Protoplasts, Spheroplasts, and L-Forms

Transcription

1 JOURNAL OF BACTERIOLOGY, Sept. 1970, p Copyright American Society for Microbiology Vol. 103, No. 3 Printed in U.S.A. Reversion to the Streptococcal State of Enterococcal Protoplasts, Spheroplasts, and L-Forms JAMES R. KING' AND HARRY GOODER Department ofbacteriology and Immunology, School ofmedicine, University ofnorth Carolina, Chapel Hill Received for publication 10 June 1970 A method is described for predetermining whether lysozyme-damaged enterococci grow as either the parent strain or as L-forms. Organisms treated with lysozyme grew as L-forms on media solidified with low concentrations of agar, or the damaged cells grew as streptococci on media solidified with high concentrations of gelatin. After induction, some of the L-forms reverted to the parent strain, but most did not during three routine subcultures. Continued spontaneous reversion occurred through approximately 30 subcultures after induction. However, subsequent progeny did not revert, even when subjected to conditions such as the gelatin medium which strongly favors growth in the streptococcal phase. Landman and Halle (9) reported the quantitative mass reversion of Bacillus subtilis protoplasts and L-forms to the bacillary state. In the B. subtilis system, reversion occurred when hard agar (2.0 to 2.5% agar) or gelatin (15 to 35%) was used as the solidifying agent in the medium. Previous reports of L-form reversion were concerned with spontaneous sporadic reversion, not under quantitative control exerted by the investigator. In other systems where the L-forms were induced by antibiotics, reversion occurred upon the removal of the inducing agent, e.g., when transferred in the absence of penicillin, B-type L-forms of Proteus quantitatively reverted in the laboratories of Dienes and Weinberger (3) and Altenbern (1). More specifically, the reversion of an occasional sporadic colony of group A streptococcal L-forms was reported by Crawford, Frank, and Sullivan (2) and Freimer, Krause, and McCarty (4), but mass reversion of all L-form strains was not found. Although the molecular events in re-formation of wall probably differ in protoplast systems compared to systems where an inhibitor of wall formation is removed from an unstable L-form culture, the term reversion has been applied to all systems. It is used here without any connotation as to the mechanism involved in re-formation of cell wall. We have developed a system for conversion of enterococci to L-forms by the action of lysozyme (7). Such L-forms could be easily subcultured, and their progeny would continue to grow as L- I Present address: Laboratory of Microbiology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Md phase variants with occasional spontaneous revertants. We were interested in developing a system in which a significant fraction of either the lysozyme-damaged enterococci or their established L-forms would revert to the parent coccus. MATERIALS AND METHODS Preparation of lysozyme-damaged streptococci. Streptococcusfaecium strain F24 and S. faecalis strain El were grown at 37 C overnight (stationary phase) in Trypticase Soy broth (BBL, Division of Bioquest Laboratories, Cockeysville, Md.). Cells from the cultures were harvested and washed three times in distilled water. Polyethylene glycol 4000 (8%, w/v) in ph 7.1 tris(hydroxymethyl)arninomethane-chloride buffer (0.01 M, ionic strength 0.01) was used as an osmotic stabilizer during lysozyme treatment. The buffer-stabilizer mixture will be abbreviated PEG. Mixtures of cells and lysozyme (Nutritional Biochemicals Corp., Cleveland, Ohio) were prepared to contain 2 X 109 streptococcal colony-forming units (CFU) and 200 jag of lysozyme per ml. The final turbidity of such mixtures was approximately 250 Klett units (Klett-Summerson photoelectric colorimeter equipped with a no. 54 filter). Strain F24 protoplasts were formed after 2 hr of incubation at 37 C. Under identical conditions, damage to the cell wall of strain El was less intense, and spheroplasts were formed (7). Composition of growth media. The L-form growth medium (LGM) was Tryptone Soy agar (Consolidated Laboratories, Chicago, Ill.) containing 0.43 M NH4Cl, 0.5% (w/v) additional glucose, and 2% (v/v) horse serum (inactivated at 56 C for 30 min and Seitz filtered). When gelatin (Eastman Kodak Co., Rochester, N.Y., purified pigskin, Batch , or Difco, Detroit, Mich.) was used as the solidifying 692

