the action of other antibiotics, and a consistent pattern of antagonism was observed to each antibiotic class in all of the strains in which this could be tested, ...
Vol. 98, No. 2 Printed in U.S.A.
JOURNAL OF BACTERIOLOGY, May 1969, p. 447-452 Copyright @ 1969 American Society for Microbiology
Erythromycin-inducible Resistance in Staphylococcus aureus: Survey of Antibiotic Classes Involved BERNARD WEISBLUM AND VERONIKA DEMOHN Department of Pharmacology, University of Wisconsin Medical School, Madison, Wisconsin 53706
Received for publication 2 December 1969
Certain erythromycin-resistant strains of Staphylococcus
aureus
remain sensitive
to other macrolide antibiotics. If these strains are exposed to low levels of erythromycin, resistance to other antibiotics is induced. The antibiotics to which resistance
is induced by erythromycin include: other macrolides as well as lincosaminide, streptogramin (group B) antibiotics but not chloramphenicol, amicetin, streptogramin (group A) antibiotics, tetracyclines, and aminoglycosides. Hence erythromycin induces resistance exclusively towards inhibitors of 50S ribosomal subunit function and, thus far, only with respect to three of six known classes of inhibitors which act on this subunit. In the four strains tested, erythromycin did not induce resistance to pactamycin or bottromycin, to fusidic acid (which inhibits a function involving both subunits), or to other antibiotics which do not inhibit ribosomal function. Thus, by inducing resistance erythromycin could antagonize the action of other antibiotics, and a consistent pattern of antagonism was observed to each antibiotic class in all of the strains in which this could be tested, as well as to other antibiotic members of the same chemical class in each bacterial strain.
RESULTS A class of erythromycin-resistant mutants of Staphylococcus aureus studied by Chabbert (3) Studies by Davies (6), Vazquez (14, 15), and was unexpectedly found to remain sensitive to Taubman et al. (13), as well as studies in our other macrolide antibiotics, e.g., spiramycin (3) or laboratory (4, 5; F. N. Chang, Ph.D. Thesis, oleandomycin (7). Garrod (7) termed this pattern Univ. of Wisconsin, Madison, 1968), have made it of response "dissociated resistance." Further possible to classify several groups of antibiotics as studies on mutants of S. aureus showing disso- inhibitors of ribosomal function and to localize ciated resistance have yielded the surprising result their site of action on either the 30S or the 50S that, in the presence of erythromycin, these strains ribosomal subunit. Table 2 is a partial listing of appeared to be resistant to spiramycin (3), linco- these antibiotics, and contains the major cateI The present mycin (1, 8), and tristinamycin (2). used in this study. and examples specific gories studies were undertaken in an attempt to define can be demonantagonizability Erythromycin action is anwhose the classes of antibiotics use of antibiotic sensitivity discs. the strated by tagonizable by erythromycin. Two discs (one containing erythromycin and the other a test antibiotic) are placed within about 1 MATERIALS AND METHODS cm of each other on a plate seeded with the approAntibiotic sensitivity assays were performed by priate test organism. For those strains in which pouring bacteria in soft agar over plates prepared antagonism can be demonstrated, a decreased from a medium containing (in grams per liter): pep- radius of the inhibition zone is seen on the side of tone, 5; yeast extract, 5; K2HPO4, 1; glucose, 2; agar, the test disc proximal to the erythromycin disc. 20. The soft agar had the same composition except that Members of the macrolide and lincosaminide 5 g of agar was used instead of 20. After the soft agar and layer hardened, the various sensitivity discs were classes were first tested; chloramphenicol of inhibitors the which amicetin likewise, are, 18 hr for at then incubated were The plates applied. 50S subunit were tested next. Four different 37 C. The S. aureus strains used are listed in Table 1. The erythromycin-resistant strains were used. One antibiotic discs together with the respective abbrevia- strain was selected in the laboratory; the other tions and the level of antibiotic contained are listed in three strains were isolated independently from patients and were found to be erythromycinTable 2. 447
