Vol. 23, No. 3INFECTION AND IMMUNITY, Mar. 1979, p. 845-8570019-9567/79/03-0845/13$02.00/0

Isolation, Characterization, and Biological Properties of anEndotoxin-Like Material from the Gram-Positive Organism

Listeria monocytogenesHANNAH WEXLER AND JOEL D. OPPENHEIM*

Department ofMicrobiology, New York University School ofMedicine, New York, New York 10016

Received for publication 20 November 1978

The bacterial component responsible for the induction of transient cold agglu-tinin syndrome in rabbits after intravenous injection of heat-killed Listeriamonocytogenes type 4B has been purified and biologically and chemically char-acterized. A purified immunoglobulin M cold agglutinin was prepared from high-titer sera resulting from the immunization of rabbits with heat-killed L. mono-cytogenes type 4B and was subsequently used to monitor the purification of thebacterial component responsible for its induction. The bacterial component wasisolated from a hot phenol-water extract of lyophilized L. monocytogenes type4B by multiple molecular sieve chromatography. Upon chemical analysis thepurified material was found to be strikingly similar in chemical composition togram-negative lipopolysaccharide endotoxins. The material contained 15% totalfatty acid (of which 50% was fB-hydroxymyristic acid), 40 to 45% neutral sugar(glucose, galactose, and rhamnose), 11.5% amino sugar, 12% uronic acid, 2.5% 2-keto-3-deoxyoctonic acid, 2% heptose, 0.87% phosphorus, and 1.6% amino acid,thereby accounting for 85 to 90% of the weight of the component. Electronmicrographs of the purified material were similar to those of lipopolysaccharidepreparations from gram-negative organisms. The purified material exists inaqueous solutions as large aggregates, but can be dissociated into a single smallersubunit (3.1S) by dialysis against sodium dodecyl sulfate buffer. The listerialcomponent was toxic and pyrogenic to rabbits, producing symptoms typical ofgram-negative endotoxins. Activity in the limulus lysate gelation assay and in thecarbocyanine dye assay provides a further link of this material with classicalgram-negative endotoxins.

The clinical significance of Listeria monocy-togenes, a gram-positive diphtheroid-like rod ofthe family Corynebacteriaceae, is becoming in-creasingly recognized (13). Infection in humansmay result in encephalitis, meningitis, septi-cemia, and abortion. A peculiar result of somesepticemic listeric infections in humans is theinduction of a transient cold agglutinin syn-drome in which antibodies are induced that canreact with the host's own erythrocytes underappropriate conditions, causing a complement-mediated in vivo lysis (17). The antibodies re-sponsible for this phenomenon are of the im-munoglobulin M (IgM) class and appear to bedirected against the I blood group system (5).

Costea and co-workers (5) developed an ani-mal model of this response by the intravenousinjection of heat-killed L. monocytogenes sero-type 4B (HKLM) into rabbits. The cold agglu-tinin syndrome so produced was found to mimicthat which occurs in humans in symptomologyand pathology and was caused by IgM antibod-

ies. Their results further indicated that this IgMantibody was not only induced by and directedagainst the HKLM, but was also able to cross-react with the erythrocytes of the host as wellas other xenogenic erythrocytes which possess Iblood group antigen(s). Costea et al. (7) weresubsequently able to show that a crude fractionobtained from HKLM by a hot phenol-waterextraction procedure typically used for extract-ing endotoxin from gram-negative bacteria wasactive in inhibiting the cold agglutinating activ-ity of the induced antibodies. They describedthis fraction as a crude lipopolysaccharide(LPS), presumably because of the nature of theextraction procedure, but did not further char-acterize the component.A toxic LPS-like component isolated from a

phenol-water extract of L. monocytogenes wasdescribed and characterized by Conklin and co-workers (4). The extraction procedure used wassimilar to the method used by Costea et al. (7),except for one major distinction: the bacteria

845

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

846 WEXLER AND OPPENHEIM

were not heat killed before extraction. Heatingin aqueous solution at 80'C may remove surfacecomponents from bacteria (such as endotoxinfrom gram-negative bacteria [30]); thus, the heattreatment used by Costea et al. (7) may haveremoved a large portion of the LPS materialbefore their phenol extraction procedure. Conk-lin and co-workers characterized this materialand found that it was toxic to chicken embryosand contained carbohydrate, including 2-keto-3-deoxyoctonic acid (KDO), lipid, protein, nucleicacid, and phosphorus.The present study was undertaken to isolate

and characterize the bacterial component re-sponsible for cold agglutinin syndrome in hu-mans and animals. Rabbits were immunizedwith HKLM, and cold agglutinin purified fromthe sera so obtained was used to monitor thepurification of the component responsible for itsinduction. The chemical composition, physicalproperties, and biological activity of this isolatedcomponent were found to bear a striking simi-larity to classical gram-negative LPS endo-toxins.

MATERIALS AND METHODSBacterial cell cultures. L. monocytogenes sero-

type 4B (American Type Culture Collection no.19115), a virulent strain (intraperitoneal inoculation of108 organisms caused death in 10 of 10 SwissHA/ICR mice within 72 h), was grown in a brain heartinfusion broth (Difco). Overnite cultures (50 ml) wereinoculated into fresh media and grown in a MicrofermLaboratory fermentor (New Brunswick Scientific Co.)at 260C. Log-phase bacteria were harvested by cen-trifugation at 8,000 rpm in a JA10 rotor in a BeckmanJ-21C refrigerated centrifuge for 15 min, washed twicein saline and twice in distilled water by centrifugation,and finally lyophilized.Immunization and serology. Antibody to HKLM

was raised in female New Zealand white rabbits (2 to3 kg) by intravenous injection with 1010 HKLM dividedinto three equal doses given at 0, 4, and 24 h. Bloodsamples were collected from the ear vein and allowedto clot at 370C. The serum was subsequently separatedby centrifugation at 370C, filtered through a 0.45-,ummembrane filter (Millipore Corp.), heated at 560C for30 min, and stored at -20'C without preservative.

Cold hemagglutination assays were performed byserial dilutions of 0.2 ml of serum (or purified IgMantibody) in phosphate-buffered saline (PBS), pH 6.8,to which was added 0.05 ml of a 2% (vol/vol) suspen-sion of either isologous or autologous rabbit erythro-cytes washed three times. After mixing, the tubes wereincubated at 4°C for 2 h. The agglutination end pointwas determined macroscopically after centrifugationfor 15 s in an Adams serofuge. Titers were recorded asthe reciprocal of the highest dilution of serum exhibit-ing definite agglutination. Hemagglutination inhibi-tion assays were performed similarly by using either a1% antigen solution or fractions directly off chroma-tography columns and an antiserum or purified im-

munoglobulin dilution with a final hemagglutinationtiter of 4 to 8. All steps of these assays were performedand read at 4°C. The erythrocytes used were eitherautologous or isologous as indicated. As Lind sug-gested (18), to evaluate the rise in cold antibody titerduring immunization, consecutive sera from one ani-mal, including nonimmune sera, were tested simulta-neously in the same titration experiment. Anti-Liste-ria antibody levels were determined by slide aggluti-nation assay with microflocculation slides. Titrationswere performed by serial dilutions of 0.1 ml of serum(or purified antibody fraction) in PBS to which anequal volume of HKLM suspension containing 1010organisms per ml in PBS was added. Slides wereincubated for 5 min at 22°C on a rotary shaker. Theend point was determined macroscopically as the lastdilution in which visible agglutination could be ob-served.To differentiate IgM from IgG antibody popula-

tions, rabbit sera or purified immunoglobulin fractionswere incubated at 37°C for 1 h with an equal volumeof 0.1 M 2-mercaptoethanol in PBS, pH 7.4, before usein the appropriate assay system. This procedure causescomplete inactivation of IgM antibodies but does notaffect IgG immunoglobulin (24).

