Tumor Growth Inhibitory and Natural Suppressor Activities of Murine Bone Marrow Cells: A Comparative Study

CELLULAR IMMUNOLOGY ARTICLE NO.

182, 12–19 (1997)

CI971218

Tumor Growth Inhibitory and Natural Suppressor Activities of Murine Bone Marrow Cells: A Comparative Study Victor I. Seledtsov, Vadim Y. Taraban, Galina V. Seledtsova, Denis M. Samarin, Ilias V. Avdeev, Vladimir V. Senyukov, and Vladimir A. Kozlov Department of Experimental Immunology, Institute of Clinical Immunology, Novosibirsk, Russia Received January 31, 1997; accepted October 1, 1997

not identical in their cell compositions, but also differ in their mechanisms of antiproliferative action; (2) a contact cell-to-cell interaction may play a significant role in BM-cell-mediated tumor growth inhibition; (3) the activated lymphocytes, through both IFN-g-mediated and IFN-g-independent pathways, are able to operatively up-regulate the cytostatic activity of BM cells; and (4) the tumor growth-inhibitory activity exhibited by the normal unmanipulated BM cells, at least in its significant part, may not be a consequence of thymus-dependent immune processes occurring normally in the body. q 1997 Academic Press Key Words: bone marrow cell; natural suppression; tumor growth inhibition.

The murine bone marrow (BM) cells having a certain phenotypic similarity to null natural suppressor (NS) cells have been previously established to be able to inhibit in vitro leukemic cell growth in a genetically unrestricted manner. In this study we found that the treatment of normal (C57BL/6 1 DBA)F1 BM cells with a lysosomotropic agent, L-leucine methyl ester (LME), largely abrogated their ability to reduce both P815 mastocytoma and L1210 lymphoma cell proliferation, as well as their NS activity tested for suppression of mitogen (Con A or LPS)-driven spleen cell proliferation. However, after being depleted of the cells binding wheat germ agglutinin (WGA), the BM cells maintained tumor growth-inhibitory activity, while demonstrating no significant NS activity. Moreover, in contrast to T-cell blastogenesis-inhibitory NS activity of BM cells, that was greatly reduced by the addition into the culture of either neutralizing anti-interferon (IFN)-g antibody (Ab) or NG-monomethyl-L-arginine, a competitive inhibitor of NO synthase, natural antitumor cytostatic activity of BM cells was not found to be dependent on the presence in medium of IFN-g and to be associated with NO production. When incubated at suboptimal numbers with tumor cells on conic, round, and flat well bottoms for 7 h, BM cells provided the most, middle, and least (or no) tumor growth inhibition, respectively, suggesting, thereby, a significance of cell density in cytostatic process. It was also found that the BM cells cultured for 20 h with the medium conditioned by mitogen-preactivated T or B lymphocytes were significantly more suppressive to tumor cell proliferation than the BM cells cultured in medium alone. The potentiation of BM-cell cytostatic activity by T-cell-conditioned medium (CM), but not that by B-cell-CM, was found to be partially reversed by anti-IFN-g Ab. Finally, a noticeable tumor growth-inhibitory activity, which could be significantly enhanced upon T-cell-CM, was shown to be also attributable to BM cells from athymic BALB/c mice. Taken together, the results suggest that (1) the tumor growth inhibitory BM cells and the NS BM cells are

CI 1218

/

6c2b$$$$21

We (1, 2) and others (3) have previously reported that bone marrow (BM)1 cells isolated from adult mice are able to suppress nonspecifically in vitro growth of different leukemic cells, while not killing them. The level of BM-cell-mediated stasis of both P815 and L1210 cells in a 24-h culture has been shown (1, 2) to be directly dependent on BM–tumor cell ratio; a tumor growth inhibition close to 100% may be achieved at a 480/1 ratio (2). The cytostatic BM cells have been found to be distinct from mature T, B lymphocytes, or macrophages (1, 3) and to mediate their tumor growth-inhibitory effect in part through producing a soluble factor (or factors) (2, 3). After separation of BM cells on a discontinuous Percoll density gradient, the cells with antitumor activity are recovered predominantly in relatively low density (1.060–1.075 g/ml) fractions (1, 3). The available data, thus, indicate an apparent resemblance of 1 Abbreviations used: Ab, Antibody (antibodies); BM, bone marrow; Con, concanavalin; CM, conditioned medium; DMMA, NG-monomethyl-D-arginine; D-PBS, Dulbecco’s phosphate-buffered saline; FCS, fetal calf serum; IFN, interferon, IL, interleukin; LME, L-leucyl methyl ester; LMMA; NG-monomethyl-L-arginine; LPS, lipopolysaccharide; NS, natural suppressor; WGA, wheat germ agglutinin.

