Zinc finger protein 831 promotes apoptosis and enhances chemosensitivity in breast cancer by acting as a novel transcriptional repressor targeting the STAT3/Bcl2 signaling pathway

Abstract

Emerging evidence suggested that zinc finger protein 831 (ZNF831) was associated with immune activity and stem cell regulation in breast cancer. Whereas, the roles and molecular mechanisms of ZNF831 in oncogenesis remain unclear. ZNF831 expression was significantly diminished in breast cancer which was associated with promoter CpG methylation but not mutation. Ectopic over-expression of ZNF831 suppressed breast cancer cell proliferation and colony formation and promoted apoptosis in vitro, while knockdown of ZNF831 resulted in an opposite phenotype. Anti-proliferation effect of ZNF831 was verified in vivo. Bioinformatic analysis of public databases and transcriptome sequencing both showed that ZNF831 could enhance apoptosis through transcriptional regulation of the JAK/STAT pathway. ChIP and luciferase report assays demonstrated that ZNF831 could directly bind to one specific region of STAT3 promoter and induce the transcriptional inhibition of STAT3. As a result, the attenuation of STAT3 led to a restraint of the transcription of Bcl2 and thus accelerated the apoptotic progression. Augmentation of STAT3 diminished the apoptosis-promoting effect of ZNF831 in breast cancer cell lines. Furthermore, ZNF831 could ameliorate the anti-proliferation effect of capecitabine and gemcitabine in breast cancer cell lines. Our findings demonstrate for the first time that ZNF831 is a novel transcriptional suppressor through inhibiting the expression of STAT3/Bcl2 and promoting the apoptosis process in breast cancer, suggesting ZNF831 as a novel biomarker and potential therapeutic target for breast cancer patients.

Abbreviations

BLCA

bladder urothelial carcinoma

BRCA

breast carcinoma or breast cancer

Cape

capecitabine

ChIP

chromatin immunoprecipitation

CI

confidence interval

COAD

colon adenocarcinoma

ER

estrogen receptor

FDR

false discovery rate

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

GBM

glioblastoma multiforme

GEM

gemcitabine

GO

gene ontology

GSEA

gene set enrichment analysis

HER2

human epidermal growth factor receptor

KEGG

Kyoto Encyclopedia of Genes and Genomes

MLD BRCA

mixed lobular and ductal breast carcinoma

m6A

N6-methyladenosine

PR

progesterone receptor

qRT-PCR

real-time quantitative polymerase chain reaction

sCNA

DNA copy-number alterations

STAT3

signal transducer and activator of transcription 3

TCGA

the Cancer Genome Atlas

ZNFs

zinc finger proteins

ZNF831

zinc finger protein 831

Introduction

Breast cancer (BRCA) keeps up the leading cancer-related cause of disease in women.¹ With an estimated 2.3 million new patients (11.7%), BRCA has exceeded lung cancer (11.4%) as the most commonly diagnosed cancer in the world.² Numerous risk factors played vital roles in BRCA progression, including genetic predisposition, exposure to specific environmental factors, microbiome, and immune microenvironment.^3,4 Not only considerable mortality could be caused by lung, bone, brain, and liver metastases,5, 6, 7, 8 but also the poor prognosis is partially attributed to chemoresistance and anti-apoptosis responses of the cancer cells.⁹ Even though much epidemiological and clinical research has been conducted for years, more mechanisms of BRCA oncogenesis about apoptosis and chemoresistance remain to be explored. Therefore, identifying novel biomarkers associated with the apoptotic process and chemosensitivity and exploring their molecular mechanism would be meaningful and crucial for early diagnosis and targeting therapy among BRCA patients.

Zinc finger proteins (ZNFs) belong to the largest transcription factor family and the zinc finger motifs have been revealed to possess interacting abilities with DNA, RNA, proteins, and lipids.10, 11, 12, 13 Up to now, a total of 8 different classes of zinc finger motifs have been reported, including Cys2His2-like, Gag knuckle, Treble clef, Zinc ribbon, Zn2/Cys6, TAZ2 domain-like, Zinc binding loops, and Metallothionein.14, 15, 16, 17, 18, 19, 20, 21 With different combinations of multiple zinc finger motifs, ZNFs may greatly expand their diverse roles in gene regulation under different cell contexts or stimulations. Generally, they are characterized by classic zinc finger motifs and can activate or repress the transcription level of target genes by binding to gene promoters.²² Moreover, one member of ZNFs may display its functions on transcriptional regulation toward thousands of downstream genes owing to the same or similar promoter binding sites. As a result, ZNFs exert significant differential biological functions and hence influence carcinogenesis and progression, including proliferation, migration, invasion, stemness, apoptosis, cell cycle, senescence, autophagy, etc.23, 24, 25, 26, 27, 28, 29 Studies about how ZNFs influence and participate in BRCA oncogenesis outbursts and come to the top in the last decade. Many members of ZNFs were exposed and explored. Zinc finger protein 831 (ZNF831) is one of them and its transcription product locates in the cytoplasm. Recent evidence has indicated that ZNF831 was involved both in immune regulation^30,31 and preeclampsia.³² It also has been reported that ZNF831 was negatively correlated with stem cell phenotype and reduced the relapse risk of BRCA.³⁰ This evidence suggested that ZNF831 may play a pivotal role in BRCA tumorigenesis and progression. Up to now, the role and molecular mechanism of ZNF831 in BRCA oncogenesis remain unclear.

In this study, we investigated the expression regulation, biological function, molecular mechanism, and clinical significance of ZNF831 gene in breast cancer. We found that ZNF831 expression down-regulation was mainly regulated by promoter methylation but not a genetic alteration in breast cancer. ZNF831 expression inhibited tumorigenesis in breast cancer in vitro and in vivo. Mechanistically, ZNF831 promoted BRCA apoptosis via transcriptionally repressing signal transducer and activator of transcription 3 (STAT3) and subsequently inhibiting Bcl2. ZNF831 enhanced chemosensitivity to capecitabine and gemcitabine in BRCA. Our study demonstrated for the first that ZNF831 gene acted as a novel tumor suppressor by regulating the STAT3/Bcl2 signaling axis in BRCA, hoping these results could contribute to BRCA tumorigenesis exploration and lay the foundation for early diagnosis and targeted therapy.

