Phenotypic and Genomic Profiling of Staphylococcus argenteus in Canada and the United States and Recommendations for Clinical Result Reporting

ABSTRACT

Staphylococcus argenteus is a newly described species, formerly known as S. aureus clonal complex 75 (CC75). Here, we describe the largest collection of S. argenteus isolates in North America, highlighting identification challenges. We present phenotypic and genomic characteristics and provide recommendations for clinical reporting. Between 2017 and 2019, 22 isolates of S. argenteus were received at 2 large reference laboratories for identification. Identification with routine methods (biochemical, matrix-assisted laser desorption ionization–time of flight mass spectrometry [MALDI-TOF MS], 16S rRNA gene analysis) proved challenging to confidently distinguish these isolates from S. aureus. Whole-genome sequencing analysis was employed to confirm identifications. Using several different sequence-based analyses, all clinical isolates under investigation were confirmed to be S. argenteus with clear differentiation from S. aureus. Seven of 22 isolates were recovered from sterile sites, 11 from nonsterile sites, and 4 from surveillance screens. While sequence types ST1223/coa type XV, ST2198/coa type XIV, and ST2793/coa type XId were identified among the Canadian isolates, the majority of isolates (73%) belonged to multilocus sequence types (MLST) ST2250/coa type XId and exhibited a high degree of homology at the genomic level. Despite this similarity, 5 spa types were identified among ST2250 isolates, demonstrating some diversity between strains. Several isolates carried mecA, as well as other resistance and virulence determinants (e.g., PVL, TSST-1) commonly associated with S. aureus. Based on our findings, the growing body of literature on S. argenteus, the potential severity of infections, and possible confusion associated with reporting, including use of incorrect breakpoints for susceptibility results, we make recommendations for clinical laboratories regarding this organism.

INTRODUCTION

Staphylococcus aureus, including methicillin-resistant S. aureus (MRSA), is one of the best-studied and -described bacterial pathogens. It is well known to colonize humans and animals and to cause a range of skin and soft tissue infections, severe invasive diseases, and toxin-mediated illnesses (1). The rise and dissemination of MRSA is a major global antimicrobial resistance threat with significant clinical and infection prevention and control (IPAC) implications (2). To properly support clinical decision making and IPAC interventions, timely and accurate identification of S. aureus/MRSA in the clinical laboratory is imperative.

There are many clonal complexes and sequence types of S. aureus which are often used to describe the molecular epidemiology of this pathogen. In 2015, results from investigations into two staphylococcal species which were almost identical to S. aureus phenotypically and by 16S rRNA gene sequences gave rise to the formal descriptions of Staphylococcus argenteus sp. nov., formerly known as S. aureus clonal complex 75 (CC75), and Staphylococcus schweitzeri sp. nov. (3, 4). S. argenteus and S. schweitzeri each have distinct multilocus sequence types (MLST), and genomic sequence analysis clearly demonstrates that they are phylogenetically divergent from S. aureus, displaying less than 95% average nucleotide identity (ANI) with one another and a predicted DNA-DNA hybridization of less than 70% between the species (3, 5–8). S. argenteus and S. schweitzeri have distinct spectral profiles on some research applications and/or databases of matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS), as well as unique cellular fatty acid signatures compared to those of S. aureus. While S. aureus, S. argenteus, and S. schweitzeri have nearly identical 16S rRNA gene sequences to one another, there are considerable differences in nuc, which is important given that some of these differences occur in areas of the gene often used as primer binding sites in PCR assays designed to detect S. aureus (3, 8–10).

S. schweitzeri is considered a zoonotic agent primarily associated with fruit bats (Eidolon helvum) and monkeys (Cercopithecus ascanius) in Africa (4, 11); that no human clinical cases have been reported to date suggests that this organism may inhabit a separate ecological niche from that of S. aureus.

In contrast, S. argenteus has now been established as an effective colonizer and pathogen of humans (6, 12–20). Initial reports of S. argenteus, both methicillin-susceptible and methicillin-resistant, were from Australia, the Pacific Islands, and Thailand and were thought to be geographically restricted (3, 5, 6, 13, 21–23). Based on early clinical observations, S. argenteus was considered less likely to cause nosocomial infections than S. aureus, appearing to be predominantly associated with community-onset superficial skin lesions (21). Based on this observation and murine studies, and because S. argenteus lacked some putative S. aureus virulence genes such as staphyloxanthin, the initial assessment was that S. argenteus was less virulent than S. aureus (7, 21, 24).

Awareness of S. argenteus was heightened after the formal description of the species as well as with the addition of the S. argenteus and S. schweitzeri spectra to the database of one commonly used commercial MALDI-TOF MS system (RUO database 2018, V8.0, 7854 Bruker; Bruker Daltonics). In the last few years, increased numbers of clinical reports and descriptions of S. argenteus resulted in increased attention on this organism and a recognition of a possible global distribution (12, 14–16, 25–30). It is now also evident that S. argenteus can cause serious invasive disease, including bacteremia, bone and joint infection, and purulent lymphadenitis, and has been directly linked to patient deaths (13, 16, 18, 19, 30). Like its relative S. aureus, S. argenteus has been shown to cause toxin-mediated foodborne illnesses (15, 20). In addition to community-associated spread, reports have documented nosocomial spread of S. argenteus (13, 24). Recent studies have also described strains of S. argenteus harboring traditional S. aureus virulence factors such as Panton-Valentine leukocidin (PVL), enterotoxins, and toxic shock syndrome toxin-1 (TSST-1), among others (12, 14, 20, 25, 28, 31, 32).

Considering the clinical and IPAC significance of S. argenteus, this organism should be correctly identified so that cases are not missed. For greater understanding of clinical, microbiological, and epidemiologic similarities and differences between S. argenteus and S. aureus, the two species should be distinguished. Despite the addition of S. argenteus spectra to at least one commercial MALDI-TOF MS database, confirmation of identification of and distinction from S. aureus and S. schweitzeri in the clinical laboratory is not straightforward. Laboratories may be identifying S. argenteus as S. aureus, as these species share equivalent reactions to most key biochemical tests traditionally used for characterization, such as catalase and tube coagulase positivity and beta-hemolysis on blood agar (3). One exception is that while S. aureus is often (but not exclusively) golden in color, S. argenteus colonies have been found to be nonpigmented, appearing white/silver due to the absence of the staphyloxanthin gene cluster (crtOPQMN) which encodes carotenoid pigment (7). As mentioned previously, 16S rRNA gene sequence analysis is not able to discriminate between S. argenteus, S. aureus, and S. schweitzeri.

