MEIS-1 level in unresectable hepatocellular carcinoma can predict the post-treatment outcomes of radiofrequency ablation

Radiofrequency ablation (RFA) is a local-ablative therapy for unresectable hepatocellular carcinoma (HCC). At present, there is no predictive marker for RFA treatment outcomes. This work aimed to valuate myeloid ecotropic viral integration site 1 (MEIS-1) in predicting post-RFA treatment outcomes of unresectable HCC patients. The time to progression (TTP) and overall survival (OS) of 81 HCC patients who received RFA treatment were measured. The protein level of MEIS-1 in tumor specimens was measured by western blot. The role of MEIS-1 in RFA-treating HCC in vivo growth nude mouse model was examined via PET/CT imaging. Higher level of MEIS-1 in tumor tissue is associated with better RFA treatment outcomes. The median TTP was 9.0 (95% confidence interval [CI]: 6.8–11.3) months in patients with high MEIS-1 expression (n = 43) versus 6.0 (95% CI: 4.6–7.4) months in patients with low MEIS-1 expression (n = 38). Moreover, in rodent HCC model we found overexpression of MEIS-1 enhanced the anti-tumor effect of RFA treatment. We conclude that high level of MEIS-1 expression predicts better RFA treatment outcome in HCC.


INTRODUCTION
High proportion of hepatitis viruses infection makes China a heavily afflicted country of hepatocellular carcinomas (HCCs); effective HCC treatment approaches will relieve this urgent medical burden [1][2][3]. Normally HCC cannot be diagnosed until developed into advancedstage which is not suitable for liver transplantation or surgical resection. There are limited treatment options available for advanced HCC patients, current HCC treatments have poor clinical outcomes and poor prognosis [4][5][6]. Radiofrequency ablation (RFA), an in situ ablative therapy which selectively destroys HCC tissues, is a promising treatment for advanced-stage HCC patients with cirrhosis and compromised liver function [7][8][9][10]. However, rapid or aggressive recurrence of HCCs after RFA treatment is a major obstacle [11]. It is urgent to study the mechanisms of the recurrence of HCCs after RFA treatment and identify predictive prognosis marker of HCC patients.
Transcription factor myeloid ecotropic viral integration site 1 (MEIS-1) is a member of the triple amino acid loop extension family, which are thought to play important roles in cell growth and differentiation during vertebrate embryogenesis [12]. MEIS-1 contains four functional domains: a N-terminal MEIS-A domain, a MEIS-B domain, a C-terminal transcription factor activity region, and a homeodomain, which links MEIS-B domain and the C-terminal region [13,14]. Previously, MEIS-1 was reported as a positive tumor inducer [15][16][17][18][19]. However, MEIS-1 can also function as a negative regulator of cancers by inhibiting cell proliferation and inducing cell cycle arrest [20][21]. A previous study reported higher expression of MEIS-1 in healthy prostate tissues than in prostate carcinoma tissues, and concluded that MEIS-1 may serve as a predictive biomarker of Research Paper www.oncotarget.com prostate cancer prognosis [21]. Similar studies reported that MEIS-1 is a suppressor of non-small-cell lung cancer, esophageal squamous cell carcinomas and clear cell renal cell carcinomas [14,[22][23][24][25]. A recent paper reported that MEIS-1's expression level can predict treatment outcome in HCC patients [26].
In this study, we examined the predictive value of MEIS-1 expression in determining post-RFA treatment outcomes in HCC patients with advanced-stage disease. MEIS-1 may function as a negative regulator of HCC growth. Overexpression of MEIS-1 enhanced the efficiency of RFA's anti-tumor effect on HCC cell proliferation in vivo. Thus, high level of MEIS-1 expression predicts better RFA treatment outcome in HCC.

Association of MEIS-1 protein level and RFA treatment outcomes
The endogenous protein level of MEIS-1 in clinical specimens was detected by western blot. The results of the quantitative analysis of the western blots are shown in Figure 1A. According to the median value of intratumoral MEIS-1's protein expression, the 81 patients were divided into two groups: those with low MEIS-1 expression (n = 38) and those with high MEIS-1 expression (n = 43), as shown in Figure 1B.

