Mechanisms of the Antimetastatic Effect in the Liver and of the Hepatocyte Injury Induced by α-Galactosylceramide in Mice

Ryusuke Nakagawa, Ikuko Nagafune, Yoshiko Tazunoki, Hiromi Ehara, Hitomi Tomura, Rieko Iijima, Kazuhiro Motoki, Masaru Kamishohara and Shuhji Seki
J Immunol June 1, 2001, 166 (11) 6578-6584; DOI: https://doi.org/10.4049/jimmunol.166.11.6578

Abstract

The role of mouse liver NK1.1 Ag+ T (NKT) cells in the antitumor effect of α-galactosylceramide (α-GalCer) has been unclear. We now show that, whereas α-GalCer increased the serum IFN-γ concentration and alanine aminotransferase activity in NK cell-depleted C57BL/6 (B6) mice and B6-beige/beige mice similarly to its effects in control B6 mice, its enhancement of the antitumor cytotoxicity of liver mononuclear cells (MNCs) was abrogated. Depletion of both NK and NKT cells in B6 mice reduced all these effects of α-GalCer. Injection of Abs to IFN-γ also inhibited the α-GalCer-induced increase in antitumor cytotoxicity of MNCs. α-GalCer induced the expression of Fas ligand on NKT cells in the liver of B6 mice. Whereas α-GalCer did not increase serum alanine aminotransferase activity in B6-lpr/lpr mice and B6-gld/gld mice, it increased the antitumor cytotoxicity of liver MNCs. The α-GalCer-induced increase in survival rate apparent in B6 mice injected intrasplenically with B16 tumor cells was abrogated in beige/beige mice, NK cell-depleted B6 mice, and B6 mice treated with Abs to IFN-γ. Depletion of CD8+ T cells did not affect the α-GalCer-induced antitumor cytotoxicity of liver MNCs but reduced the effect of α-GalCer on the survival of B6 mice. Thus, IFN-γ produced by α-GalCer-activated NKT cells increases both the innate antitumor cytotoxicity of NK cells and the adaptive antitumor response of CD8+ T cells, with consequent inhibition of tumor metastasis to the liver. Moreover, NKT cells mediate α-GalCer-induced hepatocyte injury through Fas-Fas ligand signaling.

Both NK cells and NK1.1 Ag+ T cells (NKT cells)2 are abundant in mouse liver (1, 2, 3) and are capable of producing substantial amounts of IFN-γ (3, 4, 5). The TCR of mouse NKT cells comprises the Vα14Jα281 gene product combined with a Vβ8, Vβ2, or Vβ7 chain (6, 7), and the development of these cells is dependent on the nonclassical MHC class I molecule CD1d (8). Liver NKT cells produced more IFN-γ and exhibited a greater antitumor cytotoxicity in response to stimulation with IL-12 in vivo than did NK cells (4). In contrast, the amount of IFN-γ produced by liver NK cells was greater than that produced by NKT cells in mice with bacterial peritonitis (5). Furthermore, injection of mice with LPS induced a marked increase in the antitumor cytotoxicity of NKT cells as a result of the IFN-γ produced by NK cells (3, 4, 9).

The glycolipid Ag α-galactosylceramide (α-GalCer) induces the production of IFN-γ by mouse NKT cells in a CD1d-dependent manner (10). We previously suggested that liver antitumor effectors activated by α-GalCer might include NK cells (11), whereas other studies have implicated NKT cells as important mediators of the antitumor effect of α-GalCer both in vitro and in vivo (10, 12). However, α-GalCer-induced antitumor cytotoxicity was recently shown to be mediated predominantly by NK cells in response to the IFN-γ produced by NKT cells (13, 14). Osman et al. (15) also recently showed that liver NKT cells undergo apoptosis in response to injection of α-GalCer in mice. These researchers suggested that activated NKT cells may be effectors of the hepatocyte damage induced by α-GalCer injection but that NK cells may also play a role in this process. The importance of NKT cells in both antitumor immunity and hepatocyte injury induced by α-GalCer was revealed by the observation that these phenomena do not occur in CD1-deficient or β2-microglobulin-deficient mice (15), both of which lack NKT cells. However, the effectors of the antimetastatic action of α-GalCer in the liver as well as of the hepatocyte damage induced by this glycolipid remain to be definitively identified.

We have now shown that the principal mediators of the direct antitumor effect and the antimetastatic effect of α-GalCer in the liver are NK cells that have been activated by IFN-γ produced by NKT cells. We also show that CD8+ T cells contribute to the antimetastatic effect of α-GalCer, whereas NKT cells mediate α-GalCer-induced hepatocyte damage through the Fas-Fas ligand signaling pathway. These conclusions are supported by observations with beige/beige (bg/bg) mice, in which NK cell function is impaired (16, 17), as well as with lpr/lpr (lpr) mice (2, 18, 19) and gld/gld (gld) mice (20), in which Fas and Fas ligand, respectively, are defective.

