TRANSDUCTION OF ANTIBIOTIC RESISTANCE IN STAPHYLOCOCCUS AUREUS

1107

administration of sodium phosphate. The only side-effect of this therapy was mild to moderate diarrhoea. This

therapy for idiopathic hypercalcuria is cheap, effective, and relatively non-toxic.

mode of

This work was supported in part by a grant from the National Institute of Arthritis and Metabolism, National Institutes of Health (AM-03967) and the Fund for Research and Teaching, Department of Nutrition, Harvard School of Public Health. Requests for reprints should be addressed to D. S. B., Robert Breck Brigham Hospital, Parker Hill Avenue, Boston, Massachusetts. REFERENCES

Albright, F., Reifenstein, E. C., Jr. (1948) in The Parathyroid Glands and Metabolic Bone Disease (edited by Williams and Wilkins Co.); p. 241. Baltimore.

Fiske, C. H., Subbarow, Y. (1925) J. biol. Chem. 66, 375. Fleisch, H., Bisaz, S. (1962) Am. J. Physiol. 203, 671. Folin, A., Wu, H. (1919) J. biol. Chem. 38, 81. Goldsmith, R. S., Ingbar, S. H. (1966) New Engl. J. Med. 274, 1. Hellman, D. E., Au, W. Y. W., Bartter, F. C. (1965) Am. J. Physiol. 209, 643. Henneman, P. H., Benedict, P. H., Forbes, A. P., Dudley, H. R. (1958) New Engl. J. Med. 259, 802. Howard, J. E. (1961) J. clin. Endocr. Metab. 21, 1254. Kessler, G., Wolfman, M. (1964) Clin. Chem. 10, 686. Lemann, J. Lennon, E. J., Goodman, A. D., Litzow, J. R., Relman, A. S.

(1965) J. clin.

Invest.

44, 507.

MacInyre, I., Boss, S., Thompson, I. (1963) Nature, Lond. 198, 1058. Shorr, E. (1945) J. Urol. 53, 507.

TRANSDUCTION OF ANTIBIOTIC RESISTANCE IN STAPHYLOCOCCUS AUREUS SHEILA MCDONALD M.D. Edin. LECTURER IN

UNIVERSITY OF EDINBURGH

Methods

Organisms Antibiotic-sensitive

staphylococci representing the main selected from clinical sources. Strains which were not associated with the hospital environment were collected by culturing nasal swabs from science students and preclinical medical students. A small number of phage-type 71 staphylococci were kindly supplied by Dr. M. P. Jevons. Media Tryptone-soya agar (Oxoid) or brain-heart-infusion agar (brain-heart-infusion broth, Difco, solidified with 1-2% Davis agar) was used for the poured plates and Oxoid no. 2 nutrient broth or brain-heart-infusion broth (Difco) was the fluid medium.

phage

groups

were

Mercuric-chloride-resistance Test The method described by Moore (1960) was used, but each test strain was spotted on a series of peptone-water agar plates containing different concentrations of mercuric chloride. A mercury-sensitive and a mercury-resistant strain were inoculated on each plate and the readings were taken from the plate on which these control strains gave the correct result. The controls were necessary since minor variations between batches of media influenced the results.

Phage-typing Phage-typing of the donor and recipient strains of staphylococci was carried out as described by Williams and Rippon (1952). The typing phages used were supplied by the Staphylococcus Reference Laboratory, Colindale, London

