Functional alterations of liver innate immunity of mice with aging in response to CpG-oligodeoxynucleotide†
Article first published online: 16 JUL 2008
Copyright © 2008 American Association for the Study of Liver Diseases
Volume 48, Issue 5, pages 1586–1597, November 2008
How to Cite
Kawabata, T., Kinoshita, M., Inatsu, A., Habu, Y., Nakashima, H., Shinomiya, N. and Seki, S. (2008), Functional alterations of liver innate immunity of mice with aging in response to CpG-oligodeoxynucleotide. Hepatology, 48: 1586–1597. doi: 10.1002/hep.22489
Potential conflict of interest: Nothing to report.
- Issue published online: 28 OCT 2008
- Article first published online: 16 JUL 2008
- Accepted manuscript online: 16 JUL 2008 12:00AM EST
- Manuscript Accepted: 11 JUN 2008
- Manuscript Received: 17 JAN 2008
- National Defense Medical College
Immune functions of liver natural killer T (NKT) cells induced by the synthetic ligand α-galactosylceramide enhanced age-dependently; hepatic injury and multiorgan dysfunction syndrome (MODS) induced by ligand-activated NKT cells were also enhanced. This study investigated how aging affects liver innate immunity after common bacteria DNA stimulation. Young (6 weeks) and old (50-60 weeks) C57BL/6 mice were injected with CpG oligodeoxynucleotides (CpG-ODN), and the functions of liver leukocytes were assessed. A CpG-ODN injection into the old mice remarkably increased tumor necrosis factor (TNF) production in Kupffer cells, and MODS and lethal shock were induced, both of which are rarely seen in young mice. Old Kupffer cells showed increased Toll-like receptor-9 expression, and CpG-ODN challenge augmented TNF receptor and Fas-L expression in liver NKT cells. Experiments using mice depleted of natural killer (NK) cells by anti-asialoGM1 antibody (Ab), perforin knockout mice, and mice pretreated with neutralizing interferon (IFN)-γ Ab demonstrated the important role of liver NK cells in antitumor immunity. The production capacities of old mice for IFN-γ, IFN-α, and perforin were much lower than those of young mice, and the CpG-induced antitumor cytotoxicity of liver NK cells lessened. Lethal shock and MODS greatly decreased in old mice depleted/deficient in TNF, FasL, or NKT cells. However, depletion of NK cells also decreased serum TNF levels and FasL expression of NKT cells, which resulted in improved hepatic injury and survival, suggesting that NK cells are indirectly involved in MODS/lethal shock induced by NKT cells. Neutralization of TNF did not reduce the CpG-induced antitumor effect in the liver. Conclusion: Hepatic injury and MODS mediated by NKT cells via the TNF and FasL-mediated pathway after CpG injection increased, but the antitumor activity of liver NK cells decreased with aging. (HEPATOLOGY 2008.)
