JOURNAL OF BACrERIOLOGY, Feb. 1968, p. 688-699Copyright © 1968 American Society for Microbiology

Vol. 95, No. 2Printed in U.S.A.

Isolation, Composition, and Structure of Membraneof Listeria monocytogenes1

B. K. GHOSH2 AND K. K. CARROLL3Departments of Bacteriology and Immunology and Medical Research, Uniiversity of Westernz Ontario, Lonidoni,

Ontario, Canada

Received for publication 13 November 1967

The plasma membrane of Listeria monocytogenes strain 42 was prepared by os-motic lysis of protoplasts with tris(hydroxymethyl)aminomethane (Tris) buffer, pH8.2, containing MgCl2 and glucose, followed by washing with NaCi and MgCl2 inTris buffer. Electron microscopy showed that the preparation was not contaminatedwith cytoplasmic material. The membrane preparation was composed of 55 to 60%oprotein, 1.5% ribonucleic acid, 0.1 % deoxyribonucleic acid, 1.3 to 2.3% carbohy-drate, 0.17 to 0.38% amino sugar, 0.2 to 0.4% rhamnose, 3.5 to 4.0% phosphorus,10.5 to 12.0% nitrogen, and 30 to 35% lipid. Amino acid composition of the washedmembrane showed some variation from that of the whole cells. Sulfur-containingamino acids were not present in the membrane hydrolysate. The membrane carbo-hydrate contained glucose, galactose, ribose, and arabinose. The membrane lipidwas 80 to 85% phospholipid and 15 to 20% neutral lipid. The lipid contained 2.3to 3.0% phosphorus, 2.5 to 3.0% carbohydrate, and a very small amount of nitrogen(0.2 to 0.3%). The phospholipid was of the phosphatidyl glycerol type. Electronmicrographs of the washed membrane showed three layers. The outer and innerlayers varied in thickness from 25 to 37 A and the middle layer from 20 to 25 A. Thetotal thickness varied between 85 and 100 A. These preparations contained manyvesicles which stained heavily with lead citrate. Some vesicles were also attached tothe protoplast ghosts in the form of extrusions or intrusions, or both. Membranepreparations obtained by lysis of protoplasts in the absence of MgCl2 were frag-mented and contained less lipid (20 to 22%) and ribonucleic acid (0.3 to 0.5%)than preparations prepared with MgC92.

The bacterial membrane has attracted theattention of many investigators since its isola-tion by Weibull (38, 39) as a ghost fraction oflysed protoplasts. Specialized subcellular mem-branous organelles, such as mitochondria, Golgiapparatus, and endoplasmic reticulum, have notbeen demonstrated in bacterial cells. However,mesosomes (intracytoplasmic membranous struc-tures) of varying morphology have been demon-strated frequently in the gram-positive bacteria(23), the actinomycetes (5, 36), and rarely in thegram-negative bacteria (30). The role of thesestructures as subcellular organelles has not beendemonstrated. The most widely accepted view isthat the peripheral membrane of the bacterial

1 Part of this investigation was presented at the 33rdAnnual Meeting of the Public Health Association,Toronto, Canada.

2Postdoctoral Fellow of the Medical ResearchCouncil of Canada, 1964-1966. Present address: In-stitute of Microbiology, Rutgers, The State Uni-versity, New Brunswick, N.J. 08903.3Medical Research Associate, Canada.

cell, besides controlling permeability, carriesout a wide range of metabolic work. Variousinvestigators have studied the chemical com-position (8, 17, 28, 33), the enzymatic makeup(7, 9, 13), the molecular anatomy (27), and thevariation in composition resulting from changesin nutrition and growth phase (28, 33) of themembrane.

Listeria monocytogenes is a gram-positive bac-terium, the cells of which have complex intra-cytoplasmic extensions of the plasma membrane(6, 10). The organism produces a lipid materialwhich stimulates monocytosis in certain animalsand in man (34). In gram-positive organisms,most of the lipid is concentrated in the plasmamembrane (15). Thus, it seems possible that thispharmacologically interesting lipid may be local-ized in the plasma membrane. These considera-tions prompted us to study the plasma membraneof this organism more closely. We previouslyreported (10) that some strains of L. monocyto-genes, including the one used for the present

688

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

LISTERIA MONOCYTOGENES MEMBRANE

study (strain 42), are lysozyme-resistant and be-come sensitive to lysozyme only after lipase treat-ment. This report describes the procedure forisolation of the plasma membrane of L. mono-cytogenes (strain 42), as well as the chemicalcomposition and structure of this membrane.

MATERIALS AND METHODS

Preparationi and lysis of protoplasts. L. monocyto-genes strain 42 was used throughout these experi-ments. Maintenance of the organism and conditionsof cell cultivation were described in a previous com-munication (10).

