FSTL1 contributes to tumor progression via attenuating apoptosis in a AKT/GSK-3 β – dependent manner in hepatocellular carcinoma

Abstract

BACKGROUND:

Several investigations have demonstrated that follistatin-like 1 (FSTL1) is implicated in the initiation and progression of diverse cancers. It remains unclear whether FSTL1 acted as a cancer-promoting gene through its overexpression in HCC.

PATIENTS AND METHODS:

We detected FSTL1 protein expression in 210 consecutive HCC cases curatively resected in our hospital between 2004 and 2007. The correlation between FSTL1 expression in HCC tissues and post-surgical prognosis of HCCs was analyzed. The in vitro experiments including apoptosis assessment, MTT, BrdU incorporation ELISA assay, Western immunoblotting, and qRT-PCR were performed to determine the impact of FSTL1 on apoptosis and proliferation abilities of HCC cells and the relevant mechanisms.

RESULTS:

FSTL1 protein was found aberrantly increased in 172 of 210 HCC tissues (81.9%) compared to adjacent liver tissues. FSTL1 overexpression was apparently associated with larger tumor size, advanced TNM staging, portal vein invasion, intra-hepatic metastases. Patients with higher FSTL1 expression in tumors suffered from the worse overall survival rate as assessed by comparison of Kaplan-Meier survival curves. Higher FSTL1 expression in HCC tissues was identified as a independent poor post-surgical prognostic predictor for HCC. Silencing FSTL1 by siRNA promoted cell apoptosis and leaded to suppression of cell viability and proliferation in MHCC97h cells. Furthermore, enforced expression of FSTL1 obtained the opposite results in Huh7 cells. Mechanistic investigation showed that FSTL1 repressed HCC cell apoptosis through AKT/GSK-3 $\beta$ /Bcl2/BAX/Bim signaling.

CONCLUSION:

These data proved that FSTL1 contributed to unfavorable post-surgical outcome of HCC patients via inhibiting cell apoptosis.

Keywords

FSTL1 HCC AKT/GSK-3β signaling prognostic factor cell apoptosis

Abbreviations

HCC	hepatocellular carcinoma
FSTL1	follistatin-like 1
GSK-3 $\beta$	Glycogen synthase kinase-3 $\beta$
IHC	immunohistochemistry
qRT-PCR	quantitative reverse-transcription-
	polymerase chain reaction

BrdU	5-Bromo-2-deoxyUridine
siRNA	short interference RNA
MTT	methyl thiazolyl tetrazolium
PBS	phosphate Buffered Saline
FBS	fetal bovine serum
PMSF	phenylmethylsulfonyl fluoride
TBST	tris-buffered saline buffer with tween

1. Introduction

Hepatocellular carcinoma (HCC), which is one of the most aggressive malignant cancers, has been the second most leading cause of cancer-related deaths in males and the sixth in females because of its high-risk post-surgical recurrence [1, 2]. Despite the tremendous advancement has been achieved by physicians and researchers on diagnostic and innovative therapy strategies for HCC during decades, the prognosis of most HCC cases remains very poor with a 5-year survival of 11% [3, 4]. Moreover, the incidence of HCC remains increasing in the last decade [5]. Herein, it is imperative to explore the underlying molecular mechanisms of HCC progression, establish the novel diagnostic markers, and develop the effective therapy regimen.

Follistatin-like protein 1 (FSTL1), also known as TSC-36 (TGF $\beta$ 1-stimulated clone 36) or FRP (follistatin-related protein), was first cloned from the mouse osteoblastic cell line MC3T3-E1 in 1993 [6]. FSTL1 gene located on chromosome 3 in humans, consisting of 11 exons spanning nearly 60KD. The FSTL1 transcript produces both the FSTL1 protein and microRNA-198 (miR-198) [7]. And expression of FSTL1 has been found widely in human and mice tissues. A growth bodying studies demonstrated that FSTL1 was involved in organ development [8, 9], inflammation [9], stem cell repopulation [10], and carcinogenesis [11, 12, 13] by mediating cell survival, proliferation, migration, and metastasis. There have been entirely different functions of FSTL1 found in diverse cancers. Bae and his colleague found that suppression of FSTL1 resulted in the cellular portion of G2/M phase in lung cancer cells via recruiting Cdk1-cyclin B1 complex, which indicated that FSTL1 was a novel potential therapy target for human lung cancer [13]. In breast cancer, Kudo-Saito reported that FSTL1 accelerated tumor bone metastasis via both up-regulating migration and invasion capacities of cancer cells and immune dysfunction induced by recruiting pluripotent mesenchymal stem-like CD45( $-$ )ALCAM( $+$ ) cells [12]. However, In the studies about nasopharyngeal carcinoma, renal cell carcinoma, ovarian and endometrial carcinogenesis, FSTL1 was found to exert the anti-tumor function. Therefore, it seems that there is complicated function of FSTL1 on cancer progression which have not understood completely.

To the best of our knowledge, no investigation was performed to detect the expression of FSTL1 in HCC tissues and identify its role on HCC progression. In the present study, we found that FSTL1 was over-expressed frequently in HCC tissues in contrast to adjacent liver tissues, which was correlated with the worse post-surgical prognosis of HCC positively. The further experiments revealed that FSTL1 functioned as a pro-oncogenic factor on HCC progression via inhibiting tumor cell apoptosis in a AKT/GSK-3 $\beta$ – dependent manner.

