Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

LPS is a component of the cell wall of Gram-negative bacteria that exhibits a variety of biological activities, including both toxicity and immunopotentiating activity (Homma et al., 1985; Kanegasaki et al., 1986). The lipid moiety, lipid A, is an active part of LPS. Lipid A and its analogs exhibit various immunopharmacological activities, i.e., activation of macrophages (Matsuura et al., 1995), immunoadjuvant activity (Takada et al., 1985; Ulrich and Myers, 1995), B cell mitogenicity (Kotani et al., 1983), induction of resistance to bacterial infection (Schutze et al., 1994) and antitumor activity (Nakatsuka et al., 1989). ONO-4007, sodium 2-deoxy-2-[3S-(9-phenylnonanoyloxy)tetradecanoyl]-amino-3-O-(9-phenylnonanoyl)-D-glucopyranose 4-sulfate (fig. 1), is a monosaccharide lipid A analog with low toxicity. ONO-4007 exhibits strong antitumor activity via intratumoral TNF production and induces macrophages to a tumorcidal state in MM46 mammary carcinoma-bearing mice (Yang et al., 1994). ONO-4007 also induces high levels of TNF within tumor tissue in KDH-8 hepatoma-bearing rats (Kuramitsu et al., 1997), and intratumoral injection of anti-TNF- antibody attenuates the antitumor activity of ONO-4007 (N. Matsumoto and A. Akimoto, unpublished observation). These findings suggest that the antitumor effect of ONO-4007 is mediated by endogenous TNF- production from tumor-infiltrating macrophages. TNF-inducing activity of ONO-4007 in human monocytes/macrophages, however, has not been examined.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Chemical structure of ONO-4007.

Several analogs of lipid A have been shown to display species-specific pharmacological activities. For example, GLA-60, a monosaccharide lipid A analog, activates both human monocytes (Maeda et al., 1990) and murine macrophages (Matsuura et al., 1995), whereas lipid IV_A, a disaccharide precursor of lipid A, inhibits LPS-induced activation of human monocytes although it activates murine macrophages (Golenbock et al., 1991; Delude et al., 1995), and Rhodobacter sphaeroides lipid A inhibits LPS-induced activation of both human monocytes and murine macrophages (Golenbock et al., 1991; Delude et al., 1995). In contrast to its effects in mice and rats, ONO-4007 has little or no ability to induce TNF production when injected into cynomolgus monkeys, dogs or rabbits (N. Matsumoto and A. Akimoto, unpublished observation), which suggests that the pharmacological features of ONO-4007 are species-specific.

GM-CSF and M-CSF enhance the survival (Erickson-Miller et al., 1990) and differentiation (Geissler et al., 1989; Dimri et al., 1994; Hashimoto et al., 1997) of cultured human monocytes. Moreover, pretreatment of human monocytes with GM-CSF (Cannistra et al., 1988) or M-CSF (Sampson-Johannes and Carlino, 1988; Asakura et al., 1996) enhances TNF- production in response to subsequent LPS stimulation, a result that indicates functional priming activity of GM-CSF and M-CSF on human monocytes.

In this study, to elucidate cytokine-inducing activity of ONO-4007 in human monocytes/macrophages, we used freshly isolated human monocytes and monocytes functionally primed by incubation with GM-CSF or M-CSF and investigated the induction of cytokines (TNF-, IL-1, IL-6 and IL-12) in response to ONO-4007. In comparison with ONO-4007, LPS derived from Salmonella abortus equi was used as a reference agent, because clinical trials in cancer patients have been reported (Engelhardt et al., 1991 and 1995).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. ONO-4007, synthesized at Ono Pharmaceutical Co., Ltd (Osaka, Japan), was dissolved in dimethyl sulfoxide and further diluted with culture medium for monocyte activation or added to whole blood (final dimethyl sulfoxide concentration, 0.1% and 0.5%, respectively). LPS from Salmonella abortus equi was obtained from Sigma (St. Louis, MO), LA-15-PP (Escherichia coli-type synthetic lipid A) from Daiichi Pure Chemicals (Tokyo, Japan), rhGM-CSF (2.05 × 10⁸ U/mg) from Genzyme (Cambridge, MA), rhM-CSF (6.5 × 10⁷ U/mg) from Cellular Products (Buffalo, NY) and monoclonal anti-CD14 (RMO52) from Cosmo Bio (Tokyo, Japan). The culture medium consisted of RPMI-1640 (Nissui Pharmaceutical, Tokyo, Japan) supplemented with 0.3 g/l glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin and 10% FCS (Gibco-BRL, Grand Island, NY).

