| BackgroundSporadic colorectal cancer (sporadic colorectal cancer, sCRC), deriverd from colon mucosa epithelial, is one of the five most common malignancy in China, and accounts for 70%of colorectal cancer (Colorectal cancer, CRC). In recent years, the incidence and mortality of CRC is increasing year by year, becoming a major threat to the health of people. From the aspect of molecular level, reported research on sCRC markers are mainly focused on APC, KRAS, and TP53 genes, as well as the systemetic function and molecular mechanisms of these genes. Nevertheless, reported genetic markers of these genes can only explain little for the occurrence of sCRC. With the advent of the era of genomics, screening and identifying new candidate genes genome-widely for sCRC is helpful for us to further understand the molecular pathogenesis of the disease. The emergence of the whole genome sequencing technology and the gradual reduction of the cost of sequencing make it possible for us to screen the genes associated with the disease from the mass of human genome variation data. What is more important than the sequencing, is how to dig out the candidate genes of interest from these data rapidly and accurately. Meanwhile, the function and status of genes screened is of great need to be further illustrated.Methods1. Firstly, we started from 3 sCRC cases of Han Chinese. Whole exome sequencing of colon cancer DNA and data quality analysis were performed. To determine novel somatic LoF mutations for sCRC, we carried out loss of function (Loss of Function) mutation screening strategy and sanger sequencing confirmation of tissue-blood matched samples. The candidate genes with tumor suppressor or mismatch repair function were screened out after bioinfbrmatic analysis.2. Nextly, in-depth study from the DNA mutation, RNA and protein expression levels were performed to determine whether significant changes of candidate genes happened in sCRC clinical samples. We deep-sequenced the coding regions of candidate genes in case of 100-110 independent samples, and determined the mutation frequency. To detect differential RNA expression of candidate genes,28 cases of RNA from sCRC tissues and matched normal tissues were colected and validated by real-time quantitative PCR. Immunohistochemistry was performed using tissue microarrays to determine protein positve rate in sCRC tumor tissues and adjacent normal tissues. Together with the results of three different levels of information analyzed, we finally selected bone morphogenetic protein 5 (BMP5), with high confidence, as a novel tumor suppressor gene for sCRC.3. Both gain of BMP5 and loss of BMP5 in vitro were carried out to observe the effect on the biological characteristics of CRC cells. The effect of BMP5 on the tumor growth of CRC cells in nude mice was further explored.4. Transcriptome sequencing was applied to determine differentially expressed genes caused by overexpression of BMP5. We then deeply analyzed the function phenotype of BMP5 in colorectal cancer model and further ascertained the signaling pathway BMP5 involved in to affect its downstream effector and thus to participate in the occurrence and development of CRC.Result1. We identified a total of 72 LoF somatic mutated genes in 3 cases of sCRC samples using whole exome sequencing and data quality control screening analysis. Integration of bioinformatic database query and gene function analysis, we selected 5 important genes: BMP5, DDI2, RASSF6, SARDH, and C9orf9 that may act as tumor suppressor or mismatch repair genes.2. Deep sequencing of clinical samples revealed the highest mutation frequency (7.7%, 8/104) in BMP5 coding region, including 7 cases of pathogenic mutations. The RNA expression of BMP5 in cancer tissues was significantly lower than that in adjacent normal tissues (P<0.0001), with 1.5 times fold change in 23 cases (82.1%,23/28), and the average downregulation was 5.72 times as compared with paired normal tissues. Immunohistochemistry of 129 cases of paired samples showed the positive expression of BMP5 in normal colon tissues was significantly higher than that in cancer tissues (89.1% vs 75.2%, P=0.0053). There was no correlation between expression of BMP5 and pathological information such as gender, age, and clinical stage. Integration of the above analysis, we considered the high mutation rate and low expression rate of BMP5 are the reliable basis for its candidate tumor suppressor for sCRC.3. In vitro cell experiments showed that both gain and loss of BMP5 could affect the proliferation of CRC cell line HT-29 and SW480, but the effect of the proliferation ability may not be caused by cell apoptosis. Inhibition of BMP5 in SW480 cells could promote the migration and invasion of cancer cells, but the effect of BMP5 overexpression was not significant especially in cell invasion. Xenograft model in nude mice showed, compared with the control group, BMP5 overexpressed HT-29 cells revealed a decreased tumor size and tumor weight (P=0.033).4.The results of transcriptome sequencing showed the upregulation of BMP5 in HT-29 cells could affect differential expression of the 366 genes in the genome, and most of these genes were downregulated. Pathway and co-expression analysis showed that BMP5 could inhibit the expression of epithelial stromal interaction 1 (EPSTI1) through Jak-STAT signaling pathway.ConclusionThe screening strategy based on LoF mutation can capture the critical tumor-related mutated genes efficiently. Mutation of BMP5 and its low expression in clinical samples may be used as a tissue-specific marker for sCRC. BMP5 could inhibit cell proliferation, migration and, invasion in CRC model and show an anti-tumor effect. There may exist a Smad-independent pathway for BMP5 signal transduction. BMP5 repressed the expression of EPSTI1 through the Jak-STAT signaling pathway, suggesting that dysregulation of BMP5 might be an early event of sCRC, which was not reported previously. In this study, based on the theory of omics, we systematically explored the potential biomarker for sCRC, and completion of this subject might provide new evidence and theoretical basis for sCRC pathogenesis. |
PDF Full Text Request