Insights into durability against resistance from the antibiotic nitrofurantoin
Properties of nitrofurantoin contributing to its durability against resistance
Rapid achievement of therapeutic concentrations specific to the site of infection
The favourable pharmacokinetic and pharmacodynamic (PK/PD) properties of nitrofurantoin treatment likely aid the suppression of resistance evolution18,19,20,21,22, as resistance has been experimentally demonstrated to evolve faster at low concentrations23. High therapeutic concentrations of nitrofurantoin are rapidly achieved, and specific to the site of infection in the urinary tract. After oral administration, nitrofurantoin is rapidly excreted unchanged in the urine, achieving concentrations of over 200 μg/ml3—well above the mutant prevention concentration of 64 μg/ml24. This allows nitrofurantoin to effectively kill bacteria in the urinary tract before the emergence of de novo-resistant mutants. Exploiting collateral sensitivity could further enhance killing, as mutants resistant to several antibiotics have increased susceptibility to nitrofurantoin25,26. In contrast, antibiotics that do not rapidly achieve high localised concentrations at the site of infection are more vulnerable to the development of resistance, as susceptible strains can proliferate and accumulate resistance mutations before the antibiotic reaches inhibitory concentrations. The rapid increase in concentrations impedes evolution particularly well when clinical resistance to a drug requires stepwise accumulation of multiple mutations—as is the case for nitrofurantoin27.
Low off-target concentrations
Nitrofurantoin rapidly achieves therapeutic concentrations at the site of infection (i.e. the bladder), with low off-target concentrations. While it achieves concentrations of over 200 μg/ml in urine3, concentrations in other bodily fluids are much lower; for example, typical plasma levels are less than 1 μg/ml, which is below the concentration needed to select for antimicrobial resistance mutations24. Nitrofurantoin is rapidly absorbed from the small intestine20, with only 2% of the dose recovered in faeces in a rat model28. This low concentration in the gut reduces the selection pressure for resistance within the microbiome, which can otherwise serve as a reservoir for resistant strains29. As a result, nitrofurantoin-resistant organisms are only rarely isolated from faeces30,31. Nitrofurantoin also does not cause dysbiosis of the normal intestinal microbiome32,33, a state that can result in increased colonisation by pathogenic organisms.
Multifactorial mechanism of action on bacterial physiology
An aspect of nitrofurantoin’s robustness against antibiotic resistance development likely stems from its multifactorial mechanism of action. Nitrofurantoin itself is not directly bactericidal; its activity requires the reduction of the nitro group in its molecular structure by bacterial nitroreductases after uptake into the cell34. This generates reactive intermediates that target multiple vital metabolic processes, such as protein synthesis, the citric acid cycle, and DNA/RNA synthesis3,35. These multifactorial effects within bacteria make resistance development highly complex and difficult to achieve, as changes (through mutations) to multiple mechanisms are required to concurrently overcome each damaging effect3.
Requirement for multiple independent mutations for clinical resistance
As nitrofurantoin’s mechanism of action involves several different bacterial nitroreductases, including NfsA and NfsB36, de novo resistance emergence requires changes to multiple targets. Separate inactivating mutations in both nfsA and nfsB are typically required for clinical levels of resistance and, indeed, are observed in clinically resistant organisms24,37,38. Crucially, the inactivation of just one out of nfsA and nfsB confers only a small increase in the minimum inhibitory concentration (MIC)24. Coupled with the favourable PK/PD properties of nitrofurantoin, this reduces the probability of resistance emerging via stepwise loss of nfsA and nfsB.
A single large deletion mutation that simultaneously inactivates both nfsA and nfsB is likewise an unlikely route to acquiring resistance. Despite being functionally similar, nfsA and nfsB are located far apart within the genome. In the type strain E. coli str. K-12, the distance between the two genes is 287 kbp24 (ranging between 262–337 kbp in other isolates); this genomic region contains 260 genes, of which 16 are identified as essential39. A genomic deletion inactivating both simultaneous genes would, therefore, likely be non-viable.
The proper functioning of nitroreductases is also dependent on other genes, including ribE, which is required to produce co-factors used by NfsA and NfsB. Inactivating mutations in ribE can also give rise to some degree of resistance37,40, but frequencies of ribE mutations among nitrofurantoin-resistant clinical isolates are low and are not found without nfsA and nfsB mutations36, suggesting other factors may constrain its contribution to resistance in vivo. Notably, intrinsically resistant organisms inherently lack the specific nitroreductases that convert nitrofurantoin into its active form5.
