Relationship between nosocomial Acinetobacter species occurring in two geographical areas (Greece and the UK)

Journal of Hospital Infection (2003) 54, 207–211

www.elsevierhealth.com/journals/jhin

Relationship between nosocomial Acinetobacter species occurring in two geographical areas (Greece and the UK) I.D. Dimopouloua,*, S.I. Kartalia, G.N. Kartalisb, K.I. Manolasc, K.E. Simopoulosd, B.A. Vargemezise, G. Theodoropoulou-Rodiouf, I.C.J.W. Bowlerg, D.W.M. Crookg a

Laboratory of Medical Microbiology, Democritus University of Thrace, Alexandroupolis 68100, Greece A’ Pathology Unit, Democritus University of Thrace, Alexandroupolis 68100, Greece c A’ Surgery Unit, Democritus University of Thrace, Alexandroupolis 68100, Greece d B’ Surgery Unit, Democritus University of Thrace, Alexandroupolis 68100, Greece e Nephrology Unit, Democritus University of Thrace, Alexandroupolis 68100, Greece f Microbiology Laboratory, Xanthi’s General Hospital, Xanthi 67100, Greece g Interdepartmental academic unit of Infectious Diseases and Microbiology, Oxford University, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK b

Received 26 September 2002; accepted 1 April 2003

KEYWORDS Acinetobacter spp; 16 S Ribosomal DNA sequence analysis; REP-PCR; PFGE; Nosocomial

Summary Fifty-two isolates of Acinetobacter spp. obtained from three Greek and one UK hospital, were studied using partial 16 S ribosomal DNA sequence analysis, repetitive extragenic palindromic sequence-based polymerase chain reaction (REPPCR) mediated fingerprinting and DNA macro-restriction analysis. The aim was twofold: first, to discern the major differences in the population of Acinetobacter spp. between the two countries. Second, to compare a simple PCR-based typing scheme with pulsed-field gel electrophoresis (PFGE). The multi-resistant Greek isolates were within DNA groups 2 and TU13, and clustered into three types both by REP-PCR and PFGE. By contrast, the more susceptible Oxford isolates were heterogeneous on 16 S RNA sequence analysis and distinguishable on typing. The need for studies that elucidate the phylogeny of Acinetobacter spp. inside and outside hospitals are important, as this will help clarify the relationship between organisms that are increasingly recognized as causes of severe infections. Q 2003 The Hospital Infection Society. Published by Elsevier Science Ltd. All rights reserved.

Introduction In a recent European study, Acinetobacter spp. *Corresponding author. Tel.: þ30-6945-494763; fax: þ30-2106990111. E-mail address: [email protected]

and Stenotrophomonas maltophilia causing severe nosocomial infections, especially in intensive care units (ICUs), in five European countries (Belgium, France, Portugal, Spain and Sweden) were among the most antibiotic-resistant pathogens tested.1 In Greece, Acinetobacter spp. has accounted for 7.4% of all nosocomial infections

0195-6701/03/$ - see front matter Q 2003 The Hospital Infection Society. Published by Elsevier Science Ltd. All rights reserved. doi:10.1016/S0195-6701(03)00152-X

