Fosfomycin resistance mechanisms in Enterobacterales: an increasing threat

Antimicrobial resistance is well-known to be a global health and development threat. Due to the decrease of effective antimicrobials, re-evaluation in clinical practice of old antibiotics, as fosfomycin (FOS), have been necessary. FOS is a phosphonic acid derivate that regained interest in clinical practice for the treatment of complicated infection by multi-drug resistant (MDR) bacteria. Globally, FOS resistant Gram-negative pathogens are raising, affecting the public health, and compromising the use of the antibiotic. In particular, the increased prevalence of FOS resistance (FOSR) profiles among Enterobacterales family is concerning. Decrease in FOS effectiveness can be caused by i) alteration of FOS influx inside bacterial cell or ii) acquiring antimicrobial resistance genes. In this review, we investigate the main components implicated in FOS flow and report specific mutations that affect FOS influx inside bacterial cell and, thus, its effectiveness. FosA enzymes were identified in 1980 from Serratia marcescens but only in recent years the scientific community has started studying their spread. We summarize the global epidemiology of FosA/C2/L1-2 enzymes among Enterobacterales family. To date, 11 different variants of FosA have been reported globally. Among acquired mechanisms, FosA3 is the most spread variant in Enterobacterales, followed by FosA7 and FosA5. Based on recently published studies, we clarify and represent the molecular and genetic composition of fosA/C2 genes enviroment, analyzing the mechanisms by which such genes are slowly transmitting in emerging and high-risk clones, such as E. coli ST69 and ST131, and K. pneumoniae ST11. FOS is indicated as first line option against uncomplicated urinary tract infections and shows remarkable qualities in combination with other antibiotics. A rapid and accurate identification of FOSR type in Enterobacterales is difficult to achieve due to the lack of commercial phenotypic susceptibility tests and of rapid systems for MIC detection.


Introduction
Antimicrobial resistance (AMR) is one of the major global public health threats in 21 st century that affects prevention and treatment of a wide range of bacterial infections (Prestinaci et al., 2015). In the last 20 years, several strategies have been developed and suggested to combat AMR. In 2012, World Health Organization (WHO) published The Evolving Threat of Antimicrobial Resistance -Options for Action, which presented interventions that will strength the health systems and enhance surveillance through improving the usage of antimicrobials in hospitals and communities, infection prevention, and encouraging the development of appropriate new drugs and vaccines (Prestinaci et al., 2015). In accordance with WHO report published in 2020,43 antibiotics and combinations are currently in clinical development and, since 2017, 11 new antimicrobial drugs have been approved for clinical use. However, WHO claims that none of the 43 antibiotics sufficiently address the problem of AMR in the most clinically problematic bacteria (e.g., Escherichia coli, Klebsiella pneumoniae). As the antibiotics availability is decreasing with time, the old antibiotics retaining effectiveness against some multi-drug resistant (MDR) pathogens are re-introduced (Theuretzbacher and Paul, 2015). This temporary solution allowed the renaissance of molecules such as colistin, nitrofurantoin and fosfomycin (FOS).  (Christensen et al., 1969). FOS interferes with the early stages of peptidoglycan production, inhibiting UDP-N-acetylglucosamine enolpyruvyl transferase (MurA) enzyme. MurA enzyme catalyzes the formation of peptidoglycan precursor, N-acetylmuramic acid. The binding of FOS to MurA and, thus, the inability to proceed in peptidoglycan formation result in a bactericidal activity of the drug (Candel et al., 2019). Since both Gram-positive and -negative bacteria requires the formation of Nacetylmuramic acid for peptidoglycan, FOS presents a broadspectrum antibiotic activity against the main genera in clinical practice, including carbapenemase-and/or extended-spectrum blactamase (ESbL)-producing Enterobacterales, methicillin-resistant Staphylococcus aureus (MRSA), glycopeptide-resistant enterococci and multidrug-resistant (MDR) Pseudomonas aeruginosa (Putensen et al., 2019). Chemically, FOS has a simple structure consisting in an active epoxic group bonded, through a carbon molecule, to a phosphorous (Baron et al., 1986). FOS has some unique features such as low molecular weight (138.06 g/mol) and protein binding capabilities, providing it with high tissue penetration (volume of distribution of 0.3 L/kg) (Candel et al., 2019). FOS mode of action was first described in 1974 by Kahan and colleagues, and the in vitro standardization testing was provided by Andrews et al. in 1983(Hirschl et al., 1980Andrews et al., 1983). Despite FOS advantages, intravenous use of FOS almost disappeared from clinical practice, partly due to its incongruency of in vitro results in early susceptibility testing (Barnett et al., 1969). FOS is available in three formulations: two orally used calcium salt form (C 3 H 5 O 4 PCa;194.2) and FOS tromethamione (C 7 H 18 NO 7 P;259.194), and an intravenously used disodium salt (C 3 H 5 O 4 PNa 2 ; 182.03) (Falagas et al., 2016). In 1996, Food and Drugs Administration (FDA) approved the clinical use of oral FOS (Monurol) in the treatment of uncomplicated lower urinary tract infections (UTIs), as acute cystitis. In the following years, FOS oral formulation was also approved in perioperative prophylaxis for transrectal prostate biopsy in adult man, post-operative treatment of UTIs, recurrent UTIs, acute uncomplicated UTIs in children and acute cystitis during pregnancy. In 2020, the European Medicine Agency (EMA) approved FOS for infusion in the treatment of a wide range of conditions (e.g. complicated urinary tract infections, bone and joint infections, bacterial meningitis) when the commonly recommended drugs are considered inappropriate (Figure 1). Some European countries such as Austria, France, Germany, Greece, and Spain allow the use of FOS intravenously with other antibiotics, such as b-lactam antibiotics or fluoroquinolones in critically ill patients suffering from carbapenem-resistant Enterobacterales infections (Michalopoulos et al., 2011). This is due to FOS' unique mechanism of action and to the absence of side effects as nephrotoxicity, typical of aminoglycosides or colistin (Michalopoulos et al., 2010). FOS usage in veterinary settings is forbidden in China and European countries, while in Central and South America regions, such as Brazil and Argentina, is largely administered in diseased broiler chickens and pigs (Peŕez et al., 2014;Wang et al., 2017). In 2016, WHO categorized phosphonic acid derivatives as critically important antibiotic in human medicine highlighting their high frequency use in human medicine and their role as available therapy to treat serious bacterial infections in people. Despite the relevance in human medicine, data concerning FOS susceptibility profiles have not been included yet in annual report on antimicrobial resistance by WHO or ECDC. Consequently, the global epidemiology of FOS resistant profiles and FOS-modifying enzymes is still incomplete and not well monitored.

