MMP-1 and ADAM10 as Targets for Therapeutic Intervention in Idiopathic Pulmonary Fibrosis

Abstract

Idiopathic pulmonary fibrosis (IPF), a fatal disease characterized by excessive matrix degradation and fibrosis, destroys the lung architecture and results in the inability of the lungs to absorb oxygen. The cause(s) of IPF is unknown and current treatments are palliative. Matrix metalloproteinases (MMPs) and A Disintegrin And Metalloproteinases (ADAMs) likely play roles in IPF progression. However, specific MMPs and ADAMs in IPF have not been identified due to challenges in MMP/ADAM profiling. We employed a designer affinity resin that binds exclusively to the active forms of MMPs and ADAMs and found by mass spectrometry higher levels of active MMP-1, ADAM9, ADAM10, and ADAM17 in lung tissues of IPF patients. Inhibition of MMP-1 and ADAM10 with the small-molecule inhibitor GI254023X in an in vitro lung fibrosis assay decreased the profibrotic protein α-smooth muscle actin (α-SMA). Our results indicate that inhibition of MMP-1 and ADAM10 may hold promise in treatment of IPF.

Idiopathic pulmonary fibrosis (IPF) is a fatal disease characterized by out-of-control matrix degradation and fibrosis, leading to irreversible loss of the lung’s capacity to absorb oxygen. Approximately 128 000 individuals in the United States have IPF at any given time, with 48 000 new cases occurring annually.¹ Individuals with IPF die within 3 years after diagnosis. This amounts to 40 000 deaths each year in the United States alone,¹ the same number as the annual deaths from breast cancer. The cause(s) of IPF remains unknown, and there are no biomarkers for the disease. Pirfenidone, an anti-inflammatory drug, and nintedanib, an angiogenesis inhibitor, are U.S. Food and Drug Administration approved drugs for IPF. However, these drugs treat the symptoms and not the molecular basis of the disease. The pathology of IPF is thought to result from abnormal alveolar epithelial cells that induce fibroblast-to-myofibroblast transition, which produces disproportionate amounts of extracellular matrix components, including proteolytic enzymes. This contributes to scarring and destruction of the lung architecture,² culminating into what is referred to as honeycomb lung, which is the terminal stage of the ailment.

Mechanistically and evolutionarily related proteinases matrix metalloproteinases (MMPs) and ADAMs (A Disintegrin And Metalloproteinases), are among the aforementioned extracellular components. MMPs constitute a family of zinc-dependent endopeptidases (24 MMPs in humans) that restructure the extracellular matrix. MMPs exist in three forms: proMMPs (inactive), active MMPs, and TIMP (tissue inhibitor of matrix metalloproteinase)-complexed MMPs (inactive). Only the active forms of MMPs play catalytic roles in disease pathology and repair. Dysregulation and secretion of MMPs by stimulated inflammatory cells are thought to be involved in the pathophysiology of IPF,³ but the exact mechanism has not been elucidated. The genes for MMP-1, MMP-2, and MMP-7 have been proposed to be expressed in IPF.⁴ MMP-9 has been shown to be present in lungs and bronchoalveolar lavage fluid from patients with rapid progression of IPF.⁴ However, the methods used to identify the MMPs do not distinguish among the MMP forms. Thus, the actual active MMPs involved in IPF have not been identified to date. Similarly, ADAMs are a family of 29 mammalian membrane-anchored zinc proteinases. ADAMs’ main understood roles are to process membrane-bound precursors, shedding various proteins. ADAMs play roles in diverse physiological events, such as cell adhesion, fusion, and migration. ADAMs have been implicated in a number of diseases, including cancer, diabetes, and Alzheimer’s. There are limited studies that have reported on ADAMs (ADAM8 and ADAM15) in IPF.⁵ Mast cells, which are typically present in fibrotic lungs, have been found to have increased levels of ADAM9, ADAM10, and ADAM17.⁶ However, the consequence of the upregulation of these ADAMs in IPF has not been investigated. To compound the difficulty, the very three forms mentioned above for MMPs also exist for ADAMs.

