Dissemination of VanA-Type Enterococcus faecium Isolates in Hungary

Abstract

Although vanA carrying Enterococcus faecium human clinical isolates have been rarely found in Hungary before 2012, they have been detected in continuously increasing numbers since then. To identify factors associated with their dissemination, we investigated the clonal relatedness and plasmids of 30 vanA carrying E. faecium isolates originating from different Hungarian healthcare institutions from 2012 to 2014. Molecular typing of the isolates (n = 30) was performed with pulsed-field gel electrophoresis (PFGE), multilocus sequence typing, Tn1546 polymerase chain reaction mapping, plasmid restriction fragment length polymorphism analysis, and sequencing. A single Tn1546 variant was detected in all of the isolates. It harbored IS1251 in the vanS-vanH intergenic region, had an entire deletion of the transposase gene and a partial deletion of the resolvase gene, and was located on a pRUM-like plasmid. Based on PFGE, the isolates could be grouped into 13 pulsotypes. Representative strains of these pulsotypes belonged to ST17, ST18, ST80, ST117, and ST203, which are known to be part of the hospital-adapted clades. The increase in the number of vanA carrying E. faecium clinical isolates in Hungary could be explained by the dissemination of pRUM-like vancomycin resistance plasmids in hospital-adapted clonal lineages.

Introduction

Since the first appearance in the 1980s, vancomycin-resistant Enterococcus spp. (VRE) have spread worldwide with varying rates over time and per country.¹ This expansion is largely attributed to the emergence of particular hospital-adapted Enterococcus faecium clonal lineages and to their acquisition of vancomycin-resistant determinants (vanA and/or vanB). Despite this global dissemination, only sporadic human cases of VRE and one minor VanB-type E. faecium outbreak was reported from Hungary before 2012.² Since then a continuous rise in the rates of VRE was observed in our country. The rate of vancomycin resistance among E. faecium isolates responsible for invasive infections was 0.8% in 2011, and in 2014, it reached 8.5%.³

Preliminary studies performed at the National Center for Epidemiology indicated that 76.9% (n = 268) of the vancomycin-resistant E. faecium isolates, which were sent to the Reference Laboratory between 2012 and 2013, carried the vanA gene (B. Berta, 2013, personal communication). To identify the underlying factors associated with the change in resistance rates, we studied clonal relatedness and structure of the vanA operon and of the carrying plasmids of E. faecium strains isolated in Hungary between 2012 and 2014.

Materials and Methods

Antimicrobial susceptibility testing was performed in clinical microbiology laboratories of participating Hungarian healthcare institutions. Presumably, VRE isolates were submitted for confirmation and typing to the Reference Laboratory of the National Center for Epidemiology (renamed to National Public Health Institute in 2017). The identification of the isolates to species level and the detection of vanA, vanB, and vanC genes were performed with polymerase chain reaction (PCR) using primers based on the study by Dutka-Malen et al.⁴ For molecular typing, pulsed-field gel electrophoresis (PFGE) was carried out as described elsewhere.⁵ Pulsotypes and subtypes were defined as isolates with >80% and >95% similarity in banding patterns, respectively.

Thirty, nonduplicate vanA carrying E. faecium isolates were selected for further study. The selection was done as to be representative for the participating healthcare institutions, the time period, and the distribution of pulsotypes. Multilocus sequence typing (MLST) was performed for one isolate of each pulsotype and subtype.⁶ Bayesian Analysis of Population Structure (BAPS) groups were defined according to study by Willems et al.⁷ To detect Tn1546 structural variants, PCR mapping was carried out using primers based on the study by Arthur et al.⁸ The first VanA-type E. faecium isolate was selected for Sanger sequencing of Tn1546. Based on the sequencing results, additional primers (p20–p23, Table 1) were designed to test for the integration of the IS1251-like element in the vanS-vanH region for all isolates.

Table 1.

Primers Designed for This Study

Primer	Sequence	Position at pPEC286	Position at pBUD102	Position at pBUD607
p20	5′-GGATTTGGCGTTGGATTCCG-3′	26192 … 26211	19202 … 19221	—
p21	5′-GTCGTCTGATAGCCTTCGGG-3′	27756 … 27737	20766 … 20747	—
p22	5′-CAGTGAGAAAGTCCGCTGGT-3′	26987 … 27006	19997 … 20016	—
p23	5′-CTGGAAAAGCGAGAGCAGGA-3′	28778 … 28759	21788 … 21769	—
vp1	5′-TCTTGAAGAAAGACCGTGCG-3′	14775 … 14756	7785 … 7766	—
vp2	5′-CAGTCGTCGCCTGGGTTATC-3′	6293 … 6312	6293 … 6312	—
vp3	5′-ACGTCGCAGAAACAACAAGAC-3′	31474 … 31454	24484 … 24464	—
vp4	5′-GCATCGCTAGGAGGCTGTTA-3′	30005 … 30024	23015 … 23034	—
vp5	5′-ACGAGAACAACTTGCCCTCAA-3′	13679 … 13699	—	3429 … 3449
vp6	5′-TCATTTCCATAGCTGCCACCA-3′	7740 … 7720	—	3643 … 3663
vp7	5′-GCTCTTCGTAGCTCGTTGGT-3′	10489 … 10470	—	239 … 220
vp8	5′-TGAACGAGGGAAGAAACCCG-3′	10154 … 10173	—	6077 … 6096
vp9a	5′- TCGTTCAGTCCAAGCCAACT-3′	—	—	—
vp9b	5′-CGGAGCTACGTCAAACAACC-3′	—	—	—
vp10	5′-AGAGCATAAGGAAACGTGGGA-3′	—	—	—
vp11a	5′-CATTACACCTACCCCTGCGA-3′	—	—	—
vp11b	5′-ATAAGCCCATAGGCACCACG-3′	—	—	—
vp12a	5′-CATCTGTGGTATGGCGGGTA-3′	—	—	—
vp12b	5′-CCAGCGGAATGCTTTCATCC-3′	—	—	—
vp13	5′-AGTTTCTGACCCCCGATAGA-3′	—	—	—
vp14	5′-AACAGACCAGAGCTGCAAGA-3′	—	—	—
vp15	5′-GCTGTCCCTTCTGTCCCATC-3′	—	—	—
vp16	5′-AACAAGTTTTCCGTGGTGCG-3′	—	—	—
vp17a	5′-CCATGTCGTTATTTCCGGCG-3′	—	—	—
vp17b	5′-TGCATTTCTAACCGCCACCA-3′	—	—	—
vp18a	5′-CGGCAGACGATTTGAAACGA-3′	—	—	—
vp18b	5′-GGGGTAGCTGAAAGTAACGG-3′	—	—	—

