Unlocking Nitrofurantoin: Understanding Molecular Mechanisms of Action and Resistance in Enterobacterales

Abstract

Antimicrobial resistance (AMR) is a global health crisis that has already claimed millions of lives and is projected to affect millions more unless urgent action is taken. Effective control of AMR requires the correct choice and dosage of antibiotics, as well as robust surveillance and research. Understanding the mechanisms of antibiotic action and the emergence of resistance phenotypes along with their genotypes is essential. This knowledge, combined with insights into resistance prevalence and spread, empowers clinicians to propose alternative therapies. Nitrofurantoin, a 70-year-old antibiotic, remains effective for the treatment of uncomplicated lower UTIs. Preventing emergence and spread of nitrofurantoin-resistant superbugs would preserve the efficacy of this antibiotic which is crucial for ongoing and future AMR efforts. Nitrofurantoin resistance evolves slowly, leading to low prevalence compared to other antibiotics. However, it is often linked with extensive drug resistance, complicating treatment outcomes. Even a minor percentage of nitrofurantoin-resistant bacteria can cause significant clinical challenges due to irreversible evolution. While detailed study of these mechanisms can guide the development of strategies to combat nitrofurantoin resistance, early detection of resistant infections is critical for saving lives. The current review aimed to provide a comprehensive analysis of nitrofurantoin’s mechanisms of action, resistance evolution, prevalence, and resistance prediction. Our goal is to offer valuable insights for researchers and clinicians to enhance nitrofurantoin use and address the challenges posed by AMR.

Highlights of the Study

• Repurposing old antibiotics is crucial for combating antibiotic resistance in bacteria.

• Nitrofurantoin, used for over seven decades, has relatively low resistance levels.

• This review analyzes research on the mechanisms of action of nitrofurantoin and the development of nitrofurantoin resistance in Enterobacterales.

• It also discusses phenomena like heteroresistance, co-resistance, collateral sensitivity, fitness of resistant mutants, and resistance prediction.

Introduction

Nitrofurantoin, chemically defined as 1-[(5-nitrofuran-2-yl) methylidene] amino imidazolidin-2,4-dione, is a synthetic antibacterial agent shown in Figure 1. Belonging to the class of antibiotics known as “nitrofurans,” it is known to exhibit a broad-spectrum action, demonstrating bacteriostatic properties at lower concentrations (<1 mg/mL) and bactericidal effects at therapeutic levels (i.e., 100–200 mg every 12 h) [1–3]. The nitrofuran class comprises of four parent compounds, i.e., furazolidone, furaltadone, nitrofurazone, and nitrofurantoin [4]. However, other novel antibacterial prodrugs (e.g., IITR06144) of the nitrofuran class have been described with their in vitro and in vivo properties [5]. Being orally administered, it undergoes rapid absorption in the intestine and subsequent excretion in urine, resulting in elevated therapeutic levels within the urinary tract. Consequently, it is commonly prescribed for treating uncomplicated lower urinary tract infections (UTIs). Initially approved by the US FDA in 1953 for UTI treatment, nitrofurantoin is marketed under various commercial names such as Furadantin, Furent, Niftran, and Macrobid, available as tablets, capsules, or suspension [6].

Fig. 1.

Chemical structure of nitrofurantoin.

Nitrofurantoin served as the primary antibiotic choice for treating lower UTIs for nearly two decades, spanning from the 1950s to the 1970s, until the commercialization of trimethoprim/sulfamethoxazole and the introduction of newly developed β-lactam antibiotics. Recently, various national and international guidelines including Infectious Disease Society of America (IDSA), European Society of Clinical Microbiology and Infectious Diseases (ESCMID), European Association of Urology (EAU), and Indian Council of Medical Research (ICMR) [1, 2, 7–9] have reinstated nitrofurantoin as the preferred first-line therapy for uncomplicated lower UTIs. The shift is driven by increasing resistance to new antibiotics and the rise of bacteria producing extended-spectrum β-lactamase (ESBL) [10–14]. This underscores the importance of using nitrofurantoin once again.

Nitrofurantoin is primarily used for both preventing and treating UTIs. In a typical healthy individual, its bioavailability is approximately 80%. A notable advantage of nitrofurantoin is its ability to accumulate in the lower urinary tract while maintaining low serum concentrations. Significantly, nitrofurantoin has minimal impact on gut flora, a factor that may contribute to its sustained effectiveness and low resistance rates [15–17]. Several common UTI pathogens [18] including both Gram-positive (Enterococcus, Staphylococcus saprophyticus, group B streptococci, S. aureus, S. epidermidis, etc.) and Gram-negative (Escherichia, Klebsiella, Enterobacter, Shigella, Salmonella, Citrobacter, Neisseria, Bacteroides, etc.) bacteria are susceptible to the treatment with nitrofurantoin.

Administration and Side Effects

Nitrofurantoin is exclusively administered orally. It gained approval long before the implementation of modern rigorous drug development protocols. The most recent guidelines from the IDSA suggest a dosage of 100 mg of nitrofurantoin twice daily for 5–7 days in treating lower UTIs. Deviating from this prescribed dosage and duration is not advised, as it has been observed to yield lower efficacy. For long-term prophylaxis (i.e., regular medication to prevent recurrent infection) against UTIs, a recommended dosage of 50–100 mg once daily before bedtime is suggested [19, 20].

When compared to other medications used in the treatment of lower UTIs, nitrofurantoin is regarded as the safest antibiotic option. However, a minority of patients may experience common side effects, including nausea, vomiting, loss of appetite, and diarrhea. Serious adverse effects such as pulmonary toxicity occur in approximately 0.001% of cases. These pulmonary reactions are reversible upon immediate cessation of antibiotic use [21].

