Epidemiological profiles and pathogenicity of Vancomycin-resistant Enterococcus faecium clinical isolates in Taiwan

Collection, identification and antimicrobial susceptibility testing of VRE-fm clinical isolates

Sixty clinical isolates of VRE-fm were collected from urine (46), blood (7), pus (4), body fluid (2) or stool (1) from regional hospitals in northern Taiwan. All clinical isolates were cultured overnight on a blood agar plate (Creative Media Plate, New Taipei City, Taiwan) at 37 °C. A single colony of a clinical isolate was inoculated in tryptic soy broth (TSB) (Neogen, Lansing, MI, USA). The overnight culture was adjusted to McFarland 0.5 by using a densitometer. Identification and antimicrobial susceptibility testing were performed by using the BD Phoenix 100 automatic system (BD, Franklin Laker, NJ, USA). The Taipei City Hospital Institutional Review Board granted Ethical approval to carry out the study within its facilities (TCHIRB-10703123-E, Waiver of Informed Consent).

Multilocus sequence typing (MLST)

MLST is a sequence-based method for establishing the clonal relationship among bacteria (Bello Gonzalez et al., 2017). MLST of the VRE-fm clinical isolates was carried out according to previous reports (Homan et al., 2002; Kariyama et al., 2000). In brief, sequences of seven house-keeping genes specific to E. faecium, including gdh, purK, pstS, atpA, gyd, adk, ddl, were amplified by using PCR followed by electrophoresis analysis with 1.2% Agarose gel (100 V, 30 min). PCR products were purified by a commercial Kit (Favorgen, Ping-Tung, Taiwan) and sent for DNA sequencing (Mission Biotech Co., Taipei, Taiwan). The sequencing results were compared with published alleles, and sequence types (STs) were assigned using the MLST database for E. faecium (https://pubmlst.org/organisms/enterococcus-faecium/).

Virulence gene identification

Genomic DNA extraction of the VRE-fm isolates was performed based on a standard protocol (Kariyama et al., 2000). In brief, DNA was extracted from an overnight bacterial culture by heating at 95 °C for 5 min followed by centrifugation to remove the debris. All samples were subjected to amplification of the virulence genes (asa1, cylA, esp, gelE and hyl) by multiplex PCR and DNA electrophoresis as previously described (Vankerckhoven et al., 2004). The primer sequences were listed in Table S3. In brief, the PCR mixture was prepared in a total volume of 20 μl containing 2 μl of genomic DNA, 0.5 μl of forward (F) and reverse (R) primer (10 μM), 4 μl of 5X PCR Plus Master Mix II solution (Genemark, Taichung, Taiwan), and 9 μl of distilled water. Multiplex PCR was performed in a GeneAmp PCR System 2700 (Perkin-Elmer, Waltham, MA, USA.). The template was initially denatured at 95 °C for 5 min followed by 30 cycles at 94 °C for 1 min, 60 °C for 1 min, and 72 °C for 1 min. Final extension was set at 72 °C for 10 min. The PCR products were analyzed by 1.8% agarose gel electrophoresis and visualized by staining with fluorescent dye.

Biofilm measurements

Biofilm production of the VRE-fm isolates was measured according to a standard protocol with modification (Hashem, Abdelrahman & Aziz, 2021). A single colony of clinical isolates was inoculated in TSB supplemented with 1% sucrose and grown at 37 °C. Staphylococcus epidermidis strain ATCC35984 was used as a positive control (Manandhar et al., 2018). The overnight culture was adjusted to an optical density (OD) at 600 nm with 10⁸ cells/mL in a 96-well culture plate and incubated at 37 °C for 48 h. The culture plate was gently washed thrice with 0.1 ml phosphate-buffered saline (PBS) followed by aspiration and air drying for 30 min (Weng, Ramli & Hamat, 2019). 0.1 ml of 0.1% crystal violet was added to the wells for 15 min, followed by washing thrice with 0.1 ml PBS. 0.2 ml of 95% ethanol was added to the wells for 20 min to dissolve the biofilm-associated dye (Igbinosa & Beshiru, 2019). The optical density of the samples was determined at 570 nm by using a SpectraMax Max ELISA reader (Molecular Devices, San Jose, CA, USA). A negative control was employed to reduce the background absorbance OD values. Since different clinical isolates had varied growth rates, the resulting bacterial cell number would be different after incubation, leading to inaccurate measurement of the biofilm mass. In order to justify the bias due to cell number, the ability to form a biofilm was expressed by using a biofilm formation parameter: (OD₅₇₀ nm biofilm minus OD₅₇₀ nm control)/OD₆₀₀ nm (cells) (Lin, Lin & Lan, 2020). Optical density cut-off (ODc) was defined as three standard deviations above the mean OD of the negative control as described (Stepanovic et al., 2007). Each clinical isolate was classified as follows: no-biofilm detected (ND): OD ≤ ODc; weak biofilm producer: ODc < OD ≤ 2 × ODc; moderate biofilm producer: 2 × ODc < OD ≤ 4 × ODc; and strong biofilm producer: OD > 4 × ODc.

The C. elegans culture and survival assay

Wild-type C. elegans strain, N2, in their larval stage was cultured with non-pathogenic Escherichia coli OP50 as a food source. The C. elegans survival assay was performed based on a previous report (Chen et al., 2020). Briefly, staged young adult worms were transferred to Tryptic Soy Agar (TSA, Neogen, MI, USA) seeded with a bacterial lawn of Linezolid-resistant VRE-fm clinical isolates and incubated at 20 °C. E. coli OP50 were used as a control. Live or dead worms were scored every 24 h. Worms were censored if they crawled off the plate.