12

0008-8749/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

INTRODUCTION

11-19-97 04:32:30

cias

ANTIPROLIFERATIVE ACTIVITIES OF BONE MARROW CELLS

tumor growth inhibitory BM cells to null natural suppressor (NS) cells which are known to be also resident in normal BM (4–6). However, according the findings presented in this study, the BM cells inhibitory to leukemia growth and the BM cells suppressive to lymphocyte proliferative responses are not identical in their cell compositions, but also may differ in their mechanisms of antiproliferative action. On the other hand, the results suggest that the tumor growth-inhibitory activity of BM cells, like their NS activity, may be readily modulated by lymphokines including IFN-g. MATERIALS AND METHODS Animals. (C57BL/6 1 DBA)F1 (BDF; H-2b/H-2d) mice were bred in our own facilities. Athymic Balb/c nu//nu/ (H-2d) mice were purchased from the Laboratory of Biomedical Modelling (Tomsk Scientific Center, Tomsk, Russia). The mice were both males and females and of ages ranging between 3 and 8 months. They received autoclaved food and HCl-acidified (pH 2.8) boiled water. Culture media and reagents. RPMI 1640 medium, sodium bicarbonate, Hepes, L-glutamine, gentamycin, lipopolisacharide (LPS) from Escherichia coli (serotype B5:055), wheat germ agglutinin (WGA), N-acetyl-Dglucosamine, and NG-monomethyl-D-arginine acetate (DMMA) were purchased from Sigma Chemical Co. (St. Louis, MO). Ca2/, Mg2/-free Dulbecco’s phosphate-buffered saline (D-PBS) (pH 7.2) and fetal calf serum (FCS) were from Vector Co. (Novosibirsk, Russia). FCS was decomplemented at 587C for 50 min. 2-Mercaptoethanol, L-leucine methyl ester hydrochloride (LME), NG-monomethyl-L-arginine acetate (LMMA), N-1naphthylethylenediamine dihydrochloride, and phosphoric acid were from Fluka Chemika (Buchs, Switzerland). Concanavalin (Con) A was from Pharmacia Biotechnology Inc., (Upsala, Sweden). [3H]Thymidine was obtained from Isotop Co. (Sankt-Petersburg, Russia). Affinity-purified rabbit anti-mouse IgM/IgG antibodies (Ab) were prepared at our laboratory. Anti-Thy 1,2 monoclonal Ab (clone 5a-8, ascites, Cat. No. CL8600A) and fluorescein (FITC)-conjugated rabbit anti-mouse Ig Ab (Cat. No. CL6002F) were from Cedarlane Laboratories Limited (Ontario, Canada). Monoclonal hamster anti-murine IFN-g Ab (Cat. No. 1222-00), approximately 7 ng of which would be able to inhibit 50% of Mo-activating activity exhibited by 1 IU of mouse IFNg, were obtained from Genzyme Co. (Cambridge, MA) and stored in liquid nitrogen at 01967C until use.

13

maintained in RPMI 1640 medium supplemented with 10% FCS, 2 mM L-glutamine, and antibiotics. Treatment of BM cells with LME. BM cells were obtained from the tibias and femurs exactly as described (1). The washed BM cells were incubated at 107/ml in serum-free RPMI 1640 medium with 15 mM LME for 30 min. Following the incubation, the cells were washed twice with a cold D-PBS/10% FCS and, after determination of cell viability by a trypan blue exclusion method, were used in experiments. Fractionation of BM cells with WGA. The washed BM cells (4 1 108/ml) were incubated with 50 mg/ml WGA at room temperature for 30 min. Next, the cell suspension was gently layered onto a 50% solution of FCS and D-PBS in a 12-ml conical tube. After 30 min at room temperature the top fraction, containing the buoyant BM cells, and the bottom fraction, containing the agglutinated BM cells, were collected. The middle fraction was discarded. The agglutinated BM cells were dissociated by treatment with 0.2 M N-acetyl-D-glucosamine as described (7). This procedure allowed recovery of about 40% of the original BM cell preparation with a cell viability in excess of 95%. Approximately 70 and 30% of the recovered BM cells were WGA receptor (R)/ and WGA R0, respectively. After being washed extensively with D-PBS/1% FCS, the recovered BM cells were used in experiments. Culture conditions. Cells were cultured in RPMI 1640 medium supplemented with 10 mM Hepes, 2 mM 05 L-glutamine, 5 1 10 M mercaptoethanol, 30 mg/ml gentamycin, and 7.5% FCS at 377C in a moist atmosphere of 5% CO2/95% air.

