Adaptation to Overflow Metabolism by Mutations That Impair tRNA Modification in Experimentally Evolved Bacteria.

When microbes grow in foreign nutritional environments, selection may enrich mutations in unexpected pathways connecting growth and homeostasis. An evolution experiment designed to identify beneficial mutations in Burkholderia cenocepacia captured six independent nonsynonymous substitutions in the essential gene tilS, which modifies tRNAIle2 by adding a lysine to the anticodon for faithful AUA recognition. Further, five additional mutants acquired mutations in tRNAIle2, which strongly suggests that disrupting the TilS-tRNAIle2 interaction was subject to strong positive selection. Mutated TilS incurred greatly reduced enzymatic function but retained capacity for tRNAIle2 binding. However, both mutant sets outcompeted the wild type by decreasing the lag phase duration by ~3.5 h. We hypothesized that lysine demand could underlie fitness in the experimental conditions. As predicted, supplemental lysine complemented the ancestral fitness deficit, but so did the additions of several other amino acids. Mutant fitness advantages were also specific to rapid growth on galactose using oxidative overflow metabolism that generates redox imbalance, not resources favoring more balanced metabolism. Remarkably, 13 tilS mutations also evolved in the long-term evolution experiment with Escherichia coli, including four fixed mutations. These results suggest that TilS or unknown binding partners contribute to improved growth under conditions of rapid sugar oxidation at the predicted expense of translational accuracy. IMPORTANCE There is growing evidence that the fundamental components of protein translation can play multiple roles in maintaining cellular homeostasis. Enzymes that interact with transfer RNAs not only ensure faithful decoding of the genetic code but also help signal the metabolic state by reacting to imbalances in essential building blocks like free amino acids and cofactors. Here, we present evidence of a secondary function for the essential enzyme TilS, whose only prior known function is to modify tRNAIle(CAU) to ensure accurate translation. Multiple nonsynonymous substitutions in tilS, as well as its cognate tRNA, were selected in evolution experiments favoring rapid, redox-imbalanced growth. These mutations alone decreased lag phase and created a competitive advantage, but at the expense of most primary enzyme function. These results imply that TilS interacts with other factors related to the timing of exponential growth and that tRNA-modifying enzymes may serve multiple roles in monitoring metabolic health.

Flexibility of the Rotavirus NSP2 C-Terminal Region Supports Factory Formation via Liquid-Liquid Phase Separation.

Many viruses sequester the materials needed for their replication into discrete subcellular factories. For rotaviruses (RVs), these factories are called viroplasms, and they are formed in the host cell cytosol via the process of liquid-liquid phase separation (LLPS). The nonstructural protein 2 (NSP2) and its binding partner, nonstructural protein 5 (NSP5), are critical for viroplasm biogenesis. Yet it is not fully understood how NSP2 and NSP5 cooperate to form factories. The C-terminal region (CTR) of NSP2 (residues 291 to 317) is flexible, allowing it to participate in domain-swapping interactions that promote interoctamer interactions and, presumably, viroplasm formation. Molecular dynamics simulations showed that a lysine-to-glutamic acid change at position 294 (K294E) reduces NSP2 CTR flexibility in silico. To test the impact of reduced NSP2 CTR flexibility during infection, we engineered a mutant RV bearing this change (rRV-NSP2K294E). Single-cycle growth assays revealed a >1.2-log reduction in endpoint titers for rRV-NSP2K294E versus the wild-type control (rRV-WT). Using immunofluorescence assays, we found that rRV-NSP2K294E formed smaller, more numerous viroplasms than rRV-WT. Live-cell imaging experiments confirmed these results and revealed that rRV-NSP2K294E factories had delayed fusion kinetics. Moreover, NSP2K294E and several other CTR mutants formed fewer viroplasm-like structures in NSP5 coexpressing cells than did control NSP2WT. Finally, NSP2K294E exhibited defects in its capacity to induce LLPS droplet formation in vitro when incubated alongside NSP5. These results underscore the importance of NSP2 CTR flexibility in supporting the biogenesis of RV factories. IMPORTANCE Viruses often condense the materials needed for their replication into discrete intracellular factories. For rotaviruses, agents of severe gastroenteritis in children, factory formation is mediated in part by an octameric protein called NSP2. A flexible C-terminal region of NSP2 has been proposed to link several NSP2 octamers together, a feature that might be important for factory formation. Here, we created a change in NSP2 that reduced C-terminal flexibility and analyzed the impact on rotavirus factories. We found that the change caused the formation of smaller and more numerous factories that could not readily fuse together like those of the wild-type virus. The altered NSP2 protein also had a reduced capacity to form factory-like condensates in a test tube. Together, these results add to our growing understanding of how NSP2 supports rotavirus factory formation-a key step of viral replication.

