Skip Navigation


JAC Advance Access originally published online on January 31, 2006
Journal of Antimicrobial Chemotherapy 2006 57(3):450-460; doi:10.1093/jac/dki492
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
57/3/450    most recent
dki492v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Olsen, J. E.
Right arrow Articles by Aarestrup, F. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Olsen, J. E.
Right arrow Articles by Aarestrup, F. M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org

Diversity and evolution of blaZ from Staphylococcus aureus and coagulase-negative staphylococci

John Elmerdahl Olsen1,*, Henrik Christensen1 and Frank Møller Aarestrup2

1 Department of Veterinary Pathobiology, The Royal Veterinary and Agricultural University, 4 Stigbøjlen, DK-1870 Frederiksberg C., Denmark; 2 Danish Institute for Food and Veterinary Research, 27 Bülowsvej, DK-1790 Copenhagen, Denmark

* Corresponding author. Tel: +45-35282784; Fax: +45-35282757; E-mail: jeo{at}kvl.dk

Received 6 July 2005; returned 1 November 2005; revised 13 December 2005; accepted 15 December 2005


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Transparency declarations
 Supplementary data
 References
 
Objectives: To elucidate the diversity and evolutionary history of plasmid- and chromosomally-located blaZ, to detect indications of frequent exchange of blaZ between human and bovine staphylococci and to estimate the frequency of transfer of blaZ between coagulase-negative staphylococci (CoNS) and Staphylococcus aureus of bovine origin.

Methods: blaZ was detected in 143 strains of penicillin-resistant S. aureus and CoNS from five Danish cattle herds (n = 25/23), random CoNS isolates from Denmark (n = 37), a collection of S. aureus from six different countries (n = 52), humans in Denmark (n = 3) and ß-lactamase control strains (n = 3). The sequence was determined in 105 strains and compared to published sequences by pairwise and multiple alignments. Maximum likelihood analysis was performed including bootstrap analysis. Parsimony, neighbour joining and consensus comparisons were performed for recombination. The localization of blaZ was determined by Southern blotting in 108 isolates.

Results: All penicillin-resistant strains carried blaZ and showed a similar organization of blaR1 and blaZ. The blaZ gene was localized to a plasmid in only 16 of the resistant strains. Sixty-nine sequences representing 105 isolates and sequences retrieved from public databases were compared. A phylogenetic tree showed that blaZ exists in three evolutionary lines: one group was of plasmid origin, one group was of chromosomal origin and one intermediate group. Sixty-nine sequence types were demonstrated. They translated into 11 BlaZ protein types. The major types all contained strains of both human and bovine origin, and more than one Staphylococcus species, demonstrating a shared gene pool. In a comparison of S. aureus and CoNS obtained from five Danish cattle herds, the same type of blaZ was only detected in one case.

Conclusions: Results indicated a separate evolution for plasmid- and chromosomally-encoded blaZ. Although a common gene pool seems to exist among staphylococci, exchange of blaZ between strains and species is judged to be an extremely rare event.

Keywords: mastitis , penicillin , ß-lactamases


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Transparency declarations
 Supplementary data
 References
 
Staphylococcus aureus is an important human pathogen and one of the most important causes of bovine mastitis worldwide. Penicillin has been the drug of choice for treatment of infections caused by this organism; however, resistant S. aureus were reported as early as 1944.1 Today ~90% of human S. aureus are penicillin resistant2 and prevalence of penicillin-resistant bovine isolates vary from 10 to 70% according to geographic location.3,4

Two mechanisms confer penicillin resistance in staphylococci. The most important is production of ß-lactamase, which inactivates penicillin by hydrolysis of its ß-lactam ring. The second is primarily associated with human isolates and confers resistance due to a penicillin-binding protein, PBP2a, encoded by mecA.5 blaZ-encoded penicillin resistance has been thoroughly investigated. The genes (the structural gene blaZ, its repressor gene blaI, and a signal transducer-sensor protein, encoded by blaR1) are clustered together.6 Four types of blaZ product (A, B, C, D) have been distinguished by serotyping and differences in hydrolysis of selected ß-lactam substrates.7,8 Types A, C and D are usually located on plasmids, whereas type B typically resides in the chromosome.9

blaZ has also been identified as the cause of penicillin resistance among coagulase-negative staphylococci (CoNS) as well as in porcine Staphylococcus hyicus10–13 suggesting that blaZ is the main mechanism of penicillin resistance in staphylococci. Transfer of resistance genes between CoNS and S. aureus has been reported14–18 indicating that CoNS may act as a resistance gene reservoir for S. aureus. It is thus possible that the different species of staphylococci that are present in the same microenvironment, for example on the skin of dairy cows, can exchange blaZ, if the appropriate bacterial factors are met.

In the present study, we have used sequence comparison to investigate the diversity of blaZ in mainly bovine staphylococci, and used this to give presumptive answers to three important questions: (i) has the evolutionary history of plasmid- and chromosomally-located blaZ been the same?; (ii) is there indication of transfer of human derived blaZ into the bovine population of staphylococci?; and (iii) does transfer of blaZ happen between CoNS and S. aureus of bovine origin within the same epidemiological unit?


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Transparency declarations
 Supplementary data
 References
 
Bacterial isolates

A total of 187 staphylococcal isolates (36 penicillin-susceptible CoNS, eight susceptible S. aureus, 60 penicillin-resistant CoNS and 83 penicillin-resistant S. aureus) were used in the investigation. The strains originated from four sources (Table 1). A total of 68 bovine staphylococci were isolated from quarter milk samples in five average size (57–80 cows) Danish dairy herds (herds A, B, C, D and E), geographically spread in Denmark to investigate within and between herd sequence variations in blaZ. The herds were selected on the basis of a high prevalence of penicillin resistance among both S. aureus and CoNS. Quarter milk samples were sampled by quality advisers from the Danish Cattle Health Laboratory, Ladelund, in relation to routine bacterial testing of dairy herds. Four to six penicillin-resistant S. aureus and, if possible the same number of penicillin-resistant CoNS, were selected from each herd, in addition to a few susceptible isolates. When it was possible, one S. aureus and one CoNS were selected from the same cow and preferably from quarters with the highest cell count as determined by California Mastitis Test.