Tumor cell lines. P815 mastocytoma and L1210 lymphoma cell lines, both of DBA (H-2d) origin, have been obtained from Moscow Oncologic Scientific Center of Russian Academy of Medical Sciences and were

Preparation of preactivated B cells. The cell suspension was prepared from normal spleen exactly as previously described (1). Following washes, the spleen cells (5 1 107) were cultured in 10 ml with LPS (20 mg/ ml) in a plastic flask (25 cm2, Linbro) for 20–24 h. After being cultured, the viable cells were separated from dead cells by their centrifuging on an isoosmotic Percoll solution (density 1.075 g/ml), washed three times, and then incubated at 4 1 106/ml in 7 ml culture medium in a plastic 90-mm petri dish at 47C for 2 h to deplete the cells capable of nonspecifically attaching to a plastic surface. Halfway through the incubation, the nonadherent cells were resuspended by gently swirling the dish for 30 s. Following incubation, the cells were washed, resuspended in RPMI 1640 medium containing 20 mM Hepes and 5% FCS, and set into a 90mm plastic dish, the surface of which was precoated with anti-mouse Ig Ab by the previously described method (1, 8). The positive selection of B lymphocytes was performed using a procedure of ‘‘panning’’ in the standard way (8). Briefly, the cells (2–3 1 107) were

AID

cias

CI 1218

/

6c2b$$$$21

11-19-97 04:32:30

14

SELEDTSOV ET AL.

incubated in the Ab-precoated 90-mm dish for 80 min at 47C. In the middle of the incubation period the nonadherent cells were gently resuspended for 30 s. Following incubation, the nonadherent cells were removed from the dish by extensively washing with D-PBS/1% FCS; the remaining adherent cells were collected with a special policeman (Costar), washed, and used further in experiments. As determined by a direct immunofluorescence assay using FITC-conjugated anti-mouse Ig Ab, nearly all of these cells were Ig-positive. Preparation of preactivated T cells. Normal spleen cells at 5 1 107/10 ml were cultured with Con A (5 mg/ ml) in a plastic flask (25 cm2, Linbro) for 20–24 h. After being cultured, the viable cells were separated from dead cells by centrifugation on Percoll solution, washed three times, and resuspended in RPMI 1640 medium/ 20 mM Hepes/5% FCS. To deplete adherent cells and B lymphocytes, the cells (2–3 1 107) were incubated in a 90-mm plastic dish precoated with anti-Ig Abs at 47C for 2 h as described above. The nonadherent cells, collected from a dish after the incubation, were washed and then placed in serum-free medium with anti-Thy 1.2 Ab. Following 45-min exposure to anti-Thy 1.2 Ab at 47C, the cells were washed and further subjected to a ‘‘panning’’ procedure using an Ab-precoated dish in the above-described way. After the removal of the nonadherent cells, the adherent cells were collected with a policeman, washed, and used further in experiments. As determined by an indirect immunofluorescence assay with anti-Thy 1.2 Ab and FITC-conjugated antiIg Abs, close to 100% of these cells were Thy 1.2-positive. Preparation of media conditioned with preactivated lymphocytes. The preactivated lymphocytes (B or T) at 3–5 1 106/ml were cultured in a 24-well plate (Linbro) at a volume of 1.5 ml/well for 24 h, after which the culture supernatants were harvested, pooled, filter sterilized using a 0.22 mm filter unit (Cole-Parmer Instrument Co., Chicago, IL), and stored at 0207C until use. As established by functional tests, the B cell-conditioned medium (CM) at a 25% concentration was effective in promoting the LPS-driven proliferation of splenocytes placed at a suboptimal density (4 1 104/well) in a 96-well flat-bottom culture plate (Linbro), whereas the 25% T-cell-CM was able in the presence of LPS (20 ng/ml) to induce strong macrophage activation for producing nitric oxide (NO) (data not shown).

