| Hepatocellular carcinoma (HCC) is the second most common cause of cancer-related deaths and has an increasing incidence, with a reported 750,000 new cases per year worldwide. Early researches have demonstrated the action of oncogenes and inactivation of suppressor gene are closely related with carcinogenesis of HCC. With the development of intensive studies, noncodiong RNA increasingly attracts more attentions as the significant role in gene regulation. The mutations and dysregulations of noncoding RNA lead to diverse human diseases, multiple lines of evidence support noncoding RNA as an research hotspot in HCC. It is surprisingly discovered that the human genome encodes a spot of protein-coding genes, representing<2% of the total genome sequence. It was determined that more than 90% of the noncoding RNA was actively transcribed, including mircoRNAs (miRNAs)and long noncoding RNAs (lncRNAs).LncRNAs, widely existed in the processes of genome transcription, are transcripts ranging in the length of 200 nt at least. LncRNA participates in different biological processes with a variety of specific biological fanctions, such as regulation in epigenetics, RNA transcription, translational control, RNA splicing, cell cycle control, cell differentiation and so on. Recent studies have confirmed the mutations and dysregulations of lncRNA are closely related to the hepatocarcinogenesis. LncRNA is cosidered to be a generalist as the interactions in every step of gene regulation. The dysregulation of lncRNA in each process of biology would give rise to HCC:1) Abnormal DNA methylation of promoter.2) Loss of genomic imprinting.3) MicroRNA sponge. 4) LncRNA competitively binds to the transcription complex like polycommb repressive complex 2 (PRC2) as a modular scaffold to prevent transcription of several metastasis suppressor genes. With the advances of researches, HCC associated lncRNAs have been discovered and further studied, such as HI9, MALAT1, UCA1, uc.338, MEG3, HULC, HEIH and MVIH.In addition to the importance of oncogenes and suppressors in tumorigenesis, epigenetic abnormalities induce the tumorgenesis and metastasis, it was confirmed that DNA methylation of suppressor genes causes decreases of gene expressions. Similarly, abnormal DNA methynation of lncRNA is closely associated with carcinogenesis and tumor metastasis. Nevertheless, researches on mechanisms of lncRNA are limited, it is still unknown about the interactions between DNA methylation and expressions of HCC related lncRNAs. On account of the loss of mechanism study of lncRNAs which effect on the process of occurrence and development of HCC, we applied both lncRNA microarray and lncRNA promoter microarray to screen hepatoma-specific lncRNA, then verified and went into the complex mechanism study on lncRNA expression regulated by DNA methylation and gene-regulatory roles of incRNA influencing downstream gene expression and biology function. In our present study, we analysed the regulation mechanisms of HCC specific lncRNA from the changes of DNA methylation in promoter regions to the alterations of downstream target genes. Furthermore, we prospected lncRNA to be a novel molecular marker and therapeutic target of HCC.Chapter I Selection of HCC specific lncRNAs with hypermethylation of promoter regionsObjective:Epigenetic abnormality is a vital cause of hepato-carcinogenesis and metastasis, thus DNA methylation directly interacts with gene expression. LncRNA microarray has been widely applied domestic and overseas, in contrast lncRNA promoter microarray is being get promotion and application at early stage. We combined both lncRNA profile chip and promoter microarray to screen specfic lncRNA differentially expressed in HCC tissue with hypermethylation in promoter regions.Methods:We combined Arraystar Human LncRNA