Severe Acute Respiratory Syndrome Coronavirus 2 Infection Alters Mediators of Lung Tissue Remodeling In Vitro and In Vivo

Abstract

Background

Altered mediators of airway tissue remodeling such as matrix metalloproteinases (MMPs) in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection may contribute to morbidity in coronavirus disease 2019 (COVID-19); however, the differential impact of SARS-CoV-2 variants of concern (VOCs) on MMPs is unknown.

Methods

Using both in vitro human airway cell culture model and in vivo transgenic mouse model of SARS-CoV-2 infection, we studied the differential effect of SARS-CoV-2 VOCs on expression of key MMPs and inflammatory mediators in airway cells and tissues.

Results

The most consistent findings with all SARS-CoV-2 variants in infected compared to uninfected human bronchial epithelial cell air–liquid interface cultures were the SARS-CoV-2–induced increases in MMP-12 and tissue inhibitor of MMPs. Infection with both SARS-CoV-2 wild type and SARS-CoV-2 Delta variant over 3 days postinfection (dpi) and with Beta variant over 7 dpi increased lung tissue levels of MMP-9 compared to uninfected mice. Overall, SARS-CoV-2 variants had differential dose-dependent impact on secretion of MMP-1, MMP-2, MMP-9, and MMP-12 that varied at the protein versus the gene level and in the early noninflammatory compared to late inflammatory phase of infection.

Conclusions

We provide novel mechanistic insight that the differential impact of SARS-CoV-2 variants on severity of COVID-19 may partially be attributed to unique changes in MMPs.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has greatly impacted humanity. SARS-CoV-2 replication and associated lung injury, characterized by inflammation and tissue remodeling, contribute to morbidity in coronavirus disease 2019 (COVID-19) and may lead to hospitalizations, severe COVID-19, death, and long-term respiratory complications of COVID-19 [1]. However, the mechanisms of tissue remodeling in SARS-CoV-2 infection remain unclear.

Inflammation leads to lung injury and alterations in the homeostasis of the extracellular matrix (ECM) that collectively contribute to the development of acute lung injury and acute respiratory distress syndrome (ARDS) [2]. Matrix metalloproteinases (MMPs) are tissue proteases that can degrade all components of the ECM and key inflammatory mediators of lung injury [2]. Serum MMPs may be early biomarkers of severity of COVID-19 [3, 4]. Blood levels of MMPs do not reflect tissue levels and may change in different and complex directions depending on the stage and severity of lung disease [5]. Complex confounders in human studies and limited access to human lung tissue further limit efforts to elucidate whether MMPs are causal in lung remodeling or part of the inflammatory and reparative response [6].

To address the gaps in knowledge, we studied the differential in vitro effect of various strains of SARS-CoV-2, on expression of key MMPs and inflammatory mediators in human airway cell culture models. We also utilized an established mouse model of SARS-CoV-2 infection to determine the direct impact of different SARS-CoV-2 variants on key mediators of lung tissue injury (inflammation and remodeling) in COVID-19. Herein, we provide novel mechanistic insight into the effect of SARS-CoV-2 variants on established instigators of lung tissue remodeling and damage.

METHODS

Materials

Detailed materials and methods are described in the Supplementary Material. The SARS-CoV-2 isolates 2019-nCoV/USA-WA1/2020 strain (labeled as wild type [WT]), B.1.351 (Beta variant), B.1.617.2 (Delta variant), and BA.1 (Omicron variant) were utilized.

SARS-CoV-2 Infection of Human Airway Cell Cultures

Human bronchial epithelial cells (HBECs) air–liquid interface (ALI) cultures were established as described in the Supplementary Material. After 4 weeks of differentiation, the ALI HBECs were infected with SARS-CoV-2 variants for 48 hours. Using 2 different amounts of virus, we simulated the differential impact of SARS-CoV-2 in the early noninflammatory state, when the virus amount is high, compared to the late inflammatory phase of infection (when the virus amount is low).

Mice

All of the animal studies were in accordance with federal, state, and local approved guidelines. We used K18-hACE2 mice, an established mouse model of SARS-CoV-2 infection in which SARS-CoV-2 replication is highest on 2–3 days postinfection (dpi) while associated inflammation and tissue injury is highest on 5–7 dpi [7]. In total, we utilized 37 male and female 4- to 12-week-old K18-hACE2 mice in 4 groups. Mice were infected intranasally with SARS-CoV-2 (10 000 plaque-forming units [PFU]/mouse).