cated, the soluble additions (4 mg/ml anti-IFN-g Ab, 0.5 mM LMMA or DMMA) were introduced in the cultures at the time plating. In order to evaluate a significance of cell–cell contact in cytostatic process, BM and tumor cells (the ratios are indicated below) were spun down by centrifugation on conic, round, and flat well bottoms of 96-well plates (Linbro) and incubated in this position in the completed culture medium at 377C for 7 h. To evaluate the effects of lymphocyte-soluble products on BM-derived cytostatic activity, BM cells were cultured in a 96-well round-bottom plate (BDSL) at 3 1 105/well in the absence or presence of T-cell- or B-cell-CM (25%), without or with anti-IFN-g Ab (4 mg/ml), for 20 h and further (after replacement of the medium in wells) recultured with tumor cells (104/well) in medium alone for additional 24 h. In control cultures tumor cells were grown alone or in the presence of normal thymocytes that have been previously found (1) to be unable to suppress leukemia cell growth in vitro. The controls in NS activity assay included spleen cells stimulated with mitogens in the absence of any cell addition or in the presence of irradiated (20 Gy) spleen cells or normal thymocytes. In our experimental system no suppression of cell proliferation in control cultures was observed. DNA synthesis was measured by pulsing individual wells with 0.75 mCi of [3H]thymidine during the last 3–5 h of culture. The cells were transferred into a fiberglass filter with a harvesting apparatus (Flow Lab.) and radioactivity was determined by liquid scintillation counting. The percentage suppression of cell proliferation was calculated as follows:

S

10

D

cpm test 1 100. cpm control

Measurement of NO. NO produced in the cultures was measured by nitrite accumulation in the culture supernatant as described (9). Briefly, 50 ml of each supernatant was transferred to a 96-well microassay plate (Greiner) and incubated with 50 ml of Griess reagent (1% sulfanilamide/0.1% N-1-naphthylethylenediamine dihydrochloride/2.5% H3PO4) at room temperature for 10 min. Absorbance at 540 nm was determined with a microplate reader. Nitrite concentration was calculated from a sodium nitrite standard curve. Statistics. Experiments were performed at least three times and representative data (mean of triplicate cultures / SD) are presented. Student’s t test was used to determine the significance of the differences between experimental values. The differences pointed in the following text were statistically significant (P õ 0.05).

Antiproliferative activity assays. In the experiments in which both tumor growth-inhibitory and NS cell activity was simultaneously assayed, BM cells were cultured for 48 h either with tumor cells (104/well) or with syngeneic spleen cells (3 1 105/well) plus Con A (5 mg/ml) or LPS (12 mg/ml) in a 96-well round-bottom culture plate (BDSL, Ayrshire, England). Where indi-

Characteristics of the antiproliferative activities exhibited by BM cells freshly isolated from normal BDF1

AID

cias

CI 1218

/

6c2b$$$$22

11-19-97 04:32:30

RESULTS

ANTIPROLIFERATIVE ACTIVITIES OF BONE MARROW CELLS

FIG. 1. Antiproliferative activities of unmanipulated, LMEtreated, and WGA-fractioned BDF1 BM cells. The indicated BM cells (3 1 105/well) were cultured for 48 h with P815 or L1210 cells (104/ well) or with syngeneic spleen cells (3 1 105/well) plus Con A or LPS. Percentage suppression was calculated relative to cell proliferation in the corresponding control cultures with no BM cell additions. The control [3H]thymidine incorporation in P815 cells, L1210 cells, and Con A- and LPS-induced lymphoblasts was 210,000, 350,000, 132,000, and 105,000 cpm, respectively.

mice. In the experiments described previously (1–3) the BM cells capable of reducing nonspecifically leukemic cell growth in vitro have demonstrated an apparent resemblance in their phenotypic characteristics to null NS cells. This prompted us to further investigate the relationship between tumor growth inhibitory BM cells and NS BM cells. It has been previously established that the treatment of BM cells with LME abrogates their NS activity (6, 10) and that NS BM cells can be selected positively by agglutination with WGA (6, 11). In fact, the results of our own experiments depicted in Fig. 1 indicate that, in contrast to the unmanipulated BM cells, which exhibited both tumor growth inhibitory and NS activity, the LME-treated BM cells failed to reduce the growth of both P815 and L1210 cells, as well as to suppress mitogen (Con A or LPS)-driven spleen cell proliferation. Depleting BM cells of WGA R/ cells, however, failed to diminish their antitumor activity, while decreasing dramatically their NS activity. Moreover, in three of five experiments performed with a similar pattern, the WGA R0 cells were detectably more active in inhibiting tumor cell growth than were the WGA R/ cells. Thus, the results suggest that despite a certain phenotypic similarity, the tumor growth inhibitory BM cells and NS BM cells are not be identical in their cell compositions. A strong dependency of NS-cell-mediated immunosuppressive activity on the presence of IFN-g in the medium has been previously reported in different NS systems (5, 6, 12). The available data indicate that