Materials and Methods

Mice and preparation of liver mononuclear cells (MNCs)

Male C57BL/6 (B6) mice at 6 wk of age were obtained from Nippon SLC (Hamamatsu, Japan), and male B6-bg/bg, B6-lpr, and B6-gld mice of the same age were obtained from The Jackson Laboratory Japan (Tokyo, Japan). Mice were maintained and fed under standard laboratory conditions. Hepatic MNCs were prepared essentially as described (1, 4). In brief, the liver was passed through a stainless steel mesh, and the resulting dissociated cells were suspended in HBSS, washed, resuspended in an isotonic 30% Percoll solution (Sigma) containing heparin (100 U/ml), and centrifuged at 500 × g for 15 min at room temperature. The resulting pellet was resuspended in RBC lysis solution and then washed twice in RPMI 1640 supplemented with 5% FCS.

Reagents

α-GalCer, or (2S,3S,4R)-1-O-(α-d-galactopyranosyl)-2-(N-hexacosanoylamino)-1,2,4-octadecanetriol (KRN7000), was synthesized in our laboratory (21, 22, 23). The original solution of α-GalCer (220 μg/ml) was prepared with 0.5% polysorbate 20 (Nikko Chemical, Tokyo, Japan) in saline and was subsequently diluted with this solution (vehicle) or with saline before i.v. injection at a dose of 100 μg/kg of body mass.

Flow cytometric analysis

The surface phenotypes of liver MNCs were characterized by two- or three-color flow cytometric analysis. An FITC-conjugated hamster mAb to mouse TCRαβ (H57-597, IgG) and a PE-conjugated mouse mAb to NK1.1 (PK136, IgG2a) were obtained from PharMingen (San Diego, CA). Before staining with Abs, the MNCs were incubated for 10 min with Fc-blocker (2.4 G2; PharMingen) to prevent nonspecific binding. For analysis of Fas ligand expression, liver MNCs were isolated 1 h after the injection of α-GalCer or vehicle into B6 mice. The cells were then stained with the FITC-conjugated mAb to TCRαβ, the PE-conjugated mAb to NK1.1, and a biotin-conjugated hamster mAb to Fas ligand (MFL1, IgG; PharMingen); immune complexes formed by the latter Ab were detected with Cy5-streptavidin. Flow cytometry was performed with FACSCalibur and FACScan instruments (BD Biosciences, Mountain View, CA).

In vivo cell depletion

Monoclonal Abs to CD4 (L3T4), to CD8 (Lyt2.2), or to NK1.1 were derived from GK1.5, 2.43, and PK136 hybridoma cells (IBL, Gunma, Japan), respectively. We previously showed that a single i.v. injection of an optimal dose of Abs to asialo-GM1 (AGM1) (Wako, Tokyo, Japan) resulted in depletion of NK cells alone, whereas injection of Abs to NK1.1 depleted both NK and NKT cells for at least 5 days (24). Mice were injected i.p. with 500 μg of anti-CD4 or anti-CD8, 50 μg of anti-AGM1, or 200 μg of anti-NK1.1.

Cytotoxicity assay

NK cell-sensitive YAC-1 lymphoma cells and B16 melanoma cells (of B6 origin) were used as target cells. Target cells (3 × 106) were labeled for 60 min at 37°C with 100 μCi of Na251CrO4 in 500 μl of RPMI 1640 supplemented with 10% FCS. Then, they were washed three times with medium alone and subjected to the cytotoxicity assay. Labeled targets (2 × 103 cells per well) were incubated for 4 h at 37°C in 96-well round-bottom microtiter plates containing RPMI 1640 (total volume of 100 μl) and liver MNCs obtained from mice injected 24 h previously with α-GalCer (100 μg/kg). The plates were then centrifuged, and the resulting supernatants were harvested and their content of radioactivity was determined with a gamma counter. Cytotoxicity was calculated as the percentage of released radioactivity after correction for spontaneous release, which was <15% of maximal release.

Measurement of serum IFN-γ and alanine aminotransferase (ALT)

The serum concentration of IFN-γ was measured by ELISA (Endogen, Boston, MA). The activity of ALT in serum was determined with a DRI-CHEM 3000V instrument (Fuji Medical Systems, Tokyo, Japan).

B16 model of hepatic metastasis

Hepatic metastases of B16 tumor cells were produced as described (25). In brief, the spleen of anesthetized mice was exposed to allow the direct intrasplenic injection of 3 × 106 B16 cells in 0.1 ml of medium. The spleen was then removed after clamping of the artery and vein, and the abdomen and skin were surgically sutured. This approach results in the metastasis of B16 tumor cells almost exclusively to the liver.

In vivo depletion of IFN-γ

Rat monoclonal IgG1 to mouse IFN-γ was derived from R4-6A2 hybridoma cells (IBL). Mice were injected i.p. with 1 mg of anti-IFN-γ or an isotype-control Ab (R3-34; PharMingen) at the same time that they received an i.v. injection of α-GalCer (100 μg/kg).

Statistical analysis

Data are expressed as means ± SD, and differences among groups were analyzed by the Mann-Whitney U test with StatView software. Mouse survival rates were analyzed by the log-rank test. A p value of <0.05 was considered statistically significant.

Results

Transient disappearance of NKT cells and an increase in the proportion of NK cells in the liver of mice injected with α-GalCer

Injection of B6 mice with α-GalCer (100 μg/kg, i.v.) resulted in a decrease in the number of NKT cells in the liver that was first apparent after ∼3 h, with most such cells having disappeared by 12 h after injection (Fig. 1⇓). In contrast, the proportion of NK cells gradually increased for up to 5 days after α-GalCer injection; NKT cells also began to reappear from ∼2 days after injection. Injection of α-GalCer also induced a transient disappearance of NKT cells in bg/bg mice (data not shown).