Table 1
Demographic information and clinical characteristics of 210 HCC patients

Clinicopathological characteristics No. of patients $\chi^{2}$ P

Low FSTL1 High FSTL1

Age (years)

$<$ 50 12 85 0.313 0.576

$\geqslant$ 50 17 96

Gender

Male 18 104 0.218 0.640

Female 11 77

HBV infection

Present 24 137 0.698 0.403

Absent 5 44

HBV DNA load (copies/mL)

$<$ 1000 12 51 1.399 0.237

$>$ 1000 12 86

Serum AFP level (ng/mL)

$<$ 400 9 35 2.065 0.151

$\geqslant$ 400 20 146

Tumor diameter (cm)

$<$ 5 22 32 44.30 $<$ 0.001

$\geqslant$ 5 7 149

Liver cirrhosis

Present 19 142 2.338 0.126

Absent 10 39

Edmondson-Steiner Classification

I $+$ II 14 87 0.001 0.983

III $+$ IV 15 94

TNM stage

I $+$ II 19 24 41.92 $<$ 0.001

III $+$ IV 10 157

Portal vein invasion

Present 14 49 5.351 0.021

Absent 15 132

Intra-hepatic metastases

Present 2 47 4.562 0.033

Absent 27 144

Clinicopathological characteristics	No. of patients	$\chi^{2}$	P
	Low FSTL1	High FSTL1
Age (years)
$<$ 50	12	85	0.313	0.576
$\geqslant$ 50	17	96
Gender
Male	18	104	0.218	0.640
Female	11	77
HBV infection
Present	24	137	0.698	0.403
Absent	5	44
HBV DNA load (copies/mL)
$<$ 1000	12	51	1.399	0.237
$>$ 1000	12	86
Serum AFP level (ng/mL)
$<$ 400	9	35	2.065	0.151
$\geqslant$ 400	20	146
Tumor diameter (cm)
$<$ 5	22	32	44.30	$<$ 0.001
$\geqslant$ 5	7	149
Liver cirrhosis
Present	19	142	2.338	0.126
Absent	10	39
Edmondson-Steiner Classification
I $+$ II	14	87	0.001	0.983
III $+$ IV	15	94
TNM stage
I $+$ II	19	24	41.92	$<$ 0.001
III $+$ IV	10	157
Portal vein invasion
Present	14	49	5.351	0.021
Absent	15	132
Intra-hepatic metastases
Present	2	47	4.562	0.033
Absent	27	144

2. Materials and method

2.1 Patients and tissue samples

The present study was carried out with the approval of the Ethical and Scientific Committees of Xi’an Jiaotong University according to the Helsinki Declaration of 1975. HCC and adjacent liver tissues recruited in this investigation were from 210 patients diagnosed with HCC. All HCC patients received hepatectomy for HCC in Department of Hepatobiliary Surgery at the First Hospital of Xi’an Jiaotong University from January 2004 to June 2007. Written informed consent was obtained from all patients. Diagnosis of HCC was identified by a experienced pathologist according to the criteria of the Chinese Society of Liver Cancer and Chinese Anti-Cancer Association. All patients included in this study had not received preoperative radiotherapy, chemotherapy, immunotherapy, TACE or target therapy before liver resection. Specimens of both HCC and adjacent liver tissues were achieved during surgery and kept in paraformaldehyde for the further assessment. The study involved the following clinicopathological parameters: age, gender, serum AFP level, HBV infection, tumor diameter, liver cirrhosis, Edmondson-Steiner Classification, TNM stage, portal vein invasion, and intra-hepatic metastases, which were obtained from the medical records. The histopathologic information was confirmed by two experienced pathologists. The follow-up information was gained from 175 of 210 HCCs (83.3%) and the median follow-up time was 42.5 months (from 3 to 75.4 months). The demographic and clinicopathological characteristics are presented in Table 1.

2.2 Immunohistochemistry staining assay

Immunohistochemistry (IHC) staining was cond- ucted as described in our previous study [14]. Both HCC tissues and adjacent liver tissues were formalin-fixed, paraffin-embedded and cut into 4 $\mu$ m-thick specimen slides. All slides were prepared at 63 ${}^{\circ}$ C for 1 hour and deparaffinized with xylenes. Then they were rehydrated in graded ethanol to distilled water. Endogenous peroxidase activity was blocked for half an hour with the methanol solution mixed with 0.3% hydrogen peroxide. The antigen was retrieved in citrate buffer and all slides were blocked at 4 ${}^{\circ}$ C overnight with 2% BSA. The section was incubated respectively with the primary antibody against FSTL1 (1:200 dilation, Catalog No.: ab11805 from Abcam) overnight at 4 ${}^{\circ}$ C. After washing with PBS three times, all slides were treated with the biotinylated secondary antibody (Boster, China). The slides were stained by the avidin-biotin-peroxidase complex (SABC) method, visualized with diaminobenzidine, and counterstained with hematoxylin. Finally, tissues were dehydrated in alcohol and xylene. Two experienced pathologists observed all slides and evaluated the IHC results independently. Staining intensity was classified into four grades: 0 (negative), 1 (weakly positive), 2 (moderately positive), and 3 (strongly positive). The section was scored as the percentage of positive cells (0 for 0–5%, 1 for 6–25%, 2 for 26–50%, 3 for 51–75%, and 4 for $>$ 75%). The final staining score was expressed by multiplying the staining intensity and the percentage of positive cells.

