METABOLISM OF PHOSPHOLIPIDES BY BACTERIAL ENZYMES* BY OSAMU HAYAISHIt# AND ARTHUR KORNBERGI (From the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda, Maryland) (Received for publication, July 6, 1953) The hydrolysis of lecithin and cephalin at each of the four ester linkages by enzymes derived from plant, animal, and microbial sources has been de- scribed in numerous reports and reviewed recently by Zeller (2). However, relatively little work has been done on the systematic metabolic break- down of the complex phospholipide molecule by a single cell type. Schmidt et al. (3) have demonstrated that the degradation of lecithin by pancreas and other tissues leads to the accumulation of glycerophosphorylcholine, but the mechanism of this conversion, as well as the nature of the further metabolism of this product, remains to be clarified. We have studied the metabolism of phospholipides by a strain of Serratia plymuthicum isolated from soil by the enrichment culture technique. The results derived from the use of resting cell suspensions and partially puri- fied enzyme preparatiuns obtained from these cells indicate the following sequence of reactions+ Lecithin I\ L-a-glycerophosphorylcholine + 47 Lysolecithin L-a-glycerophosphate + choline The formation of lysolecithin as an intermediate in the conversion of lecithin to GPC could not be established, owing to the excessive lysoleci- thinase activity of the lecithinase fractions. Cephalin appeared to be metabolized by a comparable mechanism in- volving glycerophosphorylethanolamine as an intermediate. * A preliminary report of this work was presented (1). t Special Research Fellow of the United States Public Wealth Service. 1 Present address, Department of Microbiology, Washington University School of Medicine, St. Louis, Missouri. 1 The following abbreviations are used throughout the paper : GPC, glycerophos- phorylcholine; GPE, glycerophosphorylethanolamine; GP, glycerophosphoric acid. 647 645 PHOSPHOLIPIDE METABOLISM BY BACTERIA Methods Materials-Egg yolk lecithin, used for enzyme studies, was purified ac- cording to Pangborn (4) and on analysis gave the following:2 Found. C 66.09, H 8.10, N 1.86, P 4.18, N:P 0.99 Pangborn. " 1.94, " 4.21, (' 1.01 The choline content was determined directly from the lecithin-iodine complex (see below), and as free choline after release by acid (1 N HC1, 1 hour at l00") or by enzymatic hydrolysis. In a sample containing 0.8 mg. of lecithin, 0.75 I.IM was determined as a lecithin-iodine complex, and 0.73 and 0.75 p~ of choline were determined after acid and enzymatic hy- drolyses, respectively. These values, which represent only about 75 per cent of the theoretical, indicate that some impurity of inorganic nature or a compound similar to choline was present. Where lecithin has been used in these investigations, the determinations have depended solely on the estimation of choline content. Lecithin emulsions were obtained by plac- ing a distilled water suspension in a Raytheon sonic oscillator, model 9 KC, for 10 to 15 minutes at approximately 20". Crmde egg lecithin (used as a growth substrate), crude animal lecithin, and cephalin were products of the Nutritional Biochemicals Corporation. Lysolecithin was isolated from rice (5) or prepared by enzymatic hydrolysis of egg lecithin (6). Sphingomyelin, cardiolipin, and dipalmitoleyl-La-lecithin were kindly sup- plied by Dr. H. E. Carter, Dr. H. D. Piersma, and Dr. D. J. Hanahan, re- spectively. L-a-Dimyristoyllecithin and L-cr-dimyristoylc~halin were ob- tained through the kindness of Dr. E. Baer. GPC was generously furnished by Dr. G. Schmidt, or prepared by his method (7). The choline content of the Schmidt preparation was 100 per cent of theory; that of our own was only 72 per cent, but the elementary analysis was approximately correct. The former sample was always used CeH22NP01-2CdClz. Calculated. C 14.97, H 3.46, N 2.18, P 4.83 " 15.06, " 3.33, " 2.21, " 4.50 Found. in crucial experiments. GPE was a gift from Dr. E. Baer. The barium salt of DL-CY-GP (8) was converted to the potassium salt by treatment with Dowex 50 (Kf) resin. The organic P content was 83 per cent of theory, and the periodate consumption was 90 per cent of theory. Diphenyl phosphate was a product of The Dow Chemical Company. Particulate GP dehydrogenase was prepared from rabbit muscle accord- ing to Green (9). This preparation did not oxidize either P-GP or D-CY-GP. 9 Microanalyses in this paper were performed by the Microanalytical Laboratory of the National Institutes of Health under the supervision of Dr. W. C. Alford. 0. HAYAISHT AND A. KORNBERG 649 Determinations-Choline was determined routinely by the method of Appleton et al. (10). This method depends on the formation of an iodine complex which is insoluble in water but which in ethylene dichloride yields a solution with strong light absorption in the ultraviolet region. The sam- ple was treated with an equal volume of cold 3 per cent perchloric acid and centrifuged in the cold. Choline-containing substances, such as lecithin and lysolecithin, are found in this precipitate. 