Article Text
Abstract
Background: Melanoma inhibitory activity 2 (MIA2) is a novel gene of the MIA gene family. The selective expression of MIA2 in hepatocytes is controlled by hepatocyte nuclear factor (HNF) 1 binding sites in the MIA2 promotor. In contrast, in most hepatocellular carcinomas (HCC) MIA2 expression is down-regulated or lost.
Aim: In this study we examined the regulation and functional role of MIA2 in hepatocancerogenesis.
Methods and results: In HCC cell lines and tissues HNF-1 expression was lower than in primary human hepatocytes (PHH) and corresponding non-tumorous tissue, respectively, and correlated significantly with the down-regulation of MIA2 expression. Re-expression of HNF-1 in HCC cells reinduced MIA2 in HCC cells to similar levels as found in PHH. Further, MIA2 was re-expressed in HCC cell lines by stable transfection, and the generated cell clones revealed a strongly reduced invasive potential and proliferation rate in vitro. In line with these findings treatment of HCC cells with recombinant MIA2 inhibited proliferation and invasion. In nude mice MIA2 re-expressing HCC cells grew significantly slower and revealed a less invasive growth pattern. Immunohistochemical analysis of a tissue microarray containing HCC and corresponding non-cancerous liver tissue of 85 patients confirmed reduced MIA2 expression in HCC. Furthermore, MIA2 negative HCC tissue showed a significantly higher Ki67 labelling index and loss of MIA2 expression correlated significantly with more advanced tumour stages.
Conclusion: This study presents MIA2 as an inhibitor of HCC growth and invasion both in vitro and in vivo, and consequently, as a tumour suppressor of HCC. Further, our findings indicate a novel mechanism, how loss of HNF-1 expression in HCC affects tumorigenicity via down-regulation of MIA2.
Statistics from Altmetric.com
Primary hepatocellular carcinoma (HCC) is the fifth most frequent cancer and the third most common cause of cancer-related deaths in the world with an increasing incidence in the US, mainly due to HCV, HBV and alcoholic liver disease.1 Histopathological and molecular findings suggest that HCC develops through a multistep process with accumulating genetic alterations. However, the molecular pathogenesis of HCC is still not well understood. While new therapeutic strategies have significantly improved survival of patients with tumours detected at early stages, the majority of patients are still diagnosed at an advanced stage and their prognosis remains poor.2 3 These findings highlight the need for discovery of molecular markers and targets for improved diagnosis and treatment, respectively, of HCC.
Recently, we identified a new gene, melanoma inhibitory activity 2 (MIA2), and mapped the gene locus to human chromosome 14q13.4 MIA2, together with MIA, OTOR and TANGO, belongs to the novel MIA gene family sharing important structural features, significant homology and similar genomic organisation.5 6 In contrast to the other family members, MIA2 contains an additional C-terminal region of 422 amino acids without any homology to known proteins.4 Interestingly, the four members of the MIA gene family differ entirely with respect to their expression patterns. In contrast to MIA, which is exclusively expressed in cartilage but not in any other non-neoplastic tissue,7 and OTOR, which shows a highly restricted expression pattern in cochlea, eye and cartilage,8 the novel MIA related gene MIA2 is expressed exclusively in the liver.4 5 Hepatocytes were identified as the cellular source of MIA2 expression and promoter studies analysing transcriptional regulation of MIA2 revealed an HNF-1 binding site controlling hepatocyte specific expression. However, in HCC cell lines and tissues we observed a strong down-regulation or even loss of expression of MIA2.9
Based on these previous findings, the aim of this study was to investigate the regulation, functional role and clinical importance of MIA2 in hepatocancerogenesis.
MATERIALS AND METHODS
Cells and cell culture
The HCC cell lines HepG2 (ATCC HB-8065), PLC (ATCC CRL-8024), HuH-7 (JCR B0403) and Hep3B (ATCC HB-8064) were used. Cells were maintained in DMEM supplemented with penicillin (400 U/ml), streptomycine (50 μg/ml), l-glutamine (300 μg/ml) and 10% fetal calf serum (FCS; Sigma, Deisenhofen, Germany) and passaged at a 1:5 ratio every 3 days.