Fitness costs associated with resistance mutations
Fitness costs imposed by resistance mutations in the absence of antibiotics may contribute to an antibiotic’s durability. When antibiotic treatment is withdrawn, resistant organisms have a reduced ability to disseminate and establish infections. While nitrofurantoin resistance has been shown to impose a fitness cost, the extent and significance of this cost are debated. A recent study on clinical isolates found a 2–10% reduction in growth rate, though the results were not statistically significant24. An earlier study reported an average 6% reduction, with some isolates experiencing considerably higher costs41. Experimental studies have demonstrated up to a 10% reduction in competitive fitness, though this cost was able to be mitigated through compensatory adaptation42. However, it is important to note that these fitness costs were measured in vitro. To fully understand the implications of resistance, fitness costs should be assessed in vivo, where environmental and host factors may influence outcomes.
Resistance typically emerges due to mutations rather than horizontal gene transfer
For many antibiotics where resistance has become common, horizontal gene transfer (HGT) is the major route to resistance acquisition. This often involves antibiotic-inactivating enzymes, target bypass genes, or novel efflux pumps43. Such genes are often carried on mobile genetic elements like plasmids or integrons, or acquired through natural transformation, all of which facilitate spread between species44,45. Unlike many other antibiotics in common use, the first two resistance mechanisms play only a small part in nitrofurantoin resistance. There are limited reports of nitrofurantoin-inactivating enzymes in clinical isolates46, although the presence of nitrofurantoin-degrading enzymes in natural isolates suggests that horizontal gene transfer could eventually allow pathogens to acquire these resistance mechanisms47. Target bypass is unlikely to result in nitrofurantoin resistance, given that its mechanism of action does not rely on inhibiting a specific gene target. Recombination is also unlikely to transfer resistance because, as previously mentioned, the genomic region between the nfsA and nfsB is large24.
In contrast, the HGT of a novel multi-drug efflux pump, OqxAB, has been linked to nitrofurantoin resistance48. OqxAB was first detected in E. coli from pigs from Denmark49,50, and is associated with the feed additive olaquindox, which had been in use from the 1980s until an EU-wide ban in 1999. oqxAB was found to be carried on a conjugative plasmid (the IncX1-type plasmid pOLA5251), which has facilitated its spread between different bacteria. oqxAB is now carried by a variety of mobile genetic elements capable of transmitting between strains. The prevalence of oqxAB can exceed 50% in animal isolates from regions where olaquindox remains in use (including Australia52 and China53,54). In human isolates, prevalence ranges from 0.4–42%54, but is higher in regions using olaquindox, particularly among farmworkers55. Though OqxAB on its own confers only a modest increase in nitrofurantoin MIC, when combined with a nfsA mutation, it can confer resistance above the clinical breakpoint48. Work in other organisms has shown that the acquisition of novel efflux pumps can increase the potential for high-level resistance to evolve by allowing additional time for resistance mutations to emerge56. The potential for OqxAB to act as a similar “stepping stone” to full nitrofurantoin resistance in clinical isolates should be assessed.
The reduced role of HGT in nitrofurantoin resistance may be a fortunate consequence of its synthetic origins. Being a fully synthetic compound, nitrofurantoin likely lacks a natural reservoir of resistance genes. Many antibiotics in common use are derived from compounds originally produced by microbes and may have initially evolved for microbe-microbe competition57. Genes for synthesising antibiotics are evolutionarily ancient58,59—certainly long predating anthropogenic use of antibiotics60, and often contemporaneous with major evolutionary events59. Consequently, there has been an “arms race” between the evolution of antibiotic synthesis genes and antibiotic resistance genes that have been running for millions of years61. Consequently, in some cases, resistance to an antibiotic is already present in bacterial populations before its introduction into commercial use62. The use of fully synthetic antibiotics may help reduce resistance, to a degree63. However, not all synthetic antibiotics are inherently durable against resistance evolution; other synthetic antibiotics, such as fluoroquinolones and trimethoprim, exhibit significantly higher resistance64,65,66,67.
Insights into factors promoting nitrofurantoin resistance
Despite many properties favouring its durability against resistance evolution, nitrofurantoin resistance does emerge in both experimental and clinical contexts. Insight into how and why nitrofurantoin resistance does emerge can help ensure its continued efficacy.
While full nitrofurantoin resistance imposes a fitness cost in the absence of the drug24,41, the acquisition of an initial inactivating mutation in nfsA or nfsB can be selectively advantageous at sub-inhibitory nitrofurantoin concentrations, even without significantly changing the MIC24. Intriguingly, when resistance has been observed to arise in vitro, mutations tended to arise first in nfsA, followed by mutations in nfsB, due to the larger fitness benefits conferred by loss of nfsA under sub-inhibitory concentrations24.
