Methicillin-resistant Staphylococcus aureus typing methods: which should be the international standard?
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Journal of Hospital Infection (2000) 44: 160–172 doi: 10.1053/jhin.1999.0701, available online at http://www.idealibrary.com on REVIEW Methicillin-resistant Staphylococcus aureus typing methods: which should be the international standard? T. M.A.Weller Department of Medical Microbiology, City Hospital NHS Trust, Dudley Road, Birmingham B18 7QH Summary: Methicillin-resistant Staphylococcus aureus (MRSA) has spread to all parts of the world. Effective control measures are dependent on a thorough knowledge of the organism’s epidemiology which requires a typing technique that can be universally applied. Many typing methods have been developed for MRSA but none has been adopted as the internationally recognized standard. This review summarizes the information available on each in order to assess their suitability as a reference procedure. The majority of phenotypic and genotypic techniques are not sufficiently discriminatory, reproducible, stable or useful in an outbreak to be acceptable. The methods which do fulfil these requirements and have a potential for standard- ization, such as pulsed-field gel electrophoresis, binary typing or a combination of more rapid techniques, require further systematic evaluation. © 2000 The Hospital Infection Society Keywords: Methicillin-resistant Staphylococcus aureus; MRSA; infection control. Introduction developed, however, it is important fully to under- stand the epidemiology of the organism. Without a Methicillin-resistant Staphylococcus aureus (MRSA) thorough knowledge of the factors which affect the has been an infection control problem ever since it acquisition of MRSA and the transmissibility of was first discovered in 1961.1 Like its susceptible different strains, it will not be clear which measures counterpart it is a common cause of skin, soft tissue should be put in place to stop it. A prerequisite for a and wound infection as well as deeper sepsis such as successful epidemiological investigation is a reliable osteomyelitis and endocarditis. Although MRSA is indicator of the relationship between the organisms thought to be no more virulent than methicillin-sus- isolated, in other words, a typing scheme. ceptible S. aureus (MSSA), infections are more diffi- At present there is no consensus regarding the cult and expensive to treat. This is partly because, as best method to use for typing MRSA. In order to be well as being resistant to all β-lactam antibiotics, effective it should be highly discriminatory, repro- resistance rates for other antibiotics are also high. ducible, standardized, based on a stable feature, Most concerning has been the detection of MRSA widely available, inexpensive and have performed with reduced susceptibility to the glycopeptides, satisfactorily in an epidemiological investigation.6 vancomycin and teicoplanin, which are widely The aim of this review is to critically assess the regarded as the definitive therapy.2,3 MRSA is now a information available on techniques that have been world-wide phenomenon and the incidence contin- used for MRSA strain differentiation in order to ues to rise.4,5 In view of this and the consequent judge which merit further consideration. implications for morbidity and cost, many hospitals have attempted to control the spread of MRSA. Before the most effective methods of control can be Phenotyping Antibiogram A set of MRSA isolates can always be compared E-mail: T.M.A. Weller@bham.ac.uk on the basis of their susceptibility to a range of 0195–6701/00/030160 + 13 $35.00 © 2000 The Hospital Infection Society
MRSA typing methods 161 antibiotics. This technique is easy to perform, gives species. Those with a narrow host range could be rapid results and, above all, is cheap and readily used in groups to produce a pattern of lysis charac- available in the routine microbiology laboratory. teristic for individual strains. The major disadvantages are poor discriminatory Uniquely amongst staphylococcal typing systems ability and lack of reproducibility. Antibiotic resis- phage typing was standardized by the International tance patterns are, to some extent, influenced by the Subcommittee on Phage Typing of Staphylococci.11 local environment. Thus, the same antibiogram may An approved set of phages was recommended be produced by unrelated strains as a consequence which, with minor modifications, has been used of the similar selective pressure upon them. It is world-wide ever since. Currently 23 standard equally possible that the antibiograms of two iso- phages are applied to an agar plate covered with the lates from the same clone may differ due to the test organism. There is also room for two locally acquisition or loss of plasmids carrying resistance selected phages, which can be useful if indigenous genes. isolates are unreactive to the International Set.12 It has been found in practice that a simple catego- Areas of lysis (plaques) caused by each phage are rization of antibiotics into susceptible and resistant noted as either a strong or a weak reaction. The for- often cannot discriminate between unrelated strains. mer, defined as more than 50 plaques, is used to dis- Antibiograms based on relative zone sizes, compared tinguish between isolates. A lower number of by direct measurement or with more complicated plaques is a weak reaction which is recorded but computer aided formulae, have also been evalu- should not influence the final phage type.11 If a ated.7,8 In one epidemiological investigation the staphylococcus shows no lysis at the Routine Test quantitative antibiogram was as useful as ribotyp- Dilution (RTD), then a concentration 100 times ing.8 However, it is not clear how much the results greater (RTD X 100) is employed.11 Unfortunately may be affected by changes in culture conditions or standardization of the phages used has not led to transfer of plasmids as no assessment of repro- uniformity of practice and interpretation. A multi- ducibility was made. centre study showed that a variety of different media Strain differentiation has also been increased by was used and the method was often modified to fit in extending the range of antibiotics and other types of with local conditions.12 Many laboratories routinely antimicrobial substances tested for each isolate.9 tested at RTD X 100 in parallel with phages at RTD Inevitably some agents will have no discriminatory and some did not use the latter concentration at all. value but this can only be assessed once all isolates Furthermore, the relative influence of weak and have been examined. The time taken to perform the strong reactions on the results was inconsistent. extra tests, in some cases for no reward, has limited For many years, phage typing was the method of the appeal of this approach. Some workers may also choice for the investigation of MRSA epidemiology. feel that the inclusion of toxic chemicals, such as Several outbreaks have been defined with this tech- mercury and ethidium bromide, into the scheme nique and its discriminatory power is demonstrably makes it inappropriate for use in a routine labora- greater than phenotypic tests such as capsular typing tory. and zymotyping.13–16 Despite its popularity, the In most circumstances, the antibiogram cannot problems inherent in phage typing have never been be used as the sole typing method for MRSA. adequately resolved. It is a time consuming and Sometimes, however, the susceptibility pattern of an technically demanding procedure which is most effi- endemic or epidemic clone may be so distinctive that ciently done on large batches. This and the necessity an antibiogram is a useful screening method. In one for keeping stocks of phages and the propagating outbreak, resistance to rifampicin was used to dis- strains, has confined it to larger laboratories and ref- tinguish a new strain from those already present in erence facilities. the hospital thereby excluding staphylococci which When used in an outbreak situation, the main dis- needed no further typing.10 advantage of phage typing is the high proportion of isolates which are non-typable. Typically this reaches 20–30% for collections of MRSA7,16–18 but in Phage typing some cases has been as high as 75%.19 The value of The concept of typing staphylococci with bacterio- any information obtained is markedly reduced as phages was first developed in the 1940s. It was there is no way of knowing whether the non-typable noticed that some S. aureus contained temperate bacteria are related. Indeed, isolates non-typable by phages which lysed other bacteria of the same phages but linked initially by epidemiological