FIGURE 1.
FIGURE 1.

Effects of injection of α-GalCer on the proportions of NKT and NK cells in the liver of B6 mice. Mice were injected with α-GalCer (100 μg/kg, i.v.), and, at the indicated times thereafter, MNCs were isolated from the liver and subjected to flow cytometric analysis with FITC-conjugated Abs to TCRαβ and PE-conjugated anti-NK1.1. The percentages of NK and NKT cells are indicated in the upper left and upper right quadrants, respectively. Data were obtained from an experiment that was repeated five times with similar results.

Increase in the number of liver MNCs and long-lasting antitumor cytotoxicity of these cells induced by α-GalCer

The total number of liver MNCs showed a 4-fold increase 3 days after injection of α-GalCer in B6 mice (Fig. 2⇓A). This effect was accompanied by an increase in antitumor cytotoxicity of liver MNCs that was apparent with either YAC-1 lymphoma or B16 melanoma cells and persisted for >5 days (Fig. 2⇓B). Such a long-lasting and marked antitumor cytotoxicity was not observed after injection of either IFN-γ or IL-12 (data not shown).

FIGURE 2.
FIGURE 2.

Effects of injection of α-GalCer on the total number (A) and cytotoxicity (B) of MNCs in the liver of B6 mice. A, Mice were injected with α-GalCer (100 μg/kg, i.v.), and, at the indicated times thereafter, the total number of MNCs in the liver was determined. Data are means ± SD of values from five mice for each time point. B, At the indicated times after injection of mice with α-GalCer, the cytotoxicity of liver MNCs was determined in vitro with YAC-1 or B16 cells at an E:T ratio of 100. Data are expressed as percentage of target cell lysis and are means ± SD of values from triplicate samples. Data were obtained from an experiment that was repeated three times with similar results.

Abrogation of α-GalCer-induced antitumor cytotoxicity of liver MNCs by injection of mice with Abs to AGM1, to NK1.1, or to IFN-γ

Depletion of NK cells by pretreatment of mice with anti-AGM1 or depletion of both NK and NKT cells by pretreatment with anti-NK1.1 (Fig. 3⇓A) almost completely abolished the effect of α-GalCer on antitumor cytotoxicity of liver MNCs apparent with either YAC-1 or B16 cells (Fig. 3⇓B, a and b). However, depletion of either CD8+ or CD4+ T cells by pretreatment of mice with anti-CD8 or anti-CD4 did not affect the α-GalCer-induced increase in antitumor cytotoxicity of liver MNCs (Fig. 3⇓B, c and d). Anti-IFN-γ also markedly inhibited the α-GalCer-induced increase in antitumor cytotoxicity of liver MNCs (Fig. 3⇓B, e and f). Depletion of liver NKT cells by injection of mice with anti-NK1.1 was also confirmed by the disappearance of CD4+ Vβ8 T cells with intermediate TCR expression (data not shown). Although injection of anti-CD8 almost completely depleted CD8+ T cells, and injection of anti-CD4 depleted CD4+ T cells with high TCR expression, the number of CD4+ NKT cells with intermediate TCR expression was not substantially affected by injection of anti-CD4 (data not shown). CD4+ NKT cells have been shown to express CD4 only at a low level (26, 27).

FIGURE 3.
FIGURE 3.

A, Depletion of NK cells and of both NK and NKT cells in B6 mice. Mice were injected i.p. with anti-AGM1 (αAGM1; 50 μg) (middle panels) or anti-NK1.1 (200 μg) (bottom panels); control mice (no Ab) were injected with PBS (top panels). After 3 days, mice were injected with α-GalCer (100 μg/kg, i.v.) (right panels) or vehicle (left panels). Liver MNCs were subjected to flow cytometric analysis with anti-NK1.1 and anti-TCRαβ 1 day after injection of α-GalCer. Data were obtained from an experiment that was repeated three times with similar results. B, Effects of depletion of NK cells, NKT cells, CD4+ T cells, CD8+ T cells, or IFN-γ on α-GalCer-induced antitumor cytotoxicity. Mice were injected i.p. with anti-AGM1 (50 μg) or anti-NK1.1 (200 μg) (a and b), anti-CD4 (500 μg) (c), anti-CD8 (500 μg) (d), or anti-IFN-γ (1 mg) (e and f); control mice were injected with PBS (a–d) or an isotype-control Ab (e and f). After 3 days, or concurrently in the case of anti-IFN-γ, mice were injected with α-GalCer (100 μg/kg i.v.) or vehicle. The cytotoxicity of liver MNCs was determined in vitro with YAC-1 or B16 cells at the indicated E:T ratios 1 day after injection of α-GalCer. Data are means ± SD of values from triplicate samples. Data were obtained from an experiment that was repeated five times with similar results.

Lack of effect of α-GalCer on antitumor cytotoxicity of liver MNCs in bg/bg mice

Given that the function of NK cells is impaired in bg/bg mice (16, 17), we next examined the effect of α-GalCer on the antitumor cytotoxicity of liver MNCs in these animals. Consistent with the results obtained by injection of B6 mice with anti-AGM1, α-GalCer did not increase the antitumor cytotoxicity of liver MNCs of bg/bg mice with either YAC-1 or B16 cells as targets (Fig. 4⇓).

FIGURE 4.
FIGURE 4.

