Female C3H/HeN and C3H/HeJ mice were purchased from CLEA Japan Inc. (Tokyo, Japan). DO11.10 transgenic mice for αβ T-cell receptor (TCR) recognizing ovalbumin (OVA) in the context of I-A^d and transgenic mice carrying an activated rat HER-2/neu oncogene driven by a mouse mammary tumor virus promoter (HER-2/neu mice) were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and Charles River Laboratories (Cambridge, MA, USA), respectively. All mice were maintained in a pathogen-free environment, and experiments were performed following the ethical guidelines of Kochi Medical School and Osaka Ohtani University. The mouse tumor MH134 (hepatocellular carcinoma; kindly provided by Dr T. Kudo, Tohoku University, Sendai, Japan) and YAC-1 (T-cell lymphoma) cell lines were maintained in RPMI-1640 medium (Sigma-Aldrich, St. Louis, MO, USA) with 10% heat-inactivated fetal calf serum (FCS, HyClone Laboratories, Logan, UT, USA), 5×10⁻⁵ M 2-mercaptoethanol (2-ME), and 50 µg/ml gentamicin (Sigma-Aldrich).

LPS from Escherichia coli (E. coli) 0111:B4 was purchased from Difco (Detroit, MI, USA). Anti-IFN-γ mAb (R4-6A2, rat IgG1; no. MM701), anti-IL-17 mAb (50104, rat IgG2a; no. MAB421), anti-CD8 mAb (53–6.7, rat IgG2a; no. 100735), and rat IgG were obtained from Endogen (Rockford, IL, USA), R&D Systems (Minneapolis, MN, USA), BioLegend (San Diego, CA, USA), and Sigma-Aldrich, respectively. Anti-CD4 mAb (GK1.5, rat IgG2b) was kindly provided by Dr F.W. Fitch (University of Chicago, Chicago, IL, USA).

S. pacifica was a generous gift from Dr Genrald Cysewski (Cyanotech Corporation, Kailua-Kona, Hawaii, USA) and Mr. Nobuyuki Miyaji (Toyo Koso Kagaku Co., Ltd., Chiba, Japan). S. pacifica has been selected from a strain of edible S. platensis in 1984 and its enzyme expression profile differs from that of the parental strain. LPS was prepared from S. pacifica freeze-dried cells as previously described (27). Briefly, cells were washed with acetone, suspended in distilled water, and then extracted by addition of 90% phenol-water and vigorous agitation at 68°C. The crude preparation was dialyzed to remove phenol and residual freeze-dried cells. The sample was dissolved in water, and the insoluble material was eliminated by centrifugation, followed by ultracentrifugation at 100,000 × g. The molecular mass of the LPS sample was estimated between approximately 1,000 and 20,000 Da by electrophoresis and mass spectral analysis.

Tumor growth was measured 3 times per week after intradermal (i.d.) injection of 1×10⁶ MH134 cells in the back of C3H/HeN or C3H/HeJ mice. The tumor volume was calculated using the following formula: Volume (mm³) = width² × length/2. In some experimental settings, Spirulina LPS or E. coli LPS in saline solution was injected intraperitoneally (i.p.) every week starting 6 days after tumor inoculation. To deplete the T-cell subsets, mice injected with MH134 tumor cells on day 0 were injected i.p. with rat IgG, anti-CD4, or anti-CD8 mAb (150 µg/mouse on days, −1, 0, +3) as previously described (28). T-cell depletion was confirmed to be >95% by fluorescence-activated cell sorting (FACS).

MH134 tumors were surgically removed 3 weeks after tumor inoculation. Mice were re-challenged i.d. with 1.5×10⁶ MH134 cells of the same tumor as previously described (29).

