Staphylococcus aureus Dependent on an Unusual Specificity of the Recipient Strain

Transcription

1 JOURNAL OF BACTERIOLOGY, Dec. 1970, p Vol. 104, No. 3 Copyright C 1970 American Society for Microbiology Printed in U.S.A. Transduction of Methicillin Resistance in Staphylococcus aureus Dependent on an Unusual Specificity of the Recipient Strain SIDNEY COHEN AND HELEN M. SWEENEY Department of Microbiology, Michael Reese Hospital and Medical Center, Chicago, Illinois Received for publication 20 July 1970 Resistance to methicillin was transduced by phage 80 or 53 from two naturally occurring methicillin-resistant strains of Staphylococcus aureus to methicillinsusceptible recipient strains at frequencies of 10-v to Ultraviolet irradiation of transducing phage and posttransductional incubation at 30 C were essential for useful frequencies of transduction. Effectiveness as a recipient for this transduction was highly specific. Strain NCTC 8325 (PS47) in its native state was an ineffective recipient but became effective after it had received by transduction one of several penicillinase plasmids. This acquired effectiveness was retained in most cases after elimination of the plasmid by ethidium bromide treatment. Like the donor strain, the progeny were heterogeneous in the degree of their resistance to methicillin, which was expressed by a higher proportion of cells as the temperature of incubation was lowered from 37 to 30 C. Separate transductants varied widely in the degree of resistance acquired by transduction. Methicillin resistance was stable in the donor and transductant strains. We favored the interpretation that methicillin resistance in our strains was determined by a single chromosomal gene, although the possibility that it was determined by two or more closely linked genes could not be excluded. The genetic determinants of resistance to antibiotics have proven to be useful markers in the distinctive features. Resistant strains are hetero- recognizable enzymatic basis but it does possess investigation of the genetic constitution of Staphylococcus aureus. Depending on the antibiotic, form colonies when they are cultivated at 37 C geneous, that is only a small proportion of cells the genes mediating resistance may occur on in ordinary media containing elevated concentrations of methicillin (24). If the temperature is plasmids, on the chromosome or, in some cases, in either position (18, 23). The most thoroughly lowered to 30 or 25 C or if high concentrations of studied marker is resistance to benzylpenicillin, inorganic salts, such as NaCl or (NH4)2SO4, are caused by hydrolytic inactivation of the antibiotic added to medium incubated at 37 C, the proportion of cells forming colonies in elevated con- by the enzyme penicillinase. The structural gene for this enzyme together with closely linked centrations of methicillin may approach 100% regulatory genes may be situated either on a (1, 5). In the clones that grow at elevated concentrations of methicillin, all the cells are highly plasmid or on the chromosome (4, 22, 23). Methicillin is the prototype of a group of resistant, but when they are subcultured in semisynthetic penicillins that are resistant to methicillin-free medium, the proportion of staphylococcal penicillinase and thereby circumvent the clinical problem posed by infections by original distribution (12). The genetic basis, if highly resistant cells often reverts to the penicillin-resistant staphylococci. Shortly after any, of this phenomenon is unknown. the advent of methicillin, staphylococci were In the past, genetic studies of the nature of detected that were resistant to these agents and methicillin resistance have been frustrated by the to the cephalosporin antibiotics as well. Although absence of any method of genetic transfer. Recently Dornbusch, Hallander, and Lofquist (11) still rare in the United States, methicillin-resistant staphylococci are being isolated with increasing reported the transduction of the genetic determinant of methicillin resistance and came to the frequency in other countries, usually in hospital patients (20). Resistance to methicillin has no conclusion that it resided on a plasmid in their 1158

2 VOL. 104, 1970 TRANSDUCTION OF METHICILLIN RESISTANCE strains. In this paper we present evidence for the phage-mediated transduction of methicillin resistance from two naturally occurring resistant strains. Transduction was dependent on an unusual specificity of the recipient strains. Methicillin resistance in our strains had properties of a chromosomal marker, in contrast to those investigated by Dornbusch et al. (11). MATERIALS AND METHODS Organisms. The nomenclature of strains of S. aureus and their genetic markers follows that of Peyru, Wexler, and Novick (21). In this system penicillinase plasmids are designated by the letter P followed by the Roman numeral I or II to designate the plasmid maintenance-compatibility locus and then by a subscript indicating the original host strain of the plasmid. The maintenance-compatibility