Journal of Antimicrobial Chemotherapy (2000) 46, 205-213
© 2000 The British Society for Antimicrobial Chemotherapy
Antibiotic resistance patterns of aerobic coryneforms and furazolidone-resistant Gram-positive cocci from the skin surface of the human axilla and fourth toe cleft
The Skin Research Centre, Division of Microbiology, School of Biochemistry and Molecular Biology, University of Leeds, Leeds LS2 9JT, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Samples of skin surface bacteria from 28 healthy subjects plated directly on to selective and non-selective media revealed that the proportion of aerobic coryneforms and furazolidone-resistant Gram-positive cocci (FURECs) resistant to erythromycin was significantly greater in the fourth toe cleft than in the axilla (P < 0.05). There were more erythromycin-resistant bacteria than tetracycline-resistant bacteria at both sites (P = 0.001 for the toe cleft; P < 0.01 for the axilla). In total, 160 distinct isolates were obtained, of which 42 were FURECs and 118 were aerobic coryneforms. Of these, 153 (96%) were resistant to erythromycin and 66 (41%) to tetracycline. All except seven of the tetracycline-resistant strains were also resistant to erythromycin. The resistant isolates belonged to a variety of species. CDC group ANF corynebacteria were most numerous and composed 31% of all isolates. The majority (76%) of FURECs were identified as Micrococcus luteus. MIC determinations on selected strains revealed that tetracycline-resistant FURECs were sensitive to doxycycline and minocycline, as were most tetracycline-resistant coryneforms. Nine coryneform isolates were cross-resistant to all three tetracyclines. Only a minority of erythromycin-resistant FURECs (21%) demonstrated a macrolide–lincosamide–streptogramin type B (MLS)-resistant phenotype with inducible or constitutive cross-resistance to clindamycin and the type B streptogramin, pristinamycin IA. Twenty-nine erythromycin-resistant FURECs had a novel phenotype distinct from MLS and macrolide–streptogramin type B resistance. In contrast, most coryneforms (79%) were MLS resistant. Among the remainder, two unusual erythromycin resistance phenotypes were apparent, both of which differed from the unusual phenotype in FURECs. This study has revealed that the non-staphylococcal aerobic flora of skin contains a considerable reservoir of tetracycline and erythromycin resistance determinants. The three unusual macrolide resistance phenotypes may be associated with novel resistance mechanisms.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The resident commensal microflora of human skin comprises several bacterial genera, including staphylococci, micrococci, corynebacteria, brevibacteria and propionibacteria.1–4 The genus formerly known as Micrococcus has been dissected into five separate genera; Micrococcus Cohn 1872 gen. emend., Kocuria gen. nov., Nesterenkonia gen. nov., Kytococcus gen. nov. and Dermacoccus gen. nov.5 These organisms form a group of furazolidone-resistant Gram-positive cocci (FUREC). The prevalence of antibiotic-resistant coagulase-negative staphylococci (CNS) at various skin sites in both healthy and hospitalized patients has received considerable attention because of the role of these organisms as nosocomial pathogens in compromised hosts. It is known that they harbour a considerable reservoir of resistance genes even in untreated individuals and that they share a gene pool with Staphylococcus aureus.2,6–8 The non-staphylococcal components of the skin flora are phylogenetically much more closely related to one another than any of them are to staphylococci. In contrast to the staphylococci, their chromosomal DNA is G+C-rich. It seemed to us likely that these organisms may also share a resistance gene pool, distinct from that in staphylococci, which has been little characterized.
The aim of this study was to determine the prevalence of different erythromycin and tetracycline resistance phenotypes in populations of aerobic coryneform bacteria and FURECs from the human axilla and fourth toe cleft. These moist skin sites are the major habitats of these organisms.9 One reason for selecting these particular antibiotics is that we already know that cutaneous populations of staphylococci are commonly resistant to them, as are propionibacteria from treated acne patients.10–12 Thus it seemed likely that the other major resident skin bacteria would demonstrate similar resistance. A second reason for looking at erythromycin resistance was to detect novel phenotypes that were incompatible with the possession of either macrolide–lincosamide–streptogramin type B (MLS) or macrolide–streptogramin type B (MS) resistance determinants (erm genes or msrA).
