. L. . 13.3.... : : , a. . 5.29.... ..:.Za...mwmr.mn .293 3.. Liana 3.34. 1- . 32.331... .. . a. s: .9 1.. . z. fin , In..- , .. . . S vwfli.~q1 . v C .31.. 1. 3 1 n .1142}. . . .. I i I ... . . . .. 511.. $214.74.... .3533... liowemmfithWMngfifi m... .5. .nmw an... a. 398.. mfiwflpfixfig .fluzduwitz.fi..ma.k 39%.? kuznflv 3...... . . v . . . u .1. . . -. rrtibv..v..vz . .! $.71?! 0.1 r I u . 5..) .. . . .. ....I.r\ ; o: . 1 .2 r” 1ton.v.v.r0r19r.1\1vvl v. vtvv-v‘xN‘fl! .. 7 V! V I . 1. a... u. . .. .. :r .: . ...v.»..I......,r.wvi 1.71. 4.9.3933... 13.? . . 3&9. L .. . . Sum . “aha..."vmgch. ...uu“m.p.fiqn¥5v.mg.m1.§.wfi§§m~ . ... ta. 2:... . .u....d...nv.fi. u..$..u.$¥ . . .. x .85.. , .2 .wbufiabw ,1 13gflag... z . n} C... .3 . .. .r . M .. .mtfi....w3.5mkfifl€» . . .. . . Haas awhhn... . . 4.... 5......?..sr§.. $32.05.... 3. . ,.¢..J....Lb...2.. 1235.): 2... v‘. -. II .....x...:.:....wd...v.xItgghdfiu. of. :IIO."3I' ‘§7'.‘, ‘fl'f’ I‘ L‘ U 9 04% . .. a .. $ 22.35 .. .. .2 Von. x®my 37.. . V .35. .?xl.)31xf.x....slzw....$...313318ra .. Viv 9.”. EL“. .. . . .J .. a. . .. .. s .ux..,...“.v.u.x.\w........\... §§Qi1245b§.1\flvwmmfl‘u ¢ .5 Lfi5AI. I... I? .. 1m. ‘ Linhufiébnfl) 13.0.8) 31.13:. 3113‘ Kit) 1234.»... 1:51.. £139)». 3. .s N. b: .0..me 7 \... . .. 1.1. .1. .11 .. o? .5 1..) 2.9.0». 3% 1......4. .. 1.31.9.1.“ n. . guyfifvxs...)\hv~krvxggrlz.5... . . ' 1...“. kiln. . .25; a-.. 3...... 3mm, 3%533325. 3.2 ilflnmfifizxam. .uhzl. 1.4.1545 . 2.1. a n .. J I 1.15 v . t '0‘ x . , ‘9. .. kn “Gin“. Hm... .. .. Inxpuxumr)‘. 22$ x>1i)l)r.13\\8n)ifilb . _. . . 041'! v it I :WVV o it: I. v .6. . . > 0.0%.. 9.1%“. . 1.. . )3 a: .. khan.“ .. all): u)8.vo«r\hfl)xhflwfl..5§ fl. ... 1': x.“ .. suntan. vhnrvtiuu ... 1 1 .. 0‘8. Lu. 1 x) xaBgix 5 Sum .. .51}... . . pugiwv)! 3315!...) .. a... . . 1 . . v . c3... . $5). 3...... §)a c) .1 Him. , .\ . )3 , u." ‘73.”! try ‘3‘“..th Lo 1 t. .. whifikibssflmn Mugrfibngua mm“... §H§s\>u$x..§h 355.»... Lulvafififi. .. 3 . . . 7 mt. . . . .\. .. . 3,... \p’hvkr I. 1.. .. (3:4... . 1 1 fig: 2 .u - .32. v. .. . up! .. v?b.. .. . 5...... “mg . . gex.\.r3m§53v$.§ mu...” my... 5 a y . u .- ... r . 1 V I J .. 1 1 >. 7. as - ..:\§a.zx>l:.flh.v; 9.... a I. crut...‘ 0:21.; . r i «9‘ llfiIJ. W51 5)) 1.5.“. n.) .1)“ . J) M. 32).. a... .... . I.’ .{I3' .I .‘a . ua‘ A .713! . . n z u l . . lav! "91,132..“ .. Ix”: ‘2. niun. 9.3.3... 1....M "9&1. ohm. .89 a), 1; safabamm. 31.1339) LvafLuf..§ h. . i. . no.1 .315. . 81...... .9. .u avfiénzxu 2.2.4.». 33:141....3... 3.; .o , . ts . 3 3 hunflrréuh.¥i.l 2:”! v. . .. . 1L... 28M. a. . . £5... 2. a. . .4 arflfidfi. v.1»...vyv‘tfinrflvbftvxf‘vir Vivi}. a I x... 5.5)...) l. 1.91. .Lfiyv.q:fivvflnh?§3¥a {'0' a... )Iq § Z. 3“ hhguu an... .. 1:12.... :2 .2 .1. .. 911 r dvvvflucefl Pvtrusl'.‘ {Reg . I '1 . V 3:. .. t 1* Rf 3 1.53.5} 12.1.... . . a). 1? Q5 RHDQV. hflaz)aaammpi 7 \L t )JHDXIXQNXfiI .. .. gimuxsq Jfimigi 11...}...“4. 11). vs 5......;‘ .5“. 1 2. agonintiti z. . .fl 1. z .v’vvv....r!. :Hotxvo.h’va{§$ {git h... 1 "a . N‘Ifl?l.'¥ Ragwgtflaubflflfzz ilk-RA. VI: . .. .. . .z... uvvfluiiklklxrbnéfi~vrp+ur|it3 7 .- “VIN?! t . . Quanta! .......... gfifit:fi. . v. v. r .. 110......» 7.3.7.... “W a «VI Mn. 1.. . vr 71...... 2... .1 n 3).... .5... . .x 2 c . .3 3.... 2.5.wqaasfia . .l. , 3. . | .u. 1 . . \yin‘agkc... sirqg.’ LWW?...\h....M..m.v1..K.fla£v.uHH. 3 .v1 .539.» .. .A. . .. . 12.. 4mg .. ubrfiec sfi....l...fi§€.b§.§§3 £..u.............2. 2. 83333... )3‘ . 5.1. . 1.. 1 JR: 3 )5) . . y)$\1 3.51.1. \ .. I l I 2.... 6y :2. I. 2. goofy. _ , , 3 .11... 1>i€a§huvl\2.uu§ ...........+. 3.5.... t7.§..)§ii>§l.§.§é.i§ 4. 1a 90 iv. \1 w) 73)....1... a .n x... .4. ah} i n.) .....d:11......).. \Crx..)l...n.l..0np1x1..)>g.. I v . i ‘5 .5‘313.‘ is. 3.45.3... xx: 13.53.11. 7!» .14. I .2 : 11:13)).... > . 7‘. 40.1... 9.. 1. . v v73>vrvvvtcgil11vv§v§vvir) ‘35.... 1.333.}... 1:... ER‘ 523.11..)3 513.4015?» .an 13..., utilauxiqx .xs . 20...”... s k . .rlvl ‘Y’l'fl.v7‘l'lltza‘:i‘ iii! If, :9 a . .\§~M.I C..\ . 245.4 . .1 1 . \ 5 . .\ . ’3‘ \ {to .. g E 1'4 s. a... .89.... . .l.‘ 3.... all..." A.- ..)w(. . (.5... >3 .. K 7) Kickfitflv v.1)... Icky"; .v wltzi.... fig}. V2§§Hfllluui \fiM)- 1h. )2...” I i 2. 3. 1.. 4%.... .fl. .. 5..” A :53}... {wager a... 23... H. J: {a J 3. gram... .2... .24....)dk3h. . .3”... Vlr N g. . .. .. r .fin. 3.2.8 {.vvhrlvzl ‘.. givil'ovlt’r I." . . .x . s. r... .. . 7311...“-..454 . a)... .i .1 y: y r ,. .. a 23 .15.!1 15.15131 .1 72.) .55 .5....\ . r r. v \ ”3.91.3.3... .5121. §.u.5w.:711h&.1..fl)1§\. 1.30.22. $311H145Rdhhawfiyzl. 0.00:; kw...) )xsvf . I? «:5 bu“ as. . ~ .1. .s $.szfix..\531.31u.m 2.520.. (0.24)) r..t......flfit?v?.e..e $3.331. filcfig 793i: . 1.. .5 548... xxguglufllii 3£§.2).512. . . $3.33: iii... . in ..$.§R.fii.. u. .3? 3.. £2.522Lzei z, . '1'.‘t vvvvvvv .9. .9; . '11 o .. 5|. 112.115... glib: R .. ‘ , 3 .... 33,. :I r0... I 1 .l. . v. .. .51 .1......<..¢)...>\slwl.).. . . «(2357).. lxixm‘na... 3v§t)\.}.l).. 7 g‘t Ox} 6313).;Itigx .3‘13. iiizurwi: v... )b..3.fiha.§,20¢.§\.§§xzfi .(2\ ..\..v.¢.:.p?u.s)~llt.‘1)32.11§a31 13%...» n S .59.)...11111uwx‘unlrnitflafilwi 12.42.. . fiflaal)>h§. ;. P}. 1.1).... 1.5. .2! )III)‘ .s. 3H... 12.x; {1.3.1: 1 .. .. .. a 9... y V... 1111 i... .‘1!»..3V.€.Lv11..2‘11. '1 3‘93.) . A 1y\\531\li)3 3.?! 2).“... .. L. .55.... .72..-) i)??? .. . I. . ... >1. . . . \i'.‘ "1"”‘7’1 'yui‘) - 1'." .. ahql ( 2.5L... ... 2.2... . . . .1 ’35 g . 5. \l .I‘ .. . \ a 8‘5321‘1‘153 LN. gainim..flat§uu “:31... R. v $31.; I ll \ I 4%.)... l .Efigiaastza..§\‘$ll§ .. 1.. . . r 1d I . 31:3“? l fi‘glfijg \. . o ..\). I)... “H 312.? 32:53.... 12.). u... “Ian-ll Lil}: I. u. .- 11... .. “NIL... $351.5!” N n 3...... . 31‘ i. I}? ‘0‘: .1! 5”}; I 3"; .4. . a . .. L. :7. 1.7... 7R . » filfiwfllfl. vast... finiwlfiml v..u....&.\yv ”kuu 5 . 1.1.3! . 3;; .k... 51.2....» K... .xrvnltfirfrrufl. 3. “1:15! it]... .95)...." .1“ ”n . .111?! l. . if’gtOILu r... . 31.. as; {$1.32 3...! i... . a 1...... {Lizfiziicfiéff . 3.1.1.1. . L 3... . . .. Igflsfiifi. L. 8.. . fif'17'7'7u - n ’13., + mum" :1'1 L... . P. I“... . on... . .1... I...) Irina“. .ntvirt mhulifhbur§£l Lpdnixi_\flrlul Wt. , jg]... . ‘ I 3 .. ..C-..(7..9 .... .. 1...! ...... 4 .Psidu..?ol§...t.\n. I . . ##1##...7 o. . . u x... $.1......-..r. . ..........J«. . .2 ‘. .§\"vlaé\r 1.115... )3: 4311?. 3.5.1.... 7 .. 5).... . y . mug y Er: University ' THESIS This is to certify that the thesis entitled Thy-l glycolipid modulation of the in vivo immune response presented by Alvin A. Gabrielsen, Jr. has been accepted towards fulfillment of the requirements for V1.5. degree in M Icra biOle Md ‘9»..le Heal Major professor Date 5‘7 7'77 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records ,‘ a." W... - 1'24"» : t-zrsagxwzzaayg! a-' \":.a““’- :‘A‘ ‘Q§§$r”{$“ g; ' an t“. .'a;“‘j.: THY—l GLYCOLIPID MODULATION OF THE ;§_vrvo IMMUNE RESPONSE BY Alvin A. Gabrielsen, Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1979 ABSTRACT THY-l GLYCOLIPID MODULATION OF THE E VIVO IMMUNE RESPONSE By Alvin A. Gabrielsen, Jr. The in 2223 functional role of the Thy-l alloantigen, known to be shed from antigenically stimulated T lymphocytes, has been investigated in this report. Mice were challenged i.v. with sheep erythrocytes (SRBC) followed 24 to 48 hours later by i.v. injections of gangliosides con- taining Thy-l activity (formulated into cholesterolzlecithin liposomes by low frequency sonication at 60 Hz). Both suppressed and enhanced anti- SRBC plaque-forming-cell (PFC) responses were observed in the Thy-l— treated mice in a hemolytic plaque assay. BlOCBFl mice (Thy-1.2) receiving 5 x 108 SRBC demonstrated suppressed IgM PFC responses of 60—81% when treated with CBA GMl ganglioside (containing Thy-1.2). CBA Gle (with some Thy—1.2 activity) and AKR GMl (containing Thy-1.1) gang- liosides also depressed PFC responses to a lesser extent in BlOC3Fl mice. Indirect responses in this system were suppressed by as much as 97%. Exposing mice to gangliosides with Thy-l glycolipid prior to SRBC immuni- zation did not increase the degree of suppression. B6C3F1 mice, given 5 x 107 SRBC, followed by Thy—l treatment demonstrated enhancement of the primary (IgM) immune response. (IgG responses by the day of assay were seldom observed using the above SRBC dose.) This was observed in mice treated with Thy-l-containing liposomes formulated by high frequency (HF; 20,000 Hz) or low frequency (LP; 60 Hz) sonication. Enhancement with HF and LF liposomes was 38% and 100% over controls, respectively, with 6lng doses of GMl (plus Thy-l) given at 24 and 48 hours after antigen. Sixteen micrograms of G incorporated into Ml HF liposomes were able to increase IgM PFC by 100% over controls when given at 24 and 48 hours, while administration of 6 ug doses at 6 and 30 hours after antigen did not significantly increase the degree of enhance- ment seen at that dose given at 24 and 48 hours. In the absence of auxillary lipids (cholesterol and lecithin), GMl (plus Thy—l) was found to exert no effect on the PFC response. The kinetics of Thy-l-dependent enhancement revealed that the effect was measurable in the direct response by day 6 after antigen injection as IgM PFC declined from peak values. In this study, a signifi- cant IgG response was measured by day 5 for Thy-l-treated mice. No significant control IgG PFC were observed until day 7, and then at only 10% of the level of the Thy-l—treated group. The relative enhancement in IgG PFC declined as the responses of both groups increased with time, but was still significantly higher by day 10 in the group treated with Thy-l. These results suggested that Thy—l may suppress or enhance the humoral immune response in mice. Since BlOC3Fl mice treated with Thy-l glycolipid displayed suppression of the PFC response, while enhancement of the response was evident in B6C3F1 mice under similar conditions, the effect was probably partially dependent on the mouse genotype. The kinetics of the enhancement effect imply a more important role of Thy-l in the secondary response than the primary response. DEDICATED TO KARLEEN AND JACOB ii ACKNOWLEDGMENTS I wish to express sincere appreciation to my advisor, Dr. Harold C. Miller, for his guidance, patience, and encouragement in my research and the preparation of this thesis. Thanks are due to the other members of my guidance committee, Drs. Walter Esselman, Filipe Kierszenbaum, and Lawrence Aronson for their helpful advise and encouragement. The suggestions and support offered by my friends and coworkers, Marlize Corréa, Prince Arora, Liang Hsu, P. Narayanan, Barbara Laughter, and Warren Coon made completion of this project a worthwhile and educa- tional experience for me. To them I express sincere gratitude. A special thanks is also due to Mrs. Bonna Davis for her tech- nical assistance in preparation of this thesis. Most importantly, I would like to express my love and appreci- ation for my wife, Karleen, and son, Jacob, who have made great sacrifices on my behalf enabling me to secure an education in the field of my choice. iii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 LITERATURE REVIEW. . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Thy-l Alloantigen. . . . . . . . . . . . . . . . . . . . . . . . . . 3 Introduction . 3 Distribution of Thy-1.. . . 3 Isolation and Characterization of Thy-l Cell Surface Antigen . . 6 Protein nature of Thy-1. . 7 Lipoprotein nature of Thy-1. . 7 Glycoprotein nature of Thy-1 . . 8 Glycolipid nature of Thy-l . . . . . . . . . . . . . . . . .12 Biological Role of Thy-l . . . . . . . . . . . . . . . . . . . . .15 Introduction . . . . . . . . . . . . . . .16 B cell modulation by shed Thy-l antigen. . . . . . . . . . . . .16 Thy-l shedding from T lymphocytes. . . . . . . . . . . . . . . .19 Role of T Lymphocytes in Humoral Immune Response Regulation . . . . . . . . . . . . . . . . . . . .23 Lymphocyte Interactions. . . . . . . . . . . . . . . . . .23 Mechanisms of T cell Regulation of the Humoral Immune Response. . . . . . . . . . . . . . . . . . . . .24 Cell- to-cell contact . . . . . . . . . . . . . . . . . . . . .24 Soluble mediators of regulation. . . . . . . . . . . . . . . . .24 Soluble T helper factors . . . . . . . . . . . . . . . . . . . .25 Soluble T suppressor factors . . . . . . . . . . . . . . . . . .27 I2_vivo applications . . . . . . . . . . . . . . . . . . . . . .30 MATERIALS AND METHODS. . . . . . . . . . . . . . . . . . . . . . . . .34 Mice . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 Sheep Erythrocytes . . . . . . . . . . . . . . . . . . .34 Hemolytic Plaque-forming Cell Assay. . . . . . . . . . . . . . . . .34 Isolation of Glycolipids . . . . . . . . . . . . . . . . . . . . . .36 Liposome Preparation . . . . . . . . . . . . . . . . . . . . .37 Electron Microscopy of Liposomes . . . . . . . . . . . . . . . . . .41 Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . .41 iv RESULTS. Attempts to Modulate Antibody Responses In Vivo with Thy—l . . . Effect of Thy-l Administration Schedule. Thy—1 Enhancement of the Anti- SRBC Humoral Immune Response in vivo. . . . Thy-1 dose response-liposome morphology. Requirement for liposome form of Thy-1 . Kinetics of Thy-l-influenced enhancement in_vivo . DISCUSSION . BIBLIOGRAPHY . . . . . . . Page .42 .42 .44 .46 .46 -52 .54 .57 .7O Table LIST OF TABLES Suppression of anti-SRBC response in_vivo by mouse brain gangliosides incorporated into liposomes. Suppressive effect of Thy-l in_vivo on anti-SRBC response when administered in a single regimen prior to, simultaneous with, and after antigen. Auxillary lipid requirement for enhancement of anti-SRBC response by Thy—1 glycolipid in vivo. vi Page .43 .45 ~53 Figure 1. Effect of sonication on liposome morphology . 2. Effect on anti-SRBC PFC of varying doses of Thy-1.2 glycolipid contained in HF and LF liposomes . 3. Dose response of HF liposomes containing Thy-1.2 glycolipid: effect of shift in post-antigen administration schedule . . . . . . 4. LIST OF FIGURES Kinetics of the effect of Thy-1.2 glycolipid on the anti-SRBC response in vivo . . . . vii Page , 39 .48 , 50 . 55 ABShu ABSm AEF ATS BGG BRBC BSA cer C:L C:M Con A CRBC CT DNP Gal GalNAc GAT Dla le Glc Ml HF LIST OF ABBREVIATIONS anti-human brain antiserum anti-mouse brain antiserum allogenic effect factor anti-thymocyte serum bovine gamaglobulin burro red blood cells bovine serum albumin ceramide cholesterol : lecithin chloroform : methanol Concanavalin A chicken red blood cells cholera toxin dinitrophenol galactosyl N-acetylgalactosaminyl L-glutamic acid6O — L - alanine3O - L — lysinelo Ga1(NANA)—GalNAc-Gal(NANA)-G1c-cer Gal-GalNAc-Gal(NANA-NANA)-G1c—cer glucosyl Gal-GalNAc-Gal(NANA)-Glc-cer high frequency (20,000 Hz) viii LF LPS MBSA MEM MIF MuMTV NANA NP-4O OVA PBS PFC PHA POL PRBC SDS-PAGE SIRS SRBC (T,G)-A--L TLC TNP TT immunoglobulin intravenously keyhole limpet hemocyanin low frequency (60 Hz) lipopolysaccharide methylated bovine serum albumin minimal essential medium migration inhibition factor mouse mammary tumor virus N—acetylneuraminyl (sialic acid) Nonidet P-40 ovalbumin phosphate buffer saline plaque forming cells phytohemagglutinin polymerized flagella pigeon red blood cells sodium dodecyl sulfate polyacrylamide gel electrophoresis soluble immune response suppressor sheep red blood cells poly (tyr-Glu)-poly DL Ala--poly-Lys thin layer chromatography trinitrophenol tetnus toxin INTRODUCTION Since its discovery, the Thy-1 alloantigen has become a conveni- ent cell membrane marker for the study of thymus-derived lymphocytes (T cells) in the murine system. It has provided a basis for identifi- cation of the cell types and the cellular interactions that appear to be necessary in order for a host to mount a successful protective response to a foreign antigen. The following literature review has been organized to provide background information on the Thy-1 antigen, summarizing the details of its discovery, the attempts that have been made to characterize it, and the ig;32532_studies that have suggested a possible functional role for Thy-1 in regulation of the immune response. In addition, the role of Thy-l bearing cells in regulation of the humoral immune response has been briefly reviewed here. This was done to put the functional role of Thy-l in perspective with the multitude of regulatory factors derived from T cells upon antigenic stimulation. The ultimate test of an in vitrg_phenomenon is the demonstra- tion of its in zizg_correlate. .Inngi££9_studies Showed that a glycolipid with Thy-1 activity exerted a temporary suppressive effect upon the antibody response to a given antigen, followed by an enhanced response to that antigen. The present study was undertaken to investigate the possibility that the Thy-1 glycolipid might play a similar role in_zizg, Initial experiments looked at the effect of Thy-1 on the development of 2 antibody-producing cells in the spleen in a regimen of antigen and Thy-l administration based on the in 31532 studies. Variations in the antigen and Thy-1 administration schedules were also investigated for their effects on antibody production. The importance of the combination of Thy-1 glycolipid with auxillary lipids for modulatory activity, and the nature of the resultant liposomes was also studied. Finally, the kinetics of the effect of the Thy-1 glycolipid on the primary and secondary anti- body responses were investigated. The results of the above studies are discussed in terms of a model of immune regulation proposed from the earlier in_vi££2_studies. The role of the auxillary lipids and liposomes into which Thy-l was incor- porated for this study are discussed in relation to the method employed for assessing the in_vivo effect of Thy-1. LITERATURE REVIEW Thy-1 Alloantigen Introduction Reif and Allen (1) first observed in 1963 that multiple injections of AKR mouse thymic lymphocytes into C3H mice elicited a strong cyto- lytic response against AKR thymocytes, despite H-Z histocompatibility (1). Further investigation, using the reverse procedure, revealed the existance of an alloantigen in C3H mice (2). These antigens have been designated Thy—1.1 and Thy-1.2 (previously G-AKR and 6 C3H, respectively - see ref. 2 and 3). Anti—Thy-l.1 and anti-Thy-l.2 alloantisera were prepared and tested against a large variety of mouse strains (2,4). Results showed that most inbred mouse strains carried the Thy-1.2 alloantigen on their thymocyte membranes, while only a few strains demonstrated Thy-1.1 specificity (2). All strains of mice so far studied, carry one or the other of these two allelic forms, which are coded for on chromosome nine of the mouse (5). Distribution of Thy-1. Reif and Allen (6) were unable to demonstrate significant amounts of the Thy-l antigen in a variety of mouse tissues examined. Other murine tissues were tested for their abilities to absorb out the _L\ anti-Thy-l activity of the alloantisera and little or none was found on bone marrow cells, in neonatal brain, appendix, lung, liver, skeletal muscle, kidney, and testes (2). Thy—1.1 and Thy-1.2 were found to increase from low levels in neonatal mouse brain and lymphoid tissues to maximum levels by 5 to 6 weeks of age (6,7,8). Since the discovery of Thy-1 in the mouse, it has been identified in a number of nonlymphoid tissues (6,9-15), on T lymphocytes in various lymphoid organs (7-9,16-18), on several lymphoblastoid cell lines (2,21), and in species other than the mouse (14,22—28). In the mouse, Thy-1 activity has been demonstrated in peripheral nervous tissue (2) and brain tissue (2,6,9—11), epidermal cells (12), normal and neoplastic mammary tissue (13), fibroblasts (14), and thymus epithelium.(15). Thy—l-bearing lymphocytes are found in varying relative quantities in the thymus (7,8,16), thoracic duct lymph (7), lymph nodes (7,9,16), spleen (7,16), peritoneal cavity (7), Peyer's patches (7), and peripheral blood (17,18). Rapid appearance of Thy-1 positive lymphocytes in murine bone marrow after treatment with thymic factor (19), Vibreo cholerae neuraminidase (19), or thymopoetin (20), suggests the presence of precursor T lymphocytes in the bone marrow carrying sequestered Thy-l antigen in their plasma membranes. In their early studies, Reif and Allen (2) discovered the presence of Thy-l on the surface of many leukemias, distinct from other known characteristic tumor antigens. Zwerner and Acton (21) identified further lymphoblastoid cell lines carrying either the Thy-1.1 or Thy-1.2 alloantigen. The presence of a Thy—l-like antigen in rats was reported in 1972 by Douglas (22). Absorption and direct complement-dependent thymocytotoxicity studies demonstrated a strong cross reactivity with the Thy-1.1 antigen of the mouse, while little Thy-1.2 activity was observed (22). Further investigation has revealed that Thy-1.1 is found in rat brain, thymic tissue (23), and on rat fibroblasts (14), comparable to the mouse system, while little is expressed on peripheral T lymphocytes (23). Thy-1.1 is abundant on rat bone marrow cells, being expressed on perhaps as many as 50 per cent of cells tested (24). Surface immunoglobulin, a B lymphocyte marker, is present on about 25 per cent of the Thy—1.1 positive rat bone marrow cells, suggesting that Thy-l is not unique to "thymus-derived" lymphocytes in the rat (24). Arndt et a1. (25), investigating human tissue, reported anti- mouse Thy-1.2 activity in rabbit anti-human