Antibiotic resistance and plasmids in Staphylococcus aureus from normal populations

Transcription

1 Journal of Antimicrobial Chemotherapy (99) 9, 3-39 Antibiotic resistance and plasmids in Staphylococcus aureus from normal populations Zarin Vlranl and W. C. NoMe* Department of Microbial Diseases, Institute of Dermatology, St Thomas' Hospital, London, SE 7EH, UK Antibiotic resistance, plasmid profiles and plasmid structures were examined in nasal Staphylococcus aureus collected from a random sample of normal individuals in the mid 96s and from women at an antenatal clinic in 989. The results were generally similar except for an increase in the resistance to penicillin in the later sample. The plasmid population had changed little though more uncategorized plasmids were evident in the 96s samples. Introduction Plasmid profiles and restriction fragment length polymorphisms (RFLP) are well established tools for the study of coagulase-negative and coagulase-positive staphylococci in hospitals (Richardson, Noble & Marples, 99). Studies on staphylococci from the general population are rather more rare, but the existence of two collections of Staphylococcus aureus strains from essentially normal populations separated by about years afforded an opportunity to examine plasmids in strains from outside hospital. Materials and methods The strains available were those collected during studies on the carriage of S. aureus in random samples of a normal population in the rural Netherlands made in the years 964/6 (Noble, Valkenburg & Wolters, 967) and those collected in a study of nasal, axillary and perineal carriage in women attending an antenatal clinic in the centre of London in 989 (Dancer & Noble, 99). Strains from the Netherlands were freezedried within a few subcultures after identification and remained in the original tubes until opened for this study, whilst those from London were used as fresh cultures within a few subcultures after isolation. All were identified as S. aureus by conventional tests, chiefly colony and microscopic appearance, slide clumping test and the tube coagulase test Antibiotic sensitivity tests were re-performed on Mueller-Hinton agar (Oxoid CM 337) using discs containing chloramphenicol 3 ng, erythromycin fig, gentamicin /ig, mupirocin fig, penicillin IU, streptomycin fig, sulphonamide 3 fig, tetracycline 3 fig and trimethoprim fig. Resistance to cadmium ( mg/l cadmium acetate) and ethidium bromide ( mg/l) was determined on blood agar base medium (Oxoid CM). Phage typing had previously been performed in strains isolated Downloaded from at Pennsylvania State University on May, 6 Correspondence to: Prof. W. C. Noble /9/3+ $./ 99 The Britiih Society for Antimicrobial Chemotherapy

2 36 Z. Vinud and W. C Noble during the first three months of the 964/6 study and on the strains from the 989 study by the Staphylococcus Reference Laboratory at Colindale. Because of changes in typing strategy between 96 and 989, strains are here classified as phage group I (9,, A, 79, 8, 8), group H (3A, 3B, 3C,, 7), group III (6, 7, 4E, 47, 3, 4, 7, 77, 83A, 8), miscellaneous (87, 4D, 94, 96, and mixed groups I, n, JH) (Misc) and non-typable (NT). Plasmid profiles were determined as described previously on strains selected from both collections (Noble & Rahman, 986). A further selection was made of strains resistant only to penicillin and bearing a single plasmid of about -3 kb. Plasmid DNA was prepared from these strains by caesium-chloride/ethidium bromide density gradient centrifugation or by extraction and precipitation by alkaline detergent as described elsewhere (Sambrook, Fritsch & Maniatis, 989). Restriction fragment length polymorphisms were determined using the enzyme EcoRl according to the makers (Gibco BRL) instructions. Escherichia coli V7 was used as a standard for intact plasmids and Lambda phage cut by HindlU (Gibco BRL) as a standard for restriction fragments. Results In 989 the overall nasal carrier rate was 4% in the women sampled; in 964/6 the mean nasal carrier rate for the whole survey was 9%, with a standard deviation of 7% for each of six random samples of approximately individuals (total population sampled 77). A total of strains was available from individuals sampled in 964/6 and 63 from 989. Antibiotic sensitivity patterns are shown in Table I. The major change is the increase in pencillin resistance in the 989 group, but other trends can be Table L Percentage distribution of antibiotic sensitivity and phage typing patterns Antibiotic resistance Sensitive to all antibiotics Resistant to penicillin only Resistant to penicillin + tetracycline only Resistant to penicillin + other 964/6 population population 7 3 Downloaded from at Pennsylvania State University on May, 6 No. Phage typing Group I Group II Group III Miscellaneous Non-typable No