locus does not figure in this paper and we have omitted its symbol. Thus, 258(P5c5) signifies strain 258 containing a plasmid originally found in strain 55Cl. The prefix N before a strain designation signifies that it has been cured of the indicated plasmid. The designation cp signifies a chromosomal penicillinase linkage group containing regulatory and structural genes for penicillinase (4, 22, 25). The strain from which a cp was derived is indicated by a subscript. The naturally occurring methicillin-resistant strains employed were isolated in Seattle by Gravenkemper, Brodie, and Kirby (14) and supplied by Benner (6). They were strain Russell, phage type 53/77, resistant to penicillin, erythromycin, tetracycline, and streptomycin and strain Villaluz, phage type 53/77/83A, resistant to penicillin and streptomycin. These strains are referred to hereafter as C4 and C5. Each strain carried a penicillinase plasmid demonstrable by cotransduction of penicillinase and resistance to Cd2+ (19). The usual recipient strain, 8325(P524), was received from R. P. Novick who had prepared it by transducing the plasmid into NCTC 8325, the propagating strain for phage 47 (16). We obtained the original strain 8325 free of a penicillinase plasmid from the Center for Disease Control, Atlanta. Strain 8325 and its plasmid-cured derivative, N8325(P524), were equally susceptible to methicillin, penicillin, streptomycin, tetracycline, erythromycin, chloramphenicol, and kanamycin. Their phage types were the same, 47/53/75. However, at 1,000 times the routine test dilution (RTD), 8325 was also lysed by phages 7, 29, 52, 52A, 54, 75 and slightly by 80, whereas N8325(P524) was lysed by phages 7 and 54. Properties of other staphylococci have been described (9, 10, 25) with the exception of strain A20, a clinical isolate that was not typable at RTD and was 47/53/54/83A at 1,000 X RTD. It was susceptible to methicillin and resistant to penicillin, streptomycin, tetracycline, erythromycin, kanamycin, and chloramphenicol. Bacteriophage. Staphylococcal phages of the international typing series were propagated in soft tryptic soy agar (Difco; reference 7). Phage was grown in lytic cycle from lysogenic organisms after ultraviolet 1159 induction. The organism was inoculated into Nutrient Broth (Difco) at an initial optical density (OD) of 0.05 measured in a 19-mm cuvette at 540 nm in a Coleman Jr. spectrophotometer. It was grown with shaking at 37 C for about 4 hr to an OD of The cells from 10 ml of culture were collected in the centrifuge, washed with 0.85% NaCl solution, and resuspended in 9 ml of the same solution. This was irradiated for 10 sec with shaking in a glass petri dish (100 mm in diameter) at a distance of 30 cm from a General Electric germicidal lamp. The energy delivered was 18 ergs per sec per mm2. The suspension was mixed with 1 ml of Nutrient Broth prepared at 10 times normal concentration and incubated at 37 C until lysis was complete. The lysate was sterilized by membrane filtration. Phage typing was performed with the set of phages and the procedures of Blair and Williams (7). Transduction. The following procedure was eventually adopted for transduction of methicillin. Phage 80 [106 plaque-forming units (PFU) per ml] was propagated for 18 hr at 37 C on strain C5 [2.5 X 107 colony-forming units (CFU) per ml] in overlays of soft tryptic soy agar (Difco) with 400 Ag of CaCl2 added per ml. The phage was harvested with 20 ml of Nutrient Broth for each 150-mm petri plate and sterilized by membrane filtration. The propagation was repeated with an input of 106 PFU/ml determined by titration on C5. The membrane filter-sterilized progeny was used for transduction. Ultraviolet irradiation was delivered for variable times to 3-ml portions of phage in Nutrient Broth in the manner described above. The recipient organism was grown overnight on Brain Heart Infusion (BHI) agar (Difco) plates. Colonies were suspended in Nutrient Broth supplemented with 400 jg of CaCl2 per ml to provide 1010 to 2 X 1010 CFU per ml. One milliliter of recipient and about 2 X 109 to 5 X 109 PFU of phage in Nutrient Broth were mixed. After 30 min of incubation with shaking at 37 C, the suspension was centrifuged at 4 C. The cells were washed twice with BHI broth and resuspended in 0.3 ml of the same broth. Two 0.1-ml portions were spread on BHI agar plates containing 12.5,g of methicillin per ml when the recipient produced penicillinase and 5 $&g per ml when it did not. Transductants were enumerated after 40 hr of incubation at 30 C. They were purified by one or two streakings on agar plates containing the selective concentration of methicillin. All experiments included controls for sterility of phage and for methicillinresistant progeny in the absence of phage. None were found in many experiments. The same procedure was used for transduction of methicillin resistance with phage 53 from C4 and C5 or from derivatives of 8325 by its ultraviolet-induced temperate phage. Penicillinase plasmids or chromosomal penicillinase linkage groups were transduced as previously described (9, 25). Transductions of the penicillinase plasmid from 8325(P524) to previously plasmid-free cultures of strain 8325 were performed either with the induced prophage or with 80 a, a variant of phage 80 which gave plaques on Streptomycin resistance was transduced from C5 by phage 80. After the recipient-phage mixture was