Microarray v2.0 and Arraystar Human 2.1M LncRNA Promoter Microarray to detect HCC and peritumor tissues from 8 HCC patients treated in Zhongshan Hospital from October to December in 2011. The microarray is designed for the global profiling of human LncRNAs and protein-coding transcripts.33,045 LncRNAs and 30,215 coding transcripts can be detected by our second-generation LncRNA microarray. The LncRNAs are carefully collected from the most authoritative databases such as RefSeq, UCSC Knowngenes, Ensembl and many related literatures. In additon, the promoter microarray was applied to test all promoter regions of lncRNA, from about-3500 bp to+1250 bp of the transcriptional start sites (TSSs). In order to aviod the data error caused by technical variability and evaluate methylation differences between samples, we performed Median-centering, quantile normalization, and linear smoothing to analyse and revise the raw data. To accurately quantify methylation levels of CpGs, we applied a novel analytical methodology, modeling experimental data with MeDIP enrichment (MEDME), which utilized the absolute methylation score (AMS) as an indicative of DNA methylation to analyse the differentially methylated regions (DMRs) of lncRNA promoters. Differentially methylated probes between two groups were identified using AMS by t-test, probes with p-value< 0.05 and ABS(AMSdif)> 8 were chosen and applied to search AMS DMRs. Differentially methylated probes based on AMS were merged into candidate DMRs. In addition, the average AMS of candidate DMRs were re-calculated and re-tested via t-test. DMRs with average AMS still significant different between two groups were selected and then defined as AMS DMRs. We combined and analysed the data from two of the microarrays, then selected the HCC specific lncRNAs with CpG island hypermethylation in promoter regions.Results:We analysed the data from two microarrays and found that 2655 lncRNAs were chosen for differentially expressed lncRNA after screening in both HCC and peritumor tissues with the standard as below:fold change (FC) should be above 2 times and p-value<0.05, at least 12 out of 16 samples had flags in present or marginal. There were 1413 up-regulated and 1242 down-regulated lncRNAs in HCC tissues. Differentially methylated DMRs were analysed which based on the result of AMS, parameters used in AMS DMRs selection:Probes in DMR>2, ABS (AMS_dif)>8, P-value by t-test<0.05. Our results confirmed that there were 1487 AMS DMRs in 1737 genes’ promoters. Finally, we combined and analysed the data from both Human LncRNA Microarray and Human 2.1M LncRNA Promoter Microarray,22 HCC specific lncRNAs with promoter hypermethylation were selected as candidates for verifying:SCARF1, AK126915, MST1P9, AK056817, SRD5A1P1, RP1-93C23.1, HTR7P1, AK091100, RP11-755B10.3, AK127534, AC139666.1, MTND4P25, MT1XP1, CTB-3601.7, CTB-3601.6, AC115617.2, AY216265, LOCI 13230, RP5-1154E9.6, MT-ATP8, AC013437.2 and AC009960.8.Conclusions:As a wildly used microarray, Human LncRNA Microarrays have been approved for the stability and sensitivity. Moreover, Arraystar Human 2.1M LncRNA Promoter Microarray, applied in detecting DMRs of lncRNA promoters, is a novel chip technology being on the rise. Our study combined strengths from two of microarrays and selected candidate lncRNAs for further verifying.Chapter ⅡVerification of HCC specific lncRNAs with hypermethylation of promoter regionsObjective:LncRNA microarrays, with advantages of universal coverage, efficiency and sensitivity, provide a convenient and fast way to screen HCC related lncRNAs. Neverthless, chips are so sensitive in the process of screening and can easily be disturbed by external factors. Therefore, data from microarrays shloud be analysed and verified. In this study, we aimed to analyse the data from two of the lncRNA