208 caused by Gram-negative bacilli;2 the majority (76%) were resistant to newer antimicrobial agents such as third-generation cephalosporins, aztreonam and fluorinated quinolones.3 It is increasingly recognized that Acinetobacter spp. account for a growing burden of nosocomial infection. However, knowledge of their origin and spread is rudimentry. Characterization of the distribution and emergence of resistance in Acinetobacter spp. is difficult because there is no simple, unequivocal means of speciating this rather biochemically non-reactive and phenotypically homogeneous organism. However, this genus is genetically heterogeneous and consists of 21 DNA groups,4 and a reliable genetic means of phylogenetically classifying this taxon is needed. Speciation using ribosomal DNA (rDNA) sequences is an improvement over biochemical or DNA hybridization methods because it offers an unequivocal result with which to compare sequences of other organisms available in databases over the Internet. Recently, Ibrahim et al. 4 reported a phylogenetic study of the relationships of this complex genus using 16 S rDNA sequence analysis, and identified a partial sequence with sequence motifs that differentiated all the type strains. Such a sequence-based method, if used, should contribute to an improved appreciation of the species-specific contribution of this genus to multi-resistance and clinical disease. The appearance of multi-resistant Acinetobacter spp. in the ICU and other units of Alexandroupolis General Hospital, Thrace, Greece, was in keeping with recent European reports describing the emergence of multi-resistant Acinetobacter spp. An observational study of acinetobacter clinical isolates collected from patient samples in three hospitals located in Thrace and East Macedonia (Greece) was undertaken. These results were compared with sporadic isolates from a large UK teaching hospital, The John Radcliffe Hospital, Oxford, where no apparent outbreak of multiresistant Acinetobacter spp. has occurred. The aim was to discern the major differences in the population of Acinetobacter spp. between these two environments using partial 16 S RNA sequence analysis and to compare a simple PCR-based typing scheme with pulsed-field gel electrophoresis (PFGE).

Materials and methods In total, 52 clinically significant, sequential isolates

I.D. Dimopoulou et al.

of Acinetobacter spp. were collected from routine clinical samples [blood, cerebrospinal fluid (CSF), respiratory and wound] obtained from 50 inpatients located in four hospitals. Thirty-five isolates were obtained from 35 patients from Alexandroupolis’ General Hospital, Xanthi’s General Hospital and Kavala’s General Hospital (Thrace and East Macedonia, Greece) during the period 15 December 1997 to 27 September 1999. Seventeen isolates (from 15 patients) were obtained from the John Radcliffe Hospital (Oxford, England) during the period 28 December 1995 to 17 September 1999. All the isolates had the features of Acinetobacter spp. as defined by Bouvet and Grimont.5 The DNA group to which the isolates belonged was determined by partial 16 S ribosomal DNA sequence analysis as suggested by Ibrahim et al.4 Briefly, the sequence between position 7 and 520 equivalent to the Escherichia coli 16 S rDNA sequence was amplified by PCR and directly sequenced using the primers 50 -AGAGTTTGATC(AC)TGGCTCAG-30 6 and 50 -GTATTACCGCGGCTGCTG-30 , respectively.7 The sequence was determined on an ABI PRISM 377 sequencer (Applied Biosystems, Foster City, CA, USA). The sequence trace files were assembled using Staden, Phrep and Phrap, and analysed with the GCG program (Genetics Computer Group Inc., Madison WI, USA). The sequences of the isolates in this study were aligned with the equivalent 16 S rDNA sequences of the 21 DNA groups reported by Ibrahim et al.4 and available via Genbank accession numbers Z 93 434 – Z 93 454. The relationship between the clinical isolates and the type strains representing the DNA groups was deduced using the neighbour-joining method with a Jukes – Cantor correction supplied by the ClaustalX program. The antibiotic susceptibilities were determined using the comparative disk diffusion method following NCCLS guidelines. Nineteen antibiotics were tested: ampicillin, ampicillin – clavulanate, amikacin, aztreonam, cefepime, cefotaxime, cefoxitin, cefpirome, ceftazidime, cephalothin, ciprofloxacin, gentamicin, imipenem, netilmicin, pefloxacin, piperacillin, piperacillin –tazobactam, tobramycin, trimethoprim –sulfamethoxazole. Repetitive extragenic palindromic sequencebased PCR (REP-PCR) analysis was performed following previously described methods using the primers: 5 0 -IIIGCGCCGICATCAGGC-30 and 5 0 ACGTCTTATCAGGCCTAC-30 .8 PFGE analysis was performed as described previously.9 The endonucleases SmaI and ApaI were used to digest the genomic DNA.