STAPHYLOCOCCi and the enterobacteriacex are groups of in which the emergence of antibiotic-resistant N.W.9. strains has become a serious problem. The investigation Transduction of Tetracycline Resistance Standard cultures of the recipient strains were prepared by of the resistance factors in the enterobacteriaceae has indicated a new and unexpected mechanism for the adding 2-5 ml. of an 18-hour broth culture to 50 ml. of broth 37°C for 3 hours. These cultures contained development of antibiotic-resistant strains in this group and shaking at109 viable bacteria per ml. 1 ml. of a standard (Datta 1962, Anderson 1965). It may be that transduction approximately culture of the recipient strain was mixed with 0-5 ml. of phage plays a similar part in the emergence of epidemic resistant filtrate to give a mixture containing slightly fewer phage strains of Staphylococcus aureus. particles than bacteria. The mixture was made up to 5 ml. with Transduction in Staph. aureus was described by Ritz nutrient broth. A control mixture of an equal number of and Baldwin (1958) who showed that typing phage 52 staphylococci in broth without phage was also set up. 0-5 ml. could carry the marker for penicillinase production from samples of these mixtures were mixed with 2 ml. of 0-7% agar penicillin-resistant strains of staphylococci to penicillin- at 46°C and layered over plates of tryptone-soya agar that sensitive staphylococci. Since then many transducible contained 12-5 g. of tetracycline per ml.; this concentration characters have been described including resistance to prevented growth of the untransduced recipient cells. All were incubated at 37°C for 48 hours. Colonies from the streptomycin (Morse 1959), tetracycline (Collins and plates transduction plates in each successful experiment were subMcDonald 1962), the macrolide group of antibiotics cultured to antibiotic-free plates. These colonies were later (Pattee and Baldwin 1962), and chloramphenicol (Collins transferred to broth for testing of antibiotic sensitivity and and Roy 1963). A property of many staphylococci resistance to mercuric chloride. responsible for epidemics of hospital infection is resistance Transduction of Penicillin Resistance to mercuric chloride (Moore 1960), and joint transduction 0-5 ml. of a standard culture of the recipient strain was mixed of this character with penicillinase production was demonwith 1 ml. of the phage filtrate. The mixture was made up to strated by Richmond and John (1964). These findings 2-5 ml. with nutrient broth. A control mixture of an equal show that characters commonly associated with epidemic number of staphylococci in broth without phage was set up. strains of staphylococci are transducible in the laboratory. These mixtures were incubated on a shaker at 37°C for 1 hour If transduction does play a part in the emergence of to allow adsorption of the phage to take place. Then the epidemic strains in vivo it might be expected that mixtures were centrifuged and the deposit from each was antibiotic-sensitive staphylococci of the same phage-type resuspended in 1 ml. of brain-heart-infusion broth. 0-2 ml. samples of these suspensions were spread over the surface of as epidemic strains would be more readily transduced than other antibiotic-sensitive staphylococci. Previous work on brain-heart-infusion agar plates containing 0-06 g. or 0-12 g. ml. The plates were incubated at 37°C for the transduction of tetracycline resistance in Staph. aureus penicillin per 48 hours. Colonies from the transduction plates in each (Collins and McDonald 1962) had indicated that the experiment were subcultured to antibiotic-free plates. Later, phage-types of the tetracycline-resistant donor strains and these colonies were transferred to broth for testing of antibiotic the tetracycline-sensitive recipient strains must be similar sensitivity, mercuric-chloride sensitivity, and penicillinase for transduction to take place. To confirm this finding production.

organisms

’,

BACTERIOLOGY,

and to determine if the ability to be transduced was restricted to certain phage-types, antibiotic-sensitive staphylococci representing the three main phage groups have been collected and tested as recipient strains in the transduction of resistance to tetracycline and penicillin.