2 VOL. 103, 1970 REVERSION OF PROTOPLASTS AND L-FORMS 693 agent in the revertant growth medium (RGM), the gelatin was added to Tryptone Soy broth containing 0.43 M NH,CI, 0.5% (w/v) additional glucose, and 2% (v/v) inactivated horse serum. Propagation of stock L-form cultures. The stable L-forms of S. faecium F24 and S. zymogenes 30 were generally maintained by daily subculture in Brucella broth (Albimi Laboratories, Flushing, N.Y.) containing 0.43 M NH4CI and 0.5% (w/v) glucose. These L-forms were also routinely subcultured on LGM. Viable counts of lysozyme-damaged streptococci. Samples (0.5 ml) were removed from the lysozymedamaged cell suspensions and were diluted in PEG. After 10-fold serial dilution, 0.05-ml samples of appropriate dilutions were spread onto the surface of LGM or RGM. Plates of medium solidified with agar (LGM) or gelatin (RGM) were incubated inside plastic bags at 37 or 25 C, respectively. RESULTS Reversion of lysozyme-damaged enterococci. Our initial attempts at reversion of S. faecium F24 protoplasts were made with LGM containing hard agar (2.0 to 2.5%). The hard agar technique of Landman and Halle (9) was tested, but the number of revertants from such a suspension was low. On this medium, visual differentiation of L-forms and young streptococcal colonies was virtually impossible. Landman and Halle also used medium with high gelatin concentrations to obtain mass reversion. Strain F24 protoplasts were plated onto RGM containing various concentrations of gelatin (Difco), added in increments of 2% in the range 15 to 25%, as a solidifying agent. The results of a typical experiment are shown in Fig. 1. No L-forms of normal morphology developed on the gelatin medium. No increase in streptococcal colony count was seen at 15 and 17% gelatin, but the colony count rose between 19 and 23 % to a plateau at 23 %. The maximum recovery of streptococci in the experiment illustrated was 2.8 x 108 CFU/ml on 23% gelatin (Difco), the threshold gelatin concentration. When samples from the same protoplast suspension were plated on LGM, 9.1 X 108 L-form CFU/ml were found. Thus, in this series of experiments, approximately 30% of the protoplasts tested were recovered as L- forms on LGM and 10% were recovered as streptococci on RGM. The results of a similar experiment with strain El lysozyme-damaged cells (spheroplasts) are shown in Fig. 2. Essentially similar results were found with this strain, except that the threshold gelatin concentration of maximum recovery of revertants was 25%. The graph of strain El revertants versus gelatin (Difco) concentration was approximately straight, instead of sigmoidal, throughout the 15 to 23% LOGO NUMBER REVERTANTdCFU per ML) % GELATIN(DIFCO) IN THE MEDIUM FIG. 1. Effect of concentration of gelatin on reversion of lysozyme-damaged Streptococcus faecium F24. A suspension of enterococci was treated with lysozyme in PEG, and samples were plated out onto medium solidified with gelatin. LOG* NUMBER REVERTANT (CFU per ML) # *// % GELATIN(DIFCO) IN THE ME[ DIUM Effect of concentration of gelatin on rever- FIG. 2. sion of lysozyme-damaged Streptococcus faecalis El. A suspension of enterococci was treated with lysozyme in PEG, and samples were plated out onto medium solidified with gelatin. gelatin concentration range, as was the case with strain F24. A similar series of experiments was performed using gelatin obtained from Eastman Kodak.