Unmethylated CpG oligodeoxynucleotide (CpG-ODN) motifs (GACGTT for mice and GTCGTT for humans) in bacterial DNA have a potential to strongly activate the innate immune system.1, 2 Unmethylated CpG motifs are common in bacterial DNA, whereas they are underrepresented and methylated in vertebrate DNA.3 This difference in the frequency of unmethylated CpG dinucleotides between bacterial and vertebrate DNA provides a structural characteristic through which vertebrate immune cells may detect and respond to bacterial infection. CpG mimics the stimulatory effect of DNA of either gram-negative or gram-positive bacteria.1, 4 The Toll-like receptor (TLR) family is a phylogenetically conserved mediator of innate immunity that is essential for microbial recognition.5 Lipopolysaccharide (LPS), a gram-negative bacterial component, and bacterial DNA (or CpG) engage TLR46 and TLR9,7 respectively. A CpG challenge, similarly to an LPS challenge, preferentially elicits a T helper 1 immune response in the host1, 4, 8, 9 and thereby induces host immune cells, including macrophages and NK cells, to produce proinflammatory cytokines such as tumor necrosis factor (TNF) and interferon (IFN)-γ. These activities are indispensable for efficiently controlling the growth and dissemination of invading bacteria and may be harnessed therapeutically in antitumor agents.10, 11 However, excessive and uncontrolled production of inflammatory cytokines caused by bacterial infections is potentially harmful to the host and may lead to severe systemic inflammatory complications such as multiorgan dysfunction syndrome (MODS), septic shock, and death.12, 13
NKT cells activated either by interleukin (IL)-12 or α-galactosylceramide (α-GalCer) produce IFN-γ and play an important role in the host defense against tumors and infections.14–20 On the other hand, we recently reported that the generalized Shwartzman reaction by consecutive IL-12 and LPS injections, which is a model of septic shock and MODS, is enhanced in mice and human beings with age.21, 22 We also reported that α-GalCer induces age-dependent increases of TNF levels and FasL expression of NKT cells,19 which caused MODS and a high mortality in the aged mice.19, 23 However, we have recently found that α-GalCer–activated NKT cells conversely accelerate the mitosis of newly regenerating hepatocytes TNF/FasL dependently as evidenced in mice after partial hepatectomy,24 suggesting that NKT cells normally regulate turnover/homeostasis of hepatocytes. Therefore, we propose that the innate immune system, including macrophages and NKT cells, can be a double-edged sword.19, 23, 24
Furthermore, it has been reported that bacterial DNA or CpG induces lethal shock,7, 25 and that bacterial DNA primes mice via IFN-γ produced by NK cells and enhances mouse sensitivity toward subsequently injected LPS, namely generalized Shwartzman reaction.26 We have empirically recognized that aged patients apparently are susceptible to septic shock and MODS after bacterial infections. In addition, complications following surgery prevail in elderly hosts but not in young adult hosts.21, 22 These findings and background led us to investigate how aging affects CpG-induced immune effects.
In the present study, we show for the first time that old mice demonstrate a high mortality and MODS after CpG challenge, which is induced by NKT cells through a TNF/FasL-dependent pathway. In contrast to enhanced TNF production, CpG-induced production of IFN-γ, IFN-α, and perforin in old mice was severely suppressed, and antitumor activity of NK cells was attenuated.
Materials and Methods
Mice and Reagents.
This study was conducted according to the guidelines of the Institutional Review Board for the Care of Animal Subjects at the National Defense Medical College, Japan. Male C57BL/6 (B6) young (6 weeks) and old (50-60 weeks) mice were purchased from Japan SLC (Shizuoka, Japan). Male gld/gld mice (young, 6 weeks; middle-aged, 25 weeks, B6 background) that are FasL-deficient were also purchased from Japan SLC. The middle-aged mice were fed standard chow and kept under clean conditions in individual cages and were thereafter used as old mice (50-60 weeks old). Because B6 CD1d−/− mice were not commercially available, CD1d−/− mice with a BALB/c background were purchased from the Jackson Laboratory and were kept until 50 weeks of age. Perforin−/− mice were kindly provided by Dr. Toru Abo, Department of Immunology, Niigata University School of Medicine.
Mouse CpG-ODN (HC4033: TCCATGACGTTCCTGATGCT) and non–CpG-ODN (HC4034: GCTTGATGACTCAGCCGGAA) were purchased from Hycult Biotechnology (Uden, Netherlands).
CpG or Non-CpG Challenge.
Young and old mice were injected intravenously with 10 μg/g body weight of CpG or non-CpG to examine the effect of CpG on cytokine-induced organ injury (young mice, ≈200 μg; old mice, 350-400 μg). To examine the effect of CpG on antitumor activity in old mice, mice were also injected intravenously with 7.5 μg/g body weight of CpG or non-CpG. We confirmed that this dose of CpG did not cause mortality in old mice.
Isolation of Mononuclear Cells or Adherent Kupffer Cells.