Three isolates from different years, regions, and sequence types were selected for plasmid sequencing. Isolates were grown in a brain-heart infusion broth containing 10 mg/L vancomycin, and plasmid DNA was extracted with alkaline lysis method and purified with phenol–chloroform extraction.⁹ Sequencing was performed on Illumina MiSeq with 2 × 250 bp paired ends. The quality of the data was assessed with FastQC 0.11.5 (Braham Institute, Cambridge, United Kingdom). PlasmidSPAdes 3.8.1 was used for de novo assembly.¹⁰ De Bruijn graphs were visualized with Bandage 0.8.0.¹¹ Paths corresponding to plasmids were chosen according to depth and comparison to plasmid sequences in GenBank. The selection of the paths (order of the contigs) was verified with PCR (primers vp1–vp18b; Table 1) and restriction digestion of plasmid DNA with Bsu15I (ClaI).

Plasmids were annotated with GeneMark (Georgia Institute of Technology, Atlanta, GA), and their sequences were uploaded to GenBank (accession Nos. KY595962–KY595970).¹² Plasmid maps were created with SnapGeneViewer (GSL Biotech, LLC, Chicago, IL). Digestion of plasmid DNA with Bsu15I (ClaI) was performed for all isolates (plasmid restriction fragment length polymorphism [pRFLP]) and the results were evaluated for bands between 1,000 and 10,000 bp. Those isolates were considered to belong to pRFLP type A1, for which the characteristic restriction fragment pattern of pBUD102/pMIS102 (5,144, 4,717, 4,474, 3,462, 3,312, 2,361, 2,009, and 1,978 bp) was shown. Isolates with one to two band differences were designated type A2–A4. Isolates that differed in more than two bands were defined as type B–D. The presence of the pRUM replicon [rep(17)-F and rep(17)-R] and the axe-txe genes (Axe-Txe-F and Axe-Txe-R) were detected with PCR using primers based on the study by Freitas et al.¹³

Results

Between 2012 and 2014, vanA carrying E. faecium was confirmed for 68% of the patients (n = 437), from whom presumably VRE isolates were submitted to the Reference Laboratory (Table 2). According to the results of PFGE, the vanA carrying E. faecium isolates were grouped into 77 pulsotypes, of which the most frequent one (designated ENTCO-002/a/c) enclosed 36% (n = 332) of the isolates and it could be subdivided into several subtypes (designated ENTCO-002, ENTCO-002a, and ENTCO-002c). Other pulsotypes represented a subtype at the same time as the similarity level was found to be >95% within those pulsotypes.

Table 2.

Confirmation and Typing of vanA Carrying Enterococcus faecium Isolates at the Reference Laboratory

	Submitted presumably VRE isolates	Confirmed vanA carrying E. faecium isolates	vanA Carrying E. faecium isolates selected for further study
No. of patients (healthcare institutions) per year
2012	71 (17)	49 (13)	5 (2)
2013	169 (28)	130 (24)	10 (8)
2014	197 (33)	120 (28)	15 (10)
No. of isolates per pulsotypes
ENTCO-002	N/A	63	2
ENTCO-002c	N/A	56	4
ENTCO-035	N/A	19	3
ENTCO-008	N/A	15	3
ENTCO-057	N/A	11	2
ENTCO-027	N/A	10	2
ENTCO-033	N/A	10	2
ENTCO-005	N/A	8	1
ENTCO-024	N/A	8	1
ENTCO-041	N/A	8	2
ENTCO-032	N/A	6	1
ENTCO-008a	N/A	5	0
ENTCO-026	N/A	5	0
ENTCO-037	N/A	5	0
ENTCO-054	N/A	5	1
ENTCO-099	N/A	5	0
Pulsotypes enclosing two to four isolates (n = 21)	N/A	53	3
Pulsotypes enclosing one isolate (n = 40)	N/A	40	3

N/A, not applicable; VRE, vancomycin-resistant Enterococcus spp.

The results of MLST, Tn1546 typing and pRFLP are summarized in Table 3. According to the results of MLST, the isolates belonged to either BAPS2.1a (ST80 n = 1, ST117 n = 7, and ST203 n = 4) or BAPS3.3a (ST17 n = 2 and ST18 n = 2). The BAPS2.1a and BAPS3.3a isolates belonged to pulsotypes that included altogether 73% (22/30) and 27% (8/30) of the isolates, respectively.

Table 3.

Results of Molecular Typing of Selected vanA Carrying Enterococcus faecium Isolates

