SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

A powerful way to identify driver genes with causal roles in carcinogenesis is to detect genomic regions that undergo frequent alterations in cancers. Here we identified 1,241 regions of somatic copy number alterations in 58 paired hepatocellular carcinoma (HCC) tumors and adjacent nontumor tissues using genome-wide single nucleotide polymorphism (SNP) 6.0 arrays. Subsequently, by integrating copy number profiles with gene expression signatures derived from the same HCC patients, we identified 362 differentially expressed genes within the aberrant regions. Among these, 20 candidate genes were chosen for further functional assessments. One novel tumor suppressor (tripartite motif-containing 35 [TRIM35]) and two putative oncogenes (hairy/enhancer-of-split related with YRPW motif 1 [HEY1] and small nuclear ribonucleoprotein polypeptide E [SNRPE]) were discovered by various in vitro and in vivo tumorigenicity experiments. Importantly, it was demonstrated that decreases of TRIM35 expression are a frequent event in HCC and the expression level of TRIM35 was negatively correlated with tumor size, histological grade, and serum alpha-fetoprotein concentration. Conclusion: These results showed that integration of genomic and transcriptional data offers powerful potential for identifying novel cancer genes in HCC pathogenesis. (HEPATOLOGY 2011;) © 147.

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related mortality worldwide. New insights into the pathogenesis of this lethal disease are urgently needed. Chromosomal copy number alterations (CNAs) can lead to activation of oncogenes and inactivation of tumor suppressors in human cancers.1 Thus, identification of cancer-specific CNAs will not only provide new insight into understanding the molecular basis of tumorigenesis but also facilitate the discovery of new cancer genes.2, 3 Using traditional methodologies, frequent DNA copy number gains at 1q, 8q, 7q, 17q, and 20q and losses at 4q, 8p, 13q, 16q, and 17p have been identified in HCC.4, 5 Although these technologies adequately detect some of the known candidate oncogenes or tumor suppressors, such as CHD1L (1q21)6 and DLC1 (8p22),7 their resolutions are limited for obtaining a comprehensive view of whole-genome copy number changes. High-density single-nucleotide polymorphism (SNP) arrays now provide the possibility of defining genome-wide copy number changes.8, 9 Additionally, there has been little progress in determining specific genes targeted by various common copy number gains and losses, in part due to limited availability of complementary transcriptional data on sufficient numbers of specimens to focus on a small list of candidate genes. Although several studies have been conducted to define potential cancer genes through combined analyses of genomic alterations and transcriptomes in HCC, they are constrained by the use of different sets and small sizes of tumor samples or by the use of relatively lower-resolution platforms.10-12 In this study we applied a whole-genome SNP 6.0 array to define a comprehensive copy number profile of 58 paired HCC and nontumor tissues. We further identified potential cancer genes by adopting a combined approach to define somatic CNAs and transcriptomes in the same set of paired HCC specimens.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Methods and any associated references are available in the Supporting Materials.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Identification of Significant Somatic CNAs in HCC Genomes.

We analyzed the hybridization signal intensities of 58 paired HCC and nontumor tissues from the same individuals to identify regions of somatically generated CNAs. A total of 2,206 CNAs were identified in the 58 HCC genomes. A genome-wide view of segmented copy numbers revealed that most chromosomal arms undergo either copy number gain or loss in a large proportion of the samples (Fig. 1A). The 2,206 CNAs spanned from 0.28 kb to 30 Mb in size (median, 6.22 Mb). There was a mean of 38 CNAs per HCC genome and copy number gains were more commonly observed than losses (1.9:1).

thumbnail image

Figure 1. Genome-wide CNAs in HCC genomes. (A) Unsupervised hierarchical clustering of Affy6 copy number data from 58 HCC specimens. Each chromosome is displayed in a separate column in genomic order, and tumors are ordered from top to bottom according to the hierarchical clustering of CNAs. Red, copy number gains; blue, copy number losses. (B) Significant regions of copy number gains and losses. Chromosomes are displayed along the horizontal axis in ascending order, and the frequency of each amplification or deletion is displayed along the vertical axis. Labels on the top or bottom denote the position of peaks of the most frequently altered regions in each chromosome.

