Differences in Fosfomycin Resistance Mechanisms between Pseudomonas aeruginosa and Enterobacterales

ABSTRACT

Multidrug-resistant (MDR) Pseudomonas aeruginosa presents a serious threat to public health due to its widespread resistance to numerous antibiotics. P. aeruginosa commonly causes nosocomial infections including urinary tract infections (UTI) which have become increasingly difficult to treat. The lack of effective therapeutic agents has renewed interest in fosfomycin, an old drug discovered in the 1960s and approved prior to the rigorous standards now required for drug approval. Fosfomycin has a unique structure and mechanism of action, making it a favorable therapeutic alternative for MDR pathogens that are resistant to other classes of antibiotics. The absence of susceptibility breakpoints for fosfomycin against P. aeruginosa limits its clinical use and interpretation due to extrapolation of breakpoints established for Escherichia coli or Enterobacterales without supporting evidence. Furthermore, fosfomycin use and efficacy for treatment of P. aeruginosa are also limited by both inherent and acquired resistance mechanisms. This narrative review provides an update on currently identified mechanisms of resistance to fosfomycin, with a focus on those mediated by P. aeruginosa such as peptidoglycan recycling enzymes, chromosomal Fos enzymes, and transporter mutation. Additional fosfomycin resistance mechanisms exhibited by Enterobacterales, including mutations in transporters and associated regulators, plasmid-mediated Fos enzymes, kinases, and murA modification, are also summarized and contrasted. These data highlight that different fosfomycin resistance mechanisms may be associated with elevated MIC values in P. aeruginosa compared to Enterobacterales, emphasizing that extrapolation of E. coli breakpoints to P. aeruginosa should be avoided.

INTRODUCTION

Pseudomonas aeruginosa is an opportunistic bacterial pathogen capable of causing nosocomial and community-acquired infections (1). This pathogen poses a considerable risk to public health due to its widespread intrinsic and acquired resistance to a large number of antibiotics (2). The Centers for Disease Control and Prevention (CDC) listed multidrug-resistant (MDR) P. aeruginosa as a serious threat requiring prompt and sustained action in their 2019 report on antibiotic resistance threats in the United States (3). The lack of therapeutic alternatives to treat the emergence of MDR and extensively drug-resistant (XDR) P. aeruginosa strains has renewed interest in older, underutilized drugs including fosfomycin (2, 4). Current options for treatment of MDR P. aeruginosa infections may include polymyxins, aminoglycosides, antipseudomonal carbapenems, cefiderocol, and the newer β-lactam–β-lactamase inhibitor combinations such as ceftolozane-tazobactam, ceftazidime-avibactam, and imipenem-relebactam (2, 5–7). However, toxicities associated with the older agents and reports of treatment-emergent resistance to the novel agents emphasize the need for alternative agents such as fosfomycin (8).

FOSFOMYCIN MECHANISM OF ACTION AND FORMULATIONS

Fosfomycin, an old antibiotic originally discovered in the 1960s, is a broad-spectrum antibiotic that is structurally unlike other antibiotic classes (Fig. 1) (9). Fosfomycin is a phosphonic acid derivative with an epoxide ring structure and a direct carbon-phosphorus bond that eliminates the usual intermediate oxygen bond found in other organophosphorus compounds (9–11). The only member of its antibiotic class (12), fosfomycin has emerged as a promising therapeutic alternative due to its low toxicity and low cross-resistance (13). In Enterobacterales, fosfomycin is taken up by both the glycerol-3-phosphate transporter (GlpT) and the hexose phosphate transporter (UhpT) (14), but in P. aeruginosa the UhpT pump is absent. The drug’s unique mode of action blocks peptidoglycan biosynthesis by binding to and inactivating the UDP-N-acetylglucosamine enolpyruval transferase (MurA) enzyme (Fig. 2) (14, 15). Fosfomycin covalently binds to the MurA active site via a thioether bond to the Cys115 residue (16), acting as a phosphoenolpyruvate (PEP) analogue that prevents the formation of N-acetylmuramic acid, a precursor to peptidoglycan, resulting in bacterial cell lysis and death (14, 17, 18).

FIG 1.

Chemical structure of fosfomycin (C₃H₇O₄P). Fosfomycin is a phosphonic acid derivative with an epoxide ring structure. It has an (R,S)-1,2-epoxypropyl group attached to phosphorus, with a direct carbon-phosphorus bond.

