J Neurosurg 73:901-908, 1990
Antigen related to cell proliferation in malignant gliomas recognized by a human monoclonal antibody TAKASHI KOKUNAI, M.D., NORIHIKO TAMAK|, M.D., AND SATOSHI MATSUMOTO, M.D. Department of Neurosurgery, Kobe University School of Medicine, Kobe, Japan u- A human monoclonal antibody (CLN-IgG) was produced from a human-human hybridoma derived from lymphocytes of a patient with cervical carcinoma. The reactivities of this antibody with various human glioma tissues and cultured glioma cells and the characterization of the antigen recognized by CLN-IgG on malignant glioma cells were analyzed and reported. CLN-IgG reacted with various human glioma cells and glioma tissues, especially glioblastoma, but did not react with normal brain tissues or fetal brain tissues. A large amount of antigen recognized by CLN-IgG was expressed on cell membranes of undifferentiated glioma cells and of glioma cells at the G2/M tumor growth phase in cycling cells. Antigen recognized by CLN-IgG was detected in only one of seven samples of cyst fluid, and was not detected in 27 serum samples or 18 samples of cerebrospinal fluid from glioma patients. CLN-IgG exhibited antibody-dependent cell cytotoxicity against U-251MG glioma cells and primary cultured cells of glioblastomas and anaplastic astrocytomas. These data suggest that the antigen recognized by CLN-lgG might be related to cell proliferation in malignant gliomas. Thus, CLN-IgG might be useful for immunotherapy or immunoimaging of malignant gliomas.
KEY WORDS
brain tumor
S
9 immu~otherapy
INCE the development of hybridoma technology 15 years ago, J2 a number of murine monoclonal antibodies against human cancer cells have been produced and many antigens have been identified. 3'~6'21'25 The clinical use of murine monoclonal antibodies produced by mouse-mouse or mouse-human hybridomas in the treatment of human cancers has inspired new concepts for cancer diagnosis and treatmerit.6,15,19 However, there are some practical problems associated with the clinical use of foreign antibodies in humans, such as the induction of the human antimouse response, possibly reducing the efficacy of the treatment. 14 Several human monoclonal antibodies against human cancer cells have been created using lymphocytes from cancer patients. 1"5"7"8"1~1,~7,2224,27,28 Hagiwara and Sato ~~ reported the production of a human monoclonal antibody, CLN-IgG, made by fusing UC 729-6, 8 a 6-thioguanine-resistant human lymphoblastoid B-cell line, with lymphocytes obtained from a patient with squamous-cell carcinoma of the cervix. It has previously been reported that CLN-IgG recognized the antigen expressed in various histological types of human cancers, including malignant gliomas. 2 In this report, we describe the precise reactivities of CLN-IgG against various human glioma tissues, normal J. Neurosurg. / Volume 73/December, 1990
glioma
9 monoclonal antibody
brain tissues, and cultured glioma cells. We define the antigen recognized by CLN-IgG on malignant glioma cells for the purpose of identifying the value of the antibody in the diagnosis and treatment of malignant gliomas.
Materials and Methods Monoclonal Antibody The derivation, properties, and purification of the human monoclonal antibody, CLN-immunoglobulin (Ig)G,* have been described in previous reportsfl ~~ Cells" and Tissues Cells from the human glioma cell lines U-251MG, U-373MG, and KNS-42, t were maintained in RPMI1640 medium supplemented with 10% fetal bovine serum and kanamycin (50 ~g/ml), and grown in monolayer culture at 37~ in a humidified chamber with 5% CO2 and 95% air atmosphere. T u m o r tissues were obtained from glioma patients undergoing surgery at Kobe University Hospital. To obtain primary cultured * CLN-IgG was kindly supplied by H. Hagiwara, Ph.D. ? Glioma cells obtained from American Type Culture Collection, Rockville, Maryland. 901
T. Kokunai, N. Tamaki, and S. Matsumoto cells, tumor tissues were mechanically cut into fragments of about 0.5 cu ram, incubated for 30 minutes at 37~ in RPMI-1640 medium supplemented with the enzyme Disperse (1000 IU/ml),~ and passed through a 100-urn stainless steel mesh. The single cells were washed three times with RPMI-1640 medium and suspended in complete medium.
Cell Treatment The U-251MG cells and primary cultured glioblastoma cells were treated with 1 mM Nr,O2-dibutyryl cyclic adenosine monophosphate (AMP) for 48 hours to induce cell differentiation,t8 Cells at the G~ tumor growth phase were obtained by treatment with serumfree medium for 5 days against confluent cultured cells. Cells at the G2/M phase were prepared by mitotic shaking. 2~ Irnmunoperoxidase Staining Frozen tissue sections (6 urn) were fixed with cold acetone for 10 minutes, then washed with phosphatebuffered saline (PBS), pH 7.4, for 20 minutes. Sections were preincubated with 0.03% H202 in methanol for 20 minutes and washed with PBS for 20 minutes. Following incubation with 5% normal goat serum in PBS for 30 minutes, sections were incubated with 10 ug/ml of CLN-IgG for 60 minutes at 30~ After another wash with PBS for 30 minutes, sections were incubated with biotinylated goat anti-human IgG for 30 minutes at 30~ After a further wash with PBS for 30 minutes, sections were incubated with avidin-biotinperoxidase complex for 30 minutes at 30~ After yet another wash with PBS for 30 minutes, peroxidase reaction was initiated using 0.06% diaminobenzidine with 0.01% H202 in 50 mM Tris-HC1 buffer, pH 7.0, for 5 minutes. The sections were briefly counterstained with hematoxylin. Negative control sections were incubated with nonimmunized human IgG in place of CLN-IgG. lmmunofluorescence Staining Semiconfluent U-251MG cells were grown on coverslips in complete medium, washed with PBS, and fixed with cold 80% ethanol for 20 minutes at 4~ After being washed with PBS, cells were incubated with I0 ug/ml of CLN-IgG for 30 minutes at 4~ After another wash with PBS, cells were stained with fluorescein isothiocyanate (FITC)-labeled goat anti-human IgG for 30 minutes at 4~ then incubated with 4 tzg/ ml of propidium iodine for deoxyribonucleic acid (DNA) staining. The coverslips were viewed under a fluorescence microscope. Flow Cytometry Analysis Cells were washed three times with PBS, and fixed with 80% ethanol for 30 minutes at 4~ Staining for CLN-IgG was performed by incubating 1 • 106 cells :~Disperse supplied by Godo Susei Co., Tokyo, Japan. 902
for 60 minutes at room temperature in CLN-IgG, 10ug/ ml. After incubation, the cells were washed three times with 0.05% Tween 20/PBS and then incubated for 30 minutes at room temperature in 20 ul of PBS containing a 1:20 dilution of FITC-labeled goat anti-human IgG. Cells incubated with nonimmune human IgG served as a control for background fluorescence. After being washed three times with 0.05% Tween 20/PBS, the cells were preincubated with 1 mg/ml of ribonuclease for 30 minutes at 30~ then stained with propidium iodine (final concentration of 4 ug/ml). Immunofluorescence and correlated dual-parameter measurements of CLN-IgG versus DNA content were performed with a fluorescence-activated cell sorter.w The green FITC fluorescence signal was proportional to CLN-IgG levels and was measured with a 514-nm bandpass filter. The red propidium iodine fluorescence signal, corresponding to DNA content, was measured using a 560-nm filter in conjunction with a 610-nm longpass filter. For analysis of the antigen on the cell membrane, semiconfluent cells were treated with 0.05% ethylenediaminetetra-acetic acid for 10 minutes at 37~ in order to obtain single cells, then washed with PBS containing 5% fetal bovine serum and 0.05% sodium azide. The cells were incubated for 60 minutes with 5 ug/ml of CLN-IgG at 4~ After being washed with PBS containing 5% fetal bovine serum and 0.05% sodium azide, the cells were stained with a 1:20 dilution of FITClabeled goat anti-human IgG for 30 minutes at 4~ After washing, the signal was measured with the fluorescence-activated cell sorter and stained cells were viewed under the fluorescence microscope. For analysis of the antigen in cytoplasm, single cells were treated with 0.25% trypsin in PBS for 60 minutes at 37~ following fixation with cold 80% ethanol. Flow cytometry analysis was performed as for the analysis of the antigen on cell membrane. Treatment with 0.25% trypsin for 60 minutes at 37~ as described inhibited the expression of antigen on cell membrane (data not shown).