Nosocomial Acinetobacter species in Greece and the UK

209

Figure 1 (a) PFGE types of representative Acinetobacter spp. from the hospitals of East Macedonia and Thrace, after digestion with ApaI. Lane 1, type D. Lane 2, subtype A2. Lane 3, subtype A1. Lane 4, type C. Lane 5, type A. Lane 6, subtype B1. Lane 7, type B. Lane 8, molecular size markers (0.1-200 kb). (b) PFGE patterns of isolates from John Radcliffe Hospital and two representative isolates from the Greek hospitals, after digestion with ApaI. Lane 1, type A (Greek). Lane 2, type B (Greek). Lanes 3– 17, strains 20, 19, 18, 17, 16, 15, 14, 13, 12, 10, 8, 7, 3, 2, 1, respectively. Lane 18, molecular size markers (0.1– 200 kb).

Results Thirty-four of the 35 Greek isolates were multiresistant (i.e. exhibited resistance to 12 or more of the antibiotics tested). By contrast, only two of the 17 Oxford isolates were multi-resistant while the remainder were susceptible (i.e. resistant to six or fewer antibiotics tested). Of the 52 strains tested, all were either resistant to 12 or more or less than seven antibiotics allowing a categorical stratifica-

tion into multi-resistant or susceptible, based on the definitions above. The 34 multi-resistant Greek isolates clustered into three types both by PFGE (A, B, D) and REP-PCR typing (II, I, III). PFGE revealed subtle banding differences within the major types regarded as subtypes (A1, B1, A2) [Figures 1(a) and 2(a) and Table I]. The Oxford isolates were distinguishable [Figures 1(b) and 2(b) and Table I]. The only two pairs of indistinguishable isolates were cultured from two

Figure 2 (a) REP-PCR types of representative Acinetobacter spp. from the hospitals of East Macedonia and Thrace. Lane 1, molecular size markers (2.1 –0.4 kb). Lane 2, type IV. Lane 3, type I. Lane 4, type II. Lane 5, type III. (b) REP-PCR types of isolates from John Radcliffe Hospital and four representative isolates from the Greek hospitals. Lane 1, type IV (Greek). Lane 2, type I (Greek). Lane 3, type II (Greek). Lane 4, type III (Greek). Lanes 5– 21, strains 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 8, 7, 5, 3, 2, 1, respectively. Lane 22, molecular size markers (2.1– 0.4 kb).

210

I.D. Dimopoulou et al.

Table I Details for acinetobacter isolates from the Greek hospitals and the John Radcliffe Hospital Country

Isolatea

Susceptibilityb

Acinetobacter clusterc

REP-PCR typed

PFGE typee

Greece

7A,31A,212A,12A,169A,148A, 248A,11A,15A,304A,468A 68A 200A,101A,352A,505A,593A, 99A,356A,363A,202A,320A,226A, 289A,372A,3861K,4243K,1A 299A,1816X,189X 283X 50A,214A 4382K

R

Cluster I

I

B

R R

Cluster I Cluster I

I II

B1 A

R R R S

Cluster I Cluster I Cluster I Cluster II

II II III IV

A1 A2 D C

10E,11E 12E 13E 14E 18E 1E 2E 3E,5E 19E 15E 16E 17E 8E 7E 20E

R S S S S S S S S S S S S S S

Cluster I Cluster I Cluster I Cluster I Cluster I Cluster II Cluster II Cluster II Cluster II Cluster III Cluster III Cluster III Cluster III Cluster III n.2

