[CANCER RESEARC II 51. 144-149. January I. 199I|

Effect of a Chimeric Anti-Ganglioside GD2Antibody on Cell-mediated Lysis ofHuman Neuroblastoma Cells'

Edward Barker, Barbara M. Mueller, Rupert Handgretinger, Michael Herter, Alice L. Yu, and Ralph A. Reisfeld2

Department of Immunology, Research Institute of Scripps Clinic, La Julia, California 9203 7[E. R., B. M. M., R. A. R./; Department of Hematology. Children 's Hospital,University of Tubingen, D-7400 Tubingen. Federal Republic of Germany /R. //., M. //./; and Department of Pediatrics, University of California Medical Center,San Diego, California 92103 fA. L. }'./

ABSTRACT

An ;imi-(.,,.. ganglioside human/mouse chimeric monoclonal antibody,chl4.l8, like its murine counterpart, 14.G2a, »asshown to bind to humannvuroblastoma cells. I his chimeric antibody proved to be more effectivethan l4.G2a in mediating the lysis of neuroblastoma cells with humaneffector cells, such as granulocytes and natural killer cells within theperipheral blood mononuclear cell population. A comparison of these twoeffector cell populations isolated from the same donor revealed granulo-cytes to be more effective than peripheral blood mononuclear cells inlysing neuroblastoma cells, which »erecoated with monoclonal antibodychi4.18. Addition of recombinant human granulocytc-macrophage colony-stimulatory factor increased chl4.18-mediated lysis of neuroblastoma cells by granulocytes but not by peripheral blood mononuclearcells. In fact, granulocytes were effective in mediating lysis of neuroblastoma cells coated with chi4.18 irrespective of whether they wereobtained from normal adults or from ncuroblastoma patients.

INTRODUCTION

Neuroblastoma is one of the most common solid tumors ofchildren that generally has a poor prognosis, since at the timeof diagnosis, the majority of patients have widespread disease(1-3). The treatment available for such patients consists ofsurgical removal of solid tumor(s) followed by massive chemotherapy with or without radiotherapy and/or bone marrowreconstitution (4-7). Such treatment protocols can result incomplete remission in some patients; however, in many suchcases, the relapse rate is relatively high. It is apparent fromthese findings that additional adjuvant treatment modalities areneeded to improve the outcome of this neoplastic disease.

In this regard, the use of mAbs' directed to neuroblastoma-

associated antigen has been pursued both in the diagnosis andtreatment of neuroblastoma (8-11). One such antigen targetedby mAbs that is frequently expressed on neuroblastoma isdisialoganglioside G|)2, a glycolipid which is expressed on tumorcells of neuroectodermal origin (10, 12, 13). The Gi)2 antigenis ideal for mAb-mediated therapy of neuroblastoma, since it isexpressed at high density on almost all neuroblastoma cells, ispoorly expressed or absent from most normal tissue, and doesnot modulate its expression upon mAb binding.

Several mouse monoclonal anti-G,,: antibodies were reportedto effectively suppress the growth of tumors of neuroectodermal

Received 8/8/90; accepted 10/9/90.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 IJ.S.C. Section 17.14solely to indicate this fact.

1The work was supported by NIH Grant CA42508. This is the ResearchInstitute of Scripps Clinic's Manuscript 6470IMM.

2To whom requests for reprints should be addressed, at Department of

Immunology. IMMI3. Research Institute of Scripps Clinic. 10666 N. TorreyPines Road. La Jolla. CA 92037.

"The abbreviations used are: mAb. monoclonal antibody: GM-CSF. granulo-cyte-macrophage colony stimulatory factor; NK. natural killer: PBMC. peripheralblood mononuclear cell; ADCC. antibody-dependent cellular cytotoxicity: FITC.fluoresccin ¡sothiocyanate: PBS. phosphate-buffered saline: M-CSF. monocytecolony stimulatory factor; E:T. effectortarget.

origin in athymic (nu/nu) mice (14, 15), and one was shown tocause partial, as well as complete, regression in Phase I clinicaltrials (8, 16). The mechanisms believed to be involved in theability of anti-Go? mAb to suppress the growth of neuroectodermal tumors in vivo are antibody-mediated cellular cytotoxicity and/or complement-dependent cytotoxicity.

There are, however, several drawbacks in using murine monoclonal antibodies for therapeutic purposes in humans. First,such antibodies induce a human anti-mouse antibody response(17, 18); second, the half-life of murine mAbs in the circulationis relatively short compared to human immunoglobulin (19);and third the Fc portion of murine antibody may not elicitADCC or complement-dependent cytotoxicity as effectively asthe Fc portion of a human antibody (20). To overcome theseproblems, recombinant DNA technologies have been appliedto develop human/mouse chimeric monoclonal antibodies, containing the variable region of the murine antibody and theconstant region of human immunoglobulin heavy and lightchains (21).

The monoclonal antibody chi4.18 is an example of such ahuman/mouse chimeric antibody that is directed to gangliosideGo; and which was developed by combining the complementaryDNA sequences encoding the constant portion of human 7!heavy chain and Klight chain with those encoding the variableportions of immunoglobulin from murine hybridoma 14.18(22). This chimeric antibody bound to melanoma tumor cellsi/i vitro and reacted with melanoma tumors in athymic nudemice equally well as the equivalent mouse antibody. In addition,while chi4.18 was equal to its murine equivalent in mediatingcomplement-dependent lysis of melanoma cells, it proved considerably better in mediating ADCC with peripheral bloodmononuclear effector cells (23).