Nitrofurantoin is contraindicated in patients with impaired renal function when prescribed for extended periods, as rare adverse effects such as peripheral neuropathy have been observed [21]. Furthermore, nitrofurantoin is also contraindicated in patients with bacterial pyelonephritis, anuria, oliguria, or significant renal impairment. This restriction stems from the fact that nitrofurantoin does not attain therapeutic levels in the upper urinary tract, and these conditions often coincide with bacteremia. Additionally, nitrofurantoin is not recommended for treating UTIs in pregnant women (38–46 weeks gestation), neonates (under one month old), men with prostatitis, and elderly patients (over 65 years old). This cautionary stance is likely due to the risk of hemolytic anemia resulting from an immature erythrocyte enzyme system, nitrofurantoin’s limited ability to penetrate prostatic tissue effectively, as well as potential pulmonary toxicity, hepatotoxicity, and peripheral neuropathy [6]. A detailed narrative of nitrofurantoin as a potent antibiotic and its side effects are available in a review by Ari et al. [22].

Mechanisms of Nitrofurantoin Action

Discovered in the 1940s, nitrofurantoin’s precise mechanism of action remains poorly elucidated to this day, primarily due to its targeting of multiple pathways involved in antimicrobial activity. The most commonly proposed mechanism involves the reduction of nitrofurantoin’s nitro group by bacterial intracellular nitroreductases, leading to the formation of its active reduced form. Intermediates generated from this reduction process, such as nitro-anion-free radicals and hydroxylamine [23], nonspecifically bind to ribosomes, inhibiting bacterial enzymes crucial for the synthesis of DNA, RNA, proteins, and other essential metabolites (Fig. 2) [24].

Fig. 2.

Mechanism of action of nitrofurantoin (adapted and modified from Khamari et al. [25]). a The primary function of nitroreductases (NfsA and NfsB) in most Enterobacterales including E. coli is to reduce nitro compounds into amino compounds using electrons from NADPH/NADH and in the presence of the cofactor FMN. b Nitrofurantoin accumulated in the bacterial cytosol undergoes stepwise reduction by oxygen-insensitive nitroreductases (product of the genes nfsA and nfsB) in the presence of cofactor FMN (synthesized by products of the genes ribE, ribC, and ribB) using electrons from NADPH/NADH. c The intermediates produced during these catalyses inhibit bacterial growth by affecting the DNA/protein synthesis or by interfering in cellular metabolic pathways.

Nitrofurantoin’s mechanism of action was originally described in the 1950s. Initial studies by Asnis et al. [26, 27] indicated that Escherichia coli differed in their ability to reduce nitrofurantoin based on the strain’s resistance and susceptibility to nitrofuran compounds. This was the first study that suggested the need for nitrofurans to get activated to deliver their antimicrobial effect. Grant and De Szöcs [28] reported that at lower concentrations, nitrofurantoin can inhibit the inducible synthesis of β-galactosidase in Klebsiella aerogenes without affecting the total protein synthesis of the bacterium. Further, Herrlich and Schweiger [29] showed that nitrofurantoin (lower concentrations) can inhibit the synthesis of β-galactosidase and galactokinase in E. coli. McCalla [30] reported that nitrofurantoin at higher concentrations can inhibit enzymes involved in the citric acid cycle in addition to interference in the DNA, RNA, and protein synthesis among bacteria. It was proposed that the mode of nitrofurantoin action involves a reaction of electrophiles (generated after the reduction of nitrofurantoin in bacteria) with nucleophilic sites present on various macromolecules of bacteria [30].

Peterson et al. [31] reported the presence of two types of nitroreductase activities in E. coli, i.e., type I insensitive to oxygen and type II inhibited by oxygen. In type I nitro reduction, the reduction of nitro-moiety of the parent compound involves a sequence of intermediates, which include a nitroso and hydroxylamine step, finally leading to a biologically inactive end product. One or more of these intermediates were thought to be responsible for the toxicity of nitrofuran antimicrobials. McOsker et al. [32] supported the mechanism of reductive activation of nitrofurantoin. Further, these reports revealed the ubiquitous presence of nitroreductase enzymes among all the major uropathogenic bacteria.

These findings suggested an inverse relationship between the levels of reductases and the minimum inhibitory concentrations (MICs) of nitrofurantoin [32]. Additionally, they suggested that at bactericidal concentrations, nitrofurantoin could hinder total protein synthesis in bacteria through a nonspecific qualitative attack on ribosomal proteins and rRNA. It was proposed that the antibacterial activity of the drug may not solely rely on reductive activation, as the rapid reduction of nitrofurans by the NfsA and NfsB nitroreductases might obscure alternative mechanisms employed by the bacterium.

Mechanisms that involve genotoxic properties of nitrofurantoin through the SOS pathway were first reported by Jenkins and Bennett [33] and Lu et al. [34] as they observed nitrofurantoin to be unusually toxic to bacterial strains that are defective in DNA recombination repair mechanisms (e.g., mutated recA 13 gene). Toxicity paralleled with the mutagenic activity of nitrofurantoin on E. coli was shown by Lu et al. [34], Kada et al. [35], and Tazima et al. [36]. The work done by Sengupta et al. [37] suggested that nitrofurantoin can induce the SOS repair pathway and cause DNA interstrand cross-links. Excision and recombination (including the SOS repair) pathways were found to be involved in the repair of DNA damage caused by nitrofurantoin.

Nitrofurantoin has been used for over 70 years, and it is notable among commonly prescribed antibiotics for its relatively low levels of clinically significant antibiotic resistance. Unlike other antibacterial agents such as ampicillin or trimethoprim/sulfamethoxazole, which target specific pathways, the reduced intermediates of nitrofurantoin impact multiple pathways. The existence of multiple mechanisms of action shown in Figure 2 for nitrofurantoin could potentially account for bacteria’s limited ability to develop resistance against it [25].

Nitrofurantoin Resistance Mechanisms

Until recently, clinically significant resistance to nitrofurantoin was seldom observed. However, resistance to nitrofurantoin among Enterobacteriaceae has been steadily rising over the past two decades in multiple parts of the world. The primary mechanisms of nitrofurantoin resistance shown in Table 1 and Figure 3 employed by prominent uropathogenic Enterobacterales have been discussed in this section.

Table 1.

