Over the last few years there has been increasing interest in the matrix metalloproteinases (MMPs) in asthma. The MMPs are a family of over 20 zinc dependent endopeptidases with a range of substrate specificities. Although originally classified into gelatinases, stromolysins, collagenases, and matrilysins by their ability to cleave extracellular matrix components, it is now clear that MMPs have an increasing range of other functions. Of potential relevance to asthma is the release and activation of growth and angiogenic factors including transforming growth factor β (TGFβ)¹ and vascular endothelial growth factor (VEGF),² apoptosis by release of Fas ligand,³ and activation of cell surface receptors—collectively resulting in effects on cell differentiation, survival, proliferation, and migration.⁴ MMP activity is tightly regulated at several levels. Gene transcription is enhanced by growth factors, cytokines and adhesion molecules and downregulated by TGFβ and corticosteroids.⁵ MMPs are secreted as zymogens which require activation by proteolytic cleavage either at the cell surface or in solution by other proteases including MMPs.⁵ Once activated, natural inhibitors including tissue inhibitors of metalloproteinases (TIMPs) bind different MMPs with varying specificities in a 1:1 ratio to inhibit proteolytic activity.⁶ There are four TIMP proteins which, in addition to their antiprotease actions, also have other functions—with TIMP-1 modulating proliferation, apoptosis, and angiogenesis in various tumour cell types.^7,8

MMPS AND TIMPS IN ASTHMA

It has been suggested that, in asthma, MMPs contribute to altered extracellular matrix turnover, influx of proinflammatory cells, epithelial repair, and angiogenesis.⁹ In animal models of asthma, allergen challenge in MMP-9 and MMP-2 knockout mice results in reduced numbers of inflammatory cells in the airways compared with wild type mice due to deranged chemokine gradients.¹⁰ The role of MMPs in asthma is complex as, paradoxically, in MMP-2 knockout mice, despite a reduction in airway inflammatory cells, there is an increase in parenchymal inflammatory cells due to impaired cell egress.¹¹ These and other data suggest specific roles for different MMPs, despite apparently equivalent substrate specificities. This level of complexity, brought about by differential activity of individual proteases both spatially and temporally within the lung, may explain the lack of efficacy of broad spectrum MMP inhibitors in models of asthma. In human asthma most studies have measured MMP and TIMP protein levels and estimated activity by the presence of the active protein or by the MMP/TIMP ratio. This approach may not truly represent overall protease activity as other MMPs and MMP inhibitors may be present (including TIMPs, α-macroglobulin, and thrombospondins⁷) which are undetected.¹² These reservations notwithstanding, studies suggest that MMP-9 is the predominant airway MMP and is increased in the sputum¹³ and bronchoalveolar lavage fluid¹⁴ of patients with asthma and further increased during exacerbations.¹⁵ The precise relationship between MMP-9 and TIMP-1 (the major inhibitor of MMP-9) levels in asthma is still being evaluated. One study has shown that the MMP-9/TIMP-1 ratio and specific MMP-9 activity is increased in asthma,¹⁶ while others have shown that a reduction in the MMP-9/TIMP ratio was associated with worsening airflow obstruction in asthma and increased airway wall thickness, suggesting that lower MMP-9/TIMP-1 ratios may be associated with a profibrotic environment with reduced extracellular matrix turnover leading to airway remodelling.¹⁷ The recent association of ADAM 33 (A Disintegrin And Metalloprotease) with asthma and bronchial hyperresponsiveness¹⁸ has further turned the focus of asthma research toward the MMPs and related proteases. The association of ADAM 33 with asthma is made all the more intriguing by the findings that most ADAM 33 transcripts expressed in fibroblasts do not contain the metalloproteinase domain, which suggests that non-proteolytic activities of the MMP/TIMP proteins may also be of relevance in asthma.¹⁹

ASSOCIATION OF A NOVEL TIMP-1 POLYMORPHISM WITH ASTHMA IN WOMEN

It is timely that, in this issue of Thorax, Lose et al²⁰ add to the evidence for involvement of MMPs and TIMPs in asthma with an association study of MMP-9 and TIMP-1 polymorphisms in 543 patients with asthma. Although polymorphisms in MMP-9 were not associated with asthma in this population, a novel TIMP-1 polymorphism in the MMP-9 binding domain was associated with asthma, but not atopy, in women. This provides further evidence that TIMP-1, and possibly the MMP-9/TIMP-1 axis, are involved in the development of asthma rather than merely being deranged in the disease. The 536C>T polymorphism in exon 6 of TIMP-1 does not result in an amino acid change but is predicted to affect RNA splicing, possibly leading to the expression of different TIMP-1 splice variants. What is not clear is how this polymorphism affects an individual’s susceptibility to asthma. As the polymorphism is present in the MMP-9 binding site, it is tempting to speculate that TIMP-1 inhibitory activity against MMP-9 is impaired, possibly allowing enhanced airway inflammatory cell recruitment in response to allergen. The investigators did not examine the level of TIMP-1 in the airways of their patients, the presence of TIMP-1 splice variants, or the function of this polymorphism. Genetic association studies of this type are frequently performed in asthma and have identified a large number of apparently linked gene loci which have not been reproduced by other groups. The strengths of the current study, however, are that the investigators have used a large population of patients with a rigorously characterised asthma phenotype and have studied a set of polymorphisms in biologically relevant genes known to be linked to other diseases categorised by deranged MMP/TIMP activity including atherosclerosis²¹ and emphysema.²² Furthermore, the relationship between the 536C>T polymorphism and asthma in women is also supported by a TIMP-1 haplotype association with asthma. However, it remains essential to confirm these findings in other asthma populations and to study the function of the newly identified polymorphism with regard to both its anti-MMP activity compared with the wild type protein and non-antiprotease TIMP functions including cellular growth and apoptosis.

With this study, Lose et al have implicated TIMP-1 as a potential player in the development of the asthmatic phenotype. It is now essential that we look beyond MMP/TIMP levels toward specific protease activity and non-protease functions of these molecules to understand if selectively affecting their function could be an effective treatment for asthma.