Effect of injection of α-GalCer on the cytotoxic activity of liver MNCs of bg/bg mice. Mice (bg/bg or B6) were injected with α-GalCer (100 μg/kg, i.v.) or vehicle, and, after 24 h, the cytotoxicity of liver MNCs was determined in vitro with YAC-1 cells (left) or B16 cells (right) at the indicated E:T ratios. Data are means ± SD of values from triplicate samples. Data were obtained from an experiment that was repeated three times with similar results.

Effects of α-GalCer and Ab pretreatment on serum IFN-γ and ALT levels in B6 and bg/bg mice

Injection of B6 mice with α-GalCer resulted in an increase in the serum concentration of IFN-γ that was apparent at 6 h and maximal at 12 h (Table I⇓); the IFN-γ concentration had decreased markedly at 24 h and had returned to basal values by 48 h (data not shown) after α-GalCer injection. Whereas depletion of NK cells alone by pretreatment of B6 mice with anti-AGM1 did not inhibit the production of IFN-γ, depletion of both NK and NKT cells by pretreatment with anti-NK1.1 markedly inhibited the α-GalCer-induced increase in the serum concentration of IFN-γ (Table I⇓). Furthermore, injection of α-GalCer induced an increase in the serum concentration of IFN-γ in bg/bg mice that was similar to that observed in B6 mice (Table I⇓). Thus, these results suggested that most of the IFN-γ produced in response to α-GalCer is derived from NKT cells rather than from NK cells.

View this table:
Table I.

Effects of α-GalCer and Ab pretreatment on the serum concentration of IFN-γ in B6 and bg/bg micea

Injection of B6 mice with α-GalCer induced an increase in the activity of ALT in serum that was apparent at 3 h and maximal at 12 h; the ALT activity had decreased by 48 h (Table II⇓) and had returned to basal values by 72 h (data not shown) after α-GalCer injection. Similar to the results obtained for serum IFN-γ, the effect of α-GalCer on serum ALT activity was not affected by depletion of NK cells alone but was significantly inhibited by the depletion of both NK and NKT cells. In addition, α-GalCer also induced a marked increase in serum ALT activity in bg/bg mice. However, treatment of B6 mice with anti-IFN-γ did not inhibit the α-GalCer-induced increase in serum ALT activity (data not shown). Thus these observations suggested that NKT cells, but not NK cells, contribute to α-GalCer-induced hepatocyte injury in an IFN-γ-independent manner.

View this table:
Table II.

Effects of α-GalCer and Ab pretreatment on the serum activity of ALT in B6 and bg/bg micea

Failure of α-GalCer to induce liver injury in lpr and gld mice and α-GalCer-induced expression of Fas ligand on NKT cells of B6 mice

To explore the mechanism by which α-GalCer induces hepatocyte injury, we examined the effects of this compound in lpr and gld mice, which are defective in the function of Fas and Fas ligand, respectively. Injection of both lpr and gld mice with α-GalCer increased the cytotoxicity of liver MNCs toward YAC-1 cells (Table III⇓). NKT cells in these mice also disappeared within 12 h of α-GalCer injection (data not shown). However, α-GalCer had no significant effect on the serum ALT activity in either lpr or gld mice (Table III⇓). Furthermore, flow cytometric analysis of liver MNCs revealed that α-GalCer induced an increase in the expression of Fas ligand on NKT cells 1 h after injection of B6 mice that was markedly greater than that apparent for either NK or T cells (Fig. 5⇓). These observations are consistent with the notion that the liver injury induced by α-GalCer is mediated predominantly by NKT cells through the Fas-Fas ligand signaling pathway.

FIGURE 5.
FIGURE 5.

Effects of α-GalCer on the expression of Fas ligand on NKT, NK, and T cells. A, One hour after the injection of B6 mice with α-GalCer (100 μg/kg i.v.) or vehicle, liver MNCs were isolated and subjected to flow cytometric analysis with the indicated combinations of FITC-conjugated anti-TCRαβ, PE-conjugated anti-NK1.1, and biotin-conjugated anti-Fas ligand (and Cy5-streptavidin). The percentage values shown represent the proportion of cells with the corresponding phenotype. B, Fluorescence profiles showing the effects of α-GalCer (or vehicle) treatment as in A on the expression of Fas ligand by gated NKT, NK, and T cells. The percentage of each cell subset that expresses Fas ligand is indicated. Data were obtained from an experiment that was repeated three times with similar results.

View this table:
Table III.

Effects of α-GalCer on the serum activity of ALT and the cytotoxicity of liver MNCs in lpr and gld micea

Effects of α-GalCer and various Abs on the survival of mice after intrasplenic injection of B16 tumor cells

We next evaluated mouse survival rates after intrasplenic injection of B16 cells under various conditions. Consistent with the results of the antitumor cytotoxicity experiments, whereas α-GalCer induced a marked increase in the survival rate of B6 mice injected with B16 tumor cells, depletion of NK cells alone by pretreatment with anti-AGM1 prevented this effect of α-GalCer (Fig. 6⇓A). Similarly, depletion of both NK and NKT cells also completely inhibited the prolongation of mouse survival induced by α-GalCer (Fig. 6⇓A). Furthermore, the survival rate of B16 cell-injected bg/bg mice was not markedly affected by α-GalCer (Fig. 6⇓B). Pretreatment of mice with anti-CD8 partially inhibited the survival-promoting effect of α-GalCer in B6 mice (Fig. 6⇓C), suggesting that CD8+ T cells contribute to the antimetastatic effect of α-GalCer in the liver. Although anti-CD4 tended to reduce the increase in survival rate induced by α-GalCer, this effect was not statistically significant (Fig. 6⇓C). Treatment of B6 mice with anti-IFN-γ also completely inhibited the α-GalCer-induced increase in survival rate (Fig. 6⇓D).