MIA2 expressing cell clones were established by stable transfection of HepG2 cells with a MIA2 expression plasmid.4 Plasmids were co-transfected with pcDNA3 (Invitrogen, NV Leek, The Netherlands) containing the selectable marker for neomycin resistance. Transfections were performed using Lipofectamine Plus (Invitrogen). Mock controls received pCDNA3 alone.
Primary human hepatocytes (PHH) were isolated and cultured as previously described.10 11
Tumorous and non-neoplastic human tissues
HCC tissue and non-neoplastic liver tissue of the same patient were obtained from 10 HCC patients undergoing partial hepatectomy. Tissue samples were immediately snap frozen and stored at −80°C. Informed consent was obtained from all patients and the study was approved by the local ethics committee.
HCC tissue microarray
A tissue microarray (TMA) was constructed as described previously12 13 and contained 85 formalin-fixed, paraffin-embedded human HCC tissues and corresponding non-neoplastic liver tissues of the same patient (in 80 cases). Samples were retrieved from the surgical pathology files of the Institute of Pathology of the University of Regensburg, covering a period of 12 years (1993 to 2005). All patients were treated with surgical resection or liver transplantation. There were 69 (81.2%) males and 16 (18.8%) females with a median age of 61 years (range 20–85 years). Alcohol-related liver disease was defined if long-lasting excessive alcohol consumption (at least 80 g/d in men and 40 g/d in women for more than 10 years) was reported. In 32 patients (37.6%) no underlying cause of liver disease was found or could be defined, respectively. In all of these patients chronic viral hepatitis was excluded by testing for hepatitis B surface antigen, anti-HBc and third-generation hepatitis C antibody ELISA. However, it may be difficult to retrieve valid information on the amount and duration of alcohol consumption in retrospective studies, especially in patients with alcohol-related diseases as cirrhosis. In most cases, data stated by the patients or their relatives tend to be lower than the actual amount consumed, making it more than likely that the percentage of HCC patients with alcohol abuse was actually higher than reported in our study. Further clinicopathological patient characteristics are summarised in table 1.
Informed consent was obtained from all patients and the study was approved by the local ethics committee.
Tumour cell inoculation and measurement of tumour growth in NMRI (nu/nu) mice
A model of inoculation of tumour cells into NMRI (nu/nu) mice to monitor tumour growth in vivo was performed as described previously.14 Briefly, HCC cell clones were harvested after incubation with PBS containing 0.05% trypsin and 0.04% EDTA (Sigma). Tumour cells were washed twice with serum free DMEM (Dulbecco’s modified Eagle medium; Gibco) at room temperature and resuspended in DMEM at a concentration of 1×107 cells/ml.
For each of the four tumour cell lines a group of 10 NMRI (nu/nu) mice with a mean body weight of 32 g was formed. All mice were injected subcutaneously with a cell suspension of 0.1 ml containing 1×106 cells of a single cell line. Tumour growth kinetics were recorded on a regular basis by measurement of tumour diameters at the inoculation site (region of the thoracic mammary fat pad) with an electronic calliper. Tumour areas were calculated as the product of two perpendicular diameters, one measured across the greatest width of the tumour. For ethical reasons the mice were killed at day 15 after the first tumours underwent ulceration, and the tumours were taken out and stored for subsequent analysis.
Expression analysis
Isolation of total cellular RNA from cultured cells and tissues and reverse transcription were performed as described previously.15 Quantitative real time-PCR was performed with specific primers for MIA29 and HNF-1 (HNF-1 forward: 5′- CCT GTC CCA ACA CCT CAA CAA; HNF-1 reverse: 5′- TCT CTC GCT CCT CCT TGC TA) applying LightCycler technology (Roche, Mannheim, Germany) as described.9
Protein analysis
Protein extraction and western blotting were performed as described previously,16 applying the following primary antibodies: polyclonal anti-MIA2 antibody,9 or anti-β-actin antibody (Sigma).