The levels of MMPs and ADAMs—especially for the active forms—in virtually none of these earlier studies have been studied at the protein level. To address this paucity of information, we synthesized a designer affinity resin with a covalently attached broad-spectrum inhibitor for MMPs and ADAMs (colored magenta in Figure 1A) to Sepharose resin (colored blue). The inhibitor attached covalently to Sepharose resin is a broad-spectrum inhibitor based on the structure of batimastat (Supplementary Table S1). Only the active forms of MMPs and ADAMs bind to this affinity resin. The presence of the prodomain in the zymogenic forms and TIMP in the TIMP complexes blocks the active sites so these forms cannot bind to the affinity resin (Figure 1B).

Figure 1.

(A) Structure of the affinity resin containing a broad-spectrum MMP/ADAM inhibitor (magenta) covalently attached to Sepharose (blue). (B) The affinity resin binds only to active MMPs/ADAMs and not to proMMPs/proADAMs or TIMP-MMPs/ADAMs complexes. After incubation of homogenized tissue with the affinity resin, cysteines in MMPs/ADAMs are reduced with dithiothreitol (DTT), followed by alkylation with iodoacetamide, digestion with trypsin, and mass spectrometry for identification and quantification of active MMPs/ADAMs.

We previously used this affinity resin to identify active MMP-8 and MMP-9 in human diabetic foot ulcer tissues.⁷ We documented that active MMP-9 levels increase with severity of the disease and with infection of the diabetic foot ulcers, thus validating MMP-9 as a target for therapeutic intervention in diabetic foot ulcers.⁸ We showed that MMP-8 was involved in the normal tissue repair processes.⁸ The IPF study undertaken herein was for the same purpose, to identify culprit proteinases involved in progression of IPF. We analyzed lungs from patients in early-, middle-, and late-stage IPF using our affinity resin, coupled to proteomics analysis. Briefly, the tissue homogenates were incubated with the resin. The bound proteins were subjected to reduction of cysteines, alkylation of cysteines, and trypsin digestion. This analysis identified six active proteinases that bound to the affinity resin. They are the active forms of MMP-1, MMP-8, MMP-14, ADAM9, ADAM10, and ADAM17, the levels of which within the lung tissue were quantified using calibration curves with known amounts of synthetic peptides for the proteinases (Figure 2). There were no discernible differences in the levels of active MMPs/ADAMs by early-, middle-, or late-stage IPF (Supplementary Figure S1). That is to say, as the disease advances, the levels of these active enzymes neither increase nor decrease as a function of the stage of the disease.

Figure 2.

Higher levels of MMP-1, ADAM9, ADAM10, and ADAM17 in IPF lungs. Concentrations of (A) active MMP-1, (B) active MMP-8, (C) active MMP-14, (D) active ADAM9, (E) active ADAM10, and (F) active ADAM17. IPF lung tissue (n = 26 patients) and healthy control lungs (n = 4) were analyzed with the affinity resin and mass spectrometry; mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001 relative to control by Student’s t test with two-tail distribution and unequal variance.