Monocyte separation. Peripheral blood mononuclear cells were obtained from heparinized venous blood of healthy adults by density centrifugation with Lymphocyte Separation Medium (Organon Teknika, Durham, NC). Mononuclear cells suspended in culture medium were incubated in autologous serum-coated dishes at 37°C for 60 min. Then nonadherent cells were removed by three washes with prewarmed culture medium, and adherent cells were collected by treatment with 1 mM EDTA. The adherent cells were found to be 93% monocytes by Wright stain.

Cell culture. All cell cultures were performed at 37°C in 5% CO₂/95% air and 100% humidity. Monocytes, freshly isolated or preincubated for 3 days in the presence of GM-CSF or M-CSF in 24-well culture plates, were washed and stimulated with ONO-4007 or LPS for 3, 6 or 24 h. Heparinized whole blood of healthy adults was stimulated with ONO-4007, LA-15-PP or LPS for 6 h. Supernatants were obtained by centrifugation and stored below 20°C until measurement of cytokines. To assess the requirement of serum in TNF- production, we compared the effect of FCS-free medium with the effect of culture medium containing 10% FCS when GM-CSF-treated monocytes were stimulated with ONO-4007 or LPS. To assess the effect of anti-CD14 on TNF- production in GM-CSF-treated monocytes, we added anti-CD14 antibody (10 µg/ml) to cultures 15 min before stimulation.

Cytokine determination by ELISA. Concentrations of TNF-, IL-1, IL-6 and IL-12 (p70) in the culture supernatants were assayed with commercially available ELISA kits (R&D Systems, Minneapolis, MN) as directed by the manufacturer. The detection limits of these assays were 4.4 pg/ml for TNF-, 1.2 pg/ml for IL-1, 2.8 pg/ml for IL-6 and 5.0 pg/ml for IL-12.

TNF bioassay. TNF activities in supernatant samples were measured by L929 cell cytotoxicity in the presence of actinomycin D (Ruff and Gifford, 1980; Flick and Gifford, 1984). L929 cells (3 × 10⁴ cells/well) in 96-well culture plates were incubated with serially diluted samples in the presence of 1 µg/ml actinomycin D for 19 h. Cell viability was assessed by incubating the cells with 0.5 mg/ml MTT for another 5 h. Then the MTT-formazan produced by viable cells was dissolved in ethanol, and optical density (570-690 nm) was read (Denizot and Lang, 1986). An international standard rhTNF- (1st International Standard for TNF- [Human rDNA], National Institute for Biological Standards and Control, Hertfordshire, UK) was used as an assay calibrant. TNF activities in culture supernatants were expressed as I.U./ml.

Statistics. All data are presented as means ± S.E. Differences in cytokine production for freshly isolated and GM-CSF-treated monocytes were analyzed by two-way analysis of variance (ANOVA). Priming effects of GM-CSF or M-CSF on TNF- production were analyzed by one-way ANOVA followed by Dunnett's test. The effect of anti-CD14 on TNF- production was analyzed by paired t test. Values less than .05 were considered statistically significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Induction of TNF-, IL-1, IL-6 and IL-12 by ONO-4007 or LPS in monocytes freshly isolated or incubated with GM-CSF. First, we compared the cytokine-inducing activity of ONO-4007 (100 µg/ml) in monocytes freshly isolated or incubated for 3 days in the presence of GM-CSF (100 U/ml) with that of LPS (0.1 µg/ml). The results indicate that cytokine induction by ONO-4007 in fresh monocytes was very weak; the levels of IL-1 and IL-12 production were marginal or undetectable, and the peak levels of TNF- (6 h) and IL-6 (24 h) production were 2.4% and 1.8% of the peak levels in monocytes stimulated with LPS, respectively (table 1). In contrast to the response to ONO-4007, fresh monocytes produced significant amounts of these cytokines in response to LPS. After preincubation with GM-CSF, monocyte production of TNF- in response to ONO-4007 increased dramatically to 12 times that in fresh monocytes (820 ± 180 vs. 69 ± 11 pg/ml). However, GM-CSF-treatment did not affect IL-6 production in response to ONO-4007, and it made IL-1 and IL-12 production almost undetectable. The level of TNF- production in response to LPS was increased 2.2-fold by GM-CSF-treatment, whereas IL-6 production and IL-12 production were not affected by the treatment, and IL-1 production was dramatically attenuated.