3 694 KING AND GOODER J. BACTERIOL. A suspension of strain F24 protoplasts was plated onto RGM containing various concentrations of Eastman Kodak gelatin. The results from such an experiment are shown in Fig. 3. The highest recovery of revertants was at 35 % gelatin (Eastman Kodak), the highest concentration tested. The same type of experiment was performed with strain El (Fig. 4). In each case, approximately 10% of the streptococcal CFU was recovered after lysozyme treatment by plating on RGM. In some experiments, concentrations of gelatin above 35% were used, but such media were very difficult to handle in the molten state because of their viscosity. Furthermore, the inoculum could not be spread evenly over the surface of these gelatin plates. In the results shown (Fig. 4), colony counts of plates containing 40% gelatin were obtained, but the value is questionable. Since the maximal recovery of revertants could be obtained with gelatin (Difco) at much lower gelatin concentrations, 25% gelatin (Difco) was used in further studies requiring RGM. Other series of experiments were performed at the 25 % Difco gelatin concentration, to assess the recovery as streptococci on RGM as compared to the number of L-forms on LGM from the same protoplast suspension. In a series of seven similar experiments in which 2 X 109 CFU of strain F24 were converted to protoplasts, the counts as LOG ONUMBER REVERTANT(CFU per ML) , I -I % GELATIN (EASTMAN) IN THE MEDIUM FIG. 3. Effect ofconcentration ofgelatin on reversion of lysozyme-damaged Streptococcus faecium F24. A suspension of enterococci was treated with lysozyme in PEG, and samples were plated out onto medium solidified with gelatin. LOGIONUMBER REVERTANT (CFU per ML) % GELATIN (EASTMAN) IN THE MEDIUM Effect ofconcentration ofgelatin on reversion FIG. 4. of lysozyme-damaged Streptococcus faecalis El. A suspension of enterococci was treated with lysozyme in PEG, and samples were plated out onto medium solidified with gelatin. L-forms on LGM were 3.5 X 108, 2.0 x 108, 9.5 x 108, 2.2 x 108, 3.0 X 108, 10.0 X 108 and 12.0 X 108; the counts as streptococci on RGM were 0.2 X 109,1.5 x 109, 3.1 x 109, 2.0 x 109, 0.9 X 109, 3.0 X 109, and 2.6 X 109. In this series, approximately 30% of the protoplasts were recovered as L-forms and more than 90% of the protoplasts were recovered as streptococci. Similar results from a smaller series of experiments were obtained with strain El after lysozyme treatment. Similar experiment-to-experiment variation was reported with the reversion of B. subtilis protoplasts (10). Reversion of stable L-forms. A lysozyme-induced S. faecium F24 L-form was maintained by repeated subculture on LGM for approximately 1 year. During that time, occasional sporadic reversion occurred in single L-forms, but no attempts were made to quantitate the rate of spontaneous reversion. Beyond 1 year after induction, the L-form did not revert on routine subculture, regardless of the medium used. The L-form was further subcultured for 1.5 to 2 years, and during that time the L-form was successively adapted to grow vigorously in the absence of horse serum and agar. Adaptation of the L-form was slow and required over 100 subcultures before turbid growth in broth was achieved. Numerous attempts to adapt freshly induced L-forms of enterococci, derived from lysozyme-treated cells, to grow in broth have failed. Samples of the stable L-form described above

4 VOL. 103, 1970 REVERSION OF PROTOPLASTS AND L-FORMS 695 were inoculated onto RGM, the effective medium for reversion of lysozyme-damaged cells. Similarly, samples from the same cultures were plated onto LGM containing 2.0 to 2.5% agar instead of the usual 1.5 % agar. There was no easily visible growth on RGM, and visual interpretation of the growth obtained on hard agar LGM was difficult. Any suspected revertant was tested by subculture onto LGM and to routine bacteriological media. Each of the suspected revertants subcultured only as the L-form. In addition to strain F24 L-forms, S. zymogenes 30 stable L- forms (courtesy of W. R. Maxted) were tested. In numerous attempts, neither of the stable L- forms reverted when samples of either strain were plated out on medium solidified with either hard agar or gelatin. Thus, lysozyme-damaged streptococci would revert on gelatin medium, but stable L-forms would not. It was of interest to look at the reversion of freshly induced L-forms when subcultured on their normal