Materials and methods

Bioinformatics analysis

The transcription level of ZNF831 was analyzed through several public databases, including TIMER 2.0 (http://timer.cistrome.org/), TCGA (https://www.cancer.gov/), and Oncomine (https://software.oncomine.com/). In detail, the TIMER 2.0 database was conducted to compare the expression of ZNF831 between tumor and adjacent normal tissues in pan-cancer by the “Gene” module. Besides, we also performed the transcriptional differences of ZNF831 between tumor and adjacent normal samples among the TCGA-BRCA cohort and the Oncomine cohorts. Furthermore, ZNF831 protein expression was searched from the Human Protein Atlas (HPA, https://www.proteinatlas.org/) database. DNA methylation data of ZNF831 was analyzed through probe-level methylation and CpG-aggregated methylation modules of the SMART. Genetic alterations of ZNF831 were explored in the cBioPortal through breast invasive carcinoma. LinkedOmics was used for ZNF831 correlated gene analysis. GSEA 4.0.3 from GSEA was used to analyze 20,530 genes in the TCGA genomic matrix grouped by the median ZNF831 expression level.

Cell culture

MDA-MB-231 and MCF-7 cells were cultured in Dulbecco's Modified Eagle Medium/High-glucose medium (HyClone, USA) containing 10% fetal bovine serum (FBS, HyClone, USA) at 5% CO₂ and 37 °C. BT474 and HCC38 cells were cultured in Roswell Park Memorial Institute-1640 medium (HyClone, USA) containing 10% fetal bovine serum (HyClone, USA) at 5% CO₂ and 37 °C. All cells were procured directly from the American Type Culture Collection. Decitabine (5-Aza-2′-deoxycytidine) was supplied as a powder (Abcam, UK) and dissolved in DMSO at a final concentration of 10 μM.

Tissues samples

The study cohort contained 14 pairs of specimens including paired adjacent tissues and BRCA tissues. All tissues were obtained from the Daping Hospital of Army Medical University (Chongqing, China). Prior patient consent and approval from the Ethics Committee of Daping Hospital were obtained. Patients treated with radiotherapy/chemotherapy before surgery were excluded from this study.

Plasmid transfection

ZNF831 over-expression associated plasmids (Pcdna3.1-3xFLAG and Pcdna3.1-ZNF831-3xFLAG), ZNF831 knockdown associated plasmids (pLVx-mCherry-T2A-puro and pLVx-ZNF831-shRNA-mCherry-T2A-puro) and STAT3 over-expression associated plasmids (pLVX-Puro-FLAG and pLVX-Puro-STAT3-FLAG) were constructed and purchased for transfection (Leqin Corporation). The primer sequences for constructing the plasmid were listed in Table S1. Transient and stable transfection was performed by ViaFect (Promega, USA) following the manufacturer's instructions.

Cell viability assay

Cells were seeded into 96-well plates 24 h ahead and were transiently transfected for 48 h. Cell proliferation was determined by the CellTiter 96 AQueous One Solution Cell Proliferation Assay system (Promega, USA) according to the manufacturer's instructions. After 1 h incubation in a humidified atmosphere (5% CO₂, 37 °C), absorbance at 490 nm was measured every 24 h using a SpectraMax i3 microplate reader (Molecular Devices, USA).

Colony formation assay

Cells were seeded into 6-well plates. After 2 weeks of incubation in a humidified atmosphere (5% CO₂, 37 °C), the colonies were fixed with 4% paraformaldehyde (Solarbio) for 20 min and stained with crystal violet (Solarbio) at room temperature. Cell colonies were imaged and the number of colonies with >50 cells was counted.

Cell apoptosis assay

Cells were seeded into 6-well plates 24 h ahead and were transiently transfected for 48 h. Cell apoptosis was analyzed using the Annexin V-APC/7-AAD (7-amino-actinomycin D) Apoptosis Detection Kit (KeyGEN, China) according to the manufacturer's instructions. The cell apoptosis data were analyzed by FlowJo 7.6.1 software (Tree Star, USA).

qRT-PCR

RNA was extracted using RNAiso Plus (TaKaRa, Japan) according to the manufacturer's instructions, and then reverse transcription was performed by the GoScript reverse transcriptional system (Promega, USA). The RNA expression level was determined using the GoTaq qPCR Master Mix (Promega, USA) on the Bio-Rad Real-Time PCR System (Bio-Rad, USA). Sequences of primers for quantitative real-time polymerase chain reaction (qRT-PCR) in this study are listed in Table S2.

Western blot analysis

The protein of the cells was obtained for Western blot analysis. Cells were collected to homogenize in ice with RIPA buffer solution (Solarbio, China) for protein extraction. The 0.5 M EDTA (pH 8.0) (Solarbio, China), protein phosphatase inhibitor mixture (Solarbio, China), and deacetylase inhibitor mixture (Solarbio, China) were added for cell lysis. The concentration of protein was measured by a BCA protein assay kit (CWBIO, China). 12% SDS-PAGE was selected for electrophoresis and the proteins were transferred onto the PVDF membrane. The membrane was incubated with 3% BSA at room temperature for 1 h. The primary antibody was then incubated at the appropriate concentration over night at 4 °C. The antibodies used in this study were as follows: STAT3 (1:1000, Proteintech), P-STAT3 (Tyr705) (1:1000, Beyoime), Bcl2 (1:1000, Proteintech), Bax (1:1000, Proteintech), GAPDH (1:1000, CWBIO). Secondary antibodies conjugated with horseradish peroxidase were used for further incubation for 2 h at room temperature and visualized using ECLTM Western Blotting Detection Reagent (Bio-Rad, USA).

Chromatin immunoprecipitation (ChIP) assay

ChIP analysis was performed on transiently transfected BRCA cells with ectopic over-expressed ZNF831 using the SimpleChIP Enzymatic Chromatin IP Kit (Magnetic Beads) (CST, USA) according to the manufacturer’s instructions. DYKDDDDK Tag (D6W5B) Rabbit mAb (binds to the same epitope as Sigma’s Anti-FLAG M2 Antibody (CST, USA) was used to drag FLAG. Sequences of primers for promoter region in this study are listed in Table S3.