162 T. M.A.Weller studies have subsequently been shown to be quite Immunoblotting distinct by other methods.17 Efforts have been made Protein analysis can be refined by reacting the elec- to reduce the number of non-reacting bacteria by trophoresed antigens with labelled antistaphylococ- using RTD X 100, incubation at 48°C prior to the cal antibodies and examination by Western blot. test18 and the development of new phages.20 The source of the antibody can either be rabbit sera, The value of phage typing is further diminished following deliberate staphylococcal exposure,26 or by its lack of reproducibility. The same isolate can pooled human sera, on the assumption that enough give a variety of different results when tested on donors have had contact with S. aureus for sufficient separate occasions,7,18 a problem which is amplified antibody to be present.27 Immunoblotting is less when screening bacteria for relatedness over a long susceptible to variation in running conditions than period of time. In view of this, it is recommended whole cell protein electrophoresis and results in that isolates for comparison should be tested simul- fewer bands, thereby aiding interpretation.24 All iso- taneously and should not be regarded as distinguish- lates are typable but in an outbreak it may not be able when there is only a single difference in their discriminatory enough to correctly exclude unre- lysis pattern.7,21 lated organisms.7,24 Multilocus enzyme electrophoresis Serotyping Typing by multilocus enzyme electrophoresis Serotyping has never been used extensively for S. (MLEE) involves extraction of enzymes from the aureus. Tests have been developed, particularly to bacterial cell, their separation by electrophoresis and detect differences in the capsular polysaccharide and examination by selective staining. The rate of the antigenic properties of coagulase. There are a enzyme migration depends on its amino acid com- total of 11 S. aureus capsular types but 85–90% of position and over 80% of single substitutions can be clinical isolates belong to just two of them. detected by a change in its electrophoretic proper- Serological surveys have shown that it is rare to find ties.28 Enzyme variation is, therefore, used as a sur- an MRSA which is not capsular type 5 or 8, with the rogate marker for differences in the genetic loci from former being more common.16 Eight coagulase which they originate. Bacteria can be assigned to an serotypes have been described and used in Japan, electrophoretic type (ET) on the basis of similarities again with a predominance of two types.22 With such between the enzymes and the degree of relatedness poor discriminatory ability this method has no role between two isolates can be assessed by the propor- in epidemiological studies. tion of loci which show differences.29 MLEE has been applied to many bacterial species including S. aureus.28 When it is used to examine Protein electrophoresis MRSA all isolates are typable and reproducibility is Whole cell protein good.7 Discriminatory ability depends to a certain Analysis of cellular proteins, produced by extent on which enzymes are included as some have lysostaphin degradation, can be performed using been found to be monomorphic even within large polyacrylamide gel electrophoresis (PAGE). The collections of MRSA.29 The total number of result is a reproducible pattern consisting of enzymes used has varied from 12 to 20 but the rela- approximately 50 bands.23–25 Despite the large num- tive discriminatory power of each one has not yet ber of fragments, differences between unrelated iso- been assessed.7,29,30 Thus, it is not possible to say lates of MRSA are small. This leads to poor what the optimum combination will be or how this discrimination, only slightly enhanced by computer would perform in comparison with other tech- analysis of band intensity as well as size.25 niques. Interpretation of protein profiles is further hindered When MLEE has been used to examine outbreak by the lack of any correlation to typing results isolates most epidemiologically linked bacteria have obtained with phages.24,25 Gel-to-gel variation, been correctly classified but there have been some caused by minor changes in the running conditions, unrelated isolates mistakenly included.7 In other also means that isolates need to be run in parallel in work studying the diversity of a given MRSA popu- order to compare them accurately.24 These problems lation the majority of isolates clustered into just one with procedure and interpretation ensure that whole or two ETs.29,30 A total of nine plasmid profiles and cell protein electrophoresis is rarely used. 10 antibiograms were found amongst the 26 isolates
MRSA typing methods 163 in one ET, casting doubt on the validity of the simple to interpret,7 but it also has some features grouping.30 that are less than ideal. Although the patterns produced by MLEE are Although more MRSA than MSSA possess plas- relatively easy to read and interpret, comparison is mids, they are still not present in every isolate leav- difficult and is best done with the aid of a computer ing many organisms non-typable.7,35,36 In addition, programme.7 This, coupled with the fact that the the plasmid DNA may exist in more than one form technique is very labour intensive, has confined (supercoiled, nicked or linear) all of which have dif- MLEE to the research laboratory.21 ferent electrophoretic properties. Thus, bands seemingly of different sizes may, in fact, represent Zymotyping the same plasmid.34 Furthermore, the presence of Zymotyping is a variation of MLEE based solely on only one or two plasmids in many S. aureus leads to differences in the electrophoretic properties of the poor strain differentiation.35 bacterial esterase enzymes.31 S. aureus possesses three Restriction endonuclease analysis of plasmid esterases, designated A, B and C in order of decreas- DNA (REAP) can improve the discriminatory abil- ing affinity for the anode. Each has been defined by ity of this method by demonstrating differences in its relative activity on five synthetic substrates (α- the position and frequency of restriction sites and β-naphthyl acetates, indoxyl acetate and α- and between two unrelated plasmids of the same size.36 β-naphthyl butyrates) and its resistance to di-iso- In some studies REAP has been more discrimina- propyl fluorophosphate.32 As for MLEE, the tory than PFGE,36,37 although in others it has not enzymes are electrophoresed on a polyacrylamide been as effective.7 agarose gel and typing is based on the observed dif- The major problem with plasmid analysis is lack ferences in their mobility in separate isolates. of stability and reproducibility. The inherent mobil- Most experience with this method has come ity of extra-chromosomal DNA means that plas- from a single research group which has used it for S. mids are easily gained or lost in vivo36 as well as in aureus susceptible and resistant to methicillin.16,32,33 vitro.38 Consequently there are frequent changes in The group have detected 26 zymotypes, of which the plasmid profile of a strain resulting in poor sen- 14 have included MRSA.32,33 All isolates are typable sitivity of the test in outbreak investigation.7 There and the pattern is both stable and reproducible. may also be transfer of DNA between unrelated iso- Although discriminatory ability is better than cap- lates leading to the false impression of a clonal out- sular typing it is significantly worse than with break if only plasmid typing is used.34 This variation phages or pulsed-field gel electrophoresis over time may, however, be useful in some instances, (PFGE).16 as it allows the sub-division of related outbreak iso- It is difficult to know whether similarities lates defined by another method.36,38 amongst the esterase enzymes are truly a reflection of a common heritage. Several genotypes have been Chromosomal DNA detected within one zymotype but the reverse is not true. In addition there is no correlation between the Restriction enzyme analysis (REA) results with this technique and the antibiotic sus- Chromosomal DNA is too big to analyse as a whole ceptibility pattern. In short, zymotyping is not dis- and must, therefore, be cut into smaller pieces. This criminatory enough to be a useful typing method can be done with restriction enzymes which recognize for MRSA on its own and is too lengthy a proce- and cleave specific sequences. Enzymes, such as BglII dure to be employed as a preliminary screening and EcoR1, which bind to a site found frequently on test. the staphylococcal chromosome, produce multiple small fragments.39 These are then separated by con- stant voltage electrophoresis and the patterns pro- Genotyping duced can be compared with those of other isolates. All MRSAs are typable by this method and it has Plasmid analysis been used successfully to distinguish epidemic iso- The first molecular technique used for epidemio- lates from sporadic cases.26,39,40 However, as conven- logical investigation of MRSA was plasmid analy- tional electrophoresis will not separate fragments sis.34 This differentiates isolates according to the larger than around 20 kb the restriction pattern con- number and size of plasmids, measured by elec- sists of a large number of overlapping bands which trophoresis. The method is easy to perform and makes consistent analysis difficult.21
