Vol. 35, No. 10

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Oct. 1991, p. 1989-1996

0066-4804/91/101989-08$02.00/0 Copyright ©) 1991, American Society for Microbiology

Characterization of Erythromycin Resistance in Campylobacterjejuni and Campylobacter coli WILLIAM YAN1t AND DIANE E. TAYLOR' 2* Departments of Microbiology2 and Medical Microbiology and Infectious University of Alberta, Edmonton, Alberta, Canada T6G 2H7

Diseases,'

Received 8 March 1991/Accepted 18 July 1991

The mechanism of resistance to erythromycin, the drug of choice in the treatment of campylobacter gastroenteritis, was investigated. Erythromycin resistance (MICs, >1,024 ,ug/ml) in three clinical isolates of Campylobacterjejuni and one C. coli isolate was determined to be constitutive and chromosomally mediated. In vivo protein synthesis in erythromycin-susceptible C. jejuni and C. coli strains was completely inhibited by low levels of erythromycin (5 ,ug/ml), whereas a high concentration of the antibiotic (100 ,ug/ml) had no effect on protein synthesis in erythromycin-resistant strains. Biological assays showed that extracellular degradation of erythromycin was not responsible .for erythromycin resistance in strains of Campylobacter species. The rates and amounts of uptake of ]'4Clerythromycin by resistant and susceptible campylobacter cells were determined to be similar. Binding assays with purified campylobacter 70S ribosomes as well as 50S ribosomal subunits showed that those from erythromycin-resistant strains bound much less ['4C]erythromycin than did those from susceptible strains. Genomic DNA from C. coli UA585 was used to transform erythromycin resistance to C. coli UA417. The erythromycin resistance marker was associated with a 240-kb SmaI fragment of the C. coli UA585 genome. Our results rule out erythromycin inactivation or efflux and are not consistent with the production of an RNA methylase, although they are consistent with a mutational mechanism of resistance due to a change in a ribosomal protein gene. This study constitutes a detailed biochemical and genetic characterization of erythromycin resistance in Campylobacter species.

Infection with Campylobacter species has emerged as one of the leading causes of bacterial diarrhea on a worldwide basis. Campylobacter jejuni causes 95 to 98% of cases of campylobacteriosis, and the related species C. coli accounts for the remaining cases (24). Erythromycin is the drug of choice for the treatment of serious cases of campylobacter gastroenteritis. The incidence of isolation of erythromycinresistant campylobacter strains ranges from less than 1% in Canada and the United Kingdom (9, 22, 41) to 8.4% in Belgium and 10% in Sweden (46, 47). In the United States, Wang et al. reported that 3% of campylobacter strains from human sources were resistant to erythromycin (48). In the same study, a much greater incidence of erythromycin resistance (70%) was noted among hog isolates, most of which were found to be C. coli (48). Other workers have also noted that the incidence of erythromycin resistance is much higher in C. coli than in C. jejuni (10, 23, 37, 44, 48). Although erythromycin resistance is believed to be chromosomally mediated in Campylobacter species (42), little is known about the mechanism involved. The goal of the present study was to determine the genetic and biochemical basis of erythromycin resistance in C. jejuni and C. coli. MATERIALS AND METHODS Bacterial strains and culture growth conditions. Erythromycin-resistant (Eryr) and erythromycin-susceptible (Erys) strains of C. jejuni and C. coli used in this study are listed in Table 1. All strains of Campylobacter species were grown in Mueller-Hinton (MH) broth or on MH agar at 37°C under 7% CO2. Strains used in serotyping and biotyping experiments * Corresponding author. t Present address: Department of Molecular Biology and Microbiology, Tufts University, Boston, MA 02111.

were grown on blood agar plates at 42°C in the presence of 7% CO2. Escherichia coli BM2571 harboring plasmid pIP1527 and Staphylococcus aureus ATCC 25923, used in the detection of erythromycin-modifying activity, were cultured on MH agar at 37°C. Antibiotics and chemicals. Erythromycin was purchased from Ayerst, Mckenna & Harrison Inc., whereas oleandomycin and lincomycin as well as antibiotic discs containing 15 jig of erythromycin per ml were purchased from Sigma Chemical Co., St. Louis, Mo. [14C]erythromycin was purchased from Du Pont, NEN Research Products, Boston, Mass. An erythromycin concentration of 32 jig/ml was used for the maintenance and selection of Eryr C. jejuni and C. coli strains. Determination of MICs. MICs of erythromycin, oleandomycin, and lincomycin against C. jejuni and C. coli strains were determined by a previously described method (40). Serotyping and biotyping of C. jejuni and C. coli strains. Serotyping was carried out by using the serotyping scheme of Lior et al. (31). Biotyping of the strains was carried out by the rapid hippurate hydrolysis test, the rapid H2S test, and the DNA hydrolysis test (30). Both serotyping and biotyping were performed at the Laboratory Centre for Disease Control, Ottawa, Ontario, Canada.Detection of erythromycin-modifying activity. Tests for the detection of extracellular erythromycin-modifying enzymes were carried out as described previously (1). S. aureus ATCC 25923 was used as the indicator strain, and E. coli BM2571(pIP1527), which contains the ereA gene encoding erythromycin esterase, was used as a positive control. Antibiotic filter discs containing 15 ,ug of erythromycin were used in the assays. Determination of inducibility of erythromycin resistance in C. jejuni and C. coli. Overnight cultures (1.0 ml) of Eryr C. jejuni and C. coli strains were used to inoculate 100 ml of 1989

1990

ANTIMICROB. AGENTS CHEMOTHER.