Immunohistochemical staining of 5 μm sections of the TMA blocks was performed using polyclonal anti-MIA2 antibody and an indirect immunoperoxidase protocol according to the LSAB2-kit (Dako, Hamburg, Germany) as described.9 A surgical pathologist (F.B.) performed a blinded evaluation of the stained slides. Staining intensity was estimated using a semi-quantitative three-step scoring system (0–2+): 0, negative; 1+, weak positive; 2+, strong positive. In consideration of the small amount of tissue of each individual tumour on the TMA, even weak MIA2 immunoreactivity (1+) was considered positive.
MIB1 was analysed applying anti-Ki67 antibody (rabbit monoclonal, clone MIB1; Dako; 1:10, final concentration 5 μg/ml).
Antibody binding was visualised using AEC-solution (LSAB2-Kit, Dako Cytomation GmbH, Hamburg, Germany). Finally, the tissues were counterstained by haemalaun.
Expression of recombinant MIA2 protein
The MIA2 cDNA was cloned into the vector pIVEX2.3-MCS (Roche Applied Science). The expression vector was used in the rapid translation system, a cell-free E coli-based protein transcription/translation system (Roche Applied Science) following the manufacturer’s description. The correct expression of rhMIA2 was analysed by western blotting.
Transfection of HCC cell lines
Applying the lipofectamin plus method (Invitrogen, Carlsbad, USA) small interfering RNA (siRNA; Hs_TCF1_1 HP siRNA from Qiagen, Hilden, Germany) was transiently transfected into HCC cells to deplete HNF-1 expression.
Cell proliferation assays
Cell proliferation was measured using the XTT assay (Roche, Mannheim, Germany) as described.15 17
Invasion assays
Invasion assays were performed using the BD BioCoat™ Matrigel™ Cell Environments, Invasion Chambers (BD Biosciences, Discovery Labware, Bedford, MA, USA) as described.17
Apoptosis assays
For detection of apoptosis, cells were stained simultaneously with FITC-conjugated annexin V and propidium iodide (PI) (both from Pharmingen) as described previously.11 Total numbers of apoptotic cells were determined by calculation of annexin V+ and PI− cells together with annexin V+ and PI+ cells. Analysis of data was performed using the software WinMDI (Version 2.8, http://facs.scripps.edu).
Cell cycle analysis
Nuclear DNA content was measured by using propidium iodide staining and fluorescence-activated cell sorting (FACS) analysis as described previously.18 Briefly, 106 cells were fixed in ice-cold methanol (70%) for 1 h on ice, washed twice with PBS, and resuspended in PBS. Fixed cells were treated with RNase (0.01 mg/ml; Roche Molecular Biochemicals) for 30 min at 37°C, stained with 50 μg/ml propidium iodine (PI; Sigma), and kept in the dark on ice for 30 min before analysis. Cell cycle analysis was carried out using an EPICS XL-MCL flow cytometer (Coulter Immunotech). Data were analysed with the Multicycle software (Becton Dickinson, Heidelberg, Germany).
Statistical analysis
p values <0.05 were considered significant. Statistical analyses were performed using SPSS version 10.0 (SPSS, Chicago, IL, USA). Results are expressed as mean ± SD (range) or per cent. Comparison between groups was made using the Student unpaired t-test. Correlation between parameters was calculated with the Spearman test. Contingency table analysis and two-sided Fisher’s exact tests were used to study the statistical association between clinico-pathological and immunohistochemical variables.