FOS target
FOS binds and inhibits the UDP-GlcNAc enolpyruvyl transferase (MurA), acting as a phosphoenolpyruvate (PEP) analogue (Brown et al., 1995;Aghamali et al., 2019). MurA is a fundamental enzyme involved in the initial steps of peptidoglycan biosynthesis (Brown et al., 1995;Aghamali et al., 2019). FOS carries out its inhibiting activity to MurA through a covalent binding between the thyol group of a cysteine and the MurA active site, Cys115 ( Figure 2). This inhibitory effect occurs in the cytoplasm and impairs an earlier stage of peptidoglycan biosynthesis when compared with that of b-lactamases or glycopeptides (Kahan et al., 1974;Eschenburg et al., 2005)..

GlpT transporter
GlpT is a member of the organophosphate phosphate antiporter (OPA) family and is highly conserved in several species such as Escherichia spp., Klebsiella spp., Salmonella spp., and Citrobacter spp (Kahan et al., 1974).. GlpT is structured into two transmembrane domains, each composed of six highly conserved a-helices, that are linked by a long central loop (Lemieux et al., 2004). The glpT gene is part of the glp regulon, that controls the catabolism of G3P, glycerol and glycerophosphodiesters (Yang and Larson, 1998) (Figure 3). The extracellular G3P enters the bacterial cell through GlpT and control the expression of GlpT itself (Castañeda-Garcıá et al., 2013). In details, G3P binds to GlpR (G3P regulon repressor) that regulates the transcription of glp regulon, including glpT (Yang and Larson, 1998;Lemieux et al., 2004;Escapa et al., 2013) (Figure 3). In absence of G3P, GlpR binds to the operators of glp regulon, located in proximity of the promotor regions, and decreases the expression levels of glp regulon, including glpT (Yang and Larson, 1998) (Figure 3). When present, G3P binds to GlpR and lower GlpR-binding affinity with glp regulon, preventing the binding of GlpR to glpT promotor. The inability to bind the operator blocks glpT repression, leading to an increase of its expression levels (Cozzarelli et al., 1968;Law et al., 2009) (Figure 3).

UhpT transporter
An alternative route for FOS influx is via UhpT transport system. UhpT is a monomer consisting of twelve transmembrane a-helical segments, which show high amino acid sequence Timeline of FOS usage and the emergence of acquired FOS resistance determinants. Red dot = year of isolation. Created with BioRender.com.
Mattioni Marchetti et al. 10.3389/fcimb.2023.1178547 Frontiers in Cellular and Infection Microbiology frontiersin.org homology with GlpT (Ambudkar et al., 1990). UhpT is a member of the Major Facilitator Superfamily (MFS) and promotes the entry of G6P, fructose-6-phosphate and mannose-6-phosphate inside bacterial cell (Hall and Maloney, 2005). The UhpT system is exclusive to Enterobacteriaceae, except for Proteus spp. and Staphylococcus spp (Silver, 2017).. In the presence of G6P, UhpT expression is highly induced , leading to an increase of FOS flow inside the cell (Xu et al., 2017).

cAMP and adenylate cyclase CyaA
The transcription of both glpT and uhpT is under the control of the adenylate cyclase CyaA. CyaA catalyzes the formation of cAMP (cyclic adenosine monophosphate) from ATP ( Figure 4). Once produced, cAMP showed high affinity with the transcriptional regulator CRP (DNA-binding transcriptional dual regulator) and binds together leading to the formation of the cAMP-CRP complex. Concerning UhpT expression, the cAMP-CRP complex binds the activated UhpA and together attach the uhpT promotor, inducing its transcription (Castañeda-Garcıá et al., 2013). Similarly, regarding GlpT expression, the cAMP-CRP complex alone attaches to glpT promotor (Castañeda-Garcıá et al., 2013) (Figure 4).

Activation of CyaA
The activation of CyaA requires the presence of G6P and of the PTS system, the carbohydrate phosphotransferase system (Postma et al., 1993) (Figure 4). The PTS system is a sugar-phosphorylating system described in E. coli and requires three different entities: Enzyme I (PtsI), the heat-stable, histidine-phosphorylatable protein HPr (PtsH) and Enzyme II (composed by the domains EIIA Glc ) (Deutscher et al., 2014). Once in the bacterial cell, G6P enters the glycolysis cycle, which leads to the production of the PEP. The formed PEP undergoes to the PTS system, transferring a P group to PtsI (Deutscher et al., 2014). Thus, PtsI activates through phosphorylation PtsH, which consequently activates EIIA Glc , transferring the P group to EIIA Glc (Saffen et al., 1987). Then, the activated EIIA Glc induces the activation of CyaA (Mazéet al., 2014) ( Figure 4).

Mechanisms of fosfomycin resistance
The recent use of FOS and co-selection phenomena have contributed to the development of FOS resistance and its dissemination. FOS R mechanisms can be divided into three major groups: (a) modification of the antibiotic target MurA, (b) reduced permeability to FOS, and (c) acquisition of AMR genes. According to the recent literature, the reduction of FOS permeability is considered as the most frequent resistance mechanism (Nilsson et al., 2003;Castañeda-Garcıá et al., 2013;Silver, 2017).