2.3 Cell culture

The human hepatocellular carcinoma cell lines Hep3B, SK Hep1 and HepG2 were obtained from Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences (Shanghai, China) and MHCC97h were from Fudan University (Shanghai, China). Prof. Kefeng Dou (Department of Hepatobiliary Surgery, Xijing Hospital, Fourth Military Medical University) provided Huh7 cells as a kind gift. All HCC cells were grown as monolayer in DMEM or RPMI medium (Invitrogen, Carlsbad, CA) with 10% FBS, and maintained in a humidified 5% CO2 incubator at 37 ${}^{\circ}$ C.

2.4 Establishment of FSTL1 stable transfectant clones

FSTL1 expressing plasmid (Origene, Beijing, China) and pCMV6-Entry control plasmid were transfected into Hep3B cells using TurboFect transfection reagent, respectively. After 2-week selection with Geneticin (500 g/mL), we got FSTL1 stable transfectant clones (Hep3B FSTL1 cells) and control stable transfectant clones (Hep3B Vector cells).

2.5 Establishment of FSTL1 knockdown cells

siRNA sequences targeting FSTL1 and the scramble siRNAs were both from Sigma-Aldrich (St. Louis, MO,USA). MHCC97H cells were plated at the concentration of 0.2 $\times$ 10 ${}^{6}$ per well in six-well plates and cultured overnight. Then MHCC97H cells in each wells were transfected with 100 nM siRNAs with the help of Lipofectamine RNAi MAX Reagent (Invitrogen, CA, USA) in accordance with the manufacturer’s instructions. After 48 h, MHCC97H cells were ready for the next experiments.

2.6 Quantitative reverse-transcription-polymerase chain reaction (qRT-PCR)

Total RNA was extracted from HCC cells using Trizol reagent (Invitrogen, USA) according to the suggested manufacturer’s protocol. the PrimeScript RT Master Mix (TaKaRa, Osaka, Japan) were used for converting mRNA to cDNA. Expression levels of FSTL1 mRNA were detected by SYBR Green qRT-PCR with the following primers: FSTL1: Forward 5’-C CAGAACTATGATAATGGAGACGCT-3’, Reverse 5’-TAAGATGAACTATGAACCTCCTGCC-3’; GAPDH: Forward 5’-GAAGGTGAAGGTCGGAGTC-3’, Reverse 5’-GAGATGGTGATGGGATTTC-3’. For each sample, the relative FSTL1 mRNA level was normalized to GAPDH expression.

2.7 Western immunoblotting

Western immunoblotting was carried out as described previously [15]. Briefly, HCC cells were lysed using RIPA protein extraction reagent mixed with PMSF. 30 $\mu$ g of protein samples were denatured and then separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). After transferred onto PVDF membrane, protein blots were probed with the primary antibodies against FSTL1 (1:100 dilation, Catalog No.: ab11805 from Abcam), AKT (1:1000 dilation, Catalog No.: 4691 from Cell Signaling Technology), p-AKT (1:1000 dilation, Catalog No.: 4060 from Cell Signaling Technology), GSK3 $\beta$ (1:1000 dilation, Catalog No.: 12456 from Cell Signaling Technology), p-GSK3 $\beta$ (1:1000 dilation, Catalog No.: 5558 from Cell Signaling Technology), Bcl-2 (1:1000 dilation, Catalog No.: 15071 from Cell Signaling Technology), BAX (1:1000 dilation, Catalog No.: 5023 from Cell Signaling Technology), Bim (1:1000 dilation, Catalog No.: 2933 from Cell Signaling Technology) and $\beta$ -actin (1:1000 dilation, Catalog No.: MA1115 from Boster Biotechnology). The blots were washed 3 times by tris-buffered saline buffer with tween (TBST) and incubated with the relevant secondary antibodies conjugated with HRP. Finally, protein levels were detected by the HyGLO HRP detection kit from Denville (Metuchen, NJ, USA). The expression of all target protein was normalized to $\beta$ -actin expression.

2.8 Cell viability and proliferation assessment

The MTT assay was carried out to determine cell viability at 24, 48, 72 and 96 h and BrdU incorporation was detected by ELISA assay to assess the impact of FSTL1 on HCC cell proliferation. HCC cells were inoculated into 96-well plate at the concentration of 1 $\times$ 104 per well and then stained with 100 $\mu$ l MTT (0.5 mg/ml) for 4 h at 37 ${}^{\circ}$ C. After 4 h, the medium was replaced with dimethylsulfoxide (DMSO) and the 96-well plate was softly shaken for 10 min. The OD value was measured at 570 nm by the spectrophotometer. For the cell proliferation measurement, HCC cells were grown into 96-well plates at the concentration of 5000 cells per well for 24 hours and then examined with the BrdU ELISA kit from Roche Co. (Indianapolis, IN, USA). All measurements were performed at six times.

2.9 Detection of HCC cell apoptosis

Cell apoptosis was examined by both Annexin V-FITC/PI double staining assay and Caspase 3/7 activity assessment. For Annexin V-FITC/PI double staining assay, approximately 1 $\times$ 10 ${}^{5}$ HCC cells were washed twice with cold PBS. Then Annexin V-FITC/PI incubation reagent in binding buffer was added to HCC cells and incubated for 15 min at room temperature in dark place. HCC cells were mixed with 400 $\mu$ l of binding buffer and analyzed by a FACSCalibur flow cytometry system (Becton Dickinson, San Jose, CA, USA). The ApoONE ${}^{@}$ Homogeneous Caspase-3/7 Assay kit was used to detect the Caspase 3/7 activity of HCC cells. The results were obtained by a Fluoroskan. At least three replicates was tested in each experiment, and all examination was repeated six times.

