JOURNAL OF BACTERIOLOGY, Dec. 1970, p. 1378-1385 Vol. 104, No. 3Copyright a 1970 American Society for Microbiology Printed in U.S.A.

Alterations of the L-Forms of a SporebearingBacillusLOUIS DIENES

Department of Medicine and Department ofBacteriology, Massachusetts General Hospital,Boston, Massachusetts 02114

Received for publication 23 September 1970

A large pleomorphic gram-negative bacillus developed as a contaminant on blood-agar. Spores were formed in one culture. L-forms were produced with penicillin onblood-agar with 2.5% NaCl; they grew well when transplanted to agar with 0.5%NaCl. After several transplants and long incubation of the L-forms without peni-cillin, in three transplants small gram-negative pleomorphic bacilli grew, but noL-forms. This occurred once on blood-agar and twice on 30% gelatin. The growthobtained from these small bacilli was similar in morphology and in the physicalproperties of the organisms to the altered L-forms of Proteus and Salmonella. Mul-tiplication of the pleomorphic organisms and development of branching filamentsfrom the round forms was apparent. The original large gram-negative bacilluswas regularly recovered from the L-forms, and was recovered several times from thedescendants of the small bacilli. These observations are essentially similar tothose made with L-forms of Proteus and with an L-form studied in 1952, indi-cating alterations in L-forms of bacteria which do not produce B type L-forms.

In one experiment with a saprophytic myco-plasma, a large gram-negative bacillus developedas a dry, spreading growth around a culture ofthe mycoplasma. Most cultures obtained fromsingle colonies of the large gram-negative bacilluswere free from mycoplasmas, and mycoplasmasnever reappeared in their descendants. Thisbacillus was designated as 2GN. The cultureunder certain conditions was extremely pleo-morphic and was further studied. The mostsignificant result of these studies is that culturesfrom the L-forms of bacillus 2GN were obtainedwhich were similar in many respects to and alsocharacteristically different from the altered Lcultures of Proteus described in the precedingpaper (3).

MATERIALS AND METHODSThe materials used and the methods of microscopic

study of the cultures were similar to those used withthe altered Proteus L-forms previously described (3).In addition, the medium recommended by Hayflickfor the cultivation of mycoplasma was also used (5).Bacillus Y, mentioned in the preceding paper, wasused in several experiments to enhance the recoveryof bacteria from the L-forms.

All of the work was done by the author personally,as in the study of the altered L-forms of Proteus. Thecultures were sealed and kept in the cold roomthroughout the whole period of study to allow forlater checks and comparison. These studies would

hardly have been possible without the permanentstained-agar preparations (2) and the agar fixationtechnique of Klieneberger (6).

RESULTSThe bacillus 2GN usually grew in bundles of

long thick or thin filaments (Fig. 1). When itgrew as large bacilli, many of the bacilli hadround swellings, usually at the ends. The bacilli atthe edge of the inoculated areas were often curvedinto a half-circle or into complete coils. Pleo-morphic filaments and bacilli penetrated theagar, producing small irregular colonies therein.The bacilli and their various descendants weregram-negative in all cultures. Spores were notproduced on blood-agar, plain agar, or in broth,even when observed for several months. Sporeswere formed only on a plain agar plate whichafter long incubation became contaminated witha mold. The plate contained transplants of 12isolated colonies, and spores developed in alltransplants. Pieces of agar with the spores weresubmerged in broth and boiled for a few minutes.The cultures obtained from them were similar tothe original cultures.

Bundles of very thin filaments developed in thebroth inoculated with spores and on plain agarwhen the bacteria reappeared in L cultures. Thefilaments broke into small thin bacilli whichproduced tiny coils similar to the coils grown

1378

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

4

-0

i t

aw3i0> ltt 1

IC N #P .

.-% I% \ V I" & .')

