THE REVERSION OF THE L FORM OF PROTEUS MIRABILISINTO THE ROD FORM'

M. A. MEDILL AND W. G. HUTCHINSONLaboratory of Microbiology, Department of Botany, University of Pennsylvania, Philadelphia, Pennsylvania

Received for publication December 29, 1953

Seveial Proteus strains have been reported toform L type colonies when inoculated onto peni-cillin agar (Dienes, 1949; Tulasne, 1949; Medilland Hutchinson, 1953). These colonies arecharacterized by a dense center imbedded in theagar and a periphery on the surface of the me-dium. They contain elements which are sphericalin shape and vary in diameter from 0.2 to 7 , ormore. Tulasne (1953) designates the smallerforms as elementary bodies and the larger onesas large bodies. In the present paper the term"L form" will be used to refer to any elementfound in an L type colony in order to distinguishthese forms from the large bodies which areformed in many bacterial species under a varietyof conditions. The identity of these latter spheri-cal forms with L forms is not established. Thepresent paper presents a description of the re-version of the L forms into bacilli, illustrated byconsecutive photographs of the same living cellsundergoing the entire process.

METHODS

The organism used in this investigation wasProteus mirabilis, strain K, which was kindlysupplied by Dr. James C. Kakavas of the Uni-versity of Delaware. This strain is an activeswarmer. The medium used to support the growthof L type colonies was prepared by solidifyingtryptose blood base broth (Difco) with 1.1 percent agar. After autoclaving, 200 units of peni-cillin per ml and one per cent "PPLO" serum(Difco) were added.

Slide cultures were prepared by removing ablock of agar containing L type colonies and in-verting and smearing this onto tryptose agarwithout penicillin. A block of this latter agar con-taining L forms was removed then to a sterileslide, covered with a sterile coverslip, and sealedwith vaspar leaving air space between the vas-

1 This investigation was supported by a fellow-ship granted by the Philadelphia Lager BeerBrewers' Association.

par and agar block. It was found that the addi-tion of a drop of broth between the agar andcoverslip greatly improved the viability of theL forms. When penicillinase was used, a drop ofbroth containing one per cent "penase" (Difco)was placed between the agar and coverslip. Thesecultures were incubated at room temperature onthe stage of the microscope, and the cells wereobserved continuously as they developed intobacilli. Time-lapse motion picture photographswere recorded on Cine-Kodak super XX filmwith a time interval of 20 seconds between ex-posures. A Spencer 95x oil immersion dark Mphase contrast objective was used with a 25x eyepiece.

RESULTS

Since small amounts of penicillin carried overfrom the inoculum conceivably might affect thedevelopment of the L forms into bacilli, penicil-linase was included in some of the preparations.Under any conditions, the larger L forms werethe only ones observed to return to bacilli. Thetransitional L forms tend to branch and becomeactively motile in the presence of abundantmoisture. With the restricted moisture necessaryfor successful time-lapse photography, branchingis inhibited but still observable. A typical rever-sion is illustrated in figures 1 to 13. As seen infigures 2 to 4 the L forms appear to initiate divi-sion. However, by the end of about two hours itis evident that division is not to be completed,but rather an elongation and branching of thecells occur (figures 5 to 13). Several divisions ofthe elongated structure ultimately break it upinto rod forms. Subsequent development of thebacilli so formed differs, depending on whetheror not penicillinase is present. With penicillinasethe bacteria continue to multiply by binary fissionin the small rod form. Swarm cells are seldomobserved. Without penicillinase the small rodsgo through a period of active division and motil-ity, which is followed by a period during whichthe cells appear to become attracted to each other.

89

on April 21, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

- a~-w

II

........

AWIF .. .... .....

...

.....

..P. P

3g

B.

r~~~~~~~~~~~~~~~~~~~~~~........

..... ....

Figur-es 1 to 13. Development of L forms into bacilli. The inoculation was made from four day old Ltype colonies. A dark M contrast objective was used, and the prints are negatives of th-, orieginal image.

Figure 1-30 minutes after inoculationFigure 2-70 minutes after inoculationFigure 3-83 minutes after inoculationFigutre 4-109 minutes after inoculationFigure 5-122 minutes after inoculationFigure 6-135 minutes after inoculationFigure 7-148 minutes after inoculationFigure 8-160 minutes after inoculationFigure 9-173 minuites after inoculationFigure 10-186 minutes after inoculationFigutre 11-199 minutes after inoculationFigure 12-212 minutes after inoculationFigure 13-225 minutes after inoculation

90

MM.

1--]; .::::W.'. '..

I,::I4 4f!

.,;O:.

'o 41,4:,:.

:..

0

on April 21, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

REVERSION OF L FORM OF P. MIRABILIS INTO ROD FORM

Small and large clumps of bacteria are observedfor about an hour. Swarm cells then appearswimming about freely. Within the next severalhours large globular swellings are developed inthe center of these long rods. Motility continuesand the culture consists almost entirely of suchswarm cells with central swellings 16 hours afterinoculation. If the slide culture is overcrowded,development stops; but if the growth is trans-ferred to new media, the ends of the swarm cellscan be observed to fragment into small rods whilethe central portion becomes vacuolated and dis-integrates. This sequence of events has been ob-served many times. If the bacilli resulting fromthe reversion in the presence of penicillinase areinoculated onto penicillin agar, L type coloniesare formed readily. Many efforts have been madeto induce the bacilli resulting from reversion inthe absence of penicillinase to form L type colo-nies. These efforts have uniformly failed unlessthe culture is allowed to grow in the rod formthrough an additional transfer before reinocula-tion onto penicillin agar.