0.5 ml. of the supernatant fluid, containing 0.05 to 0.5 PM of free choline, was treated with 0.2 mf. of iodine reagent (12.5 gm. of KI and 9.8 gm. of 12 dissolved in water to 250 ml.) and kept in an ice bath for 15 minutes. After centrifugation, the supernatant fluid was removed by aspiration, the precipitate dissolved in ethylene dichloride (reagent grade) to a volume of 10 ml., and the optical density determined at 365 mp in a Beckman model DU spectrophotometer. The molar extinction coefficient, referred to choline, is approximately 2.7 This method for choline determination was also applied to the measure- ment of lecithin and lysolecithin. The sample, at neutral pH, was treated directly with iodine reagent. The insoluble lecithin-iodine complex was collected by centrifugation, dissolved in ethylene dichloride, and measured in the spectrophotometer. The lysolecithin-iodine complex, which remains soluble at neutral pH, may be precipitated by acidifying the sclution. The extinction coefficients appeared to be the same as that of the choline com- plex; the value obtained was corrected when necessary for free choline, determined as described above on an aliquot sample. Choline was also determined in a few instances by the reineckate method of Glick (ll), which gave satisfactory agreement with the procedure de- scribed above. Lysolecithin was also determined by the red blood cell hemolysis test of Bernheimer (12). Quantitative results were obtained only when the amount of lysolecithin was adjusted within the narrow range of 0.15 to 0.25 p~; similar results have been reported by Collier (13). The hemolysis test applied to egg and rice lysolecithin preparations gave results in good agreement with those from the choline analyses. GPC, which is not determined in the choline estimation by the iodine method, was estimated in the perchloric acid supernatant fluid as free choline which appeared after acid (100' in 1 N HC1 for 20 minutes (14)) or enzymatic hydrolysis (see below). Both methods of hydrolysis released equivalent amounts of choline. Ethanolamine was determined by an unpublished method of Axelrod and Brodie.s To 2.0 ml. of sample (pH 7.0 to 8.0) containing 0.25 to 1.0 p~ 3 Personal communication. We should like to thank Dr. B. B. Brodie and Dr. J. Axelrod for making available to us their unpublished methods. x 104. 650 PHOSPHOLIPIDE METABOLISM BY BACTERIA of ethanolamine were added 0.5 ml. of 5 per cent NaHCOs solution and 0.1 ml. of dinitrofluorobenaene solution (0.1 ml. of dinitrofluorobenzene diluted to 2.0 ml. with 95 per cent ethanol). The mixture was incubated at about 80" for 1 hour in stoppered test-tubes. After the solution was cooled, 8 ml. of chloroform were added, and the mixture was vigorously shaken for 10 minutes in a mechanical shaker and centrifuged. The upper layer was removed by aspiration, and 5 ml. of the chloroform layer were transferred to a glass-stoppered container, to which were added 30 ml. of petroleum ether (b.p. 30-60") and 4 ml. of 6 N HCl. The mixture was shaken manu- ally for several minutes, and the optical density of the bottom aqueous layer was determined at 420 mp in the spectrophotometer. 1.0 PM of eth- anolamine gave a density reading of approximately 1.0 in a cell of 1 em. light path. Protein and phosphorus were determined by the methods of Lowry et al. (15) and Fiske and Subbarow (16), respectively. Periodate titrations were performed by Jackson's method (17) modified to a micro level. Isolation and Properties of Organim-The organism was isolated from soil by the enrichment culture technique on a medium containing 0.1 per cent GPC, 0.1 per cent Difco yeast extract, 0.15 per cent K2HP04, 0.05 per cent KHgPOd, and 0.02 per cent MgSOd.7H20 in distilled water. De- tails of the isolation technique were previously described (19). Four strains were isolated; one which produced red pigment and liberated free choline when grown in a lecithin-containing medium was used throughout this study. The organism is a small aerobic rod, gram-negative, non-motile, and non- acid fast. Under aerobic conditions, it produces a bright red pigment which is soluble in alcohol but only slightly soluble in water. It does not decompose alkylamines; it produces a small but definite amount of gas when grown on dextrose and for- acetylmethylcarbinol. From these observations, the isolated organism has been tentatively classified as a strain belonging to S. ptymuthicum according to Bergey's manual (20). Cultivation of Organism and Preparation of Crude Enzyme Fractions- Cells were grown in the above medium with the addition of lecithin (0.2 per cent) but lacking GPC. The lecithin was dissolved in ether, sterilized separately by filtration, and poured into the hot medium. Ether was re- moved by mechanical shaking while the whole medium was hot. The cells were cultivated in 20 liter glass carboys containing 10 liters of the medium at about 26" for 16 hours with constant, reciprocal shaking. The cells were harvested by centrifugation in a Sharples supercentrifuge, washed once with a 0.5 per cent NaC1-0.5 per cent KCl solution, and frozen. At - loo, cells retained their activity for at least 6 months. The yield of cells was a-GP was determined according to Burmaster (18). 0. HAYAISHI AND A. BORNBERG 65 1 approximately 4 gm. (wet weight) per liter of medium in a large scale cul- ture and 6 to 10 gm. per liter in smaller scale cultures. Cell-free extracts were prepared by the method of McIlwain (21). Washed cells were ground in a mortar with 4 times their weight of alumina (Alcoa A-301) at 0" for 5 minutes. The cell paste was extracted with 10 parts of glycylglycine buffer (pH 9.0, 0.02 M) ; the whole slurry was centri- fuged at 0" at about 6000 X g for 15 minutes. The supernatant fluid was a slightly turbid suspension of cell fragments and contained a negligible number of intact cells. This preparation, referred to as "bacterial ho- mogenate," contained 3.5 to 4.0 mg. of protein per ml. The homogenate was centrifuged in a Spinco centrifuge model L (rotor No. 40) at 110,OOO X g for 1 hour. The clear upper layer, representing about 90 per cent of the fluid, was removed with a syringe; the turbid fluid layer overlaying the residue was discarded. The residue was washed once with a volume of glycylglycine buffer equal to that of the original homoge- nate by centrifuging at 110,000 X g for 1 hour, and resuspended in the same volume of buffer. The clear supernatant fluid and the washed resi- due, referred to as Fractions S and R, contained approximately 2.2 and 1.3 mg. of protein per ml., respectively. Assay of Lecithinase-The assay was based on the rate of choline libera- tion from purified egg lecithin in the presence of Fraction R or S as a source of lysolecithinase (see below) and an added excess of purified GPC diester- ase (see below). The test system contained 1 PM of lecithin, 10 PM of po- tassium phosphate buffer (pH 7.0), and 0.1 ml. of crude Fraction S (when Fraction R was assayed), or 0.1 ml. of crude Fraction R (when Fraction S was assayed), 0.05 ml. of purified GPC diesterase (25 units), and water to a total volume of 0.4 ml. Incubation was carried out at 25" for 1 hour. Under these conditions the rate of reaction was proportional to the enzyme concentration, Assay of Lysolecithinase-The assay was based upon the rate of choline liberation from lysolecithin in the presence of an added excess of purified GPC diesterase. The test system contained 1 PM of lysolecithin, 10 PM of phosphate buffer (pH 7.0), 0.05 ml. of purified GPC diesterase (25 units), lysolecithinase; and water to a total volume of 0.4 ml. The incubation was carried out at 25" for 30 minutes. Assay of GPC Dieskrase-The test system contained 0.2 ml. of GPC (0.02 M, pH 7.0), 0.1 ml. of glycylglycine buffer (0.2 M, pH 8.85), and 0.1 ml. of enzyme. After 10 minutes at room temperature (25-27"), the reac- tion was stopped by the addition of 0.4 ml. of perchloric acid (6 per cent) and free choline was determined in a 0.2 ml. aliquot. A unit of enzyme was defined as the amount hydrolyzing 1 PM of GPC during a 1 hour inter- val under these conditions, and the specific activity was defined as units 652 PHOSPHOLlPIDE METABOLISM BY B.ICl'EHIA Broth medium 0.002 0.047 0.077 0.000 per mg. of protein. With about 1 unit of enzyme in this test system, the reaction was linear for more than 1 hour. With a crude, cell-free extract, proportionality between rate and the amount of enzyme was obtained be- tween 0 and 4 units, whereas with purified enzyme preparations propor- tionality was observed between 0 and 10 units; this suggests the presence of an inhibitor in crude extracts. As will be shown under "Results," sev- eral metal ions were found to be inhibitory; the inhibition at high enzyme concentration could, therefore, be due to metal impurities in the crude enzyme preparations. - Lecithin medium 0.033 0.515 0.176 0.000 TABLE I Adaptive Nature of Organism The incubation mixtures (at 26") contained 1 PM of each substrate, 20 PM of each buffer, and about 0.15 mg. of protein in a total volume of 0.4 mI. The pH and buffers were as follows: lecithin (acetate, pH 5.7), lysolecithin (acetate, pH 6.3), GPC (gly- cylglycine, pH 8.9), choline (phosphate, pH 7.0). Lecithin and choline were incu- bated for 1 hour; lysolecithin and GPC were incubated for 10 minutes. The values represent the micromoles of substrate decomposed calcuIated for a 10 minute incu- bation period and 1 mg. of protein. Substrates Lecithin. ...................... Lysolecithin. .................. GPC. ......................... Choline. ...................... Cell suspension I Cell homogenate Lecithin medium 0.005 0.310 0.105 0.000 Broth medium 0.003 0.038 0.034 O.OO0 -- Results Adaptive Nature of Organism-The activities toward lecithin, lysoleci- thin, GPC, and choline of cells grown on a lecithin-containing medium and on ordinary broth were compared with both resting cell suspensions and broken cell preparations. As shown in Table I, the values were 5 to 14 times greater with extracts from lecithin-grown cells; similar results were also observed with the resting cell suspensions. Lysolecithinase activity was by far the most potent of the activities tested. Choline was observed to accumulate in lecithin-containing culture media, and no choline-remov- ing activity was detectable with any of the preparations. The failure to utilize choline explains an observed requirement for yeast, extract as a source of nitrogen and perhaps other nutrients as well. Phosphorylcholine was not converted to choline by the intact or broken cell preparations; this suggests that this compound is not an intermediate in the formation of choline from lecithin. 0. HAYAWHI AND A. KORNBERQ 653 Metabolism of Lecithin and Lysolecithin Lecithin Breakdown-When purified egg lecithin was incubated with the bacterial homogenate, choline was liberated at a rapid rate, but no signifi- cant amounts of lysolecithin or GPC accumulated. With Fraction R or Fraction S alone, the removal of lecithin and the appearance of chcline were either absent or very slow, but, with Fractions R and S combined, as rapid a disappearance of lecithin and formation of choline were observed as with the original homogenate (Fig. 1). Heating of the bacterial homogenate (at 97" for 5 minutes at pH 9.0) did not significantly reduce its ability to remove lecithin, but decreased the METABOLISM OF LECITHIN 6 cn w -14 0 E a2 0 2 40 2 40 2 4 TIME, HOURS FIG. 1. Lecithin metaboIism. The incubation mixtures contained 0.6 ml. of lecithin (0.01 M), 0.3 ml. of 0.5 M acetate buffer, pH 5.6, and 0.1 ml. of Fraction S (1.3 mg. of protein) or 0.1 ml. of Fraction R (0.7 mg. of protein), as indicated, in a final volume of 2.4 ml. The incubation was at 26". L = lecithin; C = choline. No lysolecithin or GPC wns detected throughout the incubation period. liberaticjn of free choline and led to the accumulation of GPC (Table 11). The presence of Mg++, which inhibits the breakdown of GPC (see below), further depressed the formation of free choline by the heated preparation (Table 11). Attempts to demonstrate the formation of a lysolecithin as an inter- mediate in the conversion of lecithin to GPC did not succeed, presumably as a result of the relatively high activity of lysolecithinase. Unlike a lyso- lecithinase of animal origin (22), which was rather heat-labile as compared with the accompanying lecithinase, the bacterial enzymes were both heat- stable. Heating of the bacterial homogenate at 100" for 10 minutes at pH 7.0 caused only a 50 to 80 per cent inactivation of both lysolecithinase and lecithinase. Lysolecithin Breakdown-When lysolecithin was incubated with either Fraction S or Fractions R and S, a rapid release of free choline was observed 654 PHOSPHOLIPIDE METABOLISM BY BACTERIA - Choline. .......................... GPC. ............................. without any appreciable accumulation of GPC (Fig. 2). Incubation with Fraction R alone, however, resulted in only a slight liberation of free cho- line, although the rate of lysolecithin removal was undiminished. GPC accumulated instead of choline to an extent approximating the lysolecithin that had disappeared (Fig. 2). As in the case of lecithin breakdown, heat- fix fix fiM 0.575 0.128 0.032 0.000 0.457 0.506 TABLE 11 Accumulation of GPC As Intermediate in Lecithin Breakdown The incubation mixtures contained 1 of lecithin, 25 PM of acetate buffer (pH 5.7), and enzyme (0.6 mg. of protein) in a total volume of 0.4 ml. The incubation was for 2 hours at 25". ! Enzyme preparation Substrate Homogenate Heated homogenate H~t~d-,~!~~~te METABOLISM OF LYSOLECITHIN FR S FR. R +S - - - v 0 I 20 I 20 I 2 TIME, HOURS FIQ. 2. Lysolecithin metabolism. The conditions were the same as those de- scribed in Fig. l, except that 0.6 ml. of lysolecithin (0.01 M) was used instead of lecithin. LL = lysolecithin; C = choline. ing of Fraction S (at 100" for 5 minutes) also limited the reaction to the formation of GPC. Thus with 0.1 ml. of unheated Fraction S (incubated with 1 p~ of lysolecithin and 5 p~ of glycylglycine buffer, pH 8.5, in a vol- ume of 0.4 ml. for 2 hours at 26"), 0.442 and 0.018 p~ of choline and GPC respectively, were formed; with heated Fraction S, 0.025 and 0.461 p~ of choline and GPC, respectively, were produced. Specijkity of Reaction-Since both Fractions R and S were required to initiate 1ecit)hin degradation, but either Fraction 11 or S alone degraded 0. HAYAISHI AND A. KORNBERG 655 lysolecithin, it might be assumed that the liberation of an unsaturated fatty acid from lecithin was more complicated than the release of a saturated fatty acid. However, this assumption is not supported by the observation that dipalmitoleyllecithin, which has two unsaturated fatty acids, was readily attacked by either Fraction R or S alone; when both fractions were combined, the activity was merely additive (Table 111). Dimyristoylleci- thin was attacked by Fraction S but only slightly by Fraction R. Crude animal lecithin was metabolized either by Fraction S or R, al- though at a somewhat slower rate than was egg lecithin, and the combined activities of Fractions S and R were almost additive. TABLE I11 Specificity of Lecithinase The reaction mixtures contained 1 PM of each substrate, 5 ,UM of acetate buffer (pH 5.7), and 0.15 ml. of Fraction S or 0.15 ml. of Fraction R, as indicated, in a total volume of 0.4 ml. The incubation was carried out at 26" for 2 hours. The values represent the sum of free and acid-labile choline or ethanolamine and are expressed as micromoles. Enzyme preparation Substrate Fraction R __ I Lecithin (egg). .................. Dipalmitoleyllecithin ............ Dimyristoyllecithin .............. Lecithin (animal) ................ Dimyristoylcephalin. ............. Cephalin I' ................ Sphingomyelin ................... 0.021 0.502 0 * 001 0.275 Fraction S 0.120 0.296 0. IO4 a. 168 Fractions S + R 0.582 0.760 0.139 0.390 0.037 0.176 0.000 Ethanolamine was shown to be liberated from cephalin, although at a much slower rate when crude animal cephalin was used as a substrate. However, when dimyristoyllecithin and dimyristoylcep halin were com- pared, the amounts of choline and ethanolamine released were of about the same order of magnitude. Under the test conditions no free choline was released from phosphoryl- choline and sphingomyelin. Attempts to adapt this organism by growing it in the presence of sphingomyelin were unsuccessful; the growth was poor and there was no liberation of free choline into the medium. Cardiolipin was not metabolized by the whole bacterial homogenate. Cleavage of the diester linkages would have released phosphate monoester groups which, in the presence of added monoesterase (from human ern en),^ * Human semen was fractionated with ammonium sulfate, and a fraction collected between 0.5 and 0.8 saturation was used after dialysis. 656 PHOSPHOLIPIDE METABOLISM BY BACTERIA No Ca++ 0.083 0.044 0.003 should have resulted in the appearance of inorganic phosphate, but none was detectable. Puri$calzon and Melal Requirements of Lecithinase-Frac tion R was made soluble by butanol treatment according to Morton (23). To 5 ml. of the chilled suspension was added 0.5 ml. of n-butanol with constant mechanical stirring. After 5 minutes at O", the mixture was centrifuged in a Serval1 angle centrifuge (SS-1) at 13,000 r.p.m. for 30 minutes. AImost 100 per cent of the activity remained in the supernatant fluid if Ca++ was added to the incubation mixture; prior to butanol treatment, the activity was completely sedimented under these conditions. The supernatant fluid was dialyzed against 0.01 M KzHP04 at 3" for 4 to 6 hours (until the odor of butanol was no longer detectable). Excessive diaIysis led to inactiva- ~ ~~ 0.01 Y Caw 0.138 0.142 0.095 - TABLE IV Eflect of Ca++ on Lecithinase The assay was carried out as described under "Methods" except that GPC di- esterase was omitted. The protein content of Fraction R before butanol treatment was 1.7 mg. per ml.; after butanol treatment the preparation contained 0.62 mg. per ml., and after dialysis 0.52 mg. per ml. Steps Fraction R.. ............................... Butanol treatment.. ........................ Dialysis .................................... Choline liberation,' p~ * The values represent the sum of free and acid-labile choline. tion of the enzyme. This preparation contained 80 to 100 per cent of the original activity and only 25 to 30 per cent of the protein. (Additional purification (2- to 4-fold) was obtained by fractionation between ammo- nium sulfate saturations of 30 and 50 per cent; but this latter step was not used routinely because yields of less than 30 per cent were obtained when the removal of butanol in the previous step was incomplete.) An almost obligatory requirement for Ca+f was observed at this stage (Table IV). The saturation level was reached with 0.01 M Ca++, and inhibitory effects were observed at concentrations above 0.02 M. Mg++ and Mn++ had no demonstrable effect. Fraction S was purified by adsorption on calcium phosphate gel and elu- tion with 0.5 M dipotassium phosphate. After such treatment, the enzyme was observed to lose activity by passage through a Dowex 50 (K+) col- umn, while treatment with Dowex 1 (formate) resulted in little loss of activity, although most of the protein was removed. 0. HAYAISHI AND A. KORNBERQ 657 On the basis of these observations, various metal ions were tested. Few was found to be the most effective in substituting for Fraction S (Table V), while Few was inactive. The saturation level of Fe* was at about 0.01 M. Unlike the action of untreated Fraction S, which (in the presence of Fraction R) can lead to the complete degradation of the lecithin substrate, the action of either gei-treated Fraction S or Fe+f (in the presence of Frac- tion R) ceased after only 20 to 30 per cent of the substrate was removed, although the rate of the reaction is comparable to or even greater than the rate with Fraction S. TABLE V E$ect of Various Metal Ions on Lecithinase Action of Fraction R The test system contained 0.5 p~ of lecithin, 25 PM of glycylglycine buffer (pH 7.4), 5 PM of CaC12, 0.1 ml. of Fraction R, and 5 MM of metal ions, as indicated, in a total volume of 0.5 ml. The incubation was carried out for 1 hour at 38". With 0.1 ml. of Fraction S, 0.096 /IM of choline was liberated under the same conditions. Salt None MgCL MnClt CUClS FeNH4 (SOdt Choline liberated,* p~ 0.024 0.043 0.048 0.078 0.032 Salt Fe (NH4)a(S04) FeClz FeCls COClZ ZnCI2 Choline liberated,. p~ 0.140 0.145 0.017 0.046 0.036 * The values represent the sum of free and acid-labile choline. Metabolism of GPC Purijication of GPC Diesterase-To each 100 ml. of the crude cell-free extract (homogenate) were added 31.5 gm. of (NH&S04. After 30 min- utes at O", the precipitate was centrifuged and 30 gm. of (NH4)zS04 were added to the supernatant fluid. The precipitate was collected by centrifu- gation and dissolved in 50 ml. of glycylglycine buffer (0.02 M, pH 9.0) (Ammonium Sulfate I, Table VI). A 2-fold increase in activity at this stage was probably due to the removal of metallic inhibitors in the crude extract. To each 50 ml. of Ammonium Sulfate1 were added 18.2 gm. of (NH4)2S04. The precipitate was centrifuged and to the supernatant fluid were added 6.5 gm. of (NH4) &304. The precipitate, collected by centrifugation, was dissolved in 10 ml. of glycylglycine buffer (0.04 M, pH 9.0) (Ammonium Sulfate 11, Table VI). To 9 ml. of Ammonium Sulfate I1 in a 0" bath were added 1 ml. of 1 M sodium acetate and then 6.5 ml. of acetone (- 10") dropwise with mechani- cal stirring. After standing for 5 minutes, the precipitate was removed by 658 PHOSPHOLIPIDE METABOLISM BY BACTERIA Fraction Crude extract .............................. Ammonium Sulfate I.. ..................... 11. ...................... Acetone .................................... (I (I centrifugation, and 4 ml. of acetone were added to the supernatant fluid. After 5 minutes at O", the precipitate was collected by centrifugation and dissolved in 4 ml. of 0.01 M KZHP04 (acetone, Table VI). Stabiliiy of Enzyme-The purified preparations were stored at - 10" for at least 6 months without any appreciable loss of activity. Heating the Total activity units 4,500 10,200 5,000 3,850 Choline Periodate Hrs. titration Iodine complex Reineckate 0 0 0 97.0 2 101 93 97.0 Specific activity units per mg. protein 24 437 1135 5600 __ - ___ a-Glycerw L-u-G~ cero- phosphate* phospiatet __ 93.6 0 94 .O 105 A I +lo1 1 4-93 1 0.0 1 +0.4 I +I05 * Burmaster method which involves the estimation of acid-labile phosphate (1 t Assay with IA-a-glycerophosphate dehydrogenase. --__I_ __ -. ... hour in 1 N HC1 at 100") after periodate oxidation. enzyme preparations at pH 9.0 for 5 minutes at 50°, 75", or 100" destroyed 15, 95, or 100 per cent of the activity, respectively. Products of GPC Breakdown-The enzymatic breakdown of GPC re- sulted in the production of L-a-glycerophosphate and choline in equimolar amounts equivalent to the GPC metabolized (Table VII). Choline deter- mined by two different methods yielded similar results. The values ob- tained by periodate titration and Burmaster's method, in which a-GP and GPC cannot be differentiated, showed no change. gpecific enzymatic as- say of L-~-GP indicated reasonable agreement with the theoretical value. 0. EIAYAlSHl ,4ND A. KORNBERG 659 It is significant that the enzymatic hydrolysis of lecithin, unlike alkaline hydrolysis, yields only the CY form of GP. Substrate Specificity of GPC Diesterase-GPE was the only substrate be- sides GPC among a group of choline-containing phosphate diesters which was split by the enzyme preparation at a rate comparable with the splitting of GPC. Under test conditions, which resulted in the release of 1 ~LM of choline from GPC in 60 minutes, there was no detectable choline release from purified egg lecithin, lysolecithin, animal lecithin, sphingomyelin, ---I - -- 6 .O t // OQ I I t I I I I I I I' 1 I 3 5 7 9 I/S FIG. 3. Influence of substrate concentration (8) on rate of reaction (v), with and without competitive inhibitors. The incubation mixtures contained 0.1 ml. of 0.2 M glycylglycine buffer, pH 8.9, 0.1 ml. of GPC diesterase (ammonium sulfate Fraction 11, 12.4 y of protein), 2 /.bM of inhibitors and substrates, as indicated, in a total volume of 0.4 ml. The incubation was carried out at 26" for 10 minutes. dipalmitoleyllecithin, or dimyristoyllecithin in a 120 minute incubation period. Non-choline-containing phosphate diesters, except GPE, did not appear to be attacked. Diphenyl phosphate released neither phenol nor phos- phate monoester groups (determined by subsequent treatment with human semen phosphatase) in the presence or absence of Mg++ (0.01 M) at pH 8.9. Diphenyl phosphate (0.005 M) inhibited GPC splitting by 15 per cent, Splitting of yeast ribonucleic acid and of polynucleotides resistant to ribonuclease action was tested by the method of Kunitz (24).5 Less than 1 per cent of the rate of GPC splitting was found with either substrate in the presence or absence of Mg++ (0.03 M). 6 These tests were carried out in collaboration with Dr. 1,. A. Heppel. 660 PHOSPHOLIPIDE METABOLISM BY BACTERIA A slight contamination with monoesterase activity was present in the purified preparations. Only trace amounts of inorganic phosphate were released from a-GP (at the rate of 0.2 per cent of the rate of GPC splitting). Adenosine-5-phosphate was hydrolyzed at about 3 per cent of the rate of GPC splitting; however, this activity, in contrast to that of GPC diesterase (see below), was not inhibited by Mg++ (0.01 M) but rather increased 2-fold. Influence of Substrate Concentration; Identity of GPC and GPE Diesterase -Lineweaver-Burk plots (25) of the hydrolysis rates at varying GPC and GPE concentrations are shown in Fig. 3. From these curves Michaelis 6.0 8.0 10.0 z o 4.0 PH FIQ. 4. Rate of enzymatic hydrolysis of GPC and GPE as a function of pH. The standard assay conditions were employed with 0.05 ml. of ammonium sulfate Frac- tion I1 containing 6.22 y of protein. The buffer8 employed (0.05 M) were acetate (pH 4.2), phosphate (pH 6.2), glycylglycine (pH 7.7, 8.2, 8.9) and glycine (pH 9.2, 9.8, 10.5). constants of 1.2 X 10-3 and 2.5 X lW3 M were calculated for GPE and GPC, respectively. GPE and GPC were observed to be competitive in- hibitors of GPC and GPE splitting, respectively. The dissociation con- stants of the inhibitor-enzyme complex6 were calculated to be 2.4 X M for GPE and 1.8 X loA3 M for GPC. These results are taken as an indi- cation that a single enzyme is responsible for the hydrolysis of GPC and GPE. Influence of pH and Metal Ions-Maximal activity for the splitting of 6 KI = (&(I/& + S))/((LJ - vI)uI). Ks, dissociation constant of enzyme-sub- strate complex; KI, dissociation constant of enzyme-inhibitor complex; 8, concen- tration of substrate; I, concentration of inhibitor; v, velocity; VI, velocity in the presence of inhibitor. 0. HAYAISHI AND A. KORNBERG 661 both GPC and GPE was observed in the range pH 8 to 9 (Fig. 4). At a concentration of 0.001 M, Mg++ and Mnft inhibited GPC splitting by more than 90 per dent and Zn++ inhibited the rate by about 60 per cent. Ethyl- enediaminetetraacetate (0.0001 M) inhibited GPC splitting by 91 per cent, but this inhibition could not be overcome by any metals tested nor did prolonged dialysis against water reduce the activity of the enzyme. DISCUSSION These results demonstrate that GPC is an intermediate in the release of choline from lecithin and lysolecithin. While proof is lacking that lyso- lecithin is the intermediate in the conversion of lecithin to GPC, this is strongly suggested by the 14-f old greater lysolecithinase activity of leci- thin-grown cells than of broth-grown cells. The very high content of lyso- lecithinase activity in these extracts relative to their lecithinase content is adequate to explain the failure to accumulate lysolecithin. Monsour and Colmer (26) recently reported the formation of choline from egg yolk by a number of strains belonging to the genus Xerratia. They attributed the reaction to a lecithinase C activity by which, as demonstrated with certain plant tissues (27), choline is split directly from lecithin to yield phospha- tidic acid as the other product. Our study of lecithin metabolism in S. plymuthicum suggests another interpretation of Monsour and Colmer's findings. From an investigation of the mechanism of the lecithinase action of pan- creas preparations, Shapiro (28) concluded that the primary reaction wa,s a transfer rather than hydrolysis of a fatty acid ester 1inka.ge. While the over-all result observed in our bacterial system is the hydrolysis of both fatty acid ester bonds, it has not been excluded that a transfer reaction involving an unstable intermediate is involved. Further purification of the bacterial lecithinase is needed to resolve this question. It has generally been assumed that the action of lecithinase A is to split