Here, we report the effects of chi4.18, as compared to itsmurine equivalent 14.G2a in ADCC of neuroblastoma cells byPBMC from normal donors, as well as from neuroblastomapatients. In addition, we established the population(s) of cellsfound in the peripheral blood that are primarily involved inchl4.18-mediated lysis of neuroblastoma cells. This study alsoserved to determine whether cytokines affect the magnitude oflysis of neuroblastoma cells by peripheral blood cells in thepresence of the human/mouse chimeric antibody.

MATERIALS AND METHODS

Cell Lines and Monoclonal Antibodies. The human neuroblastomacell lines NMB-7 and 1MR-32 were a gift from Dr. S. K. Liao (Mc-Master University. Hamilton, Ontario, Canada). These cells weregrown in RPMI 1640 medium (Whittaker Bioproducts, Walkersville,MD) and supplemented with I09r fetal bovine serum (Whittaker Bio-products). The adherent cell lines were detached with 0.5 ITIMEDTA(Sigma Chemical Co., St. Louis, MO), 0.02 M 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid (Fisher Chemicals, Fair Lawn. NJ). and0.15 M NaCl (Mallinckrodt. Paris. KY), washed once, and then used asa single cell suspension.

144

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

CHIMERIC ANTI-CD! ANTIBODY-MEDIATED LYSIS OF NEIROBI,ASTOMA

The hybridoma producing 14.G2a was established in our laboratory,as reported previously (15) and was produced in large amounts byBioTechnetics. San Diego, CA. The construction of the human/mousechimeric antibody, chi4.18, has been described elsewhere (22).

Indirect Immunofluorescence. Neuroblastoma cells were incubatedfor l h at 4°Cwith 10 fig/ml of mAb. The cells were washed 3 times

with PBS (Whittaker Bioproducts) containing 0.5% bovine serum albumin (Boehringer Mannheim Biochemicals, Indianapolis, IN) and0.02% sodium azide (Mallinckrodt), followed by subsequent stainingwith either goat anti-human Fab FITC (for ch 14.18) or goat anti-mouseIgG plus IgM-FITC for l h at 4°C.Both FITC-coupled antibodies were

obtained from BMB, Indianapolis, IN. The cells were washed 3 timeswith PBS containing 0.5% bovine serum albumin and 0.02% sodiumazide prior to analyzing an aliquot of 1 x IO4 cells on a Fluorescent

Activated Cell Sorter 400 flow cytometer (Becton Dickinson, MountainView, CA).

Effector Cells. The source of effector cells was heparinized venousblood obtained from either healthy adult donors or from stage IVneuroblastoma patients at least 4 weeks after their last therapy. Thewhole blood was layered onto a Ficoll-Paque (Pharmacia, Uppsala.Sweden) gradient (1.077 g/ml) and centrifuged at 800 x g for 25 min.PBMC were collected at the interface between the Ficoll-Paque and theplasma. Subpopulations of PBMC from normal donors were isolatedby adherence of PBMC to gelatin/fibronectin-coated flasks (24). Thenonadherent cell preparation consisted of 65-80% CD3+ cells, 5-10%CD57* cells, and <1% CD14* cells, as determined by microfluoro-metric analysis of 1 x 10" cells with fluorescein-conjugated antibodies

(Becton Dickinson) directed to various cell surface markers. Adherentcells were obtained by treating PBMC, which were attached to thegelatin-fibronectin-coated flasks with EDTA. This adherent populationwas composed of >90% CD14+ cells and a small population of <1%CD3* and <0.1% CD57* cells. NK cells were isolated by depleting T-

and B-lymphocytes from nylon wool-nonadherent cells by an immuno-magnetic method. To this end, nylon wool-nonadherent cells were firstincubated with anti-CD3 and anti-CD 19 antibodies (Becton Dickinson)for 30 min at 4'C, washed twice with PBS. and then further incubated

under continuous shaking with sheep anti-mouse-coated immunobeads(Dynabeads; Dynal, Great Neck, NY) for l h at 4°Cat a ratio of 20

beads/lymphocyte. Cells were then separated by placing them into amagnetic field (25). Granulocytes were isolated from pellets after frac-tionation of blood on a Ficoll gradient. Contaminating erythrocyteswere removed by prior lysis of cell pellets with either distilled water orammonium chloride. This cell preparation contained 95-99% granu-locytes, as determined by microfluorometric analysis of an aliquot of 1x 104cells by indirect staining with the granulocyte-specific monoclonalantibody PMN-8C7 (Sigma Chemical Co.) as the primary antibody (26,27).

