Abstract
The mechanisms underlying thymoma-associated immunodeficiency are largely unknown, and the significance of increased blood γδ Τ cells often remains elusive. In this study we address these questions based on an index patient with thymoma, chronic visceral leishmaniasis, myasthenia gravis, and a marked increase of rare γδ T cell subsets in the peripheral blood. This patient showed cutaneous anergy, even though he had normal numbers of peripheral blood total lymphocytes as well as CD4+ and CD8+ T cells. Despite his chronic infection, analyses of immunophenotypes and spectratyping of his lymphocytes revealed an unusual accumulation of naive γδ and αβ T cells, suggesting a generalized T cell activation defect. Functional studies in vitro demonstrated substantially diminished IL-2 and IFN-γ production following TCR stimulation of his “untouched” naive CD4+ T cells. Biochemical analysis revealed that his γδ and αβ T cells carried an altered TCR complex with reduced amounts of the ζ-chain (CD247). No mutations were found in the CD247 gene that encodes the homodimeric ζ protein. The diminished presence of CD247 and increased numbers of γδ T cells were also observed in thymocyte populations obtained from three other thymoma patients. Thus, our findings describe a novel type of a clinically relevant acquired T cell immunodeficiency in thymoma patients that is distinct from Good’s syndrome. Its characteristics are an accumulation of CD247-deficient, hyporresponsive naive γδ and αβ T cells and an increased susceptibility to infections.
Introduction
Thymomas are rare neoplasms of the thymic epithelium with an annual incidence of one to five per million. Diverse immunologic abnormalities can complicate these tumors through mechanisms that remain poorly understood. Common examples are autoimmune manifestations, such as myasthenia gravis and giant cell myocarditis (1, 2), combined B and T cell immunodeficiency (Good’s syndrome) (3), an absolute lymphocytosis, or a relative increase in circulating naive T cells (4, 5). Additionally, an isolated T cell immunodeficiency with unknown pathogenesis has been described that may be more frequent than classical Good’s syndrome (6–8).
In healthy adults peripheral blood γδ T cells comprise <5% of the circulating CD3+ lymphocytes and belong primarily (>70%) to the Vγ9Vδ2 subtype (9, 10). Their TCR consists of a TCRγ and a TCRδ chain, instead of the TCRαβ heterodimer expressed by αβ T cells, and the signal-transducing CD3-γ/-δ/-ε chains and ζ (CD247) molecules. Vγ9Vδ2 T cells recognize widely distributed Ags such as bacterial and synthetic phosphoantigens and alkylamines in a TCR-dependent but MHC-unrestricted manner (11–13). Increased levels of blood Vγ9Vδ2 T cells have been reported in infections, such as leishmaniasis, malaria, and tuberculosis, as well as following bone marrow transplantation (9, 14). Less commonly seen are increases in circulating Vδ1 cells, which have mostly been described in viral infections such as CMV and HIV (9, 15). Vδ1 cells normally reside within epithelia as a part of immunosurveillance measures against microbes and tumors. Common effector functions of both γδ T cell subsets include cytotoxicity, cytokine production, and immunomodulation (9). Although their activities may occasionally appear to be redundant in the normal immune system, their function could become important in cases of immunodeficiency as implied by the observations of increased numbers of γδ T cells in patients with primary immune defects (16, 17). In this study we describe a novel T cell immunodeficiency, in a patient with thymoma, chronic systemic leishmaniasis, and myasthenia gravis, characterized by normal numbers of PBLs containing expanded rare γδ T cell subsets.
