Profiling daptomycin resistance among diverse methicillin-resistant Staphylococcus aureus lineages in China

ABSTRACT

Daptomycin (DAP) is effective against methicillin-resistant Staphylococcus aureus (MRSA). However, reduced susceptibility to DAP in MRSA may lead to treatment failures. We aim to determine the distribution of DAP minimum inhibitory concentrations (MICs) and DAP heteroresistance (hDAP) among MRSA lineages in China. A total of 472 clinical MRSA isolates collected from 2015 to 2017 in China were examined for DAP susceptibility. All isolates (n = 472) were found to be DAP susceptible, but 35.17% (166/472) of them exhibited a high DAP MIC (MIC >0.5 µg/mL). The high DAP MIC group contained a larger proportion of isolates with a higher vancomycin or teicoplanin MIC (>1.5 µg/mL) than the low DAP MIC group (19.3% vs 7.8%, P < 0.001; 22.3% vs 8.2%, P < 0.001). We compared the clonal complex (CC) distributions and clinical characteristics in MRSA isolates stratified by DAP MIC. CC5 isolates were less susceptible to DAP (MIC₅₀ = 1 µg/mL) than CC59 isolates (MIC₅₀ = 0.5 µg/mL, P < 0.001). Population analysis profiling revealed that 5 of 10 ST5 and ST59 DAP-susceptible MRSA isolates investigated exhibited hDAP. The results also showed that CC5 MRSA with an agrA mutation (I238K) had a higher DAP MIC than those with a wild-type agrA (P < 0.001). The agrA-I238K mutation was found to be associated with agr dysfunction as indicated by the loss of δ-hemolysin production. In addition, agr/psmα defectiveness was associated with hDAP in MRSA. Whole-genome sequencing analysis revealed mutations in mprF and walR/walK in DAP-resistant subpopulations, and most DAP-resistant subpopulations (6/8, 75%) were stable. Our study suggests that the increased DAP resistance and hDAP in MRSA may threaten the effectiveness against MRSA infections.

INTRODUCTION

Daptomycin (DAP) was the first calcium-dependent lipopeptide antibiotic approved for use against Gram-positive bacteria. It is used in patients with soft-tissue infections, bacteremia, and right-sided endocarditis, and is considered a front-line agent for treating methicillin-resistant Staphylococcus aureus (MRSA) infections (1). Although resistance to DAP is uncommon in MRSA, it can lead to treatment failure in more than 20% of cases (2, 3). It is reported that mutations in mprF, dlt, pgsA, cls2, vraSR, and yycFG (also known as walK/walR) are associated with DAP resistance (4 – 6). While DAP exhibits bactericidal activity in vitro, a high minimum inhibitory concentration (MIC) (>0.5 µg/mL) is associated with increased mortality in patients with MRSA bacteremia and is also an independent risk factor for complicated methicillin-sensitive S. aureus bacteremia (7, 8). MICs of >1.5 µg/mL for vancomycin and teicoplanin and >2 µg/mL for linezolid are also associated with treatment failure and increased mortality rates (6 – 10). However, the distribution of DAP MICs against MRSA lineages and the predictive value of DAP susceptibility in China are currently unclear.

Isolates exhibiting heteroresistance are defined as those containing one or more subpopulations with reduced susceptibility to an antibiotic compared with that of the main population (9). Detecting these subpopulations is challenging because of their instability, low frequency, and transient character (10). Population analysis profiling (PAP) is the gold-standard method for identifying heteroresistance. Although heteroresistance has been observed for various anti-staphylococcal agents, including β-lactams and vancomycin, the prevalence of DAP heteroresistance (hDAP) in different MRSA lineages has yet to be widely examined (11).

The expression of the α-, β-, and δ-hemolysins in S. aureus is controlled by a quorum-sensing system encoded by the accessory gene regulator (agr) locus (12). The δ-hemolysin is encoded by hld, which is co-transcribed with RNAIII and acts as the effector molecule of the Agr system (13). In addition, the regulation of phenol-soluble modulin (PSM) expression is mediated by the Agr system (14). Dysfunction of the Agr system is associated with persistent bacteremia and increased mortality rates (15 – 17). Moreover, it is correlated with vancomycin intermediate S. aureus (VISA) and reducing DAP susceptibility (18, 19). However, the relationship between hDAP and Agr function is unclear.