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Bacterial strains
Aerobic coryneform bacteria and FURECs were isolated from either the fourth toe cleft of the right foot or from both axillae of 28 healthy volunteers (22 male, six female; age range 21–46 years), none of whom were receiving antibiotic therapy at the time of sampling. Only three volunteers were sampled from both sites, to minimize the chance of re-isolating the same strains. Samples from the fourth toe cleft were taken, using sterile cotton-tipped swabs moistened in full-strength wash fluid, from the base of the cleft and along the sides of the toes forming the cleft.13 Axillary bacteria were collected using the detergent scrub technique of Williamson & Kligman.13 Volunteers had been instructed to stop using deodorants for 1 week before sample collection.
Growth media
Total viable counts were obtained by plating 0.1 mL aliquots of decimal dilutions of wash fluid on to coryneform agar as described by Leeming et al.,14 with the inclusion of furazolidone 6 mg/L to inhibit the growth of staphylococci.15 The same medium with the addition of either erythromycin 5 mg/L or tetracycline 10 mg/L was used to obtain differential counts of resistant bacteria.11 Plates were examined after 48 h incubation at 37°C followed by 48 h at room temperature to enhance pigmentation and make it easier to discriminate colony types. All distinct colony types from the selective media were Gram-stained to differentiate Gram-positive cocci and Gram-positive pleomorphic rods. Isolates of different morphology were excluded. Following primary isolation, strains were purified and maintained on either Direct Sensitest Agar (DST) for Gram-positive cocci or DST plus 5% sterile defibrinated horse blood (E&O Laboratories, Bonnybridge, UK) and 0.1% Tween-80 (DST-BT) for coryneforms. Strains were stored in liquid nitrogen.
Gram-positive cocci were typed using API ID 32 Staph strips (bioMérieux, Basingstoke, UK) and coryneforms using API Coryne strips (bioMérieux). In addition all isolates were screened for resistance to mupirocin using 5 µg discs.
Antibiotic susceptibility and MICs
The antibiogram of each isolate was determined by disc testing on DST or DST-BT as appropriate using 20 mL volumes of medium in 90 mm diameter Petri dishes. The antibiotics and concentrations used are shown in Table I
. Micrococcus luteus NCTC 9341 and Corynebacterium glutamicum R163 were used as controls. Erythromycin-resistant strains were further tested for erythromycin-inducible resistance to pristinamycin IA, spiramycin and clindamycin. Blunting of the inhibition zone proximal to the erythromycin disc was taken as evidence of inducibility. A minority of isolates of both FURECs and coryneforms which gave small zone diameters (
15 mm) or grew up to penicillin discs were qualitatively screened for ß-lactamase production using the chromogenic substrate, nitrocefin.
|
MICs of MLS and tetracycline antibiotics were determined for selected strains by agar dilution on DST or DST-BT as appropriate. Inocula were prepared by diluting overnight cultures to give a standard inoculum of 104 cfu/ spot delivered by a multipoint inoculator. The MIC of each antibiotic for each organism was recorded after overnight incubation at 37°C as the lowest concentration yielding no growth or a barely visible haze.
Clindamycin and spiramycin were dissolved and diluted in distilled water. Erythromycin and azithromycin were dissolved in ethanol and diluted in distilled water. Pristinamycin was dissolved in dimethylsulphoxide, and furazolidone in acetone.
Media components, chemicals and testing kits
All growth media and nitrocefin were obtained from Oxoid, Basingstoke, UK; all other chemicals and antibiotics were from Sigma–Aldrich, Poole, UK with the exceptions of pristinamycin IA and azithromycin which were gifts from Rhône–Poulenc Rorer (De Vitry-Alfortville, France) and Pfizer (Sandwich, Kent), respectively. Antibiotic-impregnated discs were from Mast Laboratories, Bootle, UK.