3 1160 COHEN AND SWEENEY J. BACTERIOL. washed with BHI broth, the cells were incubated in 10 ml of the broth for 3 hr and then plated on BHI agar containing 400,ug of streptomycin per ml. Transductants were enumerated after 40 hr and corrected for the small number of spontaneous mutants in control platings. Methicillin susceptibility. Quantitative determination of methicillin susceptibility was determined by spotting 20 to 70 CFU of the test strain in 0.02-ml drops on Heart Infusion Agar (Difco) plates containing serial twofold dilutions of methicillin, usually starting with 400 pag/ml. The inoculum was taken by suspending colonies grown on Heart Infusion Agar containing 12.5 ;sg of methicillin per ml for penicillinase-forming strains and 5 pg/ml for penicillinasenegative strains. Plates were incubated at 31.5 to 32 C to 43 hr in a water-jacketed incubator, and colonies were enumerated. The minimal inhibitory concentration (MIC), determined by interpolation, was that which reduced colony formation to 50% of the control value. In most titrations a culture of C5 was included as a control. This procedure, similar to that of Dyke (12), was based on Annear's observation of the temperature sensitivity of staphylococcal methicillin resistance (1). In seven determinations the MIC for strain C5 was 141 i 25 (standard deviation),ug/ml. Accurate control of the temperature of incubation was important for reproducibility. Methicillin-susceptible staphylococci gave no colonial growth on 1.25 or 2.5 pug of methicillin per ml. Elimination of plasmids. Log-phase cultures in tryptic soy broth were diluted to contain 104 cells per ml. Ethidium bromide was added in concentrations of 1.0 to 2.5,ug/ml (8). The cultures were incubated at 37 C for 24 hr in the dark. The tube which contained the least concentration of ethidium bromide without visible growth was plated on Heart Infusion Agar plates, incubated overnight and replicated to medium containing 2.5 X 10-4 M Cd(NO3)2. Colonies that failed to replicate were tested for loss of other plasmid markers, including penicillinase formation and resistance to arsenate, bismuth, and mercuric ions (19). The same procedure was followed in attempts to eliminate methicillin resistance. In addition, some experiments were performed with acriflavine in concentrations of 12.5 to 50 ;pg/ml incubated for 48 hr (11). Colonies were replicated to medium containing 12.5 pug of methicillin/ml. Reagents. Bovine pancreatic deoxyribonuclease was purchased from Worthington Biochemical Corp. (Deoxyribonuclease I DPFF). RESULTS Conditions of transduction. Optimal conditions of transduction of methicilin resistance were determined with phage 80 propagated on C5 as the transducing agent and 8325 (P524) as recipient. The procedure finally adopted was similar to that used for other staphylococcal transductions. Ultraviolet irradiation of the phage greatly enhanced the frequency of transduction and was essential for transduction at a useful rate (Fig. 1). Sufficient irradiation to reduce plaque formation to 1 to 10% of the original value was optimal. Irradiation for 2 min under our conditions and with reasonably fresh phage gave transduction frequencies in the range of 10-v to 10-8 per PFU of unirradiated phage. In many experiments with unirradiated phage, methicillin-resistant transductants were not obtained and when present their frequency of transduction was approximately 10-9 to Similar ultraviolet irradiation of phage 80 propagated on the methicillin-susceptible strain NCTC 9789 (PS80) did not make the phage a vector of methicillin resistance. Thus it appears that the irradiation did not generate mutations in the phage that were manifested by methicillin resistance in the recipient staphylococcus. The frequency of transduction was proportional to phage input when the multiplicity of infection was <0.2 and at higher values declined progressively (Fig. 2). Posttransductional incubation in broth for 1 to 4 hr before plating on the selective medium did not increase the proportion of transductants appreciably. Thus, no phenotypic lag in expression of methicillin resistance was detectable under these selective conditions. A selective concentration of methicillin not exceeding 12.5,ug/ml gave the maximal number of transductants with 8325 (P524). Similarly, 5,ug/ml was the limiting selective concentration 1.0 ^1 I \ o 0.1A z~~~~~~~~~~ 0~~~~~~~~~~ 10 0~~ ~ ~ ~ ~ ~~~0 R~~~~~~~~~~~~....0 \ ULTRAVIOLET IRRADIATION (MIN) FIG. 1. Effect of ultraviolet irradiation of 80 phage on the frequency oftransduction ofmethicillin resistance from strain C5 to 8325 (Pu4). The multiplicity of infection in this experiment was 0.45.

