Serotype markers in a Streptococcus agalactiae strain collection from Zimbabwe.

OBJECTIVE: Group B streptococci (GBS) from Southern African areas have been less well characterized. Our objective was to study serotype and serovariant distribution of carrier GBS strains as part of a study of the epidemiology of GBS carriage in pregnant women from Zimbabwe. MATERIALS AND METHODS: We studied GBS isolated from 121 healthy pregnant women living in Harare and surrounding areas, Zimbabwe. Capsular polysaccharide (CPS) testing for serotype determination and surface-anchored protein testing for serosubtype determination were done by gene-based serotyping (PCR), except for the proteins R3 and a novel protein called Z, which were detected by antibody-based methods. RESULTS: Strains of the CPS types Ia (15.7%), Ib (11.6%), II (8.3%), III (38.8%), V (24.0%) and NT (1.7%) were detected along with the strain-variable proteins Cί (15.7% of isolates), Cα (19.8%), Alp1 (epsilon-22.3%), Alp3 (5.0%), R4/Rib (46.3%), R3 (27.3%), Z (27.3%), and SAR5 (28.9%), which encodes the R5 protein. Up to four of the protein genes could be possessed or the gene product expressed by one and the same isolate. A total of 32 serovariants were detected. The findings assessed by us as most important were the very low prevalence of the gene Alp3 (Alp3--4.9%), high prevalence of R4 (Rib--46.2%), the proteins R3 (27.3%), Z (27.3%), and of SAR5 (R5--28.9%). The low prevalence of Alp3, notably in GBS type V strains, differed from findings with CPS type V GBS from non-African areas. Bacteria of the various CPS types showed distinct CPS/protein-marker associations. CONCLUSION: The results are of importance in relation to regional variations of GBS phenotypes and genotypes and thus, of importance in planning and research in the context of future vaccine formulations.

Antibodies against Streptococcus agalactiae proteins c(alpha) and R4 in sera from pregnant women from Norway and Zimbabwe.

Group B streptococci (GBS) express strain-variable and surface-localized proteins, which are important serotype markers and targets of protective antibodies. These include the c(alpha) and R4 proteins, one or the other of which is expressed by approximately 75% of clinical GBS isolates. These proteins have been considered vaccine candidates. In this study, the c(alpha) and R4 proteins were extracted by trypsin digestion of GBS and purified by sequential precipitation with trichloroacetic acid and ammonium sulfate followed by gel filtration chromatography. The proteins were used as antigens in an indirect enzyme-linked immunosorbent assay (ELISA) to measure the levels of c(alpha)- and R4-reactive antibodies in sera from pregnant women from Norway (n = 100) and from Zimbabwe (n = 124). Antibody levels in the Norwegian group of women were significantly higher than in the Zimbabwean group, and a higher proportion of the Norwegian women contained appreciable levels of antibodies against both proteins. The antibodies traversed the placental barrier. With individual sera, a significant correlation between the anti-c(alpha) and anti-R4 antibody levels was observed and each of the two protein antigens effectively competed for human serum antibodies both against itself and against the other antigen. Inhibition ELISA results demonstrated specificity for each of the proteins of immune antibodies raised in rabbits. These results demonstrate that (i) the majority of women of childbearing age have antibodies against c(alpha) and R4, (ii) the levels of these antibodies differ among pregnant women in different parts of the world, and (iii) the normal human serum antibodies may target a common c(alpha) and R4 protein site, whereas immune antibodies targeted a different site(s) specific for each protein.

[Erythromycin and ciprofloxacin resistant Campylobacter jejuni]. / Erytromycin- og ciprofloksacinresistens hos Campylobacter jejuni.

BACKGROUND: Diarrhoea caused by Campylobacter is normally a self-limiting disease, but treatment with antibiotics may be indicated in very severe or complicated cases or in immunocompromised patients. This makes knowledge of the susceptibility of the Campylobacter to antibiotics important. MATERIAL AND METHODS: We examined retrospectively the in vitro susceptibility to erythromycin and ciprofloxacin of 296 C. jejuni strains isolated during the 1998-99 period. Particular attention was paid to the area in which the infection was acquired. RESULTS: Ciprofloxacin-resistant isolates were found in 69 (23%) of the patients. Among 121 patients probably infected abroad, the rate of resistance was 46%, highest among patients infected in Southern Europe (59%), followed by Asian and African countries. Such resistance was not recorded among strains acquired in Norway. Only one isolate showed resistance to erythromycin. A few strains showed intermediate susceptibility to both antibiotics. INTERPRETATION: During the 1998-99 period, nearly one in four of all C. jejuni isolates in Sør-Trøndelag County, Norway, were resistant to ciprofloxacin. Resistant isolates were acquired outside Norway. Resistance of C. jejuni to erythromycin occurred very rarely. In Norway erythromycin should still be the drug of choice in campylobacteriosis in cases where antibiotic treatment is indicated.