Luciferase reporter assay

Cells were seeded in 24-well plates and human STAT3 promoter (−2000 to +200) luciferase reporter plasmid was transiently transfected with ZNF831 expression plasmids into cells. Renilla luciferase pRL-SV40 (Promega) was used to normalize the luciferase activity, which was further measured using the dual-luciferase reporter assay system (Promega).

Chemosensitivity assay

Chemotherapeutic drugs were purchased and a cell viability assay was conducted to evaluate the function of ZNF831 in chemosensitivity. The final concentration for capecitabine (RO 09-1978) (Selleck, USA) was 1000 μM for all, and gemcitabine (LY-188011) (Selleck, USA) was 500 μM for MDA-MB-231, BT474, and HCC38, and 1000 μM for MCF-7.

Cell-line-derived xenograft model

All procedures were approved by the Laboratory Animal Welfare and Ethics Committee of the Third Military Medical University and carried out in strict compliance with the relevant guidelines. Six-week-old female BALB/C nude mice were randomly assigned to the vehicle and over-expression groups. Induction of tumor xenografts was performed by subcutaneous injection of 0.1 mL cell suspensions containing 5 × 10⁶ cells in the logarithmic phase into the right flank of the animal. Tumor length (L) and width (W) were measured via vernier caliper twice a week, and the volume was calculated based on the following equation: V = L × W²/2. After one month, the mice were killed and the tumors were aseptically excised and weighed up. Tumor fragments were used for pathological analysis and IHC.

Statistical analysis

All statistical analyses were performed using SPSS 26.0 software (SPSS, Inc., Chicago, IL, USA). All data were presented as the mean ± standard deviation (SD) of at least three independent experiments. Student's t-test was used to analyze the significance of differences between the two groups. Wilcoxon test was used to analyze the significance of differential expression between tumor and adjacent normal tissues for ZNF831 across all TCGA tumor types. Pearson correlation test was selected as correlation analysis in LinkedOmics. P < 0.05 was considered statistically significant.

Results

ZNF831 is down-regulated in most cancer types and BRCA patients

To briefly explore the spectrum of ZNF831 expression across most human cancer types, we first analyzed the transcription level of ZNF831 through the TIMER 2.0 database. Results showed that ZNF831 was significantly down-regulated in many kinds of cancers, including bladder urothelial carcinoma (BLCA) (P < 0.001), breast invasive carcinoma (BRCA) (P < 0.001), colon adenocarcinoma (COAD) (P < 0.001), glioblastoma multiforme (GBM) (P < 0.001), etc (Fig. 1A). In addition, we verified the results of BRCA in unpaired (P < 0.0001) (Fig. 1B) and paired (P < 0.0001) (Fig. 1C) samples from TCGA-BRCA cohort. Furthermore, the result was enhanced again by 5 independent cohort studies for different BRCA subtypes from the Oncomine database (P < 0.05) (Fig. 1D–H). These findings were verified by qRT-PCR analysis of ZNF831 mRNA expression from 10 pairs of normal and BRCA samples in unpaired (P < 0.05) and paired (P < 0.05) ways (Fig. 1I, J). The expression difference in protein level was verified in immunohistochemistry from the HPA database and the results suggested that ZNF831 was localized in the cytoplasm and expressed poorly in BRCA tissues (Fig. S1A, B).

Figure 1.

ZNF831 is down-regulated in most cancer types and BRCA patients. (A) The differential expression between tumor and adjacent normal tissues for ZNF831 across all TCGA tumor types. The data was processed by the TIMER2.0 web server and presented as the means ± SEM (the Wilcoxon test). (B, C)ZNF831 mRNA level is diminished in BRCA tissue vs. (B) unpaired normal breast tissue and (C) adjacent paired normal breast tissue. The data was from the BRCA data set of the TCGA cohort (t-test). (D–H) The expression of ZNF831 between the normal breast tissue and (D) mucinous BRCA in TCGA breast research, (E) MLD BRCA in TCGA breast research, (F) invasive ductal BRCA in TCGA breast research, (G) tubular BRCA in Curtis breast research, and (H) ductal BRCA in situ in Ma breast 4 research. The data was from the BRCA data set of the Oncomine database (t-test). (I, J)ZNF831 mRNA expression level in BRCA tissue vs. (I) unpaired normal breast tissue and (J) adjacent paired normal breast tissue. The data was from clinical BRCA tissue samples (t-test). BRCA, breast carcinoma or breast cancer; MLD BRCA, mixed lobular and ductal breast carcinoma; qRT-PCR: real-time quantitative polymerase chain reaction. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

Down-regulation of ZNF831 correlates with its promoter hypermethylation in BRCA samples

To examine the potential mechanism of ZNF831 down-regulation in BRCA, we evaluated epigenetic and genetic alteration through SMART and the cBioPortal database. The results revealed 3 CG sites (cg25325636, cg16473762, cg09642670; Fig. S2A–C) were hypermethylated in the promoter region across many human cancers. The aggregation of cg25325636, cg16473762, and cg09642670 site methylation values showed the same trend (Fig. S2D). Among them, the methylation level of ZNF831 was all higher in BRCA tissue than normal (P < 0.05) (Fig. 2A), (P < 0.001) (Fig. 2B), (P < 0.0001) (Fig. 2C). The aggregation of cg25325636, cg16473762, and cg09642670 site methylation values showed the same trend in BRCA tissue than normal (P < 0.0001) (Fig. 2D). In addition, we found that ZNF831 expression levels were increased after demethylation treatment (decitabine) in MDA-MB-231 (P < 0.001) and MCF-7 (P < 0.05) BRCA cells (Fig. 2E), proving methylation involved in the down-regulation of ZNF831. Conversely, ZNF831 somatic mutation rates (Fig. 2F) and DNA copy-number alterations (sCNA) (Fig. 2G) were low in most cancer types. As for BRCA, alteration frequency was low in all four data subsets (Fig. 2H–K). It suggested that promoter methylation, but not mutation or sCNA, was associated with ZNF831 down-regulation in BRCA.

