Br. J. exp. Path. (1977) 58. 359
THE ROLE OF THE PLASMALOGEN IN THE CROSS-REACTION BETWEEN GROUP A STREPTOCOCCUS AND HUMAN MYOCARDIUM D L. E. GLYNN* E. KASP-GROCHOWSKA* Fram the MRC Rheumatism Unit, Canadian Red Cross M1emorial Hospital, Taplaw, Maidenhead, Berkshire
Iteceived for publication March 3, 1977
Summary.-Ethanol-soluble mycardial material which reacts with anti-streptococcal sera in a number of immunological tests has been isolated and identified as ethanolamine plasmalogen. The reactions of cardiac plasmalogen with antistreptococcal sera was specific and could be inhibited by streptococcus-derived materials. Guinea-pigs sensitized to streptococci gave positive skin reactions when challenged with myocardial plasmalogen. The pattern of the immunofluorescent staining given by antiplasmalogen sera was very much like that given by antistreptococcal sera. Nevertheless, the plasmalogen failed to compete for tissue-bound myocardial antigens when tried as an inhibitor of the immunofluorescent staining of myocardium either by antistreptococcal sera or by antiplasmalogen sera. A hypothesis of the role of the plasmalogen in the formation of complexes between streptococci and myocardium-derived material in the initiation of autoimmune processes is presented.
THE challenge of the search for the non-protein myocardium-derived antigen cross-reacting With streptococci came from work done in this laboratory in the past (Nakhla and Gl-nn, 1967; Kasp-Grochowska, Kingston and Glynn, 1972) and the study of the voluminous literature on this subject of cross-reaction. Apart from the debatable role of this cross-reaction in the aetiologvY of rheumatic fever, there is a difference of opinion on the subject of the chemical identity, number and localization of the antigens involved. The antigens have been described by Kaplan's school as proteins closely associated with the streptococcal cell wall (Kaplan, 1963; Kaplan and Suchy, 1967: Kaplan, Espinosa and Frengley, 1967) having their counterpart in acidwextractable mvocardial proteins. Zabriskie and his collaborators have found plasma *
membranes of streptococci and respective membranes of myocardial fibres to be responsible for mimetic reaction of the host and the bacteria (Zabriskie and Freimer, 1966). Antigenic similarities between the Cpolysaccharide of streptococcus and glycoprotein of human valves have been claimed bv French workers (Goldstein, Halpern and Robert, 1967) their latest report referring to cell-sap of Streptooccu8 pyogenes as containing the antigen crossreacting with valvular glycoprotein (Halpern, Parlebas and Goldstein, 1971). We became rather doubtful about this cross-reaction, as we could not confirm it in our laboratory (Kasp-Grochowska et al., 1972). Nevertheless, the recent paper from Japan (Kawakita and Iwamoto, 1975) is in support of the French workers' claim. Antibodies to cvtoplasmic antigens of haemolytic streptococci have been
Present address: The Kennedy Institute of Rheumatology, Bute Gardens, London W6 7DW.
360
E. KASP-GROCHOWSKA AND L. E. GLY.NN
described in rheumatic patients by Russian authors (Tan-Sik-Khok, Rassokhina and Mazina, 1974). Sandson et al. (1968) described yet another cross-reaction between streptococcal hyaluronate and human hyaluronate protein extracted from synovial fluid and cartilage. Nakhla and Glvnn (1967) have made observations in this laboratory that material cross-reacting with streptococci can be extracted from myocardium with ethanol. As this was in contradiction with the findings of other authors, research was undertaken to recover, isolate and characterize chemically the material concerned. As the result of this study the ethanolamine plasmalogen of human mvocardium has been found responsible for the reaction with antistreptococcal sera. MATERIALS AND METHODS
Uatera8 Group A streptococcal derived material.&-Used either for immu tion, absorption or inhibition studies were as follows: sonicates of streptococcal cell walls of Group A P-haemolytic streptococcus Type 54 and Type 6 matt; crude hot formamide extract and partially purified C polysaccharide from Type 6 matt and A variant. The C polysaccharides were those already described in detail in a previous paper (Kasp-Grochowska et al., 1972); they were prepared according to the technique described by Erwa, Maxted and Brighton
(1969). Antistreptococcal sera.-Two groups of antisera were used throughout the experiments. The first group were the sera raised in rabbits against the whole streptococei grown in a synthetic medium supplemented with tryptic digest of casein. Properties of these sera, already used in our laboratory and described in full in another paper (Kasp-Grochowska et al., 1972), are compiled in Table I. The second group of sera, so-called antistreptococcal cell wall sera (SCWV) were k-indly
supplied by Dr D. A. Rajapakse. They were raised in rabbits by intradermal injections of sonicates of cell walls of Group A P-haemolytic streptococcus Type 54. Rabbits received 2 mg of freeze-dried sonicates intradermally in incomplete Freund's adjuvant at the first inoculation, followed by 1 mg of this material 3 weeks later. The blood was collected 5 weeks after the second injection. Some of the rabbits were given booster injections a few months later, blood being taken after another 5 weeks. Preparatite techniques Tisme extracti4n.-Large pieces of human myocardium or whole hearts obtained from autopsy usually within 12 h of death were kept frozen at -30° until used. The tissue was sliced with a scalpel while still frozen, care being taken to remove fat and fibrous tissue. These slivers of myocardium were homogenized with the minimum of saline in a Waring blender. Seven parts of absolute ethanol were mixed with each 3 parts of homogenate and left for extraction overnight at room temperature (with occasional shaking). The extract was separated from tissue debris by filtration with suction on a Buchner funnel lined with filter paper. The solid cake of tissue was extracted once again with half of the original volume of 70% ethanol. The two extracts were combined and the organic solvent removed in a rotary thin film evaporator until it was possible to freeze the remaining liquid, which was subsequently freeze-dried. Lipid fracionation.-The freeze-dried residue was extracted with chloroform : methanol 2: 1 (v/v) mixture. The resulting extract was evaporated under a stream of nitrogen and the residue extracted once with 5 volumes of acetone. Acetone-insoluble lipids (AcR-Acetone residue) were then extracted with 2 volumes of cold diethyl ether. The ether extract was separated by centrifugation at -, 4°. The lipids were recovered from ether bv evaporation under a stream of nitrogen. The residue was dissolved in chloroform and stored at -30°. Ether-soluble material (ES fraction) was separated on thick (2-5-mm) silica gel-coated plates PF254 containing incorporated fluorescein
TABLE L.-Properties of Antistreptococcal Sera Rabbit No. .~~~~~~~~~~~~~~~A
Type of Str. pyogenaes Group antibodies Type antibodies
10 6 matt
6
13
14
15
16
glossy
6 glossy
12
12
_(a
)
(-)
18 24
THE ROLE OF ETHANOLAMINE PLASMALOGEN3
enabling visualization of lipid fractions. 300 mg of ES as a 1000 solution in chloroform: methanol 2: 1 was applied as a band 1 inch above the bottom edge on single plate 20 x 20 cm. The plates were dried thoroughly under a stream of nitrogen and developed in chloroform : methanol : water 65 : 25 : 4 (v/v/v). Developed plates were briefly dried under nitrogen and viewed under a u.v. lamp to outline the bands of separated lipid fractions. The fractions were named as follows: RF-O, place of application: RF1 relative mobility on the plate, 0-1-0-2; RF2 of the mobility 0-2--0-3; RF3 of the mobility 0-3-0-4 and RF4 of the mobility 0-4-0-5. The selected areas of silica were scraped off the plate and placed in conical flasks. Chloroform was added and clumps of silica broken with a glass rod and diluted into a thin slurry which was packed into small glass columns. The columns were eluted with chloroform : methanol mixtures of 2: 1, 1 : 1 and 1 : 4 (v/v) in that order. The eluates were evaporated under reduced pressure nearly to dryness and the residue dissolved in measured volumes of chloroform : methanol 2 : 1 (v/v). The samples for estimation were taken and the rest stored in air-tight containers under N. at -30.
