Editorial Thorax: first published as 10.1136/thoraxjnl-2015-208204 on 14 January 2016. Downloaded from derived proteases capable of generating Untangling the protease web in COPD: leucocyte chemotaxins and activating TGFβ.12 Measures of small airway disease metalloproteinases in the silent zone were associated with a different MMP sig- nature again, comprising MMP-3, MMP-7, Simon R Johnson MMP-8, MMP-9 and MMP-10. The stron- gest association being with MMP-8. The differing MMP profiles associated with The idea that unregulated protease activity unfortunately the reality is not that small airway obstruction and emphysema underlies the pathogenesis of COPD has straightforward. The rapidly expanding highlight that small airway remodelling is been prominent since the recognition that number of non-ECM MMP substrates has required for airflow obstruction in COPD, genetic loss of α1 antitrypsin led to highlighted a large number of biological independent of loss of elastic recoil due to unregulated neutrophil elastase activity processes modified by MMPs and created emphysema. and premature emphysema.1 Increased potential new opportunities to intervene The challenge is to understand how proteolysis by other proteases, particularly therapeutically in COPD. For example, these tissue-remodelling enzymes contrib- the matrix metalloproteinases (MMPs), proteomic-based interrogation of animal ute to the airway pathology of COPD. has since been implicated in COPD. The models of inflammation has identified This study highlights the need to consider MMPs are a family of over 20 zinc- around 150 discrete protein substrates of the roles of proteases in other aspects of dependent endopeptidases, initially identi- MMP-12 in vivo. Importantly, proteolytic the COPD other than emphysema, par- fied as extracellular matrix (ECM) degrad- processing of these substrates was shown ticularly the small airways that are emer- ing enzymes and classified according to to modify inflammation, cell migration, ging as a critical component of the COPD their ECM substrates. In individual innate and adaptive immunity in addition phenotype. In health, the small airways, studies, genetic association, overexpres- to ECM remodelling.10 These far-reaching defined as those <2 mm contribute sion, and in fewer cases, activity of indi- effects put MMPs and other proteases at <20% of lower airway resistance. The vidual proteases have been associated with the centre of maintenance of normal potential for large increases in small COPD in humans and smoke-induced homeostasis as well as mediators of airway obstruction to remain relatively emphysema in animals. While a number disease. These important protective func- asymptomatic has earned this lung of MMPs and disintegrin and metallopro- tions make the use of MMP protease inhi- compartment the title of the ‘silent teinases have been studied in COPD at bitors a risky therapeutic strategy. For zone’.13 A combination of physiological some level, most interest has focused on example, MMP-12 generates angiostatin and anatomical studies, recently incorpor- MMP-1, MMP-12 and MMP-9. MMP-1 and suppresses neovascularisation and ating micro-CT, suggests that small airway is overexpressed by alveolar macrophages, metastatic lung tumour growth in animals, obstruction can progress relatively type II pneumocytes and airway epithelial and inhibiting this protease in those who undetected initially, preceding emphy- cells. The protein is induced by cigarette smoke is likely to be an unacceptable risk sema, with the development of progres- smoke and is capable of degrading col- in COPD.11 sive airway wall thickening and complete lagens I and III, the most abundant lung In this month’s edition of the journal, loss, due to luminal obliteration, of up to http://thorax.bmj.com/ proteins.23MMP-9 expressed by neutro- Ostridge and colleagues address the associ- 90% of the small airways. In COPD, phils and macrophages is overexpressed in ation of specific aspects of the emphysema increased small airway obstruction may be COPD although MMP-9 activity, rather phenotype; namely small airway disease due to airway wall thickening by ECM than protease expression, is suppressed in and emphysema, with bronchoalveolar deposition and inflammatory cell infiltra- stable patients compared with healthy lavage MMP and cytokine levels in 24 indi- tion, luminal occlusion by mucous and smokers but activated during exacerba- viduals with COPD and 8 smoking or dynamic collapse due to loss of elastic tions.45Although MMP-9 has elastase ex-smoking controls. Emphysema, bron- recoil. However, detailed multivariate on September 27, 2021 by guest. Protected copyright. activity in vitro, its action in COPD may chial wall thickening and small airway analyses of physiological and anatomic be due to generation of