448
WEISBLUM AND DEMOHN
J. BACrERIOL.
TABLE 1. Strains of S. aureus used Strains
Copenhagen
Copenhagen ERY-R 1206
CR-27(C-)
CR-51
Obtained from
Comments
ERY1|
SPI
S
S
S
S
S
Selected from above strain by passage in R ERY-containing (20 pg/ml) medium V. Grimstad Isolated from patient in University Hos- R pitals, Madison, Wis.; (disocciated) ERY-resistant. M. W. Finland Isolated from patient in Boston, Mass.; R (dissociated) ERY-resistant, parent strain CR-27 originally CAM-resistant, other properties of these strains are described by Sabath et al. (12) M. W. Finland Isolated from patient in Boston, Mass.; R (dissociated) ERY-resistant; other properties of this strain are described by Sabath et al. (12)
R
S
S
S
S
S
S
R
S
S
S
R
S
S
R
R
J. L. Strom-
inger
LNC CAM PEN
a Abbreviations: ERY, erythromycin; SPI, spiramycin; LNC, lincomycin; CAM, chloramphenicol; PEN, penicillin; S, sensitive; R, resistant. TABLE 2. Summary of antibiotics which are inhibitors of ribosomal function" Abbreviation
AMI
Full name
Amicetin
BOT CAM CAR CLN
Bottromycin Chloramphenicol Carbomycin 7-Chloro-7-
CTC ERY FUS KAN LNC NEO NID PAC SGM-A SPA SPI STR TAO
Chlortetracycline Erythromycin Fusidic acid Kanamycin Lincomycin Neomycin Niddamycin Pactamycin Streptogramin A Sparsomycin
TET VIR
VRN-A VRN-Ba
Amt (jpg/disc)
20 5 10 10 2.5
Class
Subunit specificity Erythromycin
antagonizable
Aminoacylamino
50S
nucleoside -b
-^
Macrolide Lincosaminide
ndc S0S 50S 50S
deoxylincomycin
Spiramycin Streptomycin Triacetyl oleandomycin Tetracycline Viridogrisein Vernamycin A Vernamycin Ba
20 10 0.35 30 2 20 10
Tetracycline Macrolide b
Aminoglycoside Lincosaminide Aminoglycoside Macrolide -b
30S 50S 30S and 50S 30S 50S 30S S0S nd
+
+
+
5 S 20 10 25 15
Macrolide Macrolide
50S SOS 50S 305 50S
30 5 5 40
Tetracycline Streptogramin B Streptogramin A Streptogramin B
30S 50S 50S 50S
Streptogramin A -b
Aminoglycoside
+ + +
+
+
+
aAt the concentrations listed (pg/disc) the following antibiotics, which are inhibitors of nonribosomal function were not erythromycin-antagonizable: bacitracin (BAC), 10; cephalosporin (CPH), 15; cycloserine (CSR), 20; nitrofurantoin (NFR), S0; novobiocin (NOV), 15; rifamycin SV (RIF), 0.5; sulfisoxazole (SXL), 30; vancomycin (VAN), 30. b Single member of its chemical class tested. a Not determined.
VOL. 98, 1969
449
ERYTHROMYCIN-INDUCIBLE RESISTANCE
resistant. Erythromycin antagonizability was noted only in the erythromycin-resistant S. aureus strains isolated from patients but not in the resistant strain selected in the laboratory (Fig. 1). Erythromycin antagonized all of the microlides) spiramycin, carbomycin, triacetyl oleandomycin, and niddamycin) and lincosaminides (lincomycin and 7-chloro-7-deoxylincomycin) tested, but not chloramphenicol or amicetin. Antagonism of spiramycin in strain CR27(C-) can be seen more clearly in a later figure. In subsequent experiments (data not shown), it was found that erythromycin could also antagonize the action of the macrolide tylosin. It should also be noted that, although inhibition of strains 1206, CR-51, and CR-27(C-) by the macrolides carbomycin, triacetyl oleandomycin, and niddamycin is erythromycin-antagonizable, these strains are as sensitive as the Copenhagen wild-type strain (Copen+) to carbomycin, triacetyl oleandomycin, and niddamycin in the absence of erythromycin. Moreover, in the process of acquiring resistance to erythromycin, the Copen+ strain appears to have acquired coresistance to triacetyl oleandomycin. Streptogramin antibiotics which were tested included streptogramin A and vernamycin A which belong to the A group and viridogrisein and vernamycin Ba which belong to the B group. COPEN+
