Downregulation of RFX1 predicts poor prognosis of patients with small hepatocellular carcinoma

Accepted Manuscript Downregulation of RFX1 predicts poor prognosis of patients with small hepatocellular carcinoma Yingjun Liu, Peng Jiang, Gangcheng Wang, Xiaonyong Liu, Suxia Luo PII:

S0748-7983(18)31023-0

DOI:

10.1016/j.ejso.2018.04.017

Reference:

YEJSO 4945

To appear in:

European Journal of Surgical Oncology

Received Date: 11 December 2017 Revised Date:

8 April 2018

Accepted Date: 18 April 2018

Please cite this article as: Liu Y, Jiang P, Wang G, Liu X, Luo S, Downregulation of RFX1 predicts poor prognosis of patients with small hepatocellular carcinoma, European Journal of Surgical Oncology (2018), doi: 10.1016/j.ejso.2018.04.017. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Downregulation of RFX1 predicts poor prognosis of patients with small hepatocellular carcinoma Running title: RFX1 in sHCC

1

RI PT

Yingjun Liu1, PengJiang2, Gangcheng Wang1, Xiaonyong Liu1 and Suxia Luo3

Department of General Surgery, Affiliated Tumor Hospital of Zhengzhou University, Henan Cancer

3

Affiliated Tumor Hospital of Guangzhou Medical University, Guangzhou, China

M AN U

2

SC

Hospital, Zhengzhou, China

Department of Internal Medicine, Affiliated Tumor Hospital of Zhengzhou University, Henan Cancer

TE D

Hospital, Zhengzhou, China

Address correspondence to: Suxia Luo, MD, Department of Internal Medicine, Affiliated Tumor Hospital of Zhengzhou University, Henan Cancer Hospital, 127 Dongming Road, 450008 Zhengzhou,

AC C

EP

China. Phone/Fax: +86 371 65587009; E-mail: [email protected]

1

ACCEPTED MANUSCRIPT Abstract Objective Regulatory factor X1 (RFX1) deletion has been reported to be correlated with poor prognosis of some types of cancer. The present study aimed to investigate the prognostic value of RFX1 in HCC,

RI PT

especially in small hepatocellular carcinoma. Methods Immunohistochemical assay was used to investigate RFX1 expression in 221 HCC tissues and another validation cohort of 71 small HCC samples. We also performed in vitro experiments to if

RFX1

regulated

invasive

capacity

HCC

cells

and

expression

of

M AN U

epithelial-mesenchymal transition (EMT) markers.

of

SC

investigate

Results We found that RFX1 expression was significantly lower in HCC tissues compared to the corresponding non-tumor tissues. Further survival analysis suggested that the downregulation of RFX1 correlated with poor prognosis and a high recurrence risk in HCC patients, particularly in small HCC

TE D

patients. Furthermore, another validation cohort of small HCC samples confirmed that downregulation of RFX1 in HCC tissues predicted high recurrence risk and poor prognosis for early stage HCC patients. In vitro studies suggested that knocking down RFX1 facilitated HCC cell invasion, while

EP

overexpression of RFX1 reduced the invasion of HCC cells. Western blot assays also indicated that

AC C

RFX1 regulated expression of some EMT markers. Knocking down RFX1 decreased E-cadherin and increased vimentin expression, while RFX1 overexpression enhanced E-cadherin and decreased vimentin expression.

Conclusions Our study demonstrated that RFX1 downregulation is a new predictive marker of high recurrence risk and poor prognosis of HCC; It has potential to help guide treatment for postoperative HCC patients, especially for small HCC patients. Key words: RFX1, small hepatocellular carcinoma, prognosis, early recurrence 2

ACCEPTED MANUSCRIPT

Introduction Hepatocellular carcinoma (HCC) is the sixth most common cancer, and the third largest cause of

RI PT

cancer-associated death worldwide; the incidences of these deaths increase every year[1, 2]. In past decades, with the development of diagnostic tools and treatments, a higher proportion of patients have been diagnosed at an early stage and received curative treatments, improving HCC outcomes[3-5].

SC

Based on Milan criteria, the early-stage HCC is defined as a single HCC ≤ 5 cm in the maximum

M AN U

diameter or up to 3 nodules <3 cm[6]. Small HCC (sHCC, HCC ≤ 5 cm)[7, 8] can be fallen into the early-stage HCC. Curative surgery was one of the primary treatment for patients with small HCC, however the overall and recurrence free survival of patients with small HCC after curative surgical resection remains poor[9, 10]. Thus, it is crucial to identify patients, especially sHCC patients, with a

outcomes[11].