View this table:
[in this window]
[in a new window]
  		Table 1.. Origin of strains investigated

 
Twenty-four penicillin-susceptible CoNS and 37 penicillin-resistant CoNS from Denmark, sampled in 2000 and 2001 by a pseudo-random process at the Cattle Health Laboratory with only one isolate per herd,19 were included to cover genetic diversity within staphylococci in Denmark. These isolates were especially used for the study of the chromosomal or plasmid location of the penicillin resistance gene blaZ in CoNS.

To cover the genetic diversity of blaZ in general, 52 penicillin-resistant bovine S. aureus isolates from six countries [England (3 isolates), Iceland (3 isolates), Norway (2 isolates), Sweden (38 isolates), the USA (3 isolates) and Zimbabwe (3 isolates)] previously described4,20 were included together with three Danish human S. aureus carrier-isolates selected from a previously described collection of human healthy carrier strains.21

Three S. aureus staphylococcal ß-lactamase-producing control strains of type A (E19771), type B (E19778) and type C (E19658) isolated from blood in three Danish human patients (Robert Skov, personal communication) acted as control strains.

Bacterial identification

All isolates were identified to species level using biochemical testing as previously described.22 The 10 main groups of staphylococci were further verified by 16S rDNA sequencing according to Aarestrup et al.23

Antimicrobial susceptibility testing

Confirmation of susceptibility/resistance to penicillin was performed using Oxoid's Antimicrobial Susceptibility Test Discs containing 10 IU penicillin G. Furthermore all isolates were tested for their ability to produce ß-lactamase enzyme using ß-lactamase diagnostic tablets, according to the manufacturer's guidelines (A/S Rosco, Tåstrup, Denmark).

PCR amplification of blaZ

The oligonucleotide primers used for amplification of blaZ were designed according to previously published sequences of blaZ with accession numbers AF086644 [GenBank] , U58139 [GenBank] , X16471 [GenBank] , X52743 [GenBank] , X04121 [GenBank] and M25252 [GenBank] , M25253 [GenBank] , M25254 [GenBank] and M252525. The sequences were aligned using DNASIS software version 2.5 (Hitachi Software Engineering Co., Ltd) and primers were selected in regions without sequence variation. All isolates were tested using the primer pairs 486–488, 487–373 and 487–531 shown in Table 2 and Figure 1. As a supplement to the primer pair 487–531, two other primers (489 and 374) were designed according to the DNA sequence with accession no. X52734 [GenBank] and used on a limited number of isolates together with primer 487 to amplify the last 302 bp of blaZ. Sequences for primers used were (5'–3'): 373, TTAAAGTCTTACCGAAAGCAG; 374, CTCGAAAATAATAGAGGGAAA; 486, GTTGCGAACTCTTGAATAGG; 487, TAAGAGATTTGCCTATGCTT; 488, GGAGAATAAGCAACTATATCATC; 489, CAACATCATTTCTAGAAGCA; and 531, AATTCCTTCATTACACTCTTGG.


View this table:
[in this window]
[in a new window]
  		Table 2.. PCR methods used to characterize blaZ from penicillin-resistant staphylococci

 

Figure 1
View larger version (15K):
[in this window]
[in a new window]
  		Figure 1.. Structure of the sequenced fragment encoding the blaR1 and blaZ region and number of strains positive with each PCR method. B, bovine strains; H, human strains; DK, Denmark; Ic, Iceland; Sw, Sweden; USA, United States of America.

 
DNA for PCR detection of blaZ was obtained by a boiling lysate method as previously described.24 Each PCR reaction contained 40 µL of sterile water, 5 µL of 10x PCR buffer (H.T. Biotechnology Ltd), 0.5 µL of dNTP mix (Pharmacia), 0.5 µL of each primer, 0.1 µL of super taq DNA polymerase (H.T. biotechnology Ltd) and 2 µL of boiling lysate. The MgCl2 concentration was adjusted by adding extra 1.5 mM (primer pair 486–488), 1.25 mM (primer pair 487–373) and 1.0 mM (primer pair 487–531) MgCl2 (Roche). The PCR mixture was subjected to 35 cycles of amplification. The conditions for each cycle were: denaturation for 1 min at 94°C, annealing for 1 min at 54°C, and primer extension for 1 min at 72°C. Finally the reaction was incubated at 72°C for 10 min. The PCR products were separated by electrophoresis in a 1.5% agarose gel, stained with ethidium bromide, visualized in UV-light and image analysed on Gel-Doc 2000 (Bio-Rad).

DNA sequencing

The nucleotide sequences of all positive amplification products of blaR1 and blaZ genes were determined by cycle sequencing25 according to the manufacturer's instructions, using an AmpliTaqFS dye terminator kit and a 373A automatic sequencer (Applied Biosystems/Perkin-Elmer, Foster City, CA). The primers used for the methods listed in Table 2 were used for sequencing. Sequences were obtained in both orientations and overlapping amplicons were sequenced giving several independent sequences of the same areas. All sequences were assembled and analysed using the AutoAssemblerTM software version 2 supplied by PE Applied Biosystems. DNA sequences were deposited under accession numbers AY369343–AY369356 and DQ016039–DQ016069. Details appear in the Supplementary data vailable at JAC online.

Similarity analysis of blaR1 and blaZ

Searches for sequence data in GenBank26 were performed by BLAST.27 Pairwise comparisons for similarity were performed by Bestfit (Wisconsin Sequence Analysis Package Genetics Computer Group, Madison). Multiple alignments were constructed by ClustalX.28 The sequence with accession no. X52734 [GenBank] was used as a reference. Maximum likelihood analysis was performed by fast-DNAml including bootstrap analysis29 run on a Linux 7.2 compatible server. The transition/transversion ratio was found optimal when set to 1.6. Parsimony, neighbour joining and consensus comparisons were computed by PHYLIP.30 Tests for recombination were performed by PLATO run on Linux (http://evolve.zoo.ox.ac.uk/software/plato/manual.php).31 SimPlot 2.5 and BootScan (http://sray.med.som.jhmi.edu/SCRoftware/simplot/) were also used to analyse mosaic patterns in the R1, spacer and blaZ coding regions. Default settings of the program were used.

