JOURNAL OF BACTERIOLOGY, JUlY, 1965Copyright © 1965 American Society for Microbiology

Vol. 90, No. IPrinted in U.S.A.

Characterization of the Phospholipases of Bacilluscereus and Their Effects on Erythrocytes, Bone,

and Kidney CellsMILTON W. SLEIN AND GERALD F. LOGAN, JR.

U.S. Army Biological Laboratories, Fort Detrick, Frederick, Maryland

Received for publication 9 February 1965

ABSTRACTSLEIN, MILTON W. (Fort Detrick, Frederick, Md.), AND GERALD F. LOGAN, JR. Char-

acterization of the phospholipases of Bacillus cereus and their effects on erythro-cytes, bone, and kidney cells. J. Bacteriol. 90:69-81. 1965.-Culture filtrates of Bacil-lus cereus contain phospholipases that split phosphoryl choline, phosphorylethanolamine, and phosphoryl inositol from the phospholipids phosphatidyl choline(PTC), sphingomyelin, phosphatidyl ethanolamine (PTE), and phosphatidyl inositol(PTI). It is possible that one enzyme catalyzes the degradation of PTE and PTC, butthe other phospholipases appear to be separate entities. Some activity on phospha-tidyl serine has also been noted. Quantitative paper chromatography has been usedfor characterizing the phospholipases that are separated on N,N'-diethylamino-ethyl cellulose columns. A procedure for the analysis of inositol is included. A sensi-tive kidney cortex homogenate test is described that depends on the release of alka-line phosphatase for the measurement of phosphatasemia factor (PF) activityassociated with the phospholipases. The effects of the phospholipases on erythrocytes,kidney, and bone cells are discussed. Hemolysin activity is inhibited by crude soy-bean "lecithin," but hemolysis does not seem to be identical with PTE- or PTC-phos-pholipase activity. PF activity is also inhibited by the "lecithin." Highest PF ac-tivity is associated with PTI-phospholipase. The phospholipase fractions differ intheir sensitivities to trypsin. Phospholipases with similar properties have been ob-tained from culture filtrates of B. anthracis.

The partial purification of two phospholipasesfrom culture filtrates of Bacillus cereus has beenreported (Slein and Logan, 1963). One of themresulted in a marked phosphatasemia after intra-venous injection into rabbits; bone appeared tobe a major source of the excess serum alkalinephosphatase (Slein and Logan, 1962, 1963). Thesecond phospholipase inhibited the release ofphosphatase from epiphyseal bone slices by thefirst (Slein and Logan, 1963). Results in thepresent paper indicate that other phospholipasesmay be isolated from culture filtrates of B. cereusand that each of them probably has specificphospholipid substrate requirements. The rela-tionship of the phospholipases to hemolysis andto the release of alkaline phosphatase from boneand kidney cortex cells has been investigated.

MATERIALS AND METHODSStrain 6464 of B. cereus was originally obtained

from the American Type Culture Collection. Gen-eral procedures for the preparation of culturefiltrates and for the isolation of phospholipaseshave been described (Slein and Logan, 1963). The

69

method for measuring phosphatasemia factor(PF) activity with bone slices in vitro was in-cluded in the same report.Homogenate tests for PF activity. The bone slice

test is about 100 times as sensitive as the phos-pholipase method that depends on the liberationof acid-soluble phosphorus from lecithin. Thebone slice test for PF activity is also more specific,since it is based on the degradation of phospho-lipids in the cell membrane for the release of alka-line phosphatase. A phospholipase that splitsacid-soluble phosphorus from lecithin may notnecessarily have PF activity (Table 2 in Slein andLogan, 1963). However, the preparation of boneslices is time-consuming, and the number of sam-ples that can be tested is rather limited. It wasfound that homogenates of rabbit bone epiphysesor kidney cortex could be used for the measure-ment of PF activity. The kidney cortex waschosen for general use, but some comparisons weremade with bone homogenates that gave resultsanalogous to those obtained with bone slices.Rabbits weighing about 1 kg were anesthetizedwith chloroform and were exsanguinated. Thekidneys were chilled in ice, split lengthwise, de-medullated, and the cortex was stored in ice-cold

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SLEIN AND LOGAN

Ringer phosphate solution (pH 7.5). The cortexwas blotted, weighed, and an amount of up to 2.5g was homogenized with 25 ml of cold Ringerphosphate solution in a VirTis 45 homogenizer for0.5 min with the rheostat set on 50 ( a speed ofabout 16,000 rev/min when measured with theblades out of solution). Higher speeds or longertimes resulted in too much cell damage and lossof phosphatase. The homogenate was centrifugedfor 5 min at about 200 X g at 5 C, and the particu-late fraction was washed twice with cold Ringerphosphate solution. The precipitate was sus-pended with 5 ml of cold Ringer phosphate solu-tion for each gram of cortex that had been ho-mogenized. Samples (0.5 ml) were pipetted into1 X 10-cm tubes in an ice bath and 0.2 ml of Ringerphosphate or other solutions to be tested wasadded to each. The tubes were stoppered, placedat a 450 angle in a rack in a Dubnoff MetabolicShaker, and were shaken at about 160 cycles/minfor 30 min at 37 C. The samples were chilled inice, centrifuged at 5 C for 6 min at about 2,000 Xg, and 0.1-ml portions were diluted with cold dis-tilled water (usually 1:5 to 1:10). To prevent theinclusion of any floating particles, bits of cottonwool were wound onto the tapered, ground tips ofpipettes used for sampling the supernatant solu-tions. Duplicate 0.1-ml samples of the dilutionswere used for the assay of alkaline phosphatasewith p-nitrophenyl phosphate, as described previ-ously for blood serum (Slein and Logan, 1960).Phosphatase samples from homogenate controlsthat had been incubated without PF or with anexcess of standard crude PF (about 300 units)were incubated for exactly 4 min at 37 C. Oneunit of PF is the amount that releases sufficientalkaline phosphatase from a homogenate sampleto produce an absorbancy of 0.1 in the phospha-tase test in 12 min with a 1-cm light path at 395m,u. Other phosphatase samples from the homoge-nate treated with various phospholipase fractionswere incubated for 12 min along with samplesfrom controls that had been incubated withoutPF or with 3 to 6 units of standard PF. The ab-sorbancies were corrected for alkaline phosphatasereleased spontaneously from the control homogen-ates that had been incubated without PF. Thehomogenate samples treated with excess standardcrude PF served to determine the maximal amountof phosphatase that could be released from eachpreparation. All of the supernatant solutions werediluted so that the absorbancies obtained in thephosphatase test were not greater than about 1.3,which is within the range of proportionality forthe p-nitrophenyl phosphate released. Phospha-tase released and the amount of PF added wereproportional if the amount of phosphatase wasnot greater than one-third of maximum foundwith excess PF. The 4-min absorbancy values ob-tained with excess PF were multiplied by threeto compare the 12-min samples with them.

