anigg3monoclonalantibodyestablishedafterimmunizationwithgm...

(CANCER RESEARCH 48, 5680-5685, October 15, 1988]

An IgG3 Monoclonal Antibody Established after Immunization with GM3Lactone:Immunochemical Specificity and Inhibition of Melanoma Cell Growth in Vitroand in yivo1

Taeko Dohi, Gustavo Ã‘ores,2and Sen-itiroh Hakomori3

The Biomembrane Institute; Department of Biochemical Oncology, Fred Hutchinson Cancer Research Center; and Departments of Pathobiology and Microbiology,University of Washington, Seattle, Washington 98119

ABSTRACT

In previous studies, an IgM monoclonal antibody (M2590), establishedafter immunization of C57BL/6 mice with syngeneic B16 melanoma cells,was found to react with melanoma cells, but not with various normal cellsand tissues (Taniguchi, M., and Wakabayashi, S.. Jpn. J. Cancer Res.,75:418-426, 1984). The structure defined by this antibody was identifiedas GMJ(Hirabayashi, Y., et al., J. Biol. Chem., 260:13328-13333,1985)organized in membranes at high density, although the real immunogenwas suggested to be GMJlactone (Ã‘ores, G. A., et al., J. ImmunÂ«!.,139:3171-3176, 1987). Since GMJ lactone was found to be highly immuno-genic, we subsequently immunized C57BL/6 mice with GM3lactone coatedon Salmonella minnesotae and established hybridoma DH2, secreting anIgGj antibody showing preferential reactivity with GMJlactone over GMJunder certain conditions. The reactivity of the DH2 antibody was competitively inhibited by M2590, and it showed a preferential reactivitywith melanoma cells and displayed various immunochemical and immu-nobiological properties similar to those of M2590. However, DH2 antibody inhibited melanoma cell growth in vivo, induced antibody-dependentcytotoxicity in vitro, and showed a preferential accumulation in melanomagrowth Â¡nvivo. These properties are characteristic of the IgGj subclass,in striking contrast to IgM antibody M 2590, which does not inhibit cellgrowth in vivo or in vitro and does not induce antibody-dependentcytotoxicity. Thus, immunization with lactone forms of tumor-associatedganglioside antigens might be useful in the production of antibodies andprevention of tumor cell growth in vivo (antitumor vaccines).

INTRODUCTION

Glycosphingolipids, particularly gangliosides, have been assigned as human tumor-associated antigens defined by monoclonal antibodies (1). An IgM monoclonal antibody, M2590,previously established after immunization of C57BL/6 micewith syngeneic B16 melanoma cells, was found to react notonly with B16 melanoma, but also with human and hamstermelanoma, although it does not react with various normal cellsand tissues (2). Interestingly, the structures defined by thisantibody were identified as GMJ and IV3NeuAcnLc4 (sialyllac-toneotetraosylceramide),4 which are found ubiquitously in var

ious normal cells on chemical analysis (3). The reactivity of theantibody depends on the density of GMJ at the cell surface andshowed a clear threshold value for GMJ density. The antibodyshowed stronger affinity with GMJ lactone than with GMJ, andGMJ lactone was detectable in B16 melanoma cells (4). In fact,

Received 3/25/88; revised 7/15/88; accepted 7/19/88.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1This investigation has been supported by an Outstanding Investigator Grant,

CA42505 from the National Cancer Institute, and funds from The BiomembraneInstitute.

! Recipient of a fellowship from Consejo Nacional de Investigaciones CientÃficas y TÃ©cnicasde la RepÃºblicaArgentina (1985-1986).

5To whom requests for reprints should be addressed, at The Biomembrane

Institute, 201 Elliott Avenue West, Seattle, WA 98119.4Glycolipids are abbreviated according to the recommendations of the IUPAC-

IUB Nomenclature Committee (confer with Biochem. J., Â¡71:21-35, 1978), b.itthe suffix OseCer is omitted. Ganglioseries gangliosides are abbreviated accordingto the ill-tinn iims of Svennerholm (31). SPC is used for sialylparagloboside(IV3NeuAcnLc4). NeuAcGMj and NeuGcGMj are used for GMJ containing N-acetyl ncuraminic acid and Â¿V-glycolylneuraminic acid, respectively.

GMJlactone was much more immunogenic than GMJ in C57BL/6 mice (4). It is assumed, therefore, that the real immunogenin B16 melanoma could be a trace quantity of GMJ lactonepresent in melanoma cells that induces production of antibodiesreacting with both GMJ lactone and GMJ. This working hypothesis provides a potentially useful method to obtain antiganglio-side immune response by immunization with ganglioside lac-tones instead of native gangliosides or tumor cells. This paperdescribes a hybridoma producing an IgGj antibody, DH2, whichwas established after immunizing mice with GMJ lactone. Theantibody is directed to GMJ lactone as well as GMj and has beenshown to inhibit the growth of melanoma cells in vitro and invivo.

MATERIALS AND METHODS

Cells and Cell Culture. The B16 mouse melanoma, P36F4, andHMV-I491B10 human melanomas were kindly provided by ProfessorM. Taniguchi (Chiba University School of Medicine, Chiba, Japan).M2669, M1733, M2981, and M2291 human melanoma cell lines havebeen cultured from primary expiants in this laboratory. FUA169 is amutant cell line established in this laboratory from FM3A/F28-7 mousebreast carcinoma and is characterized by the presence of GMJ,which islacking in parental cells.' Other cell lines shown in Table 1 and SP2

mouse myeloma were donated or purchased and have been maintainedin this laboratory.