MH134 tumors taken from C3H/HeN mice treated with saline solution or Spirulina LPS 22 days after tumor implantation were embedded in O.C.T. compound (Sakura Finetec USA, Inc., Torrance, CA, USA) and frozen. Frozen tumors were sectioned and stained with anti-CD4 (RM4-5, rat IgG2a, no. 100520, BioLegend) or anti-CD8 (53–6.7) antibodies using simple stain mouse MAX-PO [F(ab)'₂ goat anti-rat Ig and peroxidase coupled to the amino acid polymer] and 3,3′-diaminobenzidine according to the manufacturer's protocol (Nichirei-Biosciences Inc., Tokyo, Japan).

Mϕ/DC or CD4 T cell fractions were prepared from whole spleen cells by positive selection using a MACS cell separation system (Miltenyi Biotec, Auburn, CA, USA) according to manufacturer's instructions. Anti-CD4 and a mixture of anti-CD11b and anti-CD11c microbeads were used for the fractionation of CD4 T cells and Mϕ/DC, respectively. The purity was usually demonstrated to be >90% by FACS.

Whole splenocytes (5×10⁶/well) depleted of red blood cells were cultured in 10% FCS RPMI-1640 medium in 24-well culture plates (Becton Dickinson Labware, Franklin Lakes, NJ, USA) at 37°C in a 5% CO₂ humidified atmosphere. After 4 days of culture, supernatants were collected to assess cytokine levels by ELISA.

Whole spleen cells (8×10⁵/well) from OVA-specific TCR transgenic (DO11.10) mice were cultured in flat-bottomed 96-well plates (Costar Corning, NY, USA) in 10% FCS RPMI-1640 medium for 5 days. In some experiments, 2.8×10⁵ CD4 T cells, prepared from C3H/HeJ mice that had been immunized i.p. with 150 µg OVA and 5 mg Alum approximately 2 months earlier, were cultured with 1.2×10⁵ splenic Mϕ/DC cells from C3H/HeN or C3H/HeJ mice in the presence of 20 µg/ml OVA for 5 days. Spirulina or E. coli LPS was added at the initiation of the culture to evaluate its effect on IL-17 production. The levels of IL-17 in the culture supernatants were evaluated by ELISA.

Whole spleen cells (8×10⁵/well) were cultured in flat-bottomed 96-well plates in 10% FCS RPMI-1640 medium in the presence or absence of graded doses of Spirulina LPS or E. coli LPS. Five days later, the cultured cells were collected and stained with FITC-conjugated anti-B220 mAb (RA3-6B2, rat IgG2a, no. 553088, BD Biosciences, San Jose, CA, USA). B220⁺ cells were counted as B cells using a FACSCalibur instrument (Becton Dickinson, San Jose, CA, USA).

Briefly, 2.5×10⁶ target cells were labeled with 25 µCi of ⁵¹Cr sodium chromate for 60 min at 37°C in 10% FCS RPMI-1640 medium. After washing, ⁵¹Cr-labeled target cells (1×10⁴) and effector cells were mixed in flat-bottomed 96-well plates at the indicated effector/target (E/T) ratio. After 4 h of incubation, the radioactivity in the cell-free supernatants was measured using a 1470 Automatic Gamma Counter (PerkinElmer, Waltham, MA, USA). Percentage-specific lysis for ⁵¹Cr release was calculated according to the following formula: % specific lysis = [(experimental - spontaneous) release] / [(maximal - spontaneous) release] ×100.

Cytokine levels in serum or culture supernatants were quantified by sandwich ELISA. The following pairs of capture and biotinylated detection rat anti-mouse mAbs were used: R4-6A2 (no. 551216) and XMG1.2 (no. 554410) for IFN-γ, TC11-18H10 (no. 555068) and TC11-8H4 (no. 555067) for IL-17, 9A5 (no. 554658) and C17.8 (no. 554476) for IL-12(p35/40), MP5-20F3 (no. 554400) and MP5-32C11 (no. 554402) for IL-6, and A75-2 (no. 555052) and A75-3 (no. 555053) for TGF-β1 (all were purchased from BD Biosciences). For IL-23(p19/40), rat anti-mouse IL-23p19 (G23-8, no. 14-7232-85, eBioscience, San Diego, CA, USA) and biotinylated anti-IL-12p40 (C17.8, BD Biosciences) were used. The ELISA assays were performed according to the manufacturer's instructions.

