Heparin-Binding Protein 17/Fibroblast Growth Factor-Binding Protein-1 Knockout Inhibits Proliferation and Induces Differentiation of Squamous Cell Carcinoma Cells

Simple Summary Fibroblast growth factor (FGF) plays an important role in tumor growth by inducing angiogenesis in addition to promoting the proliferation of squamous cell carcinoma (SCC) and oral squamous cell carcinoma (OSCC) cells. Heparin-binding protein 17/fibroblast growth factor-binding protein-1 (HBp17/FGFBP-1) purified from A431 cell-conditioned media based on its capacity to bind to FGF-1 and FGF-2 is recognized as a pro-angiogenic molecule as a consequence of its interaction with FGF-2. In this study, we have examined the functional role of HBp17/FGFBP-1 in A431 and HO-1-N-1 cells using the CRISPR/Cas9 technology. Our results showed that HBp17/FGFBP-1 knockout inhibited cell proliferation, colony formation, and cell motility compared to control. The amount of FGF-2 was decreased in culture medium conditioned by HBp17/FGFBP-1 knockout cells compared to control. We performed cDNA/protein expression analysis followed by Gene Ontology and protein–protein interaction analysis. The results demonstrate that both gene and protein expression related to epidermal development, cornification, and keratinization were upregulated in HBp17/FGFBP-1-knockout A431 and HO-1-N-1 cells. Abstract Heparin-binding protein 17/fibroblast growth factor-binding protein-1 (HBp17/FGFBP-1) has been observed to induce the tumorigenic potential of epithelial cells and is highly expressed in oral cancer cell lines and tissues. It is also recognized as a pro-angiogenic molecule because of its interaction with fibroblast growth factor (FGF)-2. In this study, we examined the functional role of HBp17/FGFBP-1 in A431 and HO-1-N-1 cells. Originally, HBp17/FGFBP-1 was purified from A431 cell-conditioned media based on its capacity to bind to FGF-1 and FGF-2. We isolated and established HBp17/FGFBP-1-knockout (KO)-A431 and KO-HO-1-N-1 cell lines using the clusters of regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated protein 9 (Cas9) gene editing technology. The amount of FGF-2 secreted into conditioned medium decreased for A431-HBp17-KO and HO-1-N-1-HBp17-KO cells compared to their WT counterparts. Functional assessment showed that HBp17/FGFBP-1 KO inhibited cell proliferation, colony formation, and cell motility in vitro. It also inhibited tumor growth in vivo compared to controls, which confirmed the significant difference in growth in vitro between HBp17-KO cells and wild-type (WT) cells, indicating that HBp17/FGFBP-1 is a potent therapeutic target in squamous cell carcinomas (SCC) and oral squamous cell carcinomas (OSCC). In addition, complementary DNA/protein expression analysis followed by Gene Ontology and protein–protein interaction (PPI) analysis using the Database for Visualization and Integrated Discovery and Search Tool for the Retrieval of Interacting Genes/Proteins showed that both gene and protein expression related to epidermal development, cornification, and keratinization were upregulated in A431-HBp17-KO and HO-1-N-1-KO cells. This is the first discovery of a novel role of HBp17/FGFBP-1 that regulates SCC and OSCC cell differentiation.