X XI XII XIII XVII V VI VII XVIII XIV XV XVI IX VIII XIX

K L M N R E F G S O P Q I H T

England

a

b c

d e

A, isolate from Alexandroupolis’ General Hospital; K, isolate from Kavala’s General Hospital; X, isolate from Xanthi’s General Hospital; E, isolate from John Radcliffe Hospital. R: strains resistant to $12 antibiotics, S: strains resistant to six or fewer antibiotics. Phylogenetic clusters of Acinetobacter spp. as described by Ibrahim et al.:4 cluster I (DNA groups 2 and TU13), cluster II (DNA groups 3, CTTU13 and 1-3), cluster III (DNA groups 8, 9 and 6), n.a.: no cluster available for DNA group 4. REP-PCR, repetitive extragenic palindromic sequence-based polymerase chain reaction. PFGE, pulsed-field gel electrophoresis.

patients with apparently separate episodes of bacteraemia (3E, 5E and 10E, 11E) [Figure 2(b), Table I]. On phylogenetic analysis of the 16 S rDNA sequences, all 34 multi-resistant Greek isolates clustered with cluster I consisting of the DNA group 2 (Acinetobacter baumannii) and group TU13, while the susceptible Greek isolate clustered with cluster II consisting of DNA groups 3, CTTU13 and 1 – 3. The Oxford isolates were heterogeneous and consisted of six strains clustering with cluster I; five with cluster II; five with cluster III consisting of DNA groups 8 (Acinetobacter lwoffii), 9 and 6, and one strain belonged to DNA group 4 with no assigned cluster as defined by Ibrahim et al.4 (Table I).

Discussion This investigation demonstrates the usefulness of molecular typing methods such as REP-PCR and PFGE for distinguishing isolates. REP-PCR typing, a simpler typing method than PFGE, provided a

similarly high degree of discrimination among the 52 strains studied. Partial 16 S rDNA sequence analysis provided a useful means of speciating the isolates, and is the first time this method has been used to identify and compare strains in such a study. The major phenotypic difference is that the Greek isolates were multi-resistant, while the Oxford strains were more susceptible. The Greek isolates came from a few wards in only three hospitals. Their clonality is consistent with spread of a few strains within and between hospitals within Thrace and East Macedonia. The heterogeneity observed in the population of Oxford strains, a setting where multi-resistant strains of Acinetobacter spp. occurs only sporadically, is consistent with strains being unrelated epidemiologically, and suggests that there is no one simple model for the occurrence of nosocomial Acinetobacter spp. Furthermore, it has been previously observed that multi-resistant strains are clonal in population structure while susceptible isolates are heterogeneous10,11 However, Webster et al.12 found that

Nosocomial Acinetobacter species in Greece and the UK

the multi-resistant Acinetobacter spp. circulating in the Chris Hani Baragwanath Hospital, South Africa, were diverse. Their heterogeneity is more in keeping with the Oxford isolates, which by contrast were relatively susceptible and included five isolates within cluster I (A. baumanii—DNA group TU13), only two of which were resistant. Therefore, a simple paradigm that multi-resistance correlates with clonality or DNA group is difficult to sustain. Furthermore, the possibility that transmissibility, virulence or invasiveness is clonally restricted, a speculative possibility advanced by Dijkshoorn et al.10 is contradicted by the recent observations of Anstey et al.13 and Wang et al.14 who describe diverse strains of A. baumannii colonizing and causing life-threatening community-acquired pneumonia in tropical northern Australia and community-acquired bacteraemia in Taiwan, respectively. Our understanding of the factors shaping the population structure and transmission of Acinetobacter spp. are poorly developed and limited by the dependence on techniques, such as those reported here, that can distinguish between strains, but offer limited scope for determining the population structure of bacterial populations. This limitation may be best addressed by measuring the diversity and its structure in the population of strains representing a comprehensive collection of Acinetobacter spp. by multi-gene sequence analysis of ‘neutral’ housekeeping genes, for example using multi-locus sequence typing. Such gene sequence data provide unequivocal data with which to compare isolates between centres over the Internet. Furthermore, the sequence variation between isolates of different species is also likely to be a powerful means of determining phylogenetic relationships and would complement or even supersede speciation using 16 S rDNA.

Acknowledgements

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

We thank Dr A. Tirologou (Kavala’s General Hospital) for providing strains of Acinetobacter. 13.