Modification of the target
FOS inactivates MurA by binding to its active site, Cys115 (Skarzynski et al., 1996). Kim and colleagues demonstrated that Cys115 substitutions in MurA, as Cys115Asp, lead to in vitro FOS R (MIC > 512 mg/ml) in E. coli (Kim et al., 1996). However, mutations in MurA are uncommon in clinical isolates and none occurred in the catalytic site of MurA (Castañeda-Garcıá et al., 2013). Indeed, the first reports of mutations occurring in MurA from clinical E. coli isolates dated to 2010 in Japan, where the substitutions Asp369Asn and Leu370Ile were suggested to lead to development of FOS R in vivo (Takahata et al., 2010). Both mutations, occurring in two highly conserved residues, decreasing the susceptibility to FOS with MIC up to 512 mg/ml (Takahata et al., 2010). Subsequently, mutations in MurA associated to FOS R profiles have been detected from clinical E. coli isolates in China Bi et al., 2017), Taiwan (Tseng et al., 2015) and South Korea (Seok et al., 2020) (Table 1). Regarding Regulation of UhpT expression by CyaA activation and UhpABC system. Blue bubbles = G6P; green bubbles = G3P; red bubbles = FOS. Created with BioRender.com.  (Bi et al., 2017) Ile28Asn NA Alteration of function >256 mg/mL (Bi et al., 2017) Phe30Leu NA Alteration of function 256 mg/mL (Bi et al., 2017) Gln59Lys NA Alteration of function 256 mg/mL  Asn67Ile NA Alteration of function 256 mg/mL (Tseng et al., 2015) Val146Ala NA Alteration of function 256 mg/mL (Tseng et al., 2015) Phe151Ser NA Alteration of function 512 mg/mL (Tseng et al., 2015) Ala154Thr NA Alteration of function NA (Seok et al., 2020) His159Tyr NA Alteration of function 256 mg/mL (Tseng et al., 2015) Pro99Ser NA Alteration of function NA (Seok et al., 2020) Cys115Asp Catalytic domain Alteration of function NA (Kim et al., 1996) Cys115Glu Catalytic domain Loss of function NA (Kim et al., 1996) Glu139Lys NA Alteration of function 128 mg/mL  Trp164Ser NA Alteration of function 256 mg/mL (Tseng et al., 2015) Asp369Asn NA MurA overexpression 512 mg/mL (Takahata et al, 2010) Leu370Ile NA MurA overexpression 256 mg/mL (Takahata et al, 2010) Val389Ile NA Alteration of function 128 mg/mL  Asp390Ala NA Alteration of function 128 mg/mL  GlpT

Permeability impairment
GlpT system Impairment in GlpT activity is one of the most common mechanisms of FOS R . Strains defective in GlpT transport are not able to grow using G3P as sole carbon source (Aghamali et al., 2019). In literature, there are several reports of common mutations in GlpT associated with reduced permeability and thus increased FOS MICs (Table 1). The deletion and/or truncation in GlpT   (Seok et al., 2020) Glu130Lys NA Alteration of the target 128 mg/mL (Lu et al., 2016) Thr214Ile NA Alteration of the target 256 mg/mL (Lu et al., 2016) Asp259Asn NA Alteration of the target 128 mg/mL (Lu et al., 2016) Asp260Tyr NA Alteration of the target 512 mg/mL (Lu et al., 2016) Arg267Leu NA Alteration of the target 128 mg/mL (Lu et al., 2016) Leu282Phe NA Alteration of the target 128 mg/mL (Lu et al., 2016) Thr287Asn NA Alteration of the target >256 mg/mL (Lu et al., 2016) Thr307Lys NA Alteration of the target >256 mg/mL (Lu et al., 2016) GlpT

Arg177Lys
Transmembrane Reducted permeability 128 mg/mL (Lu et al., 2016) Phe183Leu Transmembrane Reducted permeability 128 mg/mL (Lu et al., 2016) Phe184Ile NA Reducted permeability 128 mg/mL (Lu et al., 2016) Ser205Thr protein are associated with reduction in permeability and loss of function in E. coli strains Ohkoshi et al., 2017). In 2020 Sorlozano-Puerto and colleagues investigated the effect of several mutations in GlpT from E. coli clinical isolates from Spain. The biological impact of such mutations was predicted through bioinformatic tool and tested by carbon grow test. The study identified possible alterations with a deleterious effect on GlpT activity, such as Gly84Asp, Pro212Leu, Leu373Arg, and thus a direct involvement in FOS R (Sorlozano-Puerto et al., 2020) (Table 1). Differently, deletion W28del occurring in GlpT has been associate to FOS MICs >128 mg/mL in clinical ST131 E. coli from clinical setting in Czech Republic (Mattioni Marchetti et al., 2023). Another study evaluated mutations in GlpT from ESbLproducing K. pneumoniae from hospitals in Taiwan. In this study, Lu and colleagues identified several single amino acid substitutions, occurring in the transmembrane domains, such as Arg206Lys, Ile266Ser and Ile293Phe and associated with FOS resistance at high levels (FOS MICs = 256 mg/mL) (Lu et al., 2016) (Table 2).

UhpT system
Similar to GlpT, mutations in UhpT are likely to reduce G6P entry inside bacterial cell and thus FOS permeability. Indeed, the complete loss of UhpT peptide leads to the complete loss of the transport function and leads to FOS R at high levels (FOS MICs >128 mg/mL) (Takahata et al, 2010;Li et al., 2015;Ohkoshi et al., 2017;Falagas et al., 2019). Different mutations have been reported in both E. coli and K. pneumoniae clinical strains, occurring in both transmembrane and topological domain, associated with a wide MICs range of FOS R (64 mg/mL -512 mg/mL) (Tseng et al., 2015;Seok et al., 2020;Ortiz-Padilla et al., 2022). Interestingly, Ballestero-Teĺlez and coauthors described the in vitro effect of premature Gln345stop in UhpT, which showed FOS MICs higher than 1,024 mg/mL in E. coli (Ballestero-Teĺlez et al., 2017).

UhpABC system
Impairment in the activity of UhpABC system might reduce the effectiveness of bacterial transportation systems and, consequently, reduce FOS influx into the bacterial cell (Kadner and Shattuck-Eidens, 1983). The loss of entire UhpA portion leads to different extent of FOS R (MIC > 32 mg/mL) (Ohkoshi et al., 2017;Falagas et al., 2019), while deletion of 163-188 aa or premature stop codon in UhpA contribute to high level of FOS R (MIC = 1,024 mg/mL) (Lucas et al., 2017;Ohkoshi et al., 2017)

Regulation in cAMP levels
Despite the relevant implication of CyaA activity in GlpT and UhpT expression, investigation of mutations in CyaA and its eventual effect on FOS MICs are still not clear, with just few reports conducted in E. coli strains Ohkoshi et al., 2017) (Table 1).