361
tives on the same acid-hydrolyzed samples fractionated between water (serine) and chloroform (ethanolanine) phases as described in
detail by Long and Staples (1961b). Choline.-This was determined as described by 'Wheeldon and Collins (1958) as choline phosphomolybdate in the water soluble fraction of hydrolyzate (6 N-HCR 100°-16 h) of lipid specimen. Free amino groups of unhydrolysed lipids. These were estimated with ninhydrin in 90% methyl cello-solve as described by Lea and Rbodes (1955). Amino groups in acid hydrolysates of phospholipids were determined either by the FDNB (fluorodinitrobenzene) technique of A-xelrod, Reichenthal and Brodie (1953) or the ninhydrin method as improved by Cocking and Yemm (1954). Glycerol.-The content of glycerol was measured with chromotropic acid in acid hydroIysates of phospholipids, with a glycerol standard taken through the whole procedure as described by Renkonen (1962). Estimation of long chain fatty alehydes.-This was done exactly as described in detail by Gray (1969). Dimethylpamitate (Sulpelco Inc.) was used as the standard. Esterifed fatty acids in phospholipids were Analytical method measured as ferric hydroxamic acid complexes Phosphorus estimation.-Digestion and de- as described by Rapport and Alonzo (1955). velopment of the colour reaction were performed Commercially available high-purity lecithin in 1-7 x 13-cm Pyrex tubes. The samples (Sigma, London) and methyl stearate were used containing 5-40 FLg of P were introduced into as standards. Thin-layer chromatography of lipids was the tubes. The samples were either measured aliquots of umhydrolized lipids in organic solvent performed on commercially available ready or water hydrolyzates, or scraping of silica coated 0-25-mm thick plates (Silica Gel 60 or cellulose from respective thin-layer chromato- from 'Merk Anderman & Co. Ltd.). The plates grams. With liquid samples the solvent was were preheated before use for 30 min at 1100. evaporated to dryness under a stream of N2 Shandon tanks lined with filter paper were before addition of 0-7 ml of 72/ perchloric acid. used for development. The solvents were The digestion was carried out in an electrical always made fresh for each chromatograph, thermoblock placed in a fume cupboard. The 1-2 h being allowed for equilibration before temperature was set at 200 3--/-5 . Thirty development. The solvent systems employed either for min were usually adequate for digestion of the specimens. The tubes were cooled; to each, mono or two-dimensional development were 8 ml of distilled water was added followed as follows: (I) chloroform: methanol: water by 0-4ml of 8-3%o ammonium molybdate 65 : 25 : 4; (II) chloroform: methanol: glacial and 0-4 ml of reducing reagent as described acetic acid 65 : 25 : 8; (Im) chloroform: methby Allen (1940). The solutions were made up anol : 7 x ammonia in water 60 : 35 : 5, Ames to 10 ml, mixed and left for 20 min. Extinctions (1968). The spots were localized by charring with were read at 660 jim against blanks. Standards of 10 and 20 jig P were included in every 500o sulphuric acid. Phospholipids were identiestimation and treated the same way as un- fied with Dittmer and Lester's (1964) spray reagent specific for phosphate esters. Phosphoknown samples. Total nitrogen .-This was determined by lipids containing amino groups were stained nesslerization essentially as described by Long wlith the ninhydrin reagent (0-1% in n-butanol). and Staples (1961a), their Gallenkamp micro- Commercially available sphingomyelin (Sigma, Kjeldahl digestion apparatus being replaced by London), phosphatidyl choline and phosphatidyl ethanolamine were used as reference standards. a thermoblock. "Argentation" thin-layer chromatography of Ethanolamine and serine.-These were estimated as their respective dinitrophenyl deriva- methyl esters of polyunsaturated fatty acids,
362
E. KASP-GROCHOWSKA AND L. E. GLYNN
enabling the separation of fatty acids containing 0-6 double bonds was performed as described by Dudley and Anderson (1975). Thin-layer chromatography of the products of hydrolysis of phosphilipids an cellulose plates was carried out as described by Haworth and Heatheote (1969) for amino acids and related compounds. For identification of amino compounds the ninhvdrin-cadmium acetate reagent as recommended by Haworth and Heathcote was used. Phosphate esters were localized with Rosenberg's reagent (Rosenberg. 1959), which was originally designed for paper chromatograms but could be adapted to cellulose plates by spraying instead of dipping. Dragendorf reagent was used for detection of quaternary ammonium and related compounds.
Im-munological methods Passire haenigglutinration test.-This is our variation of the classical technique. Rabbit red cells were used; they were formolized as described by Herbert (1973) and kept refrigerated for months as a 1000 v/v saline solution with sodium merthiolate added as a preservative. For the test, 1 ml of thoroughly mixed suspension of cells wa-s centrifuged for 10 min at 1500 rpm, the supernatant drained and 3-6 ml of saline adjusted to pH 4-1 added, followed by 0-2-0-4 ml of a 0.500 solution of the antigen. The suspension was kept on a rota-table for 1 h at room temperature. The cells were spun and washed twice with 4 ml PBS pH 7-2 and finally suspended in 4 ml of PBS resulting in a 2-5%I cell suspension. The sera to be tested were absorbed with packed formolized rabbit cells in a proportion of 1 part packed cells to 4 parts of serum. The absorption was allowed to proceed for 2 h at room temperature on a rota-table The test was performed in thin-walled 7 x 100mm rotud bottom tubes. A drop dilution technique was employed. To each tube in a row of 10, 2 drops of PBS were added from a Pasteur pipette. Two drops of serum were added to the first tube, mixed with diluent and 2 drops transferred to the next tube. Two drops of diluent were added to each tube followed by 1 drop of antigen-coated cells. The controls contained 4 drops of PBS and 1 drop of coated cells. WA-hen newly bled sera were used, the controls included a row of solutions of the sera and uncoated cells; they were always negative. The contents of the tubes were mixed by shaking and left for E h at room temperature followed by overnight incubation at -4 The conditions of incubation have been found optimal in a number of experiments including variation of time and temperature.