Although the absolute amount of phosphatasereleased varied from one homogenate to another,the values obtained with any given PF fraction

relative to the standard PF were always verynearly the same. The activity of the standardcrude PF solution was stable when stored frozenfor over a year even at a concentration of only0.1 mg of protein per ml of Ringer phosphatesolution (pH 7.5). To compare the relative PFactivities of the phospholipase fractions, thestandard crude PF was given an arbitrary value of30 units per,tg of protein, i.e., 0.1 uLg releasedphosphatase from a homogenate so that the cor-rected absorbancy in the phosphatase test was0.3. For each new homogenate preparation andsupernatant fluid dilution used for phosphatasedetermination, a factor was calculated to relateunknown PF samples to the standard crude PF.The sensitivity of the kidney cortex homogenate

test for measuring PF is of the same order ofmagnitude as the previously used bone slice pro-cedure. Because of the very high PF activity ofcertain phospholipase fractions, it was necessaryto dilute them to less than 1 ,ug of protein per ml.In such cases, serum albumin (0.1 mg/ml of Ringerphosphate) was included to stabilize the PF frominactivation. Although the kidney alkaline phos-phatase released by PF was very stable at 5 C,the activities were always measured on the daythat the homogenates had been treated with PF.The same washed homogenate preparation wasused for tests in the morning and afternoon ofthe day of preparation; the spontaneous leakageof phosphatase increased during storage in ice,but the correction was not excessive. Breakdownof cells during storage of intact cortex in coldRinger phosphate solution appeared to be greaterthan with the homogenate, so that it was betterto use the same homogenate in tests over a 6-hrperiod than to prepare fresh homogenates from agiven sample of kidney cortex.

Characterization of phospholipases by the use ofquantitative paper chromatography. We had hopedto measure the degradation of purified phospho-lipids individually by the various phospholipasefractions of B. cereus, but this was not feasible,perhaps partly because of an unfavorable surfacecharge distribution with aqueous emulsions ofsingle phospholipids (Bangham and Dawson,1962). A compromise test system was used inwhich a mixture of phospholipids was incubatedwith the phospholipases, and degradation wasmeasured quantitatively after elution of thestained phospholipid spots from chromatograms.Most of the results were obtained with an emul-sion of "purified" soybean lecithin (Mann Re-search Laboratories, Inc., New York, N.Y.) sup-plemented with sphingomyelin (SPH) (GeneralBiochemicals, Chagrin Falls, Ohio), and in somecases with lysolecithin (lyso PTC) (General Bio-chemicals). The soybean "lecithin" containedsignificant amounts of phosphatidyl ethanolamine(PTE) and phosphatidyl inositol (PTI) as well asphosphatidyl choline (PTC, lecithin). It also wascontaminated with what appeared to be phospha-tidic acid, smaller amounts of lyso PTC, andother substances. Approximately 60% of the phos-

70 J. BACTERIOL.

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PHOSPHOLIPASES OF B. CEREUS

phorus was accounted for as PTC, PTE, PTI, andlyso PTC. An emulsion containing 100 mg of soy-bean "lecithin" per ml of distilled water wasprepared by treatment for 10 min in a 10-kc Ray-theon sonic oscillator while cold fluid was circu-lated around the chamber to prevent excessiveheating. The emulsion was adjusted to about pH7 by the addition of KOH. For tests with phos-pholipases, the emulsion was buffered with 0.133M phosphate. When SPH was to be included, anemulsion was prepared by adding 40 mg of SPHto 2 ml of the "lecithin" emulsion and 4 ml of0.2 M phosphate buffer (pH 7.0). Emulsions werestored at 5 C with 1:10,000 Merthiolate as pre-servative.To 0.15 ml of the buffered emulsion in 7.5 X

75-mm tubes was added 0.15 ml of phospholipaseor distilled water. The tubes were stoppered, in-cubated at 37 C for various times, and chilled inice for 1 min. Duplicate 10-,uliter samples wereplaced on Whatman no. 3MM paper that had beenimpregnated with silicic acid, and the chromato-grams were developed for 3.5 hr at about 24 C withdiisobutyl ketone-glacial acetic acid-water (40:20:3, v/v), essentially as described by Marinetti(1962). The papers were dried for 10 min in aforced-air drying oven at about 50 C. The stripswere cut in two lengthwise, and the duplicatepairs were stained for 1 hr by immersion in acidfuchsin or brilliant green solutions in tubes 2.5 X20 cm, one pair per tube. The stained chromato-grams were blotted between facial tissue papersand rinsed, with gentle agitation at intervals, for10 min each in two changes of rinsing solutions.After a final blotting, the strips stained with acidfuchsin were dried for 10 min at about 50 C, andthose stained with brilliant green were only partlydried at about 35 C or less. Areas slightly largerthan the spots stained red with acid fuchsin (PTE,PTC, lyso PTC, and SPH) were cut out withscissors and further cut into pieces about 4-mmsquare that were allowed to stand overnight atabout 24 C with 2 ml of tertiary butanol-0.2 Macetate buffer, pH 5.0 (1:1, v/v) in stopperedtubes (1 X 10 cm). PTI spots were stained bluewith brilliant green and were eluted in the sameway with 3 ml of the solution. Blank areas cor-responding in size to the various samples werealso eluted to correct for the background stainingthat was slight with acid fuchsin but was moresignificant in the case of brilliant green. Afterelution, the paper bits were stirred with a glassrod, and the samples were centrifuged at about2,000 X g. The absorbancies of the supernatantsolutions were measured at 640 m, (maximum forthe brilliant green stain under these conditions)or at 560 m,u (acid fuchsin) in a Beckman DUspectrophotometer with the eluting fluid as refer-ence solution and a 1-cm light path. The valueswere corrected for the blanks, and the percentagedegradation of each phospholipid was determinedby comparing the absorbancies with those ob-tained with control samples that had been incu-bated without phospholipase.

Phosphate buffer (pH 7.0) was selected for useafter preliminary tests indicated that the degrada-tion of phospholipids was better under these con-ditions than with other buffers or pH values tried.Strong buffering is needed to prevent a decreasein pH. The milkiness that usually develops in thereaction mixture during phosphilipase actiontends to become flocculent if the pH decreases,and uniform sampling with a micropipette be-comes impossible. Adequate activity was obtainedwith the commercial "lecithin" without addedmetal cofactors; therefore, these were omittedfrom the reaction mixture, since the tendency forflocculation seemed greater with added metalions.The procedure for the quantitative analysis of

phospholipids is based on an elegant series ofpapers (Bungenberg de Jong and van Someren,1959; Hooghwinkel, Lexmond, and Bungenbergde Jong, 1959; Hooghwinkel and van Niekerk,1960; Bungenberg de Jong, 1961). Because of itsavailability, we used acid fuchsin instead of EdicolSupra Ponceau 4 RS or other Ponceau dyes thatalso gave less intense colors. The acid fuchsinwas certified reagent quality obtained from FisherScientific Co., Pittsburgh, Pa. (C.I. no. 42685).The dye solution consisted of 0.02% acid fuchsinin 0.2% U02(NO3)2-6H20, 0.01 N HCl. The solu-tion for rinsing out excess acid fuchsin was thesame, but with the dye omitted. The brilliantgreen was certified and obtained from the NationalAniline Division of the Allied Chemical and DyeCorp. (C.I. no. 662). It was used as a 0.01% solu-tion in 0.01 N HCl after aging at about 24 C for16 hr or more. Excess brilliant green was rinsedfrom chromatograms with 0.1 M acetic acid. ThepH 5 solution that was described above for elutingboth stains from chromatograms was developedbecause brilliant green was destroyed by the moreacid solvent recommended for acid fuchsin or thePonceau dyes (Hooghwinkel and van Niekerk,1960). Acid fuchsin did not visibly stain PTI, butbrilliant green, which was strongly bound by theacidic PTI, also gave faint blue spots with PTE,PTC, lyso PTC, SPH, and phosphatidyl serine(PTS). In our tests, only SPH was of concern,since it partly overlapped PTI on chromatograms.However, the color contributed by SPH was in-significant when compared with that given byPTI in our samples, so that it was possible tomeasure PTI by staining with brilliant green.The blue color obtained with PTI approached anend point that was about 25% of the control valueunder most conditions of testing with the phos-pholipases. For this reason, a decrease of 75%from the absorbancy of the control sample wasassumed to represent complete degradation ofPTI in our samples, and the values were adjustedaccordingly. The absorbancy was proportional toconcentration when 5-, 10-, and 15-pliter samplesof the phospholipids were separated by paperchromatography.The acid-soluble products split from the phos-

pholipids were obtained by stopping enzyme ac-

71VOL. 90, 1965

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SLEIN AND LOGAN

tion by adding trichloroacetic acid to give a finalconcentration of 5%. After chilling in ice for 15min, the precipitate was removed by filtrationthrough Schleicher and Schuell no. 602 ED paper.When phosphorus determinations were to bemade, the phosphate buffer of the reaction mix-ture was replaced by 2,4,6-trimethylpyridinebuffer. The acid-soluble products were separatedwithout further treatment by descending chroma-tography on Whatman no. 1 paper for 6 hr atabout 24 C with the isopropanol-glacial aceticacid-water (3:1:1, v/v) solvent of Kerr andKfoury (1962). The phosphorus-containing spotswere made visible by spraying with the Hanes andIsherwood reagent (1949), heating for 1 min atabout 85 C, and then exposing the paper to ultra-violet radiation (Bandurski and Axelrod, 1951).The spots were cut out, ashed for about 15 min atabout 200 C until the paper was digested and thesample bleached, and phosphorus was determinedaccording to the procedure of Gerlach and Deu-ticke (1963). Total phosphorus in the acid filtrateswas determined by the method of Fiske andSubbaRow (1925) or by the procedure of Gerlachand Deuticke (1963). It was not possible to meas-ure inorganic orthophosphate directly in the acidfiltrates, because a precipitate formed duringcolor development. Therefore, it was estimatedafter separation on Whatman no. 1 paper with asolvent described by Wood (1961) [isopropylether-n-butanol-90% (w/v) formic acid (30:30:20, v/v)]. The chromatography was carried outwith the solvent descending for 16 hr at about24 C.

Inositol analysis. The occurrence of PTI in thesoybean "lecithin" was suspected on the basis ofits RF value and staining properties with Rhoda-mine 6 G (Marinetti, 1962), as well as with bril-liant green (Bungenberg de Jong, 1961). Chemicalmethods for the analysis of inositol are not spe-cific, and biological assays are not sufficientlysensitive. A specific method that uses inositoldehydrogenase isolated from Aerobacter aerogenesgrown with inositol as carbon source was reportedby Weissbach (1958). With the crude extract as asource of enzyme, the method is not reliable, be-cause reoxidation of the reduced nicotinamideadenine dinucleotide (NADH2) occurs and pre-vents the attainment of a stable end point. Thisdifficulty was overcome by using the proceduredescribed by Larner et al. (1956) modified in thefollowing way. Although partial purification ofthe dehydrogenase was reported by Larner et al.(1956) and Larner (1962), we obtained good pro-portionality between inositol concentration andabsorbancy even with a crude extract. A mucoidstrain of A. aerogenes was kindly supplied by R.A. Altenbern of Fort Detrick. The cells weregrown as described by Magasanik (1953) and wereharvested by centrifugation at about 38,000 X gat 2 C. The cells were washed twice with colddistilled water and were suspended with cold 0.05M tris(hydroxymethyl)aminomethane (Tris), pH7.5, to give approximately 20 g of cells (wet

weight) per 40 ml of suspension. The cells weredisrupted for 20 min at 1.25 amp in a 10-kc Ray-theon sonic oscillator cooled with fluid circulatingat about -5 C. The material was centrifuged for30 min at about 38,000 X g at 2 C, and the slightlyopalescent supernatant fluid containing about 9mg of protein per ml was stored frozen. Samplescontaining 0.1 and 0.2 Amole of myo-inositol andsamples with neutralized hydrolyzates were pre-pared along with a control without inositol andanother without inositol, NAD, or enzyme. Thesamples were made up to 1 ml with final con-centrations of 0.04 M Tris (pH 8.6), 0.025 MK3Fe(CN)6, and 0.002 M NAD, and were incu-bated at 37 C for a time determined to give maxi-mal Prussian blue color development and absorb-ancies proportional to the two concentrations ofinositol. An incubation of 15 to 30 min was suffi-cient with 0.9 mg of protein from one crude extractthat we prepared. The reaction was stopped bythe addition of 0.1 ml of 100% (w/v) trichloro-acetic acid. After the addition of 1.7 ml of waterand 0.2 ml of Duponol-iron reagent (Larner et al.,1956), the samples were mixed and read in a Beck-man DU spectrophotometer at 560 m,u withinabout 15 min. The control without added inositol,NAD, or enzyme served as the reference solution.The absorbancy was doubled at 620 muM and tripledat 700 m/,, so that these wavelengths might beused to increase the sensitivity. Although theabsorbancies continued to increase with time, thevalues, corrected for the control without addedinositol, remained fairly constant, especially at560 mu (an absorbancy of about 0.1 per 0.1 ,umoleof myo-inositol).

Since inositol dehydrogenase will act only onfree inositol, it is necessary to hydrolyze PTIbefore analysis. When the inositol content of aspot on a chromatogram was desired, the areawas cut out and eluted repeatedly with chloro-form-methanol (7:3, v/v); the eluates were driedin a Flash-Evaporator and transferred with asmall amount of the solvent to a tube (7.5 X 75mm) for drying. After the addition of 0.1 ml of6 N HCl, the end of the tube was sealed off, andthe sample was hydrolyzed for 6 hr at 110 C (Bohmand Richarz, 1954). The hydrolysate was neutral-ized with a measured volume of NaOH. The sam-ple was centrifuged and the supernatant solutionwas analyzed for inositol. Since the completion ofthis work, a fluorometric micromethod for theenzymatic determination of inositol has been de-scribed (Garcia-Buniuel and Garcia-Bufiuel, 1964).

Extraction of phospholipids from cells. The pro-cedure of Bligh and Dyer (1959) was used to ex-tract phospholipids from aqueous suspensions ofwashed erythrocytes or from rabbit kidney cortexhomogenates. Relatively small volumes of ma-terials were extracted in a VirTis 45 homogenizerat low speed or in a Waring Blendor with the speedreduced by means of a rheostat to avoid excessivesplattering. Homogenate (9 ml) was extracted andcentrifuged briefly to clump the particles andfacilitate filtration. After filtration, the mixture

72 J. BA~CTERIOL.

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PHOSPHOLIPASES OF B. CEREUS

was centrifuged for 10 min, and the liquid phaseswere separated with practically no film at theinterface. The upper layer was carefully removedby aspiration through a fine-tipped pipette so thatessentially no loss of the lower chloroform layeroccurred. The latter was dried within a few min-utes at 45 C in a Flash-Evaporator with the con-denser in an ice bath. The residue was carefullydissolved with 0.5 ml of isoamyl alcohol-benzene(1:1, v/v) and was transferred by pipette to asmall tube for storage at -15 C.

