Preferential enlargement of leukemia cells using cytoskeletal-directed agents and cell cycle growth control parameters to induce sensitivity to low frequency ultrasound.

Sonodynamic therapy (SDT) is a form of ultrasound therapy that has been shown to preferentially damage malignant cells based on the relatively enlarged size and altered cytology of neoplastic cells in comparison to normal cells. This study sought to determine whether cytoskeletal-directed agents that either disrupt (cytochalasin B and vincristine) or rigidify (jasplakinolide and paclitaxel) microfilaments and microtubules, respectively, affect ultrasonic sensitivity. U937 human monocytic leukemia cell populations were treated with each cytoskeletal-directed agent alone, and then sonicated at 23.5 kHz under relatively low power and intensity (20-40 W; 10-20 W/cm(2)), or at 20 kHz using moderate power and intensity (60 W; 80 W/cm(2)). In addition, human leukemia lines U937, THP1, K562, and Molt-4, and the murine leukemia line L1210 were sonicated using pulsed 20 kHz ultrasound (80.6 W; 107.5 W/cm(2)) both with and without the addition of cytoskeletal-directed agents to assess whether cytoskeletal-directed agents can potentiate ultrasonic sensitivity in different leukemia lines. Human hematopoietic stem cells (hHSCs) and leukocytes were sonicated with continuous 23.5 kHz ultrasound (20 W; 10 W/cm(2)) to determine whether this approach elicited the preferential damage of neoplastic cells over normal blood components. To determine whether ultrasonic sensitivity is exclusively dependent on cell size, leukemia cells were also enlarged via alteration of cell growth parameters including serum deprivation and re-addition, and plateau-phase subculturing. Results indicated that cytochalasin B/ultrasound treatments had the highest rates of initial U937 cell damage. The cells enlarged and partially synchronized, either by serum deprivation and re-addition or by plateau-phase subculturing and synchronous release, were not comparably sensitive to ultrasonic destruction based solely on their cell size. In addition, cytochalasin B significantly potentiated the ultrasonic sensitivity of all neoplastic cell lines, but not in normal blood cells, suggesting that preferential damage is attainable with this treatment protocol. Therefore, it is likely that ultrasonic cell lysis depends not only on cell size and type, but also on the specific molecular mechanisms used to induce cell enlargement and their effects on cell integrity. This is supported by the fact that either the microfilament-or microtubule-disrupting agent produced a higher rate of lysis for cells of a given size than the corresponding stabilizing agents.

Autophagic and apoptotic response to sonodynamic therapy induced cell damage in leukemia l1210 cells in vitro.

PURPOSE: The present study was initiated to study the autophagic and apoptotic response to sonodynamic therapy (SDT) in murine leukemia L1210 cells in vitro. METHODS: L1210 cells were examined after 6 hours of incubation following SDT induction with ultrasound at a frequency of 1.1 MHz and an intensity of 3 W/cm(2) in the presence of 1 µg/mL protoporphyrin IX. Apoptosis rate and cell survival rate were assessed through double fluorescent staining with Annexin V-PE and 7-amino-actinomycin D as determined by flow cytometry. Transmission electron microscope and fluorescence microscope were used to identify the formation of acidic vesicular organelles (autophagic vacuoles) during autophagy. Western blots were used to examine the processing of light chain 3 (LC3)-I to LC3-II and Atg5 expression levels. RESULTS: This study showed that SDT treatment reduced the numbers of viable cells to 30.17% and enhanced the apoptotic cells to 19.37% (p < 0.05 compare with control). Additionally, autophagic vacuoles formation clearly occurred after SDT and simultaneously accompanied by obvious LC3 processing and increased Atg5 expression levels. CONCLUSIONS: In L1210 cells, both apoptosis and autophagy were involved in cell damage induced by SDT treatment at the experimental conditions.

Anti-NeuGcGM3 antibodies, actively elicited by idiotypic vaccination in nonsmall cell lung cancer patients, induce tumor cell death by an oncosis-like mechanism.