In this study, we analyzed the DAP susceptibility of 472 well-typed MRSA isolates based on their genomic profiles from our previous national study (20) to determine the distribution of MICs, the predictive value of DAP susceptibility, and the mechanism of hDAP in these MRSA isolates.

RESULTS

In vitro DAP susceptibility and clonality

The DAP MIC for the MRSA isolates ranged from 0.25 to 1 µg/mL, and the MIC₅₀ and MIC₉₀ were at 0.5 and 1 µg/mL, respectively (Table 1). All MRSA isolates were susceptible to DAP with MICs of ≤0.5 and 1 µg/mL in 64.83% (306/472) and 35.17% (166/472) of isolates.

TABLE 1.

Clonality of DAP MIC distribution and the MIC range, MIC₅₀, and MIC₉₀ in MRSA isolates

CC	N	%	DAP MIC (µg/mL), N						P value a	MIC range	MIC50	MIC90
			0.25	%	0.5	%	1	%
CC59	154	32.63%	3	1.95%	108	70.13%	43	27.92%		0.25–1	0.5	1
CC239	117	24.79%	1	0.85%	79	67.52%	37	31.62%	0.438	0.25–1	0.5	1
CC5	97	20.55%	0	0.00%	45	46.39%	52	53.61%	<0.001	0.5–1	1	1
CC1	24	5.08%	0	0.00%	15	62.50%	9	37.50%	0.29	0.5–1	0.5	1
CC22	5	1.06%	0	0.00%	3	60.00%	2	40.00%	0.528	0.5–1	0.5	1
CC30	3	0.64%	0	0.00%	2	66.67%	1	33.33%	0.803	0.5–1	0.5	1
CC398	13	2.75%	5	38.46%	8	61.54%	0	0.00%	<0.001	0.25–0.5	0.5	0.5
CC45	25	5.30%	0	0.00%	18	72.00%	7	28.00%	0.881	0.5–1	0.5	1
CC7	5	1.06%	0	0.00%	3	60.00%	2	40.00%	0.528	0.5–1	0.5	1
CC88	9	1.91%	0	0.00%	4	44.44%	5	55.56%	0.075	0.5–1	1	1
Other	20	4.24%	0	0.00%	12	60.00%	8	40.00%	0.231	0.5–1	0.5	1
Total	472	100.00%	9	1.91%	297	62.92%	166	35.17%		0.25–1	0.5	1

^a P ≤ 0.05 was considered as significant compared to CC59.

According to our previous study, the top three clonal complexes (CCs) to which the MRSA isolates belonged were CC59, CC239, and CC5, representing 32.63% (154/472), 24.97% (117/472), and 20.55% (97/472) of the isolates, respectively (20). Nearly one-third (27.92%, 43/154) of CC59 isolates exhibited a DAP MIC of 1 µg/mL, and 72.08% (111/154) exhibited an MIC of ≤0.5 µg/mL. Similarly, about one-third (31.62%, 37/117) of CC239 isolates exhibited a DAP MIC of 1 µg/mL, and 68.38% (80/117) showed an MIC of ≤0.5 µg/mL. Over half of the CC5 isolates (53.61%, 52/97) exhibited a DAP MIC of 1 µg/mL and 46.39% (45/97) showed an MIC of ≤0.5 µg/mL. Notably, 100% (13/13) of the CC398 isolates exhibited an MIC of ≤0.5 µg/mL. Compared with CC59 isolates (MIC₅₀ = 0.5 µg/mL, MIC₉₀ = 1 µg/mL), CC5 isolates (MIC₅₀ = 1 µg/mL, MIC₉₀ = 1 µg/mL) were less susceptible to DAP (P < 0.001), whereas CC398 isolates (MIC₅₀ = 0.5 µg/mL, MIC₉₀ = 0.5 µg/mL) were more susceptible to DAP (P < 0.001).

