PDF
Citation
Print
The spread of vancomycin-resistant Enterococcus (VRE) is alarming owing to the limited availability of antibiotics suitable for treating infections caused by VRE. Vancomycin resistance is conferred by the acquisition of van gene clusters, such as vanA, B, C, D, E, G, L, M, and N, with vanA and vanB being the most prevalent and clinically important [1]. Among VRE strains, the frequency of vanA varies greatly in different countries [2]. The vanA gene cluster (located on the transposon Tn1546) has been extensively studied to explain the spread of high-level glycopeptide resistance among clinical Enterococcus isolates [3]. Glycopeptide-susceptible, vanA-bearing Enterococcus faecium (vancomycin variable Enterococcus [VVE]) was first discovered in 2011 in Quebec, Canada, and all six isolates identified lacked the vanR and vanS genes responsible for vanHAX expression [4]. Because VVE can acquire vancomycin resistance, resulting in treatment failure, detecting VVE is important [5].
We assessed the prevalence of VVE among vancomycin-susceptible E. faecium (VSE) isolated from clinical specimens and determined the structures of the vanA gene cluster in these isolates.
Between June 2021 and May 2022, we collected 282 consecutive VSE isolates from clinical specimens (one per patient). They were identified at the species level using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Antimicrobial-susceptibility testing was performed with the Vitek2 system (bioMérieux, Marcy 1’Etoile, France), and the vancomycin minimum-inhibitory concentration (MIC) was determined for all VSE isolates using the E-test (bioMérieux). According to CLSI M100 guidelines, the following breakpoints were used for interpretive categories: susceptible, ≤4 µg/mL; intermediate, 8–16 µg/mL; and resistant, ≥32 µg/mL [6]. The Institutional Review Board of Seoul St. Mary’s Hospital in Seoul, Korea (approval number KC21TISI0399) approved this study and waived the requirement for informed consent because only those strains that were acquired from diagnostic testing conducted for patient care purposes were utilized.
All VSE isolates were tested for the presence of vanA, vanB, and vanC using the Anyplex VanR Real-time Detection Kit (Seegene, Seoul, Korea). Transposon Tn1546-like elements in VVE isolates were further characterized based on a published set of overlapping primers [7] with some modifications (Supplemental Data Table S1). The PCR products were directly sequenced with the forward or reverse PCR primers using an ABI BigDye Terminator v3.1 Cycle Sequencing Kit and an ABI 3730xl automated DNA analyzer (Applied Biosystems, Foster City, USA). The sequencing results were compared with the prototypic Tn1546 sequence (GenBank accession number M97297, https://www.ncbi.nlm.nih.gov/nuccore/M97297). Sequence gaps were filled by performing whole-genome sequencing using the Illumina MiSeq platform. The sequence reads were trimmed with Trimmomatic software (v.0.39) and assembled with SPAdes software (v.3.15.4). Multilocus sequence typing (MLST) was performed to further analyze the VVE isolates, as described [8].
To investigate the potential reversion of VVE isolates to a vancomycin-resistant phenotype through vancomycin exposure, the VVE isolates were exposed to serial challenges in vitro with escalating vancomycin concentrations. Following culture on blood agar plates (BAPs), the colonies were cultured and adjusted to a McFarland (McF) standard of 0.5. Subsequently, 1 mL of each suspension was added to 30 mL tryptic soy broth (TSB) containing a vancomycin disk (30 μg), yielding an approximate concentration of 1 μg/mL. After 48 hrs of incubation in TSB with the vancomycin disk, the broth samples were spread onto BAPs to verify bacterial growth. Subsequently, each isolate was suspended in distilled water to an McF standard of 0.5 and exposed to increasing vancomycin concentrations by sequentially reducing the volume of TSB containing the vancomycin disk to 20 mL, 10 mL, and 5 mL, resulting in approximate concentrations of 1.5, 3, and 6 μg/mL, respectively. As a negative control, we included 10 VSE (vanA, vanB, vanC-negative E. faecium) isolates.
During the study period, 282 VSE isolates were collected from blood (N=50), urine (N=115), and other sample types (N= 117). Among them, 20 (7.1%) isolates were vanA positive, including those from blood (3/50, 6.0%), urine (8/115, 7.0%), and the other sample types (9/117, 7.7%).
All VVE isolates were resistant to ampicillin and erythromycin, and only one was resistant to linezolid. The MIC distributions for the VVE isolates were as follows: ≤0.5 µg/mL (1/20, 5.0%), 0.75 µg/mL (2/20, 10.0%), 1.0 µg/mL (12/20, 60.0%), and 1.5 µg/mL (5/20, 25.0%).
Our structural analysis results for Tn1546 are presented in Table 1. The most common change was a partial deletion in vanS with an IS1216 insertion (10 isolates), followed by a partial deletion in vanR with an IS1216 insertion (three isolates) and an IS1542 insertion (two isolates) or a partial deletion in vanH with an IS1216 insertion (two isolates).
