Estrogen inhibits hepatocellular carcinoma progression dependent on HOXA11-AS/HOXA11.

Pincheng Zhou: Writing – original draft, Validation, Investigation, Conceptualization. Fengze Sun: Investigation. Peixu Lin: Methodology, Investigation. Yan Yan: Project administration. Jiayao Liu: Validation, Investigation. Yang Zhou: Validation, Software. Ting He: Validation, Formal analysis, Data curation. Pengcheng Liu: Visualization, Formal analysis, Data curation. Jie Wang: Visualization, Validation. Huanhuan Sun: Writing – original draft, Methodology, Investigation, Funding acquisition. Haiqing Ma: Writing – review & editing, Writing – original draft, Investigation, Funding acquisition, Conceptualization.

According to the ethical guidelines of the Helsinki Declaration, an experimental plan was established and approved by the Human Ethics Committee of Guangdong Provincial People’s Hospital. Obtain written informed consent from an individual or guardian participant.

Not applicable.

The work was supported by The National Natural Science Foundation of China (82072719, 82473318), the Natural Science Foundation of the Guangdong Province (2021A1515010790, 2023A1515012872) and High-level Hospital Construction Research Project of Heyuan People’s Hospital (YNKT202203).

HCCLM3 and SNU-182 cells were purchased from the American Type Culture Collection (ATCC). LO2 cells were obtained from the Shanghai Institute of Cell Biology (SICB). The cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco, NY, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, NY, USA) in a humidified atmosphere at 37°C with 5% CO 2 . Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, USA). Reverse transcription was performed using the RevertAid First Strand cDNA Synthesis Kit K1622 (Thermo Fisher, MA, USA). The SYBR Green Mix was purchased from Vazyme (Nanjing, China). The primer sequences are listed in Table S1. Western blot analysis was performed as previously described [ 41 ]. The following antibodies were used: Anti-HOXA11 antibody (Abcam, Boston, USA), anti-ESR1 antibody (Proteintech, Wuhan, China) , anti-ESR2 antibody (Proteintech, Wuhan, China), anti-c-MET antibody (Cell Signaling Technology, Boston, USA), anti-phospho-c-MET antibody (Cell Signaling Technology, Boston, USA), anti-cleaved caspase-3 antibody (Cell Signaling Technology, Boston, USA), anti-BAX antibody (Proteintech, Wuhan, China), anti-BCL-II antibody (Proteintech, Wuhan, China), anti-AKT antibody (Abclonal, Wuhan, China), anti-phospho-AKT antibody (Abclonal, Wuhan, China), anti-mTOR antibody (Abclonal, Wuhan, China), anti-phospho-mTOR antibody (Abclonal, Wuhan, China), and anti-GAPDH antibody (Cell Signaling Technology, Boston, USA) as a loading control. Protein samples were separated by SDS-PAGE, transferred to PVDF membranes, and incubated with the respective primary antibodies. After washing, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies. The protein bands were visualized using enhanced chemiluminescence (ECL) reagents. RNA from HCCLM3 and SNU-182 cells was digested with Ribonuclease A+T cocktail (Ambion) (Thermo Fisher, USA) to degrade single-stranded RNA, but not RNA duplexes. Samples were incubated at 37°C for 30 minutes, then treated with proteinase K, and finally, RNA was extracted for cDNA synthesis and RT-qPCR as described above. The primer sequences for the 63 RPA assays are listed in Table S1. HEK293T cells were cultured in DMEM supplemented with 10% fetal bovine serum. Three luciferase reporter constructs were generated based on the PL01 vector (GeneCoqole), each containing 2000 bp, 1500 bp, and 1000 bp upstream of the c-MET gene. These luciferase reporter constructs were co-transfected with pEZ-Lv207-HOXA11 into HEK293T cells according to the Lipofectamine 3000 (Thermo Fisher, USA) protocol. Relative luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega). Four-week-old Balb/c-nu/nu mice were purchased from Charles River (Beijing, China) and randomly divided into three groups, with five mice in each group. A total of 2 × 10^6 HOXA11-AS sh1 or sh2 HCCLM3 cells or shNC (scrambled shRNA) cells suspended in 0.15 ml PBS were subcutaneously injected into the right axillary region of the mice. Tumor size was measured every three days starting from the seventh day post-injection. On the 21st day after tumor formation, the mice were sacrificed, and the tumors were excised and measured. All animal experiments were approved by the Animal Ethics Committee of Guangdong Provincial People’s Hospital. The CCK8 reagent was purchased from MCE (China). Cells were seeded at 2000 cells/100 µl into six-well plates and incubated overnight. Subsequently, 10 µl of CCK8 reagent was added to each well. Absorbance was measured at 450 nm at 2 hours, 12 hours, 24 hours, 36 hours, and 72 hours. Statistical analyses were performed using SPSS 20.0 (IBM, USA). Kaplan-Meier curves were generated by GraphPad Prism 8. A p-value of <0.05 was considered statistically significant. Data are presented as mean ± standard deviation. Differences between two independent groups were analyzed using the t -test. Differences among multiple independent groups were analyzed using One-way ANOVA with multiple comparisons. The χ² test was used to analyze the relationship between HOXA11 expression and clinical characteristics. Transfection was performed using Lipo3000 (Invitrogen) according to the manufacturer’s protocol. For each transfection, 7.5 µl of Lipo3000 was used, with a final siRNA concentration of 50 nM, in 2 ml of culture medium in a six-well plate. Stable cell lines were established using 200 µg/ml of puromycin.

In hepatocellular carcinoma (HCC) patients, the incidence in females is significantly lower than in males, with a ratio of approximately 1:3, indicating a notable gender difference. The aim is to investigate the underlying reasons for this clinical phenomenon. By analyzing LIHC-related data from TCGA database, we found that several estrogen receptors are significantly downregulated in HCC tissues ( Fig. 1 A) and are associated with poor prognosis ( Fig. 1 B). We collected tissues from 12 patients with hepatocellular carcinoma, and we used immunoblotting to detect oestrogen receptor ESR1 and ESR2 expression. We found that ESR1 expression wasreduced in tumor tissues of most patients, while ESR2 did not change significantly ( Fig. 1 C). To determine the appropriate concentration of oestrogen to use, we used the CCK8 assay to measure the viability of HCCLM3 cells after treatment with different concentrations of estrogen. CCK8 shows that estrogen can reduce cell viability, with a dose-dependent effect ( Fig. 1 D). We use 0.01 μmol/ml as the concentration for subsequent experiments, because higher concentrations cause excessive cell death, which may lead to false-positive results. Using flow cytometry and Western blotting to assess apoptosis-related markers, we found that estrogen can promote apoptosis in HCC cells HCCLM3 and SNU182 ( Fig. 1 E, F). The oncogenic role of HOXA11-AS has been reported in various tumors, with Huiping Zhang et al. indicating that the inhibition of HOXA11-AS is associated with cell apoptosis. Fig. 1 Estrogen can promote apoptosis in liver cancer. A. TCGA-LIHC database analysis shows that ESR1 and ESRRG are reduced in hepatocellular carcinoma tissues. ns, not significant; * p  ⩽ 0.05, ** p  ⩽0.01, *** p ⩽0.001; B. TCGA data indicates that hepatocellular carcinoma patients with high expression of ESR1 have a better prognosis. C. Immunoblotting shows that ESR1 was reduced in most hepatocellular carcinoma patients in tumor tissues, while ESR2 remained unchanged. D. CCK8 shows that estrogen can reduce cell viability, with a dose-dependent effect. E. Flow cytometry shows that estrogen can gradually increase the apoptosis rate of HCCLM3, with a dose-dependent effect. Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p  ⩽ 0.05, ** p  ⩽ 0.01, *** p  ⩽ 0.001, **** p  ⩽ 0.0001. F. Immunoblotting shows that estrogen can gradually increase the apoptosis rate of HCCLM3/SNU-182, with a dose-dependent effect. Fig 1 Estrogen can promote apoptosis in liver cancer. A. TCGA-LIHC database analysis shows that ESR1 and ESRRG are reduced in hepatocellular carcinoma tissues. ns, not significant; * p  ⩽ 0.05, ** p  ⩽0.01, *** p ⩽0.001; B. TCGA data indicates that hepatocellular carcinoma patients with high expression of ESR1 have a better prognosis. C. Immunoblotting shows that ESR1 was reduced in most hepatocellular carcinoma patients in tumor tissues, while ESR2 remained unchanged. D. CCK8 shows that estrogen can reduce cell viability, with a dose-dependent effect. E. Flow cytometry shows that estrogen can gradually increase the apoptosis rate of HCCLM3, with a dose-dependent effect. Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p  ⩽ 0.05, ** p  ⩽ 0.01, *** p  ⩽ 0.001, **** p  ⩽ 0.0001. F. Immunoblotting shows that estrogen can gradually increase the apoptosis rate of HCCLM3/SNU-182, with a dose-dependent effect. By analyzing data from HCC patients in the TCGA database, we found that the long non-coding RNA HOXA11-AS and the transcription factor HOXA11 are elevated in HCC patients and are associated with poor prognosis ( Fig. 2 A, B). Additionally, HOXA11-AS and HOXA11 exhibit co-expression across various cancer types (Fig. S1). To select appropriate HCC cell lines for HOXA11 knockdown experiments, we assessed the expression levels of HOXA11-AS and HOXA11 in several HCC cell lines using qPCR and Western blotting. Among these, HCCLM3 and SNU-182 cell lines exhibited the highest expression levels of HOXA11-AS and HOXA11, making them suitable candidates for specific knockdown of HOXA11-AS or HOXA11 ( Fig. 2 C, D). Fig. 2 Estrogen can promote apoptosis in liver cancer cells dependent on HOXA11-AS. A. According to TCGA data, the expression of HOXA11-AS is higher in hepatocellular carcinoma (HCC) tissues than in non-tumor tissues and is associated with poor prognosis. B. According to TCGA data, the expression of HOXA11 is higher in HCC tissues than in non-tumor tissues and is associated with poor prognosis. C. The differential expression of HOXA11-AS in HCCLM3, SNU-182, SNU423, Bel-7402, Huh7, HepG2, MHCC97H, and LO2 cell lines was detected using qPCR. D. The differential expression of HOXA11 in HCCLM3 and other liver cancer cell lines, as well as the normal liver cell line LO2, was detected using immunoblotting. E. Flow cytometry shows that knocking down HOXA11-AS reduces the sensitivity of HCCLM3 to estrogen-induced apoptosis. (C,E) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p  ⩽ 0.05, ** p  ⩽ 0.01, *** p  ⩽ 0.001, **** p  ⩽ 0.0001. F. Immunoblotting shows that knocking down HOXA11-AS reduces the sensitivity of HCCLM3/SNU-182 to estrogen-induced apoptosis. G. Immunoblotting shows that overexpression of HOXA11-AS can reverse the estrogen-induced apoptosis of HCCLM3/SNU-182. Fig 2 Estrogen can promote apoptosis in liver cancer cells dependent on HOXA11-AS. A. According to TCGA data, the expression of HOXA11-AS is higher in hepatocellular carcinoma (HCC) tissues than in non-tumor tissues and is associated with poor prognosis. B. According to TCGA data, the expression of HOXA11 is higher in HCC tissues than in non-tumor tissues and is associated with poor prognosis. C. The differential expression of HOXA11-AS in HCCLM3, SNU-182, SNU423, Bel-7402, Huh7, HepG2, MHCC97H, and LO2 cell lines was detected using qPCR. D. The differential expression of HOXA11 