Dose-dependent effect of ONO-4007 and LPS on TNF- production. Then we examined the effect of various concentrations of ONO-4007 and LPS on TNF- production in both freshly isolated and GM-CSF-treated monocytes. In fresh monocytes, slight production of TNF- was observed at 100 and 300 µg/ml of ONO-4007 (fig. 2). In GM-CSF-treated monocytes, however, ONO-4007 induced significant amounts of TNF- production in a dose-dependent manner (14 ± 3 pg/ml at 1 µg/ml and 1577 ± 314 pg/ml at 300 µg/ml). At 300 µg/ml of ONO-4007, the levels of TNF- production in GM-CSF-treated monocytes were 16 times greater than those in fresh monocytes. LPS-induced TNF- production in GM-CSF-treated monocytes was also greater than that in fresh monocytes, but it represented only a 2.0- to 2.6-fold increase.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effect of various concentrations of ONO-4007 and LPS on TNF- production. Monocytes (5 × 104 cells/ml) freshly isolated or incubated for 3 days in the presence of GM-CSF (100 U/ml) were washed and stimulated with the indicated concentrations of ONO-4007 or LPS for 6 h (n = 4). The concentrations of TNF- in culture supernatants were measured by ELISA. Two-way ANOVA confirmed that GM-CSF-treatment increased both ONO-4007-induced and LPS-induced TNF- production.

Priming effects of GM-CSF and M-CSF on ONO-4007- or LPS-induced TNF- production. Because both M-CSF and GM-CSF are known to be priming agents for monocyte/macrophage activation, we examined the dose-dependent effect of GM-CSF and M-CSF on monocyte TNF- production. Monocytes were incubated for 3 days in the presence of 1 to 1000 U/ml of either GM-CSF or M-CSF and then we stimulated with ONO-4007 (300 µg/ml) or LPS (1 µg/ml). Preincubation with GM-CSF augmented monocyte TNF- production at 10 U/ml and higher concentrations when the monocytes were stimulated with ONO-4007 and at 1 U/ml and higher concentrations when they were stimulated with LPS (fig. 3). Preincubation with M-CSF at 1000 U/ml also augmented both ONO-4007- and LPS-induced TNF- production. Next we examined the effects of GM-CSF and M-CSF on monocyte survival. Monocytes were incubated for 3 days with GM-CSF (100 U/ml) or M-CSF (1000 U/ml), and the recovered cells were counted. The survival rate of monocytes incubated with culture medium alone was 27.5%, whereas the treatment with GM-CSF and treatment with M-CSF improved the survival to 64.9% and 53.9%, respectively (table 2).

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Priming effects of GM-CSF and M-CSF on ONO-4007- or LPS-induced TNF- production. Monocytes (5 × 104 cells/ml) incubated for 3 days with the indicated concentrations of GM-CSF or M-CSF were washed and stimulated with ONO-4007 (300 µg/ml) or LPS (1 µg/ml) for 6 h (n = 3). The concentrations of TNF- in culture supernatants were measured by ELISA. * P < .05, ** P < .01 vs. medium alone.

ONO-4007-, lipid A- or LPS-induced TNF- production in whole blood. To examine the effect of the possible interaction of monocytes with other leukocytes or serum factors on TNF- production, we examined ONO-4007-, LA-15-PP (synthetic lipid A)- and LPS-induced TNF- production in whole blood. TNF- production in response to ONO-4007 (10, 30, 100 and 300 µg/ml) was marginal (13.2 ± 2.0 pg/ml at 300 µg/ml) (fig. 4). In contrast to ONO-4007, LA-15-PP and LPS induced significant amounts of TNF- production at 100 pg/ml and higher concentrations; the levels of TNF- production in response to 1 µg/ml of LA-15-PP and LPS were 15423 ± 1655 and 25014 ± 5066 pg/ml, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   ONO-4007-, LA-15-PP- or LPS-induced TNF- production in whole blood. Whole blood was stimulated with the indicated concentrations of the test compounds for 6 h (n = 3). The concentrations of TNF- in supernatants were measured by ELISA.

Requirement of FCS for ONO-4007- or LPS-induced TNF production. We next examined the requirement of FCS in culture medium for the induction of TNF activity in freshly isolated and GM-CSF-treated monocytes. When serum-free medium was used for monocyte activation, ONO-4007-induced TNF production by fresh monocytes was attenuated, but that by GM-CSF-treated monocytes was observed from 1 µg/ml of ONO-4007 and to a maximum at 10 µg/ml, at which concentration the levels of TNF production were 9 times greater than those stimulated in serum-containing medium (fig. 5). TNF production at 30 and 100 µg/ml of ONO-4007, however, declined when GM-CSF-treated monocytes were stimulated in serum-free medium. We also observed cellular swelling and decreased number of cultured monocytes when the monocytes were stimulated with 30 and 100 µg/ml of ONO-4007 in serum-free medium, which suggests cytotoxicity of ONO-4007 in these culture conditions. The levels of LPS-induced TNF production by both fresh and GM-CSF-treated monocytes were attenuated when serum-free medium was used for monocyte activation.