Download figure to PowerPoint

To find evidence of driver alterations in tumor genomes we further evaluated the recurrent regions of copy number gains and losses using the following parameters: the minimum physical length of putative CNAs was more than 100 kb; the CNAs was present in at least three tumor samples; and finally, the overlapping common regions among multiple tumors were calculated. Accordingly, a total of 1,241 significant CNAs were obtained, including 963 amplifications and 278 deletions (Fig. 1B). These regions were highly concordant with previous findings, including the recurrent gains at 1q, 6p, 7q, 8q, 11q, 17q, and 20q and recurrent losses at 4q, 8p, 16q, and 17p.13, 14 Importantly, a large number of novel CNAs were also discovered, including gains at 1p, 2p, 3q, 6q, 8p, 9p, 10p, 10q, 13q, 16p, 16q, 18p, 19p, and 22q and losses at 3q, 4p, 8q, 12p, 14q, and 18q. Notably, the major regions of copy number gains were observed to reside within 1q and 8q, accounting for 48.2% (464/963) of all gains, whereas the major regions of losses were found within 4q and 8p, accounting for 51.3% (143/278) of all copy number losses.

The most frequently amplified region observed was 8q24.21-24.22, which occurred in 53.4% of samples and targets the known oncogenes MYC, DDEF1, and MLZE. Additionally, two other recurrent amplified regions at chromosome 8q were found to be 8q21.13, which targets hairy/enhancer-of-split related with YRPW motif 1 (HEY1), and 8q24.3, which contains several genes, including SCRIB and BOP1 (Table 1). Consistent with previous studies, we found peaks of amplified regions targeting MET on 7q31.2,15 TERT on 5p15.33,16 and SRC on 20q11.2317 as well as an interstitial 11q13.2-13.3 amplification spanning CCND1.18 Other amplifications included 1q21.2-q21.3, which spans MCL1 and LASS2, in addition to ARNT, which is a previously reported target of this genomic amplification.19 The most commonly deleted loci included DLC1 at 8p23.1-8p22 and a previously unreported tripartite motif-containing 35 (TRIM35) deletion at 8p21.2-8p21.1. Several other frequent deletions were also observed, including a deletion targeting SERPINA5 at 14q31.1-32.13 and a larger 17p13.3-13.1 deletion spanning PER1, ENO3, and TP53 (Table 1).

Table 1. Top Focal Regions of Copy Number Gains and Losses in HCC
CytobandFrequency (%)StartEndLength, (Mb)Max/Min Inferred Copy No.Number of GenesKnown Oncogene/Tumor SuppressorRepresentative Candidate(s)
  1. Start = first base pair location in the CNA region. End = last base pair location in the CNA region.

Amplification
8q24.21-24.2253.4130,892,414131,657,1860.733.194  DDEF1
1q21.2-21.350148,262,590149,766,4411.433.148 ARNTMCL1, LASS2
8q21.1348.380,669,61481,158,3510.473.155  HEY1
8q24.348.3143,851,512146,268,9602.33.0796 SCRIBBOP1
1q24.1-24.244.8165,028,985166,069,92613.0315 MPZL1
7q31.212.1116,224,802116,261,5250.046.951  MET
10p12.112.125,394,30526,535,8241.093.332  
11q13.2-13.312.169,165,11169,278,0110.119.844  CCND1
19q13.4212.160962241620762231.063.0134 
22q13.1-13.212.136638926395314512.763.0464 
20q11.2310.335,169,40735,700,1690.53.0410 SRC
5p15.3310.368,5321,339,9531.213.125 TERT
6p12.310.347,835,01848,531,4030.663.235  
17q25.310.376,402,08578,643,0892.142.8474 
Deletion
8p23.1-222012,605,26313,852,1491.191.196  DLC1
8p21.2-21.117.225,122,98029,184,1353.871.1737 TRIM35
4q13.213.869,065,41569,170,2520.11.073  
14q31.1-32.138.680,694,31694,566,78913.231.25103SERPINA5
16q24.110.384,798,84585,427,6570.61.28  
17p13.3-13.16.92,866,71711,192,9507.941.21224TP53PER1, ENO3