in HCCLM3 and other liver cancer cell lines, as well as the normal liver cell line LO2, was detected using immunoblotting. E. Flow cytometry shows that knocking down HOXA11-AS reduces the sensitivity of HCCLM3 to estrogen-induced apoptosis. (C,E) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p  ⩽ 0.05, ** p  ⩽ 0.01, *** p  ⩽ 0.001, **** p  ⩽ 0.0001. F. Immunoblotting shows that knocking down HOXA11-AS reduces the sensitivity of HCCLM3/SNU-182 to estrogen-induced apoptosis. G. Immunoblotting shows that overexpression of HOXA11-AS can reverse the estrogen-induced apoptosis of HCCLM3/SNU-182. Endogenous estrogen and HOXA11 work together in the conversion of vaginal epithelium to thick glycogenized squamous epithelium [ 19 ], which can be maintained stable by HOXA11-AS. Therefore, we aimed to verify whether estrogen-induced apoptosis in HCC cells is dependent on HOXA11-AS. Uponknocking down HOXA11-AS, we found that the pro-apoptotic effect of estrogen in HCC cells was significantly inhibited ( Fig. 2 E, F). Similarly, in cells overexpression HOXA11-AS, estrogen could reverse the anti-apoptotic effect of HOXA11-AS ( Fig. 2 G). Using lentiviral infection with sh-HOXA11-AS in HCCLM3 and SNU-182 liver cancer cell lines, we found that knocking down HOXA11-AS resulted in decreased expression of HOXA11 at both RNA and protein levels, as determined by qPCR and Western blotting ( Fig. 3 A, B). Similarly, after infecting HCCLM3 and SNU-182 cell lines with OE-HOXA11 lentivirus, qPCR showed that overexpression of HOXA11 also increased the expression of HOXA11-AS ( Fig. 3 C). There is an overlapping region between HOXA11-AS and HOXA11 ( Fig. 3 D). To explore the functional mechanism between HOXA11-AS and HOXA11, we conducted an RNase protection assay (RPA) on RNA from HCCLM3 cells to detect the possibility of RNA duplex formation. RT-qPCR results showed that the overlapping region of the two transcripts was protected and not easily degraded, indicating that HOXA11-AS can form an RNA duplex with HOXA11 to prevent its degradation ( Fig. 3 E). Fig. 3 HOXA11-AS forms an RNA duplex with HOXA11 to maintain their mutual expression. A. qPCR analysis showed that knocking down HOXA11-AS in HCCLM3 and SNU182 resulted in a decrease in HOXA11 expression. B. Immunoblotting analysis showed that knocking down HOXA11-AS in HCCLM3 and SNU182 resulted in a decrease in HOXA11 expression. C. qPCR analysis showed that overexpression of HOXA11 in HCCLM3 and SNU182 resulted in an increase in HOXA11-AS expression. D. There is an overlapping region between HOXA11-AS and HOXA11. E. The overlapping region between HOXA11-AS and HOXA11 transcripts is not protected by RNase A+T and is degraded. (A-C, E) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p ⩽0.05, ** p ⩽0.01, *** p ⩽0.001, **** p ⩽0.0001. Fig 3 HOXA11-AS forms an RNA duplex with HOXA11 to maintain their mutual expression. A. qPCR analysis showed that knocking down HOXA11-AS in HCCLM3 and SNU182 resulted in a decrease in HOXA11 expression. B. Immunoblotting analysis showed that knocking down HOXA11-AS in HCCLM3 and SNU182 resulted in a decrease in HOXA11 expression. C. qPCR analysis showed that overexpression of HOXA11 in HCCLM3 and SNU182 resulted in an increase in HOXA11-AS expression. D. There is an overlapping region between HOXA11-AS and HOXA11. E. The overlapping region between HOXA11-AS and HOXA11 transcripts is not protected by RNase A+T and is degraded. (A-C, E) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p ⩽0.05, ** p ⩽0.01, *** p ⩽0.001, **** p ⩽0.0001. Knocking down HOXA11-AS significantly inhibited the proliferation of HCCLM3 cells starting from day 3, with similar results observed in SNU-182 cells ( Fig. 4 A). Similarly, wound healing assays showed that the migration abilities of HCCLM3 and SNU-182 cells were significantly