Suppression of islet allogeneic immune response by indoleamine 2,3 dioxygenase-expressing fibroblasts

Success of transplantation of pancreatic islets which is a promising way for restoring efficient insulin regulation in type 1 diabetes depends on lifelong use of immunosuppressive drugs. To eliminate the use of systemic immunosuppressive drugs for islet transplantation, we examined the potential use of a local immunosuppressive factor, indoleamine 2,3-dioxygenase (IDO). Thus, the aim of this study was to determine whether local expression of IDO in bystander syngeneic fibroblasts could prevent islet allogeneic immune response in vitro. C57BL/6 (B6) mouse fibroblasts were induced to express IDO by either IFN-gamma treatment or transduction with an adenoviral vector and were co-cultured with B6 mouse lymphocytes and BALB/c mouse pancreatic islets in the presence or absence of an IDO inhibitor. Proliferation of lymphocytes were then assessed using [(3)H]-thymidine incorporation assay. IDO-expression by co-cultured syngeneic fibroblasts resulted in a five-fold decrease in lymphocyte proliferation rate upon stimulation of lymphocytes by allogeneic mouse pancreatic islets (21.9% +/- 5.3 and 22.1% +/- 4.9 in the preparations with IFN-gamma treated and genetically modified IDO-expressing fibroblasts, respectively vs. 100% in control groups, P < 0.01). Allogeneic response was restored when IDO inhibitor was added to the culture indicating that suppression was due to IDO. In conclusion, this study shows that local expression of IDO by syngeneic bystander fibroblasts can suppress in vitro proliferation of lymphocytes in response to stimulation with allogeneic pancreatic islets. This local immunosuppressive function of IDO may be employed for development of a novel alternative strategy for preventing allogeneic islet graft rejection.     

Regulation of indoleamine 2,3-dioxygenase activity in the small intestine and the epididymis of mice

Indoleamine 2,3-dioxygenase activity was found to be ubiquitously distributed in various tissues of mice, such as brain, lung, stomach, intestine, and epididymis. The highest enzyme activity was detected in the alimentary canal and the epididymis. Developmental and daily rhythmic changes of indoleamine 2,3-dioxygenase activity and the effects of various regulatory factors were studied with the supernatant fractions derived from the small intestine and the epididymis. The enzyme activity in these two tissues was absent during the first 2 weeks (the weaning period). From the third week, there was a rapid increase in activities and a maximum was reached when the mice were 8 to 10 weeks of age (adolescence). The enzyme activity in the small intestine then gradually diminished to zero level at 30 weeks of age (prime) or later, while that in the epididymis remained at the high level throughout 69 weeks of age (senescence). The enzyme activity of the small intestine from mice fed during the hours 9:00–13:00 showed daily rhythmic changes; high in the daytime and low at night. Under night feeding (21:00–1:00), the enzyme activity was high at night and low in the daytime. The epididymal enzyme activity showed no daily fluctuations by either feeding schedule. With regard to the developmental and daily rhythmic changes, indoleamine 2,3-dioxygenase activity in the small intestine was similar to that of hepatic tryptophan 2,3-dioxygenase. However, in contrast to the hepatic tryptophan 2,3-dioxygenase activity, indoleamine 2,3-dioxygenase activity in the small intestine and the epididymis was not affected by adrenalectomy or intraperitoneal administration of adrenal steroid or tryptophan.     

IFN-gamma- and cell-to-cell contact-dependent cytotoxicity of allograft-induced macrophages against syngeneic tumor cells and cell lines: an application of allografting to cancer treatment.

