Association of Alterations in ParC and GyrA Proteins with Resistance of Clinical Isolates of Enterococcus faecium to Nine Different Fluoroquinolones

Abstract

The parC and gyrA genes of 73 ciprofloxacin-resistant and 6 ciprofloxacin-susceptible Enterococcus faecium clinical isolates were partly sequenced. Alterations in ParC and GyrA, possibly in combination with other resistance mechanisms, severely restricted the in vitro activities of the nine quinolones tested. For all isolates, clinafloxacin and sitafloxacin showed the best activities.

The prevalence of infections due to Enterococcus faecium species has been increasing over the last few years (3, 16, 18). Until now, the available fluoroquinolones have been of limited value in the treatment of enterococcal infections because of their poor efficacy (25) and the emergence of acquired fluoroquinolone resistance (9). However, new drugs, such as clinafloxacin and sitafloxacin, are showing increased activity against enterococci (1, 13, 15, 29, 31).

The purpose of the study described here was to characterize the parC and gyrA genes of 73 ciprofloxacin-resistant and 6 ciprofloxacin-susceptible isolates of E. faecium. In addition, the in vitro activities of nine fluoroquinolones were compared (see Table 1).

TABLE 1.

Amino acid changes within ParC and GyrA in 79 E. faecium isolates and corresponding MICs of 9 fluoroquinolones

Amino acid change(s) in:		Drug	No. of isolates for which the MIC (μg/ml) was:
ParC	GyrA		<0.06	0.12	0.25	0.5	1	2	4	8	16	32	64	128	256	>512
None	None	Ciprofloxacin					6	8	15	8
		Levofloxacin					6	18	8	2	1	2
		Sparfloxacin			5	15	7	3	3	3	1
		Gatifloxacin				2	6	19	6	2	2
		Grepafloxacin				1	1	5	11	13	6
		Trovafloxacin			3	1	14	12	4	3
		Moxifloxacin			2	1	8	20	3	2	1
		Clinafloxacin	1	1	7	18	7	3
		Sitafloxacin	1	2	20	4	3	3	4
Ser-80 → Ile		Ciprofloxacin											1	1	2
		Levofloxacin									2		2
		Sparfloxacin									1	2	1
		Gatifloxacin									2	1	1
		Grepafloxacin									2	1	1
		Trovafloxacin								1	2	1
		Moxifloxacin								1	2	1
		Clinafloxacin						2	2
		Sitafloxacin					2	2
Glu-84 → Lys		Ciprofloxacin												1	1
		Levofloxacin										1	1
		Sparfloxacin											2
		Gatifloxacin										1	1
		Grepafloxacin									1	1
		Trovafloxacin									1	1
		Moxifloxacin									1	1
		Clinafloxacin							1	1
		Sitafloxacin						1	1
Ser-80 → Ile	Ser-83 → Arg	Ciprofloxacin												3	5	4
		Levofloxacin										2	5	3	2
		Sparfloxacin										5	4	3
		Gatifloxacin									2	3	5	2
		Grepafloxacin									2	5	3	2
		Trovafloxacin								1	5	5	1
		Moxifloxacin								1	5	4	2
		Clinafloxacin							2	3	4	3
		Sitafloxacin						2	3	3	4
Ser-80 → Ile	Glu-87 → Leu	Ciprofloxacin												4	4	4
		Levofloxacin									2	4	4	2
		Sparfloxacin								6	3	3
		Gatifloxacin							2	6	2	1	1
		Grepafloxacin							2	4	4	2
		Trovafloxacin							6	4	1	1
		Moxifloxacin							2	6	4
		Clinafloxacin						2	6	2	2
		Sitafloxacin					2	4	4	2
Ser-80 → Arg	Glu-87 → Leu	Ciprofloxacin													2
		Levofloxacin										1	1
		Sparfloxacin									1	1
		Gatifloxacin									1	1
		Grepafloxacin										1	1
		Trovafloxacin									1	1
		Moxifloxacin									1	1
		Clinafloxacin								2
		Sitafloxacin						1	1
Glu-84 → Thr	Ser-83 → Leu	Ciprofloxacin									1	1
		Levofloxacin								1	1
		Sparfloxacin								1	1
		Gatifloxacin							1	1
		Grepafloxacin									1	1
		Trovafloxacin							2
		Moxifloxacin							2
		Moxifloxacin							2
		Clinafloxacin					1	1
		Sitafloxacin					1	1
Glu-84 → Lys	Glu-87 → Gly	Ciprofloxacin													3	3
		Levofloxacin									1	2	3
		Sparfloxacin									3	2	1
		Gatifloxacin									2	3	1
		Grepafloxacin									2	3	1
		Trovafloxacin								3	1	2
		Moxifloxacin								1	3	2
		Clinafloxacin							3	2	1
		Sitafloxacin						3	3
Ser-80 → Ile	Glu-87 → Gly; Ser-97 → Asn	Ciprofloxacin													1	1
		Levofloxacin									1	1
		Sparfloxacin									1	1
		Gatifloxacin									1	1
		Grepafloxacin									1	1
		Trovafloxacin								1	1
		Moxifloxacin								1	1
		Clinafloxacin							1	1
		Sitafloxacin					1	1

The isolates tested originated from 24 European university hospitals participating in the European SENTRY Antimicrobial Surveillance Program (8, 27). Between April 1997 and December 1998, 116 epidemiologically nonrelated E. faecium isolates were collected. Of these, 94 (81%) were intermediate or fully resistant to ciprofloxacin. An epidemiologically representative subset of 73 ciprofloxacin-resistant isolates and 6 ciprofloxacin-susceptible isolates was analyzed.