4 VOL. 104, 1970 TRANSDUCTION OF METHICILLIN RESISTANCE with its plasmid-cured derivative N8325 (P524) as recipient. More transductants were obtained when the temperature of the post transductional incubation on methicillin agar was 30 C than 32 or 37 C (Table 1). This result was to be expected from Annear's observations of the temperature sensitivity of staphylococcal methicillin resistance (1). Properties of the transducing agent. The usual transducing agent was phage 80 propagated twice on strain C5. Before using it for this purpose, we purified our stock 80 phage by three serial singleplaque passages on PS80. The efficiency of plating of stock 80 phage on strain C5 was 2.5 X 10-4 relative to PS80, whereas that of phage propagated for the second time on strain C5 was 2.6. We do not know whether the transducing phage was mutant or-host modified. The transducing capacity of the phage 80 lysate was not appreciably reduced by 30 min of incubation at 35 C with 50,ug of pancreatic deoxyribonuclease per ml. Furthermore, a rabbit antiserum to stock 80 phage inhibited plaque formation and transduction of methicillin resistance by phage propagated on C5 (Table 2). Thus, it appears that the genetic transfer of methicillin resistance was in fact phage mediated. The effectiveness of the phage stored at 4 C for transducing methicillin resistance tended to decline irregularly but appreciably more rapidly than plaque-forming ability. The useful life of most preparations was 2 to 4 weeks. Methicillin resistance could be transduced from strains C4 and C5 to 8325 (P524) by 53 phage similarly propagated on these hosts; the frequency of transduction, however, from C4 was about 2.5 X 10-8 and from C5 about 5 X These rates were distinctly less than those obtained for the transduction from C5 by 80 phage. Strain C5 carried one or more prophages inducible by ultraviolet irradiation. A phage lysate induced from C5 which contained 2.4 X 1010 PFU per ml assayed on PS7 did not transduce methicillin resistance to 8325 (P524) although it did transduce the C5 penicillinase plasmid at a frequency of 5 X 10-. However, once transduced by 80 or 53 phage to 8325 (P524), methicillin resistance could be transduced further at a frequency of 3 X 10-8 to strains 8325 (P524) or N8325 (P524) and at lower frequency to 55C1 by an ultraviolet-inducible prophage resident in 8325 (16, 17). Effect of recipients on the frequency of transduction of methicillin resistance. A striking feature of this transductional system was its highly selective requirement for a suitable recipient. The ability of a number of strains to act as recipients in this MULTIPLICITY OF INFECTION FIG. 2. Relation of multiplicity of infection with 80 phage to the number of methicillin-resistant trantsductants. The donor strain was C5 and the recipient 8325 (PF24). The frequency of transduction was 2.6 X 107 when the multiplicity of infection <0.2. The different symbols designate results of three experiments with the same lot ofphage. TABLE 1. Effect of temperature on the number of methicillin-resistant transductantsa Multiplicity of infection No. of methicillin-resistant transductants 30C 32C 37C a Transducing phage was propagated on C5. The recipient was 8325 (P524). The headings indicate the temperature of incubation of the methicillin-agar-selective plates. system is indicated in Table 3. In sharp contrast to the effectiveness of the plasmid-containing strain 8325 (P524), the original plasmid-free strain 8325 failed to yield any methicillin-resistant transductants in many attempts. The difference was not due to the presence of the penicillinase plasmid, since cured clones of N8325 (P524) were still effective recipients. Both 8325 and 8325 (P524) absorbed equally well (more than 99%) the ability of the transducing phage to form plaques and to transduce methicillin resistance. A preparation of phage 80 propagated on C5 that transduced methicilin resistance to 8325 (P524) and to N8325 (P524) but not to 8325 transduced the penicillinase plasmid with equal fre-

5 1162 COHEN AND SWEENEY J. BACTERIOL. TABLE 2. Effect of phage antisera on transduction of methicillin resistancea Phage treatment Phage ti transductants None X Normal rabbit serum (1:10) X Antiserum to 80 phage (1:10)... <5 X Antiserum to 80 phage (1:10) propagated on CS <5 X athe donor was phage 80 propagated twice on strain C5. The recipient was 8325 (P524). One milliliter of a 1:10 dilution of antiserum in Nutrient Broth was incubated in the dark at 37 C for 30 min with 1 ml of a phage lysate which had been ultraviolet-irradiated for 2 min. Samples (0.1 ml) were taken for phage assay, and, to the remainder, 1 ml of recipient bacterial suspension was added to initiate transduction. The antisera had K values of 33. b Expressed as plaque-forming units per milliliter. TABLE 3. quency to the latter two hosts, 5 X 10-6 before ultraviolet irradiation and 7 X 107 after irradiation. These experiments indicate that the ineffectiveness of 8325 as a recipient for transduction of methicillin resistance was not the result of failure to absorb the transducing phage. Streptomycin resistance in staphylococci has been considered to be a chromosomal marker since its transduction from streptomycin-resistant mutants was enhanced by ultraviolet irradiation of the transducing phage (15). The frequency of transduction of the naturally occurring streptomycin resistance in C5 was increased fourfold by ultraviolet treatment, and we concluded that it also reflected a chromosomal mutation. Streptomycin resistance was transduced from C5 to 8325 and 8325 (P524) at frequencies of 8 X 108 and 6 X 10-8, respectively, with an ultraviolettreated phage that transduced methicillin resistance to 8325 (P524) at a frequency of 9 x 10- and gave no transductants (<1.6 x 10-) with Therefore it does not appear that 8325 differed from 8325 (P524) in ability to integrate chromosomal markers generally. Effectiveness of strains of S. aureus as recipients in transduction of methicillin resistance from strain C5 by phage 80 Source of recipient strains Effective recipients Ineffective recipients No. testeda Strain designation No. testeda Strain designation Wild type, containing their native plasmids 1 55C1 1 PS80 1 A20 1 PS53 1 AS Wild type, penicillinase negative Penicillinase plasmid transductants (P268) (PA 6) (PI 47) (P524) (PcO) Strains cured of penicillinase plasmidsb 5 N8325 (P524) 2d N8325 (Pcs) 10 N8325 (Pcs) Chromosomal penicillinase transductants (cpsscs) (CPB 4) a Designates the number of genetically independent clones examined. This number is one for wildtype strains and multiple for transductants or plasmid-cured strains. b Cured of the plasmid by treatment with ethidium bromide. For each strain the cured clones were descendants of a single penicillinase plasmid transductant which was an effective recipient of methicillin resistance. c Four independently cured clones and six others from an additional experiment were tested. d These clones were obtained from the same experiment and therefore may have been sibs.