FIG 2.

Overview of fosfomycin resistance mechanisms in P. aeruginosa. Fosfomycin (denoted by the circles containing the letter “F”) mimics the structure of the substrate, glycerol-3-phosphate (G3P), which is transported by the glucose-3-phosphate (GlpT) transporter under normal conditions. When fosfomycin is present, it is transported inside the cell using the GlpT transporter. Once inside the cell, fosfomycin acts as a mimic of phosphoenolpyruvate (PEP) and blocks the action of UDP-N-acetylglucosamine enolpyruval transferase (MurA), the enzyme that catalyzes the first committed step in peptidoglycan biosynthesis, preventing the enzyme from synthesizing UDP-N-acetylmuramic acid (UDP-MurNAc), a peptidoglycan precursor, from UDP-N-acetylglucosamine (UDP-GlcNAc). The three categories of fosfomycin resistance mechanisms in P. aeruginosa are depicted. Peptidoglycan recycling enzymes salvage UDP-MurNAc, bypassing de novo synthesis of peptidoglycan and thus interfering with fosfomycin action. Fos enzymes modify the structure of fosfomycin and inactivate the antibiotic. Mutations to the glpT transporter gene cause reduced fosfomycin permeability. Figure adapted from reference 18. Created with BioRender.com.

In the United States, the oral formulation of fosfomycin (fosfomycin tromethamine) is available as a sachet and is approved as a one-time, 3-g dose for treatment of uncomplicated urinary tract infection (UTI) caused by Escherichia coli or Enterococcus faecalis (19). Single oral dosing results in high urinary concentrations despite low plasma concentrations (maximum concentration of drug in serum [C_max] of ∼25 mg/L) (20). As such, the oral formulation of fosfomycin is currently recommended by the Infectious Diseases Society of America (IDSA) as a first-line agent for acute cystitis/uncomplicated UTI only (21). Contemporary pharmacokinetic data, however, demonstrate considerable variability in urinary drug concentrations following administration of oral fosfomycin (22–24). Due to the increase in MDR and XDR pathogens, fosfomycin has also been suggested as an alternative therapeutic option for other Gram-negative and Gram-positive bacterial infections (15). The intravenous (i.v.) formulation has for years been used to treat systemic infections outside the United States (25), and an i.v. formulation is currently under review by the U.S. Food and Drug Administration (FDA) for treatment of complicated UTI (26).

FOSFOMYCIN SUSCEPTIBILITY TESTING AND BREAKPOINTS

Fosfomycin susceptibility testing presents several challenges. Both the Clinical and Laboratory Standards Institute (CLSI) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST) consider agar dilution (AD) supplemented with glucose-6-phosphate (G6P) the reference method for fosfomycin susceptibility testing; however, this method is both labor- and time-intensive and thus unlikely to be feasible for most clinical microbiology laboratories (5). Several studies have shown that supplementation with G6P may not be necessary for P. aeruginosa isolates due to the absence of the UhpT transporter (27). Furthermore, the use of broth microdilution (BMD) testing is not recommended by CLSI or EUCAST based on historical studies showing unsatisfactory precision and difficulty in reading endpoints due to trailing endpoints and skipped wells (28). Finally, automated/commercial systems employing broth-based methods often exclude fosfomycin. While select panels/cards (bioMérieux Vitek 2 and BD Phoenix) may be available outside the United States, these panels are not approved by the FDA (29). As a result, disk diffusion is generally used for testing and fosfomycin may often be tested only after special request by a clinician.

