ABAT and ALDH6A1, regulated by transcription factor HNF4A, suppress tumorigenic capability in clear cell renal cell carcinoma

Background Clear cell renal cell carcinoma (ccRCC) is a malignancy characterized by metabolic reprogramming. ABAT and ALDH6A1 are metabolic enzymes. In this study, we aim to investigate the associations of ABAT and ALDH6A1 with the malignancy of ccRCC cells. Methods The gene expression levels of ABAT and ALDH6A1 in ccRCC were analyzed from gene expression microarray datasets and RNA sequencing data. Clinical information was analyzed from The Cancer Genome Atlas (TCGA) data. The distributions of ABAT and ALDH6A1 in ccRCC clinical tissues were screened by reverse transcription-quantitative polymerase chain reaction (RT-QPCR) and immunohistochemical assays. The effect of overexpression of ABAT or ALDH6A1 was measured by detecting the cell viability, migration ability, and the ratio of lactate and nicotinamide adenine dinucleotide phosphate (NADPH). Chromatin immunoprecipitation (ChIP) and luciferase reporter assays were carried out to investigate the transcript regulation of HNF4A in ABAT and ALDH6A1. Results Remarkable downregulated ABAT and ALDH6A1 expression levels were observed in ccRCC patients and low expression of ABAT and ALDH6A1 was correlated with poor survival. Overexpression of ABAT or ALDH6A1 significantly attenuated cell proliferation and migration, and impaired lactate production. In ABAT increased ccRCC cells, the ratio of NADPH/NADP+ was reduced. Finally, we demonstrated that ABAT and ALDH6A1 were directly regulated by a tumor suppressor, HNF4A. Conclusions These observations identified HNF4A-regulated low-expressed ABAT and ALDH6A1 as promising diagnostic and prognostic biomarkers for ccRCC.


Introduction
Renal cell carcinoma (RCC) is a malignant tumor that originates from the renal parenchyma urinary tubule epithelia. Clear cell renal cell carcinoma (ccRCC) is the most common pathological type of RCC (60-85%) [1]. Surgical resection is still the first choice for treatment of localized renal cancer [2]. Because kidney cancer is not sensitive to radiotherapy and chemotherapy, some patients who cannot undergo surgical treatment need to consider targeted therapy. Since the US FDA approved sorafenib treatment of metastatic renal cell carcinoma in 2006, it has successively approved a series of drugs, mainly divided into three categories: 1 Tyrosine kinase inhibitors (TKI) categories: sunitinib, pazopanib, etc.; 2 anti-vascular endothelial growth factors: ranibizumab, bevacizumab, etc.; 3 mammalian target of rapamycin (mTOR) inhibitor: everolimus, temsirolimus, etc. [3]. Although targeted drugs prolong the patient's survival time, their high cost and prolonged use, which leads to the gradual emergence of problems such as drug resistance and toxicity and side effects, limit their widespread use in clinical practice. Therefore, the detailed investigation of the pathogenesis of renal cancer and the search for new targets for targeted therapy of renal cancer are the focus of current research on renal cancer.
Metabolic disorders are one of the important hallmarks of tumors [4]. The proliferation of tumor cells usually requires the re-adjustment of their metabolism to produce enough energy and synthetic raw materials [5]. ccRCC is a tumor characterized by abnormal glucose and lipid metabolism [6]. Numerous studies have shown that multiple pathways associated with gluconeogenesis (e.g., pyruvate metabolism, valeric acid metabolism, arginine and proline metabolism, and degradation of valine, leucine, and isoleucine) are down-regulated in ccRCC [6][7][8]. ABAT (gamma-aminobutyric acid transaminase) [9] and ALDH6A1 (aldehyde dehydrogenase 6 family, member A1) [10] are involved in the pyruvate metabolism and catabolism of valine, leucine, and isoleucine. Both of them are differential expressed genes (DEGs) of The Cancer Genome Atlas (TCGA)-kidney renal clear cell carcinoma (KIRC) data [11][12][13]. The role of these two genes in ccRCC development and their gene regulation mechanisms have not been discussed in depth. Therefore, this study will explore the expression of ABAT and ALDH6A1, their clinical significance and biological functions, correlations between ABAT and ALDH6A1 gene expression, and the transcription factor involved in the regulation of the expression of these two genes in ccRCC.