" i I

I 4 .(V'IL)l '), ". %",.-

*1 .( W,- L "-: *ocp4.;F

.0

%.- .el.. -d

t."% .0

9

FIG. 1-9. 2GN bacillus and descendants. All photographs were made either from stained-agar preparations orfrom imprints made by agar fixation. Magnification, X2,250. Fig. 1. Bacillus 2GN on blood-agar plate (BAP).Fig. 2. Edge of the cultures on BAP; thick curved pleomorphic bacilli, forming coils. Fig. 3. Pleomorphic growthinto the agar under the culture of2GN. Fig. 4. Groups ofthin filaments in the culture obtained in brothfrom spores.Fig. 5. Same culture as in Fig. 4, showing coil formation of thin filaments. Fig. 6. Thin filaments on plain agardeveloping under autolyzed L type colonies, showing coil formation. The darkly stained large round forms are theremnants of the L colony. Fig. 7. Same as Fig. 6, showing straight longfilaments. Fig. 8. L colony growing on thesurface of blood-agar with 0.5% NaCl. Irregular spreading large bodies at the edge of the colony. Fig. 9. Same asFig. 8, showing large bodiesfilled with small granules inside the colonies.

1379

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

from the large bacilli, but much smaller. Thesethin filaments were connected by transitionalforms to the thick filaments and could not becultivated separately from them. They wereevidently one aspect of the pleomorphism of thecultures.The appearance of the culture and of the bacilli

suggest that bacillus 2GN belongs to the genusBacillus. Its classification was not attempted. Ahigh degree of pleomorphism is often the prop-erty of a strain, not of the species, and is theresult of the previous history of the strain.Exposed to penicillin on blood-agar or on

serum-agar with 0.5% NaCl, the filaments andbacilli produced masses of round forms corre-sponding in appearance to protoplasts, but theyproduced no L-forms. On media with 2.5%NaCl, large bodies developed abundantly fromthe bacilli, and within a few days L colonies alsodeveloped. The L colonies remained viable andcontinued to enlarge for several weeks. L coloniesalso developed when the bacteria were inoculatedon Hayflick's medium (5). Well-developed Lcultures transferred to blood-agar or plain agarwith 0.5% NaCl produced colonies which grewmainly on the surface with little penetration ofthe agar. In the course of a few weeks, the Lforms grew to large mucoid colonies, and inheavy inoculated areas confluent masses wereproduced. The colonies consisted of large bodieswhich were often very large and irregular (Fig. 8)and contained well-stained granules (Fig. 9).

Immediately after production, L-forms trans-planted to blood-agar plate (BAP) with 0.5%NaCl without penicillin did not grow, and nobacteria were produced. The transplants madeafter 4 to 6 weeks of incubation of the L-forms tomedia without penicillin often reproduced thelarge bacilli. When either young or old L cultureswere exposed to the influence of Bacillus Y, thelarge bacilli were recovered in almost every case.

Successive transfers of the L-forms and of thevarious types of growth obtained from them areindicated in Table 1. The bacillus 2GN waspropagated in two separate lines (3 Aug and 5Aug). They were handled in a similar manner.L-forms were obtained on a BAP containing2.5% NaCl and penicillin, which was inoculatedwith the bacilli on 13 August. Transplants fromthis plate were made after 4 days of incubation toplain nutrient agar containing 17% NaCl and nopenicillin. Slow growing L colonies developed onthis plate without reappearance of the bacteria.In further transplants, BAP were used with 0.5%ONaCl and no penicillin. After 6 weeks of incuba-tion (5 October) single L colonies were trans-ferred to a BAP. In the transferred culture only

large bodies and granules, but no bacteria, werevisible, and no bacteria developed in the trans-plant. Three single colonies of the L-forms ofline 3 Aug transferred 23 October after 9 days ofincubation to a blood plate, produced smallcolonies of small pleomorphic gram-negativebacilli and no L colonies. The bacilli at the pe-riphery of the inoculated area grew on furtherincubation to quite large opaque colonies. Similartransplant of the 5 Aug line produced only growthof L-forms. On 31 October, single colonies after17 days of incubation from both of the lines,3 Aug and 5 Aug, were transferred to 30% gelatinin broth containing 0.5% NaCl, 5% horse blood,and no penicillin. Although there were no L-formson this plate, small colonies consisting of smallbacilli and pleomorphic forms developed. Col-onies developing from line 5 Aug were transferredfrom the gelatin on 4 November to BAP with0.5% NaCl. After 1 day of incubation, theyproduced a few groups of large straight gram-negative bacilli corresponding to the original2GN and small colonies (Fig. 10) consisting ofsmall pleomorphic bacilli mixed with roundforms of various sizes. Figures 11 to 21 show thevarious morphological forms present in coloniesand in their descendants. Single colonies weretransplanted from this plate again to a BAP on11 November. After three more transplants, four16-day-old colonies were transferred to BAP. Onetransplant from a larger colony gave abundantgrowth of small bacilli. The others yielded onlyslight growth; in two, branching filaments grewfrom the round forms (Fig. 19 and 20). The in-terior of the colonies consisted of round forms,some of which were elongated into short filaments(Fig. 18). After 7 days of incubation, morebranching filaments were present (Fig. 20 and21).These cultures were transplanted and examined

for 2 months. By picking isolated colonies, it wasnot possible to separate the various morphologi-cal forms. Small bacilli, granules, and roundforms predominated in the cultures. Occasionallybranching filaments and large gram-negativebacilli similar to the original 2GN appeared inthe cultures.

