Contribution of Topoisomerase IV Mutation to Quinolone Resistance in Mycoplasma genitalium

Abstract

The mechanism of quinolone resistance in Mycoplasma genitalium remains poorly understood due to difficulties with in vitro culture, especially of clinical isolates. In this study, to confirm the association between mutations in topoisomerases and antimicrobial susceptibilities to quinolones, ciprofloxacin-resistant mutant strains were selected using the cultivable type strain ATCC 33530. Sequence analysis revealed that the mutant strains harbored mutations in topoisomerase IV: Gly81Cys in ParC, Pro261Thr in ParC, or Asn466Lys in ParE. The MICs of all quinolones tested against the mutant strains were 2- to 16-fold higher than those against the wild-type strain. No cross-resistance was observed with macrolides or tetracyclines. We determined the inhibitory activities of quinolones against DNA gyrase and topoisomerase IV in order to investigate the correlation between antimicrobial susceptibility and inhibitory activity against the target enzymes, considered the primary targets of quinolones. Furthermore, using enzymatic analysis, we confirmed that Gly81Cys in the ParC quinolone resistance-determining region (QRDR) contributed to quinolone resistance. This is the first study to isolate quinolone-resistant mutant strains of M. genitalium harboring substitutions in the parC or parE gene in vitro and to measure the inhibitory activities against the purified topoisomerases of M. genitalium.

INTRODUCTION

Mycoplasma genitalium was first isolated in urethral cultures from men with nongonococcal urethritis (NGU) in 1981 (1). M. genitalium is an important cause of NGU in men and has been shown to be associated with cervicitis, endometritis, salpingitis, and pelvic inflammatory diseases in women (2–5).

M. genitalium has been reported to be susceptible to macrolides and is highly susceptible to azithromycin. The current guidelines recommend an azithromycin regimen for the treatment of NGU (6, 7). On the other hand, macrolide-resistant clinical strains of M. genitalium have emerged due to selection by the azithromycin regimen, and these clinical isolates with elevated azithromycin MICs have been shown to harbor an A2058G or A2059G substitution in the 23S rRNA gene (8).

Some quinolones, such as moxifloxacin, have potent activity against M. genitalium, and treatment with moxifloxacin is considered an effective second-line treatment for persistent or recurrent NGU (9). The antibacterial activities of quinolones are due to their inhibitory activities against type II topoisomerases, DNA gyrase (composed of two GyrA and two GyrB subunits), and topoisomerase IV (composed of two ParC and two ParE subunits). It has been reported that mutations in the quinolone resistance-determining regions (QRDR) of the genes encoding DNA gyrase and/or topoisomerase IV contribute to quinolone resistance in various bacterial species, including other mycoplasmas (10–13). We recently detected the amino acid substitution Ser83Asn, Asp87Tyr, or Asp87Val in the ParC QRDR of M. genitalium DNAs in urine specimens from men with NGU (14–16).

Due to the difficulties of isolating M. genitalium strains from clinical specimens, there have been only a few reports regarding the antimicrobial susceptibility of a small number of clinical isolates. Moreover, there have been no reports regarding the association between mutations in topoisomerases and quinolone resistance.

The inhibitory activities against DNA gyrase in Mycoplasma pneumoniae have been determined recently (17). However, there have been few reports regarding the inhibitory activities of quinolones against mycoplasmas. To elucidate the association between mutations of the target enzyme and susceptibility to quinolones, we selected quinolone-resistant mutant strains using the type strain ATCC 33530, which was available for culture in vitro using SP-4 medium. Furthermore, we determined the inhibitory activities of quinolones against DNA gyrase and topoisomerase IV in order to investigate the correlation between antimicrobial susceptibility and inhibitory activity against the target enzymes, which are considered the primary targets of quinolones. We also determined the inhibitory activities against mutant topoisomerase IV in order to investigate whether the mutation was involved in quinolone resistance.

MATERIALS AND METHODS

Antibacterial agents and bacterial strains.

The antibacterial agents tested were synthesized by Kyorin Pharmaceutical Co., Ltd. (Tokyo, Japan), or were purchased from commercial sources. M. genitalium ATCC 33530 (G37) was obtained from the American Type Culture Collection (ATCC) (Manassas, VA). M. genitalium was grown at 37°C in SP-4 medium.