The Lung Tissue Research Consortium generously supplied the diseased lung samples studied by us. Each tissue was supplied with numerous parameters that were measured on the donor, including force-vital capacity (FVC) and diffused capacity of the lungs for carbon monoxide (DLCO). The severity of IPF is classified by FVC (Table 1), the maximum amount of air expelled from the lungs after a full inhalation.⁹ However, there is no agreed upon parameter staging IPF severity.¹⁰ We also used DLCO (Table 1) to classify IPF severity. DLCO < 39 signifies advanced IPF, whereas a value of >40 is for limited IPF.⁹ No differences in the levels of active MMPs/ADAMs were observed in advanced IPF compared to limited IPF (Supplementary Figure S2). Therefore, we grouped the results for all IPF lung samples together (Figure 2). MMP-1 was present at 5-fold higher concentrations in IPF (97.0 ± 0.3 fmol/mg) relative to the healthy lung control (17.7 ± 5.3 fmol/mg, p = 0.0006, Figure 2A). In contrast, concentrations of MMP-8 and MMP-14 were lower in IPF lungs at 0.19 ± 0.02 fmol/mg (p = 0.006) and 29.6 ± 2.4 fmol/mg (p = 0.01), respectively, compared to 13.3 ± 1.9 fmol/mg and 40.5 ± 2.7 fmol/mg, respectively, in the healthy lung control (Figures 2B and C). ADAM9 was present at 18.0 ± 5.3 fmol/mg in the healthy lung control and 24-fold higher in IPF at 435 ± 28 fmol/mg (p < 0.001, Figure 2D). Levels of ADAM10 were 8.2 ± 2.7 fmol/mg in the healthy lung control and 303 ± 15 fmol/mg in IPF (37-fold higher, p < 0.001, Figure 2E). ADAM17 was found at 7.6 ± 2.8 fmol/mg in the healthy lung control compared to 71.9 ± 0.6 fmol/mg in IPF (9-fold higher, p < 0.001, Figure 2F). We underscore that these proteins are catalytically competent enzymes and hence even small differences in quantity might play an outsized role in the disease processes. As no differences in quantities of active MMP/ADAM as a function of the disease severity were discernible by FVC or DLCO, one explanation could be due to the difficulty of detecting IPF at an early stage. The symptom of early IPF is merely exercise-related breathlessness, and no specific laboratory abnormalities are detectable.¹¹ Alternatively, FVC or DLCO might not be a good indicator of progression of the disease at the molecular level. In fact, FVC has been shown to not predict IPF severity in a multivariate analysis.¹² This might be consistent with IPF typically not being diagnosed for up to 2 years before the onset of breathlessness.¹³ It is conceivable that dysregulation of MMPs/ADAMs occurs regardless of the stage of the disease and its severity is accompanied by inexorable accumulating of scarring of the lungs.

Table 1. Lung Samples Demographics.

classification by FVC
number of lung samples early stage	6
age range early stage (years)	56–76
average age ± SEM early stage	64 ± 3
gender early stage	1M/5F
FVC early range	80–96
FVC early ± SEM	86 ± 2
number of lung samples middle stage	10
age range middle stage (years)	53–66
average age ± SEM middle stage	59 ± 2
gender middle stage	8M/2F
FVC middle range	50–69
FVC middle ± SEM	58 ± 2
number of lung samples late stage	10
age range late stage (years)	37–62
average age ± SEM late stage	53 ± 3
gender late stage	8M/2F
FVC late range	26–45
FVC late ± SEM	38 ± 2

classification by DLCO (diffuse capacity for carbon monoxide)
number of lung samples advanced IPF	17
age range advanced IPF (years)	37–76
average age ± SEM	57 ± 2
gender advanced IPF	12M/5F
DLCO range advanced IPF	9–38
DLCO advanced IPF ± SEM	24 ± 2
number of lung samples limited IPF	5
age range limited IPF (years)	59–68
average age ± SEM	62 ± 2
gender limited IPF	1M/4F
DLCO range limited IPF	41–80
DLCO limited IPF ± SEM	61 ± 7

A study by Lagares et al. found mRNA for MMP-1, MMP-2, and MMP-14 and ADAM9, ADAM10, ADAM12, and ADAM17 in healthy human lung fibroblasts.¹⁴ siRNA knockdown of these individual proteinases significantly lowered sEphrin-B2 only for siRNA-mediated knockdown of ADAM10, indicating that ADAM10 is involved in ephrin-B2 shedding.¹⁴ This study measured mRNA levels and not protein concentrations. We did not observe active MMP-2 or ADAM12 at the protein level in lung tissues from IPF patients with our affinity resin and proteomics approach. Using a shotgun mass spectrometry approach, decreased abundance of ADAM9 was found in IPF lungs compared to control lungs.¹⁵ However, this study was small (5 control and 6 IPF lung samples). Another issue is that the use of sodium dodecyl sulfate (SDS) in protein extraction would have dissociated the TIMP-ADAM9 complex, so conclusions on the status of the presence of active enzymes (uncomplexed by TIMP) could not be drawn.