Figure 2.

Down-regulation of ZNF831 correlates with its promoter hypermethylation in BRCA samples. (A) The cg25325636 site methylation value in BRCA between normal and tumor. (B) The cg16473762 site methylation value in BRCA between normal and tumor. (C) The cg09642670 site methylation value in BRCA between normal and tumor. (D) The aggregation of cg25325636, cg16473762, and cg09642670 site methylation values in BRCA between normal and tumor. (E) The expression of ZNF831 was measured by qRT-PCR in MDA-MB-231 and MCF-7 after demethylation treatment. (F) Somatic mutation frequency of ZNF831 for each TCGA cancer type. (G) The relative proportion of different sCNA states of the ZNF831 for all TCGA cancer types. (H–K) Genetic alteration frequency of ZNF831 was observed from the cBioPortal online tool in four BRCA studies of TCGA ((H) TCGA, Cell 2015; (I) TCGA, Firehouse legacy; (J) TCGA, Nature 2012; (K) TCGA, Pan-cancer atlas) and genomic profiles were limited in mutations and putative copy-number alteration from GISTIC. BRCA, breast carcinoma or breast cancer; qRT-PCR: real-time quantitative polymerase chain reaction; sCNA, somatic DNA copy-number alterations. ∗P < 0.05, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

ZNF831 inhibits proliferation and promotes apoptosis of BRCA in vitro

To investigate the role of ZNF831 in BRCA oncogenesis, we analyzed the expression of ZNF831 in four BRCA cell lines (MDA-MB-231, MCF-7, BT474, HCC38) and one normal breast tissue cell line (MCF-10A). We found that ZNF831 expression was relatively higher in HCC38 cells and relatively lower in MDA-MB-231, MCF-7, and BT474 cells than that in MCF-10A (Fig. 3A). So, we expressed ectopic ZNF831 by Pcdna3.1-3xFLAG vector in MDA-MB-231, MCF-7, and BT474 (Fig. 3B) and knocked it down in HCC38 (Fig. 3C). The result from proliferation assay showed that over-expression of ZNF831 inhibited cell proliferation (Fig. 3D–F) and knockdown of ZNF831 could promote the proliferation (Fig. 3G). Over-expression of ZNF831 resulted in a significant reduction of colony formation (Fig. 3H, I). In contrast, the knockdown of endogenous ZNF831 in HCC38 increased cell viability (Fig. 3J, K). Besides, we found that ZNF831 triggered BRCA cell apoptosis (Fig. 3L–N) and representative images were presented in Figure S3A–C. The rate of cell apoptosis was decreased with ZNF831 knockdown (Fig. 2O) and representative images were presented in Figure S3D. Collectively, these data indicated that ZNF831 was a suppressor in BRCA.

Figure 3.

ZNF831 inhibits proliferation and promotes apoptosis of BRCA in vitro. (A)ZNF831 expressed relatively high in the HCC38 cell line and relatively low in MDA-MB-231, MCF-7, and BT474 cell lines compared to MCF-10A. (B) Ectopic ZNF831 expression in MDA-MB-231, MCF-7, and BT474 cell lines were measured by qRT-PCR. (C) Expression of ZNF831 knockdown in the HCC38 cell line was measured by qRT-PCR. (D–F) Proliferation assays were used to determine the proliferation effect of ZNF831 over-expression on (D) MDA-MB-231, (E) MCF-7, and (F) BT474 cells. (G) A proliferation assay was used to determine the proliferation effect of ZNF831 knockdown on HCC38 cells. (H, I) Colony formation assays were used to determine the proliferation effect of ZNF831 over-expression on MDA-MB-231, MCF-7, and BT474 cells. (H) Representative images and (I) quantification plots were presented. (J, K) A colony formation assay was used to determine the proliferation effect of ZNF831 knockdown on HCC38 cells. (J) Representative images and (K) quantification plots were presented. (L–N) Flow cytometric analysis of cell apoptosis of ZNF831 over-expression on (L) MDA-MB-231, (M) MCF-7, and (N) BT474 cells with Annexin V-APC/7-AAD double staining. Statistical plots were presented. (O) Flow cytometric analysis of cell apoptosis of ZNF831 knockdown on HCC38 with Annexin V-APC/7-AAD double staining. Statistical plots were presented. BRCA, breast carcinoma or breast cancer; qRT-PCR: real-time quantitative polymerase chain reaction; VC: vector control; NC: negative control; K.D.: knockdown. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

Genome-wide screening of potential target and pathway of promoting apoptosis for ZNF831 by bioinformatics modeling

To explore the molecular mechanism underlying tumor suppressive effects of ZNF831 in BRCA, we first analyzed the correlated genes through the LinkedOmics database and found there were 12,693 genes (P < 0.05) correlated with ZNF831 in the “LinkFinder” part, 8108 genes were positively associated and 4585 genes were negatively associated (Fig. S4A). The top 50 significant genes positively and negatively correlated with ZNF831 were shown in the heat map (Fig. 4A, B). A total description of the associated genes was detailed in Table S4. In the “LinkInterpreter” module, significant Gene Ontology (GO) annotations showed that all these associated genes were negatively associated with transcription by RNA polymerase I (FDR > 0.05), DNA-templated transcription (elongation) (FDR >0.05), and DNA-templated transcription (termination) (FDR > 0.05) in biological process (Fig. 4C). As for Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways, associated genes were negatively associated with RNA polymerase-related pathway (FDR < 0.05) and basal transcription factors (FDR >0.05) (Fig. 4D). To further forecast the potential tumor-suppressive mechanism of ZNF831, we carried out a Gene Set Enrichment Analysis (GSEA) of ZNF831 with KEGG gene sets in TCGA-BRCA cohort. With 1215 samples and 20,530 genes in the analysis, we adopted the median cut-off value of ZNF831 to divide the 1215 samples into phenotype high group and low group. RNA polymerase-associated KEGG pathway was enriched in the low group (Fig. 4E); while JAK/STAT signal pathway and apoptosis-associated pathways were enriched in the high group (Fig. 4F). These results reminded us that ZNF831 may play a tumor suppressor and participate in transcriptional regulation process and apoptotic pathways, including JAK/STAT pathway. Then, we identified differentially expressed genes between the vector group and the ZNF831 ectopic over-expressed group in the MDA-MB-231 cells by RNA sequencing. A total of 778 differentially expressed genes (441 up-regulated and 337 down-regulated mRNAs, P < 0.05) were obtained (Fig. 4G) and these genes were detailed in Table S5. Regulation of transcription (DNA-templated) (P < 0.001), negative regulation of transcription (DNA-templated) (P < 0.05), transcription (DNA-templated) (P < 0.001), and RNA splicing (P < 0.01) associated biological processes were obtained by GO enrichment in up-regulated genes (Fig. 4H). Meanwhile, activation of Janus kinase activity (P < 0.05) associated biological process was obtained by GO enrichment in down-regulated genes (Fig. 4I). Signal transduction and cancer-related pathways were enriched in a considerably high percentage (Fig. S4B). Considering that JAK/STAT pathway and Janus kinase activity-associated biological process were enriched and both of them were apoptosis-related, we aimed at figuring out the key role in ZNF831-associated JAK/STAT family members.