MICs were measured by using concentrations of antibiotics that ranged from 0.06 to 512 μg/ml and were determined by a broth microdilution method (19).

The gyrA and parC portions of the genomes of the E. faecium isolates, which are homologous to the quinolone resistance-determining region of Escherichia coli (30), were amplified and sequenced as described previously (26). The primers were designed according to the gyrA and parC sequences of E. faecium (EMBL database accession nos. AF060881 and AB017811, respectively). Primers gyrA-A (5′-CGGCGGCACCGTCACCGTCAACAG-3′; nucleotides [nt] 139 to 162), gyrA-C (5′-GAATTGGGTGTGACACCGGATAAAG-3′; nt 579 to 558), parC-A (5′-TTCCCGTGCATTTCGATCAGTACTTC-3′; nt 185 to 204), and parC-C (5′-CGTATGACAAAGGATTCCGTAAATC-3′; nt 573 to 554) were used.

Sequences with no mutations were defined as being identical to the reference EMBL sequences. The alterations found in E. faecium GyrA and ParC proteins are indicated in Table 1.

Eight different single or combined amino acid changes were detected in 42 of the 73 ciprofloxacin-resistant isolates of E. faecium. No amino acid change was detected in either ParC or GyrA of the remaining 31 ciprofloxacin-resistant isolates or the 6 susceptible isolates.

Thirty-two of the ciprofloxacin-resistant E. faecium isolates demonstrated mutations in parC, leading to an amino acid change from Ser-80 to Ile or Arg, and 10 showed a deduced amino acid change from Glu-84 to Lys or Thr. Thirty-six of the ciprofloxacin-resistant isolates showed an amino acid change in GyrA, either from Ser-83 to Arg or Leu (14 isolates) or from Glu-87 to Leu or Gly (22 isolates). Six isolates had amino acid changes in ParC alone, without an additional change in GyrA.

The MICs of each quinolone tested are shown in Table 1 for all isolates with alterations in ParC and/or GyrA. These results demonstrate an association between protein alterations and increased MICs. Indeed, for isolates with no identifiable mutations, MICs were lower than those for isolates with only a single amino acid change in ParC. Finally, for those isolates with one alteration in ParC and at least one alteration in GyrA, the highest MICs observed were, for all quinolones tested, only 1 or 2 dilutions higher than those for isolates with only one ParC alteration.

Since ciprofloxacin is still the most commonly used quinolone, our finding of six isolates with alterations only in ParC, together with the fact that no isolate with just GyrA alterations was found, suggests that topoisomerase IV is the primary target of ciprofloxacin in E. faecium. This is similar to earlier findings for Staphylococcus aureus (4) and Streptococcus pneumoniae (17, 21) and therefore supports the theory that topoisomerase IV is the primary target of ciprofloxacin in most gram-positive bacteria (14). Moreover, our data indicate a limited impact of additional GyrA alterations on ciprofloxacin resistance in E. faecium, in contrast to their proposed importance in S. aureus and S. pneumoniae (4, 7, 14, 26).

It is known for many gram-positive pathogens, including the closely related species Enterococcus faecalis (11), that ciprofloxacin selects for mutants with alterations in ParC before it selects for those with alterations in GyrA. Therefore, it is not surprising that the ciprofloxacin-resistant clinical isolates of E. faecium described here are either parC or parC-gyrA mutants and not gyrA mutants. There is evidence that gyrase is the primary target of some quinolones, such as gatifloxacin, sparfloxacin, and clinafloxacin in Streptococcus pneumoniae (5, 23) and sparfloxacin in Mycoplasma hominis (2). Thus, and on the basis of the fact that we have analyzed only clinical isolates and no in vitro mutants, our results cannot be interpreted as indicating that topoisomerase IV is the primary target of all quinolones in E. faecium.

High-level resistance (MIC of ciprofloxacin, >16 μg/ml) in E. faecalis has been associated with either a single mutation in ParC or combined mutations in ParC and GyrA (11). Concordantly, Korten et al. (12) and Tankovic et al. (28) found alterations in GyrA only in high-level ciprofloxacin-resistant strains. In the present study, we also found that all isolates with alterations in GyrA and/or ParC showed high-level resistance. However, even without any alterations in the ParC and GyrA proteins, 25 of 31 isolates showed either intermediate susceptibility or low-level ciprofloxacin resistance. As noted previously (28), this suggests the contribution of mutations in other genes. Since we have not examined the GyrB and ParE subunits, it cannot be excluded that some isolates have alterations in these proteins (22–24). Furthermore, active efflux of quinolones has been demonstrated in other gram-positive cocci such as S. aureus (10, 20) and S. pneumoniae (6).

On the basis of a breakpoint of >1 μg/ml, 2 of 79 isolates were susceptible to grepafloxacin, 6 were susceptible to ciprofloxacin and levofloxacin, 8 were susceptible to gatifloxacin, 11 were susceptible to moxifloxacin, 18 were susceptible to trovafloxacin, 27 were susceptible to sparfloxacin, 35 were susceptible to clinafloxacin, and 36 were susceptible to sitafloxacin. Moreover, sitafloxacin and clinafloxacin showed the best in vitro activities against all isolates.

These results echo the improved activities, as reported previously against genetically undefined isolates of E. faecium, of sitafloxacin (12, 29) and clinafloxacin (1, 16, 31). However, our study also indicates that alterations in ParC and GyrA severely restrict their in vitro activities. These two new quinolones therefore may be of clinical value only for the treatment of infections caused by E. faecium without alterations in ParC and GyrA proteins.