Location of blaZ in penicillin-resistant strains

Purification and digestion of chromosomal DNA. DNA was purified using the High Pure PCR Template Preparation Kit (Roche Diagnostics GmbH, Germany) according to the instruction manual (version 2, September 1999). For lysis of staphylococci, 10 µL of lysostaphin (10 mg/mL) and 15 µL of lysozyme (10 mg/mL) were added and incubated at 37°C for 45 min. Chromosomal DNA was digested with one unit of EcoRI (Phamacia) for 2 h and run on a 0.8% Tris–borate EDTA agarose gel. EcoRI has no restriction sites in the blaZ gene. HindIII-digested lambda-DNA served to determine the size of the digested chromosomal fragments.

Purification and digestion of plasmid DNA. All 187 isolates, except for 35 Swedish S. aureus, were analysed for the presence of plasmids using the Qiagen-tip 20 kit (Qiagen GmbH, Germany) according to the Qiagen Plasmid Mini Handbook. For the lysis of staphylococci 70 µL of lysostaphin (10 mg/mL) and 30 µL of lysozyme (10 mg/mL) were added and incubated at 37°C for 45 min. Purified plasmid DNA was stored at –21°C until further testing. Plasmid DNA was digested using one unit of EcoRI (Phamacia) for 2 h and run on a 0.8% Tris–borate EDTA agarose gel. Plasmids in Escherichia coli V517 and E. coli 39R861 served as molecular weight markers for size determination. HindIII-digested lambda-DNA served to determine the size of the digested plasmid fragments.

DNA-blotting and blaZ hybridization. DNA bands from chromosomal and plasmid digests were transferred from agarose gels onto a Hybond-N membrane (Amersham, Arlington Heights, IL, USA) by Southern blotting as described by Skov et al.32 using 10x SSC as transfer buffer. A probe (675 bp) for the blaZ gene was prepared by PCR amplification on the wild-type S. aureus isolate 3833 using the primers 487 and 531. Probe DNA was purified using Qiagen spin columns (Hilden, Germany) and labelled with digoxigenin-11-dUTP using a DNA labelling and detection kit (Boehringer, Mannheim, Germany). Hybond-N membrane filters were hybridized at 60°C and development of membranes was done by use of a digoxigenin staining kit according to the manufacturer's recommendations (Boehringer, Mannheim).


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Transparency declarations
 Supplementary data
 References
 
Presence of blaZ in penicillin-resistant staphylococci

As expected, no PCR products were obtained from the 44 penicillin-susceptible isolates included as controls, allowing us to conclude that positive PCR bands would be strongly indicative of the presence of blaZ. All 143 penicillin-resistant isolates showed identical size amplification products with the primer sets 486–488 and 487–373 (Figure 1), demonstrating the presence of the first 209 bp of blaR1 separated by a 106 bp non-coding intergenic region from the first 543 bp of blaZ. In contrast, only 7 of 83 Danish penicillin-resistant bovine staphylococci (60 CoNS and 23 S. aureus) were positive with the primer set 487–531. In order to amplify this last part of the gene, additional testing of the 76 remaining isolates was performed with primers 487–489 and 487–374. As summarized in Figure 1, this yielded no amplification products, and hence only the region spanning from primer 486 to primer 373 was available for sequence comparison. Thirty-one out of 38 Swedish isolates were positive by the additional primer pairs 487–489 and 487–374, three of which were positive with the primer pair 487–531, as well, demonstrating that the lack of amplification in Danish isolates with these primer pairs was caused by a strain-specific factor. All of these methods also worked with three Danish isolates from human carriers, three isolates from Iceland and two isolates from the USA.

Determining the localization of blaZ by Southern blotting

The blaZ probe hybridized to a plasmid band in 16 of the 108 penicillin-resistant strains examined (Table 3). Five of these were S. aureus from the same Danish cattle herd, and judged from the restriction profile of the plasmid and the size of the band carrying the blaZ gene, probably clonal. blaZ was located in the chromosome in the remaining 92 strains (data not shown). blaZ was located on different size plasmid and chromosomal fragments, both among S. aureus and the different CoNS species (Table 3), suggesting that blaZ was not encoded from the same genetic element in neither plasmids nor the chromosome, except for strains from one Danish cattle herd and two Danish human isolates.


View this table:
[in this window]
[in a new window]
  		Table 3.. Sizes of plasmids and EcoRI fragments in staphylococci shown by Southern analysis to carry the blaZ gene

 
Phylogenetic analysis of sequences obtained from the blaZ region

Since the sequences available for comparison were of highly variable lengths, the phylogenetic analysis was made in two steps. First the 21 longest sequences were compared and phylogenetic analysis and groups defined. All sequences were then compared and phylogenetic groups identified in relation to the first analysis.

The alignment of 21 of the longest sequences comprised 1071 positions including 184 bp of the 3' end of the R1 region, the 106 bp long spacer region and 781 bp of the blaZ gene (out of 843 bp in total). Within the R1 region, a region of 21 bp was deleted in the sequence with accession no. AY373761 [GenBank] and this region was excluded from the analysis. Three groups were identified based on sequence similarities. One group contained only isolates where the blaR1–spacer–blaZ region was located in the chromosome (Figure 2a, group C), one group contained only isolates that had the region located on plasmids (Figure 2a, group P) and the last group contained a mixture of isolates with either chromosomal or plasmid location of the blaZ region (Figure 2a, group PC). When each of the three regions, R1, spacer and blaZ were analysed separately, congruence was for the most found to the phylogeny based on the whole region shown in Figure 2(a). Five sequences did not show congruence. Sequence AE015934 [GenBank] fell between the groups when the full sequences were analysed (Figure 2a). In the phylogeny based on R1 and spacer regions, it was located close to the PC group whereas it was located in the P group when the blaZ region was analysed alone (Figure 2b). The group with Sw41, Island1, NVH96 and pUB101 was in the P group in the full alignment (Figure 2a), and in the R1 and spacer phylogenies but in the PC group in the blaZ alignment (Figure 2b). These observations were also made when data were analysed by neighbour joining and maximum parsimony analysis.