Purification of cold agglutinin. The rabbit IgMcold agglutinin antibody population was purified fromhigh-titer whole sera by a modification of the glutar-aldehyde-treated erythrocyte membrane affinity chro-matography procedure described by Tsai et al. (40) forthe isolation of human IgM cold agglutinin. Glutaral-dehyde-treated isologous rabbit papain-treated eryth-rocyte membrane was prepared as follows. Erythro-cytes were washed three times in PBS, pH 7.0, andthe buffy coat was removed; 100 ml of packed eryth-rocytes was then incubated with an equal volume of0.4% papain in PBS, pH 7.0, at 37°C for 30 min in ashaking water bath, after which the erythrocytes werewashed three additional times with PBS. The eryth-rocytes were lysed in 30 volumes of distilled water, pH5.8. The resultant membranes were harvested by cen-trifugation at 8,000 rpm for 10 min in JA10 rotor andwashed twice with the lysis solution and once withPBS; the membranes were then suspended in 100 mlof 2.5% glutaraldehyde in PBS and incubated in a37°C shaking water bath for 2 h. A one-fifth volume of0.5 M glycine in PBS was then added, and the mixturewas allowed to incubate for an additional 5 min. Thestromata were washed three times with PBS contain-ing 0.1 M glycine and twice with PBS, pelleting thestroma each time by centrifugation for 5 min at 20,000rpm. The affinity column was prepared by mixing 25ml of a loosely packed suspension of the fixed mem-branes with 50 ml of a 50% suspension of celite in PBS,then pouring the slurry over a 2-cm base of acid-washed sand in a chromatographic column (5 by 20cm) under mild suction, and finally overlaying with anadditional 2 cm of sand. The column was washedextensively with PBS at both 4 and 37°C before use.Absorption procedures. Purified listerial LPS

samples (100 ,ug) were incubated with 2.5 ml of 2%suspensions of five-times-washed sheep erythrocytesfor 1 h at 4, 25 and 37°C in a shaking water bath. Afterincubation the erythrocytes were washed three timesin PBS and resuspended to 2%. Agglutination reac-

INFECT. IMMUN.

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

LPS-LIKE COMPONENT FROM L. MONOCYTOGENES 847

tions were then performed by using the high-titer anti-Listeria sera produced in rabbits and tanned sheeperythrocytes at 4, 25, and 370C, as previously de-scribed, with autologous rabbit erythrocytes as a con-trol.

Extraction of crude LPS. A crude LPS fractionwas obtained from lyophilized cells by using a modifi-cation of the phenol extraction procedure describedby Conklin (4). Lyophilized cells (7 g) were suspendedin 170 ml of distilled water (6500), mixed with 170 mlof 90% phenol at 650C, and shaken for 8 min. Phaseswere separated by centrifugation for 30 min at 10,000x g in a JA10 rotor at 40C, and the aqueous upperlayer was removed. The phenol layer was decanted,filtered, washed with an additional 170 ml of distilledwater, and centrifuged as described above. Theaqueous phases were combined and dialyzed exten-sively, first versus cold running tap water and thenagainst distilled water. The dialysate was reduced toone-tenth of its original volume by coating the dialysistubing with Aquacide I (Calbiochem). The concen-trated solution was centrifuged at 10,000 rpm in aJA20 rotor; the supernatant was decanted and saved,and the precipitate was discarded. Crude LPS wasprecipitated by slowly adding the supernatant to 5 to10 times its volume of cold methanol (-20'C) contain-ing 1% sodium acetate. The precipitate was collectedby centrifugation at 10,000 rpm for 10 min at -10CC,dissolved in water, and lyophilized.

Purification of LPS. Initial purification of LPSwas achieved by chromatography on Sephadex G-25.All fractions were monitored for absorbance at 260 nmas well as ability to inhibit the hemagglutinating activ-ity of the purified cold agglutinin. The relevant frac-tions were pooled and lyophilized. The lyophilizedmaterial was resuspended in distilled water (20 mg/ml)and treated with 2 volumes of packed, washed Enziteagarose ribonuclease (Miles Laboratories) on a rotatorat room temperature for 3 h. The mixture was passedthrough a scintered glass filter to remove the beads,which were then rinsed with several volumes of water.The combined filtrates were lyophilized, and the re-sultant material was suspended in 0.15 M of ammo-nium carbonate and chromatographed on a SephadexG-25 fine column. The relevant fractions as deter-mined by inhibitory activity were pooled and lyophi-lized. Final purification was achieved by chromatog-raphy on Sepharose 4B. The eluant in all columns was0.15 M ammonium carbonate.

Colorimetric assays. Hexose was measured bythe anthrone method of Scott and Melvin (32). Hex-osamine was determined by the method outlined byBlumenkrantz and Absoe-Hansen (3). Uronic acidswere measured by the method of Dische (8). KDO wasmeasured by the method of Anacker et al. (1). Heptosewas measured by the procedures described by Dische(8). Methyl pentose was determined by the method ofDische and Shettles (9). Phosphorus was determinedby Shwartzkopf Microanalytical Laboratories, Wood-side, N.Y.

Total fatty acids were determined by the method ofDuncombe (10) by using a chloroform extract of ahydrolysate (2 N HC1, 3 h, 10500) of the purifiedmaterial.

Gas-liquid chromatography. Individual sugars

were identified and quantitated by gas-liquid chro-matography on a model 400 F + M biomedical gaschromatograph. Neutral sugars were identified andquantitated by chromatography as their trimethyl silylethers on a 15% Carbowax 20 M column (AppliedScience Laboratories) (27, 28). Neutral sugars, as wellas glucuronic acid and KDO, were also identified astrimethyl silyl ethers of their methyl glycosides on a3% SE-30 column (Applied Science) by the methodoutlined by Sweeley and Walker (39).Amino sugars were identified on a model 120C

amino acid analyzer (Beckman Instruments, Inc.). Thematerial was hydrolyzed (8 h, 4 N HOC, 10000), andthe HOC was evaporated on a rotary evaporator. Theresidue was dissolved in water, extracted with hexaneto remove fatty acids, dried, and redissolved in 0.02 MHOC before introduction into the analyzer. Aminosugar analysis was done in the laboratory of RichardMargolis, Department of Medicine, New York Univer-sity School of Medicine.