1E10 is a murine anti-idiotypic mAb specific for an idiotypic mAb that reacts with NeuGc-containing gangliosides, sulfatides, and Ags expressed in some human tumors. In melanoma, breast, and lung cancer patients, this anti-idiotypic Ab was able to induce a specific Ab response against N-glycosylated gangliosides, attractive targets for cancer immunotherapy as these glycolipids are not naturally expressed in humans. A clinical study with nonsmall cell lung cancer patients showed encouraging clinical benefits. Immunological studies performed in 20 of these patients suggested a correlation between the induction of Abs against NeuGcGM3 and longer survival times. The induced anti-NeuGcGM3 Abs recognized and directly killed tumor cells expressing the Ag, by a mechanism independent of complement activation. In the present work, we show that this cytotoxicity differs from apoptosis because it is temperature independent, no chromatin condensation or caspase 3 induction are detected, and the DNA fragmentation induced has a different pattern than the one characteristic for apoptosis. It is a very quick process and involves cytosqeleton reorganization. The Abs induce cellular swelling and the formation of big membrane lesions that allow the leakage of cytoplasm and the loss of the cell membrane integrity. All of these characteristics resemble a process of oncotic necrosis. To our knowledge, this is the first report of the active induction in cancer patients of NeuGcGM3-specific Abs able to induce complement independent oncotic necrosis to tumor cells. These results contribute to reinforcing the therapeutic potential of anti-idiotypic vaccines and the importance of NeuGcGM3 ganglioside as antitumor target.

Hexaminolevulinate-mediated photodynamic purging of leukemia cells from BM.

Photodynamic therapy (PDT) with porphyrin precursors has been established for tumor treatment. This study aimed at examining applicability of hexaminolevulinate (HAL) for photodynamic purging of leukemic cells from BM grafts and evaluating the clinical relevance of in vitro models. The PDT dose resulting in no colony formation by leukemic cells in vitro, in pure form or in a mixture with BM cells, was insufficient for complete killing of the leukemic cells ex vivo and for the treatment of the leukemia-bearing animals in vivo. The efficacy of HAL-PDT in cell lines in vitro should be verified in clinically relevant in vivo models.

Combination of photodynamic therapy + immunotherapy + chemotherapy in murine leukiemia.

Photodynamic therapy (PDT) is a treatment for cancer based on the photosensitization of tumor cells by photosensitive drugs and their subsequent destruction on exposure to light of particular wavelength. The combination of drug uptake in malignant tissues and selective delivery of laser-generated light provides an effective therapy with efficient tumor citotoxicity and minimal normal tissue damage. Since immune response of the host is important in the control of tumor growth and spreading, PDT is able to increase the antitumor immunity. In our laboratory we examined the antitumor effect of combination of PDT, with photoactivated M-THPC (meta-tetrahydroxyphenylchlorin, FOSCAN, Temoporphirin), adoptive immunotherapy, with immune lymphocytes, and chemotherapy on advanced murine tumors. Mice bearing L1210 tumor were treated at day +4 with Navelbine (NVB 1mg/Kg), at day +5,+6 with PDT (0.3mg/Kg of mTHPC and 100mW/cm(2) x 200'' of exposure of laser light), and at day + 7 with immune lymphocytes(IL), collected from mice pretreated with PDT(2x10(7) cells). The results show that the combination NVB + PDT + IL demonstrates a significant synergistic antitumor effect while the chemotherapy treatment with low dose of the drug is uneffective. The same positive results were obtained with the combination of Cisplatin (CDDP 0.5mg/Kg), PDT and IL, while the CDDP treatment alone is completely uneffective. In conclusion, these results suggest that it is possible to completely cure animals bearing advanced tumors, with a combined therapy, PDT + adoptive immunotherapy + low dose chemotherapy.

[Anti-tumor effect of two adjuvants of CpG ODN on leukemic tumor in mouse models].

To investigate the anti-tumor and side effect of CpG ODN 1826 and CpG ODN2006 as an adjuvants on leukemic tumor in mouse-models, an acute lymphocytic leukemic tumor in mouse model was established, then inoculated inactivated L1210 cells alone or with different vaccine adjuvants were injected subcutaneously into each DBA/2 model mouse at different times. The activities of mice, the tumor formation rate and the growth status of leukemic tumor were observed. The tumor was examined by pathologic section. The results showed that the vaccine of inactivated L1210 cells and CpG ODN 1826 could decrease the leukemic tumor formation rate, slow down the growth of leukemic tumor mass in mice and obviously cause necrosis of tumor cells, but it could not prolong the life spans of the tumor-burden mice; while CpG ODN2006 could not only decrease the tumor formation rate, slow down the growth of tumor mass in mice and result in obvious necrosis of tumor cells, but also could eliminate the existing tumor mass in mice, and prolong the life spans of the tumor-burden mice. It is concluded that using CpG ODN2006 as an adjuvant enhances the anti-tumor effect against the leukemic tumor, prolong the life span of tumor-burden mice without obvious side effect.