AID

CI 1218

/

6c2b$$$$23

11-19-97 04:32:30

15

IFN-g is able to induce the production in BM-derived NS cells of NO (6) that, in turn, can be responsible for suppressing lymphocyte proliferation (6, 13). Figure 2 shows that neutralizing anti-IFN-g Ab added into the cultures at the time of plating failed to reverse BMcell-mediated tumor growth inhibition. This inhibition also was not affected by LMMA, a competitive inhibitor of arginine-dependent NO synthase. Consistent with this finding, we were unable to detect any noticeable NO production in BM-tumor cell cocultures (data not shown). On the other hand, anti-IFN-g Ab, as well as LMMA, but not DMMA, an inactive enantiomer of LMMA, was found to be able to decrease dramatically both the BM-cell-derived NS activity directed against Con-driven spleen cell proliferation (Fig. 2) and the nitrite accumulation in the mitogen-stimulated BM– spleen cell cocultures (Fig. 3), It is noteworthy, however, that the BM-cell-derived suppression of LPS response was affected by neither anti-IFN-g Ab nor LMMA, indicating a marginal role for the NO-mediated mechanism in this suppression. Thus, the results suggest that BM cells may utilize more than one mechanism to reduce cell proliferation. A role of cell-to-cell contacts in BM-cell-mediated tumor growth suppression. In order to evaluate a significance of cell-to-cell contact in BM-cell-caused leukemia growth inhibition, BM cells were incubated at various ratios with P815 or L1210 on conic, round, and flat

FIG. 2. Antiproliferative activities of BDF1 BM cells in the presence of anti-IFN-g Ab, LMMA, or DMMA. BM cells (3 1 105/well) were cultured for 48 h with P815 or L1210 cells (104/well) or with syngeneic spleen cells (3 1 105/well) plus Con A or LPS, in the absence or presence of anti-IFN-g Ab (4 mg/ml), LMMA (0.5 mM), or DMMA (0.5 mM). The control proliferation of P815 cells, L1210 cells, and Con A- and LPS-stimulated spleen cells was 184,000, 191,000, 98,000, and 60,000 cpm, respectively. Each of the soluble additions used, by itself, lacked any significant effect on cell proliferation in the control cultures.

cias

16

SELEDTSOV ET AL.

Thus, the results suggest that by producing soluble mediators (including IFN-g), the activated lymphocytes may potentiate BM-cell-derived tumor growthinhibitory activity. It should be noticed that, as it was determined by 24 h cytolytic assay using [3H]thymidine-prelabeled L1210 cells as targets, both BM cells precultured in the medium alone and those preexposed to lymphokines lacked any significant tumoricidal activity, suggesting, thereby, no generation of cytolytically active cells in our experimental system (data not shown). From the results suggesting the existence of immune-mediated upregulation of BM-cell-derived tumor growth-inhibitory activity, it might be proposed that antitumor activity exhibited by the BM cells isolated freshly from normal mice could be a consequence of TFIG. 3. Nitrite production in mitogen-stimulated BM–spleen cell cocultures. BDF1 BM cells (3 1 105/well) were cocultured for 48 h with syngeneic spleen cells (3 1 105/well) plus Con A or LPS, in the absence or presence of anti-IFN-g Ab (4 mg/ml), LMMA (0.5 mM), or DMMA (0.5 mM), after which nitrite accumulation in culture supernatants was measured. Nitrite production by BM and spleen cells cultured separately was õ3 and õ8 mM, respectively.

well bottoms for 7 h. As can be seen in Fig. 4, BM cells at a number optimal for generating cytostasis provided equal tumor growth suppression in all wells regardless of the shape of their bottom. A strong correlation of suppression levels with cell densities in cultures was, however, observed in the samples with suboptimal numbers of BM cells. In these cases, the most suppression was found in conic-bottom wells, providing highest levels of cell-to-cell contacts. Similar results were obtained when BM cells were incubated with tumor cells for 24 h (data not shown). It is important to emphasize that, in contrast to BM cells, thymocytes failed to suppress tumor growth regardless of the level of their contact with tumor cells (data not shown). Thus, the results suggest the involvement of contact cell-to-cell interaction in the process of BM-cell-mediated tumor growth inhibition. Lymphokine-mediated upregulation of BM-cell-derived leukemia growth inhibitory activity. We have previously reported that the murine T lymphocyte activated in allogeneic mixed lymphocyte culture (MLC) were capable of producing a soluble mediator(s) able to enhance rapidly (for 20 h of culture) BM-cell-derived cytostatic activity assayed against P815 or L1210 cells (2). In fact, the results shown in Fig. 4 indicated that not only T-cell-CM, but also B-cell-CM may contain products potentiating tumor growth inhibitory ability of BM cells. Importantly, anti-IFN-g Ab was found to be capable of partly reversing the additional BM-cellmediated tumor growth inhibition induced with T-cellCM, while not affecting that induced with B-cell-CM.