Here, our objective was to verify the identifications of S. argenteus identified by MALDI-TOF MS and characterize the isolates. Using several identification methods, including whole-genome sequencing (WGS), we describe initial isolates of S. argenteus received at two large North American microbiology reference laboratories, one in Canada and one in the United States. The description of the microbiological characteristics, including antimicrobial susceptibility profiles, of these North American isolates as well as basic clinical details of the patients contributes to a growing understanding of the epidemiology and the clinical presentations of this organism. Based on our laboratory experience as well as the growing body of literature on the clinical and IPAC significance of S. argenteus, we provide recommendations as to how to report this organism in the clinical laboratory, addressing breakpoint challenges with S. argenteus.

MATERIALS AND METHODS

Specimen collections.

This study involved a collection of 22 clinical isolates of S. argenteus received between 2017 and 2019 by two large North American reference laboratories (Public Health Ontario [PHO], Toronto, ON, Canada [n = 16] and Mayo Clinic, Rochester, Minnesota, USA [n = 6]), along with strains S. argenteus DSM 28299^T/MSHR1132^T and S. schweitzeri DSM 28300^T/FSA084^T. Clinical isolates were received at the reference laboratories for identification and antibiotic susceptibility testing. Routine identification methods for Staphylococcus and related species were conducted, including tube coagulase, pyrrolidonyl aminopeptidase test (PYR), bacitracin disk, and MALDI-TOF MS (Bruker, BioTyper, databases V7.0 [7311] used 2017, V8.0 [7854] used 2018, and V9.0 [8468] used 2019) (Table 1). Of note, BioTyper RUO database V7.0 (7311) did not include S. argenteus or S. schweitzeri for which spectra were added in the V8.0 (7854) update. For closely related organisms, MALDI-TOF MS (Bruker, BioTyper) can occasionally result in multiple species scoring at or above the cutoff of >2.0 (confident to the species level). Different approaches were employed by each laboratory to avoid potential misidentifications that can occur as a result of reporting only the top-scoring identification by MALDI-TOF MS. At PHO, 16S rRNA gene PCR and sequence analysis were performed using universal primer pair 8FPL and 806R (33). 16S rRNA gene sequences were subjected to a BLAST search against the NCBI GenBank type strain and open nucleotide databases (34) with interpretation criteria described in the Clinical and Laboratory Standards Institute (CLSI) document MM18-A used to identify the isolates (35). At Mayo Clinic, if multiple species had scores of ≥2.0, a score separation of at least 10% was required between the top match and additional species in order to report the top-scoring organism to the species level. If a ≥10% score separation was not observed, organisms were only reported to the genus level. This 10% differential rule, or separation rule, has been used by other groups to aid in differentiating closely related organisms (36, 37).

TABLE 1.

Initial characteristics

Study no.	Date of isolation (mo-yr)	Region	Submitter	Site of isolation	Patient sex	Age range	Tube coagulase	PYR	Bacitracin disk	Cefoxitin 30 μg disk (mm zone)a	mecA PCRb	nuc PCRb	MALDI-TOF MS ID ≥2.0c,d	16S rRNA ≥99%e
PHL3431	10-2017	Ontario, Canada	Site 1	Rectal	M	≥75	+	−	+	13	+	−	S. aureus	NDf
PHL3432	10-2017	Ontario, Canada	Site 1	Nose	F	≥75	+	−	+	10	+	−	S. aureus	ND
PHL3433	10-2017	Ontario, Canada	Site 1	Nose	M	≥75	+	−	+	11	+	−	S. aureus	ND
PHL5740	10-2018	Ontario, Canada	Site 2	Eye wound	M	55–64	+	−	+	30	−	−	S. aureus /S. argenteus/S. schweitzeri	S. aureus /S. argenteus/S. haemolyticus/S. simiae
PHL6344	11-2018	Ontario, Canada	Site 2	Cheek	F	≥75	+	−	+	28	−	−	S. argenteus/S. schweitzeri/ S. aureus	S. aureus /S. argenteus/S. haemolyticus/S. simiae
PHL8605	12-2018	Ontario, Canada	Site 2	Wound	M	≥75	+	−	+	29	−	−	S. argenteus	S. aureus /S. argenteus/S. haemolyticus
PHL3446	01-2019	Ontario, Canada	Site 3	Toe tissue	M	45–54	+	−	+	29	−	−	S. argenteus/S. schweitzeri/ S. aureus	S. aureus /S. argenteus/S. haemolyticus
PHL2420	01-2019	Ontario, Canada	Site 2	Stump wound	M	65–74	+	−	+	29	−	−	S. argenteus/S. schweitzeri/S. aureus	S. aureus /S. argenteus/S. haemolyticus
PHL6318	01-2019	Ontario, Canada	Site 3	Sternal wound	M	55–64	+	−	+	26	−	−	S. argenteus/S. schweitzeri/ S. aureus	S. aureus /S. argenteus/S. haemolyticus/S. simiae
PHL1144	02-2019	Ontario, Canada	Site 1	Rectal & wound swab	F	0–1	+	−	+	28	−	−	S. argenteus	S. aureus /S. argenteus/S. haemolyticus/S. simiae
PHL2411	03-2019	Ontario, Canada	Site 2	Tendon tissue	M	45–54	+	−	+	30	−	−	S. argenteus/S. schweitzeri	S. aureus /S. argenteus/S. haemolyticus/S. simiae
PHL8642	03-2019	Ontario, Canada	Site 4	Eye	M	NA	+	−	+	31	−	−	S. argenteus/ S. aureus	S. aureus /S. argenteus/S. haemolyticus/S. simiae
PHL4815	03-2019	Ontario, Canada	Site 5	Head wound	M	35–44	+	−	+	26	−	−	S. argenteus/S. schweitzeri	S. aureus /S. argenteus/S. haemolyticus
PHL4226	11-2019	Ontario, Canada	Site 6	Blood	M	45–54	+	−	+	30	−	−	S. argenteus	S. aureus /S. schweitzeri/S. argenteus/S. haemolyticus
PHL4313	11-2019	Ontario, Canada	Site 6	Synovial fluid	M	55–64	+	−	+	31	−	−	S. argenteus/S. schweitzeri/ S. aureus	S. argenteus/ S. aureus /S. schweitzeri/S. haemolyticus
PHL4553	11-2019	Ontario, Canada	Site 2	Blood	F	45–54	+	−	+	32	−	−	S. argenteus/S. schweitzeri/ S. aureus	S. argenteus/S. aureus/S. schweitzeri/S. haemolyticus
WU1	08-2019	Missouri, USA	Site 7	Foot tissue	F	55–64	+	−	NAg	25	−	−	S. aureus /S. schweitzeri/S. argenteus	ND
WU2	03-2019	Missouri, USA	Site 7	Buttocks wound	M	15–24	+	−	NA	25	−	−	S. schweitzeri/S. argenteus/ S. aureus	ND
WU3	12-2018	Missouri, USA	Site 7	Deep throat swab (cystic fibrosis)	M	15–24	+	−	NA	25	−	−	S. argenteus	S. argenteus/ S. aureus
MC2	02-2019	Minnesota, USA	Site 8	Urine	M	2–14	+	−	NA	25	−	−	S. argenteus/ S. aureus	ND
MC3	03-2019	Missouri, USA	Site 7	Buttocks wound	M	45–54	+	−	NA	25	−	−	S. aureus	ND
MC4	08-2019	Colorado, USA	Site 9	Blood	M	2–14	+	−	NA	24	−	−	S. argenteus	ND
DSM 28299T S. argenteus	2006	Northern Territory, Australia	Type strain	Blood	F	55	+	−	NA	12	+	−	S. argenteus/S. schweitzeri/ S. aureus h	ND
DSM 28300T S. schweitzeri	2010	Moyen-Ogooué, Gabon	Type strain	Nasal swab (monkey)	NA	NA	+	−	NA	26	−	−	S. schweitzeri h	ND

