Published in Volume
108, Issue 1
(July 1, 2001)J Clin Invest.
Copyright © 2001, American Society for Clinical Investigation
Adenosine deaminase deficiency increases thymic apoptosis and causes defective T cell receptor signaling
1Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, Maryland, USA
2University of Texas-Houston Medical School, Department of Biochemistry and Molecular Biology, Houston, Texas, USA
Address correspondence to: Michail V. Sitkovsky, Laboratory of Immunology, National Institutes of Health, 10/11N311, 10 Center Drive, Bethesda, Maryland 20892-1892, USA. Phone: (301) 496-5495; Fax: (301) 480-7374; E-mail: firstname.lastname@example.org.
Published July 1, 2001
Submitted: May 17,
2000; Accepted: May 14,
Adenosine deaminase (ADA) deficiency in humans results in a severe combined immunodeficiency (SCID). This immunodeficiency is associated with severe disturbances in purine metabolism that are thought to mediate lymphotoxicity. The recent generation of ADA-deficient (ADA–/–) mice has enabled the in vivo examination of mechanisms that may underlie the SCID resulting from ADA deficiency. We demonstrate severe depletion of T and B lymphocytes and defects in T and B cell development in ADA–/– mice. T cell apoptosis was abundant in thymi of ADA–/– mice, but no increase in apoptosis was detected in the spleen and lymph nodes of these animals, suggesting that the defect is specific to developing thymocytes. Studies of mature T cells recovered from spleens of ADA–/– mice revealed that ADA deficiency is accompanied by TCR activation defects of T cells in vivo. Furthermore, ex vivo experiments on ADA–/– T cells demonstrated that elevated adenosine is responsible for this abnormal TCR signaling. These findings suggest that the metabolic disturbances seen in ADA–/– mice affect various signaling pathways that regulate thymocyte survival and function. Experiments with thymocytes ex vivo confirmed that ADA deficiency reduces tyrosine phosphorylation of TCR-associated signaling molecules and blocks TCR-triggered calcium increases.
J. Clin. Invest. 108:131–141 (2001). DOI:10.1172/JCI200110360.
Inherited immunodeficiencies were recognized as distinct human disease entities several decades ago. In subsequent years, the analysis of naturally occurring immunodeficiencies yielded considerable insight regarding the composition and function of the immune system. In 1972, adenosine deaminase (ADA) deficiency became the first of the immunodeficiency diseases for which a specific molecular defect was identified (1). The importance of ADA for immune function was revealed unexpectedly when it was discovered that the enzyme was absent in several immunodeficient patients being considered for bone marrow transplantation therapy. Within a few years of this discovery, it was determined that a deficiency in another purine catabolic enzyme, purine nucleoside phosphorylase (PNP), also results in a serious immunodeficiency disease in humans, although in this case the lymphotoxic impact is largely restricted to T cells (2). Together, these findings underscore the importance of these two metabolically linked purine catabolic enzymes to proper immune development and/or function. Because of its higher prevalence and more profound immunological impact, more attention has been focused on ADA deficiency.
ADA deficiency is one of the most severe of the immunodeficiencies in humans and is associated with a severe depletion of all three major categories of lymphocytes, T cells, B cells, and NK cells (3). ADA deficiency is the second-most prevalent form of severe combined immunodeficiency disease (SCID), accounting for approximately 20% of the group (3, 4). Without intervention, affected individuals die from overwhelming opportunistic infections within the first few months of life. Available evidence suggests that the metabolic basis for ADA-deficient immunodeficiency is related to the impact of its substrates, adenosine and 2′-deoxyadenosine (reviewed in ref. 4). However, the exact molecular mechanisms of lymphotoxicity remain obscure, and the relative contribution of intracellular lymphotoxic versus signaling properties of extracellular adenosine on ADA SCID have not been sufficiently explored. Adenosine may function as an extracellular signal transducer that mediates a variety of physiological effects by binding to G protein–coupled receptors present on the surface of target cells (5). In this regard, the accumulation of adenosine and the activation of adenosine receptors known to be present on T cells (6) may lead to impaired T cell development or function (7, 8). Alternatively, 2′-deoxyadenosine may act as a cytotoxic metabolite that can mediate its effects directly at the nucleoside level or after conversion to dATP. Potential downstream consequences of elevated 2′-deoxyadenosine and/or dATP include: (a) interference with cellular transmethylation reactions (9), (b) interruption of deoxynucleotide synthesis (10, 11), and (c) activation of apoptosis (12). Thus, ADA deficiency may provoke a variety of consequences either through metabolic disturbances caused by elevated 2′-deoxyadenosine or cell signaling disturbances caused by elevated adenosine. Efforts to identify the metabolic and molecular basis for the immune dysfunction associated with ADA deficiency are motivated in part by the desire to identify pharmacological targets that could be used to control the immune system. This could be of considerable benefit in treating conditions where lymphocytes proliferate or function abnormally, such as lymphocytic leukemias and autoimmune disease. To facilitate the analysis of the SCID phenotype associated with ADA deficiency, we have recently created a mouse model of ADA-deficient SCID (13).
A two-stage genetic engineering strategy was used to produce ADA-deficient mice that retain many of the features associated with ADA deficiency in humans (13, 14). These mice develop a combined immunodeficiency that is associated with profound disturbances in purine metabolism (13). The ability to analyze tissues from ADA-deficient mice, an experimental approach not possible with ADA-deficient humans, enabled us to make novel immunological and metabolic observations. Analysis of tissue-associated metabolic disturbances in these mice revealed that absolute increases in ADA substrates are greatest in the thymus and spleen (15), suggesting that metabolic disturbances in these major immune organs play a major role in the ensuing immunodeficiency (13). Here we report a detailed analysis of the impact of ADA deficiency on intrathymic T cell development and on T cell receptor–mediated (TCR-mediated) signaling in the periphery. Our results show that ADA deficiency is associated with enhanced thymocyte apoptosis and impaired TCR signaling.
