|About GSBS | FAQ | Job Opportunities | Search UMDNJ|
Swapnil C. Devarkar
Bachelor of Technology in Biotechnology, D.Y. Patil University (Mumbai), 2010
Thesis advisor: Dr. Smita Patel
Graduate Program in Biochemistry
RWJMS Research Tower
Wednesday, April 26, 2017
RIG-I (Retinoic acid Inducible Gene – I) is a helicase/ATPase that serves as a cytosolic innate immune receptor for viral dsRNAs containing 5’-triphosphate (5’ppp) and blunt-ends. Activation of RIG-I by these viral PAMP (Pathogen Associated Molecular Pattern) RNAs produces Type I interferons and cytokines that establish an antiviral state and potentiate the adaptive immune response. RIG-I has a high selectivity for PAMP RNAs and is not activated by cellular RNAs, despite their high concentrations. The mechanistic basis for this RNA selectivity and the contributions of individual RIG-I domains in RNA discrimination is not known. The role of ATP binding and hydrolysis in RIG-I function also lacks clarity. My thesis research work was aimed at addressing these questions.
I employed a combination of biochemical, biophysical, structural, and cell-based approaches with a panel of dsRNA containing PAMP and non-PAMP end-modifications to address RIG-I’s selectivity. My studies show that RIG-I binds to 5’ppp blunt-ended dsRNAs with a 100-1000 fold higher affinity as compared to dsRNAs with 5’ or 3’ overhangs. The C-terminal domain of RIG-I is primarily involved in this RNA selectivity, but the autoinhibitory CARD2-Hel2i interface interactions are also necessary for RNA selectivity. This work provides a role for the autoinhibitory CARD2-Hel2i interface in RNA selectivity, wherein it acts as a ‘gate’ to prevent non-PAMP RNAs from forming stable signaling competent complexes with RIG-I.
It was commonly thought that the 7-methyl guanosine (m7G) cap in cellular RNAs prevents RIG-I binding and activation. However, we find that capping itself is not sufficient in RIG-I evasion. The m7G capped (Cap-0) dsRNA binds to RIG-I with the same affinity as the 5`ppp dsRNA, activates RIG-I`s ATPase to the same extent, and shows the same signaling response. Crystal structures show that RIG-I can accommodate the m7G cap via conformational changes in the helicase-motif IVa without perturbing the triphosphate interactions. On the other hand, dsRNAs containing both m7G cap and 2’-O-methylation of 5’-end nucleotide ribose (Cap-1) show weak RNA binding and ATPase activity, and elicit an attenuated signaling response. The capping and 2’-O-methylation works synergistically, and the H830 residue of RIG-I is primarily responsible for preventing Cap-1 dsRNAs from activating RIG-I.
To understand the role of RIG-I’s ATPase activity, I measured the lifetimes of RIG-I-RNA complexes with PAMP and non-PAMP RNAs. In general, the PAMP dsRNAs generate long-lived complexes with RIG-I relative to non-PAMP RNAs. The lifetimes of RIG-I-RNA complexes however are greatly affected by ATP binding and hydrolysis, with opposing effects. ATP binding increases the lifetimes and ATP hydrolysis decreases the lifetimes of RIG-I-RNA complexes, and this is observed with both PAMP and non-PAMP RNAs. Thus, ATP hydrolysis is critical for recycling RIG-I complexes with PAMP and non-PAMP RNAs, and this explains the constitutively active phenotype of ATP binding but ATPase deficient RIG-I mutants, E373A and C268F, linked to the atypical Singleton Merten Syndrome. Because ATP hydrolysis does not selectively decrease the lifetimes of non-PAMPs, ATP hydrolysis does not provide a ‘proofreading’ function, which is contrary to existing models.
In summary, my studies have provided a deeper understanding of the mechanisms employed by the innate immune receptor RIG-I to distinguish self-versus-non-self RNA.