Figure 4.

Genome-wide screening of potential target and pathway of promoting apoptosis for ZNF831 by bioinformatics modeling. (A, B) Important positively (A) and negatively (B) correlated genes with ZNF831 through the LinkedOmics database, TCGA_BRCA was selected as the cancer cohort and the platform of HiSeq RNA was chosen as the search and target dataset. (C, D) Bar charts of interesting (C) biological processes and (D) KEGG pathways of ZNF831 associated genes with GO enrichment from LinkedOmics. (E, F) Bubble charts of interesting KEGG pathways enriched by the GSEA app in phenotype (E) ZNF831-low and (F) ZNF831-high group through TCGA genomic matrix. (G) Transcriptome sequence difference in ectopic ZNF831 expression group and control group of MDA-MB-231 cells. (H, I) Bubble charts of interesting biological processes enriched in (H) up-regulated and (I) down-regulated gene sets from transcriptome sequence in MDA-MB-231 cells. BRCA, breast carcinoma or breast cancer; KEGG: Kyoto Encyclopedia of Genes and Genomes; GO: gene ontology; GSEA: gene set enrichment analysis; TCGA: the Cancer Genome Atlas.

ZNF831 induces cell apoptosis through direct binding to the specific site of the STAT3 promoter region and transcriptionally suppresses STAT3 expression in BRCA cell lines

To define the underlying molecular mechanism by which ZNF831 induces apoptosis promotion, we focused on apoptosis-related star molecules of the JAK/STAT pathway. Firstly, we evaluated the mRNA level of STAT3 and protein expression of STAT3 and P-STAT3 in BRCA cell lines which differentially expressed ZNF831, on account of its vital status in the JAK/STAT pathway. Transcriptionally, STAT3 was down-regulated with the ectopic over-expression of ZNF831 in MDA-MB-231 and BT474 (Fig. 5A, B), which was consistent with protein change (Fig. 5D); Whereas, whereas the knockdown of ZNF831 increased STAT3 expression both in mRNA (Fig. 5C) and protein level (Fig. 5D) in HCC38. The phosphorylated form of STAT3 was evaluated and the variation of P-STAT3 was consistent with the STAT3 (Fig. 5D). To further explore the change of the STAT3/Bcl2 apoptotic pathway, we evaluated the protein level of Bcl2 and Bax. After the over-expression of ZNF831, Bcl2 was down-regulated when Bax was not changed in MDA-MB-231 and BT474 (Fig. 5D); whereas, knockdown of ZNF831 increased Bcl2 but no change in Bax in HCC38 (Fig. 5D). Coincidently, the variation of Bcl2 was consistent with the upstream STAT3 change. Then, we applied ectopic STAT3 over-expression to explore whether amelioration of STAT3 rescued the apoptosis-promoting function induced by enhanced expression of ZNF831 in breast cancer. Restored expression of STAT3 reversed the promotion of apoptosis in MDA-MB-231 and BT474 cell lines (Fig. 5E, F) and representative images were presented in Figure S5A and B. STAT3 and P-STAT3 were successfully ameliorated in ZNF831 and STAT3 co-expressed MDA-MB-231 and BT474 cell lines (Fig. 5G). Moreover, the attenuation of Bcl2 was offset by the restoration (Fig. 5G). One important question that still needed to be addressed was whether ZNF831 directly targeted the STAT3 promoter in BRCA cells. To answer this question, we performed ChIP-qPCR assays using anti-FLAG antibodies with five promoter primers in different sequence positions according to the upstream 2000 bp region of NC_000017.11 (42313324..42388502, complement) (NCBI STAT3 sequence) of STAT3 (Fig. 5H). Only one of these primers was enriched in ZNF831 ChIP-qPCR assay and the relative amount of STAT3 promoter was higher in ZNF831-overexpressing cells, suggesting the occupancy of ZNF831 on the precise region of STAT3 promoter (Fig. 5I). Furthermore, to confirm the suppression of STAT3 expression in BRCA cells was directly mediated by ZNF831, we constructed STAT3 promoters into the PGL3 plasmid and detected luciferase activity. We found that ectopic over-expression of ZNF831 decreased the activity of the STAT3 promoter by luciferase assay (Fig. 5J, K). These results were consistent with the negative correlation of ZNF831 and STAT3 expression in transcription and protein levels. All these data provided evidence that ZNF831 directly bound to and suppressed the endogenous STAT3 promoter.

Figure 5.

