JAC Advance Access published online on November 14, 2008
Journal of Antimicrobial Chemotherapy, doi:10.1093/jac/dkn475
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original research |
Evidence for recombination among the alleles encoding TEM and SHV β-lactamases
School of Natural Sciences, University of California, Merced, CA, USA
* Corresponding author. Tel: +1-209-228-4174; Fax: +1-209-228-4060; E-mail: mbarlow{at}ucmerced.edu
Received 16 July 2008; returned 9 September 2008; revised 17 October 2008; accepted 22 October 2008
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
Objectives: The objective of this research was to determine whether recombination occurs in class A β-lactamases.
Methods: We performed 
2 analysis of the observed and expected numbers of times that β-lactamases from the TEM, SHV and CTX-M groups co-occurred. Additionally, we performed phylogenetic analysis to detect independent occurrences of silent mutations in blaTEM and blaSHV variants.
Results: We found that the distribution of co-occurring blaTEM, blaSHV and blaCTX-M alleles in clinical microbial populations is consistent with the regular occurrence of recombination among alleles within the groups. We also found that the distribution of silent mutations in blaTEM and blaSHV alleles is inconsistent with spontaneous point mutations.
Conclusions: Our findings indicate that recombination has an important effect on the sequence evolution and population distribution of the alleles that encode class A β-lactamases.
Key Words: penicillins , antibiotic resistance , ESBLs
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
The class A β-lactamases are encountered throughout the world at high frequency among clinical microbial populations, and they confer resistance to the heavily used β-lactam antimicrobials.1 Despite numerous molecular evolution studies of alleles encoding β-lactamases, there are few reports of recombinations2 occurring among them. Recent evidence that recombination occurs regularly among plasmidic qnr resistance alleles3 and experimental evidence that recombination can affect the resistance phenotypes present in a bacterial population4 indicate that it may be a regular occurrence among allelic variants of genes that are frequently located on plasmids.
The detection of recombination events in antimicrobial resistance genes is difficult because of strong directional selection that results in multiple independent emergences of identical mutations. However, extensive sequence data for allelic variants and reports of the population distributions of plasmidic alleles have made it possible to study the occurrence and the effects of recombination among alleles encoding class A β-lactamases.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
Phylogenetic reconstruction
Maximum likelihood phylogenies of blaTEM and blaSHV alleles were generated with Phyml v2.4.5 using a general time reversible substitution matrix, a discrete gamma model with four categories and an estimated shape parameter of 0.338 (blaTEMs) or 0.424 (blaSHVs). The proportion of invariant sites was estimated as 0.448 (blaTEMs) or 0.420 (blaSHVs). The blaSHV-2 allele was used as the outgroup for the blaTEM phylogeny and blaTEM-1 was used as the outgroup for the blaSHV phylogeny.
After constructing a blaSHV phylogeny, we found that the extended spectrum blaSHV-11 was the ancestor of the blaSHV alleles, including blaSHV-1. The ancestral allele blaSHV-11 was used as a reference sequence for identifying synonymous mutations. blaTEM-1 was used as the reference sequence for identifying synonymous mutations in blaTEM alleles.
To estimate the probability that variation between alleles in the TEM and SHV families has resulted from spontaneous point mutations, we performed a statistical analysis of the occurrence of point mutations throughout the blaTEM and blaSHV phylogenies. We computed the mean number of silent mutations per silent site and used those values to compute the expected number of times that mutations would occur at a single site in either the blaTEM or the blaSHV genes according to the Poisson distribution. The observed number of times that point mutations occurred at each site was inferred from the phylogenies and the observed and expected values were compared using a G-test. This test was performed for alleles from the TEM and SHV families. However, this analysis was not possible for the CTX-M family because of insufficient silent variation.
Data were compiled from 10 surveillance studies2,5–13 in which bacteria exhibiting the extended-spectrum β-lactamase (ESBL) phenotype were screened by isoelectric focusing or sequence analysis to identify genes conferring the ESBL phenotype. Since these surveys were performed in epidemiologically similar clinical environments and within 3 years of each other, the compilation of these datasets is legitimate. While the entire dataset includes frequencies of all classes of serine β-lactamases, our analysis included only class A β-lactamases. The aggregated dataset contains 291 strains. Of those, 163 strains expressed a single class A enzyme, 102 strains expressed two and 25 strains expressed more than two. One strain expressed no class A enzymes. The 25 strains that expressed more than two class A enzymes were excluded from the analysis because the sample size was insufficient for significant statistical analysis.