Glycolipids. GMJ (NeuAcGMj) was prepared from dog erythrocytes(5), rat brain, and B16 melanoma cells. GMI, GDI., GDit>,and ( ÃMwerepurified from bovine brain. GM2 was purified from rat brain. GMJethylester (6) and GMJ gangliosidol (7) were prepared as previouslydescribed (4). Glucosylceramide was purchased from Sigma (St. Louis,MO). GMJ containing AAglycolylneuraminic acid (NeuGcXjMj;II3NeuGcLcCer) was prepared from horse erythrocytes (8); sialyllacto-neotetraosylceramide (sialylparagloboside; SPG; IV3NeuAcnLc4), sia-lylactonorhexaosylceramide (VI3NeuAcnLc6), and lactosylceramide

were prepared from human erythrocytes (9); G,j (asialo GMZ)wasprepared from guinea pig erythrocytes (10). These preparations werepurified and have been stocked as standards in this laboratory. Lactoneforms of each ganglioside were induced when ganglioside solution inchloroform:methanol:12 N HC1 60:30:4.5 was allowed to stand for 18h at room temperature and purified by DEAE-Sephadex chromatogra-phy (4).

Immunization with GMJLactone and Preparation of Hybridoma DH2.For immunizing mice, 40 fig of GMJ lactone prepared from dog eryth-rocyte GMJ was solubili/ed or suspended in 4 ml of distilled water,sonicated, and mixed with 1 mg of acid-treated Salmone Ila minnesotae.The suspension was incubated for 10 min at 37Â°Cand lyophilized. Thelyophilized material was resuspended in 4 ml of PBS,6 and a 200-^1

aliquot of the suspension, containing 2 Mgof GMJor 2 /Â¿gof GMJlactonewith 50 fig of the bacteria, was injected into BALB/c mice i.v. once perweek for 3 weeks. Three days after the last injection, spleen cells wereharvested and fused with mouse myeloma SP2 cells. Hybridomas weregrown as previously described (11), and antibodies secreted werescreened by solid-phase radioimmunoassay using gelatin-coated plasticplates (polyvinyl flexible 96-well plates; Costar Cambridge, MA).

5T. Tsuruoka and S. Hakomori, unpublished data." The abbreviations used are: PBS, phosphate-buffered saline; BSA, bovine

serum albumin; PC, phosphatidylcholine; TLC, thin-layer chromatography.

5680

on April 17, 2018. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

IgGj MoAb ESTABLISHED AFTER GM) LACTONE IMMUNIZATION

Determination of the Specificity and Other Properties of DH2 Antibody. Solid-phase radioimmunoassay was performed by two techniques,which differed in the way that the glycolipid solid phase was prepared.In brief, varied (1-25 pmol/well) quantities of glycolipid, 50 jig of PC,and 30 Â¿igof cholesterol in 50 ÃŸ\of ethanol were dried on 96-wellplastic plates (Becton Dickinson, Oxnard, CA) (12). Alternatively, thewells were incubated with 200 ÃŸ\of 0.1% gelatin in PBS for 24 h at4Â°C,washed with PBS, incubated with glycolipid solution (0.2 p.mo\/

ml of PBS) overnight at room temperature, and washed again with PBS(4). Wells were blocked with 5% BSA in PBS for 2 h and reacted withculture supernatant from hybridomas or purified monoclonal antibodiesfor 2 h at room temperature. After washing, bound antibody wasdetected using a rabbit anti-mouse IgG and IgM antibody (MilesBiochemical, Elkhart, IN) followed by detection with '"l-Protein A.

Finally, the wells were cut and radioactivity was counted with a Packardgamma counter. TLC immunostaining was performed on Baker TLCplates (J.T. Baker Chemical Co., Phillipsburg, NJ) by a slightly modified version of the procedure described by Magnani et al. (13). Glyco-lipids were applied on TLC plates for chromatography using a solventsystem of chloroform:methanol:water (50:40:10) containing 0.05%CaCl2. After drying, TLC plates were blocked for 2 h with 5% BSA inPBS and reacted with culture supernatant of DH2 hybridoma at roomtemperature overnight. After washing, bound antibody was detectedusing a rabbit anti-mouse IgG antibody (Miles), followed by detectionwith '"I-Protein A. TLC plates were tested by autoradiography.

The isotope of the antibody was determined by Ouchterlony immu-nodiffusion using rabbit anti-mouse IgG isotype antibodies purchasedfrom Miles Biochemical.

Monoclonal antibody M2590 was provided by Prof. M. Taniguchi(Chiba University School of Medicine). Both CU-1 and SH-1 monoclonal antibodies, which were established in this laboratory, are IgGj,and react with Tn-antigen and Le" antigen, respectively. Antibodieswere '"I-labeled using IODO-BEADS (Pierce Chemical, Rockford,

IL).Comparative Reactivity of DH2 and M2590 to Glycolipids and to Cell

Surface. Inhibition of M2590 binding to GMÂ¡lactone or GMJby antibodyDH2, M2590, or SH-1 was performed as follows: GMSlactone wasadsorbed onto gelatin-coated plates, and (Â¿MIwas adsorbed onto aplastic surface with cholesterol and PC, i.e., 20 pmol/well of GM3withPC and cholesterol were dried onto plastic plates as described in thepreceding section. Wells were preincubated with various concentrationsof M2590, DH2, or SH-1, followed by addition of '"I-M2590 (final

concentration, 2 ng/m\) and incubation for 1.5 h at room temperature.I25l-Labeled antibody bound onto each well was counted with a gamma

counter.The reactivity of antibody DH2 with various cell lines in comparison

with that of M2590 was tested by immunofiuorescence and by flowcytometry. Cells were harvested using 0.02% EDTA and 0.02% trypsin,washed with PBS, and incubated with 20 Mg/ml of DH2 or 10 Â¿ig/mlof M2590 for l h in ice. After several washings with PBS, cells wereincubated with fluorescein-labeled goat anti-mouse IgG + IgM, andimmunofiuorescence was analyzed by microscopy and by fluorescence-activated cytometry, using negative control cells incubated without firstantibodies.