These two groups are also inhibitors of the SOS subunit (14, 15). Erythromycin antagonizability was observed only with respect to the streptogramin B-type antibiotics (Fig. 2). This finding is consistent with the recent observations of Bourse and Monnier (2), who reported that pristinamycin I (a streptogramin B type) was antagonizable by erythromycin but pristinamycin II (a streptogramin A type) was not. Spiramycin and sparsomycin were also tested. A sufficiently wide zone of inhibition by sparsomycin could not be obtained under the conditions used. However, in strains 1206 and CR-51 there was no indication that the narrow inhibitory zone around the sparsomycin disc was in any way influenced by erythromycin. Antagonism of spiramycin was clearly seen in strain CR-27(C-). Since all of the antibiotics studied above are inhibitors of the 50S subunit, we next inquired whether inhibition by antibiotics which act at other sites could also be antagonized by erythromycin. Therefore, we tested a group of antibiotics whose site of action had been (i) localized to the 30S subunit, namely, tetracycline, chlortetracycline, streptomycin, neomycin, and kanamycin; (ii) tentatively identified as the 30S subunit, namely, pactamycin, bottromycin; or (iii) associated with a function which requires both 305 and 50S subunits, namely, fusidic acid. The
COPEN ERY-R
SPI CAR TAO IJID
1206
CR-51
ERY ERY ERY ERY
LNC CLN CAM AMI
CR-27(CV)
FIG. 1. Effect of erythromycin on the sensitivity of five S. aureus strains to antibiotics. The antibiotics used and their disposition on each plate correspond to the pattern given (upper right). Abbreviations and antibiotics which they represent are summarized in Table 2. The antibiotics surveyed are known to be inhibitors of 50S ribosomal subunit function.
450
WEISBLUM AND DEMOHN
J. BAcreRioL.
basis for these assignments of subunit specificity Several antibiotics which do not exert their pnhas recently been reviewed (17). Antagonism was mary effect on ribosomal function were also tested not observed to any of these antibiotics in any of as above. Antagonism by erythromycin could not the strains tested (Fig. 3). be demonstrated (data not shown) to cephaloCOPEN+
1206
COPEN FPY-_
[051
-C -27
Fia. 2. Effect of erythromycin on the sensitivity of five S. aureus strains to antibiotics. The antibiotics used and their disposition on each plate correspond to the pattern given (upper right). Abbreviations and the antibiotics which they represent are summarized in Table 2. The antibiotics surveyed include two members each from the streptogramin A and streptogramin B families, respectively, as well as a repeat test of spiramycin. COPENI
1206
COPEN ERY-R
CR-51
CR-27(C )
FioI. 3. Effect of erythromycin on the sensitivity offive S. aureus strains to antibiotics. The antibiotics used and
their disposition on each plate correspond to the pattern given (upper right). Abbreviations and the antibiotics which they represent are summarized in Table 2. The antibiotics surveyed include known inhibitors of the 30S ribosomal subunit (aminoglycosides and tetracyclines) as well as fusidic acid which inhibits a function requiring both ribosomal subunits. See reference 17 regarding pactamycin and bottromycin.
VOL. 98, 1969
ERYTHROMYCIN-IINDUCIBLE RESISTANCE
sporin, bacitracin, vancomycin, sulfisoxazole, nitrofurantoin, novobiocin, cycloserine, and rifamycin. Strains 1206, CR-51, and CR-27(C-) were resistant to penicillin and ampicillin. DISCUSSION The present experiments were undertaken in an attempt to gain insight into erythromycin antagonizability of antibiotic action by defining the classes of antibiotics which are involved. Indeed, we noted that only inhibitors of the 50S ribosomal subunit were involved and that they belong to only three of several different classes which act on this subunit, each being structurally different. Several groups of investigators (9-11, 16) have examined the effects of erythromycin in strains which display dissociated resistance. One interesting aspect of the behavior of these strains, not directly observable in the sensitivity-disc experiments, is the fact that erythromycin at subinhibitory concentrations (0.001 to 0.1 ,ug/ml) can "induce" resistance to itself (16). Garrod (7) attributed this to the selection of a small fraction of resistant cells (