DISCUSSION

Freundt (1950) and Pulvertaft (1953) bothdescribe reversion of other strains of Proteus Lforms by a process similar to that describedabove. Tulasne (1953) describes four methods ofthe reversion of L forms of Proteus to bacilli.The L form ruptures liberating rods, the L formdivides longitudinally into five or six filaments,the L form grows into a fusiform cell from whichrods eventually break off, or it develops by amethod which appears to be Similar to that illus-trated in the present paper. None of these work-ers has published consecutive photographs of thesame living celLs to illustrate these processes, andtherefore it is difficult to compare in detail theirobservations with the present ones.The development of Proteus large bodies (as

distinguished from L forms) into bacilli has beenillustrated by time-lapse photography of livingmaterial by Stempen and Hutchinson (1951).The large body becomes irregular and divides intoseveral bacteria, a process which is very differentfrom the elongation and branching of the L forms.Dienes and Weinberger (1951) state that the Lforms of Proteus revert to bacilli in the same man-ner as do the L forms of Bacteroides. However,the description (Dienes and Smith, 1944) dealswith the development of large bodies which have

formed from the bacilli not with the L formsfound in the L type colonies. This process inBacteroides is illustrated by photographs of livingcells and is similar to that described by Stempenand Hutchinson (1951) for Proteus large bodies,but differs from the method described here forProteus L forms.Although penicillin initiates both large body

and L type colony production, almost all thebacilli will form large bodies, but less than oneper cent forms L type colonies (Medill andHutchinson, 1953). Tulasne (1951) and Dienes(1949) state that L type colonies are formed fromlarge bodies. Although their observations arevery convincing, they are not illustrated by con-secutive photographs of the same large bodyundergoing this entire transformation. Kliene-berger-Nobel (1951) describes the initiation of Lforms by the release of small granules from therods, but she does not state whether these gran-ules initiate L type colonies.Dondero and Zelle (1953) have demonstrated

large body formation to be clonal in Azotobacteragile. On the other hand, L type colony forma-tion has been shown to be nonclonal (Medilland Hutchinson, 1953).

Since the reversion of large bodies and L formsinto bacilli differs, and since a connection betweenthese two forms has not been definitely estab-lished, it would appear to be desirable to dis-tinguish between them until further informationis obtained as to their respective natures.

SUMMARY

The reversion of the L forms of Proteus mira-bilis into bacilli was recorded by time-lapsemotion picture photography using the darkphase contrast microscope. The spherical Lforms elongate, branch, and break up into smallrods. Branching is stimulated by moisture be-tween the agar and coverslip. If penicillinase ispresent, the rods resulting from the reversionform only an occasional swarm cell, while in theabsence of penicillinase abundant swarm cellsare formed which develop central globular swell-ings.

ADDENDUM

After this paper was submitted for publication,a paper by von Prittwitz und Gaffron (Naturwis-senschaften, 40, 590, 1953) appeared. The photo-graphs in this paper show reversion of spherical

911954]

on April 21, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

M. A. MEDILL AND W. G. HUTCHINSON

forms of Proteus ulgaris to rods in a manner en-tirely similar to that shown in the present paper.The strain used by von Prittwitz und Gaffron,however, does not produce L type colonies, andtheir photographs are of spherical forms deriveddirectly from rods on penicillin agar rather thanfrom L type colonies. The photographs accom-panying the present paper are of elements re-moved from typical macroscopic L type colonieswhich have been growing on penicillin agar forfour days. The difference between those penicillininduced spherical forms which will reproduce inthe L form and those which will not remains tobe clarified.

REFERENCESDIENES, L. 1949 The development of Proteus

cultures in the presence of penicillin. J.Bact., 57, 529-546.

DIENES, L., AND SMaTH, W. E. 1944 The sig-nificance of pleomorphism in Bacteroidesstrains. J. Bact., 48, 125-153.

DIENES, L., AND WEINBERGER, H. J. 1951 TheL forms of bacteria. Bact. Revs., 15, 245-288.

DONDERO, N. C., AND ZELLE, M. R. 1953 Obser-vations on the formation and behavior of

"conjugation" cells and large bodies inAzobacter agilk. Science, 118, 34-436.

FREUNDT, E. A. 1950 Observations on Dienes'L type growth in bacteria. Acta Pathol.Microbiol. Scand., 27, 159-171.

KLENEBERGER-NOBEL, E. 1951 The L cycle inbacteria. J. Gen. Microbiol., 5, 525-530.

MEDILL, M. A., AND HUTCHINSON, W. G. 1953The nature of L type colony formation inProteus mirabilis. Bact. Proc., 38-39.

PULVERTAFT, R. J. V. 1953 The L form ofbacteria in relation to antibiotics. J. Pathol.Bacteriol., 65, 175-184.

STEMPEN, H., AND HUTCHINSON, W. G. 1951 Theformation and development of large bodies inProteus vulgaris OX-19. I. Bright phase con-trast observations of living bacteria. J.Bact., 61, 321-335.

TULASNE, R. 1949 Existence of L forms incommon bacteria and their posible impor-tance. Nature, 164, 876-877.

TULASNE, R. 1951 Les formes L des bacteries.Rev. immunol., 15, 223-241.

TULASNE, R. 1953 Le cycle L et les formesnaines des bacteries. Symposium on bac-terial cytology. Sixth Intern. Congr. Micro-biol., 144-181.

92 [VOL. 68

on April 21, 2021 by guest

http://jb.asm.org/

Dow

nloaded from