suppressed after HOXA11-AS knockdown ( Fig. 4 B). Additionally, the apoptosis-related protein cleaved caspase-3 increased ( Fig. 4 C). Flow cytometry analysis showed that the apoptosis rate significantly increased in HCCLM3 cells after HOXA11-AS knockdown ( Fig. 4 D). To evaluate the effect of HOXA11-AS knockdown on HCC in vivo, we established a tumor xenograft model by inoculating sh-HOXA11-AS cells and control cells into the bilateral axillae of 4-week-old male BALB/c nude mice ( Fig. 4 E). There was no significant difference in body weight between the different groups of mice during tumor development ( Fig. 4 F). Compared to the control group, tumors formed by HOXA11-AS knockdown HCC cells grew more slowly, were significantly smaller in volume, and lighter in weight ( Fig. 4 G-I). Fig. 4 Reducing HOXA11-AS and HOXA11 can inhibit hepatocellular carcinoma (HCC). A. CCK-8 assay showed that knocking down HOXA11-AS can reduce the proliferation ability of HCCLM3/SNU-182 cells. B. Scratch assay demonstrated that knocking down HOXA11-AS can reduce the migration ability of HCCLM3/SNU-182 cells. C. Immunoblotting analysis showed that knocking down HOXA11-AS can promote the apoptosis of HCCLM3/SNU-182 cells. D. Flow cytometry analysis showed that knocking down HOXA11-AS can promote the apoptosis of HCCLM3 cells. E. Subcutaneous tumor formation diagram of HCCLM3 cells. F. Statistical chart of changes in mouse body weight. G. Statistical chart of subcutaneous tumor growth curve in mice. H. Tumor growth is slower in the HOXA11-AS knockout group compared to the negative control group. I. Statistical chart of subcutaneous tumor weight. (A-D, G, I) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p ⩽0.05, ** p ⩽0.01, *** p ⩽0.001, **** p ⩽0.0001. Fig 4 Reducing HOXA11-AS and HOXA11 can inhibit hepatocellular carcinoma (HCC). A. CCK-8 assay showed that knocking down HOXA11-AS can reduce the proliferation ability of HCCLM3/SNU-182 cells. B. Scratch assay demonstrated that knocking down HOXA11-AS can reduce the migration ability of HCCLM3/SNU-182 cells. C. Immunoblotting analysis showed that knocking down HOXA11-AS can promote the apoptosis of HCCLM3/SNU-182 cells. D. Flow cytometry analysis showed that knocking down HOXA11-AS can promote the apoptosis of HCCLM3 cells. E. Subcutaneous tumor formation diagram of HCCLM3 cells. F. Statistical chart of changes in mouse body weight. G. Statistical chart of subcutaneous tumor growth curve in mice. H. Tumor growth is slower in the HOXA11-AS knockout group compared to the negative control group. I. Statistical chart of subcutaneous tumor weight. (A-D, G, I) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p ⩽0.05, ** p ⩽0.01, *** p ⩽0.001, **** p ⩽0.0001. To further validate the clinical relevance of HOXA11 with HCC prognosis, we conducted immunohistochemical (IHC) analysis of HOXA11 protein levels in paired paraffin-embedded HCC tissues and adjacent non-tumor tissues from 87 HCC patients treated at the Fifth Affiliated Hospital of Sun Yat-sen University between May 11, 2004, and April 15, 2015. The results showed that HOXA11 protein levels were upregulated in HCC tissues compared to adjacent non-tumor tissues ( Fig. 5 A). Fig. 5 Increased expression of HOXA11 is associated with poor prognosis in HCC. A. Immunohistochemistry (IHC) image of HOXA11 high expression in patients. B. High expression of HOXA11 is associated with poor overall survival (OS) in patients. C. High expression of HOXA11 is associated with poor disease-free survival (DFS) in patients. D. Kaplan-Meier analysis for <5cm (red) and ⩾5cm (blue) with size. E. Kaplan-Meier analysis for High differentiation (red), middle differentiation (blue) and low differentiation (grey)with size. F. Kaplan-Meier analysis for no recurrence (red) and having recurrence (blue) with