Individual fatty acids were identified and quanti-tated by gas-liquid chromatography of their methylesters on polar (10% Silar 10C) and nonpolar (3% SE-30) columns, both obtained from Applied Science.Fatty acid methyl esters were prepared from a chlo-roform extract of an LPS hydrolysate (2 N HOC, 3 h,1100C) according to the procedure described by Kates(16). Fatty acid methyl ester standards were purchasedfrom Applied Science.Amino acid analysis. Amino acids were deter-

mined and quantitated on a Durrum D-500 amino acidanalyzer by using standard hydrolysis methodology inthe laboratory of Edward Franklin, Irvington HouseInstitute, New York University School of Medicine.

Thin-layer chromatography. Thin-layer chro-matography was performed on precoated thin-layerchromatography plates (0.25 mm of Silica Gel 60 onglass supports [20 by 20 cm]; E. M. Laboratories,Merck). The chromatography was performed on alipid extract of listerial LPS obtained by the extractionprocedure of Bligh and Dyer (2). The solvent systemused was chloroform-methanol-water-acetic acid (65:25:4:1), and development of the plate was carried outby using iodine vapor. The chromatography was per-formed in the laboratory of Peter Elsbach, Depart-ment of Medicine, New York University School ofMedicine.

Analytical ultracentrifugation. Analytical ultra-centrifugation was carried out with a Beckman modelE analytical ultracentrifuge at 220C by Marylyn Hor-witz, New York Blood Center. Lyophilized, purifiedlisterial LPS was either resuspended in PBS, pH 7.0,and directly run or resuspended and dialyzed againsttris(hydroxymethyl)aminomethane (Tris)-ethylene-diaminetetraacetate buffer (0.05 M Tris-hydrochlo-ride, 0.005 M ethylenediaminetetraacetate, pH 8.0) orTris-sodium dodecyl sulfate (SDS) buffer (0.05M Tris-hydrochloride, 0.001 M SDS, pH 8.0) before centrifu-gation.

Slide and gel electrophoresis. Slide zonal elec-trophoresis was performed in 0.5% (wt/vol) agarosegels containing barbital-HCl buffer (ionic strength =0.02, pH 8.4) cast on glass slides (8.3 by 10.2 cm) witha ratio of volume to surface area of 0.132 Mi/cm2. Afterapplication of 4-ru samples, gels were subjected to

VOL. 23, 1979

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

848 WEXLER AND OPPENHEIM

electrophoresis at a 20-mA constant current per slidefor 45 min in a Behring Diagnostic water-cooled im-munoelectrophoresis chamber using the barbital-HClbuffer and Whatman no. 1 filter paper wicks. Afterelectrophoresis, slides were dried directly by a streamof hot air and then stained with toluidine blue (0.15%toluidine blue in 2% ethanol containing 1% acetic acid)and destained with 10% acetic acid in 40% ethanol.Polyacrylamide-SDS gel electrophoresis was per-formed as described by Fairbanks et al. (11). Gels werestained with either periodic acid-Schiff stain (38) orCoomassie brilliant blue (11).UV spectra. Ultraviolet (UV) spectra were deter-

mined on a Carey model 14 recording spectrophotom-eter (Varian Associates). Samples were dissolved ineither 0.1 N HCl, 0.1 N NaOH, or PBS, pH 7.0.

Infrared spectra. The infrared spectra were ob-tained on a Perkin-Elmer model 421 grating infraredspectrophotometer by using KBr pellets prepared bymixing 0.75 mg of material with 100 mg of KBr andpressed.

Biological assays. The limulus lysate assay wasperformed with a Pyrostat test kit (Worthington Bio-chemicals Corp.). Carbocyanine dye assay was per-formed by the method of Janda and Work (15).

Pyrogenicity was determined in 2-kg New Zealandwhite female rabbits according to instructions outlinedin the United States Pharmacopeia (United StatesPharmacopeia XIX, p. 613, 1975). Rectal tempera-tures were determined with a Yellow Springs Instru-ments model 425-G tele-thermometer. Listerial LPSsolutions and a saline control were injected intrave-nously in the ear vein.

Electron microscopy. Purified LPS was resus-pended in distilled water and placed directly on Form-var carbon-coated grids. The dried samples were neg-atively stained with 2% ammonium molybdate andexamined in a Siemens Elmiskop I electron micro-scope.

RESULTSProduction of cold agglutinin with

HKLM. To be able to identify and immunolog-ically monitor the purification of the bacterialcomponent responsible for the induction of thecold agglutinin syndrome, cold antibody waselicited in rabbits and subsequently purified.Rabbits injected with HKLM responded with aparallel increase in both cold agglutinin and anti-listerial activity (Fig. 1), as determined by coldhemagglutination and bacterial slide agglutina-tion assays. In all 10 rabbits tested nonimmunesera exhibited low levels of naturally occurringIgM cold antibody, with titers ranging from 16to 64. In the nonimmune sera of 2 of 10 rabbitsvery low levels (titer of 2) of an apparent IgManti-Listeria antibody were also observed. In allof the rabbits increases both in cold agglutininand in anti-listerial activity followed a 30- to 40-day cycle. Both activities increased within 72 hafter intravenous injection, peaked between 9and 20 days, and subsequently began to decline.

INFECT. IMMUN.

I-IJ

0 10 " 40 50

DAYSFIG. 1. Levels of anti-listerial and cold agglutinin

antibodies in a typical rabbit injected with HKLM.;Arrows indicate injections at days 0 and 40 with 10'0HKLM.

All initial antibody produced was 2-mercapto-ethanol sensitive. Anti-listerial IgG antibody (2-mercaptoethanol resistant) was not detectedwithin the first 7 days after inoculation but thenbegan to rise. All cold agglutinin activity wasalways found to be 2-mercaptoethanol sensitive.After 40 days cold agglutinin activity was ap-proximately back to basal levels, whereas anti-listerial activity was still significantly increased.If, at this time, the rabbits are reinjected withHKLM, a new 30- to 40-day cycle in cold agglu-tinin and anti-listerial activities was initiated.

Purification of cold agglutinin. The puri-fication of the cold agglutinin from high-titersera was achieved by using a thermal affinitychromatographic procedure (Fig. 2), followed bymolecular sieve chromatography. The affinitycolumn was prepared by using glutaraldehyde-fixed rabbit erythrocyte membrane mixed withsand and celite; high-titer sera were passedthrough the column at 4°C, and the column wasthen rinsed extensively with cold buffer. Thecolumn was then warmed to 37°C and elutedwith buffer at the same temperature. Columnfractions were monitored for both absorbance at280 nm and cold agglutinating activity in thehemagglutination assay. No cold agglutinin ac-tivity was eluted from the column at 40C, but itwas eluted at the higher temperature. The frac-tions containing cold agglutinin activity werepooled and concentrated in an Amicon Diaflocell. Upon immunodiffusion analysis of this ma-terial against anti-rabbit whole sera, it was de-termined that further purification was needed.