growth medium. Reversion of freshly induced L-forms during routine subculture. Petri plates of LGM containing approximately 200 L-forms of strain F24 and strain El were induced by the action of lysozyme (7). These colonies were subcultured three times in succession by the replica plating technique. The revertants were monitored after each subculture by replica plating from the most recent subculture onto routine bacteriological media (bile esculin medium to test for the typical group D streptococcal reaction and blood-agar to test for alpha hemolysis). Three subcultures by replica plating were chosen, to give ample opportunity for any streptococci hidden in an L-form to grow. After three subcultures, the revertant colonies were counted (Table 1). Obviously many of the revertant colonies present at that time were progeny of clones which had reverted during the previous subcultures, so that the results represent the total number of revertants arising during the three subcultures. The majority of the revertants arose after the first subculture. Of 2,275 strain F24 L-forms tested, only 89 revertants were found. This is in sharp contrast to the 185 revertants from 1,001 strain El L-forms tested. DISCUSSION A system was developed for growing lysozymedamaged group D streptococci as either L-forms or as streptococcal colonies. A similar system was described previously by Landman and his co-workers (9-11) for B. subtilis protoplasts, but there are some differences. In the B. subtilis system, quantitative (100%) recovery of protoplasts as bacillary colonies was found when either TABLE 1. Spontaneous reversion of lysozymeinduced enterococcal L-forms after three successive subcultures No. of revertant Colonies which Strain No. of L-forms streptococcal reverted colonies F24 2, El 1, elevated agar concentrations (2.0 to 2.5 %) or 25% gelatin (Eastman Kodak) was used as the solidifying agent in the medium. In the enterococcal system, quantitative recovery of lysozymedamaged cells as streptococcal colonies was not found when high agar concentrations were used to solidify the medium. When gelatin was the solidifying agent in the medium, quantitative recovery of streptococcal colonies was observed in some series of experiments but not in others. The experiment-to-experiment variation, or series of experiments compared to another series performed at a later time, illustrates a problem of frequent occurrence when working with streptococcal protoplasts. The enterococcal system requires at least 35% pig skin gelatin (Eastman Kodak), whereas 25 % gelatin (Difco) is adequate to insure maximal reversion of lysozyme-damaged cells of the two strains. This difference in threshold gelatin concentrations probably represents the difference in gel properties between the two preparations. In the B. subtilis system, 25 7 gelatin (either Eastman Kodak pigskin or Difco) was adequate for maximal recovery of protoplasts as bacilli (9-11). It appears that in both the B. subtilis and streptococcal systems firmness of the gelatin medium is the important determining factor in reversion. We found routinely that 0.1 % of the streptococci treated with lysozyme remained as streptococci and escaped any demonstrable damage, being still able to grow as streptococci at low concentrations of gelatin (Fig. 1-4). Thus it was necessary to dilute the treated suspensions (approximately 108 to 109 CFU streptococci) 10,000,000-fold or greater, to insure that the majority of the suspensions contained no streptococci. There was no major difference in the yield of revertants from either of the strains tested. This phenomenon was surprising, since strain El lysozyme-damaged cells retained a large quantity of cell wall rhamnose typical of spheroplasts, whereas strain F24 cells were plated out as protoplasts (7). The chemical data may be only a coincidental observation, but it may suggest

