T-cell-growth-factor (TCGF) activate peripheral blood lymphocytes (PBL), cultured for 14 days, showed killer cell activities against natural-killer resistant Daudi cells in a 4 h 51Cr-release assay. However, the effector cells obtained from patients with nonresectable carcinoma exhibited very much lower cytotoxicity to tumor cells. To analyze the mechanism of depression, we have attempted to examine suppressor cell activities of the TCGF-activated PBL. The assay for the suppressor cell activities was made by in vitro inhibition of cell-mediated cytotoxicity by incubating radiolabeled target tumor cells with lymphokine-activated killer (LAK) cells and TCGF-activated PBL. LAK cells were induced by cultivation with recombinant interleukin-2. TCGF-activated PBL, obtained from four out of ten patients with resectable carcinoma and nine out of ten patients with nonresectable carcinoma, significantly suppressed the LAK cell activities. However, this suppression was not observed in TCGF-activated PBL from ten normal healthy control subjects. TCGF-activated PBL with immunosuppressive reactivity were named lymphokine-activated suppressor (LAS) cells. To investigate the phenotypic characterization of TCGF-activated PBL, the cells were analyzed by two-color flow cytometry. TCGF preferentially expanded CD8+CD11- cells and decreased the growth of CD8+CD11+ cells in both normal healthy control subjects and gastric cancer (resectable and nonresectable) patients. Dominantly expressed CD8+CD11- cells on TCGF-activated PBL in patients--especially those with nonresectable gastric carcinoma--showed strong LAS cell activity, irrespective of the presence of killer cell activities of CD8+CD11- cells in TCGF-activated PBL from normal healthy control subjects. The results suggested the generation of CD8+CD11- LAS cells from cancer patients, and revealed that CD8+CD11- T-cells contained killer and/or suppressor cell function. In addition, it was found that the TCGF-activated PBL from gastric cancer patients were associated with an increased proportion of CD4+ Leu8+, HLA-DR+CD8+ and HLA-DR+CD25+ cells.

Proliferation of normal human T cells in vitro requires activation of resting T cells by lectin or antigen. This stimulation initiates a series of events which includes elaboration of T cell growth factor (TCGF), expression of TCGF receptors, and, ultimately, cellular proliferation. We sought to determine if TCGF was required for expression of the TCGF receptor in phytohemagglutinin (PHA)-stimulated normal human T cells. Utilizing dexamethasone (DEX), a known inhibitor of TCGF production, reductions in T cell proliferation, TCGF production, and TCGF receptor expression, as measured by TCGF adsorption and Tac acquisition, were demonstrated after PHA stimulation. When exogenous partially purified TCGF was added to DEX-containing cultures, the DEX inhibition of proliferation and TCGF receptor expression was completely reversed. These experiments were reproduced utilizing both highly purified TCGF from the Jurkat cell line and purified TCGF synthesized by bacteria from cloned TCGF DNA. Short-term experiments showed TCGF to be capable of restoring Tac antigen expression after DEX inhibition in the absence of cellular proliferation. These results indicate that TCGF is required for optimal expression of Tac antigen-associated TCGF receptors in PHA-activated T cells.

Stable self-reactive T-cell lines were established from the draining lymph nodes of Buffalo rats primed with rat thyroglobulin (Tg) in complete Freund's adjuvant. Following two rounds of in vitro selection with Tg or with purified protein derivative (PPD) of Mycobacterium tuberculosis, Tg- or PPD-specific lines were selected. Self-reactivity, manifested by a potent proliferative response of these lines to syngeneic stimulators in the absence of the selecting antigen (Tg or PPD), emerged by 3-4 weeks after the establishment of the lines. This proliferation could not be attributed to reactivity against heterologous serum components. Despite this syngeneic mixed-leukocyte reaction (SMLR) which was readily detectable, specific stimulation by Tg or PPD over the level of the SMLR could still be demonstrated. These lines also responded to concanavalin A but not to allogeneic stimulator cells. More than 95% of T cells of these lines displayed the rat T-helper-cell-specific (W3/25) marker and they secreted a defined lymphokine, T-cell growth factor (TCGF), upon stimulation with syngeneic stimulator cells, independent of the presence of soluble antigen. T-Cell lines such as the ones described here will be useful in elucidating the biological significance of the AMLR/SMLR and its relation to nominal antigen reactivity.

