The C57BL/6 mouse kidney cell line was obtained from American Type Cell Collection (ATCC, Manassas, VA) and was grown in Dulbecco’s Modified Eagle’s medium (DMEM) containing 5.5 mM glucose, 1% penicillin-streptomycin, and 10% fetal bovine serum.

Cell proliferation was measured using the Cell Counting Kit-8 (CCK-8, Dojindo Bio., Japan) following the manufacturer’s instruction. Briefly, the cells (2.6 × 10⁴ cells/well, 100 mL) were digested with trypsin and subsequently transferred to 96-well plates, with five parallel wells assigned for each treatment.

The FGF9 open reading frame was amplified from the cDNA of C57BL/6 mouse kidney cells by PCR and cloned into the pcDNA3.1 (+) (Cat. No. V79020; Thermo Fisher Scientific, Inc., Waltham, USA) expression vector to create FGF9 overexpression plasmids. The primer information for CTNNB1 silencing is shown as follow: FGF9 CF: ATGGCTCCCTTAGG TGAAGTTGG, FGF9 CR: TCAACTTTGGCTTAG AATATCCTTATA. The negative control siRNA (ON-TARGETplus si CONTROL non-targeting pool, D-001810) and CTNNB1 siRNA duplex (ON-TARGET plus SMART pool, L-004018) were obtained from Dharmacon RNA Technologies (Chicago, IL, USA). RNAs (shRNAs) targeting ANXA4 (shANXA4F: CGCGTccccCCGATGAAGACGCCATTATttcaagagaATAATGGCGTCTTCATCGGtttttGGAAAT; shANXA4R: cgatTTCCaaaaaCCGA TGAAGACGCCATTATtctcttgaaATAATGGCGTCTTCATCGGGGGGA) and negative control non-specific shRNA (NC shRNA) were synthesized by Songon Biotech Co. (Shanghai, China; Liu et al., 2017). Fibroblasts that reached 30%–50% confluence were used for transfection with 2 µg of the FGF9 overexpression plasmid or 30 nM of the siRNA duplex using Lipofectamine 2000 (Invitrogen) and Opti-MEM® I Reduced Serum Medium (Gibco), according to the manufacturer’s instructions.

Total RNA extracted from C57BL/6 mouse kidney cells, with or without FGF9-overexpression, was used for RNA-sequencing experiments. The RNA-seq experiments and further data analyses were performed by Songon Biotech Co., Ltd. (Shanghai, China), and data were deposited in a personal computer, which will be released upon request.

Differentially expressed genes were analyzed further based on statistical outcomes by assessing their association with biological processes using the gene ontology (GO) database (Consortium, 2006). Fisher’s exact test was performed to enrich the GO category. The false discovery rate (FDR) was further calculated to correct the P-value, with a smaller FDR indicating a smaller error in judging the P-value (Dupuy et al., 2007). The enrichment of GO terms among differentially expressed probe sets was identified using the one-tailed Fisher’s exact test (Dunnick et al., 2012). KEGG, Biocarta, and Reactome were employed to analyze pathway enrichment for differentially expressed genes. Fisher’s exact test was followed by Benjamini–Hochberg (BH) multiple testing correction to select the significant pathways, and the threshold of significance was defined by P-value and FDR (Draghici et al., 2007).

Chromatin immunoprecipitation (ChIP) assay was performed using a chromatin immunoprecipitation assay kit (Cat No. 17-295, Millipore, Billerica, MA), according to the manufacturer’s instructions. Approximately 1 µg of anti-β-catenin antibody (Abcam, ab32572) or 1 µg of anti-TCF4 antibody (Cell Signaling Technology, 2566) was added to the reaction solution to immunoprecipitate DNA fragments. The resulting immunoprecipitated (IP) DNA was compared with the DNA precipitated with anti-mouse IgG antibody (Abcam, ab190475) using quantitative PCR (qPCR) with SYBR Green (cat. no. 4472908) for F1, F2, or F3 regions; Thermo Fisher Scientific, Inc.), according to the manufacturer’s protocol. ChIP-PCR was performed with an initial denaturation at 95°C for 3 min, followed by 40 cycles of denaturation for 30 s at 95°C, annealing for 30 s at 58°C, and extension at 72°C for 30 s, followed by a final extension at 72°C for 5 min. GAPDH was used as a reference gene to normalize the data. The primer sequences for qRT-PCR are listed as follow: F1 (F: ATCTTGAGGGAGACTTGGACA, R: CTCTGGATGAATACCTGTGGC), F2 (F: CTCTGAGTGAGTCGCGAGGTTAT, R: G AGTTGGTCTCCAATGTTGTTTG), and F3 (F: TGTGATT CAAGAGCTCGAGAC, R: AAGCTCGAATGACGTA CGTTC).

