Regulation of the protein kinase PKR by higher-order RNA secondary and tertiary structures

Open Access
Heinicke, Laurie Ann
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
March 18, 2010
Committee Members:
  • Philip C. Bevilacqua, Dissertation Advisor/Co-Advisor
  • Philip C. Bevilacqua, Committee Chair/Co-Chair
  • Christine Dolan Keating, Committee Member
  • Scott A Showalter, Committee Member
  • Andrey S Krasilnikov, Committee Member
  • innate immune response
  • kinase
  • RNA
  • PKR
A single strand of RNA can fold into multiple structures, often forming complex secondary and tertiary stuctures. Many factors influence RNA folding, including proteins, small molecules, salt, temperature, and pH. Human protein kinase PKR is a component of the innate immune response, and is activated by long stretches of double-stranded RNA (dsRNA), but is inhibited by highly structured or short dsRNAs. Activation of PKR often results in inhibition of protein synthesis. The types of RNAs encountered by PKR in the cell are diverse in structure and function, including mRNA, miRNA, tRNA, rRNA and viral RNAs. While PKR regulation by human cellular RNAs has not been extensively studied, it is known that PKR is activated by viral dsRNA genomes and long dsRNA replicative intermediates. The work presented in this thesis focused primarily on trapping and characterizing a unique RNA fold and then examining PKR regulation by that RNA. Experiments were performed using a diverse collection of viral and cellular RNAs, as well as model RNAs. Viral RNAs, including a small hairpin (TAR) from human immunodeficiency virus type 1 (HIV-I) and various length sequences from hepatitis delta virus (HDV), are single-stranded, but form complex secondary and/or tertiary structures. Both TAR RNA and the ribozyme portion of HDV are reported herein to form dimers and/or aggregates, which are found to be PKR activating species. Longer HDV sequences, which are rod-shaped and do not include continuous double-stranded RNA sequence greater than 20-bp, are still found to be potent PKR activators. Cellular RNAs have not been extensively examined as PKR regulators. It was our hyphothesis that most cellular RNAs have evolved defects to avoid activating or inhibiting PKR. We examined one cellular RNA, precursor miRNA 374a, which is reported herein to be an inhibitor of PKR. Precursor miRNAs are ~70-nt and are very structured with multiple bulges and mismatches. Removing defects and replacing them with Watson-Crick GC base pairs improves the ability of partially truncated pre-miRNAs to inhibit PKR. Another finding is that PKR activation at low pH is RNA-independent and represents a novel PKR response. Surprisingly, partially truncated pre-miRNA sequences were very potent inhibitors of RNA-independent activation of PKR at low pH, while only a weak inhibitors at high pH. Lastly, three separate studies were performed using model RNAs in an effort to understand the effects of RNA helical defects and dimerization on PKR activation, as well as to attempt to crystallize a minimal RNA binding register with p20 (dsRNA binding domain of PKR). The first study examined the effects of RNA helical defect position, size, and geometry on PKR activation. Positioning a bulge in the center of a helix, increasing bulge size, and orienting RNAs in cis or trans geometry decreased PKR activation relative to perfect double-stranded RNA. These results may be useful for predicting PKR activators. The second study investigated PKR activation by ternary complexes comprised of TAR dimers linked by base pairing to a bridging DNA oligo. Although all attempts to design a PKR activating complex were unsuccessful, rationale for designing complexes and native gel analyses are provided. Lastly, a third study involved preparing a ternary complex comprised of p20/stem-loop RNAs/U1A for crystallography. Preliminary gel-shift data and purification protocols are provided. Overall, the work in this thesis provides guidelines for predicting the ability of an RNA to activate or inhibit PKR.