b-lactam resistance and b-lactamase expression in clinical Stenotrophomonas maltophilia isolates having defined phylogenetic relationships

Transcription

1 Journal of Antimicrobial Chemotherapy Advance Access published December 13, 2005 Journal of Antimicrobial Chemotherapy doi: /jac/dki453 b-lactam resistance and b-lactamase expression in clinical Stenotrophomonas maltophilia isolates having defined phylogenetic relationships Virginia C. Gould, Aki Okazaki and Matthew B. Avison* Bristol Centre for Antimicrobial Research and Evaluation, Department of Cellular and Molecular Medicine, University of Bristol, School of Medical Sciences, University Walk, Bristol BS8 1TD, UK Introduction Received 22 June 2005; returned 2 November 2005; revised 11 November 2005; accepted 15 November 2005 Aims: To test the hypothesis that Stenotrophomonas maltophilia isolates from certain phylogenetic groups have predictable b-lactamase expression and b-lactam resistance profiles. Methods: Isolates were grouped using sequences of the 16S rrna gene and smet smed intergenic region. b-lactamase activities in cell extracts were quantified spectrophotometrically and b-lactam MICs were determined using agar dilution methodology and Etest as appropriate. Results: A collection of 50 clinical S. maltophilia isolates from Europe and North, South and Central America were phylogenetically grouped. Group A (22 out of 50) includes remarkably genetically homogeneous isolates; group B (17 out of 50) includes isolates that are genetically heterogeneous and quite distinct from those of group A. Members of these two groups are, however, indistinguishable in terms of their b-lactam resistance and b-lactamase expression phenotypes. In contrast, isolates from group C, which are less common (8 out of 50), are considerably more susceptible to b-lactams owing to reduced inducibility of b-lactamase expression following b-lactam challenge. Conclusions: The majority of S. maltophilia clinical isolates behave similarly in terms of b-lactamase expression and b-lactam resistance properties, despite considerable phylogenetic variability. Keywords: expression of resistance, bacterial diversity, in vitro susceptibility In recent years, Stenotrophomonas maltophilia has become an increasing nosocomial threat to human health. Pathogenic strains are associated with a wide spectrum of diseases, though infection is typically restricted to hospitalized patients who are immunocompromised or otherwise severely debilitated. 1 3 Prior exposure to antimicrobial agents, particularly to b-lactams, significantly increases the risk of S. maltophilia infection, since the organism is commonly resistant to these compounds, 1 3 primarily because of the inducible expression of two b-lactamases, L1 and L2. 4,5 Phenotypic and genotypic variability among clinical isolates of S. maltophilia are recurring observations. 4,6,7 We have previously attempted to link genotypic parameters with important phenotypic properties. Our previous study of 10 bacteraemia-causing isolates from the Bristol Royal Infirmary, UK, revealed three 16S rrna gene sequence types. 4 This three group classification scheme was confirmed using the phylogenetic analysis of L1 and L2 b-lactamase sequences, 4 and the sequence of the intergenic region located between smed and smet, encoding a putative smedef transcriptional repressor, has been used for phylogenetic grouping of S. maltophilia isolates. 8 The three phylogenetic groups found so far are: A (16S rrna type 1), typified by isolate K279a; B (16S rrna type 2), typified by isolate N531; and C (16S rrna type 3), typified by isolate J675a. 4,8 Through analysis of these 10 isolates from Bristol, we have shown that those from phylogenetic group A express L1 and L2 b-lactamases inducibly, whereas group B isolates express L2 constitutively at low levels, but have an inducible L1. Group C isolates express both L1 and L2 constitutively at low levels. 5 In the study reported here, we aimed to determine whether these b-lactamase expression phenotypes have an effect on b-lactam resistance profiles, and to find out whether the cross section of isolates found in our locality are representative of a wider selection of clinical isolates from other countries.... *Corresponding author. Tel: ; Fax: ; Matthewb.Avison@bris.ac.uk... Page 1 of 5 Ó The Author Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Gould et al. Table 1. 