 at the junction between the propeptide and catalytic domains, preventing the activation of the zymogen (9). Surprisingly, inhibition of MMP-9 resulted in two reproducible opposite responses in the IPF ABC-like cell lines. In some of them, pSMAD2 was reduced (responders), whereas others exhibited an increase in pSMAD2 expression after anti–MMP-9 treatment (nonresponders). As expected, anti–MMP-9–treated responder ABC-like cells exhibited reduced TGFBI, a TGF-β1 inducible gene.

Transcriptional signatures through RNA sequencing identified IFN signaling as the most differentially active canonical pathway in responders treated with anti–MMP-9, whereas IFNA2 was the most highly expressed transcript. Supporting this finding, CXCL10 and CXCL11, two IFN-inducible chemokines, were significantly increased in conditioned media from cultures of responder cells compared with media from nonresponder cells. Interestingly, the addition of exogenous IFNα2 was able to reverse the response to anti–MMP-9 of the nonresponder cells, provoking a marked reduction of SMAD2 phosphorylation.

As a proof of concept, the authors explored the in vivo antifibrotic effects of anti–MMP-9 antibody treatment using a well-characterized humanized nonobese diabetic, severe combined immunodeficient IL-2 receptor γ mouse model (NSG) of lung fibrosis that they had previously described (10).

In this model, mixed explant lung cell suspensions from a nonresponder patient with IPF and a responder patient with IPF were intravenously injected into separate groups of mice. From Day 35 to Day 63 after injection of IPF cells, mice were treated with a mixture of both anti–human and anti–mouse MMP-9 monoclonal antibodies or IgG as control. Anti–MMP-9 treatment was efficacious only in the NSG group that received cells from the responder patient with IPF, whereas the NSG mice that received IPF cells from a nonresponder patient exhibited consistent lung remodeling and foci of alveolar wall thickening. Consistently, pSMAD2 was decreased in the responder group treated with anti–MMP-9 monoclonal antibody compared with the similarly treated nonresponder group. Paralleling the results observed in IPF, NSG mice that received responder IPF cells expressed higher CXCL10.

This study confirms that MMP-9 is increased in IPF and that it likely has a profibrotic effect through TGF-β1 signaling among others. The authors go further, trying to reverse this MMP-9 associated fibrogenic effect with a specific antibody both in vitro and in vivo.

The idea of treating IPF with MMP inhibitors is attractive but challenging. First, given the structural similarities in the catalytic domain of MMPs, assumed inhibitors should be highly selective for a particular MMP target and certainly able to accumulate in the fibrotic lung without eliciting adverse systemic effects. Both of these problems were overcome by Espindola and colleagues because they used a specific anti–MMP-9 antibody and obtained lung improvement in vivo suggesting local action.

Unfortunately, the authors did not analyze the effect of blocking MMP-9 on the expression or activation of other MMPs that are also expressed in IPF epithelium. This is important because it is widely recognized that inhibiting one MMP may provoke a compensatory response that may include enzymes that enhance a fibrotic response (e.g., MMP-7) (11) or have a protecting role (e.g., MMP-19) (12).

In addition, this study exhibits another critical problem with the therapeutic use of molecular targets; some patients may respond, whereas others do not at all, and moreover, as shown in this study, blocking MMP-9 may exacerbate the fibrotic response.

The reason for this paradoxical response is uncertain. The authors propose that, at least in part, the discrepancy could be related to the differential expression of IFNα because the treatment of nonresponder cells with IFNα restored responsiveness to anti–MMP-9.

However, this different response may be related to the variability of the human disease. Unfortunately, there are no data about the patients with IPF included in this study, but they clearly have two subsets of patients that confirm the biological heterogeneity of this disease and the complexity of its temporal and sequential pathogenic mechanisms. Actually, this finding exposes the following crucial question: when and why does the putative clinical benefit of a specific treatment (e.g., anti–MMP-9) sometimes turn into a harmful factor? Undoubtedly, the diverse behavior of IPF will require future individualization of diagnosis and treatment.

Despite these limitations, this study is a step forward to better understand the role of MMP-9 in the pathogenesis of this devastating disease.