TE D

high risk of recurrence in order to develop more effective and targeted treatments and improve disease

Regulatory factor X1 (RFX1) is one of seven members of the RFX family, which is characterized

EP

by a highly conserved DNA-binding domain[12]. This domain binds the X-box consensus sequence in

AC C

promoter regions to regulate gene expression[13]. Expression of RFX1 is linked to various biologic processes, such as hearing[14], transactivating the hepatitis B virus enhancer[15], and gene expression and regulation[16-19]. In addition, emerging evidence suggests that RFX1 might play an important role in cancer development. Decreased expression of RFX1 had been observed in glioblastoma and esophageal adenocarcinoma, and decreased expression indicated poor prognosis[16, 19]. RFX1 might play a key role in cancer development and recurrence. Nevertheless, these data are fragmented, and thus far, the role of the RFX1 gene in hepatocellular carcinoma has not been explored. In this study, we 3

ACCEPTED MANUSCRIPT investigated the expression status of RFX1 in HCC samples and evaluated the prognostic significance of RFX1 in HCC, especially sHCC. Moreover, we investigated the molecular function of RFX1 in HCC cells to determine if RFX1 regulated the invasion and metastatic capacity of HCC cells.

RI PT

Materials and methods Patients and specimens

Experimental cohort tumor tissues and matched adjacent non-tumorous tissues were consecutively

SC

collected from 221 HCC patients; this collection included 124 large HCC and 97 small HCC patients

M AN U

who received curative resection. Validation cohort tumor tissues were consecutively collected from 71 small HCC patients who received curative resection. The cases selected were collected consecutively and were based on the following criteria[20]: patients had undergone curative liver resection and had survived for at least 2 months after hepatectomy; patients had pathological confirmation of HCC;

TE D

patients had no anticancer treatment prior to hepatectomy; patients had no previous malignant disease or a second primary tumor; and, patients had regular follow-up visits. In addition, for inclusion in the validation cohort, the sHCC patients met the following criteria[21, 22]: they had a solitary tumor, ≤ 5

EP

cm in diameter, with no lymph node metastasis or cancer embolus. Detailed clinicopathological

AC C

parameters are listed in Table 1. Cell culture

Huh7 and SMCC7721 cells were obtained from the Liver Cancer Institute of Fudan University (Shanghai, China) and maintained in high-glucose DMEM (Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Life Technologies) at 37°C under 5% CO2. Immunohistochemistry assay The resected surgical specimens were fixed by formalin, embedded in paraffin, cut into 4-µm sections 4

ACCEPTED MANUSCRIPT and mounted on glass slides. All sections were incubated at 62°C for 2 h, then deparaffinized in xylene and rehydrated using a series of graded alcohols. The tissue slides were then treated with fresh 3% hydrogen peroxide in methanol for 15 min to block endogenous peroxidase, and antigen retrieval was

RI PT

carried out in a 0.01 M sodium citrate buffer (pH 6.0) using a microwave oven for 25 min. After sections were colored for 30 min at room temperature, non-specific staining was blocked by preincubation in 10% normal goat serum for 1 h. The specimens were incubated with mouse

SC

monoclonal antibodies against RFX1 (1:100; Santa Cruz, Dallas, TX) overnight at 4°C. Then, the tissue

M AN U

slides were incubated with HRP-labeled anti-mouse/rabbit secondary antibody (DAKO, Glostrup, Denmark) for 1 h at room temperature, and the reaction product was visualized with a non-biotin horseradish peroxidase detection system according to the manufacturer’s protocols (DAKO). Finally, specimens were counterstained with hematoxylin for 90 s. The IHC staining was scored as in previous

TE D

reports[20, 23]. Two different pathologists who specialize in liver cancer evaluated the IHC results for RFX1 and were blinded to the clinical data. Both the extent and intensity of immunostaining were taken into consideration when analyzing the data. The intensity of staining was scored from 0% to

EP

100%. The final quantification of each staining was obtained by multiplying the two scores. RFX1 and

AC C

E-cadherin expression were classified as positive if the score was higher than 1.5; if the score was 1.5 or less, the case was classified as negative. Vimentin expression was considered positive if the score reached 1.

Lentiviral vector production and infection RFX1 overexpression lentiviral vectors were produced as previously described[24]. Briefly, 293T cells were co-transfected with packaging vectors payload, psPAX2, pRSV-REV and pMD2.G plasmids. Lentiviral particles in supernatants were harvested at 48 and 72 hrs after transfection then were filtered 5

ACCEPTED MANUSCRIPT and stored at -80°C until use. Cells were infected by lentivirals with 8 µg/ml Polybrene (Sigma-Aldrich, St. Louis, MO) and were selected with 750 µg/ml G418 (Calbiochem, San Diego, CA). RNAi treatment

RI PT

Briefly, Huh7 cells were transfected with chemically synthesized siRNA using the lipofectamine RNAiMAX transfection reagent (Life Technologies) for 48 hrs. The cells were subsequently analyzed by cell invasion assays or lysed and analyzed by western blot. The siRNA against RFX1 was designed

SC

and chemically synthesized by Guangzhou Ribobio LTD (Guangzhou, China).