Of the identified active proteinases in IPF lungs, MMP-8 and MMP-14 were present in lower concentrations in diseased lungs compared to healthy control tissue. This suggests that elevated activities of these proteinases are not implicated in IPF. One distinct possibility might be a role in the lung’s response in repair of the damage in IPF. Whereas the diminution of the level of MMP-14 was small (1.4-fold), though statistically significant, that of MMP-8 was profound at 70-fold. In an earlier study, we have shown that MMP-8 facilitates healing in diabetic wounds, and one wonders if a similar role might apply in IPF lungs.^16,17 With diminution of MMP-8 in the diseased tissue being as much as 70-fold, if indeed this enzyme would engage in repair mechanisms of damage in IPF, the repair function would be undermined. Notwithstanding the smallness of the effect in the production of MMP-14 in IPF lungs, we note that MMP-14 (also known as membrane type1 (MT1)-MMP) has been shown recently to mediate DNA damage repair.¹⁸ Furthermore, our results of decreased MMP-14 levels in human IPF lungs are in agreement with a recent study in which MMP-14 deficiency exacerbated fibrosis in bleomycin-induced IPF.¹⁹ The proof of whether MMP-8 and/or MMP-14 perform mitigation functions in IPF—a “yin-yang” cellular struggle against the detrimental enzyme(s)—should await future studies. However, this is exactly the case in diabetic wound healing, where MMP-8 has a salutatory effect in wound repair and MMP-9 is detrimental (the effect of pathology of the disease).^16,17

A characteristic of IPF is the accumulation of myofibroblasts, which synthesize α-smooth muscle actin (α-SMA), a profibrotic protein that enhances fibroblast contractile activity.²⁰ In IPF, there is excessive deposition of extracellular matrix components that lead to loss of pulmonary function. Transforming growth factor (TGF)-β1 is a potent inducer of the production of extracellular matrix components.²¹ A major source of myofibroblast accumulation is fibroblast to myofibroblast transition.²² TGF-β1 induces fibroblast to myofibroblast transition through Smad3.²³ Smad3 knockout mice are resistant to TGF-β1 induced pulmonary fibrosis and attenuate bleomycin-induced pulmonary fibrosis.^24,25 Overexpression of TGF-β1 with an adenovirus vector results in prolonged and severe fibrosis in rat lungs²⁶ and in mouse lungs.²⁷ We used a fibroblast to myofibroblast transition assay as an in vitro fibrosis assay using lung fibroblast cells and TGF-β1 to induce fibrosis, followed by staining with α-SMA, imaging, and fluorescence quantification.²⁸ Accordingly, addition of TGF-β1 increased α-SMA levels 1.7-fold from (8.2 ± 0.4) × 10⁵ to (13.6 ± 1.3) × 10⁵ (p = 0.016, Figure 3A and B).

Figure 3.

In vitro lung fibrosis assay. Human lung fibroblasts were seeded and incubated with 20 ng/mL TGF-β1 in the presence of 0, 25, 100, and 500 nM GI254023X. The cells were fixed, incubated with anti--SMA antibody, and probed with goat antimouse IgG secondary antibody, with DAPI for nuclear staining. (A) Representative images of human lung fibroblasts show increased α-SMA upon treatment with TGF-β1, which decreased with increasing concentrations of GI254023X. (B) Quantification of α-SMA, n = 6 per group, *p < 0.05, **p < 0.01, ***p < 0.001 relative to TGF-β1; ^##p < 0.01 relative to control; Student’s t test with two-tail distribution and unequal variance.