Three families of class A enzymes were identified among the isolates: TEM, SHV and CTX-M. The frequency of each family was computed among the isolates. The frequencies of each family were then multiplied to predict the frequencies of co-occurring alleles from the class A families. 2 analysis was used to compute the probability that the co-occurrence of alleles results from their frequencies in populations.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
We analysed the silent variation of blaTEM alleles to determine whether recombination is detectable. Non-synonymous variation was excluded from our analysis because some recurrent amino acid replacement could be selected as a means of adapting to new substrates and could therefore appear to arise more often than expected by chance alone. Silent mutations usually occur randomly and they usually are not subject to selection14 because they do not change the amino acid. Therefore, independent occurrences of silent spontaneous point mutations should be Poisson distributed.14 When an entire phylogeny is taken into consideration, it becomes possible to detect whether the occurrence of silent mutations is consistent with the Poisson distribution because they occur randomly and infrequently. Among all blaTEM alleles, there are 295 silent sites, 61 independent occurrences of silent mutations in our phylogeny and an average of 0.2068 silent mutations per silent site. Among the blaSHV alleles, there are 277 silent sites, 88 total silent mutations and an average of 0.3177 silent mutations per silent site. Using these values, we computed the expected number of occurrences of each silent mutation according to a Poisson distribution (Table 1). We compared the observed and expected values using a G-test. Those values were significantly different for both blaTEM (P = 6.13 x 10–97) and blaSHV (P = 1.03 x 10–39) alleles. This indicates that the observed pattern of silent mutations is not consistent with the occurrence of spontaneous point mutations. To eliminate the possibility that the difference may be an artefact caused by under-sampling silent mutations due to biases against reporting blaTEM and blaSHV alleles that contain only silent variations, we increased the mean number of silent mutations per silent site by an order of magnitude. This resulted in the observed pattern of silent mutations becoming even less likely to have occurred by silent point mutations for blaTEM (P = 1.20 x 10–272) as well as blaSHV (P = 2.01 x 10–292) alleles. These results indicate that the pattern of silent variation among blaTEM and blaSHV alleles cannot be explained by spontaneous point mutations. Another possible alternative is codon bias. However, we found that most of the silent mutations changed the codons from preferred to less preferred codons (Escherichia coli codon usage was used as our reference). The most reasonable alternative is that recombination is responsible for the seemingly independent occurrence of pre-existing point mutations.
|
To determine the amount of variation in blaTEM and blaSHV alleles that has resulted from recombination, we estimated the frequency of alleles that have arisen by spontaneous point mutations and the frequency of those that have arisen by recombination. There are 61 independent occurrences of silent mutations that occur at 18 sites in the blaTEM gene. There are six sites at which 46 of the 61 silent mutations occur (Table 1). The difference between the expected and the observed number of occurrences of mutations at those sites exceeds two orders of magnitude. By assuming that all variation at those sites, except for the first mutation, arose by recombination, we can infer that 40 mutations or 66% of silent variation among blaTEM alleles resulted from recombination occurring between variants. For the blaSHV alleles, 46 mutations or 52% of silent variation resulted from recombination occurring between variant blaSHV alleles.
Phylogenetic evidence indicates that recombination occurs regularly within blaTEM and blaSHV alleles. Therefore, recombination may result in other detectable patterns among microbial populations that carry multiple alleles encoding class A β-lactamases. For example, if two blaTEM alleles co-occur on a single plasmid and are oriented in the same direction, a single recombination event between them will remove one allele from the plasmid and one will remain. Depending on the recombination site, the remaining allele may be a chimera derived from the two alleles that previously co-occurred. If the co-occurring blaTEM alleles are oriented in opposite directions, a single recombination event between them will result in the rotation of the DNA segment between the recombination sites. By assuming that blaTEM allele orientation is random, that there is an equal probability of either orientation and that recombination occurs regularly among blaTEM alleles, one would predict that approximately half of all co-occurrences of blaTEM alleles would be subject to elimination by recombination.