prolyl–glycine– obstruction were examined using validated data point towards airway wall remodel- proline (PGP), a collagen-derived neutro- quantitative CT measures. Total MMP ling as the dominant influence on small phil chemotactic matrikine rather than protein and cytokine levels were measured airway resistance.14 Taken together, these elastolysis.6 MMP-12 is an elastase asso- by Luminex multiplex microparticle array. data suggest that the association between ciated with the development of COPD by Consistent with various previous reports, small airway obstruction and MMP-8, a genome-wide association study,7 whose MMP-1, MMP-2, MMP-3, MMP-8, neutrophil-derived collagenase, is likely to presence and activity in sputum are asso- MMP-9 and MMP-10, although not be related to MMP-8’sprofibrotic actions. ciated airflow obstruction8 and in which MMP-12, were elevated in COPD when In vitro, MMP-8 is a potent type I single-nucleotide polymorphisms that compared with controls. The interest in this collagen-degrading protease, yet despite reduce enzyme activity are associated with study lies with the associations between its collagenase activity, which might be protection from emphysema in COPD.9 protein levels and specific aspects of COPD expected to reduce lung fibrosis in While the ‘classical’-matrix-degrading pathology. Emphysema in COPD was asso- response to injury: in murine airways, actions of these proteases can be easily ciated with elevation of MMP-3, MMP-7 MMP-8 inactivates macrophage inflamma- related to the pathogenesis of COPD, and MMP-10; interestingly, both MMP-3 tory protein-α in allergic airway inflam- and MMP-10 are capable of activating mation and contributes to the generation MMP-7 whereas MMP-10 itself is activated of the neutrophil chemoattractant PGP.15 Correspondence to Professor Simon R Johnson, by neutrophil elastase. Airflow obstruction, MMP-8 also promotes lung fibrosis in Division of Respiratory Medicine and Respiratory Research Unit, Queens Medical Centre, University measured using FEV1, however was asso- response to bleomycin in the mouse and of Nottingham, Nottingham NG7 2UH, UK; ciated with MMP-8, MMP-9 and MMP-12, reduces lung inflammation by cleaving [email protected] predominantly neutrophil and macrophage- interleukin-10.16

Johnson SR. Thorax February 2016 Vol 71 No 2 105 Editorial Thorax: first published as 10.1136/thoraxjnl-2015-208204 on 14 January 2016. Downloaded from The study by Ostridge and colleagues need now is to incorporate protein activity in the distal MMP-1 promoter. Am J Respir Cell Mol advances the field by examining multiple and localisation into these studies. Biol 2009;40:4–12. 4 Lowrey GE, Henderson N, Blakey JD, et al. MMP-9 proteases in patients with detailed pheno- Complementary approaches to examine protein level does not reflect overall MMP activity in type data. However, the multilevel regula- the turnover of protease targets can also the airways of patients with COPD. Respir Med tion of proteases in vivo including the provide data on overall protease activity. 2008;102:845–51. presence of endogenous inhibitors dictates Measurement of ECM protein fragments 5 Cane J, Mallia-Milanes B, Forrester D, et al. Matrix that to understand protease activity at a cel- produced by specific proteases reflect metalloproteinases -8 and -9 in the airways, blood and urine during exacerbations of COPD. COPD. lular, spatial and temporal level, this type changes in ECM degradation and depos- Published Online First: 29 Sep 2015. doi:10.3109/ of detailed protein expression and pheno- ition. Interestingly, these markers can 15412555.2015.1043522. type data need to be analysed together predict disease progression in lung fibrosis 6 Weathington NM, van Houwelingen AH, Noerager with precise localisation and overall activ- and in COPD exacerbations, show both BD, et al. A novel peptide CXCR ligand derived from extracellular matrix degradation during airway ity data for different protease classes. increases in collagen and elastin turnover inflammation. Nat Med 2006;12:317–23. Furthermore, the redundancy among sub- and simultaneous de novo synthesis of 7 Hunninghake GM, Cho MH, Tesfaigzi Y, et al. strate–protease interactions, of which the other ECM proteins.18 19 MMP12, lung function, and COPD in high-risk MMPs are only a subset, makes the signifi- The eventual aim of such work is to populations. N Engl J Med 2009;361:2599–608. cance of these data harder to interpret. For improve targeting of injurious tissue 8 Chaudhuri R, McSharry C, Brady J, et al. Sputum matrix metalloproteinase-12 in patients with chronic example, MMP-8 and neutrophil elastase remodelling and other emerging protease- obstructive pulmonary disease and asthma: are generally colocalised and have multiple dependent processes in COPD. relationship to disease severity. J