AID

CI 1218

/

6c2b$$$$23

11-19-97 04:32:30

FIG. 4. A significance of cell-to-cell contact in BM-cell-mediated tumor growth suppression. After being spun down by centrifugation, BDF1 BM cells and tumor (P815 or L1210) cells were incubated together at the indicated ratios in wells having a conic, round, or flat bottom for 7 h. The number of tumor cells in a well was 104. The control P815 and L1210 cell proliferation (with thymocytes) in all cases was close to 70,000 and 87,000 cpm, respectively.

cias

ANTIPROLIFERATIVE ACTIVITIES OF BONE MARROW CELLS

FIG. 5. Lymphokine-induced enhancement of BM-cell-derived cytostatic activity. BDF1 BM cells (3 1 105/well) were cultured without or with T-cell- or B-cell-CM (25%), in the absence or presence of anti-IFN-g Ab (4 mg/ml) for 20 h and further recultured with P815 or L1210 cells (104/well) in the medium alone for an additional 24 h. The control P815 and L1210 cell proliferation was 172,000 and 182,000 cpm, respectively.

cell-mediated or -dependent unapparent immune reactions occurring normally in the body. However, the results shown in Fig. 5 indicate that the BM cells from athymic BALB/c possessed a significant natural tumor growth-inhibitory activity. Moreover, as can be seen, the nude BM cells precultured with T-cell-CM were detectably more active in suppressing both P815 and L1210 cell growth than the nude BM cells precultured in the medium alone (Fig. 6). Similar results were obtained in experiments with the BM cells from euthymic BALB/c mice (data not shown).

17

anism, depending on the type of proliferating target cells. Indeed, the IFN-g-dependent, NO-mediated mechanism, which is involved in BM-mediated suppression of ConA-driven T-cell blastogenesis, was not found to be responsible for BM-cell-mediated suppression of both tumor growth and LPS-driven B-cell blastogenesis. It should be noted that these results in their part contrast with the previous observation by Angulo et al., demonstrating an important role for NO as a mediator in BM-derived suppression of not only Con A-driven, but also LPS-driven lymphoblastogenesis (6). The arisen inconsistence might be presumably explained by differences in the experimental systems used; Angulo et al. have examined the NS activity exhibited by the low-density BM cells, whereas we dealt with the unfractionated BM cells. Since BM cells can utilize more than one mechanism to suppress cell proliferation, it appears that the significance of NO in mediating BM-cell-derived suppression of lymphoblastogenesis might vary from critical to minimal, greatly depending on features of experimental systems used. Soluble suppressive molecules are believed to be of critical importance for mediating BM-cell-derived NS activity. We (2) and others (3) have previously reported that BM cells can also elaborate a soluble factor (or factors) capable of reducing in a nonspecific manner the proliferation of different leukemic cells. The published data raise the possibility of involvement in BM cellderived NS activity of prostaglandins (14–16), as well as TGF-b (13, 17–19). Each of these mediators is known to be characterized by a broad antiproliferative spectrum of action. However, in our experimental system, the blockade of prostaglandin synthesis with indo-

DISCUSSION Previously, by examining effects of freshly isolated normal BM, spleen, thymus, and lymph node cells on in vitro growth of P815 and L1210 cells, we have indicated that BM cells have the most leukemia growth-inhibitory potential (1). The data presented herein suggest that although the BM cell population responsible for suppressing tumor growth in vitro and that responsible for NS activity share a range of common characteristics (including their sensitivity to LME treatment), their cell compositions do not appear to be identical. Indeed, we found that depleting BM cells of WGA R/ cells may have led to a dramatical decrease in their NS activity, while not diminishing their tumor growth-inhibitory ability. It is of principal importance that BM cells may utilize predominantly that or another antiproliferative mech-