Protoplasts were formed by lipase and lysozymetreatment (10). The protoplasts [10 to 15 mg (dryweight)/ml] were washed in a solution of 0.03 Mtris(hydroxymethyl)aminomethane (Tris) buffer, pH6.8 to 7.0, 0.01 M MgCl2, and 0.5 M sucrose (Tris-Mg-sucrose) at room temperature (28 to 30 C). The washedprotoplasts were suspended in various lytic solutionsand were shaken at 37 C for 30 min to effect completelysis. After lysis, the optical densities of these solutionswere measured against a blank containing only thesuspending medium at 660 mMA. The lysate was cen-trifuged at 10,000 X g for 15 min, and the absorptionsof the resulting clear supernatant fluid were measuredat 260 and 280 m,u to assess release of intracellularsoluble materials. This method was used to choose asuitable lysing solution. The lytic solutions testedwere: (i) distilled water, (ii) 0.01 M MgCl2, (iii) 0.01M MgCl2 in 0.01 M Tris buffer (pH 8.1), (iv) 1.0 Mglycerol, and (v) 0.01 M Tris buffer (pH 8.1) containing0.02 M MgCl2 and 0.01 M glucose (Tris-Mg-glucose).

Preparation of the membrane. Lysis of the washedprotoplasts was carried out in Tris-Mg-glucose solu-tion as described above. The lysate was transferredinto a Sorvall Omnimixer bottle (Ivan Sorvall, Inc.,Norwalk, Conn.) and homogenized at 8,000 rev/minfor 10 min under ice and was then frozen at -20 Covernight. The next morning, the lysate was thawedand homogenized as above. The homogenate was cen-trifuged at 41,000 X g for 30 min and the clear super-natant fluid was kept as soluble cytoplasm. The residuewas suspended at room temperature (28 to 30 C) in asolution containing 0.01 M Tris buffer (pH 8.2), 0.05M NaCl, and 0.02 M MgCl2 (Tris-NaCl-Mg), and themixture was centrifuged at 105,000 X g for 45 min.The process was repeated twice. Finally, the residuewas washed and suspended in 0.01 M MgCl2. A flowdiagram of the preparation of washed membrane isshown in Fig. 1.

Analytical. A sample of the washed membrane sus-pensions was transferred to a cleaned and weighedaluminum dish, then heated at 70 to 80 C for 2 to 3hr, followed by overnight drying in vacuo over P205.The difference in weight after subtracting the equivalentblank of MgC12 gave the dry weight of the membranepreparation. Ribonucleic acid (RNA) and deoxyribo-nucleic acid (DNA) were extracted according to themethod of Ogur and Rosen (25), with the exceptionthat the initial precipitation was carried out with 0.2N perchloric acid (PCA) rather than with ethyl al-cohol. The amounts of separated RNA and DNA were

estimated by the determination of phosphorus (7),ribose (25), and deoxyribose (32). These results werechecked by ultraviolet absorption measurements. Pro-tein was estimated in 10% cold trichloroacetic acidprecipitate by Lowry's method (20). Total amino acidwas determined in 6 N HCl hydrolysate (for 18 hr)by Rosen's method (27). Individual amino acids wereanalyzed in a Beckman Spinco model 120 amino acidanalyzer (Beckman Instruments, Inc., Fullerton,Calif.). Carbohydrate was estimated by the anthronemethod (31) in the acid hydrolysate of the membrane(2 N HCl for 2 hr at 100 to 105 C), after removing theacid by treating with Dowex 50 ion-exchange resin.The reducing sugar was determined according to themethod of Nelson (24) and individual sugars wereidentified by paper chromatography (14). The rham-nose content was estimated by the spectrophotometriccysteine-sulfuric acid method (4), whereas glucosaminewas determined by the use of a modified Elson andMorgan reaction (26). Hydrolysis of the membranefor this determination was conducted in a sealedampoule in 6 N HCl for 2 hr at 100 to 105 C. The HCIwas removed by repeated evacuation over P205 andKOH pellets. Nitrogen was assayed by a modifiedNessler's reaction (19). For lipid analysis, a membranepreparation was lyophilized, after which the lipid wasextracted from the lyophilized membrane powder byshaking at room temperature with chloroform-methanol (2:1) for 12 to 18 hr. The extract was filteredand evaporated in a flash evaporator. The resultingresidue was dissolved in chloroform, and the quantityof residue was determined by weighing a sample ofthe solvent after evaporation under nitrogen. The lipidwas analyzed for nitrogen, phosphorus, carbohydrate,and amino acid. Phospholipid and neutral lipid weredetermined by Florisil column chromatography, andthe phospholipids were further characterized by thin-layer chromatography. (2).

Electron microscopy. Membrane structure wasexamined by use of a Phillips EM 100 electron micro-scope (Phillips Electronics and Pharmaceutical In-dustries Corp., St. Joseph, Mo.). Membrane prepara-tions in different stages of purification were prefixedin 3% glutaraldehyde and then fixed and embeddedaccording to the method used by Ryter and Kellen-berger (16). Details of the thin sectioning and stainingmethods were described in an earlier communication(10). Membrane preparations were negatively stainedwith 1% phosphotungstic acid (neutralized to pH 6.1with KOH) or were shadowed lightly (3:1 slope)with tungsten oxide in an Edward's Evaporator (model13 E6; Philips Electronic Instruments, Mt. Vernon,N.Y.