An MMP-1 inhibitor (the paper by Martin et al. referred to this as compound 9q) with IC₅₀ values of 6 nM for MMP-1, 300 nM for MMP-2, 400 nM for MMP-3, 10 nM for MMP-8, and 40 nM for MMP-13 has been reported.²⁹ We synthesized this inhibitor according to the literature, but to our surprise it did not inhibit MMP-1. As a selective inhibitor of MMP-1 is not available, we used the commercially available GI254023X for simultaneous inhibition of MMP-1 and ADAM10; IC₅₀ values reported for this compound were 108 nM for MMP-1, 280 nM for ADAM9, 5.3 nM for ADAM10, and 541 nM for ADAM17.^28,30 Prior to using GI254023X, we determined the values for the inhibition constants (K_i) for MMP-1 and ADAM10 at 131 ± 82 and 16.4 ± 0.6 nM, respectively. Whereas IC₅₀ could be used to approximate the value of the dissociation constant for the inhibitor (K_i), we accepted these values as sufficiently close to each other. Treatment with 25 nM GI254023X, a concentration that would inhibit ADAM10, reduced α-SMA fluorescence to (7.5 ± 0.5) × 10⁵ (p = 0.014 relative to TGF-β1). At a concentration of 100 nM GI254023X, 6-fold its K_i value for ADAM10, α-SMA fluorescence decreased further to (3.7 ± 0.3) × 10⁵ (p = 0.0004 relative to TGF-β1). At 500 nM GI254023X, 4-fold above the K_i value for MMP-1 and 30-fold the K_i value for ADAM-10, α-SMA fluorescence was (3.0 ± 0.1) × 10⁵ (p = 0.0005 relative to TGF-β1). These results reveal that as the two enzymes are inhibited, fluorescence of α-SMA—a profibrotic protein—is consistently decreased. Our results indicate that there might be merit to targeting MMP-1 and ADAM10 for inhibition in attenuating progression of fibrosis in IPF. ADAM9 has been found to increase in experimentally induced acute lung inflammation, and ADAM knockout mice showed protection against bleomycin-induced lung inflammation.³¹ However, a role for ADAM9 in chronic lung inflammation remains to be determined.

The ADAM10 inhibitor GI254023X has been shown to inhibit TGF-β1-induced incorporation of α-SMA into stress fibers in human lung fibroblasts.¹⁴ Accordingly, bleomycin-induced lung-injured mice administered 200 mg/kg GI254023X intraperitoneally (once a day for 14 days) showed significantly reduced mortality.¹⁴ Consistently, we found that inhibition of MMP-1 and ADAM10 with GI254023X decreases the profibrotic protein α-SMA. The present work highlights that inhibition of MMP-1 and/or ADAM10 might prove efficacious in addressing the root causes of IPF.

Methods

Affinity Resin

The affinity resin was synthesized for this study as described previously.³²

Human Lung Samples

Flash-frozen lung tissues from patients in early-, middle-, and late-stage IPF were obtained from the Lung Tissue Research Consortium. The samples were stored at −80 °C until analysis. Flash-frozen lung tissues from healthy control humans were purchased from ILSBio LLC (Chestertown, MD). The human IPF and control human samples were received without identification information. This research was reviewed and approved by the University of Notre Dame Institutional Review Board. Demographics of the patient samples are included in Table 1.