AID

CI 1218

/

6c2b$$$$23

11-19-97 04:32:30

FIG. 6. Tumor growth inhibitory activity of BM cells from athymic BALB/c mice. Nude BM cells (3 1 105/well) were cultured without or with 25% T-cell-CM for 20 h and further recultured with P815 or L1210 cells (104/well) in the medium alone for additional 24 h. The control P815 and L1210 cell proliferation was 152,000 and 165,000 cpm, respectively.

cias

18

SELEDTSOV ET AL.

methacin (1006 M, Sigma), as well as neutralizing of TGF-b activity with specific anti-TGF-b Ab (20 mg/ml, Genzyme) lacked any noticeable effect on BM-cellmediated leukemia growth inhibition, suggesting, thereby, no significant role for both prostaglandins and TGF-b in this process (data not shown). The identification of BM-cell-derived soluble molecules inhibitory to leukemia growth is one of the goals for our further studies. It is of great importance, however, that according to the data presented in this study, apart from soluble mediators, contact cell-to-cell interactions may play a significant role in BM-cell-mediated tumor growth suppression. This notion agrees with our own previous observation, indicating that under equal culture conditions, BM cells may be significantly more effective in suppressing leukemia growth, compared with their culture supernatants (2). In our view, the data concerning contact cytostasis could partially explain the abrogating effect of LME on BM-cell-derived cytostatic activity. Indeed, a normally functioning lisosoma apparatus is well known to be a critically important for the mechanism providing membrane expression of a whole range of protein molecules. It seems, therefore, quite possible that by disrupting lysosomas LME might impair the expression on BM cells of the structures responsible for contact inhibition of tumor growth. The soluble mediators produced by lymphocytes in response to an activating stimulus are believed to be extremely important for generating NS cell activity. In addition to IFN-g, interleukin (IL)-2 (12), IL 3 (5, 20), IL 4, IL 6 (5), and granulocyte–macrophage colonystimulating factor (CSF-GM) (20) have been previously reported to be capable of favoring NS cell activity development. The results presented herein suggest that IFN-g, as well as T-cell products other than IFN-g, may be also involved in up-regulating BM-cell-derived tumor growth-inhibitory activity. Consistent with this finding. We have previously indicated that recombinant human IL-2 is capable of potentiating an ability of BM cells to suppress leukemia growth in vitro (2). Interestingly, in our experiments not only the activated T, but also B cells were found to be able to secret a soluble mediator (or mediators) rendering BM cells more inhibitory to tumor growth. This B-cell mediator(s) did not appear to be IFN-g. At present its nature is still unclear and needs to be identified. It should be noted that in four of nine experiments performed with a similar pattern a somewhat similar increase in tumor growth-inhibitory activity was demonstrated by the BM cells precultured with the medium conditioned with LPS-activated macrophages. However, that increase was not as significant as that induced by T-cellCM or B-cell-CM (data not shown). The effector mechanism underlying lymphokine-induced additional cytostatic activity of BM cells is still

uncertain. We found that T-cell-CM, but not B-cell-CM, was able to induce NO production in BM cells that was significantly augmented in the presence of LPS and was largely abrogated by the addition into the medium of anti-IFN-g Ab. However, the blockade of NO production by LMMA failed to affect not only the B-cellCM-stimulated but also T-cell-CM-stimulated tumor growth-inhibitory activity of BM cells, suggesting, thereby, no significant role of NO in mediating the antiproliferative action of lymphokine-activated cytostatic effectors (data not shown). The data presented in this study indicate that the antitumor activity pattern demonstrated by the BM cells from athymic mice was similar to that exhibited by the BM cells from euthymic mice. Based on this finding, it may be suggested that BM-cell-derived natural tumor growth-inhibitory activity, at least in its significant part, may not be a consequence of thymusdependent immune processes normally occurring in the body. The interesting example of effective cooperation of BM-cell- and T-cell-mediated mechanisms in establishing antitumor immunity has been recently demonstrated by Khazaie et al. (21) who particularly revealed that in the DBA mice vaccinated with nontumorigenic doses of L5178 lymphoma cells a fraction of these cells persisted in BM in a dormant state for long periods of time and provided antigenic stimulation leading to the generation of effective T-cell-mediated antitumor immunity. Of importance, persistence of dormant tumor in BM correlated with the duration of antitumor immunity. In view of the data presented herein, there may be speculated that the T-cell production of factors potentiating BM-cell-derived cytostatic activity might play a significant role in preventing leukemia development in that experimental system. Donor T lymphocytes responding to antigens expressed on recipient cells are known to be largely responsible for inducing a long-lasting graft-versus-leukemia effect in an allogeneic BM-grafted recipient (22, 23). From the data presented in this study it seems reasonable to hypothesize that the T cells activated in such a recipient, through producing IFN-g, IL 2, and possibly other soluble molecules, might promote BMmediated tumor cell stasis, whereas the dormant tumor cells, in turn, might provide a threshold antigenic stimulation necessary for inducing tumor-specific T-cell reactions which could ultimately be responsible for final elimination of malignancy from the body. In conclusion, the results of this study suggest that antitumor and NS activities exhibited by BM cells may be based on substantially distinct antiproliferative mechanisms. This, together with the data suggesting the existence of immune-mediated up-regulation of BM cell cytostatic activity, prompts us now to seek for new immunologic approaches to leukemia treatment which