Immunoassays

Protein levels of human or murine C-C motif chemokine ligand 2 (CCL-2), interleukin (IL)-1β, IL-6, tumor necrosis factor alpha (TNF-α), MMP-1, MMP-2, MMP-9, MMP-12, and tissue inhibitor of metalloproteinases 1 (TIMP-1) were determined in cell culture supernatants of HBEC ALI cultures or lung protein lysates using magnetic Luminex performance assay kits according to the manufacturer’s instructions (R&D) and as previously described [8]. Fifty to 100 mg of lung tissue samples was mechanically dissociated in T-PER tissue protein extraction reagent as previously described [8].

Quantitative Polymerase Chain Reaction

Total RNA was isolated using the RNeasy Mini Kit (Zymo Research) and complementary DNA (cDNA) was synthesized using the RevertAid first strand cDNA synthesis kit. Quantitative polymerase chain reaction (qPCR) was performed using SYBR Green Master Mix as described in the Supplementary Material. Primers are listed in Supplementary Table 1.

Mouse Lung Histological Analysis

Paraffin-embedded murine lung tissue blocks were cut into 5-μm sections that were stained with Masson trichrome. Light microscopic scans of whole lung were examined by an experienced pathologist.

Statistical Analysis

Unless noted, error bars in all figures represent mean and standard error of the mean. The Shapiro-Wilk normality test assessed normality of data for use of parametric tests (t test). The Mann–Whitney U tests and unpaired t-tests were used to compare statistical differences between 2 groups, as appropriate. P values < .05 were considered significant. All analyses were performed with GraphPad, version 8.0 (GraphPad Holdings, San Diego, California).

RESULTS

In Vitro Impact of SARS-CoV-2 Infection on MMPs Secreted by Airway Epithelial Cells

Using a sensitive multiplex Luminex immunoassay, we quantified proteins of the secreted MMPs in HBEC ALI cultures infected with SARS-CoV-2 variants. Compared to uninfected HBEC ALI (mock) cultures, infection with SARS-CoV-2 Delta variant at multiplicity of infection (MOI) 2.5 increased levels of MMP-1 (Figure 1A), MMP-9 (Figure 1C), and MMP-12 (Figure 1D) (P < .05 for all comparisons), and infection with Beta and Omicron variants at MOI 0.5 also increased levels of MMP-12 (Figure 1D). Compared to mock HBEC ALI cultures, infection with SARS-CoV-2 Omicron variant at MOI 2.5 reduced MMP-2 (Figure 1B; P < .05). There was no impact of SARS-CoV-2 Beta variant (both MOI 0.5 and 2.5) and lower inoculum (MOI 0.5) of SARS-CoV-2 Omicron on secretion of MMP-1, MMP-2, and MMP-9 by HBEC airway ALI cultures (Figure 1A–C). Infection with MOI 0.5 of SARS-CoV-2 Delta variant drastically reduced secretion of MMP-12 by airway cells compared to infection with MOI 2.5 of SARS-CoV-2 Delta variant (Figure 1D; P < .05). In contrast, infection with MOI 0.5 of SARS-CoV-2 Omicron variant drastically increased secretion of MMP-12 by airway cells compared to infection with MOI 2.5 of SARS-CoV-2 Omicron variant (Figure 1D; P < .01). A dose-dependent effect on MMPs was consistently observed with the Omicron variant with MMP-2, MMP-9, and MMP-12 (Figure 1A–C). Overall, SARS-CoV-2 variants had differential dose-dependent impact on secretion of key MMPs (MMP-1, MMP-2, MMP-9, and MMP-12).

Figure 1.

Impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on secretion of mediators of lung tissue remodeling by upper airway cell cultures. Human bronchial epithelial cells (HBECs) air–liquid interface (ALI) cultures were infected with SARS-CoV-2 variants (Beta, Delta, Omicron) for 48 h with 5× fold gradient of high (multiplicity of infection [MOI] 2.5) versus low (MOI 0.5) virus. Protein levels of the matrix metalloproteinases (MMPs) 1 (MMP-1; A), 2 (MMP-2; B), 9 (MMP-9; C), and 12 (MMP-12; D) were determined by multiplex Luminex immunoassay as described in the Methods. The mean value of each protein measurement in SARS-CoV-2–infected HBEC ALI supernatants was normalized to the mean value of uninfected HBEC ALI supernatants. In all panels, summary data are shown as mean ± standard error of the mean of 3 biological replicates. Each data point represents an average of 3 technical replicates of a sample. Statistical comparisons between the uninfected control group and each experimental group (asterisk above each column) and between 2 noncontrol experimental groups were done using unpaired t tests. *P < .05, **P < .01.

In Vitro Impact of SARS-CoV-2 Infection on Intracellular Levels of MMPs in Airway Epithelial Cells

To better characterize the differential impact of SARS-CoV-2 on MMPs expressed in airway epithelial cells, we then assessed intracellular MMP gene levels in HBEC ALI cultures using qPCR (Figure 2A–D). Compared to uninfected HBEC ALI cultures, infection with SARS-CoV-2 Delta variant at MOI 2.5 increased gene levels of MMP-1 (Figure 2A); infection with SARS-CoV-2 Beta, Delta, and Omicron variants consistently increased MMP-12 (Figure 2D) (P < .05). Similarly to protein levels, there were differential and inverse effects of higher (MOI 2.5) versus lower (MOI 0.5) inoculums of both Delta and Omicron variants on gene levels of MMP-2 (Figure 2B), MMP-9 (Figure 2C), and MMP-12 (Figure 2D). Compared to uninfected HBEC ALI cultures, infection with SARS-CoV-2 Beta variant increased MMP-1 gene levels (Figure 2A); infection with Delta variant at MOI 2.5 did not increase gene levels of MMP2 (Figure 2B) and MMP-9 (Figure 2C), while infection with Delta variant at MOI 0.5 decreased gene levels of MMP-2 (Figure 2B) and MMP-9 (Figure 2C). Compared to uninfected and SARS-CoV-2 Omicron-infected (MOI 0.5) HBEC ALI cultures, infection with SARS-CoV-2 Omicron variant at MOI 2.5 increased MMP-1 gene levels (Figure 2A). The most notable and consistent mean magnitude of impact (>3-fold) of all 3 SARS-CoV-2 variants on MMPs was on levels of MMP-12 at both the protein (Figure 1) and gene level (Figure 2). There were no other consistent notable effects of SARS-CoV-2 variants on gene levels of MMP-1, MMP-2, MMP-9, and MMP-12 (Figure 2A–D). Thus, our studies in HBEC ALI cultures at the gene level confirmed that SARS-CoV-2 variants had differential dose-dependent impact on both the gene expression and the protein secretion of key MMPs (MMP-1, MMP-2, MMP-9, and MMP-12).

Figure 2.

Impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on gene levels of mediators of lung tissue remodeling in upper airway cell cultures. Human bronchial epithelial cells (HBECs) air–liquid interface (ALI) cultures were infected with SARS-CoV-2 variants (Beta, Delta, Omicron) for 48 h with 5× fold gradient of high (multiplicity of infection [MOI] 2.5) versus low (MOI 0.5) virus. Messenger RNA levels of the matrix metalloproteinases (MMPs) 1 (MMP-1; A), 2 (MMP-2; B), 9 (MMP-9; C), and 12 (MMP-12; D) were determined by quantitative polymerase chain reaction as described in the Methods. The mean value of each protein measurement in SARS-CoV-2–infected HBEC ALI supernatants was normalized to the mean value of uninfected HBEC ALI supernatants. In all panels, summary data are shown as mean ± standard error of the mean of 3 biological replicates. Each data point represents an average of 3 technical replicates of a sample. Statistical comparisons between the uninfected control group and each experimental group (asterisk above each column) and between 2 noncontrol experimental groups were done using unpaired t tests. *P < .05, **P < .01.

