Investigating the translation of Cobra1: canonical expression is alternatively initiated from a non-AUG codon

COBRA1, co-factor of BRCA1, is a transcriptional regulator and a subunit of the Negative elongation complex also known as NELF-B. Although this protein was first designated as a cofactor of BRCA1 and hence acts accordingly, it was found later that it elicits a battery of response genes overlapping those regulated by BRCA1 in absence of BRCA1 itself. Cobra1 deletion is embryonic lethal and results in embryonic stem cells (ESC) differentiation independent of the typical pluripotency machinery. Moreover, it was found that it has a role in suppression of tumors\u27 growth and patients with poor prognosis of breast cancer had decreased levels of COBRA1. Paradoxically, levels of COBRA1 was found elevated in some upper gastro-intestinal tract tumors. Our understanding of the regulation of gene expression has been evolving as an important venue to explain gene product\u27s diversification. Alternative initiation of translation has been observed in many important genes and showed different subsequent phenotypes. In some cases, the discovered protein isoforms are not generated from the classically recognized Kozak/ATG system (i.e. Canonical initiation). Alternatively, their expression is initiated using a non-canonical mechanism resembling viral internal ribosomal entry site (IRES) pathway. Generation of different protein isoforms has been linked to paradoxes in the associated genes\u27 functions. Among the different functions observed are resistance to degradation, altered cellular localization and regulation of different cell cycle phases. In this study we have substantiated the hypothesis that Cobra1 has two protein isoforms, which might be one of the possible reasons for the associated paradoxes. We have used in-silico prediction analyses to verify that the 5\u27 un-translated region (5\u27UTR) of Cobra1 has the required sequences and complex RNA structures for non-canonical initiation. We also could detect these isoforms in endogenous mouse tissues from different strains and ages. Finally, we were able to induce the expression of the two isoforms ex-vivo and still could recognize the isoforms in flag-tag based systems