Summary of genes associated with nitrofurantoin resistance

Gene	Event	Associated mechanism	Studied in pathogens	Implication	Reference
nfsA	Loss-of-function mutations	Reduction of nitrofurantoin into toxic intermediates	E. coli, S. enterica, K. pneumoniae	Intermediate- to high-level resistance	[25, 38–52]
nfsB	Loss-of-function mutations	Reduction of nitrofurantoin into toxic intermediates	E. coli, S. enterica, K. pneumoniae	Wild-type susceptibility to low-level resistance
ribE	Multiple nucleotide deletion	Biosynthesis of cofactor FMN for NfsA/NfsB	E. coli	Resistance in in vitro mutants	[53–55]
ahpF	Mutation/deletion/overexpression	5-Nitrofuran-activating reductase; protection from oxidative stress	E. coli	Limited expression increases resistance; overexpression increases susceptibility	[23, 56]
ribC, ribB	Base substitution; disrupted promoter	Riboflavin (precursor of FMN) biosynthesis	K. pneumoniae	Resistance in in vitro mutants	[54, 55]
oqxA/oqxB	Overexpression	Drug efflux	E. coli, S. enterica, K. pneumoniae	High-level resistance	[57–64]
acrA/acrB	Overexpression	Drug efflux	K. pneumoniae	Knockout can lead to up to 16× reduction in MIC	[54, 64]
ramA	Overexpression	Regulation of efflux pumps	K. pneumoniae	Increased resistance	[64]
ramR	Mutational inactivation	Repressor of RamA efflux pumps	K. pneumoniae	High-level resistance	[64]
oqxR	Loss-of-function mutations	Repressor of OqxA/OqxB efflux pumps	K. pneumoniae	High-level resistance in in vitro mutants	[54, 65]
recB	Loss-of-function mutations	Defective DNA repair pathway	E. coli	Increased susceptibility	[66]
soxR/soxS	Gene knockout	Protection from oxidative stress/regulation of nfsA	E. coli	Non-induction of nfsA leading to resistance	[67, 68]
mprA	Loss-of-function mutations	Multidrug resistance repressor	E. coli	Resistance in in vitro mutants	[69]
ompR/ompK36	Loss-of-function mutations	Outer membrane porin regulation	E. coli, K. pneumoniae	Repression of drug influx leading to resistance	[54, 69]
rpoA-rpoB	Loss-of-function mutations	RNA polymerase subunits	E. coli	Resistance in in vitro mutants	[69]
CTX-M-14	Gain-of-function mutation	Hydrolysis of hydantoin ring – lowering affinity to nitroreductases	E. coli	Hyper-resistance	[70]

Fig. 3.

Primary mechanism of nitrofurantoin resistance (adapted and modified from Khamari et al. [25]). a Loss-of-function mutations in genes nfsA, nfsB, ribE, ribC, or ribB can lead to the production of proteins that have lost specificity for nitrofurantoin or are truncated/faulty. This results in the production of no active intermediates that can attack the cell machinery, thus leading to resistance. Similar mutations in genes encoding repressors of efflux pumps (oqxR, ramR) can lead to overexpression of efflux pumps. b Overexpression of OqxAB, AcrAB, RamA efflux pumps further enhances the resistance by pumping the accumulated nitrofurantoin out of bacterial cell. c Mutations in housekeeping genes like nfsA, nfsB, ribE, etc., can affect the general physiological pathways of the bacterium, which may lead to compromised fitness in the resistant mutants.

Major Contributors of Nitrofurantoin Resistance

Nitrofurantoin-Activating Enzymes

In 1975, McCalla et al. [38] identified three distinct, separable nitrofuran reductases in E. coli, noting the absence of one of these enzymes in certain nitrofurazone-resistant mutant strains. In 1977, they discovered two components of nitrofuran reductase I in E. coli K-12, encoded by distinct genes linked to the galactose operon [39]. Through a single mutational event, E. coli K-12 could achieve a threefold increase in resistance to nitrofurantoin. These partially resistant mutants could subsequently undergo a second mutation, rendering them ten times more resistant than the wild type. The initial mutational step was associated with a partial loss of nitrofuran reductase activity, while the second step led to the loss of the remaining activity [39]. The genes controlling nitroreductase activity were designated as “nitrofuran sensitivity genes” (nfsA and nfsB). Thus, wild-type strains were deciphered as nfsA⁺nfsB⁺, and the resistant double mutants as nfsA⁻nfsB⁻. A variety of crosses were performed which established that both the genes are located close to the gal gene. The highly plausible genetic context was proposed to be lac-nfsB-gal-nfsA. The single-step mutants with an intermediate level of resistance to nitrofurans were nfsA⁻nfsB⁺ with only 20 to 30% of the wild-type nitroreductase I activity. In contrast, the nfsA⁺nfsB⁻ strains exhibited about 70 to 80% of the wild-type nitroreductase activity, apparently enough to confer wild-type susceptibility [39]. In 1980, Bryant et al. [40] revealed the existence of at least three type I nitroreductase components in E. coli. They observed that nitrofurazone-resistant mutants could progressively lose individual reductase components in a stepwise fashion as they acquired increased resistance. Furthermore, they determined that the primary nitroreductase (IA or nfsA) specifically relied on NADPH for its activity, whereas both IB1/nfsB and IB2/nfsC (gene not yet mapped) demonstrated activity with either NADH or NADPH as cofactors.

In 1984, Sastry and Jayaraman [41] isolated novel mutants of E. coli resistant to nitrofurantoin and mapped mutations in the two nitroreductase genes (called nfnA and nfnB, instead of nfsA and nfsB) governing resistance to nitrofurantoin on the E. coli chromosome. Subsequently, both the nfsA and nfsB genes were identified, cloned, expressed, and the respective enzymes were characterized [42, 71]. NfsA has a monomeric molecular mass of 26.8 kDa and contains flavin mononucleotide (FMN) as a prosthetic group [40, 71]. The NfsB flavoprotein also requires an FMN group and has a monomeric molecular mass of ∼24 kDa. NfsB can use either NADPH or NADH as an electron donor [72]. The crystal structure of NfsA protein was later solved by Kobori et al. [73] in 2001.