A lack of susceptibility breakpoints for fosfomycin against P. aeruginosa likely limits the use of this agent against this pathogen. The CLSI has established Gram-negative bacterial breakpoints for oral fosfomycin against E. coli only (Table 1) (30). No breakpoints have been set for i.v. fosfomycin as the sponsor has not yet submitted data for consideration to the CLSI, nor has this formulation been approved by the FDA for use in the United States. Interpretive categories have been established by EUCAST for both oral and i.v. fosfomycin against Enterobacterales (31). EUCAST does not publish breakpoints specific to P. aeruginosa but notes that wild-type isolates (epidemiological cutoff [ECOFF] value, MIC of 128 mg/L) have been treated with fosfomycin as part of combination therapy. The absence of established susceptibility breakpoints for P. aeruginosa has required extrapolation of the CLSI E. coli or EUCAST Enterobacterales susceptibility breakpoints to this organism (5, 32). However, data from several in vitro susceptibility studies have demonstrated this approach should be avoided due to higher P. aeruginosa MIC values in comparison to Enterobacterales isolates (5, 33, 34). These studies found that the majority of P. aeruginosa MIC values clustered around the CLSI E. coli susceptibility breakpoint of 64 mg/L and the intermediate breakpoint of 128 mg/L. Additionally, ECOFF values have ranged from 128 mg/L to 512 mg/L, which are several dilutions higher than the current E. coli susceptibility breakpoints of ≤64 mg/L (CLSI, E. coli for oral fosfomycin) or ≤32 mg/L (EUCAST, Enterobacterales for i.v. fosfomycin) and ≤8 mg/L (EUCAST, E. coli uncomplicated UTI for oral fosfomycin). The clustering of these MIC/ECOFF values near the current E. coli breakpoints makes the interpretation of P. aeruginosa values more difficult.

TABLE 1.

Current fosfomycin breakpoints and recommendations approved by EUCAST and CLSI for various Gram-negative organisms^a

Organization and formulation	Gram-negative organisms with approved breakpoint type:
	MIC (mg/L)	Disk diffusion zone diam (mm)
CLSI, oral formulation	Escherichia coli urinary tract isolates: ≤64, S; 128, I; ≥256, R	Escherichia coli urinary tract isolates: ≥16, S; 13–15, I; ≤12, R
EUCAST
Oral formulation	Escherichia coli uncomplicated UTI: ≤8, S; >8, R	Escherichia coli only: ≥24, S; <24, R
	Pseudomonas aeruginosa: no breakpoints but note that wild-type isolates (ECOFF: MIC, 128) have been treated with combination therapy
Intravenous formulation	Enterobacterales: ≤32, S; >32, R	Escherichia coli only: ≥21, S; <21, R
	Pseudomonas aeruginosa: no breakpoints but note that wild-type isolates (ECOFF: MIC, 128) have been treated with combination therapy

^aAbbreviations: CLSI, Clinical and Laboratory and Standards Institute; ECOFF, epidemiological cutoff value; EUCAST, European Committee on Antimicrobial Susceptibility Testing; I, intermediate; R, resistant; S, susceptible; UTI, urinary tract infection.

DATA SOURCE AND RESISTANCE MECHANISM CATEGORIZATION

The efficacy of fosfomycin is limited by both intrinsic and acquired resistance mechanisms. P. aeruginosa mechanisms of resistance to fosfomycin have not been comprehensively summarized following the identification of new mechanisms in recent years. A majority of previous reviews have focused on fosfomycin resistance mechanisms present in E. coli and other Enterobacterales (17, 18, 35–37) or P. aeruginosa mechanisms of resistance to other classes of antibiotics (1, 2). This narrative review aims to summarize the different resistance mechanisms mediated by P. aeruginosa against fosfomycin, including those considered inherent and acquired. A comparison of resistance mechanisms between P. aeruginosa and Enterobacterales will also be made.

PubMed was utilized to identify relevant studies assessing fosfomycin resistance in P. aeruginosa. Search terms included “Pseudomonas aeruginosa,” “fosfomycin,” and “resistance.” Data sources were narrowed to studies in the English language. Studies indicating fosfomycin resistance mechanisms in other species, namely, E. coli, were also considered. Resistance mechanisms were divided into broad categories. Those present in P. aeruginosa—peptidoglycan recycling enzymes and upstream processing enzymes, Fos enzymes, and transporters—were summarized first. Subsequently, resistance mechanisms absent in P. aeruginosa were reviewed, namely, those found in E. coli—transporters and associated regulatory genes, plasmid-mediated Fos enzymes, fosfomycin resistance kinases, and MurA modification—for comparison between species.

P. AERUGINOSA INHERENT FOSFOMYCIN RESISTANCE MECHANISMS

Various inherent resistance mechanisms for fosfomycin have been identified among P. aeruginosa (38–40).

(i) Peptidoglycan recycling.