Materials and Methods
Case report
A previously healthy 35-y-old Sicilian salesman, designated patient B, presented with a 2-mo history of fever and night sweats. Workup revealed generalized lymphadenopathy with a mediastinal mass, hepatosplenomegaly, and pericardial and right pleural effusions. Immunological tests demonstrated cutaneous anergy against all seven recall Ags in the Multitest Mérieux (Clostridium tetani and Corynebacterium diphtheria toxoids, tuberculin, Ags from Streptococcus group C, Proteus mirabilis, Candida albicans, and Trichophyton mentagrophytes), normal numbers of total circulating lymphocytes and CD4+ (840/μl) and CD8+ (528/μl) T cells, an expansion of γδ T lymphocytes (30% of T cells), marked NK cell cytopenia (64/μl), and a polyclonal hypergammaglobulinemia (IgM 3.25 g/l [normal values 0.40–2.30], IgG 31.50 g/l [7.00–16.00], and IgA 4.46 g/l [0.70–4.00]). No Abs against HIV or Toxoplasma gondii were detectable. Biopsies of cervical lymph nodes, bone marrow, and skin revealed abundant Leishman–Donovan bodies inside macrophages, which expanded dramatically during the following 2 y (Supplemental Figs. 1, 2). The complete absence of epithelioid cell transformation at all times was striking. Electron microscopy of the patient’s lymph nodes (Supplemental Fig. 2) confirmed massive infestation by a subspecies of Leishmania, which was identified as Leishmania infantum through culturing, PCR with parasite species determination via restriction fragment length polymorphism analyses, and serology (18, 19). Although fever, lymphadenopathy, and organomegaly rapidly subsided during and after treatment with liposomal amphotericin B, the mediastinal mass remained unchanged and the patient developed diplopia and muscle weakness. Serum acetylcholine receptor Abs were diagnostic for myasthenia gravis, and a biopsy of the mediastinal mass revealed a type B2 thymoma, with invasion of the pericardium and the right pleura. Steroids and pyridostigmine were administered, as well as six cycles of VIP-E chemotherapy (etoposide, ifosfamide, cisplatin, and epirubicin) followed by tumor resection. A histologically confirmed pleural relapse of the thymoma 1 y later was treated with two additional VIP-E courses and radiotherapy. Thereafter, the thymoma remained stable, but the patient continued to develop relapses of his leishmaniasis, despite normal PBLs (1850–3400/μl) and CD4+ and CD8+ T cell counts. During his periods of exacerbated leishmaniasis, low numbers of NK cells (<100/μl) were observed, which rose to >150–200/μl after longer remission intervals. He also suffered from approximately four acute respiratory tract infections per year, but no chronic infections other than the visceral leishmaniasis. The leishmaniasis relapses were successfully treated with liposomal amphotericin B, pentavalent antimony, INF-γ, or oral miltefosine. Approximately 9 y after the initial diagnosis, the patient developed heart failure and died of cardiogenic shock a few months later. At autopsy, giant cell myocarditis and lymphocytic myositis were noted. Leishmania amastigotes were still detectable in some of the bone marrow macrophages, although the thymoma had not progressed.
Blood samples and cell culture
Following informed consent we obtained many blood samples from patient B and from healthy volunteer donors during an interval of >7 y. PBMCs were isolated on Ficoll gradients (Lymphoprep; Progen Biotechnik, Heidelberg, Germany). Additionally, we used cryopreserved samples of thymocytes from three other patients with World Health Organization type B2 thymoma and myasthenia gravis. Purified αβ and γδ T cell subsets were cultured in IMDM (Life Technologies/Invitrogen, Karlsruhe, Germany) supplemented with heat-inactivated 10% human AB serum (PAN-Biotech, Aidenbach, Germany) as well as l-glutamine, penicillin, and streptomycin (all from Life Technologies) (“complete medium”).
CFSE proliferation assay
For the CFSE proliferation assay, freshly obtained PBMCs from patient B and healthy controls were labeled with CFSE (5 μM CFSE in complete medium at 37°C for 2 min, followed by washing in cold complete medium) and stimulated in complete medium with PHA-P (250 ng/ml; Oxoid, Wesel, Germany) and IL-2 (100 IU/ml; Roche, Mannheim, Germany) in a round-bottom 96-well plate (5 × 104/well). At 48 and 120 h cells were harvested and analyzed by flow cytometry by gating on DAPI− αβ and γδ T cells.
