HYAL1 Is Downregulated in Idiopathic Pulmonary Fibrosis and Inhibits HFL-1 Fibroblast Proliferation When Upregulated

Background Idiopathic pulmonary fibrosis (IPF), the most common interstitial lung disease, arises from transforming growth factor beta 1- (TGFβ1-) induced aberrant fibroproliferation in response to epithelial injury. The TGFβ1-) induced aberrant fibroproliferation in response to epithelial injury. The TGF Methods We first performed microarray data mining of previously published gene expression datasets to identify key gene signatures in IPF lung tissues. HYAL1 expression levels in IPF and normal lung tissues were then characterized using immunohistochemistry followed by real-time quantitative reverse transcription-PCR (qRT-PCR) and western blot analysis on isolated fibroblasts from fresh lung tissues of IPF and healthy donors. A human fetal lung fibroblast HFL-1 cell line, which was used in place of primary lung fibroblasts, was used to assess the proliferative or apoptotic effects associated with lentiviral-induced HYAL1 overexpression using CCK-8 cell proliferation assay and Annexin V-APC staining. The identification of potentially associated molecular pathways was performed using microarray analysis followed by qRT-PCR and western blot analysis. Results Lung tissue microarray data mining and immunohistochemistry revealed significantly downregulation of HYAL1 in IPF lung tissue. However, HYAL1 in IPF lung tissue. However, HYAL1 in IPF lung tissue. However, HYAL1 in IPF lung tissue. However, β1-) induced aberrant fibroproliferation in response to epithelial injury. The TGFβ1-) induced aberrant fibroproliferation in response to epithelial injury. The TGF Conclusions We showed that HYAL1 overexpression could prevent HFL-1 fibroproliferation. Furthermore, our findings suggest that transcriptional regulators and BMP receptor signaling may be involved in HYAL1 modulation in IPF therapy.HYAL1 in IPF lung tissue. However,


Introduction
Idiopathic pulmonary fibrosis (IPF) is the most common type of interstitial pneumonia with an annual incidence rate of approximately 10 out of 100,000 cases [1,2]. IPF presents with a higher prevalence among men than women (ratio of 1.5-1.7 : 1) and primarily affects individuals aged 50-70 years old [1]. Typical symptoms of IPF include gas exchange dis-ruption and respiratory failure, which inevitably leads to death within 2-5 years after diagnosis [3,4]. As the etiology of IPF is poorly characterized [5,6] and the only effective treatment is lung transplantation [7], there is an urgent need to elucidate the mechanism of action. Excessive extracellular matrix (ECM) accumulation, formation of fibroblast and myofibroblast foci, and excessive fibroproliferation are the hallmarks of IPF [8,9]. The major ECM constituent, hyaluronic acid (HA), was found to be increased in IPF patients [10,11], and its ability to promote transforming growth factor beta 1-(TGFβ1-) mediated fibroblast proliferation [3] led to IPF pathogenesis.
Hyaluronidase 1 (HYAL1), also known as lung carcinoma protein 1 (LUCA-1), is the primary endoglycosidase that degrades HA [3,9]. The HYAL1 gene resides in chromosome 3p21.3 and is associated with tumor suppression suggesting its possible role in regulating cell proliferation [12]. The role of HYAL1 as a tumor suppressor or oncogene remains unclear [12][13][14][15][16][17][18] with reports of both HYAL1 downregulation in human endometrial cancer and lung and kidney carcinomas [12,13] and upregulation in breast, epithelial ovarian, and bladder tumors [16][17][18]. The overall function is believed to be tumor type or concentration dependent [19] as with prostate cancer; cell proliferation was suppressed at a high HYAL1 concentration and vice versa [19]. Further convoluting evidence explains that the role of HYAL1 in the same cancer type but different species can differ. For example, it promotes colorectal tumor progression in humans but inhibits cancer proliferation in rat models [14,20].
Abnormal tissue damage and repair often lead to tissue fibrosis [21]. IPF is the most common lung interstitial disease, and importantly, it is irreversible [22]. Hyaluronidases have shown efficacy in lung fibrosis therapy as they can increase the number of mesenchymal stem cell-like cells and reduce collagen deposition and production [23][24][25]. HYAL1 and TGFβ1 were found to play antagonistic roles: HYAL1 overexpression in murine L929 fibroblasts inhibited TGFβ1-mediated cell growth, while TGFβ1 inhibited the HYAL1-induced tumor necrosis factor-(TNF-) mediated cytotoxicity [26]. In light of these reports, we designed this study to clarify the effect and mechanism of HYAL1 expression on lung fibroblast proliferation and apoptosis. In our previous study, utilizing microarray, we identified gene expression changes in the IPF lung tissue genetic profile compared to normal tissue [27]. In the present study, we performed further analysis of the HYAL1 gene expression profiles that have been obtained from published data mining, immunohistochemistry on paraffin-embedded lung tissues, and real-time quantitative reverse transcription-PCR (qRT-PCR) and western blot analyses on isolated primary lung fibroblasts. We also investigated the effect of HYAL1 overexpression on HFL-1 fibroproliferation and apoptosis. Moreover, we performed microarray and bioinformatics analysis to identify key genes that were significantly altered within corresponding signaling pathways.