Clinical characteristics of MRSA isolates with higher DAP MICs

Recent studies showed that MRSA bacteremia with DAP MIC >0.5 µg/mL is associated with increased mortality rates (7, 8). Therefore, we categorized the MRSA isolates in this study into the high DAP MIC group (>0.5 µg/mL) and low DAP MIC group (≤ 0.5 µg/mL). The demographic and clinical characteristics of the 472 patients are summarized in Table S1. In the univariate analysis, the proportion of MRSA isolates with vancomycin MIC (>1.5 µg/mL) or teicoplanin MIC (>1.5 µg/mL) was higher in the high DAP MIC group than in the low DAP MIC group (19.3% vs 7.8%, P < 0.001; 22.3% vs 8.2%, P < 0.001, respectively). Furthermore, in the Pearson correlation analysis, the DAP MIC showed slight positive correlated with the MICs of vancomycin (r = 0.147, P = 0.001) and teicoplanin (r = 0.264, P < 0.001) (Fig. S1). While we found statistically significant associations between DAP MIC and resistance to vancomycin and teicoplanin, the r values were small, indicating that the level of resistance to these antibiotics can only be weakly predicted by the DAP MIC. In addition, the rifampicin resistance rate of MRSA isolates was lower in the high DAP MIC group than that in the low DAP MIC group (19.9% vs 29.4%; P = 0.024).

Additionally, daptomycin-susceptible (DAP-S) MRSA rates varied in different regions in China (Fig. S2). Our data showed that a considerable proportion MRSA with higher DAP MIC for MRSA was identified in East China. Of the 166 MRSA with a DAP MIC at 1 µg/mL, 48.80% (81/166) were isolated in East China with 75.31% (61/81) from Anhui and Fujian provinces (Table S2), while there were only 19.28% (32/166) isolated in Southwest China (19.28%, 32/166).

The distributions of DAP MICs differed among MRSA isolates from different regions, demographic characteristics, and underlying medical conditions and infections, possibly because of the distinct predominant MRSA lineages among these groups.

Some DAP-susceptible ST5 and ST59 MRSA exhibit hDAP

To characterize the prevalence of hDAP in DAP-S MRSA isolates, we randomly selected 15 isolates from different hospitals representing dominant lineages in China (ST5, ST239, and ST59) and tested them for hDAP using PAP. The results showed that 5 of the 15 isolates exhibited hDAP with the frequencies of resistant subpopulations ranging from 1.83 × 10⁻⁶ to 7.69 × 10⁻⁶ (Fig. 1; Table 2).

Fig 1.

Population analysis profiling for methicillin-resistant Staphylococcus aureus isolates. (A) Five ST5 MRSA isolates; (B) five ST59 MRSA isolates; (C) five ST239 MRSA isolates. Different colors represent distinct strains. Solid and dashed lines represent heteroresistance and homogeneous responses to DAP, respectively. The bracket represents the sequence type (ST) and DAP minimum inhibitory concentration, and the gray dashed line indicates the detection limit of this experiment. The error bars represent the standard deviation of four independent cultures of each isolate.

TABLE 2.

Molecular characteristics and δ-hemolysin of clinical MRSA isolates

Strains	ST	SCCmec	Parental strain DAP MIC (µg/mL)	Frequency of heteroresistant subpopulations	δ-Hemolysin
N02HSA03	5	II	0.5	7.69 × 10−6	+
N08HSA26			1	3.85 × 10−6	−
N17HSA08			0.5	4.66 × 10−6	+
N03HSA01			1	<1 × 10−7	−
N10HSA03			0.5	<1 × 10−7	+
N02CSA21	59	IV	0.5	1.83 × 10−6	+
N05CSA02			0.5	7.51 × 10−6	+
N04CSA12			1	<1 × 10−7	+
N26CSA01			1	<1 × 10−7	+
N28CSA03			0.5	<1 × 10−7	+
N25CSA38	239	III	0.5	<1 × 10−7	−
N02HSA09			0.5	<1 × 10−7	−
N05HSA05			0.5	<1 × 10−7	−
N09HSA07			1	<1 × 10−7	−
N12HSA12			1	<1 × 10−7	−