YAN AND TAYLOR TABLE 1. Properties of bacterial strains used Strain

Original strain designation

C. jejuni UA67 C. coli UA417 C. coli UA585 C. coli UA586 C. jejuni UA695 C. jejuni UA697 C. jejuni UA709 C. jejuni UA736 E. coli BM2571 S. aureus ATCC 25923

SD2 C2633 118114r 118114S E7513 631 205224r 205224S NAc

a

b c

Plasmid

Resistance'

Nal Nal Ery Tcr

Ery, Tc Ery Ery

pIP1527

Ery, Cm, Tc

NAC

Sourceb Canada (M. A. Karmali) Canada (H. Lior) Wales (D. Ribeiro) Wales (D. Ribeiro) Canada (L. Mueller) England (I. Phillips) The Netherlands (H. Endtz) The Netherlands (H. Endtz) France (P. Courvalin) American Type Culture Collection

Cm, chloramphenicol; Ery, erythromycin; Nal, nalidixic acid; Tc, tetracycline. Country of origin (name of supplier). NA, not applicable.

fresh MH broth either with or without subinhibitory levels of erythromycin (32 ,ug/ml), lincomycin (13.6 U/ml), or oleandomycin (32 ,ug/ml). After incubation at 37°C under 7% C02, 100 mg of erythromycin was added to each culture. A culture to which no antibiotics were added was also included in each experiment as a control. Cultures were incubated for 72 h, and A6. readings were taken at various times. In vivo uptake and accumulation of ['4C]erythromycin. Cultures of Eryr and Erys C. jejuni and C. coli strains were grown overnight in MH broth, centrifuged, and resuspended to an A6. of 0.4 in MH broth. Cells were preincubated at 37°C for 15 min, and -1.5 ,Ci of [14C]erythromycin was added (zero time point). Cells were further incubated at 37°C for 3 h, and 0.5-ml samples were removed at various times. Samples were applied to Millipore filters (HAWP; 0.45-,umpore size; prewashed with 2.0 mg of unlabeled erythromycin per ml to reduce background nonspecific binding of [14C]erythromycin) under vacuum, washed three times with ice-cold 10 mM Tris (pH 7.6), and dried at 65°C, and the radioactivity was measured with a Beckman LS6800 scintillation counter. Effect of erythromycin on in vivo protein synthesis. Fresh overnight cultures of C. jejuni UA695, UA697, UA709, and UA736 as well as C. coli UA585 and UA586 were harvested, washed, and suspended in fresh MH broth to a cell density of 0.25 at A600. Portions (1.0 ml) of the cell suspensions were incubated at 37°C for 4 h with [35S]methionine (20 ,uCi/ml) and various concentrations of erythromycin (0, 5.0, 30.0, and 100.0 ,ug/ml). The radiolabeled cells were washed with ice-cold 10 mM Tris (pH 7.6) and lysed in cracking buffer (14) by being boiled for 5 min. The polypeptides were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (27) and autoradiographed as described previously (8). The extent of [35S]methionine incorporation into polypeptides was taken as a measure of in vivo protein synthesis. Isolation and purification of 70S ribosomes and 50S and 30S ribosomal subunits from Campylobacter species. Ribosomes from C. jejuni and C. coli were isolated by a modified version of a previously described method (21). Large-scale preparations of overnight C. jejuni and C. coli cultures (five 750-ml cultures) were used to harvest -3.0 g (wet weight) of cells. Cells were subjected to grinding for 20 min at 0°C with two times the wet weight of Alumina (Sigma). After removal of cell debris by low-speed centrifugation, the supernatant containing the 70S ribosomes was introduced at 25 ml/h into a Sephacryl S-200 column (3 by 50 cm). After peak A260

fractions were pooled, 70S ribosomes were precipitated with 100 mg of polyethylene glycol 8000 per ml. The ribosome pellet was resuspended and applied to a Sephadex G-75 column to remove polyethylene glycol 8000 from the ribosomal preparation. The A260 values of the peak fractions were determined, and 200-,ul aliquots of the purified 70S ribosomes were stored at -20°C. To prepare 50S and 30S ribosomal subunits, we resuspended 70S ribosomal preparations in a buffer containing low concentration of MgCl2. The dissociated ribosomes were subjected to sucrose gradient centrifugation. The A260 values of gradient fractions were determined, and the peak fractions corresponding to 50S and 30S ribosomal subunits were pooled and stored at -20°C. "14C]erythromycin binding to campylobacter 70S ribosomes and 50S and 30S ribosomal subunits. The 70S ribosomes as well as the 50S and 30S ribosomal subunits (130 pmol) from Eryr and Erys strains were incubated in 70 ,ul of reaction buffer (50 mM Tris hydrochloride [pH 7.6], 50 mM KCl, 10 mM MgCl2, 6 mM P-mercaptoethanol, 0.1 mM EDTA) containing various concentrations (20 to 100 1xM) of ['4C] erythromycin for 30 min at 370C. The ribosomes were applied to prewashed Millipore filters (HAWP; 0.22-p.m-pore size) and washed three times (5 ml each time) with ice-cold reaction buffer. The filters were dried for 30 min at 650C, and the radioactivity associated with the filters was determined. Reaction mixtures without ribosomes were treated identically, collected on filters, and washed, and the radioactivity associated with the filters was determined for each concentration of [14C]erythromycin used to obtain the background count. Chromosomal DNA isolation. Preparations of chromosomal DNA from Campylobacter species were isolated from cells harvested from 20 MH agar plates by the method of Marmur (32). Restriction endonuclease digestion and pulsed-field gel electrophoresis analysis of campylobacter genomic DNA. C. jejuni and C. coli genomic DNAs were digested with the restriction endonucleases SmaI and Sall and subjected to pulsed-field gel electrophoresis as previously described (11). Natural transformation. C. coli transformation was performed as previously described (49). Chromosomal DNA from C. coli UA585 (0.2 p.g) was used in MH agar surface transformation experiments. Transformation studies with UA585 genomic DNA fragments extracted from low-melting-point agarose were performed by the biphasic transfor-