Integrative Analysis of CNAs and Gene Expression Data to Discover Candidate Cancer Genes in HCC.

To discover candidate cancer genes in regions of CNAs, we performed an integrated analysis of CNAs and gene expression data. First, we profiled genome-wide gene expression for 49 paired HCC and nontumor tissues and a total of 1,409 differentially expressed genes (DEGs) were obtained. Subsequently, the list of genes located in the 1,241 aberrant regions was matched with the DEG list. The results showed that a set of 362 genes were differentially expressed in the aberrant regions, with 228 exhibiting increased expression in the amplified regions and 134 showing decreased expression in the deleted regions (Fig. 2A; Supporting Table 1).

thumbnail image

Figure 2. Integrated analyses of CNAs and gene expression data. (A) Clustered heatmap of expression values for 362 DEGs in the regions of CNAs. A hierarchical clustering map of the normalized gene expression values for the 362 genes identified as being dysregulated in regions of CNAs is shown for 49 paired HCC and matching nontumor tissues, with rows representing samples and columns representing genes. (B) Interaction network analysis of 362 DEGs. The 362 altered genes were connected in a network based on prior known protein-protein interactions and signaling pathways, as described in Materials and Methods. Nodes represent genes (ellipse, altered genes; diamond, linker genes), and edges indicate the biological relationship between the nodes, as annotated in the graph. Red, up-regulated genes in regions of copy number gains; blue, down-regulated genes in regions of copy number losses; gray, linker genes that are not altered in HCC but are statistically enriched for interaction with the altered genes.

Download figure to PowerPoint

To further define the cellular processes and pathways in which these 362 DEGs are involved, we performed gene ontology (GO) enrichment analysis. Overall, the 362 genes were enriched for cancer-dominant functions, such as DNA replication / messenger RNA (mRNA) processing, cell cycle/cell proliferation, protein transport/protein folding, and cell adhesion/cell motility (Supporting Fig. 1). Additionally, to determine the regulatory relationships of these genes and the key players in HCC neoplastic processes, we performed a network analysis to generate an interaction network containing relevant biological information for the 362 genes. The resulting network shows a high degree of connectivity that further supported the existence of biologically related functions (Fig. 2B). According to the connected subgraphs and their GO terms, the functional modules were enriched for the cell cycle, cell adhesion, regulation of the actin cytoskeleton, complement and coagulation cascades, long-term depression, WNT signaling pathways, PPAR signaling pathways, and pathways in cancer (Supporting Table 2). Some critical genes are located in these modules, including GNAO1, GNAZ, PLCB1, CDC25B, LAMC1, FOS, ETS1, and SHC1 (Fig. 2B), suggesting that these genes may have important roles in the pathogenesis of HCC.

Identification of Putative Oncogenes and Tumor Suppressors in HCC.

