Uterine carcinosarcoma (UCS) are aggressive neoplasms consisting of high-grade malignant epithelial and mesenchymal elements. UCS represent less than 5% of all uterine malignancies. In the US, approximately two in 100,000 women develop UCS annually. At the time of diagnosis, approximately one-third of patients have disease that has spread beyond the uterus. The survival percentage for patients with UCS projected at 5 years from the time of diagnosis is 40% to 75% for neoplasia confined to the uterus. Uterine carcinosarcoma is a cancer with high frequency of mutations specifically insertion and deletion polymorphisms such as Copy Number Alterations (CNA) linked to messenger Ribonucleic Acid (mRNA) transcripts. In this work was carried out omic data integration using CNA genomic data and transcripts from mRNA sequence counts from 57 American patients with different levels of infiltration and invasiveness of UCS analyzing 16383 genes and 60488 transcripts separately. For analyzing CNA genes, Component Principal Analysis (PCA) was carried out and for analyzing mRNA sequences counts, Differential Expression Analysis was carried out. After CNA and mRNA separately analysis, 36 genes and 96 transcripts highly significant were found, which were used in the integration analysis. Integrative analysis was carried out using Sparse Least Square (sPLS) methodology using mixOmics package in R software. Integrative analysis was based on graphical analysis from two output plots. Samples graphical representation, from RNAseq and CNA data show the clustering between samples. On RNAseq, samples showed clustering around central zero of all types of tumors, without clear separation between them. This indicates variance of different samples is not explained by the transcripts (genes). Clusters top and bottom of central zero especially tumor with most infiltration and invasiveness explained the most proportion of variance. On CNA genes, samples showed clear separated clustering’s according with types of tumors. Tumor of less infiltration and invasiveness were clustered more closely near of central zero and tumor with most infiltration and invasiveness were clustered more closely away from central zero. Many samples were clustered very closely at central zero especially samples belonging tumors with less infiltration and invasiveness indicating some CNA genes have a weak influence on tumors with less infiltration and invasiveness. Samples from both RNAseq and CNA genes showed a strong negative correlation between them. Tumors with more infiltration and invasiveness showed high dispersion under the central zero, while tumors with less infiltration and invasiveness shows moderate dispersion above the central zero. This indicates both types of genes mRNA transcripts and CNA genes are highly expressed in aggressive tumors. According to genes graphical representation, three important CNA genes were highly expressed, TPM3, tropomyosin 3, RPS27 ribosomal protein S27 genes, both located on chromosome 1 and ACTR1A, Alpha-centractin gene located on chromosome 10 was seen keeping direct positive correlation with the transcript ENG00000122145 human transcript located on Chromosome 16. On RNAseq genes, four genes5 were highly expressed, ENSG00000143028 (SYPL2, Synaptophysin-like protein 2) human gene located on Chromosome 1; ENSG00000077522 (ACTN2, alpha actinin-2) human gene also located on Chromosome 1; ENSG00000086967 (MYBPC2, Myosin-binding protein C) human gene located on Chromosome 19 and ENSG00000253767 (PCDHGA8, Protocadherin gamma-A8) human gene located on Chromosome 5. The results showed a high correlation between CNA and mRNA genes, indicating that copy number alteration also results in differential gene expression in uterine carcinosarcoma.
Author(s): Cristobal Ricardo De Leon Garcia