VOL. 35, 1991

ERYTHROMYCIN RESISTANCE IN C. JEJUNI AND C. COLI

1991

TABLE 2. Serotypes and biotypes of and MICs for strains of Campylobacter species MIC of: Strain

C. jejuni UA67 C. coli UA417 C. coli UA585 C. coli UA586 C. jejuni UA695 C. jejuni UA697 C. jejuni UA709 C. jejuni UA736

Sero38 29 8 8 2 NT 53 1

BioIII II II II II I II

II

Erythro-

Oleando-

mycin

(,ug/ml)

mycin (U/ml)

Lincomycin (,g/ml)

2 2 >1,024 2 >1,024 >1,024 >1,024 4

13.6 13.6 217.6 13.6 217.6 217.6 217.6 13.6

8 4 1,024 4 1,024 2,048 1,024 8

Determined as described by Lior et al. (31). NT, nontypeable. b Determined as described by Lior (30).

a

mation procedure (49). Eryr transformants were selected on MH agar plates containing 32 jig of erythromycin per ml. RESULTS

Serotyping, biotyping, and plasmid content of and MICs for C. jejuni and C. coli strains. The serotypes and biotypes of all strains of Campylobacter species used in this study are listed in Table 2. Eryr strains from several different geographic sources were investigated. These included C. coli UA585 (Eryr) and C. coli UA586 (Erys), which are isogenic strains, the latter being a laboratory-isolated derivative of the former. As expected, C. coli UA585 and UA586 shared the same serotype and biotype. On the other hand, C. jejuni UA709 and UA736, which were initially considered to be an isogenic pair isolated from a single patient (17), were found to have similar biotypes but different serotypes (51, 52). The serotype of C. jejuni UA697 could not be determined because of the inability of this strain to produce flagella. Despite culturing in MH broth and on semisolid agar, motility could not be restored. All Eryr strains tested were found to be plasmid free, except for C. jejuni UA695, which contained a 45-kb plasmid encoding tetracycline resistance. The MICs of the three macrolide-lincosamide-streptogramin B (MLS) antibiotics (erythromycin, lincomycin, and oleandomycin) for the different Campylobacter strains are also shown in Table 2. All the Eryr strains (C. jejuni UA695, UA697, and UA709 and C. coli UA585) were found to be resistant to high levels of the other antibiotics. Absence of erythromycin-modifying activity. Biological assays for the detection of erythromycin-modifying activity were performed on C. jejuni UA67 and UA695, C. coli UA585, and E. coli BM2571, and the results are shown in Fig. 1. Erys C. jejuni UA67 (Fig. 1A) did not grow beyond or distort the zone of inhibition on the confluent lawn of the S. aureus indicator culture. A clearly observable arrow-shaped distortion in the zone of inhibition was produced by the growth of E. coli BM2571 (Fig. 1B). Although the growth of Eryr C. jejuni UA695 and C. coli UA585 (Fig. 1C and D, respectively) was observed within the zone of inhibition, distortion of the shape of the zone was not detected. Similar results were obtained for C. jejuni UA697 and UA709 (data not shown). Therefore, these results demonstrate that Eryr C. jejuni and C. coli strains do not produce extracellular enzymes capable of modifying erythromycin. Studies of induction of erythromycin resistance. Eryr C. jejuni and C. coli strains were preincubated with subinhib-

FIG. 1. Biological assay for the detection of erythromycin-modifying enzymes. A filter disc containing 15 ,ug of erythromycin was placed on an MH agar plate inoculated with the indicator strain S. aureus ATCC 25923. An inoculum of a test strain was applied to the center of each plate; an arrow-shaped distortion in the zone of inhibition indicates a breakdown of erythromycin by enzymes secreted by the test strain. (A) Negative control C. jejuni UA67. (B)

Positive control E. coli BM2571(pIP1527). (C) C. jejuni UA695. (D) C. coli UA585.