3 Plasmkb in nan-hospital staphylococd 37 Table EL Distribution of plasmid profiles in relation to antibiotic sensitivity Number of plasmids seen on gel L L IS LS LS other 964/6 n - 84 Sensitive to all antibiotics 4 3 Resistant to penicillin only 4 Other patterns n = 3 Sensitive to all antibiotics 3 3 Resistant to penicillin only 8 6 Other patterns 3 L, Plasmid greater than circa kb; S, plasmid smaller than circa kb tee text seen. In the 964/6 strains there were nine (4-%) strains classed as resistant to streptomycin and a further (-7%) classed as 'intermediate' on the basis of zone diameter. In the 989 strains, only one (-6%) was classed as 'intermediate' to streptomycin and none as resistant Three (-4%) of the 964/6 strains were classed as resistant to tetracydine and none was resistant to any other antibiotic. In the 989 strains, three (4-8%) were resistant to tetracycline, and one each resistant to erythromycin and gentamitin. Most of these latter strains in both samples were also resistant to penicillin. None was resistant to chloramphenicol, mupirocin, sulphonamide or trimethoprim. No resistance to ethidium bromide was seen but many strains were resistant to cadmium. Overall 67/ (79%) strains were resistant to cadmium in the 964/6 collection and 8/63 (94%) in 989. This difference is chiefly due to the increase in penicillin resistant strains. Three strains (6%) were resistant to penicillin but sensitive to cadmium in 964/6 and one (%) in 989; amongst penicillin sensitive strains 9/49 (8%) were resistant to cadmium in 964/6 and / (8%) in 989. The phagc types are also shown in Table I. In 964/6 the strains were spread reasonably evenly over the phage groups but in 989 the number of non-typable strains had increased. Plasmid profiles were difficult to group meaningfully as no attempt was made to identify those plasmids which mediate antibiotic resistance. Results are recorded here on the simple scheme that a plasmid running behind the residual chromosomal DNA band in the agarose gel is described as 'large' (greater than about kb) and those running ahead of the chromosomal DNA as 'small'. The results are shown in Table n. In the 964/6 strains, (37%) of the 4 strains resistant to penicillin only (mainly from phage groups ID, Misc and NT), revealed no plasmid. In 989,8 (46%) of the 39 penicillin resistant strains (from all phage groups) had no plasmid, a difference that was not statistically significant. There were a number of uncategorized plasmids in all sensitivity patterns in both groups but the results are too diverse to merit detailed analysis. Nevertheless there were uncategorized plasmids in the 66 strains either resistant to penicillin alone or sensitive to all antibiotics in 964/6 compared with 9 in 4 similar strains from 989. RFLP patterns using cori were determined for strains resistant to penicillin only and possessing one large plasmid only. Three patterns were observed repeatedly, and Downloaded from at Pennsylvania State University on May, 6

4 38 Z. Vlmri and W. C NoMe Table UL Distribution of RFLP patterns according to phage group of staphylococcus Phage group typea typeb typec Other 964/6 989 I n m Misc NT I n m Misc NT 4 3 Type A fragments 3, 9-, 6, - kb, type B fragmenti,9-, 6, - kb, type C fragment* 9, 4-, -9 kb. strains from both studies fell into all three groups. The most frequent was a pattern with fragments of about 3, 9-, 6 and - kb which was found principally in the phage group ITI, Misc and NT groups. A second pattern with fragments of about, 9-, 6 and - kb was found mainly in group I and was indistinguishable from that of the plasmid in NCTC 9789, the phage 8 propagating strain, and the remaining pattern of about 9, 4-, and -9 kb was chiefly in phage group II strains. A few patterns did not fit in to any of these groups (Table TI). Discussion Although the two populations are not strictly comparable, there has been no other random survey of nasal staphylococci in normal populations, and no other antenatal strains from the 96s were available. The women attending the antenatal clinic can be regarded as normal as they do not represent a hospital inpatient population and, in the 989 study, those with any hospital contact had been omitted. In addition, in the 964/ 6 survey there was no difference in nasal carriage in respect of sex, hospitalization or outpatient attendance for self or family in the six months before the survey. The chief difference between the two groups of staphylococci studied lies in the increased proportion of strains resistant to penicillin in the 989 population. The other changes in resistance, though small, reflect changes in the availability and prescription of antibiotics. There is a suggestion that uncategorized plasmids were more common in the staphylococci from 964/6. Dyke & Noble (984) reported that plasmids of 3- kb mediating resistance to cadmium were more prevalent in phage group II S. aureus from the mid 96s than from the early 98s. Assuming a plasmid of -3 kb to represent the plasmid mediating penicillinase production in strains resistant only to penicillin, the ratio of chromosomal to plasmid mediated penicillin resistance is not statistically significantly different in the two groups. We have not shown that these plasmids were the penicillinase plasmids, but the plasmids studied were distributed in much the same way in the two populations. Use of only one enzyme does not, of course, indicate other than a general similarity in structure, and minor changes can be anticipated (Dyke & Noble, 984). Studies on uncategorized plasmids, and on plasmids mediating resistance 3 Downloaded from at Pennsylvania State University on May, 6

5 Plasmids in non-hospital staptiylococd 39 to tetracyclinc or streptomycin have also shown a high degree of conservation of genetic material over similar time periods (Dyke & Noble, 984; Rahman, Kent & Noble, 99). The shift in phage type to NT strains may be due to lysogenization by phage; this would also explain the shift in distribution of plasmid RFLP patterns from 964/6 to 989. Changes in antibiotic resistance between 964/6 and 989 have not greatly altered the balance of the genes mediating those resistances in normal populations outside hospital. References Dancer, S. J. & Noble, W. C. (99). Nasal, axillary, and perineal carriage of Staphyhcoccus aureus among women: identification of strains producing epidermolytic toxin. Journal of Clinical Pathology 44, Dyke, K. G. H. & Noble, W. C. (984). Plasmids of phage-group-h Staphylococcus aweus. Journal of Medical Microbiology 7, Noble, W. C. & Rahman, M. (986). Plasmid profiling of epidemic staphylococci from around 96; a comparison of epidemiological techniques. Journal of Hygiene 97, -8. Noble, W. C, Valkenburg, H. A. & Wolters, C. H. L. (967). Carriage of Staphylococcus aureus in random samples of a normal population. Journal of Hygiene 6, Rahman, M., Kent, L. & Noble, W. C. (99). Streptomycin and tetracycline resistance plasmids in Staphylococcus hyicus and other staphylococci. Journal of Applied Bacteriology 7,-. Richardson, J. F., Noble, W. C. & Marplcs, R. R. (99). Identification and typing of the staphylococci. In Methods in Microbiology. Society for Applied Microbiology, London. Sambrook, J., Fritsch, E. F. & Maniatis, T. (989). Molecular Cloning. A Laboratory Manual, nd edn, 3 vols. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. {Received February 99; revised version accepted 9 September 99) Downloaded from at Pennsylvania State University on May, 6

6 Downloaded from at Pennsylvania State University on May, 6