Characterization of a Daptomycin-Nonsusceptible Vancomycin-Intermediate Staphylococcus aureus Strain in a Patient with Endocarditis

Abstract

We analyzed the emergence of daptomycin nonsusceptibility in a patient with persistent vancomycin-intermediate Staphylococcus aureus (VISA) bacteremia. The daptomycin-nonsusceptible VISA's cell wall demonstrated a reduction in muramic acid O-acetylation, a phenotypic parameter not previously reported for VISA; some isolates also contained a single point mutation in the mprF gene.

Despite initial tests that suggest the organism's susceptibility, rare Staphylococcus aureus strains develop the vancomycin-intermediate (VISA) phenotype in patients receiving vancomycin (5). Daptomycin, a relatively new agent for the treatment of antimicrobial-resistant gram-positive infections, is a chemically unrelated lipopeptide whose proposed unique mechanism of bactericidal activity is the disruption of cell membrane integrity, depolarization, and subsequent cell death (3). In a U.S. surveillance program initiated in 2002, only 4 (0.04%) of 9,959 S. aureus clinical strains were found to be daptomycin nonsusceptible (MIC ≥2 μg/ml) (18).

However, the in vivo emergence of daptomycin nonsusceptibility has been rarely observed (4, 7, 24). In addition, VISA strains may be more likely than vancomycin-susceptible strains to demonstrate daptomycin nonsusceptibility. Among a unique set of isolates submitted to the Centers for Disease Control and Prevention, 43 (80%) of 54 S. aureus isolates with a vancomycin MIC of 4 μg/ml and 15 (94%) of 16 isolates with a vancomycin MIC of 8 to 16 μg/ml also exhibited daptomycin nonsusceptibility (15).

In this study, we characterized the evolution first of vancomycin and subsequently of daptomycin nonsusceptibility in a series of 29 S. aureus isolates recovered from one patient with a prosthetic aortic valve who, as a complication of a bacteremic pacemaker infection, developed an endocardial abscess (10). Vancomycin troughs were low early in therapy (3.5 to 6.7 μg/ml during the first 10 days), and the patient's high-grade bacteremic condition persisted for >1 month prior to surgical debridement of the endocardial abscess and valve replacement.

Strains were subjected to pulsed-field gel electrophoresis (PFGE) using conditions and interpretation criteria described previously (9, 22). MICs were determined using the CLSI broth macrodilution method (2, 12). Isolates with elevated vancomycin and daptomycin MICs were evaluated by a reference laboratory (Laboratory Specialists, Westlake, OH) using the CLSI broth microdilution and E-test methods.

Four strains were selected for molecular characterization, including detection of the Panton-Valentine leukocidin gene, the staphylococcal cassette chromosome mec type, and the agr type gene, using primers and conditions described previously (11, 13, 14). Multilocus sequence typing (MLST) analysis was performed according to a standard protocol for analyzing S. aureus (www.mlst.net).

To detect the hetero-VISA (hVISA) phenotype, the four isolates selected were subjected to population analysis which was performed in duplicate for each strain (23). An E-test macro method using a 2.0 McFarland standard, a 48-h incubation, and vancomycin and teicoplanin strips was also performed with 11 selected isolates with vancomycin MICs of 2 to 8 μg/ml by standard broth macrodilution (25).

The four selected strains also were analyzed by high-performance liquid chromatography. Cell walls were prepared by using a Mickle laboratory shaker (Mickle Laboratory Engineering Co., Gomshall, Surrey, United Kingdom) with 0.1-mm glass beads to disrupt the cell wall (1) and were then enzymatically digested and treated with hydrofluoric acid to remove teichoic acids (20). Peak identification was done by comparison with standard samples. O-Acetylation of the muramic acid residues was calculated from the relative abundance of the O-acetylated main monomer versus that of the non-O-acetylated counterpart from samples that had not been treated with hydrofluoric acid to avoid the removal of labile O-acetyl ester groups.

Seven isolates were chosen for sequencing analysis of the mprF, yycFG, and rpoBC genes, which were previously reported to be associated with daptomycin nonsusceptibility (6).