Antibody-dependent Cellular Cytotoxicity. Tumor cells (5 x IO5)werelabeled for 2 h at 37°Cwith 0.5 mCi of sodium "Cr (10-35 mCi/ml,

Amersham Corp., Arlington Heights, IL) in 0.5 ml RPMI 1640. Afterwashing these cells 3 times, aliquots of 1-2 x IO3 cells in a 25-pivolume were plated in 96-well U-bottomed microtiter plates (CorningLaboratory Sciences, Corning, NY). Human/mouse chimeric mAbchl4.18 or murine mAb 14.G2a were added each in a 50-^1 volume totriplicate wells containing tumor cells. The concentration of antibodiesranged between 1 and 0.001 Mg/ml. Effector cells were placed intomicrotiter wells in a total volume of 100 p\ at effector:target cell ratiosof 200:1, 100:1, 50:1, or 25:1. In assays in which either 10-100 ng/mlof human recombinant GM-CSF (a gift of BehringWerke, Marburg.Federal Republic of Germany) or 1000 units/ml of human recombinantM-CSF (Genzyme, Boston, MA) were admixed, the total volume ofcells added was reduced to 50 /¿I.with the other 50-^1 volume consistingof either the cytokines or media. Lysis of target cells mediated byeffector cells alone was determined for each blood donor by incubatingthe tumor cells and the effector cells together in the absence of antibody.The plates were incubated for 4 h at 37°Cand then centrifuged at 100

x g for 2 min. A lOO-^ilaliquot of each supernatant was harvested andanalyzed for radioactivity in a Comp/Gamma counter (Pharmacia LKBBiotechnology, Gaithersburg, MD). Total "Cr release was measured

by lysing the tumor cells with 10% sodium dodecyl sulfate. Spontaneous"Cr release was determined in wells that contained only tumor cells.

Percentage of specific lysis was calculated as:

Experimental "Cr release - mean spontaneous "Cr releaseMean maximal "Cr release - mean spontaneous "Cr release x 100

The data are presented as the mean percentage of specific release ofthree replicates. Although the percentage of specific "Cr release exerted

by effector cells from various individuals varied from experiment toexperiment, the pattern of the results remained consistent.

RESULTS

Binding of Human/Mouse Chimeric mAb chl4.18 to Surfaceof Neuroblastoma Tumor Cells. The NMB-7 neuroblastoma cellline used as a target cell throughout this study is known toexpress the ganglioside Gl)2 (14). Here, it was determinedwhether chi4.18 mAb could recognize Gr>2and bind to theNMB-7 cell line. As a control, the same neuroblastoma cell linewas tested for its ability to bind murine mAb 14.G2a. As shownin Fig. 1, both chl4.18 and 14.G2a stained all the NMB-7 cells,as determined by indirect immunofluorescence.

mAb chl4.18-mediated Lysis of Neuroblastoma Tumor Cellsby PBMC. Here, we compared the ability of chi4.18 and itsmurine counterpart, 14.G2a, to mediate the lysis of neuroblastoma cells by PBMC. As shown in Fig. 2, chi4.18 wasmore effective in mediating the lysis of NMB-7 neuroblastomacells by PBMC at an E:T of 100:1, since less chi4.18 wasrequired than the murine mAb to achieve the same lytic effect.Proportionately, similar results were obtained at E:T ratios of

CD

CJ

IO2 IO3 104

450-

l

3

10° IO1 102 IO1 10«

FluorescenceIntensityFig. I. Flow cytometric analysis of neuroblastoma cell line NMB-7 after

staining of cells with murine mAb l4.G2a and goat anli-mouse-FITC (Fig. \A)or human/mouse chimeric mAb chi4.18 and goat anti-human FITC (Fig. IB).

. staining with secondary antibody alone; , staining with primary andsecondary antibody.

145

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

CHIMERIC ANTI-GD¡ANTIBODY-MEDIATED LYSIS OF NEUROBLASTOMA

I01

t

Concentration of Monoclonal Antibody [..q ml]Fig. 2. Antibody-dependent cellular cytotoxicity mediated by PBMC' against

NMB-7 neuroblastoma tumor cells in the presence of either chi4.18 (•)orl4.G2a (D) antibodies. Cytotoxicity was measured in a 4-h 5lCr release assay atE:T of 100:1. D, level of lysis of NMB-7 cells mediated by PBMC in the absenceof antibody.

Table I Antibody-dependent cytotoxic activity of PBMC from neuroblastomapatients against IMR-32 neuroblastoma cells in the presence of either chi 4. IS or

l4.G2a antibodies

Concentration ofPatient antibody (¿¿g/ml)

% of specific "Cr release at an E:T of:

10:1 5:1 2.5:1 1.25:1

J.B.J.B.J.B.J.B.K.

RK.RK.RK.RF.

K.F.K.F.

K.•Chl4.02.50.040.00502.50.040.00502.50.040.00518

antibody.039-/20*50/1046/122233/3545/3853/241843/4257/4858/50030/1634/633/4522/2135/2535/26421/1830/2130/20015/725/321/2222/1525/1225/12014/1020/1024/11013/512/014/0017/1020/615/807/19/417/3*14.G2a antibody.

200:1 and 50:1 (data not shown). This observation is not limitedto the NMB-7 cell line, since LAN-1 and IMR-32 neuroblastoma cells were also lysed by PBMC more efficiently whenmediated by chi4.18 rather than 14.G2a (data not shown).

The results depicted in Fig. 2 are from experiments that madeuse of PBMC isolated from normal adults. Additional studieswere done to assess whether similar results could be achievedwith PBMC from neuroblastoma patients. Table 1 summarizesthe results that were obtained with PBMC isolated from theblood of 3 such patients that were tested individually. Overall,chi4.18 mediated the lysis of neuroblastoma tumor cells byPBMC from neuroblastoma patients more effectively than14.G2a.