No.	Year	County	Specimen	PFGE	MLST	BAPS	Tn1546 type^a	pRFLP
1	2012	Baranya	Wound	ENTCO-005	ST117	BAPS2.1a	F3	A1
3	2012	Baranya	Urine	ENTCO-002	ST203	BAPS2.1a	F3	A1
11	2012	Baranya	Blood	ENTCO-008	ST17	BAPS3.3a	F3	A2
12	2012	Baranya	Rectal	ENTCO-002a	ST203	BAPS2.1a	F3	A1
24	2013	Baranya	Urine	ENTCO-155	ST117	BAPS2.1a	F3	A1
40	2013	Baranya	Blood	ENTCO-039	ST117	BAPS2.1a	F3	A3
50	2013	Baranya	Blood	ENTCO-146	ST117	BAPS2.1a	F3	A3
79	2012	Pest	Trachea	ENTCO-002c	ST203	BAPS2.1a	F3	A1
80	2013	Fejér	Urine	ENTCO-002c	NA	NA	F3	A1
81	2013	Pest	Blood	ENTCO-027	ST203	BAPS2.1a	F3	A1
82	2013	Csongrád	Abscess	ENTCO-024	ST17	BAPS3.3a	F3	A1
83	2013	Pest	Urine	ENTCO-002c	NA	NA	F3	A1
84	2013	Pest	Urine	ENTCO-033	ST18	BAPS3.3a	F3	B
85	2013	Zala	Urine	ENTCO-041	NA	NA	F3.a	A2a
86	2013	Pest	Wound	ENTCO-002c	NA	NA	F3	A1
87	2014	Borsod-Abaúj-Zemplén	Urine	ENTCO-041	ST18	BAPS3.3a	F3	A1
88	2014	Pest	Bile	ENTCO-027	NA	NA	F3	A2
89	2014	Nógrád	Urine	ENTCO-008	NA	NA	F3	A4
90	2014	Bács-Kiskun	Blood	ENTCO-032	ST117	BAPS2.1a	F3	A1
91	2014	Pest	Wound	ENTCO-033	NA	NA	F3	A1
92	2014	Pest	Urine	ENTCO-054	ST117	BAPS2.1a	F3	A2
93	2014	Pest	Urine	ENTCO-057	ST80	BAPS2.1a	F3	A2
94	2014	Pest	Wound	ENTCO-057	NA	NA	F3	A2
95	2014	Pest	Blood	ENTCO-008	NA	NA	F3	A4
96	2014	Pest	Blood	ENTCO-035	ST117	BAPS2.1a	F3.b	C
97	2014	Csongrád	Abdominal fluid	ENTCO-035	NA	NA	F3	D
98	2014	Pest	Throat	ENTCO-035	NA	NA	F3	A2
121	2014	Baranya	Blood	ENTCO-146	NA	NA	F3	A3
128	2014	Baranya	Abscess	ENTCO-002	NA	NA	F3.a	A2a
132	2014	Baranya	Bronchus	ENTCO-146	NA	NA	F3.c	A2b

Type F3 designated by Freitas et al.¹³

MLST, multilocus sequence typing; NA, not analyzed; PFGE, pulsed-field gel electrophoresis; pRFLP, plasmid restriction fragment length polymorphism.

Typing of Tn1546 revealed that 26 isolates shared an identical variant with an ∼1,500 bp increase of p11p12 and a lack of the first 4 segments (p1p2-p7p8) (Table 4). Sanger sequencing showed integration of a copy of IS1251-like element in the vanS-vanH intergenic region (corresponding to additional DNA in p11p12), and total deletion of orf1 (transposase) and partial deletion of orf2 (resolvase), consistent with the absence of fragments p1p2-p7p8. This structure corresponded to “type F3” designated by Freitas et al. and detected in neighboring country (Serbia) in 2005.¹³ Four isolates showed additional alterations, and their transposons were designated here subtypes F3.a-c. With primers p20–p23, the integration of the IS1251-like element in the vanS-vanH intergenic region was confirmed in all isolates.

Table 4.

Tn1546 Variants Identified by Polymerase Chain Reaction Mapping


			p1p2 ^a	p3p4 ^a	p5p6 ^a	p7p8 ^a	p9p10 ^a	p11p12 ^a	p13p14 ^a	p15p16 ^a	p17p18 ^a	p19p1 ^a	p20p21 ^b	p21p22 ^b	p22p23 ^b
			IR_L-tnp	tnp	tnp-res	tnp-vanR	res-vanS	vanS-vanH	vanH-vanX	vanA-vanX	vanY-vanZ	vanZ-IR_R	vanH-IS1251	IS1251	vanS-IS1251
Tn1546 type	No. of isolates	Year of isolation	1,309 bp	1,132 bp	1,299 bp	1,274 bp	1,225 bp	1,678 bp	1,793 bp	1,942 bp	1,585 bp	428 bp	1,565 bp	770 bp	1,792 bp
F3	26	2012–2014	−	−	−	−	+	++	+	+	+	+	+	+	+
F3.a	2	2013–2014	−	−	−	−	++	++	+	+	+	+	+	+	+
F3.b	1	2014	−	−	−	−	+	++	+	+	−	−	+	+	+
F3.c	1	2014	−	−	−	−	+	++	+	++	+	+	+	+	+

Size of the PCR product and the region of Tn1546 amplified was indicated for strain BM4147.⁸

Size of the PCR product and the region of Tn1546 amplified was indicated for the isolate for which Sanger sequencing of Tn1546 was performed.

−, No PCR product was detected.

+, The size of the PCR product matched the indicated size.

++, The size of the PCR product was larger than the indicated size.

PCR, polymerase chain reaction.

The results of sequencing and assembly of plasmids are summarized in Table 5. The sequence quality (per base, per tile, and per sequence) of the short reads was good. The report indicated overrepresentation of certain sequences (warnings for “per sequence GC content,” “sequence duplication level,” and “overrepresented sequences”). Overrepresented sequences totaled upto <0.15% of total sequences and by BLAST search, they gave match with different E. faecium plasmid sequences (replication module protein, insertion sequence, hypothetical protein and noncoding sequence). The overrepresentation of certain sequences could be explained with the large variations in the copy number of plasmids.

Table 5.