6 VOL. 104, 1970 TRANSDUCTION OF METHICILLIN RESISTANCE 1163 The preliminary acquisition by transduction of P524 or Pc5 made 8325 an effective recipient in tranduction of methicillin resistance. Table 3 indicates that 23 clones of 8325 that received the C5 plasmid by transduction and 28 clones that received the 524 plasmid became effective recipients for methicillin resistance. Some naturally occurring strains that harbored plasmids (55C1 and A20) were also effective recipients. In general, the strain of 8325 (P524) originally received from R. P. Novick gave the highest frequency of transduction although this was approached by some clones of 8325 into which we had transduced the C5 or the 524 plasmid. Twelve clones of 8325 transduced to streptomycin resistance with unirradiated phage raised on C5 did not become effective recipients for transduction of methicillin resistance. In another experiment, 8325 was exposed to 80 phage raised on C5 and, after serial dilution, was plated on nonselective medium. Of 25 colonies chosen at random, none were effective recipients in transduction of methicillin resistance. These experiments showed that recipient effectiveness was not a nonspecific consequence of exposure to transducing phage but was linked to transduction of the penicillinase plasmid. Plasmid-bearing strains of 8325 which were effective recipients retained this property, for the most part, after the plasmids were eliminated by treatment with ethidium bromide. The effectiveness of independently cured clones of 8325 (P524) or 8325 (Pc5) varied, but for any single clone was reasonably constant in repeated tests (Table 4). Two cured clones of 8325 (Pc5) failed to give any transductants and, in this respect, were indistinguishable from the original Thus, the cured strains usually differed as recipients from the original 8325 even though they retained none otf the plasmid markers for which we have tested. These included penicillinase production and resistance to arsenate, lead, zinc, bismuth, cadmium, and mercuric ions. These results suggested that 8325 was altered as the result of the introduction of these penicillinase plasmids in a fashion not regularly reversible by cure of the plasmids. A similar specificity of recipient-effectiveness was observed when methicillin resistance was transduced from 8325 (P524) by induced temperate phage or from C4 by 53 phage. The basis of this unusual phenomenon is under further study. It is worth noting that the plasmid-induced effectiveness of 8325 was not a direct consequence of the presence of a penicillinase linkage group alone since the introduction of either the B4 or the 55C1 chromosomal penicillinase linkage group alone did not confer effectiveness on TABLE 4. Strains N8325 (P524) and N8325 (Pc5) as recipients of methicillin resistancea Recipient No. of methicillin-resistant transductants (date of experiment) 5/6/70 5/18/70 5/20/ (P524) N8325 (P524)b N8325 (P524)b (Pc5) N8325 (PC5) N8325 (PC5)b a Transductions were performed with the same batch of phage. b Designates strains independently cured of their penicillinase plasmids by treatment with ethidium bromide. Furthermore, a small number of clones of 8325 bearing other transduced penicillinase plasmids were not effective recipients for transduction of methicillin resistance (Table 3). Properties of transductants. All of 28 methicillin-resistant transductants of 8325 (P524) that were examined retained the recipient's phage type, and all of 117 transductants retained susceptibility to streptomycin, kanamycin, tetracycline, erythromycin, and chloramphenicol. The sole exceptions were eight transductants exhibiting high-level resistance to methicillin (MIC > 400,ug of methicillin/ml). These were lysed by 53 phage only at 1,000 X RTD in contrast to the lysis at RTD of the other transductants and the recipient strain. Transductants grew relatively slowly on the methicillin selection plates at 30 C, becoming visible as pinpoint colonies in about 20 to 24 hr. At 40 hr the transductant colonies varied in diameter from 0.3 to 2 mm. In this respect they resembled the parental strains similarly cultivated on methicillin agar. Growth in the presence of ethidium bromide or acriflavine failed to cure methicillin resistance from C5 or from two strains of 8325 (P524) transduced to methicillin resistance, with the exception of a single methicillin-susceptible clone obtained from acriflavine treatment of C5 (Table 5). This clone differed from its parent in giving coarsely clumped growth in broth. We think it may well be mutant. Ethidium bromide treatment, but not acriflavine, cured the penicillinase plasmid from C5 and the two transductants as shown by recovery of Cd2+_susceptible clones. These experiments did not suggest that the gene for