Imrnunoblotting The U-251MG cells (107 cells) were solubilized with 500 ~1 of cell-lysis buffer containing 1% Triton X-100, 1% sodium deoxycholate, 0.01% sodium dodecyl sulfate (SDS), 0.15 M NaC1, 50 mM Tris-HC1 (pH 7.4), and 2 mM phenylmethylsulfonylfluoride. Cyst fluids, sera, and cerebrospinal fluid (CSF) obtained from glioma patients were lyophilized. Samples (1 mg/ml of protein concentration) were dissolved in buffer containing 2.5% SDS, 5% 2-mercaptoethanol and 0.01 M TrisHC1 (pH 8.0), and boiled for 5 minutes at 100~ The procedure for SDS-polyacrylamide gel electrophoresis used in this study essentially followed the method of Laemmli. 13 The stacking gel contained 5% acrylamide wFluorescence-activated cell sorter was manufactured by Becton-Dickinson, Inc., Mountain View, California.
J. Neurosurg. / Volume 73 / December, 1990
Antigen related to cell proliferation
FIG. 1. Immunoperoxidase staining of glioblastoma (A), anaplastic astrocytoma (B), adult normal brain (C), and fetal brain (D) tissues with CLN-IgG. Nuclei were counterstained with hematoxylin. In A, a blood vessel is identified with an arrow. • 150.
and 0.13% N,N'-methylene-bisacrylamide (Bis), and the running gel contained 10% acrylamide and 0.1% Bis. Electrophoresis was performed at room temperature for 4 hours at 20 mA to a 0.4-mm width of gel. The migration front was followed by adding 0.002% bromophenol blue to the electrode buffer. Proteins from the gels were electrophoretically transferred onto nitrocellulose filters at a constant voltage of 50 V for 3 hours as reported by Towbin, eta/. 26 The nitrocellulose filters were immersed in 3% bovine serum albumin then reacted according to the method described for immunoperoxidase staining. Antibody-Dependent Cell Cytotoxicity The U-251MG cells and primary cultured cells from various glioma tissues for use as target cells were mixed with sodium chromate (5~Cr) at 100 uCi/106 cells/ml and incubated for 1 hour at 37~ in a CO2 incubator. The 5~Cr-labeled target cells (1 x 104/well) and effector lymphocytes were incubated with or without 50 ug of CLN-IgG for 6 hours at 37 ~ in a 96-well microplate. Human peripheral lymphocytes collected from healthy donors were used as effector cells. The radioactivity of 0.1 ml of supernatant in each well was collected, and the percentage of lysis was assessed by means of the J. Neurosurg. / Volume 73/December, 1990
formula: cytolysis (%) = ((a - c) + (b - c)) x 100, in which a is the percentage of release (expressed as counts per minute (cpm)) in the presence of effector cells, b is the percentage of maximum release (in cpm) by 1 N HC1, and c is the percentage of spontaneous cpm with medium alone. Results
Immunohistochemical Characterization Table 1 shows the immunohistochemical reactivity of CLN-IgG on various histological types of brain tumor and normal brain tissue. There were 51 brain tumors; seven of the eight glioblastomas (Fig. 1A) and all four anaplastic astrocytomas (Fig. 1B) stained positively. Benign gliomas (oligodendrogliomas, ependymomas, and astrocytomas) showed a low occurrence of CLN-IgG reactivity. The heterogeneous staining by CLN-IgG was typical in tumor cells from glioblastoma specimens (Fig. 1A). Blood vessels in tumor tissue were not stained by CLN-IgG (Fig. 1A). Benign brain tumors (meningiomas, neurinomas, pituitary adenomas, and choroid plexus papillomas) did not stain, but tissue from two of five craniopharyngiomas did. Adult normal brain tissue (Fig. IC) and fetal brain tissue (Fig. 1D) 903
T. Kokunai, N. Tamaki, and S. Matsumoto
FIG. 2. Immunofluorescence staining of fixed U-251MG cells with CLN-IgG. The reactive fluorescein isothiocyanate products with CLN-IgG, especially ruffles, were found on the cell membrane. Nuclei stained with propidium iodine, x 280. did not show any reactivity with CLN-IgG. The antigen recognized by CLN-IgG was localized in the cell membrane of U-25 IMG cells (Fig. 2), especially ruffles. The reactivity of CLN-IgG toward U-251MG cells was resistant to treatment with sodium periodate and other modes of deglycosylation (data not shown). Flow Cytometry The relationship between the expression of antigen and the cell cycle was examined by flow cytometry. Flow cytometric analysis with the use of CLN-IgG showed strong positive binding to U-251 MG and KNS-
42 cells, especially at the GriM phase compared to that at the G~ phase (Fig. 3). When U-251MG cells were treated with cyclic AMP, they showed reduced expression of the antigen recognized by CLN-IgG, especially the expression of the antigen on cell membrane (Fig. 4). The maximum expression of the antigen on the cell surface was detected in the cells at G2/M phase and synchronously cultured U-251 MG cells at the Go/G1 phase showed a markedly decreased expression (Fig. 4). In primary cultured cells from glioblastomas, the expression pattern of the antigen recognized by CLN-IgG was the same as the pattern of U-251MG cells, but more marked expression during the G2/M phase was observed (Fig. 5) than at the Go/G~ phase. I m m unoblotting
Immunoblotting analysis was used to examine the epitope of the antigen in U-251MG cells and the amount of antigen shed into cyst fluids, sera, and CSF of glioma patients (Fig. 6). Cell lysates of U-251MG cells under reduced conditions showed two bands with a molecular weight (MW) of 60 kD and 53 kd. These data suggest that the epitope of the antigen in U-251MG cells recognized by CLN-IgG is the same as that of a previously reported antigen. '~ The antigen recognized by CLN-IgG was detected in the cyst fluid of one glioblastoma, but was a single band with an MW of 60
TABLE 1 Immunohistochemical reactivities of CLN-IgG against various human brain tumors, normal brain, and cultured cells Histology glioblastoma anaplastic astrocytoma astrocytoma medulloblastoma ependymoma oligodendroglioma metastatic tumor adenocarcinoma squamous cell carcinoma craniopharyngioma pituitary adenoma meningioma neurinoma choroid plexuspapilloma normal brain (adult) normal brain (fetus) totals U-373MG U-251MG KNS-42 primary cultured cell glioblastoma anaplastic astrocytoma astrocytoma 904
No. Positive/ Total Cases 7/8 4/4 1/5 1/5 1/2 1/3 3/3 1/2 2/5 0/3 0/5 0/4 0/2 0/5 0/3 21/59 positive positive positive 5/5 3/3 2/3
FIG. 3. Two-color analysis of the vertical fluorescein isothiocyanate (FITC)-fluorescence (CLN-IgG) and the perpendicular propidium iodine-fluorescence (DNA) on KNS-42 (upper) and U-251MG (lower) cells by flow cytometry. J. Neurosurg. / Volume 73/December, 1990
Antigen related to cell proliferation
FIG. 4. Binding pattern of CLN-IgG on the antigen of U-251MG cells obtained with flow cytometric analysis. Upper Left: Exponentially growing cells. Lower Left: Cells treated with cyclic adenosine monophosphate (AMP). UpperRight: Cells synchronously accumulated at the Go/G, phase. Lower Right: Cells collected at the G2/M phase by mitotic shaking. FITC = fluorescein isothiocyanate. kD. Other cyst fluids (six cases), sera (27 cases), and CSF (18 cases) of glioma patients and the supernatant of cultured U-251MG cells showed no reactivity against CLN-IgG.