In 1998, Whiteway et al. [43] analyzed nfsA and nfsB genes in numerous nitrofuran-resistant E. coli mutants, correlating mutations in these genes with nitrofurantoin resistance. Additionally, they reported that overexpression of nfsA and nfsB genes in resistant mutants via plasmids restored original sensitivity to nitrofurans. Many mutants harbored insertion sequence (IS) elements in nfsA and nfsB genes, with IS1 integrating into both genes, IS30 and IS186 exclusively in nfsA, and IS2 and IS5 specifically in nfsB. Insertion hotspots for IS30 and IS186 were noted in nfsA, while IS5 displayed a hotspot in nfsB. These findings suggested regional and sequence-specific hotspots for IS element integration, hinting at underlying adaptive mechanisms driving nitrofurantoin resistance emergence. Later, these findings were supported by García et al. [44], who reported that genetic alterations in the nfsA (IS1 insertion in the coding region) and nfsB (duplication of two codons) genes led to nitrofurantoin resistance in Salmonella enterica strains. Shanmugam et al. [45] also reported insertions and substitution mutations in the nfsA gene of nitrofurantoin-resistant uropathogens. A 2023 study in the UK [46] precisely identified ISs interrupting nfsA and nfsB to be from the IS1 and IS4 families. Specifically, two strains had interruptions in nfsA by IS1R (768 bp, IS1 family) encoded on the reverse complementary strand. IS1R is inserted between bases 693 and 694 of nfsA, marked by 9-bp direct repeats. In another strain, nfsB was interrupted by an IS10R-like element (1,329 bp, IS4 family) with 12 nucleotide differences from IS10R, including substitutions in the right inverted repeat and missense mutations in the transposase gene, leading to four amino acid substitutions. This element is inserted between bases 327 and 328 of nfsB, indicated by flanking 9-bp direct repeats, in an opposite orientation to nfsB [47]. In one of the nitrofurantoin-resistant strains, a complete Tn10 variant (GenBank accession: AF162223.1), flanked by an IS10L-like element and IS10R, was found inserted. Although this transposon did not directly integrate into nfsB, it is possible that the replicative transposition of its IS10L or IS10R component led to the interruption of nfsB [48].

Roemhild et al. [49] described the relative importance of nfsA and nfsB gene expressions. Their findings suggest that the expression of nfsA is critical for initial survival, while the expression of nfsB has a large effect on the bactericidal activity of nitrofurantoin. A recent investigation uncovered that the two genes nfsA and nfsB are located at a considerable distance from each other (287 kb apart). This separation makes it highly improbable for both genes to be simultaneously inactivated through a single natural genetic event. Consequently, the inactivation of these genes is more likely to occur in a stepwise fashion, beginning with nfsA followed by nfsB [46]. Until recently, single-step mutations selected in vitro were found to have exclusively targeted the nfsA gene, and nitrofuran resistance was solely traced to the type I nitroreductase genes. Interestingly, resistant clinical isolates and in vitro selected mutants have exhibited heightened susceptibility to nitrofurantoin under anaerobic conditions. This phenomenon was linked to the existence of active oxygen-sensitive reduction systems prevalent in anaerobic environments [39, 43, 66].

The enzymes responsible for oxygen-sensitive nitrofuran reduction (by type II nitroreductases) remained unknown until 2019, when a group from New Zealand discovered a new nitrofuran-activating enzyme by selecting furazolidone-resistant mutants in a nfsA nfsB double knockout (ΔnfsA ΔnfsB) E. coli strain [23, 56]. All 15 of their independently isolated mutants resistant to furazolidone had mutations in the ahpF gene, which encodes part of the antioxidant alkyl hydroperoxide reductase. Enzymatic assays of purified AhpF later confirmed that this enzyme is a type II oxygen-sensitive nitroreductase. Deleting ahpF in the ΔnfsA ΔnfsB strain reduced resistance to nitrofurazone and nitrofurantoin under aerobic conditions. The mechanism of how AhpF works is quite fascinating as AhpF has a dual role: it independently activates nitrofurans while also working with AhpC to mitigate oxidative stress caused by 5-nitrofurans. The balance between these functions influences the impact of ahpF deletion on nitrofuran susceptibility. Depending on the level of oxidative stress, the rate of AhpF-mediated nitrofuran reduction, and the toxicity of the reduction products, varying effects on resistance to furazolidone, nitrofurantoin, and nitrofurazone were observed [56].

In 2014, Vervoort et al. [53] identified a 12-nucleotide deletion in the ribE gene of nitrofurantoin-resistant E. coli. The ribE gene codes for lumazine synthase, an essential enzyme involved in the FMN biosynthesis. FMN is a cofactor for NfsA and NfsB enzymes. Complementing the mutants with a functional wild-type lumazine synthase was shown to have restored nitrofurantoin susceptibility. Interestingly, nitrofurantoin-resistant E. coli isolated from stool, intestine, and urinary tract of UTI patients treated with nitrofurantoin were found to harbor mutations in nfsA and/or nfsB genes but not in the ribE gene. In Klebsiella pneumoniae, the ribC (riboflavin synthase) and ribB (3,4-dihydroxy 2-butanone 4-phosphate synthase) genes are involved in riboflavin biosynthesis, a universal precursor of FMN and flavin adenine dinucleotide [54, 55]. A resistant mutant of K. pneumoniae was found to have acquired an amino acid substitution in RibC at codon 75, and another resistant mutant had two IS1 insertions 37 bp upstream of ribC. This insertion likely disrupts the ribC promoter, as the gene was also found to be significantly under-expressed in the study mutants [54].

Efflux Pumps

In 2003, a plasmid-borne OqxAB efflux pump was identified in an MDR E. coli isolated from “swine manure” in Denmark, where olaquindox was used as a feed supplement [57]. The MDR E. coli was found to express robust nitrofurantoin resistance (MIC = 128 µg/mL). OqxAB belongs to the resistance nodulation division family of efflux pumps with broad substrate specificity against olaquindox, ciprofloxacin, chloramphenicol, trimethoprim, nalidixic acid, and other disinfectants [58]. In E. coli and S. enterica, the oqxA and oqxB genes were plasmid-borne and flanked by IS26 elements, while the genes were chromosomally encoded in Klebsiella spp. without the flanking IS elements [59, 60]. A high prevalence of oqxAB has been reported among Salmonella and E. coli isolates from animal products in China [61, 62]. In 2015, Ho et al. [63] investigated the molecular epidemiology of the OqxAB efflux pump and its contribution to nitrofurantoin resistance among E. coli isolates from UTI patients (n = 205, 2003–13) and farm animals (n = 136, 2012–13) in Hong Kong. These results suggested that plasmid-mediated oqxAB is generally associated with Tn6010 and is an important nitrofurantoin resistance mechanism. The presence of oqxAB together with nfsA mutations is sufficient for high-level nitrofurantoin resistance. Curing of plasmids carrying oqxAB genes from nitrofurantoin-resistant UTI isolates markedly reduced the MIC of nitrofurantoin [63]. Further, it was demonstrated that both AcrAB and OqxAB efflux pumps contribute significantly toward nitrofurantoin resistance in K. pneumoniae [64].