Lung Sample Processing

Lung samples (10 mg) were weighed and homogenized in 100 μL of cold lysis buffer (25 mM Tris-HCl pH 7.5, 100 mM NaCl, 1% v/v Nonidet P-40) and protease inhibitors, with the exception of metalloproteinase inhibitors (EDTA). Homogenates were centrifuged at 15 000g for 10 min at 4 °C, and the supernatants were stored at −80 °C. The tissue extracts were mixed with 100 μL of the affinity at 4 °C for 2 h. After centrifugation (15 000g, 1 min), the supernatant was removed, the resin beads were thoroughly washed with carbonate-bicarbonate (CB) buffer, and the resin-bound proteins were reduced at 65 °C for 20 min with 10 mM dithiothreitol. Iodoacetamide was added to a final concentration of 10 mM, and the alkylation was performed at room temperature for 20 min in the dark. Trypsin (0.2 μg) was added and digestion was performed for 18 h at 37 °C. Following trypsin digestion, samples were desalted using Millipore ZipTip C18 (EMD Millipore Corp., Billerica. MA) and concentrated to dryness on a miVac concentrator (Genevac Ltd., Suffolk, UK).

The residue was resuspended in 12 μL of water containing 1% formic acid and the internal standard (digested yeast enolase at a final concentration of 150 fmol/mg tissue). A 2 μL aliquot of the samples was analyzed on a nanoLC BEH130 C18 column (1.7 μm, 100 μm i.d. × 100 mm, Waters Corp., Milford, MA). The mobile phase consisted of 0–5 min, 99% A; 5–7 min, 99–90% A; 7–37 min, 90–60% A; 37–38 min, 60–15% A; 38–48 min, 15% A; 48–49 min, 15–99% A; 49–60 min, 99% A, where A = 0.1% formic acid and 2% acetonitrile in water, B = 0.1% formic acid and 2% water in acetonitrile. The flow rate was 1.2 μL/min. The eluent was analyzed by a LTQ Velos Orbitrap tandem mass spectrometer (Thermo Fisher Scientific, Waltham, MA) using positive-electrospray ionization. High resolution (60,000 resolving power), accurate mass spectra were recorded between m/z 395–2000 in ∼1.2 s on the orbitrap mass analyzer. While the next high-resolution mass spectrum was being acquired on the orbitrap, the LTQ Velos linear ion trap independently recorded CID fragmentation mass spectra of the 8 most abundant ions present in the previous orbitrap mass spectrum. During the course of a 60 min nano LC/MS/MS run, this approach typically generated ∼3000 high-resolution mass spectra and between 12 000 and 15 000 CID MS/MS spectra.

Thermo-Finnegan Proteome Discoverer 2.0 software (Thermo Fisher Scientific, Waltham, MA, USA) was used to interface with the Mascot (Matrix Science, Boston, MA, USA) protein database search engine. MS/MS spectral information was used by Mascot to search the SwissProt Protein database.

For quantification, a 2 μL aliquot of the sample was injected onto a nanoACQUITY UPLC C18 column (1.8 μm, 100 μm i.d. × 100 mm, Waters Corp., Milford, MA). The mobile phase consisted of 12 min elution at 600 nL/min with 2% acetonitrile/0.1% formic acid/water, followed by a 60 min linear gradient to 35% acetonitrile/0.1% formic acid/water. Samples were analyzed on a ABSciex QTrap 5500 mass spectrometer (ABSciex, Farmingham, MA, USA) running in ion trap IDA mode coupled to a two-dimensional Eksignet Ultra NanoUPLC system, consisting of a nanoLC ultra 2D pump and a nanoLC AS-2 autosampler (Eksignet, Dublin, CA, USA). The mass spectrometer was operated in the positive electrospray ionization (ESI) mode. The following conditions were used: curtain gas 20 units, ion spray voltage 2350 V, ion source gas 110 units, ion source gas 20 units, declustering potential 100 units, entrance potential 10 units, collision cell exit potential 40 units. Three peptides (custom-synthesized by GenScript Biotech, Pistacaway, NJ) per MMP/ADAM were used for identification and quantification (Table 2). Calibration curves containing known amounts of synthetic peptides in human plasma relative to internal standard were used to determine the levels of active MMPs/ADAMs in the lung samples. Three peptides per MMP/ADAM were used, with three transitions as qualifiers to identify the protein and three transitions as quantifiers to quantify MMPs/ADAMs.