Several studies have identified the presence of recycling pathways that bypass de novo peptidoglycan biosynthesis in P. aeruginosa (13, 41, 42) (Table 2). In particular, an anabolic recycling pathway that salvages the peptidoglycan precursor UDP N-acetylmuramic acid (UDP-MurNAc) through the action of a MurNAc 6-phosphate phosphatase (MupP) (43, 44), an anomeric MurNAc kinase (AmgK) (13, 41), and a uridylyl transferase (MurU) (45), has been confirmed. Together, these enzymes bypass the de novo biosynthesis of UDP-MurNAc, the process targeted by fosfomycin, and thus interfere with fosfomycin action. Blocking this salvage pathway reduced intrinsic resistance to fosfomycin, resulting in a 4- to 8-fold-increased susceptibility (41).

TABLE 2.

Fosfomycin resistance mechanisms documented within Pseudomonas aeruginosa

Category	Mechanism of resistance	Gene	Resistance type	Citation(s)
Transporter	Inactivation of glycerol-3-phosphate transporter. Mutation causes reduced permeability.	glpT	Acquired	Castañeda-García et al., 2009 (14)
Peptidoglycan recycling enzyme	MurNAc 6-phosphate phosphatase. Helps bypass fosfomycin-sensitive peptidoglycan de novo synthesis.	mupP	Inherent	Borisova et al., 2017 (43); Fumeaux and Bernhardt, 2017 (44)
	Anomeric MurNAc/GlcNAc kinase, forms MurNAc a1P. Helps bypass fosfomycin-sensitive peptidoglycan de novo synthesis.	amgK	Inherent	Gisin et al., 2013 (13); Borisova et al., 2014 (41)
	Uridylyl transferase, transfers uridine phosphate from UTP to the MurNAc a1P, yielding UDP-MurNAc. Helps bypass fosfomycin-sensitive peptidoglycan de novo synthesis.	murU	Inherent	Gisin et al., 2013 (13); Borisova et al., 2014 (41); Renner-Schneck et al., 2015 (45)
Upstream enzyme for muropeptide processing	Recycling N-acetylglucosaminidase that yields N-acetylglucosamine (GlcNAc) and 1,6-anhydroMurNAc (AnhMurNAc) peptides.	nagZ	Inherent	Gisin et al., 2013 (13); Borisova et al., 2014 (41); Park and Uehara, 2008 (46)
	AnhMurNAc-peptide amidase	ampD	Inherent	Gisin et al., 2013 (13); Borisova et al., 2014 (41); Hamou-Segarra et al., 2017 (47)
	AnhMurNAc kinase	anmK	Inherent	Gisin et al., 2013 (13); Borisova et al., 2014 (41); Park and Uehara, 2008 (46)
Fos metalloenzyme	Mn2+- and K+-dependent dimeric glutathione S-transferase. Opens epoxide ring of fosfomycin. Overexpression of gene.	fosA	Inherent	Ito et al., 2017 (15); Ito et al., 2017 (48); Beharry and Palzkill, 2005 (49); Rife et al., 2002 (50); De Groote et al., 2011 (52)
	Bacillithiol transferase	fosE	Acquired	Kieffer et al., 2020 (54)
	Glutathione S-transferase	fosF		Yatsuyanagi et al., 2004 (53); Kieffer et al., 2020 (54)
	Bacillithiol transferase	fosH		Kieffer et al., 2020 (54)

Numerous enzymes have been identified to act in this recycling pathway. The enzyme MupP, which converts MurNAc 6-phosphate to MurNAc, was recently confirmed in P. aeruginosa (44) with a parallel study identifying a mupP ortholog in Pseudomonas putida (43). mupP-defective strains were found to be fosfomycin hypersensitive in P. aeruginosa and P. putida. Deletion of mupP or another peptidoglycan recycling enzyme also increased ampC expression and thus β-lactam resistance (44).

This anabolic recycling pathway also requires the anomeric cell wall amino sugar kinase (AmgK) that forms MurNAc α-1-phosphate (MurNAc α1P) and the uridylyl transferase (MurU) that subsequently convert MurNAc to UDP-MurNAc (13, 41). Deletion of murU and amgK increased fosfomycin susceptibility in P. putida (13). amgK and murU orthologs from P. aeruginosa, with 66% and 71% amino acid sequence identity, respectively, were able to restore the wild-type fosfomycin susceptibility phenotypes in the respective P. putida mutants. Increased fosfomycin resistance was not observed when mupP was expressed alone in P. aeruginosa (44).