We developed a local AIT using PEL cultured with TCGF combined with preadministration of OK-432. Twenty-six patients of breast cancer with pleural effusion have been treated with this therapy since 1983. PEL expanded and tumor cells collapsed by day 9 in culture with TCGF. Cultured PEL possessed significantly higher cytotoxic activity against autologous tumor cells than PBL cultured in the same condition (p less than 0.05), but there was no difference between their cytotoxic activities against K562. The proliferation rate of PEL obtained after intrapleural administration of OK-432 was higher than that obtained before OK-432 (p less than 0.01). Moreover, the cytotoxic activities against both autologous tumor and K562 of cultured PEL obtained after OK-432 administration was significantly (p less than 0.05) higher than those cultured PEL obtained before. Cultured PEL (1 x 10(8)-6 x 10(9)) were transferred into the pleural cavity after the intrapleural administration of OK-432 (1-5 KE). The volume of pleural effusion increased temporarily after the administration of OK-432 but significantly (p less than 0.01) decreased after AIT. Tumor cells disappeared cytologically in 22 patients at the last puncture of pleural effusion. Pleural effusion disappeared completely in 19 of 26 patients and decreased by more than 50% in volume in 6 patients. Performance status improved in 22 patients. The response rate for OK-432-combined AIT in the present study was 96%. The survival period of the patients treated by OK-432-combined AIT in this trial was significantly (p less than 0.002) prolonged compared to that of the patients receiving chemotherapy alone.(ABSTRACT TRUNCATED AT 250 WORDS)

In this paper we have examined the possibility that soluble factors produced by the thecal and granulosa cells may be essential local modulators of follicular development. The observations that insulin could influence both the growth and the differentiation of granulosa cells were important in establishing the concept that peptides could act as amplifiers of the actions of gonadotrophins. Insulin alone did not influence aromatase activity significantly but acted synergistically with FSH to augment aromatase activity in rat granulosa cells. Unlike aromatase activity, insulin alone was able to significantly stimulate 3 beta-hydroxysteroid dehydrogenase (3 beta-HSD) activity, the maximum level achieved approaching that obtained with high concentrations of FSH. To determine if insulin could influence other parameters of granulosa cell function in addition to steroidogenesis, we measured a component of extracellular matrix, fibronectin, previously shown to be inhibited by FSH. Treatment with insulin independently inhibited the increase in fibronectin secretion observed in control cultures. Also, insulin alone was able to stimulate quiescent bovine granulosa cells to incorporate [3H]thymidine into DNA under serum-conditions. The concentration of insulin required in these experiments was higher than physiological levels suggesting that other insulin-like growth factors may be involved. Our work and that of others has shown that IGF1 can mimic the actions of insulin and is effective at lower concentrations. A possible source of IGF1 production in the follicle was sought initially by collecting rat granulosa cell conditioned medium, and assessing biological activity and immunoreactivity. Conditioned medium augmented the actions of FSH on aromatase activity and alone stimulated 3 beta-HSD, indicating the presence of insulin-like bioactivity. A positive reaction on immunoblots using specific antiserum confirmed the presence of immunoreactive IGF1. Conditioned medium from thecal cells contained a growth-promoting activity (TcGF) that did not augment FSH-induced aromatase activity. The production of growth factors locally within the follicle may represent the self-amplifying mechanism that enables the dominant follicle to complete its developmental program and ovulate.

Malignant pleural or peritoneal effusion-associated lymphoid (EAL) cells from 17 patients with advanced carcinoma were cultured with autologous carcinoma cells in the presence of either recombinant interleukin 2 (rIL 2) or T-cell growth factor (TCGF). Considerable cytolytic activity of the cells against allogeneic tumor cells, such as K562 and Daudi cells was induced by the cultivation. TCGF-activated EAL cells acquired higher anti-Daudi tumor cytotoxicity than rIL 2-activated EAL cells. The resultant TCGF-activated EAL cells from cancer patients significantly exceeded lytic activity of TCGF-activated EAL cells from patients with liver cirrhosis for control (p less than 0.01). Four of 6 cases examined also showed cytotoxic activity against autologous tumor. In facts, viable carcinoma cells co-cultured with EAL cells and TCGF mostly disappeared during 14 days. Similar phenomenon was not observed in rIL 2-activated EAL cells. Thus, it was suggested that more additional lymphokine other than IL 2 was necessary to generate cytotoxic activity against autologous tumor cells. The cell populations responsible for cytolytic activity to allogeneic and/or autologous tumor cells were investigated by two-color flow cytometry. The majority of killer-effector cells against allogeneic cells in rIL 2-activated EAL cells from cancer patients showed CD4+Leu8- phenotype at population level. In contrast, it was suggested that cytolytic activity against allogeneic and/or autologous tumor cells in TCGF-activated EAL cells might be mediated by CD8+ CD11- and CD8+ CD28+ effector cells.