itory levels of erythromycin, lincomycin, or oleandomycin and incubated with high concentrations of erythromycin, and the growth rates of the cultures were monitored over a 72-h period. The resulting growth curves of C. jejuni UA709 and C. coli UA585 preincubated with erythromycin are shown in Fig. 2A and B, respectively. The growth rate was highest in control cultures to which no erythromycin was added. No significant difference in growth rate was detected, however, between induced and uninduced cultures. Identical results were obtained when cultures were preincubated with oleandomycin and lincomycin as well as for the other erythromycin-resistant C. jejuni strains, UA695 and UA697 (data not shown). Therefore, Eryr in C. jejuni and C. coli is constitutive. In vivo uptake and accumulation of ['4C]erythromycin. The rates of uptake and accumulation of [14C]erythromycin by isogenic C. coli UA585 (Eryr) and UA586 (Erys) as well as the unrelated strains C. jejuni UA709 (Eryr) and UA736 (Erys) were determined. There was no significant difference in [14C]erythromycin uptake among these strains. Optimal uptake was reached by 60 min and was followed by a slight decrease in accumulation over the next 120 min. Similar results were observed for the other Eryr C. jejuni strains, UA695 and UA697, when compared with the rate of uptake by the Erys strain UA67. Therefore, although strains of C. jejuni and C. coli varied in their susceptibilities to erythromycin, their abilities to transport the antibiotic into the cell and the subsequent accumulation of the drug were similar. Effect of erythromycin on protein synthesis in Campylobacter species. The amount of [35S]methionine incorporation into campylobacter cells was taken as an indication of in vivo

ANTIMICROB. AGENTS CHEMOTHER.

YAN AND TAYLOR

1992

A

0.10 0.08 0

0

"

0.06

0

0.04 O.0Q

X)

4) Time (hours)

0.10 0

B

0.08

'0

0

0.06

0.04

Joductiou4~ 0.02I

0.00 ^f~. 0

,-

20

40 Time (hours)

'

60

*

' 80

FIG. 2. Determination of the inducibility of C. coli UA585 Ery' (A) and C. jejuni UA709 Eryr (B) by preincubatio'n with subinhibitory concentrations of erythromycin and challengte with high concentrations of the antibiotic. Symbols: D, controlFI strains with no erythromycin added; *, strains preincubated wiith subinhibitory concentrations of the antibiotic and challenged witi high concentrations of erythromycin; O, strains not preincubate4d but challenged with high concentrations of erythromycin. OD6w, optical density at 600 nm.

protein synthesis. Cultures of C. coli UA5S 85. and UA586 were incubated with different concentratidns of erythromycin (0, 5.0, 30.0, and 100.0 j±g/ml) and eqtial amounts of [35S]methionine. Figure 3 shows the autoradiiogram of polypeptides from 'each culture separated by SDS-polyacrylamide gel electrophoresis. Roth C. coli UA5585 and UA586 in the absence of erythromycin incorporaited significant amounts of [35S]methionine (lanes B and E), demonstrating that in vivo protein synthesis occurred during the incubation period. However, the addition of 5 jig of er3ythromycin per ml to C. coli UA586 (lane F) almost comp]letely inhibited [35S]methionine incorporation. No protein synthesis was detected when greater amounts of erythromiycin (30.0 and 100.0 jig/ml) were added (lanes G and H, re spectively). C.

FIG. 3. Deterniination of the effect of increasing erythromycin concentrations on the incorporation of [35S]methionine into C. coli UA585 (Eryr) and UA586 (EryS). The polypeptides were sepakated by SDS-polyacrylamide gel electrophoresis. Lanes A through D represent the amount of incorporation into UA585 cells incubated with 0, 5.0, 30.0, and 100.0 ,ug of erythromycin per ml, respectively. Lanes E through H show the amount of incorporation into UA586 cells incubated with 0, 5.0, 30.0, and 100.0 xg of erythromycin per ml, respettively. Lane I represents "'C-protein molecular weight niarkers: methylated phosphorylase b (97,400), methylated bovine serum albuifin (66,200), methylated ovalbumin (42,700), methylated carbonic ainhydrase,(31,000), methylated trypsin inhibitor (21,500), and methylated Iysozyme (14,400).

coli UA585 protein synthesis, on the other hand, was unaffected by erythromycin concentratibns as high as 100 pLg/ml (lane D). Similar results werq ol?tained for C. jejuni strains; in vivo protein synthesis in resistant straiins UA695, UA697 and UA709 was unaffected by up to 100 pLg of erythromycin pdr ml, whereas 5 ,ug of the antibiotic per ml completely inhibited protein syhthesis in susceptible strains UA67 and UA736 (data not shown). These results demonstrate that the mechanism, df erythromycin resistance in C. jejuni and C. coli is at the level of protein synthesis. [`'C]erythromycin binding to purified ribosomes from Campylobacter species. Purified campylobacter ribosomes were isolated as described in Materials and Methods. Our binding studies demonstrated that under identical assay conditions, 705 ribosomes from Eryr C. coli UA585 bound significantly less ["'C]erythromycin than did those from the Erys derivative C. coli IPA586 (Fig. 4A). Similar binding properties were

ERYTHROMYCIN RESISTANCE IN C. JEJUNI AND C. COLI

VOL. 35, 1991

0.6

.,

1993

A

0

Ae 0.5 L.

0.4

8 0.

0.3

.

ao

1r-10.2 XaO.

0

20

40

60

14C-Ery U0

0.6

80

100

120

concentration (,uM)

B

a0 a. US.

