Abstract 2225: Transposable elements are an abundant and pan-cancer source of shared tumor-specific antigens and membrane targets

• Published in: Cancer research (Chicago, Ill.) Ill.), 2021-07, Vol.81 (13_Supplement), p.2225-2225
• Main Authors: Shah, Nakul M., Jang, Hyo Sik, Maeng, Ju Heon, Tzeng, Shin-Cheng, Wu, Angela, Fan, Changxu, Basri, Noah L., Katz, Benjamin, Li, Daofeng, Xing, Xiaoyun, Evans, Bradley S., Wang, Ting
• Format: Article
• Language: English
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2021-2225

Transposable elements (TEs) represent close to half of the genome and are generally disregarded in genomic studies due to their silencing in somatic tissues and difficulty in mapping to their repetitive sequences. Recent studies have revealed that epigenetic dysregulation in cancer can unlock the regulatory potential of transposable elements (TEs), and they can play an important role in cancer progression and oncogenesis. One important consequence of this phenomenon is the pervasive activation of TEs' intrinsic promoters, which leads to generation of thousands of unique transcripts. Many of these transcripts splice into downstream genes and lead to the generation of TE-gene chimeric transcripts. These transcripts can alter the main reading frame of the original transcript to generate unique isoforms of the target gene or generate novel out-of-frame peptides that could be therapeutic targets. In this study, we analyzed the transcriptomes of 11,092 samples from 33 TCGA cancer types and 675 cancer cell lines to comprehensively profile all TE-gene fusion transcripts. Using somatic tissues from FANTOM5 and GTEx, we filtered these transcripts for tumor-specificity and discovered 2,642 tumor-specific TE-gene transcripts that promiscuously occur in nearly all TCGA tumor samples. Computational prediction of reading frames of these transcripts identified 1,202 candidates with the potential to generate tumor-specific TE-derived antigens (TS-TEAs). We further analyzed tumor mass spectrometry data from breast adenocarcinoma and ovarian cancer and confirmed that unique peptide sequences from TS-TEAs could be detected. In addition, we performed HLA-pulldown mass spectrometry and confirmed that TS-TEAs are presented on the cell surface in cancer cell lines. Given that these antigens are highly shared within and across cancer types, we assessed their potential to generate universal antigen-based therapies. Optimal combinations of 5, 10, and 20 TS-TEAs could generate unique peptides that bind to patient-specific HLA alleles for 39.2%, 50.8%, and 60.8% of all TCGA tumors respectively. Lastly, we highlight the tumor-specific membrane proteins transcribed from TE-exapted promoters that can potentially expose novel epitopes on the extracellular surface of cancer cells. These can be valuable targets of CAR-T or alternative antibody-based therapies. In conclusion, we showcase the high prevalence of TE-derived promoter activation in cancer and suggest multiple avenues by which thi