To test this expectation, we compiled data from 10 surveillance studies and computed the frequencies of alleles from TEM, SHV and CTX-M families expressed among all isolates and multiplied them to predict the frequencies of co-occurring alleles. 2 analysis revealed that the observed frequencies of co-occurring alleles from a single family are significantly lower than expected and that the frequencies of co-occurring alleles from different families are significantly higher than expected (P = 1.64 x 10–7) (Table 2). This result rejects the hypothesis that the co-occurrences of class A β-lactamases depend upon their frequencies in microbial populations. Furthermore, we predicted that 
55 isolates should co-express two TEM enzymes and found that only 31 (56.4%) did. This finding is consistent with random orientation of co-occurring alleles because 1/2 of co-occurring alleles would be oriented in the same direction. These results indicate that recombination may limit or reduce the spread of some class A alleles through bacterial populations. These results support a recent experimental finding that recombination can eliminate one blaTEM allele from bacterial populations where two blaTEM alleles co-exist on a single plasmid.4 While that study provided laboratory-based experimental evidence of the effects that recombination can have, this study provides population-based evidence that recombination is an important process affecting the evolution of class A bla alleles in clinical populations.
|
|
Funding |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
This study was supported by start-up funds from the University of California, Merced.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
None to declare.
|
Footnotes |
|---|
Present address: Laboratory of Genetics, Wageningen University, Wageningen, The Netherlands. 
|
Acknowledgements |
|---|
We thank Barry Hall for suggestions about statistical analyses.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingTransparency declarationsReferences |
|---|
1 . Paterson DL, Bonomo RA. Extended-spectrum β-lactamases: a clinical update. Clin Microbiol Rev (2005) 18:657–86.
2
.
Baraniak A, Fiett J, Mrowka A, et al. Evolution of TEM-type extended-spectrum β-lactamases in clinical Enterobacteriaceae strains in Poland. Antimicrob Agents Chemother (2005) 49:1872–80.
3 . Baquirin MH, Barlow M. Evolution and recombination of the plasmidic qnr alleles. J Mol Evol (2008) 67:103–10.[Medline]
4
.
Mroczkowska JE, Barlow M. Recombination and selection can remove blaTEM alleles from bacterial populations. Antimicrob Agents Chemother (2008) 52:3408–10.
5
.
Decre D, Burghoffer B, Gautier V, et al. Outbreak of multi-resistant Klebsiella oxytoca involving strains with extended-spectrum β-lactamases and strains with extended-spectrum activity of the chromosomal β-lactamase. J Antimicrob Chemother (2004) 54:881–8.
6
.
Brinas L, Lantero M, de Diego I, et al. Mechanisms of resistance to expanded-spectrum cephalosporins in Escherichia coli isolates recovered in a Spanish hospital. J Antimicrob Chemother (2005) 56:1107–10.
7
.
Ho PL, Ho AY, Chow KH, et al. Occurrence and molecular analysis of extended-spectrum β-lactamase-producing Proteus mirabilis in Hong Kong, 1999–2002. J Antimicrob Chemother (2005) 55:840–5.
8
.
Ho PL, Shek RH, Chow KH, et al. Detection and characterization of extended-spectrum β-lactamases among bloodstream isolates of Enterobacter spp. in Hong Kong, 2000–2002. J Antimicrob Chemother (2005) 55:326–32.
9
.
Kruger T, Szabo D, Keddy KH, et al. Infections with nontyphoidal Salmonella species producing TEM-63 or a novel TEM enzyme, TEM-131, in South Africa. Antimicrob Agents Chemother (2004) 48:4263–70.
10
.
Kim S, Kim J, Kang Y, et al. Occurrence of extended-spectrum β-lactamases in members of the genus Shigella in the Republic of Korea. J Clin Microbiol (2004) 42:5264–9.
11
.
Schlesinger J, Navon-Venezia S, Chmelnitsky I, et al. Extended-spectrum β-lactamases among Enterobacter isolates obtained in Tel Aviv, Israel. Antimicrob Agents Chemother (2005) 49:1150–6.
12 . Tasli H, Bahar IH. Molecular characterization of TEM- and SHV-derived extended-spectrum β-lactamases in hospital-based Enterobacteriaceae in Turkey. Jpn J Infect Dis (2005) 58:162–7.[Medline]
13
.
Lavigne JP, Bouziges N, Chanal C, et al. Molecular epidemiology of Enterobacteriaceae isolates producing extended-spectrum β-lactamases in a French hospital. J Clin Microbiol (2004) 42:3805–8.
14 . Kreitman M. Methods to detect selection in populations with applications to the human. Annu Rev Genomics Hum Genet (2000) 1:539–59.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||