164 T. M.A.Weller Southern hybridization distinguish them from unrelated isolates44 it was found to be the least discriminatory of all the tech- A combination of REA and Southern hybridization niques compared by Tenover et al.7 In a direct com- can decrease the number of bands produced to a parison with PFGE, only six ribotypes were more manageable level. In this technique probes are detected in a collection of MRSA comprising 26 designed to target specific sequences likely to be pulsotypes.42 Furthermore, criteria for the interpre- found in multiple copies on the chromosome in a tation of the significance of minor band differences variety of positions. They are labelled, usually has not been standardised. As it is also time consum- radioactively, and allowed to bind to these sequences ing and technically complicated, ribotyping has not within the cleaved DNA. A relatively simple pattern been widely adopted for typing of MRSA. is then produced following DNA electrophoresis and detection of the probes. Differences in the num- ber and site of the bands can then be used to dis- Insertion sequences criminate between strains. Insertion sequences (IS) are pieces of DNA that are able to insert themselves into other DNA. The posi- Ribotyping tion and copy number of each within the staphylo- The most commonly used probe is ribosomal RNA. coccal chromosome varies from strain and strain and This utilizes the fact that within the total DNA this diversity can be used as the basis for typing. there are multiple copies of rRNA transcriptional Several, such as IS431, IS256 and IS1181, have sequences which are relatively conserved between been used to investigate MRSA by Southern different species. Labelled rRNA from the same or a hybridisation following REA.7,48–51 There will always different species will hybridise with these genes be some strains which are non-typable as not all wherever they occur. Restriction site heterogeneity MRSA have acquired every IS.7,51 This is more of a should ensure that the hybridization areas of unre- problem when dealing with susceptible staphylo- lated isolates are in different positions on the elec- cocci as IS tend to be associated with resistance trophoresed DNA. The number of bands detected genes.7 Interpretation can also be complicated by IS may also change if a restriction site is present within carried on plasmids creating profile variation which the rRNA transcriptional sequence itself.41 is not related to strain difference.7,48 In epidemiolog- Ribotyping has been applied to MRSA on a num- ical investigations IS typing has been found to ber of occasions with some variation in the restric- give both stable and reproducible results.50,51 tion enzyme and type of probe employed.7,42–44 Of Discriminatory ability has depended upon the the restriction enzymes that have been studied, nature of the isolates being tested and the occur- EcoRI consistently produces a greater number of rence of the sequence being detected. In some stud- bands and fingerprints than ClaI or HindIII and ies it has been better than ribotyping and phage many others.42,43,45 There does not appear to be any typing but typing by IS1181 was poorly discrimina- benefit in using more than one endonuclease as the tory for the British EMRSAs.48,50–52 Demonstration overall discriminatory power is the same as that for of strain diversity has been consistently better with EcoRI.42,46 PFGE but a combination of the two techniques may Probes originating from Escherichia coli,42,44 be useful. As insertion sequences are acquired Bacillus subtilis47 and S. aureus43 have all been used. cumulatively during the clonal spread of a strain, When directly compared, homologous probes have the IS type is independent of the pulsotype and can been shown to hybridize more rapidly than those be used to sub-divide isolates within an out- from another species but the difference is negligible break.48,50 compared to the effect of inter-run variation.46 The probe is usually labelled with the radioisotope 32P mecA:Tn554 probe typing although one group has used a biotinylated probe A number of other probes has been examined to which they believe is safer, more stable and gives determine usefulness in typing of MRSA. Those better band definition.43 targeting the genes coding for aminoglycoside resis- All MRSAs are typable by this method and it is tance and toxic shock syndrome toxin were found to more discriminatory and reproducible than pheno- have only limited application as not all MRSAs typic methods.45 In practice, however, ribotyping exhibit these characteristics.53,54 In contrast, mecA is, of MRSA has not been very useful. Although it by definition, present in all MRSA and the trans- has been used to identify outbreak strains and poson Tn554 has been detected in over 90% of
MRSA typing methods 165 isolates.54,55 A typing system has been devised using Binary typing these targets following restriction digestion with the Binary typing is a relatively new technique based on endonuclease ClaI. Southern hybridization, so far employed by just one Only one copy of mecA occurs on the chromo- group of workers.61,62 In this method many probes some and its position is constant.56 Simple detection were made by random amplification of the staphylo- of the gene would not, therefore, be of any use in coccal chromosome followed by cloning. Each was discriminating between unrelated strains. tested for strain specificity and those binding to the Spontaneous mutation and genetic rearrangement DNA of some, but not all, of a diverse set of MRSA does, however, change the number and position of were selected for validation. A binary code was ClaI restriction sites and consequently the size of assigned to the test isolates on the basis of a simple the DNA fragment on which mecA is found. presence or absence of hybridization for each probe, Furthermore ClaI has one cutting site within the thus possibly permitting inter-laboratory compari- gene which means that two fragments are detected son of results. with the mecA probe. Original studies on a large col- All MRSAs were typable, the results were stable lection of MRSA characterised six ClaI:mecA pat- and reproducible, but the initial group of seven terns and assigned roman numerals to them.55 probes proved to be poorly discriminatory.61,62 Further work by other groups has revealed a further However, when the probe number was increased to six polymorphs.57,58 With this small number of read- 15, strain discrimination better than PFGE was ily distinguishable types, mecA hybridization may be achieved.62 Isolates from known outbreaks were also a suitable technique to compare results from multi- clustered correctly. The potential disadvantage is the ple centres. However, in order to assign isolates to amount of time needed to complete the whole the correct ‘type’ strains with all the known patterns process. As the authors speculate, however, it may must be run in parallel. The logistics of providing be possible to adapt it to an enzyme-linked reference strains to all those performing MRSA typ- immunosorbent assay-like technique or an auto- ing may prove difficult. Tn554 carries the gene for mated system in order to make binary typing avail- resistance to spectinomycin. It is present in most able in the routine lab.62 MRSA, often in multiple copies.55 There is a ClaI restriction site within the transposon ensuring that