^aFor S. aureus, cefoxitin disk (30 μg disk content) zone diameter breakpoints ≥22 mm susceptible and ≤21 mm resistant; for other Staphylococcus spp. (excluding S. aureus, S. lugdunensis, S. pseudintermedius, S. schleiferi, S. epidermidis), zone diameter breakpoints ≥25 mm susceptible and ≤24 mm resistant (CLSI M100 2020).

^bqPCR assay by Pichon et al. (10).

^cBruker MALDI BioTyper, databases: 2017, V7.0 (7,311 entries); 2018, V8.0 (7,854 entries), and 2019, V9.0 (8,468 entries). The identifications presented are those at the initial time of testing with the database that was in use at the time.

^dMatches presented by descending score (top hit listed first).

^ePartial 16S rRNA gene sequence (∼700 bp) BLAST NCBI type strain database, presented by descending max score.

^fNot done.

^gNot available.

^hResults with Bruker MALDI BioTyper database V9.0 (8,468 entries).

The S. argenteus strain DSM 28299^T (MSHR1132^T) and the S. schweitzeri strain DSM 28300^T (FSA084^T), inoculated at Mayo Clinic, were used as controls. All isolates were frozen and maintained at −80°C.

Basic characteristics of the patients, including sex, age range, month and year of collection, geographic region, and specimen type, were recorded (Table 1). This work was approved by the PHO Research Office’s Ethics Review Board, and a Privacy Impact Assessment was completed. This work was also approved by Mayo Clinic’s Institutional Review Board.

Antimicrobial susceptibility testing and determination of MICs.

Antimicrobial susceptibility testing was performed by agar dilution according to CLSI guidelines on all isolates to determine MICs (Table 2) (38). D-test for inducible clindamycin resistance and penicillin zone edge test (inducible beta-lactamase) were also performed in accordance with CLSI M100 recommendations (39). Cefoxitin disk (30 μg) diffusion and a PCR test to detect mecA (with nuc as a positive control for S. aureus) were also performed (10, 39).

TABLE 2.

Antibiotic susceptibility test results (phenotypic and genomic)

Isolate no.	MIC (μg/ml)a
	Ampicillin	Cefazolin	Ciprofloxacin	Clindamycin	Erythromycin	Gentamicin	Linezolid	Penicillin	Tetracycline	Trimethoprim-sulfamethoxazole	Vancomycin	Ceftaroline	Levofloxacin	Mupirocin	Minocycline	Nitrofurantoin	Rifampin	Doxycycline	Delafloxacin	Oxacillin	D-test	Penicillin zone edge test (inducible beta-lactamase)	CARD resistance database (genes)
PHL3431	≥16 b	≥32	≥4	≤0.5	≤0.5	≤4	≤2	8	≤4	≤2/38	1	1	≤1	≤256	≤4	≤32	≤0.5	≤4	0.12	>2	−	NAc	mgrA, arlR, mepR, lmrS, fosB, mecA
PHL3432	≥16	≥32	≥4	≤0.5	≤0.5	≤4	≤2	8	≤4	≤2/38	1	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	mgrA, arlR, mepR, lmrS, fosB, mecA
PHL3433	≥16	≥32	≥4	≤0.5	≤0.5	≤4	≤2	8	≤4	≤2/38	1	0.5	1	≤256	≤4	≤32	≤0.5	≤4	0.12	>2	−	NA	mgrA, arlR, mepR, lmrS, fosB, mecA
PHL5740	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	4	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	1	−	NA	mgrA, arlR, mepR, lmrS, fosB, blaZ
PHL6344	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	4	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	1	−	NA	mgrA, arlR, mepR, lmrS, fosB, blaZ
PHL8605	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
PHL3446	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	≤0.12	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
PHL2420	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
PHL6318	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	≤0.12	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	NA	mgrA, arlR, mepR, lmrS, fosB
PHL1144	≤8	≤8	≤1	≤0.5	≤0.5	≥16	≤2	4	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	1	−	NA	mgrA, arlR, mepR, lmrS, fosB, blaZ, AAC(6′)-Ie-APH(2′')-Ia
PHL2411	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS
PHL8642	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	0.5	≤4	≤2/38	1	≤0.12	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	NA	mgrA, arlR, mepR, lmrS, blaZ
PHL4815	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	NA	mgrA, arlR, mepR, lmrS, fosB
PHL4226	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	≤0.12	≤4	≤2/38	1	0.25	≤0.5	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	mgrA, arlR, mepR, lmrS, fosB
PHL4313	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	4	≤4	≤2/38	1	0.25	≤0.5	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	NA	mgrA, arlR, mepR, fosB, lmrS, blaZ
PHL4553	≤8	≤8	≤1	≤0.5	≤0.5	≤4	≤2	0.5	≤4	≤2/38	1	0.25	≤0.5	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	NA	mgrA, arlR, lmrS, blaZ
WU1	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	≤0.12	≤4	≤2/38	2	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
WU2	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	≤0.12	≤4	≤2/38	2	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
WU3	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	≤0.12	≤4	≤2/38	2	≤0.12	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
MC2	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	4	≤4	≤2/38	2	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	1	−	NA	mgrA, arlR, mepR, lmrS, fosB, blaZ, fusB
MC3	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	≤0.12	≤4	≤2/38	2	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, fosB
MC4	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	≤0.12	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, mexT, lmrS, fosB
S. argenteus DSM 28299T	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	8	≤4	≤2/38	1	0.5	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	>2	−	+	mgrA, arlR, mepR, lmrS, fosB, mecA, blaZ
S. schweitzeri DSM 28300T	NA	NA	NA	≤0.5	≤0.5	≤1	≤2	≤0.12	≤4	≤2/38	1	0.25	≤1	≤256	≤4	≤32	≤0.5	≤4	≤0.008	≤0.25	−	−	mgrA, arlR, mepR, lmrS, murA with mutation