Antibody-Dependent Cell Cytotoxicity For analysis of the cytotoxic activity of CLN-IgG, we examined the antibody-dependent cell cytotoxicity (ADCC) activity of CLN-IgG against U-251MG and
FIG. 5. Two-color analysis (left) and binding patterns (righO by flow eytometry of CLN-IgG on the antigen of primary cultured glioblastoma cells. FITC = fluorescein isothiocyanate; AMP = adenosine monophosphate.
J. Neurosurg. / Volume 73/December, 1990
905
T. Kokunai, N. Tamaki, and S. Matsumoto
FIG. 6. Immunoblotting analysis of the antigen in U25 IMG cell lysate (lane 1), supernatant of U-251MG cells (lane 2), cyst fluid from glioblastomas (lanes 3 and 4), sera from glioma patients (lanes 5 and 6), and cerebrospinal fluid from glioma patients (lanes 7 and 8).
primary cultured glioma cells. The ADCC activity of CLN-IgG against U-251 MG cells was l 0.81% + 1.24 % at an effector:target ratio of 40:1, and was reduced when the cells were treated with cyclic AMP (Fig. 7). Table 2 shows the ADCC activities of CLN-IgG against primary cultured cells from various glioma tissues. The ADCC activity of CLN-IgG against the primary cultured cells of glioblastoma and anaplastic astrocytoma was 36.2% _+ 12.85% and 29.4% _+ 7.65%, respectively, but against the primary cultured cells of astrocytoma and normal brain CLN-IgG ADCC activity was very low (8.9% -+ 3.65% and 4.3% _+ 2.81%, respectively). Discussion
Reactivity of CLN-IgG Antibody Human monoclonal antibody CLN-IgG reacted with malignant gliomas, especially with glioblastoma and anaplastic astrocytoma. These tumors are the most malignant and invasive among all tumors of neuroectodermal origin. By contrast, CLN-IgG did not react with normal brain tissues. There have been few reports of tumor antigens being detected by human monoclonal antibody. Irie, et al., ~ reported that the L72 antibody (designated anti-OFA-1-2) reacted only with tumor cells of neuroectodermal origin (melanomas, neuroblastomas, and gliomas). In previous studies, j'~~ CLN-IgG has been reported to react with various human cancers such as cervical carcinoma, gall bladder carcinoma, stomach carcinoma, melanoma, lung carcinoma, and prostate carcinoma. The antigen recognized by CLN-IgG had an MW of 226 kD consisting of a-subunit (MW 60 kD) and ~3-subunit (MW 53 kD) which were linked by intermolecular disulfide bond. Our immunoblotting results demonstrated that the epitope of the antigen recognized by CLN-IgG in human 906
FIG. 7. Antibody-dependent cell cytotoxicity of CLN-IgG against U-251MG cells treated without (left) or with (right) cyclic adenosine monophosphate (AMP).
glioma cells had two molecules of the same sizes as in the previously reported cancer cell antigen. L~~ It is suggested that CLN-IgG recognized the epitope of an antigen that commonly exists in a broad range of malignancies beyond the tumors of neuroectodermal origin.
Therapeutic Application There is a potential therapeutic application for monoclonal antibodies against cancer cells because cell surface expression of the antigen is thought to be a major factor in the initial reaction of antibody-mediated immunity against cancer cells. Flow cytometric analysis and immunofluorescence staining showed that the antigen was expressed on the cell membrane of U-251MG cells and primary cultured glioblastoma cells. Moreover, flow cytometric analysis revealed that the expression of the cell-surface antigen was higher in the undifferentiated cells than the differentiated cells treated with cyclic AMP. Maximum expression of this antigen on the cell membrane was observed during the G J M phase of the cell cycle. These data suggest that the function of the antigen recognized by CLN-IgG may be related to cell proliferation. The shedding of the antigen into body fluid is important for cancer immunotherapy. The antigen recognized by CLN-IgG was shed into the cyst fluid in only one case of glioblastoma and did not shed into the cyst fluid, serum, or CSF of any other case or into the supernatant of U-251MG cells. This shed antigen had a single molecule (MW 60 kD) in contrast to the two molecules detected in the cell lysate of U-251MG cells. This finding suggests the possibility that the shed antigen might be a degraded form of the antigen recognized by CLN-IgG. J. Neurosurg. / Volume 73 / December, 1990
Antigen related to cell proliferation TABLE 2 ADCC activity of CLN-IgG against primary cultured cellsfrom various tissues* No. of ADCC Cases Activity (%) glioblastoma 5 36.2 +- 12.85 anaplastic astrocytoma 3 29.4 + 7.65 astrocytoma 3 8.9 - 3.65 normal brain 3 4.3 _ 2.81 * ADCC = antibody-dependent cell cytotoxicity. The effector/ target ratio was 40:1. Means are expressed _+standard deviation of the means.
4.