Susceptibility of Zimbabwean Streptococcus agalactiae (group B Streptococcus; GBS) isolates to four different antibiotics.

OBJECTIVE: To establish the susceptibility of Zimbabwean GBS strains isolated from hospitalised patients to four antibiotics. DESIGN: Cross sectional survey. SETTINGS: Four regions of Zimbabwe (Bindura, Bulawayo, Harare, and Masvingo). SUBJECTS: 113 GBS isolates from hospitalised patients in Bindura, Bulawayo, Harare and Masvingo, of whom most were suffering from infectious diseases. MAIN OUTCOME MEASURES: All isolates were tested for their susceptibility to clindamycin, erythromycin, penicillin and tetracycline. RESULTS: All isolates were 100% sensitive to clindamycin, 98% to penicillin, 86% to erythromycin; 2% of the isolates showed intermediate susceptibility to penicillin and 100% showed resistance to tetracycline. CONCLUSION: Penicillin is still the antibiotic of choice for treatment of GBS infections and for intrapartum chemoprophylaxis in Zimbabwe. For patients who are allergic to penicillin, clindamycin will be the drug of choice for both treatment and/or chemoprophylactic use in Zimbabwe.

The putative R1 protein of Streptococcus agalactiae as serotype marker and target of protective antibodies.

The streptococcal R1 protein was studied by means of anti-R1 antibodies prepared by appropriate cross-absorption of rabbit antiserum raised against the group B streptococcal (GBS) strain ATCC 12403 (D136C), serotype III/R1. The protein was a ladder-forming antigen according to banding patterns in immunoblotting, similar to several other GBS proteins, and was susceptible to digestion by both pepsin and trypsin. Antibody-based testing revealed that 10% of Norwegian GBS isolates expressed the R1 protein, most frequently capsular antigen type V strains (72%) and less frequently type III strains (3%). None of 132 GBS strains from Zimbabwe, including 39 type V strains, expressed the R1 protein. R1-specific rabbit antibodies showed protective activity in mice challenged with a GBS type V/R1 strain. The results show that the R1 protein is an important GBS serotype marker in strains from certain geographical areas, notably for the subtyping of capsular type V strains, and that this protein is a target of protective antibodies.

Distribution and expression of bca, the gene encoding the c alpha protein, by Streptococcus agalactiae.

A total of 52 clinical isolates of group B streptococci (GBS) was tested for expression of the c protein c(alpha) by a fluorescent antibody test (FAT) and by PCR amplification of a 202-bp stretch within the repeat unit of the bca gene encoding the c(alpha) protein. The strains were categorised as follows: c(alpha) FAT positive and PCR positive with amplification products of multiple sizes (category A, n = 12); FAT negative and with PCR products of multiple sizes (category B, n = 11); FAT negative and with a single PCR product of c. 200 bp (category C, n = 5); negative in both tests (category D, n = 24). A single amplification product of minimum size and additional products of larger sizes corresponded to one and more bca repeats, respectively. Five of the 11 category B strains showed expression of low Mr c(alpha) in whole cell-based Western blotting. The results showed that a proportion of the GBS isolates harboured bca gene elements that either were not expressed or they expressed c(alpha) molecular variants which could not be detected by the whole cell-based FAT. This genotype/phenotype discrepancy should be considered in relation to GBS typing, including the selection of antibody reagents and the technical approach to c(alpha) protein detection.

Prevalence, capsular type distribution, anthropometric and obstetric factors of group B Streptococcus (Streptococcus agalactiae) colonization in pregnancy.