16S rrna group specificity of b-lactam MICs for S. maltophilia isolates from Bristol, UK MIC (mg/l) a Isolate 16S rrna group ampicillin piperacillin cefotaxime ceftazidime imipenem meropenem K279a (>256) 32 (>256) 3.0 (16) 0.38 (1.0) > (2.0) J675Ib (>256) 24 (>256) 4.0 (16) 0.38 (1.0) > (2.0) H33b (>256) 24 (>256) 2.0 (16) 0.38 (1.0) > (2.0) J530b (>256) 24 (>256) 4.0 (16) 0.5 (2.0) > (2.0) N (12) 16 (>256) 1.0 (2.0) 0.25 (0.38) > (2.0) P567II (8) 24 (>256) 0.75 (1.5) 0.25 (0.38) > (2.0) J675a (2.0) 16 (16) 0.75 (0.75) (0.125) 1.5 (1.5) (0.125) J675b (1.5) 8.0 (12) 0.5 (0.5) 0.25 (0.25) 0.5 (0.5) (0.125) O (2.0) 16 (16) 1.0 (1.0) 0.25 (0.25) 1.5 (1.5) (0.125) O132II (1.5) 12 (12) 0.5 (0.5) 0.25 (0.25) 0.5 (0.5) (0.125) a MICs were determined using Etest on Iso-Sensitest agar. Those values recorded in brackets are the MICs when colonies within the zone of clearing are taken into consideration (as per the Etest guidelines) i.e. the maximal MIC for spontaneous b-lactam-resistant mutants; the first value is the MIC as it would be recorded if these colonies in the zone were ignored (see Figure 1). Where there is no difference between these two MIC values, few, if any, resistant mutants were generated. Materials and methods Materials Nitrocefin was obtained from Becton-Dickinson (Cockeysville, USA) and imipenem was from Merck Sharpe & Dohme, West Point, USA. PCR reagents were obtained from Abgene (Epsom, UK) and PCR primers were from Qiagen (Crawley, UK). All other general reagents and b-lactams were from Sigma Chemical Co. or BDH, both of Poole, Dorset, UK. Susceptibility tests and b-lactamase assays All susceptibility data were determined using Iso-Sensitest agar or Mueller Hinton agar (Oxoid, Basingstoke, UK) as stated in the Results and discussion section. MIC values were either quantified by using Etest strips (AB Biodisk, Solna, Sweden) with an inoculum equivalent to a 0.5 McFarland standard following incubation at 37 C for 18 h, or by using agar dilution according to the BSAC protocol. 9 b-lactamase inductions and specific assays for L1 and L2 b-lactamase in crude cell extracts were carried out according to the methods described previously 5 except that nitrocefin was used as substrate instead of imipenem or ceftazidime. L1 and L2 b-lactamase activities were differentiated by determining the amount of nitrocefin hydrolysing activity in cell extracts in the absence (L1 + L2 combined activity) or presence (L2 activity only) of 50 mm (final) EDTA, which was added to cell extract for 10 min prior to the assay being carried out. One unit of enzyme activity was designated as hydrolysing 1 nmol of nitrocefin per min at 25 C. The protein concentration of each bacterial extract was determined using the Bio-Rad protein assay reagent (Bio-Rad, München, Germany) according to the manufacturer s instructions, allowing specific activity (U/mg of protein) to be calculated for each extract. Average data from three separate preparations of cells were used in later analyses. PCR protocols and analysis of sequence data The 16S rrna gene-specific PCR and sequencing were performed using the protocols and primers described previously. 4 smet smed intergenic region PCR sequencing was performed as described previously. 8 Computer-assisted sequence manipulation and alignments were made with the Lasergene software package (DNA star, Madison, WI, USA). The phylogenetic tree was produced as previously described. 10 Results and discussion The data presented in Table 1 clearly show that there is a link between the ability of the Bristol collection of isolates to produce b-lactamases inducibly and the MICs (determined using Etest on Iso-Sensitest agar) of a variety of b-lactam antimicrobials for them. Specifically, in 16S rrna group 2 (phylogenetic group B) 4 isolates, the inability to produce L2 inducibly 5 has a dramatic effect on the MICs of cephalosporins and some penicillins. Furthermore, 16S rrna group 3 (phylogenetic group C) 4 isolates, which do not produce b-lactamases inducibly, 5 are highly susceptible to all b-lactams (Table 1). Figure 1 illustrates that the inducible expression of