AID

cias

CI 1218

/

6c2b$$$$24

11-19-97 04:32:30

ANTIPROLIFERATIVE ACTIVITIES OF BONE MARROW CELLS

would promote simultaneously both tumor cell stasis and the establishment of effective long-term T-cell-mediated antitumor immunity in the body. ACKNOWLEDGMENT The authors thank E. A. Kashenko for excellent technical support.

REFERENCES 1. Seledtsov, V. I., Avdeev, I. V., Morenkov, A. V., Seledtsova, G. V., and Kozlov, V. A., Immunobiology 192, 205, 1995. 2. Seledtsov, V. I., Avdeev, I. V., Seledtsova, G. V., Samarin, D. M., Prokopenko, V. A., and Kozlov, V. A., Biomed. Pharmacother. 49, 293, 1995. 3. Sugiura, K., Inaba, M., Ogata, H., Yasumuzu, R., Sardina, E. E., Inaba, K., Kuma, S., Good, R. A., and Ikehara, S., Cancer Res. 50, 2582, 1990. 4. Dorshkind, K., and Rosse, C., Am. J. Anat. 164, 1, 1982. 5. Holda, J. H., Maier, T., and Claman, H. N., Cell. Immunol. 125, 459, 1990. 6. Angulo, I., Rodriguez, R., Garcia, B., Medina, M., Navarro, J., and Subiza, J. L., J. Immunol. 155, 15, 1995. 7. Brooks-Kaiser, J. C., Murgita, R. A., and Hoskin, D. W., J. Reprod. Immunol. 21, 103, 1992. 8. Mage, M. G., McHugh, L. L., and Rothstein, Th. L., J. Immunol. Methods 15, 47, 1977.

AID

CI 1218

/

6c2b$$$$25

11-19-97 04:32:30

19

9. Ding, A. H., Nathan, C. F., and Stuehr, D. J., J. Immunol. 141, 2407, 1988. 10. Moore, S. C., Theus, S. A., Barnett, J. B., and Soderberg, L. S. F., Exp. Hematol. 20, 1178, 1992. 11. Sugiura, K., Inaba, M., Ogata, H., Yasumizu, R., Inaba, K., Good, R. A., and Ikehara, S., Proc. Natl. Acad. Sci. USA 85, 4824, 1988. 12. Holda, J. H., Maier, T., and Claman, H. N., J. Immunol. 137, 2511, 1986. 13. Young, M. R., Wright, M. A., Matthews, J. P., Malik, I., and Prechel, M., J. Immunol. 156, 1916, 1996. 14. Choi, K. L., Maier, T., Holda, J. H., and Claman, H. N., Cell. Immunol. 112, 271, 1988. 15. Howell, C. D., Yoder, T. Y., and Vierling, J. M., Cell. Immunol. 140, 54, 1992. 16. Brooks, J. C., and Hoskin, D. W., Transplantation 58, 1096, 1994. 17. Hoskin, D. W., Bowser, D. A., and Brooks-Kaiser, J. C., J. Leukocyte Biol. 51, 649, 1992. 18. Moore, S. C., Shaw, M. A., and Soderberg, L. S. F., J. Leukocyte Biol. 52, 596, 1992. 19. Hooper, W. C., Leuk. Res. 15, 179, 1991. 20. Moore, S. C., Theus, S. A., Barnett, J. B., and Soderberg, L. S. F., Cell. Immunol. 141, 398, 1992. 21. Khazaie, K., Prifti, S., Beckhove, P., Griesbach, A., Russell, S., Collins, M., and Schirrmacher, V., Proc. Natl. Acad. Sci. USA 91, 7430, 1994. 22. Aizawa, S., and Sado, T., Transplantation 52, 885, 1991. 23. Butturini, A., and Gale, R. P., Immunol. Res. 11, 24, 1992.

cias