In allogeneic tumor or skin transplantation, the rejection process that destroys the allogeneic cells leaves syngeneic cells intact by discrimination between self and nonself. Here, we examined whether the cells infiltrating into the allografts could be cytotoxic against syngeneic immortal cells in vitro and in vivo. The leukocytes (i.e., macrophages (Mphi; 55-65% of bulk infiltrates), granulocytes (20-25%), and lymphocytes (15-20%)) infiltrating into allografts, but not into autografts, in C57BL/6 mice were cytotoxic against syngeneic tumor cells and cell lines, whereas the cytotoxic activity was hardly induced in allografted, IFN-gamma-/- C57BL/6 mice. Among the leukocytes, Mphi were the major population of cytotoxic cells; and the cytotoxic activity appeared to be cell-to-cell contact dependent. When syngeneic tumor cells were s.c. injected into normal C57BL/6 mice simultaneously with the Mphi-rich population or allogeneic, but not syngeneic, fibroblastic cells, tumor growth was suppressed in a cell number-dependent manner, and tumor cells were rejected either with a Mphi:tumor ratio of about 30 or with an allograft:tumor ratio of approximately 200. In the case of IFN-gamma-/- C57BL/6 mice, however, the s.c. injection of the allograft simultaneously with tumor cells had no effect on the tumor growth. These results suggest that allograft or allograft-induced Mphi may be applicable for use in cancer treatment and that IFN-gamma induction by the allograft may be crucial for the treatment.     

Induction of indoleamine dioxygenase by interferon in mice: a study with different recombinant interferons and various cytokines

Since it is important the availability of a specific marker for interferon induction in vivo, we investigated the effect of different recombinant interferons and various cytokines on indoleamine 2,3-dioxygenase activity. Although with different magnitude, recombinant interferon-alpha A/D (Bgl II) hybrid, interferon-gamma and tumor necrosis factor, all increase the activity of this enzyme, whereas interleukin-1, recombinant interferon-alpha A and interferon-alpha D do not induce this activity in mice lung tissue. Dexamethasone is able to inhibit indoleamine 2,3-dioxygenase induction by lipopolysaccharide or by interferon-alpha A/D but it fails to prevent the induction by interferon-gamma.     

IFN-gamma is the inducer of indoleamine 2,3-dioxygenase in allografted tumor cells undergoing rejection

The depletion of an essential amino acid, tryptophan, caused by induction of indoleamine 2,3-dioxygenase (IDO), has been shown to be a mechanism involving self-defense against inhaled microorganisms and tumor growth. We recently reported that the IDO is dramatically (approximately 50-fold) induced in allografted tumor (3-methylcholanthrene-induced ascites type tumor cells) cells undergoing rejection, and that the enzyme is induced by factor(s) released through the interaction of allografted tumor cells with infiltrating leukocytes. The culture supernatant of infiltrating leukocytes, which were harvested on day 7 after tumor transplantation, induced the highest IDO activity in the tumor cells. The inducer activity was completely neutralized by the addition of antibody to IFN-gamma but not by antibody to IFN-alpha/beta. Approximately 6 U/ml of IFN-gamma was detected by an ELISA assay in the 12-h culture supernatant with 2 x 10(6) leukocytes/ml, and rIFN-gamma at 6 U/ml induced IDO in 3-methylcholanthrene-induced ascites type tumor cells to the same extent as IFN-gamma in the culture supernatant. Moreover, i.p. administration of antibody to IFN-gamma almost completely inhibited the induction of IDO in the allografted tumor cells. These observations indicate that the factor responsible for IDO induction in the allografted tumor cells is IFN-gamma.     

Indoleamine 2,3-dioxygenase activity and L-tryptophan transport in human breast cancer cells

The activity and expression of indoleamine 2,3-dioxygenase together with L-tryptophan transport has been examined in cultured human breast cancer cells. MDA-MB-231 but not MCF-7 cells expressed mRNA for indoleamine 2,3-dioxygenase. Kynurenine production by MDA-MB-231 cells, which was taken as a measure of enzyme activity, was markedly stimulated by interferon-gamma (1000 units/ml). Accordingly, L-tryptophan utilization by MDA-MB-231 cells was enhanced by interferon-gamma. 1-Methyl-DL-tryptophan (1 mM) inhibited interferon-gamma induced kynurenine production by MBA-MB-231 cells. Kynurenine production by MCF-7 cells remained at basal levels when cultured in the presence of interferon-gamma. L-Tryptophan transport into MDA-MB-231 cells was via a Na(+)-independent, BCH-sensitive pathway. It appears that system L (LAT1/CD98) may be the only pathway for l-tryptophan transport into these cells. 1-Methyl-D,L-tryptophan trans-stimulated l-tryptophan efflux from MDA-MB-231 cells and thus appears to be a transported substrate of system L. The results suggest that system L plays an important role in providing indoleamine-2,3-dioxygenase with its main substrate, L-tryptophan, and suggest a mechanism by which estrogen receptor-negative breast cancer cells may evade the attention of the immune system.     