Extent of mAb chl4.18-mediated Lysis of NeuroblastomaCells by Different Subpopulations of PBMC. In order to test theeffect of different PBMC subpopulations on ADCC, these cellswere separated based on their ability to adhere to fibronectin-coated surfaces (24). Adherent and nonadherent cells thusobtained were assayed for their ability to mediate the lysis ofNMB-7 neuroblastoma cells with mAb chi4.18. As illustratedin Fig. 3, at E:T of 100:1, effective chl4.18-mediated lysis ofNMB-7 cells was achieved with the nonadherent PBMC population and only marginal lysis was obtained with adherentPBMC. A similar pattern of results was observed at E:T of 50:1and 25:1 with proportionately lower values (data not shown).The addition of M-CSF (1000 units/ml) to this assay did notincrease the ability of adherent PBMC to lyse NMB-7 tumor

cells in the presence of chi4.18 (data not shown). In ADCCexperiments with enriched NK cells from healthy adult donors(>90% CD56+ cells), chi4.18 was more effective than 14.G2a,

especially at lower antibody concentrations of 5-2.5 mg/ml andlower E:T ratios of 2.5:1 to 1.25:1 (Table 2).

Comparison of chi4.18-and 14.G2a-mediated Lysis of Neuro-blastoma Tumor Cells by Granulocytes and PBMC. Here, wewanted to determine whether granulocytes, as well as PBMC,can mediate the lysis of neuroblastoma with ch 14.18 better thanmurine mAb 14.G2a. The data in Fig. 4 illustrate that wheneverthe concentration of antibody used in these experiments washigh (>0.1 ¿ig/ml),granulocytes were capable of mediating thelysis of neuroblastoma equally well with both mAbs. However,at antibody concentrations of <0.1 /ig/ml, the ability of mAbchl4.18 to mediate the lysis of NMB-7 tumor cells by granulocytes exceeded that of its murine counterpart. Fig. 4 showsresults obtained at E:T of 100:1. The same experiments werealso done at E:T of 200:1 and 50:1, where essentially the samepattern of results was obtained (data not shown).

In an effort to assess whether chl4.18-mediated lysis ofNMB-7 cells was more effective with granulocytes than PBMC,these effector cells were isolated from the blood of the samedonors and compared for their ability to mediate tumor celllysis in the presence of chi 4.18. The data in Fig. 5 demonstratethat granulocytes arc superior to PBMC in mediating chi 4.18-directed lysis of NMB-7 cells, since it required less antibodyfor tumor cell lysis mediated by granulocytes than by PBMC.The data in Fig. 5 are from experiments done at an E:T ratio

Concentration of Monoclonal Antibody (i:q ml]

Fig. 3. Effect of adherent and nonadherent PBMC in mediating chi4.18antibody-dependent cytotoxic activity against NMB-7 ncuroblastoma tumor cells.Unfractionated PBMC (D). adherent PBMC (D). and nonadherent PBMC (•)were tested for their ability to mediate ADCC activity. Cytotoxic activity wasmeasured in a 4-h "Cr release assay at E:T of 100:1.

Table 2 Antibody-dependent cytotoxic activity of enriched NK celts from normaldonors against IMR-32 neurohlaswma cells in the presence of either chI4.I8 or

l4.Ga2 antibodies.% of specific 5lCr release at an E:Tof:Concentration

ofDonorantibody(^g/ml)1

02.50.040.0050.00252

05.00.160.040.020.01•CH14.I8.•

14.G2a.10:12677-/7I*74/6775/5170/484085/7477/6777/5775/5679/415:1762/5465/5659/3952/322570/6873/6273/4565/3561/412.5:1018/1632/2028/1116/51950/4258/4254/2750/2449/231.25:107/48/49/25/07ND33/2133/1528/1328/12

146

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

CHIMERIC ANTI-GDI ANTIBODY-MEDIATED LYSIS OF NEUROBLASTOMA

1.0 0.1 0.01 0.001 0.0

XI^T^iÃŽL^J-t

Concentration of Monoclonal Antibody [,.g ml]

Fig. 4. Antibody-dependent cylotoxicity mediated by granulocytes againstNMB-7 neuroblastoma tumor cells in the presence of either chi4.18 (•)or14.G2a (^) antibodies. Cytotoxicity was measured in a 4-h "Cr release assay atE:T of 100:1. D, level of lysis of NMB-7 cells mediated by granulocytes in theabsence of antibody.

¡^ 20

Q.CO

1.0 0.1 0.01 0.0075 0.0025 0.001 0.0

Concentration of Monoclonal Antibody [ug/ml]Fig. 5. Human/mouse chimeric antibody-dependent cytotoxicity of NMB-7

neuroblastoma tumor cell line mediated by either granulocytes (•)or PBMC (D).Cytotoxicity was measured in a 4-h 5lCr release assay at E:T of 100:1.

of 100:1. E:T ratios of 200:1 and 50:1 produced essentially thesame pattern of results (data not shown).

In this study, we determined whether chl4.18-mediated lysisof NMB-7 by granulocytes could also be augmented by GM-CSF in a 4-h 51Cr release assay at E:T of 50:1 (Fig. 6). Atconcentrations of 100 and 10 ng/ml of the cytokine, GM-CSFdid indeed augment the ability of the granulocytes to mediatelysis of NMB-7 tumor cells with chi4.18. In contrast, anotheraliquot of the same batch of GM-CSF was not effective inenhancing chl4.18-mediated lysis of NMB-7 tumor cells byPBMC (data not shown).

