Toshiki Yamasaki, Kouzo Moritake and George Klein
Object. Natural killer (NK) cell—mediated immunosurveillance in the brain is currently obscure, in contrast with the intracerebral immune reaction of cytotoxic T lymphocytes (CTLs) to tumor cells. The goal of this study, in which a controlled tumor model was used, was to investigate a relationship between NK cells and major histocompatibility complex (MHC) Class I gene expression in intracerebral tumor—bearing hosts.
Methods. A matched set of two cloned tumor cell lines (lymphoma+ and lymphoma−), which differ only in MHC Class I gene expression, was established from the parental YAC-1 cell line (a target widely accepted as being sensitive to murine NK cells). An in vivo rapid elimination assay (REA) was performed using tumor cells labeled with [125I] 5-iodo-2-deoxyuridine to evaluate intracerebral NK cell—mediated defense immunity. There was no difference in the in vitro growth rate and c-myc gene expression between lymphoma+ and lymphoma− cells. An in vitro cytotoxicity assay showed that the lymphoma+ cell line was sensitive to MHC Class I—restricted CTL-mediated lysis, whereas the lymphoma− line was refractory to it. Both were susceptible to NK cell—mediated lysis, comparable to the level shown by YAC-1 cells. Flow cytometry revealed that lymphoma+ reacted positively for cell-surface MHC Class I molecules, whereas lymphoma− had no reaction. Four- to 72-hour REAs, performed using either cell line, disclosed no clearance of radiolabeled tumor cells from the brain in independent groups of untreated and T cell—depleted mice; this contrasted with eradication of radioactivity from the lungs. In NK cell—depleted mice, however, there was no elimination of radiolabeled tumor cells from the brain or lungs. The MHC Class I expression on lymphoma+ cells was enhanced after intracerebral inoculation, rendering them less sensitive to NK cells. By contrast, lymphoma− cells remained negative for cell-surface MHC expression, being sensitive to NK cells and refractory to CTLs after intracerebral inoculation. These results indicate the absence of NK cell—mediated lytic activity in the brain. This allows even NK cell—sensitive tumor cells to escape intracerebral immunosurveillance.
Conclusions. These experiments have refined the information that the brain may lack NK cell—mediated defense immunity against intracerebrally growing tumors, representing a characteristic aspect of this immunologically privileged organ.
Toshiki Yamasaki, Kou-Ichi Enomoto, Kouzo Moritake and Takashi Maeno
✓ Intra- and intercellular calcium signaling in glioma cells was examined by mechanical stimulation of a monolayer cell line of methylcholanthrene-induced mouse ependymoblastoma, 203-glioma, with a fine round-tip glass needle. A fura-2 fluorescence image of the glioma revealed a four- to eightfold increase in the cytosolic calcium ion concentration in directly stimulated signal cells. The increased calcium spread to surrounding cells at a speed of 20 µm/sec for a distance of up to 200 µm. Calcium was transmitted between adjacent cells and even in cells up to 200 µm distant from the initially stimulated cell. Microinjection of Lucifer yellow dye showed no gap junctional communication between cells. Depletion of extracellular calcium ion inhibited both cytosolic calcium elevation and propagation to neighboring cells by mechanical stimulus. An intracellular calcium blocker, TMB-8, eliminated the cytosolic calcium mobilization in a mechanically stimulated cell, but had no effect on calcium diffusion to surrounding cells. Nifedipine and verapamil, antagonists of voltage-dependent calcium channels, did not act on the mechanically induced calcium response. This suggests that some stimulating factor may trigger transmission of calcium, which may be ejected directly from single stimulated cells and mediated via a membrane receptor but not through a gap junction. The calcium signaling in a mechanically stimulated cell may be related to both an influx and a redistribution of intracellular calcium from internal stores, while calcium propagation to neighboring cells may involve calcium influx alone.
Toshiki Yamasaki, Hajime Handa, Junkoh Yamashita, Yuziro Namba, Yoshihiko Watanabe, Shigeki Kuwata and Masao Hanaoka
✓ In an attempt to facilitate the long-term proliferative growth and subsequent cloning of cytotoxic T lymphocytes (CTL's) against syngeneic murine 203-glioma (20-methylcholanthrene-induced ependymoblastoma of C57BL/6 mouse origin), sensitized T lymphocytes from tumor-bearing mice were cultured in the presence of T cell growth factor (TCGF). Of five clones established by a limiting dilution technique, two clones (G-CTLL 1 and 2) exhibited tumor-specific cytotoxicity. G-CTLL 1 cells, which possessed much higher cytotoxic activity than G-CTLL 2 cells, were further analyzed. G-CTLL 1 cells were maintained in a TCGF-dependent exponential proliferative culture for over 18 months and continued to mediate an extremely high cytotoxic activity with the target specificity (50- to 100-fold increases over the peak cytotoxic activity of sensitized T lymphocytes in tumor-bearing mice). Their phenotypes of surface antigens were Thy-1+ (weak positive), Lyt-1.−2.+3+, and asialo-GM1 −, and their cytotoxicity was blocked by adding only anti-Lyt-2 monoclonal antibodies. These results indicated that the cloned cells originated from CTL's. The cloned cells were characterized by the production of immune interferon with the glioma antigen-stimulation, suggesting that the immune interferon could enhance the cytotoxic activity of the CTL clone at the site of a clone-target cell recognition event.