Human placental indoleamine 2,3-dioxygenase: cellular localization and characterization of an enzyme preventing fetal rejection

In order to test the hypothesis (Munn, Zhou, Attwood, Bondarev, Conway, Marshall, Brown, Mellor, Science 281 (1998) 1191-1193) that localized placental tryptophan catabolism prevents immune rejection of the mammalian fetus, the cellular localization and characteristics of human placental indoleamine 2,3-dioxygenase (EC 1.13.11.42) were studied. The localization of indoleamine 2, 3-dioxygenase activity was determined quantitatively using cell fractionation by differential and discontinuous sucrose gradient centrifugation. Enzyme activity was looked for in isolated brush border microvillous plasma membranes of placental syncytiotrophoblast. We found that this membrane preparation (which showed a 32.4-fold purification from the starting homogenate with reference to the activity of a membrane marker enzyme, alkaline phosphatase (EC 3.1.3.1)) was strongly negatively enriched with indoleamine 2,3-dioxygenase (which showed a one twenty-fifth decrease in its specific activity). Placental indoleamine 2, 3-dioxygenase is thus not expressed in the maternal facing brush border membrane of syncytiotrophoblast. 1-Methyl-DL-tryptophan which was used by Munn et al. as a key experimental tool for inhibiting indoleamine 2,3-dioxygenase in the murine model showed a competitive inhibition of human placental indoleamine 2,3-dioxygenase with L-tryptophan. The hypothesis, based on experiments performed in mouse, may therefore be applicable to avoidance of immune rejection of the fetus in human pregnancy.     

Indoleamine 2,3-dioxygenase activity and l-tryptophan transport in human breast cancer cells

The activity and expression of indoleamine 2,3-dioxygenase together with l-tryptophan transport has been examined in cultured human breast cancer cells. MDA-MB-231 but not MCF-7 cells expressed mRNA for indoleamine 2,3-dioxygenase. Kynurenine production by MDA-MB-231 cells, which was taken as a measure of enzyme activity, was markedly stimulated by interferon-γ (1000 units/ml). Accordingly, l-tryptophan utilization by MDA-MB-231 cells was enhanced by interferon-γ. 1-Methyl-dl-tryptophan (1 mM) inhibited interferon-γ induced kynurenine production by MBA-MB-231 cells. Kynurenine production by MCF-7 cells remained at basal levels when cultured in the presence of interferon-γ. l-Tryptophan transport into MDA-MB-231 cells was via a Na⁺-independent, BCH-sensitive pathway. It appears that system L (LAT1/CD98) may be the only pathway for l-tryptophan transport into these cells. 1-Methyl-d,l-tryptophan trans-stimulated l-tryptophan efflux from MDA-MB-231 cells and thus appears to be a transported substrate of system L. The results suggest that system L plays an important role in providing indoleamine-2,3-dioxygenase with its main substrate, l-tryptophan, and suggest a mechanism by which estrogen receptor-negative breast cancer cells may evade the attention of the immune system.     

In vitro studies on transplantation immunity. II. The migration inhibition in homograft reactions in guinea pigs

The migration of peritoneal exudate cells from guinea pigs exhibiting transplantation immunity is inhibited in the presence of donor antigens. This inhibition of migration is demonstrable whether the donor transplantation antigens are presented in the form of viable cells (peritoneal exudate cells) or as particulate subcellular antigens (spleen microsomes). A greater degree of inhibition was observed when transplantation immunity was induced with lymphoid cells in Freud's adjuvant compared to sensitization with orthotopic skin grafts. There was no inhibition of migration in mixtures of normal allogeneic cells or when peritoneal cells from guinea pigs exhibiting tuberculin hypersensitivity were mixed with similar cells from normal animals. Finally, supernatants from cultures of sensitive lymphocytes plus donor antigens inhibited the migration of normal peritoneal cells indicating the presence of migration inhibitory factor (MIF) activity.     