^aMICs were obtained by CLSI agar dilution method, with the exception of delafloxacin for which agar dilution was performed on all isolates at Mayo Clinic and D-test was performed on all isolates at Mayo Clinic by the disk diffusion method.

^bBold, underlined MICs indicate resistance according to CLSI M100 (39) using S. aureus breakpoints.

^cNA, not available.

Whole-genome sequencing.

Genomic DNA from colonies were extracted and purified using a QiaAmp DNA minikit (Qiagen, Valencia, CA) or Zymo Research Quick-DNA Fungal/Bacterial MiniPrep kit (Zymo Research Corp., CA) from an overnight culture grown aerobically on sheep blood agar according to the manufacturer’s protocol. The quality of isolated DNA was analyzed using gel electrophoresis and quantified using Qubit 2.0 (Invitrogen, Waltham, MA) fluorometer.

DNA libraries were prepared and multiplexed with a unique combination of two indexes of the Nextera XT index kit (Illumina, San Diego, CA). The sequencing library was quantified using Qubit 2.0 (Invitrogen, Waltham, MA) and qualified by Bioanalyzer (Agilent Technologies, Richardson, TX). The library was normalized and pulled together aiming for an average coverage of 100× and sequenced on the Illumina MiSeq platform, using Illumina MiSeq reagent kit v2 (2 × 150 bp), according to the manufacturer’s instructions.

Genome assembly.

FastQ files were imported into CLC Genomics Workbench version 8.0.1 (CLC bio, Germantown, MD, USA). Raw reads were trimmed to remove Nextera transposase adapter sequences and assembled using de novo assembler. Following de novo assembly, the largest contig was subjected to NCBI BLAST to select the closest matching complete S. argenteus genome in GenBank to serve as the reference for reference-based assembly using CLC Genomic Workbench 8.5.3 (QIAGEN Digital Insights).

Genome analysis.

The assembly was annotated using the RAST server (http://rast.nmpdr.org) (40) for gene prediction and annotation. Genome sequencing data of each isolate was assessed for in silico identification of MLST, SCCmec, and plasmid replicon and resistance genes using online tools such as CARD (41), MLST-1.8 server (42), ResFinder 2.1 (43), and SCCmecFinder 1.2 server (44) available by Center for Genomic Epidemiology (http://genomicepidemiology.org). Virulence factors (VFs) were identified by using VFanalyzer of the virulence factor database (VFDB) (http://mgc.ac.cn/VFs/) (45). To identify plasmids, the genome assemblies were screened using PlasmidFinder database for plasmid replicon (rep) genes (46).

Single nucleotide variant (SNV) analysis was conducted using a custom pipeline. Briefly, reads for all isolates were mapped against the chromosome of the strain MSHR1132^T (DSM 28299^T) available in GenBank (accession number NC_016941.1) for reference using SMALT software v 0.7.6 (https://www.sanger.ac.uk/tool/smalt-0/). Single nucleotide polymorphism (SNP) calling was performed using Freebayes with min-base-quality 30, min-mapping-quality 30, min-alternate-fraction 0.75, read-snp-limit 10, and min-coverage 15 (47). Additional variant confirmation was done using the SAMtools mpileup tool (48). Repetitive regions were removed by using MUMmer (49). The meta-alignment of core informative positions (SNVs) was used to create a maximum likelihood (ML) tree using MEGA 6 (50).

A whole-genome phylogenomic approach was performed by uploading the assembled sequences to Type Strain Genome Server (TYGS; https://tygs.dsmz.de) for a whole-genome-based phylogenetic analysis of isolates to the most closely related genome database (51).

ANI was calculated using EZBioCloud ANI calculator (52) between assembled genomes obtained in this study and publicly available genomes of S. argenteus MSHR1132^T (accession number NC_016941.1), S. argenteus strain XNO62 (accession number CP023076.1), S. schweitzeri NCTC 13712^T (accession number NZ_LR134304.1), and S. aureus NCTC 8325 (accession number NC_007795.1). An ANI threshold of ≥96% was considered to be the cutoff for species identification, which correlates to DNA-DNA hybridization studies (53). Whole-genome comparison was performed and visualized using the Gview tools (54).

spa typing, Sanger protocol.

All S. argenteus in this study were subjected for spa sequence-based typing using the previously published PCR primers (55). spa amplicons were sequenced with the same primers and analyzed using the spa type finder/identifier (http://spatyper.fortinbras.us) and the Ridom SpaServer (https://spaserver.ridom.de).