Materials and Methods
2.1. Data Mining. Data mining of the merged microarray data from our previous study of 48 IPF lung tissues (17 from GSE24206 data and 31 GSE10667 data) and 27 healthy controls (6 routine lung volume reduction after lung transplantation from GSE24206; 15 free margins with normal lung cancer histology from GSE10667; and 6 from the GSE16538 data) was performed [27][28][29][30]. Different genes with altered profiles were selected using an unpaired t test (P < 0:0001), and the general gene expression pattern was extracted through Euclidean distance similarity metric and complete linkage algorithm-based two-way clustering method using the Cluster 3.0 and Java TreeView software (Stanford University, USA). Gene expression data of significantly altered genes were further analyzed with the supervised classification method support vector machine (SVM) algorithm, which outputs a gene expression model using multtest and kernlab packages (New York University and Rockefeller, USA). Significantly altered genes with weight coefficients above specific cutoff values (0.9 for lung tissue; 0.1 for HFL-1 fibroblasts) were selected. Next, the gene expression data which met the selection criteria were dimensionally reduced through principal component analysis (PCA), and genes with higher loading coefficient (coefficient > 0:95 for HFL-1 fibroblasts) were collected as gene signatures. The PCA was completed using the prcomp command. Principal components with cumulative proportion of 85% were subjected to Bayesian probit regression for significance analysis. All analyses were completed under the R 3.5.0 environment.

Ethics Statement and Tissue
Specimens. Experiments and subsequent data analysis were performed with approval from the Beijing Chao-Yang Hospital Ethics Committee (2017-Science-10). The Committee waived the need to obtain informed consent from subjects whose samples were used. The IPF lung tissue samples were procured from three patients diagnosed in accordance with the criteria approved by the American Thoracic Society [31]. The healthy control lung tissue samples were obtained from one healthy donor for lung transplantation and two disease-free margins of non-small-cell lung cancer patients which showed normal histology. IPF primary lung fibroblasts were isolated from surgical lung biopsy samples from three IPF patients. Normal primary lung fibroblasts were isolated from lung transplant explants of three healthy donors. The details of all patients are summarized in Table 1. 2.3. Immunohistochemistry. IPF and normal human lung tissue specimens (4 μm thick) were formalin-fixed and paraffinembedded on microslides. Tissue sections were treated for 1 h at 72°C, deparaffinized in xylene for 10 min twice, and rehydrated by soaking in 100%, 95%, and 75% ethanolwater solutions for 2 min each. Antigen retrieval was performed by autoclaving the samples in preheated EDTA pH 8.0 at 100% power for 2.5 min. The tissue sections were then incubated in rabbit polyclonal anti-HYAL1 primary antibody (1 : 500; Abcam 203293, UK) at 4°C overnight, followed by incubation with HRP-labeled goat anti-rabbit IgG secondary antibody (ZSGB-BIO PV-6001, China) at room temperature for 20 min. Subsequently, the sections were stained with DAB (ZSGB-BIO ZLI-9017, China) at room temperature for 5 min and then counterstained with hematoxylin (ZSGB-BIO ZLI-9017, China) at room temperature for 20 min before examination under an optical microscope (Olympus CX31, USA). The HYAL1 expression in lung tissues by immunohistochemistry was analyzed by average integrated optical density (IOD) using Image-Pro Plus 6.0 software (Media Cybernetics, USA).