Anti-tumor effects of the bacterium Caulobacter crescentus in murine tumor models.

Caulobacter crescentus is a gram negative, non-pathogenic bacterium, common in aquatic and soil environments. One feature of note is a protein surface layer (S-layer) composed of a single protein, organized as a self-assembled crystalline array that coats the bacterium. In the course of efforts to express cancer-associated peptides as genetic insertions into the S-layer, we noted a tumor suppressive effect of the unmodified bacterium. C. crescentus was examined for anti-tumor activity against three transplantable tumor mouse models: Lewis lung carcinoma cells transfected with the MUC1 gene in C57BL/6, murine mammary carcinoma (EMT-6) in BALB/c (both in prophylactic and therapeutic mode) and murine leukemia cells (L1210) in DBA2. Mice were immunized three times i.p. with C. crescentus (2 x 10(7) cells/mouse). In prophylactic mode, the mice were challenged with tumor cells two weeks after the last immunization. Immunization with live C. crescentus resulted in anti-tumor activity in all three transplantable tumor models, as measured by prolonged survival, reduced tumor mass or reduced number of lung nodules, compared to saline control groups. In the Lewis lung and the EMT-6 mammary carcinoma murine models the number of lung nodules as well as the tumor weight was lower in mice treated with C. crescentus, compared to the control group; for EMT-6, this was observed in prophylactic and therapeutic modes. In the murine leukemia and Lewis lung carcinoma models prolonged survival was observed in the groups of mice immunized with Caulobacters. In most cases the live C. crescentus cells were markedly more efficacious than heat killed or formalin fixed cells, despite the fact that they do not grow or persist in mice. The results suggest that C. crescentus may be a safe, bacterial immunomodulator for the treatment of tumors.

Induction of apoptosis in hematopoietic cells with an antibody against tomoregulin-1.

Tomoregulin-1, a type-I transmembrane protein with two follistatin modules, a unique epidermal growth factor (EGF) domain and a short, highly conserved cytoplasmic tail, was studied. A number of hematopoietic cell lines (L1210, CEM, Jurkat, U937, K562, JY, THP-1 and T2) express tomoregulin-1 endogenously. In these cells, apoptosis was induced by an antiserum (C29) and purified IgG against the follistatin modules, but not by antisera against the EGF-domain or the cytoplasmic tail. Furthermore, C29 induced apoptosis in tomoregulin-1-, but not in mock-transfected cells. Apoptosis was monitored through genomic DNA fragmentation, annexin-V staining and caspase-3 activation. Treatment of the cells with C29 in the presence of H89 (a SerlThr kinase inhibitor) or 8'-bromo-cyclicAMP revealed that apoptosis was mediated by a cAMP-dependent Ser/Thr kinase. Moreover, C29 increased [cAMP]i over 5-fold. Together, these data suggest that the C29 antiserum against tomoregulin-1 induces apoptosis of hematopoietic cells.

A bivalent single-chain Fv fragment against CD47 induces apoptosis for leukemic cells.

We constructed a single-chain antibody fragment (scFv) of murine monoclonal antibody, MABL, which specifically bound to human CD47 (hCD47) and induced apoptosis of the leukemic cells. The scFv of MABL antibody with a 15-residue linker (MABL scFv-15) formed both dimer (Mr 50 kDa) and monomer (Mr 25 kDa). Both MABL scFv-15 dimer and monomer had binding activity for hCD47. MABL scFv-15 dimer strongly induced apoptosis of hCD47-introduced mouse leukemic cells in vitro and exhibited anti-tumor effect in a myeloma transplanted mice model. However, MABL scFv-15 monomer scarcely exhibited these activities. These results strongly demonstrate that the ligation of CD47 antigen by two antigen-binding sites of MABL dimer is needed for inducing apoptosis. The parent MABL antibody caused hemagglutination due to the CD47 expressed on erythrocytes. Interestingly, MABL scFv-15 dimer did not cause hemagglutination. This apoptosis-inducing dimer appears to be a lead candidate for novel leukemic therapy.

Polyclonal antibodies to xenogeneic endothelial cells induce apoptosis and block support of tumor growth in mice.