recurrence. G. Kaplan-Meier analysis for no vascular invasion (red) and having vascular invasion (blue) with vascular invasion. Fig 5 Increased expression of HOXA11 is associated with poor prognosis in HCC. A. Immunohistochemistry (IHC) image of HOXA11 high expression in patients. B. High expression of HOXA11 is associated with poor overall survival (OS) in patients. C. High expression of HOXA11 is associated with poor disease-free survival (DFS) in patients. D. Kaplan-Meier analysis for <5cm (red) and ⩾5cm (blue) with size. E. Kaplan-Meier analysis for High differentiation (red), middle differentiation (blue) and low differentiation (grey)with size. F. Kaplan-Meier analysis for no recurrence (red) and having recurrence (blue) with recurrence. G. Kaplan-Meier analysis for no vascular invasion (red) and having vascular invasion (blue) with vascular invasion. The Patients, aged 26 to 79 years, included 12 females (14.8%) and 75 males (86.2%), High HOXA11 expression was also associated with poorer overall survival (OS) and disease-free survival (DFS) ( p =0.001, p =0.002) ( Fig. 5 B, C). Univariate analysis identified gender, tumor size, tumor stage, recurrence, vascular invasion and HOXA11 expression as risk factors. We used Kaplan-Meier analysis to verify the above risk factors ( Fig. 5 D-G). All the above factors were statistically significant ( p <0.05). Multivariate Cox regression analysis was subsequently performed to determine whether the identified risk factors in the univariate log-rank test were independent predictors. The result showed that High HOXA11 expression (HR, 2.981; 95% CI, 1.392-6.384), tumor size⩾5cm (HR, 2.390; 95% CI, 1.084-5.271), low differentiation (HR, 3.203; 95% CI, 1.123-5.137), recurrence (HR, 9.228; 95% CI, 3.563-23.904) and vascular invasion (HR, 4.981; 95% CI, 1.952-12.711) were significant independent predictors of survival ( Table 1 ). AFP level, age, gender, liver cirrhosis and tumor number were no significant (Fig. S2). Table 1 Univariate and multivariate analysis of factors associated with clinical outcomes (N=87). Table 1 Variable Univariate analysis Multivariate analysis Median survival time Log-rank χ2 test p Hazard Ratio (95% CI) p HOXA11 expression 9.78 0.001 Low expression 79 (65,93) Reference High expression 17 (11,23) 2.981 (1.392-6.384) 0.005 Age 0.105 0.796 < 50 years 62 (25,99) ≥ 50 years 51 (30,72) Gender 0.48 0.65 Female 61 (50,72) Male 75 (47,104) Size (cm) 8.294 0.007 <5cm 80 (65,97) Reference ≥5cm 25 (9,41) 2.390 (1.084-5.271) 0.031 Stage 6.01 0.03 High differentiation 73 (42,104) Reference Middle differentiation 52 (21,83) 0.523 (0.197-1.389) 0.193 Low differentiation 13 (4,22) 3.203 (1.123∼5.137) 1 41 (22,59) Recurrence 22.61 <0.001 No 100 (86,115) Reference Yes 25 (4,46) 9.228 (3.563-23.904) <0.001 Liver cirrhosis 2.655 0.316 No 53 (40,66) Yes 78 (61,94) AFP 1.672 0.145 < 400 62 (39,85) ≥ 400 39 (11,67) Vascular invasion 9.506 0.002 No 69 (46,92) Reference Yes 13 (7,19) 4.981 (1.952-12.711) 0.001 Univariate and multivariate analysis of factors associated with clinical outcomes (N=87). c-MET, an important tyrosine kinase receptor, has been reported to be abnormally activated through c-MET-related pathways in various tumors. To investigate whether HOXA11 can promote c-MET transcription, we overexpressed HOXA11 in HCCLM3 and SNU-182 HCC cell lines. Compared to the control group, the expression of c-MET was significantly increased ( Fig. 6 A). We predicted several potential binding sites in the 2000 bp region upstream of c-MET ( Fig. 6 B). To further confirm the specific binding region of HOXA11 to c-MET, we conducted a dual-luciferase reporter assay and found that the binding occurred approximately at the 2000 bp region ( Fig. 6 C). As previously mentioned, HOXA11-AS can form a duplex with HOXA11, preventing its