9 -

'5

r.i

U0

0.1

I. p .I.i.i. 0

20

40

60 80 yconcentration

100

120

M)

FIG. 4. (A) Binding of [14C]erythromycin to purified C. coli UA585 (IEryr9 and UA586 (Erys) 708 ribosomes. Symbols: El, binding cuirve for UA585; *, binding curve for UA586. (B) Binding of ['4C]erythromycin to 50S ribosomal subunits isolated from C. coli UA585 (Eryr) and UA586 (Erys). Symbols are as in panel A. For both A and B, each point represents the mean of two independent determinations. The experiment was repeated twice, and identical results were obtained.

observed for 50S ribosomal subunits isolated from C. coli UA585 and UA586; 50S subunits from resistant strains bound significantly less ['4C]erythromycin than did those from susceptible strains (Fig. 4B). Binding experiments with [14C]erythromycin and ribosomes or 50S ribosomal subunits from C. jejuni UA709 and UA736 yielded similar results (data not shown). As expected, [14C]erythromycin did not bind to 30S ribosomal subunits from any of the strains tested (data not shown). Therefore, these results show that the 50S ribosomal subunits of Eryr Campylobacter species were altered so that binding of erythromycin was inhibited. Natural transformation of the Eryr marker. To confirm that erythromycin resistance is chromosomally mediated, we used genomic DNA isolated from C. coli UA585 to transform C. coli UA417 cells by the plate transformation method (49) and selected Eryr transformants. A transformation frequency of approximately 50 transformants per spot (corresponding to l0-5 transformants per viable cell) was obtained. Since neither the recipient nor the donor strains contained plasmid DNA, the success of the natural transfor-