Differences in mean values between groups were calculated using an unpaired two-tailed Student's t-test, Mann-Whitney U test, or Fischer's exact test. P-values of the Student's t tests are shown unless otherwise indicated.

To examine the antitumor activities of Spirulina LPS in comparison with that of E. coli LPS, we inoculated hepatocellular carcinoma MH134 cells i.d. into syngeneic C3H/HeN and TLR4 mutant C3H/HeJ mice (30), followed by i.p. administration of different doses of Spirulina LPS or E. coli LPS 6 days later. The injection with different doses of Spirulina LPS suppressed the tumor growth in C3H/HeN but not in C3H/HeJ mice, to the same degree as E. coli LPS treatment (Fig. 1A). This suggests that Spirulina LPS as well as E. coli LPS reduced tumor growth in a TLR4-dependent manner.

In order to assess the involvement of T cells in the Spirulina LPS-induced antitumor effect, we administered anti-CD4 and/or anti-CD8 mAbs to deplete the T-cell subsets before Spirulina LPS injection. While injection of both anti-CD4 and anti-CD8 mAbs completely abolished the antitumor activity of Spirulina LPS, anti-CD4 or anti-CD8 mAb alone did not result in a strong effect (Fig. 1B). In accordance with the results of the in vivo T-cell depletion, immunohistochemical examinations revealed the enhancement of infiltration of both CD4 and CD8 T cells in the tumor masses upon administration of Spirulina LPS (Fig. 1C). These results suggest that CD4 and CD8 T cells are both involved in the antitumor effect induced by Spirulina LPS.

To examine whether Spirulina LPS induces immunity against MH134 tumors, we reinoculated MH134 cells into C3H/HeN mice that had been implanted with an MH134 tumor and were treated with saline, E. coli, or Spirulina LPS, followed by surgical resection of primary tumors 5 days before the re-challenge. The growth rate of the reimplanted tumors reduced even in saline-treated mice, compared with that in untreated mice implanted with only new tumor cells without the first inoculum of MH134 tumor cells (Fig. 2A), implicating the induction of antitumor immunity without LPS administration when the primary tumor was removed. Spirulina LPS induced a stronger resistance to reimplanted MH134 tumors than saline (Fig. 2A). In contrast, the tumor growth rate in E. coli LPS-treated mice was comparable to that in saline-treated mice (Fig. 2A). These results suggest that while Spirulina LPS facilitated the generation of immunity against MH134 tumors, E. coli LPS did not enhance secondary antitumor immune response. Since E. coli LPS was effective in the prevention of primary tumor growth as shown in Fig. 1A and is known to exhibit antitumor activity partly through activation of NK cells, we evaluated the ability of E. coli and Spirulina LPS to activate NK cells. Indeed, spleen cells from C3H/HeN mice injected with E. coli LPS showed remarkable toxicity towards NK-sensitive YAC-1 cells and a weak but significant toxicity towards MH134 cells, whereas Spirulina LPS failed to activate NK cells (Fig. 2B). NK activation was not induced in C3H/HeJ mice, not even when E. coli LPS was administered (Fig. 2B). Thus, this result may explain why E. coli LPS elicited antitumor effects against the primary tumor, regardless of its failure to enhance adaptive immunity to tumors.