Some laboratory results and clinical situations suggest that human T cells may be important in the regulation of growth of hematopoietic cells. Since the discovery of T-cell growth factor (TCGF), systems are now available for the long-term specific in vitro propagation of mature normal or neoplastic human T cells, providing an opportunity to study the influence of T cells on hematopoiesis. Recently, 24 cell lines from patients with cutaneous T-cell lymphoma (CTCL) and T-cell acute lymphoblastic leukemia (T-ALL) were grown with TCGF and then assessed for release of humoral factors that affect hematopoiesis. Conditioned media (CM) from these cell lines were tested for erythroid burst-promoting activity (BPA) and granulocyte colony-stimulating activity (CSA). BPA was detected in CM from 3/6 cultures of T-ALL patients and 4/6 CTCL cultures. CSA was found in the CM from 6/8 cultures of T-ALL patients, 7/12 CTCL cultures, and 3/4 CTCL cell lines that become independent of exogenous TCGF for growth. The CSA from several of the neoplastic T-cell cultures stimulated high levels of eosinophil colonies, a possible source of the eosinophilia seen in these patients. The ability of continuously proliferating human T lymphocytes, which retain functional specificity and responsiveness to normal humoral regulation, to produce factors that directly or indirectly stimulate myeloid and erythroid colony formation lends further credence to the role of T lymphocytes in regulating hematopoiesis.

The T cell hybridization technique can be used to prepare continuous cell lines which express the antigen specificity and function of T cells within the milieu of a proliferating lymphoma. Technical details for the preparation and maintenance, selection and analysis of T cell hybrids are defined. Techniques for cloning of hybrid cells and production of hybrid-derived tumors are also presented. Parameters influencing the hybridization frequency and the production of functional hybrids are discussed. A variety of T cell subsets, including suppressor cells and delayed-type hypersensitivity effector cells as well as T cells maintained on TCGF, are excellent sources of primary parents for hybridization. When BW 5147 is used as the T lymphomas parent in these experiments, the resulting hybridization frequencies range between 10 and 434 X 10(-7). We have had moderate success using YAC-1; however, additional lines such as L5178Y, BALENTL 5, EL4 BU and S491TB.2 have proved ineffective as sources of T lymphoma parents. The technique of T cell hybridization is evaluation in terms of retention of differentiated functions and the stability and growth characteristics of the resultant hybrids.

Thymocytes from juevenile Xenopus laevis did not proliferate in response to commercial preparations of lipopolysaccharide (LPS), responded poorly when cultured with the T-cell mitogen, phytohaemagglutinin-P (PHA), and were not co-stimulated by PHA plus LPS. However, supernatants (SNs) from LPS-treated cultures of adult Xenopus macrophage-enriched resident peritoneal cells (PCs) enhanced the proliferative responses of thymocytes to a submitogenic dose of PHA. These SNs were incapable of supporting long-term growth of thymic lymphoblast cell lines, and thus could be distinguished from T-cell growth factor (TCGF)-rich SNs, which were essential for propagating these cells. The co-stimulatory activity was present in 0-24-hr SNs; after 48 hr, SN activity declined. No functional cross-reactivity of mammalian and Xenopus interleukin-1 (IL-1)-rich SNs was detected. These data are consistent with the proposition that a macrophage-derived factor, functionally homologous with mammalian IL-1, can enhance a T-cell proliferative response in an amphibian.

The surface antigenic phenotype of peanut lectin agglutinin (PNA)-separated, high-density ("immature") thymocytes which are the target for the T cell growth factor (TCGF) enabling this otherwise unresponsive cell population to mount a strong proliferative response to concanavalin A (Con A) has been studied. The results obtained indicate that (a) the major population of cells bearing the typical phenotype of immature thymocytes, i.e. low content of H-2, high content of Thy-1. TL, Lyt-1 and Lyt-2 antigens, is unable to respond to Con A in the presence of TCGF, and (b) that the responsive population constitutes a minor fraction (about 15%) of PNA+, high-density thymocytes with surface antigenic phenotypes similar to those of mature T cells, i.e. high content of H-2, low content of Thy-1 and differentiated expression of Lyt antigens. Such a unique combination of properties and the presence of TL antigen on some members of the responding population suggest that it contains cells at intermediate stage(s) of differentiation between immature thymocytes and T cells.