Data availability.

WGS sequence data are available in the following databases. NCBI (https://www.ncbi.nlm.nih.gov/): accession numbers QQOV00000000 (PHL3431), PUXC00000000 (PHL3432), and QQOW00000000 (PHL3433). BioProject accession number PRJNA666697: WGS numbers JADANH000000000 (PHL6344), JADANG000000000 (PHL5740), JADANF000000000 (PHL3446), JADANE000000000 (PHL2420), JADAND000000000 (PHL6318), JADANC000000000 (PHL8605), JADANB000000000 (PHL2411), JADANA000000000 (PHL1144), JADAMZ000000000 (PHL4815), JADAMY000000000 (PHL8642), JADAMX000000000 (PHL4226), JADAMW000000000 (PHL4313), JADAMV000000000 (PHL4553), JADAMU000000000 (DSM 28299^T), JADAMT000000000 (DSM 28300^T), JADAMS000000000 (MC2), JADAMQ000000000 (MC3), JADAMP000000000 (MC4), JADAMO000000000 (WU1), JADAMN000000000 (WU2), and JADAMR000000000 (WU3).

RESULTS

Initial description of isolates.

In total, 22 isolates were received at the PHO laboratory (n = 16) and Mayo Clinic (n = 6) for confirmation of identification of S. aureus, MRSA, or S. argenteus from 2017 to 2019 (Table 1). All isolates were tube coagulase positive and PYR negative.

The initial 3 S. argenteus isolates arrived at the PHO laboratory in 2017, as MRSA isolates (PHL3431, PHL3432, and PHL3433) for confirmation of identification. At the time, MALDI-TOF MS provided confident identifications of S. aureus (≥2.0), with positive tube coagulase and PYR reactions; of note, the BioTyper RUO database V7.0 (7311) used at the time did not include S. argenteus or S. schweitzeri. To determine the presence of mecA and confirm the identification of MRSA, these isolates were run on an in-house mecA PCR assay (10). All three isolates were positive for mecA; however, the positive control for S. aureus, nuc, was negative (Table 1). These results prompted further investigation of the identification of the organisms, and they were referred for WGS analysis, which determined that they were S. argenteus (details below).

In 2018, the Bruker BioTyper RUO database was updated to include spectra for S. argenteus and S. schweitzeri for the first time (RUO databases V8.0 [7854 MSP] and V9.0 [8468]), after which both reference laboratories were technically able to use MALDI-TOF MS to identify S. argenteus and S. schweitzeri. When the single S. schweitzeri strain available in this study (DSM 28300^T) was tested by MALDI-TOF MS using the new databases, it resulted in a single identification of ≥2.0 of S. schweitzeri; of note, this same strain was used to produce the spectrum in the library (Table 1). However, use of the updated databases challenged with S. argenteus typically resulted in multiple species (S. aureus, S. argenteus, and S. schweitzeri) with high (≥2.0) scores (Table 1), resulting in an inability to confidently identify the organism to species level. This prompted additional testing to confirm identification, including attempts at using 16S rRNA gene sequence analysis.

Partial 16S rRNA gene PCR and sequence analysis of these isolates were not able to reliably differentiate between S. aureus, S. argenteus, and S. schweitzeri as well as a strain of Staphylococcus haemolyticus (accession number Z26896.1); all had ≥99.0% homology to reference strain deposits within NCBI GenBank for these species (Table 1). Of note, S. haemolyticus could be ruled out as it is coagulase negative, and Z26896.1 was likely deposited into NCBI incorrectly.

Isolates were also tested with the in-house nuc quantitative PCR (qPCR) in which the primers are specific to S. aureus (10) (the nuc target is used as a control for S. aureus in a mecA PCR assay). All isolates of S. argenteus for which the described nuc qPCR assay was run were negative (Table 1), while control isolates of S. aureus were positive (data not shown). While this result is evidence that the isolates under investigation were not S. aureus, the negative results of this assay do not specifically identify S. argenteus and cannot rule out S. schweitzeri. The lack of definitive identification prompted the use of WGS.

Patient demographics and specimen descriptions.

S. argenteus isolates were cultured from patients in Ontario, Canada (n = 16), Missouri (n = 4), Minnesota (n = 1), or Colorado (n = 1) and identified between October 2017 and November 2019. Detailed patient information for these isolates was not available. Of the 22 isolates described, 7 were from sterile sites, 11 from nonsterile sites, and 4 from surveillance screens. Seventy-seven percent (n = 17) were recovered from males and 23% (n = 5) from females. Ages of patients ranged from <1 to ≥75 years (Table 1).

Phenotypic antimicrobial susceptibility.

MICs for all isolates and antibiotics tested are presented in Table 2. All S. argenteus identified in this collection demonstrated low MICs to clindamycin, erythromycin, linezolid, trimethoprim-sulfamethoxazole, and vancomycin and would be considered susceptible using S. aureus breakpoints (39). Ten of 22 (45.4%) isolates were resistant to penicillin (MIC range 0.5 to 8 μg/ml); 3 of these isolates (PHL3431, PHL3432, PHL3433) were also mecA PCR positive, demonstrating oxacillin MICs of >2 μg/ml (PHL3431, PHL3433) and resistance to ciprofloxacin. S. argenteus isolates with penicillin MICs of ≤0.12 μg/ml were negative by the penicillin zone edge test. One isolate (PHL1144) was resistant to gentamicin with an MIC of 16 μg/ml.

As there are different oxacillin breakpoints for different Staphylococcus species (39), we evaluated multiple species-specific breakpoints for oxacillin susceptibility testing. Using S. aureus breakpoints, 3 of 21 isolates (14.3%) would be considered oxacillin resistant, while when using the “other Staphylococcus species” breakpoints, 7 of 21 (33.3%) would be considered resistant. All S. argenteus isolates identified as oxacillin resistant using S. aureus breakpoints harboring mecA (as determined by mecA-specific qPCR as well as WGS analysis). PHL3432 was identified as being mecA positive but was not subjected to oxacillin susceptibility testing due to strain loss.

Genome features and comparative genomics.