brain antisera (ABShu). The ABShu cytotoxicity to CBA (Thy—1.2) thymocytes could be absorbed out with human, rat, and CBA mouse brain. Conversely, anti—mouse brain antisera (ABSm) activity could only be removed by absorption with mouse brain. Cocapping experiments using mouse anti-Thy-l.2 and ABShu antisera, followed by fluorescent labeled goat anti-mouse globulin and goat anti- rabbit globulin, demonstrated the presence of two closely linked anti- genic determinants for Thy-l, one of which was proposed to be species— specific, and the other nonspecies-specific (25). The nonspecies-specific determinant, previously described in human and mouse brain (25), was then found to be shared by human thymus (26), while the species-specific determinant could also be identified on human thymus, peripheral blood lymphyocytes, and bone marrow cells (26). Isolation and Characterization of Thy-1 Cell Surface Antigen Owing to a variety of techniques employed to isolate and charac- terize the Thy-l antigen, the exact physiochemical nature of this moiety is still somewhat in question. Researchers have attempted to isolate Thy-l from mouse brain tissue (9,29-31), T cells and/or thymocytes (4,29,30,32-36), lymphoblastoid lines (35,37-39), rat thymocytes (27,28), and rat brain (28). These tissues have been subjected to a variety of techniques designed to purify and identify Thy-l. Common procedures include detergent solubilization (27,31,35,36), followed by immunopre- cipitation (31,35) or gel filtration and affinity chromatography (27,36). Thy-l has been radiolabeled by 125I (31,34) and 3H amino acids and car- bohydrates, followed by immunoprecipitation (31). It has been extracted by organic solvents after physical disruption of cells (30,31), or been subjected to enzyme solubilization followed by organic solvent extraction (38). Further purification steps have involved density gradient (31) and zonal gradient centrifugation (9), thin layer chromatography (30,40), isoelectric focusing (29), and separation by electrophoretic mobility (27,28,34-36). Biological activity before, during, and after combinations of the above techniques has been monitored by inhibition of thymocyto- toxicity of heterologous and alloantisera to Thy-1 (30,38,41), ability to induce specific anti-Thy-1.1 or anti-Thy-l.2 activity in spleen cell cultures (42), and precipitability by specific anti-Thy—l antisera (31,34,35). In keeping with the various methods used to determine the nature of Thy-l, different conclusions have also been drawn. Thy-l has been characterized as a protein (34,38), a lipoprotein (4,29), a glycoprotein (9,27,28,36,37,41), or a glycolipid (30,31,33,40,43,44). Protein nature of Thy-l. Lactoperoxidase radioiodination of thymocyte cell surface proteins, followed by extraction with 10 M urea plus 1.5 M acetic acid, was the method employed by-Atwell et a1. (34), to begin Thy-1 isolation. Thy-1 was then immunoprecipitated in a double antibody system reacting mouse anti-Thy-l antiserum with the thymocyte extract, followed by excess rabbit anti-mouse Ig. The precipitate was examined by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) which suggested that Thy-l activity resided in a protein moiety of 60,000 molecular weight. Kucich et a1.(38) found that the supernatant of crude papain- solubilized S-49.1-TB.2.3. murine lymphoblastoid cells had anti-Thy-l.2 inhibiting properties as measured by a 51Cr-release cytotoxicity inhibition assay. Further digestion by crude papain diminished Thy-1.2 activity. Similar results with crystallized papain excluded the possibility that glycosidases or lipases were responsible for the decreased activity. Prolonged exposure to proteases destroyed the Thy-1.2 activity as did extraction of the supernatant by chloroform and methanol. This evidence was interpreted to suggest that Thy-l was protein in nature (38). Although evidence of the protein nature of Thy-l had been pre— sented, the possibility of carbohydrate or lipid involvement had not been ruled out. Lipoprotein nature of Thy-1. Possible lipid involvement in Thy-l activity was first suggested by Reif and Allen (4) in a study of the nature of the Thy-1.2 antigen present on C33 thymocytes. Solubilization of thymocytes was followed by dialysis and centrifugation at 35,000 x g. On the basis of unit nitrogen content, the microsomal fraction of the thymus homogenate was 19 times more potent than whole homogenate when tested for absorption of AKR alloantiserum against C3H thymocytes. This nondialyzable activity was not recovered after treatment with chloroform: methanol (2:1) and other lipid solvents. These results were interpreted as indicating that this membrane-associated molecule was lipoprotein in nature (4). Arndt et a1. (29) isolated Thy-l from mouse brain and thymocytes and studied it in an anti-Thy-l cytotoxicity inhibition system. Brain or thymocytes were first solubilized in the nonionic detergent Nonidet P-40 (NP-40). Subsequent gel filtration determined the molecular weight to be about 35,000 daltons. Isoelectric focusing gave two peaks at pH 4.5-5.0 and pH 5.4. Organic solvent extractions of the thymocytes in acetonezchloroform and acetone only resulted in an 80 per cent loss of activity, while extraction in acetonezwater left most of the activity. In all three extractions, the residual activity was associated with the extracted proteins, but could be linearly increased by addition of the extracted lipids or NP-40. These investigators concluded that Thy-l was a protein requiring nonspecific lipid interaction for Thy—l activity (29). Glycoprotein nature of Thy-l. A number of investigators who have studied the Thy-lantigen of murine system and its analogue in the rat system have concluded that Thy-l is a low molecular weight glycoprotein (9,27,28,35-37,4l). Trowbridge et al.(35), studying several mouse T lymphoma cell lines, isolated a molecule of 25,000 to 30,000 m.w. which they reported to be identical to Thy-1.1. The cells were lactoperoxidase radio- iodinated, detergent solubilized, treated with crude rabbit anti-thymo- cyte serum (ATS) or ATS absorbed to be made specific for T cell antigens. Biosynthetic labeling studies, sensitivity to proteolytic enzymes, and affinity for plant lectins suggested that this molecule was glycoprotein. No radioactivity was precipitated from NP-4O lysates of cells labeled with 3H-palmitate using a variety of antisera, suggesting that Thy-l was not associated with lipids. The Thy—l-active antigen was precipitated by rabbit antisera against mouse brain, rat brain, and rat thymocytes, and was not detected on the surface of Thy-1 negative variants of mouse T lymphoma cell lines. Rabbit antiserum to ”purified" rat brain Thy-1.1 demonstrated the ability to recognize and bind their "purified" molecule. Mouse anti-Thy—l antiserum failed to precipitate any 125I-labeled cell surface moiety on the T lymphomas studied, though this was attributed to low affinity antibodies in these antisera or disruption of "multivalent cell membrane interactions" by the presence of the detergent (35). Another T lymphoma cell line, S—49.l-TB.2.3 (Thy-1.2), was sub- mitted to a limited digestion by papain and the solubilized material partially purified by passage through Sephadex G-200 (37). A single peak was eluted which contained all the inhibitory activity of a 51Cr- release cytotoxicity inhibition system using 51Cr-labeled S-49.l cells and AKR anti—C3H Thy-1.2 antiserum. Trypsin digestion diminished the inhibitory activity as did treatment with neuraminidase. Since this enzyme cleaves sialic acid residues, the proposal was set forth that lO Thy-l was glycoprotein in nature. Furthermore, sialic acid alone was found to inhibit the anti-Thy-l.2 cytotoxicity for S-49.l cells, prompting the suggestion that sialic acid was part of the antigenic determinant of the Thy-1.2 alloantigen (37). Zwerner and his associates (41) also used a T lymphoma cell line in their studies on large scale production of Thy-l for detailed structural and functional analysis. They used the cell line BW 5147, a spontaneously derived AKR T lymphoma, expressing large amounts of the Thy-1.1 alloantigen. The cells were mechanically disrupted, acetone precipitated, subjected to gel filtration and SDS-PAGE. The "purified” Thy-1.1 was evaluated by a thymocytotoxicity inhibition assay. Based on this study they reported that Thy-1.1 was a glycoprotein of about 25,000 m.w. (41). In accord with previous findings, McClain (9) isolated and studied the composition of Thy-1.1 and Thy-1.2 in mouse brain tissue, reporting that about 20 per cent of the molecule was carbohydrate. Amino acid and carbohydrate composition data revealed the alloantigens to be almost identical, suggesting the antigenic distinction between Thy-1.1 and Thy-1.2 to be too subtle to be detected by compositional studies alone (9). Finally, Barclay et al.(27,28,36), isolated and characterized Thy-1 from rat brain, rat thymocytes (27,28), and mouse brain (36). Their method employed solubilization of crude thymocyte and brain cell membranes in deoxycholate, gel filtration on an upward flowing Sephadex G—200 column, and affinity chromatography on a lentil lectin column, or affinity chromatography of the deoxycholate extract of thymocyte ll membranes on an anti-brain-associated Thy—1 antibody column. A final purification was performed by SDS-PAGE. Rat brain Thy-1.1 was determined to be a glycoprotein of 24,000 m.w. Rat thymocyte Thy-1.1 was isolated in two forms: a moiety of 25,000 m.w. that bound readily to the lentil lectin affinity column, and a molecule of 27,000 m.w. that passed through the column. Amino acid analysis indicated a high degree of similarity between the three forms of Thy-1.1. All three molecules contained approximately 30 per cent carbohydrate, althOugh, in disagreement with McClain's findings (9), analysis of the carbohydrate composition showed great dissimilarities in relative amounts of various sugars (28). The Thy-l activity of these "purified" molecules, as assayed by a radio- active binding inhibition assay, was destroyed by treatment with pronase. Therefore, they proposed the antigenicity of the Thy-1.1 alloantigen to be found in the protein portion of the molecule (27,28). This evidence is inconclusive since pronase is usually contaminated with other enzymes. Employing the same techniques, the same group (36) examined mouse brain Thy-1 antigens. Antigenic activity was associated with a glyco- protein of 25,000 daltons, estimated by SDS-PAGE. No carbohydrate hetero- geneity was detected in mouse Thy-l by the separation methods employed. The amino acid composition was similar to that of the rat brain (28). Antigenic studies indicated that heteroantisera raised against the mouse brain Thy-l glycoprotein was strongly crossreactive with the rat brain Thy-l moiety and did not discriminate between Thy-1.1 and Thy-1.2 allo- antigens. As with some other studies, the presence of detergents inactivated Thy-l, destroying its ability to block the cytotoxicity of Specific anti-Thy-l.2 antisera. Removal of the detergent only partially 12 restored activity. This finding was interpreted to mean that anti-Thy-l antibodies were probably of low affinity, requiring "multivalent antigen for effective displacement of binding to a target cell where the antigen density is high" (36). It also allows for the possibility that Thy-l activity is associated with a lipid or glycolipid. Glycolipid nature of Thyel. A number of researchers, in their attempts to characterize Thy-l, have reported that Thy-l is sensitive to exposure to lipid solvents and detergents (4,29,31,35,36,38). Reif and Allen (4) described a membrane bound antigen with Thy-l activity that was sensitive to organic solvent extraction. Kucich et al.