Cytokine ELISA
For measuring cytokine release we used cryopreserved PBMCs from patient B and healthy controls that were thawed followed by purification of negatively selected (“untouched”) naive CD4+ TCRαβ T cells using magnetic microbeads (MACS; Miltenyi Biotec, Bergisch Gladbach, Germany). The quality of the cell purifications was confirmed by flow cytometry (purity >85%, data not shown). The purified cells were stimulated at a concentration of 105/well with 250 ng/ml PHA-P or with plate-coated anti-CD3 (BioCoat 96-well plates; Becton Dickinson, Heidelberg, Germany) alone or with additional soluble anti-CD28 at a concentration of 2 μg/ml for 48 h. Measurements of human IL-2 and IFN-γ were performed by ELISA (BD OptEIA kits) using the manufacturer’s protocols.
T cell proliferation assay to Leishmania Ags
γδ T cells were purified from fresh PBMCs of patient B with MACS microbeads, after which the purified γδ+ and γδ− (αβ+) T cell fractions were each cultured in flat-bottom 96-well plates in complete medium at 5 × 104 (γδ+) or 105 (αβ+) responding cells/well, with autologous irradiated PBMCs as APCs (105/well) and Leishmania donovani Ags in various concentrations. L. infantum belongs to the L. donovani species complex (20). Controls were freshly isolated PBMCs from two healthy control donors (N1, N2) that were assayed in parallel without cell separation. After 72 h the wells were pulsed overnight with tritiated thymidine (1 μCi/well), harvested, and counted on a TopCount microplate scintillation counter (PerkinElmer, Rodgau-Jügesheim, Germany).
Immunophenotyping
Standard techniques of direct and indirect immunofluorescence in flow cytometry were used with mAbs reactive against the epitopes indicated. The Abs were anti-Vδ1 (clones A13 and R9-12-6-2), anti-Vδ2 (clone BB3), anti-Vδ3 (clone P8.6B1), anti-Vγ9 (clone TiγA), anti-Vγ8 (clone R4.5.1), anti-Vγ5 (clone 56.3), and anti-Vγ4 (clone 23D12), of which the clones BB3 and A13 were gifts from L. Moretta and A. Moretta (Genoa, Italy), the clone 23D12 was a gift from D. Wesch and D. Kabelitz (Kiel, Germany), and the other mAbs were gifts from M. Bonneville (Nantes, France). The fluorescence-labeled anti-TCRγδ mAb (clone 515) as well as mAbs reactive against CD3, CD4, CD27, CD45, and CD45RA were purchased from Beckman Coulter (Krefeld, Germany).
Spectratyping
Western blotting
Freshly isolated PBMCs or PHA-P and IL-2–expanded PBMCs from patient B and healthy donors were lysed in buffer containing 1% digitonin (22). The γδ and αβ TCRs were sequentially affinity purified. In the first immunoprecipitation, 2 μg anti-TCRδ1 mAb and protein A/G–coupled Sepharose (GE Healthcare, Freiburg, Germany) were used. Then, residual Ab was cleared from the lysate by overnight incubation with excess protein A/G–coupled Sepharose. Subsequently, a second purification step with 2 μg TCRβ-chain–specific mAb Jovi1 (Abcam, Cambridge, U.K.) and protein A/G–coupled Sepharose was performed. Purified TCRs were washed three times with buffer containing 0.1% digitonin, subjected to reducing SDS-PAGE, and visualized by Western blotting using anti-CD3ε (M20, Santa Cruz Biotechnology), anti-CD3δ (M20δ, Santa Cruz Biotechnology), and anti-CD247 Abs (449, rabbit polyclonal). The 449 serum was produced as described (23). A similar analysis was performed for the CD3ε and CD247 chains after immunoprecipitation of the TCR with OKT3 mAb in cryopreserved thymocytes from thymomas and normal thymi or in MACS-purified naive and memory T cells from healthy donors.