Acquisition of antibiotic resistance genes FosA family
FosA group is a class of metalloenzymes able to disrupt the epoxide ring of FOS drug. It depends on manganese (II) and potassium as cofactors, and on glutathione (GSH) as nucleophilic molecule. Nowadays, 11 different and genetically related variants have been deposited in GenBank Database and 10 of these are reported in the global dissemination scenario (Figures 5-7). In accordance with Ito et al., 2017, fosA genes are chromosomally distributed in Providencia stuartii, Providencia rettgeri, K. pneumoniae, Klebsiella oxytoca, Serratia marcescens, Enterobacter aerogenes and Enterobacter cloacae genomes, while they are rarely reported in E. coli, Citrobacter freundii, Proteus mirabilis and Acinetobacter baumannii (Zurfluh et al., 2020).

FosA and FosA2
The first plasmid-mediated fosA was identified and isolated from a clinical sample of S. marcescens in Spain in 1980 (Mendoza et al., 1980) (Figure 1). FosA was located on a Tn2921 cassette on the plasmid pSU912 (Seoane et al., 2010) ( Figure 7A). The origin of FosA is linked with the FOS-modifying enzyme Fos EC , located on E. cloacae chromosome (100% identity) (Garcıá-Lobo and Ortiz, 1982;Ito et al., 2017).
FosA2 variant was first reported in 2011 (Xu et al., 2011) in E. cloacae chromosome from a water sample in Canada ( Figure 1). Currently, fosA2 reports are correlated with chromosomal location only.

FosA3
FosA3 is the plasmid-acquired subtype mostly disseminated and reported (Figure 8). FosA3 shows close relation (>94% identity) to the chromosomally encoded FosA KG from Kluyvera georgiana. The first report is dated 2010 from a clinical isolates E. coli in Japan (Figure 1). Shortly after, in 2013, a fosA3 plasmid-mediated dissemination among food-chain animals in Chinese region was reported (Hou et al., 2012;Ho et al., 2013). Currently, China has the highest dissemination of plasmid-mediated fosA3 among both clinical and veterinary settings (Figures 5, 8). Concerning Chinese veterinary field, several animal species have been identified as silent reservoir, ranging from pets, as dogs and cats, to food-chain animals, as pigs and bovines, and wild animals, as pigeons. FosA3 Epidemiological map of FosA among Enterobacterales. Created with mapchart.net.
First isolation of plasmid-mediated fosA3 in clinical K. pneumoniae strains was in 2012, when Lee and co-authors described the co-presence of fosA3+bla CTX-M-14 on an IncN plasmid and organized in IS26-ISEcp1-bla CTX-M-14 -DIS903D-IS26-fosA3-orf1-orf2-Dorf3-IS26 (with a spacer sequence of 1,222 bp) (Lee et al., 2012). Lately, in 2015, Jiang Y et al. reported the characterization of 94 KPC+FosA3 co-producing K. pneumoniae collected from twelve Chinese hospitals. Additionally, the authors highlighted a clonal relation between KPC-and FosA3-producers, indicating a FOS R clonal dissemination in China (Jiang et al., 2017). In K. pneumoniae plasmid-mediated fosA3 is largely associated with isolates belonging to ST 11 (Xiang et al., 2015;Nishida et al., 2020), ST37 (Taniguchi et al., 2017), ST485 . In recent years, a secondary spread of plasmid-mediated fosA3 occurred in Salmonella spp. among food-chains animals and humans in China Zhang et al., 2020). Outside Chinese settings, similar cases have been recorded from pediatric patients in Spain (Vaźquez et al., 2022), from clinical patients in USA (Turcotte et al., 2022), and from a wild bird in Germany (Villa et al., 2015). Noteworthy, Villa and colleagues described the first case of a Salmonella enterica Serovar Corvallis co-producing FosA3+NDM-1+CMY-16. FosA3 and bla NDM-1 were located on the same IncA/C2 plasmid and fosA3 included in a type A transposon (Villa et al., 2015). This report highlighted the bird's migration as route for environmental diffusion of fosA3 from norther Asia to Europe (Villa et al., 2015). Among Salmonella spp. strains, transposon Type A is the most spread fosA3 environment, located on IncFII (Lin and Chen, 2015) and IncHI2 (Wong et al., 2016), followed by Type C on IncFIB (Vaźquez et al., 2022) and type D on IncHI2 (Wong et al., 2016). Interestingly, a multi-replicon IncC-IncN plasmid, coharboring fosA3 Type A and bla CTX-M-14 have been already isolated from chickens in China . FosA3 cases occurred in Salmonella ST32 (Vaźquez et al., 2022), ST17 , ST34 , ST198 . Since 2017, few reports evaluate the occurrence of plasmidmediated fosA3 in P. mirabilis from both hospitalized patient and food-chain animals (He et al., 2017;Hua et al., 2020;Lei et al., 2020). The first case focused on the chromosomal integration of bla CTX-M-14 /bla CTX-M-65 and fosA3 in P. mirabilis collected in 2015 from diseased broilers in China, with the following compositions: a) IS26-DISEcp1-bla CTX-M-14 -DIS903-fosA3-1,222 bp-IS26; b) IS26-DtraI-fip-DISEcp1-bla CTX-M-65 -IS903D-iroN-IS26-fosA3-1758 bp-IS26. In the same study, the presence of the transposition unit "b" was detected in IncHI2 plasmid from E. coli ST117, together with the presence of minicircles that contain fosA3, bla CTX-M-65 and IS26 (He et al., 2017). Thus, the authors speculated the fosA3+bla CTX-M-65 integration into the P. mirabilis chromosome via a transposable minicircle from E. coli (He et al., 2017). Similarly, the presence of minicircles harboring IS26 and fosA3 was identified even in S. enterica from a Chinese chicken and speculations about their role in fosA3 acquisition and spread are under evaluation . Similar environments containing bla CTX-M-65 + fosA3 were identified in retail meat and aquatic products from markets (Ma et al., 2022), from diseased pig (Lei et al., 2018;Song et al., 2022) and from retail chickens  from Chinese regions, while the co-expression CTX-M-3 +FosA3 was reported from Chinese chicken (Turcotte et al., 2022). Rather worrying was the isolation of a KPC-2+CTX-M-65+FosA3 producing P. mirabilis from a Chinese 49-year-old female with a pulmonary infection (Hua et al., 2020). The bla CTX-M-65 +fosA3 was located on an IncFII-33 and the authors emphasized the successful association of IS26 and IncFII-33 in spreading antimicrobial resistance features (Hua et al., 2020).
FosA3 easily fits in different plasmid environments, including single-and multi-replicons. The major vehicle of plasmid-mediated fosA3 spread is IncFII (Hou et al., 2012), followed by IncI1 (Sato et al., 2013), IncN (Liu et al., 2022), IncHI2 , and IncP (Hameed et al., 2022). The successful and global diffusion of fosA3 could be explain by the combination of IS26 sequences and IncFII plasmids. FosA3 genes are mainly flanked by IS26, that play a fundamental role in AMR effective transposition and in their AMR dissemination among Enterobacterales (Partridge et al., 2018;Lv et al., 2020). Moreover, as mentioned elsewhere, IS26-flanked transposons are able to form circular intermediates that could accelerate the spread of fosA3 (He et al., 2015;Harmer and Hall, 2016). The IncFII plasmids are commonly low copy number plasmids and are recognized as vehicles of ESbLs dissemination among Enterobacterales (Muthuirulandi Sethuvel et al., 2019). Moreover, researchers speculate on the role of IncFII F33:A-:Band F2:A-:B-in fosA3 dissemination due to its high adaptation levels (Hou et al., 2012;Sun et al., 2012).