Table 2. Peptides Used for the Affinity Resin/Proteomics Quantification of Active ADAMs and MMPs in Human IPF Samples and Human Healthy Lung Controls^a.

protein	peptide sequence	Q1 precursor ion m/z	Q3 product ion m/z
ADAM9	FLPGGTLCR	510.77 [M + H]2+	606.30 [M + H]+ y5 663.32 [M + H]+y6 760.38 [M + H]+ y7
	DCFIEVNSK	556.26 [M + H]2+	447.26 [M + H]+ y4 689.38 [M + H]+ y6 836.45 [M + H]+y7
	FGNCGFSGNEYK	690.29 [M + H]2+	844.38 [M + H]+ y7 901.41 [M + H]+ y8 1061.44 [M + H]+y9
ADAM10	AIDTIYQTTDFSGIR	850.93 [M + H]2+	579.32 [M + H]+ y5 896.45 [M + H]+ y8 1187.57 [M + H]+y10
	FSLCSIR	441.73 [M + H]2+	535.27 [M + H]+y4 648.35 [M + H]+ y5 735.38 [M + H]+ y6
	LVDADGPLAR	513.78 [M + H]2+	456.29 [M + H]+ y4 699.38 [M + H]+ y7 814.41 [M + H]+y8
ADAM17	GEESTTTNYLIELIDR	927.46 [M + H]2+	758.44 [M + H]+ y6 871.52 [M + H]+y7 1034.59 [M + H]+ y8
	GYGIQIEQIR	588.82 [M + H]2+	786.45 [M + H]+y6 899.53 [M + H]+ y7 956.55 [M + H]+ y8
	AYYSPVGK	442.73 [M + H]2+	487.29 [M + H]+ y5 650.35 [M + H]+y6 813.41 [M + H]+ y7
MMP-1	VTGKPDAETLK	579.82 [M + H]2+	773.4 [M + H]+y7 901.5 [M + H]+ y8 958.52 [M + H]+ y9
	LTFDAITTIR	575.83 [M + H]2+	674.42 [M + H]+y6 789.45 [M + H]+ y7 936.51 [M + H]+ y8
	DIYSSFGFPR	594.79 [M + H]2+	710.36 [M + H]+ y6 797.39 [M + H]+ y7 960.46 [M + H]+y8
MMP-8	QFMEPGYPK	548.76 [M + H]2+	690.35 [M + H]+y6 821.39 [M + H]+ y7 968.45 [M + H]+ y8
	SISGAFPGIESK	596.81 [M + H]2+	630.35 [M + H]+y6 777.41 [M + H]+ y7 848.45 [M + H]+ y8
	YYAFDLIAQR	630.32 [M + H]2+	600.38 [M + H]+ y5 862.48 [M + H]+ y7 933.52 [M + H]+y8
MMP-14	VGEYATYEAIR	636.32 [M + H]2+	752.39 [M + H]+ y6 823.43 [M + H]+y7 986.49 [M + H]+ y8
	VWESATPLR	529.79 [M + H]2+	644.37 [M + H]+ y6 773.42 [M + H]+ y7 959.49 [M + H]+y8
	EVPYAYIR	505.77 [M + H]2+	685.37 [M + H]+ y5 782.42 [M + H]+y6 881.49 [M + H]+ y7
internal standard (yeast enolase)	TAGIQIVADDLTVTNPK	878.48 [M + H]2+	772.46 [M + H]+ y7 1002.51 [M + H]+ y9 1172.62 [M + H]+y11
	VNQIGTLSESIK	644.86 [M + H]2+	676.39 [M + H]+ y6 777.44 [M + H]+ y7 1075.60 [M + H]+y10
	NVNDVIAPAFVK	643.86 [M + H]2+	561.34 [M + H]+ y5 632.38 [M + H]+y6 844.53 [M + H]+ y8

^aFor quantification, three peptides per proteinase were used. Precursor ions were selected in Q1 (first quadrupole) then fragmented in Q2, and the specified product ions were monitored in Q3. In bold are the fragments used for quantification.