M AN U

Western blot

Western blot assays were performed as previously reported[20, 25]. The primary antibodies used for the western blot were as follows: mouse monoclonal antibodies against E-cadherin (BD Biosciences, San Jose, CA, USA) and vimentin (Santa Cruz); mouse monoclonal antibodies against RFX1 (Santa

Cell invasion assay

TE D

Cruz); and rabbit monoclonal antibodies against β-actin (Cell Signaling Technology, Danvers, MA).

In vitro Matrigel invasion assays were performed in transwell chambers (8-µm pore size; Costar,

EP

Corning, NY) according to the manufacturer’s instructions. Approximately 2 × 104 cells were placed

AC C

into the top chamber of each insert (BD, Franklin Lakes, NJ) and incubated at 37°C for 48 hrs. Cells that moved through the Matrigel were stained using Hoechst 33342 (Beyotime, Jiangsu, China) and quantified.

Statistical analysis

All statistical analyses were performed with SPSS statistical software (version 22.0; SPSS, IBM, Armonk, NY). Survival curves were constructed using the Kaplan-Meier method and analyzed by the log-rank test. Significant prognostic factors identified by univariate analysis were entered into 6

ACCEPTED MANUSCRIPT multivariate analysis using the Cox proportional hazards model. The two-tailed chi-squared test was used to analyze the association of RFX1 expression with various clinicopathological parameters. Student’s t-test was used for comparisons, and the Pearson correlation test (two-tailed) was used to

RI PT

calculate the correlation coefficient (r) and P-value between RFX1 and vimentin staining scores or between RFX1 and E-cadherin staining scores. Statistical significance was declared if P < 0.05. Results

SC

Downregulation of RFX1 was correlated with high recurrence risk and poor prognosis of HCC

M AN U

patients.

To explore the expression level of RFX1, 221 samples including HCC tissues and adjacent non-tumor liver tissues were subjected to immunohistochemistry staining for RFX1. The results suggested that RFX1 was expressed in the nucleus of hepatocyte and HCC cells (Fig. 1A). We detected that RFX1

TE D

expression was downregulated in HCC tissue samples compared with the corresponding non-tumor tissues (P < 0.05; Figs. 1A & B). This result revealed that RFX1 was obviously downregulated in tumor tissues. To further understand the clinicopathological significance of RFX1 expression in HCC patients,

EP

we tested the correlation between RFX1 expression status in the 221 HCC specimens with several

AC C

standard clinicopathological parameters. The results of the statistical analysis showed that low RFX1 expression correlated with vascular invasion, tumor size and early recurrence (P < 0.05, Table 1). No significant association was observed between RFX1 expression and other clinicopathological parameters, such as tumor differentiation, tumor number and tumor capsule. Moreover, we found that RFX1 expression level in HCC tissues was negatively correlated with high recurrence risk and poor prognosis of HCC patients. The 1-, 2- and 3-year cumulative recurrence rates in the HCC patients with negative RFX1 expression were significantly higher than in the patients 7

ACCEPTED MANUSCRIPT with positive RFX1 expression (54.9%, 63.1%, and 70.7% vs. 36.1%, 45.6%, and 52.6%, respectively; Fig. 1C). Similarly, the 1-, 2- and 3-year OS rates in the patients who were negative for RFX1 expression were significantly lower than in the patients who were positive for RFX1 expression (67.2%,

RI PT

50.6%, and 42.3% vs. 82.3%, 67.5%, and 60.5%, respectively; Fig. 1D). Univariate and additional multivariate analyses demonstrated that RFX1 was an independent prognostic factor in both the OS (P < 0.05) and TTR (P < 0.05) of HCC patients (Tables 2 & 3).

SC

Low RFX1 expression was correlated with high early recurrence risk in small HCC patients.

M AN U

To further demonstrate the value of RFX1 expression in predicting survival of small HCC patients, we next focused on a subset of the 221 HCC patients; we defined 97 sHCC patients as the experimental cohort. In the experimental cohort, negative RFX1 expression was observed in 39/97 cases (40.3%) by IHC. HCC patients with negative RFX1 expression were significantly associated with higher risk of

TE D

early recurrence (21/39, 53.8%), compared to patients with positive RFX1 expression (12/58, 20.7%; P < 0.05). We also employed a validation sHCC cohort to further evaluate the prognostic value of RFX1 for sHCC patients. In the validation cohort, negative RFX1 expression was observed in 45/71 (63.3%)

EP

of the sHCC samples, which was similar to the proportion observed in sHCC patients from the

AC C

experimental cohort (P > 0.05). Moreover, negative expression of RFX1 was also significantly correlated with early recurrence in the validation cohort (51.1% vs 15.4%; P < 0.05). Prognosis value of RFX1 in patients with sHCC. Survival analysis was also performed in both the experimental and validation cohorts. These results suggested that RFX1 expression status was significantly associated with TTR and OS in both the experimental (Figs. 2A & B) and validation cohorts (Figs. 2C & D). The prognosis of RFX1-negative patients was significantly worse than that of RFX1-positive patients. Moreover, the prognostic 8

ACCEPTED MANUSCRIPT predictive value of RFX1 expression was more significant for the sHCC cohort than in the entire HCC cohort. These findings suggest that RFX1 may serve as an important factor predicting potential recurrence in HCC patients, particularly in sHCC patients.