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

LPS-LIKE COMPONENT FROM L. MONOCYTOGENES 849

GRAPHY

0

4

FRACT ION

FIG. 2. Purification of cold agglutinin byaffinity chromatography. High-titer immursera were titrated to pH 6.8, membrane(Millipore Corp.), and subsequently passeda rabbit membrane celite column (5 by 20was prepared as described in the text. Thowas washed extensively with PBS at 40Celated with PBS at 370C. Fractions (10collected and monitored for absorption al(A2,o; 0) and for cold hemagglutinating acttiter; 0).The warm eluted material was therefomatographed on a Sepharose 6B column100 cm) at 300C with PBS as the eluifractions were monitored for both absor280 nm and cold agglutinin activity. Aantibody fraction was obtained, whshown to be IgM by immunoelectrophcsay when developed with anti-whole rirum.

Purification of LPS. The pumscheme of the LPS-like component fromocytogenes is shown in Fig. 3. A crufraction was obtained from lyophilizedusing hot phenol-water extraction as dabove. Purification was achieved throLries of molecular sieve chromatographdures (Fig. 4A through C) and monicold hemagglutination inhibition assayslide gel (Fig. 5) and polyacrylamide gephoresis. Initial purification of the crudwas obtained on a Sephadex G-25 colu4A). The void volume peak containedcold agglutinin inhibitory activity. Matethis peak was found to consist of tvpopulations of macromolecular speciesone apparently was high-molecular-vbonucleic acid. This material was therejected to ribonuclease treatment and timatographed on a Sephadex G-25 fin(Fig. 4B). All of the inhibitory activity v

in material from the void volume peak. Electro-phoresis of this material (Fig. 5) again showedthe presence of two molecular species. Finalpurification was achieved by chromatographyon a Sepharose 4B column (Fig. 4C). Althoughthe inhibitory material was again found in thevoid volume peak, electrophoresis (Fig. 5) indi-cated a single component which was designatedpurified LPS. The final product was able toinhibit 4 to 8 hemagglutinating doses of thepurified cold agglutinin at a concentration of 7.5,ug/ml. As opposed to the void volume peaksfrom the Sephadex G-25 and G-25 fine columns,which showed UV absorption maxima at 260nm, material in the Sepharose 4B void volumepeak showed increasing UV absorbance as thewavelength was lowered, but it did not exhibitany UV maxima. The second peak from the

thermal Sepharose 4B column apparently contained theme rabbit nucleic acid, judging by its UV absorption spec-> filtered trum (maximum at 260 nm).through Slide and gel electrophoresis. Figure 5cm) that shows the results of slide electrophoresis of lis-e column terial LPS at various stages of purification. Theand then slides were subjected to electrophoresis, dried,ml) were and stained with toluidine blue. The purifiedt (CA material migrated as a single band under the

conditions employed. The purified material wasdissolved and dialyzed against 2% SDS in Tris

re chro- buffer as described above and subjected to elec-n (2.5 by trophoresis on SDS gels. No bands were notedant, and when the gels were stained with Coomassie bril-bance at liant blue, indicating that no contaminating pro-purified tein was present. Gels stained with periodic acid-ich wasbresis as- Log phase Listeria monocytogenes type 4Babbit se-

2 X wash saline

rification 2 X wash HOLL. mon-ide LPS Lyophilized L. monocytogenescells by

Describedigh a se-ic proce-tored byrs and by,1 electro-le extractimn (Fig.all of thearial fromvo major3 (Fig. 5);eight ri-afore sub-hen chro-e columnvas found

Phenol water at 650C

Aqueous phase concentrated,precipitated in 1%16 Na-acetatein methanol

Crude LPS

G25 column

G25 void (Inhibitory fraction)

I Enzite RNAase

4 G25 fine column

G25 fine void (Inhibitory fraction)

; Sepharose 4B column

Void - "pure LPS" (Inhibitory fraction)

FIG. 3. Purification scheme for listerial LPS.

VOL. 23, 1979

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

850 WEXLER AND OPPENHEIM

4 B

0--25Fine

C-25

3

C)Dr.C

3:

I crude

I

3

oxaS-S5

FRACTION

m

C)r-

--4

I

FRACTION

FIG. 4. Chromatographic purification of listerialLPS. (A) Crude LPS (250 mg) was dissolved in 10 mlof 0.15 M ammonium carbonate and chromato-graphed on a Sephadex G-25 column (5 by 50 cm). (B)Ribonuclease-treated crude LPS (200 mg) from a

Sephadex G-25 column was dissolved in 5 to 10 ml of0.15 M ammonium carbonate and chromatographedon a Sephadex G-25 fine column (2.5 by 100 cm). (C)The Sephadex G-25 fine void volume fraction mate-

FIG. 5. Slide electrophoresis of 4-gil samples offractions from the various stages of purification ofLPS. The slide was stained directly after drying withtoluidine blue as described in the text.

Schiff stain, however, showed differing distri-bution patterns and number of bands (one orthree) depending on temperature and length ofstorage, conditions of solubilization, and modeof dialysis before electrophoresis.Analytical ultracentrifugation. Figure 6

shows the Schlieren patterns from sedimenta-tion velocity ultracentrifugation analysis of thepurifed LPS treated as described above. LPSsuspended in PBS (Fig. 6A) showed a fast-mov-ing, diffuse peak of 37S and a slower, sharperpeak of 17S. As the material sedimented, aboundary at the bottom of the cell was noted,and, when it was removed from the ultracentri-fuge, it was seen that a hard gel had accumulatedat the bottom of the ultracentrifuge cell. LPSdialyzed against Tris-ethylenediaminetetraace-tate (Fig. 6B, bottom pattern) exhibits twopeaks-a small diffuse shoulder which quicklyflattened out and a sharp peak of approximately3.9S. LPS dialyzed against Tris-SDS (Fig. 6B,top pattern) showed a single sharp peak of 3.1S.Electron microscopy. An electron micro-

scopic examination of the purified listerial LPS,in which negative staining techniques were used,revealed a heterogeneous mixture of forms, pri-marily ribbons and spheres of various sizes andshapes (Fig. 7) Such structures are very similarto those reported by a number of other investi-

rial (> 100 mg) was dissolved in 5 ml of 0.15 Mammonium carbonate and chromatographed on aSepharose 4B column (2.5 by 100 cm). All columnswere elated with 0.15 M ammonium carbonate, and10-ml fractions were collected. All fractions weremonitored for absorbance at 260 nm (A260; 0) andcold agglutinin inhibitory activity (0).

FRACTION

0

010

INFECT. IMMUN.

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

LPS-LIKE COMPONENT FROM L. MONOCYTOGENES 851

A

FIG. 6. Sedimentation veloLPS under various conditionsat 40,000 rpm. (B) Top pattermin at 52,000 rpm; bottom paminetetraacetate buffer, 104rection of sedimentation was

gators for LPS from gram-negative bacteria (14,19,33, 41).