It was characteristic of the growth of the smallgram-negative bacilli that in the colonies, espe-cially during the early development, transition inmorphological appearance was present betweenthe small bacilli and granules or round forms.The colonies often consisted of granules andround forms, with a few bacilli at the periphery.After 24 hr of growth, tiny colonies often con-sisted of very small granules with occasionallarger round forms, but otherwise they had the

1380 DIENES

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L-FORMS OF BACILLUS

TABLE 1. Successive transplants starting with bacillus 2GN and leading to the development ofsmall pleomorphic bacilli and branching filamentsa

Type of growth obtained

17 June 1968. Agar blocks with culture of asaprophytic mycoplasma exposed to Bacillus Y.

3 August and 5 August. Transplant to blood agarplate (BAP).

13 August. Transplant to BAP with 2.5% NaCland penicillin.

17 August. Transplant of L colonies to plain nu-trient agar with 1% NaCl and no penicillin.

5 October. Single L colonies transferred to BAPwith 0.5% NaCl.

14 October. Transfer to BAP.

23 October. Transfer to BAP.

31 October. Gelatin (30%) inoculated from plate14 October.

4 November from 31 October. Gelatin plate toBAP.

11 November. Single colonies of line 5 Aug trans-ferred from plate 4 November

13 November. BAP from 11 November.

27 November. Four single colonies from 13November plate.

3 August. Spreading of dry bacillary growtharound the culture of mycoplasma.

Thick and thin filaments, large straight and pleo-morphic gram-negative bacilli.

Large L colonies.

Slow-growing L colonies.

Growth of L colonies.

Growth of L colonies.

In line 3 Aug on 29 October, abundant growth ofsmall gram-negative bacilli and round forms.No L-forms.

In line 5 Aug, growth of L-forms.

In both lines, small colonies of the small gram-negative bacilli and round forms. No L-forms.

Large straight bacilli and small bacilli and roundforms.

From one colony large bacilli with coils. From twoother, small bacilli and round forms.

Growth of small bacilli and round forms.

From one colony abundant growth of small bacilli.From the other three slight growth and branch-ing filaments at the edge of colonies.

a The two lines of 2GN (3 Aug and 5 Aug) were handled similarly. Media contained 0.5% NaCl andno penicillin in all transplants except on 13 and on 17 August.

characteristics of bacterial colonies. The colonies of the colonies. However, microscopic examina-developed on the surface of agar without pene- tion did not clearly differentiate tiny bacilli fromtrating it. the granules.The granules and round forms, small and The small bacilli were thin and many were be-

large, were different from those seen in autolyzed tween 0.5 and 1.0 ,um in length (Fig. 14). Theybacterial or L-form cultures. They were deeply were less deeply stained than the granules andstained and mechanically resistant; when they seemed to be less resistant to deformation andwere smeared they retained their shape. They destruction. They were not motile. Exposed towere evidently close to bacteria in structure, like penicillin they enlarged slightly and becamethe organisms in the altered L cultures of Proteus rounded. In one experiment with increased NaCldescribed in the preceding paper. It was apparent concentration, they produced large bodies andthat the larger-sized round forms take part in L-forms similar to those produced by the bacillusmultiplication. They disintegrated into smaller 2GN. In some cases, in thickly grown areas theforms, and the branching filaments developed small bacilli grew longer (over 1 ,um) withoutfrom them. Multiplication of the small granules branching, and only a few granules and roundwas not directly observed but it was suggested by forms were present. These bacilli regained theirthe absence of bacillary forms in the early growth original properties when transplanted and pro-

Transplants

VOL. 1049 1970 1381

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ft s;.

*0

18 1i9 '. .I-<

i'.. - %

VtI

S

a... \

b.J'

io, .X

..L.e:

*4S

FIG. 10-21. Descendants of 2GN. All photographs were made as in Fig. 1 to 9. Magnification is X2,250 unlessotherwise noted. Fig. 10. Tiny colonies in transplantsfrom 30% gelatin plate inoculated on 31 October with L-formsof line Aug S; 24 hr incubation. X250. Fig. 11-16. Edge of colonies after further development, as in Fig. 10. Thesmall bacilli together with small rounded forms are best visible in Fig. 14. The different morphological forms are

intermingled in the colonies: Fig. IS, the small round granules; Fig. 13-16, the large round forms and their dis-integration into smaller forms. Fig. 17-21. Cultures on BAP inoculated on 27 November from a descendant of

1382

10 C

*

41.

14

&

b2..k ; 'w''~~~~~~~~~~~~~.0.

15

0

I T

...

'I: f

if A 14

** *

..

A

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L-FORMS OF

duced small colonies in which granules and roundforms predominated. Development of branchingfilaments was observed only from the roundforms. They did not grow separately from themand seemed to be a transitional phase in thegrowth of the cultures, as seen in Proteus andSalmonella.

Recently, a few experiments were made withBacillus Y, the large sporebearing bacillus pre-viously mentioned. Only a few large bodies andno L-forms were produced on exposure to peni-cillin. After several hours of exposure of the

22 p

..P"2_ , .X.a.. w~~~~~~~~~ji4.6V _

^sJ_ I _ - ~~~~~~~~~~~*40

N4-

d6

BACILLUS 1383

bacilli to lysozyme, large bodies were producedabundantly on plates containing 2.5% NaCl andpenicillin. A few small L colonies also developedon these plates; these were autolyzed after 24 hr.Bacteria reappearing on the plates destroyed thepenicillin, and under a few of the degenerated Lcolonies very small bacilli appeared (Fig. 22 and23). Their isolation was not successful since theplate was overgrown by Bacillus Y. The appropri-ate study of a large number of similar bacilliwould probably make it possible to reproduce theobservation made with the bacillus 2GN.

23a.kyNi..

25

.1.

I"

%..Et .- ^''

s

FIG. 22-25. All photographs were made as in Fig. 1-9. Magnification, X2,250. Fig. 22 and 23. Small bacilliproducedfrom L colonies of the large gram-positive sporebearing Bacillus Y. In Fig. 22, the small bacilli are scat-tered and in groups under a disintegrated small L colony. In Fig. 23, a colony of small bacilli is shown within theagar under a disintegrated L colony. Figures 24 and 25 show the observations made in 1952 (2). Fig. 24. Tiny bacilli,granules and larger roundforms. Fig. 25. Branching filaments gro wing from the roundforms.

the 31 October gelatin plate. In Fig. 17, the inner part ofa 2-day-old colony consists ofmedium-sized roundforms,some of which elongate to thin filaments. Fig. 18 and 19. Advanced growth offilaments, some showing branching,is visible at the edge of the colony. Fig. 20 and 21. After long incubation, the branchingfilaments are well developedand produce new large bodies from which the branchings start.

VOL. 104, 1970

I

4-

100

I 6.1

.0,

--do

.. .0 Ai4 Ai: i ,-V -

9 lp41k'N

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

DISCUSSION

The available evidence agrees with and sup-ports the derivation of the small bacilli and thevarious morphological forms from the L-formsof 2GN. The L cultures appearing after inductionby penicillin were propagated on media withoutpenicillin. Both the large and small bacilli grewwell on the media and should have been visiblein the L cultures if they were present. The smallbacilli developed in transplants from single Lcolonies from two different plates. The L-formsdid not grow in these transplants.The cultures in the case studied in 1952 be-

haved similarly (1). When granules grew intransplants made from the L-forms, there wereno L-forms mixed with them. When L-forms de-veloped from the culture of granules, the granulesgrew into large bodies, stopped multiplying, andonly the L-forms produced colonies. The or-ganisms in these cultures were apparently changedthroughout the cultures and grew either in one orthe other form. The alteration of the culturewhen N20 of Proteus was produced was similar.No colonies of B type L-forms from which N20originated were mixed with the culture of N20(3).The absence of mixed cultures and the simul-

taneous transformation of the whole culture ofL-forms do not agree with the development of acontaminant. The derivation of the altered formsfrom the L-forms is also supported by the re-production of L-forms and of the large bacilli inthe culture of the altered forms. It is also unlikelythat unusual organisms which have not been seenotherwise contaminate only the cultures of L-forms. However, it is desirable to study the vari-ous forms present in the cultures with serologicaland biochemical methods.The essential properties of the altered orga-

nisms in 2GN and Proteus and Salmonella aresimilar. They grow in the form of small bacilli orgranules, round forms, and branching filaments.They grow on the surface of agar, and theirphysical properties seem to be similar to those ofbacteria. The 2GN is the first bacillus knownwhich did not produce B type L cultures (4) andin the cultures of which altered organisms de-veloped. Besides similarities, there are two char-acteristic differences between the altered formsproduced in 2GN and those of Proteus or Salmo-nella. One is that, in cultures developing from Lforms of 2GN, well-defined small bacillary formswere usually present or were predominant in thecolonies, although absent in cultures derivedfrom Proteus or Salmonella, and were producedonly by exposure to Bacillus Y. The altered formsobtained from 2GN may be regarded as pleo-

morphic forms of the small bacillus. This pleo-morphism is different from the pleomorphism ofStreptobacillus moniliformis, Haemophilus in-fluenzae, and many other bacteria. Pleomorphismin these bacteria is the result of the weakening ofthe cell wall; this is the first step in the develop-ment of the L-form. In the case of 2GN, the roundforms and filaments, and probably the granulesmultiply. All of these organisms are mechanicallymore resistant than the small bacilli and lessinclined to autolysis. They probably represent anirregular step in the reconstruction of the fullbacterial structure. The second difference of thealtered forms in 2GN and in Proteus is in themorphology of the branching filaments. Thesame type of branching filaments and of granuleswhich developed from 2GN had been seen pre-viously only once in the culture of an L-formstudied in 1952 which has already been mentioned(1). This culture presented the most spectacularmorphological transformation observed in con-nection with L-forms. The transition betweentypical L-forms, small bacilli, and branchingfilaments was observed several times. However,the parent bacterium of this L-form is not known.It developed on a plate containing penicillininoculated from a human throat. For the sake ofcomparison with 2GN, photographs of the gran-ules and branching filaments seen in these culturesare shown in Fig. 24 and 25. One interesting re-sult of the present study is that it suggests a pos-sible identity of the parent bacterium of thisstrange L-form. It also suggests methods bywhich similar cultures may be obtained. Suchcultures would be valuable in studying the con-ditions inducing transformation and the struc-tures involved in transformation.

Less complete observations of the reproductionof bacteria from large bodies and L-forms ofother species also suggest the production ofaltered forms. In Escherichia coli and Proteus, inthe early stage of the reversion to bacteria, a fewor a large number of very small bacilli have beenmixed with bacilli of regular size (unpublisheddata). In several streptococci, the cocci recoveredwere also very small at first and gram-negative.Isolation of the small forms was not successful inthese cases.

It is probable that, in the process of reconstruc-tion of the full bacterial structure from the L-forms, organisms varying in morphology and intheir potentiality for development are produced.Probably not all of the variants produced grow inthe cultures; those that do grow are selected by thetechnique of cultivation. The observations avail-able refer to L-forms produced under the influenceof penicillin. Whether L-forms produced spon-taneously, by lysozyme, or by other enzymes can

1384 DIENES

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

VOL. 104, 1970 L-FORMS OF BACILLUS 1385

produce similar variations of morphology needs 2. Dienes, L. 1967. Permanent stained agar preparations of Myco-

to be studied. plasma and of L forms of bacteria. J. Bacteriol. 93:689-692.3. Dienes, L. 1970. Permanent alterations of the L-forms of

ACKNOWLEDGMENT Proteus and Salmonella under different conditions. J. Bac-

This investigation was supported by Public Health Service teriol. 104: 1369-1377.grant AI-05625 from the National Institute of Allergy and In- 4- Dienes, L. and Weinberger, H. J. 1951. The L forms of bacteria.

fectious Diseases. Bacteriol. Rev. 15:245-288.

LITERATURE CITED5. Hayflick, L. 1965. Cell cultures and mycoplasma. Tex. Rep.

Biol. Med. 23(Suppl. 1) :285-303.1. Dienes, L. 1953. Some new observations on L forms of bacteria. 6. Klieneberger-Nobel, E. 1962. Pleuropneumonia-like organisms

J. Bacteriol. 66:274-278. (PPLO) Mycoplasmataceae. Academic Press Inc., London.

on May 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from