FGF19 promotes cell autophagy and cisplatin chemoresistance by activating MAPK signaling in ovarian cancer

Patients

A cohort of 157 HGSOC specimens, 26 normal ovaries and 22 normal fallopian tubes was obtained from Xiangya Hospital of Central South University from September 2016 to September 2021. Among these, 118 HGSOC, 26 paired normal ovaries and 22 normal fallopian tube were used for immunohistochemistry (IHC). A total of 62 diagnosis HGSOC samples were used for next-generation sequencing (NGS). A total of 41 HGSOC samples were used for fluorescence in situ hybridization (FISH). Specimens were collected in accordance with the ethical standards of Xiangya Hospital of Central South University. The Ethical Committee of Xiangya Hospital of Central South University approved this study (Approval No. 202201021).

Cells culture

Human ovarian cancer cells, OVCAR3, HO8910, HO8910pm, SKOV3, SKOV3-IP, A2780 and normal ovarian epithelial cells IOSE were obtained from School of Basic Medical Science, Central South University. All the cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) medium (Bi, Israel Beit Haemek Ltd., Beit Haemek, Israel) containing 10% fetal bovine serum (FBS) at 37 °C under an atmosphere with 5% CO₂.

Chemicals and reagents

BIRB796 and cisplatin were purchased from Good Laboratory Practice Bioscience. IFN-γ was purchased from MedChemExpress. FGF19 and p62 antibodies were obtained from Santa. Antibodies against LC3 and Beclin 1, along with secondary antibodies, were purchased from Cell Signaling Technology. Antibodies against p38 MAPK, phospho-p38 MAPK (p-p38 MAPK) and GAPDH were purchased from Abclonal.

IHC

After the tissue sections were dewaxed in turpentine oil and hydrated in gradient alcohol (95%, 85%, 75%, 50%), the exposure of tissue antigen was repaired with citrate buffer for 3 min. Endogenous peroxidase activity was blocked by using 3% H₂O₂ for 20 min. The sections were incubated with FGF19 antibody at 37 °C for 1h, and incubated with secondary antibody for 30 min. After then, color reaction was analyzed with 3,3′-diaminobenzitine (DAB) solution.

FISH

According to the manufacturer’s recommendations, FISH was performed on tissue using two probes targeting FGF19-CCND1 and CEP11 (Abnova, Walnut, CA, USA). The slides were examined using a 100× objective lens and an Olympus BX51 fluorescence microscope (Olympus, Tokyo, Japan). Each image is obtained by Olympus cellsens software.

Western blot

The same amount of celluar protein in each sample was separated by SDS-PAGE, and then imprinted on PVDF membrane (Millipore, Burlington, MA, USA; Merck, Rahway, NJ, USA). After blocking with 5% milk, the membrane was incubated with primary antibodies overnight at 4 °C, and then incubated with HRP-conjugated secondary antibodies at room temperature for 2 h. Then, the Chemiluminescence imaging system was used to detect the immune response.

RNA extraction and qPCR

The total RNA was extracted by TRIzol reagent (Invitrogen, Waltham, MA, USA), reverse transcribed into cDNA using PrimeScript RT Kit (Abclonal, Wuhan, China) according to the manufacturer’s instructions. qRT-PCR was performed by iTaqTM Universal SYBR green Supermix (Abclonal, Wuhan, China). GAPDH was used as internal control. The forward and reverse primers were used as follows: FGF19: 5′-CGGAGGAAGACTGTGCTTTCG-3′ and 5′-CTCGGATCGGTACACATTGTAG-3′; GAPDH: 5′-CAGCAAGAGCACAAGAGGAA-3′ and 5′-TGGTACATGACAAGGTGCGG-3′. Relative expression levels were decided using the 2^−ΔΔCT method.