B, , _, TW absorption spectra. UV spectra of thepurified LPS revealed no UV maxima, but didreveal a steadily increasing optical density as thewavelength was lowered. The absence of an ab-sorption maximum at 258 nm indicated that nocontaminating nucleic acid was present.Aqueous solutions of LPS were translucent, andthe observed absorbance may have been due tolight scattering by the macromolecular aggre-gate.Infrared absorption spectra. The infrared

;'i;- spectra of purified LPS from both an Esche-richia coli K-12 rough strain and a L. monocy-togenes strain are shown in Fig. 8A and B,respectively. Although there are limitations ininterpreting infrared spectra of complex naturalproducts such as LPSs, the striking similarities

panbetween the two spectra are apparent.city patterns ofpurified Chemical composition of listerial com-

n, Tris-SDS buffer, 104 ponent. The purified material was analyzed,ttern, Tris-ethylenedia- and its chemical constituents (Table 1) weremin at 52,000 rpm. Di- carbohydrate, lipid, and phosphorus, with afrom right to left. small amount of amino acid. Sugars were quan-

w.p.

AV,

.;.-re.In

WIto;*N 't;'aI'6 a

-(i''*

rr

'C~~~~~~~~~~~~~~A

t¢ e' -'

pa - :. * . r

FIG. 7. Electron micrograph ofpurified listerial LPS negatively stained with ammonium molybdate. Bar= 0.5 !um.

* i. t

4~~~~4~~a

VOL. 23, 1979

kkw-

: _s

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

852 WEXLER AND OPPENHEIM-t r TI-

-t

t-p

3

-t'*1 1 -.I-

- -- I I +4-A

.f-100

' -' 1 - ---I

0o

V

.Jd-l l t- -1I -

WAVELENGTH (MCRONS)6 7 8I .I . . . .

L-1

tv#I,IgI8I,8;|I'Ii-;|I20FREUENCY (CM) 20

WAVELENGTH (MKRONS)

00 1S DO

9 10 12 1, 20

1A.f

-w1

1C+ 4,/ -, 'X 1 1 <

0o

tr

1 TI -

N

5 3

IF~TT7iCB

mt T- I

6 8 9 10 12 15 20

,-! - 1- i- - t - -t 7-

I I---- t1 1t I I- -~

3500 3000 2500FREQUE4CY (CM') 200 150

FIG. 8. Infrared spectra ofpurified LPS preparations. (A) LPS from a rough strain of E. coli K12. (B) LPSfrom L. monocytogenes.

TABLE 1. Chemical composition of listerial LPS'SOf TOfCompound ~ '0f %Ototal fraction

CarbohydrateHexose

Glucose 20Galactose 20

Methyl-pentose (rhamnose) 12Uronic acid (glucuronic acid) 12Amino sugars

N-acetylgalactosamine 9N-acetylglucosamine 2.5

KDO 2.5Heptose 2

Fatty acids 15C12 saturated 5C1., saturated 17Unidentified, elutes after C14 15,B-Hydroxymyristic acid 45Ci,; saturated 17

Phosphorus 0.87Amino acids 1.6

titated by both colorimetric techniques and gas-liquid chromatography. The carbohydrates con-

sisted of neutral sugars (glucose, galactose, andrhamnose), glucuronic acid, KDO, heptose, N-

acetylgalactosamine, and N-acetylglucosamine.The elution patterns in gas-liquid chromatogra-phy of the neutral sugars, uronic acid, and KDOas trimethyl silyl ethers of their methyl glyco-sides are shown in Fig. 9.The fatty acids are apparently covalently at-

tached to the molecule since they were releasedonly upon hydrolysis and could not be extractedby a total lipid extraction procedure. To ensurethat there was no contaminating phospholipidpresent, a chloroform-methanol-water extrac-tion was performed on the purified material, andthe chloroform phase was subjected to thin-layerchromatography. Development of the chromat-ogram revealed no phospholipid or free fattyacids. The fatty acids were identified by gas-

liquid chromatography of their methyl esters onboth polar and nonpolar columns. Tracings oftheir elution patterns are shown in Fig. lOA andB. Methyl laurate (C12), methyl myristate (C14),methyl palmitate (C16), and methyl-,8-hydroxy-myristate were identified by comparison of theirRf values relative to methyl stearate (C18) withthose of standard fatty acid methyl esters.When listerial LPS was hydrolyzed and the

fatty acids were extracted with chloroform andthen methylated, the fatty acids noted in Fig.

2.5

z

100

u

-60

L40

02C

. - - .111.1- - - - - -!

fl -Pt. P -H- I I I-+ L ..Jd t-d I- I I [- I . .,-,III II II --.-.- II III \-L A., II IIII,

INFECT. IMMUN.

I

L-IF7--

01t

iV~jH.

7

tt.vtuu~t t tr

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

LPS-LIKE COMPONENT FROM L. MONOCYTOGENES 853

10A and B were seen. However, when the ma-terial was subjected directly to methanolysis(which, unlike hydrolysis, does not break amidelinkages), the peak emerging after methyl my-ristate was absent, indicating that it may beconnected by an amide linkage, as is the casewith some fatty acids in classical gram-negativeLPS (12).Limulus lysate and carbocyanine dye as-

says. The purified listerial material was foundto be active in two assays normally used for

GLC LPS CarbohydratesE

FIG. 9. Gas-liquid chromatographic elution pat-tern ofpurified LPS sugars as trimethyl silyl ethersof their methyl glycosides prepared as described inthe text.

assaying gram-negative endotoxin, the limuluslysate and carbocyanine dye assays. The limuluslysate assay involves the gelation of an extractfrom Limulus polyphemus amoebocytes bynanogram amounts of endotoxin (36). Therehave been conflicting reports in the literature asto the specificity of this assay; but, althoughlimulus lysate can apparently be gelled by sub-stances other than LPS, the exquisite sensitivityseems to be unique to endotoxins. The listerialLPS-like component was able to induce gel for-mation in the same concentration range as theE. coli endotoxin standard. A 1-ng amount oflisterial LPS was as active as 0.4 ng of the E.coli reference endotoxin in inducing gel forma-tion. A lipoteichoic acid (LTA) preparation fromLactobacillus fermentum was also able to in-duce gel formation, but only at 1,000 to 2,000times the concentration at which listerial LPSand the reference endotoxin were active.Another LPS assay involves the use of a cati-

onic carbocyanine dye. This assay is based on aspectral shift in an acidic solution of a cationiccarbocyanine dye. Polyanions, such as proteins,nucleic acids, and acidic polysaccharides causethe dye, which normally absorbs at 510 nm, toundergo shifts to longer wavelengths. Spectralshifts to shorter wavelengths result from reac-tions of the dye with LPSs from gram-negativebacteria, as well as with their isolated lipid Acomponent (15). The concentration range de-tected is from 0.5 to 10 jig of LPS. As Fig. 11

GLC Nonpolar Column Fatty Acids

LPS+C,8

C18

B

GLC Polar Column Fatty Acids

LPS +C18

U

FIG. 10. Gas-liquid chromatographic elution patterns of purified LPS fatty acids as their methyl esters,with methyl stearate added as an internal standard. (A) Chromatography on a nonpolar (3%o SE-30) column.(B) Chromatography on a polar (10% Silar 10 C column). /8-OH Myr., /3-Hydroxymyristic acid.

