Involvement of the extracellular matrix proteins periostin and tenascin C in nasal polyp remodeling by regulating the expression of MMPs

Abstract Background Tissue remodeling caused by increased MMPs is involved in the pathogenesis of chronic rhinosinusitis with nasal polyposis (CRSwNP). We previously found higher levels of periostin and tenascin C in CRSwNPs, but whether they are associated with the dysregulation of MMPs is unknown. Therefore, the present study aimed to investigate the regulatory roles of these two ECM proteins in the expression of MMPs in nasal polyps. Methods The concentrations of MMP‐2, MMP‐3, MMP‐7, MMP‐8, MMP‐9, MMP‐12, MMP‐13, TIMP‐1, TIMP‐2, TIMP‐3, TIMP‐4, periostin, and tenascin C in tissue homogenates of 51 patients with chronic rhinosinusitis with and without nasal polyps and 15 control subjects were measured and were analyzed by adjusted logistic regression and spearman correlation test. Primary human nasal polyp fibroblasts and epithelial cells were stimulated ex vivo with periostin and tenascin C and the gene expression of MMPs and TIMPs was determined by means of real‐time PCR. Results The protein levels of MMP‐3, MMP‐7, MMP‐8, MMP‐9, TIMP‐1, TIMP‐2, periostin, and tenascin C were significantly higher in patients with CRSwNPs than in healthy control subjects. The adjusted logistic regression analyses showed that MMP‐3, MMP‐7, MMP‐8, MMP‐9, TIMP‐2, periostin, and tenascin C were related to the occurrence of CRSwNP. Spearman correlation test showed periostin was positively correlated with MMP‐3 and TIMP‐2, and tenascin C was positively correlated with MMP‐3, MMP‐7, MMP‐8, MMP‐9, and TIMP‐2. Periostin stimulated the gene expression of MMP‐3, MMP‐7, MMP‐8, and MMP‐9 in fibroblasts and MMP‐9 in epithelial cells ex vivo. Tenascin C stimulated the expression of MMP‐3, MMP‐7, MMP‐8, and MMP‐9 in epithelial cells. The expression of TIMPs in fibroblasts and epithelial cells was affected by neither periostin nor tenascin C. Conclusions Periostin and tenascin C might be involved in the remodeling of nasal polyps by regulating the expression of different MMPs in epithelial cells and fibroblasts. Our findings have the potential to identify key factors of tissue remodeling in CRSwNPs.


| BACKGROUND
Chronic rhinosinusitis (CRS) is a complex inflammatory disease in the upper airways characterized by 12 weeks of persistent symptoms involving nasal congestion, nasal discharge, facial pressure, loss of olfaction, cough, and fatigue. CRS affects approximately 11% of adults in Europe and about 12% of adults in the United States. 1 Phenotypically, CRS is classified as CRS with nasal polyps (CRSwNP) and CRS without nasal polyps (CRSsNP), which represents about 20% and 80%, respectively, of patients with CRS. 2 The prevalence of chronic rhinosinusitis with nasal polyps (CRSwNP) in Europe is estimated to be between 2.1% (France) and 4.4% (Finland) and is 4.2% in the United States. 1 Compared with CRSsNP, CRSwNP manifests a higher disease severity and a higher risk of asthma comorbidity, and also showed a high risk for recurrence. 3,4 CRSwNP is characterized by the formation of edematous stroma and pseudocysts. 5 Evidence shows that tissue remodeling is one of the main causes of the formation of nasal polyps. 6 Tissue remodeling is a dynamic process which results in both extracellular matrix (ECM) production and degradation. 7 Matrix metalloproteinase (MMPs) are zinc-dependent and calcium-dependent endopeptidases that are known to degrade and remodel ECM, which can be inhibited by the tissue inhibitors of metalloproteinase (TIMPs).
Physiological remodeling in nasal mucosa plays an important role in wound repair, of which the histological features involve fibroblast proliferation, angiogenesis, and increased connective tissue formation. 7,8 Abnormal ECM production and degradation may contribute to pathological reconstruction with the formation of pathological tissue.