Effect of DH2 on Cell Growth in Vitro. The effect of various monoclonal antibodies on B16 cell growth was studied as follows: B16 cellswere harvested with 0.02% EDTA, placed in 24-well culture plates(Becton Dickinson) at a density of 5 x IO4 cells/well, and grown in

RPMI medium supplemented with 3% fetal calf serum. After 24 and48 h, the indicated final concentrations of antibodies were added, andcells were counted at the indicated time (see legends for Figs. 7 and 8).For a control, PBS was added instead of the antibody solution.

Cellular Cytotoxicity Induced by DH2 Antibody. Antibody-dependentcellular cytotoxicity was studied by 4-h chromium release assay (14).Mononuclear cells from peripheral blood from healthy human donorsprepared by Ficoll-Paque (Pharmacia, Piscataway, NJ), or lymphocytesharvested from spleen of C57BL/6 mice were used as effectors. 1 x IO6B16 cells were labeled for 2 h with 100 Â¿iCiof sodium [51Cr]chromatein RPMI supplemented with 3% fetal calf serum at 37Â°Cin a CO2

incubator, washed, incubated with 50 ^g/ml of DH2 in the samemedium for 30 min at 37Â°Cin a CO2 incubator, and washed again.

Some "Cr-labeled cells not treated with DH2 antibody were used fordirect effector cell lysis. 5'Cr-Labeled B16 cells treated with or withoutDH2 antibody were placed in 96-well round-bottomed plates (Costar,Cambridge, MA) at a density of 5 x 103/well, and incubated withvarious concentrations of effector cells for 4 h at 37Â°C.The plates were

then centrifuged at 500 x g for 5 min, and radioactivity was measuredin a 125-/J aliquot of each supernatant using a gamma counter. Spontaneous 5lCr release was determined in wells that contained only labeled

target cells treated with antibody. Total release was determined bysupernatant of the wells in which cells were lysed with 2% Triton X-100 and centrifuged. Release of 51Cr by antibody DH2 alone was

determined by incubating labeled B16 cells with (experimental release)or without (spontaneous release) 50 ^g/ml of DH2; the release of "Cr

by effector cells alone was determined by incubating labeled cells with(experimental release) or without (spontaneous release) coincubatingeffector cells for 24 h. Percentage of lysis was calculated as follows:

Experimental release â€”spontaneous releaseTotal release - spontaneous release

In Vivo Effect of DH2 Antibody on B16 Melanoma Growth in Mice.To study the inhibition of B16 melanoma cells in vivo by antibodyDH2, two groups of four C57BL/6 mice were given s.c. injections of 5x IO6B16 cells at each of two separated sites on the back (Day 0). On

Days 0, 2, 4, 6, 8, 10, 12, and 14, experimental group animals wereinjected with 400 Â¿igof DH2 antibody in 400 n\ of PBS via the tailvein. Control group animals were injected with 400 p\ of PBS on thesame day. Three diameters (di, d2, and d3) of the tumor were measured,and tumor volume was calculated by the formula ir(d, -d2-d3)/2.

Distribution Pattern of DH2 Antibody in Melanoma-bearing Mice.Tissue distribution pattern of the DH2 antibody in B16 melanoma-bearing mice was determined using '"I-labeled antibody DH2. ThreeC57BL/6 mice were injected with 5 x IO6 B16 melanoma cells s.c.

Drinking water for mice was changed to 0.1% KI 5 days before antibodyinjection. Ten days after melanoma cell inoculation, 20 Â¿ig(60 MCi)ofDH2 antibody was injected via the tail vein, and mice were sacrificed72 h later. After taking a blood sample from the cardiac cavity, PBSwas injected into the heart to flush blood from the tissues. Samplesfrom tissues and tumors were weighed and radioactivity was counted.In vivo tissue distribution was expressed as the ratio of radioactivity inthe tumor to that in normal tissues [(cpm/g in tumor)/(cpm/g in normaltissue)].

RESULTS

Reactivity and Specificity of Antibody DH2. Antibody DH2showed a peculiar differential reactivity with CMS and GMÃŒlactone on solid-phase radioimmunoassay, i.e., it reactedstrongly with GMj, but not with GM3 lactone when they weredried on plastic wells from their ethanol solutions (Fig. \A),whereas it reacted strongly with GMJ lactone but weakly withGM3 when they were adsorbed on gelatin-coated plates (Fig.IB). The same differential reactivity of M2590 antibody withGMJ lactone and GM3 was observed with two kinds of solid-

3â€¢JT

o^ 2

O.O-o 1

OCD

10 2.5 .625 .156 10 2.5 625 .156

Conc.DHZ Antibody (jjq/ml)Fig. 1. Reactivity of DH2 antibody with GM3and GMj lactone as determined

by solid-phase radioimmunoassay. A, 20 pmol/well CMS(â€¢)or <>â€¢,,,lactone (O)with PC and cholesterol in ethanol dried onto plastic plates. /'. <â€¢â€¢.,,(â€¢)or (Â¿M.lactone (O) dissolved in PBS adsorbed onto gelatin-coated plates. Following stepswere performed as described in "Materials and Methods." Data are the average

of triplicate experiments.

