Mechanism of Reduced Susceptibility to Fosfomycin in Escherichia coli Clinical Isolates

In recent years, multidrug resistance of Escherichia coli has become a serious problem. However, resistance to fosfomycin (FOM) has been low. We screened E. coli clinical isolates with reduced susceptibility to FOM and characterized molecular mechanisms of resistance and reduced susceptibility of these strains. Ten strains showing reduced FOM susceptibility (MIC ≥ 8 μg/mL) in 211 clinical isolates were found and examined. Acquisition of genes encoding FOM-modifying enzyme genes (fos genes) and mutations in murA that underlie high resistance to FOM were not observed. We examined ability of FOM incorporation via glucose-6-phosphate (G6P) transporter and sn-glycerol-3-phosphate transporter. In ten strains, nine showed lack of growth on M9 minimum salt agar supplemented with G6P. Eight of the ten strains showed fluctuated induction by G6P of uhpT that encodes G6P transporter expression. Nucleotide sequences of the uhpT, uhpA, glpT, ptsI, and cyaA shared several deletions and amino acid mutations in the nine strains with lack of growth on G6P-supplemented M9 agar. In conclusion, reduction of uhpT function is largely responsible for the reduced sensitivity to FOM in clinical isolates that have not acquired FOM-modifying genes or mutations in murA. However, there are a few strains whose mechanisms of reduced susceptibility to FOM are still unclear.


Introduction
Escherichia coli is a causative agent of uncomplicated urinary tract infections in immunocompetent hosts and opportunistic infections in immunocompromised hosts. In recent years, fluoroquinolone-resistant and/or extended spectrum -lactamase-(ESBL-) producing E. coli strains have been frequently isolated from such patients [1][2][3][4]. In addition, these E. coli resistant strains occasionally show cross-resistance to aminoglycosides [5]. Thus, these multidrug-resistant E. coli have an impact on the selection of therapeutically effective drugs.
Because of this serious concern, the use of fosfomycin (FOM), an antibiotic as a bacterial cell wall synthesis inhibitor developed 40 years ago, has been reevaluated against drugresistant bacteria, especially E. coli [6,7]. Since FOM has a unique mode of action that differed from other antibiotics, it is expected to display little cross-resistance to other antimicrobial agents. E. coli is the most frequent causative pathogenic bacterium of acute cystitis [8,9]. For instance, trimethoprim/sulfamethoxazole and FOM are recommended as the first-line treatment in acute uncomplicated urinary tract infections according to the US guideline [10]. On the other hand, the guideline of the Japanese Association for Infectious Diseases and Japan Society for Chemotherapy recommends fluoroquinolones as the first-line drug and FOM as the second-line drug [11].
Nevertheless, several surveillance studies report that the rate of emergence of E. coli isolates showing FOM resistance or reduced FOM susceptibility has been markedly low [7,9,32]. Although spontaneously mutational rate to acquire FOM resistance is high, FOM resistance confers biological costs, such as reduced cell growth rate in Gram-negative bacteria [29,33]. This indicates that FOM continues to be an effective agent against E. coli infections. However, the overall up-to-date status of FOM resistance needs to be continuously surveyed and its molecular characteristics have to be understood to prevent future emergence and increase of multidrug-resistant E. coli with FOM resistance in the clinic. In this study, we screened E. coli clinical isolates from Japan showing resistance or reduced susceptibility to FOM and identified the molecular mechanisms of their reduced susceptibility.

Antibiotic
Susceptibility. FOM, imipenem (IPM), and ceftazidime (CAZ) were provided by Meiji Seika Pharma (Tokyo, Japan), MSD (Tokyo, Japan), and Glaxo SmithKline (Tokyo, Japan), respectively. Minimum inhibitory concentration (MIC) was determined by broth microdilution method or agar plate dilution method according to the recommendations of the Clinical and Laboratory Standards Institute (CLSI) with breakpoints according to CLSI guidelines [34].
a Primers designated "full" were used for direct sequencing, and "qPCR" were used for real-time RT-PCR.
The MIC of FOM was measured by the agar dilution method in the absence of G6P, in the presence of 25 g/mL G6P, and in the presence of 25 g/mL G6P and 2 mM cAMP.

Measurements of Carbohydrate Phosphate Transporter
Activity. E. coli cells were cultured for 24 h in Muller-Hinton broth and harvested by centrifugation. The cell pellet was washed twice with saline. E. coli cell suspension with saline was then used to inoculate to M9 minimum salt (Becton Dickinson, Franklin Lakes, NJ) agar supplemented with 0.2% G6P or 0.2% G3P [18]. Cell growth was observed after incubation for 24 h in the case of G6P and for 48 h in the case of G3P.
No murA coding sequence mutations that would result in  , uhpA and uhpT were not detected by PCR amplification with two distinct primer pairs ("full" and "partial" in Table 1). In one strain with MIC 32 g/mL (SRE252), cyaA was not detectable by PCR. All strains with reduced FOM MIC (≥8 g/mL) except one (SRE29), had several mutations in one or more genes leading to amino acid deletion or point mutation(s) of amino acid residues, compared with other susceptible strains.