RESULTS

Changes during lysis. Protoplasts suspended inTris-Mg-sucrose gradually swelled and liberatedsmall amounts of protein and low molecularweight material absorbing at 260 m,u.The membranes of protoplasts suspended in

distilled water broke up into fragments. Frag-mentation could be prevented by the addition of0.01 M MgCl2. This was not a suitable washingsolution, however, as it prevented the release of

689VOL. 95, 1968

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GHOSH AND CARROLL

cytoplasmic protein and nucleic acid. The lysisin Tris-Mg-glucose was gradual, and there wasvery little fragmentation of the membranematerial. At an alkaline pH (8.1), Tris bufferprevented agglutination of the protoplast lysateand allowed greater solubilization of the cyto-plasmic material.The supernatant fluids of the lysozyme incuba-

tion mixture, the disintegrated protoplasts, andthe respective washes were collected separatelyand assayed for protein and nucleic acid (Fig. 1).Figure 2 illustrates the release of protein andnucleic acids on successive washings. Only a smallamount of protein was released by the lysozymetreatment, but there was no liberation of RNAand DNA, indicating no gross lysis of protoplastsin the lysozyme incubation medium. The suspen-sion of the protoplasts in Tris-Mg-glucose causedthe liberation of 53% of the total protein, 63%oof the total RNA of the cell, but only 6%o of thetotal DNA. The wash in Tris-NaCl-Mg removedabout 90%- of the total DNA, but less RNA andprotein were released.

After the fourth wash, successive washes re-moved a small but constant amount of RNA,DNA, and protein. This may be an indication ofpartial fragmentation of the membrane ratherthan removal of unbound or weakly bound non-membrane components (Fig. 2). Analysis of theratio of 260-mu absorption to 280-m,u absorptionindicated that the main release of materialabsorbing at 260 m,u occurred during the secondwash.During the successive washes of the membrane,

it was noted that about 20%c of the protein and10% of the RNA could not be recovered, whereasnearly all of the DNA was recovered. This maybe due to ribonuclease and proteolytic activityin the degraded cell. Results of these balanceexperiments (Table 1) show that the membranecontained 8 to 10% of the total cell protein.Practically all of the cellular lipid was present inthe membrane, and the membrane also retaineda significant amount of the cell's carbohydrate.

Composition. In common with all biologicalmembranes, the plasma membrane of L. mono-cytogenes is composed mainly of protein andlipid. The results of the general analysis of theplasma membrane, prepared both in the presenceand absence of MgCl2, are given in Table 2. Datafrom the analysis of the whole cell were includedfor comparison. Removal of cell wall materialwas indicated by the absence of rhamnose andamino sugar, both of which are typical cell wallcomponents. Table 1 illustrates the fraction ofcell components found in the cell membrane.Membrane preparations made in the presence of

MgCl2 contained a considerable amount of RNAwith respect to dry weight, but this representedonly about 2 to 3%7 of the total cellular RNA.Despite the small amount of RNA present, itseems to be a significant membrane component.RNA was very firmly bound to the membraneand could only be removed by ribonuclease treat-ment. However, the amount of RNA droppedto a very low value when the membrane was pre-pared in the absence of MgCl2. A very smallamount of DNA (0.8% of total DNA) was de-tected in the membrane prepared both in presenceand absence of MgCl2, but its significance was notapparent from the present experiments. Thephosphorus content of the membrane was lowerthan that of whole cells. The latter containedmore nucleic acid phosphorus than did themembrane. The membrane, on the other hand,contained almost all of the phospholipid phos-phorus. The total amount of phosphorus deter-mined in the membrane could not be accountedfor by the lipid and nucleic acid phosphorus.Therefore, other phosphorus-containing materialsmust remain bound to the membrane. It is notunlikely that components like teichoic acid existpartly as membrane-bound materials. A largeamount of phosphoserine was present in theprotein hydrolysate of the membrane.The general composition of the lipid is shown

in Table 3. The nitrogen content of this lipid wasvery low, and amino acids and choline were absentin this preparation. The major phospholipidfraction, obtained from an acid-treated Florisilcolumn, was further fractionated by thin-layerchromatography. A chloroform solution of thephospholipid fraction was applied to the thin-layer plate as a band, and this plate was thendeveloped in a chloroform-methanol-water (60:20:3) solvent system. Three opaque bands wereobtained after spraying with water: (i) lower(near origin), (ii) middle, and (iii) upper (nearthe running front). The nitrogen, phosphorus,and carbohydrate analyses of these fractions arepresented in Table 4. The results show that thesefractions have a low nitrogen content. The middlefraction is low in phosphorus but high in carbo-hydrate. Paper chromatography of the hydroly-sate of the middle fraction showed the presenceof glucose and galactose. The phosphorus con-centration was highest in the upper fraction. Pre-liminary characterization showed that the majorphospholipid may be of a phosphatidyl glyceroltype. Studies on the detailed characterization ofthis lipid are in progress (K. K. Carroll, H. JCutts, and E. G. D. Murray, unpublished data).The amino acid composition of the membrane

is presented in Table 5. Membrane and whole

690 J. BACTERIOL.

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

Supernatant fluid,discarded

Supernatant fluid,discarded

Washed lipase-treated cells of Listeria monocytogenes (980 mg, dry wt)