FosA4
FosA4 enzyme shows 94% amino acid identity with FosA3, and speculation proposes Kluvyera georgiana as possible origin of the plasmid-mediated resistance gene fosA4 (Nakamura et al., 2014;Rodriguez et al., 2018). FosA4 epidemiology is limited and varies geographically, but it was mainly reported in E. coli isolates (Figure 8). Increasing cases of FosA4-producing E. coli have been reported among food-chain animal settings in Egypt Sadek et al., 2022) and in France (Lupo et al., 2018). Other cases, concerning clinical settings, have been described from hospitals in Madrid (Loras et al., 2021) and Australia (Mowlaboccus et al., 2020). In Southern Turkey, Cansu Önlen Güneri and co-authors described a regional diffusion of plasmidmediated fosA4 among E. coli collected from waste-water treatment plant (Güneri et al., 2022). The fosA4 gene has been reported predominantly on IncFII plasmid type and, consequently, on IncHI2 and IncI1 (Ma et al., 2015;Loras et al., 2021;Ramadan et al., 2021). IncFII and IncI1 normally harbors additional genes responsible for resistance to other antibiotics such as penicillins, sulphonamides and aminoglocosyde (Mowlaboccus et al., 2020;Ramadan et al., 2021). FosA4-harboring plasmids often coexist with bla CTX-M -and mcr-1-harboring plasmids Sadek et al., 2022). FosA4 is associated with a conserved cassette of 4,022 bp in size, consisting of: two IS26, fosA4, tetR/acrR family and a helix-turn-helix domain. In southern Turkey, a novel genetic enviroment was detected, replacing the upper IS26 with an IS4 (Güneri et al., 2022) (Figure 7B). MIC data for fosA4 have been reported in E. coli as >1,024 µg/ml (Güneri et al., 2022).

FosA5
In 2015, Ma Y et al. reported the first case of fosA5 from a clinical E. coli strain in an inpatient with hospital-acquired pneumonia in China (Ma et al., 2015). FosA5 enzyme shares 69% amino acid sequence similarity with FosA and 80% with FosA3. The K. pneumoniae chromosome has been identified as the origin of fosA5 variant and its spread is associated with pKP96 plasmid, as reported by Ho PL et al., 2013(Xu et al., 2011. The genomic enviroment of fosA5 is characterized by insA and insB and an IS10 in the opposite side ( Figures 10A, B). In 2019 Wang S and colleagues investigated the genomic enviroment of an IncHI2A plasmid (pIMP26) coharboring bla IMP-26 , bla DHA-1 and fosA5, isolated from an E. cloacae strain involved in blood infection (Wang et al., 2019). In pIMP26, the fosA5 structure was as follow: IS4, rfaY, lysR, fosA5, rfaY, ISVsa5 (IS4-like) (Wang et al., 2019). A similar organization of the fosA5 cluster has been detected in an IncHI2/2A plasmid (pEHZJ1) from an E. hormaechei of clinical origin (Gou et al., 2020). FosA5-carrying E. coli strains were found to be highly FOS R (MIC = 512 µg/ml) (Ma et al., 2015).

FosA6
FosA6 was firstly described in a clinical CTX-M-2-producing E. coli ST410 from an US hospital in 2017 (Guo et al., 2016). FosA6 was carried on a self-conjugative IncFII plasmid (69 kb) and inserted in the cassette IS10R-DlysR-fosA6-DyjiR_1-DIS26, nearly identical to those on the chromosomes of some K. pneumoniae strains ( Figure 7C). Moreover, fosA6 shared >99% sequence identity with chromosomally encoded fosA in K. pneumoniae. A point prevalence study conducted among seven Hospitals in Madrid, identified the only European case of ST354 E. coli producing FosA6 enzyme (Loras et al., 2021). FosA6-carrying E. coli had FOS MIC values of 128 to >1024 µg/ml (Guo et al., 2016).