Function and Expression of Carbohydrate Phosphate
Transporters. To determine the activity of UhpT and GlpT, we examined cell growth in M9 minimum salt solution supplemented with G6P or G3P. Nine of ten strains with FOM MIC ≥ 8 g/mL did not grow within 24 h on G6P-containing M9 minimum salt agar. On the other hand, only three of ten strains did not grow in the presence of G3P (Table 3), and the two strains (SRE91 and SRE252) showed reduced MIC to FOM by addition of cAMP. This suggested that these two strains shared insufficient intracellular concentration of cAMP for full expression of GlpT and/or UhpT. These results suggested that FOM incorporation through the UhpT system is involved to a greater extent in the reduction of E. coli FOM compared to the GlpT system. Expression of UhpT is induced by G6P [15][16][17]. We therefore determined uhpT gene induction by G6P using quantitative RT-PCR analysis (Table 3 and Figure 2). In strains with FOM MIC < 1 g/mL, uhpT expression was strongly induced (190-to 730-fold) by G6P. By contrast, in eight of ten strains with FOM MIC ≥ 8 g/mL, the uhpT induction was markedly lower (0.56-to 3.8-fold) or no uhpT signal was amplified by PCR. Of the remaining two strains, we observed a high induction (200-fold) of uhpT expression by G6P in SRE253 strain; however, the strain did not grow on G6P-containing M9 minimum agar. High induction (1200fold) of uhpT by G6P was observed in SRE29 strain, and the strain grew on G6P-containing M9 minimum agar; however, it showed reduced susceptibility to FOM (MIC 32 g/mL). These results indicated that G6P-dependent growth lacked in most strains with reduced susceptibility to FOM because of lack of uhpT or a markedly reduced uhpT induction by G6P. However, the changes in uhpT expression and UhpT activity did not account for the reduced FOM susceptibility of a few strains (Figure 2).

Discussion
The frequency of FOM resistance has been recognized to be low. And it has been expected that the frequency of crossresistance of FOM and other antibiotics is very low because of a unique mode of action. However, several reports examined FOM resistance in ESBL-producing E. coli [19,23,24,28]. A report indicates that FOM-resistant and intermediate E. coli are more frequently resistant to other types of antimicrobials than FOM-susceptible strains [28]. These reports suggest the importance for surveillance of FOM-resistant E. coli, particularly for focus on their cross-resistance of FOM in multidrug resistance.
In this study, only one resistant and two intermediate strains were found among 211 clinical isolates from Japan ( Figure 1). This indicated that FOM was a promising candidate agent against E. coli infections as generally described. However, we found seven strains susceptible according to the CLST breakpoint but reduced susceptibility (MIC ≥ 8 g/mL). Under the selective pressure of FOM usage, these strains might acquire FOM resistance more easily. In the present study, cross-resistance of strains with reduced FOM susceptibility to other antimicrobial drugs was not significantly higher than in other strains, and these strains were not concentrating on specific genotypes.
Based on previous reports, acquisition of FOMmodifying enzymes (encoded by fos genes) and mutations in murA gene results in resistance, that is, surpass of breakpoints, to FOM [18,19,28]. The present study did not identify such strains. Another resistance mechanism is altered FOM incorporation into bacterial cells. The G6P and G3P transporters contribute to incorporation of FOM into cells [13,14]. Loss of function or decreased expression levels of them leads to a reduction of FOM susceptibility [18,29,39]. We found that loss of UhpT (G6P transporter) activity is more dominant than that of GlpT (G3P transporter) in strains with decreased FOM susceptibility. In resistant (>128 g/mL) and intermediate (128 g/mL) strains, we noted amino acid deletion in the respective encoded proteins and no PCR amplification of the transporter-related genes (Table 3). In strains with FOM MICs between 8 and 64 g/mL, gene mutation(s) leading to alternations of amino acid residues were found; however, it was unclear whether these contributed to the reduced susceptibility. Among the ten strains with reduced susceptibility, nine strains did not grow on G6Psupplemented M9 minimum salt agar. This suggested that in these nine strains G6P-induced UhpT function was attenuated. We found several mutations in cyaA and ptsI in some of these strains. Dysfunction of CyaA and PtsI leads to decrease in intracellular concentration of cAMP and insufficient expression of GlpT and UhpT [16,17,31]. Only two strains (SRE91 and SRE252) with reduced susceptibility to FOM showed that MIC was decreased by the exogenous addition of cAMP, and these strains did not grow in M9 medium supplement with G3P. Certainly, SRE252 defected cyaA; however, SRE91 did not share any mutations in cyaA and pstI. Thus reduced susceptibility to FOM in the other strains was not explained by the insufficient intracellular concentration of cAMP caused by dysfunction of cyaA and pstI. The mutations in these genes found in this study (Table 3) seemed to scarcely relate with dysfunction of cAMP synthesis.
In conclusion, FOM resistance occurs with low frequency and is independent of resistance to other antimicrobials in E. coli clinical isolates from Japan. On the other hand, we identified strains with decreased FOM susceptibility. Most of them displayed fluctuated activity of the G6P transporter UhpT. However, G6P transporter function was altered even though the G6P-induced uhpT expression and amino acid sequence of UhpT were preserved in one strain (FOM MIC 8 g/mL). Another strain (FOM MIC 32 g/mL) displayed reduced susceptibility to FOM and no alteration of MIC in the presence and absence of G6P, even though G6Pinduced uhpT expression, amino acid sequence of UhpT, and growth on G6P-supplemented M9 minimum salt agar were preserved. The exact molecular mechanism of the reduced susceptibility of these strains remains unclear and requires further evaluation.