ZNF831 induces cell apoptosis through directly binding to the specific site of the STAT3 promoter region and transcriptionally suppresses STAT3 expression in BRCA cell lines. (A, B) Evaluation of the transcriptional level of STAT3 in (A) MDA-MB-231 and (B) BT474 cells which were characterized by ectopic ZNF831 expression. (C) Evaluation of the transcriptional level of STAT3 in HCC38 cells which were characterized by ZNF831 knockdown. (D) Evaluation of protein level of STAT3, P-STAT3, Bcl2, and Bax in MDA-MB-231 and BT474 cell lines which were characterized by ectopic ZNF831 expression, and in HCC38 cells which were characterized by ZNF831 knockdown. GAPDH was used as a loading control. (E, F) Flow cytometric analysis of cell apoptosis of ectopic ZNF831 and STAT3 co-overexpression on (E) MDA-MB-231 and (F) BT474 cell lines with Annexin V-APC/7-AAD double staining. Statistical plots were presented. (G) Evaluation of protein level of STAT3, P-STAT3, Bcl2, and Bax in MDA-MB-231 and BT474 cells which were characterized by ectopic ZNF831 and STAT3 co-overexpression. (H) The schematic plot showed the location of the STAT3 primer used and detected in ChIP-qPCR assays. (I) Detection enrichment percentages of the specific promoter region of STAT3 by qPCR followed by ChIP with anti-FLAG antibody. (J, K) The effect of ectopic ZNF831 expression on the transcription of STAT3 as determined by luciferase reporter activity assays. GAPDH was used as a loading control. VC: vector control; NC: negative control; K.D.: knock down; ChIP: chromatin immunoprecipitation. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

ZNF831 enhances chemosensitivity to capecitabine and gemcitabine in BRCA cell lines

To assess the potential relationship between ZNF831 and chemoresistance in BRCA cells, we next investigated whether ZNF831 affects BRCA cells' response to chemotherapy drugs. We found that ectopic over-expression of ZNF831 promoted MDA-MB-231, MCF-7, and BT474 more sensitive to capecitabine (Fig. 6A–C) and gemcitabine (Fig. 6E–G). Knockdown of ZNF831 reduced the sensitivity of HCC38 to capecitabine (Fig. 6D) and gemcitabine (Fig. 6H). In the comparison between capecitabine and gemcitabine, MDA-MB-231, MCF-7, and BT474 showed more sensitivity to gemcitabine after the augmented expression levels of ZNF831. Precisely, the degree of variation of chemosensitivity to capecitabine in MDA-MB-231 (Fig. 6A) is greater than MCF-7 (Fig. 6B) and BT474 (Fig. 6C) after the enhanced expression of ZNF831. Analogously, the degree of variation of chemosensitivity to gemcitabine in BT474 (Fig. 6G) is greater than MDA-MB-231 (Fig. 6E) and MCF-7 (Fig. 6F) after the enhanced expression of ZNF831. Taken together, these data further supported a critical function of ZNF831 in regulating the response of BRCA cells to capecitabine and gemcitabine chemotherapy and suggested that a therapeutic approach targeting ZNF831 may offer a route to overcome capecitabine and gemcitabine resistance.

Figure 6.

ZNF831 enhances chemosensitivity to capecitabine and gemcitabine in BRCA cell lines. (A–C) Proliferation assay was used to determine the chemosensitivity of capecitabine in (A) MDA-MB-231, (B) MCF-7, and (C) BT474 cells which were characterized by ectopic ZNF831 expression, and in (D) HCC38 cells which were characterized by ZNF831 knockdown. (E–G) Proliferation assay was used to determine the chemosensitivity of gemcitabine in (E) MDA-MB-231, (F) MCF-7, and (G) BT474 cells which were characterized by ectopic ZNF831 expression, and in (H) HCC38 cells which were characterized by ZNF831 knockdown. Cape: capecitabine; GEM: gemcitabine. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

ZNF831 suppresses BRCA cell proliferation in vivo

To verify the tumor-suppressive effect of ZNF831 in vivo, stable transfected BT474 cells with ZNF831 over-expression were subcutaneously injected into BALB/c nude mice. The attenuation of growth rate and tumor weight was observed in BT474-ZNF831 cells compared with the vector group (Fig. 7A–D). The volume of tumor mass formed in the over-expressed ZNF831 group was significantly lower than that of the control group (Fig. 7A, B) and the growth rate of the overexpressed ZNF831 group was significantly lower (Fig. 7C). The weight of the tumor in overexpressed ZNF831 group was significantly lower (Fig. 7D). IHC staining results verified the ectopic expression of ZNF831 and showed that over-expression of ZNF831 significantly augmented the levels of TUNEL and mitigated Ki-67 expression (Fig. 7E–H). Altogether, these results indicated that ZNF831 suppressed tumor growth in vivo, which is consistent with our results in vitro.

Figure 7.

ZNF831 suppresses BRCA cell proliferation in vivo. (A, B) The growth of tumors after subcutaneous injection of BRCA cells with ZNF831 over-expression for 30 days in nude mice. (B) Representative images of tumor volumes from xenograft mice described above. (C) The growth of the tumor was observed every 5 days after subcutaneous injection of cells with or without over-expressing ZNF831 gene. (D) After the tumor was removed, the weight of the tumor was measured. The data are shown as the means ± SD (E) Representative images of apoptosis detection in tumor sections from xenograft mice by TUNEL and immunohistochemistry staining for ZNF831 expression and cell proliferation test (Ki-67) in tumor sections from xenograft mice were described above. (F–H) Quantification plots were presented. Scale bar = 100 μm. BRCA, breast cancer. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Discussion

ZNF831, a member of the zinc finger protein family, is a protein coding gene and located in chromosome 20. It has been reported that ZNF831 was correlated with BRCA stem cell phenotype and relapse risk of patients.³⁰ Several studies have indicated that ZNF831 was involved in immune regulation^30,31 and preeclampsia.³² Notwithstanding, the role and mechanism of ZNF831 in tumorigenesis remain unclear. In this study, we demonstrated that ZNF831 was down-regulated due to promoter methylation and repressed the cell proliferation, and promoted apoptosis through transcriptional attenuation of STAT3 expression in BRCA (Fig. 8). These findings not only significantly enriched our understanding of the tumor suppressor function of ZNF831 in BRCA but also indicated a new therapeutic strategy for BRCA treatment for the first time.

Figure 8.

ZNF831 induces BRCA cell apoptosis through STAT3/Bcl2 axis. In mechanism, we proved that ZNF831 could directly bind the promoter region of STAT3 and transcriptionally down-regulated its expression to attenuate transcription of Bcl2 which promoted the apoptosis of BRCA cells. The expression of ZNF831 could be regulated by promoter CpG site methylation. BRCA, breast carcinoma or breast cancer; TF, transcriptional factor.