5681



IgGj MoAb ESTABLISHED AFTER GMJ LACTONE IMMUNIZATION

10 25 .625 156 10 2.5 625 156Cone. M2590 Antibody (pg/ml)

Fig. 2. Reactivity of M2590 antibody with GMJand GMJlactone as determinedby solid-phase radioimmunoassay. A, 20 pmol/well GMJ (â€¢)or GMJ lactone (O)with PC and cholesterol in ethanol dried onto plastic plates. K. (Â¡MI(â€¢)or GMJlactone (O) dissolved in PBS adsorbed onto gelatin-coated plates. Following stepsÂ»ereperformed as described in "Materials and Methods." Data are the average

of triplicate experiments.

oâ€” 4

Eo.e_>

Â¡2om

oO 10 20 30

Conc.GM3 (pmol/well)Fig. 3. DH2 antibody binding to (.-,,, as determined by solid-phase radio

immunoassay. Various concentrations of GMJwith PC and cholesterol in ethanolwere dried onto plastic plates. Following steps were performed as described in"Materials and Methods." Data are the average of triplicate experiments.

phase assay as well (Fig. 2, A and B). Thus, both antibodiesDH2 and M2590 may recognize a specific conformation of GMJlactone preferentially when lactone is adsorbed on a gelatin-coated surface. However, these antibodies do not recognize thelactone conformation, particularly a hydrophobic domain, whenlactone is dried from an ethanol solution (see "Discussion").

The reactivity of the DH2 antibody on ethanol-dried GMJshowed a dependence on GMJ concentration, i.e., the reactivityincreased greatly between 10 and 20 pmol/well (Fig. 3). Theantibody reacted specifically with GMJcontaining TV-acetylneuraminic acid, whereas it reacted minimally with GMJcontaining/V-glycolyl neuraminic acid or sialylparagloboside (Fig. 4). Thespecificity of the antibody assessed by thin-layer chromatogra-phy immunostaining of glycolipids separated on TLC is shownin Table 1. The antibody reacted only with GMJ containing N-acetyl neuraminic acid and its lactone, and not with any otherglycolipid tested. No obvious reactivity was observed with SPG,its lactone, GMJ ethyl ester, or the reduced form of GMJ, inwhich the carboxyl group of sialic acid was reduced to alcohol(termed GMJ gangliosidol). Since various types of lactone derived from other gangliosides, such as lactones of SPG, sialyllactonorhexaosylceramide, GMI, and Goib were all negative,antibody reactivity to lactone was limited to that of NeuAcGMs-

Comparative Reactivity of Antibodies DH2 and M2590. BothIgM antibody M2590 and IgG3 antibody DH2 showed a preferential reactivity with GM3 lactone over GM3 when adsorbedon gelatin-coated plates, but reacted only with GM3when driedfrom ethanol solution (Figs. 1 and 2). The binding of M2590on solid-phase GMJ lactone was clearly inhibited by both antibodies DH2 and M2590, but not by anti-Le" IgGj antibodySH-1, as shown in Fig. 5A. Similarly, the binding of M2590 tosolid-phase ethanol-dried GMJ was inhibited by both DH2 and

E 4Q.

Ã¤2m

NeuAcGM3

10 2.5 0.625 0.156Conc.DH2 Antibody (pg/ml)

Fig. 4. Reactivity of DH2 antibody with various glycolipids as determined bysolid-phase radioimmunoassay. 20 pmol/well of NeuAcGMj (â€¢),NeuGcGMj (O),sialylparagloboside (â€¢),or other glycolipids GMi. GDU, (Â¡i,ivG-n, lactosylcer-amide, and sialyllactonorhexaosylceramide (all are indicated as D) with PC andcholesterol in ethanol were dried onto plastic plates. Following steps wereperformed as described in "Materials and Methods." Data are the average of

triplicate experiments.

Table 1 Specificity of monoclonal antibody DH2 for glycolipids determined byTLC immunostaining

Glycolipid Reactivity"

NeuAcGM3 (dog erythrocytes)NeuAcGMj (B16 melanoma)NeuAcGMj (rat brain)NeuAcGMj lactone (dog erythrocytes)NeuGcGMjNeuGcGMj lactoneSialylparagloboside (IV'NeuAcnLc.,)

Sialylparagloboside lactoneSialyllactonorhexaosylceramide

(VI'NeuAcnLc*)

Sialyllactonorhexaosylceramide lactoneNeuAcGMj ethylesterNeuAcGMj gangliosidol

GMIGMI lactoneGDI.

Goib lactoneGT1Asialo (IM.-LactosylceramideGlucosylceramide

" +, positive reactivity; Â±,weak positive reactivity; â€”,negative reactivity.

100

co

50

50 12.5 3.12 50 12.5 3.12Cone.of Preincuboled Antibodies (pg/ml)

Fig. 5. Inhibition of M2590 binding to GM3lactone (A) or GMj (B) by DH2,M2590, and SH-1 antibodies. GMJ lactone was adsorbed onto gelatin-coatedplate. GMJwas dried onto plate with PC and cholesterol. Wells were preincubatedwith various concentrations of DH2 (â€¢),M2590 (O), and anti-Lex antibody SH-1 (D), followed by addition of '"I-labeled M2590. Details of the procedure aredescribed in "Materials and Methods."

M2590, but not by anti-Le" antibody SH-1, as shown in Fig.

5B. Thus, M2590 and antibody DH2 may recognize the sameepitope on GMJ and its lactone.