Primary Lung Fibroblast Cell
Culture. Human lung tissues of both IPF patients and healthy donors were minced and allowed to adhere to the bottom of the culture flask. Samples were cultured for 4 h after attachment to the culture flask bottom, in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS; Gibco, USA), 100 mg/mL streptomycin, and 100 U/mL penicillin at 37°C in a 5% CO 2 humidified incubator. The cell culture medium was refreshed every 3 days, and the cells were passaged at a 1 : 2 ratio every 3 days.
2.5. HFL-1 Cell Culture. Human fetal lung fibroblast HFL-1 (Type Culture of the Chinese Academy of Science, Shanghai, China) cells in F-12K medium (Invitrogen, USA) supplemented with 10% (v/v) fetal bovine serum were cultured at 37°C in a 5% CO 2 humidified incubator. The cell culture medium was replaced every 2 days, and the cells were passaged at a 1 : 2 ratio every 3 days.

Gene Expression Profiling and Data Preprocessing.
Gene expression profiling of HFL-1 fibroblasts transiently infected with the lentivirus Hyal1 and the control were performed using GeneChip® PrimeView™ Human Gene Expression Array (Affymetrix, USA). Total RNA was extracted from cells using the Total RNA Isolation Reagent Trizol (Superfect TRI™, Shanghai, China). The RNA concentration and purity were determined using the NanoDrop 2000 (Thermo Fisher, Massachusetts, USA), and the RNA integrity was assessed using the 2100 Bioanalyzer (Agilent, California, USA). In vitro transcription (IVT) was then performed to obtain fragmented biotin-labeled RNA according to the manufacturer's instructions (GeneChip 3 ′ -IVT Express Kit, Affymetrix, California, USA). Next, the fragmented biotinlabeled RNA was hybridized to the Affymetrix chip and measured with GeneChip Scanner 3000 (Thermo Fisher, Massachusetts, USA). Data were saved as .CEL files, which could be accessed through the GEO repository (http://www.ncbi .nlm.nih.gov/geo/) with the accession number GSE131443 (the data are scheduled to be released on Jun 30, 2020). We performed triplicate measurements of the test (overexpressed HYAL1: N5948-1, N5948-2, and N5948-3) and control (vector CON335-infected HFL-1 fibroblasts: N5949-1, N5949-2, and N5949-3) groups using the same Affymetrix chip. Genes with significantly altered expression profiles between the HYAL1 overexpressed group and the negative control group (false discovery rate ðFDRÞ < 0:05) were identified using empirical Bayesian linear modeling. The Benjamini-Hochberg method was employed to calculate the FDR. Supervised classification using the SVM algorithm and gene signature extraction were also performed as described in the data mining subsection.
2.8. Gene Annotation and Ingenuity Pathway Analysis. Gene set enrichment analysis (GSEA) was performed to identify genes with significantly altered expression profiles in HFL-1 fibroblasts with overexpressed HYAL1 (loading coefficients < −0:95). The online software "The Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.8" (https://david.ncifcrf.gov/) was employed. The gene interaction networks and the regulatory mechanism of HYAL1 gene expression were subsequently predicted with the Ingenuity Pathway Analysis (IPA) software (Qiagen, Germany).