In vitro reactivity of splenic lymphocytes from normal and UV-irradiated mice against syngeneic UV-induced tumors.

Skin tumors induced in mice by chronic ultraviolet (UV) irradiation are highly antigenic and are frequently immunologically rejected upon transplantation to normal syngeneic recipients. In this study we characterized this immune response with an in vitro microcytotoxicity test. Cytotoxic activity was present in the spleen cells of mice given a single injection of syngeneic UV-induced fibrosarcoma cells. After removal of adherent spleen cells, the remaining splenic lymphocytes were specifically cytotoxic for the immunizing tumor and showed no cross-reactivity with other syngeneic UV-induced or methylcholanthrene-induced tumors of similar histologic type. The level of cell-mediated reactivity against UV-induced tumors was quite high compared to that obtained with syngeneic tumors induced by methylcholanthrene, and the cytotoxicity was attributable to a population of theta antigen-bearing lymphocytes. With this in vitro test, we compared the response of normal mice, which reject a syngeneic tumor challenge, with that of UV-irradiated mice, in which the syngeneic UV-induced tumors grow progressively. After tumor cell inoculation, lymphocytes form the unirradiated (regressor) mice showed a high degree of cytotoxicity that reached a maximum level 8 days after injection. In contrast, no reactivity could be detected in the spleens of tumor-challenged UV-irradiated (progressor) mice.     

Inhibition of indoleamine 2,3-dioxygenase and tryptophan 2,3-dioxygenase by β-carboline and indole derivatives

β-Carboline derivatives inhibited both indoleamine 2,3-dioxygenase and tryptophan 2,3-dioxygenase activities from various sources. Among them, norharman is most potent for both enzymes from mammalian sources. Kinetic studies revealed that norharman is uncompetitive (K_i = 0.12 mm) with l-tryptophan for rabbit intestinal indoleamine 2,3-dioxygenase, and linearly competitive (K_i = 0.29 mm) with l-tryptophan for mouse liver tryptophan 2,3-dioxygenase. In addition, some β-carbolines selectively inhibited one enzyme or the other. Pseudomonad tryptophan 2,3-dioxygenase was inhibited by a different spectrum of β-carbolines. Such a selective inhibition by the structure of substrate analogs is more evident by the use of indole derivatives. Indole-3-acetamide, indole-3-acetonitrile and indole-3-acrylic acid exhibited a potent inhibition for mammalian tryptophan 2,3-dioxygenase, while they moderately inhibited the pseudomonad enzyme. However, they showed no inhibition for indoleamine 2,3-dioxygenase. These results suggest the difference of the structures of the active sites among these enzymes from various sources.     

A kinetic, spectroscopic, and redox study of human tryptophan 2,3-dioxygenase

The family of heme dioxygenases, as exemplified by indoleamine 2,3-dioxygenase and tryptophan 2,3-dioxygenase, catalyzes the oxidative cleavage of L-tryptophan to N-formylkynurenine. Here, we describe a bacterial expression system for human tryptophan 2,3-dioxygenase (rhTDO) together with spectroscopic, kinetic, and redox analyses. We find unexpected differences between human tryptophan 2,3-dioxygenase and human indoleamine 2,3-dioxygenase [Chauhan et al. (2008) Biochemistry 47, 4761-4769 ]. Thus, in contrast to indoleamine 2,3-dioxygenase, the catalytic ferrous-oxy complex of rhTDO is not observed, nor does the enzyme discriminate against substrate binding to the ferric derivative. In addition, we show that the rhTDO is also catalytically active in the ferric form. These new findings illustrate that significant mechanistic differences exist across the heme dioxygenase family, and the data are discussed within this broader framework.     