MLST analysis showed that all 20 VVE isolates belonged to the CC17 complex (Table 1). The predominant ST was ST17 (8/20, 40.0%), followed by ST1421 (5/20, 25.0%), ST80 (5/20, 25.0%), ST787 (1/20, 5.0%), and ST981 (1/20, 5.0%). After vancomycin exposure, none of the 10 VSE isolates showed an increased vancomycin MIC, although 18 of the 20 VVE isolates reverted to the vancomycin-resistant phenotype. The MIC was >256 µg/mL for 17 isolates, and one isolate showed a vancomycin MIC of 16 µg/mL, as determined by the E-test.
To our knowledge, this is the first report showing the prevalence of VVE isolates from consecutively collected clinical specimens in Korea. A previous study conducted in Korea from 2004 to 2008 revealed a 7.4% (18/244) prevalence of VVE among VSE isolates obtained from fecal samples [9]. In a recent study conducted in India, five (1.5%) of 340 VSE isolates from clinical samples were identified as VVE, indicating a relatively lower prevalence rate when compared with our findings (7.0%) [10].
Although previously reported VVE strains commonly exhibited a complete deletion of vanRS, the first VVE isolates documented in Korea had deletions in both vanH and vanX [9]. Consistent with our results, recent studies revealed that VVE isolates harbored an insertion sequence in the vanHAX promoter region, which might have altered the vancomycin-resistant phenotype [1, 11].
In this study, all VVE isolates were resistant to ampicillin and erythromycin and susceptible to linezolid, except for one. Consistent with our results, most VVE isolates from Korea showed resistance to erythromycin (100%) and ampicillin (88.9%) [9]. In Canada, 29 VVE isolates collected between 2012 and 2016 were all resistant to ampicillin and susceptible to linezolid, except for one [12]. The three most frequent STs (ST17, ST1421, and ST80) belong to clade A, which is referred to as a hospital-associated clade [13], as ST15 among vancomycin-intermediate resistance Staphylococcus aureus [14]. Notably, the proportion of ST17 (40.0%) was markedly higher in this study than in a previous study conducted in Korea (11.1%) [9]. ST1421 was the second most common ST, which is noteworthy considering that despite its relatively recent emergence, it has become a dominant vanA ST among E. faecium in Australia [13] and that inter-hospital transmission of ST1421 was reported in Japan [15].
Most VVE isolates (90.0%) exhibited reversion to vancomycin resistance upon serial exposure to the antibiotic. This contrasts with a previous study by Jung et al. [9], which showed that only four of 18 VVE isolates (22.2%) reverted to VRE following exposure to glycopeptides. This discrepancy might be attributable to differences in the vancomycin exposure methods employed.
Vancomycin exposure provides the selective pressure required for secondary genetic rearrangements. Such alterations can eliminate an upstream transcription block, for example, via deletion of vanRS and/or vanHAX and the introduction of a nonregulated promoter, resulting in constitutive vancomycin resistance [5]. A recent analysis of the VVE structures before and after vancomycin exposure revealed that both isolates exhibited a 5′-truncated vanR activator sequence and that the VVE isolate that reverted to the resistant phenotype had an additional 44 bp deletion upstream of the vanHAX genes. The authors proposed that the alternative promoters in the vanHAX promoter region might have been responsible for the phenotypic reversion [1]. An alternative molecular mechanism responsible for phenotypic reversion with VVE isolates is the excision of the insertion sequence (ISL3-family element), which leads to the restoration of the vanHAX promoter region [11].
The ST1421 VVE isolates that converted to VRE upon vancomycin exposure exhibited more copies of the vanA plasmid than observed before exposure [16, 17]. Elevated copy numbers of vanA plasmids can potentially facilitate increased vanHAX expression from an alternative constitutive promoter, which could contribute to the resistant phenotype in the absence of functional vanRS. An increase in the copy number of a conjugate plasmid can potentially enhance its horizontal spread by increasing its transfer rate along with plasmid-encoded antibiotic resistance [18]. Although various factors can influence the vanA plasmid-transfer rate, in the case of ST1421 VVE (where vancomycin exposure results in an increased copy number of vanA plasmids), we postulate that this phenomenon could influence the dissemination of VRE within a hospital setting.
This study has a limitation in that we did not investigate van gene clusters other than vanA, vanB, and vanC. A recent report from China showed that VVE strains carried the vanM gene [19]. Future research is needed to explore the presence of various van gene clusters in VVE strains.
In summary, we observed a notable prevalence of VVE isolates at 7.1%, most of which belonged to three hospital-associated clades (ST17, ST1421, and ST80). Furthermore, most of these isolates converted to VRE following vancomycin exposure. These findings underscore the importance of accurately detecting VVE in clinical microbiology laboratories for appropriate treatment and effective infection control.
XML Download