A human malignant glioma cell line, LN-18, has been established in monolayer culture and subcultured for more than 115 passages. LN-18 cells grow in vitro as bipolar or stellate cells with pleomorphic nuclei, have a doubling time of about 72 h and a plating efficiency of 3%. The glial nature of these cells has been assessed by ultrastructural examination. The synthesis of glial fibrillary acidic and S-100 proteins could not be demonstrated, although the initial biopsy tissue and the early cultures were positive for the former. The presence of Ia-like antigens on the surface of these cells was demonstrated using allo and xeno antisera. LN-18 cells were also shown to synthesize large quantities of fibronectin. The injection of LN-18 cells into nude mice induced the formation of solid tumor masses that could be retransplanted every 3 weeks and showed a morphology comparable to that of the initial biopsy. Karyotype analysis revealed the presence of three marker chromosomes, constantly present before and after hetero-transplantation.

HLA typing of 80 glioma patients was determined and the antigen frequencies were compared with 176 normal controls. Increased phenotypic frequencies of Bw35 and DRw1 were observed, but when P values were corrected by the number of antigens tested (35), the results were no longer significant.

Cytotoxic antibodies against glioblastoma-associated antigen(s) have been sought for in glioma patient sera. Complement-dependent cytotoxicity (CDC) I and antibody-dependent cell-mediated cytotoxicity (ADCC) assays were used to test sera from 80 patients using 51 Cr labelled target cells derived from eight different glioblastoma lines. As more positive sera were detected with ADCC than with CDC, ADCC assay was used for the remainder of the study. Cytotoxic antibodies were detected in the sera from 8 of 80 glioma patients (10%) by CDC and in 20 of 143 (14%) by ADCC. Fourteen percent of 27 meningioma patients and 16% of 25 normal donors used as controls were found to react in ADCC against the same glioblastoma cell lines. The positive serum samples showed extensive cross-reactions with the different glioblastoma cells, but the pattern of reactivity was different for each serum tested. The antibodies detected did not seem to be directed against tumor-associated antigen(s), since the positive sera were found to have a similar ADCC reactivity against unrelated tumor cells and normal fibroblasts. Moreover, their antiglioma reactivity was absorbed by cells of unrelated tumors and by normal platelets. These results do not support previous reports of specific humoral responses in glioma patients against common tumor-associated antigen(s).

Prolonged secretory stimulation of the exocrine pancreas in the rat by in vivo infusion of caerulein leads to a rapid degranulation of the organ associated with a progressive reduction in the size of the zymogen granules. During the first six to twelve hours of stimulation Golgi complexes are enlarged and several structural forms of multivesicular bodies are found indicating a lysosomal degradation of membrane material in the Golgi area. Maximum secretory activity is obtained after a 24 hour infusion, Golgi complexes appear fragmented, the secretory granules measure only 1/3 to 1/4 their normal size. Thereafter, in spite of a continuous stimulation, the exocrine cells regranulate progressively up to 72 hours of infusion. This regranulation is associated with massive enlargement of the Golgi complexes.

An antiglioma antiserum was produced by immunization of a rabbit with membrane‐enriched preparations of a human malignant glioma cell line, LN‐18. After extensive absorption, this antiserum reacted exclusively with antigenic determinants present on seven out of 16 malignant glioma cell lines tested, as shown by complement‐dependent cytotoxicity in a 51Cr‐release assay. This glioma specificity could be further confirmed by quantitative absorption experiments where cells from a glioma line, LN‐135, abolished the cytolytic reactivity of the antiserum against four other glioma lines. The step‐wise absorption of the crude antiserum consisted of: step 1, absorption with normal peripheral blood lymphocytes from 10 individual donors; step 2, absorption with a pool of cells from four different lymphoblastoid cell lines and one endometrial carcinoma; step 3, absorption with cells from a colon carcinoma; and step 4, absorption with cells from two different melanoma lines. After each absorption step the antiserum was tested by complement‐dependent cytotoxicity against a large panel of malignant glioma lines and control non‐glioma cell lines. After the absorptions from step 1 and 2 the antiserum reacted with all cell lines tested, while after step 3 absorption, it reacted only with cells from malignant glioma, melanomas and fetal brain. Quantitative absorption experiments performed at this stage with fetal brain cells showed that the reactivity for fetal brain and melanoma cells could be abolished, while the antiserum was still cytolytic for malignant gliomas. After the absorption from step 4, the antiserum reacted exclusively with seven malignant gliomas. After a further absorption with normal adult brain homogenate, the antiserum still reacted with four of the seven malignant glioma cell lines. Thus, the antiserum described here recognized three types of antigens: antigens common to cells of neuroectodermal origin, antigens shared by malignant gliomas and adult brain, and antigens expressed only on some gliomas.