7 1164 COHEN AND SWEENEY J. BACaERIOL. TABLE 5. Attempts at curing methicillin resistance No.ocolnies No. o cadium-no. of methicilclones Organism"a Curing treatment No. of colonies pno ọf cmiun- lin-susceptible replicaedsuseptibleclones 8325 (P524)b Ethidium bromide, 1 to 1.5 pg/ml 4, Acriflavine, 12.5 pg/ml (P624)c Ethidium bromide, 2 pg/ml Acriflavine, 12.5 pg/ml CS Ethidium bromide, 2 or 3 ;pg/ml Acriflavine, 12.5 pg/ml a Strains tested had been transduced to methicillin resistance from C5. b Strain with low-level methicillin resistance. c Strain with high-level methicillin resistance. methicillin resistance in these strains was located on a plasmid. A striling feature of the transductants was the great variation in their degree of methicillin resistance. Table 6 gives results of plate dilution titrations of 11 clones derived from one transduction. Their MIC varied from 12.5 to 400,ug of methicillin per ml. Table 6 also shows that most transductants resembled the parental strain in that a higher proportion of cells formed colonies in methicillin agar at 32 than 37 C. Table 7 shows that the methicillin MIC was not greatly changed in three of four transductants that were serially subcultured daily for 7 days in Heart Infusion Broth. Evidently their methicillin resistance was relatively stable. It will be noted that the methicillin resistance of transductants from strain C5 ranged from values much greater to much less than that of the parental organism. From the progeny of transduction of methicillin resistance from C5 to 8325 (P524), we selected one clone with a methicillin MIC appreciably higher and another lower than that of the parental C5. The methicillin resistance of these two transductants was transduced again by means of their ultraviolet-induced prophage to 8325 (P524). The MIC values of the progeny of the two cycles of transduction varied widely, from concentrations much lower to concentrations much higher than that of each parental strain (Table 8). DISCUSSION In their studies of transduction of methicillin resistance, Dornbusch, Hallander, and Lofquist used as vectors temperate phages which were grown first in lytic cycle after ultraviolet irradiation of their methicillin-resistant hosts and were then propagated on a methicillin-resistant donor strain (11). The methicillin-susceptible recipient strains were either wild-type or derived from methicillin-resistant strains after acriflavine treatment. They included both penicillinase-positive and -negative strains. Methicillin resistance was transduced at frequencies of 0.2 x 10-8 to 1.1 x 10-8 with no correspondence between multiplicity of infection and number of transductants. Methicillin resistance was transduced coordinately with enterotoxin formation and, in some cases,, hemolysin and penicillinase formation, and resistance to Hg2+ or Cd2+. Our methicillin-resistant strains, C4 and CS, did not hemolyze 1% sheep blood-agar, and we did not test for enterotoxin formation. We observed no cotransduction of methicillin resistance TABLE 6. Methicillin resistance of strain 8325 (P524) transductantsa Transductant designation Methicillin MIC1 32 C 37 C pg/mi pg/mi < > < >400 > > > <12.5 <12.5 aclones tested were progeny of one experiment made with various multiplicities of phage 80 propagated on C5. Clones were chosen to illustrate the range of resistance encountered, therefore they are not a random sample of the transductant population. bminimal inhibitory concentration.

8 VOL. 104, 1970 TRANSDUCTION OF METHICILLIN RESISTANCE 1165 and the P524 plasmid, whether the primary selection was for Cd2+ resistance or methicillin resistance. Furthermore, penicillinase production and methicillin-resistance were not cotransduced from host cells bearing a chromosomal penicillinase linkage group and a transduced marker for methicillin resistance. In our experiments, the big increase in frequency of transduction of methicillin resistance after ultraviolet irradiation of phage suggested a chromosomal location for this marker (2, 13). Dornbusch et al. obtained a high proportion of clones durably susceptible to methicillin after 2 days of exposure of their resistant strains to acriflavine (11). They concluded that methicillin resistance resided on a plasmid in their strains. These results are somewhat surprising since acriflavine has been relatively ineffective in curing recognized plasmids from staphylococci (23). Furthermore, the isolation of antibiotic-susceptible strains after prolonged treatment with acridines does not necessarily signify cure of a plasmid. The problems of interpretation are discussed in detail by Novick (reference 18, p ). In contrast to the results of Dornbusch et al. (11), our strains C4 and C5 gave only one questionable cure of methicillin resistance after acriflavine treatment. For the present, we can only record the differences between their results and ours and raise the possibility that there may be more than one genetic site for staphylococcal methicillin resistance, as has been found with penicillinase formation. In this connection it may be significant that C4 and CS were phage-type group III, as are most reported methicillin-resistant staphylococci. On the other hand, all but one (11) were of the strains used by Dornbusch et al. type 29 or untypable, a reaction given by many methicillin-resistant strains recently isolated in Europe (20). Dyke failed to transduce methicillin resistance by using phage 53 and three different recipients (12). He did not report the use of ultravioletirradiated phage nor any attempt to select transductants at temperatures other than 35 or 37 C. Since transduction transfers only a small segment of deoxyribonucleic acid, the reproducible transduction of methicillin resistance from strains CS and C4 suggests that either a single gene or possibly two or more closely linked genes is a sufficient determinant of this property. If the determinant should prove to be single, there would remain to be explained the wide variation