Aminoglycoside antibiotics (AGAs) are positively charged oligosaccharides that target bacterial ribosomes. AGAs induce misreading of the mRNA, resulting in the incorporation of incorrect amino acids into protein, and/or inhibit translocation, the movement of the ribosome along the mRNA. Many AGAs, such as paromomycin (Par), tobramycin (Tob), neomycin (Neo), gentamicin (Gen), kanamycin (Kan), or amikacin (Amk), are clinically important as broad-spectrum antibiotics. On the other hand, their propensity to induce translation errors can be utilized to alleviate the symptoms of human genetic diseases (e.g., Duchenne muscular dystrophy) by increasing readthrough of premature stop codons. The negative side effect of AGA treatment is its unfavorable effect on translation in mitochondria. In particular, the high oto- and nephrotoxicity due to targeting of mitoribosomes often hampers systemic administration of AGAs.
Although misreading is the key element of the AGA action, we know surprisingly little which types of errors are induced and how they affect the proteome and the fitness of bacterial cells. AGAs preferentially induce a subset of errors that result from a misreading of the third position of the mRNA codon, but errors in the first and second position can also occur. The type of most prevalent errors depends on the drug, e.g., streptomycin (Str) is less efficient than other AGAs in inducing U-G mismatches in the second position or G-U mismatches at any position of the codon–anticodon complexes. Severe mistranslation leads to growth arrest or cell death in bacteria. However, the AGA-induced error load alone cannot explain their bactericidal effects. In fact, ribosomal ambiguity (ram) mutants, with alterations in ribosomal protein S4, also cause increased misreading, but grow at almost wild-type (wt) rates. This led to the notion that there is something particular to the bactericidal effect of AGAs, but what causes their detrimental effect on bacterial fitness and cell viability remains unclear.
AGAs are polycationic molecules that bind to the negatively charged outer membrane of bacteria and enter the periplasm via porin channels. Once in the cytosol, AGAs bind to the ribosome and induce misreading by stabilizing an error-prone conformation of the decoding center of the ribosome. Accumulation of translation errors in membrane proteins leads to the disintegration of membrane structures, renders the membrane permeable for small molecules, and allows for a massive influx of AGAs. However, how sub-lethal intracellular AGA concentrations and the associated mild increase in mistranslation cause damage of membrane proteins is unknown. At bactericidal concentrations, proteotoxic stress induces the heat shock response, aggregation of specific sub-proteomes, including membrane and metabolic proteins, protein oxidation and carbonylation, and inclusion body formation. Further downstream effects of AGA treatment are dramatic changes of the cell metabolism, oxidative stress, DNA damage, and ultimately cell death. In contrast, sub-lethal AGA concentrations may even be beneficial for bacteria, as they help cells to adapt to stress conditions, change to a more drug-resistant lifestyle (e.g., biofilm formation), or acquire antibiotics resistance.
The majority of AGAs affect both decoding and translocation steps of translation. The efficacy of AGAs is often attributed to their ability to induce misreading, rather than to their effect on translocation. In fact, Str, which is bactericidal and induces misreading, has no effect on translocation. Vice versa, spectinomycin (Spc), an AGA that causes a translocation defect without inducing misreading, is bacteriostatic. One potential exception is apramycin (Apr) which is bactericidal, although recent studies suggested that it does not induce misreading. This is important, because Apr is less ototoxic than other AGAs in model organisms, presumably because its antibacterial and anti-mitoribosomal activities are uncoupled. Moreover, due to its unique structure, Apr is not susceptible to many prevalent resistance mechanisms and is thus a promising drug for the treatment of multiresistant bacteria. Somewhat paradoxically, the bactericidal effect of AGAs is lost when a translation is blocked by bacteriostatic antibiotics, such as chloramphenicol.
Fig: Fitness, stress response, and miscoding of Str-treated wtand non-treated error-prone ramcells
Here we used quantitative mass spectrometry to correlate the AGA-dependent cell growth inhibition, miscoding burden, and stress responses, and to obtain insights into the bactericidal effects of AGAs. We found that AGA binding induces not only single errors, but clusters of errors with two, three or four amino acid substitutions located close to each other in the protein sequence. We characterize this type of misreading events for a variety of AGAs and show that their prevalence depends not only on the misreading propensity of a given AGA but also on its ability to inhibit translocation. We also show that Apr, an AGA which is thought not to cause misreading, induces frequent translation errors. These results reveal an additional, unexpected aspect of AGA action in bacteria, suggest how antibiotics cause proteotoxic stress, and provide a simple explanation as to why some misreading-inducing antibiotics are bactericidal.
Wohlgemuth, I., Garofalo, R., Samatova, E. et al. Translation error clusters induced by aminoglycoside antibiotics. Nat Commun 12,1830 (2021). https://doi.org/10.1038/s41467-021-21942-6