To further determine which genes among these 362 DEGs represent novel cancer genes, we next examined the potential driver roles of the candidates in hepatic carcinogenesis. By combining the results of interaction network analysis with functional suggestions obtained from PubMed data mining, a set of 20 potential genes in frequently aberrant regions (amplification ≥12 samples and deletion ≥4 samples, respectively) were chosen for further functional assessments (Supporting Table 3). Specifically, eight genes (DDEF1, SHC1, HAX1, RAD21, MPZL1, YWHAZ, SERPINA5, and GNAO1) were selected according to the results of network analysis (Fig. 2B). Four genes selected (PEA15, ILF2, MT1G, and PER1) have been identified previously in HCC.20-23 Eight genes selected (SNRPE, C1ORF2, BOP1, HEY1, DUSP12, C8ORF4, SLC25A4, and TRIM35) have been reported in other tumors.24-31 First, the mRNA expression levels of these 20 genes were confirmed by quantitative real-time polymerase chain reaction (q-PCR) analysis in 47 of the 49 paired HCC samples (Supporting Fig. 2A). The results indicated that there was greater than two-fold up-regulation of HEY1 in 42.6% of HCC tumors (20/47), whereas there was greater than two-fold down-regulation of TRIM35 in 60% of HCC tumors (27/47) compared with matched noncancerous tissues. Additionally, in order to choose suitable cell line tools for the following functional assessments, the expression levels of these 20 genes in eight liver cancer cell lines were determined by q-PCR (Supporting Fig. 3).

To investigate whether these genes are involved in liver tumorigenesis, they were overexpressed using a lentivirus vector in SMMC-7721 and Huh-7 cells, which was confirmed by q-PCR (Supporting Fig. 4). The results showed that of the seven deleted genes, TRIM35 was capable of significantly inhibiting the in vitro cell proliferation and in vivo tumor growth of both SMMC-7721 and Huh-7 cells, whereas of the 13 amplified genes, HEY1 and SNRPE were capable of significantly promoting the proliferation of both SMMC-7721 and Huh-7 cells in vitro and in vivo (Fig. 3A-D). Consistent with the above results, knockdown of TRIM35 was observed to markedly promote HCC cell proliferation, whereas knockdown of HEY1 and SNRPE significantly suppress HCC cell proliferation, based on small interfering RNA (siRNA) analyses in HepG2 cells (Fig. 3E). Taken together, wededuced that TRIM35 may be a new tumor suppressor candidate and that HEY1 and SNRPE may be novel putative oncogenes in HCC.

thumbnail image

Figure 3. Effects of TRIM35, HEY1, and SNRPE on the in vitro cell proliferation and in vivo growth of HCC cells. (A,B) Representative results of CCK-8 assays for the effects of TRIM35, Hey1, and SNRPE genes on the in vitro proliferation of SMMC-7721 and Huh-7 cells by lentivirus-mediated overexpression. (C) Colony formation assays of the effects of three genes on the proliferation of SMMC-7721 cells by lentivirus-mediated overexpression. (D) The effects of three genes on the growth abilities of SMMC-7721 cells in xenograft models of nude mice (n = 6) as determined by tumor weight. (E) Effects of siRNA mediated-knockdown of three genes on the proliferation of HepG2 cells by CCK-8 assays. All the results are shown as the mean ± standard error of the mean (SEM). *P < 0.05; **P < 0.01; ***P < 0.001. OD, optical density.

Download figure to PowerPoint

Given that TRIM35 is often deleted and down-regulated in HCC, and functions as a novel tumor suppressor in hepatocarcinogenesis, we further sought to determine the mechanisms of how TRIM35 could inhibit the proliferation of HCC cells. As shown in Fig. 4, TRIM35 was observed to not only arrest the cell cycle at G2/M phase but also to promote cell apoptosis of HCC cells overexpressing TRIM35, indicating that one mechanism by which TRIM35 inhibits HCC cell proliferation is through impeding the cell cycle progression or inducing apoptotic cell death.

thumbnail image

Figure 4. TRIM35 arrested cell cycle in G2/M phrase and induced cell apoptosis of HCC cells. (A) Representative results of cell cycle analyses by FACS. TRIM35 arrested the cell cycle in G2/M phase of both SMMC-7721 and HUH-7 cells. (B) Statistical analyses of the results of cell cycle analyses. (C) The effects of TRIM35 on the apoptosis of both SMMC-7721 and HUH-7 cells; the levels of caspase-3/7 were measured using the Caspase-Glo 3/7 assay. All results are shown as the mean ± SEM. *P < 0.05; **P < 0.01.