Animals. ADA-deficient mice (ADA–/–) were developed using a two-stage genetic engineering strategy (13). The development of this mouse model was based on previous studies that suggested that ADA expression in trophoblast cells of the placenta is critical for fetal development in the mouse (14). Thus, to generate completely ADA-deficient postnatal mice, an ADA minigene that targeted expression specifically to the trophoblast lineage, was introduced onto the ADA-deficient background (13). This was accomplished by intercrossing mice carrying the trophoblast-specific ADA minigene with mice heterozygous for the null ada allele. Subsequent intercrosses yielded litters that contained mice harboring the ADA minigene that were also homozygous for the null ada allele. Given that the regulatory elements used targeted ADA expression only to trophoblasts (14), once born, and with the loss of the placenta, ADA enzymatic activity was not observed in any of the tissues examined in ADA–/– mice (13).
Mice were bred and maintained in the NIH pathogen-free animal facilities. ADA–/– mice were first identified using ADA zymogram analysis of blood samples, and then genotypes were conclusively determined by Southern blot analysis of tail DNA (Figure 1a). Measurements of ADA activity in blood samples from littermates born to heterozygous (ADA+/–) intercrosses were performed by zymogram analysis (13) using a modified protocol of Knudsen et al. (16). Zymogram analysis was conducted with G493 agarose gels in a temperature-controlled electrophoresis chamber (Innovative Chemistry, Marshfield, Massachusetts, USA). Heat inactivation and ADA inhibitor treatment (coformycin; Calbiochem-Novabiochem International, San Diego, California, USA) prevented ADA band visualization in control samples, thereby verifying the ADA enzymatic activity assays in zymogram gels. ADA wild-type (ADA+/+) and heterozygous (ADA+/–) littermates were used as controls for most studies, as no differences between heterozygous and wild-type mice were found in preliminary experiments (data not shown).
Demonstration of extensive cell death in thymus but not in peripheral lymphoid organs of ADA–/– mice. (a) Genetic and biochemical evidence of ADA deficiency in screening for ADA+/+, ADA+/–, and ADA–/– mice. Littermates of ADA heterozygous mice were first analyzed by zymogram assay to identify ADA–/– mice, and then tail DNA samples were analyzed by Southern blot as described in ref. 11. Arrowheads on the zymogram indicate position of ADA and hemoglobin (internal control). (b) Comparison of spleens and thymi from ADA+/+ and ADA–/– littermates. ×3. (c) Decreased cellularity of lymphoid organs of ADA–/– mice. n, number of animals analyzed.
Cells and medium. Thymocytes or spleen cells were isolated from 3-week-old mice and either directly labeled with mAb’s or incubated in RPMI-1640 (Biofluids Inc., Rockville, Maryland, USA) supplemented with 5% dialyzed FCS (heat inactivated) and 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, 1 mM HEPES, nonessential amino acids, and 5 × 10–5 M 2β-mercaptoethanol. Where indicated, cells were incubated in ADA-free serum-free media (Life Technologies Inc., Gaithersburg, Maryland, USA).
Reagents. R-phycoerythrin-conjugated (R-PE-conjugated) rat anti-mouse anti-CD25 mAb, anti-mouse CD69, and FITC-conjugated rat anti-mouse CD8a mAb’s, as well as Cy-Chrome–conjugated CD4, were purchased from PharMingen (San Diego, California, USA). Rat anti-mouse CD4 mAb’s conjugated with RED-613 fluorochrome were purchased from Life Technologies Inc. Adenosine was prepared freshly as 20 mM stock solution. Adenosine and the ADA inhibitor EHNA were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). 2′-Deoxyadenosine was purchased from Sigma Chemical Co. (St. Louis, Missouri, USA). Coformycin was purchased from Calbiochem-Novabiochem International. Annexin V-FITC was purchased from BioWhittaker Inc. (Walkersville, Maryland, USA).
Flow cytometry. Single-cell suspensions of murine thymocytes were generated by standard procedures, and cells were either directly analyzed or cultured in 96-well plates (0.5 × 106 – 1 × 106 cells per well) as described elsewhere (17). After incubation for 16–18 hours or as indicated, cells were harvested and analyzed by flow cytometry. Flow cytometric quantitation of live, apoptotic, and dead cells was done according to a modified flow cytometry procedure (18) as described previously (17). The effects of adenosine on thymocytes were studied after incubating thymocytes ex vivo in short-term culture. Determining the status of cells (live, dead, or apoptotic) was based on gating of cells by their size (side scatter and forward scatter), plasma membrane integrity (PI staining), and redistribution of plasma membrane phosphatidylserine (Annexin V–binding). The Annexin V binding assay was done as described previously (19). Briefly, 0.6 × 106 – 1 × 106 cells were resuspended in 100 μl of buffer containing 10 mM HEPES (pH 7.3), 150 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 1.8 mM CaCl2 and incubated with 0.3 μg/ml of FITC-conjugated Annexin V and 5 μg/ml propidium iodide for 15 minutes. After incubation, samples were diluted four times with buffer containing 1.8 mM CaCl2 and were analyzed by FACScan (Becton Dickinson Immunocytometry Systems, San Jose, California, USA).
Statistical analysis of triplicate sample measurements was performed using the StatView statistic program (Abacus-Concepts Inc., Berkeley, California, USA). Standard deviations of triplicate measurements within the same experiment were lower than 1%. Flow cytometry data acquisition and analysis were done on FACScan using FACScan research software and CellQuest programs (both, Becton Dickinson Immunocytometry Systems).
Ca2+ measurements. For measurements of Ca2+ flux, freshly isolated thymocytes were preloaded with indo-1 (final 3 μM) in Ca2+ buffer (1% FCS, 10 mM HEPES, HBSS) for 30 minutes at 37°C and were washed and analyzed on a FACSVantage flow cytometer (Becton Dickinson Immunocytometry Systems) equipped with an argon laser tuned to 488 nm and a krypton laser tuned to 360 nm.
Indo-1 fluorescence was analyzed at 390/20 and 530/20 nm for bound and free probe as was described elsewhere (20, 21). The percentage of cells that responded by an increase in intracellular Ca2+ after stimulation with Concanavalin A (2.5 μg/ml) (Vector Laboratories, Burlingame, California, USA) was determined using CellQuest software program.