Centrifuged at 12,000 X g, 15 min

ICells,

suspended in 0.85% saline

Centrifuged at 12,000 X g for 15 min

Cells,treated with lysozyme (100 gg/ml) in Tris-Mg-sucrose

Centrifuged at 4,500 X g for 30 min

Supernatant fluid I ProtoplastsTris-Mg-glucoseIncubated at 37 C for 30 min with shaking

LysateHomogenized at 8,000 rev/min for 15 min

HomogenateFrozen at -20 C

Frozen homogenateThawed at 37 C

Thawed lysateHomogenized at 9,000 rev/min for 5 min

HomogenateCentrifuged at 41,000 X g for 30 min

Supernatant fluid 2 Residue

Homogenized in Tris-NaCl-Mg

Supernatant fluids 3-7

Supernatant fluiddiscarded

HomogenateCentrifuged at 105,000 X g, 45 minProcess repeated five times

ResidueHomogenized in 0.01 M MgC12

_ and centrifuged at 105,000 X g for 45 min

Washed membrane residueSuspended by homogenization

in small volume of 0.01 M MgC12

Washed membrane suspension(dry wt, 90 mg)

FIG. 1. Procedure for isolation ofpure Listeria monocytogenes membrane.

691

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GHOSH AND CARROLL

0

Lu

< 70Lu

uJ

60-

~~~~~~DNAz

Lu

L 10z

0

40 RNA

0

0

z 20-LUJ

J 0o PROTEI

2 3 4 5 6 7

SUPERNATANT NUMBER

FIG. 2. Release of proteini anid niucleic acid dutrintglysozyme inicuibationi, lysis of protoplasts, anid washinigof membrane. Superniatanit fluid niumber r-epresenitsniumber of sluperniatanit fluids designiated iin Fig. 1.

cell preparations contained a very large amountof aspartic acid, glutamic acid, alanine, glycine,lysine, and leucine. Ornithine was present onlyin trace amounts in the whole-cell preparation,but the membrane preparation contained a con-

siderably greater amount of this substance.Hydroxyproline was present in very smallamounts, and there was no peak for half cystine.Methionine could not be detected in the mem-brane preparation, but the whole cell preparationcontained about 0.8%--. In the whole cell prepara-

tion, there were two peaks in the acidic region,one of which was identified as methionine sulf-oxide. There were no such peaks in the mem-brane preparation. In both whole cell and mem-

brane preparations, there were peaks of phos-phoserine and glycerylphosphoryl ethanolamine.The carbohydrate composition of the mem-

brane is shown in Table 6. There was very littleglucosamine or rhamnose present. Qualitativepaper chromatography revealed that, in additionto hexoses, there were two pentoses, ribose andarabinose, in the membrane. The hexoses were

identified as mainly glucose and galactose, in

approximately equimolar amounts.The membrane generally constituted 7 to 10%-

of the dry weight of the whole cell. The yield ofthe membrane material prepared in the absenceof MgCl2 was reduced to 4 to 5', In the latter

TABLE 1. Proportioni of whole cellttlar materialretainled by the membraniea

Cellular material Content as percentage ofwhole cellular material

Protein...................... 8-10Lipid ....................... 88-91Carbohydrate................ 6-8RNA...... . 2-3DNA ....................... 0.8-1.0

a Figures represent the range of the values ob-tained in five different experiments.

preparation, the RNA and lipid content weremuch smaller than in the preparation made inthe presence of MgCl2 (Table 2). But the com-positions of lipid, protein, and carbohydrate werecomparable in both cases.The membrane preparation made without

MgCl2 fragmented into smaller parts which re-formed into small vesicles (Fig. 7 and 8). In addi-tion, many particles which could not be separatedby centrifugation at 105,000 X g were also re-leased. Both CaCl2 and SrCl2 could partially pro-tect the membrane against fragmentation.

Structure. A negatively stained preparation ofa typical protoplast used for the preparation ofplasma membrane is shown in Fig. 3. The prep-aration shows three flagella still attached to themembrane. Cells of L. monocyttogenes grown atroom temperature (28 C) showed the presenceof flagella (1 to 3 in number), but the cells grownat 37 C contained no flagella.An enlarged picture of the plasma membrane

is presented in Fig. 4. The plasma membrane,with some amorphous cytoplasmic material andinternal membranous structures still attachedcan be seen. In such preparations, a specializedregion can be noted frequently. The outer com-ponent of the three-layered membrane can branchand surround an inner membrane, which thentakes the form of a small vesicle.