De novo assembly of sequence reads using CLC Genomics Workbench resulted in 1,610,728 to 7,425,408 bp. Between 12 and 315 contigs were obtained, with the largest contig size of 1,088,191 bp length and average coverage of 153×. The N₅₀ values of assemblies were 23,043 to 693,825 bp.

The sequences of all 23 S. argenteus (22 clinical isolates and 1 type strain, DSM 28299^T) and 1 S. schweitzeri genome (DSM 28300^T) were assembled by mapping short reads to the reference sequence XNO62 (accession number CP023076.1). Assembled genomes ranged from 2,599,989 to 2,724,271 bp in size with a GC content of 32.42%. Genome annotation by RAST showed an average of 2,665 (median 2,603) coding sequences and average of 63 RNAs (tRNAs and rRNAs) among all S. argenteus. De novo assembly of 4,014,380 reads obtained for type strain S. schweitzeri (DSM 28300^T) resulted in 37 contigs between 200 and 343,130 bp; the assembled genome was 2,784,939 bp long with 305× coverage and N₅₀ of 200,036. Using RAST annotation, 2,885 coding DNA sequences (CDSs) and 65 RNAs (tRNAs and rRNAs) were identified.

The clinical isolates’ pairwise ANI values between isolates studied and S. argenteus XNO62 (accession number CP023076.1) were 98.70 to 99.96% and were 92.61 to 92.23% and 87.58 to 89.0% compared against S. schweitzeri and S. aureus, respectively (see Table S1 in the supplemental material).

A circular alignment of genomes from this study compared with reference genomes S. argenteus XNO62 and MSHR1132^T is presented in Fig. S1. This alignment, as well as other features such as sequence similarity and distribution of GC content, demonstrated a high level of homology among members of these groups.

The WGS data for all 22 clinical isolates were interrogated for MLST and other features commonly used for typing S. aureus. Four MLSTs were identified in the collection: ST2250 (16/22) predominated, followed by ST1223 (3/22), ST2198 (2/22), and ST2793 (1/22) (Fig. 1A). MLSTs for type strains of S. argenteus (DSM 28299^T) and S. schweitzeri (DSM 28300^T) were determined to be ST1850 and ST2022, respectively, as previous described (3). Using Basic Local Alignment Search Tool (BLAST; https://blast.ncbi.nlm.nih.gov/Blast.cgi) on extracted staphylocoagulase (coa) sequences from assemblies, three coa types were identified: XId, XV, and XIV, with coa-type XId being the predominant type (17/22) (Fig. 1A). We were unable to assign spa type from in silico whole-genome data due to the repetitive region span being greater than the read length of 150 bp. Using traditional endpoint PCR and Sanger sequencing, we were able to confirm presence of at least 8 spa types among clinical isolates of S. argenteus in this study. All mecA-positive S. argenteus identified in this study (3/22) were assigned the spa type t7960, while the majority of isolates were spa type t5078 (9/22). Other spa types identified include t10900 (1/22), t9385 (2/22), and 3 novel spa types, t19456 (1/22), t19457 (1/22), and t19497 (2/22). We were unable to assign a spa type to 3 isolates, all of which belonged to ST1223 (https://spa.ridom.de/spatypes.shtml) (Table S2). Strains DSM 28299^T and DSM 28300^T were t17252 and t6705, respectively.

FIG 1.

Whole-genome sequence phylogenetic tree of S. argenteus. (A) Whole-genome sequence phylogenetic tree of S. argenteus from this study and related type strains of the genus Staphylococcus obtained from TYGS. Tree inferred with FastME 2.1.6.1 from Genome BLAST Distance Phylogeny approach (GBDP) distances calculated from genome sequences. The branch lengths are scaled in terms of GBDP distance formula d5. The numbers above branches are GBDP pseudo-bootstrap support values >60% from 100 replications. The reference sequences available in GenBank and the type strains sequenced in this study are indicated by solid black and blue dots, respectively. The tree was rooted at the midpoint. Staphylocoagulase (coa), SCC mec IV (2B), mecA, ccrA and ccrB, IS1272, PVL, TSST, MLST, and spa type are indicated for each strain. Presence of a given gene/element is identified by colored boxes. NT, nontypable. (B) Magnified main branch of phylogenetic tree containing S. argenteus representing clustering of strains possessing same MLST. *, sequences for these type strains (DSM 28229^T, DSM 28300^T) were generated in this study; #, novel spa types identified in this study.

Phylogenetic relationships of S. argenteus genomes were assessed using core genome SNV as well as whole-genome sequence analysis, with both methods yielding phylogenetic trees with similar topologies. Examination of clinical isolates as well as the strain DSM 28299^T resulted in 5 clusters of S. argenteus which coincided with the 5 MLSTs (WGS phylogenetic tree shown in Fig. 1, SNV tree not shown). The major cluster was comprised of 16 isolates (10 from Canada and 6 from the United States), all MLST 2250, with little variation in the core genome (0.2 to 0.3%).

The full lengths of S. argenteus, S. aureus, and S. schweitzeri 16S rRNA and nuc genes were extracted from the assemblies. As shown before by others, our results demonstrate that there is insufficient demarcation between the 16S rRNA gene of these three species to be used for species identification (Fig. S1). However, the nuc gene allows for clear separation of the three species (Fig. 2) (3, 8, 9).

FIG 2.

nuc phylogenetic tree. Phylogenetic dendrogram of nuc sequences of S. argenteus from this study and S. argenteus, S. schweitzeri, and S. aureus strains from GenBank constructed by neighbor-joining method. The evolutionary distances were computed using the maximum composite likelihood method and are in the units of the number of base substitutions per site. The analysis involved 125 isolates. There were a total of 426 positions in the final data set. Evolutionary analyses were conducted in MEGA6. Green text, S. argenteus; red text, S. schweitzeri; blue text, S. aureus; black diamond, genome sequence of type strains generated in this study; black square, genome sequence of type strains obtained from GenBank; red circles, sequences from PHO isolates; blue circles, sequences from Mayo Clinic isolates.

Antibiotic resistance determinants identified through WGS analysis.

