The insertion and dysregulation of transposable elements in osteosarcoma and their association with patient event-free survival

Chao Wang,Chun Liang

doi:10.1038/s41598-021-04208-5

Abstract

The dysregulation of transposable elements (TEs) has been explored in a variety of cancers. However, TE activities in osteosarcoma (OS) have not been extensively studied yet. By integrative analysis of RNA-seq, whole-genome sequencing (WGS), and methylation data, we showed aberrant TE activities associated with dysregulations of TEs in OS tumors. Specifically, expression levels of LINE-1 and Alu of different evolutionary ages, as well as subfamilies of SVA and HERV-K, were significantly up-regulated in OS tumors, accompanied by enhanced DNA repair responses. We verified the characteristics of LINE-1 mediated TE insertions, including target site duplication (TSD) length (centered around 15 bp) and preferential insertions into intergenic and AT-rich regions as well as intronic regions of longer genes. By filtering polymorphic TE insertions reported in 1000 genome project (1KGP), besides 148 tumor-specific somatic TE insertions, we found most OS patient-specific TE insertions (3175 out of 3326) are germline insertions, which are associated with genes involved in neuronal processes or with transcription factors important for cancer development. In addition to 68 TE-affected cancer genes, we found recurrent germline TE insertions in 72 non-cancer genes with high frequencies among patients. We also found that +/− 500 bps flanking regions of transcription start sites (TSS) of LINE-1 (young) and Alu showed lower methylation levels in OS tumor samples than controls. Interestingly, by incorporating patient clinical data and focusing on TE activities in OS tumors, our data analysis suggested that higher TE insertions in OS tumors are associated with a longer event-free survival time.

Highlights

The dysregulation of transposable elements (TEs) has been explored in a variety of cancers
After comparing the tumor sample with its paired normal control using a 100-bp window to identify the somatic or germline TE insertions followed by visual validation via IGV, we found that most OS patient-specific TE insertions belong to germline TE insertions
By using T Etranscripts[36], which demonstrated superior performance for TE expression measurement followed by DESeq[237], we showed that compared with samples of normal muscle tissues adjacent to osteosarcoma, LINE-1 and Alu of different evolutionary ages, several subfamilies of SVA and HERV-K and various satellite repeats were all significantly up-regulated in OS tumor samples

Summary

Introduction

The dysregulation of transposable elements (TEs) has been explored in a variety of cancers. By integrative analysis of RNA-seq, whole-genome sequencing (WGS), and methylation data, we showed aberrant TE activities associated with dysregulations of TEs in OS tumors. Chromothripsis has recently been linked to both somatic and germline TP53 gene mutations in pediatric medulloblastoma and acute myeloid leukemia[3], and TP53 and RB1 were identified as the only recurrently altered genes in a recently OS study[7] Apart from those reported mutations, it was shown that over-expression of IGF2 and IRX1 via the hypomethylation of their promoters were involved in the OS-associated metastasis[8]. A recent study focusing on TE insertions (i.e., LINE-1, Alu, SVA, and HERV) in 202 colorectal tumors showed that the average number of somatic TE insertions is 25 with wide variability among individuals[16]. A higher number of somatic TE insertions, including LINE-1, Alu, SVA, and HERV was associated with poorer diseasespecific survival in colorectal cancer patients[16]. In terms of TE expressions, it has been reported that a significant increase in LINE-1 expression levels was observed in breast invasive carcinoma, head, and neck squamous cell carcinoma, and lung adenocarcinoma when expressions of full-length LINE-1 were c onsidered[17]

Methods

Results

Conclusion