Cell lines and main reagents
HEK293T cells were purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences (Shanghai, China). ACHN and 786-O cells were provided by Shanghai GK Genentech Co., Ltd. (Shanghai, China). Cells freshly amplified and frozen were used every 2-3 months. They were routinely tested for the absence of mycoplasma contamination. Dulbecco's modified Eagles medium (DMEM) high glucose medium and RPMI 1640 medium were purchased from Hyclone (Thermo Fisher Scientific, Inc.). Fetal bovine serum was purchased from Gibco (Thermo Fisher Scientific, Inc.). Transfection reagent Lipofiter ™ was purchased from HanBio (Shanghai, China). TRIzol was purchased from Invitrogen (Thermo Fisher Scientific, Inc.). A complementary DNA (cDNA) kit was purchased from Fermentas (Thermo Fisher Scientific, Inc.), and a SYBR Green real-time polymerase chain reaction (PCR) kit was purchased from Applied Biosystems (Thermo Fisher Scientific, Inc.). ABAT and ALDH6A1 antibodies were purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). ACTIN antibody and Flag-tag antibody were purchased from Cell Signaling Technology (Danvers, MA, USA). A watersoluble tetrazolium salt (WST)-1 cell proliferation assay kit was purchased from Roche (Mannheim, Germany). Horseradish peroxidase (HRP)-tagged secondary antibody and ECL kits were purchased from Pierce (Thermo Fisher Scientific, Inc.).

Western blotting detection
SDS-PAGE electrophoresis was performed at 50 μg/well for each group of proteins quantified by BCA. After electrophoresis, the cells were transferred to PVDF membranes at a constant pressure of 60 V for 180 min. The blocking solution was blocked for 1 h and reacted with the primary antibody at 4 °C overnight; TBST washed three times; HRP-labeled goat anti-rabbit antibody reaction at room temperature for 1 h, TBST washed four times; added ECL luminescence, exposed on an X-ray film, film development, fixing, washing with water, drying, preservation, and photographing.

Immunohistochemical staining
A tissue microarray (HKid-CRCC060PG-01; Shanghai Outdo Biotech Co., China) was constructed with formalin-fixed paraffin-embedded 29 paired ccRCC and adjacent non-tumor renal tissues. The demographic information was summarized in Additional file 1: Table S2. The paraffin-embedded sections were stained with antibodies against ABAT or ALDH6A1 antibody (dilution at 1:750 or dilution at 1:3000; HPA041528; HPA029073) at 4 °C overnight and incubated with secondary antibody. Negative controls were stained with isotype control IgG. Then, the sections were counterstained with hematoxylin and eosin. Immunohistochemical staining intensity was based on the proportion of cell staining and scored from 0 to 3 following criteria described previously [24]. The slides were analyzed by standard light microscopy.

Lentivirus transfection
The lentivirus pLVX-Puro-ABAT, pLVX-Puro-ALDH6A1, and control virus pLVX-Puro were purchased from the Public Protein/Plasmid Library (PPL, Nanjing, China). For virus infection, the virus was diluted with serum-containing cell culture medium containing 8 μg/mL polybrene. The fresh medium was changed after 12 h. Puromycin (1 μg/mL, Sigma) was then added to select stable cells (10 days).

Water-soluble tetrazolium detects cell proliferation
Cells of the observation group, the control group, and the blank group were taken, and the number of cells was adjusted to 2 × 10 3 cells/ml, and 100 μL cells were inoculated into a 96-well cell culture plate. After 24 h, the cells were adhered and 10 μL of water-soluble tetrazolium (WST-1) reagent was added for 24, 48 and 72 h testing, and incubated for 45 min. Then, the optical density (D450) of each well at 450 nm wavelength was measured with a spectrophotometer (Multiskan GO, Thermo Scientific). Each group set up four complex holes.

Cell death detection
ABAT, ALDH6A1 stable overexpressed or control ACHN and 786-O cells (1 × 10 5 /well) were plated and grown in 100-mm culture dishes overnight. Cells were harvested and cell death rates were determined using a Cell Death Detection ELISA Kit (Cat. No. 11544675001, Roche, Switzerland) following the manufacturer's instructions.