In situ analysis of apoptosis in thymocytes. Frozen tissue preparations and apoptosis analysis in the spleen, thymus, and lymph nodes of ADA+/+ and ADA–/– mice were performed according to procedures used by Molecular Histology Inc. (Gaithersburg, Maryland, USA). The detection of apoptotic cells by in situ staining of single-strand breaks in nuclear DNA was preformed according to protocols described previously (22, 23).
Immunoblotting with anti-phosphotyrosine mAb’s. After isolation, cells were incubated for 10 minutes at 37°C, washed once with PBS, and placed in lysis buffer containing 1% NP 40, 10 mM Tris-HCl (pH 7.2), 140 mM NaCl, 2 mM EDTA, 5 mM iodoacetamide, 1 mM Na3VO4 (Sigma Chemical Co.), and complete protease inhibitor cocktail (Boehringer Mannheim GmbH (Mannheim, Germany) for 25 minutes on ice. After removal of nuclear debris by centrifugation, the resultant supernatants were subjected to immunoprecipitation and analyzed by immunoblotting. ZAP-70 immunoprecipitation SDS-PAGE and immunoblotting were performed as described previously (24). The following antibodies were used in these experiments: rabbit antiserum to ZAP-70 (31); 4G10, a mouse mAb to phosphotyrosine (Upstate Biotechnology Inc., Lake Placid, New York, USA) and peroxidase-linked goat antibodies to mouse and rabbit Ig (Bio-Rad Laboratories Inc., Richmond, California, USA).
Extensive apoptosis occurs in the thymus but not in lymph nodes or spleens of ADA-deficient mice. ADA deficiency in humans and mice results in lymphopenia (3, 13). Previously, the status of the immune system of ADA–/– mice was analyzed at 2 weeks of age, and, although a lymphopenia was seen, it was not severe (13). We therefore analyzed the immune status of ADA–/– mice at 3 weeks of age to determine whether the immune status was more severe at these later stages. Figure 1 demonstrates that there was a pronounced decrease in the size of immune organs in ADA–/– mice (Figure 1a). Furthermore, analysis of cells isolated from ADA–/– thymi, spleens, and lymph nodes demonstrated a large reduction in the absolute numbers of cells found in ADA–/– immune organs (Figures 1, b and c). These data demonstrated that a severe lymphopenia was present in ADA–/– mice at 3 weeks of age.
The metabolic consequences associated with ADA deficiency have been demonstrated to induce apoptosis in lymphocytes in vitro (25, 26) suggesting that apoptosis may play an important role in the lymphopenia associated with ADA deficiency. To determine whether apoptosis was increased in lymphoid tissues in association with ADA deficiency in vivo, we analyzed lymphocytes in immune organs of ADA–/– mice for the presence of live, apoptotic, and dead cells (Figure 2). Analysis of side scatter versus forward scatter in thymi from ADA–/– and ADA+/+ littermates revealed an increase in the number of dead cells among immature T cells isolated from ADA–/– thymi (Figure 2a, top), whereas the proportions of dead cells among mature T cells found in the peripheral lymphoid organs (spleen and lymph nodes) were similar between ADA+/+ and ADA–/– mice (data not shown). The double staining of cells with PI and Annexin V reveals an increase in apoptosis in the thymus of ADA–/– mice (Figure 2a, bottom), but not in peripheral lymph organs (data not shown). These data suggest that there was an increase in apoptosis in thymi of ADA–/– mice, but not in secondary immune organs such as the spleen and lymph nodes.
Increased apoptosis in the thymus but not in the lymph nodes or spleens of ADA–/– mice. (a) Top: Side scatter versus forward scatter evaluation of proportion of dead cells in thymus of ADA+/+ and ADA–/– littermates. Numbers indicate percentage of live cells in the lymphoid organ. Bottom: Propidium iodide and Annexin V–aided evaluation of proportion of dead cells in thymus of ADA+/+ and ADA–/– littermates. Numbers indicate the percentage of live cells in the lower left gate estimated using both Annexin V and PI cell death flow cytometry assay as described in Methods. (b) Cytochemical demonstration of extensive clusters of apoptotic cells in thymus of ADA–/– but not of ADA+/+ littermates. Apoptotic cells are shown in red. The frozen tissue preparations and apoptosis detection were performed as described in Methods. (c) Flow cytometry demonstration of thymocyte distribution in thymi of ADA–/– and ADA+/+ littermates. (d) Flow cytometry demonstration of normal subset distributions in lymph nodes and spleens of ADA–/– mice compared with organs of ADA+/+ littermates. Numbers indicate the percentage of live cells in different subsets estimated as described in Methods.
To determine which cells were undergoing apoptosis in ADA–/– thymi, the localization of DNA strand breaks in thymus compartments was analyzed by histochemistry in situ. As expected, there was an abundance of apoptosis in ADA–/– thymi (Figure 2b, red). Increased clusters of apoptotic cells were located predominantly in the cortical-medullary boundary (Figure 2b). These findings suggest that immature T cells that accumulate in this area of the thymus are susceptible to the consequences of ADA deficiency.
ADA-deficient mice show abnormal distribution of thymocytes but a relatively normal peripheral T cell distribution. The increased apoptosis at the cortical-medullary boundary of ADA–/– thymi prompted us to examine the distribution of lymphocyte populations in this organ. As expected, the majority (70–80%) of cells recovered from ADA+/+ thymi were CD4+CD8+ double-positive cells (Figure 2c). In contrast, the reduction in the double-positive subset was observed in ADA–/– littermates. Although total cell numbers in ADA–/– thymi were always dramatically reduced, the decrease in the proportion of DP cell ranged from moderate (Figure 2c, top) to very severe in extreme cases (Figure 2c, bottom). There was a significant increase in the proportion of DN cells (24% vs. 3.2% in normal thymus), whereas the proportion of DP thymocytes was only moderately decreased (Figure 2c). Analysis of DN thymocytes for the expression of CD44 and CD25 demonstrated no significant differences in ADA+/+ and ADA–/– mice (data not shown). Interestingly, both CD4+ and CD8+ single-positive cells were found in ADA–/– thymi, suggesting that a fraction of ADA–/– lymphocytes manage to proceed through intrathymic T cell development (Figure 2c). Consistent with this observation, relatively normal distributions of CD4+ and CD8+ lymphocytes were seen in ADA–/– lymph nodes and spleens (Figure 2d).
