and tInstitute for Experimental Animals, Tohoku Unitersity School of Medicine, Miyagi, JCentral Research Laboratories,. Chugai Pharmaceutical Co. Tokyo, and ...
Clin. exp. Immunol. (1991) 86, 413-418
Inhibition of autoimmune diabetes in NOD mice with serum from streptococcal preparation (OK-432)-injected mice H. SEINO, J. SATOH, S. SHINTANI, K. TAKAHASHI, X. P. 2HU, T. MASUDA*, T. NOBUNAGAj, M. SAITOJ, Y. TERANO§ & T. TOYOTA Third Department of Internal Medicine, *Second Department of Pathology and •\Institute for Experimental Animals, Tohoku University School of Medicine, Miyagi, XCentral Research Laboratories, Chugai Pharmaceutical Co. Tokyo, and §Biomedical Institute of Suntory Co., Osaka, Japan
(Accepted for publication 25 June 1991)
SUMMARY We have recently reported that systemic and chronic administration of recombinant tumour necrosis factor alpha (TNF-a), as well as streptococcal preparation (OK-432), inhibits development of insulin-dependent diabetes mellitus (IDDM) in NOD mice and BB rats, models of IDDM. In this study we examined whether serum containing endogenous TNF induced by OlC-432 injection could inhibit IDDM in NOD mice. Treatment twice a week from 4 weeks of age with OK-432-injected mouse serum, which contained endogenous TNF (75U), but not IL-1, IL-2 and interferon-gamma (IFN-y) activity, reduced the intensity of insulitis and significantly inhibited the cumulative incidence of diabetes by 28 weeks of age in NOD mice, as compared with the incidence in non-treated mice ( / ' < 0 01) and in mice treated with control serum ( / ' < 0 02). This inhibitory effect of the serum was diminished, although not significantly, by neutralization of serum TNF activity with anti-mouse TNF antibody. In the mice treated with the serum from OK-432-injected mice. Thy-1.2+ or CD8+ spleen cells decreased ( / ' < 0 01) and surface-Ig+ (S-Ig+) cells increased ( P < 0 05), whereas the proliferative response of spleen cells to concanavalin A ( f < 0 01) and lipopolysaccharide {P < 0 05) increased. The results indicate that the inhibition by OK-432 treatment of IDDM in NOD mice was partially mediated by serum factors including endogenous TNF. Keywords
type 1 diabetes
NOD mice
immunotherapy
tNTRODUCTtON NOD mice (Makino et al., 1980) and BB rats (Marliss et al., 1982) are useful as animal models of insulin-dependent diabetes mellitus (IDDM) which is caused by the autoimmune destruction of pancreatic ji cells (Eisenbarth, 1986). Susceptibility of these animals to insulitis and to the development of diabetes is genetically affected by class II MHC genes (Todd et al., 1988). However, genetically determined IDDM in these animals is prevented by various immune interventions after birth (Bach, 1988). We previously reported that immunostimulation by streptococcal preparation (OK-432) inhibits insulitis and the development of IDDM in NOD mice (Toyota et al., 1986) and BB rats (Satoh et al., 1988) and that OK-432 treatment suppresses the generation of effector cells necessary for fi cell destruction (Shintani et at., 1990), although the exact mechanisms of the action of OK-432 in vivo are uncertain. Furthermore, because OK-432 induces multiple cytokines in vivo (Hoshino et al., 1986; Ishida, 1986), we screened various recombinant Correspondence: Dr Jo Satoh, 3rd Department of Internal Medicine, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aobaku, Sendai, Miyagi 980, Japan.
413
OK-432
endogenous TNF
cytokines for possible preventive effects against autoimmune diabetes and recently reported that recombinant human tumour necrosis factor alpha (TNF-a) prevented the development of diabetes in NOD mice (Satoh et al., 1989) and in BB rats (Satoh etal., 1990). OK-432, which is a lyophilized preparation of a penicillintreated, low virulence Su-strain of streptococcal pyogencs, potentiates host defence mechanisms directed against infections and tumours through cytokine cascades in animals and humans (Ishida & Klein, 1985) and has been widely used for cancer immunotherapy in Japan (Hoshino & Uchida, 1985; Ishida, 1986). In this study we examined whether serum factors induced by OK-432 injection inhibit the development of NOD mouse diabetes. We found that the administration of serum from OK432 injected mice, containing endogenous TNF but not detectable amounts of IL-I, IL-2 and interferon-gamma (IFN-y) activity, as well as the administration of OK-432 or high doses of recombinant TNF, significantly suppressed insulitis and the development of IDDM in NOD mice, and that neutralization of endogenous TNF activity with anti-mouse TNF antibody diminished, although not significantly, the inhibitory effect of the serum.