FosA7
In 2015 Dhanani and colleagues investigated the resistome of four FOS R S. enterica serovars Heidelberg from broiler chickens among different commercial farms in Canada (Dhanani et al., 2015). As described later by Rehman et al., the 4 S. enterica strains produced a FosA-like enzyme, named FosA7, with a chromosomal location (Rehman et al., 2017).
Currently, 9 alleles of fosA7 genes are deposited in GenBank (fosA7.1-fosA7.9). All these variants have a chromosome location among different bacterial species. FosA7.5 and fosA7.9 are strictly linked with the chromosome of E. coli and C. freundii, respectively. In Salmonella spp. fosA7 is surrounded by two hypothetical proteins and located in an integrase cassette composed of Int-type II endonuclease-ATP helicase-type II methylase-RNA helicase-DNA helicase ( Figure 11A). In E. coli the intercellular diffusion of fosA7.5 is due to the composite transposon flanked by ISL3 and IS3 (IS911 and ISEC52) elements ( Figure 11B-E). A different composition has been highlighted for fosA7.9 in C. freundii: the fosA7.9 cassette is flanked by HNH endonuclease at both sides and organized in HNH endonuclease-fosA7.9-Fic family-type II restriction-DNA methyltransferase-AAA domain-HNH endonuclease ( Figure 11F) (Mattioni Marchetti et al., 2023).
The epidemiology of FosA7 family displays a relevant dissemination, with reports in livestock animals, clinical settings and enviroment (Balbin et al., 2020;Jovcǐćet al., 2020;Mosime et al., 2022). The Canadian and USA regions reported the larger diffusion of fosA7, followed by China (Pan et al., 2021). Recently, cases of FosA7 enzymes have been described in South Africa from Citrobacter koseri (Ekwanzala et al., 2020), in Czech Republic from C. freundii (Mattioni Marchetti et al., 2023) and in Poland from E. coli (Skarzẏńska et al., 2021). Expression of FosA7 showed high value of FOS R MIC (>512 mg/ml) (Rehman et al., 2017).

FosA9
FosA9 has been reported in 2019 by Doesschate et al. from an E. coli strain causing bacteremia in Utrecht. The patient had suffered from recurrent episodes of sepsis, with blood cultures positive for K. variicola, which was identified as the source of fosA9. The fosA9 genomic environment consisted of a ISEcp1-syrM1-fosA9-lysN2 region, flanked by 5 bp DRs (AAAAA) and identical to those found in K. variicola (Wang et al., 2019) (Figure 7E). The expression of FosA9 confers FOS R at high levels, with MIC > 1,024 µg/ml (Ten Doesschate et al., 2019).

FosA10
The FosA10 enzyme has been described by Ying Huang et al. from a local broiler meat outlet in Pakistan. A 53,736 bp IncFII plasmid harbored the fosA10, inserted in a 4,328 bp variable region, flanked by two copies of IS10 element  ( Figure 7F). Differently, the identical genomic enviroment was identified on a IncK plasmid from a clinical ST648 NDM+FosA10producing E. coli isolated in Czech Republic (Mattioni Marchetti et al., 2023). FosA10 shares highest identity with FosA6 and FosA9 (ID = 97.84%), confirming its possible origin from K. pneumoniae species . In E. coli strains FosA10 induces FOS R phenotype with MIC >128 µg/ml .

FosL1 and FosL2
FosL1 is a novel glutathione S-transferase metalloenzyme that shared a 63% identity with FosA8. FosL1 was described on a conjugative IncX1 plasmid in a E. coli strain of a Swiss patient (Kieffer et al., 2020). The genomic enviroment surrounding fosL1 consisted of a mobile insertion cassette, flanked by DIS91-like at both sides. The same fosL1 cassette, was detected on an IncQ1 plasmid from a clinical S. enterica. Subsequently, an in-silico analysis of fosL1 identified a similar gene, classified as fosL2, on an IncP-like plasmid, collected from a clinical S. enterica strain. Genomic environment of fosL2 consisted of Tn7L-like-fosL1-urk-Tn7R-like and flanked by Dhyp at both sides (Kieffer et al., 2020) (Figure 13). The ancestor source for FosL1-2 remains unknown. FosL1 induces FOS R profile at high level (MIC = 1,024 µg/ml) (Kieffer et al., 2020).

Epidemiological breakpoints and detection strategies
According to European Committee on Antimicrobial Susceptibility Testing (EUCAST) and the Clinical and Laboratory Standards Institute (CLSI), agar dilution method (ADM) is the gold standard for FOS MIC detection in both Gram-positive and -negative bacteria but the breakpoints for FOS susceptibility have been formalized for few species and are different for CLSI and EUCAST. EUCAST breakpoints for Enterobacterales define as susceptible (S) MIC ≤ 32 mg/L and resistant (R) MIC > 32 mg/L, while CLSI breakpoints for E. coli are S ≤ 64 mg/L, I =128 mg/L, R ≥ 256 mg/L (Falagas et al., 2008).. Currently, there is a lack of fast, time-saving susceptibility tests for FOS and the limited breakpoints standardization, that highlights the difficulty in monitoring FOS profiles epidemiology and in identifying FOS R strains. In this section we describe the current available methods for the investigation of FOS susceptible profiles among Enterobacterales. Structure of representative genetic environments of fosC2. (A) (AB522969) (Lucas et al., 2017), (B) (KM877517) (Guo et al., 2016). Yellow = IS, light blue = integrase, red = antimicrobial resistance genes, gray = open-reading frame, black = unknown protein.

Agar dilution method (ADM)
The reference method ADM consists in the incorporation of different concentration of FOS (generally from 0.25 mg/ml up to 1,024 mg/ml) into Mueller-Hilton (MH) agar, added with 25 mg/L of G6P; Balouiri et al., 2016). Then, a 0.5 MacFarland suspension of the studied strain is prepared and diluted, to obtain the final inoculum required of 1 × 104 CFU/spot (2 ml). When replicators with 1-mm pins that deliver 0.1 to 0.2 mL are used, dilution of the initial suspension is not recommended. After inoculation, the plates are left at room temperature until the inoculation spots are completely absorbed into the agar (no more than 30 minutes). Incubate at 35 ± 2°C for 16 to 20 hours. The MIC value corresponds to the concentration in which a growth reduction of at least 80% is obtained, as compared to the control. The method should be conducted at least in duplicate. Although ADM remains the reference method for FOS MIC evaluation, it is not used routinely in diagnostic practice due to its labor-intensity and high time requirement (16-20 h) (Croughs et al., 2022). Alternative and faster methods, as gradient and disk diffusion test, or routinely used automated systems, as Vitek2, resulted unreliable due to their poor ability in detecting FOS R isolates, with high error rates (van den Bijllaardt et al., 2018;Croughs et al., 2022). According to EUCAST guidelines, the disk diffusion test is intended only in investigating FOS profiles among E. coli strains, using 200 mg FOS disk and in presence of 50 mg of G6P.