(38) submitted their Thy-l preparations to a Folch partition and found that the "Thy—l protein” had lost its Thy-l properties. Arndt and his co-workers (29) also reported a loss of Thy-l activity from their preparations upon chloroformzacetone extraction, but activity could be restored by adding back the extracted lipids or a small amount of the nonionic detergent NP-40. In contrast to results reported by Arndt et al. (29), Trowbridge et a1. (35) were unable to specifically precipitate any Thy-l from lysates of NP-4O or deoxycholate-treated thymocytes or T lymphoma cells. Letarte and Meghji (36) encountered a similar problem where Thy-l, isolated in the presence of deoxycholate, was not inhibitory in an anti-Thy-l cytotoxicity assay. Both groups attributed this phenomenon to low affinity antibodies that required multivalent antigen for an effective precipitating complex to form (35,36). Even so, removal of the detergent restored only a small amount of the original activity (36). None of these groups tested for Thy-l activity in the lipids removed by their procedures. l3 Vitteta et a1. (31) also had difficulty with the precipitation of Thy-l in NP-4O lysates of murine thymocytes and T cells. The cell sur- faces were radioiodinated and, while specific antisera were capable of precipitating H-2 and TL antigens, Thy-l activity could not be removed from the NP—4O lysate by congenic anti-Thy-l alloantiserum. If the ly- sates were prepared by freeze—thawing, labeled Thy-1 reacted with the specific antibodies. It was proposed that the majority of the radio— activity was present in a molecule noncovalently bound to the antigenic portion of the Thy-l complex and/or the antigenicity of the complex depended on a lipid moiety dissolved by the detergent. A Folch partition of the NP-40 soluble material resulted in greater than 50 per cent recovery of the radioactivity in the lower (chloroform) phase where labeled lipids and lipoproteins were found. Sedimentation studies showed that Thy-l sedimented at a lower density than protein, suggesting lipid association with the Thy-l complex. These researchers also found that they were able to label Thy—1 with 3H—galactose, but not 3H-leucine or 3H-tyrosine. Polyacrylamide gel electrophoresis of the 3H—galactose-labeled moiety gave a broad spectrum of radioactivity peaking at 35,000 m.w., suggesting protein and glycolipid participation in the Thy-l complex. It was proposed that Thy-l might be glycolipid in nature, or that the antigenicity of Thy-l was associated with a glyco- lipid (31). In accord with this finding, Rabinowitz et al.(33) reported that boiling of the mouse thymus destroyed the antigenicity of TL and H—2 antigens but not of Thy-l, an expected result if Thy-l were a glycolipid or if its antigenicity were carried by the carbohydrate portion of the molecule. 14 More evidence for the glycolipid character of Thy-l was provided by Miller and Esselman (30,40,43,44). They reported that GMl and GD gangliosides were capable of inhibiting the cytotoxicity of specific lb anti-Thy-l antisera and anti-brain-associated-Thy-l serum (30,40). These gangliosides were found to have Thy-l activity. Further support that the antigenicity of Thy-l was carried in the carbohydrate portion of the molecule was the finding that pentasaccharides derived from the ganglio- sides were able to specifically inhibit anti-Thy-l thymocytotoxicity (45). GMl ganglioside is known to function as a cell surface receptor for cholera toxin (46). Thiele et al. (47) reported experiments in which fluorescent-labeled cholera toxin (CT) and anti-Thy-l antibodies induced cocapping on thymocyte cell surfaces. Since pretreatment of the cells with CT did not inhibit anti-Thy-l-dependent cytotoxicity, it was proposed that the CT receptor and the Thy-l antigenic determinant were on the same molecule but represented spacially separated sites (47). Contra- dicting results have since been published by DeCicco and Greaves (48) claiming that cocapping cannot be induced by treating murine thymocytes with anti-Thy-l and CT carrying fluorescent probes. Furthermore, "pure" anti-GMl antibodies have been shown to react equally well with murine thymocytes of either Thy-l allotype and any H-2 specificity (49) as well as with spleen and lymph node cells bearing Ig surface markers (32). Recent evidence from the same laboratory that implicated the biological identity of Thy-l and certain gangliosides (30) has emerged which seems to clear up this controversy. Their further studies suggest that the Thy-l-active glycolipid is present in very small quantities in and GD in the previous experiments using these ganglioside prepa- GM1 lb rations (42). In earlier experiments, brain tissue or thymocytes were 15 mechanically disrupted, submitted to a Folch partition, followed by a mild base hydrolysis of the ganglioside-rich upper phase, dialysis, and lyophilization. This material was applied to an Anasil S column utilizing a series of solvent systems of chloroform:methanol:water mixtures. Fractions were collected and submitted to thin layer chromatography (TLC) in a chloroformzmethanol:ammonia system (30). Isolation of thymocyte gly- colipids omitted the column chromatography separation step. Further fractionation in another TLC system of chloroformzmethanol:water resulted in the isolation of a minor component from G which contained all the M1 Thy-l activity. This activity was assessed in an antigenicity assay by its ability to induce an anti-Thy-l response in allogeneic spleen cell cultures (42). Disagreement still remains as to the glycolipid or glycoprotein nature of the Thy-l alloantigen. The variety of approaches to solving this problem has often increased the confusion and led to the reporting of results in conflict with those already in the literature. Whether the antigenicity lies in a protein or lipid, or a carbohydrate moiety that can be shared by either, as with the blood group antigens (50), is an area open to further investigation. Recent evidence, however, suggests that the antigenicity resides in the carbohydrate structure of Thy-l. Wang et al.(42) demonstrated an anti-Thy-l response when either Thy-l glycolipid or Thy-l glycoprotein was incubated with spleen cells of the opposite allotype. The reacting spleen cells were assayed for anti-Thy-l plaque-forming cells on a lawn of thymocytes exhibiting the Thy-l allo- type of the isolated inducing antigen. 16 Biological Role of Thyel Introduction. The discovery of a biomolecular complex unique to a system raises several interesting questions and warrants intensive research to discover their answers. The uniqueness of a molecule to a system or a cell type, its ontogeny, its biophysical properties, its location on or within a cell, represent information which must be explored for important clues to the teleology of the complex. To this point in time, the majority of those investigations of the Thy-l allo- antigen have concentrated on documenting its presence, nature, and its uniqueness as a T lymphocyte marker. It has been identified in mouse (1,4), rat (22), and possibly human tissue (25). Although it can be serologically identified in other tissues, it has been found to be a convenient T cell membrane marker in the mouse (16), and its appearance during T lymphocyte maturation has been studied (8,19,20,51). A variety of products elaborated from lymphocytes carrying this marker is reviewed in a later section of this paper. Few investigators, however, have con- fronted the problem of what biological role Thy-l may play, if it has a function beyond that of its involvement in T lymphocyte differentiation and maturation. B cell modulation by shed Thy—l antigen. An interesting model of immunugenesis has been proposed in which a T cell factor interacts with the lipid membrane of the B lymphocyte, directing its differentiation toward eventual antibody production (52). This is also considered a possible mechanism of interaction of LPS with B cells, in which the lipid A portion of the molecule interacts lipophilically with the B cell 17 membrane. Miller and Esselman (43,44) have proposed a similar mechanism for the interaction of a T cell product, a glycolipid with Thy-l activity (30), with B lymphocytes. This model begins with antigen-stimulated T cells releasing membrane fragments containing Thy-l which then interact nonspecifically with surrounding B cells and protectively block direct antigen binding to antigen-specific B cells, a condition that would result in tolerance and/or unresponsiveness. The proposed regulation is temporary until T helper factors can accumulate and become involved in the response (43). Indeed, studies have shown that injections of near- syngeneic thymocytes are able to prevent the induction of B cell toler- ance to a T-dependent antigen in nu/nu (congenitally athymic) mice (53). Similar results were reported in in_yi££g_studies (54). It was discovered that prior addition of polimerized flagella (POL) could protect B cell cultures from tolerance induction due to large doses of a T-dependent antigen (54). POL may have provided steric hindrance disallowing direct antigen binding to B cells, thus protecting them from tolerance. Miller and Esselman (40) provided additional support for their model when they extracted glycolipids that exhibited Thy-l activity from mouse brain. These glycolipids which comigrated with GMl and Gle gangliosides on TLC (using earlier technology), were then prepared into liposomes and incubated with spleen cell cultures for five days with sheep erythrocytes (SRBC) as the test antigen. Subsequent hemolytic plaque-forming