b-lactamases facilitates the selection of high-level b-lactam-resistant mutants with full data for all the isolates being summarized in Table 1. Given these important findings concerning specific phenotypic properties of members of the three known phylogenetic types of S. maltophilia isolates collected in our locality, we wanted to determine whether they represent general properties of S. maltophilia isolates from diverse geographical locations. To this end, we obtained a collection of 50 randomly selected clinical S. maltophilia isolates. Ten were from Brazil, nine from the USA, five from Turkey, four each from Chile, Mexico and Venezuela, three each from Argentina and Canada, two each from Germany, Italy and Spain, and one each from France and Belgium. The isolates were cultured from blood (31 out of 50), respiratory samples (10 out of 50), intensive care units (7 out of 50), skin swabs (1 out of 50) and urine (1 out of 50). Sequences of the 16S rrna genes of all 50 test isolates were determined. Analysis of these data (Table 2) revealed three additional 16S rrna types (named 4, 5 and 6) not seen previously. The 16S rrna type 1 isolates predominate (18 out of 50); all of the other types are represented by seven isolates or fewer. The 16S rrna sequence data revealed, in some cases, mixed sequence traces within the hypervariable region, as previously seen with Page 2 of 5

3 b-lactam resistance in S. maltophilia (a) (b) (c) Figure 1. Generation of b-lactam-resistant mutants from S. maltophilia isolates. Representative isolates from phylogenetic groups A (K279a) (a), B (N531) (b) and C (J675a) (c) were inoculated onto Iso-Sensitest agar using a 0.5 McFarland suspension. An ampicillin Etest strip was placed on the surface of each plate, which was then incubated for 18 h at 37 C. It has been previously shown that S. maltophilia colonies growing within the zone of clearing are stable b-lactam-resistant mutants. 5 Table 2. Variable sequence regions in the 16S rrna gene from the six 16S rrna groups of S. maltophilia isolates found in this study Group Isolates in group a , 12, 14, 16, 17, 23, caggagag tctct ttttt 26 32, 35 37, 39, , 25, 33, 44, 48 caggagag tctct tctgt 3 4, 9, 11, 13, 41, 42 cagaggag tcctt tctgt 4 15, 21, 38, 43, 46, 47, 49 cagtaagag tcttac ttttt 5 1, 2, 5 caaaggag tcctt tctgt 6 7, 8, 20, 50 cagaggag tcctt ttttt Mixed b 3, 6, 18, 19, 22, 34, 45 a Positions within the 500 bp PCR amplicon covering the 16S rrna gene hypervariable region. 4 b More than one possible nucleotide called at any position. S. maltophilia isolate D457, 8 meaning that 7 out of 50 isolates could not be typed using 16S rrna sequence with certainty (Table 2). Next we determined smed smet intergenic sequences for the 50 isolates, and used these data to produce a phylogenetic tree (Figure 2). This tree broadly defines four phylogenetic groups: the previously identified groups A, B and C, 8 and one additional group, named D. These data, when coupled with 16S rrna typing data in Table 2 reveal that all the isolates in 16S rrna type 1 fall into phylogenetic group A, which is the most populous group (22 out of 50 isolates) and contains isolates that are highly genetically homogeneous. Isolates from the other 16S rrna types are scattered around the four phylogenetic groups, with phylogenetic group C (8 out of 50 isolates) being most similar to group A, as previously defined by b-lactamase gene sequencing of the Bristol collection of isolates, 4 and phylogenetic group B (17 out of 50 isolate) being a highly heterogeneous group, with isolates sharing more limited smed smet intergenic sequence identity with those isolates in groups A and C. Group D is the least populous group, having 3 out of 50 isolates, and it is most similar to group B. An analysis of L1 and L2 b-lactamase inducibility was performed as described previously, 5 using nitrocefin as a substrate for b-lactamase quantification and EDTA to inhibit L1 b-lactamase and so reveal how much L2 (and by inference, L1) specific b-lactamase activity was present in each cell extract. The data for b-lactamase inducibility and per cent L2 activity in extracts K279a Group A ATCC J675a 41 Group D N531 Group C Group B of the 50 isolates revealed that the presence of a non-inducible L2 is not a general property of phylogenetically group B isolates as was previously suggested from our study of a more limited collection of isolates. 