ADA deficiency is accompanied by the presence of T cells with abnormal expression of cell surface markers. Phenotypic changes in peripheral CD4+ and CD8+ T cells from spleens of ADA–/– mice were revealed when ex vivo splenocytes or lymph nodes were analyzed for the expression of different cell surface markers (Figure 3). Surface expression of CD69 and CD44 was increased and CD62L expression was decreased in both CD4+ and CD8+ peripheral ADA–/– T cells. These data suggest that ADA–/– T cells that manage to escape apoptosis in the thymus are abnormal. Interestingly, expression of CD43 was increased in ADA–/– CD4+ T cells, but was somewhat decreased in CD8+ cells.
Changes in phenotype of peripheral CD4+ and CD8+ T cells from spleens of ADA–/– mice. Splenocytes of ADA–/– and ADA+/+ littermates were analyzed by three-color flow cytometry. Expression of different cell surface markers on CD4+ cells (a) and CD8+ cells (c) were analyzed by using CD4/CD8 gates as indicated by arrows in b. Numbers indicate proportions of cell from total cell population (b) or among gated cells (a and c).
ADA deficiency is accompanied by changes in the distribution of peripheral B cells among marginal, follicular, and newly formed zones in the spleen. To determine the status of B cells in 3-week-old ADA–/– mice, splenic B cells were analyzed using triple staining with mAb’s to B220, CD21 and CD23 (27) and to B220, IgM and IgD (data not shown). In these experiments, we found fewer mature B cells among splenocytes mice than in marginal and follicular zones of spleen of ADA–/– mice, as was indicated by CD21 versus CD23 analysis (Figure 4b) of B220+ splenocytes (Figure 4a).
Changes in phenotype of peripheral B cells from spleens of ADA–/– mice. Splenocytes of ADA–/– and ADA+/+ littermates were analyzed by three-color flow cytometry as described in Methods and indicated by arrows. (a) Evaluation of B220+ B cells in ADA–/– and ADA+/+ littermates. Numbers represent percentage of B cells from total splenocytes. (b) Distribution of B cells among marginal, follicular, and newly formed zones of spleen in ADA–/– versus ADA+/+ mice. Numbers indicate percentage of cells in different zones. Cartoon illustrates location of different B cell subsets according to staining with mAb to CD21 and CD23.
ADA deficiency is accompanied by TCR activation defects of T cells in vivo. Previously, we demonstrated that adenosine can inhibit T cell activation and expansion in vitro (6, 7, 28), suggesting that elevated extracellular adenosine levels observed in conditions of ADA deficiency may contribute to the mechanisms of the depletion and function of peripheral lymphocytes by interfering with TCR signaling. To test this hypothesis in vivo, T cell activation was analyzed in ADA–/– mice that exhibited marked increases in circulating adenosine (13). T cell activation in vivo was accomplished by the injection of anti-TCR/CD3 mAb that is known to activate T cells both in vitro (29) and in vivo (30). ADA+/+ and ADA–/– littermates were injected with anti-TCR/CD3 mAb, and 16 hours later, T cells were harvested from spleens and T cell activation was monitored by examining the upregulation of CD25 and CD69 by flow cytometry. Initial dose-response experiments in control animals suggested that a dose of 5 μg of anti-TCR/CD3 mAb was optimal for stimulation of CD25 and CD69 upregulation on T cells (data not shown). This dose was therefore used to assess T cell activation in ADA–/– mice (Figure 5). Pronounced differences in T cell activation were seen in both CD8+ (Figure 5a) and CD4+ (Figure 5b) T cells from ADA–/– mice. CD25 activation in CD8+ T cells was up to 17 times stronger in ADA+/+ mice than in ADA–/– littermates, and TCR-triggered upregulation of CD69 on CD8+ T cells was approximately two times stronger in ADA+/+ mice than in ADA–/– mice (Figure 5a). Similar differences were seen with CD4+ T cells (Figure 5b). These results demonstrated that the activation of both CD8+ and CD4+ T cells was strongly inhibited in ADA–/– mice compared with control ADA+/+ littermates.
Inhibited TCR-triggered activation of T cells in an ADA-deficient environment in vivo. ADA+/+ and ADA–/– littermates were injected intraperitoneally (i.p.) with PBS (control) or anti-CD3 mAb; 16 hours later, spleens were harvested and analyzed by flow cytometry for expression of T cell activation markers as described in Methods. Representative results of two (Exp. 1 and Exp. 2) of more than ten similar experiments using more than 20 pairs of littermates of ADA–/– and ADA+/+ mice are presented. (a) Comparison of TCR-triggered upregulation of CD25 and CD69 surface antigens on CD8+ T cells in an ADA–/– or ADA+/+ in vivo environment after injection of 5 μg of anti-CD3 mAb. (b) Comparison of TCR-triggered upregulation of CD25 and CD69 surface antigens on CD4+ T cells in an ADA-deficient or ADA-containing in vivo environment after injection of 5 μg of anti-CD3 mAb.
Inhibition of TCR-triggered activation of ADA+/+ T cells in vitro requires the addition of ADA inhibitors, suggesting the need to prevent the rapid degradation of adenosine by ADA (7). To test this interpretation, we incubated ex vivo T cells from ADA+/+ an ADA–/– mice with anti-CD3 mAb in vitro in the absence of an ADA inhibitor (Figure 6). Results demonstrated that adenosine markedly inhibited TCR-triggered activation in ADA–/– but not ADA+/+ T cells. It is shown that control incubations (0 μg of anti-CD3 mAb; Figure 6, top) had no effect on CD25 and CD69 expression in ADA+/+ or ADA–/– T cells. In contrast, activation of T cells with anti-CD3 mAb (5 μg; Figure 6, bottom) resulted in dramatic upregulation of both CD25 and CD69 activation markers. This activation was observed in both ADA+/+ and ADA–/– T cells, although strong activation of ADA–/– T cells (0.2– 64% of CD25-expressing cells) was somewhat lower than activation of ADA+/+ cells (2.6–92% of CD25-expressing cells).