Figure 2
View larger version (15K):
[in this window]
[in a new window]
  		Figure 2.. Phylogenetic analysis of sequences of the blaZ region [blaR1–spacer–blaZ (a) and blaZ only (b)] in staphylococci. P, blaZ located on plasmids in this group; C, blaZ located in the chromosome in this group, PC, blaZ located on either plasmids or in the chromosome in this group. Strains in bold have been sequenced in the present investigation and enclosure of strain names in brackets indicates sequence identity of the previous listed strain. Numbers represent bootstrap support values of 100 repeats for (a). For (b), bootstrap analysis could not be performed due to too few differences. *Isolates with deviating position when the blaZ region was analysed alone.

 
The alignment of all isolates comprised 69 sequences, representing 105 isolates, with 119 positions available for comparison of the region 101–219 of blaZ. Seventy-eight isolates were sequenced as part of the current investigation and 27 were retrieved from public databases. This included two blaZ sequences obtained from penicillin-resistant Enterococcus faecalis for comparison. Identical sequences had been grouped before analysis and each type was only represented by one sequence. Neighbour joining analysis of this alignment also identified the three groups (C, P and PC) (data not shown). Analysis of the partial sequences of all isolates could not be performed by maximum likelihood since it was impossible for the heuristic tree search algorithm to find the best tree with the few sequence differences available. The analysis was instead repeated with maximum parsimony analysis and the three groups confirmed. No significant recombination between the regions of blaR1, spacer and blaZ was detected by PLATO. To supplement this test, SimPlot and BootScan tests were used in order to analyse each of the divergent sequences pSE12228, Island1, Sweden41, NVH96 and pUB101 compared to the P, PC and C phylogenetic groups. The four sequences, Sweden41, Island1, NVH96 and pUB101 included with the P group in the phylogeny of R1, spacer and blaZ, showed pair-wise similar patterns in the SimPlot and BootScan analysis without obvious indications for divergence between regions (data not shown). SimPlot analysis of the sequence of pSE12228 included in the PC group indicated that R1 and spacer were of PC type but the blaZ region was of P type (Figure 2b). BootScan analysis confirmed this result. This documents a potential for recombination between two evolutionary different types of plasmid-located R1–spacer–blaZ regions.

Allele and signature types of the blaZ region

The isolates characterized in the present investigation, as well as published sequences of the blaZ region, were grouped according to sequence type. Sixty-nine alleles were observed, and most alleles (n = 60) were only represented by a single isolate and only three alleles with five or more isolates, indicating a high degree of sequence variation in the sequenced region of blaZ. In order to break down the sequence variation into fewer types, deduced protein signatures were produced, and isolates were grouped according to similarity of protein type. In this comparison, a conservative approach was used. Hence a signature type was made up of proteins with three or less deviations in amino acid composition. Eleven types were observed. The signature amino acids of these are listed in Table 4. Table 5 summarizes the number of allele type, strains and origin of strains within each protein sequence type. A detailed list of all strains investigated and the sequence type detected in these are available as Supplementary data at JAC online.


View this table:
[in this window]
[in a new window]
  		Table 4.. Protein signature types of BlaZ

 

View this table:
[in this window]
[in a new window]
  		Table 5.. blaZ alleles and strains grouped according to protein signature type

 
Signature type 6 was the most commonly demonstrated, covering 25 allele types and three species of staphylococci. This type only included strains belonging to the C group in the phylogenetic analysis and it contained all but three strains of this group. The plasmid group (P), on the other hand was composed of four signature types (2–5) with 22 alleles and contained four species of Staphylococcus in addition to an E. faecalis derived sequence. The PC group contained the signature type 1 only, demonstrated in three species and containing 16 different sequence types, one of which was unique to E. faecalis. The three most common signature types (1, 3, 6) all contained both human and bovine isolates. However, signature type 6 was mainly from bovine isolates in the current investigation (50 out of 54 strains), while signature type 1 was demonstrated mostly in human isolates (14 out of 16).

Distribution of blaZ in isolates obtained from Danish dairy farms

One aim of the study was to investigate whether S. aureus and CoNS from the same epidemiological unit, in this case diary cattle herds, would appear to exchange and therefore share blaZ types. Five signature types and twenty-two alleles were found in the five herds and only allele types 18 and 24 were found in more than one herd (Table 6). Allele 18 was by far the most widely distributed and represented 10 isolates from four different herds, all CoNS isolates. It was not otherwise recorded in the study and seemed to be a local clone in Denmark.


View this table:
[in this window]
[in a new window]
  		Table 6.. blaZ characteristics in isolates obtained from the same epidemiological unit (dairy cattle herds) in Denmark

 
blaZ was located on different sized EcoRI fragments in S. aureus and CoNS in three of the five herds, suggesting that blaZ was located in different genetic surroundings in these bacteria in these herds. However, the same signature/allele combination was detected in strains where the hybridizing fragment was of different size (e.g. signature type 6/allele type 18 in different herds), indicating that this was not a firm proof of lack of exchange. While strains isolated from the same herd mostly were of the same signature type, they were only in one instance of the same combined signature/allele type in S. aureus and CoNS, suggesting that S. aureus and CoNS were normally of different origin within herds (Table 6). The one instance was in herd A, where the five S. aureus carried blaZ on EcoRI chromosomal fragments of 9 or 12 kb and blaZ were of signature and allele types 6/11 and 6/14. The 6/11 type was also demonstrated in one strain of Staphylococcus haemolyticus, while the remaining three S. haemolyticus and one Staphylococcus epidermidis carried blaZ on 7 kb EcoRI fragments and were of type 6/22. One isolate of Staphylococcus xylosis carried blaZ on a plasmid and had the 5/47 type.