Three of the five ST5 MRSA isolates (N02HSA03, N08HSA26, and N17HSA08) tested exhibited hDAP (Fig. 1A). The N08HSA26 subpopulation grew at 2 µg/mL, which was eightfold higher than the highest non-inhibitory concentration (0.25 µg/mL) (frequency of resistance: 3.85 × 10⁻⁶). For ST59 MRSA, two of the five isolates (N02CSA21 and N05CSA02) tested exhibited hDAP (Fig. 1B). N02CSA21 subpopulations survived at 1 µg/mL, which was 16-fold higher than the highest non-inhibitory concentration. No hDAP was detected among the ST239 isolates tested (Fig. 1C). ST239 isolates with an MIC of 1 µg/mL exhibited a PAP curve that was right shifted by one- or twofold, compared with those of the ST239 isolates with an MIC of 0.5 µg/mL. Thus, hDAP could be common among these ST5 and ST59 MRSA isolates in China.

Agr dysfunction in CC5 MRSA isolates

Agr system plays an important role in MRSA virulence (21), and loss of the agr system enhances S. aureus survival during DAP treatment (18). To identify whether decreased DAP susceptibility in clinical CC5 MRSA isolates is associated with agr mutation, we aligned the sequences of the agrA locus using that of strain N315 as a reference (Table 3). We found that 28.87% (28/97) of CC5 MRSA isolates possessed an I238K substitution in their agrA, and nearly all of them (92.86%, 26/28) had a higher DAP MIC at 1 µg/mL.

TABLE 3.

agrA mutations of DAP MIC distribution in clinical CC5 MRSA isolates (n = 97)

agrA mutation a	DAP MIC (µg/mL)				P value b
	0.5	(%)	1	(%)
WT	40	60.61%	26	39.39%
I238K	2	7.14%	26	92.86%	<0.001
I238fs	1	100.00%	0	0.00%	1
R218 c	1	100.00%	0	0.00%	1
K136R	1	100.00%	0	0.00%	1

^a amino acid change; I, isoleucine; K, lysine; R, arginine; fs, frameshift;

^b P ≤ 0.05 was considered significant compared to WT.

^c Stop; WT, wild-type agrA compared to agrA in N315 (NC_002745.2).

To further investigate whether mutation in agrA affects Agr function, δ-hemolytic activity was analyzed for 20 randomly selected ST5 MRSA isolates (10 with wild-type agrA and 10 with agrA-I238K) (Fig. S3). All isolates with the mutation showed no δ-hemolytic activity, suggesting a loss of agr function (Fig. S3B). Wild-type agrA strains produced various zones of hemolysins, indicating agr competency (Fig. S3A).

Agr dysfunction confers hDAP to MRSA

To investigate the role of agr in hDAP development, agr-deleted and psmα-deleted MRSA strains HL1 were evaluated using PAP. At DAP concentrations of 0.125 or 0.25 µg/mL, both the wild-type and mutant strains showed significantly decreased growth (Fig. 2A). The result showed that the Δagr and Δpsmα strains displayed substantially right-shifted population curves and significantly increased resistance frequencies compared with those of wild-type strains. The HL1-Δagr and HL1-Δpsmα MRSA strains were associated with the highest inhibitory concentrations of 0.5 and 1 µg/mL, respectively, which were eightfold higher than the highest non-inhibitory concentrations (0.0625 and 0.125 µg/mL, respectively). The frequencies of resistant subpopulations in HL1-Δagr and HL1-Δpsmα were 5.40 × 10⁻⁶ and 4.00 × 10⁻⁶, respectively. These findings showed that agr or psmα deficiency contributed to hDAP in MRSA.

Fig 2.