A study found that overexpression of ramA in a nitrofurantoin-resistant K. pneumoniae strain was caused by a mutation in the ramR repressor gene [64]. Restoring ramR or deleting ramA increased susceptibility to nitrofurantoin, highlighting the key role of ramA in mediating nitrofurantoin resistance. The addition of efflux pump inhibitors was shown to increase the susceptibility of resistant K. pneumoniae to nitrofurantoin, implying the important role of efflux pumps in nitrofurantoin resistance. Recent findings identified alterations in the oqxR gene in two nitrofurantoin-resistant mutants of K. pneumoniae [54, 65]. One mutant exhibited an ISKpn26 insertion, while the other had a non-synonymous mutation (N38K). OqxR typically represses the oqxAB efflux pump; hence, disruptions in oqxR are anticipated to de-repress oqxAB, increasing efflux. High-level nitrofurantoin-resistant mutants of K. pneumoniae displayed elevated oqxB expression, with those bearing oqxR mutations showing the most significant overexpression, indicating relief from oqxR-mediated repression. Moreover, deleting acrB or oqxB genes led to a fourfold reduction in the MIC of nitrofurantoin, while knockout of both acrB and oqxB resulted in a sixteenfold decrease [54].

Several efflux pumps play a critical role in expelling antibiotics from bacterial cells, contributing to multidrug resistance. These include the resistance nodulation division superfamily (e.g., YdhE), MATE superfamily (e.g., NorE/MdtM), MFS (e.g., EmrAB-TolC, MdfA, MdtM), and SMR family (e.g., EmrE, KpnEF). While their direct link to nitrofurantoin resistance has not been specifically studied, they are likely crucial in drug efflux [74].

Studies by Zhang et al. [50], Mottaghizadeh et al. [51], and Khamari et al. [25] strengthened the understanding that mutations in nfsA and nfsB are the major nitrofurantoin resistance mechanisms, while mutations in the ribE gene were rarely observed. Overexpression of the oqxAB genes might help in increased nitrofurantoin resistance, but their absence/downregulation alone does not always imply susceptibility to nitrofurantoin. Diverse mutations in nfsA, nfsB, and ribE genes, complemented by oqxAB efflux pump genes, are responsible for high-level nitrofurantoin resistance among clinical Enterobacterales as shown in Table 1 and Figure 3 [52].

Adaptive and Compensatory Mutations

Studies indicate that even low concentrations of nitrofurantoin (sub-MIC) can be selected for resistant mutants and these are mostly linked to adaptive mutations. For example, resistance was observed at concentrations as low as 1/2000th of the MIC, highlighting the potential for rapid evolution under weak selective pressures [75]. Concurrently, mutations not directly related to antibiotic resistance can be enriched. These mutations may enhance growth or survival in specific environments, contributing to a broader adaptive strategy. A 2023 study reported that there were mutations in unrelated genes like rph, relA, ptsP, kdpD, and fimE among laboratory-selected nitrofurantoin-resistant strains under sub-MIC conditions [75]. However, adaptation to nitrofurantoin mostly proceeds through diverse combinations of mutations that may not repeat in mutants. In a comparative study, pairs of nitrofurantoin-adapted strains showed only 16.5% overlap in their sets of mutated genes, while the same figure for chloramphenicol was 38% [76].

Secondary adaptive mutations often arise alongside resistance mutations and are generally compensatory. These compensatory changes can mitigate fitness costs associated with the acquisition of resistance, allowing the resistant strains to thrive even when not exposed to antibiotics. For example, certain mutations can enhance growth rates or alter metabolic pathways beneficial for survival under antibiotic stress. A study reported mutations identified in selected resistant strains of E. coli among genes that are involved in defense against nutritional stress, modification of respiration and/or membrane potential, and transcriptional rewiring [76]. These processes may not be directly linked to nitrofurantoin resistance mechanisms but may help in the survival of fitness-compromised resistant mutants solely by compensation.

Alternate Mechanisms of Nitrofurantoin Resistance

Initial studies on nitrofurantoin resistance mechanisms that are independent of nitroreductase activity were conducted by Breeze and Obaseiki-Ebor [66]. They started with the knowledge that strains of E. coli that are defective in DNA repair exhibit high sensitivity to nitrofuran antibiotics [33, 77]. Here, nitrofurantoin-resistant mutants were examined for UV resistance to explore the possibility of “enhanced DNA damage repair” as a potential contributor to nitrofurantoin resistance. However, only one mutant of the recB gene resistant to both UV and nitrofurantoin, yet retaining full reductase activity, was observed. The findings from the study revealed that “enhanced DNA repair” could be associated with resistance to nitrofurantoin but not nitrofurazone. The potential for alternative nitrofurantoin resistance mechanisms was suggested when a significant reduction in the MIC of nitrofurantoin was observed following EDTA permeabilization of nitrofurantoin-resistant E. coli, despite minimal or absent nitroreductase activity in the study strains [66]. Subsequently, Rafii and Hansen [78] showed that the development of nitrofurantoin resistance in five mutant Clostridium strains was successful but not dependent on the loss of nitroreductase activity. Further studies by Liochev et al. [67] revealed that the nfsA gene is regulated as a member of the SoxRS regulon, and thus, it is inducible by the redox-cycling agent “paraquat” in the parental, but not in the soxR knockout strain of E. coli. Hence, NfsA may contribute to protection against oxidative stress by minimizing the redox-cycling attendant upon the univalent reduction of quinones, dyes, and nitro compounds. Paterson et al. [68] examined the regulation of nfsA gene expression by paraquat and identified the transcription start site and the SoxS binding site of nfsA gene. They demonstrated that SoxS regulates both nfsA and ybjC (the gene located directly upstream of nfsA) genes in a coordinated manner. The ybjC-nfsA promoter is identified as a class I SoxS promoter. These results suggest the potential involvement of redox-cycling pathway genes in nitrofurantoin resistance. However, additional investigations are necessary to support these hypotheses.