Measurement of lactic acid
Levels of lactic acid were measured using a Lactate Assay Kit (Sigma-Aldrich, USA) following the manufacturer's instructions. In brief, ABAT, ALDH6A1 stable overexpressed or control ACHN and 786-O cells were treated with Lactate Assay Buffer and the Reaction Mix was added. The reaction mixture was incubated at room temperature for 30 min. The samples were analysed at an absorbance at 450 nm using a spectrophotometer.

Measurement of NADPH and NADP
Levels of NADPH and NADP+ were measured using a NADP/NADPH Quantitation Kit (Sigma-Aldrich, USA) following the manufacturer's instructions. In brief, ABAT, ALDH6A1 stable overexpressed or control ACHN and 786-O cells were treated with NADP/NADPH extraction buffer and the NADPH developer was added. The reaction mixture was incubated at room temperature for 1 h. The samples were analysed at an absorbance at 450 nm using a spectrophotometer (Multiskan GO, Thermo Scientific).

Plasmid and transfection
The pHNF4A-flag and its control plasmid were purchased from the Public Protein/Plasmid Library (PPL, Nanjing, China). HEK293T and ACHN cells were transfected with pHNF4A-flag or control plasmid. At 48 h after transfection, the cells were extracted for quantitative PCR and Western blot assay.

Luciferase assay
HEK293 cells were plated in 24-well plates and were co-transfected with 100 ng of reporter construct (pGL3-Basic, pGL3-ABAT-promoter, or pGL3-ALDH6A1-promoter), 500 ng of HNF4A expression vector or control vector, and 20 ng of the internal Renilla vector using Lipofiter. After 48 h, the luciferase activity was measured using a dual luciferase reporter assay system (Promega, Madison, WI, USA).

Statistical processing
SPSS13.0 software was used for statistical analysis. The data obtained were expressed in terms of mean ± SD.
Single-factor repeated-measures analysis of variance found a statistically significant difference at p < 0.05.

Identification of ABAT and ALDH6A1 as downregulated genes in ccRCC by GEO data analysis
Data mining was performed on the Oncomine database, and four data sets were selected, all of which were detected using the Affymetrix genomic array (Human Genome U133A Array). ABAT and ALDH6A1 were determined to be the downregulated DEGs (Fig. 1a). KEGG pathway enrichment (Metascape) revealed that the top 120 down-regulated DEGs were correlated with the biosynthesis of amino acids, carbon metabolism, endocrine and other factor-regulated calcium reabsorption, valine, leucine and isoleucine degradation, and lysine degradation pathways (Fig. 1b). The expression of ABAT and ALDH6A1 was significantly reduced in human ccRCC samples among all four datasets (Fig. 2a,  ). Protein association network analysis was performed by STRING. The result shows that ABAT and ALDH6A1 were both involved in the valine, leucine, and isoleucine degradation and beta-alanine metabolism pathways (Fig. 2c).

ABAT and ALDH6A1 are downregulated in TCGA renal clear cell carcinoma data
In the TCGA-KIRC dataset, there were 72 paraneoplastic tissues and 538 tumor tissues. The results still showed decreased expression of ABAT and ALDH6A1 in tumor compared with paraneoplastic tissues (p < 0.0001) (Fig. 3a, b). The expression of ABAT and ALDH6A1 was decreased in the four pathologically graded tissues, but there was no significant difference between the individual grades, see Fig. 3c, d. Analysis of patient survival revealed that patients with relatively low expression of ABAT or ALDH6A1 tended to have shorter overall survival (Fig. 3e-f ). These results suggest that the down-expression of ABAT and ALDH6A1 could be used as biomarkers for ccRCC and that their expression correlates with the survival of patients.
Based on the pan-cancer analysis, both ABAT and ALDH6A1 were significantly down-expressed in three types of renal cell carcinomas, KIRC, kidney renal papillary cell carcinoma (KIRP), and kidney chromophobe (KICH) (Fig. 3g-h). ABAT and ALDH6A1 have similar expression profiles in cancers.