All 29 isolates from the case patient's blood and perivalvular tissue specimens collected over 36 days shared an identical PFGE profile. The evolution of decreasing susceptibilities to vancomycin and daptomycin is shown in Table 1. Before the patient underwent therapy, the methicillin-resistant S. aureus (MRSA) exhibited a vancomycin MIC of 2 μg/ml and a daptomycin MIC of 1 μg/ml. The VISA phenotype first became manifest while the patient was undergoing the second week of vancomycin therapy (vancomycin MIC of 4 μg/ml; daptomycin MIC of 1 μg/ml); while the patient was undergoing the second week of subsequent daptomycin therapy, a nonsusceptible daptomycin MIC also developed (vancomycin MIC of 8 μg/ml; daptomycin MIC of 4 μg/ml).

TABLE 1.

Analysis of 29 Staphylococcus aureus isolates, with indistinguishable PFGE patterns, collected from the case patient

Strain(s)	Source(s) of isolation	Isolation date (mo/day/yr)	MIC (μg/ml) of		Concurrent antibiotic therapies	MIC (μg/ml) analysis repeateda			Additional analyses performedb
			Vancomycin	Daptomycin		Vancomycin	Daptomycin		Macro E-test	PAP	Cell wall analysis	Sequencing and results
							Broth microdilution	E-test
1	Blood	9/28/2005	2	1	Vancomycin, rifampin, gentamicin				Y	Y	Y
12	Blood	10/11/2005	2	1	Vancomycin, rifampin, gentamicin				Y
13	Blood	10/11/2005	4	1	Vancomycin, rifampin, gentamicin				Y
14	Blood	10/12/2005	4	1	Vancomycin, rifampin, gentamicin				Y
15	Blood	10/12/2005	4	1	Vancomycin, rifampin, gentamicin				Y
16	Blood	10/13/2005	4	1	Vancomycin, rifampin, gentamicin				Y
17	Blood	10/13/2005	4	2	Vancomycin, rifampin, gentamicin	4	1	2				Y; no genetic changes found
18	Blood	10/15/2005	4	1	Daptomycin, gentamicin
19	Blood	10/15/2005	4	1	Daptomycin, gentamicin
20	Blood	10/17/2005	4	1	Daptomycin, gentamicin
21	Blood	10/17/2005	4	1	Daptomycin, gentamicin
11	Blood	10/20/2005	2	1	Daptomycin, gentamicin
22	Blood	10/20/2005	4	1	Daptomycin, gentamicin
23	Blood	10/24/2005	4	1	Daptomycin, gentamicin
24	Blood	10/25/2005	4	1	Daptomycin, gentamicin
26	Blood	10/27/2005	4	1	Daptomycin, gentamicin
25	Blood	10/27/2005	8	4	Daptomycin, gentamicin	8	4	4	Y		Y	Y; MprF mutation I420N
28	Blood	10/29/2005	8	4	Linezolid	8	4	6	Y	Y		Y; no genetic changes found
27	Blood	10/29/2005	4	1	Linezolid	4	1	1	Y	Y
29	Blood	10/30/2005	8	4	Linezolid	8	4	4				Y; MprF mutation I420N
10	Annulus vegetation	11/2/2005	8	8	Linezolid, vancomycin, gentamicin	8	4	8	Y	Y	Y	Y; MprF mutation I420N
2	Aortic valve leaflet wing	11/2/2005	2	1	Linezolid, vancomycin, gentamicin				Y
3	Aortic valve leaflet wing	11/2/2005	2	1	Linezolid, vancomycin, gentamicin	2	1	1	Y	Y	Y	Y; no genetic changes found
4, 5, 7, 8, and 9	Valve tissue/abscess	11/2/2005	2	1	Linezolid, vancomycin, gentamicin
6	Subvalvular tissue; sewing ring of the prosthesis	11/2/2005	2	1	Linezolid, vancomycin, gentamicin	2	1	1				Y; no genetic changes found

^aFor repeated MIC analyses, vancomycin analysis was performed using the broth microdilution method and daptomycin analysis was performed using the broth microdilution and E-test methods.

^bPAP, population analysis profile. Y, yes.