2.10 Statistical analysis

All data were presented as the mean $\pm$ standard errors of the mean. Chi-square or Fisher’s exact test was run for categorical clinical data, whereas comparisons between cell lines were conducted by Mann-Whitney U or Student’t test. The Kaplan-Meier survival curves between High FSTL1 group and Low FSTL1 group was compared by the log-rank analysis. $P<$ 0.05 was considered statistically significant.

3. Results

3.1 Aberrant FSTL1 over-expression in HCC tissues was found frequently and associated with the worse prognosis of HCC patients after liver resection

FSTL1 protein expression was detected by IHC staining in the HCC tissues and adjacent liver tissues from 210 HCC patients. As shown in Fig. 1(A), FSTL1 protein located predominantly in cytoplasm. And FSTL1 expression was increased in 172 of 210 HCC tumor tissues (81.9%) compared to adjacent liver tissues. Mann-Whitney U test verified that IHC score of FSTL1 protein in HCC tissues was significantly higher than that in adjacent liver tissues ( $P<$ 0.001, Fig. 1(B)). We divided all 210 patients into two group according to the ratio of FSTL1 expression in HCC/adjacent liver tissues: High FSTL1 group in which HCC patients expressed more FSTL1 protein in tumor tissues than adjacent liver tissues, and Low FSTL1 group in which patients had lower FSTL1 expression in tumor tissues. As described in Table 1, after analysis of clinical characteristics, FSTL1 overexpression was verified to be correlated with larger tumor diameter, advanced TNM stage, portal vein invasion, and intrahepatic metastasis notably. Among 175 HCC cases with the follow-up information, 118 patients expressed higher FSTL1 protein in HCC tissues than adjacent liver tissues who was classified as higher FSTL1 group, whereas lower FSTL1 group consisted of another 57 cases with lower FSTL1 expression in HCC tissues. The median post-surgical survival time in higher FSTL1 group was 26.89 months which was shorter than one in lower FSTL1 group (43.78 months). The 3-year survival rate of patients in higher FSTL1 group (37.3%) was remarkably less than one in lower FSTL1 group (54.4%). The similar results were found during analysis of 5-year survival rate (higher group vs. lower group: 7.6% vs. 17.8%). Comparison of Kaplan-Meier survival curves between higher FSTL1 and lower FSTL1 group revealed that the post-surgical outcome of HCC patients from higher FSTL1 group was magnificently worse than one from lower FSTL1 group (HR $=$ 1.843; 95% CI: 1.128, 3.010; $P=$ 0.015; Fig. 1(C)), which indicated that FSTL1 could mediate HCC progression positively.

Table 2
Univariate and multivariate analyses of predictive factors in HCC patients with liver resection

Clinicopathological features Univariate analysis Multivariate analysis

RR (95% CI) P value RR (95% CI) P value

Age, year ( $<$ 50 versus $\geqslant$ 50) 0.985 (0.523–1.698) 0.215 0.879 (0.497–1.538) 0.344

Gender (Female versus Male) 0.214 (0.085–1.052) 0.169 0.234 (0.099–1.106) 0.205

HBV infection 1.624 (0.897–1.689) 0.087 1.529 (0.745–1.429) 0.092

Serum AFP level (ng/mL) ( $\geqslant$ 400 versus $<$ 400) 1.981 (0.987–2.681) 0.062 1.748 (0.878–1.540) 0.056

Tumor diameter (cm) ( $\geqslant$ 5 versus $<$ 5) 2.158 (0.912–2.801) 0.051 1.908 (0.854–2.607) 0.077

Liver cirrhosis 1.533 (0.817–2.192) 0.084 1.301 (0.568–1.872) 0.095

Edmondson-Steiner Classification (III $+$ IV versus I $+$ II) 1.687 (0.859–2.325) 0.107 2.157 (1.155–3.255) 0.115

Advanced TNM staging 1.632 (1.026–3.015) 0.042 1.354 (0.898–2.411) 0.081

Portal vein invasion 2.321 (1.523–3.544) 0.004 2.011 (1.341–3.017) 0.017

Intra-hepatic metastases 3.215 (2.218–4.028) 0.001 3.008 (2.012–3.752) 0.005

Higher EDG2 in tumor tissue 2.689 (1.765–4.358) 0.001 2.018 (1.164–3.915) 0.001

Clinicopathological features	Univariate analysis	Multivariate analysis
Age, year ( $<$ 50 versus $\geqslant$ 50)	0.985 (0.523–1.698)	0.215	0.879 (0.497–1.538)	0.344
Gender (Female versus Male)	0.214 (0.085–1.052)	0.169	0.234 (0.099–1.106)	0.205
HBV infection	1.624 (0.897–1.689)	0.087	1.529 (0.745–1.429)	0.092
Serum AFP level (ng/mL) ( $\geqslant$ 400 versus $<$ 400)	1.981 (0.987–2.681)	0.062	1.748 (0.878–1.540)	0.056
Tumor diameter (cm) ( $\geqslant$ 5 versus $<$ 5)	2.158 (0.912–2.801)	0.051	1.908 (0.854–2.607)	0.077
Liver cirrhosis	1.533 (0.817–2.192)	0.084	1.301 (0.568–1.872)	0.095
Edmondson-Steiner Classification (III $+$ IV versus I $+$ II)	1.687 (0.859–2.325)	0.107	2.157 (1.155–3.255)	0.115
Advanced TNM staging	1.632 (1.026–3.015)	0.042	1.354 (0.898–2.411)	0.081
Portal vein invasion	2.321 (1.523–3.544)	0.004	2.011 (1.341–3.017)	0.017
Intra-hepatic metastases	3.215 (2.218–4.028)	0.001	3.008 (2.012–3.752)	0.005
Higher EDG2 in tumor tissue	2.689 (1.765–4.358)	0.001	2.018 (1.164–3.915)	0.001

Figure 1.