Other upstream enzymes for muropeptide processing are also present in Pseudomonas spp., including a recycling N-acetylglucosaminidase (NagZ) that yields N-acetylglucosamine (GlcNAc) and 1,6-anhydroMurNAc (AnhMurNAc) peptides, an AnhMurNAc-peptide amidase (AmpD), and an AnhMurNAc kinase (AnmK) (13, 41, 46, 47). Growth rates of wild-type P. aeruginosa and the respective amgK, murU, and anmK mutants on Luria-Bertani (LB) medium were identical. These findings suggest that blocking peptidoglycan recycling has no effect on the fitness of P. aeruginosa (13, 41).

(ii) Chromosomal Fos enzymes.

Some enzymes are able to modify fosfomycin structure and thus inactivate the antibiotic. Fos enzymes belong to the metalloenzyme superfamily and confer resistance to fosfomycin by opening the epoxide ring, rendering the antibiotic ineffective (15, 48–50). One type of fosfomycin resistance enzyme is FosA, a Mn²⁺- and K⁺-dependent dimeric glutathione S-transferase that inactivates the antibiotic by the addition of glutathione to the epoxide ring (15, 51). Key residues in the enzyme active site involved with Mn²⁺, K⁺, and fosfomycin binding are highly conserved among FosA proteins (15, 49, 50). Many Gram-negative species carry chromosomal fosA genes, including P. aeruginosa, in which 98.8% of 2,257 published genomes surveyed contained a fosA homologue (15). In contrast, chromosomal fosA genes were nearly absent in E. coli, found in 4.6% of 5,363 genomes.

Chromosomal fosA genes conferred inherent high-level fosfomycin resistance with MIC values of 1,024 μg/mL in most Gram-negative species, including Klebsiella pneumoniae, Klebsiella oxytoca, Klebsiella aerogenes, Enterobacter cloacae, Serratia marcescens, Morganella morganii, and Providencia stuartii, when transformed with recombinant plasmids carrying fosA from the respective species in E. coli (15). In contrast, fosfomycin exhibited a modest baseline activity against E. coli when transformed with recombinant plasmids carrying fosA from the respective species in E. coli, with an MIC of 16 μg/mL, which is considered susceptible despite a 16-fold increase from the baseline MIC (15). Overexpression of fosA was also reported to confer fosfomycin resistance (52). Other types of Fos enzymes have also been confirmed in P. aeruginosa, including FosF, a glutathione S-transferase (53), and recently reported bacillithiol transferases FosE and FosH (54).

P. AERUGINOSA ACQUIRED FOSFOMYCIN RESISTANCE MECHANISMS—TRANSPORTERS

In addition to the numerous inherent fosfomycin resistance mechanisms present in P. aeruginosa, this pathogen can also develop resistance to fosfomycin through chromosomal mutations in the glpT gene. GlpT is reported to be the only fosfomycin transporter present in P. aeruginosa (14). Thus, fosfomycin can enter P. aeruginosa cells only through GlpT, and mutation of the gene causes reduced permeability. glpT appears to be the only target gene in P. aeruginosa that, when inactivated, induces fosfomycin resistance (14, 18). All 10 independent fosfomycin-resistant mutants generated by spontaneous mutation contained mutations in the glpT gene, with even subtle changes in the glpT amino acid sequence resulting in increased fosfomycin MICs of 512 to 1,024 μg/mL compared to the wild-type MIC of 8 μg/mL (14). Fosfomycin susceptibility was subsequently recovered in the mutants with the addition of the wild-type glpT gene. Mutation of glpT has no obvious fitness cost in vitro or in vivo with the glpT null mutant having equal virulence as the glpT⁺ parent (14, 17, 55).

FOSFOMYCIN RESISTANCE MECHANISMS ABSENT IN P. AERUGINOSA

There are numerous fosfomycin resistance mechanisms absent in P. aeruginosa, which are mostly exhibited by bacteria in the Enterobacterales order (56).

(i) Transporters and associated regulators.

