Matrix metalloproteinase‐2, ‐7, and ‐9 activities in dogs with idiopathic pulmonary fibrosis compared to healthy dogs and dogs with other respiratory diseases

Abstract Background Canine idiopathic pulmonary fibrosis (CIPF) is a chronic, interstitial lung disease that mainly affects West Highland white terriers (WHWTs) and is characterized by excessive deposition of extracellular matrix (ECM) in the lung. Matrix metalloproteinases (MMPs) participate in remodeling of ECM. Objectives To compare metalloproteinase‐2, ‐7 and ‐9 activities in blood or bronchoalveolar lavage fluid (BALF) samples or both of CIPF WHWTs with healthy WHWTs, healthy dogs of other breeds, and dogs with other lung diseases and determine if these MMPs could be used as diagnostic and prognostic markers for CIPF. Animals Forty‐four CIPF WHWTs, 24 dogs with chronic bronchitis (CB), 17 with eosinophilic bronchopneumopathy (EBP), 10 with bacterial pneumonia, 39 healthy WHWTs, and 35 healthy dogs of other breeds. Methods Cross‐sectional observational study. Pro‐MMP and active MMP activities were analyzed by zymography. Results In serum, significantly higher (P < .01) pro‐MMP‐7 activities were observed in CIPF WHWTs compared to healthy dogs of other breeds, dogs with CB and dogs with EBP. In BALF of CIPF WHWTs, both pro‐MMP‐9 and pro‐MMP‐2 activities were significantly higher (P < .01) compared to healthy WHWTs, but these differences were not detected in plasma. The CIPF WHWTs had significantly higher (P < .05) activities of pro‐MMP‐9 compared to dogs with CB and of pro‐MMP‐2 compared to dogs with CB and EBP. No statistically significant prognostic factors were observed in CIPF WHWTs. Conclusions and clinical relevance Serum MMP‐7 and BALF MMP‐2 and ‐9 potentially may be useful diagnostic markers but not prognostic markers for CIPF.


| INTRODUCTION
Matrix metalloproteinases (MMPs) are a family of zinc-containing endopeptidases. They are divided into subgroups based on substrate specificity and structure. 1 Twenty-three different MMPs have been identified in humans. 2 These endopeptidases degrade both extracellular matrix (ECM) proteins, such as collagen and fibronectin, and non-ECM proteins, such as cytokines and growth factors. Matrix metalloproteinases are secreted as inactive pro-forms that require activation to form a functional active enzyme. 1 In healthy tissue, MMP activity is low or absent, but it increases in pathological conditions, including lung diseases such as idiopathic pulmonary fibrosis (IPF) in humans 3 and in physiological repair and remodeling processes. 1 Canine IPF (CIPF) is a chronic, progressive, interstitial lung disease that mainly affects West Highland white terriers (WHWTs) 4,5 and occasionally other terrier breeds. 6 This disease shares many similarities with IPF in humans, such as unknown etiology, clinical signs, and negative impact on survival. 6,7 High-resolution computed tomography (HRCT) and histopathologic findings in CIPF are similar to those of both IPF and another interstitial lung disease in humans, non-specific interstitial pneumonia. 8,9 Both CIPF and IPF are characterized by excessive deposition of ECM components in the lungs. 10,11 In humans, increased genetic expression, zymographic activity, immunoreactivity, or concentration of several MMPs are observed in lung tissue, bronchoalveolar lavage fluid (BALF), and blood of IPF patients. 3 Of these MMPs, MMP-7 is considered 1 of the best blood biomarkers for diagnosis and severity assessment of human IPF patients, 12,13 and it is a useful prognostic biomarker. 14,15 In addition to MMP-7, gelatinolytic MMP-2 and -9 also play a role in tissue remodeling in IPF of humans, 11 and an increase of MMP-9 in BALF reflects rapidly declining lung function. 16 Matrix metalloproteinases, especially MMP-9, are involved in several pulmonary diseases in dogs. In BALF, increased MMP-9 but not MMP-2 activity has been reported in dogs with eosinophilic bronchopneumopathy (EBP), 17 recurrent bronchopneumonia, bronchiectasis, 18 and in induced models of airway inflammation. 19,20 Increased pro-MMP-7 activity has been observed in the renal cortex of dogs with Alport syndrome 21 and in the endometrium of dogs with cystic endometrial hyperplasia and pyometra. 22 In addition, increased MMP-7 expression has been observed in atrial tissue in induced atrial fibrillation in a Beagle dog model. 23 To our knowledge, the only study of MMP-7 in pulmonary diseases of dogs is a microarray study that identified downregulation of MMP-7 and also MMP-9 gene expression in lung tissue of dogs with CIPF. 24 Our aim was to evaluate the diagnostic and prognostic value of MMP-7 in serum and MMP-2 and -9 in plasma and BALF in CIPF WHWTs and to compare findings with healthy dogs and dogs with other common respiratory diseases. We hypothesized that MMP activities would be increased in CIPF WHWTs and that, as in humans, especially blood MMP-7 would be a potential diagnostic and prognostic marker for CIPF.