MIC determination.

MICs were determined by the broth dilution method with SP-4 medium. The MICs were defined as the lowest concentration of antimicrobial agents without color change.

Selection of resistant mutant strains in vitro. (i) Selection with quinolones at subinhibitory concentrations.

Multistep resistance selection was performed using SP-4 medium with increasing 2-fold dilutions of quinolones as described previously (13). When a color change in the control wells without an antimicrobial agent was observed, the MICs were determined, and the cultures in subinhibitory concentrations were used for subsequent passages.

(ii) Selection with a quinolone at a MIC above the MIC for the parent strain.

Quinolone-resistant mutant strains were selected by serial transfers in SP-4 medium containing ciprofloxacin at four times the MIC. Cultures with visible growth were used for passages. After some passages in SP-4 medium containing ciprofloxacin, subcultures in an antibiotic-free medium were performed.

PCR amplifications and DNA sequencing.

Genotypic analysis of resistant mutant strains was performed by standard methods with amplification and DNA sequencing of the DNA gyrase (gyrA and gyrB) and topoisomerase IV (parC and parE) genes. The PCR primers for amplification of DNA gyrase (gyrA and gyrB) were Mg gyrBA_F (5′-TTATGGTTCCTTGTTCCAAACG-3′) and Mg gyrBA_R (5′-GTTAGCAAGTTCAAGTTGTTGC-3′), and those for amplification of topoisomerase IV (parC and parE) were Mg parEC_F (5′-CTCTGATCCTTACCATGGATCTG-3′) and Mg parEC_R (5′-TAGTAGGACCAACCAGAATGGAC-3′). These primers were designed based on the published sequence of M. genitalium ATCC 33530.

Cloning.

The gyrA, gyrB, parC, and parE genes of M. genitalium were amplified from M. genitalium ATCC 33530 by PCR with the following oligonucleotide primers (restriction sites are underlined): for gyrA, 5′-ATTCCCGGGTATGGCAAAGCAACAAGATCAAGTAGATAA-3′ (SmaI) and 5′-ATTCCCCGGGTTATTGCGTAATTTGTTTGGATCCAACATC-3′ (SmaI); for gyrB, 5′-CCGGCCGGAATTCATGGAAGAAAATAACAAAGCAAATATC-3′ (EcoRI) and 5′-GCCGCTCGAGTTAAATATCAATGTTTTTAACACTACGAGC-3′ (XhoI); for parC, 5′-CGCGGATCCATGGATCAAAAAAACAACAACCTCTTTC-3′ (BamHI) and 5′-CCGCTCGAGCTAATTAAGTTTGTTAAACCTGGTTTGSC-3′ (XhoI); and for parE, 5′-ATCCGCGGATCCATGAAAAGTAACTACAGTGCAACTAAC-3′ (BamHI) and 5′-GCCCGCTCGAGTTAGTTTTCCACACTAAAGTTAATGTTG-3′ (XhoI). PCR products were cloned into pGEX-6P-1 and were transformed into Escherichia coli DH5α.

In most mycoplasmas, the universal stop codon TGA encodes tryptophan (18, 19). Mycoplasma-specific TGA (tryptophan) codons were replaced with the universal TGG (tryptophan) codons in cloned genes (five codons in gyrB, two codons in parC, and one codon in parE) by site-directed mutagenesis using the DpnI method so that the full length of the gene could be expressed in E. coli.

To construct mutated ParC G81C, the mutation was introduced in the wild-type parC plasmid by site-directed mutagenesis with the following oligonucleotide primers (mutation sites are underlined): 5′-CCACCCCCATTGTGATAGTTCCATTTATG-3′ and 5′-CATAAATGGAACTATCACAATGGGGGTGG-3′.

All constructs were confirmed by DNA sequencing.

Preparation of DNA gyrase and topoisomerase IV.