Reverse genetic engineering of simian rotaviruses with temperature-sensitive lesions in VP1, VP2, and VP6.

Rotaviruses are 11-segmented double-stranded RNA viruses and important causes of acute gastroenteritis in young children. To investigate the functions of specific viral proteins during the rotavirus lifecycle, temperature-sensitive (ts) mutants were previously created using a cultivatable simian strain (SA11) and chemical mutagenesis. These ts SA11 mutants replicate more efficiently at the permissive temperature of 31 °C than at the non-permissive temperature of 39 °C. Prototype strains SA11-tsC, SA11-tsF, and SA11-tsG were mapped to the genes encoding structural proteins VP1, VP2, and VP6, respectively, and putative ts lesions were identified using Sanger sequencing. However, additional background mutations in their genomes had hampered validation of the ts lesions and confounded their use in mechanistic studies. Here, we employed plasmid only-based reverse genetics to engineer recombinant (r) SA11 rotaviruses containing only the putative ts lesions of SA11-tsC (L138P change in VP1), SA11-tsF (A387D change in VP2) or SA11-tsG (S10T, D13H, and A121G changes in VP6). For simplicity, we refer to these newly-engineered, isogenic viruses as rSA11-tsVP1, rSA11-tsVP2, and rSA11-tsVP6. Single-cycle growth assays revealed that these mutants indeed exhibit ts phenotypes with significantly diminished titers (>1.5-logs) at 39 °C versus 31 °C. The rSA11 ts mutants proved genetically stable at the population-level following 3 sequential passages at 39 °C, but individual revertant clones were detected in plaque assays. Heat sensitivity experiments showed that pre-incubation of rSA11-tsVP1 or rSA11-tsVP2, but not rSA11-tsVP6, at 39 °C diminished replication at 31 °C. This result indicates that the ts lesions in VP1 and VP2 affect the incoming virion but those in VP6 affect a later stage of the viral lifecycle. In silico molecular dynamics simulations predicted temperature-dependent, long-range effects of the S10T, D13H, and/or A121G changes on the VP6 structure. Altogether, our results confirm the ts lesions of the original SA11-tsC, SA11-tsF, and SA11-tsG mutants, provide a new set of isogenic strains for investigating aspects of rotavirus replication, and shed light on how the ts lesions might impact VP1, VP2, or VP6 functions.

Bacterial wobble modifications of NNA-decoding tRNAs.

Nucleotides of transfer RNAs (tRNAs) are highly modified, particularly at the anticodon. Bacterial tRNAs that read A-ending codons are especially notable. The U34 nucleotide canonically present in these tRNAs is modified by a wide range of complex chemical constituents. An additional two A-ending codons are not read by U34-containing tRNAs but are accommodated by either inosine or lysidine at the wobble position (I34 or L34). The structural basis for many N34 modifications in both tRNA aminoacylation and ribosome decoding has been elucidated, and evolutionary conservation of modifying enzymes is also becoming clearer. Here we present a brief review of the structure, function, and conservation of wobble modifications in tRNAs that translate A-ending codons. © 2019 IUBMB Life, 2019 © 2019 IUBMB Life, 71(8):1158-1166, 2019.