Takeshi Uemura, Kouzo Moritake, Yasuhiko Akiyama, Yoriyoshi Kimura, Takashi Shingu and Toshiki Yamasaki
Object. Deuterium oxide (D2O), or heavy water, affects a variety of biological activities different from those of water. The authors examined the antitumoral effect of D2O on brain neoplasms and demonstrated D2O-mediated cytotoxicity by using a Rous sarcoma virus—induced murine malignant astrocytoma cell line, RSVM. The mechanism of the observed cytotoxicity may involve D2O-induced apoptosis and cell-cycle modulation.
Methods. The authors performed an assay with methylthiazol tetrazolium bromide and a trypan blue dye exclusion test to confirm in vitro D2O-mediated cytotoxicity for RSVM cells. At D2O concentrations of 10 to 50%, the cytotoxic effect was dose and time dependent. Flow cytometry analysis revealed programmed cell death (apoptosis) and the accumulation of RSVM cells during the G2/M phase. By applying the terminal deoxynucleotidyl transferase—mediated deoxyuridine triphosphate nick-end labeling method, fluorescein isothiocyanate—annexin V and propidium iodide double staining, and caspase-family protease activity analysis, the authors demonstrated both DNA fragmentation and enhancement of caspase activity after a 48-hour treatment with D2O, thus indicating that D2O induces apoptosis in RSVM cells. Apoptotic DNA fragmentation was completely abolished by the caspase inhibitor Z-VAD-FMK (benzyloxycarbonil-Val-Ala-Aps-fluoromethylketone). The findings indicate that the caspase activation pathway may be involved in D2O-induced apoptosis.
Conclusions. The authors found that D2O is cytotoxic to malignant astrocytoma cells. The mechanism of D2O-mediated cytotoxicity involved the induction of apoptosis and cell accumulation during the G2/M phase. This D2O-induced apoptosis is modulated through the caspase activation pathway.
Toshiki Yamasaki, Hajime Handa, Junkoh Yamashita, Yoshihiko Watanabe, Masaaki Taguchi, Shigeki Kuwata, Yuziro Namba and Masao Hanaoka
✓ The authors have established a murine malignant glioma-specific cytotoxic T lymphocyte clone (G-CTLL 1) by T cell growth factor (TCGF) using 203-glioma (a methylcholanthrene-induced ependymoblastoma of C57BL/6 mouse origin). The cloned cells were found to release a large amount of gamma interferon (IFN) in response to glioma-associated antigen-specific stimulation. The authors have investigated whether the IFN produced can contribute to killing the target cells. Adding anti-mouse gamma IFN antibody to the mixed clone-target cell culture inhibited IFN production by the cloned cells but the toxicity of the cells was minimally diminished. Therefore, it is suggested that the endogenous gamma IFN produced by the TCGF-dependent cloned cytotoxic T lymphocyte line does not have direct cytotoxic action on the target cells. Furthermore, IFN production as well as cytotoxicity was blocked by anti-Lyt-2 monoclonal antibody in the absence of complement. This suggests that IFN plays a role in the process of antigen recognition of target cells because the Lyt-2 molecule is involved in an antigen-specific function on the cytotoxic T lymphocyte receptor.
The role of TCGF in gamma IFN production was also investigated. The spontaneous production of gamma IFN by the cloned cells paralleled the amounts of exogenous TCGF added to the cultures, but TCGF had no synergistic effect on IFN production in the presence of mitogen or tumor antigen. Accordingly, it is possible that TCGF stimulates the cloned cells to proliferate, causing IFN release.
Toshiki Yamasaki, Kouzo Moritake, Mikio Takaya, Takato Kagawa, Hidemasa Nagai, Yasuhiko Akiyama and Masako Kawahara
✓ An intraoperative monitoring tool is described that prevents mechanical injury to intracerebral vessels during stereotactic surgery. The method, which combines pulse Doppler ultrasonography and fiberendoscopy, allowed stereotactic biopsy to be performed without serious intracerebral bleeding in 25 patients with hypervascular malignant brain tumors, 13 with glioblastoma multiforme, five with anaplastic astrocytoma, five with metastatic tumor, and two with malignant lymphoma. The ultrasound apparatus has a built-in fast-Fourier transformation system analyzer and an improved filtering system that provide real-time measurement of blood flow velocity. The source of flow (arterial or venous) could be identified by both real-time sonography and acoustic signal frequencies. It was possible to measure the size and distance of a vessel by adjusting the Doppler signal gain dial from initially waxing to waning sounds, because the acoustic signal was adjusted to the axial flow of each vessel in 0.1-mm steps. Each of three Doppler probes (1 mm, 2 mm, and 3 mm in diameter) fit through the outer cannula of the biopsy needle. Vessels located within 7 mm from the tip of these probes could be detected easily and rapidly, so the biopsy needle could be advanced safely to the desired target in 7-mm steps. If sonograms revealed blood flow, indicating the presence of larger vessels in the intended stereotactic trajectory, the angle of the needle was changed slightly to avoid vascular injury. Because the fiberendoscope was connected to a video processor, the vessel could be visualized at a higher magnification on the video display, unless there was active bleeding. This technically simple and reliable system enhances operative safety while maintaining accuracy.