For expression, plasmids were transformed into E. coli BL21 competent cells. The GyrA, GyrB, ParC, ParE, and mutated ParC G81C proteins were purified separately as fusion proteins with glutathione S-transferase from overproducing strains of E. coli as described previously (20, 21) with minor modifications. The proteins were eluted with PreScission protease (GE Healthcare, Waukesha, WI). Neither GyrA nor GyrB alone had DNA supercoiling activity, and neither ParC nor ParE alone had decatenation activity.

Topoisomerase assays.

The activities of DNA gyrase and topoisomerase IV were determined with recombinant forms of M. genitalium enzymes reconstituted by incubation of each A and B subunit of the enzymes (GyrA-GyrB or ParC-ParE) on ice. The DNA supercoiling activity of DNA gyrase was assayed by monitoring the conversion of relaxed pBR322 to its supercoiled form. The decatenating activity of topoisomerase IV was assayed by monitoring the conversion of kinetoplast DNA (TopoGEN, Inc., Port Orange, FL) to the monomer. The reaction mixtures were electrophoresed in 1.0% agarose gels. DNA in agarose gels was quantified with an FMBIO II Multi View fluorescent image analyzer (Hitachi Software Engineering Co., Ltd., Yokohama, Japan) after ethidium bromide staining. The inhibitory activities of quinolones against topoisomerases were assayed by determining the concentrations required to inhibit 50% of the enzyme reaction (IC₅₀).

(i) Supercoiling activity of DNA gyrase.

Reaction mixtures (10 μl) containing 40 mM Tris-HCl (pH 7.5), 5 mM MgCl₂, 25 mM KCl, 10 mM dithiothreitol (DTT), 2 mM spermidine HCl, 50 μg of bovine serum albumin/ml, 10 μg of tRNA/ml, 2 mM ATP, 1 U of DNA gyrase, 50 ng of relaxed pBR322 DNA, and various concentrations of the quinolones tested were incubated at 37°C for 1 h. The reaction was terminated by the addition of 2.5 μg of proteinase K. After an additional 10 min of incubation at 37°C, a loading dye was added. One unit of enzyme activity was defined as the amount of DNA gyrase that converts 50 ng of relaxed pBR322 to the supercoiled form under the conditions described above.

(ii) Decatenating activity of topoisomerase IV.

Reaction mixtures (10 μl) containing 40 mM Tris-HCl (pH 7.5), 10 mM MgCl₂, 10 mM DTT, 100 mM potassium glutamate, 50 μg of bovine serum albumin/ml, 2 mM ATP, 1 U of topoisomerase IV, 100 ng of kinetoplast DNA, and various concentrations of the quinolones tested were incubated at 37°C for 1 h. The reaction was terminated by the addition of 2.5 μg of proteinase K. After an additional 10 min of incubation at 37°C, a loading dye was added. One unit of enzyme activity was defined as the amount of topoisomerase IV that converts 100 ng of kinetoplast DNA to the monomer under the conditions described above.

RESULTS

Selection of ciprofloxacin-resistant M. genitalium ATCC 33530 mutant strains.

First, we attempted to select quinolone-resistant mutant strains of M. genitalium ATCC 33530 by serial passages in SP-4 medium containing subinhibitory concentrations of some quinolones, as described previously for other mycoplasmas (13). However, we failed to isolate mutant strains with increased MICs, even after 10 passages.

It has been reported that mutation in the QRDR of gyrA or parC increases the MICs of several quinolones, including ciprofloxacin, against M. pneumoniae more than 4-fold (13). Therefore, we selected mutant strains with ciprofloxacin at four times the MIC rather than at a subinhibitory concentration in order to isolate stable resistant strains. The cultures containing ciprofloxacin in seven test tubes showed visible growth after 26 to 37 days of incubation. The cultures were then diluted with fresh medium containing ciprofloxacin at four times the MIC against the wild-type strain. After successive passages, the cultures were transferred to an antibiotic-free medium.

The susceptibilities of the mutant strains, which were selected with ciprofloxacin, to quinolones, as well as to macrolides and tetracyclines, are summarized in Table 1. The MICs of all quinolones tested against the ciprofloxacin-selected mutant strains were 2- to 16-fold higher than those against the wild-type strain ATCC 33530. Quinolone resistance remained stable in the mutant strains after subculture in an antibiotic-free medium. No differences in MICs were observed between wild-type and mutant strains for other classes of antibiotics, such as clarithromycin, azithromycin, tetracycline, and doxycycline.