Data analysis
Within-group differences in median bacterial counts were determined using Wilcoxon's matched pairs. Between-group differences in the proportion of resistant bacteria were determined using the Mann–Whitney U test. All computations were performed on release 8.0 of Minitab.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Table II
summarizes the bacterial counts for the axilla and fourth toe cleft. Although axillary samples were collected using a quantitative technique, such methodology was not possible for sampling toe clefts. Despite this difference in sample collection, the median number of viable bacteria detected at each site was similar and there were significantly more bacteria resistant to erythromycin than tetracycline at both sites. When the number of resistant bacteria was transformed into a percentage of the total viable count at each site, the same trend was apparent (Table III) in that there were proportionally more bacteria resistant to erythromycin than tetracycline at both sites and the proportion of bacteria resistant to erythromycin was significantly higher in the fourth toe cleft than in the axilla. In total, 160 distinct isolates were obtained, of which 42 were FURECs and 118 were aerobic coryneforms (Table IV). Of these, 157 (98%) were resistant to erythromycin and 66 (41%) to tetracycline. Only seven strains resistant to tetracycline were not also resistant to erythromycin. The resistant isolates belonged to a variety of species (Table IV). CDC group ANF corynebacteria were most numerous and composed 31% of all isolates. The majority (76%) of FURECs were identified as M. luteus.
|
|
|
Before co-resistance of the isolates to other antibiotics could be determined, it was necessary to plot frequency distributions of zone diameters obtained in disc tests for each antibiotic used. These distributions, most of which were bimodal, were used to determine the zone diameter cut-off points to distinguish sensitive and resistant isolates as shown in Table I
. The distribution of zone sizes for penicillin and cloxacillin did not reveal two populations and could not be used to discriminate between sensitivity and resistance. Using the chromogenic cephalosporin, nitrocefin, none of the isolates (corynebacterial or FUREC) with small (<15 mm) zone diameters produced penicillinase. The most commonly encountered co-resistance in FURECs was to the fluoroquinolone, norfloxacin (Table V). The only other resistances detected in more than one Gram-positive coccal isolate were to chloramphenicol and trimethoprim. As far as the coryneforms were concerned, resistance to fusidic acid, chloramphenicol and norfloxacin was the most frequent. A minority of isolates (<1%) were resistant to novobiocin or netilmicin. Nine of 60 tetracycline-resistant coryneforms (15%) were also resistant to minocycline.
|
In order to characterize the tetracycline and erythromycin resistance more fully, MIC determinations were carried out for selected strains. The six strains of tetracycline-resistant FURECs were inhibited by tetracycline 8–64 mg/L but remained sensitive to doxycycline and minocycline (MICs
1 mg/L). The MICs of various tetracyclines for sensitive and resistant coryneforms are shown in Table VI. The strains fell into two classes: those that were resistant to tetracycline only, for which the modal MIC of tetracycline was 16 mg/L, and secondly, nine strains that were cross-resistant to all three tetracyclines for which the modal MIC of tetracycline was much higher, 256 mg/L.
|
The majority of erythromycin-resistant FURECs exhibited an unusual phenotype in that they were sensitive to clindamycin and pristinamycin IA but inducibly resistant, albeit weakly, to spiramycin in disc tests. The MICs of erythromycin for these isolates (designated group A) were low in comparison with those for isolates exhibiting an MLS-resistant phenotype (Table VII
). MICs of azithromycin were four-fold higher than MICs of erythromycin. Without induction, the MICs of spiramycin were similar to those for fully sensitive strains. Of the isolates for which MICs were determined, 23 exhibited the unusual phenotype, two had a constitutive MLS-resistant phenotype and three had phenotypes which could not be classified into either group. Of the 14 isolates for which MICs were not determined, six had the unusual phenotype and seven had either an inducible or a constitutive MLS resistance phenotype in disc tests. The results for one strain suggested that it did not belong to either group. Two of the atypical strains showed reversion of erythromycin resistance in the presence of pristinamycin. There is insufficient evidence to comment reliably on the distribution of resistance phenotypes among the micrococci as most of the isolates were M. luteus and inevitably, therefore, the group A phenotype was predominantly associated with this species (data not shown).
|
In contrast to the FURECs, most of the aerobic coryneforms were MLS resistant (Table VIII
). In total, 89 isolates (75%) demonstrated an MLS-resistant phenotype. In disc tests the majority of isolates showed constitutive expression of erythromycin, spiramycin and clindamycin resistance but erythromycin-inducible expression of pristinamycin IA resistance. In addition, a minority of isolates showed inducibility of clindamycin or spiramycin resistance or were constitutively resistant to pristinamycin. The resistance phenotype was labile in that the same isolates tested on different occasions produced different results with respect to the inducibility of resistance. The MIC results revealed that the degree of resistance to erythromycin was eight to 128 times lower than that to azithromycin and/or clindamycin for 12 of the 32 isolates tested (38%). Although MICs varied between isolates, this variation was not species dependent.