The four selected strains (strains 1, 3, 10, and 25) were resistant to penicillin, ampicillin, oxacillin, levofloxacin, and rifampin but remained susceptible to trimethoprim-sulfamethoxazole, tetracycline, linezolid, and dalbavancin. The two vancomycin-susceptible strains had teicoplanin MICs of 1 μg/ml, and the two VISA strains both had teicoplanin MICs of 16 μg/ml. All four selected isolates had a multilocus sequence type 5, a staphylococcal cassette chromosome mec type IV, and an agr group II, with no Panton-Valentine leukocidin gene present. The macro E-test assay and population analysis profile showed no hVISA phenotype.

The muropeptide profile of the original daptomycin-susceptible, vancomycin-susceptible S. aureus isolate revealed a peptidoglycan composition similar to that of a standard MRSA (NCTC 8325-4 derivative carrying a mec gene) strain (data not shown) and to the daptomycin-nonsusceptible VISA strain. However, the degree of cross-linking was reduced in the latter daptomycin-nonsusceptible VISA strains (Table 2). The degree of muramic acid O-acetylation was also significantly reduced from a normal value of 41% in the original to 23 to 31% in the daptomycin-nonsusceptible VISAs.

TABLE 2.

Cell wall analysis: degree of cross-linking and O-acetylation in correlation with antimicrobial MICs

Straina	Vancomycin MIC (μg/ml)	Daptomycin MIC (μg/ml)	Cross-linking (%)	O-acetylation (%)
1	2	1	76	41
3	2	1	77	38
10	8	8	74	31
25	8	4	74	23

^aSee Table 1 for further information on the numbered strains.

A single point (T-to-A) mutation was observed at position 1259 in the mprF gene in two of the three daptomycin-nonsusceptible (MIC of 4 μg/ml) VISA (MIC of 8 μg/ml) strains (strains 25 and 29) and in another (strain 10) daptomycin-nonsusceptible (MIC of 8 μg/ml) VISA (MIC of 8 μg/ml) strain, resulting in an amino acid substitution at position 420 from isoleucine to asparagine (I420N) (Table 1).

Even though the VISA strain was initially daptomycin susceptible, during daptomycin therapy, daptomycin nonsusceptibility developed, with changes in the cell wall and in a gene that impacts membrane biosynthesis. Cui et al. suggest that the thickened cell wall found in VISA strains may directly reduce vancomycin effectiveness and also prevent the transfer of the relatively high-molecular-weight daptomycin to its site of action (3). In our study, cell wall analysis demonstrated reduced muropeptide cross-linking as has been noted in other VISA strains.

The daptomycin-nonsusceptible VISA's cell wall also demonstrated a reduction in muramic acid O-acetylation, a phenotypic characteristic that to our knowledge has not been previously investigated in VISA strains. O-Acetylation affects the susceptibility of the cell wall to peptidoglycan hydrolases, host cell wall lytic systems such as lysozyme, and the overall hydrophobicity and charge distribution of the peptidoglycan, which might be important for the polyanionic daptomycin (8).

The accumulation of mutations may contribute to the stepwise development of daptomycin nonsusceptibility. For the case reported here, three of the four daptomycin-nonsusceptible (MICs of 4 to 8 μg/ml) VISA strains contained a single point mutation in the mprF gene, the gene that encodes lysylphosphatidylglycerol (LPG) synthetase. Other amino acid substitutions in LPG synthetase have been found in laboratory-derived isolates and some, but not all, daptomycin-nonsusceptible clinical isolates (6, 17). LPG synthetase transfers lysine from lysyl tRNA to phosphatidylglycerol, a major lipid component of the S. aureus membrane, to produce LPG. Incorporation of LPG in the membrane leads to an overall positively charged cell membrane (16, 19). Changes in LPG synthetase could potentially lead to alterations in the phospholipid composition or in the net charge of the membrane and subsequently affect the activities of a variety of cell wall or cell membrane agents including daptomycin and vancomycin.

In patients receiving vancomycin therapy, rare strains of S. aureus have the capacity to develop thickened cell walls with altered biochemical properties that likely decreased susceptibility to vancomycin. As illustrated here, daptomycin therapy can select for additional cell wall changes and for mutations in a cell membrane biosynthesis enzyme that perhaps enables S. aureus to also become daptomycin nonsusceptible. While these correlations are suggestive, the underlying drivers of S. aureus antimicrobial resistance mechanisms and the connection between vancomycin and daptomycin nonsusceptibility remain elusive.