Results of Sequencing and De Novo Assembly of Plasmids

Isolate	1	81	87
Sequencing
No. of short reads	304,553	306,973	855,883
Sequences flagged as poor quality	0	0	0
Sequence length	35–251	35–251	35–251
% GC	36	37	37
De novo assembly
No. of nodes (≥50 × depth)^a	30	18	16
No. of edges (≥50 × depth)^a	39	25	21
Total length	141,081	69,350	77,409
N50	12,400	11,157	11,153
Median depth	108.8 ×	766.6 ×	1,146.6 ×
Scaffolding^b
vp1–vp2	0 (+)	∼1,493 bp (+)	∼1,493 bp (+)
vp3–vp4	∼1,470 bp (+)	∼1,470 bp (+)	∼1,470 bp (+)
vp5–vp6	0 (+)	∼235 bp (+)	∼235 bp (+)
vp7–vp8	∼336 bp (+)	∼1,153 bp (+)	∼336 bp (+)
vp1–vp5	∼1,097 bp (+)	0 (+)	0 (+)
vp2–vp6	∼1,448 bp (+)	0 (+)	0 (+)
vp9a–vp10	0 (−)	∼300 (−)	0 (+)
vp9b–vp10	0 (+)	0 (−)	0 (−)
vp11a–vp12a	∼1,885 bp (+)	0 (+)	0 (+)
vp11b–vp12b	0 (+)	0 (−)	0 (−)
vp13–vp14	∼300 (−)	∼1,222 bp (+)	∼1,244 bp (+)
vp15–vp16	0 (+)	∼800 bp (−)	0 (+)
vp17a–vp18a	0 (+)	0 (+)	0 (−)
vp17b–vp18b	0 (+)	∼300 (−)	∼300 (−)
vp17a–vp18b	0 (+)	0 (+)	0 (−)
vp17b–vp18a	0 (+)	0 (+)	0 (−)
Closed plasmids^c Name accession Size GC content Best match	pPEC286 KY59562 36124 bp 35.3% pJEG040 (KX810025) c = 78% id = 99%	pBUD102 KY595966 29,134 bp 35.2% Plasmid unnamed3 (CP014452) c = 94%, id = 99%	pMIS102 KY595966 29,134 bp 35.2% Plasmid unnamed 3 (CP014452) c = 94%, id = 99%
Name accession	pPEC011	pBUD607	pMIS005
Size	KY595963	KY595967	KY595969
GC content	4,786 bp	6,990 bp	7,672 bp
Best match	32.8%	35.6%	33.4%
	Plasmid unnamed2 (CP011830)	pB82 (AB178871)	pE1_13 (CP018068)
	c = 64%, id = 99%	c = 88%, id = 100%	c = 77%, id = 99%
Name accession	pPEC012		pMIS006
Size	KY595964		KY595970
GC content	4,465 bp		6,173 bp
Best match	32.0%		35.5%
	pHY (AB570326)		pB82 (AB178871)
	c = 49%, id = 94%		c = 100%, id = 100%
Name accession	pPEC013
Size	KY595965
GC content	3,221 bp
Best match	34.2%
	pE1_3 (CP018069)
	c = 100%, id = 100%

For isolate 87, nodes with depth ≥100 × were given.

Match between in vitro and in silico results is indicated with (+). Discrepancy between in vitro and in silico results is indicated with (−).

The name (accession number) and statistics (“c” coverage, “id” identity) of the plasmid with the highest maximum score in BLAST search are indicated.

GC, guanine-cytosine content.

After removing dead ends and nodes with low depth (<50 × for isolate 1 and 81; <100 × for isolate 87) from the assembly graph, several nodes (contigs) interconnected through a central node (corresponding to IS1216) were visible for each of the isolates (Figs. 1 –3).

FIG. 1.

Assembly graph of isolate 1 showing ≥50 × depth contigs and binding sites of scaffolding primers. Nodes are represented by black thick lines (contigs assigned to plasmids) and gray thick lines (contigs not assigned to plasmids). Edges indicate overlaps between node sequences and are shown with black thin lines.

FIG. 2.

Assembly graph of isolate 81 showing ≥50 × depth contigs and binding sites of scaffolding primers. Nodes are represented by black thick lines (contigs assigned to plasmids) and gray thick lines (contigs not assigned to plasmids). Edges indicate overlaps between node sequences and are shown with black thin lines.

FIG. 3.

Assembly graph of isolate 87 showing ≥100 × depth contigs and binding sites of scaffolding primers. Nodes are represented by black thick lines (contigs assigned to plasmids) and gray thick lines (contigs not assigned to plasmids). Edges indicate overlaps between node sequences and are shown with black thin lines.

Altogether, 14 paths predicted to be distinct plasmids were selected. Scaffolding primers vp1–vp18b were designed to bind to the end of contigs (Figs. 1 –3). For nine paths, the in vitro results of the scaffolding PCR matched the in silico predicted one (Table 5). The contigs of these closed plasmids are shown with black thick lines in Figs. 1 –3. Five paths could not be circularized and their contigs (gray thick lines on Figs. 1 –3) were not considered for further analysis. For all closed plasmids, the observed size of restriction fragments (Fig. 4) matched the in silico predicted one. This way the sequence of four, two, and three plasmids could be verified for isolate 1, 81, and 87, respectively (Table 5).

FIG. 4.

In vitro (left) and in silico (right) digestion of plasmids with Bsu15I (ClaI). Calculated fragment sizes (bp) and parent plasmids are indicated.

For the 9 plasmids, altogether 109 open reading frames and 61 unique proteins were predicted (Table 6). Maps of the vancomycin resistance plasmids pPEC286 and pBUD102 are shown on Figs. 5 and 6. Plasmids pPEC286, pBUD102, and pMIS102 harbored the Tn1546 variant-type F3 and shared 29,134 bp with each other (identity: 100%). The best match retrieved with BLAST search for this shared sequence was unnamed plasmid 3 of strain ATCC700221, which is a vancomycin resistance plasmid with a pRUM backbone. In addition, pPEC286 contained the complete sequence of the bacteriocin encoding plasmid pB82. The sequence of pB82 was also identified in the other two isolates on plasmids pBUD607 and pMIS006. Plasmids pPEC011, pPEC012, pPEC013, and pMIS005 encoded proteins with unknown functions and belonged to the Rep_3 family.

FIG. 5.

Structure of pPEC286. (IS elements—black; genes from pRUM—dark gray; genes from p3 of ATCC700221—light gray; genes from pB82—white; restriction sites and binding sites of scaffolding primers vp1–vp8 are also indicated.) IS, insertion sequence.

FIG. 6.

Structure of pBUD102. (IS elements—black; genes from pRUM—dark gray; genes from p3 of ATCC700221—light gray; restriction sites and binding sites of scaffolding primers vp1–vp8 are also indicated.)

Table 6.