OBJECTIVE: To establish the prevalence, serotype distribution, anthropometry and obstetric factors of Group B Streptococcus (GBS) colonization in pregnant women. DESIGN: Cross sectional survey. SETTING: Chinhoyi General Hospital. SUBJECTS: 206 pregnant women attending the antenatal clinic at Chinhoyi General Hospital were systematically randomly sampled. MAIN OUTCOME MEASURES: All the isolates were serotyped on the basis of capsular polysaccharide (CHO) antigen designated, Ia, Ib, II, III, IV and V. RESULTS: 65 (31.6%) were carriers of GBS. The serotypes found were, type III (41.8%), type V (37.4%), type Ia (11.0%), type IV (3.3%), type Ib (3.3%) type II (1.0%) and 2.0% of the isolates were non-typable. All isolates were sensitive to penicillin and resistant to gentamycin. Colonization was more common in women with parity 0 to 2 (4.6%) and age group 20 to 24 years (43.1%). There was some evidence (p = 0.063) to suggest that GBS was more often isolated from the vagina (12.6%) than from the rectum (6.3%). CONCLUSION: There was a high prevalence of GBS colonization among pregnant women in Chinhoyi. Types III and V were the most common serotypes found.

bca, beta gene, and gene product divergency in reference and prototype strains of streptococcus agalactiae.

Reference and prototype strains of Streptococcus agalactiae (GBS) were originally selected on the basis of phenotypic traits which, however, do not always mirror genotypic traits. A total of 14 reference and prototype GBS strains were examined by PCR designed to detect the bca and beta genes, encoding the c proteins c(alpha) and c(beta), respectively. The cognate proteins were detected by whole-cell-based fluorescent antibody testing and Western blotting. The PCR for beta gene detection and the antibody-based c(beta) protein detection showed concordant results with all of the isolates, whereas 7 of 14 strains which did not express c(alpha) protein at detectable levels contained bca gene elements, consistent with bca gene and gene product divergency in these strains. The results emphasize the importance of genetic characterization of reference and prototype strains of GBS which, in the past, have been selected on the basis of phenotypic traits.

Properties and distribution of the putative R3 protein of Streptococcus agalactiae.

Strain-variable Streptococcus agalactiae (group B streptococci; GBS) proteins exposed at the bacterial cell surface are important markers in GBS serotyping. These proteins include the c proteins c(alpha) and c(beta) and the R proteins R1 through R4, of which R1 and R4 have been studied most extensively. This study presents the characteristics of a protein which was expressed by a capsular antigen type V GBS strain shown by means of polyclonal and monoclonal antibody testing. Examination of a number of reference and prototype strains by fluorescent antibody testing and Western blotting provided evidence that the serotype V-derived protein was the R3 protein of GBS, previously defined on the basis of immunoprecipitation assays. The putative R3 protein formed ladder-like banding patterns on Western blotting with polypeptides in the 30 kDa to > or = 140 kDa range, was destroyed by pepsin digestion, and partially degraded by trypsin digestion. The protein was expressed by 10 (6.5%) of 153 clinical GBS strains tested, the expression being restricted to isolates of the capsular antigen types II, III, and V. Some isolates expressed both the c(beta) and the R3 protein. Expression in combination with c(alpha) or R4 protein synthesis was not detected. Inclusion of the anti-R3 monoclonal antibody among antibody reagents for GBS serotyping will enhance the discriminatory power of this typing method.

Comparison of three different methods in monoclonal antibody-based detection of Streptococcus agalactiae protein serotype markers.

Surface-exposed proteins are important serotype markers in Streptococcus agalactiae (group B streptococci; GBS). The proteins include the c proteins c(alpha) and c(beta), the R4 protein and a protein provisionally called P. For all of these markers, protein-specific monoclonal antibodies (MAbs) have been generated. We have compared whole-cell-based fluorescent antibody testing (FAT), ELISA, and dot blotting for MAb-based detection of these proteins by testing a panel of 52 GBS isolates of different capsular antigen types. Of a total of 208 observations with each of the tests, positive signalling in the dot assay was observed in 32.2%, with ELISA in 27.8%, and with FAT in 26.4% of the recordings. Discordant results were noted most frequently with the c(beta) and c(alpha) MAbs. In the case of c(alpha) the reason for the discordant test results was further examined and it appeared that this could be attributed to low level expression of the c(alpha) protein, although structural variations of c(alpha) proteins cannot be excluded. Our findings favour dot blotting as the method of choice although we consider all three methods acceptable for serotyping of GBS.

An epitope shared by enterobacterial and neisserial porin proteins.