Sequencing of the CD247 gene
Coding sequences and exon/intron boundaries of the genomic CD247 (CD3Z) gene were amplified using the Taq polymerase system (Qiagen). PCR products were sequenced directly using the BigDye Terminator v3.1 cycle sequencing kit on an ABI Prism 3130XL genetic analyzer (Life Technologies, Darmstadt, Germany). The primer sequences are given in Supplemental Table I.
Results
Expansion of naive γδ and αβ T cells in the peripheral blood of patient B
In repeated measurements, patient B had up to 30% or 340–680/μl circulating γδ T cells (normally <5% of circulating T cells or <100/μl) with a subset composition of ∼60% Vδ1-, 22% Vδ2-, and 18% Vδ3-expressing cells. Similarly, analysis of TCRγ variable chain expression using a set of commercially unavailable mAbs demonstrated a polyclonal γδ T cell expansion (20% Vγ9-, 24% Vγ8-, 20% Vγ5-, and 28% Vγ4-expressing cells, data not shown). Most of patient B’s γδ and αβ T cells showed a naive, nonactivated phenotype (76 and 80% CD45RA+CD27+, Fig. 1A, 1B), in contrast to healthy donors (<50% CD45RA+CD27+), as also described by others (24). His γδ T cells displayed very low expression of HLA-DR (<4%) on the cell surface, no expression of NK inhibitory receptors, such as the NKG2A-CD94 heterodimer, LIR1 (CD85j), and various killer Ig receptors, with no detectable NK-like cytotoxicity against K562 target cells in 51Cr-release assays (data not shown). By spectratyping, the expanded γδ T cells featured a highly unusual, almost Gaussian distribution of the CDR3 lengths across the complete Vγ and Vδ repertoire, in contrast to healthy donors (Fig. 1C and data not shown). Similarly, analysis of the TCRβ variable gene repertoire by spectratyping showed only minimal distortions of the Gaussian CDR3 length distributions (Supplemental Fig. 3). This very unusual Gaussian-type distribution of the γδ T cell repertoire was reproducible in six blood samples obtained from the patient during a 5-y period.
Expansion of naive γδ and αβ T cells in patient B. (A) Naive (CD45RA+CD27+) and memory subsets among γδ T cells from patient B and a healthy donor (N1). (B) Naive (CD45RA+CD27+) and memory subsets among αβ T cells from patient B and a healthy donor (N1). (C) TCRδ spectratyping of fresh lymphocytes from patient B and two healthy donors (N1, N2). The red peaks represent m.w. standards.
Expansion of naive γδ and αβ T cells in patient B. (A) Naive (CD45RA+CD27+) and memory subsets among γδ T cells from patient B and a healthy donor (N1). (B) Naive (CD45RA+CD27+) and memory subsets among αβ T cells from patient B and a healthy donor (N1). (C) TCRδ spectratyping of fresh lymphocytes from patient B and two healthy donors (N1, N2). The red peaks represent m.w. standards.
Naive T cells are functionally impaired whereas in vivo–primed αβ T cells react against Leishmania Ags
The clinically observed immunodeficiency with cutaneous anergy and multiple relapses of leishmaniasis in patient B, as well as his naive T cell repertoire and phenotype, suggested a generalized defect of T cell function. Thus, we compared the ability of untouched naive CD4+ T cells from this patient to produce cytokines after polyclonal stimulation ex vivo compared with those from healthy donors. Production of IL-2 and IFN-γ was significantly impaired after a 48-h ex vivo stimulation with PHA and plastic-coated anti-CD3 mAb (Fig. 2). Upon additional costimulation by soluble anti-CD28 mAb, the patient’s cells showed normal production of IL-2, the main cytokine produced by naive T cells, but still showed suboptimal IFN-γ responses. T cell proliferation in response to strong stimulators during longer periods of incubation appeared normal. Following 5 d of stimulation by PHA and IL-2, patient B’s CFSE-labeled PBMCs showed upregulation of HLA-DR and vigorous proliferation of the αβ and γδ T cell subsets (including the CD45RA+CD27+ fractions) similar to corresponding responses in healthy donors (Fig. 3A, 3B and not shown data). Freshly obtained T cells from patient B mediated immune responses to Leishmania Ags in vitro with secondary type kinetics. There was considerable tritiated thymidine incorporation on day 3 after stimulation, whereas unprimed normal donors showed much milder proliferative responses, which was compatible with the presence of unprimed cells (Fig. 3C). Proliferation of patient B’s PBMCs to Leishmania Ags in vitro was exclusively mediated by his αβ T cells. These results suggest a defect in TCR-dependent activation of naive T cells, which could be partially reversed by stronger stimulatory signals or longer in vitro incubations, whereas in vivo–primed αβ T cells responded to Leishmania Ags in vitro quite effectively.