RI PT

Downregulation of RFX1 changed EMT-marker expression and facilitated HCC cell invasion. Analysis of clinical parameters suggested that downregulation of RFX1 facilitated HCC invasion. Thus, we performed in vitro experiments to investigate if RFX1 regulated invasive capacity of HCC cells. We

SC

detected RFX1 expression in Huh7 and SMCC7721, two well reported HCC cell lines with different

M AN U

invasion capacity[20]. We found that RFX1 expression was higher in Huh7 cells than that of SMCC7721 cells, which had higher invasion capacity than Huh7 cells (Figs. 3A). We used siRNA against RFX1 to knock down RFX1 expression in Huh7 cells, which have high RFX1 expression level, and overexpressed RFX1 with a lentivirus vector in SMCC7721 cells, which have almost

TE D

non-detectable RFX1. Then, we performed cell invasion assays on these cells. We found that siRNA against RFX1 could significantly promote the invasion of Huh7 cell (Figs. 3B & C), while overexpression of RFX1 in SMMC7721 cells enhanced the invasion of SMMC7721 cells (Figs. 3B &

EP

C). These data suggested that low RFX1 expression could promote the invasion of HCC cells. In

AC C

addition, we detected E-cadherin and vimentin expression in the indicated cells. E-cadherin and vimentin are important epithelial-mesenchymal transition (EMT) markers that are present in epithelium-type and mesenchyme-type cells, respectively. It has been well reported that epithelial-mesenchymal transition increases the invasive capacity of tumor cells[20]. We found that knocking down RFX1 decreased E-cadherin and increased vimentin, while overexpression of RFX1 enhanced E-cadherin expression and decreased vimentin expression (Fig. 3A). This result suggests that downregulation of RFX1 expression might promote the EMT process and thus enhance the invasion of 9

ACCEPTED MANUSCRIPT HCC cells. RFX1 expression was correlated with the expression of E-cadherin and vimentin in HCC tissues. Our western blot results suggested that RFX1 expression was associated with the expression status of

RI PT

EMT markers, E-cadherin and vimentin. To further confirm whether RFX1 expression is correlated with E-cadherin and vimentin expression in HCC tissues, we implemented IHC assays for RFX1, E-cadherin and vimentin in serial sections of 118 HCC tissues (Fig. 3D). Pearson correlation tests

SC

suggested that the IHC score of RFX1 was negatively correlated with the IHC score of E-cadherin (Fig.

M AN U

3E) and positively correlated with the IHC score of vimentin (Fig. 3F). These data confirmed the result of western blot performed on HCC cells and suggest that downregulation of RFX1 might promote EMT and thus increase the invasive capacity of HCC cells. Discussion

TE D

The long-term survival of HCC patients remains unsatisfactory[26-28]. Recurrence and intrahepatic metastasis are common for HCC patients, even for small hepatocellular carcinoma (sHCC) patients who receive radical surgey. Recurrence and metastasis are the key factors that affect the prognosis of

EP

HCC patients. Thus, identifing sHCC patients with a high risk of recurrence and metastasis would be

AC C

help to develop more effective and targeted treatment strategies to improve disease outcome.In recent studies, aberrant downregulation of RFX1 expression has been observed in the progression of many kinds of tumors. Thus, we tried to explore the role of RFX1 in HCC in this study. Our study identified RFX1 as a potential predictive biomarker for invasion and prognosis in HCC and sHCC. We found that downregulation of RFX1 increased invasion of HCC cells and indicated high recurrence risk and poor prognosis. In this study, we evaluated the expression of RFX1 in 221 HCC samples and explored the 10

ACCEPTED MANUSCRIPT prognostic predictive value of RFX1 in HCC patients. By IHC assays, we found that RFX1 expression was downregulated in HCC tissues. Moreover, downregulation of RFX1 was correlated with poor prognosis of HCC patients. Kaplan–Meier analysis showed that the patients with low RFX1 expression

RI PT

exhibited shorter OS times and higher recurrence rates compared to patients with high RFX1 expression. Univariate and multivariate Cox analysis further demonstrated that the RFX1 expression was an independent predictor of survival and recurrence in HCC patients. Interestingly, the predictive

SC

value of RFX1 for recurrence risk and poor prognosis was more significant in sHCC patients. With the

M AN U

improvement of early diagnosis rates, the incidence rate of sHCC patient is increasing. However, the prognosis of sHCC is still unsatisfactory. Postoperative recurrence and intrahepatic metastasis are frequently observed in sHCC patients and result in treatment failure. Using biomarkers to identify patients with higher risk of recurrence or developing a worse prognosis may help to direct further

TE D

treatment, decrease mortality and reduce medical costs. Nevertheless, biomarkers with applied values for predicting recurrence risk and prognosis of small HCC patients are still absent in clinical practice. Some well-accepted prognostic predictive parameters such as tumor size, tumor number and cancer