The haemagglutination was read macro-
scopically holding the bottom of the tube
against the light. The results were expressed as - to depending upon the pattern of the sedimented cells. Inh ibition of paxssive hoemagglutin2ation.-To the serial dilution of inhibitor in 2 drops of PBS, 2 drops of appropriately diluted antiserum were added. The dilution of serum was chosen as
1
of the titre of the serum that gave a
good (4) positive reaction in passive haemagglutination. Inhibitors and antisera were left at room temperature for 1 h and then 1 drop of coated cells was added. Mixtures were left for w h at room temperature and then overnight at -4". Two types of controls were included: positive, containing the serum diluted as in the inhibitorv test and coated cells, and negative with the least diluted inhibitor and control cells only. The strength of the inhibitor usually emploved as a 0-.50, solution was expressed as its maximum titre still giv%ing complete inhibition. Double diffusion ini agar.-This was performed in plastic Petri dishes in 100 agar in phosphate buffer, pH 7-2. with 0-100 azide incorporated as a preservative. The holes of 3 or 5 mm were cut in a 3-mm-thick layer of agar in various patterns, the distance between holes being 21-5 mm. The bottoms of the holes were "sealed" wVith a minimum amount of hot agar solution to prevent the leakage of the solutions under the agar gel. Complementfixation.-As described for soluble virus antigens by Tyrrell (1973). Tests of the inhibition of macrphage migration.-These were carried out in collaboration with Dr D. A. Rajapakse. In. this technique human peripheral blood lymphocytes are used in a mixed system with guinea-pig peritoneal exudate cells. The details of the method are described by Rajapakse and Glynn (1970). Immunofluorescent staining of human myocardium.-This was generally carried out by the technique recommended by Holborow and Johnson (1973). Pieces of human mvocardium used for fluorescence study were usually obtained Within 12 h of death. They were "snap-frozen" in thin polythene bags in a bath of solid CO2 in ethanol. They were stored at -30° until used. 6-gm-thick crvostat sections were dried in a stream of air for j h and used without any fixation. The sera of appropriate dilutions were layered over the sections and incubated for 30 mmn at room temperature in a moist atmosphere. This was followed by a 15-min wash in PBS, pH 7-2, with constant stirring.
Commerciall- obtainable fluorescein-labelled
goat anti-rabbit IgG (Wellcome Reagents Ltd) diluted 1/40 was applied and left for 30 min. The final wash lasted 45 miM. The slides were
THE ROLE OF ETHA1NOLAMIN E PLASMALOGEN mounted in glvcerine-PBS mixture and viewed under a Reichert microscope.
Skin tests for delayed hypersensitivity Twenty-four animals diVided into 3 groups were used. The first group of 8 guinea-pigs was immunized with sonicated streptococcal cell walls in complete Freund's adjuvant, the second group received the same amount of streptococcal material but in incomplete Freund's adjuvant; the remaining 8 animals wsere left uninjected and served as controls. The mixtures for injection for 8 guinea-pigs were prepared as follows: 20 mg of freeze-dried sonicated streptococcal cell walls was suspended in 2 ml of saline and 2-2 ml of either complete Freund's adjuvant was added or the same volume of incomplete Freund's adjuvant. The mixtures were homogenized to obtain water-in oil emulsions. Animals were injected intradermally in multiple sites on their backs with 0-2 ml of emulsion per guinea-pig. They were reinjected with half of this volume 3 weeks later. Each animal received a total of about 2-3 mg of streptococcal cell walls. Guinea-pigs were tested for delayed skin reaction 1 week after booster injection. Animals were challenged with streptococcal sonicates and lipid haptens. A series of 0- 1-ml intradermal injections of different strengths of these antigens was placed on both shaved flanks. The results were read at 24 and 48 h after the injections. The redness was measured in millimetres of diameter and the thickening was assessed macroscopically and scored from - to
Antisera against myocardial planmlogen Three different techniques were chosen from current literature for immunization with lipid haptens. The hapten was enriched with another "helper" lipid: lecithin or cholesterol as well as With protein. The protein used was either bovine serum albumin (BSA) or methylated bovine serum albumin (MBSA). 1. The first group of 4 rabbits.-This was given a series of i.v. injections of a mixture consisting of antigen, cholesterol and BSA (in the ratio 1: 5 : 20). This was a slight modification of the scheme developed by Coulon-3Morelec, Faure and Mfarechal (1970) for raising antibodies against palmitic ester of glucuronic acid. The amount of antigen used u-as twice as much as the amount used for palmitic glucuronate as our antigen's molecular weight is approximately twice as high. The solutions for injections (for 2 rabbits) were prepared as follows: 10mg of fraction 0-4 RF (either '.aged" at room temperature or stored at -30 ) and 50 mg cholesterol were dissolved in 1 ml ethanol at 70". The solution, while still hot, was injected rapidly into 19 ml saline. "Aged" 0-4 (RF) mixture resulted in
363
a very fine milky suspension. but fraction 0 4 RF (-30c) produced a precipitate which had a tendency to float. Energetic dispersion with a syringe improved the suspension. 2 ml 1000 BSA in saline was added to each solution with energetic mixing. Addition of this protein improved both emulsions, particularly that with 0-4 RF (-30 ° ) fraction. The solutions were ampouled in 2-ml portions, snap-frozen in acetone-solid CO2 mixture and stored at -30 until used. Each rabbit received 10 injections of 1 ml of these solutions ever- other day. Samples of blood were taken after the 6th injection and the rabbits were bled 8 days after the last
injection.
Three out of 4 rabbits were given an extra booster injection a few weeks later, blood being taken after 8 more days. As some rabbits showed a slight shock reaction on the third injection, antihistamine treatment was employed. Each rabbit was given 10 mg of Piriton i.m. 45 min before antigen
mjection. 2. The serond group of 4 an imals.-These animals received antigen combined with lecithin and methvlated bovine serum albumin in the ratio 1: 10 : 16. The immunization scheme employed was that found best for cardiolipin out of those tried by Inoue and Xojima (1967). Four rabbits received courses of IO i.v. injections (Piriton protection being employed from the 4th injection) spaced every 2 days. The total amount of antigen given to each animal was 15 mg. The day preceding the injection the solutions were prepared as follows: for 2 rabbits I mg of fraction 0-4 RF either "aged" at room temperature or kept at -30 were added to 10 mg lecithin. All lipids were originally dissolved in chloroform : methanol mixture 2 : 1. The solvents were evaporated under a stream of nitrogen at room temperature and the residues suspended in 2 drops of water. 4 ml of 0-10o solution of MIBSA in saline was added rapidly from the svringe to each lipid suspension. Multiple aspiration and injection of fluid improved the emulsion, which was left overnight at -' 4 to be used the next day. Rabbits were bled on the 8th day after the last injection, samples of blood also being withdraw-n after the 6th injection. 3. The third group of rabbits received the antigen and helper mixtures incorporated into incomplete Freund's adjuvant injected intra-
dermallv. This was a modification of the method designed successfully by Tal, Taketomi and Yamakawa (1967) to obtain anti-lactoceramide antibodies. The ratios of helpers : lecithin and bo-ine serum albumins were as described in the original paper, but complete Freund"s adjuvant was replaced by us by incomplete