Population analysis profiling for methicillin-resistant Staphylococcus aureus isolates with or without PSMα peptides. (A) HL1 (ST72), Δagr mutant, and Δpsmα mutant MRSA isolates; (B) three ST5 hDAP MRSA isolates; (C) two ST59 hDAP MRSA isolates. The bracket represents the sequence type (ST) and DAP minimum inhibitory concentration. Solid and dashed lines represent heteroresistance and homogeneous responses to DAP, respectively. The hollow label reflects the presence of mixtures of PSMα peptides (α1–4). Genomic analyses comparing strains HL1-∆agr and HL1-Δpsmα with HL1 are presented in Table S3.

PSMα peptides restore DAP susceptibility to resistant subpopulations

Furthermore, a mixture of synthetic PSM peptides was supplemented to DAP to determine whether PSM exposure influenced hDAP development. The PAP curves of the Δagr and Δpsmα strains were left shifted one- or twofold in the presence of PSMα peptides (Fig. 2A). Similarly, all heteroresistant strains reverted to a homogeneous phenotype in the presence of PSMα peptides, suggesting that PSMα peptides can restore DAP susceptibility to resistant subpopulations (Fig. 2B and C). These revealed that agr dysfunction contributed to the hDAP phenotype in MRSA isolates, and DAP displayed a stronger bactericidal effect when combined with PSMα peptides.

Mutations identified in DAP-resistant subpopulations

To further investigate the mechanisms involved in hDAP, the DAP MIC of heteroresistant subpopulations revealed by the PAP assay was tested, and mutations were screened using whole-genome sequencing (WGS; Table 4). The subpopulations exhibited greater DAP resistance than their parental isolates, with MICs of 8 and 4 µg/mL for N08HSA26 and N17HSA08, respectively. In addition, a mutation in walK was identified with different amino acid substitutions (A567V in N08HSA26 and T357I in N17HSA08). In N08HSA26, resistant subpopulations isolated from the same strain displayed different MICs and mutations (walK and walR in subpopulations 1 and 2, respectively), indicating that subpopulations with distinct mutations and phenotypes can co-exist in one isolate. Subpopulations from N02HSA03, N02CSA21, and N05CSA02 exhibited reduced DAP susceptibility, with MICs of 4, 2, and 4 µg/mL, respectively. These resistant subpopulations exhibited mutations in mprF that caused various amino acid changes (1337–1369 delΔ33 bp in N02HSA03, P314L in N02CSA21, and L826F and R798P in N05CSA02). Additionally, purR contained point mutations in N02CSA21 (subpopulation 2). No mutations were found in any other genes.

TABLE 4.

Stability of antibiotic resistance and mutations of hDAP MRSA isolates ^a

Isolates	Subpopulation no.	Resistant subpopulations MIC (µg/mL)	Gene b	Specific function	Nucleotide/amino acid change c	MIC after 50 generations (µg/mL) d	Gene change/nucleotide change e	Additonal gene (nucleotide/amino acid change)
N02HSA03	1	4	mprF	Phosphatidylglycerol lysyltransferase	Δ33 bp/coding (1,337–1,369/2,523 nt)	4	mprF/Δ33 bp	–
N08HSA26	1	8	walK	Sensor protein kinase WalK	C1700T/A567V	8	walK/C1700T	–
	2	4	walR	Transcriptional regulatory protein WalR	C242T/A81V	4	walR/C242T	–
N17HSA08	1	4	walK	Sensor protein kinase WalK	C1070T/T357I	4	walK/C1070T	–
N02CSA21	1	2	mprF	Phosphatidylglycerol lysyltransferase	G941A/P314L	2	mprF/G941A	–
	2	1	mprF	Phosphatidylglycerol lysyltransferase	G941A/P314L	1	mprF/G941A	–
			purR	Pur operon repressor	(T)6→7/coding (91/825 nt)		purR/(T)6→7	–
N05CSA02	1	4	mprF	Phosphatidylglycerol lysyltransferase	G2476A/L826F	0.5	WT	noc (C796T/H266Y)
	2	4	mprF	Phosphatidylglycerol lysyltransferase	C2393G/R798P	0.5	WT	apt (G218T/G73V)
HL1-∆agr	1	2	mprF	Phosphatidylglycerol lysyltransferase	C1010T/S337L	/	/	/
	2	2	mprF	Phosphatidylglycerol lysyltransferase	C1010T/S337L	/	/	/
HL1-Δpsmα	1	0.5	–	–	–	/	/	/
	2	1	mprF	Phosphatidylglycerol lysyltransferase	A1279T/I427F	/	/	/

^a “–” indicates that no mutation was observed in the respective sample; “/” means the results could not be obtained or are not applicable.