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Requirement of FCS for ONO-4007- or LPS-induced TNF production. Monocytes (5 × 104 cells/ml) freshly isolated () or incubated for 3 days with 100 U/ml GM-CSF () were washed and stimulated with ONO-4007 or LPS in 10% FCS-containing or serum-free culture medium for 6 h (n = 4). TNF activities in culture supernatants were measured by L929 cell cytotoxicity. The detection limit of the assay was 1.3 I.U./ml. Two-way ANOVA confirmed that GM-CSF treatment increased both ONO-4007- and LPS-induced TNF- production when FCS-containing medium was used for monocyte activation and that it increased ONO-4007-induced TNF- production when serum-free medium was used.

Effect of anti-CD14 on ONO-4007- or LPS-induced TNF- production. Finally, we examined the effect of anti-CD14 (RMO52) on TNF- production in GM-CSF-treated monocytes. Anti-CD14 treatment completely suppressed TNF- production in response to 0.1 ng/ml of LPS, but not in response to 30 or 300 µg/ml of ONO-4007 or to 1 µg/ml of LPS (table 3).

View this table: [in this window] [in a new window]   TABLE 3 Effect of anti-CD14 antibody on TNF- production

Correlation of TNF activity with TNF- ELISA. TNF activities in monocyte supernatant samples shown in tables 1 and 3 and figures 2 and 3 were assessed by L929 cell cytotoxicity as well as by ELISA. TNF activities measured by L929 cell cytotoxicity correlated well with TNF- concentrations measured by ELISA (fig. 6).

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Correlation of TNF activity with TNF- ELISA.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This study investigated the cytokine-inducing effect of ONO-4007, a synthetic lipid A analog, on human monocytes/macrophages. ONO-4007 clearly induced TNF- production from GM-CSF- or M-CSF-treated monocytes, but not from fresh monocytes or whole blood. The present observations are consistent with reports that pretreatment of human monocytes with GM-CSF (Cannistra et al., 1988) or M-CSF (Sampson-Johannes and Carlino, 1988; Asakura et al., 1996) enhances LPS-induced TNF- production. Priming effects of GM-CSF or M-CSF were more pronounced when ONO-4007 was used for monocyte activation; the levels of ONO-4007-induced TNF- production in pretreated monocytes increased to 12 to 16 times those in fresh monocytes, whereas LPS-induced production was increased only 2.0- to 2.6-fold by the pretreatment. Present findings also showed that both GM-CSF and M-CSF improved the survival of cultured monocytes, but the cell numbers did not increase during incubation. These results suggest that the effect of GM-CSF or M-CSF on TNF- production is exerted by functional priming rather than stimulation of cell growth. Although the cytokine-inducing effect of ONO-4007 was weak in fresh monocytes, other leukocytes such as lymphocytes and neutrophils or serum factors might modulate TNF- production in response to ONO-4007. Therefore, we examined TNF- production in whole blood. In contrast to synthetic lipid A and LPS, ONO-4007 was less active in whole blood than in fresh monocytes; that is, the levels of ONO-4007-induced TNF- production in whole blood were marginal, whereas the levels of LPS-induced TNF- production were greater than those in fresh monocytes. These findings suggest that monocyte separation and adherence to a plastic plate (Ralph and Sampson-Johannes, 1990) somewhat prime monocytes for subsequent stimulation by ONO-4007 and that, therefore, ONO-4007 might not induce TNF- production in nonstimulated blood monocytes. M-CSF is present at 100 to 200 U/ml in normal human sera (Das et al., 1981; Logan et al., 1996), but the results of this study showed that 1000 U/ml of M-CSF is required for monocyte priming, which is consistent with the inability of ONO-4007 to release TNF- in whole blood. LPS-induced IL-1 production declined dramatically in monocytes pretreated with GM-CSF. This agrees with previous reports that IL-1 release in response to LPS was suppressed during monocyte maturation both in vivo (Herzyk et al., 1992) and in vitro (Hogquist et al., 1991).