2.9.
Real-Time Quantitative RT-PCR (qRT-PCR). Total RNA extracted from primary fibroblasts or HFL-1 cells was reverse transcribed to cDNA using the M-MLV kit (Promega, USA) and oligo(dT) primers (Sangon Biotech (Shanghai) Co. Ltd., China). Next, cDNA amplification was performed using the SYBR Premix Ex Taq DNA Polymerase (Takara, Japan). The housekeeping gene human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the internal control. The sequences of the primers used in RT-qPCR are shown in Table 2. qRT-PCR was performed using the Roche Light Cycler 480II (Switzerland), and the gene expression levels were normalized with reference to GAPDH using the Cq method [32].

CCK8 Cell Proliferation and Apoptosis
Assays. Cell proliferation was assessed using the Cell Counting Kit-8 (CCK8) assay (Sigma, USA) according to the manufacturer's instructions. Infected HFL-1 fibroblasts (100 μL; 2 × 10 3 cells/well) collected at the logarithmic phase were seeded in 96-well plates and then incubated for 1, 2, 3, 4, and 5 days. Triplicate measurements of the absorbance readings were performed at 450 nm daily for the tested duration. Apoptotic HFL-1 fibroblasts infected with lentivirus carrying either recombinant HYAL1 plasmid or empty vector control were detected with Annexin V-APC (eBioscience, USA) and flow cytometry (Millipore, USA).

HYAL1 Is a Key Signature Gene That Is Significantly
Downregulated in IPF Lung Tissue. In this study, we applied SVM and PCA methods to the normalized microarray data obtained from our previous study [27] to extract key gene signatures in IPF lung tissue from controls. We identified 413 key gene signatures (weight coefficients > 0:9), and they were divided into IPF and control lung tissue groups according to gene expression estimates (data not shown). Compared with our previous study, PCA identified three additional outliers: "Advanced IPF explant lower lobe rep146" samples from GSE24206 and "UIP biological rep22" and "AEx biological rep5" samples from GSE10667. Furthermore, 229 key gene signatures (loading coefficient > 0:7) were identified and perfectly distinguished 35 IPF from 20 control lung tissue samples (data not shown). From the two-way clustering heat map (Figure 1(a)), we identified the top 50 key gene signatures (25 highest positive loading coefficients and 25 highest negative loading coefficients). HYAL1 was among the top 25 negative key gene signatures with an SVM weight of 0.937 and PCA loading coefficient of -0.696. Significantly reduced HYAL1 expression was noted in IPF lung tissues compared with healthy controls (P < 0:000001; unpaired t test; Figure 1(b)).

HYAL1 Expression Was Inhibited in IPF Lung Tissues
While Upregulated in IPF Fibroblasts at the mRNA Level Only. For formalin-fixed and paraffin-embedded lung tissues, HYAL1 was predominantly expressed in the pulmonary alveoli type II alveolar cells of the negative control paracarcinoma tissue or healthy donor lung tissue (Figure 2(a)). In contrast, when fibroblast accumulation occurred in the IPF lung tissue, HYAL1 expression in the alveolar cells and in the accumulated fibroblasts was significantly reduced (Figure 2(b)). Healthy lung tissues were mainly composed of intact alveolar epithelial cells and few lung fibroblasts, unlike in IPF lung tissues where the main components were accumulated lung fibroblasts and few intact alveolar epithelial cells. The average IOD of HYAL1 in IPF was decreased compared to normal  Control13   IPF14  IPF16  IPF18  IPF24  IPF20  IPF10  IPF32  IPF30  IPF19  IPF23  IPF31  IPF3  IPF2  IPF21  IPF29  IPF13  IPF17  IPF9  IPF34  IPF12  IPF8  IPF11  IPF15  IPF7  IPF27  IPF1  IPF5  IPF25  IPF26  IPF33  IPF22  IPF4  IPF28  IPF35   TMEM100  SLC6A4  SLCO1A2  CYP3A5  STXBP6  NDRG4 Figure 2(c)). HYAL1 expression levels in IPF and healthy lung tissues typically depend upon cell types. When compared with fibroblasts from healthy lung tissues, HYAL1 was significantly upregulated in IPF lung fibroblasts at the mRNA level (P < 0:05; unpaired t test; Figure 3(a)) but not at the protein level (P > 0:05; unpaired t test; Figure 3(b)).

HYAL1 Upregulation Induces Alteration in Gene
Expression in HFL-1 Fibroblasts. A total of 1,664 differentially expressed genes (953 upregulated and 711 downregulated) were identified in HFL-1 fibroblasts (FDR < 0:05; Figure 5(a) and Table S1): 11 genes were noted to have up to a 1.5-fold change in gene expression level. Of these 1,664 genes, 888 key gene signatures showed weight greater than 0.1 ( Figure 5(b)). Of the 888 gene signatures, 339 possessed loading coefficients greater than 0.95 ( Figure 5(c)), and 90 genes were transcriptionally silent. All gene lists in each analysis are shown in Table S2.

Discussion
The role of hyaluronidases in lung fibrosis is evident from the noted therapeutic applications [33] and antifibrotic effects in bleomycin-induced lung fibrosis in mice [34]. In addition, HYAL2 was found to regulate the splicing of CD44 pre-mRNA, which determines the antifibrotic phenotype [35]. Together, these previous reports provided the motivation for our investigation into the mechanistic role of hyaluronidase in pulmonary fibrosis. Furthering our interest, HYAL1 and HYAL2 were previously reported as dual blocking agents for TGFβ1-enhanced cell growth in murine L929 fibroblasts [26]. As the antifibrotic effects of hyaluronidases in humans are largely unknown, we compared the gene expression profiles of IPF and control lung tissues. HYAL1 expression was specifically measured in lung tissues and isolated lung fibroblasts. HYAL1 inhibitory effects and its induced gene signatures as well as the corresponding signaling pathways were measured using HYAL1-overexpressed HFL-1 fibroblasts.
Through previous data mining with SVM and PCA methods, we identified three IPF lung tissue sample outliers The P values were calculated using the unpaired t test. * P < 0:05; * * P < 0:01.    [27]. In this study, clustering validation of 35 IPF (the three outlier IPF clusters were omitted) and 20 control samples was performed using the top 25 positive and 25 negative key gene signatures to evaluate the goodness of a cluster. Consistent with our previous study [27], no sample type preference was identified. HYAL1 was one of the top 25 genes that was significantly downregulated in IPF lung tissues, which supports the reported antifibrotic effect of hyaluronidases in lung fibrosis [34,36]. Our observation of reduced HYAL1 expression in IPF lung tissue was consistent with reports of the absence of HYAL1 mRNA in lung carcinoma cell lines [12] and reduced HYAL1 protein levels in endometrial carcinomas [13]. Given the similarities between IPF and cancers [37,38], HYAL1 downregulation in IPF may lead to malignancy. Moreover, HYAL1 expression level could potentially be a biomarker of IPF pathogenesis.
HYAL1 expression was reduced in IPF lung tissues; however, an upregulation in IPF fibroblasts was noted at the mRNA level, but not at the protein level. The conflicting results could probably be due to reduced alveolar epithelial cells in IPF lung tissue, which is the cell type that predominantly expresses HYAL1. Hence, despite upregulated HYAL1 mRNA in IPF fibroblasts, as well as the accumulation of fibroblasts, these could not alter HYAL1 downregulation in IPF lung tissue. In the present study, we used HFL-1 cells to mimic the mechanism of HYAL1-mediated regulation of fibroblast proliferation or apoptosis in primary lung fibroblasts.
Overexpression of HYAL1 and HYAL2 enhanced TNF cytotoxicity and ultimately reversed TGFβ1 effects in murine normal lung fibroblast cell line L929 [26]. This was consistent with our observations of inhibited fibroproliferation in HYAL1-overexpressed HFL-1 cells. The regulatory effects of HYAL2 on gene splicing [35] also indicate the potential role of HYAL1 in modulating gene expression as well as signaling pathways in lung fibroblasts. Future studies should focus on characterizing the effect of HYAL1 on IPF lung fibroproliferation. In contrast, hyaluronan synthases (HASs) that produce HA in response to tissue injury promote severe lung fibrosis [39]. HAS2, the major isoform that controls HA production, has been reported as upregulated in IPF fibroblasts [21]. HA fragments stimulate TGFβ-mediated oral mucosal fibroblast proliferation [40]. HYAL1 downregulation in IPF lung tissues in this study indicated the antagonistic actions of HYAL1 and HAS2. Therefore, HYAL1 could potentially affect the pathogenesis of lung fibrosis through counteracting HASs. Future investigations of the exact mechanism of HYAL1-mediated regulation of fibroproliferation through HA would clarify the role of HYAL1 in modulating ECM in IPF.
HYAL1-overexpressed HFL-1 fibroblasts were engineered to elucidate the mechanism of HYAL1-mediated regulation of signaling pathways in lung fibroblasts. Microarray analysis and data mining revealed genes with significantly altered expression levels in HYAL1-overexpressed HFL-1 fibroblasts compared with the normal control tissue. In addition, we identified genes that were likely inhibited by HYA-L1overexpression. Of note, seven transcription regulators were annotated and clustered in the term group of negative transcription regulation, and five of the seven identified factors were validated by real-time qRT-PCR: Dachshund 1 (DACH1), TRIM24, TBX3, KHDRBS1, and RUNX1 translocation partner 1 (RUNX1T1). Of these five transcription regulators, DACH1,