5 Indeed, only 4 out of 50 isolates have this phenotype, two from group A and one each from groups B and C. Furthermore, two isolates (both group B) express L2 inducibly, whereas L1 remains constitutive at low levels. This phenotype has not been previously reported but the fact that isolates exist with only one inducible b-lactamase adds further evidence to the hypothesis that mechanisms used to regulate L1 and L2 b-lactamase expression are separate. 5 In total, 12 out of 50 isolates were found to express b-lactamases constitutively at low levels (defined as having b-lactamase activity <20 U/mg of protein in extracts of cefoxitin-challenged cells). While phylogenetically group C isolates are most likely to express this phenotype (4 out of 8 isolates), members of groups A (2 out of 22) and B (6 out of 17) also do so. Thus again, our hypothesis Figure 2. Phylogenetic tree of S. maltophilia isolates based on smet smed intergenic sequence. Sequences of the intergenic regions from phylogenetic group-representative isolates K279a (Group A), N531 (Group B) and J675a (Group C), and the S. maltophilia type strain, ATCC 13637, have also been included for reference. The scale bar represents per cent nucleotide divergence between the smet smed sequences of the different isolates. Page 3 of 5

4 Gould et al. Table 3. b-lactam resistance phenotypes expressed by isolates from the three main phylogenetic groups of S. maltophilia found Percentage of isolates in group that are susceptible to Phylogenetic group cefuroxime aztreonam ampicillin piperacillin cefoxitin cefipime cefotaxime imipenem meropenem A(n = 22) B(n = 17) C(n = 8) concerning constitutive low-level expression being an exclusive property of group C isolates, based on a previous study of a limited number of strains 5 is disproven. Finally, constitutive overexpression of both L1 and L2 (defined as b-lactamase activity of >100 U/mg of protein in extracts from cells that have not been challenged with cefoxitin) was found in 5 out of 50 isolates, three from phylogenetic group A and two from group B. Overall, the average L1 b-lactamase activity in extracts of cefoxitin-challenged group A, B and C isolates (group D was excluded from this analysis because of the small sample size) was 54.2, 48.6 and 18.6 U/mg of protein and the average amount of L2 activity in these extracts was 78.8, 67.1 and 21.9 U/mg. Thus, the b-lactamase expression phenotypes of group A and B isolates were indistinguishable, but the levels of b-lactamase expressed in cefoxitin-challenged group C isolates were around one-third of those expressed by isolates from the other two main groups. MICs for the collection of 50 clinical S. maltophilia isolates were determined for representative b-lactams. MICs were first determined using Mueller Hinton agar (the method was agar dilution), which showed that all the isolates are resistant (according to BSAC breakpoints for Pseudomonas aeruginosa) 11 to all b-lactams tested. A strong effect of medium choice on b-lactam MIC has been reported when testing S. maltophilia isolates, 12 and using agar dilution and Iso-Sensitest agar (the medium prescribed in the BSAC methodology) 9 levels of resistance to b-lactams among the test collection of isolates were found to be dramatically reduced compared with those seen using the Mueller Hinton agar. From these Iso-Sensitest susceptibility data, the percentages of isolates from phylogenetic groups A, B and C (group D was again excluded) that are susceptible to the various antibiotics were calculated (Table 3). There is little difference between the percentages of isolates in phylogenetic groups A and B that are susceptible to each of the b-lactams tested. Group C isolates are generally more susceptible to the test b-lactams, with the exception of imipenem, to which all isolates tested were found to be resistant, and aztreonam, which is not a substrate of the L1 and L2 b-lactamases. 