In E. coli and other Enterobacterales, fosfomycin is transported via both GlpT and UhpT (14) (Table 3). The expression of these transport systems depends on their respective substrates, glycerol-3-phosphate and hexose-6-phosphate, as well as the cyclic AMP receptor protein (cAMP-CRP) complex (57, 58). Expression of glpT and uhpT is positively regulated by cAMP. Additionally, the regulatory genes uhpA, uhpB, and uhpC are needed for high-level uhpT expression (57, 58), and inactivation of any of these regulatory genes results in fosfomycin resistance (59). No uhpA, uhpB, and uhpC homologues were found in P. aeruginosa (14). Mutations in any of these structural genes confer various levels of reduced fosfomycin permeability and resistance in E. coli (14, 18, 59, 60). E. coli mutants with reduced permeability due to altered UhpT or GlpT function were estimated to have a fitness cost of around 20% (61, 62).

TABLE 3.

Fosfomycin resistance mechanisms documented within non-Pseudomonas aeruginosa species

Category	Mechanism of resistance	Gene	Bacterium(-a)	Remarks	Citation(s)
Transporter	Inactivation of glycerol-3-phosphate transporter. Mutation causes reduced permeability.	glpT	Escherichia coli		Castañeda-García et al., 2009 (14)
	Inactivation of glucose-6-phosphate transporter. Mutation causes reduced permeability.	uhpT	Escherichia coli		Castañeda-García et al., 2009 (14); Merkel et al., 1995 (57); Olekhnovich et al., 1999 (58)
Transport regulator	Affects phosphoenolpyruvate: sugar phosphotransferase transport system. Regulates expression of transporter – reduces intracellular levels of cAMP.	ptsI	Escherichia coli		Karageorgopoulos et al., 2012 (37); Cordaro et al., 1976 (63); Lévy et al., 1990 (64)
	Codes for adenylyl cyclase. Regulates expression of transporter – reduce intracellular levels of cAMP.	cyaA	Escherichia coli		Karageorgopoulos et al., 2012 (37); Sakamoto et al., 2003 (65)
	Represses glpT and uhpT	cpxR	Escherichia coli		Kurabayashi et al., 2014 (66)
	Reduces expression of uhpT	uhpA	Escherichia coli		Castañeda-García et al., 2009 (14); Merkel et al., 1995 (57); Olekhnovich et al., 1999 (58)
		uhpB
		uhpC
Fos metalloenzyme	Mn2+- and K+-dependent dimeric glutathione S-transferase (GST). Opens epoxide ring of fosfomycin.	fosA	Enterobacterales	Acquired Subtypes: fosA2, fosA3, fosA4, fosA5, fosA6, fosA7, fosA8, fosA9, fosA10	Ito et al., 2017 (15); Aghamali et al., 2019 (35); Falagas et al., 2019 (36); Ito et al., 2017 (48); Kieffer et al., 2020 (54); Wachino et al., 2010 (69); Ma et al., 2015 (70); Guo et al., 2016 (71); Rehman et al., 2017 (72); Poirel et al., 2019 (73); Huang et al., 2020 (74)
	Mg2+-dependent enzyme. Catalyzes addition of l-cysteine or bacillithiol to fosfomycin. Opens epoxide ring of fosfomycin.	fosB	Plasmid encoded: Staphylococcus spp., Enterococcus spp. Chromosomally encoded: Bacillus subtilis	Acquired and inherent	Falagas et al., 2019 (36); Zilhao and Courvalin, 1990 (78); Etienne et al., 1991 (79); Thompson et al., 2014 (80); Cao et al., 2001 (81)
	Glutathione S-transferase	fosC2	Escherichia coli	Acquired	Wachino et al., 2010 (69)
	Bacillithiol transferase	fosD	Staphylococcus rostri, Staphylococcus aureus, Staphylococcus arlettae	Acquired	Nakaminami et al., 2008 (84); He et al., 2014 (85); Liu et al., 2017 (86)
	Glutathione S-transferase	fosG	Achromobacter denitrificans	Acquired	Kieffer et al., 2020 (54)
	Bacillithiol transferase	fosI	Mycobacterium abscessus	81% amino acid identity to FosF	Kieffer et al., 2020 (54)
	Glutathione S-transferase	fosK	Acinetobacter soli	57–63% amino acid identity to FosA enzymes	Kitanaka et al., 2014 (87)
	Glutathione S-transferase	fosL1	Escherichia coli, Salmonella enterica		Kieffer et al., 2020 (54)
	Glutathione S-transferase	fosL2	Salmonella enterica	96% amino acid identity to fosL1	Kieffer et al., 2020 (54)
	Mn2+-dependent enzyme. Catalyzes addition of water to fosfomycin. Opens epoxide ring of fosfomycin.	fosX	Listeria monocytogenes	30–35% identity to fosA and fosB	Falagas et al., 2019 (36); Fillgrove et al., 2003 (88); Fillgrove et al., 2007 (89)
Fos kinase	Catalyzes the phosphorylation of fosfomycin to fosfomycin monophosphate.	fosC	Pseudomonas syringae	fomA homologue	Díez-Aguilar and Cantón, 2019 (17); García et al., 1995 (83)
	Catalyzes the phosphorylation of fosfomycin to fosfomycin monophosphate.	fomA	Escherichia coli		Kobayashi et al., 2000 (90); Kuzuyama et al., 1996 (91)
	Catalyzes the phosphorylation of fosfomycin monophosphate to fosfomycin diphosphate.	fomB	Escherichia coli		Kobayashi et al., 2000 (90); Kuzuyama et al., 1996 (91)
Peptidoglycan biosynthesis enzyme	Cysteine-to-aspartate mutation in the MurA active site.	murA	Escherichia coli		Karageorgopoulos et al., 2012 (37); Kim et al., 1996 (94); Horii, 1999 (95)
	Overexpression of gene.