EP

embolus do not have a prognostic predicative value for sHCC. Thus, sensitive prognostic molecular

AC C

markers of early stage HCC patients, especially for small HCC patients are still needed, and it is important to identify a prognosis predicator for sHCC. In this study, we found that survival differences between patients with high and low RFX1 expression were much more significant in 97 of the 221 HCC patients who had sHCC (solitary tumor ≤ 5 cm in diameter, with no lymph node metastasis and cancer embolus) than the difference in the overall HCC patients. In addition, verification in another cohort of 71 sHCC patients confirmed this result. This result suggests that RFX1 has potential to be a prognosis biomarker for small HCC and might help to direct further therapy. 11

ACCEPTED MANUSCRIPT We also investigated the molecular function of RFX1 in HCC cells. We found that RFX1 affected the expression of epithelial–mesenchymal transition markers and the invasion capacity of HCC cells. Overexpression of RFX1 increased E-cadherin expression, decreased vimentin expression and reduced

RI PT

invasion of SMCC7721 cells, while knocking down RFX1 expression produced the opposite. It has been well reported that EMT is a critical event in cancer development and metastasis, and EMT can increase the invasion of cancer cells[29, 30]. Our result suggested that RFX1 might inhibit EMT, thus

SC

reducing the invasion capacity of HCC cells. Downregulation of RFX1 promoted the EMT process and

M AN U

enhanced the invasion of HCC cells, thereby increasing the recurrence risk and resulting in poor prognosis of HCC patients. However, more studies are needed to understand the specific molecular mechanisms of EMT and the invasion regulation function of RFX1. Conclusion

TE D

In conclusion, our study showed that RFX1 was a novel predictive biomarker for prognosis and recurrence risk in HCC patients, specifically for sHCC patients. RFX1 has the potential to help predict prognosis of small HCC and to direct further treatment plan development.

EP

Conflict of interest statement

AC C

All of our authors declare that we have no conflicts of interest concerning this article. Ethical statement

This study has been performed in accordance with the ethical standards as laid

down in the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by the Clinical Research Ethics Committee of Affiliated Tumor Hospital of Guangzhou Medical University.

References 1.

Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012.

CA: a cancer journal for clinicians. 2015;65(2):87-108. Epub 2015/02/06. doi: 10.3322/caac.21262. 12

ACCEPTED MANUSCRIPT PubMed PMID: 25651787. 2.

Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet (London, England).

2012;379(9822):1245-55. Epub 2012/02/23. doi: 10.1016/s0140-6736(11)61347-0. PubMed PMID: 22353262. 3.

Zhu WW, Guo JJ, Guo L, Jia HL, Zhu M, Zhang JB, et al. Evaluation of midkine as a diagnostic

serum biomarker in hepatocellular carcinoma. Clinical cancer research : an official journal of the American Association for Cancer Research. 2013;19(14):3944-54. Epub 2013/05/31. doi: 4.

RI PT

10.1158/1078-0432.ccr-12-3363. PubMed PMID: 23719264.

Vitale A, Peck-Radosavljevic M, Giannini EG, Vibert E, Sieghart W, Poucke SV, et al.

Personalized Treatment of Patients with very Early Hepatocellular Carcinoma. Journal of hepatology. 2016. Epub 2016/09/30. doi: 10.1016/j.jhep.2016.09.012. PubMed PMID: 27677712. 5.

Jin GZ, Yu WL, Dong H, Zhou WP, Gu YJ, Yu H, et al. SUOX is a promising diagnostic and

SC

prognostic biomarker for hepatocellular carcinoma. Journal of hepatology. 2013;59(3):510-7. Epub 2013/05/15. doi: 10.1016/j.jhep.2013.04.028. PubMed PMID: 23665285. 6.

Lan T, Chang L, Rahmathullah MN, Wu L, Yuan YF. Comparative Efficacy of Interventional

Therapies for Early-stage Hepatocellular Carcinoma: A PRISMA-compliant Systematic Review and Meta-analysis.

Medicine.

10.1097/md.0000000000003185.

PubMed

PMCPMC4839802. 7.

2016;95(15):e3185.

Epub

PMID:

PubMed

M AN U

Network

27082558;

2016/04/16. Central

doi: PMCID:

Hocquelet A, Aube C, Rode A, Cartier V, Sutter O, Manichon AF, et al. Comparison of no-touch

multi-bipolar vs. monopolar radiofrequency ablation for small HCC. J Hepatol. 2017;66(1):67-74. Epub 2016/07/17. doi: 10.1016/j.jhep.2016.07.010. PubMed PMID: 27422750. 8.

Gong Q, Qin Z, Hou F. Improved treatment of early small hepatocellular carcinoma using

TE D

sorafenib in combination with radiofrequency ablation. Oncology letters. 2017;14(6):7045-8. Epub 2018/01/19. doi: 10.3892/ol.2017.7174. PubMed PMID: 29344133; PubMed Central PMCID: PMCPMC5754899. 9.