Effect of adenosine on TCR-triggered upregulation of CD25 and CD69 activation markers in vitro. Ex vivo spleen cells from ADA–/– and ADA+/+ littermates were incubated in 96-well plates with immobilized anti-CD3 mAb (5 μg/ml mAb) or with serum-free, ADA-free media alone (0 μg/ml mAb) in the presence or absence of adenosine (100 μM); 16 hours later, TCR-triggered upregulation of the T cell activation markers CD69 and CD25 was evaluated by flow cytometry.
Importantly, no ADA inhibitors were required to observe the block of TCR-triggered CD25 and CD69 upregulation by adenosine in ADA–/– T cells, as the proportion of CD25+ cells decreased from 64% to 2.1% in the presence of adenosine alone. In contrast, ADA+/+ T cells were affected very little (92% vs. 88% of CD25+ cells in the presence of adenosine). Similar observations of the effects of ADA deficiency and extracellular adenosine were made in evaluations of the CD69 activation marker. These findings demonstrated that activation of ADA–/– T cells are indeed susceptible to inhibition by elevated adenosine that occurs in ADA-deficient environments, and that effects of ADA inhibitors were indeed due to their ADA-inhibiting activities. Another notable observation made in these studies was that a much higher proportion of ADA–/– T cells were activated in vitro by anti-CD3 mAb than after anti-CD3 mAb incubation in vivo (cf. Figures 5 and 6). For example, about 64% of CD4+ T cells from ADA–/– mice were activated by anti-CD3 mAb in vitro, whereas only 25–36% of ADA–/– CD4+ T cells were activated by anti-CD3 mAb in vivo. These findings suggested that the degree of inhibition of TCR-triggered activation of T cells in vivo was likely related to the severity of the metabolic disturbances seen. Collectively, these studies present the first demonstration that TCR-triggered activation of T cells was inhibited in an ADA-deficient environment in vivo.
Adenosine, but not 2′-deoxyadenosine, inhibits TCR-triggered signaling as reflected in the rescue of thymocytes from TCR-induced apoptosis. Adenosine and 2′-deoxyadenosine levels are elevated in ADA-deficient humans and mice (4, 13). We proposed recently that TCR-inhibiting properties of extracellular adenosine signaling might contribute to the mechanisms of pathogenesis in ADA SCID (7). The availability of ADA–/– mice enabled us to directly test one of the predictions of that model without using ADA inhibitors. Thus, to determine the relative impact of adenosine and 2′-deoxyadenosine on TCR-triggered signaling we selected an assay in which the TCR-inhibiting effects of adenosine would be reflected in protection of thymocytes from cell death. Such a “cell survival” read-out of the experiment on the effects of adenosine was expected to simplify the interpretation of our observations, in that it would allow us to discount possible toxic effects of intracellular adenosine and of its metabolites. Figure 7 demonstrates the effects of exogenous adenosine or 2′-deoxyadenosine on T cells during their incubation with anti-CD3 mAb in vitro. Incubation of ADA+/+ thymocytes in vitro resulted in some “spontaneous” cell death (Figure 7a), and the addition of anti-CD3 mAb increased the amount of cell death seen. Addition of adenosine did not affect the outcome of CD3-induced apoptosis in ADA+/+ thymocytes, most likely because the added adenosine was degraded by ADA. In contrast, the same concentration of exogenously added adenosine was very efficient in blocking TCR-signaling, and protected ADA–/– thymocytes from both spontaneous and TCR-triggered apoptosis (Figure 7a, –/– graph).
Extracellular adenosine inhibits TCR-induced signaling in thymocytes. (a) Adenosine alone inhibits TCR-triggered apoptosis in ADA–/– but not in normal ADA+/+ thymocytes. (b) Demonstration of opposite effects of adenosine and 2′-deoxyadenosine on spontaneous and TCR-triggered apoptosis in ADA–/– thymocytes. Thymocytes from wild-type or ADA–/– mice were incubated for 16 hours in 96-well plates with apoptosis-inducing immobilized anti-CD3 mAb in the presence or absence of added adenosine (100 μM) or 2′-deoxyadenosine (100 μM). The effect of adenosine and 2′-deoxyadenosine on thymocyte survival (proportion of live cells) was evaluated using Annexin V assay as described in the Methods.
In contrast to adenosine, we found that 2′-deoxyadenosine (Figure 7b) not only did not inhibit TCR-induced apoptosis of thymocytes, but it strongly exacerbated both spontaneous and TCR-triggered apoptosis. These data confirmed the well-documented observations of 2′-deoxyadenosine lymphotoxicity and suggested that 2′-deoxyadenosine was not involved in the inhibition of TCR-triggered events in surviving T cells in ADA–/– mice or in experiments in vitro.
ADA deficiency is accompanied by defects in TCR signaling pathways in vivo and in vitro. Proper TCR signaling plays a critical role in survival and development of thymocytes in vivo. It has been extensively shown that TCR molecules trigger a complex series of early and late biochemical events including tyrosine phosphorylation and Ca2+ mobilization (20, 21, 24, 31). It was therefore important to determine whether the inhibitory function of adenosine in ADA–/– thymocytes was the result of inhibition in TCR-triggered biochemical events in vivo. A number of studies have demonstrated the importance of the TCR ζ chain phosphorylation for initiating of TCR cell signaling. In primary thymic and lymph node T cells, the TCR ζ chain appears to be preassociated with ZAP-70, which plays a central role in the T cell signal transduction (32, 33). Using approaches developed in studies of phosphotyrosine phosphorylation of TCR-associated signaling CD3 chains (24, 31), we compared the level of ζ p21 tyrosine phosphorylation in ex vivo thymocytes from ADA–/– and ADA+/+ control littermates. Partial phosphorylation of CD3 ζ chain was notably reduced in nonactivated ex vivo thymocytes from ADA–/– mice (Figure 8a), suggesting that an ADA-deficient environment in vivo is not conducive to normal levels of TCR signaling. An internal biochemical control was provided by blotting of the same samples with anti–ZAP-70 mAb’s to show similar amount of ZAP-70 in immunoprecipitated samples from both ADA–/– and ADA+/+ littermates (Figure 8a).