FIG. 5. Pulsed-field gel electrophoresis analysis of SmnaI restriction fragments of C. ccli UA417 and its Eryr transformants. The bromide-stained 1% agarose gel contained restniction fragments separated by electrophoresis for 24 h at 175 V with a 12-s pulse time. (A) C. coli UA417. (B) C. coli UA585. (C to F) Eryr transformants of UA417. (G) X DNA concatamers, the sizes of

~~~~~~~ethidium

SmnaI

which are shown (in kilobases) on the right-hand side of the figure. Arrows show two new fragments generated by transformation of the Eryr marker in the isolate shown in lane D (see text).

mation procedure reconfirms that erythromycin resistance in C. coli is chromosomally mediated. Furthermore, of the 13 C. coli UA585 genomic DNA fragments produced by SmaI digestion (Fig. 5, lane B), only one was capable of transforming erythromycin resistance to C. coli UA417. Thus, the Eryr determinant of C. coli UA585 is located within the 240-kb SmaI fragment. Several Eryr transformants of C. coli UA417 were analyzed by pulsed-field gel electrophoresis after SmaI digestion. All except one possessed restriction patterns identical to that of the recipient (Fig. 5). In one derivative, SmaI fragment 3 was missing (Fig. 5, lane D). In its place two new fragments were generated. Additional transformation studies demonstrated that the larger of the two new SmaI fragments from this transformant was able to transform erythromycin resistance to C. coli UA417. The MICs for all Eryr transformants were similar to those for both UA585 and the original Eryr transformant of UA417. DISCUSSION

Although erythromycin resistance in campylobacters was first reported 13 years ago (9, 47), little work has been published on either the genetic mechanism involved or the biochemical basis behind macrolide resistance in this genus.

1994

YAN AND TAYLOR

ANTIMICROB. AGENTS CHEMOTHER.

16S

23S F 11 I

16S

16S

11

I

r, Eryr rI

12

1101

23S

n

11

5

1

2

8

3

1___6 r 9 1

7

1

4

1

ISmaI

0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 FIG. 6. Genomic map of C. coli UA417. The Sall and SmaI restriction sites are shown. Numbers are in kilobase pairs. Approximate locations of 16S and 23S rRNA genes were determined by hybridization with DNA probes (43), whereas the position of the erythromycin resistance marker (Eryr) was determined by natural transformation (53).

Before discussing the mechanism responsible for erythromycin resistance in campylobacters, it is instructive to consider the types of mechanisms found in other genera. Five different mechanisms have been reported. In gram-positive organisms (Staphylococcus, Streptococcus, Bacillus, and Clostridium species), as well as in Bacteroides species, the so-called MLS resistance mechanism is prevalent (6, 33). MLS resistance is specified by a methylase which produces N6N6-dimethylation of specific adenine residues within 23S rRNA (28, 29), resulting in a reduced affinity of the antibiotic for the target site, 50S ribosomes. The genes for erythromycin resistance are presumed to have evolved from the erythromycin-producing organism Streptomyces erythreus (7). Although in S. erythreus the genes are transcribed constitutively, in many organisms resistance is inducible (7, 20). In the absence of erythromycin, the mRNA for the methylase is inactive and 23S rRNA remains unmethylated. In the presence of subinhibitory concentrations of the antibiotic, the methylase mRNA is subjected to an attenuation mechanism during translation and active methylase is made (15, 16, 20). Seven classes of highly conserved genes coding for the RNA methylases, designated ermA to ermG, as well as ermP, have been identified by DNA hybridization studies (2, 6, 13, 33). High-level resistance to erythromycin in members of the family Enterobacteriaceae has been shown to be due to macrolide-inactivating enzymes of two different types, erythromycin esterases and macrolide 2'-phosphotransferases (1-3, 5, 34). Resistance to macrolides and streptogramin B in Staphylococcus spp. was shown to depend on the efflux of these antibiotics (35). Protein MsrA, encoded by the msrA gene, was shown to contain two ATP-binding domains and to confer resistance by a mechanism of active efflux. Resistance mediated by plasmid pNE24 in Staphylococcus epidermidis was also shown to result from energy-dependent efflux of macrolides (19). The three mechanisms described above constitute acquired resistance, whereas the following two can be considered mutational resistance. In E. coli, erythromycin resistance can result from a mutation in one of the seven copies of the 23S rRNA gene (18, 39). This situation only occurs if a single rrn operon is amplified after being cloned into a multicopy plasmid. Mutations which mapped to the 23S rRNA gene of the rrnH operon cloned into a multicopy cloning vehicle resulted in resistance of host cells to erythromycin (18, 39). Eryr E. coli mutants can be isolated in the laboratory by selection on plates containing erythromycin. E. coli EryA mutants have an alteration in the large ribosomal subunit protein L4 (50) encoded by the rplD gene at 73 min on the E. coli map (4). Similarly, Bacillus stearothermophilus mutants resistant to erythromycin (ery-l) have altered L4 proteins, as

determined by two-dimensional polyacrylamide gel electrophoresis of ribosomal proteins (38). Although the presence of an rRNA methylase in Eryr campylobacters was not sought directly, several pieces of evidence suggest that this is unlikely to be the mechanism responsible for erythromycin resistance. First, the Eryr determinant from Campylobacter species has never been successfully cloned in E. coli, despite extensive efforts in three different laboratories (12, 45, 51). On the other hand, RNA methylase genes from several different species have been successfully cloned in E. coli (6, 33). Failure to clone the Eryr determinant from Campylobacter species suggests that the determinant has a recessive phenotype, such as that due to a direct change in the target, 50S ribosomal subunits. Second, DNA probes for several of the methylase genes (ermA, ermAM, ermC, ermD, and ermE) do not hybridize to DNAs from Eryr C. jejuni and C. coli, even under conditions of low stringency (12, 36, 51). Third, erythromycin resistance in C. jejuni and C. coli is constitutive, unlike