In this study, we demonstrate that vaccination of rabbits with murine endothelial cells yields polyclonal immunoglobulin (IgG) with potent antiangiogenic activity. The mechanism of this response appears to be through apoptosis of endothelial cells in vitro. Induction of polyclonal IgG in a xenogeneic host may be useful in passive immunotherapy of a variety of cancers. In fact, the antibody showed antitumor activity in three mouse tumor models (murine B16F10 melanoma, murine SVR angiosarcoma, and human DLD-1 colorectal adenocarcinoma). The polyclonal antibody generated here demonstrated utility in radioimaging of tumors in vivo, using positron emission tomography (PET) imaging, and suggested an antitumor effect in vivo. The results suggest that the antitumor effect in vivo may be related to antiangiogenic effects. Furthermore, anti-endothelial cell antibodies such as these could be useful reagents in isolating specific targets that comprise and induce the antiangiogenic effect.

Purging in autologous hematopoietic stem cell transplantation using adenosine triphosphate (ATP) and 4-hydroperoxycyclophosphamide (4-HC).

In this study, we show potent in vitro purging induced by adenosine triphosphate (ATP) for leukemic cells. The treatment of murine L1210 leukemic cells with 2mM of ATP in vitro for 3h was able to reduce the number of leukemic clonogenic cells by about one order of magnitude presumably by changing the permeability of the leukemic cell membrane. Furthermore, the incubation of L1210 cells with ATP (2mM) and low dose 4-hydroperoxycyclophosphamide (4-HC, 2 microg/ml) for 3h resulted in at least a four-log reduction of clonogenic L1210 cells. Only a slight degree of toxicity to pluripotent hematopoietic stem cells (CFU-S) was observed in both treatment protocols. To determine the efficacy of pharmacological purging by ATP, we designed a murine system to mimic the conditions expected in the clinical setting of autologous transplantation using simulated partial remission marrow (SPRM) which was prepared by mixing normal marrow cells and L1210 cells at a ratio of 9:1. After the SPRM cells were incubated in vitro at a concentration of 1 x 10(6)/ml with both ATP (2mM) and low dose 4-HC (2 microg/ml) for 3h, 5 x 10(4) of the cells were then injected into lethally irradiated 9 weeks male BDF1 mice. All the mice given untreated-SPRM died of leukemia by day 27, whereas none of the recipients transplanted treated-grafts had died by day 70, thus suggesting that the combination use of ATP and 4-HC may be a potentially effective way to purge leukemic cells in autologous stem cell transplantation. The mechanism of the selective killing of leukemic cells is assumed that 4-HC is effectively incorporated into leukemic cells by increasing the permeability of the cell membrane by ATP. Taken together, this simple and rapid procedure is able to purge leukemic cells from autologous bone marrow grafts.

Biological investigation of the platinum(II)-[*I]iodohistamine complexes of potential synergistic anti-cancer activity.

Cisplatin chemotherapy in combination with external irradiation or with low-dose continuous internal radiotherapy produces significant supra-additive treatment effects towards several tumor cells. The purpose of our research is to develop a new class of platinum-based anticancer drugs containing moieties of synergistic potency such as platinum core and a radiotherapeutic isotope which, delivered directly to the tumorous cells by a specifically designed vectors, should produce a local enhancement of therapeutic dose. Thus, we have synthesized a new platinum-iodohistamine complex and its radioactive analogues labeled with I-125 and I-131. In the present study some biological properties of those compounds have been investigated. The in vitro screening study pointed out that non-radioactive platinum-iodohistamine complex possesses high cytostatic activity against COLO-205 cells, and moderate activity against HL-60 cell line. No cytotoxicity was observed against MOLT-4 and L-1210 cells, as well as against VERO normal cells. The biodistribution of intravenously administered radioactive platinum-[131I]-iodohistamine complex to normal rats revealed the highest accumulation in the liver (c.a. 40%ID). Intraperitoneal injections of the complex to tumor-bearing C3H mice resulted in scattering of the dose in the organs (mainly in GIT, liver, kidney). The retention of radioactive complex in neoplastic tissue was 3-4 times higher than in normal muscular tissue, although exhibited the tendency to decrease with time post injection. The results of the present study show promising features of the newly developed platinum-iodohistamine complexes and justify prospective investigation of in vivo anticancer potency on animal models of solid tumors.

Potentiatied antitumor effectiveness of combined chemo-immunotherapy with interleukin-12 and 5-fluorouracil of L1210 leukemia in vivo.