Histology of Tissue
It is important for cancer i m m u n o t h e r a p y that CLNIgG had antibody-dependent cell cytotoxicity against cells and primary cultured cells o f glioblastomas and anaplastic astrocytomas. Furthermore, synergistic augmentation was reported in complement-dependent cytotoxicity against cancer cells using C L N - I g G and another h u m a n monoclonal antibody. 9 In addition, during the several years since its establishment, the clone has been stable both in C L N - I g G secretion and in specificity against cancer cells? ~ Radioimaging Mouse monoclonal antibodies to h u m a n cancer cells have been used for i m m u n o t h e r a p y 6'25'~9 and radioimaging? There are some practical problems ~ to using mouse monoclonal antibodies in clinical practice, including the induction of anti-mouse IgG after multiple administrations and the potential adverse reactions against certain normal tissues. The accumulation of anti-mouse IgG in the circulation could neutralize subsequently administered monoclonal antibodies and diminish their effectiveness. Circulating mouse m o n o clonal antibody-anti-idiotype complexes might lead to immunological suppression, enhancement o f t u m o r growth, and renal dysfunction. For the above reasons, the h u m a n - h u m a n monoclonal antibody (CLN-IgG) is thought to be of possible use for i m m u n o t h e r a p y or radioimaging of malignant gliomas, and for analysis o f the humoral i m m u n e response against h u m a n malignant gliomas such as the anti-idiotype network.
5. 6. 7.
8.
9. 10. 1 I.
12. 13. 14. 15. 16.
17.
18.
Acknowledgment
We thank H. Hagiwara, Ph.D., for making it possible for us to use CLN-IgG.
19.
References
20.
1. Andreasen RB, Olsson L: Antibody-producing humanhuman hybridomas. III. Derivation and characterization of two antibodies with specificity for human myeloid cells. J Immunol 137:1083-1090, 1986 2. Aotsuka Y, Hagiwara H: Identification of a malignant cell associated antigen recognized by a human monoclonal antibody. Enr J Cancer Clin Oncol 24:829-838, 1988 3. Brown JP, Hewick RM, Hellstr6m I, et al: Human reelJ. Neurosurg. / Volume 73/December, 1990
21. 22.
23.
anoma-associated antigen p97 is structurally and functionally related to transferrin. Nature 296:171-173, 1982 Carrasquillo JA, Krohn KA, Beaumier P, et al: Prognosis of and therapy for solid tumors with radiolabeled antibodies and immune fragments. Cancer Treat Rep 68: 317-328, 1984 Cote RJ, Morrissey DM, Houghton AN, et al: Generation of human monoclonal antibodies reactive with cellular antigens. Proe Natl Aead Sei USA 80:2026-2030, 1983 Dillman RO, Shawler DL, Sobal RE, et al: Murine monoclonal antibody therapy in two patients with chronic lymphocytic leukemia. Blood 59:1036-1045, 1982 Edwards PAW, Smith CM, Neville AM, et al: A humanhuman hybridoma system based on a fast-growing mutant of the ARH-77 plasma cell leukemia-derived line. Eur J Immunol 12:641-648, 1982 Glassy MC, Handley HH, Hagiwara H, et al: UC 729-6, a human lymphoblastoid B-cell line useful for generating antibody-secreting human-human hybridomas. Proc Natl Acad Sci USA 80:6327-6331, 1983 Hagiwara H, Aotsuka Y: Cytotoxicity of mixed human monoclonal antibodies reacting to tumor antigens. Acta Paediatr Jpn 29:552-556, 1987 Hagiwara H, Sato GH: Human x human hybridoma producing monoclonal antibody against autologous cervical carcinoma. Mol Biol Med 1:245-252, 1983 Irie RF, Sze LL, Saxton RE: Human antibody to OFA-I, a tumor antigen, produced in vitro by Epstein-Barr virustransformed human B-lymphoid cell lines, Proc Natl Acad Sci USA 79:5666-5670, 1982 K6hler G, Milstein C: Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256: 495-497, 1975 (Letter) Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685, 1970 Miller RA, Maloney DG, Mckillop J, et al: In vivo effects of murine hybridoma monoclonal antibody in a patient with T-cell leukemia. Blood 58:78-86, 1981 Miller RA, Maloney DG, Warnke R, et al: Treatment of B-cell lymphoma with monoclonal anti-idiotype antibody. N Engl J Med 306:517-522, 1982 Mitchell KF, Fuhrer JP, Steplewski Z, et al: Biochemical characterization of human melanoma cell surfaces: dissection with monoclonal antibodies. Proc Natl Aead Sci USA 77:7287-7291, 1980 Olsson L, Kaplan HS: Human-human hybridomas producing monoclonal antibodies of predefined antigenic specificity. Proc Natl Acad Sci USA 77:5429-5431, 1980 Prasad KN, Hsie AW: Morphological differentiation of mouse neuroblastoma cells induced in vitro by dibutyryl adenosine 3 ',5 '-cyclic monophosphate. Nature 233 (New Bioi):141-142, 1971 Ritz J, Pcsando JM, Sallan SE, et al: Serotherapy of acute lymphoblastic leukemia with monoclonal antibody. Blood 58:141-152, 1981 Robbins E, Marcus PI: Mitotically synchronized mammalian ceils: a simple method for obtaining large populations. Science 144:1152-1153, 1964 Saito H, Uchiyama K, Nakamura I, et al: Characterization of a human monoclonal antibody with broad reactivity to malignant tumor cells. JNCI 80:728-734, 1988 Schlom J, Wunderlich D, Teramoto YA: Generation of human monoclonal antibodies reactive with human mammary carcinoma cells. Proc Natl Acad Sci USA 77: 6841-6845, 1980 Sikora K, Alderson T, Ellis J, et al: Human hybridomas 907
T. Kokunai, N. Tamaki, and S. Matsumoto
24. 25.
26.
27.
908
from patients with malignant disease. Br J Cancer 47: 135-145, 1983 Sikora K, Wright R: Human monoclonal antibodies to lung-cancer antigens. Br J Cancer 43:696-700, 1981 Soule HR, Linder E, Edgington TS: Membrane 126kilodalton phosphoglycoprotein associated with human carcinomas identified by a hybridoma antibody to mammary carcinoma cells. Proc Natl Acad Sci USA 80: 1332-1336, 1983 Towbin H, Staehelin T, Gordon J: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350-4354, 1979 Usuba O, Fujii Y, Miyoshi I, et al: Establishment of a human monoclonal antibody to Hanganutziu-Deicher antigen as a tumor-associated carbohydrate antigen. Jpn J Cancer Res 79:1340-1348, 1988
28. Warenius HM, Taylor JW, Durack BE, et al: The production of human hybridomas from patients with malignant melanoma. The effect of pre-stimulation of lymphocytes with pokeweed mitogen. Eur ,/Cancer Clin Oncol 19:347-355, 1983
Manuscript received September 12, 1989. Accepted in final form June 18, 1990. This work was supported by a Grant-in-Aid for the Comprehensive 10-Year Strategy for Cancer Control from the Ministry of Health and Welfare of Japan, and by the Hagiwara Institute of Health, Japan. Address reprint requests to: Takashi Kokunai, M.D., Department of Neurosurgery, Kobe University School of Medicine, 7-5-1, Kusunoki-cho, Chuo-ku, Kobe 650, Japan.