FSTL1 was found increased frequently in HCC tissues, which was associated with poor prognosis of HCC patients after liver resection. A. the FSTL1 IHC staining in adjacent liver tissues (a) was stronger apparently than that in HCC tissues (b); B. Mann-Whitney U test revealed that the score of FSTL1 IHC staining in HCC tissues was significantly higher that one in adjacent liver tissues; C. Comparison of the Kaplan-Meier survival curves demonstrated that patients with higher FSTL1 expression in HCC tissues (higher FSTL1 group) had the worse post-surgical prognosis compared to those with lower FSTL1 expression in HCC tissues (lower FSTL1 group).

As shown in Table 2, univariate analyses demonstrated that the unfavorable predictive factors consisted with higher FSTL1 expression in HCC tissues, advanced TNM staging, portal vein invasion, as well as intra-hepatic metastases. Furthermore, higher FSTL1 expression in HCC tissues, portal vein invasion, and intra-hepatic metastases were established as the independent post-surgical prognostic factors for HCC patients by multivariate analysis. These data implied strongly that FSTL1 promoted HCC progression potentially. To confirm the oncogenic function of FSTL1 on liver carcinogenesis and explore the underlying mechanism.

3.2 Silencing FSTL1 induced cell apoptosis and repressed cell growth in MHCC97h cells

As shown in Fig. 2(A), we examined FSTL1 expression in 5 kinds of HCC cell lines by both qRT-PCR and Western immunoblotting assays and found that MHCC97h cells expressed the highest level of FSTL1 among all 5 kinds of HCC cell lines. Transfection with siRNA sequences against FSTL1 resulted in a sharp decrease of FSTL1 expression in MHCC97h cells (Fig. 2(B)). Annexin V-FITC/PI double staining assay showed that knockdown of FSTL1 (MHCC97h FSTL1 siRNA cells) reinforced tumor cell apoptosis apparently (Fig. 2(C)). And Caspase 3/7 activity measurement also found that FSTL1 deletion expedited cell apoptosis of MHCC97h cells (Fig. 2(D)). As assessed by MTT, cell viability of MHCC97h cells was attenuated by getting rid of FSTL1 expression (Fig. 2(E)). Consistently, cell proliferation of MHCC97h cells were found inhibited after silencing FSTL1 expression by BrdU incorporation examination (Fig. 2(F)). Therefore, it demonstrated that FSTL1 contributed to suppressing cell apoptosis and accelerating cell growth in HCC cells.

Figure 2.

Knockdown of FSTL1 in MHCC97h cells accelerated cell apoptosis and promoted cell viability and proliferation. A. MHCC97h cells expressed the highest level of FSTL1 among 5 kinds of HCC cell lines, whereas Huh7 cells had the lowest level of FSTL1, as assessed by both qRT-PCR and Western immunoblotting; B. Both qRT-PCR and Western immunoblotting verified that siRNA sequences inhibited FSTL1 expression in MHCC97h cells notably; C. Annexin V-FITC/PI double staining assay displayed that there were more apoptotic tumor cells in MHCC97h FSTL1 siRNA cells than MHCC97h Scr siRNA cells; D. ELISA assessment showed that the MHCC97h FSTL1 siRNA cells had significantly higher Caspase3/7 activity than MHCC97h Scr siRNA cells; E. MTT assay displayed that the cell viability of MHCC97h cells was remarkably restrained by FSTL1 silencing at 48, 72 and 96 h; F. ELISA assay showed that MHCC97h FSTL1 siRNA cells possessed less BrdU incorporation than MHCC97h Scr siRNA cells notably.

3.3 Ectopic expression of FSTL1 in Huh7 cells restrained cell apoptosis and enhanced cell growth

As shown in Fig. 2(A), wild type Huh7 cells had the lowest FSTL1 expression among the all 5 HCC cells. Thus, we enhanced the expression of FSTL1 in Huh7 cells by transfecting with FSTL1 expressing plasmid, which was validated by both qRT-PCR and Western immunoblotting (Fig. 3(A)). Huh7 Vector cells had the higher Caspase 3/7 activity than Huh7 FSTL1 cells (Fig. 3(B)). And more apoptotic cells were found in Huh7 Vector cell group than Huh7 FSTL1 cell group during Annexin V-FITC/PI double staining assay (Fig. 3(C)), which illuminated that FSTL1 over-expression lead to suppression of Huh7 cell apoptosis. MTT assay revealed that cell viability of Huh7 FSTL1 cells was significantly higher than one of Huh7 Vector cells (Fig. 3(D)). BrdU incorporation examination showed that enhanced expression of FSTL1 promoted cell proliferation of Huh7 cells (Fig. 3(E)). Thus, the present results further illustrated that FSTL1 overexpression was involved closely in HCC progression.

Figure 3.