Figure 5 shows a thin section of a membranepreparation after the third wash. A substantialamount of cytoplasmic material is still retainedby the membrane material. There are many con-centric vesicular structures of various sizes. Thesemay be reformed structures from the fragmentedplasma membrane and do not have any specialsignificance. There are some smaller vesicularstructures, very heavily stained with lead citrate,which may be different in character from thereformed vesicles. There are also many smallvesicles attached to the membrane which couldbe derived from the small invaginated structuresseen in the membrane of the whole cell. A mem-brane preparation washed eight times is illustrated

692 J. BACTERIOL.

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

LISTERIA MONOCYTOGENES MEMBRANE

TABLE 2. Cheemical composition of membranle of Listeria momiocytogenies strait? 421,

AMembrane

Whole cell (5sets)........

Membranewashed inpresence ofMgCI2 (7sets) .......

Membranewashed inabsence ofMgCI2 (3sets) ......

Protein

54-61

55-60

61-62

Totalaminoacid

RNA DNA

47-52 5.0-5.4 1.7-2.1

51-52 1.5-1.6 0.1-0.2

53-55 0.3-0.5 0.1-0.2

a Results expressed as percentage dry weight, whereasdifferent sets of experiments.

Totalnitrogen

10. 1-11.4

PhoUs- (hCa rbo-phors- hydratephrs,(anthrome)

5.

Rhamnose Amino sugarI Lipid

4.4-4.5 4-5

!| 10.5-12.0| 3.5-4.0~1.3-2.3| 0.2-0.4 0.17-0.381 30-35

3.7-4.15 1.7-2.0 .06-0.11 0.15-0.2 20-22

figures represent the range of variation in

TABLE 3. Gross conmpositioni of membrante lipidof Listeria moniocytogenies strait? 42',

Lipid As percentage of whole lipid

Phospholipid ............ 80-85Neutral Lipid ............ 15-20Phosphorus.............. 2.3-3Nitrogen................ 0.2-0.3Carbohydrate ............ 2.5-3.0

a Figures represent the range of variation in fivedifferent experiments.

TABLE 4. Gross compositionl of the threephospholipid fractions of' the membrante

lipid of Listeria molnocytogeniesseparated oni thini layer

CarbolhydrateFraction Nitrogen (antlhrone Phosphorus

method)

Upper. 0.45a 0.67 2.40Middle 0.16 12.00 0.65Lower . 0.37 1.00 1.10

aResults expressed as a percentage.

in Fig. 6. Here we see the complete disappearanceof cytoplasmic material, although the heavilystained vesicular structurs are still present. Cleanpreparations like this were used for chemicalanalysis.

Figures 7 and 8 present shadowed preparationsof membrane material, prepared with and withoutMgCl2, respectively. This clearly demonstrates thefragmentation of large protoplast ghosts intosmaller vesicles in the absence of MgCI2.Thin sections presented in the above illustra-

tions show that the membrane of L. monocyto-

TABLE 5. Aminio acid compositioni of the cytoplasmmembranie atnd whole cells of Listeria

molnocytogenies strailt 42

Amino acid

Alanine ........Arginine........Aspartic acid...Glutamic acid..Glycine.........Histidine.......Hydroxyproline.Isoleucine ......Leucine ........Lysine ..........Methionine .....Ornithine.......PhenylalanineProline .........Serine ..........Threonine......Tyrosine........Valine ..........

Concn [as %(dry wt) ofImembrane]

3.332.505.885.942.500.73trace1.993.826.23

1.551 .801.802.242.901 .221.93

Molar Conon las all Mlolarratio" (dry wvt) raiof wholecell] ai

12.54.313.713.414.01.7

6.610.89.8

1.04.56.59.68.22.36.9

4.62.373.545.11.870.86trace1.672.885.150.81trace1.311.571.621.711.162.1

9.32.54.86.24.51.0

2.34.06.31.0

1.52.52.82.61.13.2

{l Determined by dividing all the amino acidconcentrations (in micromoles) by the smallestamount.

genes has three layers (outer and inner, electron-dense; middle, electron-lucid) similar to otherbiological membranes. The total thickness of thethree layers varied from 85 to 100 A. The thick-nesses of individual components were: outer andinner, 25 to 37 A; middle, 20 to 25 A.

DISCUSSIONEstimates of the nucleic acid content of bac-

terial membranes made by various investiga-

693VOL. 951 1968

1-5.51 6.0-6.2~1.4-2.01

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GHOSH AND CARROLL

TABLE 6. Carbohydrate composition ofmembrane of' Listeria monzocytogenes

straini 42

Amt as per-Component centage of mem-

brane (dry wt)

Total sugar by anthrone method(as glucose) ...................... 2.3

Reducing sugar by Nelson Somyogimethod (as hexose) ............... 1.2

Rhamnose by cysteine sulfuric acidmethod (as methyl pentose) ....... 0.4

Total pentose by orcinol method(as Ribose) ....................... 0.82

Hexosamine (as glucosamine HCl).. 0.15