Inhibition of indoleamine 2,3 dioxygenase activity by H2O2

Indoleamine 2,3-dioxygenase is the first and rate limiting enzyme of the kynurenine pathway of tryptophan metabolism, has potent effects on cell proliferation and mediates antimicrobial, antitumorogenic, and immunosuppressive effects. As a potent cytotoxic effector, the mechanisms of indoleamine 2,3-dioxygenase inhibition deserve greater attention. The work presented here represents the first systematic study exploring the mechanisms by which low levels of hydrogen peroxide (10-100 microM) inhibit indoleamine 2,3-dioxygenase in vitro. Following brief peroxide exposure both enzyme inhibition and structural changes were observed. Loss of catalysis was accompanied by oxidation of several cysteine residues to sulfinic and sulfonic acids, observed by electrospray and MALDI mass spectrometry. Enzyme activity could in part be preserved in the presence of sulfhydryl containing compounds, particularly DTT and methionine. However, these structural alterations did not prevent substrate (l-tryptophan) binding. Some enzyme activity could be recovered in the presence of thioredoxin, indicating that the inhibitory effect of H(2)O(2) is at least partially reversible in vitro. We present evidence that cysteine oxidation represents one mechanism of indoleamine 2,3-dioxygenase inhibition.     

Tryptophan degradation in mice initiated by indoleamine 2,3-dioxygenase

Tryptophan degradation in mice initiated by indoleamine 2,3-dioxygenase was characterized, taking advantage of its induction by bacterial lipopolysaccharide. Our results demonstrated that in various tissues, N-formylkynurenine produced by the dioxygenase from tryptophan was rapidly hydrolyzed into kynurenine by a kynurenine formamidase, but it was not further metabolized. The localization in the liver and kidney of the kynurenine-metabolizing enzymes suggested that kynurenine thus formed was transported by the bloodstream to those two organs to be metabolized. In fact, the plasma kynurenine level increased in parallel with the induction of the dioxygenase by lipopolysaccharide, and kinetic analysis indicated that at the maximal induction of the enzyme there was a 3-fold increase in the kynurenine production. The major metabolic route of kynurenine was excretion in urine as xanthurenic acid. This increase in the kynurenine production was not explained by L-tryptophan 2,3-dioxygenase in the liver, because during the induction of indoleamine 2,3-dioxygenase, the hepatic enzyme level was substantially suppressed. These findings indicated that indoleamine 2,3-dioxygenase actively oxidized tryptophan in mice and that its induction resulted in an increase in tryptophan degradation.     

The in vitro killing of syngeneic cells by peritoneal cells from adjuvant-stimulated mice.

The cytotoxicity of peritoneal exudate cells from mice which had been injected with anaerobic coryneform organisms which have adjuvant activity was assessed by measuring the release of radioactive chromium from monolayers of whole mouse embryo cells. It was found that the peritoneal cells from adjuvant-stimulated mice were more cytotoxic than cells from normal mice. The increased cytotoxicity was present as early as 2 days after injection of the organisms, and was abolished by trypsin treatment of the peritoneal cells. The cytotoxic effect requires the presence of live peritoneal cells, and is more marked as the ratio of effector to target cells is increased. The plastic adherent cells of the peritoneal cell population are more effective in the cytotoxic reaction than are the non-adherent cells. The stimulated peritoneal cells can kill both syngeneic and allogeneic mouse embryo cells. Consideration is given to the possible mechanisms by which the increased cytotoxicity might be induced.     

Selective cytotoxicity of peritoneal leucocytes for neoplastic cells

Peritoneal leucocytes of nonspecifically stimulated rats (about 70% large mononuclear cells) were found to display a high degree of cytotoxicity in vitro. This cytotoxicity appeared to be exerted through a selective mechanism. Growth inhibition and cell destruction were observed when peritoneal leucocytes were cultivated with syngeneic or allogeneic neoplastic cells. Less or no cytotoxicity was observed when peritoneal leucocytes were cultivated with syngeneic or allogeneic normal kidney cells or normal embryo cells. Lymph node cells of stimulated rats were not cytotoxic towards syngeneic or allogeneic normal kidney cells. These findings may reflect a mechanism for surveillance of neoplastic cells more primitive than that ascribed to specifically committed T cells. This mechanism may be relevant to tumor regression occurring at sites of induced delayed hypersensitivity reactions.     