Suh SW, Lee KW, Lee JM, You T, Choi Y, Kim H, et al. Prediction of aggressiveness in

early-stage hepatocellular carcinoma for selection of surgical resection. Journal of hepatology.

EP

2014;60(6):1219-24. Epub 2014/02/20. doi: 10.1016/j.jhep.2014.01.027. PubMed PMID: 24548529. 10. Zhai XF, Liu XL, Shen F, Fan J, Ling CQ. Traditional herbal medicine prevents postoperative recurrence of small hepatocellular carcinoma: A randomized controlled study. Cancer. 2018. Epub

AC C

2018/03/03. doi: 10.1002/cncr.30915. PubMed PMID: 29499082. 11. Hong J, Hu K, Yuan Y, Sang Y, Bu Q, Chen G, et al. CHK1 targets spleen tyrosine kinase (L) for proteolysis in hepatocellular carcinoma. The Journal of clinical investigation. 2012;122(6):2165-75. doi: 10.1172/JCI61380. PubMed PMID: 22585575; PubMed Central PMCID: PMC3589177. 12. Feng C, Zuo Z. Regulatory factor X1-induced down-regulation of transforming growth factor beta2 transcription in human neuroblastoma cells. The Journal of biological chemistry. 2012;287(27):22730-9. Epub 2012/05/15. doi: 10.1074/jbc.M111.338590. PubMed PMID: 22582395; PubMed Central PMCID: PMCPmc3391092. 13. Aftab S, Semenec L, Chu JS, Chen N. Identification and characterization of novel human tissue-specific RFX transcription factors. BMC evolutionary biology. 2008;8:226. Epub 2008/08/05. doi:

10.1186/1471-2148-8-226.

PubMed

PMID:

18673564;

PubMed

Central

PMCID:

PMCPmc2533330. 14. Elkon R, Milon B, Morrison L, Shah M, Vijayakumar S, Racherla M, et al. RFX transcription 13

ACCEPTED MANUSCRIPT factors are essential for hearing in mice. Nature communications. 2015;6:8549. Epub 2015/10/16. doi: 10.1038/ncomms9549. PubMed PMID: 26469318; PubMed Central PMCID: PMCPmc4634137. 15. Li MS, Lau TC, Chan SK, Wong CH, Ng PK, Sung JJ, et al. The G1613A mutation in the HBV genome affects HBeAg expression and viral replication through altered core promoter activity. PloS one. 2011;6(7):e21856. Epub 2011/08/05. doi: 10.1371/journal.pone.0021856. PubMed PMID: 21814558; PubMed Central PMCID: PMCPmc3140978. 16. Watts JA, Zhang C, Klein-Szanto AJ, Kormish JD, Fu J, Zhang MQ, et al. Study of FoxA pioneer adenocarcinoma

development.

10.1371/journal.pgen.1002277.

PLoS

genetics.

PubMed

RI PT

factor at silent genes reveals Rfx-repressed enhancer at Cdx2 and a potential indicator of esophageal 2011;7(9):e1002277.

PMID:

21935353;

PMCPmc3174211.

Epub

PubMed

2011/09/22.

Central

doi:

PMCID:

17. Su JC, Tseng PH, Hsu CY, Tai WT, Huang JW, Ko CH, et al. RFX1-dependent activation of

SC

SHP-1 induces autophagy by a novel obatoclax derivative in hepatocellular carcinoma cells. Oncotarget. 2014;5(13):4909-19. Epub 2014/06/24. doi: 10.18632/oncotarget.2054. PubMed PMID: 24952874; PubMed Central PMCID: PMCPmc4148109.

18. Jiang J, Zhang N, Shiba H, Li L, Wang Z. Spermatogenesis associated 4 promotes Sertoli cell

M AN U

proliferation modulated negatively by regulatory factor X1. PloS one. 2013;8(10):e75933. Epub 2013/10/23. doi: 10.1371/journal.pone.0075933. PubMed PMID: 24146794; PubMed Central PMCID: PMCPmc3795713.

19. Feng C, Zhang Y, Yin J, Li J, Abounader R, Zuo Z. Regulatory factor X1 is a new tumor suppressive transcription factor that acts via direct downregulation of CD44 in glioblastoma. Neuro-oncology. 2014;16(8):1078-85. Epub 2014/02/15. doi: 10.1093/neuonc/nou010. PubMed PMID: 24526308; PubMed Central PMCID: PMCPmc4096175.

TE D

20. He L, Zhou X, Qu C, Hu L, Tang Y, Zhang Q, et al. Musashi2 predicts poor prognosis and invasion in hepatocellular carcinoma by driving epithelial-mesenchymal transition. J Cell Mol Med. 2014;18(1):49-58. Epub 2013/12/07. doi: 10.1111/jcmm.12158. PubMed PMID: 24305552; PubMed Central PMCID: PMCPMC3916117.

