European Journal of Clinical Microbiology and Infectious Diseases

, Volume 23, Issue 1, pp 15–19

Obvious Lack of Association Between Dynamics of Epidemic Methicillin-Resistant Staphylococcus aureus in Central Europe and agr Specificity Groups

Authors

  • B. Strommenger
    • Wernigerode BranchRobert Koch Institute
  • C. Cuny
    • Wernigerode BranchRobert Koch Institute
  • G. Werner
    • Wernigerode BranchRobert Koch Institute
    • Wernigerode BranchRobert Koch Institute
Article

DOI: 10.1007/s10096-003-1046-8

Cite this article as:
Strommenger, B., Cuny, C., Werner, G. et al. Eur J Clin Microbiol Infect Dis (2004) 23: 15. doi:10.1007/s10096-003-1046-8
  • 100 Views

Abstract

During the past 8 years, changes in the prevalence and spread of different epidemic methicillin-resistant Staphylococcus aureus (MRSA) have been observed in central Europe, with the emergence of new strains possessing fewer resistance characters. This has also been demonstrated at the level of particular hospitals. Since variation in agr specificity type has been proposed as a possible reason for population dynamics in Staphylococcus aureus, the agr specificity groups of different epidemic MRSA strains were investigated by PCR using agr group-specific primers. Four of the “old” as well as two “new” epidemic strains exhibited agr specificity group I. One group of epidemic MRSA strains, which has been observed since the beginning of the 1990s, exhibited the agr specificity group II. Sequencing the variable part (agrB-D-C) of the agr locus revealed only six relevant nucleotide changes within this region, with three of them modifying the Shine-Dalgarno sequence region of agrC. On the basis of the results obtained, it is proposed that the dynamics observed in the population of MRSA in Germany is not due to different agr group specificities in “old” and “new” epidemic clones.

Introduction

Staphylococcus aureus is a major nosocomial pathogen that causes a variety of infections including septicemia, endocarditis, pneumonia, and osteomyelitis. During the past 15 years, the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) has increased in many parts of the world. As shown by several studies, this is due mainly to the emergence and spread of epidemic strains [1, 2, 3]. In central Europe, a particular dynamic of epidemic strains has been observed: strains widely disseminated during the beginning of the 1990s have become less frequent, whereas an increasing spread of newly emerging strains has been recorded [4]. Changes in the population of bacteria can occur for different reasons. One putative reason is the impact of selective pressure due to the use of antibiotics, which results in the acquisition of new resistance genes. Since the newly emerging strains are even less broadly resistant [4, 5, 6], acquisition of resistance genes may not play a role in the spread of new epidemic MRSA in Germany.

As already suggested and elucidated in previous studies, the staphylococcal quorum-sensing system mediated by the accessory gene regulator (agr) is probably involved in competition between strains for sites of colonization or infection [7, 8, 9]. Quorum sensing by the agr global regulatory system is mediated by a post-transcriptionally processed autoinducing peptide that is excised from the agrD gene product and that interacts with a transmembrane receptor encoded by the agrC gene. As a part of a two-component regulatory system, AgrC mediates phosphorylation of the agrA gene product, which autoactivates the transcription of the agr operon from two divergent promotors, P2 and P3. This results in downregulation of surface proteins, including putative colonization factors, and in upregulation of genes encoding extracellular proteins via the regulatory RNA III transcribed from promotor P3 [10]. Until now, four different groups of autoinducer-receptor specificity have been described [7, 9]. The secreted autoinducing peptide can activate the agr response in strains in the same specificity group and can inhibit it in strains from other groups.

Epidemic Staphylococcus aureus strains have to compete successfully with resident staphylococcal microflora. If the agr quorum-sensing system plays a role in this competition, this should be reflected by the agrC-agrD polymorphism of epidemic strains. Therefore, the aim of this study was to compare the agr group specificity of older epidemic MRSA strains with that of the newer ones that have replaced them in recent years.