A total of 9 antibiotic resistance genes were identified among the 22 S. argenteus clinical isolates; each isolate harbored multiple genes associated with resistance (Table 2). Genes identified include mecA (3/22), mgrA (norR) and arlR (both 22/22) (encode regulatory components for a major facilitator superfamily [MFS] antibiotic efflux pump associated with fluoroquinolone, tetracycline, penicillin, cephalosporin, and peptide antibiotic resistance) (56, 57), mepR (21/22) (encodes a repressor that helps to regulate expression of a multidrug and toxic compound extrusion [MATE] efflux pump [58, 59]), blaZ (7/22) (encodes class A β-lactamase) (60), fosB (18/22) (encodes a fosfomycin thiol transferase which inactivates fosfomycin) (61), lmrS (22/22) (encodes a component of an MFS efflux pump associated with resistance to several antibiotics, including macrolides, aminoglycosides, diaminopyrimidines, and oxazolidinones) (62), and aac(6′)-Ie-aph(2′')-Ia (1/22) (encodes an aminoglycoside-modifying enzyme which inactivates aminoglycoside, including gentamicin) (63). As demonstrated in Table 2, many of these molecular determinants of resistance correlated with phenotypic resistance. Further analysis of isolates with mecA showed that their SCCmec type IV possesses the class B mec gene complex, harboring the mecA, mecR1, IS1272, and ccrA/ccrB gene complex (Fig. 1A).

Given a recent report of daptomycin resistance in S. argenteus associated with a point mutation in mprF, we analyzed the predicted protein of this gene in all isolates of this study and found all to be wild type (S337, data not shown) (30).

As plasmids often carry antibiotic resistance determinants, we investigated the presence of plasmids using genomic analysis. Using the PlasmidFinder, a total of 7 plasmid replicon types were detected in our isolates, characterized as rep16, rep5, repUS5, rep5a, repUS9, rep20, and rep21 with accession numbers BX571858, NC005011, NC003265, AP003139, AF203376, FN433597, and NC007790, respectively, with an alignment similarity ranging from 99.24 to 100%. The S. argenteus isolates in this study were found to harbor multireplicons in various combinations, including rep16/rep5, rep16/rep5a, rep16/repUS5, rep16/rep5/repUS9, and rep21/rep20 (Table S2).

Putative virulence determinants.

Analysis of putative virulence gene composition based on the VF database revealed a number of virulence determinants in all S. argenteus isolates from this study, including autolysin (atlE), elastin binding protein (ebp), fibrinogen binding protein (efb), intercellular adhesion (icaA, icab, icaC, and icaR), staphylokinase (sak), staphylococcal complement inhibitor (scn), staphylococcal protein A (spa), cysteine protease (sspB), hyaluronate lyase (hysA), lipase (geh and lip), staphylocoagulase (coa), thermonuclease (nuc), capsular (cap5 and cap8), type VII secretion system (esaG, essA, essB, essC, and esxA), alpha hemolysin (hly/hla), delta hemolysin (hld), exfoliative toxin type A (eta), and gamma hemolysin (hlgA, hlgB, and hlgC) genes (Table S3). Additionally, one clinical S. argenteus isolate (WU3, a throat swab isolate) harbored the gene that encodes Panton-Valentine leukocidin (PVL), and another single isolate (MC4, a blood isolate) harbored the gene for toxic shock syndrome toxin (TSST-1).

DISCUSSION

Through many reports over the past several years, including this one, it is clear that S. argenteus is a distinct species from S. aureus and that it is clinically relevant as a cause of skin and soft tissue infections as well as severe invasive disease (3, 7, 12–14, 16–19, 23, 24, 64). Several reports have indicated international detection of S. argenteus (12, 14–16, 25–29, 65–68); here, we report the first large collection of clinical isolates of S. argenteus in North America, from both Canada and the United States. As large reference laboratories, we each receive isolates to confirm bacterial identification. From 2017 to 2019, collectively, we received and identified 22 isolates of S. argenteus.

We encountered some of the same initial challenges as other investigators in the differentiation of S. argenteus from S. aureus using both traditional biochemical assays and a commercial MALDI-TOF MS (Bruker Biotyper) system, despite having spectra for both S. argenteus and S. schweitzeri in the databases. Although not investigated here, another commercial MALDI-TOF MS system (Vitek MS, bioMérieux) does not currently have spectra for S. argenteus or S. schweitzeri, and users of this system may erroneously call these organisms S. aureus. Further investigation must be undertaken concerning the ability of commercial MALDI-TOF MS platforms and database/software versions to differentiate among the S. aureus complex members. At this time, current routine methods used in clinical microbiology laboratories will not conclusively differentiate S. argenteus from S. aureus.

All S. argenteus encountered during this investigation period were definitively identified using a WGS approach. Our WGS data are consistent with previous reports that S. argenteus is distinct from S. aureus and S. schweitzeri (3, 7, 14, 25, 69). All isolates of S. argenteus for which the described nuc PCR assay was run were negative, while isolates of S. aureus were positive (data not shown) (10); however, depending on where the primers are, this may not always be the case. The nuc sequences examined demonstrate considerable sequence diversity between S. aureus, S. argenteus, and S. schweitzeri, consistent with what other investigators have found (3, 8, 9, 14, 70). This sequence diversity can be exploited as with the appropriate primers the nuc gene can be a target for discriminating S. aureus from other members of the complex, thus avoiding the need for WGS to distinguish S. aureus from S. argenteus/S. schweitzeri. An example of such an assay is the one developed by the Mayo Clinic (71). As many laboratories already include a nuc target in their mecA PCR as a control, this may not require significant workflow changes. Other groups have identified crtM and sodA as useful targets for differentiating these species (29, 72).

S. argenteus from this collection were isolated from a range of sites, including normally sterile and nonsterile specimen sources, and from surveillance swabs. Patients ranged in age from <1 to ≥75 years, with most being older than 45 years of age. One limitation of this study is that we do not have detailed information on the patients, including if they were inpatients or outpatients and whether the S. argenteus was likely acquired in the community or via a nosocomial route. It is possible that 3 isolates in this study were part of a nosocomial transmission event, as these isolates were collected from patients being investigated as part of a MRSA cluster in a health care facility (PHL3431, PHL3432, and PHL3433). These three isolates were identified through surveillance screens and are all temporally related from the same health care facility; all were mecA-positive ST2250, spa type t7960, SCCmec type IV (2B) with similar antimicrobial susceptibility patterns and were ≥99.94% similar to one another by ANI comparison. Another limitation of this work is that we have no treatment or clinical outcome data on these patients. This type of information needs to be collected to better understand the transmission dynamics, reservoirs, and best treatment options for patients; however, in order for this to be done, widespread, accurate identification of S. argenteus must be available.