A T-cell growth factor (TCGF) is produced by antigen- or mitogen-stimulated T lymphocytes from the South African clawed frog Xenopus laevis. This study further defines the physical and biological properties of this cytokine and demonstrates that TCGF is biochemically similar to mammalian interleukin-2 (IL-2). Biologically active TCGF eluted from SDS-PAGE displays a M(r) of 16 kD and lectin-affinity chromatography indicates that the three-dimensional configuration of carbohydrates on TCGF and human IL-2 is similar. Secretion of TCGF is detectable 1 day after stimulation of splenocytes with the T-cell mitogen phytohemagglutinin (PHA) and peaks following 2 to 3 days of stimulation. Finally, despite the biological and physical similarities between Xenopus TCGF and mammalian IL-2, anti-human IL-2 monoclonal antibodies do not recognize Xenopus TCGF.

Following a previous observation that moderate physical training (running) of rats did not impair T-cells, in this study moderately trained Wistar rats were run to exhaustion on 2 consecutive days: in one case (T-dex) this was preceded by an intraperitoneal injection of 0.5 mg.kg-1 of dexamethasone (dex) and in the other case there was no prior injection (T). Similarly one group of sedentary control rats, was injected with dex (C-dex) and the other group was not (C). Rats were killed 24 h after the last treatment (dex, exercise). Compared with the C rats, the T rats exhibited a decreased number of thymocytes (75%), in particular CD4+CD8+ thymocytes and splenocytes (55%), notably CD4+CD8- splenocytes (P < 0.01). Also noted in the T rats was a lower (45%) in vitro (+mitogen) percentage of IL2r+CD4- splenocytes (expressing the IL2 receptor), and reduced (40%, P < 0.01) or unchanged in vitro production of T-cell growth factor (TCGF) by splenocytes or blood mononucleated cells (BMC), respectively. The dex decreased the number of thymocytes and splenocytes in the same way in T-dex rats (compared to T rats) and in C-dex rats (compared to C rats, P < 0.01). In T-dex rats compared with C-dex rats, on the other hand, dex had little effect on in vitro TCGF production by BMC, and no effect on other in vitro parameters. These results would indicate that physical exhaustion was responsible for an alteration in T-cells in the moderately trained rat. This alteration was in part enhanced by dex.

We demonstrated the efficacy of a long-term cultured cytotoxic T-lymphocyte line, CTLL-D4, on tumour growth inhibition using athymic nude mice as recipients. CTLL-D4, specific for a unique surface determinant on a radiation-induced leukaemia RL male 1 of BALB/c origin, was obtained from the limiting dilution culture of MLTC cells performed between spleen cells of a CB6F1-nu/+ mouse immunized in vivo and RL male 1 stimulator cells, and cultured for several months in the absence of added TCGF as described in our preceding paper (Kuribayashi, 1985). The specific inhibition of tumour growth by CTLL-D4 was demonstrated both in Winn-type neutralization assay and in systemic transfer experiments. A subcutaneous inoculation of the mixture of CTLL-D4 and RL male 1 cells resulted in the complete inhibition of tumour growth, even at the effector to tumour cell ratio of 1:1, whereas non-cytolytic D4f, which was self-Ia antigen(s)-reactive, composed entirely of Lyt-1+23- T cells and derived originally from CTLL-D4 but completely lost its cytotoxic activity during culture with the irradiated syngeneic feeder cells alone, had no inhibitory effect at all. In the adoptive transfer studies, the subcutaneously established tumours were rejected by the single i.v. transfer of 2 X 10(7) CTLL-D4 cells into CB6F1-nu/nu mice. However, D4f was ineffective again in this systemic transfer system. When the effect of CTLL-D4 cells on tumour rejection in vivo was compared to that of non-cultured spleen cells hyperimmunized with RL male 1 cells, the former exhibited more rapid rejection in nude mice after i.v. transfer than the latter did, suggesting that CTLL-D4 cells also attack the tumour cells much more effectively as effectors in vivo. Thus, it is conceivable that CTLs are mainly involved in tumour rejection in this adoptive transfer system using RL male 1 tumour cells and athymic nude mice.