A murine monoclonal antibody (MAb F9-16) raised against a porin protein epitope called Po I of an E. coli 055 strain showed broad cross-reactivity with bacteria within the Enterobacteriaceae, and also recognized neisseriae and moraxellae. In an immunodot assay, the antibody was bound by 32/33 strains of neisseriae and moraxellae after SDS treatment of the bacteria. Testing intact bacteria, 11/33 isolates showed definite MAb binding, including serogroup A and B meningococci. In Western blotting, the anti-Po I MAb targeted the gonococcal porin proteins PIA and PIB, and class 1, class 2, and class 3 porins of meningococci. The MAb showed no reactivity against decapeptides which corresponded to the whole length of a meningococcal class 1 porin protein of the subtype P1, 7, 16. These findings accord with the inference that enterobacterial, neisserial and moraxellae porin proteins share an epitope (Po I) which is determined by the three-dimensional rather than by the primary structure of the proteins and that this epitope is shielded in most isolates but surface-exposed in some isolates, including some strains of meningococci. Since Po I is broadly distributed among commensal and pathogenic bacteria and has demonstrated immunogenicity in humans, this epitope may play a role in elicitation of "normal" antibodies with immunoprotective activity.

Streptococcus agalactiae beta gene and gene product variations.

Streptococcus agalactiae (group B streptococci; GBS) are serotyped on the basis of the capsular polysaccharide antigens and subtyped on the basis of the strain-variable and surface-localised c proteins c alpha, c beta, and R proteins. This study compared c beta protein detection and the polymerase chain reaction (PCR) for beta gene detection, by examining 50 clinical GBS strains. The c beta protein was detected by antibody-based immunofluorescence in a GBS whole-cell assay and Western blotting by probing with the anti-c beta antibody or human IgA. Absorption experiments were performed to test for surface-anchoring of c beta; and bacterial supernates were examined to test for c beta production. Primers for the PCR target regions resulted in a 620-bp product that included beta gene-encoding IgA-binding domains. The results demonstrated four categories of GBS with respect to the beta gene and the c beta protein: (1) strains (16 of 50) that harboured the beta gene and regularly expressed normal surface-localised c beta with a M(r) of 120 kDa; (2) strains (5 of 50) that harboured the gene but did not express the protein; (3) strains (2 of 50) that harboured the gene but expressed a c beta that was not surface-localised and had reduced M(r); (4) strains (27 of 50) without beta gene and c beta expression. One strain amongst the third group generated a PCR product of 1330 bp. These results demonstrate considerable strain variability of the beta gene of GBS and of its product the c beta protein.

Mechanisms of methicillin resistance in staphylococci.

The continuously high prevalence of methicillin-resistant staphylococci (MRS) throughout the world is a constant threat to public health, owing to the multiresistant characteristics of these bacteria. Methicillin resistance is phenotypically associated with the presence of the penicillin-binding protein 2a (PBP2a) not present in susceptible staphylococci. This protein has a low binding affinity for beta-lactam antibiotics. It is a transpeptidase which may take over cell wall synthesis during antibiotic treatment when normally occurring PBPs are inactivated by ligating beta-lactams. PBP2a is encoded by the mecA gene, which is located in mec, a foreign DNA region. Expression of PBP2a is regulated by proteins encoded by the plasmid-borne blaR1-bla1 inducer-repressor system and the corresponding genomic mecRl-mecl system. The blaRl-blal products are important both for the regulation of beta-lactamase and for mecA expression. Methicillin resistance is influenced by a number of additional factors, e.g. the products of the chromosomal fem genes which are important in the synthesis of normal peptidoglycan precursor molecules. Inactivation of fem-genes results in structurally deficient precursors which are not accepted as cell wall building blocks by the ligating PBP2a transpeptidase during antibiotic treatment. This may result in reduced resistance to beta-lactam antibiotics. Inactivation of genes affecting autolysis has shown that autolytic enzymes are also of importance in the expression of methicillin resistance. Methicillin resistance has evolved among earth microorganisms for protection against exogenous or endogenous antibiotics. Presumably the mec region was originally transferred from coagulase negative staphylococci (CNS) to Staphylococcus aureus (SA). A single or a few events of this kind with little subsequent interspecies transfer had been anticipated. However, recent data suggest a continuous horizontal acquisition by S. aureus of mec, being unidirectional from CNS to SA. Methicillin resistance may also be associated with mechanisms independent of mecA, resulting in borderline methicillin resistance. These mechanisms include beta-lactamase hyperproduction, production of methicillinases, acquisition of structurally modified normal PBPs, or the appearance of small colony variants of SA. Most MRS are multiresistant, and the mec region may harbour several resistance determinants, resulting in a clustering of resistance genes within this region.