ADA deficiency is accompanied with defects in TCR signaling pathways in vivo and in vitro. (a) Partial phosphorylation of CD3 ζ chain is notably reduced in nonactivated ex vivo thymocytes from ADA–/– mice, whereas ZAP-70 levels are shown to be the same in parallel samples of immunoprecipitates from ex vivo thymocytes harvested from ADA–/– and ADA+/+ littermates. (b) Intracellular Ca2+ mobilization upon ConA stimulation is inhibited by adenosine in normal thymocytes in the presence of the ADA inhibitor EHNA. Thymocytes from ADA+/+ or ADA–/– mice were used for isolation of ZAP-70 immunocomplexes followed by immunoblotting analysis of TCR ζ chain phosphorylation with an anti phosphotyrosine mAb as described in Methods. Anti ZAP-70 mAb’s were used in control immunoblotting. For measurements of calcium flux ADA+/+ thymocytes were preloaded with indo-1 and analyzed on a FACSVantage flow cytometer as described in Methods. The percentages of cells that increase intracellular calcium after stimulation with Concanavalin A are shown on the graphs. Adenosine (100 μΜ) alone or in combination with EHNA (10 μM) was added a minute before the ConA stimulation. Arrow indicates time of injection of T cell–activating stimuli.
Observations of inhibited TCR-CD3 chain tyrosine phosphorylation most likely reflects inhibited TCR signaling in the in vivo thymic environment of ADA–/– mice and predicts that other early events of TCR signaling (e.g., intracellular Ca2+ accumulation) may be inhibited by adenosine in conditions of ADA deficiency. This was examined by in vitro studies of intracellular Ca2+ accumulation in activated ADA+/+ thymocytes in the presence of the ADA inhibitor EHNA (Figure 8b). The addition of adenosine or EHNA alone did not have an inhibitory effect on Ca2+ levels in this short term Ca2+ measurement assay (Figure 8b and data not shown). However, the addition of adenosine in the presence of EHNA inhibited Ca2+ accumulation (Figure 8b, bottom). Adenosine alone was able to inhibit Ca2+ mobilization in a similar experiment with ADA–/– thymocytes, thereby confirming that effects of EHNA were due to its interactions with ADA (data not shown). Thus, early events of TCR-mediated signaling in thymocytes are inhibited in an ADA-deficient environment.
ADA deficiency in humans results in an immunodeficiency characterized by a severe reduction in T, B, and NK cells (3). The generation of ADA–/– mice (13) has allowed the examination of molecular mechanisms that underlie the immunodeficiency seen in ADA-deficient humans. In this study, we demonstrated that there was a severe lymphopenia in ADA–/– mice at 3 weeks of age. Furthermore, there was an abundance of apoptosis occurring in thymi of ADA–/– mice, but not in the spleen and lymph nodes of these animals. These findings provided evidence that the metabolic disturbances associated with ADA deficiency induced apoptosis in developing thymocytes in vivo. Peripheral T and B cells were abnormal in ADA–/– mice as reflected in the expression of cell surface markers and localization in different zones of lymphoid organs. In addition, mature T cells recovered from spleens of ADA–/– mice were defective with regard to their ability to functionally signal through the TCR. Finally, ex vivo experiments on ADA–/– T cells demonstrated that elevated adenosine was responsible for inhibited and abnormal TCR signaling.
The immunodeficiency seen in ADA-deficient patients can be severe (3, 4), suggesting mature lymphocytes and/or their precursors are sensitive to the metabolic disturbances associated with ADA deficiency. Correspondingly, studies have demonstrated that apoptosis can be induced in lymphocytes after direct exposure to adenosine or 2′-deoxyadenosine in vitro (25, 26). Our ability to analyze apoptosis in the immune system of ADA–/– mice has provided the first in vivo evidence that apoptosis is abundant in ADA–/– thymi. Most of the apoptosis seen in ADA–/– thymi was localized to the cortical-medullary boundary, a location where CD4+CD8+ double-positive thymocytes should accumulate. Consistent with these findings, analysis of thymocyte populations in these and previous studies (13) show a reduction of in the number of CD4+CD8+ thymocytes in ADA–/– mice. Similar findings have been described in mice treated with the ADA inhibitor deoxycoformycin (26). Interestingly, increased apoptosis was not noticed in histological studies and lymphocytes collected from ADA–/– spleens or lymph nodes (data not shown), suggesting that mature lymphocytes are less sensitive to the cytotoxic properties of adenosine or 2′-deoxyadenosine. Indeed, in vitro observations suggest that mature CD4+ T cells do not undergo apoptosis in response to 2′-deoxyadenosine exposure (26). These data suggest that CD4+CD8+ double-positive thymocytes, or their immediate precursors, are sensitive to the metabolic disturbances seen in ADA deficiency, and the depletion of this population is likely to account for the reduction of mature T cells found in the periphery of ADA-deficient mice (13) and humans (3, 4). Individual variations in ADA levels during early ADA–/– mice development may account for the severity of apoptotic and phenotypic changes in thymus. Drastic reduction with almost disappearance of double-positive thymocytes in some litters (Figure 2c) after 3 weeks of postnatal development could be the result of low level expression of ADA minigene in the progeny from particular breeders, and the exact mechanism of it is under study (S.G. Apasov, work in progress).
Both adenosine and 2′-deoxyadenosine exhibit potent biologic activities that may impact T cell development and function. Adenosine serves as a extracellular signaling molecule by engaging G protein–coupled receptors on the surface of target cells (5), whereas 2′-deoxyadenosine is cytotoxic to cells through mechanisms that include the inhibition of cellular transmethylation reactions (34), the disruption of cellular growth (10, 11) or differentiation (35) and the activation of apoptosis (12). The large amount of apoptosis detected in the thymi of ADA–/– mice, recent studies in fetal thymic organ cultures (36) and the observation of 2′-deoxyadenosine accumulation in the thymus of ADA–/– mice (13, 15) support the view that ADA substrates generated from apoptotic thymocytes contribute to the cytotoxicity and phenotypic outcomes of ADA deficiency.