Immunofluorescence

HO8910pm and SKOV3-IP cells were seeded into 6-well plates. After transfection and/or adding BIRB796, the cells were fixed with 4% formaldehyde for 20 min at room temperature. After staining with 2-(4-Amidinophenyl)-6-indolecarbamidine dihydrochloride (DAPI) and sealing the slide, the formation of GFP-LC3 puncta in cells were analyzed by confocal microscope.

CCK-8 assays

Cells (1 × 10³ cells/well) were inoculated in 96 well plates after treatment with cisplatin. Then, after adding CCK-8 test solution (b34304, Bimake, Houston, TX, USA), the cells were cultured at 37 °C in 5% CO₂ for 1 h. Subsequently, the optical density (OD) of the sample was measured at 450 nm using a Microplate Reader (Biotech, Camarillo, CA, USA).

Statistical analysis

SPSS 22.0 and GraphPad Prism 8.0 were used for all statistical analyses. Data have been expressed as means ± Standard Error of Mean. The two groups were compared by Student’s t-test. The correlations between the FGF19 expression and FGF19 amplification were analyzed by Spearman’s correlation analysis. All experiments were performed at least three times and are reported as mean ± SD. Statistical significance is shown as with *p-value < 0.05, **p-value < 0.01, ***p-value < 0.001 and ****p-value < 0.0001.

Identification of FGF19 amplification in ovarian cancer

Using NGS technology, we detected and analyzed the gene mutation profiles in HGSOC patients (Table S1). Overall, 621 mutations and 273 mutated genes were detected. Among them, the most frequently mutated genes were TP53 (90%), followed by Myc (35%), neurofibromatosis type 1 (NF1, 18%) and RecQ like helicase 4 (RECQL4, 18%) (Fig. 1A). The mutation types were classified into eight categories: missense mutation, deletion mutation, splicing mutation, frameshift mutation, stop-gained variation, copy number variation, gene fusion, large genomic rearrangement and promoter mutation. Among them, copy number amplification was the most type (44.28%), followed by missense mutation (33.49%), splicing mutation (5.48%) and frameshift mutation (4.03%) (Fig. 1B).

Studies have shown that aberrant FGF19 can be a driver of malignant behavior, contributing to the oncogenesis and progression of human cancers (Kanzaki et al., 2021). In our analysis of 62 paraffin-embedded ovarian cancer tissues, we identified FGF19 amplification (Fig. 1A). Additionally, the cBioPortal database revealed FGF19 amplification in a variety of cancers (Fig. 1C) and four ovarian cancer datasets (Fig. 1D). Furthermore, FISH analysis was conducted to validate the amplified FGF19 in HGSOC tissues (Fig. 1E). This showed that FGF19 is highly amplified in ovarian cancer.

FGF19 was overexpressed in ovarian cancer and correlated with poor prognosis

The expression profiles of FGF19 were analyzed by several bioinformatics databases. First, the analysis results from TNMplot database showed that the expression of FGF19 mRNA in ovarian cancer tissues was higher than that in normal tissues (Fig. 2A). From TCGA-OC, we found that the expression of FGF19 was significantly upregulated in ovarian cancer tissues (Fig. 2B). Next, we analyzed the association between FGF19 expression and patients’ clinical characteristics. IHC was performed to explore the FGF19 level in 118 ovarian cancer tissues, 22 normal ovarian and 26 normal fallopian tube. FGF19 was expressed in both nucleus and cytoplasm, and the expression of FGF19 in tumor tissues was higher than that in normal tissues (Fig. 2C). Next, we studied the correlation between FGF19 expression and the clinical characteristics, such as age, clinical stage and lymphatic metastasis (Table 1). The expression of FGF19 was related to lymphatic metastasis and abdominal and distant metastasis (p = 0.001, p = 0.028, respectively), and high expression of FGF19 was more prone to tumor metastasis. The Kaplan Meier plotter database suggested that the patients with high level of FGF19 displayed poor prognosis in GSE26193 (Fig. 2D) and GSE19829 (Fig. 2E). Collectively, the overexpressed FGF19 in ovarian cancer could be utilized as a potential biomarker for prognosis.