microarray and verify 22 candidate lncRNAs by testing the differential expressions in HCC and peritumoral tissues and DMRs of promoters.Methods:We analysed the data from chapter Ⅰ and selected 22 candidate lncRNAs for further verification. Quantitative Reverse Transcriptase Polymerase Chain Reaction (qRT-PCR) was used to detect the differential expressions of lncRNA in 8 tissue samples from our preliminary study. Both Methylation specific PCR (MSP) and Bisulphite modification combining sequencing PCR (BSP) were applied to verify DNA methylation of lncRNA promoters. With the purpose of confirming the veracity of preliminary screening, Two independent sample cohorts were enrolled:54 pairs and 30 pairs of HCC and peritumoral tissue samples are from HCC patients who were diagnosed or phisical checked in Zhongshan Hospital in 2009. Both quantitative RT-PCR and MSP were used to re-test the differential expressions and promoter methylation of candidate lncRNAs.Results:According to the detection of 8 pairs of HCC and peritumoral tissues from preliminary study, we found 20 out of 22 candidate lncRNAs were significantly decreased in HCC:lncRNA-SRD5A1PA (P=0.162), lncRNA-RP1-93C23.1 (P=0.091), AC115617.2 (P=0.009), SCARF1 (P=0.011), MST1P9 (P<0.001), HTR7P1 (P=0.002), AK056817 (P=0.007), MTND4P25 (P=0.021), LOC113230 (P=0.005), MT1XP1 (P<0.001), AC013437.2 (P=0.010), CTB-360.1.6 (P=0.009), CTB-360.1.7 (P=0.017), RP11-755B10.3 (P=0.007), AK127534 (P<0.001), AY216265 (P＜0.001), AK091100 (P<0.001), RP5-1154E9.6(P<0.001), AC139666.1 (P=0.002), AK126915 (P=0.004), MT-ATP8 (P=0.020) and AC009960.8 (P=0.035). By analysing the sequences of lncRNA promoters, we ensured the presence of CpG island in 13 out of 22 candidates. In addition, we verified the DMRs of the 13 cadidates by BSP and found dense CpG island hypermethylation in all described lncRNAs (>40%) with significant differences between HCC and peritumor (P<0.01). In accordance with previous result, MSP was used to confirm the presence of promoter methylation. Furthermore, both qRT-PCR and MSP were applied to verify the results in two independent cohorts. Other than lncRNA-MT-ATP8 (P=0.072), expressions of 21 lncRNAs were significantly decreased in HCC:SCARF1, AK126915, LOC113230, MST1P9, AC009960.8, AK091100, MTND4P25, HTR7P1, MT1XP1, AC139666.1, RP11-755B10.3, AK127534, AK056817, AY216265, AC115617.2, AC013437.2 (P＜0.001), CTB-360.1.6 (P=0.033), SRD5A1P1 (P=0.036), CTB-360.1.7 (P=0.046), RP5-1154E9.6 (P=0.009) and RP1-93C23.1 (P=0.022). The presence of promoter methylation in 13 cadidates were demonstrated as below:SCARF1 (76.7% vs 26.7%), AK126915 (63.3% vs 20.0%), MSTIP9 (80% vs 40%), AK056817 (66.7% vs 30%), SRD5A1P1(RP1-93C23.1) (73.3% vs 43.3%), HTR7P1 (56.7% vs 16.7%), RP11-755B10.3 (63.3% vs 26.7%), AK127534 (53.3% vs 26.7%), CTB36O1.6-1.7 (83.3% vs 36.7%), AY216265 (70.0% vs 46.7%) and LOCI 13230 (60% vs 16.7%). Data from microarrays, qRT-PCR, MSP and BSP were combined and analysed, lncRNA-SCARF1 (P=0.004) was selected and proposed as a role of suppressor lncRNA with hypermethylated promoter for further mechanism study.Conclusions:In this study we demonstrated 20 out of 22 lncRNAs were significantly decreased in HCC in the preliminary verification. In another cohort of 54 pairs of samples, the expressions of all candidates were confirmed to be down-regulated in HCC tissues. On the other hand,13 lncRNAs were verified the presences of promoter hypermethylation by BSP and MSP. Consequently, data from both Array star Human LncRNA Microarray v2.0 and Arraystar Human 2.1M LncRNA Promoter Microarray were demonstrated to be stable and accurate.Chapter Ⅲ Function research of lncRNA-SCARFl and its promotor methylation signature for Hepatocellular CarcinomaObjective:On account of the verification in two cohorts of HCC samples, we