Introduction
Fibroblast growth factor (FGF) is a cell signaling protein originally isolated from the pituitary gland as a growth factor that promotes fibroblast proliferation [1,2]. So far, 23 members of the FGF family have been identified [3]. Fibroblast growth factor is believed to play an important role in tumor growth and invasion by inducing angiogenesis, in addition to promoting the proliferation of squamous cell carcinoma (SCC) cells [4,5]. The expressions of FGF-1 and FGF-2 increased with the transformation to SCC, and FGF-like activity is also present in SCC cell and salivary gland tumor cell culture mediums. However, since there is no signal peptide in FGFs, its secretory mechanism is unclear [6,7].
Heparin-binding protein 17 (HBp17) was originally co-purified with FGF-2 from culture media conditioned by A431 human epidermoid carcinoma cells, and the molecular weight of HBp17 is 17 kDa [8]. Heparin-binding protein 17, also known as fibroblast growth factor-binding protein-1 (FGFBP-1), binds to FGF-1 and FGF-2 in a noncovalent and reversible manner to induce their release from the extracellular matrix (ECM). The expression of HBp17/FGFBP-1 is observed in epithelial cells, and its expression increased with the severity of epithelial dysplasia [9,10]. In addition, HBp17/FGFBP-1 binds reversibly to FGF-1 and FGF-2 in vitro, so it may be related to the secretion and activation of FGFs and may control cancer cell growth [11]. Liu et al. reported that HBp17/FGFBP-1 complementary DNA (cDNA)-transfected A431-#4, a nontumorigenic A431 clonal variant that does not express HBp17/FGFBP-1, forms palpable tumors in nude mice but parental cells do not [12].
In our previous studies, HBp17/FGFBP-1 expression could be reduced but not knocked out completely in cells and in the culture supernatant. In this study, to determine the role of HBp17/FGFBP-1 in SCC and OSCC cells, we investigated the effect of HBp17/FGFBP-1 knockout (KO) using the clusters of regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated protein 9 (Cas9) gene editing technology on SCC and OSCC cells in vitro and in vivo. For the sake of simplicity, HBp17/FGFBP-1 will be referred to as HBp17 in the remainder of this manuscript.
These cell lines are free from mycoplasma contamination and have been authenticated using short tandem repeat profiling (BEX CO., LTD. Tokyo, Japan), and the profiles of A431 and HO-1-N-1 matched with the publicly available reference profiles, as mentioned previously [15,16].

SDS-PAGE and Western Blot
A431-HBp17-KO, A431-WT, HO-1-N-1-HBp17-KO, and HO-1-N-1-WT cells were collected at designated times by scraping them from the surface of the plates in the presence of radioimmunoprecipitation assay (RIPA) buffer (10 mM Tris-HCL, 150 mM of NaCl, 5 mM of ethylenediaminetetraacetic acid (EDTA), 1% Triton X-100, pH 7.4 (Nacalai Tesque Inc., Kyoto, Japan)). The cell suspension was sonicated on ice for 30 s (Taitec Ultra S Homogenizer VP-5S; Taitec, Tokyo, Japan). To prepare the conditioned medium (CM), A431 and HO-1-N-1 cells were cultured in 10 mL of DF6F medium in 100 mm dishes (BD Falcon). The medium was changed to the DF nutrient medium when the cells grew to 80% confluence. Next, the cells were further cultured for 48 h and the CM was collected. The CM was centrifuged at 10,000× g for 30 min at 4 • C to remove cells and debris and then concentrated 10 times using an Amicon Ultra-15 Ultracel-5K (Merck).

Immunofluorescent Analysis
One thousand cells/chamber of A431-HBp17-KO, A431-WT, HO-1-N-1-HBp17-KO, and HO-1-N-1-WT cells cultured in DF6F on the LAB-TEK Chamber Slide (Nulgen Nunc International, Naperville, IL, USA) were fixed with 4% paraformaldehyde for 20 min, washed twice with phosphate-buffered saline (PBS), and blocked with 1% bovine serum albumin (BSA) in PBS for 30 min at room temperature. The cells were stained with anti-IVL at a dilution of 1:100 in 1% BSA in PBS overnight at 4 • C, washed twice with PBS, and then incubated with fluorescent goat anti-rabbit Alexa 488 secondary antibody (Molecular Probes, Inc., Eugene, OR, USA) in 1% BSA in PBS at a dilution of 1:400 for 1 h at room temperature. Next, the cells were covered with VECTASHIELD Antifade Mounting Medium with 4 ,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Inc., Burlingame, CA, USA). Finally, images were taken using the EVOS FL Auto 2 Imaging System (product #AMAFD2000) with an Olympus Super Apochromat objective (×40; product #AMEP4754) and a Nikon Eclipse E800 microscope at ×400 magnification and analyzed using Adobe Photoshop Elements (Adobe Inc., San Jose, CA, USA). The images contained an overlay of IVL (green) and nuclei (blue).