cell (PFC) responses to SRBC were evaluated and sub- stantial suppression of the anti—SRBC PFC response in the treated cultures was noted when compared to the controls. The effect was most dramatic when Thy—l glycolipid was added on the first day of the culture, 1 | 18 although some suppression was seen if it was added on days 2 and 3. Fur- ther investigations were conducted (43) in which mice were immunized with a high dose of antigen (109 SRBC), their spleens removed two days later, enriched for T cells by passage over glass and nylon wool columns, and cultured for three days. When added to bone marrow-thymocyte cultures, the 3-day culture medium was able to suppress anti-SRBC PFC responses. Nonspecificity of this phenomenon was demonstrated in normal spleen cell cultures stimulated by burro red blood cells (BRBC) or trinitrophenylated- SRBC (TNP-SRBC). Addition of the suppressor medium from the T lym- phocytes, originally stimulated in_yizg_by SRBC, abrogated the PFC response to both heterologous antigens. Extraction of the glycolipids from the medium resulted in the isolation of a glycolipid with GMl mobility on TLC that had nonspecific suppressive properties in a PFC response assay. The bone marrow-thymocyte cultures were maximally sup- pressed by day 12 for the IgM response, and day 11 for the IgG response. When PFC responses were measured 48 hours later, they had increased to peak levels beyond the peak responses of the control cultures. Hence, the suppression was only temporary and was followed by an increased response in the Thy-l-treated cultures (43). The suppressor activity in the supernatants of the SRBC-primed T cell cultures could be absorbed out with anti-Thy-l and, to a lesser extent, with anti-GMl antisera, allowing the PFC response to return to the level of the controls (43). Corréa et. al. (55) have studied a system of antigenic competition in which Thy-l glycolipid appears to play a key role. Spleen cell cultures from TNP-bovine gamma globulin (TNP-BGG)-primed mice were immunized with ovalbumin (OVA) and then given a second antigen, TNP-BGG, 19 24 hours later. This resulted in a weak anti-TNP hemolytic plaque response compared to controls that did not receive OVA. This apparent state of antigenic competition was overcome by absorbing the culture medium with anti-Thy-l or anti—GMl antisera. The culture media of various treated and control groups were submitted to a lipid extraction, and a glycolipid with Thy-l activity was found to be responsible for the suppressive activity (55). Further investigation showed that if B cell cultures were allowed to react directly with large amounts of antigen (SRBC), followed by addition of T cells, they were unresponsive in an anti-SRBC PFC assay. Hewever, if the cultured B cells were treated with Thy-l glycolipid first, followed by antigen and then T cells, or the T cells were allowed to react with the antigen before antigen and T cells were added to the B cell cultures, a near normal PFC response ensued. The results presented by these workers provided support for the proposal that antigen-stimulated T cells release Thy-l as a modulatory signal, pre- paring B cells for further stimuli leading to terminal differentiation (55). The ultimate test of an in zitrg_phenomenon is its applicability to an in vizg_system. While the above studies on the functional role of Thy-l are instructive, it has yet to be established whether this is an in_vitro artifact or, indeed, an in_vivo reality. Thyél shedding from T lymphocytes. Shedding, the phenomenon of selective release of membrane constituents into the extracellular environment from viable cells (56), is seen as an important process in release of Thy-l antigen from lymphocytes (43,44,57,58). Whether mem- brane elimination is a normal maintenance process in the turnover of membrane constituents, or serves a more important role is not fully 20 understood. Little is known about the mechanism of shedding. It has been suggested that surface proteases are involved or that clasmatosis, the pinching off of microvilli, is important in the process (56). The temperature dependence of plasma membrane shedding implies enzymatic or metabolic control (56). Release of membrane—bound constitutents has been described in a number of malignant (39,59-63,69) and nonmalignant cell types (43,44,58, 64-68). Shedding of surface complexes has been documented in human (59) and murine (63) cells. Several mouse T lymphoma cell lines (39) and at least one human T lymphoma (69) are known to rapidly shed their surface antigens. Correlations between rates of shedding, rates of tumor growth, immune response suppression, and tumor metastises have led to the sug— gestion that shedding is an important mechanism for tumor escape from immune destruction (60-62). Shedding has been described in some nonmalignant cells and appears to be a normal process under certain conditions. Embryonic and fetal cells shed cell surface antigens rapidly, perhaps to avoid immune destruction by maternal antibodies before they can bind to embryonic tissue (61). Lutz et a1. (64) reported that under conditions of ATP depletion, aged human red blood cells (HRBC) shed surface complexes in the form of small liposomes. Selectivity of the process was evidenced by the fact that normal amounts of all membrane proteins were present in the liposomes except spectrin, a major HRBC membrane-bound protein, which was absent (64). Considerable attention has been given to a variety of biomole- cules elaborated from various lymphoid cell types. Vitteta and Uhr (68) 21 used lactoperoxidase radioiodination of cell surface antigens to observe the release of cell surface Ig noncovalently bound to a fragment of plasma membrane. Shedding was selective as H-Z antigens were not shed during the short term incubation used to detect the shed Ig. Cone et a1. (65), used the same labeling technique with thymocytes, bone marrow cells, and thoracic duct cells to study lymphocyte shedding. Use of metabolic inhibitors indicated that cellular respiration and protein synthesis were indispensible for this process (65). In yet another study, radioiodinated T cells were shown to release 90% of the surface label in the first hour of incubation in culture, while maintaining 85% viability (67). The radioactive shed complexes could be absorbed by Phytohemagglutinin (PHA) or Concanavalin A (Con A) treatment of the medium. Treatment of the cells with PHA or Con A increased the rate of shedding. Absorption of the media with anti-PEA or anti-Con A antisera resulted in the precipitation of complexes consisting of PHA or Con A and their respective shed mem- brane receptors (67). Differential shedding from T lymphocytes was described in radio- labeling studies conducted by Vitteta et. al. (58). By labeling murine thymocytes with 125I, 3H-galactose, or 3H-leucine, they were able to show preferential shedding of Thy—l antigen from T cells while H—2 antigen was either not shed or was eliminated at a much lower rate. These authors proposed Thy-l to be a peripheral membrane moiety, while H-2 was prob- ably a more integral membrane constituent (58). Lake (57) reported the induction of a primary anti-Thy-l response in_zi££g_by medium from a murine T cell culture of one allotype when it was introduced to a spleen cell culture of the other allotype, as assayed by an anti—thymocyte 22 plaque technique. This was interpreted as evidence that Thy—l antigen was shed by thymocytes in culture (57). Finally, Miller and Esselman (43,44)) reported that medium from a T cell culture, prepared from mice primed in_yiyg_with a high dose of 9 SRBC), contained Thy-l activity in glycolipid form. Further- antigen (10 more, in antigenic competition studies, medium from OVA-stimulated cul— tures of spleen cells previously primed in_zizg with TNP-BGG, exhibited the same activity which could be attributed to the presence of the same glycolipid moiety, suggesting the selective release of Thy-l from T cells during antigen stimulation (55). The nature of the shed complexes is still an area for continuing research. Electron microscope studies of Lutz et al. (64) on shedding from ATP-depleted human red blood cells indicate that membrane fragments were released as small vesicles of liposomes of uniform size and shape. Surface antigen shedding in liposome form has also been observed in mouse mammary tumor virus (MuMTV) studies of GR Ascites leukemia cells (61). Release of the Thy-l and H-2-containing vesicles could be stimu- lated by treatment of the cells with rabbit anti-MuMTV antiserum (61). In their investigation of Thy-l dependent modulation of B cell differentiation, Miller and Esselman (40,43,44,55) found it necessary to incorporate Thy-l glycolipid into liposomes made up of auxillary lipids, cholesterol and lecithin (C:L). When Thy-l glycolipid or C:L liposomes were used alone in their culture systems, no significant effect on B cells was seen. However, when Thy-l was incorporated into the CiL liposomes, B cell modulation again became evident (40). They proposed a model in which Thy-l was shed from activated T cells in complex with 23 other membrane lipids, forming liposomes. The liposome form appeared to be necessary for Thy-l glycolipid activity (43,44). Role of T Lymphocytes in Humoral Immune Response Regulation Lymphocyte Interactions It has been demonstrated in recent years that regulation of the humoral immune response, both in_yi££2_and in_zivg, requires interaction between thymus-derived lymphocytes (T lymphocytes) and precursors of antibody-producing cells, B lymphocytes (70-74). By reconstitution of irradiated or irradiated/thymectomized mice with combinations ofthymo- cytes, bone marrow cells, and thoracic duct cells, Claman et al.(70) and Miller and Mitchell (71,73) demonstrated T cell - B cell synergism for host antibody responses to most antigens. T cells appeared to be regu- latory while B cells eventually became the antibody-producing cells. The first regulatory function attributed to T cells was that of a helper function as exhibited in the above studies. In contrast, Gershon and Kondo (75) observed that specific immunological unresponsiveness in mice given high doses of sheep erythrocytes could be passively transferred to normal mice by the transfer of T cells from the tolerant mice to the unimmunized mice. 'This-observation established that, under suitable con- ditions, antigens could induce the differentiation of a subpopulation of T lymphocytes capable of suppressing the formation of antibody-producing cells (75). Thus, T cells could be categorized as to their ability to exert a positive (helper) or negative (suppressor) influence on the humoral immune response. Development and function of T lymphocyte sub- populations have been reviewed in detail (76,77). 