Additionally, our findings suggest that the metabolic disturbances associated with ADA deficiency impact both intrathymic T cell development and mature T cell functions. We had shown previously that T cells treated with a combination of ADA inhibitors and adenosine in vitro exhibit defects in TCR induced activation (6, 7, 8, 28). In this study, we extended these observations to show similar effect on ADA–/– T cells, which challenged with anti-CD3 mAb in vivo. Both CD4+ and CD8+ cells in ADA–/– mice had a decreased capacity to upregulate CD25 and CD69 expression in response to anti-CD3 mAb in vivo. In important control experiments, these cells were activated by anti-CD3 mAb in vitro similarly to wild type, whereas added adenosine was inhibitory to the activation of ADA–/– but not ADA+/+ T cells. These findings suggest that the increased levels of adenosine found in peripheral lymphoid organs such as the spleen (13) can affect normal T cell signaling. Disruptions in normal T cell signaling could in turn affect the function of existing peripheral T cells in ADA–/– mice, by preventing T cell activation during an immune response.
These observations also suggest that the T cell depletion in ADA SCID may be at least partially due to blocks in TCR-driven thymocyte maturation by adenosine, as well as to direct apoptotic effects of intracellular adenosine, 2′-deoxyadenosine, and dATP. Thus, we propose that there may be at least two alternative or simultaneously operating mechanisms of T cell depletion: (a) intracellular lymphotoxicity of intracellularly accumulated adenosine, 2′-deoxyadenosine, and dATP (4, 9, 10) and (b) inhibition of TCR signaling and, hence, the inhibition/block of TCR-driven processes of T cell selection (7, 8). Both of these mechanisms have potential to cause T cell depletion in our studies in ADA–/– mice and in patients with ADA SCID.
The challenging task for testing the signaling model of ADA SCID was in demonstrating that the inhibition of TCR signaling by extracellular adenosine was not due to translocation of exogenously added adenosine into the cytoplasm followed by intracellular lymphotoxicity of adenosine catabolites. One way to exclude considerations of lymphotoxicity of exogenously added adenosine was to use an assay system in which adenosine was rescuing T cells from death by virtue of acting as a TCR-antagonizing ligand. Accordingly, we used an assay in which incubation of TCR-transgenic thymocytes with antigenic peptides and antigen-presenting cells resulted in their activation followed by their apoptosis (7). The demonstrated rescue of thymocytes from TCR-induced death (7) by adenosine was evidence of the ability of extracellular adenosine to antagonize the TCR-signaling and TCR-driven processes in conditions of ADA deficiency without being toxic in vitro. It was, however, still possible, that there was not enough extracellular adenosine to block T cell activation in an ADA-deficient environment in vivo. This issue could not be addressed without ADA–/– mice, and in this report, evidence in support of this was provided by showing that there was a strong inhibition of a T cell response in ADA–/– mice after injection with activating anti-TCR-CD3 mAb (Figure 5). Taken together, these findings suggest that the metabolic disturbances seen in ADA-deficient mice are capable of affecting various signaling pathways and intracellular processes that regulate thymocyte survival and function.
In conclusion, we have demonstrated that ADA deficiency in mice severely effects intrathymic T cell development and TCR-mediated signaling in the periphery. Our results show that ADA deficiency is associated with impaired TCR signaling and enhanced thymocyte apoptosis. This observations suggest a model that inhibition of TCR signaling in developing thymocytes may block positive selection resulting in the increase of cell death among immature thymocytes. This, in turn, will increase the accumulation of extracellular adenosine and 2′-deoxyadenosine facilitating apoptotic process in the thymus. Although the exact mechanisms by which the ADA substrates adenosine and 2′-deoxyadenosine promote immunodeficiency as a whole are still not clear, the data presented in this in vivo model have provided insight that may not have been previously appreciated. The observation that 2′-deoxyadenosine accumulation is greatest in the thymus (15), and the visualization of massive apoptosis in this organ, provide strong correlative evidence that this metabolite mediates apoptosis in this organ. The widespread accumulation of adenosine, together with the observation that adenosine inhibits TCR-triggered activation events in immature and mature T cells, suggest that systemic accumulations of adenosine may prevent the TCR signaling–dependent thymocyte differentiation and normal function of peripheral T cells. This model lends itself well to additional genetic manipulations that will help to determine the specific mechanisms involved in these processes. In addition, the ability to perform ADA enzyme therapy and gene therapy in this model (37) will provide a means to further assess the lymphoid specificity of this disease and the efficacy of these novel therapies.
The authors thank Andrea Barnes, Michelle Kress, and Charlene Shaw (National Institute of Allergy and Infectious Diseases, NIH) for help with breeding and screening of mice, Brenda Marshall for editing, and Shirley Starnes for preparation of the manuscript. Special thanks to I. Stefanova for help with immunoblotting. This work was supported in part by NIH grants (DK46207 and DK54443 to R.E. Kellems; and AI43572 and HL61888 to M.R. Blackburn).
Giblett, ER, Anderson, JE, Cohen, F, Pollara, B, Meuwissen, HJ. Adenosine-deaminase deficiency in two patients with severely impaired cellular immunity. Lancet 1972. 2:1067-1069.
Giblett, ER, Ammann, AJ, Wara, DW, Sandman, R, Diamond, LK. Nucleoside-phosphorylase deficiency in a child with severely defective T-cell immunity and normal B-cell immunity. Lancet 1975. 1:1010-1013.
Buckley, RH, et al. Human severe combined immunodeficiency: genetic, phenotypic, and functional diversity in one hundred eight infants. J Pediatr 1997. 130:378-387.
Hershfield, M.S., and Mitchell, B.S. 1995. Immunodeficiency diseases caused by adenosine deaminase deficiency and purine nucleoside phosphorylase deficiency. In The molecular and metabolic bases of inherited disease. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle, editors. McGraw-Hill. New York, New York, USA. 1725–1768.
Olah, ME, Stiles, GL. Adenosine receptor subtypes: characterization and therapeutic regulation. Annu Rev Pharmacol Toxicol 1995. 35:581-606.
Koshiba, M, Kojima, H, Huang, S, Apasov, S, Sitkovsky, MV. Memory of extracellular adenosine A2A purinergic receptor-mediated signaling in murine T cells. J Biol Chem 1997. 272:25881-25889.
Apasov, SG, Sitkovsky, MV. The extracellular versus intracellular mechanisms of inhibition of TCR- triggered activation in thymocytes by adenosine under conditions of inhibited adenosine deaminase. Int Immunol 1999. 11:179-189.