FIGURE 6.
FIGURE 6.

Effects of α-GalCer and various Abs on survival rate after intrasplenic injection of B16 tumor cells in B6 and bg/bg mice. A, B6 mice were injected i.p. with anti- AGM1 (αAGM1; 50 μg), anti-NK1.1 (200 μg), or PBS 3 days before intrasplenic injection of B16 cells. After an additional 24 h, the mice were injected with α-GalCer (100 μg/kg i.v.) or vehicle. Survival was evaluated at the indicated times after injection of tumor cells. ∗, p < 0.05 vs other groups. B, B6 and bg/bg mice were injected with α-GalCer or vehicle 1 day after injection of B16 tumor cells. ∗, p < 0.05 vs other groups. C, B6 mice were treated as in A with the exception that the Abs administered were anti-CD4 (500 μg) or anti-CD8 (500 μg). ∗, p < 0.05 vs other groups with the exception of that treated with anti-CD4 and α-GalCer. D, One day after intrasplenic injection of B16 cells, B6 mice were injected with 1 mg of anti-IFN-γ (or an isotype-control Ab) i.p. in the absence or presence of α-GalCer (or vehicle) (i.v.). ∗, p < 0.05 vs other groups. All experimental groups in A–D included five to eight mice.

Discussion

We have now shown that NK cells are the principal effectors of the antimetastatic action of α-GalCer in the liver. Most NKT cells had disappeared from the liver within 12 h of α-GalCer administration, probably as a result of apoptosis (15). However, IFN-γ produced by α-GalCer-activated NKT cells was required for the antitumor and antimetastatic actions of NK cells. In addition, CD8+ T cells were also shown to contribute to the antimetastatic effect in the liver and the prolongation of mouse survival induced by α-GalCer. Our observations that α-GalCer did not induce hepatic injury in lpr and gld mice also indicate that NKT cells mediate the hepatocyte damage induced by this compound through the Fas-Fas ligand signaling pathway.

NKT cells are activated by IL-12 to produce IFN-γ and thereby exert an antimetastatic effect in the liver of both normal and bg/bg mice (28). In contrast, α-GalCer did not induce a marked increase in the antitumor cytotoxicity of liver MNCs of bg/bg mice and did not exert an effective antimetastatic action in the liver of these animals injected with B16 tumor cells. Thus, these observations indicate that the function of NK cells is impaired in bg/bg mice, whereas that of NKT cells is intact. Indeed, α-GalCer injection resulted in the depletion of NKT cells from the liver and increased the serum concentration of IFN-γ in bg/bg mice to extents similar to those apparent in B6 mice. The role of IFN-γ produced by NKT cells in the antitumor action of α-GalCer in the liver was demonstrated by the observation that treatment of mice with anti-IFN-γ greatly reduced both the antitumor cytotoxicity of liver MNCs and the antimetastatic effect induced by this glycolipid. It was also recently shown that α-GalCer induces NK cell proliferation and the acquisition by these cells of antitumor activity in vitro and in vivo, and that the antitumor activity of NK cells was, at least in part, dependent on IFN-γ produced by NKT cells (13, 14).

The marked increase in the antitumor cytotoxicity of liver MNCs induced by α-GalCer persisted for up to 10 days. Neither IFN-γ nor IL-12 induced such a long-lasting increase in the antitumor cytotoxicity of liver MNCs. Indeed, we previously showed that administration of α-GalCer to mice that had received an intrasplenic injection of colon-26 tumor cells (which metastasize to the liver) increased the survival rate more effectively than did treatment with IL-12 (11). Furthermore, we have now also shown that CD8+ T cells participate in the antimetastatic effect of α-GalCer. Although IFN-γ produced by NKT cells was essential for theantitumor action of α-GalCer, the serum concentration of this cytokine had returned to basal values within 48 h after α-GalCer injection. Therefore, other factors may also be required for the long-lasting antitumor effect of α-GalCer. α-GalCer also inhibited liver metastasis of i.v. injected EL4 cells (murine T lymphoma) and increased the survival rate of the injected mice (29). However, in contrast to IL-12 (28), α-GalCer did not effectively inhibit lung metastasis of i.v. injected 3LL Lewis tumor cells (which are highly resistant to NK cells) or liver metastasis of 3LL cells injected into the spleen, nor did it significantly increase the survival rate of the injected mice (R. Nakagawa and K. Motoki, unpublished data). Thus, the antitumor function of activated NKT cells appears to be important for inhibiting the metastasis of NK cell-resistant tumors (28).