For instance, excessive collagen deposition results in fibrosis in
CRSsNP and higher levels of MMP-induced tissue remodeling contribute to expansive histologic changes in CRSwNP. 9 MMP-induced tissue remodeling is strongly associated with ECM proteins, which play diverse roles and modulate cell-matrix interactions to control cellular metabolism within the ECM. 10 The ECM proteins are readily up-regulated under pathological conditions. 11 We previously found higher levels of ECM proteins, tenascin C and periostin, in NPs compared with controls. 12 Periostin is a confirmed novel biomarker for the formation of nasal polyps and tenascin C is an indicator of inflammation. 10,13 However, whether the two ECM proteins contribute to the formation of CRSwNP via tissue remodeling is unknown. Previous studies have demonstrated higher levels of MMPs in NP tissues 9,14 and the association between the two ECM proteins (periostin and tenascin C) and the expression of MMPs beyond nasal mucosa. 15,16 Together with evidences of the expression of MMPs in fibroblasts and epithelial cells isolated from nasal mucosa 17 Paper on Rhinosinusitis and Nasal Polyps (EPOS2020). 19 None of the study subjects had any history of malignancy, cystic fibrosis, ciliary dyskinesia, allergic fungal sinusitis, maxillary antrochoanal polyps, upper or lower respiratory tract infections within 2 weeks preoperatively, or autoimmune diseases. Subjects who had taken systemic or local glucocorticosteroids or antibiotics within the last 2 weeks were excluded from the study. Subjects undergoing rhinoseptoplasty because of anatomic variations were recruited as control subjects.
Tissue samples were obtained from the inferior turbinates of the control subjects, the ethmoid mucosae of patients with CRSsNP, and the NPs of patients with CRSwNP. These samples were frozen and stored at −80°C until used for immunoassays.

| Clinical data
A complete blood count was taken, and a computed tomography (CT) scan was performed for the enrolled patients. The percentage of eosinophils and neutrophils in circulating blood was determined using

| Tissue eosinophils and neutrophils counting
Samples of nasal tissues from patients with CRSwNP, CRSsNP and control were processed for histological evaluation by H&E stain. All stained samples were observed by two independent pathologists, who were blinded to the clinical diagnosis and characteristics of the patients. Eosinophils and neutrophils were assessed by bright-field light microscopy (BX51, Olympus) at �400 magnification. The counts were recorded as the mean of the counts for 10 nonoverlapping fields in the lamina propria.

| Immunoassay
Tissue homogenates were prepared as previously described. 21 Briefly, frozen nasal tissues were weighed and homogenized with an

| Primary human nasal polyp fibroblast cell culture and ex vivo stimulations
Fibroblasts in nasal polyps were isolated based on a previously described method. 22 Briefly, fresh nasal polyp samples were obtained from CRSwNP patients and were washed several times with phosphate-buffered saline (PBS), supplemented with 200 U/mL penicillin and 200 U/mL streptomycin. Samples were diced and plated in 100-mm tissue culture plates containing Hyclone RPMI 1640 medium with 10% fetal bovine serum (FBS; Gibco) and 100U/mL of penicillin and streptomycin. When a monolayer of fibroblast-like cells was found to be confluent, the cells were digested and passaged.
After three passages, the cells were stimulated with 1 ug/ml, 2 ug/ml, and 5 ug/ml of periostin or tenascin C for 24 h, with the culture medium used as the control. The collected cells were stored at −80°C for further gene expression detection. The cells were characterized by flow cytometry using anti-human CD90-FITC and anti-human CD45-PerCP antibodies (Miltenyi Biotech) as previously described. 23 The purity was more than 96%.

| Primary human nasal polyp epithelial cells culture and ex vivo stimulations
Nasal polyp epithelial cells were established according to a previously described method. 24 Briefly, fresh nasal polyp samples were rinsed with PBS and digested in 1 mg/ml protease (protease from Streptomyces griseus, Type XIV; Sigma-Aldrich) for 1 h at 37°C. Cell suspensions were centrifuged at 800 rpm for 5 min and resuspended in bronchial epithelial growth medium (BEGM, Lonza, Basel). Cells were plated for 1 h on 100-mm tissue culture plates to remove contaminating fibroblasts. The isolated cells were seeded on rat-tail collagencoated tissue culture plates at 37°C with 5% CO 2 . The culture medium was changed every other day until the cells reached confluence.
The isolated nasal epithelial cells were trypsinised and seeded into 12-well culture plates at a concentration of 5 � 10 5 cells/mL in 1 ml BEGM with 10% FBS and 100 U/mL of penicillin and streptomycin.
After reaching 80% confluence, the cells were stimulated with 1, 2, and 5 ug/ml of periostin or tenascin C for 24 h, with the culture medium used as the control. The collected cells were stored at −80°C for further gene expression detection.