This study is unique in that it includes a large number of S. argenteus isolates from North America, from both Canada and the United States. Although it has been postulated to have spread globally, this is the first report to demonstrate that S. argenteus is present in North America beyond a single case report and is likely circulating (30). Like many global studies, the majority of isolates (73%) in this study belong to MLST ST2250 (14, 16, 18, 19, 23, 32, 65, 73). These ST2250 isolates were also part of coa XId, described as the predominant type by Aung et al. in their study (69). There was a high degree of homology among these ST2250/coa XId isolates; however, 5 different spa types were identified, showing some diversity between strains.

Three additional previously reported but less frequently described MLSTs (ST1223, ST2198, and ST2793) were identified from the Canadian specimens, with corresponding coa types XV, XIV, and XId, respectively, thus demonstrating some diversity in the North American isolates (14, 16, 23, 25, 31, 32, 69, 72, 74–77). Most isolates in this study belonged to the coa types XIV and XV. spa typing is based on the highly variable and repetitive X-region of a single gene (spa) that encodes protein A. Use of reads generated by WGS in order to infer spa type has been done for S. aureus; however, factors such as length of the reads and repetitive sequence regions affect the de novo assembly and may result in incorrect assignments of spa type. To overcome these challenges, here we relied on traditional Sanger sequencing to confirm spa types and identified 7 different spa types, including 3 previously undescribed types (all in ST1223 isolates), adding evidence of the diversity of this collection of S. argenteus. A wide array of virulence determinants commonly associated with S. aureus were also detected, with PVL and TSST-1 of note. Further detailed investigation of isolates is needed to truly understand the diversity and epidemiology of the organism in North America.

It is interesting to consider why this organism appears to be emerging at this time. It is possible that S. argenteus, as part of the S. aureus clonal complex, has always been a human pathogen and had gone undetected until recently due to implementation of newer, more specific identification methods (e.g., WGS). However, others have provided evidence that supports a fairly recent host adaption which has allowed it to colonize and infect humans (the same does not appear to have happened with the closely related S. schweitzeri, for which there are no reports of human infection) (25, 75). Regardless of whether it is a newly emergent pathogen or an organism that we can only now differentiate from S. aureus, it is clear that S. argenteus is pathogenic and possesses many of the same virulence determinants and resistance mechanisms (including mecA) as S. aureus. Based on our experience as well as the emerging findings reported in the literature, we support the proposal that S. argenteus should be treated in a similar manner as S. aureus in terms of IPAC measures, as well as clinically, for the following reasons: (i) evidence points to similar disease presentations (including invasive disease), (ii) there is evidence of both community and nosocomial spread, and (iii) S. argenteus, like S. aureus, can carry mecA, PVL, TSST-1, and other virulence genes, suggesting it has similar pathogenic potential (78).

Clinical laboratories and other members of the health care teams should recognize that S. argenteus and S. schweitzeri are members of the S. aureus complex; this is significant, as classification guides appropriate clinical interpretation of culture as well as oxacillin susceptibility. If the oxacillin MIC breakpoints of the non-S. aureus group (S ≤ 0.25 μg/ml; R ≥ 0.5 μg/ml) are assigned to S. argenteus instead of S. aureus oxacillin MIC breakpoints (S ≤ 2 μg/ml; R ≥ 4 μg/ml), isolates may be inappropriately reported as oxacillin resistant (39). Likewise, cefoxitin disk diffusion breakpoints differ between these two groups. Furthermore, allowable methods for detection of methicillin resistance differ according to Staphylococcus species; for example, cefoxitin MIC is an appropriate testing method for S. aureus, while it is not appropriate for the “other Staphylococcus spp.” group. Given that S. argenteus appears to be more like S. aureus than other staphylococci applying the genomic, phenotypic, and clinical data to date, we propose that S. aureus antimicrobial breakpoints be applied for S. argenteus and S. schweitzeri as members of the S. aureus complex. At the January 2020 meeting of the CLSI Antimicrobial Susceptibility Testing Subcommittee, CLSI members voted to approve a motion to accept the recommendation, when a definitive identification cannot be made, to report S. argenteus as “S. aureus complex” or, when identified to species level, to report it as “S. aureus complex (S. argenteus)” so that it is not overlooked as a less pathogenic/nonpathogenic species (79). Additionally, it was approved to use S. aureus breakpoints and interpretive categories for reporting. This change will ensure accurate reporting of susceptibility results in all laboratories, including those currently unable to differentiate species within the complex, as well as provide guidance for laboratories that can identify these bacteria to the species level.

The availability of enhanced technology has dramatically improved our ability to accurately describe and identify microorganisms, but access to this technology as well as changing taxonomy can be challenging for the clinical laboratory as well as health care providers. If both organisms result in similar clinical presentations, some may ask if clinical laboratories should distinguish between S. aureus and S. argenteus. A recent position paper by the ESCMID Study Group for Staphylococci and Staphylococcal Diseases (ESGS) has proposed that due to known pathogenic and clinical similarities between S. aureus and S. argenteus, there is currently no need to distinguish within the S. aureus complex (78). However, in order to better understand the population epidemiology and pathogenicity of S. argenteus (and S. schweitzeri), it is important to distinguish the members of the S. aureus complex from a research and surveillance perspective. Based on the data to date, we recommend that clinical laboratories, where possible, (i) clearly report the organism as “Staphylococcus argenteus, member of Staphylococcus aureus clonal complex” or something similar in order to capture both the unique nature of the organism as well as its close link to the well-known pathogen (see Table S4 in the supplemental material) and (ii) apply “S. aureus/S. aureus complex” breakpoints for antibiotics, including oxacillin. This method of reporting will hopefully allow for education of clinical staff and better data collection as well as larger studies on the clinical and treatment outcomes of patients with these infections.