Interferon: a mediator of indoleamine 2,3-dioxygenase induction by lipopolysaccharide, poly(I) X poly(C), and pokeweed mitogen in mouse lung

When C3H/He mice were treated with lipopolysaccharide, poly(I) X poly(C), or pokeweed mitogen, the serum interferon titer increased almost instantaneously (100-2000 units/ml), and then the pulmonary indoleamine 2,3-dioxygenase was induced 50- to 140-fold. The peaks corresponding to interferon induction always preceded (approximately 24 h) those corresponding to dioxygenase induction. In C3H/HeJ (lipopolysaccharide-nonresponder) mice, however, lipopolysaccharide was totally inert in induction of both interferon and dioxygenase, although treatment with poly(I) X poly(C) and pokeweed mitogen led to a remarkable increase in the serum interferon titer and the enzyme activity. When lymphocytes of C3H/HeJ mice were inactivated by X irradiation and then reconstituted by the transfer of spleen cells from C3H/He mice, both enzyme and interferon from C3H/HeJ mice thus treated were induced almost normally after the lipopolysaccharide treatment. In addition, murine interferon alpha/beta, which was injected intravenously in C3H/He or C3H/HeJ mice, almost instantaneously and dose-dependently induced the pulmonary enzyme, and at a dose of 10(5) units per mouse the enzyme activity was enhanced 20- to 26-fold in these two strains of mice. These results suggest that interferon, which is generated by the interaction of lymphocytes with lipopolysaccharide, poly(I) X poly(C), or pokeweed mitogen, is a mediator of indoleamine 2,3-dioxygenase induction in the mouse lung by these agents.     

Characteristics of interferon induced tryptophan metabolism in human cells in vitro

Interferon-gamma-induced tryptophan metabolism of human macrophages was compared to ten human neoplastic cell lines of various tissue origin and to normal dermal human fibroblasts. Tryptophan and metabolites were determined in supernatants of cultures, after incubation for 48 h, by high-performance liquid chromatography with ultraviolet and fluorescence detection. With the exception of two cell lines (Hep G 2, hepatoma and CaCo 2, colon adenocarcinoma) in all of the ten other cells and cell lines tryptophan degradation was induced by interferon-gamma. Five of these ten formed only kynurenine (SK-N-SH, neuroblastoma; T 24, J 82, bladder carcinoma; A 431, epidermoid carcinoma; normal dermal fibroblasts), three formed kynurenine and anthranilic acid (U 138 MG, glioblastoma; SK-HEP-1, hepatoma; A 549, lung carcinoma). Only one line, A 498 (kidney carcinoma) showed the same pattern of metabolites as macrophages (kynurenine, anthranilic acid and 3-hydroxyanthranilic acid). Interferon-gamma regulated only the activity of indoleamine 2,3-dioxygenase. All other enzyme activities detected were independent of interferon-gamma, as shown by the capacity of the cells to metabolize L-kynurenine or N-formyl-L-kynurenine. Increasing the extracellular L-tryptophan concentration resulted in a marked induction of tryptophan degradation by macrophages. Contrarily, a significant decrease of the tryptophan degrading activity was observed when the extracellular L-tryptophan concentration was increased 2-fold with SK-N-SH, T 24 and J 82, 4-fold with A 431 and A 549 and 10-fold with U 138 MG and SK-HEP-1. The activity was unaffected by extracellular L-tryptophan with dermal fibroblasts and A 498. Though interferon-gamma was the most potent inducer of tryptophan metabolism, interferon-alpha and/or -beta showed small but distinct action on some of the cells. In all cells which reacted to interferon-gamma by enhanced expression of class I and/or class II major histocompatibility complex antigens tryptophan degradation was also inducible. These results demonstrate that induction of indoleamine 2,3-dioxygenase is a common feature of interferon-gamma action, that the extent of this induction is influenced by extracellular L-tryptophan concentrations and that indoleamine 2,3-dioxygenase is the only enzyme in the formation of 3-hydroxyanthranilic acid from tryptophan which is regulated by interferon-gamma.