24 Mechanisms of T Cell Regulation of.the Humoral Immune Response Extensive work has been done in an effort to elucidate the role of the T cell subpopulations in control of the immune response, although the precise mechanisms of this system of "fine tune” regulation remain to be defined. Cell-to-cell contact. Several investigators have shown that cell- to-cell contact or close association is necessary for the induction, pro- liferation, and differentiation of B cells into mature antibody producing cells (78-80). Working independently, Mosier (78) and Pierce and Benacerraf (79) produced evidence of antigen-induced lymphocyte-macro- phage clustering in 23532, This interaction was specific (80), mediated by antigen and/or immunoglobulin (Ig) on the surface of the interacting cells, and was required at least during the early stages of the primary response (78,79). The cells could then be dispersed and activated cells could continue to develop into antibody—forming cells, independent of other cells and clusters (79). McIntyre et al. (80) produced ultra- structural studies confirming septate-like junctional complexes between lymphoid cells antigenically stimulated in culture, although no group distinction between the lymphoid populations was possible. Soluble mediators of regulation. Considerable evidence points to the importance of soluble mediators of immunoregulation produced and released by macrophages (77,81) and T lymphocytes (43,44,69,76,84—90,92,94-101) in directing the proliferation and differentiation of B cells without cell- to-cell direct contact. The in yiyg_induction of antibody-forming cells probably requires a combination of specific cell-to—cell contacts among 25 macrophages and lymphocytes and elaboration of soluble mediators effective over a short distance in the microenvironment surrounding the cell types active in the response. Soluble T helper factors. A number of researchers have isolated and studied cell-free helper substances released from antigen-stimulated spleen cell cultures (82,83). These factors have been shown to replace T cells both the in ziyg_and ig;yi££2_induction of the immune response. Rubin and Coons (84) first described a factor isolated from the super— natant of spleen cell cultures or mice primed 30 to 60 days before cul- ture with tetnus toxin (TT). Addition of nanogram amounts of TT to cul- tures being stimulated with sheep red blood cells (SRBC), enhanced the anti-SRBC response significantly over the control cultures. The respon- sible factor was nondialyzable, stable for 30 minutes at 56 C, and inacti- vated by proteases but not DNase or RNase (84). Gorczynski et a1. (85) and watson (86), using slightly different methods, also described the presence of a similar factor in antigen-stimulated cultures of spleen cells from irradiated mice (86). This factor was proposed to be released from specifically activated T cells, although nonspecific in its ability to enhance the humoral response to an unrelated antigen (84—86). Taussig (87) described a factor which was present in the medium of cultures made from "T cell spleens" from irradiated, thymocyte-recon— stituted mice, subsequently immunized with a T-dependent antigen. The medium containing the cell-free factor was mixed with bone marrow cells and antigen and transferred to lethally irradiated recipients. The humoral response was measured at a later time and found to be equal to 26 that of the control mice which received either the "educated" T cells (from mice already exposed to the antigen) or normal T cells. This helper substance was found to be antigen-specific in its effect, as transfer of the supernatant combined with bone marrow cells and an antigen unrelated to the primary antigen failed to elevate the response to the unrelated antigen compared to the controls (87). Armerding et al. (88) have studied a factor elaborated by allo- antigen-activated T cells in a mixed lymphocyte culture. Short term incubation of T cells, previously primed with histoincompatible lympho-' cytes in vivo, with the appropriate target cells produced an "allogenic effect factor" (AEF), elaborated into the culture medium capable of non- specifically restoring the anti-SRBC and anti-dinitrophenol responses in T cell-depleted spleen cell cultures (88). This was attributed to its ability to replace the need for carrierfspecific helper T cells. AEF was further characterized as being a glycoprotein of two subunits, 40,000 and 12,000 daltons, both necessary for activity (89). Production of this factor was dependent on H—2 differences of the cells in the culture (90). DNA replication and proliferation were not necessary, while glycolysis, protein synthesis, and electron transfer were prerequisites to the syn- thesis and release of AEF (90). Phenotypic studies on T cell subpopu- lations indicated that T cells exhibiting Ly-l and Ly-2,3 cell surface antigens were required for AEF production (90). By using specific anti-Ly antisera, Jandinski et a1. (91) confirmed Ly-l positive T cells as those responsible for helper activity and identified T cells exhibiting the Ly-2,3 phenotype as those responsible for suppressor functions. 27 Soluble T suppressor factors. A multitude of soluble T cell- generated immune response suppressor substances have been described. Suppressor T cells and their products have been characterized as antigen- specific (76,92,93) or antigen-nonspecific (43,44,69,86,94,96-101) in their inhibitory activities. Suppression of the immune response by soluble T cell factors may provide at least partial explanations for antigenic competition (55), tolerance (75), mixed lymphocyte reaction (101), and modulation of the humoral immune response (43,44,69,76,86, 94-101). The apparent diversity of factors in their biological as well as their physiochemical properties can probably be attributed to the variety of methods used by investigators to isolate and study them. Takemori and Tada (92) extracted a factor from physically disrupted thymocytes and spleen cells of mice formerly immunized with the soluble hapten car- rier, keyhole limpet hemocyanin (KLH). Injection of this factor into syngenic mice immunized with dinitrophenol-KLH (DNP-KLH) significantly suppressed the secondary (IgG) anti-DNP response (92) as measured by a hemolytic plaque technique. The activity could be removed by absorption with heterologous anti-thymocyte serum, alloantibodies against the donor strain, and antigen, but not by anti-mouse immunoglobulin antibodies. This antigen-specific, heat labile protein of 35—60,000 daltons was suppressive in mice with same H-Z specificity and had no effect in allo- geneic mice (92). Further studies showed that this substance exerted its inhibitory influence on T helper cells rather than B cells, providing evidence for interactions among T cell subpopulations as well as T cell-B cell communication (76). 28 A factor similar in many respects to that described by Takemori and Tada (92), was investigated by Kapp et al. (93) using mice classified as nonresponders to the synthetic polypeptide L-glutamic acid6O - L- alanine30 - L-lysinelo (GAT). Priming with GAT specifically decreased the ability of these mice to respond to CAT-methylated bovine serum albumin (CAT-MESA) or GAT-pigeon red blood cells (CAT-PRBC), two forms of the antigen that were normally immunogenic (93). T cells from these mice were extracted according to the procedure of Takemori and Tada (92). The T cell extract was then introduced into normal syngeneic mice together with CAT-MESA or CAT-PRBC. Again the response was suppressed. The antigen specific factor in the T cell extract could be absorbed by GAT- Sepharose (but not BSA-Sepharose), was removed by alloantisera to the Ir gene products, and eluted from a Sephadex G-100 column with a molecular weight of 45,000 daltons (93). Following a protocol similar to that described for his helper factor (87), Taussig (94) used the spleens from'irradiated thymocyte-recon- stituted mice which were subsequently primed with a T-dependent antigen in 231g and removed 7 days later for culture. The cultures were exposed to the primary antigen and found to develop suppressive function not observed in previous cultures not exposed to the immunizing antigen in ‘ziggg (94). He attributed the suppressive activity to the possible presence of a soluble suppressor factor elaborated by a subpopulation of T cells, distinct from the helper T cell population, which outlasts the helper factor. That is, the suppressor factor was more stable or it eventually overrode the helper factor effect (94). 29 Others have elicited production of nonspecific suppressor factors, employing a host of similar methods of antigen or mitogen activation of T cells. Most methods involved priming the experimental mice with antigen several days before collecting the spleens for culture (43,93-97). In some studies, the spleen cells required incubation with additional antigen in culture (94,96). In one report, a single exposure to antigen in 1129 produced a helper effect, and a second exposure in_ .XEEES elicited the suppressor activity (Taussig - compare 87 and 94). Another required "double priming" with two antigens for the suppressive factor release; priming with a single antigen gave an enhancing factor (96). Yet another showed that priming in_vigg_with a low dose of antigen (1068RBC) influenced the release of helper activity from T cells in culture, while prior exposure to a high dose (109 SRBC) resulted in a suppressive factor being released into the medium from the cultured T cells (43). Many of the substances investigated fell in the molecular weight range of 35-60,000 daltons (93,95,97,98), with some exceptions where the molecular weights were considerably less, i.e., less than 10,000 daltons (43,99). When characterized, most of these were found to be glycoprotein (95,97) but no further properties were available. In one case (43), the molecule responsible for suppression was defined as a glycolipid with the mobility of G ganglioside by thin layer chromatog- raphy and was associated with activity of the T cell surface antigen, Thy-l (43). Con A has been used for multiclonal activation of spleen cells in culture to elicit suppressor factor release into the medium (98,101). Rich and Pierce (100) isolated and characterized a "soluble immune 30 response suppressor" (SIRS) and showed its ability to suppress the primary response in spleen cell cultures to heterologous erythrocytes without affecting viability (lOO). SIRS presence was only required for the initial 24 hours of the response and the suppression was evident by day 5 of the culture. SIRS was determined to be a glycoprotein of 48-67,000 m.w. It was not reactive with anti-IgG or anti-u chain, could not be absorbed by antigen, and lacked I region determinants (98). In addition, SIRS could not be separated from migration inhibition factor (MIF) activity (98). Further studies determined that SIRS was elaborated from a T cell subpopulation bearing the Ly—2,3 phenotype (91). Reinertsen and Steinberg (99) isolated a molecule from Con A-stimulated spleen cell culture supernatants which contained suppressor activity, was less than 10,000 m.w., and lacked MIF activity. This suggests separation of the active part of the SIRS molecule from that with MIF activity. Work has been done on a human T lymphoid cell line, MOLT-4 (69), which forms SRBC rosettes, has no surface immunoglobulins, no comple- ment receptors, and no receptors for antibody-antigen complexes. Anti— MOLT cytotoxicity could be absorbed with fresh human thymocytes. When exposed to SRBC, but not chicken erythrocytes (CRBC), MOLT elaborated nondialyzable immunosuppressive factors into the medium which were cap- able of suppressing the anti-SRBC and the anti-CRBC responses in mouse spleen cell cultures (69). In vivo applications. The ultimate test of any helper or suppres- sor factor is demonstration of its role in_vivo. As fulfillment of this criterion, Taussig (94) reported replacement of T cells with a T cell 31 factor in an in 3139 response. Irradiated, thymocyte-reconstituted mice were_primed with the highly T-dependent synthetic polypeptide poly (Tyr-Glu)- poly DL Ala--poly—Lys, abbreviated (T,G)-A-—L. Their spleens were subse- quently removed and cultured. The culture supernatant, bone marrow cells, and (T,G)-A--L were injected into lethally irradiated syngeneic recipients. Twelve days later, spleens were removed and subjected to a direct (IgM) hemolytic plaque-forming-cell (PFC) assay for evaluation of antibody- forming cells. A significant increase in PFC over controls was observed. Induction of the helper activity was antigen—specific while its effect was antigen-nonspecific (87). Takemori and Tada (92) reported in_viyg_suppression of the indirect (IgG) anti-DNP response, using a factor extracted from carrier-primed T cells. This antigen-specific factor was coded for by the I-J region of the H—2 complex (102) and was active only in H-Z histocompatible hosts (92). A further demonstration of in xiyg_biological activity was pro- vided by Kapp and her associates (93) in experiments with a factor extracted from antigen-primed T cells. This factor, extracted from GAT- primed T cells from nonresponder mice, was injected into normal syngeneic mice, suppressing the immune response to GAT-MBSA and CAT-PRBC, antigens which would normally elicit a response. In an interesting set of experiments, Krakauer et al. (102) drew a connection between SIRS production (100) and age-dependent development of an autoimmune syndrome by NZB/NZW mice. They attempted to elicit SIRS production by Con A stimulation of T cells from these mice at various ages and measured its effect on a defined immune response. 32 A strong correlation between declining SIRS production, increase in age, and increase in development of autoimmunity was reported (102). These authors proposed that this was due to fewer SIRS-producing cells with the increase in age and the animal's incapacity to suppress "forbidden clones" acting against "self" tissue antigens (103). Until this battery of helper and suppressor substances is purified and biochemically characterized, it is difficult to do more than specu- late about the mechanisms of action on the molecular level. More under- standing is required about the nature of the mediators, their active site(s), their interactions with cell membranes, their effects on cyclic nucleotides, intracellular divalent cations, microtubules and micro- filaments (83). Regulation of the immune response appears to be a com- plex interrelated network of finely tuned systems of which soluble medi- ators are only one important part. Useful manipulation of the immune system will require a much greater understanding of all facets of the network. In summary, the discovery of the dichotomy of the cell types par— ticipating in the humoral immune response has stimulated a great deal of research directed toward elucidating the nature of the cellular inter- actions which occur. Efforts have been aimed at methods of identifying the specific cell types involved and the mechanisms of their cooperation. The Thy-l alloantigen, a convenient cell surface marker for identifying T lymphocytes in the mouse, has been helpful in studying the subpopu- lations of lymphocytes which appear to play a supervisory role in control- ling and fine tuning the humoral immune response. Controversy still exists as to its biochemical nature, though it may be that it exists in 33 both a glycolipid and glycoprotein form with its antigenicity residing in the carbohydrate portion. In_vitro studies suggest B cell modulation as the biological role of Thy-l. An in_vivo functional role for Thy-l has yet to be established. MATERIALS AND METHODS M139 C3H x c57BL/10 Cz (Health Research, Inc., Buffalo NY) and C3H x C57Bl/6 He (Cumberland View Farms, TN) female mice were used to evalu- ate antibody responses following imposed treatments. Mice of both sub- strains were referred to as BC3Fl by their respective suppliers. For clarity, they will be referred to asBlOCBFl and B6C3Fl in this report. Mice were 8 to 15 weeks old and were age-matched for a given experiment. Sheep Erythrocytes Sheep erythrocytes (SRBC) to be used as antigens were obtained from a single tested sheep (735 Blk) at Grand Island Biological 00., Madison, WI. SRBC were stored in Alsever solution before use and washed three times in sterile phosphate-buffered saline plus 0.2% glucose (PBS plus glucose), the buffy coat removed and discarded, and the cells resuspended in PBS plus glucose for i§_yigg_immunization or in Minimal Essential Medium (MEM, Hank's base, Grand Island Biological 00., Grand Island NY, pH 7.2-7.5) for use in the hemolytic plaque assay. Hemolytic Plaque-Forming Cell Assay The in vivo induction of primary and secondary humoral immune responses in mice was measured by the slide modification of the hemolytic plaque assay as described by Miller and Cudkowicz (104). Mice were 34 35 primed with 5 x 107 or 5 x 108 SRBC by intravenous injection in the lateral tail vein. Each individual spleen was removed 3 to 11 days later, minced with forceps in 1.0 ml of MEM, followed by gentle aspiration with a syringe and needles of increasing gauge (21 to 27) to obtain a single cell suspension, and placed on ice for immediate use. The cells were diluted in MEM to permit detection of 50 to 500 PFC/0.1 ml. One tenth ml of the appropriate dilution was combined with melted agarose (53 C, Indubiose, L'Industrie Biologique Francaise, S.A.) and SRBC and plated onto microscope slides previously coated with 0.1% agarose and dried. The slides were inverted on special lucite trays and incubated l-l% hours at 8% CO2 tension in a 37 C humid incubator. Assessment of IgM-producing cells (direct PFC) was carried out using guinea pig complement (fresh frozen, Colorado Serum Laboratories, Denver, CO, and Suburban Laboratories, Silver Spring, MD) diluted 1:10 or 1:20 in MEM, and added to the trays which were incubated for an addi- tional 2 to 3 hours as described above. A duplicate set of slides was prepared for detection of IgG-producing cells (indirect PFC) to which goat anti-mouse IgG (Meloy, Springfield, VA) was added (diluted 1:100 in MEM) after the initial 1 hour incubation. This antiserum was 70% to 80% inhibitory for direct PFC, so appropriate adjustments were made for indirect PFC counts. The slides exposed to anti-mouse IgG were incu- bated for 1 hour, the antiserum drained off and replaced with complement, and incubated for an additional 2 to 3 hours before reading indirect PFC. Both direct and indirect anti-SRBC PFC were determined using duplicate slides and the average taken as the direct or indirect PFC/ spleen for a given animal. The PFC/spleen for a given treatment group 36 was calculated from groups containing 3 to 6 mice/group. Treatment groups were of equal size for a given experiment. Occasional discrepancies from experiment to experiment between PFC readings of control groups receiving comparable treatment were encountered, therefore, data for each experiment was standardized to the SRBC control given the value of 100.0, with the exception of PFC data for Figure 4. PFC response for each mouse of each group was adjusted relative to this value, then group means and standard errors calculated. These values are referred to as ”adjusted PFC/spleen." For reference, the actual group mean PFC/spleen for the control group is indicated for each experiment. Isolation of Glycolipids Glycolipids were prepared by extraction of AKR, C3H, or CBA mouse brain tissue with choloroformrmethanol mixtures (105). Extracts were submitted to a Folch partition (106) and the upper ganglioside-rich phase dialyzed against several changes of distilled water for 24 to 48 hours at 4 C. The remaining material was lyophilized, subjected to mild alkali hydrolysis with 0.6N NaOH in methanol, and dialyzed against several changes of distilled water for 24 to 48 hours at 4 C. The remaining sample was lyophilized and separated by column chromatography on Anasil S (Analabs, Inc., North Haven, CT) with chloroformimethanol: water (C:M:H20) mixtures. The C:M:H20 mixtures were prepared in the following proportions: 37 Solvent C : M : H20 (v/v/v) A 65: 25: 4 B 65: 30: 6 C 60: 35: 8 D 60: 40: 10 This sample was applied to the column and 400 ml of solvent A was allowed to flow through the column as void volume at the rate of 20-27 ml/hour. The next 200 ml were collected in 10 ml fractions. Sol- vent B was then substituted for solvent A and 500 ml collected in 10 ml fractions. Solvents C and D were substituted in turn and 500 ml of each collected in 7.5 ml fractions. Samples of the fractions collected were chromatographed on Silica Gel G thin layer plates (0.25mm thickness, E Merck, Darmstatt, W. Germany) in a solvent system comprised of chloro- formzmethanolz 2.5 N NH4OH, 100:42:6, v/v/v. The glycolipids were then visualized with resorsinal and heat and compared to a standard. Fractions containing GMl ganglioside plus Thy-l-active glycolipid were pooled. This material and other gangliosides were quantitated by the method of Svennerholm (107). Several of the glycolipid preparations were provided Dr. walter Esselman of the Dept. of Surgery, Michigan State University, East Lansing, MI. Liposome