Statistical analysis

Statistics of virulence genes, antibiotic susceptibility and biofilm production were performed with a Chi-square test by using SPSS v. 21 software (SPSS Inc., Chicago, IL, USA). The C. elegans lifespan was analyzed by a log-rank test using GraphPad 8 software (GraphPad, San Diego, CA, USA). Statistical significance was indicated by a P value (*P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant P > 0.05).

Antibiotics susceptibility, molecular typing and virulence gene carriages of VRE-fm clinical isolates

VRE-fm clinical isolates were subjected to analysis of antibiotic susceptibility (Table S1) (Fig. 1). These isolates confer resistance to the most commonly used antibiotics, while being sensitive to a few antibiotics, including Daptomycin, Doxycycline, Linezolid and Quinupristin-dalfopristin. This indicates that a majority of VRE-fm isolates exhibited multi-drug resistance, while a few isolates tolerate Linezolid. Molecular typing based on MLST revealed that three major ST types in clinical isolates, including ST17, ST78 and ST262 (Table 1). The prevalence of ST17 and ST78 is consistent with the findings in a longitudinal study of VRE-fm blood isolates in Taiwan (Kuo et al., 2018).

Five virulence genes were tested, including asa1, gelE, cylA, esp and hyl to determine which enterococcal virulence factor is present in clinical isolates. asa1, gelE and cylA were absent in all isolates, whereas esp and hyl were detected (Table S2). These isolates can be further classified as esp⁺ (40.0%, n = 24), hyl⁺ (33.3%, n = 20), esp⁺/hyl⁺ (13.3%, n = 8) or esp⁻/hyl⁻ (13.3%, n = 8) (Table 2). Based on the isolation sites, VRE-fm carrying esp⁺ and/or hyl⁺ were isolated mainly from urine (76.6%), followed by blood (11.6%), pus (6.6%), body fluid (3.3%) and stool (1.6%). In urine samples, the percentage of esp⁺ isolates (39.1%) was higher than that of hyl⁺ strains (32.6%). The percentage of esp⁺/hyl⁺ isolates (15.2%) was higher than esp⁻/hyl⁻ (13.0%). Nearly all ST17 isolates (95.8%), encoded esp⁺ or hyl⁺ (54.2% vs 58.3%). In ST78 isolates, esp⁺ isolates comprised the major group (58.8%), whereas hyl⁺ isolates were the major group in ST262 isolates (68.4%) (Table 1).

The relationship between antibiotic resistance/susceptibility and VRE-fm isolates

Isolates carrying esp and/or hyl were resistant to antibiotics, including Ampicillin, Penicillin, Levofloxacin, Ciprofloxacin, Erythromycin, and Vancomycin. Some isolates carrying esp and/or hyl showed either resistance or susceptibility to Doxycyclin, Gentamycin-Synerge, Linezolid, Minocycline, Nitrofurantoin, Quinupristin-dalfopristin, Teicoplanin, and Tetracyclin (Table S2). The esp and/or hyl were predominant in isolates sensitive to Doxycyclin (68.3%), Quinupristin-dalfopristin (88.3%), Linezolid (90.0%) and Daptomycin (100%) (Table S1). These findings indicate that despite the high degree of the virulence gene carriage, there is a lack of a clear association between antibiotic susceptibility and esp/hyl genes in VRE-fm clinical isolates.

Biofilm production in VRE-fm isolates with an esp or hyl carriage

Nearly all isolates displayed moderate to strong biofilm production (96.6%, 58/60), while only two isolates showed weak biofilm production. Among the isolates with moderate to strong biofilm production, a majority of isolates carried either esp (42%) or hyl (42%). While both virulence genes were detected in some isolates (esp/hyl, 16%), the absence of both genes were found in a few isolates (esp⁻/hyl⁻, 16%). These results indicate that esp and/or hyl carriage are/is dominant in biofilm producing VRE-fm isolates.

The effect of VRE-fm clinical isolates on C. elegans survival

Since multi-drug resistant VRE-fm isolates confer resistance to many clinical antibiotics, this limits treatment options and requires Linezolid as a last resort. The emerging Linezolid-resistant VRE-fm is a pressing issue. To this end, Linezolid-resistant VRE-fm clinical isolates (D11, D12, B51, A68, A126) and a control strain (E. coli OP50) were examined in a C. elegans lifespan assay (Fig. 2). Two hyl⁺ strains (D11 and D12) reduced C. elegans lifespan (P < 0.001) (Fig. 2A) and survival at the 8^th day post infection (Fig. 2B). While the C. elegans lifespan was unaffected by a hyl⁺ strain (A68) (P = 0.086), this isolate increased C. elegans lethality at the 8^th day post infection (Fig. 2B). Two esp⁺ strains (B51 and A126) were mostly harmless to C. elegans.

A esp⁺/hyl⁺ strain with Linezolid susceptibility (A127) was used to determine whether or not Linezolid resistance in VRE-fm enhances virulence (Fig. 2C). This isolate did not affect C. elegans survival compared to the control (P = 0.88), suggesting that resistance to certain antibiotics, such as Linezolid, as well as the virulence gene, may be required for VRE-fm pathogenesis.