Expression of ABAT and ALDH6A1 is downregulated in ccRCC clinical samples
From 2013 to 2015, our center collected a total of 50 pairs of pathologically identified and surgically removed ccRCC tumor tissues and their adjacent tissues. Fluorescent quantitative PCR assays showed that ABAT expression was reduced in 94% of tumor tissues (Fig. 4a). The paired analysis showed that ABAT mRNA levels were significantly decreased in renal clear cell carcinoma tissue (p < 0.0001) (Fig. 4b). Similar results were observed in ALDH6A1 expression (Fig. 4c, d).
A ccRCC paraffin tissue microarray was used for immunohistochemical detection. Results showed that ABAT protein levels were reduced in ccRCC tissues (Fig. 4e), which reduction was statistically significant (p < 0.0001) (Fig. 4f ). Similar results were observed for ALDH6A1 protein levels ( Fig. 4g-h). These data indicate significant decreases in ABAT and ALDH6A1 expression in the ccRCC clinical samples.

Overexpression ABAT or ALDH6A1 suppresses renal clear cell carcinoma cell line growth and migration
We next sought to examine the functions of ABAT and ALDH6A1 in ccRCC cells using a lentivirus-mediated overexpression system. The high expression of ABAT and ALDH6A1 was confirmed in ACHN cells (Fig. 5a). The results of WST-1 assay showed that the proliferation of ABAT and ALDH6A1 overexpression group was significantly lower than that in the control group at 72 h (p < 0.05, Fig. 5b). Compared with the control group, the ACHN cells with stable ABAT or ALDH6A1 expression had a significantly reduced ability to form colonies (Fig. 5c). We further assessed the effects of ABAT and ALDH6A1 on cell death. Using cell death ELISA assay, it was observed that overexpression of ABAT and ALDH6A1 increased death rates in ACHN cells (Fig. 5d).  Those data were verified in the 786-O cell line, another ccRCC cell line (Fig. 5e-h). Moreover, cell migration was significantly decreased in the ABAT or ALDH6A1 overexpression ACHN and 786-O cells (Fig. 5i). These data collectively indicate that ABAT and ALDH6A1 act as tumor suppressors of ccRCC.

ABAT and ALDH6A1 overexpression impairs metabolic modeling in ccRCC cells
To investigate the effects of metabolic enzymes ABAT and ALDH6A1 on ccRCC metabolism and the role of gene abnormalities in driving ccRCC tumorigenesis, TCGA data were used to analyze gene set enrichment analysis (GSEA) of ABAT and ALDH6A1. Besides the valine, leucine, and isoleucine degradation pathway, the citrate cycle (TCA cycle) and fatty acid degradation correlated with downregulated levels of ABAT and ALDH6A1 (Fig. 6a, b). These metabolism pathways are known to be dysregulated in cancers, especially in ccRCC. Cell-proliferation-related pathways were found to be negatively correlated with ABAT and ALDH6A1, such as the cell cycle pathway with ABAT ( Fig. 6c) and DNA replication pathway with ALDH6A1 (Fig. 6d). Increased production of lactic acid and NADPH are prominent characteristics of the metabolic remodeling of tumor cells [25]. To further confirm the link between ABAT and ALDH61 expression and metabolic modeling in ccRCC, we measured lactate levels and the NADPH/ NADP + rate in ACHN and 786-O cells. We found that upregulated ABAT or ALDH6A1 impaired lactate production (Fig. 6e), and ABAT overexpression decreased the ratio of NADPH/NADP + (Fig. 6f ). These data suggest that overexpression of ABAT or ALDH6A1 impairs cell oncologic-metabolism.