Apasov, SG, Chen, JF, Smith, PT, Sitkovsky, MV. A2A receptor dependent and A2A receptor independent effects of extracellular adenosine on murine thymocytes in conditions of adenosine deaminase deficiency. Blood 2000. 95:3859-3867.
Hershfield, MS, Kredich, NM, Ownby, DR, Ownby, H, Buckley, R. In vivo inactivation of erythrocyte S-adenosylhomocysteine hydrolase by 2’-deoxyadenosine in adenosine deaminase-deficient patients. J Clin Invest 1979. 63:807-811.
Ullman, B, Gudas, LJ, Cohen, A, Martin (Jr), DW. Deoxyadenosine metabolism and cytotoxicity in cultured mouse T lymphoma cells: a model for immunodeficiency disease. Cell 1978. 14:365-375.
Carson, DA, et al. Metabolic basis for immune dysfunction in adenosine deaminase deficiency. Ann NY Acad Sci 1985. 451:34-41.
Liu, X, Kim, CN, Yang, J, Jemmerson, R, Wang, X. Induction of apoptotic program in cell-free extracts: requirement for dATP and cytochrome c. Cell 1996. 86:147-157.
Blackburn, MR, Datta, SK, Kellems, RE. Adenosine deaminase-deficient mice generated using a two-stage genetic engineering strategy exhibit a combined immunodeficiency. J Biol Chem 1998. 273:5093-5100.
Blackburn, MR, Wakamiya, M, Caskey, CT, Kellems, RE. Tissue-specific rescue suggests that placental adenosine deaminase is important for fetal development in mice. J Biol Chem 1995. 270:23891-23894.
Blackburn, MR, Datta, SK, Wakamiya, M, Vartabedian, BS, Kellems, RE. Metabolic and immunologic consequences of limited adenosine deaminase expression in mice. J Biol Chem 1996. 271:15203-15210.
Knudsen, TB, Blackburn, MR, Chinsky, JM, Airhart, MJ, Kellems, RE. Ontogeny of adenosine deaminase in the mouse decidua and placenta: immunolocalization and embryo transfer studies. Biol Reprod 1991. 44:171-184.
Apasov, SG, Koshiba, M, Chused, TM, Sitkovsky, MV. Effects of extracellular ATP and adenosine on different thymocyte subsets: possible role of ATP-gated channels and G protein-coupled purinergic receptor. J Immunol 1997. 158:5095-5105.
Darzynkiewicz, Z, et al. Features of apoptotic cells measured by flow cytometry. Cytometry 1992. 13:795-808.
Martin, SJ, et al. Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl-2 and Abl. J Exp Med 1995. 182:1545-1556.
Rabinovitch, PS, June, CH, Grossmann, A, Ledbetter, JA. Heterogeneity among T cells in intracellular free calcium responses after mitogen stimulation with PHA or anti-CD3. Simultaneous use of indo-1 and immunofluorescence with flow cytometry. J Immunol 1986. 137:952-961.
Rabin, RL, et al. Chemokine receptor responses on T cells are achieved through regulation of both receptor expression and signaling. J Immunol 1999. 162:3840-3850.
Modak, SP, Bollum, FJ. Detection and measurement of single-strand breaks in nuclear DNA in fixed lens sections. Exp Cell Res 1972. 75:307-313.
Ben-Sasson, SA, Sherman, Y, Gavrieli, Y. Identification of dying cells: in situ staining. Methods Cell Biol 1995. 46:29-39.
Hemmer, B, Stefanova, I, Vergelli, M, Germain, RN, Martin, R. Relationships among TCR ligand potency, thresholds for effector function elicitation, and the quality of early signaling events in human T cells. J Immunol 1998. 160:5807-5814.
Kizaki, H, Suzuki, K, Tadakuma, T, Ishimura, Y. Adenosine receptor-mediated accumulation of cyclic AMP-induced T-lymphocyte death through internucleosomal DNA cleavage. J Biol Chem 1990. 265:5280-5284.
Benveniste, P, Cohen, A. p53 expression is required for thymocyte apoptosis induced by adenosine deaminase deficiency. Proc Natl Acad Sci USA 1995. 92:8373-8377.
Oliver, AM, Martin, F, Gartland, GL, Carter, RH, Kearney, JF. Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses. Eur J Immunol 1997. 27:2366-2374.
Huang, S, Apasov, S, Koshiba, M, Sitkovsky, M. Role of A2a extracellular adenosine receptor-mediated signaling in adenosine-mediated inhibition of T-cell activation and expansion. Blood 1997. 90:1600-1610.
Ferran, C, Bach, JF, Chatenoud, L. In vivo T cell activation properties of anti-T cell monoclonal antibodies. Exp Nephrol 1993. 1:83-89.
Bluestone, JA, Bruce, DS, Newell, KA. Activation of T cells in vivo using anti-CD3 and staphylococcal enterotoxins. Int J Cancer Suppl 1992. 7:39-41.
Burkhardt, AL, et al. Temporal regulation of non-transmembrane protein tyrosine kinase enzyme activity following T cell antigen receptor engagement. J Biol Chem 1994. 269:23642-23647.
Sebzda, E, et al. Selection of the T cell repertoire. Annu Rev Immunol 1999. 17:829-874.
Bolen, JB, Brugge, JS. Leukocyte protein tyrosine kinases: potential targets for drug discovery. Annu Rev Immunol 1997. 15:371-404.
Hershfield, MS, Kredich, NM. S-Adenosylhomocysteine metabolism in adenosine deaminase deficient cells. Adv Exp Med Biol 1980. 122A:421-425.
Benveniste, P, Zhu, W, Cohen, A. Interference with thymocyte differentiation by an inhibitor of S- adenosylhomocysteine hydrolase. J Immunol 1995. 155:536-544.
Thompson, LF, et al. Metabolites from apoptotic thymocytes inhibit thymopoiesis in adenosine deaminase-deficient fetal thymic organ cultures. J Clin Invest 2000. 106:1149-1157.
Blackburn, MR, et al. The use of enzyme therapy to regulate the metabolic and phenotypic consequences of adenosine deaminase deficiency in mice: differential impact on pulmonary and immunologic abnormalities. J Biol Chem 2000. 275:32114-32121.