Rapid fosfomycin/E. coli NP test
Nordmann and co-authors reported the description of a rapid test for FOS susceptibility profiles in E. coli . The rapid test is based on the microbial ability to metabolize glucose, that induce a colorimetric change of a specific pH indicator (culture medium, 2.5% MHB-CA powder, 0.005% phenol red indicator, and 1% D(+)-glucose). The test consists in preparing two solutions, named NP solutions: one solution with 25 mg/ml G6P and 40 mg/ ml FOS, and one without. For bacterial suspension, a 3.0 to 3.5 McFarland solution for each tested isolate is prepared in 5 ml of sterile NaCl (0.85%). A 96-well polystyrene microtest plate is filled with both NP solutions and the bacterial suspension is directly inoculated in the presence or absence of FOS. After an incubation of 1 h 30 min at 35 ± 2°C, color changes are visually detected. FOSresistant E. coli triggers a color change from orange to yellow, while FOS-susceptible remains orange .
isolates tested, all showed a positive result to the test . Similarly, Mueller and co-authors revealed a 100% correlation between susceptibility and resistance strains after screening 1,225 clinical ESbL-producing E. coli . The rapid fosfomycin/E. coli NP test has the potential to be used as a rapid and first-step screening of FOS-resistant E. coli, thanks to its good performance and rapidity. A more recent evaluation on the accuracy of this rapid method was conducted on 149 clinical E. coli isolates, showing high rate of sensitivity and specificity (94.2% and 98.75%, respectively) and highlighting the reliability of the technique (Yunus et al., 2021). Differently, Kansak and colleagues found similar rate of sensitivity and specificity (95.9% and 100%, respectively) but a Very major Error (VME) of 22.2%, limiting the possibility to use the rapid test instead of ADM (Kansak et al., 2021). Despite the potential offered, the use of the rapid fosfomycin/E. coli NP test is still limited due to its applicability on E. coli only, the difficult in the interpretation of the results and the inability to distinguish between chromosomal and plasmid-acquired resistance mechanisms .

SuperFOS selective medium
The SuperFOS selective medium provide a first line screening for FOS resistant Enterobacterales.
The SuperFOS medium combines the differentiation features of the CHROMagar orientation medium with an optimal concentration of FOS (16 mg/ml) and G6P (25 mg/ml). To avoid any contamination by eventual Gram-positive organism and fungi, the SuperFOS medium is enriched with vancomycin (20 mg/ml) and amphotericin B (5 mg/ml).
This medium provides several advantages due to its ease in preparation, the low cost, and the excellence performance, with both sensitivity and specificity at 100%. Moreover, the medium allows a first step screening of both chromosomal and plasmid mediated FOS R mechanisms among Enterobacterales from clinical specimens .

Disk potentiation testing with PPF
The disk potentiation testing with sodium phosphonoformate (PPF) is an agar-based diffusion test requiring the presence of FOS, G6P and PPF. PPF, commercially named Foscarnet, is an anti-viral compound used primarily in the treatment of CMV infections with inhibitory properties against FosA and FosC2 enzymes (Schreiber et al., 2009;Nakamura et al., 2014). PPF is able to bind FosA/FosC2 enzymes interacting with the residue MnII(+) and Thr9 that are present in the active site of FosA/FosC2-like enzymes, leading to a inhibitory effect and, thus, restoring the FOS susceptibility . The test requires MH agar plates with 25 mg/L G6P, 0.5 MacFarland solution of the isolate to investigate, two disk of FOS (50 mg) and PPF (1 mg). The cutoff is set to a 5 mm enlargement in the inhibition zone of FOS+PPF disk compared with the FOS disk alone (Nakamura et al., 2014). This agar-based method shows 100% sensitivity and specificity, and successfully detects the producing of enzymes FosA/A2 (Rigsby et al., 2004), FosA3, FosA4, FosA6 (Loras et al., 2021), FosA7 (Mattioni Marchetti et al., 2023), FosA8 (Biggel et al., 2021), FosA10 (Mattioni Marchetti et al., 2023), FosC2 (Nakamura et al., 2014), and FosL1 (Kieffer et al., 2020). However, the PPF test has been validated for E. coli strains only.

Carbon source growth test
The carbon source growth test evaluates the ability of a bacterial strain to grow with G3P or G6P as the sole source of carbon. The inability to grow in presence of G3P and/or G6P is the result of a functional deficiency of the transporters GlpT and UhpT, respectively (Huang et al., 2021). This method requires the inoculation of the bacterial isolate on a M9 minimal medium agar supplemented with G3P or G6P at 0.2% (w/v) (Sorlozano-Puerto et al., 2020). After an incubation phase at 36°C for 48 h, the poor or total absence of growth is associated to an impairment in the transporter's activity (Sorlozano-Puerto et al., 2020). The limitation of this growth test is mainly represented by the time required to perform it (72 h for results) and restricted results only on direct impairment of GlpT and UhpT activity.

Limitations
This review presents several limitations. Few studies evaluate the prevalence of amino acidic mutations in proteins involved in FOS influx and their possible effect in FOS MIC increase (Kim et al., 1996;Takahata et al, 2010;Li et al., 2015). Whereby, the knowledge on specific mutations affecting FOS influx is not clear and incomplete.
Considering plasmid-mediated mechanisms for FOS R , the update global epidemiology of fosA/fosC2/fosL1-2 gene is not completely and clearly monitored, mainly due to the lack of national surveillance plan, of fast methodology for the investigation of FOS R profiles and the lack of general interest. Moreover, the characterization of fosA-like gene variants is so far only through molecular investigations and/or WGS. These point together, explain the difficulty to draw the updated epidemiology of FosA/C2/L1-2 enzymes and to clearly specific mutation decreasing FOS MICs.
Additionally, this review describes the FOS R mechanisms that has been investigated and reported in literature among Enterobacterales only, while does not consider other relevant FOS R sources, as S. aureus and Enterococcus faecium.