VOL. 23, 1979

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

854 WEXLER AND OPPENHEIM

50-

60500 600 7

WAVELENGTH nm

FIG. 11. Spectra of increasing quantities10, 15pg) ofpurified listerial LPS in the carbondye assay performed by the method of JaWork (15).

shows, at the lower end of the conceitrange, the listerial component shifted tto 630 nm, a shift which could possibleplained by the presence of uronic aciclisterial material. At the higher end of 1centration range, 10 to 15 jig of a purifiedLPS caused a spectral shift to 470 nm.Biological activity. When 2-kg rabb

injected intravenously with the purifieddosages ranging from 10 to 400 jig, a vaphysiological responses were observed,from a simple pyrogenic response acconby malaise to severe diarrhea and deatiseveral hours. All rabbits (four) injected,ug doses exhibited a classical endotoxinpyrogenic response (23) (Fig. 12) butmore severe. These rabbits showed amalaise and a 2.50C rise in rectal temp4which peaked within 3 h and gradually r4to normal within 6 to 8 h. In one case a ltemperature pattern was observed,peaked at both 1 and 3 h (Fig. 12). Arabbit similarly injected with pyrogen-fiile saline showed no response. When givedoses, rabbits became much sicker, anchibited severe diarrhea from which the,

-I recovered or died within 24 h. The rate of deathin these animals increased as dosage was in-creased: when given 50-jig doses, one of foursuccumbed within 2 h; at a 200-jig dosage two offour died, one after 1 h and the other after 24 h;and with 400 jig three of four expired within afew hours. In those rabbits which died within 2h autopsy revealed only gastrointestinal andstomach distension, most likely a result of thediarrhea. In the rabbit which survived 24 hbefore expiring, autopsy also revealed fluid inthe lungs.The purified LPS could also cause localized

edema and erythema in rabbits when injectedsubcutaneously in doses as low as 2.5 jug. Sub-sequent intravenous injection of the purifiedmaterial caused enhanced inflammation at theoriginal subcutaneous injection site but did notproduce the hemorrhagic or necrotic appearancethat would be expected for a localized Shwartz-man reaction.Amphipathic nature. Similar to gram-neg-

ative LPS and gram-positive LTA macromole-cules, listerial LPS is amphipathic. Incubationofrelatively nonreactive sheep erythrocytes with

0oo the purified listerial LPS at 4, 25, or 370C re-sulted in sensitized erythrocytes which subse-

(0, 1 5 quently could be readily agglutinated by im-socyanine mune anti-listerial rabbit sera or purified IgMnda and fractions. The degree of agglutination was de-

pendent on the time of the incubation of theLPS with the sheep erythrocytes, requiring a

ntration minimum of 1 h for adsorption, but did not seemhe peak to be temperature dependent, since similar re-

y be ex- sults were obtained whether the incubation wasi in the performed at 4, 25 or 370C.the con- DISCUSSIONlisterial

The listerial component responsible for thetits wereLPS atriety ofrangingnpaniedi withinwith 10-Ltype ofnothinggeneralerature,eturnedbiphasicwhich

controlree ster-,n largerI all ex-y either

42

o 41

X 40Lu

-J< 39

m 38

0 60 120 180 240 300 360

TIMEFIG. 12. Pyrogenic response of rabbits to intrave-

nous injection with 10 pg ofpurified listerial LPS (-and *) and with saline (0). Time is expressed inminutes.

INFECT. IMMUN.

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

LPS-LIKE, COMPONENT FROM L. MONOCYTOGENES 855

induction of the cold agglutinin syndrome inrabbit appears to bear a striking similarity, bothchemically and biologically, to classical LPS orendotoxin from gram-negative organisms.Chemical analysis of the purified material re-vealed the presence of 15% total fatty acid (ofwhich 50% was 83-hydroxymyristic acid), 40 to45% neutral sugar (glucose, galactose, and rham-nose), 11.5% amino sugar, 12% glucuronic acid,2.5% KDO, 2% heptose, 0.87% phosphorus, and1.6% amino acids (Table 1). KDO, heptose, andB8-hydroxymyristic acid are often consideredmarkers for LPS. Whereas KDO and heptosehave been identified in bacterial sources otherthan LPS, ,8-hydroxymyristic acid has only in afew cases been encountered in bacterial lipidsother than lipid A (12), and the presence of allthree LPS markers together in this listerial mac-romolecule is surprising. Physical characteristicsof the purified listerial component, such as struc-tural appearance in electron micrographs (14,19, 33, 41), sedimentation velocity in analyticalultracentrifugation analysis (26), and infraredspectra, are similar to those reported for gram-negative LPSs. Toxic effects in rabbits, as wellas activity in two assays specific for LPS (thelimulus lysate and carbocyanine dye assays) pro-vide a further link of this material to other gram-negative endotoxins.To test the specificity of the two LPS assays

employed in this study, a comparison was madebetween the reactivities of listerial LPS and anLTA preparation from L. fermentum. A com-parison of the properties of gram-negative LPS,listerial LPS, and LTA may be seen in Table 2.Both LPS and listerial LPS are pyrogenic andlethal to rabbits; LTA is not. All three sub-stances are amphipathic and can sensitize eryth-rocytes to agglutination with specific antibody.LPS and LTA both give positive localizedShwartzman reactions. With listerial LPS, intra-venous injection of the material produced en-hanced inflammation of the originally subcuta-neously injected site, but the area did not be-come hemorrhagic or necrotic.LPS and listerial LPS are both active in the

limulus lysate assay. The LTA preparation wasable to activate the assay, but only at 1,000 to2,000 times the concentration range required byLPS and the listerial component. In the carbo-cyanine dye assay, LTA caused a spectral shiftof the dye to a longer wavelength, characteristicof the shift caused by polyanions, but did notcause a spectral shift to 470 nm, even at 10 timesthe concentration at which listerial LPS wasactive. It is interesting to note the types of shiftscaused by the various concentrations of LPS(Fig. 11). At the lower concentrations it appearsthat the acidic moieties of the macromolecule

TABLE 2. Comparison of characteristics ofbacterial cell surface components

Liste-Characteristic LPSa rial LTAb

LPS

Pyrogenicity in rabbits + + _CLethality in rabbits + + _CAmphipathic (sensitizes erythro- + + +C

cytes)Localized Shwartzman reaction + ± +CLimulus lysate gelation + + _dCarbocyanine dye assay (spectral + + _d

shift)Presence of acylated hydroxy + + -c

fatty acidPresence of KDO, heptose + + ND

a From gram-negative organism.b From gram-positive organism.c From Wicken (42).d LTA from L. fernentum.eND, Not determined.

comprise the available sites for reaction with thecarbocyanine dye. At the higher end of the con-centration range, it appears that the lipid A-likecomponent (assuming that lipid A, as suggestedby Janda and Work [15], is the responsibleagent) is available for reaction with the dye.These results suggest that the aggregation of themacromolecule may be concentration depend-ent, a characteristic which has been noted byothers working with various LPS preparations(26). The formation of aggregates also appearsto be time and temperature dependent, since thesame preparation exhibited various SDS gel pat-terns, depending on temperature and time ofstorage, conditions of solubilization, and modeof dialysis before electrophoresis.