Enforced expression of FSTL1 restrained cell apoptosis and increased cell viability and proliferation. A. Both qRT-PCR and Western immunoblotting assays confirmed that FSTL1 expressing plasmid transfection leaded to up-regulation of FSTL1 expression in Huh7 cells; B. As examined by ELISA assay, Huh7 Vector cells had significantly more Caspase3/7 activity than Huh7 FSTL1 cells; C. Annexin V-FITC/PI double staining assay also demonstrated that Huh7 FSTL1 cells had the lower percentage of apoptotic cells in contrast to Huh7 Vector cells; D. MTT assay displayed that cell viability of Huh7 FSTL1 cells were obviously stronger than one of Huh7 Vector cells at 48, 72 and 96h, respectively; E. BrdU incorporation of Huh7 cells was found up-regulated by FSTL1 overexpression, as measured by ELISA assay.

Figure 4.

FSTL1 up-regulated Bcl-2 expression and inhibited the expression of BAX and Bim simultaneously via activating AKT/GSK3 $\beta$ signaling. A. Western immunoblotting showed that there was more expression of p-AKT, p-GSK3 $\beta$ and Bcl-2 and less expression of both BAX and Bim in Huh7 cells in Huh7 FSTL1 cells than Huh7 Vector cells; B. As measured by Western immunoblotting assay, it was found that knockdown of FSTL1 repressed phosphorylation of AKT and GSK3 $\beta$ , and consequently down-regulated Bcl-2 and up-regulated BAX and Bim.

3.4 The oncogenic function of FSTL1 on HCC was attributed to aberrant activation of AKT/GSK-3

\beta

signaling

To determine the mechanism that FSTL1 inhibited HCC cell apoptosis, we examined whether FSTL1 mediated AKT/GSK-3 $\beta$ signaling which has been proven to repress cell apoptosis via modulating the expression of Bcl-2, BAX, and Bim [1]. As shown in Fig. 4(A), there were more expression of p-AKT and p-GSK-3 $\beta$ in Huh7 FSTL1 cells than Huh7 Vector cells, which indicated that enforced expression of FSTL1 sensitized the activation of AKT/GSK-3 $\beta$ signaling in Huh7 cells. Western immunoblotting assay also showed that Bcl-2 expression was remarkably increased by FSTL1 overexpression, whereas the expression of both BAX and Bim was down-regulated apparently. To confirm the findings from FSTL1 overexpression in Huh7 cells, we examined the expression of Bcl-2, BAX and Bim in MHCC97h cells transfected with FSTL1 siRNA or Scr siRNA sequences. And it was found that FSTL1 knockdown leaded to down-regulation of both p-AKT and p-GSK-3 $\beta$ . Consequently, after knockdown of FSTL1, Bcl-2 expression was decreased while both BAX and Bim were up-regulated in MHCC97h cells (Fig. 4(B)). Thus, these data here supported strongly that FSTL1 up-regulated Bcl-2 and suppressed both BAX and Bim via activating AKT/GSK-3 $\beta$ signaling and then constrained HCC cell apoptosis.

4. Discussions

FSTL1 was found as a secreted glycoprotein which widely expressed in various human tissues [16, 17]. It was characterized by the presence of a follistatin-like domain with the similar sequence to the endometrial protein follistatin [18]. Previous study revealed that follistatin-like protein 1 promoted follistatin binding and inactivating activin which is a member of the TGF-b family. However, the role of FSTL1 on human pathology remain unclear. Several investigation displayed that FSTL1 was involved closely in human inflammatory diseases via regulating multiple signal pathway. Li group reported that serum FSTL1 were notably up-regulated in patients with different inflammation diseases, including rheumatoid arthritis, ulcerative colitis, systemic lupus erythematosus, Sjogren’s syndrome, systemic sclerosis, and polymyositis [19]. FSTL1 was found to mediate negatively proinflammatory cfos and IL-6, while repressing the expression of matrix metalloproteinases and prostaglandin E2 in synovial fluid of rheumatoid arthritis [20, 21]. FSTL1 was found by Gorelik group to be aberrantly over-expressed in acute macrophage activation syndrome (MAS) and treatment of MAS decrease its expression to the normal level [22].

There were different results about the role of FSTL1 on the progression of various cancer. Wilson group revealed that FSTL1 was up-regulated in prostate cancer and play a critical role during androgen receptor promoting prostate cancer progression [11]. Yoon reported that FSTL1 facilitated cellular proliferation and inhibited cell apoptosis in lung cancer cells via increasing phosphorylation of Erk1/2 [13]. Kawakami et al. reported that FSTL1 expedited bone metastasis of breast cancer by driving immune incapacity [12]. However, there were also some investigations finding out that FSTL1 exerted the anti-tumor function in diverse cancers. FSTL1 promoter was found frequently hypermethylated in nasopharyngeal carcinoma cells and clinical tumor biopsies, which leaded to down-regulation of FSTL1 in nasopharyngeal carcinoma by Zhang group. The further investigation revealed that FSTL1 down-regulation resulted in dysfunctional innate responses to surrounding macrophages which accelerating nasopharyngeal carcinoma immune evasion, though repression of FSTL1 inhibited the proliferation and migration of nasopharyngeal carcinoma cells [23]. It seemed that FSTL1 inhibited carcinogenesis of nasopharyngeal carcinoma via mediating tumor immune positively. Zhang et al. found by gene expression microarray analysis that FSTL1 up-regulation was closely associated with inhibition of the invasion and metastasis ability of pulmonary giant cell carcinoma cells [24]. The study about ovarian and endometrial cancer reported that FSTL1 functioned as a tumor suppressor through inducing cell apoptosis [25].