Pulsed-field gel electrophoresis at least two bands are produced following hybridisa- tion. A larger number of bands are detected in the Interpretation of restriction endonuclease patterns presence of more than one copy or genetic can be aided by the use of enzymes which produce rearrangement within the transposon. At least thirty fewer, larger fragments. These ‘rare cutters’ have different patterns have been produced from MRSA, longer recognition sequences which naturally occur each of which is stable over approximately 1000 less often within the chromosome. Thus, instead of generations.55 hundreds of fragments ranging from 0·5 to 50 kb in It has been found that there is no constant rela- length, a simpler pattern consisting of between 10 tionship between the position of Tn554 insertion and 30 fragments, 10–800 kb in length, is produced. and the mecA pattern confirming that these are Although small segments of DNA separate quite independent evolutionary events and increasing the easily by conventional electrophoresis due to the potential strain diversity.57 In isolation, neither of sieving effect of the agarose gel, it is not possible to these probes provides a technique with adequate split large fragments. This is because mobility of discriminatory ability and even in combination, DNA above about 50 kb is independent of molecu- changes in the dominant clone can be missed.59 The lar weight. Adaptation of PFGE, a technique devel- most successful use of mecA:Tn554 hybridization to oped to examine chromosomal DNA in yeasts has provide epidemiological data has been in association provided a method.63 with PFGE. The three tests together have a demon- In conventional electrophoresis, negatively strably better discriminatory ability than when they charged DNA migrates in a straight line from the are used in isolation.60 The major disadvantage of cathode to the anode. For PFGE, however, the employing such a typing scheme is the length of direction of the charge is only kept constant for time needed to produce definitive results, estimated short ‘pulses’ of time. This continual shift in the at a month and a half for a single technician working direction of the field causes the DNA to regularly on 80 isolates.60 stop and re-orientate to a new course. Larger
166 T. M.A.Weller fragments take longer to achieve this switch and, made unless each isolate is run in exactly the same therefore, progress more slowly than smaller sec- conditions.70,71 In essence this means that running tions. Consequently PFGE allows comparison of two isolates next to each other on the same gel the entire chromosome without the complicated pat- remains the best way to assess whether they are terns produced by frequent cutting restriction related. It is not surprising, therefore, that inter- enzymes. The principle of PFGE has been applied centre comparison of results has not been success- in several different ways. Alternating the electric ful.70,71 Until an objective numerical system can be field to occur on just two bearings is known as field devised to describe PFGE patterns this technique inversion gel electrophoresis (FIGE). Progression is will be classed as a fingerprinting rather than a typ- achieved by allowing a longer pulse in the forward ing system. direction. The most commonly used variation, con- These shortcomings with regard to inter-centre tour clamped homogenous electrophoresis (CHEF), comparison do not, however, prevent PFGE from uses an electrophoresis chamber consisting of six being an extremely useful technique for the investi- electrodes arranged in a hexagonal pattern. The cur- gation of outbreaks. It has been used extensively to rent is applied in each of three directions, 120° delineate the epidemiology of both endemic and apart, in turn, for short periods of time. Overall epidemic staphylococci.14,72–75 Interpretation of pat- migration of DNA is in a straight line, but by a cir- terns produced by PFGE in these situations is aided cuitous route.64 by published guidelines.76 These are based on the PFGE has been used to investigate MRSA and number of changes in fragment size that can be compared to other techniques in numerous stud- expected following a genetic event. Thus, if more ies.7,16,17,42,65–67 Although a variety of restriction than two differences have occurred in the chromo- enzymes has been used, none has been found to be some (which corresponds to more than six distinct better than SmaI.42,68,69 All isolates are typable and bands) then the two isolates are deemed unrelated. the pattern is reproducible even after many sub-cul- This objective measurement of degree of related- tures.7,16,17,65 Discriminatory ability is high and has ness is virtually unique amongst typing methods for been shown superior to bacteriophage typing, any species. antibiograms, RAPD, ribotyping and zymotyp- ing.16,42,66,67 Results are also more reliable than for PCR typing standard REA as there is no interference from plas- mid DNA, the fragments being too small to affect In view of the length of time and technical ability the pattern.67 Thus PFGE has many of the features need to produce good typing results with PFGE or associated with the ideal typing method and it has Southern hybridization, a less demanding and more been proposed that it be regarded as the ‘gold stan- rapid method has been sought. The polymerase dard’ for MRSA.17 chain reaction (PCR) is known to have these quali- There are some features that stop PFGE from ties and has been used extensively for diagnostic being accepted as the universal technique of choice. purposes. The technique is easy to master and it is Detailed preparation procedures mean that results estimated that 50 isolates can be typed in 48 hours are not available for several days.7 There is also a once the assay has been standardized.77 Potential financial implication as the cost has been estimated disadvantages lie in its discriminatory ability and to be over three times as expensive as bacteriophage reproducibility. These features have been assessed in typing even before the initial cost of specialized a number of studies employing a variety of different equipment is taken into account.17 FIGE is quicker approaches, targeting either specific staphylococcal and cheaper than CHEF and may resolve more low sequences or random areas of the chromosome. molecular weight fragments but despite this it is the latter which is used more often.7 Coagulase gene typing Although the total number of bands produced is Production of coagulase is one of the defining char- relatively small interpretation of results can still be acteristics of S. aureus and is thought to be an problematic. Slight differences in running condi- important virulence factor. A typing technique tions can alter the distance travelled by each band, based on the heterogeneity within a specific segment complicating the comparison of isolates run on dif- of the coagulase gene has been developed.78 The 3′ ferent gels. Computer aided analysis and the use of a region consists of a series of 81 base-pair tandem control organism or molecular weight standard on repeats which vary in number from four to eight. In each gel can help comparison but mistakes are still this method, therefore, PCR is used to amplify the 3′