Practically all of the phospholipids of maturemammalian erythrocytes are located in the cellmembrane (Ways and Hanahan, 1964). Thereforeit is possible to measure the phospholipid contentof the membranes by the use of whole erythro-cytes. Erythrocytes were washed three to fourtimes with isotonic NaCl; leukocytes were re-moved by aspiration after each centrifugation.The packed cells were suspended with an equalvolume of Ringer phosphate solution (pH 7.5),and as little as 6 ml was used for the extraction ofphospholipids. It was not necessary to centrifugethe suspension of extracted erythrocytes beforefiltration. After filtration, the mixture was cen-trifuged, and the upper layer, with as much inter-facial precipitate as possible, was carefully re-moved by aspiration. The chloroform layer wasdried, and the residue was dissolved, as describedabove for kidney.Hemolysin activity. Sheep and rabbit erythro-

cytes were washed and leukocytes were removed,as described in the preceding paragraph. Thepacked cells were diluted 1:40 with Ringer phos-phate (pH 7.5). To 4.8 ml of the diluted erythro-cytes in a 12-ml centrifuge tube at 37 C was added0.2 ml of the solution to be tested for hemolysinactivity. The tube was closed with a Saran-wrapped stopper; the contents were mixed severaltimes by inversion and incubated at 37 C for 30min. The tube was quickly chilled in ice, and wascentrifuged at 5 C; 1 ml of the supernatant fluidwas mixed with 1 ml of Drabkin's reagent. After10 min, the cyanmethemoglobin color was read at540 m, in a Beckman DU spectrophotometer withDrabkin's reagent and diluted with an equal vol-ume of water, as the reference solution (Wintrobe,1956). Readings were corrected for slight spon-taneous hemolysis of control samples. Completehemolysis resulted in absorbancies of about 2.0 to2.5 in control samples with excess crude PF. Whenthe destruction of hemolysin activity by trypsinwas studied, the percentage of residual activitywas calculated from a curve obtained with gradedamounts of untreated hemolysin.The relationship between hemolysis and the

phospholipid content of erythrocyte membraneswas studied by suspending the packed washedcells with an equal volume of Ringer phosphate(pH 7.5) and incubating with phospholipase at37 C. After incubation, 0.2-ml samples were di-luted with 3.8 ml of Ringer phosphate or water,centrifuged, and 1 ml of each supernatant fluidwas mixed with 1 ml of Drabkin's reagent for

measuring hemolysis. The sample that had beendiluted with water gave the value for completehemolysis. The bulk of the phospholipase-treatederythrocyte suspension was used for the extrac-tion and analysis of phospholipids.

Treatment with trypsin. Crystalline trypsin wasobtained from Worthington Biochemical Corp.,Freehold, N.J. Soybean trypsin inhibitor (STI)was a product of the Armour Laboratories. Thetrypsin was dissolved in distilled water to give asolution of 2 mg/ml. The same concentration ofSTI was preapred with 0.01 M Tris (pH 7.5). Phos-pholipase fractions were treated with approxi-mately 0.1 mg of trypsin per ml in 0.02 M Tris(pH 7.1) for various times at 37 C. The trypsinactivity was stopped by adding an equal weightof STI. Control samples were incubated withSTI added before the trypsin.

Preparation of lysophosphatides. Approximately100 mg of commercial soybean "lecithin" wereincubated with 0.5 mg of Crotalu.s adamanteusvenom (Sigma Chemical Co., St. Louis, Mo.) for2 hr at 37 C as an emulsion buffered with 0.05 Mphosphate (pH 7.0). Most of the PTC and PTEof the "lecithin" was converted to the correspond-ing lysophosphatides.

RESULTSEnzymatic degradation of phospholipids. Tests

with crude PF preparations, such as the prota-mine-treated ammonium sulfate precipitate de-scribed previously (Slein and Logan, 1962), show-ed that PTE, PTC, PTI, and SPH were readilydegraded. The effect on lyso PTC was much lessmarked, if at all significant. Since the commercialsoybean "lecithin" contained too little lyso PTCfor accurate testing, it was supplemented with acommercial product. A mixture produced by theaction of snake venom on soybean "lecithin"contained both lyso PTE and lyso PTC (seeMaterials and Methods). Neither of these lyso-phosphatides appeared to be significantly affectedby crude PF, although the residual parent com-pounds and PTI were degraded by the PF. Thedegradation of PTI, PTC, PTE, and SPH byvarious amonts of crude PF preparation in 30min is shown in Fig. 1. Because of the complexityof the test system and the analytical procedure,the activities of various fractions (see below) weremerely compared by calculating the percentage ofdegradations obtained under conditions that ledto partial decomposition of the various phospho-lipids, and no activity units have been defined.Since activities were not linearly proportional toenzyme concentration under these conditions, itwas necessary to construct curves like those inFig. 1 with intact enzymes to calculate the per-centage of inactivation caused by trl)psin.The PTI in the commercial "lecithin" was

identified by analyzing the brilliant green-stained

VOL. 90, 1965 7,3

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SLEIN AND LOGAN

PTI

c0

-0ah.0V0C-

4

Protein (U!

FIG. 2. Distribution of protein and of sheep andrabbit hemolysins in fractions obtained by the elu-tion of a crude preparation of phosphatasemia factorfrom a DEAE cellulose column with increasing

8 12 concentrations of Tris buffer (pH 8.6). Symbols:A = molarity of Tris; 0 = protein concentration;

9) E = rabbit hemolysin; 0 = sheep hemolysin.FIG. 1. Degradation of phospholipids in an emul-

sion by different amounts of a crude phospholipasepreparation in 30 min at pH 7 and 37 C. For ab-breviations, see footnote of Table 1.

material after it was eluted from a paper chro-matogram. The mono PTI spot accounted forabout 50% of the total inositol found in thedirectly analyzed "lecithin" emulsion. A secondinositol-containing spot, which was assumed to bea diphosphatide, had a lower RF than the prin-cipal PTI spot. Approximately 6 mg of monoPTI were present in 1 ml of the "lecithin" emul-sion that contained 100 mg (dry weight).The acid-soluble phosphorus compounds ac-

counted for practically all of the phospholipidsdegraded by the phospholipases in crude PF. Forexample, soybean "lecithin" was incubated withcrude PF for 1 hr so that 90 to 98% of the PTI,PTE, and PTC were degraded. Direct analysis ofthe trichloroacetic acid filtrates showed that 1.688mg of acid-soluble phosphorus had been releasedfrom 1 ml of emulsion. Analysis of the phosphorylcholine, phosphoryl ethanolamine, and phos-phoryl inositol products from paper chromato-grams yielded 1.754 mg of phosphorus. Thus,104% of the acid-soluble phosphorus released wasrecovered in the products that represented theprincipal phospholipids of the "lecithin" emul-sion. When another preparation of PF wasincubated with soybean "lecithin" supplementedwith SPH, 89% of the acid-soluble phosphorusreleased was accounted for by the same products.In either case, inorganic orthophosphate ac-counted for only about 2% of the total acid-soluble phosphorus released. These results indi-cate that the phospholipase of culture filtrates ofB. cereus liberate diglycerides and acid-soluble

phosphate esters from the phosphatides, andphosphoryl choline from SPH. Very littlephosphomonoesterase activity is present in crudePF under these conditions at pH 7. Other testshave shown that crude PF also has no significantalkaline phosphatase activity at pH 10.5 withp-nitrophenyl phosphate as substrate. The factthat no lysophosphatides were ever found toaccumulate on chromatograms after incubationof phospholipids with crude PF or any of thefractions, although the lysophosphatides were notsignificantly degraded by these preparations,indicates that no significant "phospholipase A"activity exists in them.A preparation containing 19 mg of crude PF