Download figure to PowerPoint

Correlation of TRIM35 Expression with HCC Clinicopathologic Features.

To further determine the clinicopathologic significance of TRIM35 in HCC, we performed immunohistochemical (IHC) analysis of TRIM35 in a tissue array that included an independent set of 207 paired HCC and adjacent noncancerous tissues as well as 10 normal liver and 16 cirrhosis tissues. As shown in Fig. 5A,B, not only was the protein level of TRIM35 sharply decreased in HCC tissues compared with matched noncancerous tissues, normal liver samples or cirrhosis tissues, but the proportion of HCC specimens with TRIM35 underexpression was also predominant, which is consistent with our genomic and transcriptional results (Fig. 5C,D). Importantly, the expression level of TRIM35 was negatively correlated with tumor grade, tumor size, and serum alpha-fetoprotein (AFP) level (Table 2). However, the down-regulation of TRIM35 expression has no significant correlation with overall survival of HCC patients (data not shown). Taken together, these results suggested that loss of TRIM35 expression is a critical event in the development and progression of HCC.

Table 2. Correlation of TRIM35 Expression with Various Clinicopathological Features of HCC Specimens by Immunostaining Assays
 nTRIM35 IHCχ2P-Value
Negative/Weak Positive n (%)Strong Positive n (%)
  1. P value represents the probability from χ2 test for TRIM35 staining pattern between variable subgroups.

  2. AFP, alpha-fetoprotein; HBsAg, hepatitis B virus surface antigen. Negative, score = 0; Weak positive, score = 1 or 2; Strong positive, score = 3 or 4. *, χ2 test, P < 0.05.

Gender      
 Male17078 (45.88)92 (54.12)00.994
 Female37 17 (45.95)20 (54.05)  
Age (Year)    0.4920.483
 ≤5513866 (47.83)72 (52.17)  
 >5568 29 (42.65)39 (57.35)  
Edmondson grading      
 I11 2 (18.18)9 (81.82)9.1720.027(*)
 II95 40 (42.11)55 (57.89)  
 III62 28 (45.16)34 (54.84)  
 IV39 25 (64.10)14 (35.90)  
Tumor size (cm)    5.9880.014(*)
 <510338 (36.89)65 (63.11)  
 ≥598 53 (54.08)45 (45.92)  
Serum AFP (ng/ml)    5.770.016(*)
 <2067 23 (34.33)44 (65.67)  
 ≥2013671 (52.21)65 (47.79)  
Serum HBsAg    0.7090.4
 Absent38 20 (52.63)18 (47.37)  
 Present16273 (45.06)89 (54.94)  
Cirrhosis    1.1950.274
 Absent35 19 (54.29)16 (45.71)  
 Present17276 (44.19)96 (55.81)  
Intrahepatic metastasis    0.1390.709
 Absent14063 (45.0)77 (55.0)  
 Present67 32 (47.76)35 (52.24)  
thumbnail image