Since our purified LPS-like material ac-counted for approximately 6% of the dry weightof the cell and from the immunological dataappeared to be located on the outer surface, anelectron microscope study was initiated withKwang Shin Kim, Department of Microbiology,New York University School of Medicine, todetermine whether there is (are) any unique cellsurface structures) associated with the Listeriastrain used and also whether any structuralchanges are induced in cells subjected to hotphenol-water extraction. Electron micrographsof cells taken from an actively growing culturerevealed typical gram-positive morphology, withno indication of a gram-negative outer envelope.Micrographs of the same cells subjected to theextraction procedure showed drastically alteredinternal morphology, but cell wall structuresappeared intact with approximately the samethickness as in untreated cells.Although the isolated LPS-like component

VOL. 23, 1979

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

856 WEXLER AND OPPENHEIM

was not able on its own to induce a cold agglu-tinin response in rabbits, when rabbits were firstinjected with HKLM and subsequently injectedwith 50 ,jg of LPS after a specified time intervalhad elapsed, a rise in both cold agglutinin titerand anti-listerial antibody was observed. In thiscase, only IgM antibodies were induced; with abooster of HKLM, on the other hand, both IgMand IgG anti-listerial antibodies, as well as IgMcold agglutinins, were induced (Fig. 1). Absorp-tion studies carried out on the test bleed andimmune sera indicated that two different IgMpopulations are induced by the HKLM: the coldagglutinating antibody, which can be absorbedby both erythrocytes and HKLM, and an anti-listerial antibody, which can be absorbed onlyby HKLM. Nonimmune rabbit sera have lowlevels of naturally occurring cold agglutinins,with titers ranging from 16 to 64; HKLM wasable to absorb out cold agglutinating activity atboth 4 and 370C in all the immune sera tested,but only down to the naturally occurring levels,suggesting that HKLM may be able to absorbout only the induced antibody. Rabbit erythro-cytes (autologous and isologous) were able toabsorb out all of the cold agglutinating activity,both naturally occurring and induced, at 40Cbut not at 370C.

Other workers, while studying listerial cellwalls as well as crude preparations from phenolextracts of L. monocytogenes serotype 4B (4, 21,34), found that these preparations did have someendotoxic activities, such as dermal edema anderythema in rabbits and death in chicken em-bryos. Indications of KDO and heptose werealso noted by these workers in listerial cell wallpreparations and phenol extracts. The presentstudy indicates that a component isolated inapparently "pure" form (i.e., with no detectablecontaminating nucleic acid, protein, or phospho-lipid) which migrates as a single band on slideelectrophoresis and as a single, small homoge-neous subunit in an analytical ultracentrifuge isextremely similar to gram-negative endotoxinsin terms of its chemical composition, behaviorin LPS assays, and biological activity.Many workers have attempted to isolate the

toxic component of L. monocytogenes (4, 20-22,25, 29, 31, 35, 37), and various cell fractions(lipoidal, carbohydrate, and proteinaceous) havebeen reported to be biologically active. Our pu-rified LPS-like substance, apparently located onthe surface of this gram-positive organism, isextremely toxic to rabbits, may account for someof the toxic effects due to infection with L.monocytogenes, and has potent biological activ-ity. The biological properties of listerial LPSinclude: (i) pyrogenicity; (ii) activity in limulus

INFECT. IMMUN.

amoebocyte lysate assay; (iii) death within fewhours after injection, if given in sufficiently largedoses (death is accompanied by severe diarrhea);(iv) edema and erythema in rabbit skin uponsubcutaneous injection; (v) rise in anti-listerialantibody titer; and (vi) rise in cold agglutininantibody titer.

ACKNOWLEDGMENTSThis research was supported by Public Health Service

grant HL 21328, a Young Investigator Research Award fromthe National Heart, Lung and Blood Institute and by fundsfrom the Public Health Service Biomedical Research Supportgrant RR 05399 awarded to New York University MedicalCenter.We are grateful to Richard Margolis (Department of Med-

icine) for the amino sugar analysis, Edward C. Franklin (De-partment of Medicine) for amino acid analysis, Peter Elsbach(Department of Medicine) for the thin-layer chromatography,Kwang Shin Kim (Department of Microbiology) for help withthe electron microscopy, Gisela Witz (Department of Environ-mental Medicine) for the infrared spectra, and Marylyn Hor-witz (New York Blood Center) for analytical ultracentrifuga-tion. We also thank Julie Siegel (University of Florida MedicalCenter) for the kind gift of E. coli K-12 LPS and AnthonyWickens (University of New South Wales, Australia) forkindly supplying L. fermentum LTA. We are indebted toMartin S. Nachbar (Departments of Medicine and Microbi-ology) and Milton R. J. Salton (Department of Microbiology)for their advice and assistance throughout the course of thework. We express our appreciation. tq Stan Willard for pho-tography and Josephine Markiewicz for typing.

LITERATURE CITED

1. Anacker, R. L., W. D. Bickel, W. T. Haskins, K. C.Milner, E. Ribi, and J. A. Rudbach. 1966. Frequencyof occurrence of native hapten among enterobacterialspecies. J. Bacteriol. 91:1427-1433.

2. Bligh, E. G., and W. J. Dyer. 1959. A rapid method oftotal lipid extraction and purification. Can. J. Biochem.Physiol. 37:911-917.

3. Blumenkrantz, N., and G. Asboe-Hansen. 1966. Anassay for total hexosamine and a differential assay forglucosamine and galactosamine. Clin. Biochem. 9:269-274.

4. Conklin, A. R., Jr., and I. H. Siddique. 1976. Certainchemical and biological properties of phenol extractsfrom Listeria monocytogenes. Am. J. Vet. Res. 37:1331-1335.

5. Costea, N., V. J. Yakulis, and P. Heller. 1965. Experi-mental production of cold agglutinins in rabbits. Blood26:323-340.

6. Costea, N., V. J. Yakulis, and P. Heller. 1971. Themechanism of induction of cold agglutinins by Myco-plasma pneumoniae. J. Immunol. 106:598-604.

7. Costea, N., V. J. Yakulis, and P. Heller. 1972. Inhibi-tion of cold agglutinins (anti-I) by M. pneumoniae an-tigens. Proc. Soc. Exp. Biol. Med. 139:476-479.

8. Dische, Z. 1955. New color reactions for determination ofsugars in polysaccharides, p. 313-358. In D. Glick (ed.),Methods of biochemical analysis, vol. 2. IntersciencePublishers, Inc., New York.