^b Gene name derived from Prokka annotation.

^c nt, nucleotide; C, cystine; A, alanine; V, valine; T, threonine; I, isoleucine; P, proline; L, leucine; G, glycine; F, phenylalanine; R, arginine; H, histidine; Y, tyrosine.

^d Resistant subpopulations MIC after 50 generations in antibiotic-free medium (µg/mL).

^e WT means no mprF mutation.

We also identified mprF mutations in the subpopulations of both HL1-∆agr and HL1-Δpsmα strains. Specifically, the mprF mutation observed in the HL1-∆agr subpopulation was S337L, while the HL1-Δpsmα subpopulation exhibited an I427F mutation in mprF (see Table 4). Our data demonstrated that the DAP resistance in subpopulations was due to the mutations in mprF and walR/walK.

Stability and genetic alterations of DAP-resistant subpopulations

The subpopulations were serially passaged without antibiotic pressure for 50 generations. Four out of five cases retained their MIC values. One case (N05CSA02) exhibited instability; in this case, the MIC reverted to that of the parental isolate (from 4 to 0.5 µg/mL; Table 3).

Genetic alterations were identified between stable and unstable subpopulations in all eight samples through whole-genome sequencing (Table 4). The amino acid substitutions were preserved in all strains except for in unstable subpopulations (N05CSA02), in which the mprF mutation was replaced with point mutations in noc (H266Y) and apt (G73V).

DISCUSSION

In this study, we investigated DAP susceptibility and DAP heteroresistance in clinical MRSA isolates collected from Chinese hospital patients in a nationwide surveillance during 2015–2017. We found that although all MRSA isolates were susceptible to DAP, a significant number of them from different clonal lineages had decreased DAP susceptibility. This is alarming as it has been reported that increased resistance for DAP (MIC >0.5 µg/mL) in MRSA leads to higher mortality rates and poor clinical outcomes (7, 8, 22).

We observed that DAP MIC showed a marginal positive correlation with the MICs of teicoplanin and vancomycin, a trend that aligns with some previous reports (8, 23). A similar association was reported by Hsieh et al., where a weak correlation between vancomycin MICs and daptomycin MICs was identified (r = 0.26; P < 0.001) (23). However, it is essential to bear in mind that these correlations, while statistically significant, are relatively weak. As such, these findings should be cautiously interpreted. Given these subtle and complex relationships, further clinical studies are needed to more fully explore and understand the implications of a DAP MIC >0.5 µg/mL in patients infected with MRSA.

agr-defective clones, including ST5 and ST239 MRSA, are prevalent among hospital-associated MRSA worldwide (24, 25). This study identified an I238K amino acid substitution in AgrA in 28.87% (28/97) of the CC5 MRSA. This mutation led to agr dysfunction in these isolates as none of them had δ-hemolytic activity. We found that agr functional deficiency may also contribute to the decreased susceptibility to DAP in specific MRSA lineages such as CC5. Importantly, we had also established the correlation between agr deficiency and hDAP phenotype in MRSA using Δagr and Δpsmα strains. Compared to the wild-type strains, these deletion mutants displayed substantially right-shifted population curves and significant increases in their resistance frequencies.