| Real-time PCR
Total RNA from unstimulated and stimulated fibroblasts and epithelial cells was extracted using TRIzol reagent (Ambion-Life Technologies). The RNA was reverse transcribed into first-strand cDNA with random primer, and real-time polymerase chain reaction (PCR) was subsequently performed using the ABI7500 PCR system (Applied Biosystems, Foster City, Calif). Primer sequences are given in Table S1. The PCR conditions were as follows: a 95°C denaturation step for 10 min followed by 40 cycles of 95°C DU ET AL.
-3 of 11 denaturation (15 s) and 60°C annealing (1 min). Gene expression was normalized to the housekeeping gene β-actin. The comparative cycle threshold (Delta Delta Ct) method was used for relative gene expression analysis.

| Statistical analysis
All statistical data were analyzed using IBM SPSS Statistics, Version

| Demographic and clinical characteristics
The clinical characteristics of the patients and controls are presented in the Table 1. There were no significant differences in the age, sex, smoker, asthma, aspirin-exacerbated respiratory disease (AERD), and tissue and blood neutrophils among the three groups. The proportion of patients with atopy was significantly higher in patients with CRSsNP and CRSwNP than in controls. LMS scores and the levels of tissue and blood eosinophil (Eos) were significantly higher in patients with CRSsNP and CRSwNP than in controls. The counts of tissue Eos were significantly higher in patients with CRSwNP compared with CRSsNP.

| Expression of MMPs, TIMPs and ECM proteins in nasal tissue homogenates from patients with CRSsNP and CRSwNP
The protein levels of MMP-2, MMP-3, MMP-7, MMP-8, MMP-9, MMP-12, MMP-13, TIMP-1, TIMP-2, TIMP-3, TIMP-4, periostin and tenascin C were examined in tissue homogenates from patients with CRS. Figure 1 shows the results. In comparison with healthy control subjects, no difference was found in patients with CRSsNP for all these MMPs and TIMPs, while increased levels of MMP-3, MMP-7, MMP-8, MMP-9 and TIMP-1, and TIMP-2 were found in patients with CRSwNP. The levels of MMP-8 and MMP-9 were higher in patients with CRSwNP than in patients with CRSsNP, but the level of TIMP-1 was lower. As significant differences for atopy, LMS and the levels of tissue and blood Eos were found among the three groups, we investigated whether the MMP levels were related with these parameters. No relationship between the atopy and MMPs was found (data not shown). We identified that tissue and blood eosinophilic inflammation was significantly related with the levels of MMP-3 and MMP-7 ( Table 2). Additionally, significant correlations of LMS with MMP-3, MMP-7, MMP-8, MMP-9, TIMP-1, and TIMP-2 were found ( Table 2).
In comparison to healthy control subjects, both periostin and tenascin C were up-regulated in patients with CRSwNP, but not in patients with CRSsNP ( Figure 2). Additionally, a significant higher levels of tenascin C were found in patients with CRSwNP versus patients with CRSsNP.

| The association of MMPs, TIMPs, and ECM proteins with CRSwNP
Further, we investigated whether the levels of MMPs, TIMPs, and ECM proteins that showed significant differences between CRSwNPs, CRSsNPs and controls, were associated with the formation of nasal polyps. Using univariate logistic regression analysis, we found that MMP-3, MMP-7, MMP-8, MMP-9, TIMP-2, periostin, and tenascin C were associated with CRSwNPs (

| Correlations between MMPs and TIMPs and ECM proteins
Next, we investigated the correlations between these MMPs and TIMPs and the ECM proteins. As shown in Table 4 TIMPs at the gene level in the fibroblasts (Figure 3).

F I G U R E 2
The protein levels of the extracellular matrix proteins, periostin and tenascin C, in tissue homogenates from healthy control subjects and patients with CRSsNP and CRSwNP. Control subjects, n = 15; patients with CRSsNP, n = 14; patients with CRSwNP, n = 37. Data are presented as median and interquartile range. CRSsNP, chronic rhinosinusitis without nasal polyposis; CRSwNP, chronic rhinosinusitis with nasal polyposis   31,32 In this study, we found higher levels of MMP-9 in CRSwNP versus CRSsNP and identified that periostin induced the expression of MMP-9 in fibroblasts, suggesting that periostin might be able to mediate, at least in part, the pathogenesis of CRSwNP via regulating the production of MMP-9 in nasal fibroblasts.