In our study, the transcriptome data from the Timer database showed ZNF831 was diminished in most cancer types, including BRCA. The expression difference was confirmed by Oncomine and TCGA BRCA database, which suggested ZNF831 might play a vital role in breast carcinogenesis. It has been confirmed many ZNFs were down-regulated or disabled in BRCA as tumor suppressors. The GATA-type zinc finger transcription factor TRPS1 was a context-dependent tumor suppressor and loss of function promoted oncogenesis in lobular breast carcinoma.³³ Restraint of UBR7 enhanced the invasiveness, induced epithelial-to-mesenchymal transition, and promoted metastasis in BRCA.³⁴

As one of the major ways of gene expression regulation, epigenetic modification contained many methods including DNA methylation, histone modification, chromatin remodeling, and RNA regulation.³⁵ DNA methylation down-regulated ZNFs expression was a common way in cancer. ZNF377 was down-regulated by DNA methylation and suppressed tumor growth by altering cellular metabolism and inducing apoptosis and pyroptosis.³⁶ ZNF471 was an important tumor suppressor and was frequently silenced by promoter CpG methylation in esophageal cancer.³⁷ We figured out promoter methylation but not mutation or copy number variation as a promising reason and a potential researching candidate for restraint of ZNF831 in BRCA. Of course, many new-found ways regulated the expression of ZNFs, like lncRNA and miRNA modulation. For example, lnc01060 directly interacted with the transcription factor myeloid zinc finger 1 and enhanced its stability to regulate the progression of glioma.³⁸ MicroRNA-181a modulated gene expression of ZNFs by directly targeting their coding regions.³⁹ Post-transcriptional regulation of gene expression is a hotspot of scientific research in cancer. As N6-methyladenosine (m6A) RNA methylation, YTHDF3 enhanced the translation of m6A-enriched transcripts to promote BRCA brain metastasis.⁴⁰ So, RNA modification and post-transcriptional regulation of ZNF831 would be the potential target for our future study.

Functional studies revealed that ectopic over-expression of ZNF831 decreased the proliferation and colony formation and promoted cell apoptosis in BRCA. Many cancer suppressors played roles in antitumor progression in similar ways. Endothelial miR-30c suppressed tumor growth by promoting fibrin degradation and inhibiting blood vessel formation in the tumor microenvironment.⁴¹ ALOX12-mediated ferroptosis pathway was critical for p53-dependent tumor suppression.⁴² In this way, combining the low expression of ZNF831 in BRCA and anti-tumor functions, ZNF831 is strong enough to be a suppressor in BRCA. But the potential roles of ZNF831 in other aspects are to be explored. Considering ZNF831 was an immune regulator and CD8⁺ T cells regulate tumor ferroptosis during cancer immunotherapy, ZNF831 might play a role in the ferroptosis of BRCA.⁴³ Or like other reported ZNFs, E4F1 promoted DNA double-strand break repair.⁴⁴ ZNF322A promoted stem cell-like properties in lung cancer by acting as a transcription suppressor of c-Myc.⁴⁵ ZDHHC1 exerted significant pyroptosis-promoting effects through metabolic regulation.³⁶ Autophagy, senescence, oxidative stress, and tumor micro-environment were also influenced by ZNFs.^28,46, 47, 48

To figure out the details of the tumor-suppressive mechanism of ZNF831 in BRCA, we explored the ZNF831-related KEGG pathways and biological processes through public databases and transcriptome sequencing in the BRCA cell line. Apoptosis, JAK/STAT pathway, and transcriptional regulation were enriched in many items. Therefore, we supposed that ZNF831 regulated BRCA cell apoptosis through JAK/STAT pathway. STAT3 was a key member of the STAT family and a cytoplasmic transcription factor that regulates cell proliferation, stem cell-like characteristics, differentiation, apoptosis, angiogenesis, inflammation, and immune responses.49, 50, 51, 52, 53, 54 It was a point of convergence for numerous oncogenic signaling pathways and played an important role in the transcription of downstream target genes during oncogenesis.⁵⁵ STAT3 mitigated BRCA cell apoptosis through transcriptional activation of Bcl2 expression.⁵⁶ Importantly, STAT3 was central in regulating the anti-tumor immune response and played a role in inhibiting the expression of crucial immune activation regulators and promoting the production of immunosuppressive factors.⁵⁷ As mentioned before (Fig. 1, 3), ZNF831 was diminished in breast cancer, and enhanced ZNF831 attenuated cell proliferation and colony formation abilities, and augmented apoptosis of cells. Through RNA sequencing and gene expression level detection, we speculated that STAT3 is a potential target after the augmentation of ZNF831. So, we aimed at STAT3 to explore the relationship between ZNF831 with its classic STAT3/Bcl2 apoptotic pathway.

Mechanistically, we figured out that ZNF831 transcriptionally inhibited STAT3 expression by binding to the specific site of the STAT3 promoter region to regulate the oncogenesis of BRCA. Moreover, it has been reported that SD-36, a small-molecule degrader, induced the degradation of STAT3 protein and achieved complete tumor regression.⁵⁸ Resistance to durvalumab and durvalumab plus tremelimumab induced by functional STK11 mutations in non-small-cell lung cancer patients could be reversed by STAT3 knockdown.⁵⁹ Therefore, we could see the importance of the regulation between ZNF831 and STAT3 and there were many unknown functions induced by ZNF831/STAT3 pathway to be explored.

Except for transcriptional regulation, there were many other possible regulation mechanisms to influence STAT3 like methylation, acetylation, ubiquitination, and palmitoylation.⁵³ Sulfhydration of SIRT1 suppressed phosphorylation and acetylation of STAT3.⁶⁰ Protein–protein interaction to induce phosphorylation or acetylation was an important way to regulate protein activity. Especially, STAT3 is activated by the phosphorylation of residues Y705 and S727 to mediate the transcription of downstream genes.⁶¹ Palmitoylation was another way to influence STAT3 by ZNFs, as ZDHHC19 promoted STAT3 activation through S-palmitoylation.⁶² Considering ZNFs were associated with the production of non-coding RNA,⁶³ FLANC, long non-coding RNA in colorectal cancer, up-regulated and prolonged the half-life of phosphorylated STAT3.⁶⁴ Furthermore, according to the self-regulation of STAT3 through transcriptional regulation by binding to the promoter region of STAT3 gene, we raise a presumption that ZNF831 interacts with STAT3 to form a transcription–initiation complex to inhibit transcription of STAT3.⁶⁵ Of course, the protein–protein interaction analysis and co-immunoprecipitation assay are needed to support the corresponding presumption.