Table 1.

Susceptibilities of ciprofloxacin-selected mutant strains to various antimicrobial agents

Antimicrobial agent	MICa for parent strain ATCC 33530	MICa (ratio to MIC for the wild type) for the following ciprofloxacin-selected mutant strainb:
		CP-1	CP-2	CP-3
Ciprofloxacin	2	32 (16)	16 (8)	16 (8)
Garenoxacin	0.063	0.5 (8)	0.5 (8)	0.5 (8)
Moxifloxacin	0.125	0.5 (4)	0.5 (4)	0.5 (4)
Gatifloxacin	0.25	1 (4)	1 (4)	1 (4)
Sparfloxacin	0.25	2 (8)	2 (8)	1 (4)
Tosufloxacin	0.25	2 (8)	1 (4)	1 (4)
Levofloxacin	0.5	8 (16)	4 (8)	4 (8)
Fleroxacin	4	16 (4)	16 (4)	16 (4)
Norfloxacin	32	128 (4)	128 (4)	128 (4)
Clinafloxacin	0.125	0.25 (2)	0.25 (2)	0.25 (2)
Nadifloxacin	0.5	1 (2)	1 (2)	1 (2)
Pazufloxacin	16	32 (2)	32 (2)	64 (4)
Clarithromycin	0.004	0.004 (1)	0.004 (1)	0.004 (1)
Azithromycin	0.001	0.001 (1)	0.001 (1)	0.001 (1)
Tetracycline	0.125	0.125 (1)	0.125 (1)	0.125 (1)
Doxycycline	0.063	0.063 (1)	0.063 (1)	0.063 (1)

^aGiven in micrograms per milliliter.

^bCP-1 has Gly81Cys in ParC; CP-2 has Asn466Lys in ParE; CP-3 has Pro261Thr in ParC. None of the strains has any mutations in GyrA or GyrB.

Sequence analysis of the quinolone-resistant mutant strains.

We considered the possibility that multiple mutations conferred quinolone resistance on mutant strains selected with ciprofloxacin. The entire coding sequences of each of the four topoisomerase subunits (gyrA, gyrB, parC, and parE), which are known quinolone targets, from the wild-type and mutant strains were determined. The sequence analysis revealed that the quinolone-resistant mutant strain CP-1 harbored a Gly81-to-Cys substitution in ParC, which was within the QRDR; no mutations in other subunits (GyrA, GyrB, and ParE) were observed. The quinolone-resistant mutant strains CP-2 and CP-3 harbored an Asn466-to-Lys substitution in ParE and a Pro261-to-Thr substitution in ParC, respectively. The other four strains harbored one of the substitutions in topoisomerase IV observed in these three strains, CP-1, CP-2, and CP-3. Since the mutations were found only in parC or parE among quinolone target topoisomerases, it was suggested that the mutation in topoisomerase IV contributes to quinolone resistance.

Inhibitory activities of quinolones against topoisomerases of M. genitalium.

To determine the effects of the quinolone activities against the target enzymes exclusively, the IC₅₀ of various quinolones were determined using purified recombinant M. genitalium DNA gyrase and topoisomerase IV. The quinolones showed dose-dependent inhibition, with IC₅₀ of 21.0 to >1,000 μg/ml against DNA gyrase and 11.1 to 43.5 μg/ml against topoisomerase IV (Table 2). The IC₅₀ for DNA gyrase showed greater differences among the quinolones than those for topoisomerase IV. We determined the correlations between the inhibitory activities of quinolones against M. genitalium DNA gyrase or topoisomerase IV and antimicrobial susceptibilities. No significant correlations were found between the IC₅₀ for DNA gyrase and MICs or between the IC₅₀ for topoisomerase IV and MICs (correlation, 0.449 or 0.330, respectively) (Fig. 1).

Table 2.