364
E. KASP-GROCHOWSKA AND L. E.
Freund's adjuvant. This was done to avoid introducing yet another antigenic system, as we have already shown that anti-tuberculin antibodies react with streptococci (KaspGrochowska et al., 1972). Out of the 6 rabbits, 3 were injected with a 04 RF "aged" fraction and the other 3 with a fraction stored at -30° until used The mixtures for immunization of 3 animals were prepared as follows: 15 mg of fraction 0.4 RF and 30 mg lecithin were dissolved together in hexane. The hexane was evaporated, lipids dissolved in a small amount (0-2-0 3 ml) of ethanol, and 1-5 ml saline was injected rapidly into it. The mixture was agitated by aspiration and expulsion from the syringe until a fine emulsion was obtained. 0-6 ml of a 10% saline solution of BSA was added followed by 2-3 ml incomplete Freund's adjuvant (Difco) and the mixture was homogenized to obtain a water-inoil emulsion. Each animal received in total 1 ml of the emulsion into 5 sites on the back by intradermal injection. This procedure was repeated 3 or 4 times every 5 weeks, the blood being taken before each reinjection and 5 weeks after the last. Animals were protected with Piriton before injections, starting from the 4th. The total amount of antigen given to each rabbit was 15-20 mg. RESULTS
The ethanol-soluble material of mvocardium recovered from the solution and subsequentlv fractionated has been assessed for antigenic cross-reactivity with antistreptococcal sera in either all or some of the following tests: double diffusion in agar, complement fixation, passive haemagglutination and its inhibition, as well as bv inhibition of macrophage migration in mixed culture of lymphocvtes of rheumatic fever patients with guinea-pig macrophages. The finally chosen purification procedure has been detailed in the Methods section. Problems arose once the material had been recovered from ethanol solution because it was no longer dispersable in water. This could however be facilitated by mixing with 2-5 parts of lecithin or cholesterol in an organic solvent, evaporating the solvent and making a watery emulsion with resulting lipid mixture. The other alternative was to
GLYNN
'.age" the material at room temperature in a solid state. This resulted in changed mobility of the antigen in TLC on silica gel but rendered the antigen perfectlv water-soluble. A. The estimation of cardiac lipid as the antigen in readcion uith antistreptococcal sera The results of the assessment of the ether-soluble material in passive haemagglutination, complement fixation and double diffusion in agar are presented in Table II. Commerciallv available grouping serum and two groups of antistreptococcal sera obtained in our laboratories were used. The sera Nos. 10, 13, 14, 15, 16 and 18 were raised against whole streptococci; the names of the strains are given in the second column. Sera 11 and 12 served as controls; thev came from rabbits which were inoculated with growth medium onlv. Sera 1, 2, 3, 4, 5 and 6 were from rabbits which received the sonicates of streptococcal cell walls (SCW). The sera with the highest titres (Nos. 1 and 2 in the SCWV group) were consequentlv used in most tests of passive haemagglutination or its inhibition in estimating the fractions recovered from silica gel chromatography. Out of the group of sera raised against whole streptococci, Sera 10 and 15, and the commercially available grouping serum were the strongest in passive haemagglutination of the cells coated with mvocardial antigen. WVhen tested in double diffusion in agar, 3 sera out of 6 raised against streptococcal sonicates gave a distinct line of precipitation. These were Sera Nos. 1, 2 and 3, the sera with the highest titres in the passive haemagglutination test. Of the sera raised against the whole streptococcal organisms onlv one, Serum No. 10, gave a faint line of precipitation. The same sera were also the most efficient in fixing complement in the presence of mvocardial antigens. Cardiac fractions at different stages of purification and streptococcus-derived
365
THE ROLE OF ETHANOLAMIN PLASMALOGENE
TABLE II.-The Rests of Cardiac Hapten (ES) Reactions uith Rabbit Antistreptococcal Sera Passive haemagglutination No. of tube A
The titre m comLine in 4x 8A 16x 32x Mx 128x 256x 512x 1024A plement double 4x 8 x 16 x 32 x 64 x 128 x 256 x 512 x 1024 x fixation diffusion 1/128 Strong 1/256 Strong 1/64 Strong 1/128 5 1/16 sCw 6 1/36 sCw 10 631 Weak 1/64 11 gmt nd 12 nd gm 13 nd 6G 14 nd 6G 12 15 1/64 16 nd 12 18 24 nd nd Grouping serum The cells used for passive haemagglutination were coated with 0 -4 ml of 0 50O ES per 0 1 ml of packed formolized erythrocytes. The concentration of ES in complement fixation test was 200 LAg and 0-5% in double diffusion test. * SCW-sonicated streptococcal cell walls; f gm-growth medium; nd = not done. 1
2
3
TmmunNo. of ing serum antigen 1 ScW* 2 SCw 3 sCw 4 sCw
4
6 Serum dilution
8
7
5
9
TABLE III.-The Actitity of Different Fractions of Ethanol Extracts of Human MIyocardium and some of the Streptococcal-derived Material Measured in the Test of Inhibition of Passive Haemagglutination -No. of tube 2
1
3
4
Mlaterial
5
6
8
7
9
10
Dilution of fraction
The
used for Fraction tested 0 20o Neat 2x coating 1 SCew ES ES 50x ENS diluted Acetone extract 2 SCW AcR AcR L Ph 100x diluted U Ph O 1' sonicated streptococcal cell walls 2 SCW ES ES alkali resistant 100 x Fuller (formamide extract of 6M streptodiluted sera
4x
8x
16x
32x -
64x 128x 256x 512x -
-,r-
I
cocci)
1 SCW 50 x diluted
ES
6 M C polysaccharide A' C polysaccharide
ES-Alkali resistant, treated with NaOH 0 -1 AcR Acetone residue.
N
for 30' at 37' and recovered
materials were used as inhibitors in passive haemagglutination of cardiac antigen-coated cells by antistreptococcal sera. Tables III and IV contain results of
on
silica column.
manv of these tests done in the process of identification of cardiac hapten and evaluation of its relationship to streptococci.
E. KASP-GROCHOWSKA A-ND L. E. GLYN\--
366
TABLE IV.-The Activity of the Fractions Recovered from Silica TIC' Plates, Measured in the Inhibition Test of Passive Haemagglutination No. of tube )~~~~~~~~~
The fraction used for inhibition 0-4RF 0-3RF
10
11
256 x
512 x
1024 x
-
-
-
9
8 7 Dilution of inhibitor
1
2
3
4
a
Neat
2x
4X
8
THE ROLE OF THE PLASMALOGEN IN THE CROSS-REACTION BETWEEN GROUP A STREPTOCOCCUS AND HUMAN MYOCARDIUM D L. E. GLYNN* E. KASP-GROCHOWSKA* Fram the MRC Rheumatism Unit, Canadian Red Cross M1emorial Hospital, Taplaw, Maidenhead, Berkshire
Iteceived for publication March 3, 1977
Summary.-Ethanol-soluble mycardial material which reacts with anti-streptococcal sera in a number of immunological tests has been isolated and identified as ethanolamine plasmalogen. The reactions of cardiac plasmalogen with antistreptococcal sera was specific and could be inhibited by streptococcus-derived materials. Guinea-pigs sensitized to streptococci gave positive skin reactions when challenged with myocardial plasmalogen. The pattern of the immunofluorescent staining given by antiplasmalogen sera was very much like that given by antistreptococcal sera. Nevertheless, the plasmalogen failed to compete for tissue-bound myocardial antigens when tried as an inhibitor of the immunofluorescent staining of myocardium either by antistreptococcal sera or by antiplasmalogen sera. A hypothesis of the role of the plasmalogen in the formation of complexes between streptococci and myocardium-derived material in the initiation of autoimmune processes is presented.
THE challenge of the search for the non-protein myocardium-derived antigen cross-reacting With streptococci came from work done in this laboratory in the past (Nakhla and Gl-nn, 1967; Kasp-Grochowska, Kingston and Glynn, 1972) and the study of the voluminous literature on this subject of cross-reaction. Apart from the debatable role of this cross-reaction in the aetiologvY of rheumatic fever, there is a difference of opinion on the subject of the chemical identity, number and localization of the antigens involved. The antigens have been described by Kaplan's school as proteins closely associated with the streptococcal cell wall (Kaplan, 1963; Kaplan and Suchy, 1967: Kaplan, Espinosa and Frengley, 1967) having their counterpart in acidwextractable mvocardial proteins. Zabriskie and his collaborators have found plasma *
membranes of streptococci and respective membranes of myocardial fibres to be responsible for mimetic reaction of the host and the bacteria (Zabriskie and Freimer, 1966). Antigenic similarities between the Cpolysaccharide of streptococcus and glycoprotein of human valves have been claimed bv French workers (Goldstein, Halpern and Robert, 1967) their latest report referring to cell-sap of Streptooccu8 pyogenes as containing the antigen crossreacting with valvular glycoprotein (Halpern, Parlebas and Goldstein, 1971). We became rather doubtful about this cross-reaction, as we could not confirm it in our laboratory (Kasp-Grochowska et al., 1972). Nevertheless, the recent paper from Japan (Kawakita and Iwamoto, 1975) is in support of the French workers' claim. Antibodies to cvtoplasmic antigens of haemolytic streptococci have been
Present address: The Kennedy Institute of Rheumatology, Bute Gardens, London W6 7DW.
360
E. KASP-GROCHOWSKA AND L. E. GLY.NN
described in rheumatic patients by Russian authors (Tan-Sik-Khok, Rassokhina and Mazina, 1974). Sandson et al. (1968) described yet another cross-reaction between streptococcal hyaluronate and human hyaluronate protein extracted from synovial fluid and cartilage. Nakhla and Glvnn (1967) have made observations in this laboratory that material cross-reacting with streptococci can be extracted from myocardium with ethanol. As this was in contradiction with the findings of other authors, research was undertaken to recover, isolate and characterize chemically the material concerned. As the result of this study the ethanolamine plasmalogen of human mvocardium has been found responsible for the reaction with antistreptococcal sera. MATERIALS AND METHODS
Uatera8 Group A streptococcal derived material.&-Used either for immu tion, absorption or inhibition studies were as follows: sonicates of streptococcal cell walls of Group A P-haemolytic streptococcus Type 54 and Type 6 matt; crude hot formamide extract and partially purified C polysaccharide from Type 6 matt and A variant. The C polysaccharides were those already described in detail in a previous paper (Kasp-Grochowska et al., 1972); they were prepared according to the technique described by Erwa, Maxted and Brighton
(1969). Antistreptococcal sera.-Two groups of antisera were used throughout the experiments. The first group were the sera raised in rabbits against the whole streptococei grown in a synthetic medium supplemented with tryptic digest of casein. Properties of these sera, already used in our laboratory and described in full in another paper (Kasp-Grochowska et al., 1972), are compiled in Table I. The second group of sera, so-called antistreptococcal cell wall sera (SCWV) were k-indly
supplied by Dr D. A. Rajapakse. They were raised in rabbits by intradermal injections of sonicates of cell walls of Group A P-haemolytic streptococcus Type 54. Rabbits received 2 mg of freeze-dried sonicates intradermally in incomplete Freund's adjuvant at the first inoculation, followed by 1 mg of this material 3 weeks later. The blood was collected 5 weeks after the second injection. Some of the rabbits were given booster injections a few months later, blood being taken after another 5 weeks. Preparatite techniques Tisme extracti4n.-Large pieces of human myocardium or whole hearts obtained from autopsy usually within 12 h of death were kept frozen at -30° until used. The tissue was sliced with a scalpel while still frozen, care being taken to remove fat and fibrous tissue. These slivers of myocardium were homogenized with the minimum of saline in a Waring blender. Seven parts of absolute ethanol were mixed with each 3 parts of homogenate and left for extraction overnight at room temperature (with occasional shaking). The extract was separated from tissue debris by filtration with suction on a Buchner funnel lined with filter paper. The solid cake of tissue was extracted once again with half of the original volume of 70% ethanol. The two extracts were combined and the organic solvent removed in a rotary thin film evaporator until it was possible to freeze the remaining liquid, which was subsequently freeze-dried. Lipid fracionation.-The freeze-dried residue was extracted with chloroform : methanol 2: 1 (v/v) mixture. The resulting extract was evaporated under a stream of nitrogen and the residue extracted once with 5 volumes of acetone. Acetone-insoluble lipids (AcR-Acetone residue) were then extracted with 2 volumes of cold diethyl ether. The ether extract was separated by centrifugation at -, 4°. The lipids were recovered from ether bv evaporation under a stream of nitrogen. The residue was dissolved in chloroform and stored at -30°. Ether-soluble material (ES fraction) was separated on thick (2-5-mm) silica gel-coated plates PF254 containing incorporated fluorescein
TABLE L.-Properties of Antistreptococcal Sera Rabbit No. .~~~~~~~~~~~~~~~A
Type of Str. pyogenaes Group antibodies Type antibodies
10 6 matt
6
13
14
15
16
glossy
6 glossy
12
12
_(a
)
(-)
18 24
THE ROLE OF ETHANOLAMINE PLASMALOGEN3
enabling visualization of lipid fractions. 300 mg of ES as a 1000 solution in chloroform: methanol 2: 1 was applied as a band 1 inch above the bottom edge on single plate 20 x 20 cm. The plates were dried thoroughly under a stream of nitrogen and developed in chloroform : methanol : water 65 : 25 : 4 (v/v/v). Developed plates were briefly dried under nitrogen and viewed under a u.v. lamp to outline the bands of separated lipid fractions. The fractions were named as follows: RF-O, place of application: RF1 relative mobility on the plate, 0-1-0-2; RF2 of the mobility 0-2--0-3; RF3 of the mobility 0-3-0-4 and RF4 of the mobility 0-4-0-5. The selected areas of silica were scraped off the plate and placed in conical flasks. Chloroform was added and clumps of silica broken with a glass rod and diluted into a thin slurry which was packed into small glass columns. The columns were eluted with chloroform : methanol mixtures of 2: 1, 1 : 1 and 1 : 4 (v/v) in that order. The eluates were evaporated under reduced pressure nearly to dryness and the residue dissolved in measured volumes of chloroform : methanol 2 : 1 (v/v). The samples for estimation were taken and the rest stored in air-tight containers under N. at -30.
361
tives on the same acid-hydrolyzed samples fractionated between water (serine) and chloroform (ethanolanine) phases as described in
detail by Long and Staples (1961b). Choline.-This was determined as described by 'Wheeldon and Collins (1958) as choline phosphomolybdate in the water soluble fraction of hydrolyzate (6 N-HCR 100°-16 h) of lipid specimen. Free amino groups of unhydrolysed lipids. These were estimated with ninhydrin in 90% methyl cello-solve as described by Lea and Rbodes (1955). Amino groups in acid hydrolysates of phospholipids were determined either by the FDNB (fluorodinitrobenzene) technique of A-xelrod, Reichenthal and Brodie (1953) or the ninhydrin method as improved by Cocking and Yemm (1954). Glycerol.-The content of glycerol was measured with chromotropic acid in acid hydroIysates of phospholipids, with a glycerol standard taken through the whole procedure as described by Renkonen (1962). Estimation of long chain fatty alehydes.-This was done exactly as described in detail by Gray (1969). Dimethylpamitate (Sulpelco Inc.) was used as the standard. Esterifed fatty acids in phospholipids were Analytical method measured as ferric hydroxamic acid complexes Phosphorus estimation.-Digestion and de- as described by Rapport and Alonzo (1955). velopment of the colour reaction were performed Commercially available high-purity lecithin in 1-7 x 13-cm Pyrex tubes. The samples (Sigma, London) and methyl stearate were used containing 5-40 FLg of P were introduced into as standards. Thin-layer chromatography of lipids was the tubes. The samples were either measured aliquots of umhydrolized lipids in organic solvent performed on commercially available ready or water hydrolyzates, or scraping of silica coated 0-25-mm thick plates (Silica Gel 60 or cellulose from respective thin-layer chromato- from 'Merk Anderman & Co. Ltd.). The plates grams. With liquid samples the solvent was were preheated before use for 30 min at 1100. evaporated to dryness under a stream of N2 Shandon tanks lined with filter paper were before addition of 0-7 ml of 72/ perchloric acid. used for development. The solvents were The digestion was carried out in an electrical always made fresh for each chromatograph, thermoblock placed in a fume cupboard. The 1-2 h being allowed for equilibration before temperature was set at 200 3--/-5 . Thirty development. The solvent systems employed either for min were usually adequate for digestion of the specimens. The tubes were cooled; to each, mono or two-dimensional development were 8 ml of distilled water was added followed as follows: (I) chloroform: methanol: water by 0-4ml of 8-3%o ammonium molybdate 65 : 25 : 4; (II) chloroform: methanol: glacial and 0-4 ml of reducing reagent as described acetic acid 65 : 25 : 8; (Im) chloroform: methby Allen (1940). The solutions were made up anol : 7 x ammonia in water 60 : 35 : 5, Ames to 10 ml, mixed and left for 20 min. Extinctions (1968). The spots were localized by charring with were read at 660 jim against blanks. Standards of 10 and 20 jig P were included in every 500o sulphuric acid. Phospholipids were identiestimation and treated the same way as un- fied with Dittmer and Lester's (1964) spray reagent specific for phosphate esters. Phosphoknown samples. Total nitrogen .-This was determined by lipids containing amino groups were stained nesslerization essentially as described by Long wlith the ninhydrin reagent (0-1% in n-butanol). and Staples (1961a), their Gallenkamp micro- Commercially available sphingomyelin (Sigma, Kjeldahl digestion apparatus being replaced by London), phosphatidyl choline and phosphatidyl ethanolamine were used as reference standards. a thermoblock. "Argentation" thin-layer chromatography of Ethanolamine and serine.-These were estimated as their respective dinitrophenyl deriva- methyl esters of polyunsaturated fatty acids,
362
E. KASP-GROCHOWSKA AND L. E. GLYNN
enabling the separation of fatty acids containing 0-6 double bonds was performed as described by Dudley and Anderson (1975). Thin-layer chromatography of the products of hydrolysis of phosphilipids an cellulose plates was carried out as described by Haworth and Heatheote (1969) for amino acids and related compounds. For identification of amino compounds the ninhvdrin-cadmium acetate reagent as recommended by Haworth and Heathcote was used. Phosphate esters were localized with Rosenberg's reagent (Rosenberg. 1959), which was originally designed for paper chromatograms but could be adapted to cellulose plates by spraying instead of dipping. Dragendorf reagent was used for detection of quaternary ammonium and related compounds.
Im-munological methods Passire haenigglutinration test.-This is our variation of the classical technique. Rabbit red cells were used; they were formolized as described by Herbert (1973) and kept refrigerated for months as a 1000 v/v saline solution with sodium merthiolate added as a preservative. For the test, 1 ml of thoroughly mixed suspension of cells wa-s centrifuged for 10 min at 1500 rpm, the supernatant drained and 3-6 ml of saline adjusted to pH 4-1 added, followed by 0-2-0-4 ml of a 0.500 solution of the antigen. The suspension was kept on a rota-table for 1 h at room temperature. The cells were spun and washed twice with 4 ml PBS pH 7-2 and finally suspended in 4 ml of PBS resulting in a 2-5%I cell suspension. The sera to be tested were absorbed with packed formolized rabbit cells in a proportion of 1 part packed cells to 4 parts of serum. The absorption was allowed to proceed for 2 h at room temperature on a rota-table The test was performed in thin-walled 7 x 100mm rotud bottom tubes. A drop dilution technique was employed. To each tube in a row of 10, 2 drops of PBS were added from a Pasteur pipette. Two drops of serum were added to the first tube, mixed with diluent and 2 drops transferred to the next tube. Two drops of diluent were added to each tube followed by 1 drop of antigen-coated cells. The controls contained 4 drops of PBS and 1 drop of coated cells. WA-hen newly bled sera were used, the controls included a row of solutions of the sera and uncoated cells; they were always negative. The contents of the tubes were mixed by shaking and left for E h at room temperature followed by overnight incubation at -4 The conditions of incubation have been found optimal in a number of experiments including variation of time and temperature.
The haemagglutination was read macro-
scopically holding the bottom of the tube
against the light. The results were expressed as - to depending upon the pattern of the sedimented cells. Inh ibition of paxssive hoemagglutin2ation.-To the serial dilution of inhibitor in 2 drops of PBS, 2 drops of appropriately diluted antiserum were added. The dilution of serum was chosen as
1
of the titre of the serum that gave a
good (4) positive reaction in passive haemagglutination. Inhibitors and antisera were left at room temperature for 1 h and then 1 drop of coated cells was added. Mixtures were left for w h at room temperature and then overnight at -4". Two types of controls were included: positive, containing the serum diluted as in the inhibitorv test and coated cells, and negative with the least diluted inhibitor and control cells only. The strength of the inhibitor usually emploved as a 0-.50, solution was expressed as its maximum titre still giv%ing complete inhibition. Double diffusion ini agar.-This was performed in plastic Petri dishes in 100 agar in phosphate buffer, pH 7-2. with 0-100 azide incorporated as a preservative. The holes of 3 or 5 mm were cut in a 3-mm-thick layer of agar in various patterns, the distance between holes being 21-5 mm. The bottoms of the holes were "sealed" wVith a minimum amount of hot agar solution to prevent the leakage of the solutions under the agar gel. Complementfixation.-As described for soluble virus antigens by Tyrrell (1973). Tests of the inhibition of macrphage migration.-These were carried out in collaboration with Dr D. A. Rajapakse. In. this technique human peripheral blood lymphocytes are used in a mixed system with guinea-pig peritoneal exudate cells. The details of the method are described by Rajapakse and Glynn (1970). Immunofluorescent staining of human myocardium.-This was generally carried out by the technique recommended by Holborow and Johnson (1973). Pieces of human mvocardium used for fluorescence study were usually obtained Within 12 h of death. They were "snap-frozen" in thin polythene bags in a bath of solid CO2 in ethanol. They were stored at -30° until used. 6-gm-thick crvostat sections were dried in a stream of air for j h and used without any fixation. The sera of appropriate dilutions were layered over the sections and incubated for 30 mmn at room temperature in a moist atmosphere. This was followed by a 15-min wash in PBS, pH 7-2, with constant stirring.
Commerciall- obtainable fluorescein-labelled
goat anti-rabbit IgG (Wellcome Reagents Ltd) diluted 1/40 was applied and left for 30 min. The final wash lasted 45 miM. The slides were
THE ROLE OF ETHA1NOLAMIN E PLASMALOGEN mounted in glvcerine-PBS mixture and viewed under a Reichert microscope.
Skin tests for delayed hypersensitivity Twenty-four animals diVided into 3 groups were used. The first group of 8 guinea-pigs was immunized with sonicated streptococcal cell walls in complete Freund's adjuvant, the second group received the same amount of streptococcal material but in incomplete Freund's adjuvant; the remaining 8 animals wsere left uninjected and served as controls. The mixtures for injection for 8 guinea-pigs were prepared as follows: 20 mg of freeze-dried sonicated streptococcal cell walls was suspended in 2 ml of saline and 2-2 ml of either complete Freund's adjuvant was added or the same volume of incomplete Freund's adjuvant. The mixtures were homogenized to obtain water-in oil emulsions. Animals were injected intradermally in multiple sites on their backs with 0-2 ml of emulsion per guinea-pig. They were reinjected with half of this volume 3 weeks later. Each animal received a total of about 2-3 mg of streptococcal cell walls. Guinea-pigs were tested for delayed skin reaction 1 week after booster injection. Animals were challenged with streptococcal sonicates and lipid haptens. A series of 0- 1-ml intradermal injections of different strengths of these antigens was placed on both shaved flanks. The results were read at 24 and 48 h after the injections. The redness was measured in millimetres of diameter and the thickening was assessed macroscopically and scored from - to
Antisera against myocardial planmlogen Three different techniques were chosen from current literature for immunization with lipid haptens. The hapten was enriched with another "helper" lipid: lecithin or cholesterol as well as With protein. The protein used was either bovine serum albumin (BSA) or methylated bovine serum albumin (MBSA). 1. The first group of 4 rabbits.-This was given a series of i.v. injections of a mixture consisting of antigen, cholesterol and BSA (in the ratio 1: 5 : 20). This was a slight modification of the scheme developed by Coulon-3Morelec, Faure and Mfarechal (1970) for raising antibodies against palmitic ester of glucuronic acid. The amount of antigen used u-as twice as much as the amount used for palmitic glucuronate as our antigen's molecular weight is approximately twice as high. The solutions for injections (for 2 rabbits) were prepared as follows: 10mg of fraction 0-4 RF (either '.aged" at room temperature or stored at -30 ) and 50 mg cholesterol were dissolved in 1 ml ethanol at 70". The solution, while still hot, was injected rapidly into 19 ml saline. "Aged" 0-4 (RF) mixture resulted in
363
a very fine milky suspension. but fraction 0 4 RF (-30c) produced a precipitate which had a tendency to float. Energetic dispersion with a syringe improved the suspension. 2 ml 1000 BSA in saline was added to each solution with energetic mixing. Addition of this protein improved both emulsions, particularly that with 0-4 RF (-30 ° ) fraction. The solutions were ampouled in 2-ml portions, snap-frozen in acetone-solid CO2 mixture and stored at -30 until used. Each rabbit received 10 injections of 1 ml of these solutions ever- other day. Samples of blood were taken after the 6th injection and the rabbits were bled 8 days after the last
injection.
Three out of 4 rabbits were given an extra booster injection a few weeks later, blood being taken after 8 more days. As some rabbits showed a slight shock reaction on the third injection, antihistamine treatment was employed. Each rabbit was given 10 mg of Piriton i.m. 45 min before antigen
mjection. 2. The serond group of 4 an imals.-These animals received antigen combined with lecithin and methvlated bovine serum albumin in the ratio 1: 10 : 16. The immunization scheme employed was that found best for cardiolipin out of those tried by Inoue and Xojima (1967). Four rabbits received courses of IO i.v. injections (Piriton protection being employed from the 4th injection) spaced every 2 days. The total amount of antigen given to each animal was 15 mg. The day preceding the injection the solutions were prepared as follows: for 2 rabbits I mg of fraction 0-4 RF either "aged" at room temperature or kept at -30 were added to 10 mg lecithin. All lipids were originally dissolved in chloroform : methanol mixture 2 : 1. The solvents were evaporated under a stream of nitrogen at room temperature and the residues suspended in 2 drops of water. 4 ml of 0-10o solution of MIBSA in saline was added rapidly from the svringe to each lipid suspension. Multiple aspiration and injection of fluid improved the emulsion, which was left overnight at -' 4 to be used the next day. Rabbits were bled on the 8th day after the last injection, samples of blood also being withdraw-n after the 6th injection. 3. The third group of rabbits received the antigen and helper mixtures incorporated into incomplete Freund's adjuvant injected intra-
dermallv. This was a modification of the method designed successfully by Tal, Taketomi and Yamakawa (1967) to obtain anti-lactoceramide antibodies. The ratios of helpers : lecithin and bo-ine serum albumins were as described in the original paper, but complete Freund"s adjuvant was replaced by us by incomplete