Under deep anesthesia with ether, the mice were euthanized and their livers and spleens were removed. Liver and spleen mononuclear cells (MNCs) were obtained as described.21, 27 Briefly, the liver was minced and suspended in Hanks' balanced salt solution containing 0.05% collagenase (Wako, Osaka, Japan). After shaking incubation for 20 minutes at 37°C, the liver specimen was passed through a 200-gauge stainless steel mesh, suspended in 33% Percoll solution, and centrifuged at 500g for 20 minutes at room temperature. After red blood cell lysing, the liver MNCs were washed twice in 10% fetal bovine serum with Roswell Park Memorial Institute-1640 (RPMI-1640) medium and were used for experiments. To obtain Kupffer cells or lymphocytes, liver MNCs were cultured in collagen-coated plastic plates for 2 hours at 37°C in 5% CO2. Thereafter, nonadherent lymphocytes were obtained through frequent gentle pipetting of plates, and adherent Kupffer cells were subsequently obtained with a cell scraper (≥90% purity by microscopic morphology and ≥80% purity by F4/80 staining).
Flow Cytometric Analysis of MNCs.
A two- or three-color immunofluorescence test was performed. MNCs were incubated for 10 minutes at 4°C with Fc-blocker (2.4 G2; BD Pharmingen, San Diego, CA) to prevent nonspecific binding. The cells were stained with fluorescein isothiocyanate (FITC)-conjugated anti-mouse F4/80 antibody (Ab), phycoerythrin (PE)-conjugated anti-NK1.1. Ab, PC5-conjugated anti-βTCR Ab, α-GalCer–loaded CD1d-mouse immunoglobulin G1 (IgG1) dimers followed by PE-conjugated anti-mouse IgG1, biotin-conjugated anti–TNF receptor 1 (TNFR1) Ab, anti–TNF receptor 2 (TNFR2) Ab, anti-FasL Ab, and isotype-matched Ab (anti-rat IgG1); immune complexes formed by the later biotin-conjugated Abs were detected with FITC-streptavidin. CD1d-mouse IgG1 dimers were purchased from BD Pharmingen and other Abs from eBioscience (San Diego, CA). The percentage of fluorescence-positive cells was analyzed using EPICS XL (Coulter, Miami, FL).
Intracellular Staining with TLR9, TNF, or Perforin.
For intracellular staining with TLR9, TNF, or perforin, we used a commercially available kit (BD Cytofix/Cytoperm solution, BD Pharmingen) and Abs (PE-conjugated anti-mouse TLR9 Ab and anti-mouse TNF Ab [eBioscience], anti-mouse perforin rat IgG2a [Kamiya Biochemical Co., Seattle, WA], and FITC-conjugated anti-rat IgG2a [eBioscience]).
Cell Cultures and In Vitro CpG Stimulation or IL-12 Stimulation.
After counting cells, 5 × 105 MNCs in 200 μL of 10% fetal bovine serum with RPMI-1640 medium were cultured in 96-well flat-bottomed plates in 5% CO2 at 37°C for 24 hours, and the culture supernatants were then stocked at −80°C until assays. For in vitro CpG stimulation, the cells were cultured with 20 μg/mL CpG for 1, 6, or 12 hours to measure TNF, IL-12, or IFN-γ levels, respectively. For in vitro stimulation with IL-12, the cells were cultured with 25 ng/mL IL-12 (R&D Systems, Minneapolis, MN) plus added low amounts of IL-2 (5 U/mL) (Shionogi Co., Osaka, Japan) and IL-15 (5 ng/mL) (R&D Systems).
Neutralization of TNF, FasL, or Depletion of NK or NK/NKT Cells to Neutralize.