Cytokine production of purified untouched naive CD4+ cells from patient B and healthy donors. (A) The concentrations of IL-2 produced by untouched naive CD4+ cells from patient B and two healthy donors after 48 h of stimulation with PHA-P (open bars) and coated anti-CD3 mAb with (black bars) or without (gray bars) soluble anti-CD28 mAb measured in the cell culture supernatants. (B) The concentrations of IFN-γ produced by untouched naive CD4+ cells from patient B and two healthy donors after 48 h of stimulation with PHA-P with (shaded bars) or without (open bars) IL-2 as well as with coated anti-CD3 with (black bars) or without (gray bars) soluble anti-CD28 measured in the cell culture supernatants. The values following stimulation with medium alone were negligible (between 0 and 1 pg/ml). Assays were performed in triplicates and statistical comparisons were with the Student t test, with error bars depicting the SEM.
Cytokine production of purified untouched naive CD4+ cells from patient B and healthy donors. (A) The concentrations of IL-2 produced by untouched naive CD4+ cells from patient B and two healthy donors after 48 h of stimulation with PHA-P (open bars) and coated anti-CD3 mAb with (black bars) or without (gray bars) soluble anti-CD28 mAb measured in the cell culture supernatants. (B) The concentrations of IFN-γ produced by untouched naive CD4+ cells from patient B and two healthy donors after 48 h of stimulation with PHA-P with (shaded bars) or without (open bars) IL-2 as well as with coated anti-CD3 with (black bars) or without (gray bars) soluble anti-CD28 measured in the cell culture supernatants. The values following stimulation with medium alone were negligible (between 0 and 1 pg/ml). Assays were performed in triplicates and statistical comparisons were with the Student t test, with error bars depicting the SEM.
Ex vivo proliferation of γδ and αβ T cells from patient B after polyclonal stimulation and exposure to Leishmania Ags. (A) Proliferation (CFSE dilution) of γδ T cells from patient B and a healthy donor (N1) after 5 d of ex vivo stimulation with PHA-P plus IL-2 by flow cytometry, gated on DAPI− cells. (B) Proliferation (CFSE dilution) of αβ T cells from patient B and a healthy donor (N1) in the same experiment as described above. (C) Proliferation ([3H]thymidine incorporation) of γδ and αβ T cells from patient B and two healthy donors (N1, N2) after 72 h of exposure to increasing concentrations of Leishmania Ags. Assays were performed in quadruplicates, with error bars depicting the SEs of the mean.
Ex vivo proliferation of γδ and αβ T cells from patient B after polyclonal stimulation and exposure to Leishmania Ags. (A) Proliferation (CFSE dilution) of γδ T cells from patient B and a healthy donor (N1) after 5 d of ex vivo stimulation with PHA-P plus IL-2 by flow cytometry, gated on DAPI− cells. (B) Proliferation (CFSE dilution) of αβ T cells from patient B and a healthy donor (N1) in the same experiment as described above. (C) Proliferation ([3H]thymidine incorporation) of γδ and αβ T cells from patient B and two healthy donors (N1, N2) after 72 h of exposure to increasing concentrations of Leishmania Ags. Assays were performed in quadruplicates, with error bars depicting the SEs of the mean.