MRSA typing methods 167 region and to produce a primary product, usually a of a constant size but the length of the largest frag- single fragment, but on some occasions two.79 The ment varies according to the number (between 3 and size of the fragment is dependent upon the number 15) of 24 base pair repeats it contains.84 Initially it of repeats within the area amplified. Although this was considered possible that strains harbouring region is quite well conserved, point mutations do more than seven repeats were more likely to be epi- occur which affect the number of restriction enzyme demic MRSA, but later work does not support binding sites. The amplification product can, there- this.81 fore, be characterized further by digestion with a As a typing technique, PCR of the protein A restriction enzyme. AluI, which is the one most gene has two major problems.81 Firstly, it is poorly commonly used, produces up to four bands of vari- discriminatory as there are only a limited number able length depending on the number of restriction of types possible. Secondly, results do not corre- sites present. late with the clonal relationship defined by other Although the vast majority of MRSAs are methods. typable by coagulase gene PCR, no primary product was obtained from a small number of isolates when RAPD using the original primer set.22,78 The problem has Typing by random amplified polymorphic DNA not recurred in later studies employing different tar- (RAPD), also known as arbitrarily primed-PCR get sequences so it can now be considered to be uni- (AP-PCR), has been applied to many bacterial versally applicable to S. aureus.79–81 Results are stable species. Low stringency PCR of genomic DNA is even after many sub-cultures in vitro79–81 and coagu- performed with single short primers with arbitrary lase gene typing is sensitive enough to group sequences. It is assumed that there will always be epidemiologically linked isolates together.7,78,79 some areas of DNA sufficiently similar to the However, isolates, including MSSA, unrelated to primers for them to anneal in a random fashion. The the outbreaks under investigation have also shown nucleic acid between the primer sites is amplified the same coagulase type, making it an unsuitable and the resulting fragments are separated by technique to use when good epidemiological data are electrophoresis. They form a fingerprint based on not available.22 In this respect it is less discrimina- the number and size of the amplified regions. tory than PFGE7,79,81,82 although it does offer the RAPD will produce a genetic fingerprint with all advantages of greater speed and lower cost.82 In MRSA and the results have been reproducible addition, as all bands are multiples of the 81bp within the same laboratory.85 The advantage of this tandem repeats it has been possible to devise a method over others is the relative speed and simplic- numerical code for each pattern which may facilitate ity of the technique.86 Discriminatory power, how- comparison between different laboratories.81 ever, can be variable and depends heavily on the Attempts to increase the discriminatory value by number and sequence of the primers employed. using a different restriction enzyme or more than Poor differentiation is achieved when a single primer one in combination, have not met with success.80 is used but the application of three or more primers The use of coagulase gene typing to differentiate considerably increases the time needed to perform between epidemic and sporadic MRSA has also the test and still does not make RAPD as discrimi- been proposed. Distinct patterns were found for the natory as PFGE.66,77 On the other hand, RAPD has three most common British epidemic strains80 and clustered epidemiologically linked organisms and this method may, therefore, offer the option of a excluded unrelated isolates when used in an out- rapid screening test for a known strain. However, no break situation.85,86 particular distinguishing feature for EMRSA has Comparison of results obtained in different cen- been elucidated81 and the sensitivity of the test when tres performing a semi-standardised technique performed directly on clinical specimens is too low revealed that reproducibility of RAPD for MRSA for it to be used alone.83 is very poor.85 This may be due to inconsistent interpretation of the significance of weak bands66,77 Protein A gene typing or variation in the electrophoretic conditions which Another target for PCR based typing of S. aureus affect band separation and differentiation.85 Even has been the X region, a hypervariable area of the within a single laboratory, batch to batch variation spa gene coding for protein A.84 The primers used in the primers has been noted to affect the consis- form a product which is cut into three fragments tency of results.77 This lack of reproducibility led with the restriction enzyme RsaI. Two of these are van Belkum et al. to conclude that ‘inter-institute
168 T. M.A.Weller standardisation will be very hard to achieve’ and methods.20 A problem with stability still remains. that RAPD is not a suitable reference method.85 Some isolates, repeatedly subcultured, changed type By using specific primers for a known local endemic on subsequent testing.89 strain it may, however, have a role as a rapid screen- Another promising rep-PCR assay, tar916-shida, ing test. used transposon Tn916 as one target and the Shine- Dalgarno sequence with its common neighbouring rep-PCR nucleotides as the other.90 The tar916-shida patterns produced were more discriminatory than phage typ- Repetitive element sequence-based PCR (rep-PCR) ing and identified German epidemic MRSA as effi- is another method to sample the whole chromosome. ciently as PFGE. Before a full assessment of its Primers are included which hybridize to sequences usefulness can be made, further work is needed on known to be repeated throughout the chromosome MRSA from other sources. but with variable number and position. When the Other primer sets have not lead to successful typ- DNA between these binding sites is amplified, the ing. Amplification with the enterobacterial repeti- length polymorphism produces a fingerprint. tive intergenic consensus sequences (ERIC) does Reproducibility is much better than RAPD as tar- occur under low-stringency conditions but discrimi- geting specific sequences allows the use of high- natory ability is poor compared to PFGE unless stringency PCR.87 The disadvantage is a need to other primers are used simultaneously.77,91 Similarly, include an extra DNA purification step into the detection of staphylococcal dru sequences had a high method which increase the procedure time. It is not incidence of non-typable strains and a poor discrim- yet clear which of the numerous primers employed inatory ability.91 for this technique will prove to be the most useful for MRSA typing. RepMP3, a repetitive sequence from Conclusions Mycoplasma pneumoniae used as a single primer, has The plethora of techniques available for typing been compared directly with primers targeting MRSA serves to emphasize that none has yet been inter-IS256 sequences.87 Although all MRSA were recognized as the definitive method. Unfortunately typable with both primer sets the former had some many have been investigated without any objective distinct advantages. The RepMP3 patterns created measurement of their performance. Thus the indi- were reproducible, easy to compare and stable fol- vidual studies will show that one technique is better, lowing subculture. In contrast, amplification of or worse, than another when used in a particular cir- inter-IS256 required a low annealing temperature cumstance with a selected group of isolates. There which adversely affected reproducibility unless pat- are no recognized standards by which to assess the terns with one band difference were regarded as level of reproducibility or stability of a method. A indistinguishable.88 Furthermore, a larger number variety of different approaches has been used, mak- of strains were defined with RepMP3 than by inter- ing comparison difficult. There is a standard mea- IS256 PCR. Neither primer set proved to be as dis- sure of discriminatory ability, Simpson’s Index of criminatory as PFGE but both performed well Diversity.92,93 This requires the testing of a large when compared to other PCR-based tech- number of unrelated organisms and is consequently niques.20,87,88 rarely employed. The conclusions drawn from the Primers amplifying the spacer region between the investigations included in this review are, therefore, 16S rRNA gene and the 23S rRNA gene have been only tentative and further systematic study of the assessed.89 This sequence varies in length and copy most promising techniques is required. number, so that a single isolate may contain up to 15 Only one method, phage typing, has been stan- different alleles ranging from 906 to 1223 bp. The dardized for international use and this suffers from a discriminatory ability of 16S–23S spacer PCR was high proportion of non-typable isolates and poor called into question in one large study. Only nine reproducibility. Other phenotypic techniques are too ‘ribotypes’ were defined from a collection of 274 unreliable to be acceptable. Genotyping methods MRSA and 97% of isolates belonged to just two.89 have tended to be more discriminatory and repro- However, variation was much greater amongst the ducible but none of them fulfils all the required cri- methicillin- and penicillin-susceptible S. aureus teria. In particular the development of a technique tested and in other work 16S–23S spacer amplifica- capable of being standardized for inter-laboratory tion has distinguished more strains than other PCR comparison has been elusive.
