RNA-binding activity of TRIM25 is mediated by its PRY/SPRY domain and is required for ubiquitination

Nila Roy Choudhury; Gregory Heikel; Maryia Trubitsyna; Peter Kubik; Jakub Stanislaw Nowak; Shaun Webb; Sander Granneman; Christos Spanos; Juri Rappsilber; Alfredo Castello; Gracjan Michlewski

To uncover the RNA-related function of TRIM25 and the reason behind binding to its own mRNA, we focused on the well-documented molecular event of TRIM25 K117 autoubiquitination (Inn et al., 2011) . We confirmed that TRIM25 undergoes autoubiquitination by co-overexpression and co-IP of T7-tagged TRIM25 together with HAtagged ubiquitin ( Figure S8a-S8b) . This was also confirmed by the in vitro ubiquitination experiments described below. Next, we asked if binding of TRIM25 to its own mRNA augmented TRIM25 ubiquitination. To address this, we overexpressed EGFP-TRIM25 or EGFP-TRIM25ΔRBD in HeLa cells and assayed for TRIM25 ubiquitination. We used EGFP fusion proteins to distinguish overexpressed and endogenous TRIM25. The overexpression of EGFP-TRIM25 led to the appearance of modified EGFP-TRIM25 ( Fig. 5a-5b) . However, upon EGFP-TRIM25ΔRBD overexpression, the intensity of the modified band was noticeably reduced (Fig. 5a-5b ). Of note, we had to use more EGFP-TRIM25ΔRBD plasmid to achieve protein expression comparable to that of EGFP-TRIM25. Surprisingly, overexpression of EGFP-TRIM25, but not EGFP-TRIM25ΔRBD, led to significant accumulation of modified forms of endogenous TRIM25, with much higher ratios of the modified forms compared to the non-modified form ( Fig. 5a-5b) . These results suggest that the intact RBD is required for efficient ubiquitination. They also show that endogenous TRIM25 can be modified by EGFP-TRIM25 to a much greater extent than EGFP-TRIM25 modifies itself. One important difference between endogenous TRIM25 mRNA and the mRNA coding for EGFP-TRIM25 is the presence of 5' and 3' UTRs, which were shown to bind TRIM25 in the CLIP-seq experiments ( Figure S3b ).