Full expression of glpT and uhpT in Enterobacterales depends on high levels of cAMP. Several enzymes have been identified that regulate these transporters (37, 63, 64). Synthesis of cAMP is dependent on the cyaA gene, which codes for adenylyl cyclase (65). The phosphotransferase PtsI, which affects the phosphoenolpyruvate:sugar phosphotransferase transport system, also regulates cAMP levels (63, 64). Mutations in either cyaA or ptsI can reduce intracellular levels of cAMP and subsequently expression of glpT and uhpT (37, 60, 61, 63, 64). Inactivation of cyaA and ptsI decreases fosfomycin sensitivity in E. coli, but their mutant homologues do not confer resistance in P. aeruginosa (14).

The two-component signal transduction system CpxAR was also confirmed to control fosfomycin resistance in enterohemorrhagic E. coli (66). Deletions in the cpxA gene resulted in constitutive expression of its regulator cpxR. In these mutants, CpxR induced resistance to fosfomycin by directly repressing glpT and uhpT expression 8- and 2.5-fold relative to the parent, respectively, thus decreasing fosfomycin transport. Fosfomycin was also found to activate the CpxAR system and repress glpT expression, affirming the role of glpT in CpxAR-induced fosfomycin resistance. A reduction in fitness of the cpxA mutant in comparison to the parent strain was detected, suggesting a fitness cost resulting from fosfomycin resistance. While cpxR was identified in P. aeruginosa, no role in fosfomycin resistance was established (67).

(ii) Plasmid-mediated Fos enzymes.

There are numerous additional plasmid-mediated Fos enzymes present in species other than P. aeruginosa. The fosA gene is also found in the genomes of many other Gram-negative pathogens including Enterobacter spp., Klebsiella spp., and S. marcescens (15). Several subtypes of FosA have been identified and frequently associated with plasmids in MDR Enterobacterales (15, 17, 18, 36, 68), including FosA2 and FosA3 (69), FosA4 and FosA5 (70), FosA6 (71), FosA7 (72), FosA8 (73), FosA9 (54), and FosA10 (74). fosA3 is the most widespread plasmid-carried fosA gene among MDR Enterobacterales (15, 48, 69), reported in almost 90% of fosfomycin-resistant extended-spectrum-β-lactamase (ESBL)-producing E. coli isolates (n = 57) in China (75). It shares 60% identity with fosA in P. aeruginosa (15). In addition, FosA3, as well as FosA^KP, was found to confer higher fosfomycin resistance (MIC of >1,024 μg/mL) than FosA^PA (MIC of 16 μg/mL) due to increased enzymatic activity and expression (76). Transferable plasmid-encoded fosfomycin resistance was not present among a collection (n = 67) of fosfomycin-resistant P. aeruginosa isolates (77).