References 1. Hanberger H, Garcia-Rodriguez JA, Gobernado M, Goossens H, Nilsson LE, Struelens MJ, French and Portuguese ICU Study Groups. Antibiotic susceptibility among aerobic Gram-

14.

211

negative bacilli in intensive care units in five European countries. JAMA 1999;281:67—71. Douboyas J, Tzouvelekis LS, Tsakris A. In-vitro activity of ampicillin/sulbactam against multi-resistant Acinetobacter calcoaceticus var. anitratus clinical isolates. J Antimicrob Chemother 1994;34:298—300. Vatopoulos AC, Kalapothaki V, Greek network for the surveillance of antimicrobial resistance, Legakis NJ. Bacterial resistance to ciprofloxacin in Greece: results from the National Electronic Surveillance System. Emerg Infect Dis 1999;5:471. Ibrahim A, Gerner-Smidt P, Liesack W. Phylogeneticrelationships of the twenty-one groups of the genus Acinetobacter as revealed by 16 S ribosomal DNA sequence analysis. Int J Syst Bacteriol 1997;47:837—841. Bouvet PJM, Grimont PAD. Taxonomy of the genus Acinetobacter with the recognition of Acinetobacter baumannii sp. Nov., Acinetobacter haemolyticus sp. nov., Acinetobacter johnsonii sp. nov., and Acinetobacter junii sp. nov. and emended description of Acinetobacter calcoaceticus and Acinetobacter lwoffii. Int J Syst Bacteriol 1986;36:228—240. Choi BK, Paster BJ, Dewhirst FE, Gobel UB. Diversity of cultivable and uncultivable oral spirochetes from a patient with severe destructive periodontitis. Infect Immun 1994; 62:1889—1895. Kroes I, Lepp PW, Relman DA. Bacterial diversity within the human subgingival cervice. Proc Natl Acad Sci 1999;96: 14547—14552. Snelling AM, Gerner-Smidt P, Hawkey PM, et al. Validation of use of whole-cell repetitive extragenic palindromic sequencebased PCR (REP-PCR) for typing strains belonging to the Acinetobacter calcoaceticus—Acinetobacter baumannii complex and application of the method to the investigation of a hospital outbreak. J Clin Microbiol 1996;34:1193—1202. Maslow JN, Slutsky AM, Arbeit RD. In: Persing DH, Smith TF, Tenover FC, White TJ, editors. Application of pulsed-field gel electrophoresis to molecular epidemiology. Diagnostic molecular microbiology principles and applications, Washington, DC: American Society for Microbiology; 1993. p. 563—572. Dijkshoom L, Aucken H, Gerner-Smidt P, et al. Comparison of outbreak and non-outbreak Acinetobacter baumannii strains by genotypic and phenotypic methods. J Clin Microbiol 1996;34:1519—1525. Nemec A, Janda L, Melter O, Dijkshoom L. Genotypic and phenotypic similarity of multi-resistant Acinetobacter baumannii isolates in the Czech Republic. J Med Microbiol 1999; 48:287—296. Webster CA, Towner KJ, Saunders GL, Crewe-Brown HH, Humphreys H. Molecular and antibiogram relationships of acinetobacter isolates from two contrasting hospitals in the United Kingdom and South Africa. Eur J Clin Microbiol Infect Dis 1999;18:595—598. Anstey NM, Currie BJ, Hassell M, Palmer D, Dwyer B, Seifert H. Community-acquired bacteremic acinetobacter pneumonia in tropical Australia is caused by diverse strains of Acinetobacter baumannii, with carriage in the throat in atrisk groups. J Clin Microbiol 2002;40:685—686. Wang JT, McDonald LC, Chang SC, Ho M. Community-aquired Acinetobacter baumannii bacteremia in adult patients in Taiwan. J Clin Microbiol 2002;40:1526—1529.