IFN-γ plays a crucial role in the prevention of tumor development (6), whereas IL-17 and IL-23 are considered to promote tumor growth by inducing inflammation and by regulating the expansion/survival of Th17 cells, respectively (11,14–16). We measured serum levels of IFN-γ, IL-17, and IL-23 in tumor-bearing C3H/HeN and C3H/HeJ mice treated with saline, E. coli, or Spirulina LPS. E. coli and Spirulina LPS (100 µg) markedly increased serum IFN-γ levels in C3H/HeN mice with a peak response at days 7 and 14 after tumor inoculation, respectively (Fig. 3A). A low dose of Spirulina LPS (20 µg) induced slight but significant IFN-γ production from days 7 to 14. However, the serum levels of IFN-γ were not elevated in C3H/HeJ mice at any time, not even after injection of either LPS. The treatment with anti-CD4 mAb or a combination of anti-CD4 and anti-CD8 mAbs abrogated the increase in serum IFN-γ in C3H/HeN mice receiving Spirulina LPS, whereas anti-CD8 mAb alone slightly diminished the activity of Spirulina LPS to induce IFN-γ (Fig. 3D). On the other hand, serum levels of IL-17 and IL-23 in saline-treated C3H/HeN mice gradually increased during tumor progression and reached a maximum on day 21 (Fig. 3B and C). However, serum levels of IL-6 only showed an approximately 2-fold increase in saline-treated tumor-bearing mice even on day 21 as compared to non-treated mice (data not shown). Noteworthy, Spirulina LPS significantly reduced IL-17 and IL-23 levels, while E. coli LPS hardly suppressed IL-17 production on day 21 (Fig. 3B and C). On the contrary, E. coli LPS enhanced serum levels of IL-23 in tumor-bearing C3H/HeN mice on day 7 and 14. Taken together, these results support the notion that Spirulina LPS induces antitumor immune responses through the induction of IFN-γ mostly by CD4 T cells and suppressed serum levels of IL-17 and IL-23, whereas E. coli LPS exerts its antitumor effect primarily through activation of NK cells.

We have previously shown that antigen-presenting cells (APCs)-expressing tumor Ag and tumor-reactive T cells are both stimulated in vivo in tumor-bearing mice, and that in vitro culture of spleen cells from tumor-bearing mice at early stages leads to cytokine production without exogenous addition of tumor Ag as a result of the collaboration between antitumor T cells and APCs (29,31). We tested whether in vivo treatment with E. coli or Spirulina LPS affects the in vitro production of IFN-γ by culturing spleen cells from MH134 tumor-bearing mice. Spirulina LPS clearly enhanced IFN-γ production in spleen cells of tumor-bearing C3H/HeN mice, whereas E. coli LPS was unable to upregulate IFN-γ production (Fig. 3D). These results implicate that Spirulina LPS enhances IFN-γ production through the generation of memory T cells.

Since Spirulina LPS reduced serum IL-17 levels in tumor-bearing mice while increasing IFN-γ levels and because IFN-γ negatively regulates the generation of Th17 cells (32,33), we investigated whether IL-17 downregulation by Spirulina LPS occurs via inhibition of the induction of IL-17-producing cells by facilitating IFN-γ production, or alternatively by directly downregulating IL-17-producing cells. To address this question, spleen cells from OVA-specific TCR transgenic DO11.10 mice were stimulated by OVA with or without E. coli or Spirulina LPS and in the presence or absence of anti-IFN-γ mAb. Spirulina LPS significantly suppressed IL-17 production by DO11.10 spleen cells in a dose-dependent manner, regardless of the presence of anti-IFN-γ mAb. However, E. coli LPS augmented IL-17 production (Fig. 4A), possibly because of its ability to induce IL-6 (Fig. 5A). In addition, it is noteworthy that anti-IFN-γ mAb enhanced IL-17 production by DO11.10 spleen cells in response to OVA without the addition of either LPS (Fig. 4A), indicating that IFN-γ regulates IL-17-producing cells. Thus, Spirulina LPS suppresses the in vivo production of IL-17 with or without involvement of IFN-γ. Moreover, although Spirulina LPS inhibited IL-17 production when CD4 T cells from OVA-primed C3H/HeJ mice were co-cultured with C3H/HeN APCs in the presence of OVA, it failed to suppress IL-17 production when OVA-primed C3H/HeJ CD4 T cells were stimulated with OVA in the context of C3H/HeJ APCs (Fig. 4B). These results suggest that Spirulina LPS acted on APCs to inhibit the generation of Th17 cells in a TLR4-dependent manner.