Comparison of chl4.18-mediated Lysis of Neuroblastoma Tumor Cells by PBMC and Granulocytes from NeuroblastomaPatients. Experiments were performed to determine whethergranulocytes from neuroblastoma patients are able to lyseNMB-7 neuroblastoma cells in the presence of chi 4.18 equallywell as granulocytes from normal adult donors. In addition, theextent of chi4.18-mediated lysis of NMB-7 by granulocytesfrom neuroblastoma patients was compared with that achievedby PBMC isolated from the same patient. These analyses weredone at E:T ratios of 75:1 to 25:1 and at an antibody concentration of 1 ¿ig/m'-As demonstrated in Fig. 7. similar togranulocytes from normal adults, granulocytes from neuroblastoma patients proved superior to PBMC in chi4.18-mediated lysis of NMB-7 tumor cells. In the presence of 100 ng/

ml of GM-CSF, a further increase in chi 4.18-mediated lysis bygranulocytes was obtained, but only at low E:T ratios. Incontrast, an aliquot of the same batch of GM-CSF failed toincrease the ability of PBMC to lyse NMB-7 tumor cells in thepresence of chi4.18. Taken together, these data indicate thatin the presence of chi4.18, granulocytes from both neuroblastoma patients and normal adults have a greater ability tolyse NMB-7 neuroblastoma cells than PBMC isolated from thesame individuals, and that the magnitude of this lysis is increased even further following stimulation of these effector cellswith GM-CSF.

DISCUSSION

The human/mouse chimeric antibody ch 14.18, like its murinecounterpart, 14.G2a, binds to neuroblastoma tumor cells. Thischimeric antibody also proved to be more efficient than 14.G2ain mediating the lysis of neuroblastoma tumor cells with sucheffector cells as NK cells within nonadherent PBMC population, as well as granulocytes. When these two effector cellpopulations were compared, granulocytes were far better than

a.a»

0.1 0.01 0.0

Concentration of Monoclonal Antibody [ug/ml]Fig. 6. Chi 4.18-mediated lysis of NMB-7 neuroblastoma tumor cells by either

granulocytes in the absence (Q) or presence of either 100 ng/ml (D) or 10 ng/ml(•)recombinant human GM-CSF. Cytotoxicity was measured in a 4-h "Crrelease assay at E:T of 50:1.

Ü

CO

25/1 50/1 75/1

Effector to target cell ratioFig. 7. Ch 14.18-mediated lysis of NMB-7 ncuroblastoma tumor cells by either

granulocytes (O. •)or PBMC (A. A) from the same neuroblastoma patient in theabsence (D. A) or presence (•.A) of recombinant human GM-CSF (100 ng/ml).Cytotoxicity was measured in a 4-h 51Cr release assay at various E:T ratios of

75:1. 50:1. and 25:1. The concentration of chimeric antibody used was 1 i/g/ml.

147

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

CHIMERIC ANTI-GDI ANTIBODY-MEDIATED LYSIS OF NEUROBLASTOMA

PBMC in lysing neuroblastoma cells in the presence of ch 14.18.Human recombinant GM-CSF further increased chi4.18-me-diated lysis of neuroblastoma tumor cells by granulocytes, aneffect which was not obtained with PBMC.

We were interested in determining the phenotype of the cellswithin the PBMC population, which are primarily responsiblefor chl4.18-mediated lysis of neuroblastoma cells. It appearsfrom our data that the prevalent effector cells in this case aremost likely lymphocytes and not monocytes. as evident fromour results with enriched NK cells (Table 2). This contention isstrengthened by the observation that nonadherent PBMC, suchas those that failed to adhere to fibronectin, were more effectivein lysing neuroblastoma cells coated with ch 14.18 than adherentPBMC. Since fibronectin receptors are present on monocytes,but not lymphocytes (24), the more effective nonadherent cellsare likely to be lymphocytes, rather than monocytes. Thiscontention was confirmed by microfluorometric analysis, indicating that cells within the adherent population are 2:90%CD14+, <1% CD3+ cells, and <0.1% CD57+. while the non-adherent lymphocytes consist of 65-80% CD3* and 5-10%CD57* cells with <1% CD 14* monocytes.

However, the participation of monocytes in the chi 4.18-mediated lysis of neuroblastoma cells cannot be completelyruled out, since only a 4-h lytic assay was used in our studies.In this regard, others have reported monocytes as effector cellsin chimeric antibody-mediated lysis of colorectal carcinomacells in lytic assays involving incubation times of at least 18 h(25). Even though examples can be found where monocytesserve as effector cells in ADCC of certain tumor cell lines, itappears that at least in our study of neuroblastoma, monocytesplay a very minor role as effector cells in ADCC. This contention is supported in part by findings of Munn and Cheung (28),who were also unable to lyse neuroblastoma tumor cells withmonocytes in the presence of murine anti-Go: mAb 3F8, evenin an 18-h lytic assay. Only after exposing monocytes to M-CSF for at least 4 days did these investigators observe monocytes to significantly lyse neuroblastoma cell lines in the presence of antibody. Thus, it is possible that cytotoxicity mediatedby monocytes that is sufficient to effect lysis of neuroblastomacells may depend on M-CSF or other cytokines for their activation and/or maturation to macrophages.