| Effect of periostin and tenascin C treatment on the gene expression of MMPs and TIMPs in nasal polyp-derived primary epithelial cells ex vivo
Tenascin C, as an ECM protein, could be increased in parallel with MMPs in some pathological states. 16 However, the regulatory role of tenascin C on MMPs in nasal polyps remains unknown. In the present study, we found that tenascin C induced the expression of MMP-3, MMP-7, MMP-8, and MMP-9 in the nasal epithelium. There is substantial evidence that tenascin C contributes to tissue remodeling via the upregulation of MMP-9 in the mouse model of the cardiovascular system (e.g., cardiac remodeling, hepatic ischemia/reperfusion, subarachnoid hemorrhage, etc.). 16,29,33,34 Recently, Kanagala et al. 35  IgE also could induce remodeling via fibroblasts and anti-IgE treatment inhibited the process. 38 It indicates that targeting the upstream signaling could control tissue remodeling. However, which factors regulates MMP-induce remodeling within nasal polyps is still unclear.
We previously found that IL-4 and IL-13 could induce periostin expression in nasal polyps. 12 In the present study, we verified that periostin might promote tissue remodeling via inducing the expression of MMPs. Our results indicates that anti-IL-4/IL-13 treatment might have utility for controlling the remodeling process within nasal polyps, which has been confirmed to effectively treat severe CRSwNPs. 37 Given the roles of periostin and tenascin C in promoting the expression of MMPs in CRSwNPs, our results indicate that antiperiostin or anti-tenascin C treatment might be able to inhibit tissue remodeling within nasal polyps, which still needs to be further verified by in vivo animal studies.
To our knowledge, this is the first study to describe the role of periostin and tenascin C in regulating the expression of MMPs in NP.
DU ET AL.
-9 of 11 The degradation of ECM in polyp tissues induced by MMPs is a risk factor for the morbidity of CRSwNP. 5,13 In addition, the mechanisms by which MMPs are synthesized need to be further investigated. We revealed that fibroblasts and the nasal epithelium are both sources of MMPs within nasal polyps. TIMP-1 might be able to reduce the effects of the expression of MMP-9; however, the levels of MMP-9 and TIMP-1 were negatively correlated with disease severity in CRS. 39,40 Although higher levels of TIMP-1 and TIMP-2 were found in CRSwNPs, we failed to verify the regulatory roles of periostin and tenascin C on TIMPs. In this regard, further studies are required to illuminate how TIMPs are induced in nasal polyps. It has been known that CRSwNP patients prominently manifest eosinophilic inflammation 5 and periostin has been reported to have a role in orchestrating eosinophil infiltration. 41,42 Our data provide new insight into the role of periostin in the pathogenesis of CRSwNP, which promotes tissue remodeling via MMP production by fibroblasts during inflammation. Tenascin C contributes to tissue remodeling in a different manner. Tenascin C stimulates the production of MMPs mainly in the nasal epithelium, not in fibroblasts, and it mainly stimulates the production of MMP-8. Among the up-regulated MMPs in this study, MMP-9 is extensively confirmed to be increased in NPs. 43,44 We showed that periostin and tenascin C are both able to stimulate the expression of MMP-9 via fibroblasts and the nasal epithelium. There are several limitations to our study. First, the assessment of the tissue remodeling molecules was based only on the mRNA expression, which needs to be verified according to the levels of corresponding proteins in NP tissues. Second, the number of cases involved in the ex vivo experiments is relatively small. Third, we failed to study the role of periostin and tenascin C in fibroblasts and nasal epithelial cells in controls and CRSsNP, which could help to prove whether the remodeling ex vivo was observed only in CRSwNPs.
Finally, we failed to investigate the regulatory role of periostin and tenascin C on patients with different endotypes, including eosinophilic and non-eosinophilic CRSwNPs, which need further investigation.

| CONCLUSIONS
Our study showed the correlation of periostin and tenascin C with some members of the MMP family, which were confirmed to be increased in patients with CRSwNP and identify that both periostin and tenascin C have regulatory roles in the expression of these MMP, acting on different types of cells, mainly fibroblasts for periostin and nasal epithelium for tenascin C. We provide evidence for the pathogenic roles of periostin and tenascin C in the formation of nasal polyps. Our findings have the potential to identify key factors enhancing tissue remodeling, which will be necessary to further uncover the pathogenesis of CRSwNPs.