21. Jiang P, Tang Y, He L, Tang H, Liang M, Mai C, et al. Aberrant expression of nuclear KPNA2 is

EP

correlated with early recurrence and poor prognosis in patients with small hepatocellular carcinoma after hepatectomy. Medical oncology. 2014;31(8):131. doi: 10.1007/s12032-014-0131-4. PubMed PMID: 25031071.

AC C

22. Du M, Chen L, Zhao J, Tian F, Zeng H, Tan Y, et al. Microvascular invasion (MVI) is a poorer prognostic predictor for small hepatocellular carcinoma. BMC cancer. 2014;14:38. Epub 2014/01/28. doi:

10.1186/1471-2407-14-38.

PubMed

PMID:

24460749;

PubMed

Central

PMCID:

PMCPmc3922738.

23. Gao YF, Li T, Chang Y, Wang YB, Zhang WN, Li WH, et al. Cdk1-phosphorylated CUEDC2 promotes spindle checkpoint inactivation and chromosomal instability. Nature cell biology. 2011;13(8):924-33. Epub 2011/07/12. doi: 10.1038/ncb2287. PubMed PMID: 21743465. 24. Salloum S, Holmes JA, Jindal R, Bale SS, Brisac C, Alatrakchi N, et al. Exposure to human immunodeficiency virus/hepatitis C virus in hepatic and stellate cell lines reveals cooperative profibrotic transcriptional activation between viruses and cell types. Hepatology (Baltimore, Md). 2016;64(6):1951-68. Epub 2016/08/18. doi: 10.1002/hep.28766. PubMed PMID: 27531241; PubMed Central PMCID: PMCPMC5116014. 25. Hong J, Yuan Y, Wang J, Liao Y, Zou R, Zhu C, et al. Expression of variant isoforms of the 14

ACCEPTED MANUSCRIPT tyrosine kinase SYK determines the prognosis of hepatocellular carcinoma. Cancer research. 2014;74(6):1845-56. doi: 10.1158/0008-5472.CAN-13-2104. PubMed PMID: 24477596; PubMed Central PMCID: PMC3959256. 26. Seki E, Schwabe RF. Hepatic inflammation and fibrosis: functional links and key pathways. Hepatology (Baltimore, Md). 2015;61(3):1066-79. Epub 2014/07/30. doi: 10.1002/hep.27332. PubMed PMID: 25066777; PubMed Central PMCID: PMCPMC4306641. 27. Sandhu DS, Baichoo E, Roberts LR. Fibroblast growth factor signaling in liver carcinogenesis.

RI PT

Hepatology (Baltimore, Md). 2014;59(3):1166-73. Epub 2014/04/10. PubMed PMID: 24716202.

28. Lok AS, McMahon BJ, Brown RS, Jr., Wong JB, Ahmed AT, Farah W, et al. Antiviral therapy for chronic hepatitis B viral infection in adults: A systematic review and meta-analysis. Hepatology (Baltimore, Md). 2016;63(1):284-306. Epub 2015/11/14. doi: 10.1002/hep.28280. PubMed PMID: 26566246.

SC

29. Xie K, Ye Y, Zeng Y, Gu J, Yang H, Wu X. Polymorphisms in Genes Related to Epithelial-Mesenchymal Transition and Risk of Non-Small Cell Lung Cancer. Carcinogenesis. 2017. Epub 2017/10/03. doi: 10.1093/carcin/bgx079. PubMed PMID: 28968839.

30. Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nature reviews Cancer.

AC C

EP

TE D

M AN U

2002;2(6):442-54. Epub 2002/08/22. doi: 10.1038/nrc822. PubMed PMID: 12189386.

15

ACCEPTED MANUSCRIPT Figure legend: Fig. 1 Downregulation of RFX1 was correlated with poor prognosis of HCC. (A) Representative images of RFX1 Immunohistochemistry staining in HCC and adjacent non-tumor tissues. Scale bar,

with high recurrence risk and poor overall survival in HCC patients.

RI PT

100 µm. (B) RFX1 was downregulated in HCC tissues. (C & D) Downregulation of RFX1 correlated

Fig. 2 Downregulation of RFX1 was correlated with poor prognosis of small HCC. (A & B)

SC

Downregulation of RFX1 was correlated with high recurrence risk and poor overall survival of small

M AN U

HCC patients in the experimental cohort. (C & D) Downregulation of RFX1 was correlated with high recurrence risk and poor overall survival of small HCC patients in the validation cohort. Fig. 3 RFX1 correlated with E-cadherin and vimentin expression, and regulated the invasion capacity of HCC cells. (A) Downregulation of RFX1 decreased E-cadherin expression and increased

TE D

vimentin expression in Huh7 cells, while overexpression of RFX1 increased E-cadherin expression and decreased vimentin expression in SMCC7721 cells. (B & C) Downregulation of RFX1 increased invasion capacity of huh7 cells, while overexpression of RFX1 decreased the invasion of SMCC7721

EP

cells. (D) Representative immunohistochemistry staining images of RFX1, E-cadherin and vimentin in

AC C

serial sections of human HCC tissues. Scale bar, 100 µm. (E & F) Scatter plots showed the results from IHC assays performed on serial sections for the correlation between the expression of RFX1 and EMT markers, E-cadherin and vimentin.