TNF or FasL, anti-TNF Ab (0.5 mg/mouse) (MP6-XT3, BD PharMingen,) or anti-FasL Ab (0.5 mg/mouse) (MFL4, BD PharMingen) was injected intravenously into old mice 1 hour before CpG challenge. To deplete NK cells or both NK and NKT (NK/NKT) cells, anti-asialo GM1 (AGM1) Ab (50 μg/mouse) or anti-NK1.1 Ab (PK136; 200 μg/mouse) was injected intravenously into the mice 3 days before CpG challenge.18, 28
Measurements of Cytokine and Alanine Aminotransferase Levels.
Blood samples were obtained from the retroorbital sinus after CpG injection. The sera were stored at −80°C. Cytokine levels of the sera or the culture supernatants were measured using cytokine-specific enzyme-linked immunosorbent assay kits (TNF, IFN-γ, [BD Pharmingen], IL-12 [Endogen, Woburn, MA], IL-18 [Medical & Biological Laboratories, Nagoya, Japan]). The serum alanine aminotransferase (ALT) levels were measured using a DRICHEM 3000V system (Fuji Medical System, Tokyo, Japan).
In Vitro Antitumor Cytotoxicity Assay.
In Vivo Antimetastatic Effect.
The old (50 weeks) and young (6 weeks) mice were injected intravenously with CpG (7.5 μg/g of body weight) or non-CpG (7.5 μg/g). Subsequently, the mice were inoculated intravenously with 1 × 105 of EL-4 to monitor the survival.
The data are presented as the mean ± standard error. Statistical analyses were performed using an iMac computer (Apple, Cupertino, CA) and the Stat View 4.02J software package (Abacus Concepts, Berkeley, CA). The survival rates were compared using the Wilcoxon rank test, and other statistical evaluations were compared using the standard one-way analysis of variance followed by the Bonferroni post hoc test. P < 0.05 was considered to indicate a significant difference.
Age Affects Response to CpG Challenge.
Relative to young mice, the old mice had a decreased survival rate after CpG challenge (Fig. 1A). Old mice also showed increased serum ALT levels (Fig. 1B) and remarkably increased serum TNF levels 1 hour after the challenge (Fig. 1C). The serum IL-12 levels of old mice did not increase at 1 hour (Fig. 1D). Neither old nor young mice had increased serum IL-18 levels after CpG challenge (Fig. 1E). In contrast to young mice, serum IFN-γ in old mice did not increase at 6-12 hours (Fig. 1F). In young mice, NK cells are a main IFN-γ producer upon CpG stimulation, because depletion of NK cells by anti-asialo GM1 Ab almost completely suppressed the elevation of IFN-γ (data not shown). Serum IFN-α in old mice also did not increase compared with that of young mice (Fig. 1G).
A histologic examination of the old mice showed focal hepatocyte degeneration, which was rarely found in young mice (Fig. 2a, indicated by arrows). The old mice also showed septum thickness and moderate intra-alveolar edema in the lung (Fig. 2B, arrow) and moderate acute tubular injury in the kidney (Fig. 2C, arrow); young mice did not show such findings (Fig. 2D-F). We confirmed that nontreated old mice, as well as young mice, showed no significant histologic changes in these organs (not shown).
TNF and IFN-γ Production from Liver MNCs.
The liver MNCs of old mice produced a significantly larger amount of TNF but a lower amount of IFN-γ than those of young mice (Fig. 3A,B). Spleen MNCs did not show such an age-dependent difference (Fig. 3A,B). We further examined whether adherent Kupffer cells or hepatic lymphocytes produced TNF. The liver MNCs were separated into plastic adherent Kupffer cells and nonadherent lymphocytes and then cultured with CpG for 1 hour. The adherent Kupffer cells (but not the lymphocytes) of old mice produced a large amount of TNF (Fig. 3C).
Intercellular TLR9 and TNF of Kupffer Cells.
Old-mouse liver F4/80+ macrophages (Kupffer cells) strongly expressed TLR9 relative to young-mouse Kupffer cells (Fig. 3D), although no difference was observed for spleen F4/80+ macrophages (data not shown). In addition, Kupffer cells of old mice showed a higher TNF mean fluorescence intensity (Fig. 3E) than that of young mice, and an increased part of a high intensity of TNF (indicated by an arrow). Any other F4/80− cells in the liver did not show significant intracellular TNF (data not shown).
TNF Receptors and FasL on Liver NKT Cells.
Old mice markedly increased the expression of both TNF receptors on NKT cells after CpG challenge, relative to young mice (Fig. 4B,E); they also showed a little up-regulation of TNFR2 on the NK cells (Fig. 4D). Although both old and young mice increased Fas expression of the NK and NKT cells after CpG challenge, no significant difference was observed between them (data not shown). FasL expression from NKT cells (but not NK cells) was up-regulated after CpG challenge, particularly in old mice, as shown by either mean fluorescence intensity or percentage of positive cells (indicated by an arrow) (Fig. 4G,H), and TNF-related apoptosis-inducing ligand was slightly up-regulated in NKT cells of old mice (not shown), although mice of both ages did not express FasL or TNF-related apoptosis-inducing ligand in the NKT cells before CpG challenge (Fig. 4H). Because CD1d-dimer+ NKT cells are not exactly the same phenotype as NK1.1+ NKT cells, we examined the expression of TNF receptors and FasL on the CD1d-dimer+ NKT cells. Consistent with NK1.1+ NKT cells, old CD1d+ NKT cells similarly showed a significant up-regulation of the expression of TNF receptors and FasL (Fig. 4C,F,I).
Increased Survival Rates after CpG Challenge in Old Mice Pretreated with Various Abs.
We further investigated the roles of TNF, NK/NKT cells, and FasL in lethal shock and hepatic injury. Pretreatment with neutralizing anti-TNF Ab dramatically increased the survival rate of old mice and decreased the serum ALT levels after CpG challenge (Fig. 5A,B). Neutralization of TNF mostly abrogated the increase in FasL expression of NKT cells (Fig. 5C), but it did not affect the IFN-γ levels after CpG challenge (data not shown). Furthermore, neutralization of FasL also increased the survival rate of old mice and decreased liver injury (Fig. 5A,B). Anti-FasL Ab-injected mice showed TNF and IFN-γ levels similar to those in control B6 mice after CpG injection (not shown). Depletion of NK/NKT cells significantly increased the survival of old mice (Fig. 5D) and decreased their serum ALT levels (Fig. 5E). It also decreased the elevation of TNF (Fig. 5F) and strongly suppressed the serum IFN-γ levels (Fig. 5G). In addition, depletion of NK cells by anti-AGM1 Ab pretreatment also greatly increased survival rates and decreased serum ALT, TNF, and IFN-γ levels following CpG injection (Fig. 5D-G). Depletion of NK cells also decreased enhanced-FasL expression of NKT cells following CpG injection (Fig. 5H). These results suggest that TNF, FasL, NK cells, or NKT cells are profoundly involved in the CpG-induced pathophysiology.
Effect of CpG Injection on FasL-Deficient gld/gld Mice and NKT Cell-Deficient CD1d−/− Mice.
To further confirm the roles of FasL, NK cells, and NKT cells in CpG-induced death and liver injury, we examined gld mice and CD1d−/− mice. As expected, the age-dependent increase of mortality and liver injury were not obvious after CpG challenge in the gld mice (Fig. 6A,B). However, serum TNF levels in the old gld mice were unexpectedly and remarkably increased compared with young gld mice (Fig. 6C) and control old/young B6 mice (not shown) after CpG injection, which may explain some deaths of old gld mice. Furthermore, CpG-induced lethal shock and liver injury (Fig 6D,E), and the elevation of serum TNF levels (Fig. 6F) but not serum IFN-γ levels (Fig. 6G), were inhibited or alleviated in old NKT cell–deficient CD1d−/− mice. Similarly to young B6 mice, both young BALB/c and CD1d−/− (BALB/c background) mice showed no mortality (Fig. 6D) and higher serum IFN-γ levels (Fig. 6G), with lower serum ALT (Fig. 6E) and TNF (Fig. 6F) levels. These results (Figs. 5 and 6) suggest that NKT cells expressing FasL are a direct effector for hepatic injury and death after CpG injection and NKT cells are closely involved in TNF production from Kupffer cells, whereas NK cells are also profoundly involved in TNF production from Kupffer cells and indirectly responsible for hepatic injury and death. Age-related Kupffer cell–NK/NKT cell interactions may be important for the Kupffer cells to produce a sufficient amount of TNF. NK cells are also suggested to be main IFN-γ producers.
CpG-Induced Antitumor Cytotoxicity and Perforin Expression of NK Cells.
We next examined age-dependent effects on CpG-induced antitumor activity. Liver MNCs were obtained from old and young mice 24 hours after CpG treatment, and their antitumor cytotoxicity against EL-4 and YAC-1 was examined. Although antitumor cytotoxicity of liver MNCs was present in both old and young mice, the cytotoxicity against either EL-4 or Yac-1 was lower in old mice, and neutralization of TNF did not significantly decrease it (Fig. 7A,B). Consistently, CpG-induced perforin production by NK cells in young mice was more obvious than that in old mice, while perforin production by NKT cells was less obvious and did not differ between young and old mice (Fig. 7C,D). Anti-TNF Ab pretreatment did not decrease perforin production by liver NK cells after CpG injection, whereas the cytotoxicity of liver MNCs greatly decreased in perforin knockout mice or mice depleted of NK cells by anti-AGM1 Ab (data not shown). Pretreatment of young mice with neutralizing IFN-γ Ab also significantly reduced antitumor cytotoxicity of liver MNCs against EL-4 cells after CpG-ODN injection (data not shown).
Although CpG treatment significantly prolonged the survival time after EL-4 inoculation in both old and young mice, a CpG-induced antitumor effect was more obvious in young mice (Fig. 7E). However, neutralization of TNF did not further decrease the survival of EL-4–inoculated old mice after CpG injection (Fig. 7F).
Kupffer Cells and Suppressed IFN-γ Production.
We next examined why old mice produced a low amount of IFN-γ upon CpG stimulation. Adherent liver MNCs and nonadherent liver lymphocytes were obtained from old and young mice. Lymphocytes alone stimulated by CpG did not produce IFN-γ (data not shown). When young lymphocytes were cultured with old Kupffer cells and stimulated with CpG, the young lymphocytes decreased IFN-γ production (Fig. 8A). If old lymphocytes were also cultured with young Kupffer cells, old lymphocytes produced a substantial amount of IFN-γ (Fig. 8A).
We then examined the potential of old Kupffer cells to produce IL-12 and IL-18 upon CpG stimulation. The results showed that old Kupffer cells produced a significantly lower amount of IL-12 than young Kupffer cells (Fig. 8B). Both old and young liver MNCs produced little IL-18, and no significant difference in IL-18 production was observed between them (old versus young: 7.6 ± 5.9 versus 3.9 ± 1.8 pg/mL). When old and young liver MNCs were cultured with IL-12 (adding IL-2 and IL-15) for 12 hours, both showed similar production of IFN-γ (Fig. 8C), suggesting that old liver NK cells and NKT cells have a potential to respond to IL-12.