A review of 30 cases
Toshiki Yamasaki, Hajime Handa, Junkoh Yamashita, Jonathan T. Paine, Yuzuru Tashiro, Akira Uno, Masatsune Ishikawa and Reinin Asato
✓The authors review 30 documented cases of intracranial and orbital cavernous angiomas treated at their institution between 1965 and 1984. The diagnosis was based on computerized tomography (CT) or surgery; three patients were treated in the pre-CT era (1965 to 1976) and 27 since the advent of CT. The number of cases diagnosed preoperatively markedly increased after the introduction of CT, and 22 cases were verified histopathologically at surgery. Six cases were in children (aged 2 months to 17 years) and 24 in adults (aged 19 to 73 years). There was no significant sex difference (male:female ratio was 14:16). Nineteen lesions were intraparenchymal, five were intraventricular, three were in the middle fossa, two were intraorbital, and one originated from the tentorium. Symptoms varied according to the site of the lesion; hemorrhage occurred in 11 cases. Calcifications were seen on CT scans in all cases, but on plain skull films in only two. Angiography revealed hypovascular masses in all cases excluding those with lesions in the middle fossa; in two cases, tumor stain could be detected only with prolonged-injection angiography. Radionuclide brain scanning showed a dense hot area in eight of 19 patients. Recent experience has shown that magnetic resonance imaging clarified anatomic relationships that were obscure on CT. The overall outcome was favorable except for one patient who died in the postoperative period. The clinical results in this series are summarized and some diagnostic and therapeutic problems are discussed.
Atsushi Keyaki, Hajime Handa, Junkoh Yamashita, Yasuhiko Tokuriki, Shin-Ichi Otsuka, Toshiki Yamasaki and Hidefuku Gi
✓ The effects of prostaglandin D2 (PGD2) on the growth of mouse malignant glioma cells were studied in vitro and in vivo. The in vitro studies consisted of various concentrations of prostaglandins (PG's) being added to cultures of mouse glioma cells. At concentrations above 2.5 µg/ml, PGD2 strongly inhibited the proliferation of glioma cells, whereas PGE2 had no effect at the same value. Exposure to 5.0 µg/ml PGD2 for more than 2 hours resulted in inhibition of glioma cell proliferation. This growth-inhibitory effect of PGD2 was related to the inhibition of DNA synthesis of the cells. The in vivo studies were performed with a subcutaneously transplanted mouse glioma model. Injection of 0.5 mg/kg PGD2 into the tumor was more effective than the same concentration given by intraperitoneal injection. In mice with intracranially transplanted glioma, daily intraperitoneal injection of 0.5 mg/kg PGD2 had no significant effect on survival time.
Toshiki Yamasaki, Haruhiko Kikuchi, Kouzo Moritake, Seiichi Nagao, Kouichi Iwasaki, Jonathan T. Paine, Takato Kagawa and Yuziro Namba
✓ Morphological and ultrastructural changes in normal mouse brain tissue were investigated after intracerebral stereotactic injections of tumor necrosis factor (specific activity: 2.0 × 106 U/mg protein) into the right frontal lobe. The mice received either a single infusion or multiple tumor necrosis factor infusions in three different dose groups (10, 100, or 500 U). Compared with sham-treated control mice that received adjusted intracerebral injections of purified albumin, the tumor necrosis factor-treated mice in all dose groups did not show any specific in vivo behavioral abnormalities during the 2 months of study following the infusions.
Histological studies revealed hemorrhage attributable to the mechanics of the intracerebral infusions, a thickening of the arachnoid membranes, a reactive gliosis, and neutrophilic and/or mononuclear cell infiltration along the infusion pathway. A local neutrophilic response was prominent 1 day after tumor necrosis factor injection. An immunohistochemical analysis indicated that the mononuclear cell infiltration consisted of lymphocytes and macrophages. Except for the transient neutrophilic infiltration, these histological alterations did not differ from those seen in the sham-treated control groups, and most nonspecific reactive changes disappeared within 8 weeks after the injections. Furthermore, an ultrastructural study showed no apparent pathological changes in the cytoplasmic organelles of neuronal, glial, and endothelial cells in the tumor necrosis factor-injected mouse specimens. These results suggest that the tumor necrosis factor injections caused no specific toxicity and did not alter the parenchymal and stromal cells comprising normal mouse brain tissue.