RESULTS
Reduced HNF-1 expression in HCC causes down-regulation of MIA2
Initially, we wanted to study the molecular mechanisms responsible for the strong reduction or loss of MIA2 expression that we have recently described in human HCC cell lines in vitro and HCC tissue in vivo.9 Since we have previously shown the regulation of MIA2 expression by HNF-1,4 we started to analyse the expression of this transcription factor in four different HCC cell lines compared to primary human hepatocytes (fig. 1A). Interestingly, strong reduction of HNF1 mRNA expression was found in all four HCC cell lines. Similarly, analysis of HNF1 mRNA expression in tumour tissue of ten HCC patients revealed a reduction of HNF-1 expression in all cases as compared to corresponding non-neoplastic liver tissue. Overall, a 6.4±2.1-fold reduction was observed in HCC-tissue (fig. 1B). As previously shown, MIA2 was also down-regulated in all ten HCC samples (9.9±4.5-fold), and reduction of HNF1 expression correlated significantly with the down-regulation of MIA2 expression in HCC tissues (r = 0.71; p = 0.023) (fig. 1C). To further analyse whether reduced HNF-1 expression levels are responsible for the down-regulation of MIA2 expression in HCC, we transiently transfected HCC cell lines with an expression plasmid for HNF-1, and, subsequently, analysed MIA2 mRNA expression in comparison to mock transfected and non-transfected controls. Transfection of HNF-1 was sufficient to re-induce MIA2 mRNA expression in HCC cells to similar levels as found in primary human hepatocytes. Figure 1D depicts results obtained with HepG2 cells. In a complementary approach, depletion of HNF-1 expression in HepG2 cells using small interfering RNA (siRNA) resulted in a significant down-regulation of MIA2 expression compared to siRNA control and cells non-transfected controls (fig. 1E). According data were obtained also with the other HCC cell lines studied (data not shown). In summary, these findings indicate, that reduced MIA2 expression in HCC may be caused by loss of HNF-1 expression in HCC.
Re-expression of MIA2 in HCC cells
To gain insight into the functional role of the down-regulation of MIA2 in HCC, MIA2 was re-induced in the HCC cell line Hep3B by stable transfection with a MIA2 expression vector containing the full-length MIA2 cDNA under the control of the cytomegalovirus (CMV) promotor (clones 1 and 2). The vector without an insert was used as a control (mock). Quantitative RT-PCR analysis revealed a strong induction of MIA2 mRNA expression in both cell clones, whereas no (change of) MIA2 expression was seen in mock transfected cell clones (fig. 2A). Up-regulation of MIA2 expression in the cell clones was confirmed on the protein level by western blot (fig. 2B). Of note, changes in MIA2 mRNA expression match changes observed on the protein level. Immunofluorscence analysis confirmed re-expression of MIA2 in clone 1 (fig. 2C) and clone 2 (data not shown).
Functional changes after MIA2 re-expression in HCC cells in vitro
To further characterise the role of MIA2 in HCC cells, we performed functional in vitro assays with MIA2 expressing cell clones in comparison to mock-transfected cells and the parental cell line. Proliferation was significantly impaired in MIA2 re-expressing cell clones (fig. 3A).
Based on this finding, we performed FACS analysis of propidium iodide-stained MIA2 expressing cell clones and control cells to investigate the cell cycle. It appeared that the fraction of MIA2 cell clones that retained in the G0/G1 phase (clone 1 (c1): 35.9±1.9% and clone 2 (c2): 38.8±2.2) was significantly higher than the fraction found within both mock-transfected (24.1±1.1%; p = 0.032 vs. c1, and p = 0.026 vs. c2) and non-transfected control cells (26.2±1.2%; p = 0.048 vs. c1, and p = 0.036 vs. c2). Further, FACS analysis revealed that fewer MIA2 re-expressing cells were in the S-phase (c1: 32.2±1.2% and c2: 25.0±0.6%) as compared to mock-transfected controls (39.4±1.2%; p = 0.047 vs. c1, and p = 0.008 vs. c2), while the percentage of cells detected in the G2/M-phase did not differ significantly between MIA2 cell clones and controls.
Further, annexin V/propidium iodide–FACS analysis revealed only marginal apoptosis (<3%) in both MIA2 expressing cell clones and control cells, respectively, indicating that MIA2 expression does not affect apoptosis in HCC cells in vitro (fig. 3C).
To gain further insight into the functional role of MIA2 we performed matrigel invasion assays. Interestingly, HCC cells stably transfected with MIA2 displayed significantly decreased invasiveness compared to mock transfected and non-transfected control cells (fig. 3D).