Figure 5. Down-regulation of TRIM35 expression in an independent set of HCC specimens. (A) Representative IHC images of TRIM35 in two HCC and paired noncancerous tissues, one normal liver, and one cirrhosis tissue. Original magnification 50× (insert, 400×). (B) IHC analysis of TRIM35 in an independent set of 207 paired HCC and matching nontumor tissues as well as 10 normal liver and 16 cirrhosis tissues. The IHC signal intensities were scored as 0, 1, 2, 3, or 4. The P-values were determined by a χ2 test. The pie chart represents the proportions of HCC samples showing underexpression (blue) and no change (yellow) of IHC staining for TRIM35. (C) Deletion of a 3.87-Mb region of chromosome 8 (8p21.2-21.1) including the TRIM35 gene. Log2 ratio dot plots were generated from data obtained from the Affymetrix 6.0 arrays. The y-axis shows the log2 ratio of paired tumor/normal DNA, and the x-axis represents the scatter distribution of the 58 samples. The pie chart shows the proportions of HCC samples in the copy number gain (red), loss (blue), and normal (yellow) categories. (D) The expression level of TRIM35 in 47 paired HCC and matched nontumor tissues were determined by q-PCR. The data are expressed as the log2 fold change (ΔCt [HCC/Non.]). Significant down-regulation of TRIM35 in paired HCC/nontumor samples was defined as a log2 fold change < –1 (i.e., two-fold). The pie chart shows the proportions of HCC samples showing overexpression (red), underexpression (blue), and no change (yellow).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Copy number alterations represent a substantial category of genetic variations. During carcinogenesis, cancer genomes often acquire somatic copy number changes, which can alter the dosage of oncogenes and tumor suppressors.1 Although several previous studies have applied copy number analyses for HCC,12-14 in the present study we mainly focused on identifying CNAs and their associated oncogenes and tumor suppressors in HCC genomes using high-resolution SNP 6.0 arrays, which can provide high sensitivity and specificity in detecting subtle CNAs. In addition, we used 58 paired tumor and adjacent nontumor tissues from the same individuals, which allowed us to identify highly recurrent CNAs other than accidental CNAs in HCC.

Chromosomal instability, including copy number gains or losses in genomic DNA, is commonly observed in HCC, and several aberrant chromosomal loci have been frequently reported in association with this disease.32 For example, a gain at 1q is one of the most frequently detected alterations in HCC (58%-86%) and has been suggested as an early genomic event in the process of HCC development, whereas a loss at 8p is well documented in HCC, with a frequency of 29% to 77%.14 However, there is little doubt that additional CNAs and targeted genes within these loci exist in HCC. In this study we identified a total of 1,241 significant regions of CNAs. Although previous studies of HCC genomes have implicated several of the CNAs that we identified here (gains of 1q, 7q, 8q, 11q, 17q, and 20q and losses of 1p, 4q, 6q, 8p, 16q, and 17p),4, 13, 14 a large number of new CNAs were also found, including gains of 1p, 10q, 13q, 16p, 16q, 18p, 19p, and 22q and losses of 3q, 4p, 8q, 12p, 14q, and 18q, which provide opportunities to identify new cancer genes in HCC. For example, a frequent amplification target is COL4A1 on 13q34, and a frequent deletion target is SERPINA5 on 14q32.13. Moreover, the differential expressions of eight DEGs in several of these new CNAs were also validated by q-PCR (Supporting Fig. 2B). Additionally, SERPINA5 was also observed to inhibit the migration ability of HCC cells in this study (Supporting Fig. 7). To the best of our knowledge, this is the first study to use high-resolution copy number analysis of a relatively large numbers of paired specimens to create a comprehensive catalog of CNAs in HCC genomes.