4 Of those compounds tested, the most broadly efficacious were piperacillin and meropenem, though it should be noted that all five constitutive high-level b-lactamase expressing isolates identified in this study (above) are resistant to these drugs. In conclusion, despite measurably high levels of phylogenetic heterogeneity between the two most populous phylogenetic groups of clinical S. maltophilia isolates (A and B) collected from distinct sites from Europe and North, South and Central America, b-lactamase expression and b-lactam resistance properties were, overall, indistinguishable. It was intriguing to find that group A isolates represent a highly homogeneous cluster, suggesting that these isolates have some property that makes them particularly suited to causing clinical infections. In contrast, group B isolates appear to represent sporadic instances where this common environmental bacterium has entered a clinical niche. The group A isolates were distributed equally on all continents of the study, and there is no obvious bias to any clinical source. Hence, the reason for the predominance of this cluster of isolates is worthy of future investigation and is one of the most important findings to come out of this study. It is tempting to speculate that group C isolates are considerably rarer than those from groups A and B as a cause of clinical infection because they are more susceptible to b-lactam antimicrobials, and that this increased susceptibility is due to the fact that a greater proportion of group C isolates express b-lactamases constitutively at low levels. We do not have antimicrobial prescribing data for any of the isolates in this study, so we cannot know the level of b-lactam use, and so test the theory that b-lactam resistance is important for determining successful infection by S. maltophilia. Others have reported prior b-lactam therapy as a risk factor for developing an S. maltophilia infection, 1 3 which does add evidence for our hypothesis that group C isolates are less common clinically because they are less b-lactam resistant. Acknowledgements b-lactamase induction research at the Bristol Centre for Antimicrobial Research and Evaluation is funded by the British Society for Antimicrobial Chemotherapy. We are grateful to Drs Mark Toleman and Tim Walsh (University of Bristol, UK) for providing the collection of S. maltophilia isolates. Transparency declarations None to declare (applies to all authors). References 1. Denton M, Kerr KG. Microbiological and clinical aspects of infection associated with Stenotrophomonas maltophilia. Clin Microbiol Rev 1998; 11: Senol E. Stenotrophomonas maltophilia: the significance and role as a nosocomial pathogen. J Hosp Infect 2004; 57: Sader HS, Jones RN. Antimicrobial susceptibility of uncommonly isolated non-enteric Gram-negative bacilli. Int J Antimicrob Agents 2005; 25: Avison MB, Higgins CS, von Heldreich CJ et al. Plasmid location and molecular heterogeneity of the L1 and L2 b-lactamase genes of Stenotrophomonas maltophilia. Antimicrob Agents Chemother 2001; 45: Avison MB, Higgins CS, Ford PJ et al. Differential regulation of L1 and L2 b-lactamase expression in Stenotrophomonas maltophilia. J Antimicrob Chemother 2002; 49: Page 4 of 5

5 b-lactam resistance in S. maltophilia 6. Berg G, Roskot N, Smalla K. Genotypic and phenotypic relationships between clinical and environmental isolates of Stenotrophomonas maltophilia. J Clin Microbiol 1999; 37: Valdezate S, Vindel A, Martin-Davila P et al. High genetic diversity among Stenotrophomonas maltophilia strains despite their originating at a single hospital. J Clin Microbiol 2004; 42: Gould VC, Okazaki A, Howe RA et al. Analysis of sequence variation among smedef multi drug efflux pump genes and flanking DNA from defined 16S rrna subgroups of clinical Stenotrophomonas maltophilia isolates. J Antimicrob Chemother 2004; 54: AndrewsJM. Determination of minimum inhibitory concentrations. J Antimicrob Chemother 2001; 48 Suppl S1: Higgins CS, Avison MB, Jamieson L et al. Characterization, cloning and sequence analysis of the inducible Ochrobactrum anthropi AmpC b- lactamase. J Antimicrob Chemother 2001; 47: MacGowan AP, Wise R. Establishing MIC breakpoints and the interpretation of in vitro susceptibility tests. J Antimicrob Chemother 2001; 48 Suppl S1: Bonfiglio G, Livermore DM. Effect of media composition on the susceptibility of Xanthomonas maltophilia to b-lactam antibiotics. J Antimicrob Chemother 1991; 28: Page 5 of 5