Although treatment of mice that had received an intrasplenic injection of colon-26 tumor cells with α-GalCer induced the complete rejection of the tumor cells in some mice and rendered them resistant to subsequent s.c. challenge with the same tumor cells, such treatment did not affect susceptibility to s.c. injection with Meth A or other tumor cells (11). Given that the inhibitory effect of α-GalCer on s.c. growth of tumors in nonsensitized mice is weak compared with that on tumor metastasis (11), NK and NKT cells may be important for innate immunity against tumor metastasis, whereas cytotoxic CD8+ T cells (possibly with the help of CD4+ T cells) may be important either for inhibition of s.c. tumor growth or for adaptive antitumor immunity (as memory T cells). Thus, a cascade of NKT, NK, and cytotoxic CD8+ T cells may be responsible for the antitumor action of α-GalCer. Both NK and NKT cells were recently shown to be important antitumor effectors in response to the i.p. injection of tumor cells (30). Furthermore, cooperation of NK and CD56+ T cells (which may be a human functional counterpart of mouse NKT cells; Refs. 3 and 31) in human liver appears important in inhibition of the generation of hepatocellular carcinoma (31). These various observations not only support our previous proposal that the lymphocyte subsets primarily responsible for antitumorigenesis may differ among tissues (1) but also suggest that the interaction of lymphocyte subsets is important in antitumor immunity.

Our results indicate that NKT cells are the effectors of hepatocyte damage induced by α-GalCer, given that prior depletion of NK cells alone did not inhibit the increase in serum ALT activity in α-GalCer-treated mice whereas depletion of both NK and NKT cells did. In addition, α-GalCer induced a marked increase in the serum activity of ALT in bg/bg mice. Furthermore, our observations that α-GalCer increased the expression of Fas ligand most markedly on NKT cells in B6 mice and that this glycolipid did not trigger hepatic injury in lpr and gld mice indicate that NKT cells activated by α-GalCer induce hepatocyte damage through the Fas-Fas ligand pathway.

Injection of IL-12 was previously shown to induce a marked increase in antitumor cytotoxicity of liver NKT cells, and IFN-γ produced by activated NKT cells was shown to be responsible for the hepatocyte damage and lethal shock induced by IL-12 and subsequent injection (after 24 h) of LPS (4). In contrast, NK cells were shown to be responsible for hepatic failure induced by injection of LPS in Propionibacterium acnes-primed mice, and IL-12, IL-18, or IFN-γ was essential for this hepatic failure (32). Bacterial superantigens also activate liver NK and NKT cells through the IL-12 produced by Kupffer cells (33). These observations suggest that, although liver Kupffer, NKT, and NK cells are essential leukocyte subsets for protection against both malignant tumors and bacterial infection, as a result of their induction of the Th1 cell-mediated immune response, inadequate activation of this defense mechanism can lead to hepatic dysfunction and multiple organ failure (3, 4).

α-GalCer was originally isolated from a marine sponge (21). No disease or condition has been shown to induce this molecule in mice, suggesting that it is not likely an endogenous ligand in these animals. The initial rapid disappearance of NKT cells from the liver of α-GalCer-treated mice, probably as a result of apoptosis, is followed by the reappearance of these cells around 2 days after α-GalCer injection; these presumably newly produced NKT cells are likely not hepatotoxic. Injection of a second dose of α-GalCer into B6 mice 7 days after the first administration did not result in an increase in the serum levels of IFN-γ or ALT or in the disappearance of NKT cells from the liver (data not shown), suggesting that the mice had developed tolerance to this glycolipid.

LPS was recently shown to induce a marked increase in the antitumor cytotoxicity of NKT cells (4, 9), whereas NK cells produced more IFN-γ in response to LPS than did NKT cells (3). Given that the LPS-induced increase in antitumor cytotoxicity of liver MNCs and the extent of tumor metastasis to the liver were reduced by pretreatment of mice with either anti-IFN-γ or anti-IL-12 (3, 9), IL-12 produced by Kupffer cells and the subsequent production of IFN-γ by NK cells appear important for the LPS-induced antitumor cytotoxicity of NKT cells. This relationship between NK and NKT cells is opposite to that revealed by our present results showing that α-GalCer induces NKT cells to produce IFN-γ, which then increases the antitumor cytotoxicity of NK cells; IL-12 was also shown not to contribute to the α-GalCer-induced antitumor cytotoxicity of liver MNCs (14). Cells of the hepatic monocyte lineage, including Kupffer (or dendritic), NK, NKT, and CD8+ T cells, thus appear to interact with each other and to modulate precisely the immune response in various immunopathologic states in the liver.

Acknowledgments

We thank Naomi Otake for her help and cooperation in experiments.

Footnotes

  • ↵1 Address correspondence and reprint requests to Dr. Shuhji Seki, Department of Microbiology, National Defense Medical College, 3-2 Namiki, Tokorozawa 359-8513, Japan. E-mail address: btraums{at}res.ndmc.ac.jp

  • ↵2 Abbreviations used in this paper: NKT cells, NK1.1 Ag+ T cells; α-GalCer, α-galactosylceramide; MNC, mononuclear cell; AGM1, asialo-GM1; ALT, alanine aminotransferase; B6, C57BL/6; bg/bg, beige/beige; lpr, lpr/lpr; gld, gld/gld.

  • Received January 18, 2001.
  • Accepted March 29, 2001.