Several findings have emerged from our studies, mainly based on the opportunity provided by integrated analysis of genomic and transcriptional profiles. One finding is that several regulatory modules were identified as functioning in a concerted manner, including involved in cell adhesion, cell cycle, regulation of the actin cytoskeleton, and WNT signaling pathways, which have all been implicated in HCC.33, 34 Another finding is the identification of three novel cancer genes related to HCC, including one tumor suppressor candidate TRIM35 and two possible oncogenes HEY1 and SNRPE. TRIM35 is a member of the Ring finger, B box, coiled-coil (RBCC), or Tripartite motif (TRIM) family.35 It was originally isolated as a gene up-regulated during an erythroid-to-myeloid lineage switch, and independently as a proapoptotic gene activated during macrophage maturation.31, 35 It is notable that enforced expression of TRIM35 in HeLa cells could inhibit cell proliferation and tumorigenicity.31 However, the functions of this gene in HCC are largely unknown. In this study we found that TRIM35 was located in a frequently deleted region of 8p21.2-21.1. Consistently, the mRNA and protein levels of TRIM35 were also significantly down-regulated in HCC specimens. However, it is worth noting that additional regulatory mechanisms other than its genomic loss for TRIM35 down-regulation in HCC exist. Therefore, we examined the methylation status of CpG islands within TRIM35 promoter using quantitative real-time methylation-specific PCR on 31 out of 58 paired HCC and nontumor tissues. We found that the frequency of hypermethylation was approximately 45.2% (14/31) in HCC tissues compared with the nontumor tissues, which might account for the down-regulation of TRIM35 mRNA and protein level in HCC tissues in addition to that caused by genomic loss of 14q32.13 loci (17.2%). Furthermore, we found that TRIM35 could significantly suppress the in vitro cell proliferation and in vivo tumorigenicity of HCC cells. Most important, the expression level of TRIM35 was negatively correlated with the tumor grade, tumor size, and serum AFP level of HCC patients. It is conceivable that TRIM35 is one of the novel tumor suppressors located among loci at 8p21. However, these results are not sufficient to indicate that TRIM35 can be considered a candidate biomarker for HCC.

HEY1 encodes a nuclear protein belonging to the hairy and enhancer of split-related (HESR) family, which plays important roles in blood vessel formation and is involved in proliferation, migration, and network formation in endothelial cells.36 However, the roles of HEY1 in HCC have not been reported previously. Here, HEY1 was identified as a significant target gene in the 8q21.13 amplificon and the resulting up-regulation of HEY1 was obviously observed in HCC. Functional experiments showed that enhanced expression of HEY1 could significantly promote in vitro and in vivo proliferation of HCC cells. Additionally, SNRPE appears to function in RNA metabolism and has been shown to interact with DDX20.37 It was previously reported that SNRPE was amplified and up-regulated in malignant gliomas and oral squamous cell carcinomas.24, 38 In the present study we identified it as a new oncogene candidate for HCC, as it was widely amplified and up-regulated in HCC and was capable of significantly enhancing HCC cell proliferation and tumorigenicity.