In another study, whole genome sequencing of in vitro selected nitrofurantoin-resistant strains identified loss-of-function mutations in mprA (a multidrug resistance repressor), ompR (an outer membrane porin regulator), rpoA-rpoB (RNA polymerase subunits) genes, in addition to the mutations in nfsA and nfsB genes [69]. A 2020 report from a hospital in North Wales, UK, identified nitrofurantoin-resistant E. coli clinical isolates carrying a mutated CTX-M-14 β-lactamase enzyme, with three non-synonymous mutations (T55A, A273P, and R277C) [70]. When overexpressed in a laboratory E. coli strain, the mutated enzyme conferred both hyper-resistance to nitrofurantoin and β-lactam resistance. In vitro assays showed the enzyme could hydrolyze nitrofurantoin, likely targeting amide bonds in the hydantoin ring such that the hydrolytic products have lower affinity to the activation enzymes or to the essential targets of nitrofurantoin. Although rare and preliminary, this discovery raises serious concerns about the spread of these mutated CTX-M-14 β-lactamases conferring resistance to both the antibiotics at once.

Resistance Variability of Nitrofurantoin

A study by Chevereau et al. [69] revealed that extensive genetic perturbations do not lead to increased resistance variability in the context of nitrofurantoin as compared to other antibiotics. The evolution of nitrofurantoin resistance was found to be consistently occurring in two phases. An early phase with rapid increase in resistance (∼twenty-fold) involved large-effect mutations followed by a second phase where evolution occurs at an extremely slow pace (∼two-fold increase even after 21 days) due to sluggish adaptations. Whole genome sequencing revealed that the evolution of nitrofurantoin resistance involves highly reproducible mutational paths characterized by loss-of-function mutations in nfsA, nfsB, mprA, and ompR genes in the early phase. Mutations in rpoA, rpoB, and other genes acquired in the nitrofurantoin-resistant clones during the second phase were found to be irreproducible. Interestingly, these irreproducible mutations yielded only a marginal increase in the resistance.

The low mutational diversity associated with nitrofurantoin resistance is noteworthy because the reduction of nitrofurantoin damages the DNA and triggers the SOS response. This may result in increased mutation frequency leading to accelerated adaptation and diversification of mutational paths. However, observation of low levels of nitrofurantoin resistance is plausible, probably due to the deleterious nature of the accumulated mutations. Globally, nitrofurantoin resistance remains astoundingly low despite its clinical use for over 70 years.

Heteroresistance

In a study involving two E. coli blood strains, population analysis profiling revealed subpopulations with reduced susceptibility to nitrofurantoin (MICs recorded at 64 mg/L) [79]. The presence of resistant subpopulations was confirmed through genomic sequencing, which showed significant deletions linked to large-scale deletions in chromosomal regions harboring the nfsB gene. These deletions are adjacent to ISs of the IS1 family, suggesting a mechanism of genetic instability that can lead to varying levels of susceptibility within a single population. The emergence of heteroresistant strains indicates that even low-prevalence resistance can complicate treatment scenarios [80]. Particularly in UTIs, it poses challenges for infection management, as standard susceptibility testing may not detect these resistant subpopulations. This could lead to treatment failures if clinicians rely solely on conventional susceptibility results.

Fitness Cost

Andersson and Levin [81] explored the involvement of biological cost associated with antibiotic resistance. The rates and frequency of emergence and dissemination of antimicrobial resistance (AMR) in bacterial populations were found to be directly linked to the extent of antibiotic use and inversely associated with the fitness of bacteria. The majority of resistance-conferring mutations were found to result in compromised fitness, but subsequent evolution may help in amending this scenario [82]. Sandegren et al. [83] reported that the emergence of nitrofurantoin resistance is associated with fitness costs to the host bacterium. They observed a reduction in bacterial growth rates when the in vitro selected nitrofurantoin-resistant mutants and clinical isolates were cultured at therapeutic levels of nitrofurantoin. The mutations acquired during the process may help the bacterium gain resistance, but may also cause severe compromise in other essential metabolic pathways as shown in Figure 3c and hence prohibit the establishment of an infection.

Analyses of the population structure of nitrofurantoin-resistant E. coli showed high genetic diversity. The emergence of nitrofurantoin resistance in uropathogens is associated with low epidemiological fitness and may be considered an evolutionary dead end [84]. Vervoort et al. [53] compared nitrofurantoin resistance determinants and growth rates of in vitro generated mutants with nitrofurantoin-resistant E. coli isolated from UTI patients previously treated with nitrofurantoin. In the absence of nitrofurantoin, resistant clinical isolates exhibited better fitness compared to in vitro generated mutants. However, no significant variations were observed in the two groups of bacteria when cultured in the presence of nitrofurantoin. Another study suggested that the acquisition of spontaneous mutations in nfsA and nfsB genes indeed affected the fitness of in vitro selected resistant strains [25]. The resistant bacteria were found to survive under therapeutic concentrations of NIT (although at a slower growth rate), but recovered when cultured without the antibiotic. An improvement was observed in the growth of a resistant clinical isolate in the absence of NIT. This suggested that clinically resistant strains may acquire sufficient compensatory mutations and adaptive strategies to thrive in the host environment. While the fitness cost was significant in some studies, others argued that even though there were small changes in the doubling times (2–10%), statistical analysis (ANOVA and pairwise t tests) did not validate these growth changes to be significant [46]. It is crucial to recognize that sub-MIC selective pressure might play a pivotal role in facilitating the establishment of antibiotic-resistant phenotypes within a population. This contrasts with the conventional belief that bacterial resistance is solely determined by the ability of isolates to thrive in antibiotic concentrations surpassing the MIC of sensitive isolates.