|
Those isolates that did not express MLS resistance could be assigned to one of two phenotypic classes. One group (designated B1) comprised 11 strains that expressed low-level resistance to erythromycin and azithromycin but which were sensitive to clindamycin and/or spiramycin (Table VIII
). In disc tests, the strains were sensitive to pristinamycin IA (Table IX). The second group (designated B2 because of their similarity to B1 strains) comprised 11 strains that demonstrated low-level resistance to erythromycin, spiramycin, clindamycin and, usually but not always, azithromycin (Table VIII). In disc tests, group B2 strains were usually sensitive to pristinamycin IA (Table IX). Neither group demonstrated erythromycin-inducible resistance to any other antibiotic. The phenotypic characteristics of group B coryneforms are summarized in Table IX which also shows how their resistance phenotype differs from that of group A FURECs. Among the coryneforms, the B1 and B2 resistance phenotypes were not associated with particular species (data not shown). Four tetracycline-resistant isolates of Corynebacterium aquaticum selected as controls for MIC determinations with MLS antibiotics demonstrated low levels of resistance to clindamycin and spiramycin but were fully sensitive to erythromycin and azithromycin (Table X).
|
|
One unusual feature of many of the isolates with low-level resistance to one or more MLS antibiotics was that disc zone diameters could not be used reliably as predictors of resistance. This can be seen, for example, by comparing the zone diameters in Table IX
with the MICs in Table VIII (see also Table X). Isolates that required 16 mg/L of antibiotic to inhibit their growth in MIC determinations demonstrated considerable inhibition zones around discs impregnated with lower concentrations of the corresponding antibiotic. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This study has shown that, even in the absence of selective pressure, the axilla and fourth toe cleft of healthy individuals support large populations of antibiotic-resistant aerobic coryneforms and FURECs. Significantly more of these bacteria were resistant to erythromycin than to tetracycline at both body sites. This is consistent with our previous findings for the cutaneous propionibacteria, the other major constituent of the skin flora with G+C-rich DNA,16 but contrasts with the low G+C cutaneous CNS in which prevalence of resistance to tetracycline is greater than that to erythromycin.2 Comparison between the two main groups of cutaneous Gram-positive cocci revealed an even more marked difference between CNS and FURECs. The prevalence of tetracycline resistance in the FURECs was unusually low at 14.3% (6/42) compared with our previous report where 87% of untreated subjects carry tetracycline-resistant CNS.10 It was also significantly lower than the prevalence of tetracycline resistance in aerobic coryneforms at 50.8% (60/118) (Table IV
). More generally, antibiotic-resistant bacteria could be isolated in greater numbers from the fourth toe web than in the axilla. Thus a picture emerges where, in the absence of selective pressure, the skin surface carries a significant reservoir of antibiotic-resistant bacteria and in which the prevalence of individual resistance determinants is dependent on both bacterial type and body site.
Although breakpoint data are not given for cutaneous aerobic coryneforms and FURECs in the recent NCCLS guidelines,17 comparison of data given for other organisms, e.g. Staphylococcus spp., Streptococcus spp. and Enterococcus spp., with the zone sizes used in this study to assign co-resistance shows that the criteria used to define resistance were rigorous (see Table V). Analysis of co-resistance adds an extra dimension of complexity to the picture outlined above. Although all the antibiotic resistance identified could be traced either to agents used topically or to antibiotics that reach intact skin, the resistance profiles were highly dependent upon bacterial type. For example, the most common co-resistance in the cutaneous aerobic coryneforms was to fusidic acid and chloramphenicol, antibiotics that are used topically. The resistance pattern in the FURECs differed from that in both the CNS10 and the cutaneous propionibacteria,16 with a higher prevalence of resistance to chloramphenicol and trimethoprim. Notably, the most common co-resistance in FURECs was to the fluoroquinolone, norfloxacin. In the other groups of skin bacteria, the aerobic Propionibacterium spp. and the CNS, there is a paucity of information on the prevalence of fluoroquinolone resistance in untreated subjects. Further studies are also required to quantify the effects of fluoroquinolone treatment on the acquisition of fluoroquinolone resistance by propionibacteria, aerobic coryneforms and FURECs, although it has been shown that the cutaneous CNS acquire resistance to ciprofloxacin rapidly as a result of excretion of the drug in sweat.18 This route of delivery of quinolones to the skin surface may be significant in that FURECs preferentially colonize moist skin surface sites.9 Unfortunately, it was not possible to discriminate between sensitivity and resistance to penicillins in either cutaneous aerobic coryneforms or FURECs. Measurement of zone sizes of all the sensitive and resistant strains included in this study failed to reveal a bimodal distribution to distinguish the two populations. A lack of breakpoint data for coryneform suseptibility to ß-lactams has also been noted by other workers19 and this, together with the observation that none of the isolates with smaller zones of inhibition produced penicillinase, may point to an alternative resistance mechanism in these isolates. One possibility could be a succession of mutations leading to steadily increasing resistance with no clear distinction between sensitive and resistant strains.