blaZ was mostly found to be located in the chromosome (see Supplementary data available at JAC online). In herd D, however, all S. aureus strains carried blaZ on a plasmid of ~24 kb with identical restriction profiles, with blaZ located on an 18 kb EcoRI fragment and with signature type 3 (four different allele types, including type 37). This plasmid size and location of blaZ on an 18 kb EcoRI fragment was also found in two Danish human strains (see Supplementary data available at JAC online). blaZ in these strains were of signature/allele types 3/37 and 3/38. In contrast, the three S. haemolyticus and one S. epidermidis isolate from this herd carried blaZ on chromosomal fragments.

In two cases it was possible to isolate penicillin-resistant S. aureus and CoNS from the same cow teat. Strains A1358-1 (S. aureus) and A1358-2 (S. xylosis) from herd A fell in the PC and P phylogeny groups, were of two different signature types and S. xylosis carried blaZ on a plasmid while the gene was located in the chromosome in S. aureus. Strain D3950-1 (S. aureus) and D3950-2 (S. haemolyticus) from herd D fell in the P and C phylogeny groups, S. aureus carried blaZ of signature type 3 on a plasmid, and S. haemolyticus carried blaZ of signature type 6 in the chromosome.


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Transparency declarations
 Supplementary data
 References
 
The aim of this study was to answer three basic questions with regard to evolution of the penicillin resistance gene blaZ in staphylococci: (i) has the evolutionary history of plasmid- and chromosomally-located blaZ been the same?; (ii) is there indication of transfer of human derived blaZ into the bovine population of staphylococci?; and (iii) does transfer of blaZ happen between CoNS and S. aureus of bovine origin within the same epidemiological unit?

Since the penicillin resistance-encoding gene blaZ can be located on mobile elements, such as transposons, insertion sequences (IS) and plasmids,8 a comparison of penicillin-resistant isolates obtained from different sources by molecular typing would not give answers to the questions. In such a situation, investigation of spread of genes across species and between strains of the same species had to rely on comparison of the gene itself and its immediate surroundings. We, therefore, used bioinformatics tools combined with traditional microbiological techniques to obtain answers.

All penicillin-resistant strains carried blaZ, and the beginning of blaR1 was separated from blaZ with the same highly homologous spacer region. This indicates that blaZ-encoded penicillin resistance is due to the same gene in all strains and organized in a similar way as has previously been described.33 For 16 out of the 21 sequences available for analysis of the R1, spacer and blaZ regions, phylogenies were congruent for all three regions. Exceptions were observed for five sequences. Out of these, only one (pSE12228) could be confirmed by SimPlots and BootScan analysis for recombination. In this case, recombination might have occurred between two plasmids in the region between the spacer and the blaZ gene resulting in a sequence being a chimera of R1–spacer of PC type and blaZ of P type. For the four other divergent sequences results could also be explained by recombination in the region between the spacer and the blaZ gene, and in this case as chimeras of R1 and spacer regions of P type and blaZ of PC type. Based on the alignments of parts of the blaZ gene and blaR1 gene and the complete intergenic region, it is suggested that the blaZ gene in S. aureus, S. haemolyticus and S. epidermidis isolated from humans, cows and pigs (data not shown) has a common ancestor gene.

The initial amplification step before sequencing suggested that divergent sequences exist. Amplification of the last part of blaZ was not possible in the majority of Danish strains, even after several modifications of the primers based on published sequences. The reason for the inability to amplify this region remains unsolved. The strains were penicillin resistant, and we assume that this is due to blaZ-encoded ß-lactamase. It may be possible to develop a PCR-based typing system based on this region, but such work was beyond the scope of this article.

The majority of isolates characterized were of bovine origin, and a large proportion were epidemiologically related, that is isolated from the same cattle herds. With the strain material in mind, conclusions must be taken with some care. However, a surprisingly high number of sequence types were observed (n = 69), despite the epidemiological relation between some strains, indicating a high degree of variability in the sequenced area.

The bovine S. aureus strains were found mostly to carry blaZ in the chromosome. A similar observation was made by Vesterholm-Nielsen et al.11 and Yazdankhah et al.12 This is in contrast to a report on human isolates, where a shift from chromosomally-encoded to plasmid-encoded blaZ apparently has happened. Skov et al.32 found that isolates of phage group II isolated prior to 1977 harboured blaZ on the chromosome, whereas it was plasmid-encoded in 84% of isolates in 1990. Phage group II has been the most predominant phage group in human isolates of S. aureus in Denmark since the late 1980s,2 whereas it is rarely found among bovine isolates.34 A large number of types can cause bovine mastitis, but there is a tendency that a few successful clones dominate over a large geographical area.34 It may therefore be that a clone-shift in the future, as has been observed among human strains, may change the tendency from a chromosomal to a plasmid location in bovine strains. Only 6 out of 63 penicillin-resistant CoNS carried blaZ on a plasmid, and the chromosomal location thus presently seems the most frequent in all bovine staphylococci. Misclassification might possibly happen in cases where blaZ was carried on plasmids too large to be isolated by the used protocol, since fragments from such plasmids might be present in the preparation of chromosomal DNA. However, plasmids in S. aureus are usually <25 kb and very rarely up to 100 kb,35,36 and as discussed below, plasmids carrying blaZ are usually ~20 kb. According to the Qiagen Plasmid Mini Handbook for the Qiagen-tip 20 kit, plasmids up to ~150 kb can be purified using Qiagen plasmid purification protocols, and we have ourselves successfully used the Qiagen kit for purification of plasmids of sizes up to ~175 kb.37