In further delineating other potent effector cells capable oflysing neuroblastoma cells coated with chi4.18, it became evident from the results of our study that granulocytes are moreeffective than PBMC in ADCC of neuroblastoma cells. Anexplanation is still lacking for the augmented ability of granulocytes to lyse neuroblastoma cells in the presence of chi4.18.One possible explanation for this finding may be that structuralvariations in the Fc 7 receptor of granulocytes and PBMCcould potentially influence their functional activities (29), making granulocytes better effector cells in ADCC. The possibilityalso exists that Fc -yreceptor may be expressed at higher densityon granulocytes than on PBMC or that granulocytes have amore efficient mechanism than PBMC for the lysis of neuroblastoma cells coated with chi4.18. One additional factor to beconsidered in interpreting the data presented in Fig. 5 is thatthe purification protocol used in this study could activate granulocytes. In this regard, other investigators have demonstratedthat isolation of granulocytes on Ficoll gradients can enhanceactivation of these cells (30). Thus, even though the intrinsicproperties of PBMC and granulocytes could explain the difference observed in their ability to lyse antibody-coated neuroblastoma cells, it is also possible that granulocytes could be

activated during their purification.Granulocytes, unlike PBMC, were able to lyse neuroblastoma

cells equally well, whether the cells were coated with murine orchimeric anti-Go: antibody, provided that the antibody concentration was sufficiently high, i.e., >0.1 pg. One explanation forthis finding could be that granulocyte-mediated lysis in thepresence of ch 14.18 has reached its plateau at the high antibodyconcentrations so that the murine anti-Go? antibody appears tobe equally capable of lysing neuroblastoma cells as its human/mouse chimeric counterpart. Thus, only at lower antibodyconcentrations is the human/mouse chimeric anti-Go: antibodysuperior to the murine antibody.

The magnitude of chl4.18-mediated lysis of neuroblastomawas augmented further in the presence of GM-CSF. Thesefindings are in agreement with the findings of other investigators who have shown that GM-CSF can enhance granulocyte-mediated lysis of neuroblastoma, which were coated with otheranti-Go: mAbs (31, 32). One mechanism by which GM-CSFcan enhance ADCC of tumor cells is by increasing the production of toxic oxygen metabolites by granulocytes; however,granulocyte-mediated lysis of neuroblastoma in the presence ofantibody ¡snot dependent on these oxygen products for tumorcell lysis (32). Another possible explanation for the enhancingeffect of GM-CSF on granulocyte-mediated cytotoxicity ofmAb-coated neuroblastoma may be an increase in the numberof Fc y receptor on granulocytes, thus resulting in greaterantibody binding.

Although most of our experiments involved the use of bloodfrom normal adult individuals as a source of effector cells, weconsidered it important for two reasons to also evaluate effectorcells obtained from neuroblastoma patients. First, it has beenreported that effector cell functions are usually suppressed inneuroblastoma patients (33, 34). Second, many of the chemo-therapeutic drugs used in the treatment of neuroblastoma induce immunosuppression, as well as neutropenia. In this study,we could demonstrate that PBMC from neuroblastoma patientsare capable of lysing neuroblastoma cells with a greater efficiency when the tumor cells are coated with chi4.18 ratherthan with murine mAb 14.G2a. However, granulocytes fromneuroblastoma patients can lyse neuroblastoma cells coatedwith chi4.18 far more effectively than PBMC obtained fromthe blood of these same patients. These findings indicate thateffector cells of neuroblastoma patients can carry out similartasks as those obtained from normal adults. Therefore, whateverfactors may down-regulate other components of the immunesystem, they apparently do not influence the ability of neuroblastoma patients' effector cells to mediate lysis of tumor cells

that are coated with chi4.18.Since this study demonstrated that GM-CSF enhances

chi4.18-mediated lysis of neuroblastoma cells by granulocytesfrom neuroblastoma patients, this augmentation of lysis couldbe important in increasing the ability of chi4.18 to ablate thegrowth of neuroblastoma. Although Massuci et al. (35) reportedthat GM-CSF can enhance mAb 17-lA-mediated lysis of human colorectal carcinoma cells by both lymphocytes and monocytes, GM-CSF was ineffective in our hands in enhancing theability of PBMC to lyse chl4.18-coated neuroblastoma cells.The reason for the discrepancy between these two studies is notknown.

Aside from enhancing antibody-mediated lysis of tumor cellsby granulocytes, GM-CSF can also affect the migration ofgranulocytes (36). Therefore, a combined therapy of neuroblastoma patients with the human/mouse chimeric antibody

148

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

CHIMERIC ANTI-Gâ„¢ ANTIBODY-MEDIATED LYSIS OF NEl'ROBLASTOMA

chi4.18 and GM-CSF could be beneficial, since GM-CSF may

focus activated granulocytes to the tumor and enhance theirability to suppress the growth of neuroblastoma cells, thushelping to optimize the antitumor effect of chl4.18. In thisregard, preliminary clinical studies demonstrated that granulocytes isolated from the blood of neuroblastoma patients thathad been treated with murine anti-GD2 mAb 14.G2a did preferentially migrate to the tumor site following a brief ex vivoactivation with GM-CSF. Considering these effects of GM-CSF, together with its known ability to increase the productionof neutrophils in patients suffering from neutropenia as aconsequence of chemotherapy (37), it may be of considerablefuture interest to test the effect of a combined therapy withGM-CSF and anti-GD2 mAb in neuroblastoma patients.

ACKNOWLEDGMENTS

The authors wish to thank Dr. Stephen D. Gillies, Abbott Biotech,Needham Heights, MA, for making mAb chi4.18 available to us. Wealso wish to acknowledge the help of Lynne Kottel in the preparationof this manuscript.

REFERENCES

1. Breslow, N., and McCann, B. Statistical estimation of prognosis for childrenwith neuroblastoma. Cancer Res.. 31: 2098-2103. 1971.