MEIS-1 enhanced RFA-induced inhibition of HCC cells in vivo growth
To validate whether MEIS-1 intrinsicly expression affects tumor growth, we next established HCC cell lines with low or high MEIS-1 expression level. We first selected a HCC cell line MHCC97-H which has very low endogenous MEIS-1 expression (Figure 2A), MEIS-1 was overexpressed via infection of its adenovirus -vectors in MHCC97-H cells ( Figure 2B). We found the overexpression of MEIS-1 increased the expression of E-cadherin, an epithelial marker, and decreased the protein level of N-cadherin and Vimentin, two mesenchymal markers ( Figure 2C and 2D). Therefore, MEIS-1 inhibits the epithelial-mesenchymal transition of MHCC97-H cells.
Next, control MHCC97-H cells or MEIS-1 overexpressing MHCC97-H cells were injected into nude mice subcutaneously. We observed that overexpression of MEIS-1 reduced tumor growth, and the RFA treatment resulted in shrinkage of the tumors ( Figure 3A-3C). Moreover, RFA induced the EMT process of HCC tumors, MEIS-1 inhibited the EMT induced by RFA ( Figure 3D-3G). Thus, overexpression of MEIS-1 enhances the antitumor effect of RFA treatment.
Next, HCC cells were isolated from the subcutaneous tumors ( Figure 3) that had been treated with RFA or infected with vectors; transwell assays (in vitro invasion or migration) and intrahepatic growth assay (injected into the liver lobe) were performed. As shown by the results of the soft agar assay, overexpression of MEIS-1 or RFA attenuated in vitro invasion ( Figure 4A) and migration ( Figure 4B) of HCC cells.
To confirm the results in in vivo tumor growth, the effect of MEIS-1 on RFA was examined by inducing expression model. MHCC97-H cells infected with control virus or TET-on virus which express MEIS-1 were injected into nude mice subcutaneously. When the volumes of subcutaneous tumors reach 1200mm 3 , the RFA was performed. Next, mice received solvent control or tetracycline (Tetracycline does not affect HCC tumor growth, shown in Supplementary Figure 1). As expected, tetracycline treatment induced the overexpression of MEIS-1 and reduced tumor growth ( Figure 5); and the RFA treatment resulted in tumor size reduction ( Figure 5).
Moreover, we validated our hypothesis in in situ HCC cell growth model in which the intraliver tumors were measured by PET/CT. As shown in Figure 6A, RFA treatment decreased the intrahepatic growth of HCC cells. Overexpression of MEIS-1 enhanced the inhibitory effect of RFA on HCC cells' nodule formation in liver ( Figure  6A). The liver-to-blood radioactive data, tumor foci in the whole liver, and PET imaging of the whole liver confirmed the PET imaging of whole animals ( Figure 6B). Thus, overexpression of MEIS-1 enhanced the anti-tumor effect of RFA treatment in in situ HCC.

DISCUSSION
It is well known that advanced HCC patients have a poor prognosis due to the ineffectiveness of systemic chemotherapy and local radiation therapies [27,28]. Recent studies reported that Sorafenib, an oral kinase inhibitor, improved the OS and life quality of HCC patients [29][30][31]. However, intrinsic or acquired Sorafenib resistance in tumor tissue restricts the effect of this promising approach [29][30][31]. On the other hand, local-ablative therapy has becoming a new hope for patients suffering from unresectable HCC. RFA is widely used to treat advanced HCC patients, as it presents better therapeutic outcomes compared to other local ablative methods [32]. Nevertheless, there is not an effective indicator of post-RFA prognosis. www.oncotarget.com   MEIS-1 play a critical role in several important physiological processes, such as organ development and stem cell differentiation [33,34]. It also participates in the development and progression of cancers. For example, aberrant expression of MEIS-1 promoted the development of acute myeloid leukemia [15][16][17][18][19]35]. However, MEIS-1 has been reported as a negative regulator of some other cancers, such as non-small-cell lung cancer, esophageal squamous cell carcinomas, clear cell renal cell carcinomas and prostate cancer [22][23][24][25]. In these studies, the authors proposed that MEIS-1 inhibits tumor cell proliferation and induces cell cycle arrest [22][23][24][25]. In the present study, the patients with high MEIS-1 expression show better post-RFA outcomes than those with low MEIS-1 expression. The median TTP was also longer in the high MEIS-1 group than in the low MEIS-1 group. Furthermore, the CER and CDR of the patients with high MEIS-1 expression were better than those of the patients with low MEIS-1 expression. These results indicated that MEIS-1's expression level indicates the anti-tumor effect and prognosis of RFA therapy.
RFA is one of the most suitable treatment option for advanced-stage HCC, particularly when a patient's liver functional reserve precludes radiotherapy. However, HCC can recur after RFA, and the phenotype, e.g. Epithelialmesenchymal transition, of the tumor cells often changes [36]. Previous research suggested that incomplete ablation in RFA may induce cellular stress and lead to pathological changes [37, 38]. Zhu et al. suggested that overexpression of MEIS-1 inhibits the EMT process and decreases the expression of pro-survival genes in clear cell renal cell carcinomas [25]. Yasui et al. [26] indicated that MEIS-1 may be a suppressor of TGFβ signaling pathway, which is one of the foremost mediators of EMT. TGFβ/Smad signaling pathway plays a crucial role in tumorigenesis and tumor development, included HCC [39,40]. It has been previously reported that TGFβ/Smad can be a tumor suppressor and mediate the expression of tumor suppressor p15 and p21 [41][42][43][44]. Zhou et al. showed that Smad3 can sensitize HCC cells to cisplatin treatment by repressing Akt phosphorylation [44]. At same time, other papers also reported that TGFβ/Smad  [45][46][47]. It is well known that TGFβ/Smad pathway plays a central role in EMT process [48].
In the present study, we established a subcutaneous tumor model and then performed RFA at 65-70°C for 3-5 min to attenuate the growth of HCCs. Overexpression of MEIS-1 enhanced the anti-tumor effect of RFA. In the future, in addition to RFA treatment, anti-tumor agents could be injected into tumors to mimic TACE. In addition, intrahepatic growth of HCC cells in nude mice could be achieved by directly ablate the lesion in liver using abdominal surgeries in mice. Local ablation of intrahepatic HCC by RFA could also be performed in some larger animals, e.g. immunodeficient rat, guided by small animal molecular imaging system, e.g. small animal ultrasound systems.