A retrieved sequence from a human isolate of S. aureus with plasmid-encoded blaZ (pMW2) (accession no. AP004832 [GenBank] ) aligned perfectly with four of the blaZ of the Danish strains from herd D. These strains carried blaZ on a 24 kb plasmid. The pMW2 plasmid is 20.6 kb and was isolated from a community-acquired methicillin-resistant S. aureus from the USA.38 Furthermore, Doebbeling et al.39 showed apparently identical 22 kb plasmids conferring penicillin resistance in isolates of S. aureus from 20 hospitals in Europe, the USA and Brazil representing 20 hospitals sampled on each of the three continents. Together these findings suggest that certain plasmids, including the plasmid from the Danish herd, have been successfully spread and are a common reason for plasmid-encoded penicillin resistance. Plasmids that encoded blaZ in S. aureus were found not to be identical to plasmids encoding resistance in CoNS, nor was blaZ of the same allele type or carried on the same size EcoRI fragments, indicating that different species of staphylococci carry individual species of blaZ-encoding plasmids.

A remarkable low DNA similarity of 94.5% between the blaZ sequence obtained from the chromosomes and the plasmids was observed. Due to this, the phylogenetic analysis clustered plasmid-encoded and chromosomally-encoded blaZ separately. The two groups were joined by a smaller group, which contained strains with either chromosomally- or plasmid-encoded resistance. This was demonstrated by analysis of nearly full-length sequences of blaZ in addition to a high number of shorter sequences, and was found to hold true for strains isolated over a large time span and from a wide geographic origin. This is a strong indication that the evolution of the plasmid-encoded and chromosomally-encoded blaZ has been separate for a time period. The practical implication of this is that transfer of blaZ in either direction is apparently an extremely rare event. It is noteworthy that 20 out of 22 previously published blaZ sequences aligned with the plasmid-like sequences. This probably reflects that the focus has been on spread of blaZ among human nosocomial isolates. The focus of the current study was on dairy cows and human isolates were not systematically included. In future studies it will be relevant to include isolates from community-acquired infections in humans including isolates from humans in contact with cows.

The high number of sequence types translated into a much more limited number of blaZ protein types, indicating a strong selection pressure for maintenance of the protein structure. All types produced functional proteins according to analysis of open reading frames within the sequenced region. Each of the dominating types contained strains of both bovine and human origin and were demonstrated in several species of staphylococci. As already mentioned above, this indicates that the gene pool is probably shared between all staphylococci, and related bacteria. In this investigation, blaZ sequences of signature types 1 and 5 were also retrieved from E. faecalis. This confirms previous reports that the blaZ carried by some strains of E. faecalis is identical to staphylococci blaZ of Tn552 or blaZ of pUB10,40 yet with a tendency to be constitutively expressed due to truncation of the repressor gene.41 Thus, the distinct evolutionary lines of blaZ relate to plasmid or chromosomal location, but neither to the host from which the staphylococci were isolated nor to the species of staphylococci or related bacteria.

The sequences of ß-lactamase reference strains aligned with signature type 11 (PC group) and type 3 (P group) for type A and C, known to be chromosomally-encoded, and with type 6 (C) for the type-B ß-lactamase, previously shown to reside on the chromosome of phage group II isolates.42 A type D reference strain was not included. The published sequence (M25257 [GenBank] ) represents a strain that produces staphylococcal type D ß-lactamase.43 This sequence was of signature type 5 and clustered in the plasmid group by phylogenetic analysis.

The possible role of transposons and IS is evident from the alignments of previously published blaZ sequences (accession numbers X52734 [GenBank] , AF270127 [GenBank] , NC_002382, X16471 [GenBank] , AB074882 [GenBank] , AJ426835 [GenBank] , AJ400722 [GenBank] and AJ302698 [GenBank] ). In a study by Yazdankhah et al.12 based on 10 bovine staphylococci they found that the transposition of the blaZ cluster was mainly associated with the IS elements IS1181 and IS257 rather than the Tn552. Studies on human isolates have shown that IS elements are present in most staphylococci and are associated with the spread of antimicrobial resistance within and between staphylococcal species.17 It has been know for a long time that a conjugative transfer gene (tra) is present in low frequency in several staphylococcal species.44 Other conjugative systems have been described since,45 but the role of these transmission systems in animal reservoirs of staphylococci and between human and animal reservoirs of staphylococci has yet to be elucidated.

A special aim of the study was to investigate the possible sharing of a gene pool of blaZ in a more closed environment, in this case dairy cattle herds in Denmark. Only in one case was the same allele type of blaZ demonstrated in S. aureus and CoNS in the same herd. The sample size within each herd was limited, and the observation does not allow us to rule out that other isolates in the herds may have been of other types, especially since the allele type from isolates within herds was found to vary. However, seen over the five herds investigated, the total number of isolates compared, in two instances even from the same teat on the same animal, and the very few cases, where S. aureus and CoNS were found to share allele type in the full investigation, it seems safe to conclude that S. aureus and CoNS generally do not exchange blaZ, even when present in the same herd. The low prevalence of bovine staphylococci with blaZ on conjugative plasmids and the fact that most isolates obtained from the Danish herds were of signature type 6, suggest a clonal spread of staphylococci in the bovine population. This is further supported by the fact that we failed to transmit blaZ in conjugation trials (data not shown), supporting the assumption that horizontal spread of blaZ is a rare event.

In conclusion, this study has shown low similarity between chromosomal- and plasmid-encoded blaZ indicating that the blaZ gene is vertically (clonal) spread in Staphylococcus and that exchange of blaZ between strains is an extremely rare event. Based on the grouping of the blaZ sequences in a plasmid-like group and a chromosomal group it seems likely that the gene has followed two different evolutionary paths. Bovine staphylococci primarily were shown to carry blaZ on the chromosome, but where blaZ was located on plasmids, it showed high homology to previously published sequences of human origin. Likewise, chromosomally-located genes in S. aureus were similar to such genes in bovine isolates, indicating that human and bovine strains share the same gene pool. Also, the genes were shown to be of the same signature type across species boundaries. Within the same epidemiological unit, however, we did not find indications that S. aureus and CoNS isolates normally share the blaZ gene pool.