5 696 KING AND GOODER J. BACTERIOL. that the presence of attached fragments of cell wall offers no advantage to a damaged cell as to whether it will ultimately revert to a viable coccus. Such fragments could materially affect the rate at which reversion occurs. The common feature of the B. subtilis and the enterococcal systems is that lysozyme-damaged cells can be selectively grown on solid medium as either L-forms on plates with low concentrations of agar or as bacterial colonies on medium solidified with appropriate concentrations of gelatin. In contrast to the B. subtilis system, the continued growth of streptococcal protoplasts as L-forms does not require a reversion inhibitor such as D-methionine to be added to the medium. The reversion of freshly induced L-forms during a course of routine subculture under apparently nonselective conditions was therefore of interest. Our results suggest that the majority of enterococcal L-forms induced by the action of lysozyme do not revert immediately. The strain differences are shown in Table 1. With both strains, 80% or more of the L-forms induced from a suspension of enterococci remained as L-forms for at least three subcultures. Any hypothesis to explain the initial stability of the enterococcal L-forms must also account for sporadic reversion beyond three subcultures. Strain differences are apparent in this aspect of reversion. Only an occasional sporadic revertant was found in a long history of routine subculture of strain F24 L-forms, and no revertants were found beyond approximately 30 subcultures. Reversion was so frequent in strain El L-form subcultures that a pure cell line of L-forms could not be established. Neither S. zymogenes 30 L-forms derived and maintained in another laboratory nor our adapted strain of F24 L-forms reverted under the optimum conditions found effective for reversion of a suspension of lysozyme-damaged enterococci. Landman (8) has invoked the idea of commitment, the event whereby the ability of a cell with damaged wall to regrow on normal media as a bacterial form is lost. In B. subtilis, commitment occurs as the last trace of wall is removed by lysozyme. The enterococcal and group A streptococcal systems (5) may be similar to the B. subtilis system in this respect, but differ in that, once the cell is committed and grown as an L-form, it appears to be much more difficult to reverse the process and return to the coccal state. Continuous subculture and transfer to broth of the now stable L-form probably selected out variants better adapted to growt-h under these conditions. Such selected variants may well differ genetically from the original parent strain. One such selected variant was studied by Hoyer and King (6), who suggested that there is a 4 to 6%y deletion of the genome when the L-form is compared to the parent S. faecium F24. It is not known which specific cell functions have been impaired as a result of this deletion. ACKNOWLEDGMENTS This investigation was supported by Public Health Service grants AI from the National Institute of Allergy and Infectious Diseases and GM from the National Institute of General Medical Sciences, and by a contract (no. DR MD 2352) from the U.S. Army Medical Research and Development Command under the sponsorship of the Commission on Acute Respiratory Diseases, Armed Forces Epidemiology Board. LITERATURE CITED 1. Altenbern, R. A Reversion of L forms and spheroplasts of Proteus mirabilis. J. Bacteriol. 85: Crawford, Y. E., P. F. Frank, and B. Sullivan Isolation and reversion of [-forms of beta-haemolytic streptococci. J. Infec. Dis. 102: Dienes, L., and H. J. Weinberger The L forms of bacteria. Bacteriol. Rev. 15: Freimer, E. H., R. M. Krause, and M. McCarty Studies of L-forms and protoplasts of group A streptococci. 1. Isolation, growth, and bacteriologic characteristics. J. Exp. Med. 110: Gooder, H., and W. R. Maxted External factors influencing structure and activities of Streptococcus pyogelnes p In G. G. Meynell and H. Gooder (ed.), Microbial reaction to environment. Sym. Soc. Gen. Microbiol., vol. 11, Cambridge University Press, Cambridge, England. 6. Hoyer, B. H., and J. R. King Deoxyribonucleic acid sequence losses in a stable streptococcal L-form. J. Bacteriol. 97: King, J. R., and H. Gooder Induction of enterococcal [-forms by the action of lysozyme. J. Bacteriol. 103: Landman, 0. E Protoplasts, spheroplasts and L forms viewed as a genetic system, p In L. B. Guze (ed.), Microbial protoplasts, spheroplasts and L forms. The Williams & Wilkins Co., Baltimore. 9. Landman, 0. E., and S. Halle Enzymically and physically induced inheritance changes in Bacillus subtilis. J. Mol. Biol. 7: Landman, 0. E., A. Ryter, and C. Frehel Gelatininduced reversion of protoplasts of Bacillus subtilis to the bacillary form: electron microscopic and physical study. J. Bacteriol. 96: Miller, I. L., W. Wiebe, and 0. E. Landman Gelatininduced reversion of protoplasts of Bacillus subtilis to the bacillary form: photomicrographic study. J. Bacteriol. 96: