• Open Access

Enhancement of antitumor natural killer cell activation by orally administered Spirulina extract in mice

Authors

  • Yuusuke Akao,

    1. Department of Microbiology and Immunology, Hokkaido University Graduate School of Medicine, Kita-15, Nishi-7, Kita-ku Sapporo 060-8638
    Search for more papers by this author
    • 6

      These authors contributed equally to this work.

  • Takashi Ebihara,

    1. Department of Microbiology and Immunology, Hokkaido University Graduate School of Medicine, Kita-15, Nishi-7, Kita-ku Sapporo 060-8638
    Search for more papers by this author
    • 6

      These authors contributed equally to this work.

  • Hisayo Masuda,

    1. Department of Immunology, Osaka Medical Center for Cancer, Nakamichi 1-3-2, Higashinari-ku, Osaka 537-8511
    Search for more papers by this author
    • 5

      Present address: Research and Education Center for Genetic Information, Nara Institute for Science and Technology, Ikoma, Nara 631-0101, Japan.

    • 6

      These authors contributed equally to this work.

  • Yoshiko Saeki,

    1. Department of Immunology, Osaka Medical Center for Cancer, Nakamichi 1-3-2, Higashinari-ku, Osaka 537-8511
    Search for more papers by this author
  • Takashi Akazawa,

    1. Department of Immunology, Osaka Medical Center for Cancer, Nakamichi 1-3-2, Higashinari-ku, Osaka 537-8511
    Search for more papers by this author
  • Kaoru Hazeki,

    1. The Division of Molecular Medical Science, Graduate School of Biomedical Sciences, Hiroshima University, Minami-ku, Hiroshima 734-8551, Japan
    Search for more papers by this author
  • Osamu Hazeki,

    1. The Division of Molecular Medical Science, Graduate School of Biomedical Sciences, Hiroshima University, Minami-ku, Hiroshima 734-8551, Japan
    Search for more papers by this author
  • Misako Matsumoto,

    1. Department of Microbiology and Immunology, Hokkaido University Graduate School of Medicine, Kita-15, Nishi-7, Kita-ku Sapporo 060-8638
    2. Department of Immunology, Osaka Medical Center for Cancer, Nakamichi 1-3-2, Higashinari-ku, Osaka 537-8511
    Search for more papers by this author
  • Tsukasa Seya

    Corresponding author
    1. Department of Microbiology and Immunology, Hokkaido University Graduate School of Medicine, Kita-15, Nishi-7, Kita-ku Sapporo 060-8638
    2. Department of Immunology, Osaka Medical Center for Cancer, Nakamichi 1-3-2, Higashinari-ku, Osaka 537-8511
    Search for more papers by this author

To whom correspondence should be addressed. E-mail: seya-tu@pop.med.hokudai.ac.jp

Abstract

Oral administration of hot-water extract of Spirulina, cyanobacterium Spirulina platensis, leads to augmentation of NK cytotoxicity in humans. Here, we applied to syngeneic tumor-implant mice (C57BL/6 versus B16 melanoma) Spirulina to elucidate the mechanism of raising antitumor NK activation. A B16D8 subcell line barely expressed MHC class I but about 50% expressed Rae-1, a ligand for NK activation receptor NKG2D. The Rae-1-positive population of implant B16 melanoma was effectively eliminated in the tumor mass progressed in mice. This antitumor activity was induced in parallel with IFN-γ and abolished in mice by treatment with asialoGM-1 but not CD8β Ab, suggesting the effector is NK cell. NK cell activation occurred in the spleen of wild-type mice medicated with Spirulina. This Spirulina-mediated enhanced NK activation was abrogated in MyD88 –/– mice but not in TICAM-1 –/– mice. The NK activating properties of Spirulina depending on MyD88 were confirmed with in vitro bone marrow-derived dendritic cells expressing TLR2/4. In D16D8 tumor challenge studies, the antitumor effect of Spirulina was abolished in MyD88 –/– mice. Hence, orally administered Spirulina enhances tumoricidal NK activation through the MyD88 pathway. Spirulina exerted a synergistic antitumor activity with BCG–cell wall skeleton, which is known to activate the MyD88 pathway via TLR2/4 with no NK enhancing activity. Spirulina and BCG–cell wall skeleton synergistically augmented IFN-γ production and antitumor potential in the B16D8 versus C57BL/6 system. We infer from these results that NK activation by Spirulina has some advantage in combinational use with BCG–cell wall skeleton for developing adjuvant-based antitumor immunotherapy. (Cancer Sci 2009)

The immune system has innate and acquired arms to eliminate foreign cells from the host. The innate system recognizes pathogenic microbes which possess pattern molecules serving as ligands for pattern-recognition receptors.(1) These receptors reside in immune competent cells and trigger signaling to activate transcription factors in these cells.(2) Many cytokines and cellular effectors are consequently induced to orchestrate host defense. We have a variety of foods that may contain many kinds of microbial patterns. Since intestinal and colon epithelial cells express pattern-recognition receptors,(3,4) mucosal immune responses may be modulated by oral uptake of microbial material.

The cyanobacterium Spirulina platensis (Spirulina) has been taken as a supplemental food for more than 15 years without any undesirable side-effects.(5) Its safety for human consumption has been confirmed through toxicological studies.(5) Spirulina contains a lipopolysaccharide (LPS)-like constituent that structurally differs from the bacterial LPS.(6) Spirulina also contains high contents of protein, vitamins (especially A and B12) and minerals. It is rich in phenolic acids, tocopherols and γ-linolenic acid.(5,7) Since Spirulina lacks cell walls, it is easily digested.(7) Spirulina contains unique proteins, sugars and lipids(5,8) and these moieties of Spirulina are reported to participate in raising host immune responses including enhancement of Ab production,(9) cytokine liberation,(10) T-cell response and NK activation.(11) However, the exact molecular base responsible for these immune responses has not been clearly identified yet.

We previously reported that hot-water extract of Spirulina when taken orally in adult human enhances NK activation.(11) This Spirulina activity is unique since most of the reported bacterial adjuvants facilitate CTL induction but not NK activation.(12) For example, BCG–cell wall skeleton (CWS) subcutaneously injected induces IL-12, IFN-γ and local skin erosion but no NK activation.(13) TLR2 and TLR4 on myeloid dendritic cells (DCs) sense the peptidoglycan (PGN) of BCG (BCG–PGN) and induces DC maturation via the MyD88 pathway.(13,14) BCG–PGN drives DCs to elicit a CTL-inducing state.(15)

Here we show that hot-water extract of Spirulina activates NK in mice even by oral administration. Ingestion of Spirulina confers both IFN-γ production and NK-mediated Rae-1-positive cell killing activity in mice. Consequently, Spirulina administration leads to retardation of implant tumor growth in mice. This Spirulina antitumor activity depends on MyD88 but not the other adaptor TICAM-1 (TRIF). Thus, Spirulina evokes a unique MyD88-dependent NK activation through mucosal immune responses. We offer a possible immune therapy of BCG–CWS in combination with Spirulina in this communication.

Materials and Methods

Reagents and antibodies.  The following materials were obtained: fetal calf serum (FCS) from Bio Whittaker (Walkersville, MD), mouse granulocyte–macrophage colony-stimulating factor (GM-CSF) and mouse IL-2 (mIL-2) from PeproTech EC, Ltd (London, UK), polymyxin B and lipopolysaccharide (LPS) (Escherichia coli O111:B4) from Sigma-Aldrich (St. Louis, MO), synthetic macrophage-activating lipopeptide 2 (MALP-2) from Amersham Pharmacia Biotech (Piscataway, NJ) and Lympholyte-M from Cedarlane (Ontario, Canada). Enzyme-linked immunosorbent assay (ELISA) kits for mouse (m)IFN-γ were obtained from Amersham Biosciences.

The following antibodies were used: antimouse NKG2D polyclonal Ab and antipan-Rae-1 polyclonal Ab (rabbit serum) were established in our laboratory(16) and anti-asialoGM-1 Ab was obtained from Seikagaku Kogyo Co., Ltd. or Wako Pure Chemical Industrials, Ltd (Tokyo, Japan). Monoclonal Abs against mouse CD4 and CD8β were kindly provided by Drs T Takahashi and K Tsujimura (Aichi Cancer Center, Nagoya, Japan) as previously described.(17) Fluorescein isothiocyanate (FITC)-conjugated goat antirabbit and rat IgG F(ab′)2 were obtained from American Qualex (San Clemente, CA), rat IgG1κ control FITC from BD PharMingen (San Diego, CA), and hamster IgG isotype control FITC from eBioscience (San Diego, CA).

Preparation of BCG–CWS and hot-water extract of Spirulina.  Spray-dried powder of Spirulina platensis propagated under basic conditions (pH 11) in outdoor open tanks was extracted with water in an autoclave for 1 h at 120°C. In some experiments, citric acid was added to the hot water extract to adjust the pH to 4.0.(18) The water-soluble extract was prepared by removal of insoluble fractions by centrifugation. The Spirulina extract was added to the mouse food at 1 mg/g (Clea Japan, Tokyo). In other experiments to administer accurate amounts of the extract to mice, the soluble extract of Spirulina was condensed to 1 mg/mL for oral administration as described previously.(18) In in vitro experiments, Spirulina extract was treated with polymyxin B (final concentration, 5 µg/mL) for 1 h at 37°C before cell stimulation to exclude the effect of possible contaminating LPS.

BCG–CWS was prepared in Dr I Azuma's laboratory (Research Institute for Immunology, Hokkaido University) as described previously.(19) The lot used in this study (Lot 10-2) consisted of mycolic acid, arabinogalactan and peptidoglycan with greater than 97% purity, and LPS contamination less than the detection limit (data not shown). Since BCG–CWS is insoluble in water and organic solvents, the oil-in-water emulsion form of BCG–PGN micelles (BCG-emulsion) was used throughout the in vivo study. B16 cell debris conjugated to BCG–CWS (BCG–CWS/Ag) was prepared as described below and detailed elsewhere.(15)

Mouse and cell lines.  TLR2 –/–, TLR4 –/– and MyD88 –/– mice were gifts from Dr S Akira (Osaka University, Osaka) as previously reported.(15) TICAM-1 (TRIF) –/– mice were established in our laboratory.(17) Female C57BL/6 mice were purchased from Clea Japan. Mice were maintained in our institute under specific pathogen-free conditions. All animal work was performed under guidelines established by the Osaka Medical Center Institutional Animal Care and Use Committee. Mice (12-week-old female C57BL/6) were housed four per cage and allowed food and water ad libitum. Animal studies were carefully performed without ethical problems. In some experiments, mice were fed for 3 weeks with pathogen-free food containing 1 mg/g Spirulina extract and control food with no Spirulina, which were prepared by Clea Japan. For more precise studies 600 µg/head of the Spirulina extract was directly administered to the mouse stomach via an injector.

The B16D8 cell line was established in our laboratory as a subline of the B16 melanoma cell line.(20) This subline was characterized by its low MHC levels with no metastatic properties when injected s.c. into syngeneic C57BL/6 mice.(16,21) These cell lines were cultured in RPMI 1640/10% FCS. Tumor challenge studies were performed as previously described.(15)

Preparation of BMDCs and spleen NK cells of mice.  Mouse bone marrow-derived DCs (BMDCs) were prepared as described previously.(15) Spleen NK cells were prepared by a reported method(22) with minor modifications.(23) In brief, spleen cells were passed through a nylon wool column to remove B-cells. In some cases, anti-asialoGM-1 Ab (100 µg) was injected intravenously into mice to eliminate NK cells.(17) The nylon wool-nonadherent cells were incubated in tissue culture plates for 1 h to remove adherent cells. Nonadherent cells were subjected to the MACS system (Miltenyi Biotec, Auburn, CA) for the preparation of NK cells.(23) The population was used as NK cells within 3 days.

Immunization.  On days –28, –21, –14, –1 and +7 relative to the day of B16D8 challenge (Fig. 1), 5 × 106 B16D8 cells (in 90 µL) were in freeze–thaw cycles three times in PBS to prepare ‘debris’ and the debris was mixed with 10 µL of 1 mg/mL BCG–CWS in emulsion buffer (BCG–CWS/Ag).(15) Wild-type mice (>12 weeks old) were s.c. immunized with 30 µL of this mixture per head at the base of the tail. The administration protocol is shown in a previous report.(15) As controls, either tumor debris (Ag) only or emulsion only was used. Since BCG–CWS alone in emulsion buffer possesses immune potentiation activity upon subcutaneous (s.c) administration,(15,18) it could not be used as control for BCG–CWS/Ag.

Figure 1.

AsialoGM-1 Ab-sensitive retardation of MHC-low tumor growth by oral administration of Spirulina. (a) Spirulina induces antitumor NK activation. C57BL/6 mice were separated into three groups (n = 8): one pretreated with anti-asialoGM-1 Ab and two with control saline, and except one control, fed with Spirulina from day 0, when B16 D8 subline was s.c. inoculated. Tumor volume was measured at intervals described in the Methods and statistical analysis was performed as described.(15) One mouse with saline only and one mouse given Spirulina and anti-asialoGM-1 Ab died of tumor after 5 weeks. We applied statistical analysis to seven live mice at the 5-week point of the two groups and declared the significance (P < 0.05). Represent mean ± SD. (b) CD8+ T-cells barely participate in Spirulina-mediated retardation of tumor growth. Wild-type C57BL/6 mice were grouped (n = 5). Spirulina extract (600 µg/600 µL) (inline image, inline image) or control saline (inline image) was orally administered to mice every other day from day –14. One group (inline image) was pretreated with anti-CD8β Ab as described in the text. Then, B16D8 cells (6 × 105/head) were subcutaneously inoculated into the mice at day 0. Tumor volume was measured at indicated intervals until day 27 when the mice still survived. We declared the significance (P < 0.05) through statistical analysis.(15)

Tumor challenge.  B16D8 cells (2 × 106 cells in Fig. 1) were s.c. inoculated in the hind flank of wild-type mice. One group of the mice was i.v. injected with anti-asialoGM-1 Ab (100 µg/100 µL) every week and the other was with mouse IgG. The tumor-challenged mice were fed with Spirulina-containing food from day 0 to the end of the study. In the next experiments, B16D8 cells (5 × 105 cells) were s.c. inoculated into mice, either preimmunized or non-immunized with BCG CWS/Ag. Mice were grouped into four, each consisting of 20 mice: group 1 with BCG–CWS/Ag and Spirulina, group 2 with BCG–CWS/Ag only, group 3 with Spirulina only and group 4 with no adjuvant. Likewise, mice were i.v. injected with anti-asialoGM-1 or anti-CD8β Abs (100 µg/100 µL) every week,(17) challenged with B16D8 and fed with Spirulina (600 µg/600 µL) every other day. The tumor sizes were compared with those of control mice. Tumor volumes were measured at regular intervals using a caliper. Tumor volume was calculated using the formula: Tumor volume (cm3) = (long diameter) × (short diameter) × (short diameter) × 0.4.

Statistical analysis.  A Student's t-test was used to examine the significance of the data when applicable in animal studies. Comparisons with more than two groups were done using anova with appropriate post hoc testing. Differences were considered to be statistically significant when P < 0.05.

ELISA, flow cytometric (FACS) analysis of cell surface antigens.  The levels of IFN-γ were determined by sandwich ELISA (Amersham Pharmacia Biotech, Buckinghamshire, UK) or the message levels assessed by quantitative PCR.(23) The practical methods for FACS were described previously.(24)

Quantitative RT-PCR.  BMDCs were harvested 4 h after treatment with Spirulina extracts (50 µg/mL). The total RNA was extracted by RNeasy mini kit (Qiagen, Bothell, WA). Total RNA (0.5 µg) was incubated at 70°C for 5 min and then kept on ice for 2 min, and RT was performed as described previously.(25) The following primers were used for quantitative PCR: IFN-β forward, 5′-CCAGCTCCAAGAAAGGACGA-3′, and reverse 5′-CGCCCTGT AGGTGAGGTTGA-3′; IP-10 forward 5′-GTGTTGAGATCATT GCCACGA-3′, and reverse 5′-GCGTGGCTTCACTCCAGTTAA-3′. β-Actin was used as an internal control to normalize reactions.

Assessment of in vitro cytolytic activity.  The cytolytic activity of spleen NK cells was determined by 51Cr assay as described previously.(23) Effector lymphocytes were prepared from the spleen of intact C57BL/6 mice. A B16 subline (D8) or YAC-1 was used as a target cell. Target cells (0.4 × 104 cells/well) were co-incubated with the effector splenocytes at the indicated lymphocyte to target (E/T) cell ratio (typically 1, 5 and 20) in V-bottomed 96-well plates in a total volume of 200 µL of 0.5% BSA/RPMI-1640 medium at 37°C. Four hours later, the liberated 51Cr in the medium was measured using a scintillation counter. Specific cytolytic activity was obtained by the formula: Specific cytotoxic activity (%) = [(experimental 51Cr activity – spontaneous 51Cr activity)/(total 51Cr activity – spontaneous 51Cr activity)] × 100. Each experiment was done in triplicate to confirm reproducibility of the results, and representative results are shown. A Student's t-test was used to examine the significance of the data.

Results

Tumor regression in mice given Spirulina.  Retardation of tumor growth was observed in implant B16D8 tumor in mice when given Spirulina. The suppression of tumor growth by Spirulina was abrogated if the mice were treated with anti-asialoGM-1 Ab to eliminate NK cells(17) (Fig. 1a). The effect of Spirulina on tumor regression was somewhat variable in a mouse-to-mouse fashion, but was significant (n = 8). The mouse group with Spirulina all survived 6 weeks after tumor challenge although some died in the control and NK-depleted groups by 6 weeks. The results were confirmed with additional experiments, where NK activation occurred in mice in response to being fed Spirulina and disrupted by administration of anti-asialoGM-1 Ab (data not shown). We further confirmed that depletion of CD8+ T-cells had virtually no effect on Spirulina-mediated antitumor activity (Fig. 1b). NK activation may have occurred in mice given Spirulina to retard tumor growth.

NK activation in mice given Spirulina.  We confirmed that the hot-water extract of Spirulina induces NK activation by direct NK assay. Mice (wild-type) were orally administered with the extract of Spirulina or just saline (control) every other day for 2 weeks. The NK fraction of the spleen cells was prepared by MACS beads and cultured in vitro in medium only (Fig. 2a). NK-mediated cytotoxicity was determined using an NK-target, YAC-1. NK activation was enhanced in the group with Spirulina compared to those without Spirulina (Fig. 2a). Similar results were obtained with the B16D8 cells as targets (data not shown). We next tested what molecular mechanisms participate in the Spirulina-mediated NK-enhanced activation. We used MyD88 –/– and TICAM-1 –/– mice to test the Spirulina NK-enhancing effect (Fig. 2b). In MyD88 –/–mice NK activation was not enhanced by administration of Spirulina, while in TICAM-1 –/– mice NK activation was enhanced by Spirulina as in wild-type mice. Hence, the TLR pathway involving MyD88, but not TICAM-1, participates in NK activation. NK cytotoxicity was already high in control mice given no Spirulina, irrespective of disruption of the TLR pathways (Fig. 2b), suggesting that other factors than TLRs also participate in in vivo NK activation. We found that BMDCs prepared in vitro elicit NK-enhancing activity in response to the extract of Spirulina (Fig. 2c,d). The ability of Spirulina to drive NK activation in BMDCs was not abrogated with TICAM-1 –/– BMDCs but abrogated with MyD88 –/– BMDCs (Fig. 2d). Consistent with the result that MyD88 rather than TICAM-1 is crucial for Spirulina-mediated NK activation by BMDC, TICAM-1-dependent mediators IFN-β and IP-10 were barely induced in Spirulina-stimulated BMDCs (Fig. 2e). Our interpretation of these results is that part of the in vivo NK-enhancing activity by Spirulina is attributable to the MyD88 pathway in BMDCs resulting in BMDC–NK reciprocal activation. Using this in vitro system, we tested if TLR2 and TLR4 are involved in the Spirulina NK-enhancing activity. TLR2/4-double deficient mice severely abrogated the response to Spirulina to reduce the Spirulina-mediated NK enhancing activity (Fig. 2c), although either one of TLR2 or 4 deficiency exhibited only a marginal effect on NK-enhancing activity by Spirulina. Thus, TLR2/4 and MyD88 participate in Spirulina-mediated BMDC-NK activation.

Figure 2.

Orally administered Spirulina enhances NK activation in mice. (a) YAC-1 killing activity of spleen NK cells harvested from mice with or without Spirulina. Spirulina extract (600 µg/600 µL) was orally administered into mice every other day. Two weeks later, spleen cells were harvested to isolate NK cells by MACS beads from wild-type mice. NK activity against YAC-1 or B16D8 cells (not shown) were determined at the indicated E/T ratio. (b) Participation of MyD88 in enhanced NK activation by Spirulina. NK cells were prepared from wild-type, TICAM-1 –/– or MyD88 –/– mice which had been treated with saline or Spirulina extract as in (a). NK cytotoxic activity was determined using YAC-1. (c) Spirulina effect on BMDC-NK activation was abrogated with TLR2/4 –/– BMDCs. BMDCs from wild-type and TLR2/4-double deficient mice were stimulated with Spirulina extract for 4 h and mixed with NK cells for 24 h. The mixture was incubated with 51Cr-labeled target B16 cells for 4 h at the E/T ratio indicated (left panel). In the right panel, BMDCs of the indicated KO mice were admixed with NK cells as in the left panel. NK activity was determined using 51Cr-labeled B16 target at a fixed E/T ratio (1:25). Percentage cytotoxicity was determined as in (a). Only slight abrogation of the Spirulina-mediated NK activation was observed in either TLR2 –/– or TLR4 –/– BMDCs under the same conditions (data not shown). (d) Spirulina acts on BMDC and augments MyD88-mediated NK reciprocal activation by BMDC. BMDCs with TICAM-1 –/– (left panel) and MyD88 –/– (right panel) were stimulated with Spirulina extract for 4 h and mixed with NK cells for 24 h. The mixture was incubated with 51Cr-labeled target B16 cells for 4 h at the E/T ratio indicated. Percentage cytotoxicity was determined as (a). (e) The mRNA levels of IFN-β and IP-10 in BMDCs were determined by quantitative PCR 4 h after Spirulina stimulation. These experiments were performed three times and similar results were obtained. Representative analyses are shown.

Additive tumor regression by BCG–CWS and Spirulina.  BCG–CWS is an agonist of TLR2/4 to activate the MyD88 pathway, but does not activate NK cells,(12,13) in contrast to the Spirulina extract. MyD88 may have two arms to drive CTL and NK cells in this context. BCG–CWS/Ag subcutaneously administered effectively matures antigen-presenting dendritic cells to induce tumoricidal CTL depending upon antigens selected.(15) We next checked whether BCG–CWS/Ag and Spirulina elicit a synergistic effect on tumor regression in B16D8 melanoma-bearing mice (Fig. 3a). In this experiment, tumor burden was controlled for mice to survive greater than 6 weeks after tumor challenge. Four mouse groups (n = 20 each) were made to test if the tumor suppression activity depended on the host response to BCG–CWS/Ag and/or Spirulina in the same B16 implant system. Although either BCG–CWS/Ag or Spirulina alone exerted ability to regress the implant tumor, the combination of both most effectively reduced tumor sizes (P < 0.05) (Fig. 3a). Thus, Spirulina and BCG–CWS/Ag additively suppress tumor progression. Survival rate of mice challenged with B16D8 is also shown in Fig. 3(b). The group treated with BCG–CWS and Spirulina survived the longest, since 45% of mice in this group were alive after 9 weeks, by the time more than 80% of control mice died. Mice treated with anti-asialoGM-1 Ab all died by 9 weeks (Fig. 3b), suggesting the importance of NK cells for antitumor immune response and long survival.

Figure 3.

Implant tumor retardation by BCG–CWS and Spirulina in mice. (a) C57BL/6 mice were separated into four groups (n = 20): group 1 with saline only, group 2 with BCG–CWS/B16 Ag, group 3 with Spirulina and group 4 with BCG–CWS/Ag and Spirulina. BCG–CWS/Ag was administered four times from day –21 to day 0,(15) while Spirulina was fed from day –21 to the end of the survey (see inset). B16D8 subline was s.c. inoculated at day 0. Tumor volume was measured at intervals and statistical analysis was performed as described.(15) (b) Mice were grouped into four (n = 20) as indicated in (a), and the effect of asialoGM-1 Ab on the survival of tumor-implant mice was analyzed. The survival rate was plotted for each group of mice.

Levels of IFN-γ in Spirulina/BCG–CWS-treated mice.  IFN-γ is an effector that may be associated with prognosis of patients with BCG–CWS adjuvant immunotherapy.(26) IFN-γ was secreted in the blood of mice with tumor burden if they were treated with BCG–CWS and Ag (Fig. 4). Either BCG–CWS/Ag or Spirulina alone slightly induced IFN-γ. The levels of IFN-γ induced by Spirulina appeared low in mice compared to human volunteers.(11) Tumor antigen had no effect on the level of Spirulina-mediated IFN induction (data not shown). Notably, significantly high levels of IFN-γ were detected in mice treated with both BCG–CWS/Ag and Spirulina (Fig. 4). Skin reaction reflecting BCG hypersensitivity was observed in the relevant group (data not shown). No or less skin reaction was observed in the group with BCG–CWS/Ag and Spirulina. Spirulina alone did not induce skin reaction.

Figure 4.

Serum level of IFN-γ in tumor-bearing mice given BCG–CWS and/or Spirulina. Mice were treated with BCG–CWS/B16 Ag and/or Spirulina and inoculated with B16 tumor at day 0 as in Fig. 3. Forty-eight hours after tumor implantation, mouse sera were obtained from their tails and the levels of IFN-γ were determined by ELISA.

Rae-1-positive B16 cells were eliminated by Spirulina.  B16D8 cells in implant tumor of C57BL/6 mice consisted of Rae-1-positive and -negative cells. The cells were inoculated into the mice which were fed and/or treated with the material indicated (Fig. 5). Implanted B16 tumor cells were extracted from the tumor-bearing mice 5 weeks after tumor challenge, stained with the indicated Abs and analyzed by FACS (Fig. 5), where mean fluorescence intensities are shown in the insets. In the group given no Spirulina, tumor cells consisted of Rae-1-positive and -negative populations, whereas in the group given Spirulina, the Rae-1-positive population was selectively diminished. The group treated with Spirulina and anti-asialoGM-1 Ab possessed the Rae-1-positive population. Other markers including MHC class I and class II were neither detectable nor different among the groups. Although other NK-activation ligands were detected in message levels, they still remained in the Rae-1-negative cells (data not shown). Thus, Rae-1-positive tumor cells are selectively eliminated by Spirulina-derived NK cells in this mouse system.

Figure 5.

Rae-1 and MHC levels in the implant tumor of mice. Mice were treated with anti-asialoGM-1 Ab, Spirulina, B16 tumor Ag and/or BCG–CWS as in Fig. 3, and inoculated with B16 cells; prescriptions are indicated in the left column. Five weeks later, the cells were harvested and dispersed in PBS–EDTA. The levels of Rae-1, MHC class I, MHC class II and Qa-1b were assessed by FACS using their specific Abs.(16) Specific mean fluorescence intensities of Rae-1 are indicated in the fluorograms as described elsewhere.(16)

This issue was confirmed using an in vitro assay. NK cells reciprocally activated by Spirulina-treated BMDCs damaged B16D8 cells (Fig. 6a). The NK-mediated B16D8 killing was largely blocked by the addition of anti-NKG2D Ab (Fig. 6a). Without BMDCs, Spirulina extract treatment barely activated NK cells (data not shown). Thus, NK-mediated B16D8 killing was attributable to interaction between tumor cell Rae-1 and NK cell NKG2D, which is supported by Spirulina-dependent maturation of BMDC, as there is no direct route for Spirulina-mediated NK activation.

Figure 6.

Spirulina NK cells damage B16D8 tumor cells through MyD88 and NKG2D receptor. (a) NK cells kill B16D8 cells through NKG2D after incubation with Spirulina-treated BMDCs. BMDCs from wild-type mice were stimulated with Spirulina extract for 4 h and mixed with NK cells for 24 h (DC:NK = 1:3). The mixture was incubated with anti-NKG2D Ab or control IgG and 51Cr-labeled target B16 cells for 4 h at the E/T ratio indicated. (b) MyD88 is crucial for Spirulina-mediated tumor growth retardation. Wild-type and MyD88 –/– mice were grouped as shown. Spirulina extract (600 µg/600 µL) (inline image, inline image) or control saline (inline image, inline image) was orally administered to wild-type and MyD88 –/– mice (n = 5) every other day from day –14. B16D8 cells (6 × 105/head) were subcutaneously inoculated into the mice at day 0. Tumor volume was measured at indicated timed intervals until day 27 when the mice survived, and statistical analysis was performed as in Fig. 1 (P < 0.05).

Spirulina-mediated B16D8 growth was largely abrogated in MyD88 –/– mice (Fig. 6b). Antitumor NK activation predicted in this study may occur in tumor-bearing mice through being given Spirulina.

Discussion

We demonstrated that tumor growth is retarded in mice given Spirulina. Unique points in this study are: (i) Rae-1-positive tumor cells are selectively eliminated in Spirulina-fed mice; and (ii) marked tumor regression is observed in mice with combinational administration of BCG–CWS/Ag and Spirulina. Synergistic effects on tumor regression and increase of serum IFN-γ level were observed in mice by combination treatment with BCG–CWS/Ag and Spirulina. BCG–CWS/Ag induces CTL via TLR2/4 in myeloid DCs in a MyD88-dependent manner.(15) In this study, Spirulina activates NK cells also via TLR2/4 in BMDCs in a MyD88-dependent manner. These reports suggest that simultaneous attack by CTL and NK more effectively regresses tumor cells in mouse tumor implant models.

Our in vitro findings on Spirulina allowed us to interpret that the Spirulina extract activates the MyD88 pathway in BMDCs to evoke activation of mouse NK cells (Fig. 2). What mechanisms activate the MyD88 pathway to drive NK cells in Spirulina-stimulated BMDCs is an intriguing issue, since other TLR2/4 ligands mainly engage CTL induction, but not NK activation, via the MyD88 pathway in BMDCs.(27,28) Difference in downstream signal events may cause the different outcomes of NK/CTL induction in BMDCs. Studying the Spirulina-activated MyD88 pathway will be an important issue to clarify how it is possible to have two TLR2/4 agonists activate NK and CTL.

There appear several routes for activation of NK cells. The BMDC-mediated NK activation serves an important route for NK-mediated tumor clearance.(29–31) We demonstrated that the TICAM-1 (TRIF) pathway in BMDCs participates in the DC–NK reciprocal activation.(17,27,32) In contrast, most of the non-DNA/RNA adjuvants currently available originate from bacteria and activate the MyD88 pathway of TLR2 and/or TLR4 in DCs, but they barely induce NK activation.(17,28) In this context, BCG–CWS expressed high tumor-killing activity (Fig. 3a), but this activity was barely abrogated by administration of anti-asialoGM-1 Ab (data not shown). Further, Spirulina appears to differ from LPS in its adjuvant activity, since LPS simultaneously activates the MyD88 and TICAM-1 pathways to evoke the systemic cytokine storm. The Spirulina extract is unique because it has an ability to activate NK cells without inducing type I IFNs.