Cell Growth and Colony Formation Assays
We examined the growth of cells in serum-free defined culture. Briefly, A431-HBp17-KO, A431-WT, HO-1-N-1-HBp17-KO, and HO-1-N-1-WT cells were cultured in DF6F serum-free defined medium in 24-well plates (BD Falcon) at a density of 1 × 10 4 cells/well, and then the cell numbers were counted using a Coulter Counter (Coulter Electronics Inc., Hialeah, FL, USA) daily for 6 days.
For the colony formation assay, A431-HBp17-KO, A431-WT, HO-1-N-1-HBp17-KO, and HO-1-N-1-WT cells were seeded in DF6F medium in 6-well plates (BD Falcon) at a density of 200 cells/well. The cells were cultured for 14 days, and the culture medium was changed every 5 days. The cells were visualized with Giemsa staining solution (Wako Pure Chemical Industries, Ltd., Osaka, Japan), and the number of colonies consisting of more than 50 cells were counted.

Cell Motility Assay
Cell motility was analyzed with a modified Boyden chamber assay using Transwell inserts (6.5 mm diameter) with 8 µm pores (Coaster, Cambridge, MA, USA), as described previously [21,22]. The filters were coated with 100 µg/mL of gelatin (Merck, Darmstadt, Germany) to enhance cell attachment. A431-HBp17-KO, A431-WT, HO-1-N-1-HBp17-KO, and HO-1-N-1-WT cells (5 × 10 5 ) were seeded in DF nutrient medium containing 0.1% BSA to the upper compartment of each Transwell insert. After 21 h culture at 37 • C, Transwell inserts were fixed with methanol and stained with Diff Quick (Dade Behringen, Duedingen, Switzerland). The cells on the front surface of the filter were moved with a cotton swab. The cells migrated on the back surface of the filter were stained and cell numbers were counted using light microscopy under a high-power field (×200). The cell numbers in the four fields were counted in each of the three different experiments. Results were expressed as the mean number of migrating cells/mm 2 ± standard deviation (SD).

Animal Experiments
We used 4-week-old male athymic Balb/c nude mice (Charles River Japan, Tokyo, Japan). The mice were maintained under specific pathogen-free conditions. All animal procedures were performed in accordance with the Institutional Animal Care and Use Committee guidelines of Hiroshima University (permission #A20-85). All experiments were conducted after the mice were allowed to acclimate to their surroundings for 1 week. One million A431-HBp17-KO and A431-WT cells resuspended in 0.2 mL of DF nutrient medium were inoculated into the flanks of male nude mice, and tumor size was measured twice a week. Tumor volume was calculated by the formula of (1/2 × (major axis) × (the minor axis) 2 ). Experiments were performed with five mice.

Microarray Analysis and Gene Ontology Enrichment Analysis
We examined the effect of HBp17 KO on the cDNA profile of A431-HBp17-KO2 and A431-WT cells. Total RNA was isolated from the cells using TRIzol reagent. The samples were outsourced for cDNA microarray analysis using 3D-Gene (Toray Industries, Inc., Tokyo, Japan). DEGs whose fold changes were <1/2 or >2 in HBp17-KO-2 A431 cells underwent Gene Ontology (GO) functional enrichment analysis (Toray Industries, Inc.).

Proteomic Analysis
We examined the effect of HBp17 KO on the protein profile of A431-HBp17-KO2 and A431-WT cells. Cell extracts were isolated with RIPA buffer. The samples were outsourced for proteomic analysis using DIA Proteome Analysis (Kazusa Genome Technologies Inc., Chiba, Japan).

Protein-Protein Interaction Network Construction
We next investigated the functional interactions among DEGs using the online Search Tool for the Retrieval of Interacting Genes/Proteins (STRING, available at https://stringdb.org/, accessed on 23 April 2021) database (confidence score > 0.900) [23].