erythromycin resistance due to RNA methylases in most species examined, which is inducible (5, 6, 15, 16, 20, 33). Fourth, MLS resistance is characteristic of RNA methylases in other species (29). However, only macrolide and lincosamide resistance can be seen in Ery' campylobacters, since all campylobacter strains appear to be intrinsically resistant to streptogramin B (42). Our studies have ruled out erythromycin inactivation in Eryr campylobacter strains, since bioassays failed to detect any extracellular erythromycin-modifying enzyme. Our work has also shown that the ribosomes are affected directly, since 50S subunits from Eryr strains failed to bind ["4C]erythromycin. An active efflux system able to transport erythromycin out of cells was shown not to be involved in campylobacters, since Eryr and EryS strains took up and accumulated ['4C]erythromycin in an identical fashion. Elimination of the acquired Eryr mechanisms leaves us with those due to mutations. This idea is consistent with the ability of the Eryr marker from C. coli UA585 to be transformed into C. coli UA417 by genomic DNA. C. jejuni and C. coli contain three copies of both 16S and 23S rRNA genes (25, 26, 51). It would be necessary to envisage simultaneous mutations in two or three of the 23S rRNA genes to override the effect of the other, wild-type 23S rRNA gene(s). In addition, hybridization with a DNA probe containing the 23S rRNA gene has shown that the 23S rRNA gene is not located in the fragment associated with the transformation of erythromycin resistance (43) (Fig. 6). Hybridization of a DNA probe for 16S rRNA has given similar results (43, 51). Therefore, Eryr in campylobacters is unlikely to be due to a mutation(s) in an rRNA gene(s). To determine whether Eryr in Campylobacter species is due to an altered ribosomal protein, we compared ribosomal proteins in 70S ribosomes from Eryr and Erys C. coli UA585

VOL. 35, 1991

ERYTHROMYCIN RESISTANCE IN C. JEJUNI AND C. COLI

and UA586 by two-dimensional gel electrophoresis. Consistent differences could be detected between the protein profiles of ribosomes from the Eryr and Erys strains (43), although further studies are required to determine exactly which protein(s) is altered. This evidence lends further support to the hypothesis that macrolide resistance in Campylobacter species results from a mutation in a ribosomal protein gene. ACKNOWLEDGMENTS We thank P. Courvalin for E. coli BM2571(pIP1527) and for providing unpublished information, F. Jurnak for advice on the isolation and purification of ribosomes, H. Lior for hospitality and for providing the antisera and reagents for serotype and biotype determinations, F. C. Tenover for providing unpublished information, B. Weisblum for streptogramin B and for advice, and L. E. Bryan for interest in this study. This work was supported by an operating grant from the Natural Sciences and Engineering Research Council of Canada. W.Y. was a recipient of an Alberta Heritage Foundation for Medical Research Studentship, and D.E.T. is a Heritage Scientist. REFERENCES 1. Andremont, A., G. Gerbaud, and P. Courvalin. 1986. Plasmidmediated high-level resistance to erythromycin in Escherichia coli. Antimicrob. Agents Chemother. 29:515-518. 2. Arthur, M., A. Andremont, and P. Courvalin. 1987. Distribution of erythromycin esterase and rRNA methylase genes in members of the family Enterobacteriaceae highly resistant to erythromycin. Antimicrob. Agents Chemother. 31:404-409. 3. Arthur, M., and P. Courvalin. 1986. Contribution of two different mechanisms to erythromycin resistance in Escherichia coli. Antimicrob. Agents Chemother. 30:694-700. 4. Bachmann, B. J. 1987. Linkage map of Escherichia coli K-12, edition 7, p. 807-876. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology. American Society for Microbiology, Washington, D.C. 5. Barthelemy, P., D. Autisser, G. Gerbaud, and P. Courvalin. 1984. Enzyme hydrolysis of erythromycin by a strain of Escherichia coli. A new mechanism of resistance. J. Antibiot. 37:1692-1696. 6. Berryman, D. I., and J. I. Rood. 1989. Cloning and hybridization analysis of ermP, a macrolide-lincosamide-streptogramin B resistance determinant from Clostridium perfringens. Antimicrob. Agents Chemother. 33:1346-1353. 7. Bibb, M. J., G. R. Janssen, and J. M. Ward. 1985. Cloning and analysis of the promoter region of the erythromycin resistance gene (ermE) of Streptomyces erythraeus. Gene 38:215-226. 8. Bonner, W. M., and R. A. Laskey. 1974. A film detection method for tritium-labelled proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem. 46:83-88. 9. Brunton, W. A. T., A. A. M. Wilson, and R. M. Macrae. 1978. Erythromycin-resistant campylobacters. Lancet ii:1385. 10. Burridge, R., C. Warren, and I. Phillips. 1986. Macrolide, lincosamide and streptogramin resistance in Campylobacter jejunilcoli. J. Antimicrob. Chemother. 17:315-321. 11. Chang, N., and D. E. Taylor. 1990. Use of pulsed-field agarose gel electrophoresis to size genomes of Campylobacter species and to construct a Sall map of Campylobacterjejuni UA580. J. Bacteriol. 172:5211-5217. 12. Courvalin, P. (Institut Pasteur, Paris, France). Personal communication. 13. Courvalin, P., H. Ounissi, and M. Arthur. 1985. Multiplicity of macrolide-lincosamide-streptogramin antibiotic resistance determinants. J. Antimicrob. Chemother. 16:91-100. 14. Dougan, G., and M. Kehoe. 1984. The minicell system as a method for studying expression from plasmid DNA, p. 233-258. In P. M. Bennett and J. Grinsted (ed.), Methods in microbiology, vol. 17. Plasmid technology. Academic Press, Inc., New York.

1995