The ability of DH2 and M2590 to immunostain various typesof cells was compared, and the results are shown in Table 2.Quantitative reactivities of a few cell lines with M2590 andDH2 were compared by flow cytometry; the results are shownin Fig. 6. It is clear that the two antibodies showed nearly

5682




Table 2 Immunofluorescence tesi of various cell lines with monoclonal antibodiesDH2andM2590

CelllineB16B16F1BI6F10M2669M

1733M2981M2291P36-F4HMV-I491B10FM3A/F28-7FU

A169HTB19A431SW403MKN45K562A-204BUKNRKFisher

rat embryocellsDogerythrocytesRabbiterythrocytesHuman

erythrocytesOriginMouse

melanomaMousemelanomaMousemelanomaHumanmelanomaHumanmelanomaHumanmelanomaHumanmelanomaHumanmelanomaHumanmelanomaMouse

breastcarcinomaMousebreastcarcinomaHumanlungcarcinomaHumanepidermoidcarcinomaHumancoloncarcinomaHumangastriccarcinomaHuman

erythroleukemiaHumanrhabdomyosarcomaHamsterfibroblastRatfibroblastRat

fibroblastReactivity"

withDH2

M2590+++

+++++++++++++++++++_+ND*+

â€”â€”â€”â€”â€”++++Â±NDâ€”â€”â€”â€”â€”â€”NDâ€”

â€”Â±+++++++++++â€”â€”-

-

"+++, almost 100% of cells were positive; ++, more than 50% of cells were

positive; +, less than 50% of cells were positive; Â±,less than 1% of cells werepositive; -, negative.

* ND, not determined.

X ^.Control

FRE . ^Control

M2590

816 / :^.Control

DH2

B16 f

M2590

Relative fluorescence intensity

Fig. 6. Fluorescence-activated flow cytometric pattern of a few cell lines withantibodies DH2 and M2590.

identical reactivities. Those cells showing strong immunofluo-rescence with both antibodies were B16 mouse melanoma andits variants, mouse breast carcinoma FU A169, and dog erythrocytes; all of these highly reactive cells have been characterizedby a high concentration of CMS-

Inhibition of B16 Melanoma Cell Growth by Antibody DH2in Vitro. Cell growth of B16 melanoma was inhibited by thepresence of antibody DH2 at a concentration of 50 /ng/ml ascompared with the same concentration of antibodies M2590and CUI, as shown in Fig. 7. In a similar experiment, cell

growth of human colonie carcinoma cell line SW403, whichdoes not express GM.Iat the cell surface, was not affected byantibody DH2 (data not shown).

The cell growth inhibition induced by DH2 was clearly dosedependent. A clear inhibition was only observed at high concentrations of antibody (50-100 Â¿ig/ml). Interestingly, cellgrowth was significantly enhanced at low antibody concentration, as shown in Fig. 8.

The inhibition of B16 melanoma cell growth caused by DH2can be reversed if the cells are exposed to normal media withoutDH2 (data not shown).

Antibody-dependent Cellular Cytotoxicity. Antibody-dependent cellular cytotoxicity was clearly demonstrated by a lysis oftarget cells at a high effector/target ratio, as shown in Table 3.This lysis was observed in both human and mouse effector cells.

The release of 5lCr observed in this experiment was indeeddue to lysis of target cells by antibody-dependent cytotoxiceffector cells, since antibody alone without effector cells, oreffector cells alone without antibody, did not cause significantrelease of 5lCr under the same conditions. Release of 51Cr by

antibody DH2 alone during 24 h was 3.0%, and that by mouseor human effector cells without preincubation of antibody (atan effector/target ratio of 200) was 0.5 and 0.9%, respectively(see Table 3).

Inhibition of B16 Melanoma Cell Growth in Vivo by AntibodyDH2. Under the conditions and criteria of B16 melanoma

10

.o

ZA 48 72

Time (hour)

Fig. 7. Inhibition of B16 melanoma cell growth by antibody DH2 in vitro.B16 melanoma cells were seeded in 24-well plates, cultured as described in"Materials and Methods," and 50 >ig/ml of antibodies DH2 (â€¢),M2590 (A), ami-Cu-1 (â€¢),and PBS for control (O) were added 24 and 48 h later. The number ofcells was counted 43, 55, and 72 h after the beginning of the culture. Each datapoint represents the average of triplicate experiments. Standard deviation wasless than 15%.

15

â€” 10Ifl

â€¢!5

24 36 48 72Time (hour)

Fig. 8. Effect of various concentrations of antibody DH2 on B16 melanomacell growth. BI6 melanoma cells were seeded in 24-well plates and cultured asdescribed in "Materials and Methods." One hundred Mg/ml (â€¢),50 jig/ml (â€¢),

25 Mg/ml (A), and 12.5 Mg/ml (O) of DH2 antibody, and PBS as for control (O),were added 24 and 48 h later. The number of cells was counted 24, 36, 48, and72 h after beginning of the culture. Each data point represents the average oftriplicate experiments. Standard deviation was less than 15%.

5683



IgG, MoAb ESTABLISHED AFTER GMJ LACTONE IMMUNIZATION

Table 3 Antibody-dependent cellular cytotoxicity induced by monoclonal

antibody DH2 against B16 melanoma cells

Origin of effectorcellsHumanEffector/target

ratioEffector cell plus DH2

antibody (% lysis)Effector cell plus M2590

antibody (% lysis)Effector cell alone

(% lysis)DH2 antibody alone

(% lysis)200

20.32.00.9100507.0

0<1.0

00.5

ND"3.0*25

00

NDC57BL/6

mouse200

12.30.510050

6.32.90

0

0.5ND3.0Â»255.30ND

" ND. not determined.* Target cells labeled with "Cr and treated with 50 Â»ig/mlof DH2.

injection

l t M t I MJOOO

E,Â§ 2000â€¢

o>

3

injection

(titilli

1000oE

10 15 20 25 OTime (doys)

IO 15 20

Fig. 9. Effect of antibody DH2 on the growth of B16 melanoma tumor invivo. BI6 melanoma cells were injected s.c. into two sites on the back of C57BL/6 mice. Animals received i.v. injections of 400 ng of DH2 antibody in 400 , I ofPBS (A) or 400 Â»ilof PBS alone (B) on Days 0, 2, 4. 8, 10, 12, and 14, and thevolume of the tumor was measured. Details of the procedure are described in"Materials and Methods."