16

ACCEPTED MANUSCRIPT Table 1 Correlation of RFX1 protein expression with clinicopathological parameters Characteristics

n

RFX1

p value

Negative

Positive

≤ 50

113

50 (44.2%)

63 (55.8%)

> 50

108

52 (48.1%)

56 (51.9%)

≤ 20

64

23 (35.9%)

41 (64.1%)

> 20

157

79 (50.3%)

78 (49.7%)

Negative

27

9 (33.3%)

18 (66.7%)

Positive

194

93 (47.9%)

101 (52.1%)

≤ 50

59

45 (76.2%)

> 50

162

111 (68.5%)

A

203

94 (46.3%)

SC

Age (years)

B

18

0.561

Serum AFP (µg/l)

GGT (U/I)

Tumor capsule No/incomplete

173

Complete

48

Tumor differentiation I-II

138 83

Vascular invasion No Yes Liver cirrhosis Yes

109 (53.7%) 10 (55.6%)

82 (47.4%)

91 (52.6%)

20 (41.7%)

28 (58.3%)

62 (44.9%)

76 (55.1%)

40 (48.2%)

43 (51.8%)

165

69 (41.8%)

96 (58.2%)

56

33 (58.9%)

23 (41.1%)

95

40 (42.1%)

55 (57.9%)

126

62 (49.2%)

64 (50.8%)

EP

No

0.154

0.263

51 (31.5%)

8 (44.4%)

TE D

III-IV

14 (23.8%)

M AN U

Child-Pugh score

RI PT

HBsAg

0.052

0.879

0.481

0.673

0.026

0.294

Tumor size (cm)

96

35 (36.5%)

61 (63.5%)

125

67 (53.6%)

58 (46.4%)

Solitary

157

69 (43.9%)

88 (56.1%)

Multiple

64

33 (51.6%)

31 (48.4%)

0/A

91

38 (41.8%)

53 (58.2%)

B

98

44 (44.9%)

54 (55.1%)

C

32

20 (62.5%)

12 (37.5%)

No

104

39 (37.5%)

65 (62.5%)

Yes

117

63 (53.8%)

54 (46.2%)

AC C

≤5

>5

0.011

Tumor number

0.303

BCLC stage

0.122

Early recurrence

Bold values (p < 0.05) are statistically significant

0.015

ACCEPTED MANUSCRIPT Table 2 Univariate analysis of factors associated with survival and recurrence OS

TTR

Variables Hazard radio (95%CI)

p value

Hazard radio (95%CI)

p value

0.645 (0.448-0.931)

0.019

0.618 (0.442-0.864)

0.005

AFP(µg/L)

2.309 (1.451-3.674)

<0.001

2.201 (1.452-3.335)

<0.001

HBsAg

2.136 (1.041-4.381)

0.038

1.961 (1.059-3.633)

0.032

GGT(U/L)

1.594 (0.993-2.559)

0.053

1.721 (1.154-2.567)

0.008

Child-Pugh score

1.590 (0.893-2.832)

0.115

1.704 (0.997-2.912)

0.051

Tumor size(cm)

2.621 (1.761-3.901)

<0.001

2.755 (1.915-3.964)

<0.001

Tumor number

1.865 (1.285-2.706)

0.001

2.191 (1.555-3.087)

<0.001

Tumor capsule

0.708 (0.445-1.125)

0.144

Tumor differentiation

1.728 (1.201-2.478)

0.003

Vascular invasion

2.783 (1.907-4.060)

<0.001

BCLC stage

2.193 (1.696-2.834)

Liver-cirrhosis RFX1 expression

SC

RI PT

Age(years)

0.513

1.595 (1.142-2.228)

0.006

2.226 (1.559-3.178)

<0.001

<0.001

1.923 (1.523-2.427)

<0.001

1.363 (0.941-1.976)

0.102

1.182 (0.844-1.655)

0.330

0.627 (0.436-0.900)

0.011

0.598 (0.429-0.834)

0.002

AC C

EP

TE D

M AN U

0.874 (0.584-1.308)

ACCEPTED MANUSCRIPT Table 3 Multivariate analysis of factors associated with survival and recurrence Variables

OS Hazard

radio

TTR p value

(95%CI)

Hazard radio (95%CI)

p value

1.659 (1.137-2.421)

0.009

1.927 (1.357-2.736)

<0.001

Tumor differentiation

1.640 (1.136-2.367)

0.008

1.453 (1.037-2.036)

0.030

RFX1

0.660 (0.458-0.951)

0.026

0.669 (0.477-0.937)

0.019

AC C

EP

TE D

M AN U

SC

RI PT

Tumor number

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT