Phospholipid transmethylation in the membrane of human neutrophils and lymphocytes

ARCHIVESOFBIOCHEMISTRYAND BIOPHYSICS Vol. 234, No. 1, October, pp. 7-14, 1984

Phospholipid

Transmethylation in the Membrane Neutrophils and Lymphocytes’

YUKIE NIWA,*” TSUYOSHI SAKANE,* AND SHINKICHI

of Human TANIGUCHIt

*Department of Internal Medicine, Shim.ane Medical University, and ~Department of Phnrmudogg, Kgoto University, Japan Received February 21,1984, and in revised form June 8, 1984

Phospholipid transmethylation in the microsomal fraction of stimulated and unstimulated human leukocytes was measured in a recently developed assay system. Microsomal fraction was prepared from neutrophils, unseparated lymphocytes, T lymphocytes, and non-T lymphocytes by sonication and subsequent ultracentrifugation. Two hundred micrograms of microsomal protein was reacted with Sadenosyl-L[methyZ-3H]methionine. In unstimulated cells, incorporation of m&hyZ-3H into phospholipid was 0.60 f 0.06 pmol min-’ mg protein in neutrophil membrane, 0.34 + 0.075 in unseparated lymphocytes, 1.23 +. 0.17 in T lymphocytes, and 0.71 f 0.035 in non-T lymphocytes (mean + SE). Stimulation of neutrophils with opsonized zymosan or concanavalin A (Con A), and of lymphocytes with Con A, phytohemagglutinin, or pokeweed mitogen increased 15 to 30%. The resulting methylated phospholipids were identified and quantitated by two-dimensional thin-layer chromatography. The inhibitor 5’-S-isobutyl-5’-deoxyadenosine (SIBA) inhibited transmethylation 4’7-55s. This assay system appears to measure specifically the activity of methyltransferases which mediate the transmethylation of membrane phospholipid; the assay should find important applications in the study of membrane lipid metabolism in human health and disease. 0 1994 AcndemicPress,Inc. It has been proposed that the transduction of receptor-mediated signals through the membranes of various types of cells is induced mainly by the enzymatic transmethylation of phospholipids; as a result of this transmethylation, translocation of phospholipids from the cytoplasmic side to the outer side of the membrane occurs, leading to reduced membrane viscosity and enhanced membrane fluidity (l-9). Among signals thought to be transduced in this way are the chemotactic stimulation of neutrophils (1, lo-12), mitogenesis of lymphocytes (1, 13, 14), and histamine release in basophils and mast cells (1, 4, 15, 16).

In view of these concepts, it is possible that the assay of phospholipid transmethylation in the membrane of neutrophils and lymphocytes might reflect biochemical or immunopathological changes at the sites of inflammation in patients with inflammatory disorders. To date, however, there have been no reported studies of phospholipid methylation in the membranes of neutrophils or lymphocytes in any human disorder; furthermore, a reported assay of Con As-induced trans’ Abbreviations used: EGTA, ethylene glycol bis(j3aminoethyl ether)-N,N,N’,N’-tetraacetic acid, LPC, lysophosphatidylcholine; PC, phosphatidylcholine; PDME, phosphatidyldimethylethanolamine; PE, phosphatidylethanolamine; LPE, lysophosphatidylethanolamine; PHA, phytohemagglutinin; PI, phosphatidylinositol; PMME, phosphatidylmonomethylethanolamine; PS, phosphatidylserine; PWM, pokeweed mitogen; SIBA: 5’-S-isobutyl-5’deoxyadenosine; [methyl-3H]SAM, S-adenosyl-L-[mdhyZ-aH]methionine; Con A, concanavalin A.

1Supported in part by grant of Behcet’s Disease Research Committee of Japan, Ministry of Welfare. ‘To whom correspondence should be addressed, at Niwa Institute for Immunology, 4-4 Asahimachi, Tosashimizu, Koehi-ken 787-03, Japan. 7

0003-9861/84 $3.00 Copyright All rinhts

Q 1984 by Academic Press, Inc. of rewoduction in any form reserved.

NIWA, SAKANE,

8

methylation in the lymphocyte membrane (1, 13, 14) has been the object of criticism (17, 18). In the present study, using microsomal fraction, we have developed a clinically applicable assay method to measure methyltransferase activity in the membranes of human neutrophils, unseparated lymphocytes, T lymphocytes, and non-T lymphocytes. The specificity of our assay was confirmed by analysis of the methylated products by thin-layer chromatography (TIC), and by addition of a methyltransferase inhibitor, and our assay of phospholipid transmethylation is considered to be able to bear the criticism, supporting the previous investigations on lymphocyte transmethylation reported by others (1, 13, 14). Potential applications of this assay in the study of human disease are discussed. MATERIALS

AND METHODS

Harvest of microsomul fractions. Heparinized venous blood was obtained from healthy volunteers; either neutrophils or lymphocytes were separated from each sample by Ficoll-Hypaque centrifugation. In some experiments, lymphocytes were separated into T and non-T cell populations by the sheep erythrocyte rosetting technique (19). Cells were suspended in 0.25 M sucrose, and disrupted by sonication at 24 W for 10 s on ice (with a Bronson sonifier cell disruptor 200). The crude eonicates were centrifuged at 14,000~for 10 min at 4’C to remove mitochondria, nuclei, and debris. The microsomal fraction was then recovered from the supernatant by centrifugation at 104,OOOgfor 60 min at 4’C. Preliminary experiments demonstrated these conditions to afford the greatest yield of microsomes, approximately 400500 pg protein from 50 ml venous blood (3-5 x 10’ cells). Phxxptiipid tmnsmethylation way. Since 5’-nucleotidase activity was found to be high in our microsomal fraction, a considerably greater amount of plasma membrane was considered to be contained in the microsomes. Therefore, we refer to crude membrane fractions used in this study as the “microsomal” fraction. The standard assay mixture contained 100 mrd Trie-HCI (pH 8.0), 0.1 rnrd EGTA, 50 PM S-adenosyl-L-[methyZ-8H]methionine (from Amersham, Englrmd; 2 &i), and microaomal fraction containing approximately 200 pg protein in a final volume of 200 pl. Since various effects of calcium ion on phospholipid transmethylation have been discussed (20), EGTA was included for chelating

TANIGUCHI that ion. After incubation at 37°C for 30 min, the reaction was stopped by adding 0.6 ml 0.25 N HCI. In order to extract phospholipids from the incubation mixture, 3 ml chloroform/methanol (l/2) was added and vortexed, and then 1 ml 1% KC1 and 1 ml chloroform were further added and vortexed according to the method described by Bligh and Dyer (21). After the two water-soluble layers were discarded, 2 ml 0.5% KC1 in 50% methanol was added, and the mixture was vortexed and centrifuged. A l-ml volume of the chloroform phase was removed and transferred into a counting vial; the radioactivity was measured after drying at 70-80°C in an oven, and the addition of universal gel (Nakarai Chemicals, Kyoto, Japan) or aqua gel liquid scintillation cocktail. meth#H Incorporation was assessed in a liquid scintillation spectrometer (Packard Tri-Carb), and expressed as the picomoles of [methyl-aH]SAM incorporated into the phospholipids per minute per milligram protein. Control tubes, lacking only the microsomal fraction, were treated identically throughout the assay. It was confirmed that methylation occurred proportionally to the enzyme concentration of less than at least 500 gg and linearly for at least 30 min, when the substrate concentration was more than 20 pmol. To ascertain whether the present phospholipid transmethylation assay specifically represents the actual changes of membrane phospholipids of the cells, experiments were conducted in which the methyltransferase inhibitor SIBA (10 pM) was preincubated with cells for 60 min before sonication. Transmethylation assay in stimulated cell Cell activation experiments were conducted to determine the conditions for producing maximal phospholipid transmethylation by stimulation with various stimulants before sonication. Neutrophils were incubated for 10 min (13) with varying concentrations of opsonized zymosan or Con A, and lymphocytes with varying concentrations of Con A, PHA, or PWM. Furthermore, in order to examine the time-course effect of these stimulants on tranemethylation, neutrophils were incubated for varying times with 0.2 mg/ml opsonized zymosan or 2 pg/ml Con A, and lymphocytes with 2 pg/ml Con A, 0.2 pg/ml PHA, or 0.4 pg/ml PWM. Analysis of the readion prcducta A 1.4-ml volume of washed chloroform layer was evaporated to dryness under a stream of NZ. Samples were resuspended in chloroform/methanol (2/l) containing 50 pg each of authentic phospholipids as carriers, and were analyzed on silica gel plates by two-dimensional TLC (Kieselgel 60 plate; 20 x 20 cm, 0.25 mm thickness; E. Merck), with chloroform/acetone/ methanol/acetic acid/water (5/2/1/l/0.5, V/V) 88 the solvent in the first dimension and n-propyl alcohol/ ProPionic acid/chloroform/water (3/2/2/l) in the second (22, 23). The location of each phospholipid was ascertained by staining with iodine vapor. The

PHOSPHOLIPID

METHYLATION

spotsof methylated products corresponding to individual phospholipids were scraped off and transferred into counting vials for measurement of radioactivity after the addition of universal gel. The individually identified radioactive methylated phospholipid was expressed as a percentage of the total radioactivity (cpm) f SE recovered from the plate. No significant radioactivity was observed in each portion of individual phospholipid when [methyE ‘HISAM was chromatographed without incubation with membrane. In advance, the impurities of labeled SAM were examined by ascending TLC in of n-butanol/acetic acid/water (12/3/5), and only the labeled SAM with greater than 97% purity was used in the present experiment.

OF HUMAN

LEUKOCYTES

9

Con A were optimal concentrations for stimulating neutrophil membranes, and

that 2 pg/ml Con A, 0.2 pg/ml PHA, and 0.4 pg/ml PWM were optimal for stimulating the lymphocyte membranes. These values of optimal concentration of each stimulant were l/5-2/5 of the concentrations usually used to stimulate oxygen radical generation or lysosomal enzyme release by neutrophils, or the mitogenic responses of lymphocytes (24-27). Stimulants at these concentrations enhanced transmethylation activity by 15-30s (Figs. 1 and 2). As for a time-course effect of these stimulants on phospholipid methylation, RESULTS the methyltransferase activity in human Meth&ransferase activity assay. In un- neutrophil or lymphocyte membranes was stimulated cells, [methyGaH]SAM incor- stimulated within 2 min after the addition poration was 0.60 + 0.06 pmol min-’ mg of the above described stimulants. Maxiprotein-’ in neutrophil membrane, 0.34 mal stimulation was obtained by lo-min + 0.075 in lymphocytes, 1.23 + 0.17 in T preincubation with these stimulants, and lymphocytes, and 0.71 f 0.085 in non T a decrease in methylation was observed lymphocytes (Table I). with further incubation (Figs. 3 and 4). When neutrophils and lymphocytes were The preincubation of the cells with the stimulated with various concentrations of methyltransferase inhibitor, SIBA (10 FM), stimulants for 10 min before addition of for 60 min inhibited the incorporation of [methal-‘HISAM, it was found that 0.2 the methyl-‘H group into cell lipids by mg/ml opsonized zymosan and 2 fig/ml 47-55s (Table I). TABLE I METHYLTRANSFERASE ACTIVITIES OF NEUTROPHILS, LYMPHOCYTES, AND LYMPHOCYTE SUBSETS WITH OR WITHOUT STIMULATION

Stimulated with

Unstimulated

Opsonized zymosan (0.2 mg/ml)

Neutrophils

0.60 f 0.060 (0.31 + 0.035)

0.81 * 0.078 (0.40 + 0.029)

Lymphocytes

0.84 +- 0.075 (0.40 + 0.041)

T Lymphocytes non-T Lymphocytes

PHA (0.2 pg/ml)

PWM (0.4 pg/ml)

0.78 f 0.074 (0.37 + 0.030)

n.d.

n.d.

n.d.

1.20 + 0.10 (0.63 + 0.05)

1.16 + 0.10 (0.55 + 0.041)

1.13 * 0.087 (0.60 + 0.042)

1.43 f 0.17 (0.20 2 0.020)

n.d.

n.d.

n.d.

n.d.

0.71 + 0.085 (0.37 + 0.034)

n.d.

n.d.

n.d.

n.d.

ConA

(2 k&ml)

. -1 of [methyl-%]sAM incorporated into phospholipids. Note Each value is expressed as pmol min’ mg protem n.d., not done. Parentheses denote the values preincubated with methyltransferase inhibitor, 10 pM SIBA, for 60 min. Cells were preincubated with each stimulator for 10 min.

10

NIWA, SAKANE,

I

1 0

0.05 0.5

0.1 1

0.2 2 concentration

TANIGUCHI

0.4 4 of

1 10

2 w/ml 20 pg/n1

opsonized con

zymosan

A

stimulants

FIG. 1. Effect of various concentrations of stimulants on phospholipid transmethylation of human neutrophil membranes. Each plot presented in the figure was composed of mean values of 12 experiments, and the ranges of each standard error were always within one-thirteenth of mean value.

Analysis of methylated phosphdipids. Two-dimensional thin-layer chromatography (Fig. 5) demonstrated that most of the methyl-‘H incorporated into phospholipid was recovered as phosphatidyl-

choline (PC) (Table II). Modest amounts of phosphatidyldimethylethanolamine (PDME), lysophosphatidylcholine (LPC), and phosphatidylmonomethylethanolamine (PMME) were labeled, while only

0.6

0.5

1

L

I 0

0.5

1

0.05 0.1

0.1 0.2

2 0.2 0.4

Concentration

4 0.4 0.8

10 * 2

20 jIgha1 2 pgh1 4 pgh1

Con A PKA PW

Of stinu1ants

FIG. 2. Effect of various concentrations of stimulanta on phospholipid transmethylation human lymphocyte membrane. See Fig. 1.

of

PHOSPHOLIPID

METHYLATION

OF HUMAN

LEUKOCYTES

11

considerably by [meth@H]SAM. This incorporation was inhibited by the addition of the methyltransferase inhibitor SIBA, and the actual presence of methylated products was demonstrated by two-dimensional TLC. These findings indicate strongly that our assay specifically measures the activity of methyltransferases that mediates the translocation of membrane phospholipids, representing the dynamic nature of the cell membrane. Furthermore, in this study labeled SAM was used after purification, as indicated under Materials and Methods, and it is well known that highly purified blood cell populations can be obtained by the separation method used in our study. These findings FIG. 3. Effect of time-course effect of the stimulants on phospholipid transmethylation of human neutroseem to bear up against most of the raised phi1 membrane. criticism on phospholipid transmethylation (17, 18). As for the reports on stimulated lymnegligible amounts of label were recovered as phosphatidylethanolamine (PE). The phocytes, Hirata et al (13,14) have already data displayed in Table II represent the demonstrated the methyltransferase activity in the membrane of live cells withfindings from unstimulated neutrophils from all subjects tested. Similar results out sonication, while the authors assessed for lymphocytes and stimulated neutro- the enzyme activity of the cells after phils were obtained (data not shown). sonication, which required far smaller LysoPE (LPE), phosphatidylinositol (PI), amount of radioactive isotope. Our results, and phosphatidylserine (PS) were not obtained by stimulation of the cells, products of methylation, as was PE. showed behavior similar to theirs alTherefore, the amount of label recovered though our levels of stimulated methylas LPE, PI, or PS was also negligible, as ation were lower than theirs (13), which confirmed with PE (data not shown in may be due to our use of microsomal Table II except for PE), while Table II shows that 2.6 to 19 times the control level of label was recovered as each of the other methylated phospholipids. This finding supports the validity of the assay as a measurement of methyltransferase activity mediating the conversion of membrane phospholipid from PE to PC and LPC. The cpm counted in the control was 17, while that in the PE portion was only 25, which indicates that the cpm counted in PE fractions was due to a small amount of contamination by impure labeled SAM, and was negligible. DISCUSSION In the present study, the microsomal phospholipids of human neutrophils and lymphocytes were shown to be labeled

FIG. 4. Effect of time-course effect of the stimulants on phospholipid transmethylation of human lymphocyte membrane.

12

NIWA, SAKANE,

TANIGUCHI

FIG. 5. Thin-layer chromatographic separation of human neutrophilic lipid extract after exposure to iodine. Samples of 50 pg each of authentic phospholipids were chromatographed on a silica gel plate with extracted phospholipid from human neutrophils, as described under Materials and Methods. OR, origin; 1, LPC; 2, LPE; 3, PC, 4, PDME, 5, PMME, 6, PE; 7, PI; 8, PS.

fractions instead of whole cells and to the difference of the species used. Regarding the time-course effect of these stimulants TABLE II DISTRIBUTION OF RADIOACTIVE METHYLATED PHOSPHOLIPIDS IN TWO-DIMENSIONAL THIN-LAYER CHROMATOGRAMS Net cpm” LPC PC PDME PMME PE

33.5 + 298 + 65 f 23.5 f 8 k

3.7 31.1 7.2 2.1 0.7

438 (= 523 - 17 x 5)

Percentage distribution’ 38.5 X 109 = 8.8 zk 0.6 E > .68 2 5.8 14.9 + 1.2 6.5 k 0.5 1.8 + 0.1 100%

Note. Control (grass vial containing universal gel alone) cpm = 17 f 0.23. “The net cpm of reaction products is the difference in cpm between gross cpm and control cpm. bExpressed as a percentage of the total radioactivity (cpm) f SE in scraped-off methylated groups in the membrane of human neutrophils.

on phospholipid transmethylation, maximal stimulation was obtained by a lomin preincubation, which is consistent with Hirata’s study (13). And, after a lomin incubation, a decrease in methylation was also observed in our study, indicating that further metabolism followed although we did not simultaneously assess the release of arachidonate incorporated into phospholipid to investigate the correlation of the subsequent changes between methylation and arachidonate release (13). Except for Hirata’s investigations (13, 14), most studies of membrane transmethylation of lymphocytes and neutrophils have been carried out in rodent spleen cell preparations. Several modifications were necessary to adapt these methods to the study of human peripheral blood leukocytes; these modifications were necessitated by the relatively low methyltransferase activity in human leukocytes, and by the limited amount of blood

PHOSPHOLIPID

METHYLATION

that can ethically be drawn from any one individual subject. Successful methyltransferase assay in human leukocytes was dependent upon (i) sonication at 24 W for 10 s to obtain the greatest yield of the microsomal fraction; and (ii) use of 170-200 pg microsomal protein for each assay (as opposed to the 20-50 pg used in animal experiments). It seems that the necessary first step for the cellular responses involves the binding of various stimulants to the receptors on the surface membrane of the cells. Within a minute after the binding of stimulants, it has been shown that recognizable alterations in lymphocyte surface membrane will occur. These biochemical changes in the membrane include increased fluxes of ionic potassium (28) and calcium (29), and increased uptake of nucleotides (30), sugars (31), and amino acids (32). In addition, Con A (33) has been reported to increase the fluidity of the lymphocyte membrane after binding, and phosphatidylinositol turnover was markedly stimulated (34). It may be considered that phospholipid transmethylation can play a triggering role in these biochemical changes in the leukocyte membrane. However, it still remains to be elucidated how the binding of Con A to the membrane receptor actually correlates to transmethylation. The physiological function of neutrophils or lymphocytes is considered to be stimulated in the inflammatory process and, therefore, increased methyltransferase activity is also to be suspected in such cells from patients with inflammatory disorders. Our study has demonstrated, in human cells, the specificity of phospholipid methylation by [methyZ-8H]SAM adenylmethionine. In studies to be reported 4 Y. Niwa, T. Sakane, T. Kanoh, and Y. Miyachi, “Methyltransferase and phospholipase AZ activity in the membrane of neutrophile and lymphocytes from the patients with Behcet’s disease;” Submitted for publication. 6 “Methyltransferase and phospholipaee AP activity in the membrane of neutrophils and lymphocytes from the patients with syetemic lupus erythematosus and rheumatoid arthritis;” Submitted for publication.

OF HUMAN

LEUKOCYTES

13

elsewhere by the authors (35),4-6we have observed that methyltransferase activity is increased in both neutrophils and lymphocytes from patients with inflammatory diseases in direct proportion to the intensity of the inflammatory states. ACKNOWLEDGMENT We thank Dr. Satoshi Toyoshima, Pharmacology Department Tokyo University, for his kind help and suggestions in techniques. REFERENCES HIRATA, F., AND AXELROD, J. (1980) Science (Washington, D. C.) 209,1082-1090. EDELMAN, G. M. (1976) Science (Washington, D. C.) 192.218-228. HIDALGO, C., THOMAS,D. D., AND IKEMOTO,N. (1978) J. BioL Chem 253,6879-6887. ISHIZAKA,T., HIRATA, F., AND ISHIZAKA,K. (1980) Proa Nat1 Acad Sci USA 77,1903-1906. RIMON, G., HANSKI, E., AND BRAUN, S. (1978) Nature (Lmdm) 276.394-396. HIRATA, F., STRIITMATTER,W. J., AND AXELROD, J. (1979) Proc. Nat1 Acad Sci USA 76, 368372. 7. HIRATA, F., AND AXELROD,J. (1978) P~OC NatL Acad Sci USA 75,2348-2352. 8. HIRATA, F., AND AXELROD,J. (1978) Nature (kmah) 275, 219-220. 9. STRITTMATCER, W. J., HIRATA, F., AND AXELROD, J. (1979) Science (Washington, D. C) 204,12051207. 10. PIKE, M. C., KREDICH, N. M., AND SNYDERMAN, R. (1978) Proc NatL Ad Sci USA 75, 39283931. 11. HIRATA, F., CORCORAN,B. A., AND VENKATASUBURAMANIAN,K. (1979) Proc NatL Acd Sci USA 76.2640-2643. 12. SCHIFFMANN,E., CORCORAN,B. A., AND ASWANIKUMAR,S. (1978) in Leukocyte Chemotaxie (Gallin, J. I., and Quie, P. G., eds.), pp. 97103, Raven Press, New York. 13. HIRATA, F., TOYOSHIMA,S., AXELROD, J., AND WAXDAL, M. J. (1980) Proc NatL Acad Sci USA 77.862-865.