Another type of Fos metalloenzyme is FosB, a Mg²⁺-dependent enzyme that catalyzes the addition of l-cysteine or bacillithiol to fosfomycin (36, 78–80). FosB has 38% identity to FosA (78). Expression of fosB requires the extracytoplasmic sigma factor SigW, a regulator involved in inducing fosfomycin resistance (81). The fosB gene is primarily found in Gram-positive bacteria and can be either plasmid (Staphylococcus spp. and Enterococcus spp.) or chromosome (Bacillus subtilis) carried (36, 80, 82).

Several other Fos enzymes have been reported. FosC from Pseudomonas syringae catalyzes the addition of ATP, converting fosfomycin to fosfomycin monophosphate (83). FosC2 is a plasmid-borne glutathione S-transferase found in E. coli with 51% amino acid sequence identity to FosA^PA (69). FosD is a bacillithiol transferase identified in Staphylococcus aureus and has 78.9% nucleotide and 74.1% amino acid sequence identity to FosB (84). This enzyme has also been found in Staphylococcus rostri (85) and Staphylococcus arlettae (86). Another reported glutathione S-transferase is FosK, which has 81% identity to FosF in P. aeruginosa (87). The fosK gene was detected in Acinetobacter soli and conferred very high-level fosfomycin resistance (MIC of >8,192 μg/mL). FosX is a Mg²⁺-dependent hydrolase found in Listeria monocytogenes that catalyzes the addition of water to fosfomycin (88, 89) and shares between 30 and 35% identity to fosA and fosB (36). Recently, additional Fos enzymes have been identified, including FosI, FosG, FosL1, and FosL2 (54).

(iii) Kinases.

Another mechanism of fosfomycin resistance that occurs via antibiotic modification is conferred by fosfomycin kinases. These enzymes which originate from Streptomyces spp. are encoded by the fomA and fomB genes and inactivate fosfomycin by phosphorylation (90, 91). FomA and FomB catalyze the sequential phosphorylation of fosfomycin to fosfomycin monophosphate and subsequently fosfomycin diphosphate, respectively. Both reactions involve ATP and Mg²⁺ for catalysis. Together, FomA and FomB were found to confer high-level resistance (MIC of >800 μg/mL) in E. coli (90). FosC from P. syringae is a phosphotransferase homologous to FomA with 25.8% identity that exhibits a similar mechanism of resistance (36, 92). While ATP-dependent fosfomycin resistance has been identified in P. aeruginosa clinical isolates (77), it has yet to be determined whether the fomA and fomB genes play a role in this resistance mechanism, as no enzyme characterization has been performed (93).

(iv) MurA modification.

Lastly, modification of murA can also confer fosfomycin resistance. Amino acid substitution in the MurA active site decreases fosfomycin binding affinity and thus its activity (37). In E. coli, fosfomycin normally blocks peptidoglycan biosynthesis by covalently binding to the cysteine-115 residue of MurA (94). A cysteine-to-aspartate mutation in the MurA active site results in resistance to fosfomycin. Overexpression of murA was also shown to confer fosfomycin resistance (95).

CONCLUSIONS

Fosfomycin has emerged as an alternative therapeutic against the increasing threat of MDR P. aeruginosa. However, clinical efficacy data are minimal, and fosfomycin is usually administered only in combination therapy. Despite its favorable properties of low toxicity and low cross-resistance, fosfomycin use is limited due to the lack of specific susceptibility breakpoints for P. aeruginosa, which forces extrapolation of E. coli or Enterobacterales breakpoints without supporting evidence, as well as inherent and acquired resistance mechanisms.

While numerous resistance mechanisms present in Enterobacterales, including mutations in uhpT and associated transport regulators, plasmid-mediated Fos enzymes, Fos kinases, and murA modification, are absent in P. aeruginosa, a plethora of other resistance mechanisms are found only in P. aeruginosa. These P. aeruginosa-specific mechanisms include peptidoglycan recycling enzymes and their upstream processing enzymes and chromosomal Fos enzymes. Several in vitro susceptibility studies have demonstrated P. aeruginosa MIC values were frequently higher than those of Enterobacterales isolates and clustered near or above the E. coli susceptibility breakpoints. Differences in fosfomycin resistance mechanisms may be associated with elevated MIC values in P. aeruginosa. More pharmacokinetic/pharmacodynamic (PK/PD) and clinical outcomes data are needed to aid in setting breakpoints and establishing the efficacy of fosfomycin against P. aeruginosa.