Patients
This prospective study consisted of 81 consecutive patients who underwent RFA between April 2014 and May 2016, and the baseline clinical data of patients with advanced HCC were shown in Table 2. The inclusion criteria were as follows [7,[49][50][51]: (A) a diagnosis of HCC and stage B or C unresectable cancer; (B) the presence of portal hypertension and Child's class A or B cirrhosis according to endoscopy or imaging; (C) a life expectancy of at least 12 weeks according to the clinical presentation; (D) Eastern Cooperative Oncology Group performance status 0, 1, or 2; and (E) histological grade of HCC differentiation classed as well differentiated, intermediately differentiated, or poorly differentiated according to the criteria of Edmondson. The collection of the HCC specimens and study protocol were approved by the Ethics Committee of the 302nd Hospital, and informed consent was obtained from all the patients.
None of the HCC patients had received any prior treatment. Clinical specimens were obtained by a puncture biopsy using a coaxial needle (Cat. no.: MCXS1815BP, RITA Company, Crystal Lake, IL, USA) immediately after the RFA treatment. The primary endpoint was the time to progression (TTP) post-RFA, and the secondary endpoint was overall survival (OS). The association of the protein expression of MEIS-1 with post-RFA OS and TTP was assessed in the same cohort of patients. The efficacy of  RFA was based on the clinical efficacy response (CER)/ overall response rate or disease-control rate (DCR), in accordance with the protocols described in our previous work [7,51]. The overall response rate was defined as: complete response (CR) + partial response (PR) [7,51]. The disease-control ate was defined as: CR + PR + stable disease (SD) [7,51].

Cell culture and Western blot analysis
A hepatic nontumor cell line, L-02, and HCC cell lines (HepG2, MHCC97-H, MHCC97-L, and Hu7) were cultured in Dulbecco's modified Eagle medium (DMEM) with 10% fetal bovine serum at 37° C in an atmosphere of 5% CO 2 [51,52]. Tetracycline (Cat. no.: S2574) was purchased from Sellck Corporation, Houston, Texas, USA. An empty vector adenovirus and MEIS-1 were purchased from Vigene Company (Jinan, Shandong, China). A TETon Lentivirus of MEIS-1 was generated and prepared by Vigene Company (Jinan, Shandong, China). Cells were harvested for Western blot analysis. Total protein was extracted from cells or clinical specimens and subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis. The proteins were then transferred to a polyvinylidene fluoride film. Next, the blots were blocked by 5% bovine serum albumin and incubated with the no.: sc-71009, Santa Cruz, USA) in a 1:5000 dilution, anti-N-Cadherin immunoglobulin G (IgG) (Cat. no.: sc-59987, Santa Cruz, USA) in a 1:1000 dilution, anti-Vimentin immunoglobulin G (IgG) (Cat. no.: sc-73258, Santa Cruz, USA) in a 1:1000 dilution or anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) IgG (Cat. no.: sc-47724, Santa Cruz, USA) in a 1:5000 dilution. After washing three times in tris-buffered saline, the blots were incubated with a secondary antibody (1:5000 dilution) and developed by the addition of enhanced chemiluminescence reagents (Qiangen, Beijing, China) and X-ray film exposure. The quantitative protein level of MEIS-1 was determined by gray-scanning analysis using Alpha Innotech analysis software (San Leandro, CA, USA). The protein level of MEIS-1 was normalized to that of a loading control (GAPDH). What was calculated using the following formula: gray scale of MEIS-1 band)/(gray scale of GAPDH).