’ “Effect of administration of glucocorticoeteroids on methyltransferase and phospholipaae As activity in the membrane of neutrophile and lymphocytes from the patients with various disease;” Submitted for publication.

14

NIWA,

SAKANE,

14. TOYOSHIMA, S., HIRATA, F., IWATA, M., AXELROD, J., OSAWA, T., AND WAXDAL, M. J. (1982) MoL Immund 19,467-475. 15. CREWS, F. T., MORITA, Y., AND HIRATA, F. (1980)

Biochem Biophya Rea Commun. 93,42-49. 16. ISHIZAKA, T., FOREMAN, J. C., AND STERK, A. R. (1979) Proc Nat1 Acad Sci USA 76, 58585862. 17. VANCE, D. E., AND KRUIJFF, B. (1980) Nature (London) 288.277-279. 18. PAROLA, A. H., KAPLAN, J. H., LOCKWOOD, S. H., AND UZGIRIS, E. E. (1981) Biochim Biophya Acta 649, 616-624. 19. LOHRMANN, H. P., NOVIKOVS, L., AND GRAW, R. G. (1974) Nature (London) 250, 144-149. 20. MORI, K., TANIGUCHI, S., KUMADA, K., NAKAZAWA, K., FUJIWARA, M., AND FUJIWARA, M. (1933) Thromb. Res. 29, 215-224. 21. BLIGH, E. G., AND DYER, W. J. (1959) Can& .I Biochem Physiol 37,911-917. 22. YAVIN, E., AND ZUTRA, A. (1977) Ad Biochem &x0,430-437. 23. TANAKA, Y., DOI, O., AND AKAMATSU, Y. (1979) Biochem Biophys. Rea Commun. 87, 11091115. 24. SKOSEY, J. L., CHOW, D. C., MUSINOW, S., MAY,

TANIGUCHI J., GESTAUTAS, V., AND NIWA, Y. (1981) J. CeU

BioL 88, 358-363. 25. NIWA, Y., MIYAKE, S., SAKANE, T., SHINGU, M., AND YOKOYAMA, M. (1982) Clin. Exp. Immund

49.247-255. 26. NIWA, Y., AND KANOH, T. (1979) Clin Exp. Immund 37.470-476. 27. SAKANE, T., STEINBERG, A. D., REEVES, J. P., AND GREEN, I. (1979) J. Clin. Invest. 63,954-965. 28. QUASTEL, M. R., AND KAPLAN, J. B. (1970) Exp.

Cell Rea 63, 230-233. 29. ALLWOOD, G., AND ASHERSON, G. L. (1971) Immunology 21.509-516. 30. PETERS, J. H., AND HAUSEN, P. (1971) Eur. J. Biochem 19.502-508. 31. PETERS, J. H., AND HAUSEN, P. (1971) Eur. J. Biochem 19.509-513. 32. VAN DEN BERG, K. J., AND BETEL, I. (1974) Exp.

Cell Rea 84.412-418. 33. BARNETT, R. E., SCOTT, R. E., FURCHT, L. T., AND KERSEW, J. H. (1974) Nature (London) 249, 465-466. 34. MASUZAWA, Y., OSAWA, Y., INOUE, K., AND NOJIMA, S. (1973) Biochim Biophys Acta 326,339-344. 35. NIWA, Y., SAKANE, T., YAMAMOTO, S., KANO, T., AND MIYACHI, Y. (1934) I@lammation, in press.