Although previous reports suggest that reversal of antibiotic resistance phenotype is possible upon reduction in antibiotic usage, the persistence of resistance-associated mechanisms even without antibiotic selection pressure is being increasingly observed [85]. Adaptive and compensatory mutations that are cost-free and genetically linked (co-selection) to the evolution of resistance phenotype are reported to play a role in the survival and fitness of bacteria [82].

Co-Resistance

Resistance against more than one class of antibiotics by the same bacterial strain is generally referred to as co-resistance. Analysis of a large cohort of female outpatients (n = 286,187) in the USA between 1995 and 2001 revealed that nitrofurantoin susceptibility (98.3 to 99.1%) remained mostly unchanged. Importantly, co-resistance to nitrofurantoin (3.5%) in the study was found to be less frequent compared to other antibiotics [86]. Further, a study involving 377,852 patient samples revealed that co-resistance to nitrofurantoin was 1.9% among trimethoprim/sulfamethoxazole-resistant E. coli and the overall resistance rate to nitrofurantoin was 0.1 to 0.2% [87]. Assessment of AMR patterns of 1,076 urine samples from female outpatients aged >14 years from Brazil revealed that nitrofurantoin co-resistance (3.4%) among ciprofloxacin-resistant E. coli was lower compared to other antibiotics [88]. In contrast, a study from Assam, India, involving 200 UTI patients reported that 26 of 42 (62%) ESBL-producing Enterobacteriaceae were co-resistant to nitrofurantoin [89]. Nitrofurantoin co-resistance among antibiotic-resistant uropathogens is relatively rare, especially outside the Indian subcontinent. Therefore, nitrofurantoin remains one of the preferred antibiotics in the treatment of MDR bacteria associated with UTIs. However, newer studies showing the presence of a mutated CTX-M-14 gene conferring resistance to both β-lactam antibiotics and nitrofurantoin may soon change this scenario [70].

A study in India involving 100 clinical Enterobacteriaceae isolates strongly supported nitrofurantoin resistance to be linked with other antibiotic resistance. Nitrofurantoin resistance was identified to be a potential indicator of the extensively drug-resistant (XDR) phenotype among Enterobacteriaceae, harboring multiple AMR and efflux pump genes. Tigecycline and colistin were, on the other hand, found to be independent of nitrofurantoin resistance [90]. Following this study, a vast study was carried out in England with over one million bacterial isolates reported between 2015 and 2019 [91]. Based on their study, MDR and XDR phenotypes were found more frequently in nitrofurantoin-resistant E. coli urinary isolates compared to nitrofurantoin-susceptible ones. Among UTI-causing E. coli, MDR occurrence was consistently 15–20% higher in nitrofurantoin-resistant isolates. Similarly, the proportion of isolates with an XDR phenotype was elevated among nitrofurantoin-resistant ones (8.7% vs. 1.4% in 2019), especially in male patients, though variations occurred by age group in both genders [91].

Collateral Sensitivity

Szybalski and Bryson [92] coined the term “collateral sensitivity” and defined it as “increasing the sensitivity to one antibiotic by increasing resistance to another.” Hancock [93] stated that the principle of “collateral sensitivity” may serve in identifying potential tools that aid in limiting the emergence of resistance until new antibiotic therapies and strategies are developed. Subsequently, Pál et al. [94] suggested that potential mechanisms of collateral sensitivity may include mutations that cause multidrug resistance in bacteria while simultaneously enhancing the sensitivity to other unrelated drugs. The study highlighted that single-drug treatment may promote the evolution of multidrug resistance by pleiotropic effects, while combination therapy could potentially prevent the spread of resistant pathogens.

Imamovic and Somer [95] proposed a new treatment framework called “collateral sensitivity cycling” using complex collateral sensitivity networks of in vitro generated E. coli strains resistant to 23 different antibiotics. They observed that E. coli that acquired aminoglycoside resistance exhibited increased sensitivity to nitrofurantoin. Interestingly in most cases, the differences in collateral sensitivity and co-resistance patterns involved nitrofurantoin-resistant strains or changes in the susceptibility to nitrofurantoin. A recent study on E. coli isolates (n = 888,207) revealed that prescription of nitrofurantoin was associated with lower levels of resistance to amoxicillin, trimethoprim, and ciprofloxacin. Further, they observed that the prescription of trimethoprim was associated with higher levels of nitrofurantoin and ciprofloxacin resistance [96].

Roemhild et al. [49] postulated multiple molecular mechanisms associated with collateral sensitivity to nitrofurantoin. Spontaneous mutants resistant to mecillinam (spoT mutation), tigecycline (lon mutation), and protamine (hemL mutation) were found to exhibit collateral sensitivity to nitrofurantoin (3× to 35× decrease in MIC). This hyper-susceptibility toward nitrofurantoin was attributed to overexpression of nitroreductase enzymes NfsA and NfsB, increased drug toxicity (result of lon mutation), and increased drug uptake (result of hemL mutation) in the resistant mutants. The increased nitrofurantoin toxicity was further explained by the interference of nitrofurantoin’s innate drug-response system (the SOS response) with the growth of the bacterium. Hence, nitrofurantoin was suggested to be an efficient last-resort antibiotic in a drug-switching regime [49].

History and Current Prevalence of Nitrofurantoin Resistance

Nitrofurantoin has faced varying levels of resistance in different parts of the globe over the years as shown in Table 2. Early studies in South Africa, the USA, and Europe from the late 1980s to the late 1990s indicated low resistance rates of well under 10% among uropathogens to nitrofurantoin [97–99]. However, resistance began to emerge, with reports showing resistance rates of around 10% in the late 1990s and early 2000s [100]. In the early 2000s, studies in the USA highlighted the effectiveness of nitrofurantoin against uropathogens, with resistance rates ranging from 0.4% to 10.4% among different isolates. By the mid-2010s, resistance levels remained relatively low in Europe, with rates around 1% in E. coli and 4% in ESBL-producing E. coli [11, 100, 101]. However, resistance varied globally, with higher rates reported in some regions. For instance, studies in Brazil, China, and Kenya showed resistance rates ranging from 15.7% to 23.3% among different bacterial strains [100].

Table 2.