1108 I-TRANSDUCTION OF TETRACYCLINE RESISTANCE BY PHAGE TO STAPHYLOCOCCI OF THE THREE MAIN PHAGE GROUPS ISOLATED FROM CLINICAL SOURCES OR FROM NASAL CARRIERS UNASSOCIATED WITH A HOSPITAL ENVIRONMENT

TABLE

AK72

Penicillinase Production This was demonstrated by a modification of the screening method described by Foley and Perret (1962). The methods used for antibiotic-sensitivity tests, for the induction of the transducing phages AK72 and 15844 from naturally lysogenic antibiotic-resistant staphylococci, and for the titration of these phages have been described elsewhere (Collins and McDonald 1962). ,

transduce any of the 46 strains tested from group II. Since these strains had been isolated from clinical sources and from healthy carriers, they included a variety of group-n staphylococci but there was a subgroup of group 11 that was not well represented in the collection. Staphylococci of phage-type 71 differ in some respects from other members of group 11; they have been associated with outbreaks of impetigo and are frequently resistant to penicillin. But attempts to transduce resistance to 14 tetracycline-sensitive type-71 staphylococci gave entirely negative results so that these strains resemble other group-ll staphylococci in their inability to be transduced by phages induced from group i or group III donor staphylococci. Table ill records the changes in the recipient strains after transduction of tetracycline resistance. These transductants were subcultured from each experiment in

failure

to

TABLE II-TRANSDUCTION OF TETRACYCLINE RESISTANCE BY PHAGE 15844 TO STAPHYLOCOCCI OF THE THREE MAIN PHAGE GROUPS ISOLATED FROM CLINICAL SOURCES OR FROM NASAL CARRIERS UNASSOCIATED WITH A HOSPITAL ENVIRONMENT

Results

Transduction of each recipient strain was attempted with phage AK72 induced from a staphylococcus of phage-groupI and with phage 15844 induced from a staphylococcus of phage-group III. Both these donor staphylococci produced penicillinase and were resistant to 100 µg. penicillin per ml., to 50 µg. tetracycline per ml., and to mercuric chloride. All the recipient strains used in the experiments in which the transduction of tetracycline resistance was attempted were sensitive to 6 g. tetracycline per ml. The recipient strains selected for the transduction of penicillin resistance did not produce penicillinase and were sensitive to 0-06 µg. penicillin per ml. and to mercuric chloride. Transduction of Tetracycline Resistance The results of the transduction experiments with phage AK72 are recorded in table i. These results show that this phage was able to transduce tetracycline resistance to 51 out of 62 group-i staphylococci and to 10 out of 36 group-in staphylococci. The ability to be transduced by phage AK72 was not associated with any particular phage-type within groups I or ill. No positive results were obtained with any of the 46 staphylococci tested from group 11.

The results of the transduction experiments with phage 15844 are recorded in table 11 and show that only 29 out of the 62 group-I staphylococci tested were transduced in comparison with the 51 strains transduced by phage AK72. 21 of the recipient strains that gave positive results with phage 15844 were sensitive to typing phage 29 and it seems that this phage-type was associated with the ability to be transduced by phage 15844. A greater number of group-m staphylococci, 17 out of 36 strains, were transduced by phage 15844 which had been induced from a group-m donor strain, than by phage AK72 which had been induced from a group-I donor strain. It seems that, although similarity between the phage-type of the donor and recipient strain is not essential for transduction to take place, such a similarity increases the likelihood of transduction. The attempts to transduce group-II staphylococci with phage 15844 gave the same negative results as had been obtained with phage AK72. The most striking result of these experiments is the

*

Minimum inhibitory concentration > 50

µg.

tetracycline per ml.

which the recipient strain had been sensitive to both penicillin and tetracycline. All of the 640 transductants tested were resistant to more than 50 µg. tetracycline per ml. There was no evidence of co-transduction of the marker for tetracycline resistance and the markers for resistance to penicillin or mercuric chloride. Transduction of Penicillin Resistance The results in table IV show that the majority of penicillin-sensitive group-l staphylococci that gave positive results when phage AK72 was used to transduce tetracycline resistance gave positive results when penicillin resistance was the marker; phage AK72 transduced penicillin resistance to 35 out of 46 group-I strains. The number of penicillin-rensitive group-lll strains isolated was so small that it is difficult to generalise from the findings; phage AK72 transduced penicillin resistance to 3 out of the 9 strains available for testing. Negative results were obtained with 31 group-ll staphylococci

tested. The results of the experiments with phage 15844 are recorded in table v. There is a marked difference in the ability of the two phages to transduce penicillin resistance. Phage 15844 transduced this marker to only 2 out of 55