Feasibility of Helicobacter pylori identification by a slide agglutination test.

Culture isolation and identification of Helicobacter pylori represents a considerable work load in clinical microbiology. The aim of this study was to test if antibody-mediated bacterial agglutination could be used for rapid identification of H. pylori. Rabbit antiserum against H. pylori strain I and against another strain, H. pylori 330, which was very weakly agglutinated (1+) by anti-H. pylori I serum, were mixed and used in a slide agglutination test. Of 107 consecutive clinical isolates tested, 101 (94%) strains showed 2+ or 3+ reaction using the antiserum mixture, whereas 6 (6%) strains could not be evaluated owing to autoagglutinability. Bacteria of a variety of other species, including Campylobacter spp., showed no agglutination with the antiserum mixture. The results support the notion that reliable identification of the majority of cultured H. pylori strains should be possible in less than 3 min by agglutination testing.

Relationship between the efficacy of amoxicillin and intragastric pH for the treatment of Helicobacter pylori infection.

BACKGROUND: Proton pump inhibitors are reported to enhance the efficacy of antibiotics in the treatment of Helicobacter pylori infection. An elevated intragastric pH is considered to be an important factor for this increased antimicrobial efficacy. The aim of this study was to assess the effect of different doses of lansoprazole on 24-hour intragastric pH and to correlate the effect of amoxicillin on the cure rate for H. pylori infection with the intragastric pH obtained during lansoprazole treatment. PATIENTS AND METHODS: Thirty-six duodenal ulcer patients who tested positively for H. pylori as assessed by a rapid urease test, culture, and histological evaluation were allocated randomly to dual treatment with amoxicillin, 3 gm/day, and lansoprazole in different doses ranging between 30 and 180 mg/day for 2 weeks. A 24-hour intragastric pH measurement was taken in all patients on the fifth day of treatment. H. pylori status was determined by culture and histological workup 6 weeks after cessation of the amoxicillin-lansoprazole medication. RESULTS: The H. pylori infection was treated successfully in 19 of 32 patients who completed the dual therapy (per protocol, 59.4%). The median intragastric pH in patients who were treated successfully was 4.4 (95% confidence interval [CI] = 3.7-4.7), as compared to 4.0 (95% CI = 3.5-4.5) in patients who were not treated successfully (p = .47, Wilcoxon's rank sum test). The median percentage of time that the intragastric pH exceeded 4 was not different in the two groups (p = .77). Administration of lansoprazole in doses exceeding 30 mg induced only a moderate additional increase in intragastric pH. CONCLUSIONS: Profound inhibition of gastric acid secretion seems not to be necessary to improve the effect of amoxicillin on the cure rate for H. pylori infection in patients with duodenal ulcers.

A Streptococcus agalactiae R protein analysed by polyclonal and monoclonal antibodies.

Unexpected cross-reactivity between two Streptococcus agalactiae (GBS) isolates formed the basis for purification of a GBS protein called the Ra antigen, and raising of murine monoclonal antibody (MAb) against Ra. The Ra protein was resistant to trypsin digestion, susceptible to pepsin digestion, formed a ladder-like pattern of lines with a periodicity of approximately 8 kD on immunoblotting, was surface-localized in GBS strains, and was variably expressed by GBS. These characteristics provided evidence that the Ra antigen belonged to the R proteins of GBS. By testing of reference GBS isolates and antiserum, including an anti-R4 protein serum, cross-reactivity was recorded consistent with the assumption that Ra is a R4 protein. The Ra/R4 protein also showed cross-reactivity with a previously described GBS protein called protein Rib (J. Exp. Med. 177: 1593-1603, 1993). Several characteristics of the Ra/R4 protein were similar to those of the GBS protein c alpha, but the two proteins showed no cross-reactivity. The anti-Ra/R4 MAb has proved useful in serosubtype determination of GBS of known serotype and should be a valuable tool for studying the immunobiological function of antibodies targetting the surface-localized Ra/R4 protein.

Immunogenicity expressed in patients with bacteraemia of an epitope shared by enterobacterial and neisserial porin proteins.