Splenic T cells obtained from tumor-bearing mice could be cultured with T cell growth factor (TCGF) for over 12 months. The TCGF-dependent lymphoid cell line strongly inhibited cell-mediated anti-tumor immunity directed against syngeneic tumor. However, the suppression was non-specific for the given tumor. The cell line expanded with TCGF expressed a phenotypic characterization of T cells defined by monoclonal anti-Thy-1.2 antibody.

Peripheral blood lymphocytes (PBL) from 38 normal donors and from 27 cancer patients were propagated in bulk cultures for 3-6 weeks using T cell growth factor (TCGF). In addition, cultures derived from lymphocyte preparations enriched for or depleted of natural killer (NK) cells and several clones of cultured cells were studied. The following main observations were made: (a) PBL of both patients and healthy donors could be expanded to large numbers (up to 2500-fold); (b) CLC derived from unfractionated PBL exhibited intermediate levels of cytotoxic activity against autologous and allogeneic fresh lung tumor cells and strong cytotoxicity toward several cultured adherent tumor cells; (c) whereas cultures originated from populations enriched for NK cells were highly cytotoxic against both adherent tumor target cells and against an NK-sensitive leukemic cell line (K562), cultures derived from populations depleted of NK cells were preferentially cytotoxic to adherent target cells; (d) clones of CLC were also strongly cytotoxic, but 2 out of 3 clones tested showed a narrower spectrum of target cytotoxicity than that of uncloned CLC; (e) CLC, when mixed with two carcinoma cell lines, were able to inhibit tumor growth in nude mice.

To confirm the phenotypic characteristics of lymphokine-activated suppressor (LAS)effector cells, we isolated CD8+CD11b- and CD8-CD11b- cells from T-cell growth factor (TCGF)-activated peripheral blood lymphocytes (PBL) in 7 patients with gastric carcinoma (4 non-resectable and 3 resectable carcinoma) and 3 healthy controls. Sorted CD8+CD11b- cells from 3 of the patients with non-resectable carcinoma and from 1 of the patients with resectable carcinoma showed LAS cell activity. However, the LAS cell activity could not be observed in CD8+CD11b- cells from healthy controls. In addition, a sorted CD8-CD11b- subset of cells from both cancer patients and control did not express any suppressive activity. These facts clearly show that the cell populations responsible for suppression of cell-mediated antitumour immunity reside within CD8+CD11b- T-cells, at least in patients with advanced carcinoma.

Alloreactive T cell clones primed in vivo were tested for the expression of T cell differentiation antigens CD2, CD3, CD4, and CD8. Each of 29 different clones were found to express CD2 and CD3, but were variable in their expression of CD4 (7 positive clones) and CD8 (15 positive clones). Six clones were positive for both CD4 and CD8. One of the 29 clones expressed neither CD4 or CD8. Over a period of 12-18 weeks of culture, these clones began to lose their alloreactivity but acquired NK-like activity. By changing the concentration of TCGF, the "allo" and "NK-like" lytic activities could be modulated. After 18 weeks of culture, these clones lost their alloreactive specificity, but not their NK activity. The expression of surface markers was unchanged. CD2 and CD3 molecules were determined to play a role in both the alloreactive and NK activity of these clones.

The T cell lymphoma line EL-4 G-12 produces large quantities of TCGF (IL 2) shortly after exposure to anti-Thy-1 and cross-linking reagents. Early surface events that initiate nonconstitutive synthesis and secretion have been examined by fluorescent microscopy, EM analysis, flow cytometry of anti-Thy-1, and measurements of membrane polarization. The data show that anti-Thy-1 binding leads to immediate and rapid endocytosis and shedding of the antibody molecules, but with no change in membrane potential. Cross-linking of the anti-Thy-1 results in stabilization and accumulation of the antibody on the surface, and rapid hyperpolarization. The data suggest that cross-linking of ligand antibody interactions at the surface is a regulatory event involved in initiation of the signal for TCGF synthesis and secretion.

We have investigated the effects of cleavage of factor B by its activating enzyme, factor D, as well as its activation fragments Bb and Ba, on the growth of mouse spleen B lymphocytes preactivated by LPS. Neither factor B nor factor D show any growth-supporting activity when tested alone. The coaddition of factor B and factor D to serum-free cultures of LPS-preactivated B cell blasts increased the proliferation of the responding cells up to the level obtained by restimulation with LPS. Such growth-supporting activity was shown to be mediated by Ba, whereas Bb did not show any significant effect. Furthermore, this effect was not restricted to the LPS-preactivated B cell blasts; in fact, Ba also supported the growth of in vivo, activated B cell blasts of unprimed mice of the LPS-nonresponder C3H/HeJ strain. In contrast, Ba did not maintain growth of Con A-activated T cells or TCGF-dependent CTL cells. Taken together, these results describe the first biological activity of human Ba as a B cell stimulatory factor.