In our study, the role of Agr system dysfunction in mediating hDAP was not uniformly evident across clinical isolates. This lack of consistency may be attributable to the genetic heterogeneity of the investigated strains. Previous research has demonstrated that distinct STs and agr types display divergent patterns of virulence gene expression, thereby implicating that the influence of the Agr system on hDAP may be contingent upon the specific genetic background of each strain (26). Our observations align with the existing literature, underscoring that the underlying factors contributing to hDAP are still not fully elucidated and may involve a complex interplay of multiple variables.

hDAP in MRSA can lead to treatment failure (22). A previous study found that 2 of 27 clinical ST5/ST105-SCCmecII MRSA isolates exhibited hDAP and suggested that the prevalence of hDAP was likely underestimated (27). In this work, we speculated that hDAP could be common in ST5 MRSA strains as agr dysfunction in these strains was common. Although the relationship between hDAP and agr function has not been previously reported, the heterogeneous VISA (hVISA) phenotype was found to be strongly associated with compromised agr function (28 – 31). agr dysfunction was observed in 58% of hVISA strains, indicating that it is advantageous to clinical MRSA isolate (28). In addition, agr dysfunction is associated with persistent bacteremia and thrombin-induced platelet microbicidal protein (31, 32), suggesting that hDAP MRSA isolates with agr dysfunction are associated with poor clinical outcomes.

The mechanisms underlying hDAP in MRSA remain understudied. Previous studies found that rpoB(A621E)-mediated dual heteroresistance to DAP and vancomycin is accompanied by a thickened bacterial cell wall and reduced negative cell surface charge (33), and vancomycin exposure may induce hDAP through cell wall changes (34). Here, we observed mprF and walR/walK mutations in DAP-resistant subpopulations of heteroresistant MRSA isolates. These mutations were associated with DAP resistance (35, 36).

This study had a few limitations. First, the small number of isolates that have undergone PAP assay limits the generalizability of the PAP results. The prevalence of hDAP should be studied in larger sets of isolates. Additionally, the methods for the clinical detection of heteroresistance should be improved, as PAP is both time- and labor intensive.

In conclusion, we performed a surveillance study of DAP susceptibility and heteroresistance among MRSA isolates in China. Although all MRSA investigated were susceptible to DAP, a large number of them had a higher DAP MIC and a reduced vancomycin and teicoplanin susceptibility. We further demonstrated that hDAP is mediated by multiple factors, including dysfunctional agr and subpopulations carrying mutated DAP resistance-associated genes. However, further studies are needed to evaluate the relationship between DAP MIC and clinical outcomes and the possibility of employing whole-genome sequencing to inform antibiotic use during treatment for MRSA infections with DAP.

MATERIALS AND METHODS

Media and culture conditions

For standard culture procedures, S. aureus was cultured at 37°C using Mueller-Hinton agar (MHA; Oxoid) or Columbia blood agar (Oxoid), or in tryptone soy broth (Oxoid).

Bacterial isolates

A total of 472 non-duplicate clinical MRSA isolates were collected from 22 tertiary hospitals in 18 provinces of China between 2015 and 2017 (20). Strain HL1 is a clinically isolated community-associated MRSA strain. The ∆agr and ∆psmα strains of HL1 have been described previously (37).

Antimicrobial susceptibility testing

The susceptibility of the isolates to DAP was determined using micro-dilutions in Mueller-Hinton broth (MHB; Oxoid), based on guidelines provided by the Clinical and Laboratory Standards Institute (38). For DAP (LY146032, MedChemExpress, NJ, USA), cultures in MHB were supplemented with 50 µg/mL Ca²⁺. Concurrently, S. aureus ATCC 29213 was tested as a quality control organism. Isolates with a DAP susceptibility breakpoint of ≤1 µg/mL were categorized as susceptible, while those with >1 µg/mL were categorized as resistant. MIC₅₀ refers to an antimicrobial agent’s minimum inhibitory concentration that inhibits the growth of 50% of bacterial strains. In comparison, MIC₉₀ refers to the MIC that inhibits the growth of 90% of bacterial strains in a given population.