in the degree of resistance of different progeny of the same transduction. The presence of transductants less resistant than their parent might be explained by segregation of multiple genes for TABLE 7. Stability of methicillin resistance in serially subcultured transductantsa Methicillin MIC Transductant designation From methicillin- From serial agar plate (12.5 subculture in pg/ml) broth > a Four clones of 8325 (Ps24) transduced to methicillin resistance by 80 phage propagated on C5 were serially subcultured seven times daily in Heart Infusion Broth at 37 C. Their methicillin minimal inhibitory concentration (MIC) values were determined from the progeny of the last subculture and also from the same clones which had been subcultured once at 32 C on methicillinagar (12.5,g/ml) and then stored at 4 C. TABLE 8. Methicillin MIC of donors and transductantse Methicillin MIC of transductants Clones selected by Methicillin Clones chosen at ability to grow on MIC of donors random methicillin plates (120 to 720 pg/ml) No. of Cnn No. of Coc clones Concn clones Concn pg/mi 8g/mi pg/ml >400 > >400 a Methicillin resistance was transduced from strain CS to 8325 (P524). From the transductants, two clones with the methicillin minimal inhibitory concentration (MIC) indicated in the left column were chosen to be donors in a second-stage transduction also to 8325 (P524). Twelve transductants were chosen at random for MIC determination after two further single-colony passages on agar containing 12.5,ug of methicillin per ml, with results indicated in the center column. All transductant colonies were replicated from the primary selection plates to plates containing graded concentrations of methicillin. Clones that replicated on medium containing 120 to 720 Mg of methicillin per ml were subcultured once from the master plate on medium containing 12.5 pg of methicillin per ml. Results of determinations of their MIC are given in the right-hand column. methicillin resistance. However, the derivation by a second transduction of progeny more resistant than the first stage transductant would not be readily explained on this basis. The simplest hypothesis is that one gene determines methi-

9 1166 COHEN AND SWEENEY J. BACrRIOL. cillin resistance but that the degree of its expression is determined by other unknown factors. These might be secondary mutations which modify the degree of expression of a primary methicillin-resistance mutation or, perhaps, some type of phenotypic variation. A possibly analogous finding is the observation that staphylococci transduced to resistance against streptomycin exhibited a variable degree of resistance (15). The difference in effectiveness of the strains tested as recipients for transduction of methicillin resistance is also difficult to explain. Both 8325 and N8325 absorbed transducing phage and expressed other transduced markers with equal facility. Their difference as recipients in transduction of methicillin resistance may have been caused by a type of host-modified restriction, active in 8325 and highly specific for the gene for methicillin resistance. Different genes in X phage are restricted at unequal rates by nonpermissive hosts (3, 26). Nevertheless, the lack of any evidence of restriction in 8325 of the other markers examined makes this hypothesis unlikely. The simplest explanation is that the gene for methicillin resistance was transduced into 8325 but failed to become integrated into the 8325 chromosome. This interpretation would suggest that there might be abortive transductants for methicillin resistance in 8325 that might produce microcolonies on the methicillin selection plates. We have found none, even on the plates incubated for 3 or 4 days. Other possibilities are that the gene for methicillin resistance was integrated in 8325 but not expressed or that it was lethal. In either case, transductants would not give rise to colonies on methicillin agar. Introduction of Pc5 or P524 into 8325 appeared to convey coordinately recipient effectiveness for transduction of methicillin resistance. The number of clones tested as yet by this rather tedious procedure is still small in terms of genetic experiments and requires supplementation to confirm the obligatory association of these properties. Recipient-effectiveness, once gained by 8325 through acquisition of the appropriate penicillinase plasmid was not lost, in most cases, after apparently complete elimination of the plasmid. We have no plausible explanation of this finding. Possibly the transduction of the penicillinase plasmid may have triggered some secondary genetic event, for example loss of a prophage, that affected recipient-effectiveness for methicillin resistance. Alternatively, the effective recipients may have retained some unrecognized fragment of the penicillinase plasmid. This hypothesis may find some support in observations of Watanabe and Ogata (27). They found that some drug-susceptible segregants derived from Salmonella typhimurium bearing an R factor became unusually effective recipients for the transduction of that R factor. They postulated that these unusual clones retained the RTF moiety of the R factor which rescued otherwise abortive transductants by recombination. Although we have not detected any plasmid markers for resistance to inorganic ions or penicillinase formation in our cured strains, this does not exclude the