Approximately 2 µg of total RNA from C57BL/6 mouse kidney cells was reverse-transcribed using the GoScript Reverse Transcription Kit (Reverse Transcription System, Promega) as per the manufacturer’s methods. Gene expression was quantified as described previously via qRT-PCR assay (Xuan et al., 2016). The primers information are listed as follow: GAPDH (F: GCCAAGGTCATCCATGACAACT, R:GAGGGGCCATCCACAGTCTT), FGF1 (F: TGCTCTACTGCAGCAACG, R: C TAGTCAGAAGACACCGG), FGFR2 (F: CAAGAAC GGCGGCTTCTTC, R: GGAAGAAACAGTATGGCCT) FGR3 (F: CAAGCTCTACTGCGCTACC, R: GTCCAC CTGTATGCAGCT), FGF4 (F: TACTGCAACGTGG GCATC, R: GGAAGTGGGTTACCTTCA), FGF5 (F: GAAGTAGCGCGACGTTTTC, R: GGCTTAACACA CTGGCTTC), FGG6 (F: CTCTACTGCAACGTGGG C,R:GGAAGTGAGTGACAGTCA), FGF7 (F: AGAC TGTTCTGTCGCACC, R: CCGCTGTGTGTCCATTT AG), FGF8 (F: ACCTACCAGCTCTACAGCC, R: GG CGGGTAGTTGAGGAACT), FGF9 (F: CTGCAGGA CTGGATTTCATTT, R: GTTCAGGTACTTTGTCA GGG), FGF10 (F: TGTCCGCTGGAGAAGGCTGT TC, R:CTATGTTTGGATCGTCATGG), FGF11 (F: ATCGTCACCAAACTGTTCTG, R: CAGGAACAC TGTGGAGAGAA), FGF12 (F: TCAGCCAGCAGG GATATTTC, R: CACGACTTTGCCTCCATTCA), FGF13 (F: TAACCTCATCCCTGTGGG, R: GAGAA CTCCGTGAGATCG), FGF14 (F: CAACCTCATCC CAGTGGGA, R: GGGACTGTTTCACCAACATC), FGF15 (F: ACTCCGCTGGTCCCTATGTC, R: CTAC ATCCTCCACCATCCT), FGF16 (F: GCTTCCACC TTGAGATCTTC, R: GAGATCTCTGGACATGGAG) FGF17(F:CCAGCTCTACAGCCGGAC, R: GGGGC GGAGCCCACAAAT), FGF18 (F: CCAGCTCTATA GCAGGAC, R: GCTTGGTGACTGTGGTGT), FGF 19 (F: AACTTTATCCCCATATTTCACC, R: GAAG CTGGGACTCTTCACT), FGF20 (F: TCAGAGAA ATTGACTTCTG, R: GTGTACATCAGTAGGTCTT), FGF21 (F: GATGACGACCAAGACACTG, R: CGGC CCTGTAAAGGCTCT), FGF22 (F: GCCTCTTCTC CTCCACTC, R: CGAGACCAAGACTGGCAG). FG F23 (F: ACAGCCAGGACCAGCTATC, R: CTCGCGAGAGCAGGATACA), ANXA4 (F: ACCAGCAGCAATATGGACGG, R: TTCGGTTCCGGGAACAGAG), CTNNB1 (F: TCGCCAGGATGATCCCAGC, R: GC CCATCCATGAGGTCCTG), SMO (F: ACCTATGCC TGGCACACTTC, R: AGGAAGTAGCCTCCCACGAT, PTCH (F: CAAACTCCTGGTGCAAACCG, R: CCGGGATTCTCAGCCTTGTT), and GLI1 (F: CCAGAGTTCAAGAGCCTGG. R: CC TCGCTCCATAAGGCTCAG).

Protein was extracted from cells using a lysis solution (7 M urea, 2 M thiourea, 40 mM Trizma base, 2% CHAPS, 1% protease inhibitor, and 40 mM dithiothreitol (DTT)). The cell lysates were centrifuged at 15,000 × g for 15 min, and the supernatants were harvested for total protein concentration measurement via the Bradford protein assay (Bio-Rad, Richmond, CA, USA). Next, 20 µg of protein from each sample was resolved using the SDS-PAGE and subsequently transferred onto Immobilon-P Transfer Membranes (MILLIPORE JAPAN, Tokyo, Japan). The membranes were incubated in 1× TBS containing 0.05% Tween-20 and 5% milk for 2 h and incubated further with the primary antibodies at 25°C for 1 h. Anti-GAPDH (1:2500, Abcam, ab9485), anti-phospho-SAPK/JNK (Thr183/Tyr185) antibody (1:1000, Cell Signaling Technology, 4668), anti-JNK1 + JNK2 + JNK3 antibody (1:1000, Abcam, ab179461), anti-phospho-AKT (Ser473) antibody (1:2000, Cell Signaling Technology, 4060), anti-p-IκBα antibody (1:500, Santa Cruz Biotechnology, sc-8404), anti-β-catenin antibody (1:1000, Cell Signaling Technology, 8480), anti-AKT antibody (1:1000, Cell Signaling Technology, 4691), anti-β-catenin antibody (1:2000, Abcam, ab32572), and anti-ANXA4 antibody (1:2000, Abcam, ab256456) were used as primary antibodies. After incubation, the membranes were washed with 1× TBS twice and incubated for 1 h with an anti-mouse or anti-rabbit HRP-linked secondary antibody (1:2000, Cell Signaling Technology). Antigen-antibody complexes were detected with an electrochemiluminescence (ECL) kit (GE healthcare). The protein band density was quantified using the ImageJ software (National Institute of Health, Bethesda, MD, USA), and the relative levels were normalized to GAPDH levels.