9. Dische, Z., and L. B. Shettles. 1948. A specific colorreaction of methyl pentoses and a spectrophotometricmicromethod for their determination. J. Biol. Chem.175:595-603.

10. Duncombe, W. G. 1963. The colorimetric micro-deter-mination of long-chain fatty acids. Biochem. J. 88:7-10.

11. Fairbanks, G., T. L. Steck, and D. F. H. Wallach. 1971.Electrophoretic analysis of the major polypeptides of

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

LPS-LIKE COMPONENT FROM L. MONOCYTOGENES 857

the human erythrocyte membrane. Biochemistry 10:2606-2617.

12. Galanos, C., 0. Luderitz, E. T. Rietschel, and 0.Westphal. 1977. Newer aspects of the chemistry andbiology of bacterial lipopolysaccharides, with specialreference to their lipid A component, p. 239-335. In T.W. Good Win (ed.), International review of biochemis-try, biochemistry of lipids II, vol. 14. University ParkPress, Baltimore.

13. Gray, M. L., and A. H. Killinger. 1966. Listeria mono-cytogenes and listeric infections. Bacteriol. Rev. 30:309-382.

14. Jackson, S. W., and P. N. Zey. 1973. Ultrastructure oflipopolysaccharide isolated from Treponema pallidum.J. Bacteriol. 114:838-844.

15. Janda, J., and E. Work. 1971. A colorimetric estimationof lipopolysaccharides. FEBS Lett. 16:343-345.

16. Kates, M. 1972. Techniques of lipidology, p. 362. In T. S.Work and E. Work (ed.), Laboratory techniques inbiochemistry and molecular biology. North-HollandPublishing Co., Amsterdam.

17. Korn, R. J., V. J. Yakulis, C. E. Lemke, and B.Chomet. 1957. Cold agglutinins in Listeria monocyto-genes infections. Arch. Intern. Med. 99:573-580.

18. Lind, K. 1973. Production of cold agglutinins in rabbitsinduced by Mycoplasma pneumoniae, Listeria mono-cytogenes, or Streptococcus M. G. Acta Pathol. Micro-biol. Scand. 81:487-496.

19. Lopes, J., and W. E. Inniss. 1970. Electron microscopicstudy of lipopolysaccharide from an avian strain ofEscherichia coli 018. J. Bacteriol. 103:238-243.

20. McIlwain, P., D. F. Eveleth, and J. A. Doubly. 1964.Pharmacologic studies of a toxic cellular component ofListeria monocytogenes. Am. J. Vet. Res. 25:774-781.

21. Mara, M., J. Julak, and F. Patocka. 1974. Isolation andchemical properties of biologically active fractions ofthe surface of Listeria monocytogenes. J. Hyg. Epide-miol. Microbiol. Immunol. 18:359-364.

22. Mara, M., F. Patocka, J. Julak, and B. Potuznikova.1974. Contribution to knowledge of factors participatingin virulence of Listeria monocytogenes. J. Hyg. Epide-miol. Microbiol. Immunol. 18:29-41.

23. Milner, K. C., and R. A. Finkelstein. 1966. Bioassay ofendotoxin: correlation between pyrogenicity for rabbitsand lethality for chick embryos. J. Infect. Dis. 116:529-536.

24. Mollison, P. L. 1972. Blood transfusion in clinical medi-cine. p. 174. Blackwell Scientific Publications, Oxford,England.

25. Njoku-Obi, A. N., and J. C. Njoku-Obi. 1964. Studieson Listeria monocytogenes: the nature of avirulence inthe mouse. Bacteriol. Proc., p. 82.

26. Olins, A. L., and R. C. Warner. 1967. Physicochemicalstudies on a lipopolysaccharide from the cell wall of

Azotobacter vinelandii. J. Biol. Chem. 242:4994-5001.27. Owen, P., and M. R. J. Salton. 1975. Isolation and

characterization of a mannan from mesosomal mem-brane vesicles of Micrococcus lysodeikticus. Biochim.Biophys. Acta 406:214-234.

28. Owen, P., and M. R. J. Salton. 1975. Use of the lectinconcanavalin A in the preparation of mesosomal mem-brane fraction from Micrococcus lysodeikticus. Micro-bios 13:27-39.

29. Patocka, F., M. Mara, B. Potuznikova, A. Jirasek,and E. Mencikova. 1974. Biological properties of sur-face components of Listeria monocytogenes (Ei factor).J. Hyg. Epidemiol. Microbiol. Immunol. 18:365-368.

30. Roberts, R. S. 1966. Preparation of endotoxin. Nature(London) 209:80.

31. Robinson, B. B., and A. N. Njoku-Obi. 1964. Prepara-tion and characterization of a toxic polysaccharide fromListeria monocytogenes. Bacteriol. Proc., p. 82.

32. Scott, T. A., Jr., and E. H. Melvin. 1953. Determinationof dextran with enthrone. Anal. Chem. 25:1656-1661.

33. Shands, J. W., Jr., J. A. Graham, and K. Nath. 1967.The morphologic structure of isolated bacterial lipo-polysaccharide. J. Mol. Biol. 25:15-21.

34. Siddique, I. H., and K. K. Srivastava. 1973. Cell walland cytoplasmic components of Listeria monocyto-genes. Am. J. Vet. Res. 34:567-570.

35. Siddique, I. H., L. C. Ying, and R. A. Chung. 1970.Studies on diphosphopyridine nucleotidase and plateletdamaging factor in an extracellular product of Listeriamonocytogenes. Can. J. Microbiol. 16:909-916.

36. Siegal, S. E., and R. Nachum. 1977. Use of the limuluslysate assay (LAL) for the detection and quantitationof endotoxin, p. 61-86. In A. W. Bernheimer (ed.),Perspectives in toxicology. John Wiley & Sons, NewYork.

37. Silverman, S. J., L. Elwell, and D. A. Kautter. 1960.A mortality enhancing extract isolated from Listeriamonocytogenes. J. Immunol. 86:669-674.

38. Steck, T. L., and J. Yu. 1973. Selective solubilization ofproteins from red blood cell membranes by proteinperturbants. J. Supramol. Struct. 1:220-232.

39. Sweeley, C. C., and B. Walker. 1964. Determination ofcarbohydrates in glycolipids and gangliosides by gaschromatography. Anal. Chem. 36:1461-1466.

40. Tsai, C.-M., D. A. Zopf, R. Wistar, Jr., and V. Gins-burg. 1976. A human cold-agglutinin which binds lacto-N-neo-tetraose. J. Immunol. 117:717-721.

41. Weckesser, J., G. Drews, and R. Ladwig. 1972. Local-ization and biological and physicochemical propertiesof the cell wall lipopolysaccharide of Rhodopseudomo-nas capsulate. J. Bacteriol. 110:346-353.

42. Wicken, A. J., and K. W. Knox. 1975. Lipoteichoicacids: a new class of bacterial antigen. Science 187:1161-1167.

VOL. 23, 1979

on January 7, 2020 by guesthttp://iai.asm

.org/D

ownloaded from