Spirulina-mediated IFN-γ production is greatly enhanced in mice by simultaneous administration of BCG–CWS/Ag, similar to humans.(33) How Spirulina participates in IFN-γ production is a future question. Spirulina ultimately activates T- and NK cells in mice, although IFN-γ is poorly induced in mice compared to human volunteers with only Spirulina.(11) Direct addition of Spirulina to splenic lymphocytes neither results in NKG2D up-regulation nor induces enhanced B16D8 cell killing (data not shown), suggesting that cell populations other than DCs in the spleen barely participate in this event. There is a report that the MAPK pathway of BMDCs participates in promotion of NK activation by BMDCs.(34) If this is the case, the NK-enhancing effect by Spirulina in human patients can be evaluated in the mouse model and IFN-γ production is a good marker for NK activation.

Previous studies on patients with cancer suggested that Spirulina constituents serve as protective agents against oral cancers(35) in terms of tumor progression and metastasis,(36) although the results are from non-randomized trials. These reports were reminiscent of the antitumor functions of Spirulina, some of which would be derived from β-carotene, an effective antioxidant.(37) Indeed, the serum levels of β-carotene are consistently low in patients with cancer.(37,38) Our present study may offer additional evidence that NK activation is a representative of the Spirulina-mediated antitumor immunity.

However, a point remains unsettled that the constituents of Spirulina taken up via the intestine and colon do not always correspond to those of the extract of Spirulina. In vitro studies where Spirulina was added to NK cells or BMDCs appear not to be simply comparable with the in vivo studies using mice. In other reports, Spirulina orally taken significantly reduced IL-4 levels in individuals with allergic rhinitis by a randomized double-blinded trial.(39) Another report suggested that Spirulina enhances IgG1 and IgA production but not IgE production by modulating the mucosal immune system.(9) Spirulina may skew the Th2 polarization to a Th1-like state in allergic patients. We should identify the Spirulina constituents associated with activation of mucosal immunity and absorbed into the circulation.

Since cancers are usually established by circumventing the host immune system including NK and CTL, tumor cells generally possess poor immunogenicity by expressing low levels of MHC class I and ligands for NK-activating receptors including NKG2D.(40) It is notable that these receptor levels are regulated by IFN-γ. It has been reported that during BCG adjuvant therapy the increased serum level of IFN-γ 18 h after s.c. injection of BCG–CWS is a marker for evoking the innate immune response.(26) In mouse experimental models using syngeneic transplantable tumors, MHC class I-expressing tumor cells were selectively eliminated by BCG–CWS/Ag s.c. injection,(15) and the serum levels of IFN-γ were increased in the case of Spirulina, too. Nevertheless, tumor cells with low MHC expression remain, resulting in MHC-negative tumor progression. This study clearly shows that tumor growth is suppressed in an MHC-negative/Rae-1-positive population of tumor cells by treatment of tumor-implant mice with Spirulina, which can induce NK activation to damage tumor cells via NKG2D receptors.

In human cancer patients receiving BCG–CWS therapy, tumor cells have not completely disappeared from the primary region,(41) although patients’ quality of life (QOL) scores are increased in response to the BCG–CWS therapy. Similar observations were reported in patients with bladder cancer who selected BCG immunotherapy.(42) Growing the low-MHC tumor cells may account for the incomplete remission of tumors in patients with cancer.

We offer the possible immune therapy in combination with BCG–CWS and Spirulina in this communication. Additive tumor cytotoxicity based on BCG–CWS/Ag and Spirulina suggests that they elicit different effectors, putative CTL and NK. Their combinational function merely targets the MyD88 pathway and is clearly distinct from that of LPS that induces toxic shock. Although which constituents of the Spirulina extract are responsible for NK enhancement, and why Spirulina and BCG–CWS differentially activate the TLR2/4-mediated MyD88 pathway in DCs should be further investigated, this is the fist report predicting that the combination of BCG–CWS/Ag and Spirulina is applicable to immunotherapy for patients with tumor mass of variable MHC levels. We favor the premise that Spirulina is a candidate for NK activator applicable to cancer patients by oral usage.

Acknowledgments

This work was supported in part by CREST, Japan Science and Technology Corporation, by Grants-in-Aid from the Ministry of Education, Science, and Culture (Specified Project for Advanced Research) and the Ministry of Health, Labor, and Welfare of Japan, by the Akiyama Foundation and Yakulto Foundation. Financial support by the Sapporo Biocluster ‘Bio-S’ Knowledge Cluster Initiative of the MEXT, and the Program of Founding Research Centers for Emerging and Re-emerging Infectious Diseases, MEXT are gratefully acknowledged. We are grateful to Drs K Funami, M Shingai, M Sasai, A Matsuo, H Shime and H Oshiumi for their critical discussions and Ms Hatsugai and Ms Sato for technical support. Thanks also to Drs S Akira (Osaka University, Osaka) and T Takahashi (Aichi Cancer Center, Nagoya) for providing knockout mice and specific Abs, respectively.

Ancillary