Transcription factor HNF4A regulates ABAT and ALDH6A1 gene expression
In UALCAN, ALDH6A1 is the most expression-related gene of ABAT through Pearson correlation analysis, with a Pearson correlation coefficient (Pearson-CC) = 0.66 (Fig. 7a). By analyzing the transcription factors binding to the ABAT and ALDH6A1 promoters, we found that transcription factor HNF4A could simultaneously regulate ABAT and ALDH6A1, and HNF4A has a positive correlation with ABAT and ALDH6A1 expression (Fig. 7b). HNF4A is a well-known suppressor of ccRCC [26]. The expression of HNF4A is significantly downregulated in ccRCC and loss of HNF4A promotes tumorigenesis in the kidney [27]. We analyzed the promoter sequence of 2000 bp before the transcription initiation site on Jaspar. Four common binding sequences of HNF4A were selected (Fig. 7c). In each of the promoter regions of ABAT and ALDH6A1, there are two regions with more transcription factor binding sites (Fig. 7d).
To examined whether ABAT and ALDH6A1 expressions in renal cell lines are regulated by exogenous HNF4A, HEK293T and ACHN cells were transfected with HNF4A expression vector or control plasmid. 786-O cells were not used in the transfection experiments because of the inability to achieve high transfection efficiency in 786-O cells. As shown in the Fig. 8a, exogenous overexpression of HNF4A induced the mRNA levels of endogenous ABAT and ALDH6A1 in HEK293T and ACHN cells. ABAT and ALDH6A1 protein levels were also significantly increased in HNF4A overexpressing cells (Fig. 8b). To further verify that HNF4A directly targets the promoter of ABAT and ALDH6A1, we performed a ChIP assay to assess the HNF4A-DNA interaction. Figure 8c demonstrates that Flag antibody co-immunoprecipitated the DNA sequence from the ABAT and ALDH6A1 promoter region by ordinary PCR. Real-time quantitative PCR showed similar results. Luciferase-based assays were subsequently used to verify the regulated role of HNF4A. As expected, in co-transfection experiments with pGL3-ABAT-promoter or pGL3-ALDH6A1-promoter and pHNF4A-Flag in HEK293T cells, we found that HNF4A increased the promoter activities of ABAT and ALDH6A1 (Fig. 8d). Together, these data indicate that HNF4A directly regulate expression of ABAT and ALDH6A1.

Discussion
In this study, we first analyzed four of the GEO datasets and TCGA-KIRC data and found that the expression of ABAT and ALDH6A1 was decreased in ccRCC. The patients with low ABAT or ALDH6A1 gene expression had a worse survival period. Two ccRCC cell lines, ACHN and 786-O, were infected with ABAT or ALDH6A1 overexpressing lentivirus. The ability of cell clone formation was decreased in ABAT or ALDH6A1 overexpression cells, and the percentage of cell death was higher than in the control group. Transwell chamber experiments found that the cell migration ability of ABAT or ALDH6A1-overexpressed cells was inhibited. In order to understand the mechanism of overexpressing ABAT or ALDH6A1 in renal cancer cells, database analysis shows that ABAT and ALDH6A1 are mainly involved in oncologic metabolism. By measuring the lactic acid production and NADP/NADPH ratio of renal cancer cells, the results showed that after ABAT or ALDH6A1 overexpression, oncologic metabolism was impaired. We also demonstrated that ABAT and ALDH6A1 are directly regulated by a well-known tumor suppressor, transcription factor HNF4A.
In the present study, low ABAT expression is significantly associated with basal-like breast cancer aggressiveness by activating gamma-aminobutyric acid (GABA) signaling. Patients with ABAT deficiency display poor chemotherapy treatment outcome [28]. ABAT and ALDH6A1 are proved to be hub genes in association with metastasis risk and prognosis in hepatocellular carcinoma (HCC) [29]. Both of them are down-regulated in HCC [30,31]. Our study identified that loss of ABAT and ALDH6A1 contributes to ccRCC tumor growth. Although the specific mechanisms of ABAT and ALDH6A1 in glutamate metabolism, glutathione GSH/GSSG determination, glucose uptake, and oxygen consumption should be determined in future research, our findings suggest that ccRCC patients might benefit from their influence on amino acid metabolism. HNF4A inhibits proliferation in the stomach [32], colon [33] and kidney [27]. High expression of HNF4A leads to decreased growth of cancer cells. HNF4Amediated inhibition of cell proliferation involves The p-value was measured using Student's t-test; *p < 0.05, compared with the control cells. b HEK293T and ACHN cells were transiently transfected with pHNF4A-Flag or a control vector for 24 h. The protein levels of HNF4A-Flag, ABAT, and ALDH6A1 were detected by Western blotting. The experiments were repeated three times. c ChIP assay was used to analyze HNF4A binding to the two regions of ABAT and ALDH6A1 promoter in HEK293T cells. The p-value was measured using Student's t-test; *p < 0.05, compared with IgG control. The experiments were repeated three times. d HEK293T cells were transiently transfected with pRL-TK and pGL3-Basic-control, pGL3-ABAT-promoter or pGL3-ALDH6A1-promoter and together with the empty vector or HNF4A expression vector for 24 h. Cell lysates were analyzed for luciferase activity. The experiments were repeated three times. The p-value was measured using Student's t-test; *p < 0.05, compared with that of the empty-vector-transfected control cells