1109 strains of staphylococci tested from groups i and III. It is true that only a few penicillin-sensitive group-in staphylococci were available for testing, but phage 15844 was shown to transduce tetracycline resistance frequently to both groupI and group in staphylococci. It may be that a closer relation between donor and recipient strain is required for the transduction of penicillin resistance than for the transduction of tetracycline resistance. The donor strains of staphylococci from which the transducing phages were induced were investigated by Dr. M. H. Richmond. He found that each of these strains produced penicillinase type A. Therefore the difference in the ability of phages AK72 and 15844 to transduce penicillin

resistance could

TABLE V-TRANSDUCTION OF PENICILLIN RESISTANCE BY PHAGE

15844

TO STAPHYLOCOCCI OF THE THREE MAIN PHAGE GROUPS ISOLATED FROM

CLINICAL SOURCES OR FROM NASAL CARRIERS UNASSOCIATED WITH A HOSPITAL ENVIRONMENT

be explained by a difference in the produced by the staphylococci from which they had been derived. The inability of group-n staphylococci to be transduced, already noted when tetracycline resistance was used as the marker, was also found when attempts were made to transduce penicillin resistance with phages induced from groupI or group III donor not

enzyme

strains. Table VI records the changes in the recipient strains after transduction of penicillin resistance. These transductants were subcultured from each experiment in TABLE IV-TRANSDUCTION OF PENICILLIN RESISTANCE BY PHAGE AK72 TO STAPHYLOCOCCI OF THE THREE MAIN PHAGE GROUPS ISOLATED FROM CLINICAL SOURCES OR FROM NASAL CARRIERS UNASSOCIATED WITH A HOSPITAL ENVIRONMENT

*

Inadequate numbers

to

allow percentage estimations.

which the recipient strain had been sensitive to tetracycline as well as to penicillin. Only 355 of the colonies tested were resistant to more than 100 µg. penicillin per ml. It seems likely that the penicillin-sensitive colonies were picked from the transduction plates containing 0-06 µg. penicillin per ml. On these plates there was always a background growth of penicillin-sensitive colonies that could have contaminated subcultures. Of the 355 penicillin-resistant transductants tested for penicillinase production, 344 gave unequivocally positive results. Thus the acquisition of penicillin resistance generally depended on the acquisition of the ability to produce penicillinase. It was confirmed that co-transduction of the markers for penicillinase production and resistance to mercuric chloride takes place. Transduction of the marker for tetracycline resistance did not accompany transduction of the markers for mercuric chloride resistance or

penicillinase production. Discussion The aim of the present work has been to investigate the occurrence of transduction among staphylococci of different phage-types and isolated from different sources.

The results obtained after testing representative collections of staphylococci show that

Phage

phage AK72, which resembles transduce tetracycline and penicillin resistthe majority of group-i staphylococci isolated

80,

ance to

can

* Minimum colonies

inhibitory concentration were

shown to be

> 100

µg.

penicillin

per

ml.; 344

penicillinase producers

from clinical or nonclinical sources. Differences in the phage-types of these recipients and in their immunity to lysis by the transducing phage were not related to their ability to be transduced. A higher proportion of the strains from nonclinical sources gave positive results but the number of strains in each group was small and the difference was found to be barely significant. Fewer of the group-ill strains were transduced by phage AK72 and there was no difference in results between the strains isolated from clinical and nonclinical sources in this group. From these findings it seems that a phage is most likely to transduce recipient strains that are sensitive to lysis by closely related phages. However, phase 15844, although resembling a group-m typing phage, also transduced more group-l staphylococci than group-m staphylococci. The ability of phage 15844 to transduce penicillinase production was very limited. This finding is similar to that obtained by Pattee and Baldwin (1961) with phage 53. The most outstanding result with both phages was the failure to transduce any of the group-ll recipients tested. The failure to transduce antibiotic resistance to group-ll staphylococci has been noted by other workers. A few positive results with such strains have been recorded by Mitsuhashi et al. (1965) but the number of tranductants recovered were extremely small. Richmond (1965a) has shown that penicillinase type B produced by group-ll staphylococci differs from the penicillinases produced by group I and group ill strains. This could account for the failure to transduce penicillin resistance to group-ll strains by phages derived from staphylococci which produce type A