only the ester linkage with the unsaturated fatty acid (2). From the pres- ent investigation, it is apparent that a fully saturated lecithin (dimyristoyl- lecithin) and an unsaturated one (dipalmitoleyllecithin) are both attacked at comparable rates. Although it cannot be inferred from these results that a single enzyme is responsible for these several activities, it is never- theless clear that such activities do exist. With the availability of these highly purified and well characterized lecithins, it should become possible to establish whether the a- or 8-fatty acid ester linkage is the primary and exclusive site of attack. Belief in the existence of naturally occurring @-lecithin (which yields @-glycerophosphate on chemical hydrolysis) has been virtually abandoned as the result of the work of a number of investigat.ors. Particularly note- 662 PHOSPHOLIPIDE METABOLISM BY BACTERIA worthy are the recent contributions of Baer and Kates (29) demonstrating that the synthetic a-lecithins also yield &glycerophosphate under the con- ditions employed for the hydrolysis of naturally occurring lecithin. The present work adds further confirmation by specifically identifying ~-a- glycerophosphate as the exclusive product of the bacterial metabolism of egg lecithin. SUMMARY 1. The metabolism of certain phospholipides has been studied with en- zyme preparations from a strain of Serralia plymuthicum. This organism was isolated by enrichment culture on a medium containing glycerophos- phorylcholine (GPC) and was routinely grown on a lecithin-yeast extract medium. 2. Extracts from cells cultured on lecithin released free choline from lecithin, lysolecithin, and GPC at a rate 5 or more times greater than that obtained with extracts from broth-grown cells. 3. Heat-treated extracts split lecithin and lysolecithin with the accumu- lation of GPC. Fractionation of the extract by high speed centrifugation yielded residue (R) and supernatant (S) fractions, both of which were essen- tial for the splitting of lecithin. Fraction R was made soluble by butanol treatment, partially purified, and shown to require Ca*. Fraction S was replaced by Fe++ to a limited extent. Lysolecithinase activity was present in both Fractions R and S. The action of these enzyme preparations QTI saturated and unsaturated lecithin, cephalin, and other related compounds has been studied. 4. An enzyme hydrolyzing GPC to L-a-glycerophosphate and choline, named GPC diesterase, has been partially purified. Kinetic data indicate that glycerophosphorylethanolamine (GPE) is also a substrate for this enzyme. 5. These results suggest the following pathway for the bacterial metabo- lism of lecithin (or cephalin): Lecithin and lysolecithin are converted to GPC, which is then split into L-cx-GP plus choline. Cephalin is metabo- lized to L-~-GP and ethanolamine by way of GPE. BIBLIOGRAPHY 1. Hayaishi, O., Federation Proc., 12, 216 (1953). 2. Zeller, E. A., in Sumner, J. B., and Myrback, K., The enzymes, 1, pt. 2,986 (1951). 3. Schmidt, G., Hecht, L., Fallot, P., Greenbaum, L., and Thannhauser, S. J., J. 4. Pangborn, M. C., J. Biol. Chem., 188, 471 (1951). 5. Iwata, M., Biochem. Z., 224, 430 (1930). 6. Fairbairn, D., J. Biol. Chem., 167, 633 (1945). 7. Schmidt, G., Hershman, H., and Thannhauser, S. .J., J. RioE. (`hem., 161, 523 Biol. Chem., 197, 601 (1952). (1945). 0. HAYAISHI AND A. KORNBERG 663 8. BailIy, M. C., Compt. rend. Acad., 206, 1902 (1938). 9. Green, D. E., Biochem. J., 30, 629 (1936). 10. Appleton, H. D., Levy, B. B., Steele, J. M., and Brodie, B. B., Federation Proc., 11. Glick, D., J. Biol. Chem., 166, 643 (1944). 12. Bernheimer, A. W., J. Gen. Physiol., 30, 337 (1947). 13. Collier, H. B., J. Gen. Yhysiol., 36, 617 (1952). 14. Schmidt, G., Hecht, L., Strickler, N., and Thannhauser, S. J., Federation Proc., 15. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., J. Biol. Chem., 16. Fiske, C. H., and Subbarow, Y., J. Biol. Chem., 66, 375 (1925). 17. Jackson, E. L., in Adam, R., Organic reactions, New York, 2, 341 (1944). 18. Burmaster, C. F., J. Biol. Chem., 164, 233 (1946). 19. Hayaishi, O., and Kornberg, A., J. Biol. Chem., 197, 717 (1952). 20. Breed, R. S., Murray, E. G. D., and Hitchens, P., Bergey's manual of determina- 21. McIlwain, H., J. Gen. Microbiol., 2, 288 (1948). 22. Noguchi, S., J. Biochem., Japan, 36, 113 (1944). 23. Morton, R. K., Nature, 166, 1092 (1950). 24. Kunitz, M., J. Gen. Physiol., 24, 15 (1940). 25. Lineweaver, H., and Burk, D., J. An2. Chem. SOC., 66, 658 (1934). 26. Monsour, V., and Colmer, A. R., J. Bact., 63, 597 (1952). 27. Hanahan, D. J., and Chaikoff, I. I,., J. Biol. Chem., 169, 699 (1947); 172, 191 28. Shapiro, B., Nature, 169, 29 (1952). 29. Baer, E., and Kates, M., J. Biol. Chem., 175, 79 (1948); 186, 615 (1950). 9, 146 (1950). 8, 249 (1949). 193, 265 (1951). tive bacteriology, Baltimore, 6th edition (1948). (1948).