RFA treatment
A multipolar 15 cm-long RFA needle (Cat. no.: UniBlate 700-103597, RITA Company, Crystal Lake, IL, USA), with a maximum ablation diameter of 5 cm was used in all cases. Percutaneous RFA was performed using

Animal experiments
All the animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee of the 302nd Hospital, People's Liberation Army of China. Nude SCID (severe combined immune deficiency) mice aged 4-6 weeks were purchased from Si-Bei-Fu Biotechnology Corporation, Beijing China. Figure 7 shows the workflow of the animal experiments.
To produce a subcutaneous tumor model [53][54][55], MHCC97-H cells infected with a control or MEIS-1 virus vector were injected into nude mice (1 × 10 6 cells per animal). After 4-6 weeks (26 days), the tumoral volume had reached almost 1200 mm 3 . RFA of the subcutaneous tumors was performed using a thyroid ablation needle (Cat. no.: UniBlate 700-103587 17G, RITA Company). The ablation time was 3-5 min, and the temperature Figure 6: Intrahepatic growth of cells separated from subcutaneous tumors. The subcutaneous tumors described in Figure 3 were harvested. Then, single cells were separated from the tumors and injected into the right lobe of the liver. After 4-8 weeks, 18 F-FDG/ PET images (n = 6) were obtained (A). (B) The results of the PET/CT were confirmed by the radioactivity of ablated livers and H&E staining. The arrows indicate intrahepatic tumor nodules. * P < 0.05. www.oncotarget.com setting was 65-70° C. After the RFA treatment, the tumoral volume was calculated every 4 days using the following formula: width 2 × length/2. Tumors were harvested 16 days after the RFA treatment, and their weights were measured.
To produce a liver in situ tumor model [56][57][58], HCC cells were separated from subcutaneous tumors formed by MHCC97-H cells and directly inoculated into the right lobe of the liver (1 × 10 5 cells per animal). After 4-8 weeks, nude mice were injected intravenously with 100 μCi of 18F radiolabeled fluorodeoxyglucose ( 18 F-FDG), and the animals were examined using a positron emission tomography/ computed tomography (PET/CT) scanner (Philips Corp., Holland). Two-minute CT and 10-min PET scans were performed 45 min after the FDG injection. A NaI (Tl) well counter (China Atom Corp., Beijing China) was used to measure the radioactivity of organs (liver) and blood.

Transwell analysis
MHCC97-H cells were injected into nude mice to produce subcutaneous tumors. Subsequently, the tumors were divided into an RFA-treatment group and a non-RFA treatment group. Next, single cells (3000 per well) were separated from the subcutaneous tumors and were analyzed by transwell assays performed in 24-well plates chamber (Corning, Lowell, MA, USA) fitted with a polyethylene terephthalate filter membrane with 8-μm pores. The invasion cells or migration cells were measured following the methods descripted by Zhou et al. and Yang et al. [61,62].

Statistical analysis
All the statistical analyses were performed using SPSS software (version 16.0). Data are presented as the median and range. The OS rate was estimated by the Kaplan-Meier method, and groups were compared by the log-rank test. Pearson's Chi-square test was used to test the difference between the CER and DCR of the MEIS-1 high and MEIS-1 low groups. In all the analyses, a P-value < 0.05 was considered statistically significant.

CONCLUSIONS
The endogenous level of MEIS-1 seems to be associated with post-RFA treatment outcomes. The median post-RFA TTP of patients with a low MEIS-1 expression level was significantly shorter than patients with a high MEIS-1 expression level. To validate this, we used rodent tumor growth model. Overexpression of MEIS-1 enhanced the tumor rejection effect of RFA in preventing subcutaneous growth of HCC cells. The findings indicate that MEIS-1 and RFA may exert a synergetic effect, resulting in a reduction in the tumor volume. Our data suggest that the expression level of MEIS-1 may serve as a clinical predictor of post-RFA treatment outcomes in HCC patients.