Table 4 Distribution ofnil-labeled monoclonal antibody DH2 in tissues ofB16melanoma-bearingmiceOrgan

ortissueTumorBloodBone

marrowThymusSpleenSkinMuscleBoneHeart

muscleThyroidand adjacenttissueLungLiverKidneyIntestineIntestinal

mesenteryBrainUrinary

bladderUterusand attached tissue(cpm/g

oftumor)/

(cpm/g oforgan)'1.001.04

Â±0.0242.00Â±19.2814.10

Â±0.0810.00Â±0.395.55

Â±0.3213.51Â±1.7314.62Â±1.657.81

Â±0.306.11Â±1.694.79Â±0.5414.53Â±

1.964.13+0.8812.32Â±1.1514.21

Â±0.7390.60Â±4.753.94Â±0.994.44Â±0.55%

Injecteddose/g oftissue"2.51

Â±0.372.21Â±0.010.08

Â±0.030.18Â±0.040.24

Â±0.050.45Â±0.090.19

Â±0.030.08Â±0.030.29Â±0.060.46Â±0.140.52Â±0.020.19

Â±0.050.66+0.130.21

Â±0.050.18Â±0.020.03

+0.000.68Â±0.090.56+0.05*

Mean value from triplicate experiments, Â±SE.

development in C57BL/6 mice described in "Materials andMethods," tumor growth was obviously inhibited in the exper

imental group as compared to the control group (Fig. 9). Itshould be noted, in particular, that growth of two tumors in theexperimental group was minimal. The lifespan after melanomainoculation in the experimental group was 36.8 Â±8.4 days(mean Â±SD); that for the control group was 25.5 Â±1.7 days.

Antibody distribution was determined in three B16 melanoma-bearing mice after injection of I25l-labeled antibody DH2(the procedure is detailed in "Materials and Methods"). The

data, as shown in Table 4, are values taken 72 h after injectionof 20 tig (60 nC\) of U5I-labeled DH2. The highest level of

activity was observed in original melanoma grown s.c. and in

blood samples, followed by urogenital tissue. The lowest activitywas found in bone marrow and brain.

DISCUSSION

(i\n is the most common widely distributed basic gangliosidein animals and humans. It is surprising, therefore, that monoclonal antibody established after syngeneic immunization withmelanoma (2) was directed to GMJ (3). However, the IgMantibody M2590, previously described (2), recognizes the density of GMJ on solid-phase, as well as at the cell surface; thereal immunogen on B16 melanoma cells is assumed to be GMJlactone, or GMJ having a lactone-like conformation. This assumption is based on the fact that antibody M2590 showedhigher affinity with (IMI lactone than with ( ;\]t. B16 melanomacontains GMJlactone, and the lactone had clear immunogenicitybased on its ability to induce hybridomas, while GMJ had nosuch ability (4). Recently, six monoclonal antibodies were established that were found to react with (.,,.<ganglioside lactoneas well as purified GD2 ganglioside (15). Some antibodies reacted only with the lactone. These previous results also suggestthe interesting possibility that ganglioside lactones, rather thannative gangliosides, would be more effective immunogens toelicit immune responses. The use of ganglioside lactones maybe an important way to establish antibodies directed to ganglioside antigens, which constitute the major tumor-associated andcell-type-specific antigens. Thus, in the present study immunization with GMJ lactone led to establishment of a hybridomaproducing an IgG3 antibody, DH2, which showed preferentialreactivity with GMJ lactone adsorbed on gelatin-coated plate.The antibody, however, did not show reactivity with GMj lactonedried from ethanol solution or coated with cholesterol-lecithin.Such a reactivity showed a close resemblance with M2590, andthe reactivity of M2590 to GMj lactone can be competitivelyinhibited by DH2. Thus, both M2590 and DH2 should bedirected primarily to GMJ lactone, while both antibodies areable to react with GMJ preferentially under certain conditions,such as solid phase adsorbed on plastic surface and dried fromethanol solution.

In striking contrast to M2590, which is an IgM antibody,DH2 displayed properties characteristic of an IgGj antibody,as listed below, despite the fact that the two antibodies couldbe directed to the same epitope.

(a) DH2 displayed a cytotoxic effect on both human andmouse effector cells, in contrast to M2590;

(b) DH2 showed a striking in vitro growth inhibition of B16melanoma cells at high concentration, but stimulated growthat low concentration. M2590 did not show such an effect.

(c) DH2 administered intravenously in C57BL/6 mice significantly inhibited B16 melanoma growth in vivo.

These results are comparable to the effect of IgG3 anti-GDjantibody on human melanoma cell growth /// vitro (16), itscytotoxic effect in vitro (17), and in vivo tumor growth (18).After administration of radiolabeled DH2 antibody to tumor-bearing mice, the antibody accumulates significantly in tumors,while circulating blood contains the same range of specificactivity. Accumulation of the antibody in various visceral organsis relatively low and is extremely low in brain. It is possible thatantibody DH2 reacts preferentially with cell surfaces having ahigh density of GMJ. Although GMJ is present in essentially alltissues, DH2 may not bind to GM3at the cell surface when GMJconcentration is low.

