Aminoglycoside antibiotics have broad-spectrum bactericidal activity, efficaciously treating life-threatening bacterial infections in neonates, infants, and adults.
These drugs remain essential for prophylaxis in patients with bone marrow transplants and other complex surgical histories, blast injuries, or cystic fibrosis, as well as for the resolution of tuberculosis and sepsis cases in developing countries because of the medications’ low cost and superior shelf life at ambient temperatures. However, aminoglycoside therapy also can damage host tissues, causing permanent ototoxicity, acute nephrotoxicity, and, less frequently, peripheral neuropathy.
In addition to being extraordinarily cheap, aminoglycosides are readily available and have a very low rate of drug-related allergic complications compared with more expensive, non-ototoxic, narrow-spectrum bactericidal antibiotics. Aminoglycosides are also effective against multidrug-resistant bacterial strains associated with tuberculosis and hospital-acquired, or nosocomial, infections.
The prevention of ototoxicity is critical to maintaining auditory function and quality of life in patients treated with these lifesaving drugs. The ototoxicity of aminoglycosides is most often attributed to the toxic generation of reactive oxygen species, though the source remains controversial, as well as protein mistranslation during ribosomal synthesis and hypersusceptibility to ototoxicity in people with specific mitochondrial mutations, namely A1555G.
APRAMYCIN: LESS OTOTOXICITY
Naturally occurring apramycin, an aminoglycoside widely used in veterinary medicine, has cytotoxicity properties distinct from those of aminoglycosides prescribed for human patients, like gentamicin and kanamycin, reported investigators based in the United States, England, and Switzerland (Proc Natl Acad Sci U S A 2012;109:10984-10989 http://www.pnas.org/content/109/27/10984.long).
In particular, apramycin had excellent antimycobacterial activity, Tanja Matt and colleagues reported. This finding presents particular implications for the treatment of patients who have tuberculosis caused by a Mycobacterium species, selected respiratory infections associated with cystic fibrosis, or Gram-negative bacterial infections.
The critical mechanistic evidence reported by these researchers is that apramycin has significantly less affinity for eukaryotic ribosomes than for bacterial ribosomes. This decreased affinity was especially evident for mitochondrial ribosomes of eukaryotes with the A1555G mitochondrial mutation. The reduced ribosomal binding resulted in less mitochondrial protein mistranslation, a leading contributor to eukaryotic cytotoxicity.
These points were emphasized by the significant reductions in threshold shifts and in hair cell loss from cochlear explants after chronic treatment with apramycin, even at higher dosing levels, compared with gentamicin.
Identification and Evaluation of Improved 4’-O-(Alkyl) 4,5-Disubstituted 2-Deoxy-streptamines as Next-Generation Aminoglycoside Antibiotics
Duscha S, Boukari H, Shcherbakov D, et al
In their latest paper, the researchers used the same ribosomal screening strategy for another aminoglycoside critical for treating protozoal parasitic infections, paromomycin. Crucially, two 4’-O-(alkyl) derivatives of paromomycin retained substantial bactericidal activity but had significantly reduced eukaryotic mitochondrial and cytosolic ribosomal dysfunction compared with the parental paromomycin.
This dichotomy was verified when the same two derivatives significantly reduced in vivo bacterial loads of multidrug-resistant Staphylococcus aureus in immunosuppressed mice and showed very little hearing loss and no systemic toxicity even at excessive doses in a guinea pig model of chronic ototoxicity. The derivatives also led to negligible hair cell loss after chronic dosing, even at higher dosing levels, compared with gentamicin, further validating the utility of the in vitro screening methods used.
These mechanistic and empirical data suggest that apramycin and derivatives of paromomycin may be useful for treating merging bacterial strains that become resistant to the aminoglycosides currently in widespread use.
Designer Aminoglycosides Prevent Cochlear Hair Cell Loss and Hearing Loss
Huth ME, Han K-H, Sotoudeh K, et al
J Clin Invest
An alternative screening strategy for modified aminoglycosides was also reported recently in the Journal of Clinical Investigation.
Markus E. Huth and colleagues modified a biosynthetic precursor of gentamicin, called sisomicin, and screened nine derivatives for their effects on hair cell survival in cochlear explants. Two derivatives showed negligible killing of hair cells.
All derivatives were also screened for their bactericidal efficacy. Those with modifications to the third ring had negligible bactericidal activity, demonstrating the importance of this ring for antibacterial activity.
Sisomicin derivatives with modifications to the second ring retained their bactericidal efficacy, and one derivative induced negligible hair cell death in vitro. This lead compound also displayed substantial broad-spectrum bactericidal activity against Escherichia coli and Klebsiella pneumoniae (including multidrug-resistant strains) but was less effective against Pseudomonas aeruginosa and S. aureus, principal bacterial flora in cystic fibrosis subjects, when compared with sisomicin, gentamicin, and tobramycin.