Compounds

GI254023X was purchased from Aobious, Inc. (Gloucester, MA).

MMP Kinetics

Human recombinant MMP-1 was purchased from Enzo Life Science (Farmingdale, NY). Human recombinant ADAM17 was purchased from R&D Systems (Minneapolis, MN). Fluorogenic substrate Mca-KPLGL-Dpa-AR-NH₂ for MMP-1 and Mca-PLAQAV-Dpa-RSSSR-NH₂ for ADAM17 were purchased from R&D Systems (Minneapolis, MN). The K_m values used for MMP-1 and ADAM17 were calculated before the inhibition studies and were similar to previous published values.³³ Inhibitor stock solution (5 mM) was prepared freshly in DMSO before the enzyme-inhibition assays. We followed the same methodology for enzyme-inhibition studies as reported before by Ikejiri et al.(34) Enzyme-inhibition studies were carried out using a Cary Eclipse fluorescence spectrophotometer (Varian, Agilent Technologies, Santa Clara, CA). The experiments were done in duplicate.

Cell Culture

Human lung fibroblasts were obtained from ATCC (Manassas, VA). Cells were maintained and cultured in fibroblast medium (ScienCell Research Laboratories, San Diego, CA) containing fibroblast growth supplement, 2% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C in a humidified incubator with 5% CO₂.

In Vitro IPF Assay

Human lung fibroblasts were counted, seeded onto a sterilized cover glass (5 × 10⁵ cells per well), and incubated for 24 h. The culture medium was replaced by Minimum Essential Media (MEM), Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma-Aldrich Corp.) with or without GI254023X, followed by the addition of 20 ng/mL TGF-β1 (Peprotech, Inc., East Windsor, NJ). Cells were fixed with 4% paraformaldehyde, permeabilized with 0.5% Triton X-100, blocked with 10% fetal bovine serum, and incubated with anti-α-smooth muscle actin (α-SMA) antibody (Abcam, Cambridge, MA) for 2 h at room temperature. The cells were then probed with goat anti-mouse IgG secondary antibody–Alexa Fluor 488 conjugate (Abcam, Cambridge, MA). DAPI (4′,6-diamidino-2-phenylindole) was used for nuclear staining. The cells were imaged on a fluorescence microscope (Nikon A1R-MP laser scanning confocal microscope, Nikon Instruments Inc., Melville, NY). The assay was done in replicates of six. Fluorescence quantification was performed using 10 images per slide, ImageJ, and the Fiji platform.³⁵ Fluorescence intensity was measured by outlining the cells in the image using the freehand selection tool and calculation of area, integrated density, and mean gray value. Background regions were measured for each desired cell in parallel and the mean fluorescence was calculated. Data from each desired cell and mean fluorescence of background regions were used to calculate the corrected total cell fluorescence (CTCF = desired cell integrated density – (area of desired cell × mean fluorescence of background regions).³⁶ CTCF data from the 10 images per slide were averaged.

Statistical Analysis

Differences in the levels of active MMPs/ADAMs were analyzed for statistical significance using a Student’s t test with two-tailed distribution and unequal variance; p < 0.05 was considered statistically significant. Differences in α-SMA were analyzed similarly, except using equal variance.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.2c00050.

• Levels of active MMP-1, MMP-8, MMP-14, ADAM9, ADAM10, and ADAM17 in IPF lungs by disease severity as classified by FVC and DLCO; inhibition of MMPs and ADAMs by batimastat (PDF)

Author Contributions

Z.P. conducted the affinity resin analysis with proteomics of human IPF lungs and control lungs. M.M.K. synthesized chemical tools under the direction of S.M. M.W. determined K_i values for GI254023X. Z.P., L.A., and M.W. performed the fibroblast to myofibroblast assays. M.C. designed and directed the studies and wrote the manuscript with contributions from the authors.

The authors declare no competing financial interest.