   Transparency declarations
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Transparency declarations
Supplementary data
 References
 
This study was partly supported by the Danish Dairy Board in the interest of investigating the role of CoNS as a reservoir of blaZ in bovine staphylococci. None of the investigators have received money directly from the Dairy Board as all funds were channelled through University Administration.


   Supplementary data
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Transparency declarations
 Supplementary data
References
 
Supplementary data are available at JAC online (http://jac.oxfordjournals.org).


   Acknowledgements
 
This work formed a part of fulfilment of the PhD degree by Jan Vintov. Sadly Jan Vintov passed away during his study period. The authors wish to acknowledge his work, which would otherwise have resulted in first authorship. The paper is dedicated to his memory. The present study was funded by the Danish Veterinary Institute, the Royal Veterinary and Agricultural University, and the Danish Dairy Board. We are especially grateful to Dorte Stegmenn Nielsen and Berith Kummerfeldt from the technical staff in the section for antimicrobial resistance and environmental microbiology at the Danish Food and Veterinary Institute for technical assistance.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Transparency declarations
 Supplementary data
 References
 
1. Kirby WMM. Extraction of a highly potent penicillin inactivator from penicillin resistant staphylococci. Science 1944; 99: 452–3.[Abstract/Free Full Text]

2. Zinn CE. Annual Report on Staphylococcus aureus 1999. Denmark: Staphylococcus Laboratory, Statens Serum Institut, 2000.

3. Aarestrup FM, Jensen NE. Development of penicillin resistance among Staphylococcus aureus isolated from bovine mastitis in Denmark and other countries. Microbiol Drug Resist 1998; 4: 247–56.

4. Vintov J, Aarestrup FM, Zinn CE et al. Association between phage types and antimicrobial resistance among bovine Staphylococcus aureus from 10 countries. Vet Microbiol 2003; 95: 133–47.[CrossRef][ISI][Medline]

5. Hartman BJ, Tomasz A. Low-affinity penicillin-binding protein associated with ß-lactam resistance in Staphylococcus aureus. J Bacteriol 1984; 158: 513–6.[Abstract/Free Full Text]

6. Firth N, Skurray, RA. Genetics: Accessory elements and genetic exchange. In: Fischetti VA, Novick RP, Ferretti JJ et al., eds. Gram-Positive Pathogens. Washington, DC: ASM Press, 2000; 326–38.

7. Kernodle DS, Stratton CW, McMurray LW et al. Differentiation of ß-lactamase variants of Staphylococcus aureus by substrate hydrolysis profiles. J Infect Dis 1989; 159: 103–8.[ISI][Medline]

8. Dyke KGH. Resistance to ß-lactam antibiotics: resistance mediated by ß-lactamases. In: Crossley KB, Archer GL, eds. The Staphylococci in Human Diseases. New York: Churchill Livingstone Inc., 1997; 139–57.

9. Voladri RK, Kernodle DS. Characterization of a chromosomal gene encoding type B ß-lactamase in phage group II isolates of Staphylococcus aureus. Antimicrob Agents Chemother 1998; 42: 3163–8.[Abstract/Free Full Text]

10. Devriese LA, Baele M, Vaneechoutte M et al. Identification and antimicrobial susceptibility of Staphylococcus chromogenes isolates from intramammary infections of dairy cows. Vet Microbiol 2002; 87: 175–82.[CrossRef][ISI][Medline]

11. Vesterholm-Nielsen M, Ølholm Larsen M, Olsen JE et al. Occurrence of the blaZ gene in penicillin resistant Staphylococcus aureus isolated from bovine mastitis in Denmark. Acta Vet Scand 1999; 40: 279–86.[ISI][Medline]

12. Yazdankhah SP, Sørum H, Oppegaard H et al. Comparison of genes involved in penicillin resistance in staphylococci of bovine origin. Microbiol Drug Resist 2000; 6: 29–36.

13. Aarestrup FM, Jensen LB. Trends in antimicrobial susceptibility in relation to antimicrobial usage and presence of resistance genes in Staphylococcus hyicus isolated from exudative epidermitis in pigs. Vet Microbiol 2002; 89: 83–94.[CrossRef][ISI][Medline]

14. Jaffe W, Sweeney HM, Nathan C et al. Identity and interspecific transfer of gentamicin-resistance plasmids in Staphylococcus aureus and Staphylococcus epidermidis. J Infect Dis 1980; 141: 738–47.[ISI][Medline]

15. Forbes BA, Schaberg DR. Transfer of resistance plasmids from Staphylococcus epidermidis to Staphylococcus aureus: evidence for conjugative exchange of resistance. J Bacteriol 1983; 153: 627–34.[Abstract/Free Full Text]

16. Naidoo J. Interspecific co-transfer of antibiotic resistance plasmids in staphylococci in vivo. J Hyg 1984; 93: 59–66.

17. Archer GL, Thanassi JA, Niemeyer DMK et al. Characterization of IS1272, an insertion sequence-like element from Staphylococcus haemolyticus. Antimicrob Agents Chemother 1996; 40: 924–9.[Abstract]

18. Udo EE, Jacob LE, Mokadas EM. Conjugative transfer of high-level mupirocin resistance from Staphylococcus haemolyticus to other staphylococci. Antimicrob Agents Chemother 1997; 41: 693–5.[Abstract]

19. Anonymous. DANMAP 2001—Use of antimicrobial agents and occurrence of antimicrobial resistance in bacteria from food, animals and humans in Denmark. Danish Veterinary Institute, 2002.

20. De Oliveira AP, Watts JL, Salmon SA et al. Antimicrobial susceptibility of Staphylococcus aureus isolated from bovine mastitis in Europe and the United States. J Dairy Sci 1999; 83: 855–62.