Open Reading Frames of Plasmids

			Location
Plasmid	No.	Gene	Strand	5′	3′	Size (bp)	Protein size (aa)	Best match in NCBI protein database (identity)	Description
pPEC286	1	prgN	c	1	291	291	96	WP_000026576 (100%)	Replication control protein
	2	orf2	c	603	980	378	125	WP_079999421 (100%)	Hypothetical protein
	3	uvrA	c	946	2271	1,326	441	WP_002311906 (100%)	UV resistance protein
	4	sin		2784	3335	552	183	WP_010782821 (100%)	Recombinase
	5	orf5	c	3500	4324	825	274	WP_050540778 (100%)	Hypothetical protein
	6	txe	c	4692	5219	258	85	WP_000588503 (100%)	Toxin component of TA system
	7	axe	c	5212	5481	270	89	EEV43771 (100%)	Antitoxin component of TA system
	8	orf8		5823	6125	303	100	WP_002317170 (100%)	Hypothetical protein
	9	orf9		6168	6386	219	72	WP_011117471 (100%)	Hypothetical protein
	10	tnp (IS1216)		6808	7494	687	228	WP_014748744 (100%)	Transposase
	11	orf11		7755	7925	171	56	WP_085843745 (100%)	Hypothetical protein
	12	bacA		8380	8604	225	74	WP_002363584 (100%)	Bacteriocin precursor
	13	bacB		8624	8911	288	95	WP_002363589 (100%)	Bacteriocin immunity protein
	14	mobC		9361	9741	381	126	WP_010782660 (100%)	Mobilization protein
	15	mobA		9723	10637	915	304	WP_010782659 (100%)	Mobilization protein
	16	orf16		10429	10875	447	148	BAF36630 (100%)	Hypothetical protein
	17	orf17		10882	11571	690	229	WP_002369882 (100%)	Hypothetical protein
	18	repA		12415	13359	945	314	WP_002303559 (100%)	Replication protein
	19	repB		13380	13729	350	116	WP_002363576 (100%)	Replication protein
	20	tnp (IS1216)		13798	14484	687	228	WP_014748744 (100%)	Transposase
	21	orf21	c	14507	15985	1479	492	WP_000139222 (100%)	Hypothetical protein
	22	corA	c	16323	17711	1389	462	WP_000443455 (100%)	Mg/Ni/Co transporter
	23	int	c	17775	18728	954	317	WP_000222572 (100%)	Integrase
	24	orf24	c	18811	18942	132	43	EJX39873 (100%)	Hypothetical protein
	25	res		19371	19934	564	187	WP_000017403 (100%)	Resolvase
	26	relE	c	20213	20500	288	95	WP_000312841 (100%)	Toxin component of TA system
	27	relB	c	20490	20819	330	109	WP_000629052 (100%)	Antitoxin component of TA system
	28	cadE		21124	21453	330	109	WP_068561828 (100%)	Cadmium efflux system accessory protein
	29	vanZ	c	21740	22225	486	161	AOO95698 (100%)	Teicoplanin resistance protein
	30	vanY	c	22378	23289	912	303	WP_000732308 (100%)	Carboxypeptidase
	31	vanX	c	23717	24325	609	202	EEV43759 (100%)	d-Ala-d-Ala dipeptidase
	32	vanA	c	24331	25362	1032	343	WP_001079843 (100%)	d-Ala-d-Lac ligase
	33	vanH	c	25355	26323	969	322	WP_001059542 (100%)	Pyruvate dehydrogenase
	34	vanS	c	28044	29198	1155	384	WP_002305818 (99%)	Histidine kinase
	35	vanR	c	29176	29871	696	231	WP_001280781 (100%)	Regulator
	36	tnp (ISL3)		31357	32658	1302	433	WP_002302440 (100%)	Transposase
	37	repA	c	33267	34307	1041	346	WP_000997689 (100%)	Replication protein
	38	orf41	c	34656	34982	327	108	EOM22171 (100%)	Hypothetical protein
	39	soj	c	34969	35772	804	267	WP_002311905 (100%)	Partitioning protein
pPEC011	1	repB		570	1538	969	322	WP_002322956 (100%)	Replication protein
	2	orf2		1548	2018	471	156	WP_002322957 (100%)	Hypothetical protein
	3	orf3		2034	2204	171	56	EHM33532 (100%)	Hypothetical protein
	4	orf4		2668	3768	1101	366	WP_044383358 (100%)	Peptidase C47
	5	orf5		3762	4184	423	140	EEW64221 (100%)	Hypothetical protein
pPEC012	1	orf1	c	663	1097	435	144	WP_002351079 (100%)	Hypothetical protein
	2	orf2	c	1502	1675	174	57	EOG20234 (100%)	Hypothetical protein
	3	orf3	c	1752	1919	168	55	BAF44074 (88%)	Hypothetical protein
	4	orf4	c	2360	2899	540	179	WP_002351084 (100%)	Hypothetical protein
	5	repB	c	2915	3838	924	307	WP_002351085 (100%)	Replication protein
pPEC013	1	orf1		2	91	90	29	EOI68470 (100%)	Hypothetical protein
	2	repB		84	803	720	239	WP_002311815 (100%)	Replication protein
	3	orf3	c	940	1356	417	138	WP_001232859 (100%)	Hypothetical protein
	4	orf4	c	1359	1643	285	94	WP_000353844 (100%)	Hypothetical protein
	5	orf5		1988	2233	246	81	WP_002311817 (99%)	Hypothetical protein
pBUD102/pMIS102	1	prgN	c	1	291	291	96	WP_000026576 (100%)	Replication control protein
	2	orf2	c	603	980	378	125	WP_079999421 (100%)	Hypothetical protein
	3	uvrA	c	946	2271	1326	441	WP_002311906 (100%)	UV resistance protein
	4	sin		2784	3335	552	183	WP_010782821 (100%)	Recombinase
	5	orf5	c	3500	4324	825	274	WP_050540778 (100%)	Hypothetical protein
	6	txe	c	4692	5219	258	85	WP_000588503 (100%)	Toxin component of TA system
	7	axe	c	5212	5481	270	89	EEV43771 (100%)	Antitoxin component of TA system
	8	orf8		5823	6125	303	100	WP_002317170 (100%)	Hypothetical protein
	9	orf9		6168	6386	219	72	WP_011117471 (100%)	Hypothetical protein
	10	tnp (IS1216)		6808	7494	687	228	WP_014748744 (100%)	Transposase
	11	orf11	c	7517	8995	1479	492	WP_000139222 (100%)	Hypothetical protein
	12	corA	c	9333	10721	1389	462	WP_000443455 (100%)	Mg/Ni/Co transporter
	13	int	c	10785	11738	954	317	WP_000222572 (100%)	Integrase
	14	orf14	c	11821	11952	132	43	EJX39873 (100%)	Hypothetical protein
	15	res		12381	12944	564	187	WP_000017403 (100%)	Resolvase
	16	relE	c	13223	13510	288	95	WP_000312841 (100%)	Toxin component of TA system
	17	relB	c	13500	13829	330	109	WP_000629052 (100%)	Antitoxin component of TA system
	18	cadE		14134	14463	330	109	WP_068561828 (100%)	Cadmium efflux system accessory protein
	19	vanZ	c	14750	15235	486	161	AOO95698 (100%)	Teicoplanin resistance protein
	20	vanY	c	15388	16299	912	303	WP_000732308 (100%)	Carboxypeptidase
	21	vanX	c	16727	17335	609	202	EEV43759 (100%)	d-Ala-d-Ala dipeptidase
	22	vanA	c	17341	18372	1032	343	WP_001079843 (100%)	d-Ala-d-Lac ligase
	23	vanH	c	18365	19333	969	322	WP_001059542 (100%)	Pyruvate dehydrogenase
	24	vanS	c	21054	22208	1155	384	WP_002305818 (99%)	Histidine kinase
	25	vanR	c	22186	22881	696	231	WP_001280781 (100%)	Regulator
	26	tnp (ISL3)		24367	25668	1302	433	WP_002302440 (100%)	Transposase
	27	repA	c	26277	27317	1041	346	WP_000997689 (100%)	Replication protein
	28	orf31	c	27666	27992	327	108	EOM22171 (100%)	Hypothetical protein
	29	soj	c	27979	28782	804	267	WP_002311905 (100%)	Partitioning protein
pBUD607	1	orf1		179	625	447	148	BAF36630 (100%)	Hypothetical protein
	2	orf2		632	1321	690	229	WP_002369882 (100%)	Hypothetical protein
	3	repA		2165	3109	945	314	WP_002303559 (100%)	Replication protein
	4	repB		3130	3479	350	116	WP_002363576 (100%)	Replication protein
	5	orf5		3678	3848	171	56	WP_085843745 (100%)	Hypothetical protein
	6	bacA		4303	4527	225	74	WP_002363584 (100%)	Bacteriocin precursor
	7	bacB		4547	4834	288	95	WP_002363589 (100%)	Bacteriocin immunity protein
	8	mobC		5284	5664	381	126	WP_010782660 (100%)	Mobilization protein
	9	tnp (IS1216)		6250	6936	687	228	WP_014748744 (100%)	Transposase
pMIS005	1	orf1	c	502	924	423	140	WP_002368072 (99%)	Hypothetical protein
	2	orf2	c	918	2018	1101	366	WP_002368070 (100%)	Peptidase C47
	3	tnp (ISEfa4)		2253	2657	405	134	WP_002287522 (100%)	Transposase
	4	tnp		2674	3822	1149	382	WP_012274918 (100%)	Transposase
	5	orf5	c	4579	4995	417	138	WP_002368067 (100%)	Hypothetical protein
	6	orf6	c	4997	5278	282	93	WP_002368064 (100%)	Hypothetical protein
	7	orf7	c	5574	6062	489	162	WP_060763418 (100%)	Hypothetical protein
	8	repB	c	6072	7040	969	322	WP_002368293 (100%)	Replication protein
pMIS006	1	bacB	c	443	730	288	95	WP_002363589 (100%)	Bacteriocin immunity protein
	2	bacA	c	750	974	225	74	WP_002363584 (100%)	Bacteriocin precursor
	3	orf3	c	1429	1599	171	56	WP_085843745 (100%)	Hypothetical protein
	4	repB	c	1614	2147	534	177	WP_002363576 (100%)	Replication protein
	5	repA	c	2168	3112	945	314	WP_002303559 (100%)	Replication protein
	6	orf6	c	3956	4645	690	229	WP_002369882 (100%)	Hypothetical protein
	7	orf7	c	4652	5098	447	148	BAF36630 (100%)	Hypothetical protein
	8	mobA	c	4890	5804	915	304	WP_010782659 (100%)	Mobilization protein
	9	mobC	c	5786	6166	381	126	WP_010782660 (100%)	Mobilization protein