2. Evans. A. E., D'Angio, G. J., and Randolph. J. A proposed staging for

children with neuroblastoma: children cancer study group A. Cancer (Phila.),27:374-378, 1970.

3. Seeger, R. C. Neuroblastoma: clinical perspective, monoclonal antibodies,retinoic acid. Ann. Intern. Med.. 97: 873-884. 1982.

4. August. C. S., Serota, F. T.. Koch. P. A.. Burkey, E.. Schlesinger, H.. Elikins.W. L., Evans, A. E., and D'Angio. G. J. Treatment of advanced neuroblastoma with supralateral chemotherapy, radiation, and allogenic or autol-ogous bone marrow reconstitution. J. Clin. Oncol.. 2: 609-615. 1984.

5. Evans, A. E. Staging and treatment of neuroblastoma. Cancer (Phila.), ^5:1799-1802, 1980.

6. Philips, T., Bernard, J. L.. Zucker. J. M.. Pinkerton, R.. Lutz, P., Boridigoni.P., Plouvier, E., Robert, A., Carton, R.. Philip. !.. Chauvin. F., and Favarot.M. High dose chemoradiotherapy with bone marrow transplantation asconsolidation treatment in neuroblastoma: an unselected group of stage IVpatients over 1 year of age. J. Clin. Oncol., 5: 266-271, 1987.

7. Seeger, R. C., Moss, T. J., Feig, S. A., Lenarsky. C., Selch, M., Ramsay. N.,Harris, R., Wells, J.. Harland. S.. and Reynolds. P. C. Bone marrow transplantation for poor prognosis of neuroblastoma. In: A. E. Evans, G. J.D'Angio, A. G. Knudsen, and R. C. Seeger (eds.). Advances in NeuroblastomaResearch, Ed. 2, pp. 203-213. New York: Alan R. Liss, Inc., 1988.

8. Cheung, N. K. V., Munn. D.. Kushner, B. H., Usmani, N., and Yeh, S. D. J.Target radiotherapy and immunotherapy of human neuroblastoma with <•,,•specific monoclonal antibodies. Nucí.Med. Biol., 16: 111-120. 1989.

9. Kemshead, J. T., Fritschy, J., Goldman, A., Malpas, J. S., and Pritchard, J.Use of panels of monoclonal antibodies in the differential diagnosis ofneuroblastoma and lymphoblastic disorders. Lancet. /: 12-14. 1983.

10. Schultz. G.. Cheresh, D. A., Varki, N. M.. Yu, A.. Staffileno. L. K., andReisfeld, R. A. Detection of ganglioside GDJ in tumor tissue and sera ofneuroblastoma patients. Cancer Res.. 44: 5914-5920. 1984.

11. Treleaven, J. G., Ugelstad. J.. Phillip. T.. Gibson, F. M., Rembaum. A.,Caine, G. D., and Kemshead, J. T. Removal of neuroblastoma cells frombone marrow with monoclonal antibodies conjugated to magnetic micro-spheres. Lancet, /: 70-73, 1984.

12. Gomiuchi, M., Katano, M., Saxton, R. E., Lake, R., Irie, K., Morton, D. L.,and Irie. R. F. Human monoclonal antibody to tumor-associated gangliosideCos- ''«v'vo reactivity. Proc. Am. Assoc. Cancer Res.. 25: 248. 1984.

13. Reisfeld, R. A., and Cheresh. D. A. Human tumor antigens. Adv. Immunol..Â¥0:323-377, 1987.

14. Mujoo, K., Cheresh, D. A.. Yang, H. M.. and Reisfeld. R. A. Disialoganglio-side (¡i,..on human neuroblastoma cells: target antigen for monoclonalantibody-mediated cytolysis and suppression of tumor growth. Cancer Res.,47: 1098-1104, 1987.

15. Mujoo. K.. Kipps. T. J.. Yang. H. M.. Cheresh, D. A.. Wargalla, U., Sander,D. J., and Reisfeld. R. A. Functional properties and effect on growth of

human neuroblastoma tumors by isotype switch variants of monoclonal anti-ganglioside Gn: antibody 14.18. Cancer Res.. 49: 2857-2861. 1989.

16. Munn. D. H.. Cordon-Cardo. C'.. Yeh, S. D. J., Kushner. B. H., and Cheung.

N. K. V. Therapy of neuroblastoma with G,>; specific monoclonal antibody3F8. In: H. F. Oettgen (ed.). Gangliosidcs and Cancer, pp. 261-271. Heidelberg: Springer-Verlag. 1989.

17. Courtney-Luck, N. S.. Epcnetos. A. A.. Moore. R., Lärche,M.. Pecasides,D.. Dhokia, B.. and Ritter. M. A. Development of primary and secondaryimmune response to mouse monoclonal antibodies used in the diagnosis andtherapy of malignant neoplasms. Cancer Res.. 46: 6489-6495. 1986.

18. Shawler. P. L.. Bartholomew. R. M.. Smith. L. M.. and Dillman. R. O.Human immune response to multiple injections of murine monoclonal im-munoglobulin. J. Immunol., 135: 1530-1537. 1985.

19. LoBuglio. A. F.. Wheller. R. H.. Trang. J.. llaynes. A.. Rogers, K., Harvey,E. B., Sun, L. K.. Ghrayeb. J. K.. and Kha/aaeli. M. B. Mouse/humanchimeric monoclonal antibody in man: kinetics and immune response. Proc.Nati. Acad. Sci. USA. 86: 4220-4224. 1989.

