Distribution of Fusidic Acid Resistance Determinants in Methicillin-Resistant Staphylococcus aureus

Abstract

The prevalence of resistance to fusidic acid in clinical isolates of Staphylococcus aureus, including methicillin-resistant S. aureus (MRSA), has increased in the past 2 decades. However, there are limited data regarding the relative importance in this process of the different staphylococcal determinants that mediate resistance to fusidic acid. Furthermore, the roles played by clonal dissemination of fusidic acid-resistant strains versus horizontal transmission of fusidic acid resistance determinants have not been investigated in detail. To gain insight into both issues, we examined fusidic acid resistance in 1,639 MRSA isolates collected in Denmark between 2003 and 2005. Resistance to fusidic acid (MIC, >1 μg/ml) was exhibited by 291 (17.6%) isolates. For the majority of these isolates (∼87%), resistance was attributed to carriage of fusB or fusC, while the remainder harbored mutations in the gene (fusA) encoding the drug target (EF-G). The CC80-MRSA-IV clone carrying fusB accounted for ∼61% of the resistant isolates in this collection, while a single CC5 clone harboring fusC represented ∼12% of the resistant strains. These findings emphasize the importance of clonal dissemination of fusidic acid resistance within European MRSA strains. Nonetheless, the distribution of fusB and fusC across several genetic lineages, and their presence on multiple genetic elements, indicates that horizontal transmission of fusidic acid resistance genes has also played an important role in the increasing prevalence of fusidic acid resistance in MRSA.

Fusidic acid prevents bacterial protein synthesis by inhibiting the ribosomal translocase elongation factor G (EF-G) (2, 26). This antibiotic has been used since the 1960s for both topical and systemic treatment of staphylococcal disease. The prevalence of resistance to fusidic acid in Staphylococcus aureus remained low for decades after its clinical introduction, with rates of 1 to 2% reported at the start of the 1990s from countries employing fusidic acid extensively, such as the United Kingdom and Denmark (8, 27). More recently, a dramatic increase in fusidic acid resistance has been observed in several clinical settings in Europe. This has been particularly striking among strains of S. aureus causing impetigo, a phenomenon closely tied to the clonal expansion of a highly successful fusidic acid-resistant strain designated the epidemic European fusidic acid-resistant impetigo clone (EEFIC; sequence type 123 [ST123], clonal complex 121 [CC121]) (21). However, increasing rates of fusidic acid resistance have also been observed in S. aureus strains causing other skin and soft tissue infections. Most of these infections are due to methicillin-susceptible S. aureus (MSSA), but methicillin-resistant S. aureus (MRSA) isolates resistant to fusidic acid have also been reported increasingly throughout Europe and in the Maghreb (1, 13, 16, 25, 28). This spread of fusidic acid-resistant MRSA has been attributed in part to clonal expansion of a fusidic acid-resistant community-acquired (CA) clone, CC80-MRSA-IV (9, 13, 19, 25).

Resistance to fusidic acid in clinical isolates of S. aureus occurs through two mechanisms, designated fusA- and fusB-type resistance (22, 24). The former involves spontaneous mutation in the gene (fusA) encoding EF-G, leading to modification of the drug target and reduced susceptibility (18). The latter involves the recruitment of an exogenous resistance determinant whose product binds EF-G and protects it from fusidic acid (23). Although fusB was initially thought to be the only determinant encoding a protein capable of protecting EF-G, it now appears that fusB-type resistance is mediated by a family of related FusB-like proteins (22). Thus, in addition to fusB, two genes (fusC and fusD) encoding representatives of this family occur in staphylococci (4, 22).

Determinants encoding fusB-type resistance are borne on several genetic elements. The fusB gene was first identified on plasmid pUB101 (23) and also resides on a closely related plasmid (pUB102) in the CC80-MRSA-IV clone (15). In the EEFIC, fusB is present on a pathogenicity island-like element, designated SaRI_fusB (S. aureus resistance island harboring fusB) (21). The fusC gene has been found only in association with staphylococcal chromosomal cassettes: SCC₄₇₆, an element that lacks mecA (11), and the mecA-containing cassette SCC_mecN1 (6).