Global prevalence of nitrofurantoin resistance: a historical account

Country/region	Study period	Sample size	Nitrofurantoin resistance rate	References
London, UK	1985 to 1992	527 E. coli and 69 K. pneumoniae	4% and 1%	[98]
South Africa	1986 to 1991	264 Enterobacteriaceae	9.8%	[97]
Washington, USA	1992 to 1996	3736 E. coli	0.2–2%	[99]
Ontaria, Canada	1997 to 2000	2000 Gram-negative UTI isolates	4.6%	[102]
USA	2001	29,198 MDR E. coli	2.2%	[100]
USA	2003 to 2007	10,417 E. coli	2.3%	[101]
Hong Kong	2006 to 2008	271 E. coli	6.6%	[103]
UAE	2008 to 2010	292 ESBL Enterobacteriaceae	10%	[11]
USA	2010	32,742 MDR E. coli	2.1%	[100]
Australia	2009 to 2013	5333 E. coli and Klebsiella sp.	2.7%	[104]
Europe	2011 to 2019	973 E. coli and 313 Klebsiella sp.	4.8% and 46%	[105]
England	2015	24 ,821 E. coli	2.9%	[91]
Europe including Russia	2015–2017	775 E. coli	1.2%	[106]
England	2019	22 ,393 E. coli	2.3%	[91]
China	2015 to 2021	122,384 E. coli	2–4.4%	[107]
		30,154 Klebsiella spp.	27.2–60.9%
India	2022	214 Enterobacteriaceae	14%	[108]

Recent data from diverse geographic regions, including India, the Middle East, and Europe, have shown varying levels of nitrofurantoin resistance. Studies in India reported resistance rates of around 12.6% [108], while reports from the Middle East indicated resistance rates ranging from 10% to 27% [109]. In Europe, resistance levels were generally low, with resistance rates of 37% among mecillinam-nitrofurantoin-resistant isolates. However, resistance rates in Poland and South Africa were notably higher, exceeding 30% and 86.1%, respectively [100].

The emergence of XDR bacteria has posed challenges, with studies linking nitrofurantoin resistance to XDR phenotypes [90]. Co-resistance between nitrofurantoin and other antibiotics has been observed, emphasizing the need for continuous surveillance and effective antimicrobial stewardship practices. The evolving landscape of nitrofurantoin resistance underscores the importance of ongoing research, surveillance, and global collaboration to combat AMR effectively and preserve the efficacy of this valuable antibiotic.

Prediction of Nitrofurantoin Resistance

Efforts to develop tools that detect or predict nitrofurantoin resistance often focus on genetic markers, aiming to improve treatment outcomes by targeting the right antibiotics. One study utilized PCR amplification with statistical analysis of 100 clinically isolated Enterobacteriaceae, identifying the oqxB-nfsB-oqxA-ribE combination as a strong predictor of nitrofurantoin resistance. The proposed ordinal logistic regression model achieved high accuracy (>91%) in segregating resistant and susceptible variants, though it was less accurate (47.6%) for intermediate-resistant strains [25]. Another approach involved a decision-tree algorithm developed from five loci (nfsA, nfsB, ribE, oqxA, and oqxB) associated with nitrofurantoin resistance. Using data from 12,412 E. coli genomes, this algorithm successfully predicted nitrofurantoin susceptibility based on genotypes [47]. These advancements in predictive tools are promising steps toward developing rapid detection tools that give faster results and cut down the usual turnover time of 48 h at diagnostic laboratories. Early detection can lead to better management of antibiotic resistance, allowing for more precise and effective treatment strategies.

Conclusion

Nitrofurantoin resistance mechanisms and the importance of preserving this antibiotic in the fight against AMR are crucial topics in the current medical landscape. The objective of this review was to consolidate available information on the mechanisms of action of nitrofurantoin and the resistance mechanisms employed by bacteria. With the scarcity of new antibiotics and the increasing reports of drug-resistant superbugs, it is crucial to repurpose existing antibiotics and use them judiciously. The scientific evidence presented in this review aims to facilitate further research into the mechanisms of nitrofurantoin action and resistance, as well as the heteroresistance, co-resistance, and collateral sensitivity associated with its use. This knowledge will help clinicians verify their antibiotic prescriptions and create evidence-based patient-specific prescriptions. Nitrofurantoin is an important antibiotic that has stood the test of time, and its preservation is essential in the fight against AMR. Understanding bacterial metabolism of nitrofurantoin, studying resistance patterns, and using it with caution are key strategies to achieve this goal.

Based on the current literature, clinicians are encouraged to consider nitrofurantoin as a viable treatment option for uncomplicated lower UTIs, particularly in regions where resistance rates are relatively low (under 3%), such as the USA, Australia, England, and Russia. In contrast, in areas where resistance is more prevalent, it is advisable to limit the empirical use of nitrofurantoin. Instead, practitioners should adopt an evidence-based approach by prescribing nitrofurantoin only after identifying the causative organism and reviewing antimicrobial susceptibility test reports.

The detection of drug-resistant superbugs through molecular diagnosis has revolutionized medicine and healthcare. The urgent need for rapid diagnostic tools to detect nitrofurantoin resistance cannot be overstated. Significant efforts must be directed toward this endeavor to ensure effective treatment strategies and to combat the rising tide of antibiotic resistance. This proactive approach will not only enhance patient care but also contribute to the broader fight against multidrug-resistant infections, ensuring that nitrofurantoin remains a key player in our therapeutic arsenal against UTIs.

Statement of Ethics

Statement of ethics is not applicable because the present study includes neither human nor animal experiments.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

The present study was supported by ICMR EMR Adhoc Grant No. OMI/27/2020-ECD-I and AMR/Adhoc/281/2022-ECD-II.

Author Contributions

Balaram Khamari and Eswarappa Pradeep Bulagonda conceived and designed the study and prepared the final draft. Eswarappa Pradeep Bulagonda supervised the collection of data and relevant articles, while Balaram Khamari prepared the manuscript and figures. Eswarappa Pradeep Bulagonda reviewed and edited the manuscript.

Funding Statement

The present study was supported by ICMR EMR Adhoc Grant No. OMI/27/2020-ECD-I and AMR/Adhoc/281/2022-ECD-II.

Data Availability Statement

All data generated by the present study are included in this article. Further inquiries can be directed to the corresponding author.