Materials and Methods

Bacterial Strains

Staphylococcus aureus isolates investigated in this study originate from the strain collection of our laboratory at the National Reference Center for Staphylococci in Germany. Criteria for selection of MRSA for molecular typing have been described previously [3, 4]. Epidemic MRSA had been identified by molecular typing on the basis of SmaI macrorestriction patterns and supplementary polymerase chain reaction (PCR) typing [3] and cluster analysis [3, 5]. Representative isolates from epidemic MRSA in Germany had been included into the European HARMONY collection [11] and further subjected to MLST typing [2]. The Staphylococcus aureus strains 8325-4 (agr group I), 705/92 (agr group II), PS29 (agr group III), PS55 (agr group IV), and 282/94 and 306/94 (for clonal group V of Staphylococcus aureus) were selected as reference strains.

Polymerase Chain Reaction

DNA isolation and PCR reactions were performed as described previously [5]. Primers for resistance genes have been reported previously [12], and primers used for identification of agr specificity are listed in Table 1. The agr class-specific primer pairs agrI, agrII, agrIII, and agrIV were used to determine the agr class of the isolates. The primers B1 and C2 were used to amplify the variable part of the agr operon (1060 bp; Fig. 1). Annealing temperature for all primer pairs was 50°C. Primers used for the detection of resistance genes are indicated elsewhere [12].
Table 1

Primers used in the study

Primer

Sequence

Reference or source of sequence

agrI forward

5′-CAC TTA TCA TCA AAG AGC C-3′

[8]

agrI reverse

5′-CCA CTA ATT ATA GCT GG-3′

agrII forward

5′-GTA GAG CCG TAT TGA TTC-3′

[8]

agrII reverse

5′-GTA TTT CAT CTC TTT AAG G-3′

agrIII forward

5′-CAA GCT ATT ACA TTA CTA CCA-3′

AF00178 pos. no. 510–530

agrIII reverse

5′-AAT GCT TCC ACT TAC TATC-3′

AF00178 pos. no. 728-743

agrIV forward

5′-GCT CAA TTC ATG CAA TTA-3′

AF288215 pos. no. 1207–1223

agrIV reverse

5′-ATG GTA CTG TAA ACA TTA-3′

AF288215 pos. no. 1502–1520

agrB1

5′-TAT GCT CCT GCA GCA ACT AA-3′

[12]

agrC2

5′-CTT GCG CAT TTC GTT GTT TGA-3′

Fig. 1

Sequence polymorphism within the variable part of the agr locus of agr specificity group I. The nucleotide sequences of PCR-derived amplimers from different European epidemic MRSA were compared with the sequence of Staphylococcus aureus 8325 (SAAGRAB) as reference strain. Bold letters indicate nucleotide changes. B1 and C2 show the position of primers for PCR amplification of the variable part of the agr locus (Table 1). SD Shine-Dalgarno sequence

Restriction of Polymerase Chain Reaction Products

Amplicons were purified using the Qiagen PCR Purification Kit (Qiagen, Germany) and digested with DraI (Roche Biochemicals, Germany) according to the manufacturer’s instructions. Resulting restriction fragments were analyzed on a 1.4% agarose gel.

DNA Sequencing and Analysis of Sequence Data

Sequencing reactions were carried out using the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer Applied Biosystems Division, USA) according to the manufacturer’s protocol. Sequence alignment and analysis of sequence polymorphism among the different agr genes were performed with the DNASTAR software package (DNASTAR, USA).

Results

Emergence and Spread of Epidemic Methicillin-Resistant Staphylococcus aureus

Changes in the emergence and spread of epidemic MRSA in German hospitals from 1993 to 2000 are shown in Table 2. While Northern German epidemic MRSA and Hannover area epidemic MRSA dominated in the beginning of the 1990s, in 2000 we mainly recorded the spread of Berlin epidemic MRSA, Southern German epidemic MRSA, and the newly emerged Barnim epidemic MRSA. Since 2001, the incidence of the Southern German epidemic MRSA has been declining, whereas that of the Barnim epidemic MRSA and the recently emerging Rhine-Hesse MRSA is increasing. The situation for a particular hospital is demonstrated in Table 3.
Table 2

Dynamics of epidemic MRSA in German hospitals

Grouping according to molecular typing

Resistance phenotypesa and genotypesb

Frequency of occurrence of epidemic MRSA in different German hospitalsc

agr specificity groupd

1994 (n=121)

1996 (n=140)

1998 (n=337)

2000 (n=567)