We previously demonstrated that the function of mouse liver NKT cells stimulated with α-GalCer enhances age-dependently.19, 23 Enhanced TNF production is closely involved in liver injury mediated by FasL-expressing NKT cells in aged mice, which is independent of IFN-γ.23 TNF exerts its effects by binding to two cell surface receptors, TNFR1 and TNFR2. TNFR1 has a death domain that is responsible for signaling cytotoxicity and is thereby considered to mediate cell death.29, 30 TNFR2 may enhance TNFR1-induced cell death or promote cell activation, migration, or proliferation.30, 31 FasL-bearing cells mediate the apoptosis of hepatocytes constitutively expressing Fas.32
FasL expression was also remarkably up-regulated on NKT cells by CpG stimulation in old mice (Fig. 9), and old FasL-deficient or CD1d-deficient mice—as well as old mice pretreated with neutralizing anti-NK1.1 Ab, TNF Ab, or FasL Ab—showed decreased liver injury/mortality after CpG challenge. NKT cells thus induce lethal shock/MODS through a TNF/FasL/Fas pathway (Fig. 9). However, NK cells may be indirectly involved in the deleterious effect induced by FasL-expressing NKT cells after CpG injection (Fig. 9), which is in marked contrast to the case of α-GalCer (NKT cell ligand) injection into mice, in which depletion of NK cells did not improve hepatic injury at all induced by FasL-expressing NKT cells.18 The mechanism of how NK cells interact with NKT cells and affect NKT cell function following CpG injection will be the focus of future study. It is also noteworthy that TNF but not FasL was reportedly involved in LPS-induced lethality,33 suggesting that the mechanisms of LPS/TLR-4- and CpG/TLR-9-mediated lethality and MODS are distinct.
We have previously demonstrated that α-GalCer–stimulated NK cells participated in antitumor immunity via IFN-γ, while TNF- and FasL-expressing NKT cells cause hepatocyte injury in old mice,19 in which cytokine production (IFN-γ, IL-4, and TNF), antitumor cytotoxicity, and hepatocyte toxicity (by FasL of NKT cells) induced by α-GalCer were all enhanced with aging. Thus, both beneficial and deleterious effects induced in mice by α-GalCer were enhanced with aging. In marked contrast, IL-12 production by CpG-stimulated Kupffer/dendritic cells—as well as IFN-γ and perforin production from NK cells—decreased with mouse age, whereas TNF production from Kupffer cells and FasL expression of NKT cells were augmented (Fig. 9). Basically, CpG-induced IFN-γ production by NK cells is triggered by IL-12 produced by Kupffer cells/macrophages.26, 34, 35 It has also been reported that IFN-α produced mainly by macrophages/dendritic cells is an important cytokine for CpG-induced antitumor effects,36 and IFN-α/β-induced antitumor cytotoxicity is reportedly dependent on perforin,37 which was found in the present study to be low in old mice after CpG injection (Fig. 9). Consistently, in old mice, antitumor cytotoxity and antitumor effects induced by CpG decreased, and hepatotoxicity and lethal shock deteriorated (Fig. 9).
It should be noted that low IL-12, IFN-α, or IFN-γ production capacity of old mice in response to CpG strongly indicates that aged mice are susceptible to infections. In fact, aged humans are indeed susceptible to bacteria infections.38 Furthermore, aged hosts are also susceptible to TNF-related shock/MODS induced by either CpG, LPS, or bacterial infections.21, 22 However, the neutralization of TNF, which alleviated CpG-induced liver injury/mortality, did not attenuate increased antitumor immunity in old mice. The TNF/FasL system and the antitumor mechanism are thus mutually independent. Therefore, TNF-neutralizing antibodies can be used together with CpG immunization therapy for elderly hosts. This may be the same for α-GalCer injection into tumor-inoculated old hosts.23
It has been reported that hepatocytes in old mice showed hyperresponsiveness to LPS,39, 40 and the TLR4 messenger RNA level is also reportedly increased in the liver tissue homogenate of old mice after LPS stimulation.40 These findings raise the possibility that the hepatocytes of old mice may also be more susceptible to direct CpG stimulation than those of young mice.
Collectively, the immune function via IL-12, IFN-α, IFN-γ, perforin, and NK cells against tumors and infections declines with age, whereas the immune function via TNF, FasL, and NKT cells (which is primarily important for defense) and its autoreactivity enhance with age. Our results demonstrate a dilemma of the immune senescence but provide a new clue to investigating immunity, aging, and therapeutic strategies against tumors and MODS associated with infections.