J. Neurosurg. / Volume 73 / December, 1990
Antigen related to cell proliferation in malignant gliomas recognized by a human monoclonal antibody TAKASHI KOKUNAI, M.D., NORIHIKO TAMAK|, M.D., AND SATOSHI MATSUMOTO, M.D. Department of Neurosurgery, Kobe University School of Medicine, Kobe, Japan u- A human monoclonal antibody (CLN-IgG) was produced from a human-human hybridoma derived from lymphocytes of a patient with cervical carcinoma. The reactivities of this antibody with various human glioma tissues and cultured glioma cells and the characterization of the antigen recognized by CLN-IgG on malignant glioma cells were analyzed and reported. CLN-IgG reacted with various human glioma cells and glioma tissues, especially glioblastoma, but did not react with normal brain tissues or fetal brain tissues. A large amount of antigen recognized by CLN-IgG was expressed on cell membranes of undifferentiated glioma cells and of glioma cells at the G2/M tumor growth phase in cycling cells. Antigen recognized by CLN-IgG was detected in only one of seven samples of cyst fluid, and was not detected in 27 serum samples or 18 samples of cerebrospinal fluid from glioma patients. CLN-IgG exhibited antibody-dependent cell cytotoxicity against U-251MG glioma cells and primary cultured cells of glioblastomas and anaplastic astrocytomas. These data suggest that the antigen recognized by CLN-lgG might be related to cell proliferation in malignant gliomas. Thus, CLN-IgG might be useful for immunotherapy or immunoimaging of malignant gliomas.
KEY WORDS
brain tumor
S
9 immu~otherapy
INCE the development of hybridoma technology 15 years ago, J2 a number of murine monoclonal antibodies against human cancer cells have been produced and many antigens have been identified. 3'~6'21'25 The clinical use of murine monoclonal antibodies produced by mouse-mouse or mouse-human hybridomas in the treatment of human cancers has inspired new concepts for cancer diagnosis and treatmerit.6,15,19 However, there are some practical problems associated with the clinical use of foreign antibodies in humans, such as the induction of the human antimouse response, possibly reducing the efficacy of the treatment. 14 Several human monoclonal antibodies against human cancer cells have been created using lymphocytes from cancer patients. 1"5"7"8"1~1,~7,2224,27,28 Hagiwara and Sato ~~ reported the production of a human monoclonal antibody, CLN-IgG, made by fusing UC 729-6, 8 a 6-thioguanine-resistant human lymphoblastoid B-cell line, with lymphocytes obtained from a patient with squamous-cell carcinoma of the cervix. It has previously been reported that CLN-IgG recognized the antigen expressed in various histological types of human cancers, including malignant gliomas. 2 In this report, we describe the precise reactivities of CLN-IgG against various human glioma tissues, normal J. Neurosurg. / Volume 73/December, 1990
glioma
9 monoclonal antibody
brain tissues, and cultured glioma cells. We define the antigen recognized by CLN-IgG on malignant glioma cells for the purpose of identifying the value of the antibody in the diagnosis and treatment of malignant gliomas.
Materials and Methods Monoclonal Antibody The derivation, properties, and purification of the human monoclonal antibody, CLN-immunoglobulin (Ig)G,* have been described in previous reportsfl ~~ Cells" and Tissues Cells from the human glioma cell lines U-251MG, U-373MG, and KNS-42, t were maintained in RPMI1640 medium supplemented with 10% fetal bovine serum and kanamycin (50 ~g/ml), and grown in monolayer culture at 37~ in a humidified chamber with 5% CO2 and 95% air atmosphere. T u m o r tissues were obtained from glioma patients undergoing surgery at Kobe University Hospital. To obtain primary cultured * CLN-IgG was kindly supplied by H. Hagiwara, Ph.D. ? Glioma cells obtained from American Type Culture Collection, Rockville, Maryland. 901
T. Kokunai, N. Tamaki, and S. Matsumoto cells, tumor tissues were mechanically cut into fragments of about 0.5 cu ram, incubated for 30 minutes at 37~ in RPMI-1640 medium supplemented with the enzyme Disperse (1000 IU/ml),~ and passed through a 100-urn stainless steel mesh. The single cells were washed three times with RPMI-1640 medium and suspended in complete medium.
Cell Treatment The U-251MG cells and primary cultured glioblastoma cells were treated with 1 mM Nr,O2-dibutyryl cyclic adenosine monophosphate (AMP) for 48 hours to induce cell differentiation,t8 Cells at the G~ tumor growth phase were obtained by treatment with serumfree medium for 5 days against confluent cultured cells. Cells at the G2/M phase were prepared by mitotic shaking. 2~ Irnmunoperoxidase Staining Frozen tissue sections (6 urn) were fixed with cold acetone for 10 minutes, then washed with phosphatebuffered saline (PBS), pH 7.4, for 20 minutes. Sections were preincubated with 0.03% H202 in methanol for 20 minutes and washed with PBS for 20 minutes. Following incubation with 5% normal goat serum in PBS for 30 minutes, sections were incubated with 10 ug/ml of CLN-IgG for 60 minutes at 30~ After another wash with PBS for 30 minutes, sections were incubated with biotinylated goat anti-human IgG for 30 minutes at 30~ After a further wash with PBS for 30 minutes, sections were incubated with avidin-biotinperoxidase complex for 30 minutes at 30~ After yet another wash with PBS for 30 minutes, peroxidase reaction was initiated using 0.06% diaminobenzidine with 0.01% H202 in 50 mM Tris-HC1 buffer, pH 7.0, for 5 minutes. The sections were briefly counterstained with hematoxylin. Negative control sections were incubated with nonimmunized human IgG in place of CLN-IgG. lmmunofluorescence Staining Semiconfluent U-251MG cells were grown on coverslips in complete medium, washed with PBS, and fixed with cold 80% ethanol for 20 minutes at 4~ After being washed with PBS, cells were incubated with I0 ug/ml of CLN-IgG for 30 minutes at 4~ After another wash with PBS, cells were stained with fluorescein isothiocyanate (FITC)-labeled goat anti-human IgG for 30 minutes at 4~ then incubated with 4 tzg/ ml of propidium iodine for deoxyribonucleic acid (DNA) staining. The coverslips were viewed under a fluorescence microscope. Flow Cytometry Analysis Cells were washed three times with PBS, and fixed with 80% ethanol for 30 minutes at 4~ Staining for CLN-IgG was performed by incubating 1 • 106 cells :~Disperse supplied by Godo Susei Co., Tokyo, Japan. 902
for 60 minutes at room temperature in CLN-IgG, 10ug/ ml. After incubation, the cells were washed three times with 0.05% Tween 20/PBS and then incubated for 30 minutes at room temperature in 20 ul of PBS containing a 1:20 dilution of FITC-labeled goat anti-human IgG. Cells incubated with nonimmune human IgG served as a control for background fluorescence. After being washed three times with 0.05% Tween 20/PBS, the cells were preincubated with 1 mg/ml of ribonuclease for 30 minutes at 30~ then stained with propidium iodine (final concentration of 4 ug/ml). Immunofluorescence and correlated dual-parameter measurements of CLN-IgG versus DNA content were performed with a fluorescence-activated cell sorter.w The green FITC fluorescence signal was proportional to CLN-IgG levels and was measured with a 514-nm bandpass filter. The red propidium iodine fluorescence signal, corresponding to DNA content, was measured using a 560-nm filter in conjunction with a 610-nm longpass filter. For analysis of the antigen on the cell membrane, semiconfluent cells were treated with 0.05% ethylenediaminetetra-acetic acid for 10 minutes at 37~ in order to obtain single cells, then washed with PBS containing 5% fetal bovine serum and 0.05% sodium azide. The cells were incubated for 60 minutes with 5 ug/ml of CLN-IgG at 4~ After being washed with PBS containing 5% fetal bovine serum and 0.05% sodium azide, the cells were stained with a 1:20 dilution of FITClabeled goat anti-human IgG for 30 minutes at 4~ After washing, the signal was measured with the fluorescence-activated cell sorter and stained cells were viewed under the fluorescence microscope. For analysis of the antigen in cytoplasm, single cells were treated with 0.25% trypsin in PBS for 60 minutes at 37~ following fixation with cold 80% ethanol. Flow cytometry analysis was performed as for the analysis of the antigen on cell membrane. Treatment with 0.25% trypsin for 60 minutes at 37~ as described inhibited the expression of antigen on cell membrane (data not shown).