In conclusion, we produced a comprehensive copy number profile and found 1,241 somatic CNAs in HCC genomes using whole-genome SNP 6.0 arrays. By integrating genomic, transcriptional, and functional data we further identified one novel tumor suppressor candidate and two novel potential oncogenes for HCC, which will facilitate better understanding of the molecular mechanisms of hepatocarcinogenesis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Bin Cai from CapitalBio Ltd., Co. (Beijing, China) for help with SNP array analysis and data processes. We also thank Qi Li from the Invitrogen part of LifeTech for helping with data analysis, and Lin Li, Feng Su, Rui Li, and Amy Ai from Genminix Informatics Ltd., Co. for technical assistance. We thank Dr. T. Didier for gifts of the pWPXL, psPAX2, and pMD2.G lenti-virus plasmids.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_24495_sm_SuppInfoFig1.tif3381KSupporting Information Figure 1. Gene ontology enrichment of 362 DEGs within the regions of 1,241 somatic CNAs.
HEP_24495_sm_SuppInfoFig2.tif2026KSupporting Information Figure 2. q-PCR analyses of mRNA levels of differentially expressed genes within the aberrant regions in 47 paired HCC and adjacent non-tumor tissues. A. q-PCR analyses of expression levels of 20 candidate genes selected for further functional assessments, including 13 upregulated genes (DDEF1, SHC1, HAX1, RAD21, MPZL1, YWHAZ, PEA15, ILF2, SNRPE, C1ORF2, BOP1, HEY1 and DUSP12) and 7 downregulated genes (SERPINA5, GNAO1, MT1G, PER1, C8ORF4, SLC25A4 and TRIM35). p < 0.05 was considered significant. B. q-PCR analyses of expression levels of eight differentially expressed genes within eight new CNAs, including four upregulated genes in four regions of copy number gain (COL4A1 in 13q33.3-34, CLN3 in 16p12.1-11.2, BMI1 in 10p12.31-12.1 and GNAZ in 22q11.22) and four downregulated genes in four regions of copy number loss (FOS in 14q23.2-31.1, HGFAC in 4p16.2-16.1, CDH19 in 18q22.1 and SERPINA6 in 14q31.1-32.13). p < 0.05 was considered significant.
HEP_24495_sm_SuppInfoFig3.tif5633KSupporting Information Figure 3. q-PCR analyses of mRNA levels of 20 candidate genes in eight liver cancer cell lines q-PCR analyses of relative expression levels of 20 candidate genes (DDEF1, SHC1, HAX1, RAD21, MPZL1, YWHAZ, PEA15, ILF2, SNRPE, C1ORF2, BOP1, HEY1, DUSP12, SERPINA5, GNAO1, MT1G, PER1, C8ORF4, SLC25A4 and TRIM35) in eight liver cancer cell lines, including Hep3B, PLC/PRF/5, HUH-7, HepG2, SK-Hep-1, SMMC-7721, MHCC-97L and HCCLM3.
HEP_24495_sm_SuppInfoFig4.tif1766KSupporting Information Figure 4. q-PCR analyses of overexpression of the 20 candidate genes in SMMC-7721 and HUH-7 cells A and B. q-PCR analyses of overexpression of TRIM35, HEY1 and SNRPE in SMMC-7721 and HUH-7 cells, respectively. C. q-PCR analyses of overexpression of 17 candidate genes (DDEF1, SHC1, HAX1, RAD21, MPZL1, YWHAZ, PEA15, ILF2, C1ORF2, BOP1, DUSP12, SERPINA5, GNAO1, MT1G, PER1, C8ORF4 and SLC25A4) in HUH-7 cells.
HEP_24495_sm_SuppInfoFig5.tif982KSupporting Information Figure 5. q-PCR analyses of siRNA knockdown of TRIM35, HEY1 and SNRPE in HepG2 cells.
HEP_24495_sm_SuppInfoFig6.tif1010KSupporting Information Figure 6. Detection and quantitation of TRIM35 methylation in 31 paired HCC and adjacent paracancer tissues by real-time methylation-specific PCR. The methylation status of CpG islands within TRIM35 promoter was examined by quantitative real-time methylation-specific PCR on 31 out of 58 paired HCC and matched paracancer tissues. Totally, 18 out of 31 HCC tissues were methylation-positive, whereas 9 out 31 non-tumor tissues were methylation-positive. Noticeably, among the 18 methylation-positive HCC cases, the methylation of TRIM35 promoter was also observed in 7 corresponding non-tumor tissues, however, the methylation levels in 6 HCC tissues was higher than those in the corresponding non-tumor tissues. Additionally, among the remaining 11 methylation-positive HCC tissues, 3 also harbored copy number loss of the 14q32.13 loci, which targets the TRIM35 gene. In summary, the frequency of hypermethylation was approximately 45.2% (14/31) in HCCs, which accounted for the downregulation of TRIM35 mRNA and protein level in HCC tissues in addition to that caused by genomic loss of 14q32.13 loci (17.2%).
HEP_24495_sm_SuppInfoFig7.tif2438KSupporting Information Figure 7. SERPINA5 inhibited HCC cell migration, but not cell growth. A. Representative result of CCK-8 assays for the effects of SERPINA5 gene on the in vitro proliferation of SMMC-7721 and Huh-7 cells by lentivirus-mediated overexpression. B. The effects of the SERPINA5 gene on the growth abilities of SMMC-7721 cells in xenograft models of nude mice (n=6) as determined by tumor weight. C. The effects of the SERPINA5 gene on the migration abilities of SMMC-7721 and Huh-7 cells by Trans-well migration assays. All the results are shown as the mean ± s.e.m. **, p<0.01; ***, p<0.001.
HEP_24495_sm_SuppInfo.doc650KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.