| Animals
In this cross-sectional observational study, 44 Table S1. Ages of dogs and sample storage times are shown in Table S2. All dogs were pri-

| Sample collection
Serum and plasma samples were separated by centrifugation and stored at −80 C. Bronchoscopy was performed and BALF samples were processed as previously described. 27 In brief, physiological saline (2 mL/kg/lobe divided in 2 aliquots) was used for collection of BALF samples from left and right caudal lung lobes during bronchoscopy.
The supernatant was separated by centrifugation (100g, 10 minutes) and stored at −80 C. were analyzed using optimal substrates in zymography (i.e., casein for MMP-7 and gelatin for MMP-2 and-9). 28 Casein zymography (modified from a previous study 21  The gels then were rinsed, stained with Page Blue Protein staining solution (Thermo Scientific, Vilnius, Lithuania), and destained.

| Statistical analysis
Statistical analyses were performed using SAS System for Windows,     Figure 5A,B.

| Survival and prognostic factor analysis
No statistically significant prognostic factors associated with survival in CIPF WHWTs were identified (Table 1). When only CIPF WHWTs with PaO 2 ≤ 60 mmHg were evaluated, serum pro-MMP-7 activity was significantly associated with increased risk of death in all-

| DISCUSSION
We compared the activities of MMP-7 in blood and MMP-2 and -9 in BALF of CIPF dogs, dogs with other lung diseases, and healthy dogs.
In addition, we evaluated plasma MMP-2 and -9 activities in CIPF and healthy WHWTs. Our findings suggest that especially serum MMP-7 but also BALF MMP-2 and -9 may be potential diagnostic but not prognostic markers for CIPF.
Matrix metalloproteinases are secreted in latent pro-forms. Activation requires removal of a cysteine residue (ie, cysteine switch), which creates fully active MMP with molecular weight approximately 10 kDa lower than for the pro-form. Synthesis and activity of MMPs are regulated at several stages starting from gene transcription and ending with active MMP degradation. 28  when compared with healthy controls, patients with other interstitial lung diseases, or those with chronic obstructive pulmonary disease. 12,14,15,30 In humans, MMP-7 is considered a profibrotic mediator 3 and a potential blood prognostic and diagnostic biomarker for IPF. [12][13][14][15] The profibrotic nature of MMP-7 also is supported by findings in animal models and in vitro. Mice that are MMP-7 deficient are protected from bleomycin-induced lung fibrosis. 31 Matrix metalloproteinase-7 degrades several ECM components and regulates transforming growth factor beta (TGF-β) bioactivity, 32 which is a key mediator in CIPF of WHWTs. 33,34 Inconsistent with our hypothesis, we did not observe an association between PaO 2 (ie, severity of CIPF) and pro-MMP-7 activity in correlation analysis in CIPF WHWTs. This observation might indicate that serum pro-MMP-7 activity has limited utility as a marker of disease severity. This finding is in contrast to IPF in humans, where increasing total MMP-7 concentrations are associated with decreasing lung function. 12 Our BP dogs also had high pro-MMP-7 activities, which may be related to the diverse functions of MMP-7. One function is regulating neutrophil recruitment at the site of acute lung injury. 35 Serum pro-MMP-7 activity was significantly higher in CIPF WHWTs than in healthy dogs of other breeds, but not when compared with healthy WHWTs.
One explanation might be that even though healthy WHWTs were thoroughly examined and thought to have healthy lungs, they may have had subclinical, early phase fibrosis with undetectable findings on HRCT that have not yet caused hypoxemia. In humans, serum MMP-7 is considered potentially valuable for detection of subclinical interstitial lung disease. 30,36 In addition, the possible effect of other organ fibrosis on pro-MMP-7 activity cannot be excluded. In humans, increased blood total MMP-7 concentrations are associated with severe fibrosis of other organs, such as kidney 37 and liver. 38 In our study, necropsy was performed in 27/44 CIPF WHWTs, but only in a few dogs in other groups. Therefore, we were not able to evaluate whether fibrosis in other organs might have affected serum pro-MMP-7 activity. The lack of difference in serum pro-MMP-7 activities between CIPF and healthy WHWTs also might be explained as a breed specific finding. In addition, our study may have been underpowered to identify a significant difference. A previous study observed downregulation of MMP-7 gene expression by microarray in lung samples of dogs with CIPF compared to healthy controls. 24 However, the analysis method and sample type (lung tissue) were different and it is known that posttranscriptional, translational, and degradation regulation affect protein abundance. 