This thesis is organized into three sections. The first section characterizes the biochemical properties of the enzymatic activity of angiogenin (ANG), a human RNase. The second part identifies the RNA target of VapC13, a toxin belonging to a type II toxin–antitoxin (TA) system in Mycobacterium tuberculosis (Mtb). The third part explores the possibility of toxin-mediated ribosomal frameshifting triggered by stalled ribosomes. Although these topics are conceptually distinct, they have one common theme: enlisting 5′ RNA-seq analysis to identify the RNA target and precise site of cleavage of endoribonucleases, a powerful approach developed in our laboratory that informs function. ANG is a vertebrate-specific RNase so named due to its first assigned function in promoting blood vessel formation. However, it also influences stress adaptation and survival in a variety of cell types. Consequently, altered ANG levels or mutations have been associated with certain cancers and neurodegenerative diseases. Although ANG is widely believed to act as a tRNase, we sought to define its enzymatic properties in vivo to better inform therapeutic approaches that seek to modulate its activity. ANG and its well characterized relative, RNase A, generate a 5’-hydroxyl (OH) end following RNA cleavage. Therefore, we applied our specialized 5’ RNA-seq method that defines cleavage specificity of endoribonucleases based on the moiety at the 5’ end of its product—5’-OH or 5’-monophosphate (P)—to unambiguously pinpoint the ANG cleavage recognition sequence, precisely map its position of cleavage within this sequence, and identify its RNA targets in vitro and in vivo. We performed 5’-OH RNA-seq on HEK293 and Escherichia coli RNA incubated with recombinant ANG as well as RNA isolated from HEK293 or E. coli cells overexpressing ANG. While all four approaches uncovered a strong ANG preference for RNA cleavage between CA residues, 5’-OH RNA-seq of ANG-overexpressing HEK293 cells was most revealing. First, the ANG recognition site begins with a strictly conserved CA flanked by sequences that tend to form a hairpin-like secondary structure. Second, there were no statistically significant tRNAs cleaved by ANG in 5’-OH RNA-seq datasets. Finally, tRNA cleavage was identified upon ANG overexpression in the 5’-P datasets only, suggesting that tRNA cleavage associated with ANG is a downstream, indirect effect. In the second section we study toxins RelE1/RelE2 and VapC13 which are type II TA systems in Mtb. Mtb genome encodes ~80 TA type II systems and they are thought to participate in the survival in different types of stresses. First, using 5’ RNA-seq we observed that RelE toxins in Mtb act the same way as in E. coli (which cleave RNA sequences in a sequence-independent manner). Second, using 5’-P RNA-seq we also demonstrated that tRNAAsn is the VapC13 target. Expression of VapC13 remodels Mtb transcriptome and translatome, leading to upregulation of genes involved in ribosome synthesis and translation as well as the pathway for phthiocerol dimycocerosate (PDIM). PDIM is a specialized lipid that resides on outermost surface the Mtb cell membrane that is essential for Mtb virulence. In the final section we report the first evidence suggesting that type II toxin–mediated depletion of a specific tRNA can induce ribosomal frameshifting at “hungry” codons. Using 5′ RNA-seq we detect signatures consistent with frameshifting, although protein-level validation remains challenging due to the low frequency of the event and limited proteomic depth following toxin expression. These findings raise the question of whether chimeric proteins produced during VapC/MazF toxin activation are functional. We provide one published example supporting the existence of a frameshifted IniB protein chimera—also predicted by our 5’ RNA-seq datasets—in a clinical Mtb strain resistant to the antibiotic isoniazid (INH) used to treat Mtb. IniB is induced by INH; thus stress-induced translational reprogramming might enable synthesis of alternative proteins that impart a selective advantage for Mtb stress survival.
Mikhail Bass (Thu,) studied this question.