The determinants of fusB-type fusidic acid resistance, and the genetic elements on which they reside, have only recently been characterized (22, 23). Thus, until now it has not been possible to examine the relative prevalences of the different fusidic acid resistance mechanisms in S. aureus, the genetic elements that carry them, and the relative importance for increasing fusidic acid resistance of clonal expansion of successful fusidic acid-resistant clones versus horizontal dissemination of fusidic acid resistance determinants. In the present study, we sought to investigate these matters by performing a molecular genetic analysis of fusidic acid resistance in a large and defined collection of Danish MRSA isolates.

MATERIALS AND METHODS

Bacterial strains and their characterization.

All MRSA isolates (n = 1,639) collected nationwide in Denmark from 2003 to 2005 were included in this study. Isolates were defined as MRSA on the basis of cefoxitin resistance and mecA carriage (13) and were tested for susceptibility to fusidic acid by a disc diffusion assay. Antibiotic susceptibility testing by agar dilution was carried out as described previously (13). Isolates were subjected to pulsed-field gel electrophoresis (PFGE) analysis according to the Harmony protocol (17), and profiles were analyzed with BioNumerics software, version 4.6 (Applied Maths, Sint-Martens-Latem, Belgium), using a similarity coefficient of ≥80% to define a clonal complex (CC). Typing by spa analysis used established protocols (10), and types were assigned using Ridom Staphtype (version 1.4.11). Multilocus sequence typing (MLST) was performed as described previously (7), and sequence types (STs) were assigned at http://www.mlst.net.

Detection and genetic location of fusidic acid resistance determinants.

Staphylococcal DNA purification, PCR-based detection of fusB and fusC, and PCR amplification and DNA sequencing of fusA were performed using previously established methods (14, 21, 22). The genetic locations of the fusB and fusC genes were determined by PCR using oligonucleotide primers specific for these genes and/or primers hybridizing to regions flanking the various genetic elements associated with these determinants (Table 1). For fusB, these primers aimed to detect SaRI_fusB and plasmids pUB101 and pUB102. For fusC, primers were designed to probe for SCC₄₇₆ and SCC_mecN1. The specificity of primers for the amplification of their intended targets was established through use of both positive and negative controls (DNA from strains harboring or lacking resistance-associated genetic elements, respectively).

TABLE 1.

Oligonucleotide primers used in this study

Determinant and location	Target gene	Primer sequence (5′ to 3′)	Expected amplicon size (kb)
fusB in SaRIfusB	SAS1815	TAAAAAAATCTACTCAAAAC	2.7
	fusB	TAAATGATAAAGAAACCGTC
fusB in pUB101 or pUB102	blaZ	TAGATACTAAAAGTGGTAAGG	3.4 (pUB101), 2.6 (pUB102)
	fusB	CTATAATGATATTAATGAGATTTTTGG
fusC in SCC476	SAS0040	AGGTGGATTGATGGGAGTTAAG	1.3
	fusC	ATTATTTATATCATCTAGGTTCTG
fusC in SCC476	SAS0045	AGTTCTATACTGAAGGTTATGG	2.6
	fusC	TTAAAGAAAAAGATATTGATATCTCGG
fusC in SCCmecN1	Upstream of fusC within SCCmecN1	CTAATATGTTGGCGCTGATAT	10.1 or 3.7a
	fusC	ACAAACGATATGAATTCCCA
fusC (SCCmecN1)b	ccrAB4-1CHE482-ccrAB4-2CHE482	CAAATGATTGAAACAGAGGT	0.8
	ccrAB4-1CHE482-ccrAB4-2CHE482	CACGTTTTCTACAATAACGT

^aDepending on the SCC_mecN1 variant (6).

^bThe primers do not detect fusC directly but probe for the fusC-associated element SCC_mecN1 (6).

RESULTS AND DISCUSSION

Prevalence of fusidic acid resistance.