protein was fractionated on N,N'-diethylamino-ethyl (DEAE) cellulose as described previously(Slein and Logan, 1963). Approximately 80% ofthe protein was recovered in 119 fractions ofabout 5 ml each (Fig. 2). Protein was concen-centrated intwo principal regions: fractions 30 to50 and fractions 109 to 115. The latter sharp peakresulted from the sudden increase of Tris concen-tration from about 0.8 to 1 M to elute most of theresidual protein from the column. Very littlephospholipase activity was found below fraction30. After a preliminary survey, fractions wereselected for further tests to determine their rela-tive phospholipase activities. The results are pre-sented in Table 1. The original crude PF attackedall four phospholipids. The separation of threediscrete activities is indicated by the data. Frac-tions 35 to 50, which were eluted by 0.1 to 0.2 MTris (pH 8.6), degraded PTE and PTC. It is notclear whether one or two enzymes are involved inthe degradation of these two phospholipids. Aphospholipase that attacked SPH began to ap-

74 J. BACTERIOL.

Fraction Number

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PHOSPHOLIPASES OF B. CEREUS

TABLE 1. Relative phospholipase activities offractions obtained by elution from a DEAE

cellulose column with increasingconcentrations of Tris

buffer (pH 8.6)

Fraction Amt of Per cent degradationtproteinno. tested PTE PTC SPH PTI

pg

Original 23.6 46 66 56 100material

38 17.1 29 26 0 044 10.6 20 17 0 046 7.3 15 18 0 051 4.8 12 7 10 057 3.3 0 0 35 1662 2.2 0 0 55 2966 4.0 0 0 51 5272 4.9 0 0 26 8380 3.4 0 0 6 97100 4.3 0 0 9 55110 39.4 0 0 7 61

*See Fig. 2.

t Samples were incubated for 1 hr at 37 C.PTE = phosphatidyl ethanolamine, PTC =

phosphatidyl choline, SPH = sphingomyelin, andPTI = phosphatidyl inositol.

pear at about fraction 50 and had highest activityin fraction 62. The enzyme that degraded PTIoverlapped the preceding, but was concentratedin fractions 70 to 80 that were eluted by about0.4 M Tris. A similar distribution of activities wasobtained with two other fractionations of crudePF.

Hemolysin activity. The data in Fig. 2 show thedistribution of hemolysin activities as measuredwith sheep and rabbit erythrocytes. No signifi-cant hemolysin activity was detected among thefirst 29 fractions that were omitted from thefigure. No attempt was made to determine specificactivities, and the results merely indicate theabsorbancies obtained after the incubation oferythrocytes for 30 min at 37 C with 0.1 ml ofeach fraction. In some cases, excess hemolysinmay have been present when hemolysis was es-sentially complete (absorbancies of about 2.0).Sheep erythrocyte hemolysin appeared in twobroad series of fractions, whereas rabbit eryth-rocyte hemolysin activity was found in oneregion that coincided roughly with one of theprincipal sheep hemolysins. Approximately 1j,g of protein of fraction 62 was sufficient toproduce the marked hemolysis noted in Fig. 2.

Changes in the two major phospholipid com-

ponents of sheep erythrocyte membranes thatoccurred during partial hemolysis by crude phos-pholipase preparations are presented in Table 2.

TABLE 2. Loss of phospholipids from sheeperythrocyte membranes during partial

hemolysis by crude phospholipase

Per centExpt* Proteint Time Hem.ol- degradation$

PTE SPH

pg min %1 . .6.8 A 30 34 15 9

None 30 5 - -

2... 68 A 30 61 40 3368 A 60 71 65 62None 60 0.3 -

3... 14.2 B 15 32 18 1314.2 B 60 42 34 36None 60 1 -

* In experiments 1 and 2, the volume of washederythrocyte suspension was 25 ml; in experiment3 it was only 6 ml.

t A and B were different preparations of crudephospholipase.

t For abbreviations, see footnote of Table 1.

Both preparations were able to degrade PTC,PTE, PTI, and SPH in emulsions of the phos-pholipids. Increasing the concentration of phos-pholipase 10-fold in experiment 2 as comparedwith experiment 1 increased the degradation ofphospholipids about threefold to fourfold, buthemolysis was not quite doubled. Incubation fora second 30-min period in experiment 2 resultedin a much greater degradation of phospholipidsrelative to the increase in hemolysis. Similarly,with another preparation in experiment 3, a smallchange in hemolysis was accompanied by muchgreater phospholipid destruction.

Prior to testing the sensitivity of hemolysin totrypsin, the effect of soybean "lecithin" on hemol-ysin activity was studied for possible use as anagent to protect against trypsin in the way thatPF was protected (Slein and Logan, 1963). How-ever, hemolysin was found to be markedly in-hibited by such low concentrations of "lecithin"that the latter could not be used as a protectiveagent. The inhibition by "lecithin" is shown inTable 3. When 1 ,ug of crude phospholipaseprotein was mixed with "lecithin" at 0 C immedi-ately before adding to the erythrocytes (experi-ment 1), the inhibition of hemolysin activitydecreased with decreasing concentrations of"lecithin" and was not complete even with 100,ig. In experiment 2, however, incubation of thephospholipase with only 7 ,ug of "lecithin" for15 min at 37 C before adding it to the red cellsinhibited hemolysis completely. Inhibition wasmarked even after 20 min at 0 C.

75VOL. 90, 1965

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SLEIN AND LOGAN

TABLE 3. Effect of soybean "lecithin" on thehemolysis of sheep erythrocytes

by crude phospholipasea

Expt Incubationb "Lecithin"c Hemolysis

Ag %1.... None 250 1.2

None 100 6.9None 50 11.0None 40 13.5None 30 15.5None 20 18.6None 10 28.9None 0 37.6None 0d 0.8

2.... 15 min 37 C 7 0.815 min 37 C 0 27.120 min 0 C 7 6.420 min 0 C 0 28.1

None 7 21.5None 0 31.2None Od 0.8

a In each case (except for the controls) 1 ,ugof crude PF protein was present during hemolysisfor 30 min at 37 C.bPF was incubated with "lecithin" before add-

ing to the erythrocyte suspension. None = PFwas mixed with "lecithin" at 0 C immediatelybefore adding to the erythrocytes.

c Besides PTC, the "lecithin" contained PTE,PTI, and other contaminants (see Materialsand Methods). For abbreviations, see footnoteof Table 1.

d Control sample without PF.

Effect of PF fractions on rabbit bone and kidneycortex homogenates. The phospholipase that waseasily eluted from DEAE cellulose had no PFactivity and inhibited the PF activity of a frac-tion eluted by about 0.4 M Tris when tested withthe bone slice assay method (Slein and Logan,1963). Unlike the results with bone slices andsimilar results obtained with bone homogenates,either fraction was able to release phosphatasefrom kidney cortex homogenates. Furthermore,when both fractions were present with the kidneyhomogenate, the release of alkaline phosphatasewas stimulated rather than inhibited. A com-parison of results obtained with the two rabbittissues is shown in Table 4. The fraction A phos-pholipase was a combination of fractions elutedfrom DEAE cellulose by low concentrations ofTris and degraded only PTC and PTE in ourphospholipase tests. The PF was a crude prepa-ration. With the bone homogenate, as had alsobeen noted with bone slices, fraction A inhibitedthe spontaneous leakage of phosphatase (negativecorrected value) as well as PF activity. With the

TABLE 4. Effect of the phospholipase that is readilyeluted from DEAE cellulose (fraction A) on

the release of alkaline phosphatase fromrabbit epiphyseal bone and kidney

cortex homogenates by crudephosphatasemia factor (PF)

Homogenate Additions Alkaline phosphatase(absorbancy)*

Bone......... PF 0.519Fraction A -0.115Both 0.203

Kidney ........ PF 0.345Fraction A 0.275Both 1.261

* Average of duplicate values corrected for thespontaneous release of phosphatase that occurredin the absence of PF or fraction A.

TABLE 5. Relative phosphatasemia factor (PF)activities of fractions obtained by elution froma DEAE cellulose column with increasing

concentrations of Tris buffer (pH 8.6)

Fraction no.* PF actiyityt(units/pg of protein)

Original material 98.538 0.1744 0.4446 0.5251 0.6657 2.4662 4.6066 7.9772 302.080 1338.0100 71.2110 10.6

* See Fig. 2.t Average of two or more determinations made

with rabbit kidney cortex homogenates.

kidney homogenate, either material caused therelease of phosphatase, but the combinationresulted in a stimulation twice that expected fromthe sum of the effects of each alone. This is dis-tinct from the proportionality of phosphataserelease to PF concentration that is obtained withpreparations that have high PF activity (seefractions 72 and 80 in Table 6).Kidney cortex homogenates were used for

comparing the PF activities of the same fractionsfor which the phospholipase activities are givenin Table 1. The results are presented in Table 5.Of the fractions tested, PF activity was mosthighly concentrated in fraction 80, which alsohad the greatest PTI-phospholipase activity(Table 1). The effects of combinations of certain

76 J. BACTERIOL.

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PHOSPHOLIPASES OF B. CEREUS

TABLE 6. Effect of combinations of phospholipasefractions on the release of alkaline phosphatase

from rabbit kidney cortex homogenatesPhosphatase

Expt Fraction Protein released(absorbancy)

jug1 38 8.55 0.110

38 17.10 0.07472 0.0075 0.33572 0.015 0.66438 8.55

plus 1.00672 0.0075

2 .. 38 8.55 0.18280 0.0015 0.27180 0.030 0.53138 8.55

plus 0.81980 0.0015

* Values corrected for the spontaneous releaseof alkaline phosphatase that occurred in theabsence of added phospholipase.

fractions on the PF activity with kidney homoge-nates are shown in Table 6. The low PF activityand lack of proportionality with two concentra-tions of fraction 38 are typical of results obtainedwith the phospholipase that is easily eluted fromDEAE cellulose, in contrast to material havinghigh PF activity, such as fraction 72 or 80. Onlya few millimicrograms of the proteins of fractions72 and 80 were needed for a marked release ofalkaline phosphatase from kidney cortex cells.Combinations of fraction 38 with fraction 72 or80 resulted in approximately twofold stimulationsof phosphatase release over that expected fromthe sum of the effects of each alone.The presence of 1 mg of commercial soybean

"lecithin" in 0.7 ml of the kidney homogenatereaction mixture completely inhibited the releaseof alkaline phosphatase by a crude PF prepara-tion. This agrees with the effect reported for thebone slice test system (Slein and Logan, 1963).A few tests were made to relate the liberation

of phosphatase and the degradation of phospho-lipids naturally present in kidney cortex homoge-nates by phospholipases. A combination of severalfractions eluted from DEAE cellulose by 0.05 MTris (fraction A) was compared with a combina-tion eluted by 0.4 M Tris (fraction D). Althoughfraction D released about 10 times as much phos-phatase as did fraction A, the latter degradedpractically all of the PTE and PTC in thehomogenate, whereas fraction D degraded onlyPTI significantly. In another test with crude PF,analysis for inositol revealed that the phospho-

lipid extract of the control homogenate containedabout 2.8 A.moles of inositol per ml of phospholip-ids, whereas the sample treated with PF had only0.8 ,umole. These results might be interpreted toindicate that the degradation of a minor phospho-lipid component (PTI) in the kidney cell mem-brane is more critical for the damage that resultsin the release of alkaline phosphatase than is thedegradation of PTE and PTC. However, unlikethe mature erythrocytes, kidney cells have phos-pholipids in components other than the externalmembranes (e.g., mitochondria) so that nocorrelation between membrane damage and phos-pholipid degradation can be made with certaintyuntil methods for the isolation of mammalian cellmembranes are available.

Effect of trypsin on the enzymatic activities ofcombined phospholipase fractions. To have suffi-cient material for analysis, it was necessary tocombine the residues of groups of fractions, con-centrate them in Carbowax (polyethylene glycol),and dialyze against 0.01 M Tris (pH 7.5) at 5 C.Groups with similar activities were selected forcombination as summarized in Table 7. The dis-tribution of activities corresponded, in general, tothat expected from tests with individual fractionsas shown in Fig. 2 and in Tables 1 and 5. Thehemolysin activities of fraction III in Table 7were obtained with 8.2 ,ug of protein, whereas nosignificant hemolysis occurred with about 2 ,ug offractions 70 to 76 when originally tested (Fig. 2).The effects of trypsin on the various activities

are presented in Table 8. Control tests withequivalent amounts of trypsin plus STI weremade to be sure that these substances themselvesdid not have any of the activities to be measured.Neither the phospholipase activity of fraction Inor its slight PF activity was significantly in-activated by trypsin. The phospholipase thatdegraded SPH also seemed to be relativelyresistant to trypsin. PTI-phospholipase and PFactivities were rather sensitive to inactivation bytrypsin. Fraction II hemolysin was also relativelysensitive to trypsin, but the other fractions hadintermediate values.

DISCUSSIONThe results obtained with preparations from

culture filtrates of B. cereus have been duplicatedin most respects with B. anthracis, strain Sterne(Sterne, 1937). The phospholipase activity isweaker in culture filtrates of B. anthracis (Sleinand Logan, 1962), but fractionation of the crudematerial separates the various phospholipaseactivities in a manner entirely analogous to thatobtained with B. cereus. The interaction betweenfractions in the bone slice and kidney homogenatetests for PF is also the same for the two species.

77VOL. 90, 1965

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

TABLE 7. Enzymatic activities of groups of fractions that were combinedafter separation on a DEAE cellulose column

Hemolysinb Per cent degradation of phospholipidsc FPFd activity

Fractiona (units/pg of pro-Protein PE TCI tein)Sheep Rabbit tested PTE PTC SPH PTI

MgOriginal material. 0.530 0.514 23.6 46 66 56 l00 98.5

(2.84) (18-9)I (37-43) ............ 0.005 0.009 38.5 46 66 0 9 0.22

(19) (38.5)11 (59-66)............ 1.309 1.223 7.13 0 6 58 47 7.2

(1.0) (2.14)III (70-76)... .... 0.219 0.814 1.0 0 17 80 1170

(8.2) (8.2)IV (80-86) .. . 0.760 0.085 1.18 0 0 19 81 936

(0.6) (5.9)

a Numbers in parentheses represent the fractions in Fig. 2 that were combined.b Absorbancy values corrected for the slight spontaneous hemolysis of control samples. Complete

hemolysis would be indicated by an absorbancy of about 2.0. The,ug of protein tested are shown inparentheses.

c Samples were incubated for 1 hr at 37 C with the amounts of protein shown. For abbreviations, seefootnote of Table 1.

d Phosphatasemia factor (PF) activity was determined with rabbit kidney cortex homogenates.

TABLE 8. Effect of trypsin on enzymes present in groups of fractions that were combinedafter separation on a DEAE cellulose column

Hemolysin inactivation Phospholipasec inactivationFractiona trypsinb -PFd inactivation

Sheep Rabbit PTE PTC SPH PTI

min % % N% % %Original material 30 40 34 7 13 16 70 83

1 (37-43) ........ 20 .e . 0 0 060 -- 0 0 - - 0

II (59-66) ............ 30 80 80 - - 11 72 >90III (70-76) ............ 15 39 56 - _ 27 84 97IV (80-86) ............ 15 52 - - 18 81 100

30 63 - 28 89 100

a Numbers in parentheses represent the fractions in Fig. 2 that were combined.I Approximately 0.1 mg of trypsin per ml except for fraction I, which was treated with about 0.3

mg per ml.c For abbreviations of phospholipids, see footnote of Table 1.d Phosphatasemia factor (PF) measured with kidney cortex homogenates.e Dashes indicate that activities were absent or were too low for accurate measurement and were not

tested here (see Tahlp 7\

It seems likely that the phospholipases may bepart of the toxin complex of B. anthracis as wellas that of B. cereus.The "lecithinase" of B. cereus has been found

to comprise a group of phospholipases that areable to split phosphoryl choline, phosphorylethanolamine, and phosphoryl inositol from thephospholipids PTC, SPH, PTE, and PTI. Apossible lack of specificity has been ascribed tothe "lecithinase" of B. cereus by others (Chu,1949; Robinson, Harris, and Poole, 1957; Kush-

ner and Feldman, 1958). We also noted slightactivity with phosphatidyl serine (PTS), but thiswas not pursued further, because it was notpossible to separate PTS from PTE sufficientlyfor accurate determination with the chromato-graphic procedure that we used. The traceamounts of PTS in the materials that we analyzeddid not interfere significantly with the determina-tion of PTE. The phospholipases that degradePTI and SPH have been separated from thosethat attack PTE and PTC. The enrichment of

78 SLEIN AND LOGAN J. BACTERIOL.

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PHOSPHOLIPASES OF B. CEREUS

the PTI- and SPH-phospholipases in separatefractions, as well as their different susceptibilitiesto attack by trypsin, indicates that they are in-dividual enzymes. We have not succeeded inseparating the PTE- and PTC-splitting activitiesto any degree that would suggest the presence oftwo enzymes. The fact that these activities areboth resistant to trypsin (Table 8) supports theidea that one phospholipase ("lecithinase") maydegrade both substrates. The "lecithinase" of B.anthracis was also found to attack PTE and PTC(Costlow, 1958). The phospholipase preparationsof B. cereus appeared to have little, if any, phos-phomonoesterase or phospholipase A activitiesunder our conditions of testing.Hemolysin activity does not seem to be associ-

ated with the PTE- or PTC-phospholipase activ-ity (fraction I, Table 7). This agrees with ourprevious conclusion with regard to the "inhibitor"phospholipase (Slein and Logan, 1963) and withthe findings of Ottolenghi, Gollub, and Ulin(1961) that hemolysin could be separated fromphospholipase. The separation of hemolysin and"egg yolk turbidity factor" was reported byFossum (1963). Our results also indicate that thehemolysins are not identical with the SPH- orPTI-phospholipases. Although SPH-phospholip-ase was concentrated in fraction II that includedfractions having good hemolysin activities forboth sheep and rabbit erythrocytes (Fig. 2 andTable 7), the hemolysin activities appeared to bemore sensitive to trypsin than did the SPH-phos-pholipase (Table 8). On the other hand, the PTI-phospholipase was generally more susceptible toinactivation by trypsin than were the hemolysins(fractions III and IV, Tables 7 and 8). Further-more, fraction III had much more PTI-phospho-lipase activity than fraction II, but the latter hadgreater hemolysin activity (Table 7). Althoughfractions III and IV had rather similar SPH- andPTI-phospholipase and PF activities, fractionIV was much more hemolytic for sheep than forrabbit erythrocytes (Table 7). Other data inTable 1 and Fig. 2 support the independence ofphospholipase and hemolysin activities.

In agreement with others, our chromatogramsshowed that the principal phospholipids of rabbiterythrocyte membranes are PTE, PTC, andSPH; those of sheep red cells are PTE and SPH(de Gier and van Deenen, 1961; Klibansky,Condrea, and de Vries, 1962). Hemolysin activitymay be primarily caused by damage to someminor phospholipid or other constituent that wehave not measured. The attack of membranephospholipids by the phospholipases may occursecondarily after their exposure by the action ofa true hemolysin in the manner reported forsnake venoms (Condrea, de Vries, and Mager,

1964). The degradation of phospholipids in cellsthat have already hemolyzed may also accountfor the loss of phospholipids being relativelygreater than the increases in hemolysis, as notedin Table 2. However, it is possible that degrada-tion of the principal phospholipid components ofthe erythrocyte membrane contributes to hemoly-sis. The mechanism by which a crude soybean"lecithin" inhibits hemolysin activity is notknown. The augmentation of the inhibition byincubation of the "lecithin" with the crudehemolysin indicates that phospholipase activitymay be involved. Previous results have shownthat "lecithin" markedly inhibited the PF activ-ity on bone slices (Slein and Logan, 1963). Thiswas interpreted as a competition between phos-pholipids of the "lecithin" emulsion and thenatural substrates in the bone cell membrane forthe site of the phospholipase responsible for PFactivity. A similar inhibition of PF activity hasbeen noted with kidney cortex homogenates.PTE- and PTC-phospholipase activity does

not seem to be essential for PF activity, althoughit stimulates the release of phosphatase fromkidney cells by fractions that have relatively highPF activities. SPH-phospholipase also does notappear to be important for PF activity, since itwas concentrated in fraction 62 that had rela-tively low PF activity (Tables 1 and 5), and,unlike PF, it was rather resistant to trypsin(Table 8). On the other hand, PTI-phospholipasewas associated with high PF activity (Tables 1and 5) and was sensitive to trypsin (Table 8).The differences in the effects of various phos-

pholipases on the release of alkaline phosphatasefrom bone and kidney cells are very interesting.The material that had only PTE- and PTC-phos-pholipase activity not only failed to liberatephosphatase from bone cells but also suppressedthe rate of spontaneous leakage of phosphatasefrom the cells. Although it had slight PF activitywith kidney cells, the activity tended to remainthe same or to decrease slightly with increasingconcentrations of the protein. This may be inter-preted as an indication of the presence in themembranes of no, or relatively few, PTE andPTC molecules that are available to attack bythe specific phospholipase(s) and that these sitesmay become saturated with the enzyme. Withbone cells, the attachment of these phospholipaseproteins may hinder the approach of other phos-pholipases to their specific phospholipid sub-strates that may be more important for the in-tegrity of the membrane. On the other hand, suchphospholipids as PTI may be readily available tothe highly active PF phospholipase(s), or becomeso during degradation of both bone and kidneycell membranes, so that damage to the mem-

VOL. 90, 1965 79

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SLEIN AND LOGAN

branes increases directly with the amount of PFadded and results in a proportional increase inphosphatase liberation. In the case of the kidneycell, the two types of enzymes may not interferewith each other but may result in larger "holes"in the membrane, so that a stimulation of phos-phatase release is obtained. Other explanationsmight be invoked, based on differences in specificactivities of the phospholipases, differences incharge distributions of phospholipids in the cellmembranes, etc.