tors (11, 37, 40, 41) differ widely. The reasons forthis arise from the variety of preparative pro-cedures, particularly in washing procedures andin the use of hydrolytic enzymes (such as deoxy-ribonuclease or ribonuclease). The chief criterionof purity used in the present study was the appear-ance of electron micrographs of membranes aftereach treatment. We accepted the possibility thatmembranes may have nucleic acid and carbohy-drate bound to them in a special way, and there-fore we desisted in the use of injurious techniques,such as deoxyribonuclease or ribonuclease treat-ment.Our electron micrographs clearly indicate that

Mg is required to preserve membranes duringosmotic lysis. However, the slower lysis obtainedin Tris-Mg-glucose gives a superior membraneto those made with Tris-Mg.

Successive washing of the isolated membranewith various media indicated that the cytoplasmicmaterial contained in the membrane can be re-moved. The extreme viscosity of DNA in pro-toplast lysate was overcome by two cycles offreezing and thawing. Residual free DNA wasremoved most effectively in media containingNaCl. By use of an alkaline pH, aggregation ofthe membrane fraction was substantially reduced.The requirement for MgCl2 in the maintenance

of the structural and functional integrity of themembrane was reported by various workers.Lukoyanova (21) reported that succinic oxidaseactivity of Bacillus megaterium membrane de-pended upon optimal MgCl2 concentration.Abrams (1) reported that membrane-boundadenosine triphosphatase was released fromStreptococcus faecalis membrane if washed in theabsence of MgC92. Lampen (18) reported that therelease of membrane-bound penicillinase fromB. lichenijormis depended upon the integrity ofmembrane structure, and this structural stabilitycould be retained by an optimal concentration of

MgSO4. The role of Mg++ in relation to RNAcontent of the membranes has been discussed byMizushima (22). Our results also demonstratedthat Mg++ must be present to maintain the chemi-cal and physical integrity of the membrane. It isquite possible that the divalent cations act ascross-linking components between protein andvarious other constituents of membrane.There was very little DNA in the washed mem-

brane, and it is not known whether this smallamount of membrane-associated DNA (about0.8%W of wholeDNA content) had any significance.It is clear, however, that it is quite strongly bound.Although most of the RNA is free in the cyto-

plasm, 1.5 to 1.6% of the total is very stronglybound to the membrane. This binding, however,is markedly dependent on the presence of Mg,and in its absence the binding drops to 20 to 30%of this figure.The membrane of L. monocytogenes, prepared

by the method described, did not contain cell wallmaterial since the typical cell wall componentslike hexosamine or rhamnose were absent. Ab-sence of cell wall material was also demonstratedby electron microscopy. Therefore, the carbohy-drate material of the membrane appears to be asignificant membrane component. There aremany conflicting reports concerning the presenceof carbohydrate in the bacterial membrane.Weibull and Bergstrom (39, 40) reported 1 to10%o carbohydrate in a B. megaterium membrane.In the same organism, however, Godson et al.(12) reported about 40%7, carbohydrate. Gilby(11) showed the presence of 20%C mannose inMicrococcus lysodeikticus membrane.The protoplast membrane contained about

90% of the lipid of L. monocytogenes. The lipidwas very poor in nitrogen and contained glucoseand galactose. Phospholipid was themajor fractionof the total lipid. The neutral lipid content in-creased when the lipid was stored for some timebefore analysis. Presumably, this was due tospontaneous degradation of the phospholipids.Analysis of the lipid immediately after extractionshowed a 15 to 20% neutral lipid content. How-ever, the possibility of degradation and formationof neutral lipid during the extraction procedurecannot be excluded. The presence of considerableamounts of phosphorus and nitrogeninthe chloro-form-insoluble, but methanol-soluble, fractionindicates that proteolipid may be a component.No evidence is yet available concerning the

significance of glycolipid in the membrane. It wasrecently reported that membranes of L forms ofbacteria contain very large amounts of glycolipidwhen compared with the original protoplast.Elevated glycolipid and decreased phospholipid

694 J. BACTERIOL.

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

3

.~ ~ ~ ~ ~ ~ ~ 05;me4X>;1 , .

Sr @ ~I il F i

Or~~N,~~~~~~~~~~~~~~~Y

0.111

FIG. 3. Negatively stained preparations ofprotoplast with PTAFIG. 4. Thini sectioni of plasmai membranie stainied wit/h lead citrate. Note the branichinig of plasma membranie