21. Eriksen NH, Espersen F, Rosdahl VT et al. Carriage of Staphylococcus aureus among 104 healthy persons during a 19-month period. Epidemiol Infect 1995; 115: 51–60[Medline]

22. Aarestrup FM, Wegener HC, Rosdahl VT. Evaluation of phenotypic and genotypic methods for epidemiological typing of Staphylococcus aureus from bovine mastitis in Denmark. Vet Microbiol 1995; 45: 139–50.[CrossRef][ISI][Medline]

23. Aarestrup FM, Agersø Y, Ahrens P et al. Antimicrobial susceptibility and presence of resistance genes in staphylococci from poultry. Vet Microbiol 2000; 74: 353–64.[CrossRef][ISI][Medline]

24. Jensen LB, Ahrens P, Dons L et al. Molecular analysis of Tn1546 in Enterococcus faecium isolated from animals and humans. J Clin Microbiol 1998; 36: 437–42.[Abstract/Free Full Text]

25. Sears LE, Moran LS, Kissinger C et al. Thermal cycle sequencing and alternative sequencing protocols using the highly thermostable VentR (exo) polymerase. Biotech 1992; 13: 626–83.

26. Benson DA, Karsch-Mizrachi I, Lipman DJ et al. GenBank: UPDATE. Nucleic Acids Res 2004; 32: D23–6.[Abstract/Free Full Text]

27. Altschul SF, Gish W, Miller W et al. Basic local alignment search tool. J Mol Biol 1990; 215: 403–10.[CrossRef][ISI][Medline]

28. Thompson JD, Gibson TJ, Plewniak F et al. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 1997; 25: 4876–82.[Abstract/Free Full Text]

29. Olsen GJ, Matsuda H, Hagstrom R et al. FastDNAml: a tool for construction of phylogenetic trees of DNA sequences using maximum likelihood. Comput Appl Biosci 1994; 10: 41–8.[Abstract/Free Full Text]

30. Felsenstein J. PHYLIP (phylogeny inference package) version 3.5c. Department of Genetics, University of Seattle, 1995.

31. Grassly NC, Holmes EC. A likelihood method for the detection of selection and recombination using nucleotide sequences. Mol Biol Evol 1997; 14: 239–47.[Abstract]

32. Skov RL, Williams TJ, Pallesen L et al. ß-Lactamase production and genetic location in Staphylococcus aureus: introduction of a ß-lactamase plasmid in strains of phage group II. J Hosp Infect 1995; 30: 111–24.[CrossRef][ISI][Medline]

33. Rowland S-J, Dyke KGH. Tn552, a novel transposable element from Staphylococcus aureus. Mol Microbiol 1990; 4: 961–75.[ISI][Medline]

34. Vintov J, Aarestrup FM, Zinn CE et al. Phage types and antimicrobial resistance among Danish bovine Staphylococcus aureus isolates since the 1950s. Vet Microbiol 2003; 97: 63–72.[CrossRef][ISI][Medline]

35. Picinini R, Zecconi A. Relationship among plasmids recovered from Staphylococcus aureus, milk leukocytes and antimicrobial resistance. J Dairy Sci 2001; 84: 2641–8.[Abstract]

36. Matthews KR, Jayarao, BM, Oliver SP. Plasmid pattern analysis of Staphylococcus species isolated from bovine mammary secretions. J Dairy Sci 1992; 75: 3318–23.[Abstract]

37. Hasman H, Aarestrup FM. tcrB, a gene conferring transferable copper resistance in Enterococcus faecium: occurrence, transferability, and linkage to macrolide and glycopeptide resistance. Antimicrob Agents Chemother 2002; 46: 1410–6.[Abstract/Free Full Text]

38. Baba T, Takeuchi F, Kuroda M et al. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 2002; 25: 1819–27.

39. Doebbeling BN, Pfaller MA, Hollis RJ et al. Restriction endonuclease analysis of Staphylococcus aureus plasmid DNA from three continents. Eur J Clin Microbiol Infect Dis 1992; 11: 4–8.[CrossRef][ISI][Medline]

40. Tomaylo JF, Zscheck KK, Singh KV et al. Comparison of the ß-lactamase gene cluster in clonally distinct strains of Enterococcus faecalis. Antimirob Agents Chemother 1996; 40: 1170–4.

41. Zsceck KK, Myrray BE. Genes involved in the regulation of ß-lactamase production in enterococci and staphylococci. Antimicrob Agents Chemother 1993; 37: 1966–70.[Abstract/Free Full Text]

42. Rosdahl VT. Localisation of the penicillinase gene in naturally occurring Staphylococcus aureus strains. Acta Pathol Microbiol Immunol Scand 1985; 93: 383–8.

43. East AK, Dyke KGH. Cloning and sequence determination of six Staphylococcus aureus ß-lactamases and their expression in Escherichia coli and Staphylococcus aureus. J Gen Microbiol 1989; 135: 1001–15.[Medline]

44. Archer GL, Scott J. Conjugative transfer genes in staphylococcal isolates from the United States. Antimicrob Agents Chemother 1991; 35: 2500–4.[Abstract/Free Full Text]

45. Thomas WDJ, Archer GL. Mobilization of recombinant plasmids from Staphylococcus aureus into coagulase negative Staphylococcus species. Plasmid 1992; 27: 164–8.[CrossRef][ISI][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 J Antimicrob ChemotherHome page
E.-J. Yoon, A.-R. Kwon, Y.-H. Min, and E.-C. Choi
Foggy D-shaped zone of inhibition in Staphylococcus aureus owing to a dual character of both inducible and constitutive resistance to macrolide-lincosamide-streptogramin B
J. Antimicrob. Chemother., March 1, 2008; 61(3): 533 - 540.
[Abstract] [Full Text] [PDF]

Home page
 Antimicrob. Agents Chemother.Home page
A. M. Queenan, W. Shang, M. Kania, M. G. P. Page, and K. Bush
Interactions of Ceftobiprole with {beta}-Lactamases from Molecular Classes A to D
Antimicrob. Agents Chemother., September 1, 2007; 51(9): 3089 - 3095.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
57/3/450    most recent
dki492v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted