INDUCED COLONIAL VARIATION OF A TOTAL POPULATION AMONG CERTAIN LACTOBACILLI MORRISON ROGOSA AND JOYCE A. MITCHELL National Institutes of Health, National Institute of Dental Research, Bethesda, Maryland Received for publication October 13, 1949 Usually colonial variation is considered from the point of view of smooth (S) to rough (R) variation. This variation may be accompanied by changes in (1) chemical constitution of the cell, such as loss of polysaccharide or other component; (2) antigenicity associated with a specific structure of the cell, e.g., polysaccharide or flagella; (3) virulence; and (4) miscellaneous factors. This report concerns itself with a change from R to S, rather than with the more conventional S to R variation. Evidence will be presented to show (1) that the R to S shift of the total population can be induced at will; (2) that the change is not permanent; (3) that the reverse S to R change of the total population can be controlled; and (4) that the results described cannot reasonably be.ttributed to mutation. EXPERIMENTAL METHODS Cultures. The organisms for which data are presented in this paper are Lactobacillus acidophilus strains 200 and 203, Lactobacillus helveticus 80x, Lactobacillus H404, and Lactobacillus brevis Ld9. The latter two strains were isolated from the oral cavity and are heterofermentative. The media used. The formula per liter for the stock medium in which the source cultures were grown is as follows: tryptone, 20 g; tryptose, 5 g; yeast extract, 5 g; tomato juice, 200 ml; liver (Wilson's 1/20), 5 g; glucose, 3 g; lactose, 2 g; and sorbitan monooleate, 50 mg. The tomato juice and liver were combined and corrected to ph 6.5 with NaOH and then mixed with the remaining ingredients without further ph adjustment. The cultures were plated on Hadley's (1933) tomato juice agar prepared by the method of Jay and Arnold (1946). This medium was the same as that used in the original isolation of the oral strains. In this tomato juice agar the strains studied regularly developed rough surface and subsurface colonies. Medium no. 24, on which the organisms were plated and grew as smooth colonies, is of the following composition: solution I, 200 ml; trypticase, 2.5 g; salt solution, 10 ml; adenine, guanine, xanthine, and uracil, each, 10 mg; sorbitan monooleate, 1 g; glucose, 20 g; beef spleen (Wilson's extract), 2.5 g; sodium thioglycolate, 1 g; sodium formaldehyde sulfoxalate, 125 mg; vitamin supplement, 10 ml; agar, 20 g; water to 1 liter; and ph to 5.4 with acetic acid. Solution I: Acid-hydrolyzed casein, 50 mg per ml; DL-asparagine, 2.5 mg per ml; DL-tryptophan, 1 mg per ml; L-cystine, 1 mg per ml; Na-acetate -3H20, 165 mg per ml; K2HPO4, 60 mg per ml; and (NHI)2HC,H507, 20 mg per ml. Salt solution: MgSO4-7H20, 11.5 g; MnSO4*4H20, 2.4 g; FeSO4c4H20, 0.68 g; and water to 100 ml. 303
MITCHELL 304 MORRISON ROGOSA AND JOYCE A. [VOL. 59 Vitamin supplement: Nicotinic acid, 40 mg; thiamine HCI, 0 20 mg; Capantothenate, 20 mg; riboflavin, 20 mg; pyridoxamine di-hci, 8 mg; pyridoxal. HCI, 4 mg; pyridoxine, 8 mg; inositol, 400 mg; choline, 400 mg; p-aminobenzoic acid, 4 mg; biotin, 200 jig; pteroylglutamic acid, 200,g; and water to 200 ml. Cultures less than 24 hours old were diluted in sterile distilled water so that plates containing 100 to 300 colonies could be obtained. To compare the effect of tomato agar and medium 24 on colonial morphology, all plates were inoculated at the same time with the same diluted suspension of cells. The plates were incubated for 48 to 72 hours at 37 C. The colonies were then counted and the morphology was observed visually and also microscopically at magnifications ranging from 9.9 to 45 X. The photographs, which are shown below, were all made at a magnification of 46 X. RESULTS The colonial count for strain H404 was 86 X 107 on tomato agar and 108 X 107 on medium 24. Since the numbers of organisms growing on the medium producing smooth colonies was at least as great if not greater than the total rough population on tomato agar, it seems highly improbable that there was a selection of a smooth type of strain from the original culture. Rather, it would appear that the bacterial populations are genetically alike on the two media and that the change in colonial morphology results from the difference in environmental conditions. Microphotographs illustrating the differences in colonial morphology are presented in figures 1 to 4. Figure 1 represents the subsurface colonial types exhibited by strain H404 on tomato agar. The colony is representative of all the 86 colonies present on the plate and is characteristically rough or woolly. Furthermore, repeated platings over a period of months have yielded no other colony types on this medium. Figure 2 illustrates the smooth type of subsurface colony found constantly in medium 24. Figure 3 is a microphotograph of colonies of the same organism grown on the surface of tomato agar. The colonies are characteristically thin, flat, and irregular. Figure 4 represents the smooth and relatively large colonial forms after streaking on medium 24. Since the colonies were raised and compact, difficulties were encountered in focusing and lighting, and the colonies appear more granular than they are in reality. We wish to emphasize that the platings from which smooth or rough colonies grew on medium 24 and tomato agar, respectively, were made from the same bacterial suspension at the same time. The colonies of the entire populations of L. brevis Ld9 and L. acidophilus 203, as in the case of strain H404, were rough on tomato agar and smooth on medium 24. To conserve space, photographs of these colonies will not be shown. In order to determine whether the colonial types from both media were stable, six isolations of strain Ld9 were made from each medium, inoculated into broth, and replated within 24 hours. The rough isolates from tomato agar developed only rough colonies when replated on tomato agar. However, when they were replated on medium 24 they grew only as smooth colonies. Smooth isolates from medium 24 were replated immediately on tomato agar and grew
Figure 1. Typical rough subsurface colony of strain H404 in tomato agar. Figure 2. Typical smooth subsurface colonies of strain H404 in medium 24. 305
Fire' 3. Typical rough ;Surfac colonies of Strain H4 on to.. Figlzre 3. Typical rough surface colonies of strain H404 on tomato agar. Figure 4. Typical smooth surface colonies of strain H404 on medium 24. 306
1950] INDIUCED COLONIAL VARIATION OF A TOTAL POPULATION 307 only as rough colonies in at least as great numbers as occurred on the plates in which the colonies were originally smooth. The question as to what factor or factors in the medium are responsible for this alteration in colonial morphology is obviously significant. The experiences reported here recall the work of Dubos et al. (1946) with Mycobacterium tuberculo8is and their statement: The cells growing on the surface of the agar medium also appear typical in morphology and staining characteristics. The colonial morphology, however, deserves special mention. Cultures of laboratory strains of human bacilli of high virulence (H37RV) and cultures directly isolated from human pathological material give on the surface of agar media (containing 0.05 per cent Tween 80 and 0.5 per cent bovine albumin) colonies characterized by smooth contour and glistening surface. When, on the other hand, the same cultures are grown on the surface of the agar media of the same composition, but without Tween, the colonies appear much smaller and with an extremely ragged outline... Since medium 24, on which the colonies were smooth, contained 0.1 per cent of sorbitan monooleate, experiments were performed with strains Ld9, 200, 80x, and H404 to show what effect this compound might have on these organisms. Cultures were plated on tomato agar and also on tomato agar supplemented with 0.1 per cent sorbitan monooleate. The colonies from strains Ld9, 200, and 80x were rough as usual on tomato agar and consistently smooth in 6 platings when the medium contained sorbitan monooleate. Furthermore, medium 24 has produced rough colonies in the absence of sorbitan monooleate. Strain H404 has been smooth in 7 of 9 cases in the presence of sorbitan monooleate. Its colonies have always been rough on tomato agar in 9 observations over a period of many months. DISCUSSION Colonial variation among the lactobacilli in which R to S variations were reported has been observed by Barber and Frazier (1945). The present authors have often noted colonial variation on some media, particularly with such species as Lactobacillus lactis and Lactobacitlus acidophilus. Previous workers have also observed mixed colonial types on a variety of media and have designated them as X or Y forms. Tittsler (1949) has induced the growth of one strain of L. lactis as an entire population of smooth colonies by including whey in the medium. However, isolations from either the smooth or the rough type were not stable. When they were plated on certain media, such as tomato agars, usually only rough colonies were encountered. Braun (1947) has discussed the effect of environmental factors on bacterial dissociation, but from the point of view that the environment induces modifications of inherent factors controlling the establishment of mutants in a population. In the present work it is obvious that the alteration of colony morphology of entire populations is not the result of genetic changes, but rather is a difference in the phenotypic expression of genetically similar populations in two different environments. Although the mode of action of sorbitan monooleate is not known, this com-
MITCHELL[ 308 MORRISON ROGOSA AND JOYCE A. [VOL. 59 pound stimulates the growth of many lactobacilli by serving as a source of an essential fatty acid so that many organisms may dispense with a requirement for biotin. In addition, the wetting action of sorbitan monooleate may alter the surface of the cell so that certain nutrients are more readily available than they are without it. The evidence that this may be so is exemplified by the greater size and rate of development of the coloniesin the presence ofsorbitan monooleate. The data we have accumulated indicate that the addition of sorbitan monooleate, and probably some unknown factors, modify the medium both physically and nutritionally so that smooth colonies are formed by some unknown mechanism requiring further study. SUIMMARY Strains representing Lactobacillus helveticus, Lactobacillus acidophilus, Lactobacillus brevis, and certain unidentified species that normally appear as rough colonies when growing on an adequate tomato agar can be shifted in toto to smooth colonial forms. It has been shown that sorbitan monooleate can alter the colonial morphology of these strains from 100 per cent rough to 100 per cent smooth colonies, that colonies of the two types appear in equal numbers, and that the size and rate of development of smooth colonies developing in the presence of sorbitan monooleate are greater than those of rough colonies developing in its absence. Since the described change is not permanent and can be reversed in either direction at will, we believe that these results can be attributed to variation of physical and nutritional factors in the mediuim and that no mutations are involved. REFERENCES BARBER, F. W., AND FRAZIER, W. C. 1945 Dissociants of lactobacilli. J. Bact., 60, 637-649. BRAUN, W. 1947 Bacterial dissociation. A critical review of a phenomenon of bacterial variation. Bact. Revs., 11, 75-114. DuBos, R. J., DAVIS, B. D., MIDDLEBROOK, G., AND PIERCE, C. 1946 Effect of water soluble lipids on the growth and biological properties of tubercle bacilli. Am. Rev. Tuberc., 64, 204-212. HADLEY, F. P. 1933 A quantitative method for estimating Bacillus acidophilus in saliva. J. Dental Research, 13, 415-428. JAY, P., AND ARNOLD, F. A., JR. 1946 Dental caries and fluorine. Science Press, Lancaster, Pa. Refer to p. 43. TITTsLER, R. P. 1949 Personal communication.