2001 (n=849)

2002 (n=747)

Northern German epidemic MRSA ST247

PEN, OXA, GEN, ERY, CLI, OTE, SXT, RIF, CIP

22.0%

17.0%

13.0%

1.3%

0.5%

0.4%

I

mecA, aph2”-aac6’, ermA, tetM, dfrA

Southern German epidemic MRSA ST228

PEN, OXA, ERY, CLI, CIP, (GEN), (OTE)e

16.0%

34.0%

29.0%

36.8%

24.0%

15.2%

II

mecA, aph2”-aac6’, ermA, tetM, dfrA

Hannover area epidemic MRSA

PEN, OXA, GEN, ERY, CLI, SXT, CIP

32.0%

16.0%

15.0%

3.5%

1.3%

0.7%

I

mecA, aph2”-aac6’, ermA, tetM, dfrA

Vienna epidemic MRSA ST254

PEN, OXA, GEN, ERY, CLI, SXT, CIP, OTE, (FUS)

4.0%

1.0%

1.0%

0.1%

0.2%

0.2%

I

mecA, aph2”-aac6’, ermA, tetM

Berlin epidemic MRSA ST45

PEN, OXA, CIP, (GEN, ERY, ERY-CLI, SXT)

11.0%

22.0%

26.0%

26.7%

24.6%

18.2%

I

mecA, grlA, (aph2”-aac6’, ermC, dfrA)f

Barnim epidemic MRSA ST22

PEN, OXA, ERY, CLI, CIP, GEN

9.0%

19.8%

22.0%

28.0%

I

mecA, ermC, grlA, (aph2”-aac6’, rpoB)f

Rhine-Hesse epidemic MRSA ST5

PEN, OXA, ERY, CLI, CIP

12.5%

23.3%

II

mecA, ermA, grlA

Other MRSA

PEN, OXA, (ERY, CLI)

1.0%

2.0%

1.5%

0.9%

0.7%

ND

Other MRSA

variable

10.0%

8.0%

5.0%

10.3%

14.0%

13.3%

ND

No. of hospitals with epidemic MRSA

98

130

241

309

389

333

PEN, penicillin; OXA, oxacillin; GEN, gentamicin; ERY, erythromycin; CLI, clindamycin; OTE, oxytetracycline, SXT, sulfamethoxazole/trimethoprim; RIF, rifampicin; CIP, ciprofloxacin; FUS, fusidic acid; ND, not determined

aResults of PCR for 4 isolates originating from different hospitals

bPhenotypes and genotypes in brackets occur only rarely

cHospitals from which MRSA were sent for typing

dOnly in a few number of isolates

eDecrease after 1994

fOnly in a minor fraction of isolates

Table 3

Dynamics of epidemic MRSA in tertiary care hospital G

Epidemic MRSA

No. of patients affected

1995

1996

1997

1998

1999

2000

2001

2002

Northern German epidemic MRSA

14

18

4

2

2

1

1

2

Southern German epidemic MRSA

1

2

9

6

1

8

8

Hannover area epidemic MRSA

19

13

4

1

Berlin epidemic MRSA

2

2

3

1

12

10

Barnim epidemic MRSA

6

9

11

13

18

22

Determination of agr Specificity Group and Resistance Genotype

The agr specificity group and acquired resistance genes were determined by PCR for four independent isolates of each epidemic strain. The clones dominating in the year 2000 are clearly less broadly resistant compared to the older strains. Except for the Southern German epidemic MRSA and the Rhine-Hesse epidemic MRSA, which belong to specificity group II, the strains all belong to agr specificity group I (Table 2).

Sequence Comparison of the agr Locus

Sequencing the most variable part of the agr locus that comprises the 3′-terminal part of agrB and agrD and the 5′-terminal part of agrC (Fig. 1) revealed that the sequence of the older epidemic strains (northern German epidemic MRSA and Hannover area MRSA) is almost identical to the sequence of Staphylococcus aureus 8325 (SAAGRAB sequence; GenBank accession no. X52543; Fig. 1). The corresponding sequences of the Barnim epidemic MRSA and the Berlin epidemic MRSA differ from the SAAGRAB sequence in three nucleotides located in the intergenic space between agrD and agrC (pos. no. 2484, 2485 and 2493 in X52543). These changes have led to the loss of a DraI cleavage site (Fig. 2) and a modification of the ribosomal binding site of agrC. Three additional nucleotide changes (pos. no. 2227, 2229, and 2631 in X52543) were observed in the Berlin epidemic MRSA and have resulted in two amino acid changes in agrB and one in agrC, respectively.
Fig. 2

DraI restriction pattern of agrC-agrB amplimers of epidemic MRSA: 1000/93 for Hannover area epidemic MRSA; 1450/96 for northern German epidemic MRSA; 635/93 for Vienna epidemic MRSA; 1678/96 for Barnim epidemic MRSA; 1150/93 for Berlin epidemic MRSA, and 282/94 and 306/94 for methicillin-sensitive Staphylococcus aureus as reference strains for clonal group V of Staphylococcus aureus. S molecular mass standard

It was of particular interest to check whether the reported nucleotide changes are specific for both groups of newly emerging MRSA or whether they are already present in supposed methicillin-susceptible ancestors of the Berlin epidemic MRSA (Staphylococcus aureus of group V) [6, 13]. The corresponding agr sequence of these strains is identical to that obtained for isolates of the Berlin epidemic MRSA (Fig. 1).

Comparison of the nucleotide sequence of the Southern German epidemic MRSA, which belong to agr specificity type II, with the corresponding sequence of the agr specificity type II reference strain Staphylococcus aureus SA 502A (GenBank accession no. AF001782) revealed no relevant nucleotide changes.

Discussion

All agr sequences from Staphylococcus aureus strains belonging to different clonal groups can be classified in one of the agr groups defined by Ji et al. [7]. Except for the Southern German epidemic MRSA and the Rhine-Hesse epidemic MRSA (agr group II), all epidemic MRSA belong to agr group I, with the new emerging ones included. Similar findings were reported in a previous study by Van Leeuwen et al. [14]. In a worldwide collection of MRSA, the majority of the strains belonging to different clonal lineages were also attributed to agr group I, with minor nucleotide polymorphism within this group. Therefore, agr specificity does not seem to play a prominent role in the dynamics of epidemic MRSA. Otherwise, we would expect the emergence of MRSA strains differing in agr specificity from their predecessors or exhibiting at least amino acid changes in either agrD or agrC and leading to a modified agr specificity as described by Mayville et al. [15] for modifications of AgrD that resulted in inhibition of the agr response. Modifications of agr specificity by such kind of mutations are obviously rare, as indicated by identical sequences of the older epidemic MRSA and strain 8325, which was first described in 1945 [16].

In this study, we could identify only three relevant point mutations that resulted in amino acid modifications. The nucleotide changes in the Shine-Dalgarno sequence region (ribosomal binding site) between agrD and agrC are in thus far remarkable in that they are identical in the Barnim epidemic MRSA and the Berlin epidemic MRSA, both of which belong to the dominating clones in 2000 on the one hand, but can be classified into quite different clonal groups on the other [5, 6]. The same genotype has recently been described for a Staphylococcus aureus strain representing a subtype of the Canadian epidemic CMRSA-1 (7; GenBank accession no. AF210055). As Shine-Dalgarno sequences in Staphylococcus aureus include the “classical” -GAGA- part [17], it seems possible that a mutation resulting in -GAAA- affects the expression of the receptor protein AgrC. A change in the expression pattern of AgrC could lead finally to a reduced autoactivation of the agr locus. As a result, the “switch” from the expression of surface proteins to the expression of secreted proteins, mediated by the expression of RNA III from the promotor P3, would be diminished, which could be advantageous for colonization. Such a balance of virulence factors expression was also discussed by Papakyriacou et al. [18], who suggested that the protein expression pattern could be influenced by point mutations in the agr locus. It should also be kept in mind that these mutations were already present in methicillin-susceptible Staphylococcus aureus from which the Berlin epidemic MRSA had very likely originated and which became increasingly prevalent by the beginning of the 1990s [13].

However, to verify our hypothesis, further experiments are required to show whether the nucleotide changes concerning the agrC Shine-Dalgarno sequence region affect the expression pattern of agrC.

Copyright information

© Springer-Verlag 2004