414
H. Seino et al. MATERIALS AND METHODS
Mice NOD mice and Balb/c mice used in the experiment were maintained at the Institute for Experimental Animals, Tohoku University School of Medicine. For the experiment on prevention of diabetes, only female NOD mice were used because of their high incidence of diabetes. On the other hand, both male and female NOD mice and Balb/c mice were used for induction of endogenous TNF and collection ofsera. Induction of endogenous TNF Mice were intravenously injected in the tail vein with 0-02 and 0-2 mg of streptococcal preparation (OK-432; Chugai Pharmaceutical Co., Tokyo, Japan) at 3 h intervals as previously reported (Inagawa, Oshima & Mizuno, 1987). Two hours after the second injection, mice were anaesthetized with ether, then killed by cutting axillary artery; blood was then collected. Serum was isolated and pooled until the experiment. Recombinant mouse TNF (mTNF) and rabbit anti-mTNF antibody Recombinant mTNF and rabbit anti-mTNF antibody were kindly provided by Suntory Institute for Biomedical Research (Osaka, Japan). Specific activity ofthe mTNF was 8-06 x 10" U/ ml. An emulsion composed of a mixture of Freund's complete adjuvant (Carbiochem-Behring Co., La Jolla, CA) and mTNF (Suntory Institute) was injected subcutaneously to 5-month-old female New Zealand White rabbits (0-5 mg/rabbit). Furthermore, 2,4, 8, 12, 21 and 24 weeks later, the rabbits were boosted subcutaneously with an emulsion composed of a mixture of Freund's incomplete adjuvant (DIFCO Laboratories, Detroit, MI) and mTNF. After two weeks from the last immunization, the whole blood were taken from the rabbit. Antibodies were purified from serum by affinity chromatography on a protein A-Sepharose gel after salting out with 50% saturated (NH4)2SO4 (pH 7-4) and depletion of lipids by trichloro-trifluoroethane. Elution of antibody from a protein ASepharose column was accomplished with 0-1 M sodium acetate (pH 3-0). The eluted fractions were concentrated and dialysed against phosphate-buffered saline (PBS), pH 7-2-7-4. Concentration of the resulting antibody fraction was estimated by spectrophotometry at 280 nm. Titres of the antibody against mTNF (R86-24) were determined by ELISA and neutralization of TNF activity. The antiTNF antibody diluted to 1/312 500 reacted with mTNF-coated ELISA plate. These reactivities were absorbed with excess amount of mTNF, but not recombinant human TNF and lipopolysaccharide (LPS). The anti-mTNF antibody diluted to both 1 /1000 and 1 /10 neutralized the L-929 cell-lytic activity and lipoprotein lipase-inhibitory activity of 100 ng/ml of recombinant mouse TNF, but not human TNF and human lymphotoxin. Assay of cytokines TNF, IL-1, IL-2 and IFN-y activity in serum was bioassayed as previously reported elsewhere and briefly described below. Cytotoxieity against LM cells, a subline of TNF-sensitive mouse fibroblast (L-929), was used to measure TNF activity in the presence of 1 /(g/ml of actinomycin D as described previously
(Satoh era/., 1989). Two-fold diluted serum was introduced to a monolayer culture of LM cells in a 96-well microculture plate. After 24 h the cultured cells were stained with crystal violet and optical density (OD) at 540 nm was measured with a Multiscan (Titertek; Flow Laboratories, Mclean, VA). Recombinant mTNF was used as a standard. Bioactivity (U) of recombinant TNF had already been determined by the definition that one unit of TNF causes 50% lysis of L-929 cells in vitro. On a standard curve, ODs were inversely proportional to logarithm of TNF activities (U) or concentrations. Units of TNF of the serum samples were calculated in comparison with a standard curve of recombinant TNF. For neutralization of serum TNF activity, serum was mixed with rabbit anti-mTNF antibody, incubated overnight at 4°C and then introduced to TNF assay. IL-1 activity was assayed by measuring the mitogenic activity of sera for a murine thymocyte population nonagglutinated with peanut agglutinin as previously reported (Abo et al., 1986). Thymocytes (5 x 10') from young male C3H/ He mice, not agglutinated with peanut agglutinin, were suspended in 0-1 ml of medium and cultured with 0-1 ml of serially diluted serum in a 96-well microculture plate. After 48 h the cells were labelled with 0-5 /iCi per well of pH]-thymidine for 16 h and the incorporated radioactivity was counted by a liquid scintillation counter. Recombinant human IL-la (kindly provided by Dainippon Pharma. Co., Osaka, Japan) was used as a standard. IL-2 activity was assayed by using an IL-2-dependent murine lymphoid cell line, CTLL-2. CTLL-2 cells (1 x 10") in 0-1 ml medium per well were cultured with 0-1 ml of serially diluted serum samples for 24 h. For the final 4 h, 0-5 /(Ci of [^H]thymidine was added to each well and the radioactivity was measured. The bioactivity (U/ml) was calculated as reported elsewhere (Suzuki et al., 1986). The IFN-y titre of serum samples was determined by antiviral activity against L-929 cells infected with vesicular stomatitis virus (VSV). The 50% cytopathic effect reduction was used for calculation of IFN-y titre as previously reported (Saito et al., 1983). Definition of diabetes Urine glucose was tested twice a week with Testape (Eli Lilly & Co, Indianapolis, IN). When urine glucose persistently showed a positive reaction of greater than (-I-) with Testape or nonfasting blood glucose levels of more than 200 mg/dl, the mice were defined as having diabetes. Histology Pancreata were removed, fixed with 10% formaldehyde solution, embedded in paraffin, cut into 3 iim sections, stained with H.&E. solution, and then observed by light microscopy. The intensity of insulitis was defined as previously reported (Satoh et al., 1989). Flow cytometry Spleen was minced and passed through cotton gauze, cells were suspended in PBS, and then erythrocytes were lysed with TrisNH4CI solution. After washing with PBS supplemented with 0-2% bovine serum albumin (BSA) and 0-01% NaN.,, one million spleen cells suspended in 50 ^1 of PBS were incubated on ice for 30 min with PBS as a control or with 5 //I of fluoreseeinconjugated anti-Thy-1.2, fluorescein-conjugated anti-Lyt-2 (CD8), PE-conjugated anti-L3T4 (CD4) (Becton Dickinson,
Prevention of ID DM in NOD mice with serum Table 1. OK-432-induced cytokine production in NOD and Balb/c mice
Activity in serum (U/ml) Mouse strain
Priming Eliciting (OK-432) (OK-432)
NOD
TNF
IL-1
IL-2
IFN-).
+