Statistical Analysis
Statistical analysis was performed using BellCurve for Excel (Social Survey Research Information Co., Ltd., Tokyo, Japan). All data were presented as the mean ± SD of at least three independent experiments. Student's t-test was used to compare the difference between two groups. The differences were considered significant at p < 0.05.

Screening and Identification of HBp17 Gene Knockout
Two kinds of pCas-Guide-HBp17 guide RNAs (gRNAs) and the functional cassette were co-transfected into A431 and HO-1-N-1 cells by electroporation. The transfected cells were screened by adding puromycin and further cultured. The expression of HBp17 was verified by Western blot. We have successfully knocked out HBp17 at the protein level in cellular protein and in CM of A431 and HO-1-N-1 cells (Figure 1). Clones of A431 cells knockout HBp17 were designated as A431-HBp17-KO1 and A431-HBp17-KO2 cells. That of HO-1-N-1 cells knockout HBp17 were designated as HO-1-N-1-HBp17-KO1 and HO-1-N-1-HBp17-KO2 cells. Parental cells used as controls were designated as A431-WT and HO-1-N-1-WT cells.  Western blot analysis revealed that there was no difference in cellular FGF-2 levels between HBp17-KO cells (A431-HBp17-KO1, -KO2, HO-1-N-1-HBp17-KO1, -KO2) and A431-WT and HO-1-N-1-WT cells (Figure 2A). However, quantification of FGF-2 in the CM by ELISA revealed that the amount of soluble FGF-2 significantly decreased in HBp17-KO cells compared to the WT cells ( Figure 2B).

HBp17 Knockout Inhibits Proliferation and Colony-Forming Ability of A431 and HO-1-N-1 Cells In Vitro
We examined the effect of HBp17 knockout on A431 and HO-1-N-1 cell proliferation in serum-free defined culture. The proliferation rate of A431-HBp17-KO and HO-1-N-1-HBp17-KO cells was significantly decreased compare to that of A431-WT and HO-1-N-1-WT cells ( Figure 3A).
The effect of HBp17 knockout on colony-forming ability was also evaluated, and it has been revealed that the ability of A431-HBp17-KO and HO-1-N-1-HBp17-KO cells significantly decreased compared to that of A431-WT and HO-1-N-1-WT cells ( Figure 3B).

HBp17 Knockout Inhibits Proliferation and Colony-Forming Ability of A431 and HO-1-N-1 Cells In Vitro
We examined the effect of HBp17 knockout on A431 and HO-1-N-1 cell proliferation in serum-free defined culture. The proliferation rate of A431-HBp17-KO and HO-1-N-1-HBp17-KO cells was significantly decreased compare to that of A431-WT and HO-1-N-1-WT cells ( Figure 3A).
The effect of HBp17 knockout on colony-forming ability was also evaluated, and it has been revealed that the ability of A431-HBp17-KO and HO-1-N-1-HBp17-KO cells significantly decreased compared to that of A431-WT and HO-1-N-1-WT cells ( Figure 3B).

HBp17 Knockout Inhibits Tumor Growth in Athymic Nude Mice
We evaluated the effect of HBp17 knockout on tumor growth of nude mouse xenografts from the A431-HBp17-KO1, A431-HBp17-KO2, and A431-WT cells. The cells were transplanted into athymic nude mice, and tumor sizes were measured twice a week. The tumor growth of A431-HBp17-KO1 cells clearly decreased compared to that of A431-WT cells ( Figure 5). No tumor formation was observed in mice transplanted with A431-HBp17-KO2 cells.

HBp17 Knockout Inhibits Tumor Growth in Athymic Nude Mice
We evaluated the effect of HBp17 knockout on tumor growth of nude mouse xenografts from the A431-HBp17-KO1, A431-HBp17-KO2, and A431-WT cells. The cells were transplanted into athymic nude mice, and tumor sizes were measured twice a week. The tumor growth of A431-HBp17-KO1 cells clearly decreased compared to that of A431-WT cells ( Figure 5). No tumor formation was observed in mice transplanted with A431-HBp17-KO2 cells.
We focused on the protein-protein interaction (PPI) of upregulated differentially expressed proteins in A431-HBp17-KO2 cells. The analysis of PPI using Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) provided 323 nodes (proteins) and 154 edges (interactions) in the PPI network and identified a main module associated with the formation of a cornified envelope such as Filaggrin (FLG), small SPRR3, Involucrin (IVL), SPRR1A, and SPRR1B (Supplementary Figure S1).