LITERATURE CITED

BANDURSKI, R. S., AND B. AXELROD. 1951. Thechromatographic identification of some biologi-cally important phosphate esters. J. Biol. Chem.193:405-410.

BANGHAM, A. D., AND R. M. C. DAWSON. 1962.Electrokinetic requirements for the reactionbetween Cl. perfringens a-toxin (phospholipaseC) and phospholipid substrates. Biochim. Bio-phys. Acta 59:103-115.

BLIGH, E. G., AND W. J. DYER. 1959. A rapidmethod of total lipid extraction and purifica-tion. Can. J. Biochem. Physiol. 37:911-917.

BOHM, P., AND G. RICHARZ. 1954. Zur quantita-tiven Bestimmung von Inosit in Phosphatiden.Z. Physiol. Chem. 298:110-120.

BUNGENBERG DE JONG, H. G. 1961. Improveddetection of phosphatides on paper chromato-grams with a mixed staining solution. Effectof contaminating substances in the paper.Koninkl. Ned. Akad. Wetenschap. Proc. Ser. B64:445-469.

BUNGENBERG DE JONG, H. G., AND G. R. VANSOMEREN. 1959. Acid Fuchsin and uranyl ni-trate in staining chromatograms of phospha-tides. Koninkl. Ned. Akad. Wetenschap. Proc.Ser. B 62:150-160.

CHU, H. P. 1949. The lecithinase of Bacillus cereusand its comparison with Clostridium welchii a-toxin J. Gen. Microbiol. 3:255-273.

CONDREA, E., A. DEVRIES, AND J. MAGER. 1964.Hemolysis and splitting of human erythrocytephospholipids by snake venoms. Biochim. Bio-phys. Acta 84:60-73.

COSTLOW, R. D. 1958. Lecithinase from Bacillusanthracis. J. Bacteriol. 76:317-325.

FISKE, C. H., AND Y. SUBBAROW. 1925. Thecolorimetric determination of phosphorus. J.Biol. Chem. 66:375-400.

FossuM, K. 1963. Separation of hemolysin andegg yolk turbidity factor in cell-free extracts ofBacillus cereus. Acta Pathol. Microbiol. Scand.59:400-406.

GARCIA-BUN'UEL, L., AND V. M. GARCIA-BURUEL.1964. Enzymatic determination of free myo-inositol in human cerebrospinal fluid andplasma. J. Lab. Clin. Med. 64:461-468.

GERLACH, E., AND B. DEUTICKE. 1963. Eine ein-fache Methode zur Mikrobestimmung von Phos-

phat in der Papierchromatographie. Biochem.Z. 337:477-479.

GIER, J. DE, AND L. L. M. VAN DEENEN. 1961.Some lipid characteristics of red cell membranesof various animal species. Biochim. Biophys.Acta 49:286-296.

HANES, C. S., AND F. A. ISHERWOOD. 1949. Separa-tion of the phosphoric esters on the filter paperchromatogram. Nature 164:1107-1112.

HOOGHWINKEL, G. J. M., M. J. LEXMOND, ANDH. G. BUNGENBERG DE JONG. 1959. Spot testsfor phospholipids and their use in paper chro-matography. Koninkl. Ned. Akad. Wetenschap.Proe. Ser. B 62:222-235.

HOOGHWINKEL, G. J. M., AND H. P. G. A. VANNIEKERK. 1960. Quantitative aspects of thetricomplex staining procedure. The staining oflecithin spots applied on chromatographicpapers. Koninkl. Ned. Akad. Wetenschap. Proe.Ser. B 63:258-276.

KERR, S. E., AND G. A. KFOURY. 1962. Chroma-tographic separation and identification of myo-inositol phosphates. Arch. Biochem. Biophys.96:347-353.

KLIBANSKY, C., E. CONDREA, AND A. DE VRIES.1962. Changes in plasma phospholipids afterintravenous phosphatidase A injection in therabbit. Am. J. Physiol. 203:114-118.

KUSHNER, D. J., AND D. FELDMAN. 1958. Charac-terization of the bacterial enzyme 'thrombo-plastinase.' Biochim. Biophys. Acta 30:466-475.

LARNER, J. 1962. Inositol dehydrogenase fromAerobacter aerogenes, p. 326-328. In S. P. Colo-wick and N. 0. Kaplan [ed.], Methods in en-zymology, vol. 5. Academic Press, Inc., NewYork.

LARNER, J., W. T. JACKSON, D. J. GRAVES, ANDJ. R. STAMER. 1956. Inositol dehydrogenasefrom Aerobacter aerogenes. Arch. Biochem. Bio-phys. 60:352-363.

MAGASANIK, B. 1953. Enzymatic adaptation inthe metabolism of cyclitols in Aerobacter aer-ogenes. J. Biol. Chem. 205:1007-1018.

MARINETTI, G. V. 1962. Chromatographic separa-tion, identification, and analysis of phospha-tides. J. Lipid Res. 3:1-20.

OTTOLENGHI, A., S. GOLLUB, AND A. ULIN. 1961.Studies on phospholipase from Bacillus cereus.I. Separation of phospholipolytic and hemolysinactivities. Bacteriol. Proc., p. 171.

ROBINSON, D. S., P. M. HARRIS, AND J. C. F.POOLE. 1957. Further studies on ethanolaminephosphatide and blood coagulation: The effectof a Bacillus cereus phosphatidase on the cal-cium time of citrated rat plasma. Quart. J. Exptl.Physiol. 42:285-294.

SLEIN, M. W., AND G. F. LOGAN, JR. 1960. Mecha-nism of action of the toxin of Bacillus anthracis.I. Effect in vivo on some blood serum compo-nents. J. Bacteriol. 80:77-85.

SLEIN, M. W., AND G. F. LOGAN, JR. 1962. Mecha-nism of action of the toxin of Bacillus anthracis.

80 J. BACTERIOL.

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PHOSPHOLIPASES OF B. CEREUS

II. Alkaline phosphatasemia produced by cul-ture filtrates of various bacilli. J. Bacteriol.83:359-369.

SLEIN, M. W., AND G. F. LOGAN, JR. 1963. Partialpurification and properties of two phospho-lipases of Bacillus cereus. J. Bacteriol. 85:369-381.

STERNE, M. 1937. Variation in Bacillus anthracis.II. Some correlations between colony variationand pathogenicity in strains of Bacillus anthra-cis. Onderstepoort J. Vet. Sci. Animal Ind. 8:279-349.

WAYS, P., AND D. J. HANAHAN. 1964. Characteri-zation and quantification of red cell lipids innormal man. J. Lipid Res. 5:318-328.

WEISSBACH, A. 1958. The enzymatic determina-tion of myo-inositol. Biochim. Biophys. Acta27:608-611.

WINTROBE, M. M. 1956. Clinical hematology, p.384. Lea & Febiger, Philadelphia.

WOOD, T. 1961. A procedure for the analysis ofacid-soluble phosphorus compounds and relatedsubstances in muscle and other tissues. J.Chromatog. 6:142-154.

VOL. 90, 1965 81

on June 25, 2020 by guesthttp://jb.asm

.org/D

ownloaded from