Imrnunoblotting The U-251MG cells (107 cells) were solubilized with 500 ~1 of cell-lysis buffer containing 1% Triton X-100, 1% sodium deoxycholate, 0.01% sodium dodecyl sulfate (SDS), 0.15 M NaC1, 50 mM Tris-HC1 (pH 7.4), and 2 mM phenylmethylsulfonylfluoride. Cyst fluids, sera, and cerebrospinal fluid (CSF) obtained from glioma patients were lyophilized. Samples (1 mg/ml of protein concentration) were dissolved in buffer containing 2.5% SDS, 5% 2-mercaptoethanol and 0.01 M TrisHC1 (pH 8.0), and boiled for 5 minutes at 100~ The procedure for SDS-polyacrylamide gel electrophoresis used in this study essentially followed the method of Laemmli. 13 The stacking gel contained 5% acrylamide wFluorescence-activated cell sorter was manufactured by Becton-Dickinson, Inc., Mountain View, California.
J. Neurosurg. / Volume 73 / December, 1990
Antigen related to cell proliferation
FIG. 1. Immunoperoxidase staining of glioblastoma (A), anaplastic astrocytoma (B), adult normal brain (C), and fetal brain (D) tissues with CLN-IgG. Nuclei were counterstained with hematoxylin. In A, a blood vessel is identified with an arrow. • 150.
and 0.13% N,N'-methylene-bisacrylamide (Bis), and the running gel contained 10% acrylamide and 0.1% Bis. Electrophoresis was performed at room temperature for 4 hours at 20 mA to a 0.4-mm width of gel. The migration front was followed by adding 0.002% bromophenol blue to the electrode buffer. Proteins from the gels were electrophoretically transferred onto nitrocellulose filters at a constant voltage of 50 V for 3 hours as reported by Towbin, eta/. 26 The nitrocellulose filters were immersed in 3% bovine serum albumin then reacted according to the method described for immunoperoxidase staining. Antibody-Dependent Cell Cytotoxicity The U-251MG cells and primary cultured cells from various glioma tissues for use as target cells were mixed with sodium chromate (5~Cr) at 100 uCi/106 cells/ml and incubated for 1 hour at 37~ in a CO2 incubator. The 5~Cr-labeled target cells (1 x 104/well) and effector lymphocytes were incubated with or without 50 ug of CLN-IgG for 6 hours at 37 ~ in a 96-well microplate. Human peripheral lymphocytes collected from healthy donors were used as effector cells. The radioactivity of 0.1 ml of supernatant in each well was collected, and the percentage of lysis was assessed by means of the J. Neurosurg. / Volume 73/December, 1990
formula: cytolysis (%) = ((a - c) + (b - c)) x 100, in which a is the percentage of release (expressed as counts per minute (cpm)) in the presence of effector cells, b is the percentage of maximum release (in cpm) by 1 N HC1, and c is the percentage of spontaneous cpm with medium alone. Results
Immunohistochemical Characterization Table 1 shows the immunohistochemical reactivity of CLN-IgG on various histological types of brain tumor and normal brain tissue. There were 51 brain tumors; seven of the eight glioblastomas (Fig. 1A) and all four anaplastic astrocytomas (Fig. 1B) stained positively. Benign gliomas (oligodendrogliomas, ependymomas, and astrocytomas) showed a low occurrence of CLN-IgG reactivity. The heterogeneous staining by CLN-IgG was typical in tumor cells from glioblastoma specimens (Fig. 1A). Blood vessels in tumor tissue were not stained by CLN-IgG (Fig. 1A). Benign brain tumors (meningiomas, neurinomas, pituitary adenomas, and choroid plexus papillomas) did not stain, but tissue from two of five craniopharyngiomas did. Adult normal brain tissue (Fig. IC) and fetal brain tissue (Fig. 1D) 903
T. Kokunai, N. Tamaki, and S. Matsumoto
FIG. 2. Immunofluorescence staining of fixed U-251MG cells with CLN-IgG. The reactive fluorescein isothiocyanate products with CLN-IgG, especially ruffles, were found on the cell membrane. Nuclei stained with propidium iodine, x 280. did not show any reactivity with CLN-IgG. The antigen recognized by CLN-IgG was localized in the cell membrane of U-25 IMG cells (Fig. 2), especially ruffles. The reactivity of CLN-IgG toward U-251MG cells was resistant to treatment with sodium periodate and other modes of deglycosylation (data not shown). Flow Cytometry The relationship between the expression of antigen and the cell cycle was examined by flow cytometry. Flow cytometric analysis with the use of CLN-IgG showed strong positive binding to U-251 MG and KNS-
42 cells, especially at the GriM phase compared to that at the G~ phase (Fig. 3). When U-251MG cells were treated with cyclic AMP, they showed reduced expression of the antigen recognized by CLN-IgG, especially the expression of the antigen on cell membrane (Fig. 4). The maximum expression of the antigen on the cell surface was detected in the cells at G2/M phase and synchronously cultured U-251 MG cells at the Go/G1 phase showed a markedly decreased expression (Fig. 4). In primary cultured cells from glioblastomas, the expression pattern of the antigen recognized by CLN-IgG was the same as the pattern of U-251MG cells, but more marked expression during the G2/M phase was observed (Fig. 5) than at the Go/G~ phase. I m m unoblotting
Immunoblotting analysis was used to examine the epitope of the antigen in U-251MG cells and the amount of antigen shed into cyst fluids, sera, and CSF of glioma patients (Fig. 6). Cell lysates of U-251MG cells under reduced conditions showed two bands with a molecular weight (MW) of 60 kD and 53 kd. These data suggest that the epitope of the antigen in U-251MG cells recognized by CLN-IgG is the same as that of a previously reported antigen. '~ The antigen recognized by CLN-IgG was detected in the cyst fluid of one glioblastoma, but was a single band with an MW of 60
TABLE 1 Immunohistochemical reactivities of CLN-IgG against various human brain tumors, normal brain, and cultured cells Histology glioblastoma anaplastic astrocytoma astrocytoma medulloblastoma ependymoma oligodendroglioma metastatic tumor adenocarcinoma squamous cell carcinoma craniopharyngioma pituitary adenoma meningioma neurinoma choroid plexuspapilloma normal brain (adult) normal brain (fetus) totals U-373MG U-251MG KNS-42 primary cultured cell glioblastoma anaplastic astrocytoma astrocytoma 904
No. Positive/ Total Cases 7/8 4/4 1/5 1/5 1/2 1/3 3/3 1/2 2/5 0/3 0/5 0/4 0/2 0/5 0/3 21/59 positive positive positive 5/5 3/3 2/3
FIG. 3. Two-color analysis of the vertical fluorescein isothiocyanate (FITC)-fluorescence (CLN-IgG) and the perpendicular propidium iodine-fluorescence (DNA) on KNS-42 (upper) and U-251MG (lower) cells by flow cytometry. J. Neurosurg. / Volume 73/December, 1990
Antigen related to cell proliferation
FIG. 4. Binding pattern of CLN-IgG on the antigen of U-251MG cells obtained with flow cytometric analysis. Upper Left: Exponentially growing cells. Lower Left: Cells treated with cyclic adenosine monophosphate (AMP). UpperRight: Cells synchronously accumulated at the Go/G, phase. Lower Right: Cells collected at the G2/M phase by mitotic shaking. FITC = fluorescein isothiocyanate. kD. Other cyst fluids (six cases), sera (27 cases), and CSF (18 cases) of glioma patients and the supernatant of cultured U-251MG cells showed no reactivity against CLN-IgG.
Antibody-Dependent Cell Cytotoxicity For analysis of the cytotoxic activity of CLN-IgG, we examined the antibody-dependent cell cytotoxicity (ADCC) activity of CLN-IgG against U-251MG and
FIG. 5. Two-color analysis (left) and binding patterns (righO by flow eytometry of CLN-IgG on the antigen of primary cultured glioblastoma cells. FITC = fluorescein isothiocyanate; AMP = adenosine monophosphate.
J. Neurosurg. / Volume 73/December, 1990
905
T. Kokunai, N. Tamaki, and S. Matsumoto
FIG. 6. Immunoblotting analysis of the antigen in U25 IMG cell lysate (lane 1), supernatant of U-251MG cells (lane 2), cyst fluid from glioblastomas (lanes 3 and 4), sera from glioma patients (lanes 5 and 6), and cerebrospinal fluid from glioma patients (lanes 7 and 8).
primary cultured glioma cells. The ADCC activity of CLN-IgG against U-251 MG cells was l 0.81% + 1.24 % at an effector:target ratio of 40:1, and was reduced when the cells were treated with cyclic AMP (Fig. 7). Table 2 shows the ADCC activities of CLN-IgG against primary cultured cells from various glioma tissues. The ADCC activity of CLN-IgG against the primary cultured cells of glioblastoma and anaplastic astrocytoma was 36.2% _+ 12.85% and 29.4% _+ 7.65%, respectively, but against the primary cultured cells of astrocytoma and normal brain CLN-IgG ADCC activity was very low (8.9% -+ 3.65% and 4.3% _+ 2.81%, respectively). Discussion
Reactivity of CLN-IgG Antibody Human monoclonal antibody CLN-IgG reacted with malignant gliomas, especially with glioblastoma and anaplastic astrocytoma. These tumors are the most malignant and invasive among all tumors of neuroectodermal origin. By contrast, CLN-IgG did not react with normal brain tissues. There have been few reports of tumor antigens being detected by human monoclonal antibody. Irie, et al., ~ reported that the L72 antibody (designated anti-OFA-1-2) reacted only with tumor cells of neuroectodermal origin (melanomas, neuroblastomas, and gliomas). In previous studies, j'~~ CLN-IgG has been reported to react with various human cancers such as cervical carcinoma, gall bladder carcinoma, stomach carcinoma, melanoma, lung carcinoma, and prostate carcinoma. The antigen recognized by CLN-IgG had an MW of 226 kD consisting of a-subunit (MW 60 kD) and ~3-subunit (MW 53 kD) which were linked by intermolecular disulfide bond. Our immunoblotting results demonstrated that the epitope of the antigen recognized by CLN-IgG in human 906
FIG. 7. Antibody-dependent cell cytotoxicity of CLN-IgG against U-251MG cells treated without (left) or with (right) cyclic adenosine monophosphate (AMP).
glioma cells had two molecules of the same sizes as in the previously reported cancer cell antigen. L~~ It is suggested that CLN-IgG recognized the epitope of an antigen that commonly exists in a broad range of malignancies beyond the tumors of neuroectodermal origin.
Therapeutic Application There is a potential therapeutic application for monoclonal antibodies against cancer cells because cell surface expression of the antigen is thought to be a major factor in the initial reaction of antibody-mediated immunity against cancer cells. Flow cytometric analysis and immunofluorescence staining showed that the antigen was expressed on the cell membrane of U-251MG cells and primary cultured glioblastoma cells. Moreover, flow cytometric analysis revealed that the expression of the cell-surface antigen was higher in the undifferentiated cells than the differentiated cells treated with cyclic AMP. Maximum expression of this antigen on the cell membrane was observed during the G J M phase of the cell cycle. These data suggest that the function of the antigen recognized by CLN-IgG may be related to cell proliferation. The shedding of the antigen into body fluid is important for cancer immunotherapy. The antigen recognized by CLN-IgG was shed into the cyst fluid in only one case of glioblastoma and did not shed into the cyst fluid, serum, or CSF of any other case or into the supernatant of U-251MG cells. This shed antigen had a single molecule (MW 60 kD) in contrast to the two molecules detected in the cell lysate of U-251MG cells. This finding suggests the possibility that the shed antigen might be a degraded form of the antigen recognized by CLN-IgG. J. Neurosurg. / Volume 73 / December, 1990
Antigen related to cell proliferation TABLE 2 ADCC activity of CLN-IgG against primary cultured cellsfrom various tissues* No. of ADCC Cases Activity (%) glioblastoma 5 36.2 +- 12.85 anaplastic astrocytoma 3 29.4 + 7.65 astrocytoma 3 8.9 - 3.65 normal brain 3 4.3 _ 2.81 * ADCC = antibody-dependent cell cytotoxicity. The effector/ target ratio was 40:1. Means are expressed _+standard deviation of the means.
4.
Histology of Tissue
It is important for cancer i m m u n o t h e r a p y that CLNIgG had antibody-dependent cell cytotoxicity against cells and primary cultured cells o f glioblastomas and anaplastic astrocytomas. Furthermore, synergistic augmentation was reported in complement-dependent cytotoxicity against cancer cells using C L N - I g G and another h u m a n monoclonal antibody. 9 In addition, during the several years since its establishment, the clone has been stable both in C L N - I g G secretion and in specificity against cancer cells? ~ Radioimaging Mouse monoclonal antibodies to h u m a n cancer cells have been used for i m m u n o t h e r a p y 6'25'~9 and radioimaging? There are some practical problems ~ to using mouse monoclonal antibodies in clinical practice, including the induction of anti-mouse IgG after multiple administrations and the potential adverse reactions against certain normal tissues. The accumulation of anti-mouse IgG in the circulation could neutralize subsequently administered monoclonal antibodies and diminish their effectiveness. Circulating mouse m o n o clonal antibody-anti-idiotype complexes might lead to immunological suppression, enhancement o f t u m o r growth, and renal dysfunction. For the above reasons, the h u m a n - h u m a n monoclonal antibody (CLN-IgG) is thought to be of possible use for i m m u n o t h e r a p y or radioimaging of malignant gliomas, and for analysis o f the humoral i m m u n e response against h u m a n malignant gliomas such as the anti-idiotype network.
5. 6. 7.
8.
9. 10. 1 I.
12. 13. 14. 15. 16.
17.
18.
Acknowledgment
We thank H. Hagiwara, Ph.D., for making it possible for us to use CLN-IgG.
19.
References
20.
1. Andreasen RB, Olsson L: Antibody-producing humanhuman hybridomas. III. Derivation and characterization of two antibodies with specificity for human myeloid cells. J Immunol 137:1083-1090, 1986 2. Aotsuka Y, Hagiwara H: Identification of a malignant cell associated antigen recognized by a human monoclonal antibody. Enr J Cancer Clin Oncol 24:829-838, 1988 3. Brown JP, Hewick RM, Hellstr6m I, et al: Human reelJ. Neurosurg. / Volume 73/December, 1990
21. 22.
23.
anoma-associated antigen p97 is structurally and functionally related to transferrin. Nature 296:171-173, 1982 Carrasquillo JA, Krohn KA, Beaumier P, et al: Prognosis of and therapy for solid tumors with radiolabeled antibodies and immune fragments. Cancer Treat Rep 68: 317-328, 1984 Cote RJ, Morrissey DM, Houghton AN, et al: Generation of human monoclonal antibodies reactive with cellular antigens. Proe Natl Aead Sei USA 80:2026-2030, 1983 Dillman RO, Shawler DL, Sobal RE, et al: Murine monoclonal antibody therapy in two patients with chronic lymphocytic leukemia. Blood 59:1036-1045, 1982 Edwards PAW, Smith CM, Neville AM, et al: A humanhuman hybridoma system based on a fast-growing mutant of the ARH-77 plasma cell leukemia-derived line. Eur J Immunol 12:641-648, 1982 Glassy MC, Handley HH, Hagiwara H, et al: UC 729-6, a human lymphoblastoid B-cell line useful for generating antibody-secreting human-human hybridomas. Proc Natl Acad Sci USA 80:6327-6331, 1983 Hagiwara H, Aotsuka Y: Cytotoxicity of mixed human monoclonal antibodies reacting to tumor antigens. Acta Paediatr Jpn 29:552-556, 1987 Hagiwara H, Sato GH: Human x human hybridoma producing monoclonal antibody against autologous cervical carcinoma. Mol Biol Med 1:245-252, 1983 Irie RF, Sze LL, Saxton RE: Human antibody to OFA-I, a tumor antigen, produced in vitro by Epstein-Barr virustransformed human B-lymphoid cell lines, Proc Natl Acad Sci USA 79:5666-5670, 1982 K6hler G, Milstein C: Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256: 495-497, 1975 (Letter) Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685, 1970 Miller RA, Maloney DG, Mckillop J, et al: In vivo effects of murine hybridoma monoclonal antibody in a patient with T-cell leukemia. Blood 58:78-86, 1981 Miller RA, Maloney DG, Warnke R, et al: Treatment of B-cell lymphoma with monoclonal anti-idiotype antibody. N Engl J Med 306:517-522, 1982 Mitchell KF, Fuhrer JP, Steplewski Z, et al: Biochemical characterization of human melanoma cell surfaces: dissection with monoclonal antibodies. Proc Natl Aead Sci USA 77:7287-7291, 1980 Olsson L, Kaplan HS: Human-human hybridomas producing monoclonal antibodies of predefined antigenic specificity. Proc Natl Acad Sci USA 77:5429-5431, 1980 Prasad KN, Hsie AW: Morphological differentiation of mouse neuroblastoma cells induced in vitro by dibutyryl adenosine 3 ',5 '-cyclic monophosphate. Nature 233 (New Bioi):141-142, 1971 Ritz J, Pcsando JM, Sallan SE, et al: Serotherapy of acute lymphoblastic leukemia with monoclonal antibody. Blood 58:141-152, 1981 Robbins E, Marcus PI: Mitotically synchronized mammalian ceils: a simple method for obtaining large populations. Science 144:1152-1153, 1964 Saito H, Uchiyama K, Nakamura I, et al: Characterization of a human monoclonal antibody with broad reactivity to malignant tumor cells. JNCI 80:728-734, 1988 Schlom J, Wunderlich D, Teramoto YA: Generation of human monoclonal antibodies reactive with human mammary carcinoma cells. Proc Natl Acad Sci USA 77: 6841-6845, 1980 Sikora K, Alderson T, Ellis J, et al: Human hybridomas 907
T. Kokunai, N. Tamaki, and S. Matsumoto
24. 25.
26.
27.
908
from patients with malignant disease. Br J Cancer 47: 135-145, 1983 Sikora K, Wright R: Human monoclonal antibodies to lung-cancer antigens. Br J Cancer 43:696-700, 1981 Soule HR, Linder E, Edgington TS: Membrane 126kilodalton phosphoglycoprotein associated with human carcinomas identified by a hybridoma antibody to mammary carcinoma cells. Proc Natl Acad Sci USA 80: 1332-1336, 1983 Towbin H, Staehelin T, Gordon J: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350-4354, 1979 Usuba O, Fujii Y, Miyoshi I, et al: Establishment of a human monoclonal antibody to Hanganutziu-Deicher antigen as a tumor-associated carbohydrate antigen. Jpn J Cancer Res 79:1340-1348, 1988
28. Warenius HM, Taylor JW, Durack BE, et al: The production of human hybridomas from patients with malignant melanoma. The effect of pre-stimulation of lymphocytes with pokeweed mitogen. Eur ,/Cancer Clin Oncol 19:347-355, 1983
Manuscript received September 12, 1989. Accepted in final form June 18, 1990. This work was supported by a Grant-in-Aid for the Comprehensive 10-Year Strategy for Cancer Control from the Ministry of Health and Welfare of Japan, and by the Hagiwara Institute of Health, Japan. Address reprint requests to: Takashi Kokunai, M.D., Department of Neurosurgery, Kobe University School of Medicine, 7-5-1, Kusunoki-cho, Chuo-ku, Kobe 650, Japan.
J. Neurosurg. / Volume 73 / December, 1990