39 In addition, the previous used pooled samples, reported of low probe signal intensities and did not confirm results by qRT-PCR. 24 In BALF, pro-MMP-2 activity was significantly higher in CIPF WHWTs compared with other groups (healthy WHWTs, CB, and EBP groups). In a previous study, pro-MMP-2 activity was not significantly higher in dogs with EBP as compared to healthy controls. 17 Increased pro-MMP-2 activity has been detected in BALF of rats with bleomycin-induced fibrosis. 40 In addition, increased total MMP-2 concentrations have been observed in BALF of human patients with IPF compared with healthy controls. 16 The MMP-2 activity may be linked with basement membrane degradation 41 and angiogenesis, 42 which in turn may promote alveolar fibrosis. 16 In humans with IPF, MMP-2 is mainly expressed in reactive airway epithelial cells and myofibroblasts, but also in inflammatory cells such as macrophages. 43 The activity of pro-MMP-9 in BALF was significantly increased in CIPF WHWTs compared with healthy WHWTs and dogs with CB. Total MMP-9 concentrations also increased in BALF in humans with IPF compared to healthy controls. 16 We did not detect a significant difference between CIPF WHWTs and dogs with EBP. Increased pro-MMP-9 and active MMP-9 activity previously has been detected in dogs with EBP compared with healthy controls. 17 Similarly, increased MMP-9 activities have been detected in dogs with recurrent bronchopneumonia and bronchiectasis 18 and in canine models of airway inflammation. 19,20 As was the case with MMP-7, no correlations were found between pro-MMP-2 or pro-MMP-9 activity in BALF and PaO 2 when only the CIPF group was evaluated, indicating that these MMPs also may have limited utility as severity markers. In humans, high BALF total MMP-9 concentrations were found in IPF patients with rapid disease progression in a study with small groups of both diseased and control subjects. 16 Bleomycin-induced pulmonary fibrosis in a rat model showed that MMP-2 and -9 may have important roles in the early phase of the disease. In the later phase, these MMPs also may have a role in the repair process. 40 Active MMP-9 was detected only in 5 CIPF WHWTs and in 5 dogs with EPB, and active MMP-2 was only detected in 1 CIPF WHWT. An experimental study of rats with lung fibrosis found that active forms of MMPs more often are present in lung tissue, where protein degradation occurs, than in BALF. 40 In addition, active MMP in BALF might be too low to be detected by zymography. Matrix metalloproteinase-9 promotes abnormal epithelial repair in humans with IPF because it is mostly expressed in alveolar macrophages, metaplastic airway epithelial cells, and neutrophils 44 and because it activates latent TGF-β. 45 A microarray study observed downregulation of the MMP-9 gene in the lungs of dogs with CIPF. 24 As described earlier, the method and matrix differed from those used in our study.
In contrast to the BALF results, we did not observe any significant difference in plasma pro-MMP-2 and -9 activities between CIPF and healthy WHWTs. Active MMP-9 was detected both in CIPF and healthy WHWTs, but no significant difference was observed between groups with limited numbers of dogs. However, blood and BALF MMP concentrations are not always comparable with each other. 46 In humans, increased blood total MMP-2 and -9 concentrations have been detected in IPF patients compared to healthy controls. 13 Furthermore, increased pro-MMP-9, active MMP-9, and active MMP-2 activity has been observed in patients with severe IPF compared to healthy controls, but no significant difference was observed when patients with moderate IPF were compared with healthy controls. 47 An ANCOVA model of WHWTs showed that sample storage time may decrease pro-MMP-2 and -9 activities in plasma. In humans, plasma MMP-9 concentrations have been shown to be stable after 9 years of storage at −80 C. 48 Conflicting results have been published, 49  WHWTs. 25 We did not observe that MMP activities or PaO 2 were prognostic for CIPF-related death. Another study had similar findings regarding PaO 2 . 25  Our study had some limitations. Some groups were small, especially with respect to plasma samples. Samples were not available for complete analysis both from BALF and blood in all groups.
Because of the preliminary nature of our study investigating MMP activities for the first time in CIPF, power analysis was not performed. Healthy WHWTs may have been in the early stages of CIPF, which was not detected in HRCT, but their inclusion was of interest as a predisposed breed. Additionally, human recombinant proteins and antibodies used in Western blotting have not been validated for dogs.
In conclusion, our results suggest that especially circulating MMP-7 but also BALF MMP-2 and -9 may be potential diagnostic markers for CIPF. However, these MMPs do not seem to be potential prognostic markers for CIPF.

CONFLICT OF INTEREST
Authors declare no conflict of interest.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of this article.