MRSA typing methods 169 PFGE has been promoted as the ‘gold standard’ 10. Archer GL, Mayhall CG. Comparison of epidemio- as it has proved to be highly discriminatory and per- logical markers used in the investigation of an out- break of methicillin-resistant Staphylococcus aureus forms well in the context of a local outbreak.17 infections. J Clin Microbiol 1983; 18: 395–399. Although it is time consuming and attempts at inter- 11. Parker MT. Phage-typing of Staphylococcus aureus. laboratory standardisation have failed, it does In: Norris JR, Ribbons DW, eds. Methods in microbi- remain a useful reference point against which other ology. London: London Academic Press, 1972; 1–28. methods can be judged.70 The only other system 12. Marples RR, Rosdahl VT. International quality con- trol of phage typing of Staphylococcus aureus. J Med with sufficient discriminatory power to be used as a Microbiol 1997; 46: 511–516. single method is binary typing.62 It is too early in the 13. Kerr S, Kerr GE, Mackintosh CA, Marples RR. A development of this technique to gauge how useful survey of methicillin-resistant Staphylococcus aureus it will be in practice. Ultimately it may be advanta- affecting patients in England and Wales. J Hosp Infect geous to select a group of more rapid assays that can 1990; 16: 35–48. 14. Richardson JF, Reith S. Characterization of a strain of achieve similar strain discrimination when the methicillin-resistant Staphylococcus aureus (EMRSA- results are combined. It is in these areas that future 15) by conventional and molecular methods. J Hosp work to select an international MRSA typing Infect 1993; 25: 45–52. scheme should be concentrated. 15. Cox RA, Conquest C, Mallaghan C, Marples RR. A major outbreak of methicillin-resistant Staphylococcus aureus caused by a new phage-type Acknowledgement (EMRSA-16). J Hosp Infect 1995; 29: 87–106. 16. Schlichting C, Branger C, Fournier JM et al. Typing I thank Dr Adam Fraise for his useful comments on of Staphylococcus aureus by pulsed-field gel elec- the manuscript. trophoresis, zymotyping, capsular typing, and phage typing: resolution of clonal relationships. J Clin Microbiol 1993; 31: 227–232. References 17. Bannerman TL, Hancock GA, Tenover FC, Miller 1. Jevons MP. ‘Celbenin’-resistant Staphylococci. BMJ JM. Pulsed-field gel electrophoresis as a replacement 1961; 1: 124–125. for bacteriophage typing of Staphylococcus aureus. J 2. Hiramatsu K, Hanaki H, Ino T, Yabuta K, Oguri T, Clin Microbiol 1995; 33: 551–555. Tenover FC. Methicillin-resistant Staphylococcus 18. Lundholm M, Beregendahl B. Heat treatment to aureus clinical strain with reduced vancomycin sus- increase phage typability of Staphylococcus aureus. ceptibility. J Antimicrob Chemoth 1997; 40: 135–136. Eur J Clin Microbiol Infect Dis 1988; 7: 300–302. 3. Martin R, Wilcox KR, Campbell C, Ellis H. Update: 19. Khalifa KI, Heiba AA, Hancock G. Nontypable bac- Staphylococcus aureus with reduced susceptibility to teriophage patterns of methicillin-resistant vancomycin – United States, 1997. Morb Mortal Wkly Staphylococcus aureus involved in a hospital outbreak. Rep 1997; 46: 813–815. J Clin Microbiol 1989; 27: 2249–2251. 4. Wenzel RP, Nettleman MD, Jones RN, Pfaller MA. 20. Richardson JF, Chittasobhon N, Marples RR. Methicillin-resistant Staphylococcus aureus: implica- Supplementary phages for the investigation of strains tions for the 1990s and effective control measures. Am of methicillin-resistant Staphylococcus aureus. J Med J Med 1991; 91: 221S–227S. Microbiol 1988; 25: 67–74. 5. Witte W, Kresken M, Braulke C, Cuny C. Increasing 21. Mulligan ME, Arbeit RD. Epidemiologic and clinical incidence and widespread dissemination of methicillin- utility of typing systems for differentiating among resistant Staphylococcus aureus (MRSA) in hospitals in strains of methicillin-resistant Staphylococcus aureus. central Europe, with special reference to German hos- Infect Contr Hosp Epidemiol 1991; 12: 20–28. pitals. Clin Microbiol Infect 1997; 3: 414–422. 22. Kobayashi N, Taniguchi K, Kojima K et al. Analysis 6. Aber RC, Mackel DC. Epidemiologic typing of noso- of methicillin-resistant and methicillin-susceptible comial microorganisms. Am J Med 1981; 70: 899–905. Staphylococcus aureus by a molecular typing method 7. Tenover FC, Arbeit R, Archer G, et al. Comparison based on coagulase gene polymorphisms. Epidemiol of traditional and molecular methods of typing iso- Infect 1995; 115: 419–426. lates of Staphylococcus aureus. J Clin Microbiol 1994; 23. Thomson-Carter FM, Pennington TH. 32: 407–415. Characterisation of methicillin-resistant isolates of 8. Blanc DS, Petignat C, Moreillon P, Wenger A, Bille J, Staphylococcus aureus by analysis of whole-cell and Francioli P. Quantitative antibiogram as a typing exported proteins. J Med Microbiol 1989; 28: 25–32. method for the prospective epidemiological surveil- 24. Gaston MA, Duff PS, Naidoo J et al. Evaluation of lance and control of MRSA: Comparison with molec- electrophoretic methods for typing methicillin-resis- ular typing. Infect Contr Hosp Epidemiol 1996; 17: tant Staphylococcus aureus. J Med Microbiol 1988; 26: 654–659. 189–197. 9. Gillespie MT, Lyon BR, Skurray RA. Typing of 25. Costas M, Cookson BD, Talsania H, Owen R. methicillin-resistant Staphylococcus aureus by antibi- Numerical analysis of electrophoretic protein pat- otic resistance phenotypes. J Med Microbiol 1990; 31: terns of methicillin-resistant strains of Staphylococcus 57–64. aureus. J Clin Microbiol 1989; 27: 2574–2581.