possibility that a portion of the plasmid with a less readily detectable function may have survived. In this connection we have investigated one plasmid-cured derivative of 8325 (P524) for evidence of a residual maintenance-compatibility locus and found none (21). Against the hypothesis of a surviving fragment of the plasmid is the fact that the methicillin resistance gene was apparently chromosomal in our strains and therefore need not be expected to have any unusual genetic homology with the penicillinase plasmid. Methicillin resistant staphylococci are almost always penicillinase positive. In addition, they are usually multiply resistant to other antibiotics and fall into a relatively restricted range of phage types. These associations have been ascribed by Parker and Hewitt to the fact that methicillin resistance confers a selective advantage not only against methicillin but also against penicillins (20). This resistance would be especially protective for small numbers of cells in low temperature sites, such as the nose or skin, where penicillinase would be less effective. Selection for methicillin resistance would affect primarily hospital strains of staphylococci that would also be likely to be resistant to other antibiotics. Our observations suggest another possibility. Methicillin resistance genes may be available to many strains of staphylococci through the medium of transducing phage but may become established only in a small proportion of strains that are effective recipients. ACKNOWLEDGMENTS We are indebted to Carole J. Gibson for excellent assistance. This investigation was supported by Public Health Service research grant Al from the National Institute of Allergy and Infectious Diseases and by a contribution from the Betty Jane Erman Schultz Fund. LITERATURE CITED 1. Annear, D The effect of temperature on resistance of Staphylococcus aureus to methicillin and some other antibiotics. Med. J. Aust. 1: Arber, W Transduction of chromosomal genes and episomes in Escherichia coli. Virology 11: Arber, W., and S. Linn DNA modification and restriction. Annu. Rev. Biochem. 38: Asheshov, E. H Chromosomal location of the genetic elements controlling penicillinase production in a strain of Staphylococcus aureus. Nature (London) 210:

10 VOL. 104, 1970 TRANSDUCTION OF METHICILLIN RESISTANCE Barber, M Naturally occurring methicillin-resistant staphylococci. J. Gen. Microbiol. 35: Benner, E. J., and V. Morthland Methicillin-resistant Staphylococcus aureus. New Engl. J. Med. 277: Blair, E., and R. E. 0. Williams Phage typing of staphylococci. Bull. W.H.O. 24: Bouanchaud, D. H., M. R. Scavizzi, and Y. A. Chabbert Elimination by ethidium bromide of antibiotic resistance in enterobacteria and staphylococci. J. Gen. Microbiol. 54: Cohen, S., and H. M. Sweeney Constitutive penicillinase formation in Staphylococcus aureus owing to a mutation unlinked to the penicillinase plasmid. J. Bacteriol. 95: Cohen, S., E. G. Vernon, and H. M. Sweeney Differential derepression of staphylococcal plasmid and chromosomal penicillinase genes by a class of unlinked chromosomal mutations (R2-). J. Bacteriol., 103: Dornbusch, K., H. 0. Hallander, and F. Lofquist Extrachromosomal control of methicillin resistance and toxin production in Staphylococcus aureus. J. Bacteriol. 98: Dyke, K. G. H Penicillinase production and intrinsic resistance to penicillins in methicillin-resistant cultures of Staphylococcus aureus. J. Med. Microbiol. 2: Garen, A., and N. D. Zinder Radiological evidence for partial genetic homology between bacteriophage and host bacteria. Virology 1: Gravenkemper, C. F., J. L. Brodie, and W. M. M. Kirby Resistance of coagulase-positive staphylococci to methicillin and oxacillin. J. Bacteriol. 89: Korman, R. Z., and D. T. Berman Genetic transduction with staphylophage. J. Bacteriol. 84: Novick, R. P Analysis by transduction of mutations affecting penicillinase formation in Staphylococcus aureus J. Gen. Microbiol. 33: Novick, R. P Properties of a cryptic high-frequency transducing phage in Staphylococcus aureus. Virology 33: Novick, R. P Extrachromosomal inheritance in bacteria. Bacteriol. Rev. 33: Novick, R. P., and C. Roth Plasmid-linked resistance to inorganic salts in Staphylococcus aureus. J. Bacteriol. 95: Parker, M. T., and J. H. Hewitt Methicillin resistance in Staphylococcus aureus. Lancet 1: Peyru, G., L. F. Wexler, and R. P. Novick Naturally occurring penicillinase plasmids in Staphylococcus aureus. J. Bacteriol. 98: Poston, S. M Cellular location of the genes controlling penicillinase production and resistance to streptomycin and tetracycline in a strain of Staphylococcus aureus. Nature (London). 210: Richmond, M. H The plasmids of Staphylococcus aureus and their relation to other extrachromosomal elements in bacteria, p. 43. In A. H. Rose and J. F. Wilkinson (ed.), Advances in microbial physiology, vol. 2. Academic Press Inc., New York. 24. Sutherland, R., and G. N. Rolinson Characteristics of methicillin-resistant staphylococci. J. Bacteriol. 87: Sweeney, H. M., and S. Cohen Wild-type strain of Staphylococcus aureus containing two genetic linkage groups for penicillinase production. J. Bacteriol. 96: Terzi, M Physiology of bacteriophage X in the restrictive host. J. Mol. Biol. 34: Watanabe, T., and Y. Ogata Abortive transduction of resistance factor by bacteriophage P22 in Salmonella typhimwrtum. J. Bacteriol. 102: