APPUED MICROBIOLOGY, Nov. 1968, p. 1655-1662Copyright © 1968 American Society for Microbiology

Vol. 16, No. 11Printed in U.S.A.

Modified Biochemical Tests for Characterizationof L-Phase Variants of Bacteria'

RICHARD L. COHEN, RUTH G. WI1TLER, AND JOHN E. FABERDepartment of Microbiology, University of Maryland, College Park, Maryland 20740 and Department of

Bacteriology, Walter Reed Army Institute of Research, Washington, D. C. 20012

Received for publication 16 August 1968

Biochemical capabilities of bacterial and L-phase organisms of 15 bacterial strainswere examined by using a variety of modified routine diagnostic biochemical tests.The results demonstrated that 13 of the tests examined were suitable for use withL-phase variants, that L-phase variants and revertant bacterial phase organismsretained diagnostically significant capabilities of the respective bacterial or L-phaseorganisms from which they were derived, and that the 13 biochemical tests could beusefully employed to relate a given L-phase variant to a given bacterial phase organ-ism, to distinguish L-phase variants of different species, and to aid in the identifica-tion of nonreverting L-phase variants.

The literature records diagnostically significantbiochemical activities for only a very few strainsof bacterial L-phase variants. These includeProteus spp. (4, 6, 12, 13), streptococci (7, 15),Streptobacillus moniliformis (11), Staphylococcusaureus (14), Neisseria meningitidis (8, 9), and afew others. Most of these studies were not per-formed primarily to identify L-phase variants;therefore, the methods employed were not neces-sarily suitable for routine diagnostic application.As interest mounts in L-phase organisms, both

as useful laboratory tools and as possibly signifi-cant clinical entities, the need for standardizedidentification procedures increases. The pro-cedures required must help resolve several typesof problems, i.e., that of relating an L-phasevariant to a particular bacterial phase organism,that of distinguishing L-phase variants of differ-ent bacterial species or strains when present inmixed culture, and that of identifying the speciesof a nonreverting L-phase variant.

This study was undertaken (i) to design simple,rapid, reliable methods for determining bio-chemical capabilities of L-phase variants, (ii) totest these methods simultaneously on a varietyof L-phase variants and their respective presumedparent or revertant bacterial phase organisms,and (iii) to determine from the results whethercharacterization on the basis of biochemicalactivities could yield data significant for diag-nostic use with L-phase variants.

1 Submitted by R. L. Cohen to the University ofMaryland, College Park, Md., in partial fulfillment ofthe requirements for the Ph.D. degree.

MATERIALS AND METHODS

Organisms. The organisms used in this study andtheir sources are listed in Table 1. The term, "L-phasevariant" or "L-phase organism," is used in this paperto designate self-replicating, cell wall-defective bac-terial variants that grow in colonies having the so-called "fried egg" configuration. "L-phase variant" isused in preference to "L-form," but otherwise is inconformance with the definition of Wittler et al. (15).

Culture media. The culture medium for stockmaintenance of all bacterial phase organisms, as wellas L-phase variants that did not require osmoticstabilization (Table 1), was prepared by dissolving40 g of dehydrated Heart Infusion Agar (Difco) in850 ml of water. (Deionized distilled water was usedfor all media and test solutions.) Media for culture ofosmotically sensitive L-phase variants (Table 1) wereprepared by dissolving 25 g of dehydrated HeartInfusion Broth, 13.5 g of agar (Difco), and either 20or 30 g of NaCl or 100 g of sucrose in 850 ml of water.Ingredients were dissolved by gentle boiling; then thepH was adjusted to 7.6 with 5 N NaOH. The agar wasdispensed in 85-ml volumes into screw-cap bottlesand was sterilized by autoclaving for 15 min at 121 C.Immediately prior to pouring plates, the agar in eachbottle was enriched with 10 ml of sterile horse serum(heated for 30 min at 60 C) and with 5 ml of a 10%(w/v) stock solution of Oxoid yeast extract (Consoli-dated Laboratories, Inc., Chicago Heights, Ill.) thathad been adjusted to pH 7.0 and sterilized by filtra-tion through a 0.01-,um pore-size filter pad (RepublicSeitz Filter Corp., Milldale, Conn.). Potassium peni-cillin G was incorporated at a final concentration of1,000 units per ml in media for unstable L-phasevariants (Table 1). Sterile, disposable 60 by 15 mmplastic petri dishes (Falcon Plastics, Los Angeles,Calif.) were used for routine culture of all bacterialand L-phase cultures.

1655

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

COHEN, WITITLER, AND FABER

TABLE 1. Organisms employed, their sources, and the additives requiredfor stabilization ofL-phase cultures

Additives for stabilizationof L-phase

Bacterial strain Phase designation Source' Penicillin NaCi %, Sucrose

(1,000 W/() (%, WV)units/mi) w/) (,/v

Proteus mirabilis 9

Streptobacillus moniliformis

Neisseria meningitidis 47 VI

N. meningitidis 55 III

Corynebacterium sp. D Campo A

Corynebacterium sp. D Campo B

Corynebacterium sp. D-5

Streptococcus faecalis F-24

S. faecalis G-K

S. faecium D-TC

S. pyogenes Richards

S. sp. Bruno

Staphylococcus aureus Smith

S. aureus ATCC

Gaffkya tetragena Gil

9_pbL-9cS. mon.-PdLi Rat 30c47_pb

47-Lc47-Re55-pb55-Lc55-ReD Campo A-PdD Campo A-LcD Campo B-PbD Campo B-LcD SpbD-5-LCF-24-PbF-24-LcG-K-PbG-K-LcD-TC-LCD-TC-ReRich-PbRich-LcRich-ReBruno-PdBruno-LcSmith-PbSmith-LcATCC-PbATCC-LcGil-LcGil-ReGaff-Bd

ATCC 14273ATCC 14168ATCC 14647ATCC 14075ATCC 23252ATCC 19577ATCC 23248ATCC 23251ATCC 19578ATCC 23247ATCC 15933ATCC 15938S. MadofffS. MadofffATCC 23256ATCC 23257ATCC 19634ATCC 19635ATCC 23241ATCC 23242J. PitchergC. WilliamsgY. CrawfordhATCC 19563ATCC 15185ATCC 19615ATCC 19617ATCC 19636ATCC 19640ATCC 6538PR. ColeiATCC 14105'ATCC 15292ATCC 10875

+

+

3

3

3

3

33

3

2

3

3

10

10

aATCC number from 8th edition, American Type Culture Collection Catalogue of Strains.bParent (or presumed parent) bacterial phase.c L-phase variant.d Bacterial phase, not the parent of the L-phase strain.e Revertant bacterial phase.f Massachusetts General Hospital, Boston, Mass.g Walter Reed Army Institute of Research, Wash., D.C.hU.S. Naval Hospital, Great Lakes, Ill.i NIAID, Bethesda, Md.; Dr. Cole obtained the culture from M. Boris, North Shore Hospital, Man-

hasset, N.Y., who obtained-it from B. Kagan, Cedars-Sinai Medical Center, Los Angeles, Calif.i ATCC number from 7th edition, American Type Culture Collection Catalogue of Cultures.

Test media. Basal media used for the tests were At the same time, sterile substrate, reagent, or indi-essentially the same as the media used for stock cator stock solutions were added to the basal mediamaintenance of cultures, except that for each 1,000 ml in the amounts given in Table 2 to achieve the finalof complete test agar, the basal constituents were concentrations. Sufficient water was then added todissolved in 750 ml of water. The substrates and re- each bottle to bring the volume of the completeagent for phenylalanine deamination and esculin hy- medium to 25 ml. Unless otherwise specified, stockdrolysis tests were dissolved with the basal constituents solutions were prepared in water. When solutions wereof the respective test agar prior to autoclaving. The to be incorporated in test media, they were sterilizedagar was dispensed in 19-ml volumes into 2-oz (59 ml) by positive-pressure filtration through Swinnex-25prescription bottles. Immediately prior to peuring plastic filter units (Millipore Corp., Bedford, Mass.)plates, 2.5 ml of heated horse serum and 1.2 ml of fitted with (25 mm diameter) 0.22-,um pore-size mem-10% (w/v) yeast extract were added to each bottle. branes (Millipore Corp.). Sterile, disposable 35 by 10

1656 APPL. MICROBIOL.

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

VOL. 16, 1968 TESTS FOR CHARACTERIZATION OF L-PHASE VARIANTS

TABLE 2. Concentrations ofsubstrates, indicators, and reagents employed in testsfor biochemical reactions

Stock Final concn Amt added perTest Substrate Indicator or reagent solution (%, w/v) 25 ml of medium(%, W/V)

Oxidase activity

Catalase activityPhosphatase activity

Oxidation-fermenta-tion

Phenylalanine

Esculin hydrolysis

Arginine hydrolysis

Urea hydrolysis

Nitrate reduction

Tetrazolium reduc-tion

Tellurite reductionAMC production

Carbohydrate break-down

Phenolphthalein di-phosphate Na salt

Glucose

DL-phenylalanine

Esculin

L-Arginine hydrochlo-ride

Urea

Potassium nitrate

2, 3, 5-Triphenyltetra-zolium chloride

Potassium telluriteGlucose

Inulin and salicinAll other carbohy-

drates

N, N-dimethyl-p-phenylenedia-mine hydrochlorideHydrogen peroxide

NaOH

Phenol reda

Ferric chloride

Ferric citrate

Phenol red

Phenol red

Sulfanilic acidba-Naphthylamineb

CreatinePotassium hydrox-ide

Phenol red

1.0

301.0

20100.25

10

100.25

200.251.00.80.51.0

1.0101.0

40

510

0.25

0.01

1.00.00250.2

0.10.051.00.0025

2.00.00250.1

0.005

0.0050.5

0.51.0

0.0025

0.25 ml

2.5 ml0.25 ml0.05 g

0.025 g0.0125 g2.5 ml0.25 ml

2.5 ml0.25 ml2.5 ml

0.125 ml

0.125 ml1.25 ml

2.5 ml2.5 ml

0.25 ml

a Prepared by dissolving 0.25 g of phenol red in 70.5 ml of 0.1 N NaOH and then bringing the volumeto 100 ml with water.

b In 5 N acetic acid.

mm plastic petri dishes (Falcon Plastics) were usedfor all tests.

Test controls. For all tests, controls included un-inoculated test plates and inoculated plates containingwater in place of test substances. Certain tests werecontrolled by including bacteria known to be positiveor negative (1). Test bacterial and L-phase organismsthat consistently demonstrated either positive or nega-tive reactions for specific tests were used as controlorganisms in subsequent tests (Table 3).

Inoculation of test media. The inoculum for eachbacterial phase organism consisted of one drop of aheavy suspension made with organisms swabbed offa single culture plate and suspended in 2 ml of 0.9%(w/v) NaCl. The inoculum for each L-phase variantconsisted of a small agar block cut from an area ofdense growth on the appropriate culture medium.Each block was inverted on the test plate, pushedabout one-half the distance across the plate, and lefton the agar in the inverted position. No attempt wasmade to standardize the number of organisms in theinocula.

Conditions of incubation. All plates were incubatedupright in wide-mouth screw-cap glass jars at 37 C.

Bacterial phase cultures of both strains of N. men-ingitidis, as well as all L-phase cultures, were grownunder reduced oxygen tension obtained by includinga lighted candle in each jar.

Biochemical tests. The tests described in this paperare modifications of routine diagnostic biochemicaltests (1, 2, 3). The pH of tests employing phenol redindicator was estimated by comparison with indicatorstandards (LaMotte Chemical Products Co., Balti-more, Md.).

Oxidase activity was demonstrated by flooding 1-to 3-day-old colonies, grown on maintenance culturemedia, with freshly prepared N,N-dimethyl-p-phenyl-enediamine (Eastman Organic Chemicals, Rochester,N.Y.). The reagent was then poured off. Coloniesgiving a positive oxidase test turned deep red within5 min and black within 30 min.

Catalase activity was demonstrated by droppingseveral drops of 30% H202 (Mallinckrodt ChemicalWorks, St. Louis, Mo.) directly onto 2-day-oldcolonies growing on maintenance culture media. Apositive reaction was indicated by evolution of gasbubbles from the colonies.

Phosphatase activity was demonstrated by growing

1657

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

1658 COHEN, WITTL

TABLE 3. Suggested parent bacterial and L-phasecontrol organisms used to test biochemical

activities

Parent bacterial and L-phase organismsyielding

Positive reactions Negative reactions

Oxidase N. meningiti- P. mirabilis 9activity dis 55 III S. faecalis

G-KCatalase P. mirabilis 9 S. faecalis

activity S. aureus G-KATCC

Phosphatase P. mirabilis 9 S. faecalisactivity S. aureus G-K

ATCCPhenylalanine P. mirabilis 9 S. faecalisdeamination G-K

S. aureusATCC

Esculin S. faecalis P. mirabilis 9hydrolysis G-K S. pyogenes

RichArginine S. faecalis P. mirabilis 9

hydrolysis G-KUrea P. mirabilis 9 S. faecalis

hydrolysis S. aureus G-KATCC

Nitrate P. mirabilis 9 S. faecalisreduction S. aureus G-K

ATCCTetrazolium P. mirabilis 9 None avail-reduction S. faecalis able

G-KTellurite P. mirabilis 9 None avail-

reduction S. faecalis ableG-K

organisms on test media containing the sodium salt ofphenolphthalein diphosphate (K & K Laboratories,Inc., Plainview, N.Y.). After incubation for 4 days at37 C, the pH of the medium was raised by floodingeach plate with 5 N NaOH. A positive reaction wasindicated by a red color and a weak reaction by a faintpink color in the area surrounding growth.

Oxidation or fermentation of glucose was demon-strated on test media that contained glucose andphenol red (Coleman & Bell Co., Norwood, Ohio).The agar was dispensed in 2-ml volumes into steriledisposable glass test tubes (12 by 75 mm). Afterinoculation, a sterile platinum needle was stabbedthrough the inoculum to the bottom of the tube.Duplicate tubes of test and control media were used.In one of each of the test and control tubes, the agarwas overlayered with 1.5 ml of sterile Vaspar, whichwas prepared by thoroughly mixing equal parts ofVaseline and paraffin (melting range, 53.5 to 56.5 C)and by sterilizing in screw-cap test tubes in a hot airoven at 170 C for 2 to 3 hr. Overlayered tubes were

tightly stoppered with corks, and the remaining tubeswere stoppered with gauze plugs. The pH was readat 24-hr intervals for approximately 1 week. Fermenta-

,IER, AND FABER APPL. MICROB'OL.

tive organisms produced acid throughout the agar inaerobic and anaerobic tubes, whereas oxidative organ-isms produced acid throughout the agar in the aerobictube only. Weak reactions were indicated by pHreductions of less than 0.4 unit.

Phenylalanine deamination was tested by growingorganisms on test media containing DL-phenylalanine(Eastman Organic Chemicals, Rochester, N.Y.). Afterincubation for at least 2 days at 37 C, one drop ofFeCl3 solution was placed on the colonies. A positivereaction was indicated by appearance of a green toblue-green color in the area of growth.

Esculin hydrolysis was demonstrated by growingorganisms on media containing esculin (BBL) andferric citrate. A positive test was indicated by blacken-ing of the test agar.

Arginine or urea hydrolysis was demonstrated bygrowing organisms on media containing L-arginine(Nutritional Biochemicals Corp., Cleveland, Ohio) orurea (Mallinckrodt Chemical Works) and phenol red.A positive test was indicated by a pH rise of greaterthan 0.2 unit and a weak reaction by a rise of lessthan 0.2 unit, when compared with appropriate con-trol plates.

Nitrate reduction was demonstrated by growingorganisms for at least 2 days at 37 C on test mediacontaining KNO3. After incubation, each plate wasflooded with 0.5 ml of sulfanilic acid (Merck & Co.,Inc., Rahway, N.J.) solution followed by 0.5 ml ofa-naphthylamine (Fisher Scientific Co., Fairlawn,N.J.) solution. Appearance of a red color indicated apositive test. Negative tests were confirmed by placinga small piece of mossy zinc or a pinch of zinc powderon the agar in the area of growth. Appearance of ared color indicated that the nitrate had not beenreduced.

Tetrazolium reduction was indicated by formationof pink to intense red color in the colonies grown onmedia containing 2,3, 5-triphenyltetrazolium chloride(Paul-Lewis Laboratories, Inc., Milwaukee, Wis.) orin the agar surrounding the colonies.

Tellurite reduction was indicated by appearance ofa gray to black color in colonies grown on mediacontaining K tellurite (Fisher Scientific Co.).

Acetylmethylcarbinol (AMC) production wasdemonstrated by growing organisms at least 4 days at37 C on test media containing glucose. Then 0.05 mlof creatine (Nutritional Biochemicals Corp.) solutionand 1 ml of KOH solution were dropped on the colo-nies. Development of a red color indicated a positivereaction, and development of a faint pink colorindicated a weak reaction.

Breakdown of 24 carbohydrates was tested byusing phenol red as the indicator of acid production.The carbohydrates incorporated in test media wereadonitol, arabinose, cellobiose, dulcitol, fructose,galactose, glucose, glycerol, inositol, inulin, lactose,maltose, mannitol, mannose, melezitose, melibiose,raffinose, rhamnose, salicin, sorbitol, sorbose, sucrose,trehalose, and xylose. Because of the presence ofmaltase in horse serum, 10% (v/v) heated rabbitserum was used in the test for breakdown of maltose.Acid production was read at 24-hr intervals for thefirst 3 days and then at 4-day intervals through thethird week. A positive test was shown by reduction of

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

VOL. 16, 1968 TESTS FOR CHARACTERIZATION OF L-PHASE VARIANTS 1659

TABLE 4. Results of biochemical tests performed on parent, L-, revertant, and independentlyisolated bacterial phases of selected strains of organismsa

Coryne- Coryne-P. mirabilis S. moniliformis N. meningi- N. meningitidis bacterium bacterium Corynebacterium

9 lidis 47 VI 55 III sp. D sp. D sp. D-5Test Campo A Campo B

P L B L P L R P L R B L P L P L

Oxidase activity. . + + + + + + .._Catalase activity...+ + WWWWWW+ + + + +Phosphatase activ-ity. ++ W W _____ _

Oxidation-fermen-tation.. F F F F 0 0 0 0 0 0 V F V V NR NR

Phenylalanine de-amination ..... + +_

Esculin hydroly-si51........ - - W W - +

Arginine hydroly-sisis_........._.- - + +-++ + V + V

Urea hydrolysis... .+ + .+ - + V + +Nitrate reduction + + ..._Tetrazolium re-

duction..........+ + WV WV +++++++++ + +Tellurite reduction. + + + + ++ W + + + + + + +AMC production.. ....._

CarbohydratebreakdownAdonitol........- - _ _ _____ _Arabinose....... -- - _ _ _ _ _ _ _ WCellobiose.......- - _ _ _ ___ _ W + W W

Dulcitol ......... - - _ _ _ _ _ _ _ _ _Fructose......... + + - .+ -_ ___ _ +Galactose ....... + + W + w_ _ _ W + wGlucose .... + + + + + + + + + + W +W W - WGlycerol.r.o.I - - _ _ _ _ _ _ _ _ W + W W - wInositol.......--...---Inulin......-- _ _ _ _ _ _ _ _ _ _ _ _ _Lactose.......- - _ _ _ _ _ __ _ _ + - - -

Maltose...... -- W + + + + + +W +WW - WMannitol.......- -_ W + w w - wMannose......- - W W W+ w w - wMelezitose......- _ +Melibiose......- - _ _Raffinose......-- _ _Rhamnose...... w- __ _ w w WWSalicin ...... - - W ++ - -Sorbitol....- - _ - _ + -Sorbose....-- _ _Sucrose.....- _ _ _ _ _ _ _ _WV + - - -Trehalose..... + + - ._____ W + - -

Xylose .... + + - ._ _ _ _ W W WV W

a p, parent bacterial phase; L, L-phase variant; B, bacterial phase, not the parent of L-phase; R,revertant bacterial phase; +, positive reaction; -, negative reaction; W, weak reaction; V, variablereaction; F, fermentation; 0, oxidation; NR, no reaction.

pH throughout the medium of the test plate by at RESULTS AND DISCUSSIONleast a 0.4 pH unit compared with the controls. Aweak reaction was shown either by reduction of less In this study, 29 biochemical test proceduresthan a 0.4pH unit throughout the plate or by reduc- used in diagnostic bacteriology laboratories weretion of pH only in the area immediately surrounding examined. Tests were considered suitable forgrowth. use with L-phase organisms if (i) they were

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

1660 COHEN, WITTLER, AND FABER APPL. MICROBIOL.

TABLE 5. Results of biochemical tests performed on parent, L-, revertant, and independently isolatedbacterial phases of selected strains of organismsa

S. faecalis S. faecalis S. faecalis S. pyogenes Strepto- S. aureus S. aureus G. letragenaF-24 G-K D-TC Richards coccus Smith ATCC GilTest sp.BrunoGi

P L P L L R P L R B L P L P L B L R

Oxidase activity.. ..... -Catalase activity.. . . . + +Phosphatase ac-

tivity. .. W + + W+ + + +Oxidation-fer-mentation ...... F F F F F FF F F F F F F F F F NR F

Phenylalaninedeamination.... ....._

Esculin hydroly-sis.......+ + + +-

Arginine hydroly-sis......... + ++ +++++- -++±+- +

Ureahydrolysis..±+ + + + +Nitratereduction...±+ + + + W VTetrazolium re-duction. +W++++++ W+W + +++++++++++ NG +

Tellurite reduc-tion. + W + + +++NG + + + + + + + ++ +

AMC production.+ WV W WV WV+. W - W - _ _

CarbohydratebreakdownAdonitol..._Arabinose......+ + + +Cellobiose...... + + + + ++ +Dulcitol ........Fructose........++++ + ++ ++++±+++ - +Galactose. + + + + ++++ _+ ++++Glucose ........ + + + + + + + + + + + + + + ++ - +Glycerol. + ++ + + + ++ + + + +_Inositol....-- -

Inulin...-.- -

Lactose.... + + ++ + ++++ - +Maltose .... + + ++ + + + + + + + + + +Mannitol...... + ++ + + + _ + + + +_ _Mannose...... ++ + ++ + +-+±+++±- - +Melezitose..+ -_ -_Melibiose......- + +-Raffinose...-..Rhamnose...... W W W W WW .Salicin ..... + WV + + + + + -+Sorbitol ..... + + +Sorbose..._Sucrose... _ __ ++++++±++++- +Trehalose .... + + + + + + - ++++ - +Xylose...... - -+ + + + -

a P, parent bacterial phase; L, L-phase variant; R, revertant bacterial phase; B, bacterial phase, notthe parent of L-phase; +, positive reaction; -, negative reaction; W, weak reaction; V, variable re-action; NG, no growth; F, fermentation; NR, no reaction.

sufficiently sensitive to detect reactions of most of spective parent revertant bacterial phase or-the L-phase organisms; (ii) they yielded repro- ganisms.ducible results; and (iii) they yielded parallel For use with L-phase variants, 13 tests wereresults for most L-phase organisms and the re- found to be suitable. The most reliable of these

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

VOL. 16, 1968 TESTS FOR CHARACIERIZATION OF L-PHASE VARIANTS

included tests for oxidase, catalase, and phospha-tase activities, glucose oxidation or fermentation,phenylalanine deamination, esculin hydrolysis,and nitrate, tetrazolium, and tellurite reduction.Tests for breakdown of carbohydrates were alsoin this group, but because these abilities maymore readily undergo mutational changes, carbo-hydrate breakdown was not considered to be assignificant as the other tests in this group. Theremaining tests, those for arginine and ureahydrolysis and AMC production, yielded variableresults for certain L-phase variants. These threetests were apparently not sensitive enough to de-tect weak reactions consistently; therefore, onlypositive results for these tests could be consideredsignificant.The L-phase or revertant bacterial phase, or

both, of P. mirabilis 9, N. meningitidis 47 VI and55 III, Streptococcus faecalis F-24 and G-K, S.faecium D-TC, and Staphylococcus aureus ATCCretained essentially all the biochemical activitiesof the respective parent bacterial phase (Tables 4and 5). These results were in accord with recentresults establishing genetic relationships amongthe respective phases of P. mirabilis 9, N. menin-gitidis 47 VI and 55 III, and S. faecalis F-24 (10,15). The L-phase of S. pyogenes Richards also re-tained the activities of its parent bacterial phase.However, the revertant bacterial phase of thisstrain attacked only 2 out of 10 carbohydrates at-tacked by the parent bacterial and L-phaseorganisms.Minor differences were exhibited among four