170 T. M.A.Weller 26. Burnie JP, Matthews RC, Lee W, Murdoch D. A compar- 40. Cookson BD, Phillips I. Epidemic methicillin-resis- ison of immunoblot and DNA restriction patterns in tant Staphylococcus aureus. J Antimicrob Chemoth characterising methicillin-resistant isolates of 1988; 21: 57–65. Staphylococcus aureus. J Med Microbiol 1989; 29: 255–261. 41. Stull TL, LiPuma JJ, Edlind TD. A broad-spectrum 27. Mulligan ME, Kwok RYY, Citron DM, John JF, probe for molecular epidemiology of bacteria: Smith PB. Immunoblots, antimicrobial resistance, Ribosomal RNA. J Infect Dis 1988; 157: 280–286. and bacteriophage typing of oxacillin-resistant 42. Prevost G, Jaulhac B, Piemont Y. DNA fingerprinting Staphylococcus aureus. J Clin Microbiol 1988; 26: by pulsed-field gel electrophoresis is more effective 2395–2401. than ribotyping in distinguishing among methicillin- 28. Selander RK, Caugant DA, Ochman H, Musser JM, resistant Staphylococcus aureus isolates. J Clin Gilmour MN, Whittam TS. Methods of multilocus Microbiol 1992; 30: 967–973. enzyme electrophoresis for bacterial population 43. Preheim L, Pitcher D, Owen R, Cookson B. Typing genetics and systematics. Appl Env Microbiol 1986; of methicillin resistant and susceptible Staphylococcus 51: 873–884. aureus strains by ribosomal RNA gene restriction pat- 29. Musser JM, Kapur V. Clonal analysis of methicillin- terns using a biotinylated probe. Eur J Clin Microbiol resistant Staphylococcus aureus strains from intercon- Infect Dis 1991; 10: 428–436. tinental sources: association of the mec gene with 44. Hadorn K, Lenz W, Kayser FH, Shalit I, Krasemann divergent phylogenetic lineages implies dissemination C. Use of a ribosomal RNA gene probe for the epi- by horizontal transfer and recombination. J Clin demiological study of methicillin and ciprofloxacin Microbiol 1992; 30: 2058–2063. resistant Staphylococcus aureus. Eur J Clin Microbiol 30. Opal SM, Mayer KH, Stenberg MJ et al. Frequent Infect Dis 1990; 9: 649–653. acquisition of multiple strains of methicillin-resistant 45. Blumberg HM, Rimland D, Kiehlbauch JA, Terry Staphylococcus aureus by healthcare workers in an PM, Wachsmuth IK. Epidemiologic typing of endemic hospital environment. Infect Contr Hosp Staphylococcus aureus by DNA restriction fragment Epidemiol 1990; 11: 479–485. length polymorphisms of rRNA genes: Elucidation 31. Goullet P. An esterase zymogram of Escherichia coli. J of the clonal nature of a group of bacteriophage-non- Gen Microbiol 1973; 77: 27–35. typable, ciprofloxacin-resistant, methicillin-suscepti- 32. Branger C, Goullet P. Esterase electrophoretic poly- ble S. aureus isolates. J Clin Microbiol 1992; 30: morphism of methicillin-sensitive and methicillin- 362–369. resistant strains of Staphylococcus aureus. J Med 46. Richardson JF, Aparicio P, Marples RR, Cookson BD. Microbiol 1987; 24: 275–281. Ribotyping of Staphylococcus aureus: an assessment 33. Branger C, Goullet P. Genetic heterogeneity in methi- using well defined strains. Epidemiol Infect 1994; 112: cillin-resistant strains of Staphylococcus aureus 93–101. revealed by esterase electrophoretic polymorphism. J 47. De Buyser M, Morvan A, Grimont F, El Solh N. Hosp Infect 1989; 14: 125–134. Characterization of Staphylococcus species by riboso- 34. McGowan JE, Terry PM, Huang TSR, Linstrom mal RNA gene restriction patterns. J Gen Microbiol Houk C, Davies J. Nosocomial infection with gentam- 1989; 135: 989–999. icin-resistant Staphylococcus aureus: Plasmid analysis 48. Yoshida T, Kondo N, Hanifah YA, Hiramatsu K. as an epidemiologic tool. J Infect Dis 1979; 140: Combined use of ribotype, PFGE typing and IS431 864–872. typing in the discrimination of nosocomial strains of 35. Melo Cristino JAG, Torres Pereira A. Plasmid analy- methicillin-resistant Staphylococcus aureus. Microbiol sis of 219 methicillin-resistant Staphylococcus aureus Immunol 1997; 41: 687–695. strains with uncommon profiles isolated in Lisbon. J 49. Wei M, Udo EE, Grubb WB. Typing of methicillin- Hosp Infect 1989; 13: 133–141. resistant Staphylococcus aureus with IS256. FEMS 36. Hartstein AI, Phelps CL, Kwok RYY, Mulligan ME. Microbiol Lett 1992; 99: 175–180. In vivo stability and discriminatory power of methi- 50. Morvan A, Aubert S, Godard C, El Solh N. cillin-resistant Staphylococcus aureus typing by Contribution of a typing method based on IS256 restriction endonuclease analysis of plasmid DNA probing of SmaI-digested cellular DNA to discrimi- compared with those of other molecular methods. J nation of European phage type 77 methicillin-resis- Clin Microbiol 1995; 33: 2022–2026. tant Staphylococcus aureus strains. J Clin Microbiol 37. Lui PY, Shi Z, Lau Y et al. Use of restriction endonu- 1997; 35: 1415–1423. clease analysis of plasmids and pulsed-field gel 51. Symms C, Cookson B, Stanley J, Hookey JV. Analysis electrophoresis to investigate outbreaks of methi- of methicillin-resistant Staphylococcus aureus by cillin-resistant Staphylococcus aureus infection. Clin IS1181 profiling. Epidemiol Infect 1998; 120: 271–279. Infect Dis 1996; 22: 86–90. 52. Arpin C, Lagrange I, Gachie JP, Bebear C, Quentin 38. Locksley RM, Cohen ML, Quinn TC et al. Multiply C. Epidemiological study of an outbreak of infection antibiotic-resistant Staphylococcus aureus: Intro- with Staphylococcus aureus resistant to lincosamides duction, transmission, and evolution of nosocomial and streptogramin A in a French hospital. J Med infection. Ann Intern Med 1982; 97: 317–324. Microbiol 1996; 44: 303–310. 39. Jordens JZ, Hall LMC. Characterisation of methi- 53. Monzon-Moreno C, Aubert S, Morvan A, El Solh N. cillin-resistant Staphylococcus aureus isolates by Usefulness of three probes in typing isolates of restriction endonuclease digestion of chromosomal methicillin-resistant Staphylococcus aureus (MRSA). J DNA. J Med Microbiol 1988; 27: 117–123. Med Microbiol 1991; 35: 80–88.