Screening of all 1,639 isolates for fusidic acid resistance by the disc diffusion assay identified 313 resistant isolates (19.1%). Of these, 291 (17.8%) were confirmed to be resistant by agar dilution, according to the EUCAST fusidic acid resistance breakpoint (MIC, >1 μg/ml). By use of PFGE, augmented by spa and MLST analysis of representative isolates within each PFGE cluster, the fusidic acid-resistant isolates were assigned to 11 clonal complexes (Fig. 1). More than half (178 [61%]) of the resistant isolates belonged to CC80. Since a detailed analysis of Danish CC80 isolates has been reported recently (13), only six of the CC80 isolates in this collection were selected for further evaluation. In total, 119 isolates were selected for analysis of their fusidic acid resistance determinants.

FIG. 1.

Distribution of fusidic acid resistance determinants by clonal complex. Isolates designated “other” comprise those in which more than one resistance determinant was identified, as well as those that carried none of the known fusidic acid resistance determinants (see the text).

Nature of fusidic acid resistance.

PCR was used to probe resistant isolates for the presence of the fusB and fusC determinants. Where neither gene could be detected, or where the level of fusidic acid resistance was higher than that usually associated with these determinants (>16 μg/ml) (23, 24), PCR amplification and DNA sequencing of fusA were performed. The fusB determinant was detected in 24 isolates, including all 6 representatives of CC80. However, based on previous work, it is highly likely that all, or nearly all, of the untested CC80 isolates (n = 172) also carry fusB (13, 16), in which case 196 (67.4%) of the fusidic acid-resistant isolates in this collection were fusB positive. The fusB gene was detected in isolates from 9 of the 11 CCs and was the only fusidic acid resistance determinant found in CC45 and CC97 (Fig. 1). The fusC gene was detected in 57 isolates (19.6%) across six CCs and was the only fusidic acid resistance determinant associated with CC59 and CC78 (Fig. 1).

Overall, genes for FusB-type resistance (fusB or fusC) were present in ∼87% of resistant isolates. In contrast, resistance mutations in fusA were detected in only 39 (13.4%) isolates across six CC groups (Table 2). These results suggest that, as for other staphylococcal isolates (3, 14), horizontally acquired resistance determinants make the major contribution to fusidic acid resistance in MRSA, and mutational resistance is of only limited importance.

TABLE 2.

Summary of fusidic acid-resistant isolates harboring polymorphisms in EF-G

EF-G polymorphism(s)	CC	No. of isolates	Fusidic acid MIC (μg/ml)
V90I	5	1	1-8
	8	15
	80	1
V90I, L461S	8	2	16-64
	121	1
V90I, E233Q, L461S	8	1	16
D373N, H457Y	8	1	64
P404L	5	1	8
P404Q	22	1	2
L461K	5	2	128
	8	7
L461S	22	1	2
H457Y	5	1	32-128
	22	1
	30	1
H457Q	8	1	4
T656K	22	1	4

Three isolates harboring genes encoding FusB-type resistance (two with fusB, one with fusC) also carried mutations in fusA (encoding EF-G substitutions H₄₅₇Y, H₄₅₇Y/D₃₇₃N, and L₄₆₁K, respectively). We have shown previously that in laboratory strains of S. aureus containing both fusA resistance mutations and fusB, these different resistance mechanisms do not act synergistically or additively (23). Instead, the level of fusidic acid resistance observed for a strain carrying both determinants corresponds to that encoded by the determinant conferring the highest level of resistance (23). This phenomenon was also seen in these clinical isolates: all three exhibited a level of fusidic acid resistance that could be accounted for solely by their fusA genotypes.

Two isolates belonging to CC22 for which fusidic acid displayed MICs of 2 μg/ml did not harbor fusB, fusC, or resistance mutations in fusA. They were subsequently found by PCR to be negative for fusD as well (22). Since disruption of ribosomal protein L6 confers resistance to fusidic acid in laboratory-generated mutants of S. aureus (20), we subjected the gene encoding L6 (rplF) to PCR amplification and sequencing. However, in both cases, this gene was found to be intact and wild type in sequence. Consequently, the genetic basis for fusidic resistance in these two strains remains undefined, suggesting the existence in S. aureus of one or more additional determinants of resistance.