other L-phase variants and their respective bac-terial phase organisms. The L-phase variants of S.moniliformis and of Streptococcus sp. Bruno werenot isolated from the particular strains used torepresent the respective bacterial phase; therefore,differences between biochemical capabilities ofthe two phases of these species may have beendue to strain variation. On the other hand, theL-phase of Corynebacterium sp. D Campo B andof S. aureus Smith was presumably deriveddirectly from the respective bacterial phase;therefore, differences may have been caused bymutation or contamination with a closely relatedspecies.Major differences were exhibted among the re-

maining three variants and their presumed parentor revertant bacterial phase organisms. The re-sults for the Corynebacterium sp. D Campo A setsuggested that the L-phase variant was a con-

taminant, probably derived from a group Dstreptococcus. The results for the Corynebac-terium sp. D-5 set suggested that the variant was

a contaminant, probably derived from a relatedCorynebacterium. Finally, the results for theGafftkya tetragena Gil set suggested that the

variant was a mycoplasmal contaminant. Thesebiochemical results were in accord with conclu-sions obtained from molecular genetic data forCorynebacterium sp. D Campo A and Coryne-bacterium sp. D-5 (5) and with conclusionsobtained from serological data for the Gil-Lorganism (M. C. Norman, unpublished data).The methyl red test and the test for indole

production were included in this study, but theydo not appear in the tables. No test organismsdemonstrated positive results for these tests;therefore, their usefulness for L-phase variantscould not be determined.The remaining 14 test procedures that were

examined were eliminated as unsuitable for usewith L-phase variants. Tests for growth on glu-cose phenolphthalein agar (pH 9.6), hydrolysisof hippurate and gelatin, utilization of citrate assole carbon source, and production of deoxy-ribonuclease were eliminated because the con-trols consistently yielded erroneous results.Tests for production of H2S, decarboxylation ofarginine, histidine, lysine, and ornithine, hemoly-sis of horse, sheep, human, chicken, dog, andrabbit red blood cells, and utilization of malonatewere eliminated because results varied from testrun to test run. Finally, tests for determiningtolerance to tellurite, methylene blue, increasedNaCl, bile, and ethylhydrocuprein were elimi-nated because results varied from run to run ordid not necessarily parallel those for the respec-tive bacterial phases, or both. Results of thelatter tests are discussed more fully by R. L.Cohen (Ph.D. Thesis, 1968, Univ. of Maryland,College Park, Md.).

It appears that biochemical characterizationcan be usefully employed to provide neededinformation on various occasions. For instance,the need to relate a specific L-phase organism to aspecific bacterial phase organism can arise afterisolation of an unstable L-phase variant fromclinical specimens, tissue cultures, or other bio-logical material, or during a periodic check foridentity of laboratory L-phase cultures. Theprocedure for establishing relatedness of anL-phase to a specific bacterial phase organismshould be performed by use of tests for oxidase,catalase, and phosphatase activities, glucoseoxidation or fermentation, phenylalanine deami-nation, esculin hydrolysis, nitrate, tetrazolium,and tellurite reductions, and carbohydratebreakdown. Tests for arginine and urea hydroly-sis and AMC production can also be used pro-vided results are carefully read and interpreted.The need to distinguish different species of

L-phase variants can arise when a laboratoryculture is suspected of containing a mixture ofL-phase organisms or when a specimen contains

1661

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

COHEN, WITILER, AND FABER

mixed flora. To distinguish different L-phasevariants, mixtures first should be purified byfiltration through membranes (Millipore Corp.)to remove clumps of cells, followed by repeatedsingle-colony isolation from the filtrates. Isolatesso obtained then should be characterized bio-chemically by using the above 13 tests.

Finally, the need to identify an unknown, non-reverting L-phase variant can arise upon isola-tion of such a variant from clinical specimens,tissue cultures, or other biological material.The solution of this problem requires utilizationof the 13 tests described above. The results oftests performed in this project indicated thatbiochemical characteristics alone were sufficientfor identifying only one of the L-phase variantsstudied, P. mirabilis L-9. Therefore, other testssuch as those utilizing molecular genetic, sero-logical, or disc-electrophoretic techniques, mustbe employed to identify a nonreverting L-phasevariant.