MRSA typing methods 171 54. Kreiswirth BN, McGeer A, Kornblum J et al. The 68. Carles-Nurit MJ, Christophle B, Broche S, et al. use of variable gene probes to investigate a multihos- DNA polymorphisms in methicillin-susceptible and pital outbreak of MRSA. In: Novick RP, ed. methicillin-resistant strains of Staphylococcus aureus. Molecular biology of the staphylococci. New York: J Clin Microbiol 1992; 30: 2092–2096. VCH Publishers, 1990; 521–530. 69. Ichiyama S, Ohta M, Shimokata K, Kato N, 55. Kreiswirth B, Kornblum J, Arbeit RD et al. Evidence Takeuchi J. Genomic DNA fingerprinting by pulsed- for a clonal origin of methicillin resistance in field gel electrophoresis as an epidemiological marker Staphylococcus aureus. Science 1993; 259: 227–230. for study of nosocomial infections caused by methi- 56. Kuhl SA, Pattee PA, Baldwin JN. Chromosomal map cillin-resistant Staphylococcus aureus. J Clin Microbiol location of the methicillin resistance determinant in 1991; 29: 2690–2695. Staphylococcus aureus. J Bacteriol 1978; 135: 460–465. 70. Cookson BD, Aparicio P, Deplano A, Struelens M, 57. Dominguez MA, de Lencastre H, Linares J, Tomasz Goering R, Marples R. Inter-centre comparison of A. Spread and maintenance of a dominant methi- pulsed-field gel electrophoresis for the typing of cillin-resistant Staphylococcus aureus (MRSA) clone methicillin-resistant Staphylococcus aureus. J Med during an outbreak of MRSA disease in a Spanish Microbiol 1996; 44: 179–184. hospital. J Clin Microbiol 1994; 32: 2081–2087. 71. van Belkum A, van Leeuwen W, Kaufman ME et al. 58. Oliveira D, Santos-Sanches I, Mato R, et al. Virtually Assessment of resolution and intercenter repro- all methicillin-resistant Staphylococcus aureus ducibility of results of genotyping Staphylococcus (MRSA) infections in the largest Portuguese teaching aureus by pulsed-field gel electrophoresis of smaI hospital are caused by two internationally spread mul- macrorestriction fragments: a multicenter study. J tiresistant strains: the ‘Iberian’ and the ‘Brazilian’ Clin Microbiol 1998; 36: 1653–1659. clones of MRSA. Clin Microbiol Infect 1998; 4: 72. Trzcinski K, Hryniewicz W, Claus H, Witte W. 373–384. Characterization of two different clusters of clonally 59. Santos-Sanches I, Aires de Sousa M, Cleto L, Baeta related methicillin-resistant Staphylococcus aureus de Campos M, de Lencastre H. Tracing the origin of strains by conventional and molecular typing. J Hosp an outbreak of methicillin-resistant Staphylococcus Infect 1994; 28: 113–126. aureus infections in a Portuguese hospital by molecu- 73. Maslow JN, Brecher S, Gunn J, Durbin A, Barlow lar fingerprinting methods. Microb Drug Res 1996; 2: MA, Arbeit RD. Variation and persistence of methi- 319–329. cillin-resistant Staphylococcus aureus strains among 60. Tomasz A, de Lencastre H. Molecular microbiology individual patients over extended periods of time. Eur and epidemiology: Coexistence or alliance. In: Wenzel J Clin Microbiol Infect Dis 1995; 14: 282–290. RP, ed. Prevention and control of nosocomial infections. 74. Hu L, Umeda A, Kondo S, Amako K. Typing of Baltimore: Williams and Wilkins, 1997; 309–322. Staphylococcus aureus colonising human nasal carriers 61. van Leeuwen W, Sijmons M, Sluijs J, Verbrugh HA, by pulsed-field gel electrophoresis. J Med Microbiol van Belkum A. On the nature and use of randomly 1995; 42: 127–132. amplified DNA from Staphylococcus aureus. J Clin 75. Roman RS, Smith J, Walker M et al. Rapid Microbiol 1996; 34: 2770–2777. geographic spread of a methicillin-resistant 62. van Leeuwen W, Verbrugh HA, van der Velden J, van Staphylococcus aureus strain. Clin Infect Dis 1997; 25: Leeuwen N, van Belkum A. Validation of binary typ- 698–705. ing for Staphylococcus aureus strains. J Clin Microbiol 76. Tenover FC, Arbeit RD, Goering RV et al. 1999; 37: 664–674. Interpreting chromosomal DNA restriction patterns 63. Schwartz DC, Cantor CR. Separation of yeast chro- produced by pulsed-field gel electrophoresis: criteria mosome-sized DNAs by pulsed field gradient gel for bacterial strain typing. J Clin Microbiol 1995; 33: electrophoresis. Cell 1984; 37: 67–75. 2233–2239. 64. Chu G, Vollrath D, Davis RW. Separation of large 77. Struelens MJ, Bax R, Deplano A, Quint W, van DNA molecules by contour-clamped homogenous Belkum A. Concordant clonal delineation of methi- electric fields. Science 1986; 234: 1582–1585. cillin-resistant Staphylococcus aureus by macrore- 65. Tenover FC, Arbeit RD, Goering RV. How to select striction analysis and polymerase chain reaction and interpret molecular strain typing methods for epi- genome fingerprinting. J Clin Microbiol 1993; 31: demiological studies of bacterial infections: A review 1964–1970. for healthcare epidemiologists. Infect Contr Hosp 78. Goh S, Byrne S, Zhang JL, Chow AW. Molecular Epidemiol 1997; 18: 426–439. typing of Staphylococcus aureus on the basis of coagu- 66. Saulnier P, Bourneix C, Prevost G, Andremont A. lase gene polymorphisms. J Clin Microbiol 1992; 30: Random amplified polymorphic DNA assay is less 1642–1645. discriminant than pulsed-field gel electrophoresis for 79. Schwarzkopf A, Karch H. Genetic variation in typing strains of methicillin-resistant Staphylococcus Staphylococcus aureus coagulase genes: Potential and aureus. J Clin Microbiol 1993; 31: 982–985. limits for use as epidemiological marker. J Clin 67. Struelens MJ, Deplano A, Godard C, Maes N, Microbiol 1994; 32: 2407–2412. Serruys E. Epidemiologic typing and delineation 80. Hookey JV, Richardson JF, Cookson BD. Molecular of genetic relatedness of methicillin-resistant typing of Staphylococcus aureus based on PCR Staphylococcus aureus by macrorestriction analysis of restriction fragment length polymorphism and DNA genomic DNA by using pulsed-field gel electrophore- sequence analysis of the coagulase gene. J Clin sis. J Clin Microbiol 1992; 30: 2599–2605. Microbiol 1998; 36: 1083–1089.
You can also read