The six tetracycline-resistant FURECs (MICs 8–64 mg/L) were sensitive to the other tetracyclines, doxycycline and minocycline. In contrast, the aerobic cutaneous coryneforms fell into two groups: those that were resistant to tetracycline alone (modal MIC 16 mg/L), and those (comprising nine strains) that were resistant to all three tetracyclines, for which the modal MIC of tetracycline was much higher (256 mg/L). Resistance to tetracycline alone would suggest carriage of either determinants encoding tetracycline efflux systems of the major facilitator type20 or, as recently reported in Corynebacterium striatum, an ABC transporter mediating drug efflux encoded by tetAB.21 Alternatively, coryneform resistance to all three tetracyclines appears to mirror the phenotype identified in staphylococci mediated via ribosomal protection in which TetO/TetM encodes a homologue of elongation factor G.22
Analysis of the MLS resistance phenotypes suggested the presence of several different antibiotic resistance mechanisms, some of which may be novel. Most cutaneous aerobic coryneforms and a number of FURECs showed classic MLS resistance, easily explained either by expression of inducible/constitutive erythromycin ribosomal methylase (erm) genes23 or by a mutation at base 2058 in the 23S ribosomal RNA.24 However, some isolates exhibited unusual phenotypes which could not be placed in these categories and are not compatible with other known erythromycin resistance mechanisms. Examples of these included (i) the group A FURECs which were erythromycin resistant, sensitive to clindamycin and pristinamycin IA and weakly inducibly resistant to spiramycin (the majority of erythromycin-resistant FURECs fell into this group); (ii) the group B1 aerobic cutaneous coryneforms which expressed low-level resistance to erythromycin and azithromycin, but were sensitive to pristinamycin IA, clindamycin and/or spiramycin; and (iii) the group B2 aerobic cutaneous coryneforms which expressed low-level resistance to erythromycin, spiramycin and clindamycin and usually to azithromycin and pristinamycin IA. Resistance to the MLS antibiotics in the latter two groups was non-inducible and, because of the low level of some of the resistance identified, it was found that MIC determinations provided a more reliable indicator of resistance phenotype than the use of disc tests. It is likely that genetic analysis of these strains will identify previously unreported antibiotic resistance genes and that this approach would be facilitated by the development of genetic technology using C. glutamicum.21
There was strong evidence for a link between tetracycline and erythromycin resistance in the aerobic cutaneous coryneforms. In both axilla and toe cleft the vast majority of these bacteria were sensitive to erythromycin. Erythromycin-resistant strains represented <10% of the median count in the toe cleft and <1% in the axilla (Table II). In the absence of linkage, a majority of tetracycline-resistant strains would, therefore, be expected to be erythromycin sensitive. Only seven isolates were tetracycline resistant without also being erythromycin resistant (Table IV). This undoubtedly reflects both antibiotic usage as well as genetic linkage of the resistance determinants themselves. For example, in C. striatum, the plasmid pTP10, which carries the tetracycline resistance determinant tetAB, also carries genes encoding resistance to kanamycin, erythromycin and chloramphenicol.21 There are clearly many examples of linked antibiotic resistance genes in cutaneous CNS.25 Interestingly, mobile genetic elements have yet to be identified in the cutaneous propionibacteria and resistance to both erythromycin and tetracycline has been shown to be caused by mutations in genes encoding ribosomal RNA.24,26
In addition to the short-term use of antibiotics for the treatment of acute infections, the long-term use of antibiotics in dermatology, especially for the treatment of acne vulgaris, represents a significant source of selective pressure for the evolution of antibiotic-resistant organisms.16 The results of this study show that the axilla and toe cleft of untreated subjects harbour a significant reservoir of G+C-rich antibiotic-resistant bacteria. In addition many of these express novel antibiotic resistance phenotypes. It is certain that during antibiotic treatment the proportion of antibiotic-resistant cutaneous aerobic coryneforms and FURECs will increase significantly. Clearly they share a common niche with the G+C-rich cutaneous propionibacteria and the A+T-rich CNS. Thus it is likely that, under conditions of prolonged selection, gene exchange between these groups of bacteria will provide a significant impetus for the further evolution and spread of antibiotic resistance among cutaneous bacteria.