References

  1. ↵
    Hashimoto, W., K. Takeda, R. Anzai, K. Ogasawara, H. Sakihara, K. Sugiura, S. Seki, K. Kumagai. 1995. Cytotoxic NK1.1 Ag+ αβ T cells with intermediate TCR induced in the liver of mice by IL-12. J. Immunol. 154: 4333
    OpenUrlAbstract
  2. ↵
    Tsukahara, A., S. Seki, T. Iiai, T. Moroda, H. Watanabe, S. Suzuki, T. Tada, H. Hiraide, K. Hatakeyama, T. Abo. 1997. Mouse liver T cells: their change with aging and in comparison with peripheral T cells. Hepatology 26: 301
    OpenUrlCrossRefPubMed
  3. ↵
    Seki, S., Y. Habu, T. Kawamura, K. Takeda, H. Dobashi, T. Ohkawa, H. Hiraide. 2000. The liver as a crucial organ in the first line of host defense: the roles of Kupffer cells, natural killer (NK) cells and NK1.1 Ag+ T cells in T helper 1 immune responses. Immunol. Rev. 174: 35
    OpenUrlCrossRefPubMed
  4. ↵
    Ogasawara, K., K. Takeda, W. Hashimoto, M. Satoh, R. Okuyama, N. Yanai, M. Obinata, K. Kumagai, H. Takada, H. Hiraide, S. Seki. 1998. Involvement of NK1+ T cells and their IFN-γ production in the generalized Shwartzman reaction. J. Immunol. 160: 3522
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Seki, S., S. Osada, S. Ono, S. Aosasa, Y. Habu, T. Nishikage, H. Mochizuki, H. Hiraide. 1998. Role of liver NK cells and peritoneal macrophages in γ interferon and interleukin-10 production in experimental bacterial peritonitis in mice. Infect. Immun. 66: 5286
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Makino, Y., H. Koseki, Y. Adachi, T. Akasaka, K. Tsuchida, M. Taniguchi. 1994. Extrathymic differentiation of a T cell bearing invariant Vα14J α281 TCR. Int. Rev. Immunol. 11: 31
    OpenUrlCrossRefPubMed
  7. ↵
    Lantz, O., A. Bendelac. 1994. An invariant T cell receptor α-chain is used by a unique subset of major histocompatibility complex class I-specific CD4+ and CD4−8− T cells in mice and humans. J. Exp. Med. 180: 1097
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Bendelac, A.. 1995. Positive selection of mouse NK1+ T cells by CD1-expressing cortical thymocytes. J. Exp. Med. 182: 2091
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Takahashi, M., K. Ogasawara, K. Takeda, W. Hashimoto, H. Sakihara, K. Kumagai, R. Anzai, M. Satoh, S. Seki. 1996. LPS induces NK1.1+ αβ T cells with potent cytotoxicity in the liver of mice via production of IL-12 from Kupffer cells. J. Immunol. 156: 2436
    OpenUrlAbstract
  10. ↵
    Kawano, T., J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, K. Motoki, H. Ueno, R. Nakagawa, H. Sato, E. Kondo, et al 1997. CD1d-restricted and TCR-mediated activation of Vα14 NKT cells by glycosylceramides. Science 278: 1626
    OpenUrlAbstract/FREE Full Text
  11. ↵
    Nakagawa, R., K. Motoki, H. Ueno, R. Iijima, H. Nakamura, E. Kobayashi, A. Shimosaka, Y. Koezuka. 1998. Treatment of hepatic metastasis of the colon26 adenocarcinoma with an α-galactosylceramide, KRN7000. Cancer Res. 58: 1202
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Kawano, T., J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, H. Sato, E. Kondo, M. Harada, H. Koseki, T. Nakayama, et al 1998. Natural killer-like nonspecific tumor cell lysis mediated by specific ligand-activated Vα14 NKT cells. Proc. Natl. Acad. Sci. USA 95: 5690
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Carnaud, C., D. Lee, O. Donnars, S. H. Park, A. Beavis, Y. Koezuka, A. Bendelac. 1999. Cross-talk between cells of the innate immune system: NKT cells rapidly activate NK cells. J. Immunol. 163: 4647
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Eberl, G., H. R. MacDonald. 2000. Selective induction of NK cell proliferation and cytotoxicity by activated NKT cells. Eur. J. Immunol. 30: 985
    OpenUrlCrossRefPubMed
  15. ↵
    Osman, Y., T. Kawamura, T. Naito, K. Takeda, L. Van Kaer, K. Okumura, T. Abo. 2000. Activation of hepatic NKT cells and subsequent liver injury following administration of α-galactosylceramide. Eur. J. Immunol. 30: 1919
    OpenUrlCrossRefPubMed
  16. ↵
    Talmadge, J. E., K. M. Meyers, D. J. Prieur, J. R. Starkey. 1980. Role of NK cells in tumour growth and metastasis in beige mice. Nature 284: 622
    OpenUrlCrossRefPubMed
  17. ↵
    Kiessling, R. W., R. M. Welsh, Jr. 1980. Killing of normal cells by activated mouse natural killer cells: evidence for two patterns of genetic regulation of lysis. Int. J. Cancer 25: 611
    OpenUrlCrossRefPubMed
  18. ↵