A T-cell line (H3) was established by culturing human peripheral blood mononuclear cells with influenza virus A/X31 and maintained in long term culture with Interleukin-2 (TCGF). Supernatants were prepared by culturing these cells overnight in the absence of Interleukin-2 but with A/X31 and irradiated autologous E rosette negative cells as a source of antigen presenting cells, and harvesting by centrifugation. The supernatants were shown to replace T cells in helping E- (B) cells to produce antibody specific to A/X31 which was measured by enzyme immunoassay (EIA). Although maximal help was obtained with autologous or semi allogeneic B cells (in the latter case bearing HLA-DR 3 loci) there was still significant antibody production with allogeneic combinations. The supernatants were subsequently fractionated into specific and non-specific helper activities by gel filtration, giving an approximate mol. wt of 50-70,000 and 10-30,000 for each respectively. The specific HF was shown to be genetically restricted in its action upon B cells and also to generate antibody to A/X31 only. The lower molecular weight material acted on any responding B cell regardless of HLA-DR type and produced antibody non-specifically in culture with E- cells even in the absence of antigen. The apparent lack of restriction was therefore due to the masking effect of non-specific and non-restricted HF(s) on the genetically restricted specific HF produced by this line.

The present study provides first evidence for the presence of a thymic cell growth factor (TCGF) in supernatants (20 hrs) of mitogen (Con A) stimulated chicken spleen cells. 2 out of 3 batches of supernatant prepared according to a procedure originally described for the mouse system (10) induced a vigorous proliferative response of Con A prestimulated chicken spleen cells without significant mitogenic effect on unstimulated lymphocytes. No crossreactivity of this chicken-TCGF was observed with prestimulated murine lymphocytes, nor did potent mouse-TCGF preparations exhibit any proliferative effect on chicken cells. The implications of these data for both phylogenetic as well as practical experimental aspects are discussed.

The effect of T cell growth factor (TCGF) on the cell cycle of four T cell lines that differed in their response to TCGF was studied. Activated normal marmoset T cells (OH-1) were totally dependent on the addition of TCGF for long-term proliferation. In the absence of TCGF, the cells were incapable of traversing the cell cycle and became arrested in G1. The addition of TCGF to arrested OH-1 cells stimulated them to enter the S phase after a lag phase of 24 to 33 hr. The TCGF-stimulated cells reached a maximum of cells in the S phase by 12 hr after the initiation of DNA synthesis. TCGF was required for a minimum of 18 hr before cells would enter the S phase. In the absence of TCGF, activated owl monkey (8I) and human (RG) lymphocytes displayed a TCGF-sensitive block; addition of TCGF stimulated these cells to enter the S phase after 12 and 16 hr, respectively. In contrast, an owl monkey tumor-derived T cell line (OMT-1), not dependent on exogenous TCGF for proliferation, was able to progress slowly through the cell cycle without a TCGF-sensitive block. These cells responded to TCGF by showing an initial increase in cells that entered the S phase after a lag of 6 hr and a continuing movement of additional cells into S over the course of the next several hours.

The in vitro cell-mediated immune reactivities of mononuclear cells separated from spleen (SPL) of gastric cancer patients were studied and, these were compared to those of peripheral blood mononuclear cells (PBL) from the same patient. The level of NK activity of SPL was slightly higher than that of PBL (p less than 0.2, by paired t test), as measured by 51Cr release assay using K562 cells as target. TCGF preparations, generated in cultures of PHA stimulated SPL or PBL, were also compared by quantitative assay. SPL produced in significantly larger amounts than PBL (p less than 0.05). Then, the generation of cell-mediated cytotoxicity in mixed cell culture was investigated in both cell populations. When these cells were cultured with B-lymphoblastoid cell line Raji, SPL had much more capacity to generate cytotoxic cells, compared to PBL (p less than 0.1). Moreover, SPL had significantly increased ability of induce concomitant cytotoxicity during sensitizing to normal allogeneic PBL in mixed lymphocyte culture as compared to PBL (p less than 0.005). These results appeared to demonstrate that SPL from gastric cancer patients had much more increased in vitro cell-mediated reactivities, when compared to those of PBL from these patients.