Prism 5 (GraphPad Software, San Diego, CA) software was used for statistical analysis. Student’s t-test was used to analyze significant differences between two groups.

qRT-PCR was performed to evaluate the expression levels of 23 FGF family member genes in C57BL/6 mouse kidney cells. The results indicate that FGF1, 5, 7, 8, 9, 10, 11, 12, 16, and 17 exhibited the highest expression levels among the 23 FGFs. Among these, FGF9 exhibited the highest expression, followed by FGF1. Meanwhile, FGF2, FGF6, FGF13, FGF14, FGF19, FGF20, FGF21, and FGF23 were found to be rarely expressed in the kidney cells (Fig. 1).

As FGF9 was the predominantly expressed FGF in kidney cells, its function was further investigated via overexpression in C57BL/6 mouse kidney cells. The qRT-PCR results indicate that the FGF9 expression levels, but not those of FGF1 or FGF21, were higher in the overexpressed cells compared to those in the non-transfected control cells (Fig. 2A). Using the FGF9-overexpressed cells, the proliferation rate was examined by CCK-8, and cell density measurement showed that FGF9 overexpression significantly promoted the proliferation of C57BL/6 cells (Fig. 2B). Next, AKT and JNK levels in the control and FGF9 overexpressing cells were analyzed. Western blotting results indicate that FGF9 overexpression induced phosphorylation of AKT and JNK without affecting the total AKT or JNK levels (Figs. 2C and 2D).

To screen FGF9-regulating genes and pathways, RNA-seq based transcriptome analysis was performed by using the FGF9-overexpressing and control C57BL/6 mouse kidney cells. The RNA-seq results indicate that a total of 641 genes were differentially expressed following the overexpression of FGF9 (1.5-fold change; P < 0.05). Among these, 365 downregulated genes and 276 upregulated genes were identified (Fig. 3A). Further, qRT-PCR revealed that the expression of WNT pathway genes (SMO, CTNNB1, and PTCH), hedgehog pathway genes (GLI1), and ANXA4 was upregulated upon FGF9 overexpression, which was consistent with the RNA-seq data (Fig. 3B). The differentially expressed genes were further analyzed for their association with biological process GO terms. GO analysis indicated that 31 terms were enriched for these differentially expressed genes (P < 0.01) and that they were involved in diverse biological processes, including multicellular organismal development, nucleosome assembly, DNA-template transcription, and positive regulation of interleukin (IL)-10 secretion (Fig. 4A). In addition, out of the 641 differentially expressed genes, 102 were classified into known pathways (P < 0.05). Pathway analysis using the KEGG database was then performed to classify the differentially expressed genes. The dendrograms, a portion of which are presented in Fig. 4B, demonstrate the significance of the seven up- and nine downregulated pathways, including WNT, NF-κB, and hedgehog signaling pathways.

As FGF9 overexpression activated the expression of WNT pathway genes (SMO, CTNNB1, and PTCH), the expression of the key WNT signaling transcription factor, i.e., β-catenin was analyzed by western blot analysis. In agreement with the qRT-PCR results, β-catenin accumulated to a greater extent in the FGF9-overexpressing cells as compared to that in the control cells (Figs. 5A and 5B). Further, the role of β-catenin in kidney cell proliferation was examined using a specific shRNA against β-catenin (CTNNB1) and NC siRNA (control) in C57BL/6 cells (Fig. 5C). Moreover, to verify the qPCR results, Western blot analysis was performed to analyze β-catenin protein levels. The results indicate that β-catenin level was higher in FGF9 OX than that in the control, while it was significantly lower in CTNNB1 siRNA than in NC siRNA transfected cells (Fig. 5D). In addition, CTNNB1 knockdown significantly inhibited kidney cell proliferation (Fig. 5E).

ANXA4 was induced by FGF9 overexpression; therefore, its expression level was examined in CTNNB1 knockdown cells. The qRT-PCR results show that CTNNB1 suppression reduced ANXA4 expression (Fig. 6A). Moreover, as ANXA4 was previously reported to play a role in the proliferation of hepatocellular carcinoma cells (Liu et al., 2017), its function in kidney cell proliferation was further examined. The NC shRNA and specific shRNA were confirmed to significantly downregulate the expression of ANXA4 at transcript (Fig. 6B) and protein levels (Fig. 6C), whereas ANXA4 knockdown significantly inhibited kidney cell proliferation (Fig. 6D). As the β-catenin complex functions as a transcription activator to regulate downstream gene expression, the potential binding between the transcription factor complex, containing TCF4 and β-catenin, to the ANXA4 promoter region was investigated. The TCF4 or β-catenin antibody immunoprecipitation and subsequent ChIP-PCR showed that the TCF4 and β-catenin complex bound to F2, not the F1 or F3 regions, in the ANXA4 promoter (Fig. 6E).