‘

penicillinase. Direct proof that transduction plays a development of antibiotic-resistant strains

role in the of staphylococci is difficult to obtain but the results of the present work support such a theory. The antibiotic-resistant strains associated with epidemics of staphylococcal infection have phage-typing patterns belonging to the 52/52A/80/81 complex of group i or to group III. These are the strains that have been shown to be readily trans-

1110 The very small number of antibiotic-resistant staphylococci found in group 11 could be a result of their inability to be transduced. The group II staphylococci that are found to be resistant to penicillin usually belong to phage-type 71. Type-71 staphylococci have been associated with outbreaks of impetigo and have certain properties that differentiate them from other group-ll strains (Parker 1958). However, these type-71 strains do not show multiple antibiotic resistance and are not associated with hospital infections. I have tested these as recipient strains in transduction experiments, and they gave the same negative results as did other group-ll strains. It is evident, therefore that staphylococci of the phage-types associated with outbreaks of staphylococcal infection in hospitals are also those which acquire the ability to produce penicillinase and resistance to other antibiotics by transduction. Richmond (1965b) described several different extrachromosomal elements or plasmids present in staphylococci. These plasmids carry other genetic markers as well as the penicillinase marker and he has divided the penicillinase plasmids in staphylococci of phage groupsI and III into two groups depending on the compatibility of the plasmids with each other. The penicillinase marker that has been studied in the present work is linked to the marker for mercury resistance, but not to the markers for resistance to other antibiotics, and type A penicillinase is produced. If this penicillinase marker is associated with a plasmid then these characters indicate that the plasmid would correspond to the &agr; plasmid described by Richmond (1965b). It would be of value to determine if any incompatibility exists between the penicillinase plasmids and extrachromosomal elements carrying other antibiotic-resistance markers. It may be that the presence of a particular penicillinase plasmid in a staphylococcus prevents that organism from acquiring other antibiotic-resistance markers. If this mechanism were found to operate it would explain why some strains of staphylococci remain sensitive to most antibiotics while other strains are well known to develop multiple antibiotic resistance.

Preliminary Communications

ducible.

AMENORRHŒA AFTER TREATMENT WITH ORAL CONTRACEPTIVES

period of 24 months eighty-six patients with secondary arnenorrhoea of at least 12 months duration were investigated sufficiently well for a firm diagnosis to be IN

a

Nine gave histories of the onset of amenorrhoea after stopping oral contraceptives. We cannot yet be certain whether the association is causal or casual, and we present our findings to stimulate future clarification.

made.

PATIENTS AND METHODS

All nine

patients had been referred for investigation of

clinical details are shown in table I. None organic disease. Urinary 17-oxosteroids and 17-hydroxysteroids were normal. One woman (patient 3) was mildly hirsute and had had acne since the onset of amenorrhoea. Another (patient 5) had galactorrhoea. X-rays of skull, pituitary fossa, and chest were normal in all patients, and all were euthyroid. There was no evidence of local pelvic lesions and curettage revealed either proliferative endometrium or nothing. Withdrawal bleeding could be induced by the administration of exogenous gonadal steroids. Seven of the nine patients underwent diagnostic gonadotrophin stimulation. The technique used has been described previously.’-2 Urinary oestrone was assayed by the method of Brown.3

amenorrhoea;

some

had evidence of

RESULTS

The results of diagnostic shown in table 11.