Discussion
An association between FGF or the FGF receptor and tumorigenicity has been reported in salivary gland cancer, hepatocyte cell carcinoma, and melanoma [24][25][26]. Fibroblast growth factor-1 and -2, which play an important role in SCC and OSCC cell growth,

Discussion
An association between FGF or the FGF receptor and tumorigenicity has been reported in salivary gland cancer, hepatocyte cell carcinoma, and melanoma [24][25][26]. Fibroblast growth factor-1 and -2, which play an important role in SCC and OSCC cell growth, are secreted extracellularly, although there is no signal sequence in FGFs [27][28][29][30]. The secretory mechanism of FGFs is still unknown [31,32]. Therefore, HBp17, which was originally co-purified with FGF-2 from the A431 CM and was found to bind to FGF-1 and FGF-2 in a noncovalent and reversible manner, is considered to be a key molecule regulating the extracellular availability of FGFs [33][34][35].
Clusters of regularly interspaced short palindromic repeats (CRISPR) and CRISPRassociated protein 9 (Cas9) system, a gene-editing technology, can cut double strands of DNA and delete, replace, or insert any place in the genome sequence. It has been used for genome editing or modification not only in humans and mice but also in bacteria, parasites, and zebrafish [36][37][38]. In this study, to analyze the function of HBp17, we used the CRISPR/Cas9 technique to delete HBp17 from the A431 and HO-1-N-1 cell lines. No morphological changes in A431-HBp17-KO and HO-1-N-1-HBp17-KO cells were observed compared to parental WT cells. The HBp17-KO cells decreased growth rates and a lower colony-forming ability compared to A431-WT and HO-1-N-1-WT cells in serum-free culture. In addition, A431-HBp17-KO1 cells produced much smaller tumors compared to A431-WT cells an in vivo assay using immunodeficient nude mice. Further, A431-HBp17-KO2 cells lost their tumorigenic potential. These results are consistent with our previous reports that decreased HBp17 expression inhibits tumor growth by inhibiting angiogenesis [16]. Liu et al. reported that HBp17 overexpression in the A431 subclone, designated A431-#4, which has no HBp17 expression and no tumorigenic ability in nude mice, induced tumorigenicity [12]. Tumors derived from A431-WT cells had large variability in tumor growth compared to those from A431-HBp17-KO2 cells. Gross findings showed that A431-WT cell tumors were more prone to tumor necrosis and invasive into the skin (data not shown). It is considered that the invasive manner (tumor growth externally or internally) of each tumor makes a difference in tumor size. Taken together, the results show that loss of HBp17 expression in SCC and OSCC cells clearly inhibits tumor cell growth in vitro and in vivo.
The amount of FGFs in the CM resulted from the ability of HBp17 to release bound FGF from the ECM. By treatment with high-salt or heparinase-like enzymes, FGF-2 can be released from the ECM, including perlecan which is served as a reservoir [39]. Fibroblast growth factors released by HBp17 play an important biological role in normal and pathological tissues in an autocrine or paracrine manner [40]. Rosli et al. reported that treatment of OSCC cells with 1α,25(OH) 2 D 3 inhibits HBp17 expression and decreases FGF-2 secretion into the CM [14]. Furthermore, higher FGF-2 levels are present in the CM of A431-#4 cells overexpressing HBp17 compared to parental cells [12]. In this study, the loss of HBp17 resulted in a decrease in the amount of soluble FGF-2 in the CM, although intracellular FGF-2 levels did not change. This result clearly shows that HBp17 is a key molecule related to extracellular secretion of FGF-2 lacking a signal sequence, as mentioned previously [14,15].
We next focused on the altered global gene and protein expressions in A431-HBp17-KO2 cells compared to A431-WT cells. Cell differentiation molecules (FABP5, SPRR1A, AKR1C3, AKR1C2, and SPRR1B) and immune response molecules (S100A9 and S100A8) were upregulated in A431-HBp17-KO2 cells. Siegenthaler et al. reported that cellular FABP5 expression is twice as high under normal Ca 2+ concentrations, which induce differentiation, compared to a low-Ca 2+ medium, which inhibits differentiation [41]. A proteome profiling study on bladder SCC showed a decrease in FABP5 expression in less differentiated tumors [42]. Analysis of gene expression profiles in esophageal SCC showed that genes involved in squamous cell differentiation, including SPRRs and calcium-binding S100 proteins (S100A8, S100A9), are coordinately downregulated compared to their normal counterparts [43]. Phorbol ester 12-O-tetradecanoylphorbol-13 acetate (TPA), a potent tumor promotor, induced a significant increase in the formation of a cornified envelope as an indicator of terminal differentiation of epidermal keratinocytes [43][44][45]. From GO enrichment analysis of cDNA microarray data, some cascades related to epidermal development, keratinization, and cornification emerged as potential targets of HBp17. A series of epidermal structural proteins, including IVL, LOR, FLG, and the class of SPRRs, are synthesized to reinforce the cornified envelope [46]. We confirmed that these corni-fied envelope-related molecules were upregulated in A431-HBp17-KO2 cells compared to A431-WT cells. When we constructed PPI networks of molecules elevated by cDNA microarray and proteomic analysis, both networks were similar (Supplementary Figures S1  and S2). Although there are many reports on the induction of cancer cell differentiation and growth inhibition [47,48], there are almost no reports on the involvement of HBp17. It was reported that FLG expression increases in HBp17-KO mice [49]. Our results revealed that HBp17 KO induces terminal differentiation of the squamous epithelium.
The expression of HBp17, which is highly expressed in SCC and OSCC cells, increases during the course of malignant transformation of squamous epithelial cells [10,12]. In this study, using cytological, DNA microarray, and proteomic analyses, we found for the first time that HBp17 KO inhibits growth and motility of SCC and OSCC cells through induction of terminal differentiation. This is the first discovery of the novel role of HBp17, which inhibits differentiation of SCC and OSCC cells. In this regard, well-differentiated cancer cells, such as adenocarcinoma cells, show no HBp17 expression. At this point, the mechanism by which HBp17 inhibits cell differentiation is unclear. Further studies need to elucidate HBp17-interacting molecules involved in terminal differentiation of squamous epithelial cells. Our findings strongly suggest that HBp17 KO can be applied to differentiation inducing therapy for SCC and OSCC.

Conclusions
In this study, we have examined the functional role of HBp17 in A431 and HO-1-N-1 cells using the CRISPR/Cas9 technology. We have revealed that HBp17 knockout inhibited cell proliferation, colony formation, and cell motility of the SCC cells, and that soluble FGF-2 levels in the medium conditioned by HBp17-knockout A431 and HO-1-N-1 cells were decreased compared to the parental cells. In addition, the molecules related to epidermal development, cornification, and keratinization were upregulated in the HBp17-KO cells. It was suggested that HBp17 KO inhibits growth and motility of SCC and OSCC cells through induction of terminal differentiation. Based on these findings, the HBp17-targeting therapy in treating SCC and OSCC will be expected in the future.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/cancers13112684/s1, Figure S1: Predicted PPI networks of upregulated DEGs in A431-HBp17-KO2 cells from microarray analysis data. Figure S2: Predicted PPI networks of upregulated DEGs in A431-HBp17-KO2 cells from proteome analysis data. Figure S3: Original Western blots and densitometric analyses relative to Figure 1. The figure shows the original Western blots shown in Figure 1, as indicated, as well as the corresponding densitometric analyses. Figure S4

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki, and the Institutional Animal Care and Use Committee guidelines of Hiroshima University (permission #A20-85).

Data Availability Statement:
The data presented in this study are available from the corresponding author upon reasonable request.