20. Liu, A. Y.. Robinson. R. R., Murray. E. D., Jr., Ledbetter, J. A., Hellström.I., and Hellström,K. E. Production of a mouse-human chimeric monoclonalantibody to CD20 with potent Fc-dependcnt biologic activity. J. Immunol..139: 3521-3526. 1987.

21. Morrison, S. L., and Oi, V. T. Genetically engineered antibody molecules.Adv. Immunol., 44: 65-92. 1989.

22. Gillies. S. D., Lo, K. M.. and Wesolowski, J. High-level expression ofchimeric antibodies using adapted cDNA variable region cassettes. J. Immunol. Methods. 125: 191-202. 1989.

23. Mueller. B. M.. Romerdahl, C. A.. Gillies. S. D.. and Reisfeld. R. A.Enhancement of antibody-dependent cytotoxicity with chimeric anti-GD2antibody. J. Immunol.. 144: 1282-1286. 1990.

24. Freundlich. B., and Avdalovic, N. Use of gelatin/plasma coated flasks forisolating human peripheral blood monocytes. J. Immunol. Methods, 62: 31-37, 1983.

25. Masucci. G.. Lindelmalm. C.. Frodin. J. E.. Hagström.B., and Mcllstedl. H.Effect of human blood mononuclear cell populations in antibody dependentcellular cytotoxicity (ADCC) using two murine (C0171-A. Br55-2) and onechimeric (17-1 A) monoclonal antibodies against a human colorectal carcinoma cell line (SW948). Hybridoma. 7: 429-440. 1988.

26. Melnick, D. A.. Nauseef, W. M.. Markowitz. S. D., Gardner, J. P., andMalcch, H. L. Biochemical analysis and subcellular localization of a neutro-phil-specific antigen PMN-7 involved in the respiratory burst. J. Immunol..134: 3346-3355. 1985.

27. Melnick. D. A.. Meshulam, T.. Manto. A., and Malech. H. L. Activation ofhuman neutrophils by monoclonal antibody PMN7C3: cell movement andadhesion can be triggered independently from the respiratory burst. Blood.67: 1388-1394. 1986.

28. Munn. D. H.. and Cheung. N. K. V. Antibody-dependent anlitumor immunitycytotoxicity by human monocytes cultured with recombinant macrophagecolony stimulating factor. Induction of efficient antibody-mediated antitumorcytotoxicity not detected by isotope release assays. J. Exp. Med.. 170: 511-526. 1989."

29. Lanier, L. L.. Ruitenburg. J. J.. and Phillips. J. H. Functional and biochemical analysis of CD16 antigen on natural killer cells and granulocytes. J.Immunol".. 141: 3478-3482. 1988.

30. Haslett. C.. Guthrie. L. H.. Kopaniak, M. M., Johnson, R. B.. and Hcnson.P. M. Modulation of multiple neutrophil functions by preparative methodsor trace concentrations of bacterial lipopolysaccharide. Am. J. Pathol.. 119:101-110,1985.

31. Bruchelt. G., Handgretinger. R.. Kimmig. A., Goeke. B.. Siedncr. R.. Reisfeld. R. A., Niethammer, D., and Treuner. J. Effect of granulocytes on humanneuroblastoma cells measured by chemiluminescence and chromium-51 release assay. J. Biolumin. Chemilumin...?: 93-96. 1989.

32. Kushner. B. H.. and Cheung. N. K. V. GM-CSF enhances 3F8 monoclonalantibody-dependent cellular cytotoxicity against human melanoma andneuroblastoma. Blood. 73: 1936-1941. 1989.

33. Bill. A. H. Immune aspects of neuroblastoma. Current information. Am. J.Surg., 122: 142-147, 1971.

34. Floutsis. G.. Ulsh. L.. and Ladisch. S. Immunosupprcssive activity of humanneuroblastoma tumor gangliosides. Int. J. Cancer. 43: 6-9. 1989.

35. Masucci. G., Wersall, P., Ragnhammar, P.. and Mellstedt. H. Granulocyte-manocyte-colony-stimulating factor augments the cytotoxic capacity of lymphocytes and monocytes in antibody-dependent cellular cytotoxicity. CancerImmunol. Immunother.. 29:288-292. 1989.

36. Gasson, J. C. Weisbart. R. H.. Kaufman. S. E., Clark, S. C., Hewick. R. M..Wong, G. G., and Golde. D. W. Purified human granulocyte-macrophagecolony stimulating factor: direct action on neutrophils. Science (WashingtonDC). 226: 1339-1342, 1984.

37. Lindemann. A.. Herrmann. F.. Gamm. H.. Oster. W., and Mertelsmann. R.Cancer treatment with cytokines: concepts and first clinical experience: In:H. G. Beger. M. Büchler,and R. A. Reisfeld (eds.). Cancer Therapy, pp. 73-84. Berlin: Springer-Verlag, I989.

149

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

1991;51:144-149. Cancer Res   Edward Barker, Barbara M. Mueller, Rupert Handgretinger, et al.   Cell-mediated Lysis of Human Neuroblastoma Cells

Antibody onD2Effect of a Chimeric Anti-Ganglioside G

  Updated version

  http://cancerres.aacrjournals.org/content/51/1/144

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

  .pubs@aacr.orgDepartment 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/51/1/144To request permission to re-use all or part of this article, use this link

on March 27, 2020. © 1991 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from