    Ohteki, T., S. Seki, T. Abo, K. Kumagai. 1990. Liver is a possible site for the proliferation of abnormal CD3+4−8− double-negative lymphocytes in autoimmune MRL-lpr/lpr mice. J. Exp. Med. 172: 7
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Watanabe-Fukunaga, R., C. I. Brannan, N. G. Copeland, N. A. Jenkins, S. Nagata. 1992. Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356: 314
    OpenUrlCrossRefPubMed
  20. ↵
    Takahashi, T., M. Tanaka, C. I. Brannan, N. A. Jenkins, N. G. Copeland, T. Suda, S. Nagata. 1994. Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76: 969
    OpenUrlCrossRefPubMed
  21. ↵
    Morita, M., K. Motoki, K. Akimoto, T. Natori, T. Sakai, E. Sawa, K. Yamaji, Y. Koezuka, E. Kobayashi, H. Fukushima. 1995. Structure-activity relationship of α-galactosylceramides against B16-bearing mice. J. Med. Chem. 38: 2176
    OpenUrlCrossRefPubMed
  22. ↵
    Yamaguchi, Y., K. Motoki, H. Ueno, K. Maeda, E. Kobayashi, H. Inoue, H. Fukushima, Y. Koezuka. 1996. Enhancing effects of (2S,3S,4R)-1-O-(α-d- galactopyranosyl)-2-(N-hexacosanoylamino)-1,3,4-octadecanetriol (KRN7000) on antigen-presenting function of antigen-presenting cells and antimetastatic activity of KRN7000-pretreated antigen-presenting cells. Oncol. Res. 8: 399
    OpenUrlPubMed
  23. ↵
    Motoki, K., K. Maeda, H. Ueno, E. Kobayashi, T. Uchida, H. Fukushima, Y. Koezuka. 1996. Antitumor activities of combined treatment with a novelimmunomodulator, (2S,3S,4R)-1-O-(α-d-galactopyranosyl)-2-(N-hexacosanoylamino)-1,3,4-octadecanetriol (KRN7000), and radiotherapy in tumor-bearing mice. Oncol. Res. 8: 155
    OpenUrlPubMed
  24. ↵
    Seki, S., W. Hashimoto, K. Ogasawara, M. Satoh, H. Watanabe, Y. Habu, H. Hiraide, K. Takeda. 1997. Antimetastatic effect of NK1+ T cells on experimental haematogenous tumour metastases in the liver and lungs of mice. Immunology 92: 561
    OpenUrlCrossRefPubMed
  25. ↵
    Kopper, L., T. Van Hanh, K. Lapis. 1982. Experimental model for liver metastasis formation using Lewis lung tumor. J. Cancer Res. Clin. Oncol. 103: 31
    OpenUrlCrossRefPubMed
  26. ↵
    Seki, S., D. H. Kono, R. S. Balderas, A. N. Theofilopoulos. 1994. Vβ repertoire of murine hepatic T cells: implication for selection of double negative αβ+ T cells. J. Immunol. 153: 637
    OpenUrlAbstract
  27. ↵
    Takeda, K., S. Seki, K. Ogasawara, R. Anzai, W. Hashimoto, K. Sugiura, M. Takahashi, M. Satoh, K. Kumagai. 1996. Liver NK1.1+ CD4+ αβ T cells activated by IL-12 as a major effector in inhibition of experimental tumor metastasis. J. Immunol. 156: 3366
    OpenUrlAbstract
  28. ↵
    Anzai, R., S. Seki, K. Ogasawara, W. Hashimoto, K. Sugiura, M. Sato, K. Kumagai, K. Takeda. 1996. Interleukin-12 induces cytotoxic NK1+ αβ T cells in the lungs of euthymic and athymic mice. Immunology 88: 82
    OpenUrlCrossRefPubMed
  29. ↵
    Nakagawa, R., K. Motoki, H. Nakamura, H. Ueno, R. Iijima, A. Yamauchi, S. Tsuyuki, T. Inamoto, Y. Koezuka. 1998. Antitumor activity of α-galactosylceramide, KRN7000, in mice with EL-4 hepatic metastasis and its cytokine production. Oncol. Res. 10: 561
    OpenUrlPubMed
  30. ↵
    Kawamura, T., S. Seki, K. Takeda, J. Narita, Y. Ebe, M. Naito, H. Hiraide, T. Abo. 1999. Protective effect of NK1.1+ T cells as well as NK cells against intraperitoneal tumors in mice. Cell. Immunol. 193: 219
    OpenUrlCrossRefPubMed
  31. ↵
    Kawarabayashi, N., S. Seki, K. Hatsuse, T. Ohkawa, Y. Koike, T. Aihara, Y. Habu, R. Nakagawa, K. Ami, H. Hiraide, H. Mochizuki. 2000. Decrease of CD56+ T cells and natural killer cells in cirrhotic livers with hepatitis C may be involved in their susceptibility to hepatocellular carcinoma. Hepatology 32: 962
    OpenUrlCrossRefPubMed
  32. ↵
    Tsuji, H., N. Mukaida, A. Harada, S. Kaneko, E. Matsushita, Y. Nakanuma, H. Tsutsui, H. Okamura, K. Nakanishi, Y. Tagawa, et al 1999. Alleviation of lipopolysaccharide-induced acute liver injury in Propionibacterium acnes-primed IFN-γ-deficient mice by a concomitant reduction of TNF-α, IL-12, and IL-18 production. J. Immunol. 162: 1049
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Dobashi, H., S. Seki, Y. Habu, T. Ohkawa, S. Takeshita, H. Hiraide, I. Sekine. 1999. Activation of mouse liver natural killer cells and NK1.1+ T cells by bacterial superantigen-primed Kupffer cells. Hepatology 30: 430
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section