In this study we investigated the efficacy of a combination of IL-12 and 5-FU, a chemotherapeutic exerting several immunomodulatory effects, in murine L1210 leukemia. Mice inoculated with 1 x 10(5) leukemia cells were treated with a single dose of 5-FU (50 mg/kg) and seven daily doses of IL-12 (100 ng/dose), and were observed for survival. Treatment with IL-12 or 5-FU given alone produced moderate anti-leukemic effects. However, combination of both drugs resulted in a significant prolongation of mouse survival time. Importantly, there were 70% of long-term (>60 days) survivors among mice treated with both agents simultaneously. Moreover, we observed 100% of long-term survivors when mice were treated with a minimally increased dose of IL-12 (170 ng) in combination with 5-FU (50 mg/kg). The antileukemic effects were completely abrogated in scid/scid mice and in mice depleted of peritoneal macrophages and significantly decreased after administration of anti-CD3+, anti-CD4+ or anti-CD8+ monoclonal antibodies. Administration of anti-NK1.1 antibodies did not decrease the antileukemic effects indicating that NK cells are not important effectors of this treatment regimen. Collectively, these results indicate that the combination of IL-12 and 5-FU is inducing strong antileukemic responses that are dependent on the presence and activity of macrophages and T lymphocytes and warrant further studies of combined chemo-immunotherapy with IL-12.

Protection of hematopoietic cells against combined O6-benzylguanine and chloroethylnitrosourea treatment by mutant forms of O6-methylguanine DNA methyltransferase.

This report demonstrates that expression of the P140A O6-methylguanine DNA methyl transferase (MGMT) mutant via retrovirus-mediated gene transfer leads to significant, but modest, resistance of cells to both 6-benzylguanine (6-BG) depletion and treatment with 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU). Expression of the P140A/G156A double mutant appeared to be associated with reduced or unstable protein in hematopoietic cells.

Cisplatin-resistant derivatives of murine L1210 leukemia cells are not susceptible to growth-inhibiting and apoptosis-inducing actions of transforming growth factor-beta1.

Murine L1210 leukemia cells possessing an increased resistance to cisplatin were found to be refractory to transforming growth factor (TGF)-beta1-induced growth inhibition, while the parental L1210 cells were strongly inhibited by this cytokine. Growth inhibition was estimated on the basis of [3H]thymidine incorporation, cell counting and colony-forming assay. Cisplatin-resistant L1210 cells were also shown to be much more resistant than the parental cells to both cisplatin- and TGF-beta1-induced apoptosis. These results suggest the existence of cross-resistance to cisplatin and TGF-beta1 in the studied leukemia cells.

A membrane-lytic immunoconjugate selective for human tumor T-lymphocytes.

An immunoconjugate was constructed from a monoclonal antibody recognizing human T-lymphoblastoid cells and a membrane-lytic cytotoxin purified from the venom of the Thailand cobra. Activities of this novel immunoconjugate were compared using human and murine T-lymphocyte cell lines. The ability of the conjugate to inhibit human T-cell proliferation, as measured by incorporation of [3H]thymidine, was three to four times higher than its ability to inhibit proliferation of mouse L1210 cells. The immunoconjugate EC50 for human CEM cells was equivalent to 0.1 nmoles per 2 x 10(5) target cells. Immunoconjugate selectivity paralleled the monoclonal antibody's binding characteristics. Preincubation with free antibody blocked the effect of the conjugate, but only upon the human target cells. This study supports the feasibility of directing a toxic moiety to the surface of a cancer cell to accomplish cell destruction without requiring prior toxin internalization and uncoupling from its antibody carrier.

Effective chemo-immunotherapy of L1210 leukemia in vivo using interleukin-12 combined with doxorubicin but not with cyclophosphamide, paclitaxel or cisplatin.