In conclusion, modified biochemical tests havebeen shown to have value for characterizationand identification of bacterial L-phase variants.These tests, redesigned for use with L-phaseorganisms by simple modifications of existingmethods, are easy to perform and to read. There-fore, they can be of considerable value for theclinical bacteriologist, as well as for the researchscientist. Although these tests must be used inconjunction with other techniques for identifyingmost nonreverting L-phase variants, they arehighly useful by themselves for relating anL-phase organism to its parent or revertant bac-terium and also for distinguishing differentL-phase variants.

ACKNOWLEDGMENTSWe are grateful to M. C. Norman, American Type

Culture Collection and Walter Reed Army Instituteof Research, Washington, D. C., who provided well-grown cultures of strains from the ATCC, and toR. Cole, S. Madoff, J. Pitcher, and C. Williams whoprovided other strains used in this study.

This investigation was supported by Public HealthService research grant Al 02332 from the NationalInstitute of Allergy and Infectious Diseases.

LITERATURE CITED1. Cowan, S. T., and K. J. Steel. 1965. Manual for

the identification of medical bacteria. Cam-bridge University Press, London.

2. Department of the Army. 1963. Laboratory pro-cedures in clinical bacteriology. TM 8-227-5.U.S. Government Printing Office, Washington,D.C.

3. Gibbs, B. M., and F. A. Skinner. 1966. Identifica-tion methods for microbiologists. AcademicPress, Inc., New York.

4. Kandler, G., and 0. Kandler. 1955. Ernahrungs-und Stoffwechselphysiologische Untersuchun-gen an Pleuropneumonieahnlichen Organismenund der L-phase der Bakterien. Zentr. Bakter-iol. Parasitenk. Abt. II 108:383-397.

5. McGee, Z. A., M. Rogul, and R. G. Wittler.1967. Molecular genetic studies of relationshipsamong mycoplasmas, L-forms, and bacteria.Ann. N.Y. Acad. Sci. 143:21-30.

6. Medill-Brown, M., W. G. Hutchinson, and E.Cocklin. 1960. The L forms of Proteus mirabilis.Ann. N.Y. Acad. Sci. 79:374-379.

7. Panos, C. 1962. Streptococcal L-forms. IV. Com-parison of the metabolic rates of a Streptococcusand derived L-form. J. Bacteriol. 84:921-928.

8. Roberts, R. B. 1968. The in vitro production,cultivation, and properties of L-forms ofpathogenic Neisseria, p. 230-238. In L. B. Guze(ed.), Microbial protoplasts, spheroplasts andL-forms. Williams & Wilkins Co., Baltimore.

9. Roberts, R. B., and R. G. Wittler. 1966. The Lform of Neisseria meningitidis. J. Gen. Micro-biol. 44:139-148.

10. Rogul, M., Z. A. McGee, R. G. Wittler, and S.Falkow. 1965. Nucleic acid homologies ofselected bacteria, L forms, and Mycoplasmaspecies. J. Bacteriol. 90:1200-1204.

11. Warren, J. 1942. Observations on some biologicalcharacteristics of organisms of the pleuro-pneumonia group. J. Bacteriol. 43:211-228.

12. Weibull, C., and H. Beckman. 1961. Chemicaland metabolic properties of various elementsfound in cultures of a stable proteus L form.J. Gen. Microbiol. 24:379-391.

13. Weibull, C., and K. Hammarberg. 1962. Occur-rence of catalase in pleuropneumonia-likeorganisms and bacterial L forms. J. Bacteriol.84:520-525.

14. Williams, R. E. 0. 1963. L forms of Staphylococ-cus aureus. J. Gen. Microbiol. 33:325-334.

15. Wittler, R. G., Z. A. McGee, C. 0. Williams,C. Burris, R. L. Cohen, and R. B. Roberts.1968. Identification of L-forms: problems andapproaches, p. 333-339. In L. B. Guze (ed.),Microbialprotoplasts, spheroplastsand L-forms.Williams & Wilkins Co., Baltimore.

1662 APPL. MICROBIOL.

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from