(arrow) anid initernial membranie (IM) in the amorphouis cytoplaismic material retainied by the ghost.695

-

. .14I .. ., 4v V' '.5.lap" '.

Ili2r.'il.sl;

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

5o.2 II

V.N

;'

3'>'

'I,

4it:w:'.

.. .1 ,4, .,

Is ,... ..

.e :

.. s

§S :0',S

FIG. 5. Thinl sectioni ofplasma membrcanie preparationi, cafter the third was/I, stainied wit/i lead citrate. Note t/iecoiitaminiation wit/i ribosomes (r), amorp/iois cytoplasmic material (a), anid t/e fibrilar material (f). There are some/ieavily stainied vesicidlcar straictaires (arrow) atid some small vesicailar stractaires (v) origiliatitig from plasmna mem-

branze.696

I

.'iX

s:ii

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

,. v.3.. R ..a.. ..

, y_.t ... :tL: ,. ¢.

,,4. 4' .8

_'' * i' sSssX ... Ki'.'a..e_ ,, :.".

.. _,

.-,

*: .

I..

FIG. 6. Thliin sectioil of plasma membrane preparatioll, after sevenith wash/, stainied wit/i lead citracte. CAytoplasmicmaterial is completely remioved. Note tile heavily stainied vesicutlar strlctitres (arrow).

FIG. 7. Membrane prepared uwithl MgC/2. Preparation shadowed wit/i tiligsteit oxide.

697

0. 2 1J

-1.'.."4" 4,.-

l*. atrstl I.:!.s

4Mt

kft,

nf.

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GHOSH AND CARROLL

,Pl::

sis.iO.. 3....e' b

*. :.,.,e.

.; .Wd,.

-,&;,

s.S ¢ ..*

,%tf.... .. .: 4

iL6r ....F ::ffii' t.::*.:hg J.. St1 ' .ep.-:: . ,.

w ;we 4

'' .t ;r *X.

.,

"\,

74? ..

'T

*: .,2.s,:

...

*:AS-i.

::::':iA ':ri i

FIG. 8. Menbbranle prepared in absenice of MgCI2. Preparationshld(lowed1 wit/i tiigsteln oxide.

were thought to explain the inability of L-formcells to synthesize cell wall material (3). A phos-pholipid factor is required in a penicillin-insensi-tive but vancomycin-sensitive, step of cell-wallbiosynthesis (35). There is evidence which sup-ports the view that the lipids of membranes arepassive components of the permeaLbility barrier,as well as components of diverse biosyntheticprocesses of the cell.

Electron microscopic observations indicate theimportance of extensive washing to obtain a cleanmembrane preparation. However, since a smallamount of firmly bound cytoplasmic materialmay be present even after extensive washing,minor components, such as nucleic acids, shouldbe viewed, with reservations, as membrane com-ponents.

ACKNOWLEDGMENTSWe appreciate the support ol the Medical Research

Council of Canada. We are gratelul to R. G. E.Murray for his advice during the course of the re-search, to W. C. McMurray for pioviding facilities foramino acid analysis, and to John Marak for expertelectron microscopy. The authors also acknowledgethe assistance of Pamela W. Higgins, A. Ghosh, andM. G. Sargent in preparing the manuscript. Margaret

Ceneviva and D. Groot Obbink provided excellenttechnical assistance.

LITERATURE CITED

1. ABRAMS, A. 1965. The release of bound ATI'asefrom isolated bacterial membranes anid theproperties of solubilized enzyme. J. Biol. Chem.240:3675-368 1.

2. CARROLL, K. K. 1963. Acid-treated florisil as anadsorbent for column chromatography. J. Am.Oil Chemists' Soc. 40:413-419.

3. COHEN, M., AND C. PANOS. 1966. Membrane lipidcomposition ol Streptococcius pjIogenes andderived L form. Biochemistry 5:2385-2392.

4. DISCHE, Z., AND B. L. SHERLES. 1948. A specificcolor reaction of methyl pentoses and a spectro-photometric micromethod for their determina-tion. J. Biol. Chem. 175:595-603.

5. EDWARDS, M. R., AND M. R. GORDON. 1962.Membrane system of Actenomnyces bolvis. Intern.Congr. Electron Microscopy, 5th, Philadelphia,2:3.

6. EDWARDS, M. R., AND R. W. STEVENS. 1963. Finestructure of Listeria monocytogenes. J. Bacteriol.86:414-428.

7. ERNSTER, L., R. ZETTERSTROM, AND 0. LINDBERG.1950. A method for the determinationi of tracerphosphate in biological material. Acta. Chem.Scand. 4:942 947.

698 J. BACTLRI()L.

W;,L.::,....:.

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

LISTERIA MONOCYTOGENES MEMBRANE

8. FREIMER, E. H. 1963. Studies of' L forms andprotoplasts of group A Streptococci. II. Chemi-cal and immunological properties of cell mem-brane. J. Exptl. Med. 117:377-399.

9. GEL'MAN, N. S., AND M. A. LUKOYANOVA. 1962.Electron carriers in the respiratory chain andtheir relation to structures in the bacterial cell.Microbiology (USSR) 31:453-463.

10. GHOSH, B. K., AND R. G. E. MURRAY. 1967. Finestructure of Listeria moniocytogenes in relationto protoplast formation. J. Bacteriol. 93:411-426.

11. GILBY, A. R., A. V. FEW, AND K. MCQUILLEN.1958. The chemical composition of protoplastmembrane of Micrococcus lysodeikticus. Bio-chim. Biophys. Acta 29:21-29.