In Vitro Impact of SARS-CoV-2 Infection on TIMP-1 in Airway Epithelial Cells

To better characterize the differential impact of SARS-CoV-2 on MMPs expressed in airway epithelial cells, we then assessed intracellular gene levels of TIMP-1 in HBEC ALI cultures using qPCR. Compared to uninfected HBEC ALI cultures, infection with SARS-CoV-2 Omicron variant induced a major increase in levels of TIMP-1 in a dose-dependent manner (P < .001 for all comparisons; Supplementary Figure 1). Compared to uninfected HBEC ALI cultures, infection with SARS-CoV-2 Delta variant at MOI 2.5 and with SARS-CoV-2 Beta variant at MOI 0.5 increased levels of TIMP-1 (P < .01 for all comparisons). Similarly to gene and protein levels of MMPs, dose-dependent effects of SARS-CoV-2 on TIMP-1 was found with all 3 variants (P < .05 for all comparisons between MOI 2.5 vs MOI 0.5). There was no impact of SARS-CoV-2 Beta variant (MOI 2.5) and lower inoculum (MOI 0.5) of SARS-CoV-2 Delta variant on gene levels of TIMP-1 in HBEC airway ALI cultures (Supplementary Figure 1). Overall, the Omicron variant had higher impact on gene levels of TIMP-1 in HBEC airway cells compared to the Beta and Delta variants.

In Vitro Impact of SARS-CoV-2 Infection on Cytokines Secreted by Airway Epithelial Cells

To obtain insight into the differential impact of SARS-CoV-2 variants on lung tissue remodeling and damage, we also assessed secretion of cytokines and chemokines by airway epithelial cells in infected HBEC ALI cultures. We assessed secretion of IL-1β, IL-6, TNF-α, and CCL-2 (Figure 3A–D) since these proinflammatory mediators predict adverse clinical outcomes in COVID-19 and are also implicated in MMP pathways and lung tissue damage. Compared to uninfected HBEC ALI cultures and unlike infection with the SARS-CoV-2 Omicron variant, infection with the SARS-CoV-2 Delta variant induced an increase in secreted protein levels of IL-1β (MOI 2.5) (Figure 3B), IL-6 (MOI 0.5) (Figure 3B), TNF-α (MOI 2.5) (Figure 3C), and CCL-2 (both MOI 2.5 and 0.5) (Figure 3D) (P < .05 for all comparisons). Thus, similarly to MMPs and compared to the Omicron variant, the SARS-CoV-2 Delta variant had higher impact on secretion of proinflammatory mediators by airway epithelial cells.

Figure 3.

Impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on protein levels of proinflammatory mediators secreted by upper airway cell cultures. Human bronchial epithelial cells (HBECs) air–liquid interface (ALI) cultures were infected with SARS-CoV-2 variants (Beta, Delta, Omicron) for 48 h (multiplicity of infection [MOI] 0.5). Protein levels of interleukin (IL)-1β (A), IL-6 (B), tumor necrosis factor alpha (TNF-α; C), and C-C motif chemokine ligand 2 (CCL-2; D) were determined by multiplex Luminex immunoassay as described in the Methods. The mean value of each protein measurement in SARS-CoV-2–infected HBEC ALI supernatants was normalized to the mean value of uninfected HBEC ALI supernatants. In all panels, summary data are shown as mean ± standard error of the mean of 3 biological replicates. Each data point represents an average of 3 technical replicates of a sample. Statistical comparisons between the uninfected control group and each experimental group (asterisk above each column) and between 2 noncontrol experimental groups were done using unpaired t tests. *P < .05, **P < .01.

In Vivo Impact of SARS-CoV-2 Infection on Molecular Mediators of Lung Tissue Damage and Remodeling

To further characterize the in vivo impact of SARS-CoV-2 infection on molecular mediators of lung tissue damage and remodeling, we infected K18-hACE2 mice with SARS-CoV-2 and assessed the impact of SARS-CoV-2 on these mediators during both early noninflammatory (3 dpi) and late inflammatory (7 dpi) phases of infection (Figure 4A). SARS-CoV-2 infection in K18-hACE2 mice led to detectable viral lung titer on 3 dpi (range, 10⁴–10⁶ PFU/mL) but not on 7 dpi, consistent with prior studies in K18-hACE2 mice [7].

Figure 4.

Impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on mediators of lung tissue remodeling in mouse model of SARS-CoV-2. K18-hACE2 mice were infected intranasally with SARS-CoV-2 (10 000 plaque-forming units/mouse) wild-type (ancestral strain; WT) or Beta and Delta variants and lung tissue was harvested 3–7 days postinfection. Lung tissue protein lysates were prepared and protein was determined as described in the Methods. The mean value of each protein measurement in lung tissue lysates of mice was normalized to the mean value of protein content in each lysate. A, Study design for mouse cohorts A (n = 5), B (n = 10), C (n = 10), and D (n = 12). B–D, Protein levels of the matrix metalloproteinases (MMPs) 2 (MMP-2; B), 9 (MMP-9; C), 12 (MMP-12; D) were determined by multiplex Luminex immunoassay as described in the Methods. The mean value of each protein measurement in SARS-CoV-2–infected mice was normalized to the mean value of uninfected mice. In all panels, summary data are shown as mean ± standard error of the mean of at least 2 replicates. Each data point represents 1 biological sample. Statistical comparison between the uninfected control group and each experimental group (asterisk above each column) was done using 2-tailed Mann–Whitney test. Statistical comparison between 2 noncontrol experimental groups was done using 2-tailed Mann–Whitney test. *P < .05, **P < .01, ***P < .001.

Infection with SARS-CoV-2 WT but not SARS-CoV-2 Delta variant increased lung tissue levels of MMP-2 on 3 dpi compared to uninfected mice (P < .01; Figure 4A). Notably, infection with SARS-CoV-2 Beta variant decreased lung tissue levels of MMP-2 on 7 dpi compared to uninfected mice (P < .05; Figure 4B). Infection with both SARS-CoV-2 WT and SARS-CoV-2 Delta variant over 3 dpi and with Beta variant over 7 dpi increased lung tissue levels of MMP-9 compared to uninfected mice (P < .05 for all comparisons; Figure 4C). The Delta variant upregulated MMP-9 by >2 orders of magnitude compared to SARS-CoV-2 WT (P < .01; Figure 4C). Infection with SARS-CoV-2 Beta variant decreased lung tissue levels of MMP-12 on 7 dpi compared to uninfected mice (P < .01; Figure 4D). K18-hACE2 mice infected with both SARS-CoV-2 WT and SARS-CoV-2 Delta variant over 3 dpi had increased lung tissue levels of TIMP-1 compared to uninfected mice (P < .05 for all comparisons; Supplementary Figure 2). Infection with SARS-CoV-2 WT and Delta variant over 3 dpi (Figure 4D) and Beta variant over 7 days (Supplementary Figure 2) did not impact lung tissue levels of MMP-12 and TIMP-1, respectively, compared to uninfected mice. Thus, SARS-CoV-2 variants induced unique changes in lung tissue levels of MMPs and TIMP-1 in a mouse model of SARS-CoV-2 that were different in the early noninflammatory compared to late inflammatory phase of infection.

Using a sensitive multiplex Luminex immunoassay, we found that early infection with SARS-CoV-2 tended to increase IL-6 on 3 dpi (Delta variant only, P = .07) but did not increase levels of IL-1β, IL-6, TNF-α, or CCL-2 in lung lysates compared to uninfected mice (Figure 5A–D) that had low detectable protein levels (Supplementary Figure 3). SARS-CoV-2 infection with Beta variant on 7 dpi induced increase in lung tissue levels of IL-1β, IL-6, and TNF-α compared to uninfected mice (P < .05 for all comparisons; Figure 5A–C). SARS-CoV-2 infection did not impact protein levels of CCL-2 compared to uninfected mice at all timepoints (Figure 5D). Histopathologic assessment of lung inflammation showed lung inflammation in SARS-CoV-2–infected K18-hACE2 mice on 7 but not on 3 dpi compared to uninfected mice (Supplementary Figure 4). Thus, only infection with the Beta variant over 7 dpi induced increase in lung inflammation in the K18-hACE2 mice.

Figure 5.

Impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on mediators of lung inflammation in mouse model of SARS-CoV-2. K18-hACE2 mice were infected intranasally with SARS-CoV-2 (10 000 plaque-forming units/mouse) wild type (ancestral strain; WT) or Beta and Delta variants, and lung tissue was harvested 3–7 days postinfection. Lung tissue protein lysates were prepared and protein was determined by bicinchoninic acid protein assay kit as described in the Methods. The mean value of each protein measurement in lung tissue lysates of mice was normalized to the mean value of protein content in each lysate. The proinflammatory cytokines interleukin (IL) 1 beta (IL-1β) (A), IL-6 (B), tumor necrosis factor alpha (TNF-α) (C), and C-C motif chemokine ligand 2 (CCL2) (D) were determined by multiplex Luminex immunoassay as described in the Methods. The mean value of each protein measurement in SARS-CoV-2–infected mice was normalized to the mean value of uninfected mice. In all panels, summary data are shown as mean ± standard error of the mean of at least 3 replicates. Each data point represents 1 biological sample. Statistical comparison between the uninfected control group and each experimental group (asterisk above each column) was done using 2-tailed Mann–Whitney test. Statistical comparison between 2 noncontrol experimental groups was done using 2-tailed Mann–Whitney test. *P < .05, **P < .01.

It is not known whether SARS-CoV-2 infection can directly induce in vivo mediators of fibrosis that may drive acute lung injury and ARDS in severe COVID-19. Infection with SARS-CoV-2 Beta variant on 7 dpi but not with Delta variant on 3 dpi increased mRNA levels of instigators of fibrosis such as transforming growth factor beta (TGF-β) (Supplementary Figure 5) and collagen I (Supplementary Figure 6) in lung tissue of infected compared to uninfected K18-hACE2 mice (P < .01 for all comparisons). Despite the presence of lung inflammation, there was no interstitial lung fibrosis based on Masson trichrome staining of collagen in SARS-CoV-2–infected mice on 7 dpi (Supplementary Figure 4).

DISCUSSION

Herein, we describe for the first time the differential in vitro and in vivo impact of independent SARS-CoV-2 variants on established instigators of lung tissue remodeling (MMPs, TIMP-1, TGF-β). Using HBEC ALI airway cultures, we showed that SARS-CoV-2 variants had differential and unique dose-dependent impact on both the gene expression and the protein secretion of key MMPs. Our studies uniquely dissected the complex effects of different SARS-CoV-2 variants on tissue MMPs in the setting of higher amount of virus that is typically present in the early noninflammatory phase of infection compared to the lower amount of virus that is typically present in the late inflammatory phase of infection. The most consistent findings with all SARS-CoV-2 variants in HBEC ALI cultures were the SARS-CoV-2–induced increases in MMP-12 and TIMP-1. We further confirmed that SARS-CoV-2 variants induced unique changes in lung tissue levels of MMPs and TIMP-1 in mouse model of SARS-CoV-2 that were different in early noninflammatory compared to late inflammatory phase of infection. Our data provide novel mechanistic insight that the differential impact of SARS-CoV-2 variants on severity of COVID-19 may be attributed to unique in vivo changes in lung tissue remodeling in COVID-19, independently of effects on lung inflammation. Our data are also consistent with prior human studies have shown that patients with acute lung injury/ARDS have elevated levels of MMP-2, MMP-9, and TIMP-1 in the bronchoalveolar lavage compared to patients with lung disease without acute lung injury/ARDS [5]. Our findings may improve our understanding why patients with SARS-CoV-2 infection with Beta and Delta variants had higher incidence of severe lung disease compared to the WT and Omicron strain [1].

A notable finding was that the Omicron variant at higher amounts of virus consistently reduced the protein levels of MMP-12 and increased gene levels of TIMP-1 without increasing proinflammatory mediators compared to the Delta and Beta variants. As MMPs are tissue proteases that can degrade all components of the ECM and are key inflammatory mediators of lung injury, our findings of decreased levels of MMP-12 provide mechanistic insight into the significantly reduced rates of hospitalizations and deaths among Omicron-infected patients when compared to Beta and Delta variants [2]. However, these effects of Omicron were not consistently observed at the gene level, consistent with prior evidence that gene and protein levels of MMPs have complex regulation depending on the context and stage of a disease [9].

Another important observation was that the Omicron variant had quite different effect on MMP-1 levels compared to MMP-2, MMP-9, and MMP-12 in upper airway cell cultures. In contrast to MMP-2, MMP-9, and MMP-12, infection with higher amount of SARS-CoV-2 Omicron variant did not reduce MMP-1 protein levels compared to uninfected cultures. MMP-2 is mainly localized to alveolar macrophages and is a gelatinolytic MMP that interferes with chemotaxis of immune cells [4, 5]. Thus, the impact of SARS-CoV-2 variants on secretion of MMP-2 seems to be more relevant in macrophages rather than airway epithelial cells during SARS-CoV-2 infection. These data are novel since MMP-1 has a major role in tissue remodeling and can be upregulated in bacterial infections, but it has not been shown to be altered in viral infections [9].