Inhibitory activities of quinolones against DNA gyrase and topoisomerase IV

Quinolone	IC50a for:		IC50a for mutant topoisomerase IVb (fold increase over the IC50 for the wild type)
	DNA gyrase (wild type)	Topoisomerase IV (wild type)
Ciprofloxacin	>1,000	16.3	200 (12)
Garenoxacin	25.9	11.1	117 (11)
Moxifloxacin	32.4	11.6	126 (11)
Gatifloxacin	21.0	7.95	96.3 (12)
Sparfloxacin	589	11.6	109 (9.4)
Tosufloxacin	400	43.5	601 (14)
Levofloxacin	157	20.3	295 (15)
Norfloxacin	757	37.6	344 (9.1)

^aIn micrograms per milliliter.

^bMutant topoisomerase IV consisted of mutant ParC Gly81Cys and wild-type ParE.

Fig 1.

Correlation between inhibitory activities and MICs. (A) MIC versus IC₅₀ for DNA gyrase (correlation, 0.449); (B) MIC versus IC₅₀ for topoisomerase IV (correlation, 0.330).

To confirm the contribution of mutation in the parC gene to quinolone resistance with purified enzymes, a mutant form of M. genitalium, with ParC Gly81Cys, was constructed by site-directed mutagenesis of the parC gene. The mutated ParC Gly81Cys was expressed in E. coli, purified, and combined with the purified wild-type ParE subunit to produce mutant topoisomerase IV. The IC₅₀ of all quinolones were increased against the mutant topoisomerase IV; the increases ranged from 9.1-fold to 15-fold (Table 2).

DISCUSSION

It has been determined that mutations in the target proteins contributed to antimicrobial resistance in various bacterial species. Recently, we detected amino acid substitutions in ParC in M. genitalium DNAs in urine specimens from men with NGU (15, 16). We also reported an amino acid substitution in ParC in urine specimens of a patient with recurrent NGU after levofloxacin treatment (14). These mutations in ParC were located within the QRDR, which is related to quinolone resistance in various bacterial species, including other mycoplasmas.

To confirm the association between mutations in topoisomerases and susceptibility to quinolones in M. genitalium, we selected quinolone-resistant mutant strains from the cultivable type strain ATCC 33530 in vitro because of the difficulties in analyzing clinical isolates. Quinolone-resistant mutant strains were not isolated even after 10 serial passages in the presence of subinhibitory concentrations of some quinolones. These results suggested the possibility that the resistant subpopulation represented a small proportion of the cultures at subinhibitory concentrations. In contrast, quinolone-resistant mutant strains were selected from cultures containing ciprofloxacin at four times the MIC. The susceptible strains were able to multiply at subinhibitory concentrations, whereas the quinolone-resistant mutant strains were likely to grow at quinolone concentrations well above the MIC for the wild-type strain. The method used in this study is useful for selecting antibiotic-resistant mutant strains of M. genitalium.

The quinolone-resistant mutant strains selected with ciprofloxacin harbored a mutation in ParC (Gly81Cys or Pro261Thr) or ParE (Asn466Lys). These mutations have not yet been reported in clinical isolates or laboratory-derived strains of M. genitalium. However, mutation at ParC position 81 (position 78 for E. coli) has been described previously for E. coli, Salmonella enterica, Staphylococcus aureus, and M. pneumoniae (13, 22–27). For E. coli and Salmonella, mutation at ParC 78 has been reported in both clinical isolates and laboratory-derived quinolone-resistant mutant strains (22–26). The mutation at ParE position 466 (position 458 for E. coli) was described previously in S. aureus (corresponding to position 470) (28). Since mutation at this site was responsible for quinolone resistance in S. aureus, mutation at ParE 466 in M. genitalium seems to contribute to quinolone resistance. The other mutation, at ParC position 261 (position 258 for E. coli), was located outside the QRDR and has not been described previously for other bacterial species. Antibacterial activity is influenced by various factors, such as influx and efflux. We are aware of the potential limitations of our study, including the lack of analysis of other mechanisms of quinolone resistance. CP-3, which harbored a mutation at ParC 261, showed no cross-resistance with macrolides or tetracyclines, and the MICs of quinolones for CP-3 were approximately equivalent to those for CP-1 and CP-2. Despite the limitations of the present study, the observations indicate that topoisomerase IV mutation at ParC 261 would contribute to quinolone resistance. Quinolone resistance mutations outside the QRDR have been found in various bacterial species. However, the prevalence of these mutations is unclear, because such regions are rarely analyzed. In M. genitalium, further evaluation will confirm the clinical significance of the mutants selected with ciprofloxacin.