Treatment noncompliance could play a significant role in contributing to nitrofurantoin resistance. As subinhibitory nitrofurantoin concentrations can facilitate resistance23,41, non-adherence with treatment regimens could allow these first-step mutants to proliferate and acquire second-step mutations. Ensuring consistent nitrofurantoin levels through full adherence is crucial to preventing the development of resistance. A patient-focused intervention emphasising the importance of adherence could help preserve the efficacy of nitrofurantoin, benefiting both the individual patient and the broader community by reducing the risk of resistance development and maintaining effective treatment options.
The use of antibiotics as prophylactics is a potential risk factor for increasing the prevalence of antibiotic resistance. Prophylactic use can expose bacteria to sub-therapeutic levels of antibiotics, promoting the selection and spread of resistant strains68. Existing evidence suggests the risk of nitrofurantoin resistance developing during prophylaxis is relatively low. A meta-analysis of controlled trials has shown that rates of nitrofurantoin resistance reported during trials were low69. In one study, nitrofurantoin prophylaxis did not increase resistance in children with urinary tract abnormalities70. However, a second study observed that E. coli was often replaced by resistant Klebsiella spp. and (intrinsically) resistant Pseudomonas spp71. Low-dose prophylaxis in adult women has generally shown low resistance rates72,73, although one study found nitrofurantoin prophylaxis increased resistance frequency relative to control (24% vs. 9%74). Generally, nitrofurantoin had lower resistance rates than alternative antibiotics72,73,74. While current evidence suggests prophylaxis poses a low risk for nitrofurantoin resistance, continued surveillance is recommended, given the emergence of new resistance mechanisms46,48,49,54.
A problematic development is the rise of nitrofurantoin resistance alongside multi-drug resistant (MDR) and extensively-drug resistant (XDR) isolates, particularly among the carbapenem-resistant Enterobacteriaceae (CRE). High-level nitrofurantoin resistance in CRE presents significant clinical challenges, complicating the treatment of UTIs. Nitrofurantoin-resistant E. coli isolates from the UK are more likely to also be MDR or XDR, compared to susceptible isolates75. Nitrofurantoin resistance is also found in non-E. coli CRE, including Klebsiella pneumoniae, Enterobacter spp. (including E. cloacae), Citrobacter freundii and others37,76. In addition to typical mutations in nfsA and nfsB, overexpression of efflux pumps37,77, and reduced drug uptake due to disruption of outer membrane porins77 likely contribute to nitrofurantoin resistance in CRE. The association between nitrofurantoin resistance and CRE is especially concerning, as nitrofurantoin has been reintroduced alongside carbapenems and colistin to counter multi-drug resistant infections76.
Other genomic and environmental factors affecting the evolvability of nitrofurantoin resistance warrant consideration. Recent experimental work has shown that “mutators” with defects in DNA replication fidelity can more easily acquire stepwise resistance mutations78,79 and overcome fitness deficits through compensatory adaptation80. Mutators pose a specific risk for nitrofurantoin resistance, increasing both the probability of resistance evolution and the specificity of resistance mutations81. The increasing prevalence of OqxAB efflux pumps also provides a potential mechanism that may make high-level nitrofurantoin resistance easier to evolve. The case of OqxAB, which was selected for in response to a different compound, also emphasises the need to better test for potential cross-resistance between existing and new drugs and chemicals prior to their release on the market.
Despite an increasing frequency of usage, nitrofurantoin has remained a durable antibiotic option for treating urinary tract infections. This durability is likely due to a few key factors, including targeted delivery to the site of infection and favourable PK/PD properties, multitarget effects on bacterial physiology, a requirement for multiple independent mutations for resistance to be acquired, fitness costs associated with resistance in the absence of nitrofurantoin, and the near-absence of horizontally acquired resistance. Continued judicious use of nitrofurantoin will hopefully continue to preserve this important antibiotic option for treating UTIs in the face of increasing antimicrobial resistance. However, more insight needs to be gained about factors that can increase the evolvability of nitrofurantoin resistance, including treatment non-adherence, the use of nitrofurantoin as a prophylactic treatment for recurrent UTI, bacterial mutation rate evolution, and the horizontal acquisition of novel efflux pumps. More generally, insight into mechanisms favouring the durability of nitrofurantoin may prove useful in the development of other durable antibiotics and treatment options.
Responses