Chemotherapeutic agents, such as capecitabine and gemcitabine, hold the backbone of treatment for many solid tumors including BRCA and hepatocellular carcinoma, but the development of resistance is a major stumbling block of treatment.⁶⁶ In clinics, BRCA patients are frequently resistant to chemotherapy, especially advanced ones. Figuring out the mechanism of chemoresistance and identifying the biomarker of resistance will be the top priority of modern medicine for overcoming this limitation and for the development of effective therapy. Gemcitabine and capecitabine were the cornerstone of cancer therapy for the improvement of overall survival.^67,68 Importantly, we found out that the chemical sensitivity of capecitabine and gemcitabine could be ameliorated by enhanced expression of ZNF831. This provided us with a new insight into chemical therapy in BRCA patients. But the mechanism of the chemical sensitivity amelioration is yet to be explored. It has been reported that JAK/STAT3 pathway and its downstream fatty acid β-oxidation are critical for BRCA stem cell chemoresistance.⁶⁹ Circadian melatonin disruption mediated STAT3-driven paclitaxel resistance in breast cancer.⁷⁰ So, there would be a potential mechanism of chemoresistance that stayed behind the ZNF831/STAT3 pathway. What's more, the degree of variation of chemosensitivity to capecitabine in MDA-MB-231 and the degree of variation of chemosensitivity to gemcitabine in BT474 were respectively greater in three BRCA cell lines. It showed us that ZNF831-induced chemosensitivity enhancement was tumor type-specific and the mechanisms behind played in diversity.

In summary, we first showed that ZNF831 was a BRCA suppressor that prevented BRCA cell proliferation via transcriptionally diminishing the expression of STAT3 to promote cell apoptosis. The multitude of the biological function of ZNF831 is beyond what we have studied. Further studies are warranted to investigate the scope of mechanistic insights into the role of ZNF831 in other malignancies. Precise ZNF831-STAT3 promoter binding site sequence and mutation verification test could provide valuable information for target therapy of the ZNF831/STAT3 pathway. The mechanism of ameliorated chemosensitivity in an enhanced level of ZNF831 expression in BRCA cells supports the clinical development of novel therapies targeting the ZNF831 for the treatment of human BRCA. Notably, data from public databases and transcriptome sequencing provided us with a new clue that ZNF831 is probably a potentially key role in tumor immunity. So, research about BRCA immunity has been placed on the agenda.

Ethical statement

The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was approved by the Ethics Committee of Daping Hospital and the Laboratory Animal Welfare and Ethics Committee of the Third Military Medical University.

Author contributions

Concept and design: Jun Fan, Zhe Zhang, Wenbin Liu and Yan Xu; Administrative support: Wenbin Liu and Yan Xu; Provision of study materials or patients: Zhe Zhang, Dongjiao Chen, Hongqiang Chen, Wenbo Yuan, Jingzhi Li, Yong Zeng, Shimeng Zhou, Shu Zhang, Gang Zhang, Jiashen Xiong, Lu Zhou, and Jing Xu; Data acquisition, analysis, and statistical analysis: Jun Fan and Zhe Zhang; Experimental studies: Jun Fan; Manuscript preparation, editing, review, and literature search: all authors; Final approval of manuscript: all authors.

Conflict of interests

The authors have no conflict of interests to declare.

Funding

This work was supported by the National Natural Science Foundation of China (No. 81472482, 82173556), the Clinical Technology Innovation and Cultivation Project of the Army Medical University of China (No. CX2019LC120), and the National Key Clinical Specialty Military Construction Project (China, No. 425Z8).

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Figure S1.

ZNF831 expresses low in BRCA. (A, B) Representative images of ZNF831 expression in (A) normal and (B) BRCA tissues by IHC using the Human Protein Atlas database. Scale bar = 100 μm. BRCA, breast carcinoma or breast cancer.

Figure S2.

ZNF831 expression was regulated by promoter CpG methylation. (A) The cg25325636 site methylation value across all samples between normal and tumor. (B) The cg16473762 site methylation value across all samples between normal and tumor. (C) The cg09642670 site methylation value across all samples between normal and tumor. (D) The aggregation of cg25325636, cg16473762, and cg09642670 site methylation values across all samples between normal and tumor. BRCA, breast carcinoma or breast cancer. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

Figure S3.

ZNF831 suppresses BRCA cell apoptosis. (A–C) Flow cytometric analysis of cell apoptosis of ZNF831 over-expression on (A) MDA-MB-231, (B) MCF-7, and (C) BT474 cells with Annexin V-APC/7-AAD double staining. Representative images were presented. (D) Flow cytometric analysis of cell apoptosis of ZNF831 knockdown on HCC38 cells with Annexin V-APC/7-AAD double staining. Representative images were presented. Scale bar = 100 μm. BRCA, breast carcinoma or breast cancer.

Figure S4.

ZNF831 co-expressive genes for probable mechanism in BRCA. (A) The volcano plot contains all of the ZNF831-associated and statistically significant genes from LinkedOmics. (B) Pathway enrichment classification from transcriptome sequence in ectopic ZNF831 expression MDA-MB-231 cell. BRCA, breast carcinoma or breast cancer.

Figure S5.

STAT3 was the key to apoptotic induction of ZNF831 in BRCA. (A, B) Flow cytometric analysis of cell apoptosis of ZNF831 and STAT3 co-overexpression on (A) MDA-MB-231 and (B) BT474 cells with Annexin V-APC/7-AAD double staining. Representative images were presented. Scale bar = 200 μm. BRCA, breast carcinoma or breast cancer.