In this study, we found for the first time that FSTL1 was frequently up-regulation in HCC tissues in contrast to adjacent liver tissues. And overexpression of FSTL1 in HCC tissues was related significantly with unfavorable clinical features including larger tumor diameter, advanced TNM stage, portal vein invasion, and intrahepatic metastasis. Analysis of the follow-up information displayed that HCC patients with higher FSTL1 expression in HCC tissues had the worse post-surgical outcome. Multivariate analysis by Cox proportional hazards model identified higher FSTL1 expression in tumor tissues as the independent predictive biomarker for HCC patients. Next, we sought to address whether FSTL1 played a critical role on HCC progression and could be a candidate therapy target for HCC recurrence. Silencing FSTL1 expression by siRNA sequences induced cell apoptosis and inhibited cell viability and proliferation capacities of MHCC97h cells. On the other hand, enhanced expression of FSTL1 in Huh7 cells repressed cell apoptosis and strengthened cell viability, proliferation and colony-formation ability. These data supported strongly that FSTL1 promoted HCC cell growth via restraining cell apoptosis. The further mechanistic study showed that FSTL1 increased Bcl-2 expression and down-regulated BAX and Bim via inducing hyperactivation of AKT/GSK-3 $\beta$ pathway and consequently restrained tumor cell apoptosis and promoted HCC progression.

Taken together, this was the first report to demonstrate that FSTL1 was frequently up-regulated in HCC tissues compared to the adjacent liver tissues and accelerated HCC progression via inhibiting cell apoptosis in the AKT/GSK-3 $\beta$ pathway-dependent manner. The data here indicated that FSTL1 was a potential prognostic factor for HCC patients undergoing liver resection. Although it is not reliable to predict prognosis of HCC patients after surgical resection using only one kind of cancer biomarker because that individuals with HCC have several distinct biological phenotypes, the cogent data from FSTL1 knockdown experiment supported strongly that FSTL1 was a candidate target for HCC therapy.

Footnotes

Acknowledgments

This study was supported by grants from National Natural Scientific Foundation of China (81301743 and 81572733 to Xin Zheng), Research Fund for the doctoral Program of High Education of China from Ministry of Education (No. 20120201120090 to Xin Zheng), Key Science and Technology Program of Shaanxi Province (No. 2014K11-01-01-21 to Xin Zheng) and the Fundamental Research Funds for the Basic Research Operating Expenses Program of Central College sponsored by Xi’an Jiaotong University to Xin Zheng.

Conflict of interest

No conflicts of interest exist.

References

Wang

Yao

and Zheng

, RIP1 upregulation promoted tumor progression by activating AKT/Bcl-2/BAX signaling and predicted poor postsurgical prognosis in HCC, Tumour Biol 37 (2016), 15305–15313.

Torre

L.A.

Bray

Siegel

R.L.

Ferlay

Lortet-Tieulent

and Jemal

, Global cancer statistics, 2012, CA Cancer J Clin 65 (2015), 87–108.

Bruix

Gores

G.J.

and Mazzaferro

, Hepatocellular carcinoma: clinical frontiers and perspectives, Gut 63 (2014), 844–855.

Zheng

Yao

Wang

Jia

and Liu

, IL-6/STAT3 axis initiated CAFs via up-regulating TIMP-1 which was attenuated by acetylation of STAT3 induced by PCAF in HCC microenvironment, Cell Signal 28 (2016), 1314–1324.

Keckeisen

and Jayanti

V.K.

, Effectiveness of HCC treatments in real-world experiences, J Clin Oncol 30 (2012), 61.

Shibanuma

Mashimo

Mita

Kuroki

and Nose

, Cloning from a mouse osteoblastic cell line of a set of transforming-growth-factor-beta 1-regulated genes, one of which seems to encode a follistatin-related polypeptide, Eur J Biochem 217 (1993), 13–19.

Sundaram

G.M.

Common

J.E.

Gopal

F.E.

Srikanta

Lakshman

Lunny

D.P.

Lim

T.C.

Tanavde

Lane

E.B.

and Sampath

, ’See-saw’ expression of microRNA-198 and FSTL1 from a single transcript in wound healing, Nature 495 (2013), 103–106.

van Rooij

, Cardiac Repair after Myocardial Infarction, N Engl J Med 374 (2016), 85–87.

Wei

Serpooshan

Hurtado

Diez-Cunado

Zhao

Maruyama

Zhu

Fajardo

Noseda

Nakamura

Tian

Liu

Wang

Matsuura

Bushway

Cai

Savchenko

Mahmoudi

Schneider

M.D.

, van den Hoff Butte

M.J.

Yang

P.C.

Walsh

Zhou

Bernstein

Mercola

and Ruiz-Lozano

, Epicardial FSTL1 reconstitution regenerates the adult mammalian heart, Nature 525 (2015), 479–485.

10.

Holmfeldt

Ganuza

Marathe

Hall

Kang

Moen

Pardieck

Saulsberry

A.C.

Cico

Gaut

McGoldrick

Finkelstein

Tan

and McKinney-Freeman

, Functional screen identifies regulators of murine hematopoietic stem cell repopulation, J Exp Med 213 (2016), 433–449.

11.

Parris

A.B.

Grossman

Mohler

J.L.

Wang

and Wilson

E.M.

, Up-Regulation of Follistatin-Like 1 By the Androgen Receptor and Melanoma Antigen-A11 in Prostate Cancer, Prostate (2016).

12.

Kudo-Saito

Fuwa

Murakami

and Kawakami

, Targeting FSTL1 prevents tumor bone metastasis and consequent immune dysfunction, Cancer Res 73 (2013), 6185–6193.

13.

Bae

Park

K.E.

Han

Kim

and Yoon

K.A.