Thus far, our results are consistent with the notion that the antitumor effect of Spirulina LPS is caused by the downregulation of IL-17 production. To examine whether neutralization of IL-17 by anti-IL-17 mAb would result in a reduction of tumor growth in mice in the absence of Spirulina LPS, we injected anti-IL-17 or anti-IFN-γ mAb into C3H/HeN mice 1 day before and 4 days after MH134 tumor implantation, and monitored tumor development. As expected, anti-IL-17 mAb markedly suppressed tumor growth compared to control rat IgG antibodies, whereas anti-IFN-γ mAb slightly enhanced tumor development (Fig. 4C). Importantly, mice receiving anti-IL-17 mAb exhibited high levels of serum IFN-γ (Fig. 4D). In addition, in vitro culture of spleen cells from tumor-bearing mice treated with anti-IL-17 mAb resulted in the production of large amounts of IFN-γ (Fig. 4D). These results indicate that IL-17 provides an environment suitable for tumor growth partly by inhibiting the generation of IFN-γ-producing T cells.

It was recently reported that IL-17 is mainly produced by Th17 cells, and that both IL-6 and TGF-β are indispensable for the generation of Th17 cells (20). To test the ability of Spirulina LPS to induce IL-6 and TGF-β, normal C3H/HeN and C3H/HeJ mice were injected i.p. with E. coli or Spirulina LPS and serum cytokine levels were measured 4 h later. C3H/HeN mice produced high levels of IL-6 in response to E. coli LPS, but showed only a small response upon stimulation with Spirulina LPS (Fig. 5A). Contrary to IL-6, the substantial TGF-β levels present in serum of untreated groups of both strains were not significantly elevated when stimulated with E. coli LPS (Fig. 5A). However, Spirulina LPS slightly reduced TGF-β levels, but only in C3H/HeN mice (Fig. 5A). Noteworthy, E. coli LPS elicited a considerable increase in IL-23 levels, only in C3H/HeN mice, while Spirulina LPS showed almost no induction of IL-23 even in C3H/HeN mice (Fig. 5A). Thus, Spirulina LPS seems to be inferior to E. coli LPS in terms of stimulating immune cells. The possibility that Spirulina LPS has a general defect in the stimulation of immune system was excluded, because both E. coli and Spirulina LPS induced B cell proliferation in a dose-dependent manner (Fig. 5B).

Since IL-12 and IFN-γ are known to play a crucial role in the differentiation of IFN-γ-producing Th1 cells (4), we measured the serum levels of IL-12 in tumor-bearing mice that were treated with saline, E. coli, or Spirulina LPS. Spirulina LPS augmented IL-12 production more than E. coli LPS. In contrast to a transient increase of IL-12 by E. coli LPS, the enhanced IL-12 production by Spirulina LPS was still observed after 14 days (Fig. 5C). Of note, these high IL-12 serum levels on day 14 decreased when C3H/HeN mice were injected with anti-CD4, but not anti-CD8 mAb (Fig. 5D), implicating CD4 T cell-dependent IL-12 production. Taken together, these findings support the notion that Spirulina LPS facilitates the priming of CD4 T cells with tumor cells and the subsequent CD4 T cell-dependent activation of APCs, leading to IL-12 production, which in turn induces IFN-γ-producing T cells.

We finally examined whether Spirulina LPS is also effective in suppressing the spontaneous development of mammary tumors in female transgenic mice carrying the activated HER-2/neu oncogene. HER-2/neu transgenic mice display apparent hyperplasia in the mammary glands at 10 weeks of age and develop palpable mammary tumors around 24 weeks (28). Female HER-2/neu mice were injected with Spirulina LPS or PBS once per week between 120 and 240 days after birth. Spirulina LPS delayed the appearance of tumors and significantly reduced both tumor incidence and growth (Fig. 6).