11
BioMed Research International a DNA-binding transcription factor that regulates gene expression and cell fate, showed the greatest decrease in expression (fold change ðFCÞ = −2:088; P < 0:01; unpaired t test). DACH1 expression exhibits cancer-type dependency wherein it was found to be downregulated in lung adenocarcinoma [41] but upregulated in ovarian cancer [42]. DACH1 All P values were calculated using the unpaired t test. * * P < 0:01.

12
BioMed Research International inhibits lung cancer through binding to the p53 gene [43]. As HYAL1 was reported to have no effect on p53-mediated cell death [26], our finding suggests a novel and indirect association between HYAL1 and p53 through DACH1 regulation. TRIM24, TBX3, and KHDRBS1 were upregulated in lung cancer [44][45][46]. Hence, we noted their significant downregulation in HYAL1-overexpressed HFL-1 cells, which is consistent with their role in inducing fibroproliferation. The association between RUNX1T1 and lung cancer is unclear, but its binding partner RUNX1 was overexpressed in nonsmall-cell lung cancer [47]. Hence, HYAL1-induced inhibition of lung fibroblast proliferation may occur through the downregulation of these transcription regulators.
Of note, IPA analysis predicted that the transcription factor SMARCA4 could be a central node in the gene interaction network in HYAL1-overexpressed HFL-1 fibroblasts. SMARCA4 had a 1.102-fold downregulation in our microarray data (FDR < 0:01; Benjamini-Hochberg method), which was further confirmed using qRT-PCR to be a 1.550fold downregulation (P < 0:01; unpaired t test). Further investigation demonstrated that SMARCA4 played important roles in cell proliferation as evident from its upregulation in human IPF lung tissue [48,49] and its downregulation in adenocarcinoma NCI-H522 cells [50]. SMARCA4 could potentially regulate cell proliferation at the transcriptional level through chromatin remodeling [51,52]. SMARCA4 downregulation and relative position as the central node in the network suggests its potential in inhibiting lung fibroblast proliferation downstream of HYAL1. Future studies on SMARCA4 expression including the overall effect on fibroblast proliferation should be performed to elucidate the mechanistic role in HYAL1-mediated inhibition of fibroblast proliferation. These prescribed finding will aid in obtaining new insights into the effect of hyaluronidases in IPF therapy.

Conclusions
This is the first study demonstrating the association between HYAL1 and IPF and the negative correlation between HYAL1 overexpression and HFL-1 fibroproliferation. In addition, our findings suggest that HYAL1-induced inhibition of fibroblast proliferation may involve changes in transcription factor levels as well as BMPR2 signaling.

Data Availability
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.