Preparation Liposomes were prepared for in_vivo use and electron microscopy by combining the isolated glycolipid, cholesterol, and lecithin (choles- terol and lecithin from Supelco, Inc., Bellefonte, PA) in the ratio of 1:5:5 (w/w/w). This mixture was dried and sterile 0.85% saline added to 38 obtain the desired concentrations. The combined lipids were then sonicated at 60 Hz (low frequency - LF liposomes - see Figure la) in. an ultrasonic cleaner (Mettler Electronics Corp., Anaheim, CA) or at 20,000 Hz (high frequency - HF liposomes - see Fibure lb), using a probe sonicator (Branson Sonifier Cell Disruptor 200, Branson Sonic Power Co., Danbury, CT) for 10 minutes, under N2(g) in an ice bath. For comparison in the electron microscope, liposomes were pre- pared in PBS plus glucose and sonicated at 20,000 Hz as before (Figure 1c). The resultant liposome suspensions were used immediately and/or after 24 hours depending on the experiment. In one case, liposomes sonicated at 20,000 Hz were diluted and resonicated for 1 minute at 60 Hz before use. Liposomes prepared at low frequency (Figure 1a) were generally larger than high frequency liposomes (Figures 1c and 1d), ranging in diameter from 0.05 um to 1.65 um or more, while HF liposomes ranged from 0.02 um up to 0.30 um; Although size heterogeneity was observed for both preparations, the majority of the liposomes fell within the ranges indi- cated. HF liposomes prepared in PBS plus glucose consisted of some uni- lamellar and many multilamellar vesicles, although more than three bilayers were rare. All liposomes in this preparation were within the expected range for an HF preparation, but appeared to be slightly smaller on the average and more homogeneous in size than those prepared in physiological saline. Both LF and HF liposome preparations displayed- occasional tightly formed multilamellar concentric vesicles (Figure 1b), though these were rare. Multilamellar liposomes were present in both HF and LF prepa- rations. LF liposomes appeared to have more concentric superficial Hll Figure 1. la. lb. 1c. 1d. 39 Effect of sonication on liposome morphology. Liposomes were prepared from cholesterol, lecithin, and G gang- lioside, sonicated, dried, and negatively staingd with 2% ammonium molybdate for observation under the electron microscope. Low frequency (LF) liposomes formulated in 0.85% saline, sonicated at 60 Hz. (Magnification: 48,600 x). High frequency (HF) liposomes formulated in 0.85% saline, sonicated at 20,000 Hz. (Magnification: 130,500 x). HF liposomes formulated in PBS plus glucose. (Magnifi- cation: 134,500 x). "Free lipids" (non—liposome lipids, present in LF liposome preparation. (Magnification: 167,500 x). 4O Figure l. .fi ~ . I . ' ~ - _ ., .r ~.3£ . . \J, .- . '7 ‘ , "‘ v‘ 1 ~. I .4 -{ . -. 7;"; s 7"! ‘ ‘~ w” -- -. £2le a ‘W A . . ‘x , L' ". (3 ‘ r -n._ _ - s ._ .. - ~V .. .1. I; J 73‘ I 3"1'L. Hub... ‘1 f" ' I‘D. . - Q‘ -‘ I- 9. 43'}: 23".) .Ya—t ;- - ‘ 1f ' .1. c- r r "e '. i. *1.- L ~J ~~Jv tron - \f— "‘ 1' .L. . _ QNKIJFZBQSTC ‘ ‘ D ‘1 D. _: v C I}? Oil-P it {’IA '- -s-“o——- ~98“ a - van-vane- ~_. I. ' .. , , , - . ' .4113 C cqfil‘)Cc.§- “ Q -" ' I - ' q(-' a- go I\'(4~L. ‘ ’ (.15-: £1.43." t..' ,L. 41 layers and a greater number of inner nonconcentric vesicles than HF liposomes. Length of sonication generally affected the amount of "free" lipids detectable as nonliposome structures (Figure 1d), decreasing some- what with increased sonication time. Despite the length of sonication, «{ "_ _ ' _. free lipids were always present in LF preparations and to a lesser degree in HF preparations. Electron Microscopy of Liposomes Liposomes obtained as described were prepared for electron micro- scopy by allowing a drop of the suspension to dry on a 300 mesh copper grid coated with a film prepared from 0.5% parlodian (Mallinckrodt Chemical Work, St. Louis, MO) in amyl acetate (108). Each grid was negatively stained (109) with a 2% solution of ammonium molybdate (pH 7.65) for 5 seconds. The specimens were examined and photographed in a Philips 300 microscope. Statistical Analysis Comparisons of mean PFC responses of experimental groups of mice were made employing a two-tailed Student's t test. p." m- . .— -fl- -- LAT‘ -, rr r . 1.. 4.4L.._( .“ 5 y. _.,, ‘vi‘-' I a -t \ .\.., -v -v i I q CU»: RESULTS Attempts to Modulate Antibody Responses In Vivo with Thy-l It has previously been shown that GMl ganglioside, containing Thy-l—active glycolipid (40), has a modulatory effect on B cell differen— tiation in 31533 (42). Initial experiments were performed to test the possibility of an in vivo effect of the Thy-l glycolipid. B10C3Fl mice (Thy-1.2) were injected i.v. on day 0 with 5 x 108 SRBC, followed 24 and 48 hours later by i.v. injections of various glycolipids formulated into cholesterol-lecithin liposomes. Six days after receiving SRBC, spleens were removed and evaluated for the number of anti-SRBC plaque- forming cells (PFC)/spleen. The control group received SRBC followed by PBS injections in place of glycolipids and the response was assigned the arbitrary group mean value of 100.0. All other data were adjusted relative to this value. Results are given in Table l. CBA GMl(containing Thy-1.2 glycolipid) was found to suppress the anti-SRBC plaque response‘ by about 60% in the direct (IgM) PFC assay compared to the control group in Experiment 1. CBA G b ganglioside demonstrated suppression of 38%, D1 while AKR GMl (containing Thy-1.1 glycolipid) was much less effective in suppressing the PFC response, decreasing the response by 24% from control values. The indirect (IgG) response was suppressed by almost 98% in the CBA G -treated group and was suppressed to a lesser, but signi- Ml ficant, degree with both CBA Gle and AKR GMl gangliosides. Another preparation of brain GMl from C3H mice was found to be as effective as the CBA preparation in suppressing the plaque response, 42 1. ‘ ‘ .u. \h F5 _.. .2». u ,.. .. . e». .u.\. ~fi_-\nvxL . . . ,. w t .v -,..n\. .. L n (X x. .w . u . x a. at. W H... o. L; .. ”J”. l. \ - n. u .. .v- .. \ .. .Hw .........m I .\ t . . \ .. h“. .. my .4 .H./;.nrh 7.. u y. 1 . .Y. , .3 7 -7 . 43 .aooamm\omm Hmnuo< .AN .mxmv moonm Ho.ovm«** Hoo.ovm «a o~.oum « Amaouucoo ommm ou wounnfioov amoulu m.uaow=um on \oofis m mam Aa .mxmv esopw\oofia a How wound .cum HnAmuoauoz mam mamfiuoumz oomv some woumnnw< Anv .mcowuoonafi omMm noumm when w moumsam>o mes soofimm\0hm mo Rosana may .Amvonuoz can mamauoumz ommv moeomomwa mg ouafi woumuomuooaw woumofiwufi mm mwfiawaooham wmumawwmow msu mo Am .exmv w: H no AH .mme m: N van Ao hmov .mu: o um .>.« ommm moa x m wo>aooou euros zoom Amv I. m as: + Tea 8.2 .a.z Se Eu 8% «.34 H mg: .o.z Adz 3.72.... 3 Eu :8 mesa .e.z £3.53 H .3: new H CS 3419:. +v Eu 52 2% . . . .I . . I. . nan a z ««m m + m m .m «a + m Ho 0 «mo ommm «a: H 2: is; H o.~ «Na H Tea @4129 +v Eu «mo 8% vomq.wv I. I. UAOOm.¢mv.I m.m + o.ooa o.- + o.oo m.wq + c.ooa . I omdm Hz 5.0 H ed Na H Co I I can uoouwm ummyuoouwvcH 0mm uoouam N van H.mmwa o zen aaeeaaa\oem unemuaaam .nea aaeeaam\oam unamnaaaa seamenee N ucoseuomxm H unmafiwomxm unoaumouH Q mofiomomwa out“ voumuoduooaa movwmoaawsmw samun omsoa an o>«>.mw uncommon ommmnfiucm mo sofimmoumdnm H mafia. ’0, III.) 2‘ \r' C.» .0. ,..s (2‘ . . v: n .r. . \ . 1. ./ I . p. .i 5 L ._ r. a. a . . r. .I;. . II: or I p. .I, I T. I . —. I. e u m. e .I. WI 9. e . .«u I 4 u. w . . . I. I ,a . . ..~ . o J . ..I .I‘ In. If Ill O F x ’r‘ . u . 1. \- I. I . ._ I a». l- '1" SJ ~4- . ’— u ,. § '-. :7 .. . F I J .A. .1 3‘. a a ... J r; . . I a. O . . Q .o. 9 . .... dd. 0:. . a , u~ i \l. x .._ 44 while CBA GDla ganglioside did not deviate from control values (Experi- ment 2, Table l). The suppression by GMl treatment was much more dra- matic, abrogating more than 80% on the response. It is notable that the unadjusted direct PFC/spleen values in Experiments 1 and 2 of Table 1 show a great disparity. Experiment 1 produced a direct PFC control response of 39,900 PFC/spleen compared to 8,430 PFC/spleen in the control group of Experiment 2. Control mice in the study shown in Table 2, which received identical treatment as the control mice in Table 1, demonstrated a direct response of only 4,490 PFC/spleen (see discussion). Effect of Thy—1 Administration Schedule The treatment schedule used for the experiments in Table l was chosen because earlier in 21532 experiments had demonstrated that Thy-1 glycolipid was most effective when added on days 1, 2, or 3 of culture, with maximum suppressive influence when added on day l (40). Next, the possibility that the suppression seen in_ziyg_(Table 1) could be increased was tested using a treatment regimen which includes giving Thy-l glycolipid prior to antigen exposure as well as after. B10C3F 1 mice were injected with CBA (containing Thy-1.2) in liposome form for Gm 6 consecutive days, beginning 3 days prior to SRBC administration and continuing 2 days after immunization (Table 2). Again the direct anti- SRBC response was suppressed by 60%. The indirect response for the con— trol group was only 38% of that of the direct response (compared to an indirect control response of 98% of the direct PFC response in Experiment 1, Table l) but was also suppressed by approximately 60% in the GMl- treated group. 45 om.ovm «x mo.ovm « Aaouucoo omMm ou woummaoov ummulu m.uaonsum .cmoamm\omm coma Hmauo< on .mnoum\oofla m Mom Houuo .mum.H.Ammo:umz mam mamaumumz mmmv Gama wmumSHU< Anv . .meoaeemflea omMm Houmm mane o moumsam>o mus sooamm\umm mo nomads ona .Amwonuoz mam mamwuoumz oomv moaomomfla mg ouaw neuraseuom mmz H20 .Aaomwucm Hoummv N are weavsaocfl was ou >mem wcflasauaoo use Anowwucm ou uofiuav mu >n© so mafissawon %mv Mom AN.NI%:H mawsflmuaoov sz m: a man 0 ham so ommm mod N m co>fiooou mmsoa comm Amv aes.m + o.~a . «m.~ me.Ne Am.ausae+v axe emu omMm a.o~ + m.mm aaoaeev a.sa + o.ooa .IIIII omgw one neeaaeeH use aoeeaa N+ ea mu mama o sea a amaaam\oam unemuaaaa eeuaamea a eeaaameaa cowwucm Hmuwm was .sufiz mnooamuassam .ou HOHHQ .coafimou oawcfim m :H monoumficfiswm :msB onconmon ommmlfiusm do o>fi>.mfl HIkfiH mo uoowwm m>flmmmummnm N mqmo was amoamm\omm mo Hogans ens .GONumofisom Aux ooo.ONv horoscouw awe: saws venomous mums moaowoqu .Awmonumz tam addendum: ommv A m: mm .m m: mm so>Hw mums gnu ao>wm woe: .N can a whom so some w: ma co>fim onm3 H U wsfi>flmoou due: .0 hum co omMm moa x m mo>fiooou wows Ammoom Adv «m.a~ m ~.oma auo + qu emu anew A.“ H m.ea saee Am.ausae aeaav Hzo emu ommm e.ea fl H.¢HH sane masomoaaa gnu ummm eaooo.eav a.om.w.o.ooa ommm 0mm uooufio N was H mama o awn a eeeaam\omm omemuaeaa emaaamee a aamauaaas o>H>.mM mwmfiaoozaw HI%:H an uncommon omMmIHuam mo ucoaoocmscm pom uaoEouwsvou VNQHH xumaafixar m m4m