c, Complementary strand; TA, toxin-antitoxin; UV, ultraviolet.

According to the results of pRFLP, the characteristic restriction fragment pattern of pBUD102/pMIS102 was shown for 13 isolates (type A1, Figs. 7 and 8). For 14 isolates, additional variations of type A1 were detected (type A2–A4). Types A2, A2a, and A2b had alteration of the 3,462 and 3,312 bp fragments. This could correspond to alterations in the repA restriction site. In type A2a, the 4,474 bp fragment carrying the vanH, vanS, and vanR genes seemed to be enlarged, which corresponds to the increment in p9p10 that was observed in Tn1546 PCR mapping (type F3a). In type A2b, an increment of the 4,717 bp fragment (relBE, vanZYX) was detected, which is in line with the enlargement of p15p16 (Tn1546 type F3.c). For type A3 and A4, the lack of 4,717 and 5,144 bp fragments was characteristic, respectively. Three isolates had unique pRFLP patterns (type B–D).

FIG. 7.

Restriction fragment analysis of plasmids with Bsu15I (ClaI) and pRFLP types of isolates 1–87. A schematic drawing of pRFLP types A1–A4 is shown on the right. pRFLP, plasmid restriction fragment length polymorphism.

FIG. 8.

Restriction fragment analysis of plasmids with Bsu15I (ClaI) and pRFLP types of isolates 88–132. A schematic drawing of pRFLP types A1–A4 is shown on the right.

With primers rep(17)-F–rep(17)-R and Axe-Txe-F–Axe-Txe-R, the presence of the pRUM replicon and the Axe-Txe toxin-antitoxin system were shown for all isolates.

Discussion

To our knowledge, this was the first study to analyze the Tn1546 structure and vanA carrying plasmids of E. faecium human clinical isolates in Hungary. The results indicate that the widespread dissemination of VanA-type E. faecium in our country after 2012 can be attributed to the spread of pRUM-like plasmids in hospital-adapted clonal lineages (BAPS 2.1a and BAPS3.3a). These plasmids carried a Tn1546 variant “F” with insertion of IS1251 in the vanS-vanH intergenic region.

The pRUM plasmids belong to the RepA_N family, have a narrow host range, and show differences in their structure (replicon sequence, Tn1546 structure, and absence of Axe-Txe toxin-antitoxin system).¹³ The structure of the vancomycin resistance plasmids (pPEC286, pBUD102, and pMIS102) identified in this study highly resembled the pRUM variant, which was originally widespread in the United States and was later found worldwide, including a neighboring country, Serbia.^13,14

Although the aim of the plasmid next-generation sequencing was to identify the vancomycin resistance plasmids, additional small-sized Rep_3 plasmids (pBUD607, pMIS005, pMIS006, pPEC011, pPEC012, and pPEC013) could be assembled in each of the isolates. However plasmids belonging to Rep_3 family were shown to be prevalent in hospital-adapted E. faecium clonal lineages, their exact contribution to the dissemination has not yet been fully elucidated.¹⁵ Besides, other plasmids might have remained unassembled, considering that several contigs could not be assigned to plasmids (Figs. 1 –3) and multiple bands in pRFLP could not be associated with the assembled plasmids.

Although the plasmids of only three isolates were sequenced, we presume that the majority of the isolates also carried pRUM-like vancomycin resistance plasmids. This is supported by the following findings: (1) the replicon and the Axe-Txe toxin-antitoxin system of pRUM were shown with PCR for all isolates. (2) All isolates had the Tn1546 variant with insertion of IS1251 in the vanS-vanH intergenic region, and this variant was mostly identified on pRUM-like plasmids.¹³ (3) The exact pRFLP pattern of pBUD102/pMIS102 (type A1, 13/30) or a pattern with slight differences (type A2–A4, 14/30) could be recognized altogether in 90% of the isolates. In the remaining isolates, the variations in the patterns could be explained by alteration through mutation or recombination; integration into the chromosome; or presence of different vancomycin resistance plasmids.

Several clonal lineages are known to contribute to the spread of VanA-type E. faecium isolates worldwide.¹³ In Hungary, the BAPS3.3a group was detected in sporadic human isolates between 2004–2005.² In this study, most of the isolates belonged to BAPS2.1a and only a smaller portion clustered within BAPS3.3a. However, MLST was performed for a limited number of isolates; we presume that this distribution should highly resemble the sequence type and BAPS group distribution of all isolates, considering that (1) PFGE, MLST, and whole-genome sequencing (WGS) were shown to be highly concordant and PFGE at 88% similarity level supported the WGS subgrouping,¹⁶ and (2) that in our study, the selection criteria for MLST were based on >95% similarity in PFGE banding patterns.

The distribution observed in this study might reflect the global tendency, which shows a shift from BAPS3.3 (predominant in 1990s) to BAPS2.1 (predominant after 2005).⁷ The fact that the pRUM-like plasmids were identified in two separate BAPS groups indicates that not only clonal expansion but also horizontal transfer of plasmids was involved in the dissemination.

Although pRUM-like plasmids were identified in many countries since 1992, and in a neighboring country (Serbia) in 2005, their presence in Hungary was first indicated by this study.^2,13 We suggest that these plasmids could have been imported to Hungary from abroad. Considering the difficulties in establishing phylogenetic relationships among plasmids, the exact source and time point of the importation could not be determined.

Several factors were detected on these plasmids that could have facilitated the long-term stability of the vanA operon in the E. faecium population in Hungary. The Axe-Txe addiction system present on the pRUM backbone ensures stable maintenance.¹⁴ Bacteriocin 43, encoded on pPEC286, pBUD607, and pMIS102, has a broad spectrum of activity. The bacteriocin 43 coding genes were shown to be co-mobilized with vancomycin resistance plasmids, and it might play a role in colonization and invasion.^17,18

Besides the aforementioned microbial features, environmental factors might influence the dissemination of VanA-type E. faecium in Hungary. Several studies showed that the incidence of Clostridium difficile infections (CDI) and VRE colonization are associated, and that oral vancomycin and oral metronidazole given for the treatment of CDI promote the overgrowth of VRE in the gut.^19–21 In Hungary, the CDI incidence rate increased considerably recently (2008: 2/10,000 patient days; 2011–2012: 12.3/10,000 patient days).^22,23 The increment was largely attributed to the spread of hypervirulent 027 PCR ribotype.²⁴ The unfavorable change in the epidemiology of C. difficile might contribute to the spread of VRE in our country.

The worldwide success of pRUM-like vancomycin resistance plasmids suggests that vancomycin resistance will continue to circulate among E. faecium strains in Hungary. Therefore, effective measures to prevent or dismiss colonization, or preclude the progression to infection should be enforced.

Footnotes

Acknowledgment

We thank P. Courvalin (Institut Pasteur, Paris, France) for providing strain BM4147.

Disclosure Statement

No competing financial interests exist.

References

Werner

, Coque

T.M.

, Hammerum

A.M.

, Hope

, Hryniewicz

, Johnson

, Klare

, Kristinsson

K.G.

, Leclercq

, Lester

C.H.

, Lillie

, Novais

, Olsson-Liljequist

, Peixe

L.V.

, Sadowy

, Simonsen

G.S.

, Top

, Vuopio-Varkila

, Willems

R.J.

, Witte

, and Woodford

. 2008. Emergence and spread of vancomycin resistance among enterococci in Europe. Euro Surveill. 13:pii:19046.

Libisch

, Lepsanovic

, Top

, Muzslay

, Konkoly-Thege

, Gacs

, Balogh

, Fuzi

, and Willems

. 2008. Molecular characterization of vancomycin-resistant Enterococcus spp. clinical isolates from Hungary and Serbia. Scand. J. Infect. Dis. 40:778–784.

European Centre for Disease Prevention and Control.. 2017. Surveillance atlas of infectious diseases. Available at atlas.ecdc.europa.eu/public/index.aspx (accessed January 31, 2018).

Dutka-Malen

, Evers

, and Courvalin

. 1995. Detection of glycopeptide resistance genotypes and identification to the species level of clinically relevant enterococci by PCR. J. Clin. Microbiol., 33:24–27.

Turabelidze

, Kotetishvili

, Kreger

, Morris

J.G.

, and Sulakvelidze

. 2000. Improved pulsed-field gel electrophoresis for typing vancomycin-resistant enterococci. J. Clin. Microbiol. 38:4242–4245.

Homan

W.L.

, Tribe

, Poznanski

, Li

, Hogg

, Spalburg

, Van Embden

J.D.A.

, and Willems

R.J.L.

