Abstract IA31: Novel nuclear and mitochondrial RNAs that are linked to key pathways and depend on sex, population origin, race, tissue, and disease

MicroRNAs (miRNAs), the best-known short regulatory noncoding RNAs (ncRNAs), were reported for the first time exactly 25 years ago. Extensive research efforts since then have firmly established their roles as key regulators of cellular processes, in health and in disease. MiRNAs modulate protein abundance by targeting the corresponding messenger RNAs (mRNAs) in a sequence-dependent manner. Ever since the discovery of miRNAs, the community has largely adhered to a paradigm according to which a single stretch of genomic DNA is transcribed into a miRNA precursor molecule that is then processed to produce a single (regulatory) miRNA product, approximately 22 nucleotides (nts) in length. As next-generation sequencing (NGS) became widespread, scientists observed that each miRNA precursor gives rise to multiple miRNA products with slightly different endpoints and different abundances. These “variants” were called “isomiRs.” For a given miRNA precursor arm, the most abundant among the isomiRs arising from it was considered to be the “actual” product of the arm whereas all remaining isomiRs were initially dismissed as inconsequential bystanders. By analyzing datasets from thousands of healthy individuals and patients, we were first to show that human isomiRs are produced constitutively and that the identities and abundance levels of the isomiRs arising from a given miRNA precursor depend on a person's sex, population origin, and race/ethnicity. We also showed that the isomiR profiles additionally depend on the tissue at hand, the tissue's state (health vs. disease), and the disease type/subtype. From a functional standpoint, we showed experimentally that distinct isomiRs from the same miRNA precursor arm can target different groups of mRNAs. Notably, our analyses revealed that for a given miRNA arm the identity of the most abundant isomiR can change from tissue to tissue. An analogous story played out in the context of transfer RNAs (tRNAs). For more than 50 years, the community viewed tRNAs as ancillary molecules that participated in the translation of codons to amino acids. Here too, NGS revealed that both precursor and mature tRNAs give rise to multiple “tRNA-derived fragments” or “tRFs” with different endpoints and abundances. Work by several labs including our own showed for the shorter (~22 nts) tRFs that they act like miRNAs, i.e., they control the abundance of mRNAs and their respective protein products through RNAi. As in the case of isomiRs, we were first to show that human tRFs are produced constitutively and that the identities and abundances of tRFs from a given mature tRNA depend on a person's sex, population origin, and race/ethnicity. We also showed that the tRF profiles additionally depend on the tissue at hand, the tissue's state, and the disease type/subtype. Preliminary data that we generated also suggest that distinct tRFs that are produced by the same tRNA have generally distinct roles. Particularly intriguing was our finding that the mitochondrially encoded tRNAs are a very rich source of tRFs. In fact, sequence for sequence, the 22 mitochondrial tRNAs produce 15x as many tRFs as their nuclearly encoded counterparts. Having established their constitutive nature, we also investigated the possibility of cooperation and competition between isomiRs and tRFs. Specifically, we studied these molecules in normal breast and in triple-negative breast cancer (TNBC) samples from The Cancer Genome Atlas. In normal breast, we found that isomiRs and tRFs regulate genes and pathways in a cooperative manner. However, in TNBC the regulatory events that are effected by isomiRs and tRFs, respectively, are aligned with the race of the patient. Specifically, we found that in White TNBC patients mRNAs are regulated preferentially by tRFs whereas in Black/African American TNBC patients mRNAs are regulated preferentially by isomiRs. The evidence also shows that both isomiRs and tRFs regulate core cancer pathways but do so differently in White and Black/African American TNBC patients. In other words, in TNBC the key regulators and the affected genes/pathways differ in patients who differ by race/ethnicity. We carried out similar analyses in prostate cancer and a few other cancers, reaching analogous conclusions. Our findings have several implications. First, they highlight that the mRNA abundance profiles in a given tissue (and, by extension, the abundance profiles of their protein products) generally differ in different people, in both health and disease. Second, the relationships we uncovered between short ncRNAs and mRNAs provide initial mechanistic evidence that who we are determines how we progress from health to disease and the exact trajectory that we will follow. Third, the findings suggest the feasibility of a pan-cancer biomarker that is based on isomiRs and tRFs and is both sensitive and specific. Fourth, our findings on isomiRs, and on the nuclear and mitochondrial tRFs, suggest that there are many putative regulators that matter more for women than they do for men; similarly, other putative regulators matter more for men, not as much for women; yet others matter more for European populations than they do for African or Asian populations, and vice versa, etc. It is safe to assume that in the short term these findings will force us to go back to the proverbial drawing board. However, it is also reasonable to expect that in the long run these findings can provide a different vantage point from which to approach disease. Doing so will in turn lead to exciting new and powerful approaches for diagnosing disease, help establish whether disease has recurred, determine whether a therapy is effective, and fuel the pursuit of new therapeutic targets. Most importantly, these new approaches will allow us to tune diagnostics and therapeutics to each specific patient and each specific disease.

Citation Format: Isidore Rigoutsos. Novel nuclear and mitochondrial RNAs that are linked to key pathways and depend on sex, population origin, race, tissue, and disease [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr IA31.