364
E. KASP-GROCHOWSKA AND L. E.
Freund's adjuvant. This was done to avoid introducing yet another antigenic system, as we have already shown that anti-tuberculin antibodies react with streptococci (KaspGrochowska et al., 1972). Out of the 6 rabbits, 3 were injected with a 04 RF "aged" fraction and the other 3 with a fraction stored at -30° until used The mixtures for immunization of 3 animals were prepared as follows: 15 mg of fraction 0.4 RF and 30 mg lecithin were dissolved together in hexane. The hexane was evaporated, lipids dissolved in a small amount (0-2-0 3 ml) of ethanol, and 1-5 ml saline was injected rapidly into it. The mixture was agitated by aspiration and expulsion from the syringe until a fine emulsion was obtained. 0-6 ml of a 10% saline solution of BSA was added followed by 2-3 ml incomplete Freund's adjuvant (Difco) and the mixture was homogenized to obtain a water-inoil emulsion. Each animal received in total 1 ml of the emulsion into 5 sites on the back by intradermal injection. This procedure was repeated 3 or 4 times every 5 weeks, the blood being taken before each reinjection and 5 weeks after the last. Animals were protected with Piriton before injections, starting from the 4th. The total amount of antigen given to each rabbit was 15-20 mg. RESULTS
The ethanol-soluble material of mvocardium recovered from the solution and subsequentlv fractionated has been assessed for antigenic cross-reactivity with antistreptococcal sera in either all or some of the following tests: double diffusion in agar, complement fixation, passive haemagglutination and its inhibition, as well as bv inhibition of macrophage migration in mixed culture of lymphocvtes of rheumatic fever patients with guinea-pig macrophages. The finally chosen purification procedure has been detailed in the Methods section. Problems arose once the material had been recovered from ethanol solution because it was no longer dispersable in water. This could however be facilitated by mixing with 2-5 parts of lecithin or cholesterol in an organic solvent, evaporating the solvent and making a watery emulsion with resulting lipid mixture. The other alternative was to
GLYNN
'.age" the material at room temperature in a solid state. This resulted in changed mobility of the antigen in TLC on silica gel but rendered the antigen perfectlv water-soluble. A. The estimation of cardiac lipid as the antigen in readcion uith antistreptococcal sera The results of the assessment of the ether-soluble material in passive haemagglutination, complement fixation and double diffusion in agar are presented in Table II. Commerciallv available grouping serum and two groups of antistreptococcal sera obtained in our laboratories were used. The sera Nos. 10, 13, 14, 15, 16 and 18 were raised against whole streptococci; the names of the strains are given in the second column. Sera 11 and 12 served as controls; thev came from rabbits which were inoculated with growth medium onlv. Sera 1, 2, 3, 4, 5 and 6 were from rabbits which received the sonicates of streptococcal cell walls (SCW). The sera with the highest titres (Nos. 1 and 2 in the SCWV group) were consequentlv used in most tests of passive haemagglutination or its inhibition in estimating the fractions recovered from silica gel chromatography. Out of the group of sera raised against whole streptococci, Sera 10 and 15, and the commercially available grouping serum were the strongest in passive haemagglutination of the cells coated with mvocardial antigen. WVhen tested in double diffusion in agar, 3 sera out of 6 raised against streptococcal sonicates gave a distinct line of precipitation. These were Sera Nos. 1, 2 and 3, the sera with the highest titres in the passive haemagglutination test. Of the sera raised against the whole streptococcal organisms onlv one, Serum No. 10, gave a faint line of precipitation. The same sera were also the most efficient in fixing complement in the presence of mvocardial antigens. Cardiac fractions at different stages of purification and streptococcus-derived
365
THE ROLE OF ETHANOLAMIN PLASMALOGENE
TABLE II.-The Rests of Cardiac Hapten (ES) Reactions uith Rabbit Antistreptococcal Sera Passive haemagglutination No. of tube A
The titre m comLine in 4x 8A 16x 32x Mx 128x 256x 512x 1024A plement double 4x 8 x 16 x 32 x 64 x 128 x 256 x 512 x 1024 x fixation diffusion 1/128 Strong 1/256 Strong 1/64 Strong 1/128 5 1/16 sCw 6 1/36 sCw 10 631 Weak 1/64 11 gmt nd 12 nd gm 13 nd 6G 14 nd 6G 12 15 1/64 16 nd 12 18 24 nd nd Grouping serum The cells used for passive haemagglutination were coated with 0 -4 ml of 0 50O ES per 0 1 ml of packed formolized erythrocytes. The concentration of ES in complement fixation test was 200 LAg and 0-5% in double diffusion test. * SCW-sonicated streptococcal cell walls; f gm-growth medium; nd = not done. 1
2
3
TmmunNo. of ing serum antigen 1 ScW* 2 SCw 3 sCw 4 sCw
4
6 Serum dilution
8
7
5
9
TABLE III.-The Actitity of Different Fractions of Ethanol Extracts of Human MIyocardium and some of the Streptococcal-derived Material Measured in the Test of Inhibition of Passive Haemagglutination -No. of tube 2
1
3
4
Mlaterial
5
6
8
7
9
10
Dilution of fraction
The
used for Fraction tested 0 20o Neat 2x coating 1 SCew ES ES 50x ENS diluted Acetone extract 2 SCW AcR AcR L Ph 100x diluted U Ph O 1' sonicated streptococcal cell walls 2 SCW ES ES alkali resistant 100 x Fuller (formamide extract of 6M streptodiluted sera
4x
8x
16x
32x -
64x 128x 256x 512x -
-,r-
I
cocci)
1 SCW 50 x diluted
ES
6 M C polysaccharide A' C polysaccharide
ES-Alkali resistant, treated with NaOH 0 -1 AcR Acetone residue.
N
for 30' at 37' and recovered
materials were used as inhibitors in passive haemagglutination of cardiac antigen-coated cells by antistreptococcal sera. Tables III and IV contain results of
on
silica column.
manv of these tests done in the process of identification of cardiac hapten and evaluation of its relationship to streptococci.
E. KASP-GROCHOWSKA A-ND L. E. GLYN\--
366
TABLE IV.-The Activity of the Fractions Recovered from Silica TIC' Plates, Measured in the Inhibition Test of Passive Haemagglutination No. of tube )~~~~~~~~~
The fraction used for inhibition 0-4RF 0-3RF
10
11
256 x
512 x
1024 x
-
-
-
9
8 7 Dilution of inhibitor
1
2
3
4
a
Neat
2x
4X
8