degradation; thus, overexpression of HOXA11-AS indirectly leads to an increase in c-MET transcription. c-MET phosphorylation is a crucial regulatory mechanism in the c-MET/AKT/mTOR pathway. We knocked down HOXA11-AS in HCCLM3 and SNU-182 cells and treated them with estrogen. Using Western blotting to detect the expression of proteins in the MET/AKT/mTOR pathway, we observed a significant increase in the phosphorylation levels of the related proteins in this pathway ( Fig. 6 D). Fig. 6 HOXA11 promotes the transcription of c-MET, and estrogen can reverse this process. A. qPCR analysis showed that overexpression of HOXA11 in HCCLM3 or SNU-182 cells resulted in increased expression of c-Met. B. Possible binding sites of HOXA11 that we predicted in the 2000bp upstream region of c-met. C. Predicted binding sites of HOXA11 are located in the upstream region of c-Met within 2000bp. (A, C) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p ⩽0.05, ** p ⩽0.01, *** p ⩽0.001, **** p ⩽0.0001. D. Immunoblotting showed that estrogen acts through the c-Met-AKT-mTOR pathway. Fig 6 HOXA11 promotes the transcription of c-MET, and estrogen can reverse this process. A. qPCR analysis showed that overexpression of HOXA11 in HCCLM3 or SNU-182 cells resulted in increased expression of c-Met. B. Possible binding sites of HOXA11 that we predicted in the 2000bp upstream region of c-met. C. Predicted binding sites of HOXA11 are located in the upstream region of c-Met within 2000bp. (A, C) Results are shown as a mean ± SD of at least 3 independent experiments performed in duplicate. * p ⩽0.05, ** p ⩽0.01, *** p ⩽0.001, **** p ⩽0.0001. D. Immunoblotting showed that estrogen acts through the c-Met-AKT-mTOR pathway.

Estrogen can activate the c-MET/AKT/mTOR pathway in a HOXA11-AS-dependent manner, promoting apoptosis of tumor cells and thereby inhibiting the proliferation of HCC. HOXA11-AS and HOXA11 are concordantly upregulated in HCC through the formation of an RNA duplex. HOXA11-AS can promote the progression of HCC and is associated with poor prognosis in patients.

In this study, we found that Estrogen can promote apoptosis in HCC cells by activating the c-MET/AKT/mTOR pathway in a HOXA11-AS-dependent manner, thereby suppressing the development of hepatocellular carcinoma (HCC). With the advancement of molecular biology techniques, some risk factors and molecular mechanisms of liver cancer have been elucidated, and several targeted therapies are being applied [ [20] , [21] , [22] , [23] , [24] , [25] , [26] , [27] , [28] , [29] , [30] ]. However, the understanding of HCC is still incomplete, and the treatment outcomes remain unsatisfactory [ 28 , 31 , 32 ]. The incidence and mortality rates of HCC in males are 2-3 times higher than in females, suggesting that sex-related hormones may play a significant role in HCC. Estrogen and progesterone are involved in the expression of HOXA11 in the uterine endometrium during human implantation and in the development of the female reproductive tract in human fetuses [ 19 , 33 ]. In this study, we found that estrogen significantly inhibits HCC cell proliferation and the c-MET/AKT/mTOR pathway by suppressing HOXA11. To deepen our understanding of HCC, we analyzed data on lncRNAs, miRNAs, and protein-coding RNAs from HCC patients in the TCGA database. We found that the transcription factor HOXA11 and its antisense lncRNA HOXA11-AS are upregulated in HCC patients and associated with poor prognosis. These findings from the TCGA were further validated in primary HCC and adjacent non-tumor tissues. In addition, HOXA11 and HOXA11-AS are also highly expressed in the HCCLM3 and SNU-182 cell lines. Recent studies have shown that abnormal expression of HOX genes is associated with various cancers, such as leukemia, prostate cancer, and cervical cancer [ [34] , [35] , [36] ]. It