Analysis of genetic elements harboring fusidic acid resistance determinants.

To examine the prevalence and spread of genetic elements containing fusB and fusC, all isolates carrying fusB or fusC were probed for the known resistance-associated elements. Of 57 isolates carrying fusC, none yielded a PCR amplicon corresponding to fusC in either SCC₄₇₆ or SCCmec_N1. Further PCR analysis of these isolates using oligonucleotide primers specific for SCCmec_N1 revealed that four (two from CC5 and one each from CC8 and CC78) contained this or a closely related element, although this finding does not necessarily imply linkage of this element to fusC in these strains. The location of fusC in the other 53 isolates remained completely undefined, suggesting that in these strains, fusC is associated with genetic elements other than those previously reported. Of 24 isolates carrying fusB, 17 carried one of the known resistance-associated genetic elements (SaRI_fusB, pUB101, or pUB102), with pUB102 predominating, while in 7 isolates, none of these elements could be detected (Table 3).

TABLE 3.

PCR analysis of the genetic locations of fusB in isolates belonging to different clonal complexes

CC	No. of isolates tested	No. of isolates carrying:
		pUB101	pUB102	SaRIfusB	Othera
1	2		2
5	2		2
8	7		3		4
22	1				1
30	3	1			2
45	2	2
80	5		5
97	1		1
121	1			1

^aIsolates for which none of the genetic elements known to harbor fusB could be detected.

Clonal expansion versus horizontal transmission of fusidic acid resistance.

The prevalence of fusidic acid resistance can increase by the spread of successful clones harboring fusB (12, 13, 21). The present study also provides evidence for clonal expansion of strains carrying fusC: this resistance determinant was strongly associated with strains from CC5 (34 of the 57 isolates carrying fusC [60%]) and, to a lesser extent, from CC8 (15 of the 57 isolates carrying fusC [25%]) (Fig. 1). Indeed, the CC5 clone carrying fusC accounts for ∼12% of all fusidic acid-resistant MRSA isolates in this collection. Whether there has also been clonal dissemination of fusidic acid resistance mediated by fusA mutations is not clear, since there is no straightforward way to distinguish independent mutational events from clonal spread. Nonetheless, clonal expansion is a plausible explanation for the existence in this collection of isolates of multiple representatives of CC8 harboring the same mutations in fusA (Table 1).

Our results also point to horizontal transmission of fusidic acid resistance determinants, with fusB and/or fusC detected in strains of all lineages tested (Fig. 1). Indeed, both determinants occurred in four genetic lineages (CC1, CC5, CC8, and CC22). In the case of fusB, horizontal transfer appears to be associated primarily with plasmid pUB102 (Table 3), which was found in 5 different CCs. In contrast, the other known genetic elements that carry fusB were found almost exclusively in the lineage with which they were first associated: plasmid pUB101 in CC45 and SaRI_fusB in the EEFIC (ST123, CC121) (21) (Table 3). Two CCs (CC8 and CC30) included strains carrying fusB on different genetic elements (Table 3), a finding that is also indicative of horizontal dissemination of fusidic acid resistance genes.

Concluding remarks.

The occurrence of resistance to fusidic acid in MRSA strains in Denmark is considerable: fusidic acid resistance was detected in approximately 18% of the strains examined in this study. It is clear that clonal expansion of a single successful clone carrying fusB (CC80-MRSA-IV) has made a major contribution to the prevalence of fusidic acid resistance among MRSA isolates in Europe (13). However, the distribution of the fusB and fusC determinants within a variety of defined and uncharacterized genetic elements across several genetic lineages indicates that multiple independent acquisitions of fusidic acid resistance genes have also played an important role in increasing the prevalence of fusidic acid resistance in MRSA strains.