Electrophysiological data revealed greatly reduced binding kinetics to the mechanotransducer (MET) channel. Combining that finding with the reduced charge of the lead compound, N1MS, the authors concluded that N1MS had reduced entry into hair cells compared with sisomicin.
To demonstrate that N1MS displayed reduced ototoxicity and better hair cell survival compared with sisomicin, the authors used a well-established, acute paradigm of hair cell ablation via an aminoglycoside–loop diuretic combination that kills most murine cochlear hair cells within 48 hours.
Sisomicin elevated auditory thresholds and distortion product otoacoustic emissions (DPOAEs) within one week, and those measures remained consistently elevated five weeks later, indicating permanent threshold shifts. N1MS, on the other hand, induced minor high-frequency auditory brainstem response (ABR) threshold shifts with negligible changes in DPOAE thresholds at one week and five weeks after treatment.
These remarkable functional data were verified by the in vivo finding of negligible cochlear hair cell loss among N1MS-treated mice, boosting the potential for N1MS to be translated to market as a non-ototoxic antibiotic.
In preclinical models of urinary tract infections, N1MS was as effective as sisomicin in eradicating E. coli from urine and from bladder and kidney tissues, especially at higher doses.
As demonstrated by these data, aminoglycosides that do not effectively bind to eukaryotic ribosomes (especially the A1555G mutation) or are modified with a methylsulfonyl group on the central ring display reduced risks of hair cell death and permanent hearing loss. Both of the highlighted studies significantly advance our understanding of the mechanisms of bactericidal activity, showing strategies to ameliorate ototoxicity substantially (MBio 2014;5:e01827-14 http://mbio.asm.org/content/5/5/e01827-14 and J Clin Invest 2015;125:583-592 http://www.jci.org/articles/view/77424).
Nonetheless, bacteria continue to evolve and can develop resistance to currently used aminoglycosides by acquiring and adapting aminoglycoside-modifying enzymes. Bacteria resistant to gentamicin by enzymatic means will also be resistant to sisomicin and both drugs’ derivatives—likewise for paromomycin and streptomycin, and their derivatives.
Interestingly, apramycin and paromomycin are generated by bacterial species belonging to the Streptomyces genus, while gentamicin and sisomicin are generated by species of the Micromonospora genus, providing essential overlap in broad-spectrum bactericidal efficacy until resistance becomes widespread for both classes of aminoglycosides.
The collaborative studies by Erik C. Böttger's and Jochen Schacht's groups show a rational approach to identifying aminoglycosides with little or negligible ototoxic potential, particularly through the comparison of mitochondrial and cytosolic ribosomal binding kinetics (Proc Natl Acad Sci U S A 2012;109:10984-10989 http://www.pnas.org/content/109/27/10984.long and MBio 2014;5:e01827-14 http://mbio.asm.org/content/5/5/e01827-14). An added advantage of this research is that the chronic aminoglycoside dosing paradigm in rodents more closely reflected clinical conditions.
Unfortunately, the group's expertise did not extend to determining the binding affinity of apramycin or paromomycin derivatives for the MET channel or resolving whether these compounds entered hair cells at a reduced rate compared with other aminoglycosides. The dose-dependency of cochlear hair cell death suggests reduced entry and/or reduced intracellular cytotoxicity (via ribosomal dysfunction).
The alternative hair cell survival strategy by Anthony J. Ricci and Alan G. Cheng's group that first screened and then rationally identifed lead compounds was robustly validated using a combined aminoglycoside–loop diuretic dosing paradigm (J Clin Invest 2015;125:583-592 http://www.jci.org/articles/view/77424).
Accurate interpretation of the combined drug–loop diuretic data will be confounded by the reduced valency of N1MS, which would decrease endolymph loading and subsequent hair cell uptake/death compared with sisomicin. This mechanistic detail is minor compared with the illustration of retained bactericidal efficacy and, importantly, auditory function in preclinical models of bacterial infection.
The next crucial milestone for N1MS is a demonstration of ameliorated ototoxicity after chronic administration, a formidable task given the need to biosynthetically purify sufficient quantities of the compound to dose mice or guinea pigs (with increased weight) on a longer-term basis.
Another important milestone for novel aminoglycosides is determining whether they retain their reduced ototoxic potential in preclinical models of systemic infection and inflammation, as aminoglycosides are used clinically in these types of cases. Experimental endotoxemia, which has been shown to accelerate both cisplatin- and aminoglycoside-induced ototoxicity, is used as a model for sepsis-mediated or radiotherapy-induced inflammatory responses.
This milestone can also be adapted to validate the otoprotective properties of new compounds.