Mutation at ParC position 81, which is located within the QRDR, has been reported in other mycoplasmas (13). The importance of this mutation in clinical settings was suggested; therefore, the contribution of the mutation at ParC 81 in M. genitalium to quinolone resistance was examined with purified enzymes. The MICs of all quinolones tested against the mutant with Gly81Cys in ParC were 2- to 16-fold higher than those against the wild-type strain. Furthermore, the inhibitory activities of the quinolones were decreased because of the same substitution in ParC in the mutant strain. These results suggest that the mutation at ParC 81 contributes to quinolone resistance.

In this study, topoisomerase IV mutant strains were selected with ciprofloxacin. Moreover, the susceptibilities of these strains to ciprofloxacin and the inhibitory activities of ciprofloxacin against the mutant topoisomerase IV were decreased. These results suggested that the primary target of ciprofloxacin is topoisomerase IV.

A significant correlation between the inhibitory activities of quinolones against the primary target and MICs has been reported, e.g., for E. coli DNA gyrase and S. aureus topoisomerase IV (correlations, 0.972 and 0.926, respectively) (29, 30). In this study of M. genitalium, however, the correlation between the IC₅₀ for DNA gyrase and MICs or between the IC₅₀ for topoisomerase IV and MICs was markedly lower (0.449 or 0.330, respectively) than those reported previously for other bacteria. These results raise the possibility that the primary target in M. genitalium differs for different quinolones. It has been reported previously that the primary target differs for different quinolones in various bacterial species, including other mycoplasmas. In Mycoplasma hominis, DNA gyrase is the primary target for sparfloxacin, whereas topoisomerase IV is the primary target for ciprofloxacin, ofloxacin, pefloxacin, and trovafloxacin (31–33). Furthermore, some quinolones have been reported to have different primary targets in different mycoplasmas. For example, the primary target for sparfloxacin is DNA gyrase in M. hominis but topoisomerase IV in M. pneumoniae (13, 31, 32). These reports support our hypothesis. It was suggested that the correlation between IC₅₀ and MICs would increase if one used quinolones with the same primary target in M. genitalium.

It has been reported that the rates of eradication of M. genitalium from NGU patients treated with levofloxacin or ofloxacin tend to be inferior to those obtained with gatifloxacin or moxifloxacin (34). In this study, gatifloxacin and moxifloxacin were shown to exhibit more-potent inhibitory activities than levofloxacin against all topoisomerases tested, including mutant topoisomerase IV. Although the primary target of quinolones in M. genitalium is still unclear, a quinolone with potent inhibitory activities against all topoisomerases is expected to be effective for the treatment of M. genitalium infection. Our enzymatic data revealed one of the reasons for the clinical efficacy of some quinolones.

The emergence of macrolide-resistant clinical strains of M. genitalium has been reported recently, and persistent M. genitalium-positive NGU following treatment failure with azithromycin regimens has been shown to be cured with moxifloxacin (8, 9, 35). The increased usage of quinolones as alternatives to azithromycin is likely one of the key factors in the emergence of quinolone-resistant M. genitalium. Although moxifloxacin showed high activity against the wild type, its inhibitory activity against the mutant topoisomerase IV was decreased, and the mutations in topoisomerase IV increased the MICs 4-fold. Inappropriate treatment with quinolones with moderate activity against M. genitalium and the application of suboptimal dosage regimens of quinolones could increase the risk of emergence of resistance to quinolones, including moxifloxacin.

This is the first study to isolate in vitro quinolone-resistant mutant strains of M. genitalium harboring substitutions in the parC or parE gene and to measure the inhibitory activities of quinolones against the purified topoisomerases of M. genitalium. The present findings provide information useful for understanding the mechanism of quinolone resistance in M. genitalium and for choosing promising quinolones for the treatment of M. genitalium infections, including infections with topoisomerase IV mutant strains.