gonadotrophin stimulation are

COMMENT

Although amenorrhoea during treatment with oral contraceptive is well recognised, it has not, in recent reviews,4been described after treatment. None of the nine patients in this series had amenorrhoea during treatment. It is usually stated that after cessation of treatment the patient’s pretreatment cycle will be re-established promptly. Goldzieher et al.6 stated " ... that the disruption of normal pituitary-ovarian relationships ... is immediately reversible even after 2 years of cyclic treatment". Experience in this department is that Summary patients always return rapidly to a normal ovulatory The ability of various staphylococci to acquire resistance steroid excretion pattern.’8 to penicillin and tetracycline by transduction was investiR. P. Br. med. J. 1964, ii, 1115. gated. A phage induced from a staphylococcus belonging 2.1. Shearman, Shearman, R. P., Cox, R. I. Obstetl. gynec. Surv. 1966, 21, 1. to phage-group I transduced tetracycline resistance to 51 3. Brown, J. B. Personal communication, 1963. 4. Handbook on Oral Contraception (edited by E. Meares). London, 1965. of 62 group-i staphylococci and to 10 of 36 group-in 5. Shearman, R. P. Med. J. Aust. 1965, i, 68. staphylococci. This phage also transduced penicillin 6. Goldzieher, J. W., Rice-Wray, E., Schulz-Contreras, M., ArandaRosell, A. Am. J. Obstet. Gynec. 1962, 84, 1474. resistance to 38 of 55 staphylococci from groups I and III. 7. Shearman, R. P. Lancet, 1963, i, 97. A different phage induced from a group-III staphylo8. Shearman, R. P. ibid. 1964, ii, 557. coccus transduced tetracycline resistance to 29 of the DR. MCDONALD: REFERENCES group- staphylococci and to 17 of the group-in staphylococci but transduced penicillin resistance to only 2 of 55 Anderson, E. S. (1965) Br. med. J. ii, 1289. McDonald, S. (1962) J. Path. Bact. 83, 399. staphylococci from groups i and III. Both phages failed Collins, A. M., T. E. (1963) Can. J. Microbiol. 9, 541. — Roy, to transduce any of 60 strains belonging to group n. Datta, N. (1962) J. Hyg., Camb. 60, 301. These results support the hypothesis that transduction Foley, J. M., Perret, C. J. (1962) Nature, Lond. 195, 287. Mitsuhashi, S., Oshima, H., Kawaharada, U., Hashimoto, H. (1965) J. Bact. may play a part in the emergence of antibiotic-resistant 89, 967. strains of staphylococci in nature. B. 453. investigation formed part of a thesis submitted for the degree of M.D. of Edinburgh University. I thank Dr. M. P. Jevons of the Cross-Infection Reference Laboratory, Colindale, London N.W.9, for sending me a number of phage-type 71 staphylococci and to Dr. M. H. Richmond of the Department of Molecular Biology, Edinburgh University, for determining the type of penicillinase produced by certain of the staphylococci used in this work. Pfize Ltd. kindly provided a supply of tetracycline. This

References

at

foot of next

column

Moore, (1960) Lancet, ii, Morse, M. L. (1959) Proc. natn. Acad. Sci., U.S.A. 45, 722. Parker, M. T. (1958) J. Hyg., Camb. 56, 238. Pattee, P. A., Baldwin, J. N. (1961) J. Bact. 82, 875. (1962) ibid. 84, 1049. Richmond, M. H. (1965a) Biochem. J. 94, 584. (1965b) Br. med. Bull. 21, 260. - John, M. (1964) Nature, Lond. 202, 1361. Ritz, H. L., Baldwin, J. N. (1958) Bact. Proc. p. 40. Williams, R. E. O., Rippon, J. E. (1952) J. Hyg., Camb. 50, 320. -

-

-