15. Dubnau, D. 1984. Translational attenuation: the regulation of bacterial resistance to the macrolide-lincosamide-streptogramin B antibiotics. Crit. Rev. Biochem. 16:103-128. 16. Dubnau, D., and M. Monod. 1986. The regulation and evolution of MLS resistance. Banbury Rep. 24:369-387. 17. Endtz, H. P. (University Hospital, Rotterdam, The Netherlands). Personal communication. 18. Ettayebi, M., S. M. Prasad, and E. A. Morgan. 1985. Chloramphenicol-erythromycin resistance mutations in a 23S rRNA gene of Escherichia coli. J. Bacteriol. 162:551-557. 19. Goldman, R. C., and J. 0. Capobianco. 1990. Role of an energy-dependent efflux pump in plasmid pNE24-mediated resistance to 14- and 15-membered macrolides in Staphylococcus epidermidis. Antimicrob. Agents Chemother. 34:1973-1980. 20. Horinouchi, S., and B. Weisblum. 1980. Post transcriptional modification of mRNA conformation: mechanism that regulates erythromycin-induced resistance. Proc. Natl. Acad. Sci. USA 77:7079-7083. 21. Jelenc, P. C. 1980. Rapid purification of highly active ribosomes from Escherichia coli. Anal. Biochem. 105:369-374. 22. Karmali, M. A., S. A. DeGrandis, and P. C. Fleming. 1981. Antimicrobial susceptibility of Campylobacter jejuni with special reference to resistance patterns of Canadian isolates. Antimicrob. Agents Chemother. 19:593-597. 23. Karmali, M. A., S. A. DeGrandis, D. E. Taylor, and P. C. Fleming. 1982. On the association between erythromycin resistance and failure to hydrolyse hippurate in Campylobacter jejuni, p. 218-220. In D. G. Newell (ed.), Campylobacter. Epidemiology, pathogenesis and biochemistry. MTP Press, Lancaster, United Kingdom. 24. Karmali, M. A., J. L. Penner, P. C. Fleming, A. Williams, and J. N. Hennessy. 1983. The serotype and biotype distribution of clinical isolates of Campylobacter jejuni and Campylobacter coli over a three year period. J. Infect. Dis. 147:243-246. 25. Kim, N. W., and V. L. Chan. 1989. Isolation and characterization of the ribosomal RNA gene of Campylobacterjejuni. Curr. Microbiol. 19:247-252. 26. Labigne-Roussel, A., P. Courcoux, and L. Tompkins. 1988. Gene disruption and replacement as a feasible approach for mutagenesis of Campylobacterjejuni. J. Bacteriol. 170:1704-1708. 27. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685. 28. Lai, C. J., and B. Weisblum. 1971. Altered methylation of ribosomal RNA in an erythromycin-resistant strain of Staphylococcus aureus. Proc. NatI. Acad. Sci. USA 68:856-860. 29. Lai, C. J., B. Weisblum, S. R. Fahnestock, and M. Nomura. 1973. Alteration of 23S ribosomal RNA and erythromycin induced resistance to lincomycin and spiramycin in Staphylococcus aureus. J. Mol. Biol. 74:67-72. 30. Lior, H. 1984. New, extended biotyping scheme for Campylobacter jejuni, Campylobacter coli, and "Campylobacter laridis." J. Clin. Microbiol. 20:636-640. 31. Lior, H., D. L. Woodward, J. A. Edgar, and L. J. Laroche. 1981. Serotyping by slide agglutination of C. jejuni and epidemiology. Lancet ii:1103-1104. (Letter.) 32. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3:208-218. 33. Murphy, E. 1985. Nucleotide sequence of ermA, a macrolidelincosamide-streptogramin B determinant in Staphylococcus aureus. J. Bacteriol. 162:633-640. 34. O'Hara, K., T. Kanda, K. Ohmiya, T. Ebisu, and M. Kono. 1989. Purification and characterization of macrolide 2'-phosphotransferase from a strain of Escherichia coli that is highly resistant to erythromycin. Antimicrob. Agents Chemother. 33: 1354-1357. 35. Ross, J. I., E. A. Eady, J. H. Cove, W. J. Cundliffe, S. Baumberg, and J. C. Wootton. 1990. Inducible erythromycin resistance in staphylococci is encoded by a member of the ATP-binding transport super-gene family. Mol. Microbiol. 7:1207-1214. 36. Ryan, C. A., and F. C. Tenover. 1988. Characterization of erythromycin-resistant isolates of Campylobacter jejuni and

1996

37.

38.

39. 40. 41.

42. 43.

44.

YAN AND TAYLOR

Campylobacter coli. Program Abstr. 28th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1260, p. 337. Sagara, H., A. Mochizuki, N. Okamura, and R. Nakaya. 1987. Antimicrobial resistance of Campylobacterjejuni and Campylobacter coli with special reference to plasmid profiles of Japanese clinical isolates. Antimicrob. Agents Chemother. 31:713-719. Schier, J., H.-S. Gewitz, S.-E. Behrens, A. Lee, C. Ginther, and T. Leighton. 1990. Isolation and characterization of Bacillus stearothermophilus 30S and 50S ribosomal protein mutations. J. Bacteriol. 172:7306-7309. Sigmund, C. D., and E. A. Morgan. 1982. Erythromycin resistance due to a mutation in a ribosomal RNA operon of Escherichia coli. Proc. Natl. Acad. Sci. USA 79:5602-5606. Steers, E., E. L. Foltz, B. S. Graves, and J. Riden. 1959. An inocula replicating apparatus for routine testing of bacterial susceptibility to antibiotics. Antibiot. Chemother. 9:307-311. Taylor, D. E., N. Chang, R. S. Garner, R. Sherburne, and L. Mueller. 1986. Incidence of antibiotic resistance and characterization of plasmids in Campylobacterjejuni strains isolated from clinical sources in Alberta, Canada. Can. J. Microbiol. 32:2832. Taylor, D. E., and P. Courvalin. 1988. Mechanisms of antibiotic resistance in Campylobacter species. Antimicrob. Agents Chemother. 32:1107-1112. Taylor, D. E., W. Yan, M. Simons, and B. Bablitz. Unpublished data. Taylor, D. N., M. J. Blaser, P. E. Echeverria, C. Pitarangsi, L. Bodhidatta, and W.-L. L. Wang. 1987. Erythromycin-resistant Campylobacter infections in Thailand. Antimicrob. Agents Chemother. 31:438-422.

ANTIMICROB. AGENTS CHEMOTHER. 45. Tenover, F. C. (Centers for Disease Control, Atlanta, Ga.). Personal communication. 46. Vanhoof, R., M. P. Vanderlinden, R. Dierickx, S. Lauwers, E. Yourassowsky, and J. P. Butzler. 1978. Susceptibility of Campylobacter fetus subsp. jejuni to twenty-nine antimicrobial agents. Antimicrob. Agents Chemother. 14:553-556. 47. Walder, M., and A. Forgren. 1978. Erythromycin-resistant campylobacters. Lancet ii:1201. 48. Wang, W.-L. L., L. B. Reller, and M. J. Blaser. 1984. Comparison of antimicrobial susceptibility patterns of Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 26:351-353. 49. Wang, Y., and D. E. Taylor. 1990. Natural transformation in Campylobacter species. J. Bacteriol. 172:949-955. 50. Wittman, H. G., G. Stoffier, D. Apirion, L. Rosen, K. Tanaka, M. Tomaki, R. Takata, S. Dekio, E. Otaka, and S. Osawa. 1973. Biochemical and genetic studies on two different types of erythromycin resistant mutants of Escherichia coli with altered ribosomal proteins. Mol. Gen. Genet. 127:175-189. 51. Yan, W. 1990. Characterization of erythromycin resistance in Campylobacter spp. Ph.D. thesis. University of Alberta, Edmonton, Alberta, Canada. 52. Yan, W., N. Chang, and D. E. Taylor. 1991. Pulsed-field gel electrophoresis of Campylobacter jejuni and Campylobacter coli genomic DNA and its epidemiological application. J. Infect. Dis. 163:1068-1072. 53. Yan, W., and D. E. Taylor. 1991. Sizing and mapping of the genome of Campylobacter coli strain UA417R using pulsed-field gel electrophoresis. Gene 101:117-120.