FGF19 knockdown inhibited cell autophagy

We screened the expression profiles of FGF19 in several ovarian cancer cells, and found that FGF19 were over-expressed in multiple ovarian cancer cells, including HO8910pm and SKOV3-IP (Figs. 3A and 3B). Recent study displayed that cell autophagy-associated signaling affects the clinical outcomes and therapeutic responses in ovarian cancer (Zhang et al., 2022a). To investigate the roles of FGF19 on cell autophagy, we knocked down FGF19 in HO8910pm and SKOV3-IP cells, and found that FGF19 knockdown down-regulated the expression of LC3 and Beclin 1 and up-regulated the expression of SQSTM1/p62 (Fig. 3C). Then, immunofluorescence results displayed that FGF19 knockdown reduced the GFP-LC3 puncta formation in HO8910pm and SKOV3-IP cells (Figs. 3D–3G). Furthermore, autophagy-associated biomarkers and GFP-LC3 puncta formation displayed the opposite results after overexpression of FGF19 in HO8910 and SKOV3 (Fig. S1). The transmission electron microscopy showed that downregulation of FGF19 inhibited the formation of autophagosomes (Figs. 3H–3K). The above-mentioned results suggested that abnormally overexpressed FGF19 could improve cell autophagy in ovarian cancer cells.

FGF19 knockdown inhibited cell autophagy via downregulation MAPK signaling

MAPKs, in particular p38 MAPK, have dual roles in the regulation of cell autophagy (Webber & Tooze, 2010). Some studies have shown that FGF family members, such as FGF2, regulate cell autophagy by inhibiting MAPK signaling (Yuan et al., 2017). Here, we explored whether MAPK signaling participates in the regulatory effect of FGF19 on cell autophagy. Western blot showed the decreased LC3 and Beclin 1 and increased p62 after treatment with FGF19 knockdown, while increased LC3 and Beclin 1 and decreased p62 after treatment with IFN-γ, a potential p38 MAPK activator (Matsuzawa, Fujiwara & Washi, 2014), in HO8910pm and SKOV3-IP cells (Figs. 4A and 4B). Moreover, FGF19 knockdown also inhibited p38 MAPK phosphorylation, whereas IFN-γ did not affect the expression of FGF19, suggesting that p38 MAPK was the downstream target of FGF19. In contrast, FGF19 overexpression up-regulated the expression of LC3 and Beclin 1 and down-regulated the expression of p62, while treated with p38 MAPK inhibitor BIRB796 had the opposite effects in HO8910 and SKOV3 cells (Figs. 4C and 4D). We then transfected GFP-LC3 into ovarian cancers cells, and found that FGF19 downregulation and BIRB796 treatment decreased the formation of GFP-LC3 puncta, while FGF19 upregulation and IFN-γ treatment increased the formation of GFP-LC3 puncta (Figs. 4E–4H). Taken together, these findings suggested that FGF19 knockdown inhibited cell autophagy through deactivating p38 MAPK signaling pathway.

FGF19 enhanced cisplatin resistance by activating p38 MAPK signaling

Multiple studies have demonstrated that activation of p38 MAPK is related to cisplatin resistance in a variety of cancers, including ovarian cancer (Chen et al., 2018; Li et al., 2021; Liao et al., 2020). Here, CCK8 was used to detect the potential roles of FGF19-p38 MAPK axis on the anti-proliferation effect of cisplatin in ovarian cancer cells. The results displayed that both FGF19 knockdown and BIRB796 treatment increased cisplatin-induced proliferation inhibition of OC cells; however, FGF19 overexpression and IFN-γ treatment reduced cisplatin-induced proliferation inhibition (Figs. 5A–5D). These results demonstrated that FGF19 might promote the cisplatin resistance of ovarian cancer cells by activating p38 MAPK signaling.