There were several differences between the effects of ONO-4007 and LPS on monocyte activation: 1) More than 1 µg/ml of ONO-4007 was required for monocyte TNF- production, whereas as little as 0.1 ng/ml of LPS was sufficient for TNF- production. 2) The effect of ONO-4007 on the induction of IL-1, IL-6 and IL-12 production was much weaker than that of LPS. 3) ONO-4007 induced TNF- production only in monocytes pretreated with GM-CSF or M-CSF, whereas LPS induced TNF- production in both freshly isolated and pretreated monocytes. These differences suggest that cell activating pathways of ONO-4007 are different from those of LPS. It has been shown that LPS at very low levels (<1 ng/ml) binds to the serum protein LBP via the lipid A moiety (Schumann et al., 1990). This complex then binds to a receptor, CD14, present on the surface of monocytes/macrophages and activates these cells to produce cytokines, including TNF- (Wright et al., 1990; Kirkland et al., 1993). Monocyte activation by very low levels (<1 ng/ml) of LPS is inhibited by anti-CD14 treatment (Wright et al., 1990) or by elimination of LBP in the culture medium (Meszaros et al., 1995). It has also been observed that TNF- synthesis in response to 10 ng/ml of LPS occurs in the absence of both LBP (Schumann et al., 1990) and CD14 (Heumann et al., 1992), which implies that LPS and lipid A can activate monocytes/macrophages via LBP/CD14-independent pathway(s). The present study showed that low concentrations (1, 3 and 10 µg/ml) of ONO-4007 induced greater TNF- production in FCS-free medium than in FCS-containing medium, whereas LPS-induced TNF- production was attenuated by serum elimination. It was reported that bovine LBP in FCS was able to interact with human CD14 (Heumann et al., 1992). Moreover, anti-CD14 antibody treatment did not influence ONO-4007-induced TNF- production. However, this treatment attenuated TNF- production in response to 0.1 ng/ml, but not to 1 µg/ml, of LPS, in accordance with previous observations (Schumann et al., 1990; Heumann et al., 1992). These results suggest that ONO-4007 induces TNF- production in monocytes via LBP/CD14-independent pathways, whereas LPS uses both LBP/CD14-dependent and -independent pathways. It has been observed that ONO-4007 binds to human serum albumin and that bovine serum albumin added to culture attenuates the cytotoxic effect of ONO-4007 on endothelial cells (N. Matsumoto and A. Akimoto, unpublished observation). This indicates that serum albumin binds to ONO-4007 and inhibits the activity of ONO-4007. Elevated TNF- production at 1, 3 and 10 µg/ml and cytotoxicity at 30 and 100 µg/ml of ONO-4007 in FCS-free medium could be due to the absence of serum albumin in the culture medium.

GM-CSF is secreted by various cell types (activated T cells, activated macrophages, endothelial cells and fibroblasts) in response to antigenic or inflammatory stimulus (Dipersio, 1990; Monroy et al., 1990). M-CSF is constitutively secreted by fibroblasts, endothelial cells and epithelial cells (Ralph and Sampson-Johannes, 1990). Human monocytes produce M-CSF for their own use in response to inflammatory stimulus (Ralph and Sampson-Johannes, 1990). Moreover, tumor cells often produce M-CSF at high levels, as has been demonstrated in ovarian (Bauknecht et al., 1994), endometrial (Ramakrishnan et al., 1989), pulmonary (Horiguchi et al., 1988), pancreatic (Ralph et al., 1986) and breast cancer (Kacinski et al., 1991). Tang et al. (1990) reported that M-CSF expression of tumor cells strongly correlated with the infiltration of macrophages and lymphocytes in human breast carcinoma. These reports suggest that tumor-infiltrating macrophages in cancer patients would be primed for ONO-4007 stimulation if some immune responses were generated in their tumor tissues or if the cancer cells themselves secreted M-CSF.

In the present study, we used as a reference agent LPS from Salmonella abortus equi, which has been used in clinical phase I and II trials in cancer patients (Engelhardt et al., 1991 and 1995). In these studies, tumor response was observed in several patients with colorectal carcinoma; however, side effects such as fever, chill, hepatic toxicity and hypotension were also observed. Serum TNF- levels in patients treated with a therapeutic dose (4 ng/kg b.wt.) of LPS were elevated to 3 to 10 ng/ml. The toxic side effects of LPS are thought to be exerted via systemic activation of monocytes/macrophages, because LPS strongly activates monocytes to release proinflammatory cytokines, including TNF-, in both fresh and primed monocytes. Because ONO-4007 activates human monocytes/macrophages to release TNF- only in a primed state, it would activate only tumor-infiltrating primed macrophages in cancer patients, not blood monocytes or resting tissue macrophages. Therefore, ONO-4007 might show potent antitumor activity with relatively low toxicity compared with LPS in cancer patients.

In conclusion, this study shows that ONO-4007 activates human monocytes/macrophages to release TNF- only in a primed state and suggests that ONO-4007 would activate these cells via different pathways than would LPS.