, Mitotic cell death caused by follistatin-like 1 inhibition is associated with up-regulated Bim by inactivated Erk1/2 in human lung cancer cells, Oncotarget 7 (2016), 18076–18084.

14.

Zheng

Song

Dou

Jia

and Liu

, CtBP2 is an independent prognostic marker that promotes GLI1 induced epithelial-mesenchymal transition in hepatocellular carcinoma, Oncotarget 6 (2015), 3752–3769.

15.

Gai

Liang

and Zheng

, Caveolin-1 is up-regulated by GLI1 and contributes to GLI1-driven EMT in hepatocellular carcinoma, PLoS One 9 (2014), e84551.

16.

Geng

Dong

Zhang

Yan

Sun

Qiao

Geng

Nakajima

Furuichi

Ikegawa

Gao

Chen

Y.G.

Jiang

and Ning

, Follistatin-like 1 (Fstl1) is a bone morphogenetic protein (BMP) 4 signaling antagonist in controlling mouse lung development, Proc Natl Acad Sci U S A 108 (2011), 7058–7063.

17.

Wilson

D.C.

Marinov

A.D.

Blair

H.C.

Bushnell

D.S.

Thompson

S.D.

Chaly

and Hirsch

, Follistatin-like protein 1 is a mesenchyme-derived inflammatory protein and may represent a biomarker for systemic-onset juvenile rheumatoid arthritis, Arthritis Rheum 62 (2010), 2510–2516.

18.

Florio

Reis

F.M.

Torres

P.B.

Calonaci

Abrao

M.S.

Nascimento

L.L.

Franchini

Cianferoni

and Petraglia

, High serum follistatin levels in women with ovarian endometriosis, Hum Reprod 24 (2009), 2600–2606.

19.

Wang

Wei

Zheng

Sun

Jin

Zhang

Liao

and Zhang

, Follistatin-like protein 1 is elevated in systemic autoimmune diseases and correlated with disease activity in patients with rheumatoid arthritis, Arthritis Res Ther 13 (2011), R17.

20.

Kawabata

Tanaka

Fujii

Umehara

Fujita

Yoshifuji

Mimori

and Ozaki

, Ameliorative effects of follistatin-related protein/TSC-36/FSTL1 on joint inflammation in a mouse model of arthritis, Arthritis Rheum 50 (2004), 660–668.

21.

Tanaka

Ozaki

Kawabata

Kishimura

Osakada

Okubo

Murakami

Nakao

and Mimori

, Potential preventive effects of follistatin-related protein/TSC-36 on joint destruction and antagonistic modulation of its autoantibodies in rheumatoid arthritis, Int Immunol 15 (2003), 71–77.

22.

Gorelik

Fall

Altaye

Barnes

M.G.

Thompson

S.D.

Grom

A.A.

and Hirsch

, Follistatin-like protein 1 and the ferritin/erythrocyte sedimentation rate ratio are potential biomarkers for dysregulated gene expression and macrophage activation syndrome in systemic juvenile idiopathic arthritis, J Rheumatol 40 (2013), 1191–1199.

23.

Zhou

Xiao

Huang

Wang

Murata

Wen

Huang

Zeng

and Zhang

, Epigenetic inactivation of follistatin-like 1 mediates tumor immune evasion in nasopharyngeal carcinoma, Oncotarget 7 (2016), 16433–16444.

24.

Zhao

Han

H.B.

and Zhang

Z.Q.

, Suppression of lung cancer cell invasion and metastasis by connexin43 involves the secretion of follistatin-like 1 mediated via histone acetylation, Int J Biochem Cell Biol 43 (2011), 1459–1468.

25.

Chan

Q.K.

Ngan

H.Y.

P.P.

Liu

V.W.

Xue

W.C.

and Cheung

A.N.

, Tumor suppressor effect of follistatin-like 1 in ovarian and endometrial carcinogenesis: a differential expression and functional analysis, Carcinogenesis 30 (2009), 114–121.

Clinicopathological features	Univariate analysis		Multivariate analysis
	RR (95% CI)	P value	RR (95% CI)	P value
Age, year ( $<$ 50 versus $\geqslant$ 50)	0.985 (0.523–1.698)	0.215	0.879 (0.497–1.538)	0.344
Gender (Female versus Male)	0.214 (0.085–1.052)	0.169	0.234 (0.099–1.106)	0.205
HBV infection	1.624 (0.897–1.689)	0.087	1.529 (0.745–1.429)	0.092
Serum AFP level (ng/mL) ( $\geqslant$ 400 versus $<$ 400)	1.981 (0.987–2.681)	0.062	1.748 (0.878–1.540)	0.056
Tumor diameter (cm) ( $\geqslant$ 5 versus $<$ 5)	2.158 (0.912–2.801)	0.051	1.908 (0.854–2.607)	0.077
Liver cirrhosis	1.533 (0.817–2.192)	0.084	1.301 (0.568–1.872)	0.095
Edmondson-Steiner Classification (III $+$ IV versus I $+$ II)	1.687 (0.859–2.325)	0.107	2.157 (1.155–3.255)	0.115
Advanced TNM staging	1.632 (1.026–3.015)	0.042	1.354 (0.898–2.411)	0.081
Portal vein invasion	2.321 (1.523–3.544)	0.004	2.011 (1.341–3.017)	0.017
Intra-hepatic metastases	3.215 (2.218–4.028)	0.001	3.008 (2.012–3.752)	0.005
Higher EDG2 in tumor tissue	2.689 (1.765–4.358)	0.001	2.018 (1.164–3.915)	0.001