We found that SARS-CoV-2 infection with the Delta variant induced a major early upregulation of MMP-9 compared to WT virus and uninfected mice but did not impact levels of TGF-β and collagen I in lung tissue of K18-hACE2 mice on 3 dpi. Notably, MMP-9 levels were reduced while TIMP-1, TGF-β, and collagen I were all increased in lung tissue of K18-hACE2 mice on 7 dpi compared to 3 dpi. However, we observed less prominent in vitro effects of Delta variant on levels of MMP-9 upper airway cell cultures, suggesting that SARS-CoV-2 has a less relevant impact in upper airway epithelial cells compared to lung cells. Consistent with this hypothesis, higher amounts of Omicron variant, which induce higher infection in upper airway cells compared to lower airway cells, downregulated levels of MMP-9 in HBEC cultures. MMP-9 is a gelatinolytic MMP that is mostly localized to neutrophils, macrophages, and epithelial cells; promotes alveolar capillary destruction and lung injury; and is implicated in the pathogenesis of pulmonary edema [2, 5]. Experimental in vitro and in vivo studies have shown that MMP-9 activates TGF-β and induces TGF-β production in lung epithelial cells [10]. Consistent with our data, prior human and experimental studies have shown that circulating MMP-9 is upregulated early in the course of lung diseases including other viral infections [9] and COVID-19 [4, 6]. Collectively, our study suggests that SARS-CoV-2 induces secretion of gelatinolytic MMP-9 during early SARS-CoV-2 infection by innate immune cells and lower airway cells that may further induce fibrogenic responses like TGF-β production in lung epithelial cells. These data are consistent with prior evidence that MMP-9 levels in plasma are altered and associated with mortality in COVID-19 patients [11].

Notably and paradoxically, both SARS-CoV-2 Delta and Omicron variants had inverse dose-dependent effects on secretion of MMP-12. Infection with SARS-CoV-2 Beta variant decreased lung tissue levels of MMP-12 on 7 dpi compared to uninfected mice, while lower amounts of Delta variant drastically downregulated MMP-12 compared to uninfected HBEC cultures. These data suggest that lower expression of MMP-12 may be more relevant during the inflammatory phase of SARS-CoV-2. MMP-12 appears to play a consistent and important role in several lung diseases [2, 5, 12] while it has been shown to be upregulated by respiratory syncytial virus [9].

We also found that SARS-CoV-2 infection with WT and Delta variant upregulated TIMP-1 in the lung tissue of K18-hACE2 mice on 3 dpi compared to uninfected mice. In contrast, high amount of Omicron variant upregulated TIMP-1 compared to lower amount of Omicron variant in HBEC cultures, suggesting that TIMP-1 is more relevant in pathogenesis of the early noninflammatory phase of SARS-CoV-2 infection (when the virus is higher). Following upregulation of cytokines and MMPs during lung injury, TIMPs are secreted to maintain the matrix equilibrium and to regulate cytokine shedding [2]. Our data are consistent with prior experimental studies that have shown preferential elevation in TIMP-1 circulating levels and mRNA expression in brain tissue in a mouse model of infection with betacoronavirus, suggestive of a TIMP-1–mediated host defense mechanism against virus-induced upregulation of different metalloproteases [13]. Given that most viruses are known to either not impact or even downregulate TIMP-1 [9], the upregulation of TIMP-1 induced by Omicron may be a unique mechanism that contributes to reduced host tissue damage by SARS-CoV-2 Omicron variant.

Our study has limitations. We did not study all possible MMPs, TIMPs, and mediators of tissue remodeling and the impact of all possible SARS-CoV-2 variants on tissue remodeling. Instead, we focused on key mediators of tissue remodeling based on their role in tissue remodeling in viral infections and availability of a multiplex Luminex immunoassay, cells, and tissues from experimental studies of SARS-CoV-2 infection with WT, Delta, Beta, and Omicron variants. Additionally, the K18-hACE2 mouse model is not favorable for modeling long-term tissue injury and fibrosis in SARS-CoV-2 infection (which is relevant to acute lung injury/ARDS).