It has been well established that chemo-immunotherapy using cytotoxic drugs and appropriate cytokines offers a new approach to increasing the therapeutic index in the treatment of neoplastic diseases. This study investigates the efficacy of combinations of interleukin-12 with cyclophosphamide, paclitaxel, cisplatin or doxorubicin in the murine L1210 leukemia model. Mice inoculated i.p. with 1 x 10(3) or 1 x 10(5) leukemia cells were treated with interleukin-12 and/or chemotherapeutics, and were observed daily for survival. Immunosuppression with X-irradiation or macrophage depletion with injections of silica were used to examine the dependence of the therapeutic effects on the efficiency of the immune system. Treatment with interleukin-12 or one of the studied chemotherapeutics given alone resulted in moderate antileukemic effects. Combination of interleukin-12 with cyclophosphamide or paclitaxel produced no augmentation of anti-leukemic effects in comparison with these agents given alone. Combination of interleukin-12 with cisplatin resulted in prolongation of the survival time; however, in the experiment with mice inoculated with 1 x 10(5) leukemia cells, no long-term survivors (>60 days) were observed; on the contrary, combination of interleukin-12 with doxorubicin resulted in 100% long-term survivors. This effect was completely abrogated either by X-irradiation of mice or by macrophage depletion. We also found that doxorubicin augments IL-12-stimulated production of interferon-gamma in vivo. Our observations demonstrating potentiation of the antileukemic effects of the IL-12 and doxorubicin combination suggest that the combined use of these 2 agents could be beneficial in leukemia therapy.

Interactions of 2',2'-diflurodeoxycytidine (gemcitabine) with tumor necrosis factor alpha or its mutein VI in murine leukemias L1210 and P388.

The interactions of 2',2'-difluorodeoxycytidine (gemcitabine) with tumor necrosis factor alpha (TNF-alpha) or its mutein VI on the survival time of mice bearing L1210 and P388 leukemia was investigated. Four hundred eighty male CD2F1 mice were used in the experiments. They were given gemcitabine (20 mg/kg) on days 1, 4, 7 and 10 after i.p. inoculation with leukemic cells (day 0). Cytokines were administered i.p. at a dose of 250 microg/kg as daily injections for 8 days or at a dose of 400 microg/kg given 2, 4, 6 and 8 days after day 0. Drugs were administered alone and in combination. Our study indicates that TNF-alpha and its mutein VI decrease the antileukemic effect of gemcitabine on murine leukemias L1210 and P388.

Synergistic effects of Actinobacillus suis cells in combination with 5-fluorouracil on mice bearing murine L1210 leukemia cells.

The mean survival rates of female BDF1 mice transplanted intravenously (i.v.) with murine L1210 leukemia cells were significantly prolonged by intraperitoneal (i.p.) pretreatment (before i.v. transplantation) or by i.p. pre- and post-treatment (before and after the i.v. transplantation) with heat-killed Actinobacillus suis cells ATCC 15557 (AS 15557) alone, as compared with untreated L1210-leukemia-cell-bearing control mice. However, significant prolongation of the mean survival rates was not elicited by the i.p. post-treatment with AS 15557 alone. When 5-fluorouracil (5-FU) was applied i.p. to mice receiving post-treatment with AS 15557 alone, the mean survival rates of the L1210-leukemia-cell-bearing mice were significantly prolonged. The antileukemic action of AS 15557, alone or in combination with 5-FU, against L1210 leukemia was superior to that of a streptococcal preparation (OK-432) and was almost the same as of bacillus Calmette-Guérin with or without 5-FU. The results suggest the possibility that the synergism of AS 15557 in combination with 5-FU may be dependent on the relationship between the indirect immunological function of AS 15557 and the direct cytotoxic action of 5-FU on L1210 leukemia cells.

Augmentation of antitumor immunity using genetically M-CSF-expressing L1210 cells.

Macrophage colony-stimulating factor (M-CSF) enhances tumoricidal activities of macrophages. We transduced human M-CSF cDNA into the mouse lymphoid cell line, L1210, and examined the antitumor effect of the locally expressed M-CSF. Mice injected with the M-CSF-producing subline showed improved survival in comparison with the mock-transfected cell line or parental cell line plus M-CSF administration (20 microg/kg for 3 days) at inoculated cell numbers of 10(2) or 5 x 10(3). The survival rate at 50 days after injection of 10(6) high M-CSF-expressing cells was 80%, significantly higher than that after injection of the mock-transfected cells, which killed all the mice by day 23. The survival rate appeared to depend on the amount of M-CSF produced. Moreover, all surviving mice after intravenous injection of the M-CSF-expressing sublines were rechallenged with 10(6) parental L1210 cells at day 50, and all survived up to day 100, demonstrating that M-CSF-expressing cells induced immune protection against the parental cells. The same improvement of survival was observed in mouse M-CSF-expressing cell lines. These observations imply that M-CSF cDNA is a candidate gene for use in gene therapy in leukemia.