Assessment of δ-hemolytic activity

To assess the δ-hemolytic activity of S. aureus, we performed the CAMP (Christie-Atkins-Munch-Petersen) test using a Columbia sheep blood agar plate. First, a pure culture of the S. aureus test strain and RN4220 control strain was obtained, and a sheep blood agar plate was prepared by streaking the RN4220 control strain in a straight line down the center of the plate. Next, using a sterile cotton swab, a heavy growth of the S. aureus test strain was obtained and streaked perpendicular to the RN4220 control strain, starting at the edge of the plate and extending across the control strain. The plate was then incubated at 37°C for 18–24 hours. A positive CAMP test was indicated by the presence of a triangular zone of enhanced hemolysis formed between the two strains, indicating the presence of δ-hemolysin produced by S. aureus in response to the β-hemolysin produced by RN4220. Conversely, a negative result was indicated by the absence of a triangular zone of enhanced hemolysis, as described previously (39, 40).

PAP and PSM peptide assays

PAP assays were performed as previously described (22, 41, 42), with some modifications. Bacterial strains were obtained and used to initiate four independent overnight cultures in Mueller-Hinton broth, with each culture diluted 1:1,000 in PBS buffer. After overnight incubation and serial dilutions, 5 µL culture suspensions (10⁷–10⁸ colony-forming units) were spread onto plain MHA plates or MHA plates supplemented with a twofold dilution of DAP ranging from 0.0315 to 16 µg/mL (in the presence of 50 µg/mL Ca²⁺). Colonies were counted after 48 hours of incubation at 37°C, and the frequency of bacterial growth associated with each concentration was determined (42). We defined heteroresistance as the frequency of subpopulation greater than 1 × 10⁻⁷ at an MIC at least eightfold higher than that of the main population (10).

The mixture of synthetic PSM peptides (10 µM) was supplemented with DAP and directly incorporated into the agar to determine the influence of PSM exposure on the hDAP strains. Synthetic PSM peptides were generated by Sangon Biotech Ltd. (Shanghai, China) at ≥95% purity, as previously described (14, 18).

DAP-resistant subpopulation stability

To investigate the stability of daptomycin-resistant subpopulations, clones were collected and incubated in MHB without antibiotic exposure for 50 generations, as previously described (42). Briefly, the cultures were grown for five additional overnight periods (10 generations per day) in the absence of antibiotics, with 1 µL of inoculum into 1 mL of MHB. Aliquots were collected from each culture, pelleted, and stored at −80°C, or subjected to DNA extraction and WGS analysis. To estimate the stability of DAP-resistant subpopulations, their minimum inhibitory concentrations for DAP were compared with those of the original parental clones. If the DAP MIC decreased or reverted to that of the original parental isolate, the resistance was considered unstable.

Whole-genome sequencing and mutation screening

The eight resistant subpopulations of five heteroresistant isolates were analyzed using an Illumina HiSeq X ten platform in 2 × 150-bp paired-end mode (San Diego, CA, USA), as described previously (22, 43). Illumina sequence reads were trimmed and assembled with Shovill v0.9.0 with default settings (https://github.com/tseemann/shovill). The genomes were annotated using Prokka version 1.11 (44), and mutations were detected using the BRESEQ version 0.34.1 pipeline (22, 45), which were subsequently confirmed by Sanger sequencing.

Statistical analysis

PAP curves were constructed using GraphPad Prism 7.0 software (GraphPad, Inc., La Jolla, CA, USA). The Wilcoxon rank-sum test, χ ² test, and Fisher’s exact test were used to compare categorical variables. To assess the strength of the relationship between the MICs of DAP and those of other antibiotics, the Pearson correlation coefficient (r) was calculated using SPSS software (version 21.0, SPSS, Inc., Chicago, IL, USA). Statistical significance was set at P ≤ 0.05.

DATA AVAILABILITY

The sequence data utilized in this study are publicly accessible at the National Genomics Data Center under the BioProject no: PRJCA004763.

ETHICS APPROVAL

This study was approved by the local ethics committees of Sir Run Run Shaw Hospital with a waiver of informed consent (Approval No.20190426-2).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aac.00563-23.

Supplemental file 1. aac.00563-23-s0001.pdf.

Tables S1 to S3 and Fig. S1 to S3

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.