A monoclonal antibody (MAb) against an epitope (Po I) on an Escherichia coli O55 porin protein has shown broad cross-reactivity with other Enterobacteriaceae and with both pathogenic and non-pathogenic Neisseriaceae. In this study, we have measured antibody levels against the Po I site in patients with bacteraemia in order to examine the immunogenicity of the Po I domain in humans. A MAb-based competition ELISA (cELISA) was used. Only 20% of healthy controls had detectable levels of anti-Po I antibodies in serum. Of patients bacteraemic with enterobacteria (n = 45), 11% and 58% showed elevated antibody levels compared to healthy controls with the first and second serum specimens, respectively, and 73% of these patients showed > or = 10% increase in the antibody levels. Of patients bacteraemic with N. meningitidis (n = 20), only 30% showed > or = 10% increase in the antibody levels when paired serum specimens were tested. Levels of competing antibodies were similar in the cELISA with N. meningitidis (B: 15: P1, 7, 16) OM coat or E. coli O55 OM coat. The results demonstrated that the highly conserved porin protein domain Po I expressed immunogenicity in humans when present in bacteria which caused bacteraemia. This finding represents a challenge in further investigations on the immunobiological role of the cross-reacting antibodies.

Distribution of serovariants of group B streptococci in isolates from England and Norway.

The distribution of capsular polysaccharide antigen (CHO) types, surface-exposed c proteins alpha (c alpha) and beta (c beta) and an R-protein antigen was examined in 334 group B streptococci (GBS) isolates from three groups of patients hospitalised in England and Wales or Norway. The isolates were from 108 carriers, 67 cases of neonatal infection and 154 cases of adult infection. Each group contained all CHO types (Ia, Ib, II, III, IV, V and NT); type III strains predominated except in the adult infected group. Strains within each CHO type could be further subdivided by the protein markers into five subtypes by a combined typing system. The proportion of type Ib and type III strains in the neonatal infection cases and of type Ib strains in the adult infection cases significantly outnumbered isolates of these serotypes among the carrier strains. Twenty-nine different serovariants were identified; 24, 13 and 23 serovariants among the carrier, neonatal infection and adult infection isolates, respectively. Certain CHO antigen-protein associations were identified, notably those between Ia/c alpha, Ib/c alpha beta and III/R. The proportion of invasive isolates that expressed protein was not higher than in the carrier isolates. All CHO-type Ib isolates contained a c protein, but 7% of the Ib isolates did not contain any of these proteins. These findings indicate that this combined typing approach may be useful in examining epidemiological problems associated with GBS.

Direct identification of Staphylococcus aureus in blood cultures by detection of the gene encoding the thermostable nuclease or the gene product.

This study compares methods for direct identification of S. aureus in blood cultures by detection of the thermonuclease (TNase) of this bacterium or the nuc gene encoding it. The protein was detected by an enzyme diffusion test in o-toluidine blue DNA agar with a test time of at least 4 h, by a monoclonal antibody (MAb)-based sandwich enzyme-linked immunosorbent assay (sELISA) with a test time of approximately 4 h, and by a MAb-based sandwich enzyme-linked immunofiltration assay (sELIFA) with a test time of 25-30 min (sample preparation included). The nuc gene was amplified by a polymerase chain reaction (PCR) with a test time (amplification plus detection) of approximately 3.5 h. The tests were optimized for direct examination of blood-containing cultures. All tests were positive with 67/67 blood cultures which grew S. aureus, negative with 35/35 cultures which grew coagulase-negative staphylococci, and negative with 37/37 cultures with various other bacteria. These results showed positive agreement with those of the commercial AccuProbe test but not with the StaphAurex agglutination kit. With an artificially seeded blood culture, minimum total times required (incubation plus testing) were as follows: nuc-PCR, 9.5 h; sELIFA, 12.5 h; enzymatic test, 16-36 h; AccuProbe, 14 h. Direct examination of both the nuc gene and the mecA gene encoding methicillin resistance demonstrated the mecA gene in all the coagulase-negative staphylococci (48.6%) which showed oxacillin resistance. The sELIFA had the particular advantage of its short test time, the PCR its high sensitivity and the possibility of simultaneous detection of the species-specific nuc gene and genes encoding other clinically important characters of the bacteria. These tests offer the prospect of direct application to a variety of clinical specimens for rapid diagnosis of S. aureus infection.