12. GODSON, G. N., G. D. HUNTER, AND J. A. VBUTLER. 1961. Cellular components of Bacillusmegaterium and their role in protein biosyn-thesis. Biochem. J. 81:59-68.

13. HUGHES, D. E. 1962. The bacterial cytoplasmicmembrane. J. Gen. Microbiol. 29:39-46.

14. JERMYN, M. A., AND F. A. ISHERWOOD. 1949.Improved separation of sugars on paper parti-tion chromatography. Biochem. J. 44:402-407.

15. KATES, M. 1964. Bacterial lipids, p. 17-90. In R.Paoletti and D. Kritchevsky [ed.], Advancesin lipid research, vol. 2. Academic Press, Inc.,New York.

16. KELLENBERGER, E., A. RYTER, AND J. SECHAUD.1958. Electron microscope study of DNA-con-taining plasms. II. Vegetative and mature phageDNA as compared with normal bacterialnucleoids in different physiological states. J.Biophys. Biochem. Cytol. 4:671-676.

17. KODICEK, E. 1962. Aspects of the constitution ofbacterial membranes p. 23-37. In Recent prog-ress in microbiology, Intern. Congr. Micro-biol., 8th, Montreal, Univ. of Toronto Press,Toronto.

18. LAMPEN, J. 0. Release of penicillinase by Bacilluslicheniformis. J. Gen. Microbiol. 48:261-268.

19. LANG, C. A. 1958. Simple microdetermination ofKjeldahl nitrogen in biological material. Anal.Chem. 30:1692-1693.

20. LOWRY, 0. H., N. J. ROSEBROUGH, A. L. FARR,AND R. J. RANDALL. 1951. Protein measure-ments with the Folin phenol reagent. J. Biol.Chem. 193:265-275.

21. LUKOYANOVA, M. A., N. S. GEL'MAN, AND A. I.OPARIN. 1963. Succinate oxidation in Micrococ-cus lysodeikticus structural elements with refer-ence to deoxyribonucleic acid and Mg ions.Biochemistry (USSR) 28:334-339.

22. MIZUSHIMA, S., M. ISHIDA, AND K. KITAHAVA.1966. Chemical composition of protoplastmembrane of Bacillus megaterihm. J. Biochem.(Tokyo) 59:374-381.

23. MURRAY, R. G. E. 1963. The organelles of bac-teria, p. 28-41. In D. Mazia and A. Tyler [ed.],General physiology of cell specialization.McGraw-Hill, Book Co., Inc., New York.

24. NELSON, N. 1944. A photometric adaptation ofthe Somogyi method for the determination ofglucose. J. Biol. Chem. 153:375-380.

25. OGUR, M., AND G. ROSEN. 1950. The nucleicacids of plant tissues. I. The extraction andestimation of desoxypentose nucleic acid andpentose nucleic acid. Arch. Biochem. Biophys.25:262-276.

26. RANDLE, C. J. M., AND W. T. J. MORGAN. 1955.The determination of glucosamine and galac-tosamine. Biochem. J. 61:586-589.

27. ROSEN, H. 1957. A modified ninhydrin colori-metric analysis for amino acids. Arch. Biochem.Biophys. 67:10-15.

28. SALTON, M. R. J., AND J. H. FREER. 1965. Com-position of membrane isolated from severalgram-positive bacteria. Biochim. Biophys.Acta 107:531-533.

29. SALTON, M. R. J., AND K. NETSCHEY. 1965. Physi-cal chemistry of isolated bacterial membrane.Biochim. Biophys. Acta 107:539-545.

30. SCHNAITMAN, C., AND J. W. GREENAWALT. 1966.Intracytoplasmic membranes in Escherichiacoli. J. Bacteriol. 92:780-783.

31. ScoTT, A. T., AND E. H. MELVIN. 1953. Determina-tion of dextran with anthrone. Anal. Chem.25:1656-1661.

32. SEIBERT, F. B. 1940. Removal of the impurities,nucleic acid and polysaccharide, from tuber-culin protein. J. Biol. Chem. 133:593-604.

33. SHOCKMAN, G. D., J. J. KOLB, B. BAKAY, M. J.CONOVER, AND G. TOENNIES. 1963. Protoplastmembrane of Streptococcus faecalis. J. Bac-teriol. 85:168-176.

34. STANLEY, N. F. 1949. Studies on Listeria mono-cytogelies. I. Isolation of monocyte producingagent (MPA). Australian J. Exptl. Biol. Med.Sci. 27:123-131.

35. STROMINGER, J. L. 1965. Biochemistry of the cellwall of Staphylococcus aureus. Ann. N.Y. Acad.Sci. 128:59-61.

36. STUART, D. C., JR. 1959. Fine structure of thenucleoid and internal membrane systems ofStreptomyces. J. Bacteriol. 78:272-281.

37. VENNES, J. W., AND P. GERHARDT, 1956. Immuno-logic comparison of isolated surface membranesof Bacillus megaterium. Science 124:535-536.

38. WEIBULL, C. 1953. The isolation of protoplastsfrom Bacillus megaterium by controlled treat-ment with lysozyme. J. Bacteriol. 66:688-695.

39. WEIBULL, C. 1953. Characterization of the proto-plasmic constituents of Bacillus megaterium.J. Bacteriol. 66:696-702.

40. WEIBULL, C., AND L. BERGSTROM. 1958. The chem-ical nature of the cytoplasmic membrane andcell wall of Bacillus megaterium strain M. Bio-chim. Biophys. Acta 30:340-351.

41. YUDKIN, M. D., AND B. DAVIS. 1965. Nature ofthe RNA associated with the protoplast mem-brane of Bacillus megaterium. J. Mol. Biol.12:193-204.

VOL. 95, 1968 699

on August 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from