To confirm the functional effects of MIA2 on HCC cells and to get first insights into a potential therapeutic application of MIA2, HCC cells were treated with recombinant MIA2 (rMIA2). Addition of rMIA2 revealed the same inhibitory effect on cell proliferation (fig. 3E) and invasion (fig. 3F) as found in the stably MIA2 expressing cell clones (fig. 3A and D).
Tumorigenicity of MIA2 re-expressing HCC cells in vivo
To test the effect of MIA2 on tumour growth in vivo HCC cells stably expressing MIA2 were injected subcutaneously into nude mice, revealing significantly impaired growth compared to mock or non-transfected cells (fig. 4A). Ten days after injection of the HCC cells all mice from the mock and non-transfected control groups but only 6 of 10 (clone 1) and 4 of 10 (clone 2), respectively, animals from the MIA2 re-expressing groups developed tumours. After 15 days, also all mice from the clone 1 group and six mice from the clone 2 group developed tumours, although these tumours were significantly smaller than from the other two groups (17.7±5.5 mm3 (clone 1) and 7.9±3.1 mm3 (clone 2) vs. 57.5±12.1 mm3 (non-transfected) and 59.8±14.9 mm3 (mock); p = 0.017 (clone 1) and p = 0.003 (clone 2) in comparison to mock, and p = 0.008 (clone 1) and p = 0.0009 (clone 2) in comparison to non-transfected cells). These data indicate that rescue of MIA2 expression in HCC cells inhibits proliferation of HCC cells in vivo.
Furthermore, histopathological analysis revealed that tumours developed from MIA2 re-expressing HCC cells showed diffuse and invasive growth, while tumours from mock or non-transfected cells showed a more nodular growth. Representative pictures are shown in fig. 4B.
In summary and in line with the in vitro results, data obtained in nude mice indicate that reduced MIA2 expression in HCC cells induces tumorigenicity of HCC cells.
MIA2 expression in human HCC
Next, we analysed MIA2 protein expression in a series of 85 HCC patients and corresponding non-tumorous tissue of the same patients (n = 81) using tissue microarray technology. Investigation of MIA2 protein expression was informative in all 85 HCC tissue specimens and in 98.8% (80/81) of non-tumorous tissue samples, respectively. The clinico-pathological features and immunohistochemical results of the HCC tumour cohort are summarised in table 1.
In 32.9% (28/85) of the HCC, no MIA2 expression was detectable. In contrast, MIA2 protein expression (1+ or 2+) was found in all (n = 80) non-cancerous tissue samples. It is noteworthy that in 62.5% (50/80) of cases the MIA2 staining signal was lower in HCC as in corresponding non-tumorous liver tissue. Representative MIA2 immunostaining patterns are shown in fig. 5A. Figure 5B summarises the immunohistochemistry results for HCC and surrounding non-cancerous liver tissue on the TMA. MIA2 was significantly reduced in HCC (p<0.0001) compared to non-cancerous tissue.
Matched data of mRNA expression and semi-quantitative protein expression analysed on the microarray were available from ten HCC patients. In all cases MIA2 mRNA expression was reduced in HCC compared to non-tumorous liver tissue (41.0±9.6; p = 0.0002). MIA2 mRNA expression was significantly more down-regulated in five HCC cases, where also immunohistochemistry (IHC) revealed down-regulation of MIA2 protein expression (21.1±11.4%) compared to five cases, where the MIA2 immunosignal appeared similar in HCC and non-tumorous tissue (60.9±9.0%; p = 0.026). These findings indicate that strong down-regulation of MIA2 in HCC is accurately detected by IHC. On the other hand, less pronounced differences between HCC and non-tumorous liver tissue may be missed, and probably, MIA2 is down-regulated in even more cases than now shown by immunohistochemistry.
For descriptive data analysis, all relevant variables were compared with MIA2 IHC (table 1). Loss of MIA2 immunoreactivity was significantly associated with higher tumour stage (p = 0.036).