|
Acknowledgments |
|---|
Some of the findings contained in this manuscript were presented as a poster at the Ninety-Seventh General Meeting of the American Society for Microbiology, 1997, Miami, USA. This work was supported by the Leeds Foundation of Dermatology and a Research Studentship to A. H. R. from the Government of Pakistan.
|
Notes |
|---|
* Corresponding author. Tel: +44-113-2335581; Fax: +44-113-2335638; E-mail: e.a.eady{at}leeds.ac.uk
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
1 . Kloos, W. E., Tornabene, T. G. & Schliefer, K. H. (1974). Isolation and characterisation of micrococci from human skin, including two new species: Micrococcus lylae and Micrococcus kristinae. International Journal of Systematic Bacteriology 24, 79–101.
2 . Larson, E. L., McGinley, K. J., Leyden, J. J., Cooley, M. E. & Talbot, G. H. (1986). Skin colonisation with antibiotic-resistant (JK group) and antibiotic-sensitive lipophilic diphtheroids in hospitalised and normal adults. Journal of Infectious Diseases 153, 701–6.[ISI][Medline]
3 . Leyden, J. J. & McGinley, K. J. (1993). Coryneform bacteria. In The Skin Microflora and Microbial Skin Disease, (Noble, W. C., Ed.), pp. 102–17. Cambridge University Press.
4 . Roth, R. R. & James, W. D. (1989). Microbiology of the skin: resident flora, ecology, infection. Journal of the American Academy of Dermatology 20, 367–90.[ISI][Medline]
5
.
Stackebrandt, E., Koch, C., Gvozdiak, O. & Schumann, P. (1995). Taxonomic dissection of the genus Micrococcus: Kocuria gen. nov., Nesterenkonia gen. nov., Kytococcus gen. nov., Dermacoccus gen. nov., and Micrococcus Cohn 1872 gen. emend. International Journal of Systematic Bacteriology 45, 682–92.
6
.
Roberts, M. C., Leonard, R. B., Briselden, A., Schoenknecht, F. D. & Coyle, M. B. (1992). Characterisation of antibiotic-resistant Corynebacterium striatum strains. Journal of Antimicrobial Chemotherapy 30, 463–74.
7 . Soriano, F., Zapardiel J. & Nieto, E. (1995). Antimicrobial susceptibilities of Corynebacterium species and other non-spore-forming gram-positive bacilli to 18 antimicrobial agents. Antimicrobial Agents and Chemotherapy 39, 208–14.[Abstract]
8 . von Eiff, C., Herrmann, M. & Peters, G. (1995). Antimicrobial susceptibilities of Stomatococcus mucilaginosus and of Micrococcus spp. Antimicrobial Agents and Chemotherapy 39, 268–70.[Abstract]
9 . Noble, W. C. (1981). In Microbiology of Human Skin, (Rook, A., Ed.). Lloyd Luke, London.
10
.
Cove, J. H., Eady, E. A. & Cunliffe, W. J. (1990). Skin carriage of antibiotic-resistant coagulase-negative staphylococci in untreated subjects. Journal of Antimicrobial Chemotherapy 25, 459–69.
11 . Eady, E. A., Cove, J. H., Blake, J., Holland, K. T. & Cunliffe, W. J. (1988). Recalcitrant acne vulgaris. Clinical, biochemical and microbiological investigation of patients not responding to antibiotic treatment. British Journal of Dermatology 118, 415–23.[ISI][Medline]
12 . Leyden, J. J., McGinley, K. J., Cavalieri, S., Webster, G. F., Mills, O. H. & Kligman, A. M. (1983). Propionibacterium acnes resistance to antibiotics in acne patients. Journal of the American Academy of Dermatology 8, 41–5.[ISI][Medline]
13 . Williamson, P. & Kligman, A. M. (1965). A new method for the quantitative investigation of cutaneous bacteria. Journal of Investigative Dermatology 45, 498–503.[ISI][Medline]
14 . Leeming, J. P., Holland, K. T. & Cunliffe, W. J. (1984). The microbial ecology of pilosebaceous units isolated from human skin. Journal of General Microbiology 130, 803–7.[Medline]
15 . Marshall, J. (1988). The microbial ecology of the human foot. Ph.D. Thesis, Department of Microbiology, University of Leeds, UK.
16 . Eady, E. A. (1998). Bacterial resistance in acne. Dermatology 196, 59–66.[ISI][Medline]
17 . National Committee for Clinical Laboratory Standards. (1997). Performance Standards for Antimicrobial Susceptibility Testing—Eighth Informational Supplement: Approved Standard M100-S8. NCCLS, Wayne, PA.
18 . Høiby, N., Jarlov, J. O., Kemp, M., Tvede, M., Bangsborg, J. M., Kjerulf, A. et al.(1997). Excretion of ciprofloxacin in sweat and multiresistant Staphylococcus epidermidis. Lancet 349, 167–9.[ISI][Medline]
19 . Funke, G., Pünter, V. & von Graevenitz, A. (1996). Antimicrobial susceptibility patterns of some recently established coryneform bacteria. Antimicrobial Agents and Chemotherapy 40, 2874–8.[Abstract]
20
.
Levy, S. B. (1992). Active efflux mechanisms for antimicrobial resistance. Antimicrobial Agents and Chemotherapy 36, 695–703.
21 . Tauch, A., Krieft, S., Pühler, A. & Kalinowski, J. (1999). The tetAB genes of the Corynebacterium striatum R-plasmid pTP10 encode an ABC transporter and confer tetracycline, oxytetracycline and oxacillin resistance in Corynebacterium glutamicum. FEMS Microbiology Letters 173, 203–9.[ISI][Medline]
22
.
Burdett, V. (1996). Tet(M)-promoted release of tetracycline from ribosomes is GTP dependent. Journal of Bacteriology 178, 3246–51.
23 . Weisblum, B. (1985). Inducible resistance to macrolides, lincosamides and streptogramin type B antibiotics: the resistance phenotype, its biological diversity, and structural elements that regulate expression—a review. Journal of Antimicrobial Chemotherapy 16, Suppl. A, 63–90.
24 . Ross, J. I., Eady, E. A., Cove, J. H., Jones, C. E., Ratyal, A. H., Miller, Y. W. et al. (1997). Clinical resistance to erythromycin and clindamycin in cutaneous propionibacteria isolated from acne patients is associated with mutations in 23S rRNA. Antimicrobial Agents and Chemotherapy 41, 1162–5.[Abstract]
25 . Skurray, R. A. & Firth, N. (1997). Molecular evolution of multiply-resistant staphylococci. In Antibiotic Resistance: Origins, Evolution, Selection and Spread, (Chadwick, D. J. & Goode, J., Eds). Ciba Foundation Symposium Vol. 207, pp. 167–83. John Wiley, Chichester, UK.
26
.
Ross, J. I., Eady, E. A., Cove, J. H. & Cunliffe, W. J. (1998). 16S rRNA mutation associated with tetracycline resistance in a gram-positive bacterium. Antimicrobial Agents and Chemotherapy 42, 1702–5.
Received 3 August 1999; returned 20 December 1999; revised 31 January 2000; accepted 24 February 2000
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  W. Liebl, W. E. Kloos, and W. Ludwig Plasmid-borne macrolide resistance in Micrococcus luteus Microbiology, August 1, 2002; 148(8): 2479 - 2487. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||