Further perspective
FOS is still a valid option against MDR Enterobacterales, but this molecule is not always monitored routinely in clinical practice or in surveillance plans and, thus, the resistance mechanisms involved are not further investigated. In a scenario of increasing FOS R , time-saving and user-friendly methods for detecting such resistance profiles turn out to be fundamental. Implementation of faster testing would allow to conduct wide surveillance studies and to monitor FOS in clinical routine.
Time-saving methodologies aforementioned are validated for E. coli only. Therefore, the validation of these methods to further species would extend the pool of strains that can be tested, providing a more in-depth knowledge about FOS R epidemiology. Moreover, a faster detection of FOS resistant bacteria and thus a further molecular characterization, could provide more information even on rarely reported FosA-like enzymes, such as FosC2, FosA4, FosA8 and FosA9, and could supply a more update epidemiology on fosA/C2/L1-2 genes spread.

Conclusion
Even though FOS is an old antimicrobial drug, it has unique and favorable features that lead in the last 20 years it to be considered as an additional resource in the treatment of MDR microorganisms' infections (Michalopoulos et al., 2011). This review described the different mechanisms, identified so far, leading to FOS MIC increase among Enterobacterales genus. The FOS influx inside bacterial cell, that is regulated by different transporters and associated regulators, has also been described. Impairment in FOS transporters GlpT and UhpT is the most common mechanisms leading to the increase in FOS MICs, reported both in vitro and in vivo (Nilsson et al., 2003). The scientific community identified specific hotspot mutations in GlpT associated to a FOS resistance at high levels (FOS MICs > 128 mg/mL), such as W28del and Pro212Leu in E. coli, and as Arg206Lys and Ile293Phe in K. pneumoniae (Lu et al., 2016;Sorlozano-Puerto et al., 2020;Mattioni Marchetti et al., 2023). Compared with mutation frequency in GlpT and UhpT, modification in the target MurA were uncommon in vivo and no reports identified mutations in the active site (Cys115) in clinical isolates. In clinical E. coli strains the mutations Asp369Asn and Leu370Ile in MurA can likely develop FOS resistance profiles with MICs up to 512 mg/ml, while in clinical K. pneumoniae isolates the modifications Asp260Tyr and Thr307Lys has been associated to FOS MICs = 128 mg/mL (Takahata et al, 2010;Lu et al., 2016). The study of specific mutations in proteins involved in FOS influx and their eventual effect on FOS MICs is not in deep investigated and required further investigations.
Regarding acquired FOS R mechanisms, in the last twelve years there has been a global diffusion of metallo-enzymes, named FosAlike, FosC2 and FosL1-L2 (Zurfluh et al., 2020). The Chinese clinical and veterinary environments show the highest frequency of FosA/ C2 enzymes but, recently, many other countries as Brazil, Japan, Spain, and USA have reported such enzymes as well (Wachino et al., 2010;Jiang et al., 2017;Loras et al., 2021;Ewbank et al., 2022;Turcotte et al., 2022). To date, 11 variants of FosA enzymes has been identified, contributing to FOS resistance at different extents. In the global scenario, fosA3 is the predominant type and it is widely reported in humans and veterinary settings. The wide and fast diffusion of fosA3 has been facilitated by the combination of IS26-mediated transposons with epidemic broad-host-range plasmids as IncFII plasmids. The versatility of these fosA3-harboring plasmids has allowed the acquisition of fosA3 genes in several clinically important ST such as E. coli ST10, E. coli ST69, E. coli ST131, K. pneumoniae ST11 and S. enterica ST32 (Xiang et al., 2015;Falagas et al., 2019;Seok et al., 2020). FosA3 is commonly co-expressed with other ESbLs, as CTX-M-65, and even with carbapenemases as KPCs, NDMs and VIMs (Villa et al., 2015;Xie et al., 2016;Jiang et al., 2017;Tang et al., 2020). Worryingly, the co-occurrence of fosA3 + mcr-type genes in carbapenemases-producing Enterobacterales has been already described in the literature Peng et al., 2019;Tian et al., 2020).
Originated from K. pneumoniae chromosome, FosA5 and FosA6 can be considered among the most frequent metalloenzyme leading to FOS R . However, their epidemiology has not been widely investigated in strains other than K. pneumoniae and the few reported cases are confined to countries as China and Spain (Guo et al., 2016;Wang et al., 2019). The diffusion of both fosA5 and fosA6 in E. coli is linked to IS10 flaking cassettes (Xu et al., 2011;Dhanani et al., 2015).
Since the discover in 2015, FosA7 has rapidly spread among Enterobacterales, with high predominance among Salmonella spp. So far, nine alleles of fosA7 have been described and deposited in the GenBank. FosA7-like genes are strictly located on Salmonella spp. chromosome, except for fosA7.5 and fosA7.9 that are associated to E. coli and C. freundii chromosome, respectively. The current spread of fosA7-like genes includes countries as Canada and China (Dhanani et al., 2015;Pan et al., 2021).
Concurrence of impairing mutations in FOS influx and acquisition of fosA/C2/L1-2 together with ESbLs and carbapenemases genes, is worrying and could strongly affect the use of FOS in severe infections treatment.
ADM is the reference methods for FOS MICs evaluation and the few rapid methods available have been validated for E. coli only or are prone to error. The increase of surveillance plans and the implementation of new rapid approaches for the detection of FOS R Enterobacterales, would favorite a better and in-depth knowledge on the prevalence of FOS R mechanisms. Moreover, a clearer information on such mechanisms and their dissemination results of priority importance to halt eventual FOS R dissemination and to optimize therapeutic strategies.

Author contributions
VM, IB, and JH played an important role in searching the relevant literature, writing and correcting the manuscript. All authors contributed to the article and approved the submitted version.

Funding
The study was supported by research project grants NU20J-05-00033 and NU23J-09-00067 provided by the Czech Health Research Council and by the project National Institute of Virology and Bacteriology (Program EXCELES, ID project no. LX22NPO5103), funded by the European Union-Next Generation EU.

Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.