confirm that lncRNA-SCARF1 is down-regulated in HCC tissue with promoter hypermetylation and plays a role of suppressor lncRNA in the process of carcinogenesis. In this study we aimed at investigating the regulation of promoter methylation in lncRNA-SCARF1, clarifying the signaling pathway which lncRNA-SCARF1 regulates.Methods:We tested the expressions of lncRNA-SCARF1 in HCC cell lines including HepG2, Huh7, SMMC-7721, MHCC-97L, MHCC-97H, MHCC-LM3 and normal hepatocyte LO2. To demonstrate the transcriptional silencing of lncRNA-SCARF1 in HCC cell lines in association with the presence of CpG hypermethylation we treated the HCC cell lines with 5-aza-deoxycytidine (5’-Aza-dc) and measured the alterations of lncRNA expressions. In addition, the presences of CpG island metylation were ensured by BSP and MSP. As a suppressor lncRNA in HCC, lncRNA-SCARF1 was stably expressed and transfected with a lentiviral system in MHCC-97H and HepG2 cell lines to exam the effects on proliferation, migration and apoptosis of HCC cell lines. Moreover, we tested the growth of MHCC-97H and HepG2 cell lines tranfected with lncRNA-SCARF1 overexpressing lentiviral system via xenografted nude mice models in vivo. Downstream target genes were anticipated by Gene Set Enrichment Analysis (GSEA) according to the data from microarrays. Taqman pathway arrays for human molecular mechanism of cancer were applied to search for the potential target genes, qRT-PCR and western blot analysis were used to verify the signaling pathway.Results:Compared with HCC cell lines, expression of lncRNA-SCARFl was significant higher in LO2. We interfered HepG2 and MHCC-97H, with the lowest expression of lncRNA-SCARF1, with the DNA demethylating agent 5’-Aza-dc and ensured the presence of CpG island methylation by BSP and MSP. We detected the lncRNA expression by qRT-PCR and found a restoration of lncRNA expression with the treatment of the DNA demethylation agent. We established the stable lncRNA-SCARF1 overexpressing transfectants with a lentiviral system in HepG2 and MHCC-97H cell lines and observed the alterations in biologicl function of HCC cell lines in vitro. Both the proliferations and migrations of HepG2 and MHCC-97H cell lines were inhibited by overexpressing lncRNA-SCARF1. Besides, overexpression of lncRNA-SCARF1 promoted HCC cell lines apoptosis. Both HepG2 and MHCC-97H cell lines tranfected with table overexpressing LncRNA-SCARF1 vector were s.c. injected into nude mice, and the tumor inhibition was demonstrated in this vivo array. According to the data from two of microarrays, we established the coding-non-coding gene co-expression network and analysed the potential correlations of lncRNA-SCARF1 and target coding genes.16 coding genes were found to be positively correlated with lncRNA-SCARF1, while 15 genes were negatively correlated with. Futhermore, 3 human cancer related pathways and a apoptosis pathway were anticipated to be positively correlated with lncRNA-SCARFl via Gene Set Enrichment Analysis (GSEA). We applied Taqman pathway arrays for human molecular mechanism of cancer to verify signaling pathway and downsteam target genes. 7 coding genes including AKT1, PIK3CA, BCL2, BCL2L1 et al were tested to be down-regulated in HCC cell lines transfected with lncRNA-SCARF1 vector, while 4 were found to be elevated including BAX, ELK1, HRAS and PTEN. We verified the potential target genes by qRT-PCR and western blot and ensured that lncRNA-SCARF1 were able to influence the expressions of PIK3CA, AKT1, BAX and BCL-2 in HCC.Conclusions:The transcriptional silencing of lncRNA-SCARF1 was in association with the presence of CpG island hypermethylation in HCC. LncRNA-SCARF1 plays a role of suppressor in the process of HCC inhibitions and may be functions in the PIK3CA/AKT signaling pathway. |
PDF Full Text Request