Recent analysis of tumor antigens with the monoclonal antibody approach clearly indicates that a number of gangliosidesconstitute tumor-associated antigens, e.g., GDJ in melanoma

5684




(19, 20); GD2in melanoma and neuroblastoma (21, 22); disialylparagloboside in colonie cancer (23); sialyl Le" (24, 25), disialylLe" (26), and sialyl Le" in various adenocarcinomas (27-29);

and CMS in melanoma as discussed in this paper. With a fewexceptions, these antibodies were obtained after extensive immunization of mice with tumor cells or tumor cell membranescontaining ganglioside as a major component. It is not inconceivable that lactones presumably present and associated withthese gangliosides may play a role in enhancing immunogenic-ity, in view of the results of our previous (4) and present studies.The data also suggest that gangliosides in lactone form aremore immunogenic than native gangliosides, and one wouldexpect to observe a stronger immune response. GMJ lactone wasdetectable in B16 melanoma cells by treatment of the cells withKB[3H]4 followed by autoradiographic demonstration of 'Hi-

labeled GMS ganglioside. Goib lactone was previously demonstrated in brain by mass spectrometry (30), although there isongoing debate whether such a lactone represents an artifact ofthe preparation process.

Even if the lactones are artifacts, immune response directedto gangliosides can be enhanced by immunization with lactonesrather than native gangliosides. Gangliosides in lactone formcould be useful immunogens for active immunization againsttumors. Such a study is in progress.

ACKNOWLEDGMENTS

We are grateful to Professor M. Taniguchi for the gifts of antibodyand cell lines. We also thank Drs. N. Hanai, H. Takahashi, A. Singhai,and T. Greene for their invaluable advice and assistance, and S. Anderson for expert preparation of the manuscript.

REFERENCES

1. Hakomori, S. Tumor associated carbohydrate antigens. Ann. Rev. Immunol.,2:103-126, 1984.

2. Taniguchi, M., and Wakabayashi, S. Shared antigenic determinant expressedon various mammalian melanoma cells. Jpn. J. Cancer Res., 75: 418-426,1984.

3. Hirabayashi, Y., Hanaoka, A., Matsumoto, M., Matsubara. T., Tagawa, M.,Wakabayashi, S., and Taniguchi, M. Syngeneic monoclonal antibody againstmelanoma antigen with interspecies cross-reactivity recognizes GM3, a prominent ganglioside of B16 melanoma. J. Biol. Chem., 260: 13328-13333,1985.

4. Ã‘ores, G., Dohi, T., Taniguchi, M., and Hakomori, S. Density-dependentrecognition of cell surface CM] by a certain anti-melanoma antibody, andGM] lactone as a possible immunogen: requirements for tumor-associatedantigen and immunogen. J. Immunol., 139: 3171-3176, 1987.

5. Klenk, E., and Heuer, K. Ãœberdie Ganglioside von Hundeserythrozyten.Zeitschr. Verdaunungs Stoffwechselkrankheiten, 20: 180-183, 1960.

6. Handa. S., and Nakamura, K. Modification of sialic acid carboxyl group ofganglioside. J. Biochem. (Tokyo), 95: 1323-1329. 1984.

7. MacDonald. D. L., Patt, L. M., and Hakomori, S. Notes on improvedprocedures for the chemical modification and degradation of glycosphingo-lipids. J. Lipid Res., 21: 642-645, 1980.

8. Yamakawa, T., and Suzuki. S. The chemistry of lipids of posthemolyticresidue or stroma of erythrocytes: I. Concerning the ether-insoluble lipids oflyophilized horse blood serum. J. Biochem. (Tokyo), 38: 199-212, 1951.

9. Watanabe. K., Powell, M. E., and Hakomori. S. Isolation and characterization of gangliosides with a new sialosyl linkage and core structure: gangliosides of human erythrocyte membranes II. J. Biol. Chem., 254: 8223-8229,1979.

10. Seyama, Y., and Yamakawa, T. Chemical structure of glycolipid of guineapig red blood cell membrane. J. Biochem. (Tokyo), 75: 837-842, 1974.

11. Young, W. W., Jr., MacDonald. E. M. S., Nowinski, R. C. and Hakomori,S. Production of monoclonal antibodies specific for distinct portions of the

glycolipid asialo GMÂ¡(gangliotriosylceramide). J. Exp. Med., 750: 1008-1019, 1979.

12. Kannagi, R., Stroup, R., Cochran, N. A., Urdal, D. L., Young, W. W., Jr.,and Hakomori, S. Factors affecting expression of glycolipid tumor antigens:influence of ceramide composition and coexisting glycolipid on the antige-nicity of gangliotraosylceramide in murine lymphoma cells. Cancer Res., 43:4997-5005, 1983.

13. Magnani, J. L., Smith, D. F., and Ginsburg, V. Detection of gangliosidesthat bind cholera toxin: direct binding of '"I-labeled toxin to thin-layerchromatograms. Anal. Biochem., 109: 399-402, 1980.

14. Grabstein, K. Cell-mediated cytolytic responses. In: Mishell, B. B., and Shiigi,S. M. (eds.), Selected Methods in Cellular Immunology, pp. 124-137. SanFrancisco: Freeman & Co., 1980.

15. Tai, T., Kawashima, I., Tada, N., and Ikegami. S. Different reactivities ofmonoclonal antibodies to ganglioside lactones. Biochim. Biophys. Acta, 958:134-138, 1988.

16. Dippold, W. G., Knuth, A., and zum Buschenfelde, K-H. M. Inhibition ofhuman melanoma cell growth in vitro by monoclonal anti-GD3-gangliosideantibody. Cancer Res., 44: 806-810, 1984.