Allergy Clin overlapping substrates. The higher potency Broadspectrum MMP inhibitors in cancer Immunol 2012;129:655–63.e8. of neutrophil elastase for most of these were not efficacious and led to unforeseen 9 Haq I, Lowrey GE, Kalsheker N, et al. MMP-12 SNP affects MMP activity, lung macrophage infiltration substrates makes the expression and contri- side effects. The complexity of protease and protects against emphysema in COPD. Thorax bution of MMP-8 in COPD more difficult and inhibitor networks, redundancy 2011;66:970–6. to interpret. The next steps will be to among protease functions and the mul- 10 Bellac CL, Dufour A, Krisinger MJ, et al. Macrophage describe how these individual changes in tiple beneficial effects of MMPs at differ- matrix metalloproteinase-12 dampens inflammation and neutrophil in flux in arthritis. Cell Rep 2014;9: protease expression and activity contribute ent sites at different times currently make – fi 618 32. to biological changes at speci c anatomical the safe targeting proteases unworkable. 11 Houghton AM, Grisolano JL, Baumann ML, et al. and cellular sites at specific times. Targeting specific processes downstream Macrophage elastase (matrix metalloproteinase-12) Emerging proteomic techniques have of these proteases in the future is likely to suppresses growth of lung metastases. Cancer Res begun to identify multiple previously be safer and more achievable. 2006;66:6149–55. 12 Yu Q, Stamenkovic I. Cell surface-localized matrix unrecognised MMP substrates in vivo, Competing interests None declared. metalloproteinase-9 proteolytically activates revealing an increasingly complex ‘web’ of TGF-[beta] and promotes tumor invasion and protease interactions.10 While MMPs acti- Provenance and peer review Commissioned; angiogenesis. Genes Dev 2000;14:163–76. vate or inactivate individual proteins such internally peer reviewed. 13 Macklem PT, Mead J. Resistance of central and as cytokines whose actions may be predict- peripheral airways measured by a retrograde catheter. J Appl Physiol 1967;22:395–401. able, altering the activity of a single prote-

14 Hogg JC, Chu F, Utokaparch S, et al. The nature of http://thorax.bmj.com/ ase also degrades multiple protease small-airway obstruction in chronic obstructive inhibitors and activates multiple propro- pulmonary disease. N Engl J Med 2004;350:2645–53. teases affecting their substrates and their To cite Johnson SR. Thorax 2016;71:105–106. 15 Quintero PA, Knolle MD, Cala LF, et al. Matrix metalloproteinase-8 inactivates macrophage targets and so on. For example, proteomic inflammatory protein-1 alpha to reduce acute lung analysis of pharmacological inhibition of inflammation and injury in mice. J Immunol MMP-14 in a cell line affected 30 other 2010;184:1575–88. proteases and protease inhibitors all of 16 García-Prieto E, González-López A, Cabrera S, et al. fi which may have had downstream effects.17 ▸ http://dx.doi.org/10.1136/thoraxjnl-2015-207428 Resistance to bleomycin-induced lung brosis in MMP-8 deficient mice is mediated by interleukin-10. on September 27, 2021 by guest. Protected copyright. This complexity makes interpreting the Thorax 2016;71:105–106. PLoS One 2010;5:e13242. biological consequences of changes in indi- doi:10.1136/thoraxjnl-2015-208204 17 Butler GS, Dean RA, Tam EM, et al. vidual enzymes difficult. Combining the Pharmacoproteomics of a metalloproteinase detailed protein level data reported by hydroxamate inhibitor in breast cancer cells: dynamics of membrane type 1 matrix Ostridge et al with techniques such as in REFERENCES 1 Laurell C-B, Eriksson S. The electrophoretic α; metalloproteinase-mediated membrane protein situ zymography to localise changes in 1-globulin pattern of serum in α; 1-antitrypsin shedding. Mol Cell Biol 2008;28:4896–914. overall proteolytic activity at specific ana- deficiency. Scand J Clin Lab Invest 1963;15: 18 Jenkins RG, Simpson JK, Saini G, et al. Longitudinal tomical locations should be the next phase 132–40. change in collagen degradation biomarkers in fi of protease research to understand tissue 2 Imai K, Dalal SS, Chen ES, et al. Human collagenase idiopathic pulmonary brosis: an analysis from the prospective, multicentre PROFILE study. Lancet Respir remodelling in disease. Large clinical (matrix metalloproteinase-1) expression in the lungs of patients with emphysema. Am J Respir Crit Care Med 2015;3:462–72.. cohorts are allowing detailed phenotyping Med 2001;163(3 Pt 1):786–91. 19 Sand JM, Knox AJ, Lange P, et al. Accelerated of those with COPD and the correlation of 3 Mercer BA, Wallace AM, Brinckerhoff CE, et al. extracellular matrix turnover during exacerbations of genetic and protein expression data: the Identification of a cigarette smoke-responsive region COPD. Respir Res 2015;16:69.

106 Johnson SR. Thorax February 2016 Vol 71 No 2