has been reported that HOXA11 may be involved in gene fusion, thereby promoting the development of leukemia [ 37 ]. However, the role of HOXA11 in other types of tumors is less studied and remains controversial. For instance, HOXA11 is methylated in gastric cancer andlung cancer, suggested it might act as a tumor suppressor gene [ 38 , 39 ]. However, according to TCGA data, we found that HOXA11 is upregulated in gastric cancer and HOXA11 is overexpressed in lung adenocarcinoma and associated with a higher HR in TCGA [ 40 ]. Interestingly, we found that in most cancer types in TCGA database, the methylation level of HOXA11 is higher in tumor tissues than in normal tissues, yet HOXA11 is still significantly elevated in these tumors. We believe that the expression level of HOXA11 cannot be solely determined by its methylation level. In our study, IHC results from 87 HCC patients indicated that high HOXA11 expression was significantly associated with poor prognosis in both OS and DFS. The relationship between estrogen and hepatocellular carcinoma is poorly reported. We obtained in vitro and in vitro experimental results to demonstrate the inhibitory effect of estrogen in hepatocellular carcinoma. And weakened outcomes have been observed with estrogen treatment following the knockdown of HOXA11/HOXA11-AS. Knockdown of HOXA11-AS reversible estrogen-induced c-MET/AKT/mTOR pathway activation in hepatocellular carcinoma cells may be the underlying mechanism. However, the mechanism of estrogen inhibition of hepatocellular carcinoma dependent on HOXA11-AS/HOXA11 needs to be further explored in vivo experiments. In conclusion, we elucidated the mechanism by which estrogen promotes apoptosis in HCC, which could lead to the discovery of a new potential treatment.

The long-term therapeutic efficacy of hepatocellular carcinoma (HCC) remains unsatisfactory, especially for patients with advanced unresectable disease [ 1 , 2 ]. The impact of sex hormones on the development and progression of various types of tumors has been reported extensively, particularly in cancers with a significant gender-related incidence. For instance, androgens are implicated in prostate cancer [ 3 , 4 ] and liver cancer [ [5] , [6] , [7] , [8] ], while estrogens play roles in breast cancer [ 9 , 10 ] and ovarian cancer [ 11 ]. Liver cancer is a disease with a significantly higher incidence in males compared to females. Numerous studies have reported on the role of androgens and androgen receptors in the development and progression of liver cancer [ [5] , [6] , [7] , [8] ]. However, the potential inhibitory effects of estrogens and their receptors on liver cancer, as well as the underlying mechanisms, remain unclear. Homeobox A11 (HOXA11) belongs to the homeobox gene family, which encodes a highly conserved family of transcription factors that play critical roles in embryonic implantation, evolution, and morphogenesis in all multicellular organisms [ [12] , [13] , [14] ]. Dysfunction of HOXA11 can lead to abnormal embryonic development, endometriosis, infertility, and even cancer, such as hematological malignancies and endometrial adenocarcinoma [ [14] , [15] , [16] , [17] , [18] ]. Endogenous estrogen and HOXA11 work together in the conversion of vaginal epithelium to thick glycogenized squamous epithelium [ 19 ]. Analysis of data from The Cancer Genome Atlas (TCGA) on HCC patients revealed that both HOXA11 and its antisense RNA, HOXA11-AS, are consistently upregulated in HCC. Furthermore, HCC patients with higher expression levels of HOXA11 and HOXA11-AS have poorer prognoses. We discovered that estrogen can promote apoptosis in liver cancer cells through HOXA11-AS/HOXA11 dependency, thereby inhibiting the progression of liver cancer.

The authors declare that there are no conflicts of interest.