In conclusion, despite limitations, our mechanistic in vitro and in vivo studies with assessment of mediators of tissue remodeling in both the early noninflammatory and the inflammatory phases of SARS-CoV-2 infection provide novel insight into the pathogenesis of early lung tissue remodeling in SARS-CoV-2 infection. These data are consistent with a conceptual model that increases in MMPs are causal in lung remodeling in early SARS-CoV-2 infection rather than part of the inflammatory and reparative response [6] and may be directly linked to upregulation of fibrogenic responses such as TIMP-1, TGF-β, and collagen I during the inflammatory phase of SARS-CoV-2 infection (Figure 6). Excess MMP activity following viral infections may be linked to host morbidity or mortality and increased viral or persistence [14]. Developing treatments like pentoxifylline [5] and pirfenidone [15] to manipulate SARS-CoV-2–altered MMPs and increases in TGF-β could reduce the burden of lung injury in COVID-19 and decrease hospitalizations, development of acute lung injury and ARDS, and the incidence of long-term complications. Further (pre)clinical studies are needed to understand the impact of different SARS-CoV-2 variants on tissue remodeling and matrix equilibrium and whether therapeutic targeting of MMPs and TGF-β can attenuate fibrotic progression in severe COVID-19.

Figure 6.

Overall hypothesis. Emerging severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants induce unique in vivo changes in lung tissue remodeling in coronavirus disease 2019 (COVID-19). Compared to lung tissue of persons with no SARS-CoV-2 infection, SARS-CoV-2 infection and particularly variants that are associated with severe SARS-CoV-2 infection, such as the Delta variant, induce the gelatinases matrix metalloproteinases 1 and 9 (MMP-1 and MMP-9) in the early noninflammatory phase of SARS-CoV-2 infection when the virus replication is the highest and there is no significant inflammation in the lung tissue. SARS-CoV-2 and particularly variants that are associated with less severe SARS-CoV-2 infection, such as the Omicron variant, independently upregulate the tissue inhibitor of metalloproteinases 1 (TIMP-1) that may also greatly influence inflammatory responses and lung injury in SARS-CoV-2 infections by promoting the maintenance of the extracellular matrix and downregulation of MMP-12 in the late inflammatory phase of SARS-CoV-2 infection when the virus replication is the lowest and there is significant inflammation in the lung tissue. MMP-9 may directly induce the profibrotic cytokine transforming growth factor beta (TGF-β) that promotes formation of collagen I and lung fibrosis. The SARS-CoV-2–induced alterations in lung levels of MMPs and TIMP-1 in early SARS-CoV-2 infection are independent of development of lung inflammation and fibrosis and may contribute to development of severe COVID-19, lung fibrosis, acute lung injury, and acute respiratory distress syndrome (ARDS). Therapeutic targeting of MMPs and TGF-β can attenuate fibrotic progression in severe COVID-19.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Acknowledgments. This work was supported in part by the National Institutes of Health (grant number R01AG059501 to T. K.) and the California HIV/AIDS Research Program (OS17-LA-002 to T. K.). The isolate hCoV-19/USA/MD-HP01542/2021 (lineage B.1.351, Beta variant), NR-55282, was contributed by Andrew S. Pekosz. The hCoV-19/USA/PHC658/2021 (lineage B.1.617.2; Delta variant), NR-55611 was contributed by Dr Richard Webby and Dr Anami Patel.

Author contributions. Conceptualization: T. K. Methodology: M. W., C. H., M. B. S., H. V., M. D., L. F., G. A. F., T. K. Investigation: M. W., C. H., M. B. S., H. V., M. D., L. F., G. A. F., T. K. Visualization: M. W., T. K., L. F. Funding acquisition: T. K. Project administration: T. K. Supervision: T. K. Writing—original draft: M. W., C. G., C. H., T. K. Writing—review and editing: T. K., L. F.

Data availability. All data needed to understand and assess the conclusions of this research are available in the main text and supplementary materials. Datasets supporting the findings of this study are available from the corresponding author on reasonable request.