17. Welt, S. W., Carswell, E. A., Vogel, C-W., Oettgen, H. F., and Old, L. J.Immune and nonimmune effector functions of IgG3 mouse monoclonalantibody R24 detecting the disialoganglioside GD3 on the surface of melanoma cells. Clin. Immunol. Immunopathol.. 45: 214-229, 1987.

18. Houghton, A. N., Mintzer, D., Cordon-Cardo, C.. Welt, S. W., Fliegel, B.,Vadhan, S., Carswell, E., Melamed, M. R., Oettgen, H. F., and Old, L. J.Mouse monoclonal IgG3 antibody detecting <.,,. ganglioside: a phase I trialin patients with malignant melanoma. Proc. Nati. Acad. Sci. USA, 82:1242-1246, 1985.

19. Dippold, W. G., Lloyd. K. O., Li, L. T., Ikeda, H., Oettgen, H. F., and Old,L. J. Cell surface antigens of human malignant melanoma: definition of sixantigenic systems with monoclonal antibodies. Proc. Nati. Acad. Sci. USA,77:6114-6118, 1980.

20. Nudelman, E., Hakomori, S., Kannagi, R., Levery, S. B., Yeh. M-Y., Hells-trom, K. E., and Hellstrom, I. Characterization of a human melanoma-associated ganglioside antigen defined by a monoclonal antibody. 4.2. J. Biol.Chem., 257: 12752-12756, 1982.

21. Cahan, L. D., Irie, R. F., Singh, R., Cassidenti, A., and Paulson, J. C.Identification of a human neuroectodermal tumor antigen (OFA-1-2) asganglioside GD2.Proc. Nati. Acad. Sci. USA, 79: 7629-7633, 1982.

22. Tai. T., Paulson, J. C., Cahan, L. D., and Irie, R. F. Ganglioside GM2 as ahuman tumor antigen (OFA-I-1). Proc. Nati. Acad. Sci. USA, SO: 5392-5396, 1983.

23. Hansson, G. C., Karlsson, K-A., Larson, G., Stromberg, N., Teneberg, S.,Thurin, J., Brodin, T., Sjogren, H-O., Hellstrom, I., and Hellstrom, K-E.Characterization of glycolipid antigens of rat colon carcinoma including theuse of monoclonal antibodies directed against human melanoma gangliosides(abstract). In: M. A. Chester, D. Heinegard, A. Lundblad, and S. Svensson(eds.), Proceedings of the 7th International Symposium on Glycoconjugates,pp. 854-855. Lund-Ronneby, Sweden, 1983.

24. Magnani, J. L., Nilsson, B., Brockhaus, M., Zopf, D., Steplewski, Z.,Koprowski, H., and Ginsburg, V. A monoclonal antibody-defined antigenassociated with gastrointestinal cancer is a ganglioside containing sialylatedlacto-A'-fucopentaose II. J. Biol. Chem., 257: 14365-14369, 1982.

25. Chia, D., Terasaki, P. I., Suyama, N., Gallon, J., Hirota, M., and Katz, D.Use of monoclonal antibodies to sialylated Lewis* and sialylated Lewis* forserological tests of cancer. Cancer Res., 45:435-437, 1985.

26. Nudelman. E., Fukushi, Y., Levery, S. B., Higuchi, T., and Hakomori, S.Novel fucolipids of human adenocarcinoma: IV. Disialosyl Le* antigen(III4FucIII6NeuAcIV3NeuAcLc6) of human colonie adenocarcinoma and the

monoclonal antibody (FH7) defining this structure. J. Biol. Chem., 261:5487-5495, 1986.

27. Fukushi, Y., Nudelman, E., Levery, S. B., Rauvala, H., and Hakomori,S. Novel fucolipids accumulating in human cancer: III. A hybridoma antibody (FH6) defining a human cancer-associated difucoganglioside(VI3NeuAcV3III3Fuc2nLCi). J. Biol. Chem., 259: 10511-10517, 1984.

28. Fukushima, K., Hirota, M., Terasaki, P. I., Wakisaka, A., Togashi, H., Chia,D., Suyama, N., Fukushi, Y., Nudelman, E., and Hakomori, S. Characterization of sialosylated Lewis" as a new tumor-associated antigen. CancerRes., 44: 5279-5285, 1984.

29. Fukushi, Y., Kannagi, R., Hakomori, S., Shepard, T., Kulander, B., andSinger, J. W. Localization and distribution of difucoganglioside(VI3NeuAcV3III3FuC2nLc6)in normal and tumor tissues defined by its monoclonal antibody FH6. Cancer Res., 45: 3711-3717, 1985.

30. Riboni, L., Sennino, S., Acquotti, D., Malesci, A., Ghidoni, R., Egge, H.,Minutino. S.. and Tettamanti, G. Natural occurrence of ganglioside lactones:isolation and characterization of GDÂ¡binner ester from adult human brain. J.Biol. Chem., 261: 8514-8519, 1986.

31. Svennerholm, L. Chromatographie separation of human brain gangliosides.J. Neurochem., 10:613-623, 1963.

5685



1988;48:5680-5685. Cancer Res Taeko Dohi, Gustavo Nores and Sen-itiroh Hakomori

in Vivo and in Vitroof Melanoma Cell Growth Lactone: Immunochemical Specificity and InhibitionM3with G

Monoclonal Antibody Established after Immunization3An IgG

Updated version

http://cancerres.aacrjournals.org/content/48/20/5680

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/48/20/5680To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



anigg3monoclonalantibodyestablishedafterimmunizationwithgm...

Documents