Furthermore, MIA2 negative HCC had a significant higher proliferation rate (MiB-1 Index) compared to MIA2 positive HCCs (p = 0.008; fig. 5C).
No correlation could be found with MIA2 expression and age, gender, histological grade, size and aetiology of the underlying liver disease.
DISCUSSION
Recently, we identified a novel member of the MIA gene family, MIA2 and described tissue-specific expression in hepatocytes.4 Furthermore, we showed that MIA2 is down-regulated or lost in HCC.9 In this study, we demonstrate that loss of the transcription factor HNF-1 causes at least, in part, the transcriptional down-regulation of MIA2 in HCC. Furthermore, we show that restoration of MIA2 expression in HCC cells markedly inhibits growth and invasion both in vitro and in vivo. Additionally, we find that the loss of MIA2 expression is associated with higher tumour stage and tumour cell proliferation. Thus, this work provides direct evidence that MIA2 can be considered as a tumour suppressor in HCC. Furthermore, our data indicate that MIA2 is a potential marker for HCC prognosis as well as an attractive therapeutic target.
So far, we can only speculate on the molecular mechanisms how MIA2 exhibits its anti-tumorigenic effects. MIA2 is exclusively expressed in hepatocytes and analysis of the amino acid sequence of MIA2 as well as preliminary experiments analysing the supernatant of hepatocytes indicate that MIA2 is a secreted protein (unpublished observation). It is noteworthy that treatment of HCC cells with recombinant MIA2 exhibited similar effects on proliferation and invasiveness as re-expression of MIA2 by stable expression. This finding indicates that MIA2 acts at least in part in an autocrine fashion. In further studies, potential receptor(s) as well as downstream signalling pathways have to be determined. Furthermore, the therapeutic potential of MIA2 has to be further investigated.
Previously, we have shown that hepatic MIA2 expression is increased in patients with liver fibrosis or cirrhosis, respectively.4 Furthermore, we demonstrated that transforming growth factor-beta (TGF-β) induced MIA2 expression in hepatocytes in vitro,4 and it is well known that TGF-β signalling is deranged in HCC. Therefore, it may be speculated that impaired TGF-β signalling contributes to the reduced MIA2 transcriptional activity in cancerous liver tissue.
Furthermore, we have demonstrated that hepatocyte-specific expression of MIA2 is controlled by HNF-1 binding sites in the MIA2 promotor. HNF-1 belongs to the liver-enriched transcription factors (LETFs) which have been identified to bind to the promotor or enhancer regions of various hepatic functional genes. Several studies have indicated that HNF1 participates in the control of cytodifferentiation and transcriptional alteration of this factor has been observed between normal and cancerous tissue.19 20 Our finding of reduced HNF-1 expression in HCC cell lines and tissues is in accordance with these previous studies. However, it was once reported that expression of HNF-1 was increased in four of six HCC samples.21 The discrepancy between our results and this previous report may be explained by the relatively small number of the tumour samples analysed. A tumour-suppressor function for HNF-1 is supported by the phenotypes observed in two HNF1-deficient mouse strains. These mice have either a pronounced liver enlargement caused by an increased proliferation of hepatocytes22 or a liver enlargement with central lobular hypertrophy and degeneration of individual hepatocytes.23 Furthermore, germline mutations of transcription factor (tcf) 1 gene encoding HNF1 have been found in hepatic adenomas.24
However, so far, a direct effect of HNF-1 on the tumorigenicity of cells has not been suggested. Here, our findings indicate a novel mechanism, how (loss of) HNF-1 expression in HCC affects tumorigenicity via (down)regulation of MIA2.
Acknowledgments
We are indebted to Sibylla Lodermeyer for excellent technical assistance.
REFERENCES
Footnotes
Funding: The study was supported by grants from the German Research Association (DFG) to A.K.B., and the Medical Faculty of the University of Regensburg (ReForM) to C.H., A.H. and A.K.B.
Competing interests: None.
Linked Articles
- Digest