. 2002. Multilocus sequence typing scheme for Enterococcus faecium. J. Clin. Microbiol. 40:1963–1971.

Willems

R.J.L.

, Top

, van Schaik

, Leavis

, Bonten

, Siren

, Hanage

W.P.

, and Corander

. 2012. Restricted gene flow among hospital subpopulations of Enterococcus faecium. MBio. 3:e00151-12.

Arthur

, Molinas

, Depardieu

, and Courvalin

. 1993. Characterization of Tn1546, a Tn3-related transposon conferring glycopeptide resistance by synthesis of depsipeptide peptidoglycan precursors in Enterococcus faecium BM4147. J. Bacteriol., 175:117–127.

Sambrook

, and Russell

D.W.

. 2001. Molecular Cloning: A Laboratory Manual. 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

10.

Antipov

, Hartwick

, Shen

, Raiko

, Lapidus

, and Pevzner

P.A.

. 2016. plasmidSPAdes: assembling plasmids from whole genome sequencing data. Bioinformatics, 32:3380–3387.

11.

Wick

R.R.

, Schultz

M.B.

, Zobel

, and Holt

K.E.

. 2015. Bandage: interactive visualization of de novo genome assemblies. Bioinformatics, 31:3350–3352.

12.

Besemer

, and Borodovsky

. 2005. GeneMark: web software for gene finding in prokaryotes, eukaryotes and viruses. Nucleic Acids Res., 33:W451–W454.

13.

Freitas

A.R.

, Tedim

A.P.

, Francia

M.V.

, Jensen

L.B.

, Novais

, Peixe

, Sánchez-Valenzuela

, Sundsfjord

, Hegstad

, Werner

, Sadowy

, Hammerum

A.M.

, Garcia-Migura

, Willems

R.J.

, Baquero

, and Coque

T.M.

. 2016. Multilevel population genetic analysis of vanA and vanB Enterococcus faecium causing nosocomial outbreaks in 27 countries (1986–2012). J. Antimicrob. Chemother. 71:3351–3366.

14.

Halvorsen

E.M.

, Williams

J.J.

, Bhimani

A.J.

, Billings

E.A.

, and Hergenrother

P.J.

. 2011. Txe, an endoribonuclease of the enterococcal Axe-Txe toxin-antitoxin system, cleaves mRNA and inhibits protein synthesis. Microbiology, 157:387–397.

15.

Mikalsen

, Pedersen

, Willems

, Coque

T.M.

, Werner

, Sadowy

, van Schaik

, Jensen

L.B.

, Sundsfjord

, and Hegstad

. 2015. Investigating the mobilome in clinically important lineages of Enterococcus faecium and Enterococcus faecalis. BMC Genomics, 16:282.

16.

Pinholt

, Larner-Svensson

, Littauer

, Moser

C.E.

, Pedersen

, Lemming

L.E.

, Ejlertsen

, Søndergaard

T.S.

, Holzknecht

B.J.

, Justesen

U.S.

, Dzajic

, Olsen

S.S.

, Nielsen

J.B.

, Worning

, Hammerum

A.M.

, Westh

, and Jakobsen

. 2015. Multiple hospital outbreaks of vanA Enterococcus faecium in Denmark, 2012–2013, investigated by WGS, MLST and PFGE. J. Antimicrob. Chemother. 70:2474–2482.

17.

Todokoro

, Tomita

, Inoue

, and Ike

. 2006. Genetic analysis of bacteriocin 43 of vancomycin-resistant Enterococcus faecium. Appl. Environ. Microbiol., 72:6955–6964.

18.

del Campo

, Tenorio

, Jiménez-Díaz

, Rubio

, Gómez-Lus

, Baquero

, and Torres

. 2001. Bacteriocin production in vancomycin-resistant and vancomycin-susceptible Enterococcus isolates of different origins. Antimicrob. Agents Chemother. 45:905–912.

19.

Al-Nassir

W.N.

, Sethi

A.K.

, Li

, Pultz

M.J.

, Riggs

M.M.

, and Donskey

C.J.

. 2008. Both oral metronidazole and oral vancomycin promote persistent overgrowth of vancomycin-resistant enterococci during treatment of Clostridium difficile-associated disease. Antimicrob. Agents Chemother., 52:2403–2406.

20.

Fujitani

, George

W.L.

, Morgan

M.A.

, Nichols

, and Murthy

A.R.

. 2011. Implications for vancomycin-resistant Enterococcus colonization associated with Clostridium difficile infections. Am. J. Infect. Control, 39:188–193.

21.

Sethi

A.K.

, Al-Nassir

W.N.

, Nerandzic

M.M.

, and Donskey

C.J.

. 2009. Skin and environmental contamination with vancomycin-resistant enterococci in patients receiving oral metronidazole or oral vancomycin treatment for Clostridium difficile-associated disease. Infect. Control Hosp. Epidemiol., 30:13–17.

22.

Bauer

M.P.

, Notermans

D.W.

, van Benthem

B.H.

, Brazier

J.S.

, Wilcox

M.H.

, Rupnik

, Monnet

D.L.

, van Dissel

J.T.

, Kuijper

E.J.

, and ECDIS Study Group. 2011. Clostridium difficile infection in Europe: a hospital-based survey. Lancet, 377:63–73.

23.

Davies

K.A.

, Longshaw

C.M.

, Davis

G.L.

, Bouza

, Barbut

, Barna

, Delmée

, Fitzpatrick

, Ivanova

, Kuijper

, Macovei

I.S.

, Mentula

, Mastrantonio

, von Müller

, Oleastro

, Petinaki

, Pituch

, Norén

, Nováková

, Nyč

, Rupnik

, Schmid

, and Wilcox

M.H.

. 2014. Underdiagnosis of Clostridium difficile across Europe: the European, multicentre, prospective, biannual, point-prevalence study of Clostridium difficile infection in hospitalised patients with diarrhoea (EUCLID). Lancet Infect. Dis. 14:1208–1219.

24.

Tóth

, Urbán

, Osztie

, Benczik

, Indra

, Nagy

, and Allerberger

. 2016. Distribution of PCR ribotypes among recent Clostridium difficile isolates collected in two districts of Hungary using capillary gel electrophoresis and review of changes in the circulating ribotypes over time. J. Med. Microbiol., 65:1158–1163.