Biochemical analyses of the TCR from patient B and additional thymoma patients
This T cell activation defect might be caused by an aberrant TCR complex, which normally has a TCRαβCD3εγεδζζ or TCRγδCD3εγεδζζ stoichiometry (25, 26). To test this possibility, we isolated the TCR complexes and quantified the amount of copurified CD3 subunits by Western blotting separately for TCRγδ and TCRαβ (see Fig. 5A, 5B). Larger amounts of CD3ε and CD3δ were obtained with anti-TCRγδ immunopurification from lymphocytes of patient B (Fig. 4A, lanes 1 and 4) than from healthy donors (lanes 2, 3, 5, and 6), which reflects the higher percentage of γδ T cells in patient B. Remarkably, the TCRγδ complex from the PBMC preparations of patient B contained less ζ (CD247) protein than did the controls relative to the amount of copurified CD3ε protein (lanes 1–3). Additionally, PBMC-derived CD3δ from the controls was detected in two bands (closed and open arrows, lanes 2 and 3), whereas only the slower migrating, higher m.w. band of CD3δ was present in the patient’s γδ T cells (open arrow, lane 1). Because these bands most likely represent different CD3δ glycosylation isoforms, practically all γδ T cells of patient B contained a highly glycosylated CD3δ isoform. Importantly, the CD247 association defect, but not the peculiar glycosylation, was also detected in the patient’s αβ T cells (Fig. 4B, lane 1). All alterations of γδ and αβ TCR-CD3 components were restored following T cell activation by PHA-P and IL-2 treatment for 5 d (Fig. 4A, 4B, lanes 4–6).
TCR subunit and flow cytometric analysis of thymocytes from thymoma patients and healthy donors. (A) Western blotting of the CD3ε and CD247 chains from the thymocytes of three patients with World Health Organization type B2 thymomas and myasthenia gravis (Pt1–3, other than Pt B) and three healthy donors (N1–3). The intensities of the precipitated CD3ε and CD247 subunits were quantified using the LI-COR Odyssey system, and the CD247/CD3ε ratios were compared using the Student t test, with error bars depicting the SEs of the mean. This experiment was repeated with similar results. (B) Cryopreserved thymocytes from the same three thymomas and normal thymi were analyzed by flow cytometry to quantify percentages of γδ+ T cells within the CD3+ T cells (Student t test, with the bars depicting mean values).
TCR subunit and flow cytometric analysis of thymocytes from thymoma patients and healthy donors. (A) Western blotting of the CD3ε and CD247 chains from the thymocytes of three patients with World Health Organization type B2 thymomas and myasthenia gravis (Pt1–3, other than Pt B) and three healthy donors (N1–3). The intensities of the precipitated CD3ε and CD247 subunits were quantified using the LI-COR Odyssey system, and the CD247/CD3ε ratios were compared using the Student t test, with error bars depicting the SEs of the mean. This experiment was repeated with similar results. (B) Cryopreserved thymocytes from the same three thymomas and normal thymi were analyzed by flow cytometry to quantify percentages of γδ+ T cells within the CD3+ T cells (Student t test, with the bars depicting mean values).
TCR subunit analysis of primary and expanded γδ and αβ T cells from patient B and healthy donors. (A) Western blotting of the CD3ε, CD3δ, and CD247 chains contained in immunoprecipitated γδ TCRs from freshly isolated (PBMC) or PHA- and IL-2–expanded PBMCs (expan.) of patient B (Pt B) and healthy donors (N1 and N2). The open and closed arrows indicate two different glycosylation forms of CD3δ in the γδ TCR. This experiment was repeated three times. (B) Western blotting of the CD3ε, CD3δ, and CD247 chains contained in immunoprecipitated TCRαβ from freshly isolated (PBMC) or PHA- and IL-2–expanded PBMCs (expan.) of patient B (Pt B) and healthy donors (N1 and N2). (C) Western blotting of the CD3ε and CD247 chains contained in immunoprecipitated TCRs from purified naive and memory T cells from two healthy donors.
TCR subunit analysis of primary and expanded γδ and αβ T cells from patient B and healthy donors. (A) Western blotting of the CD3ε, CD3δ, and CD247 chains contained in immunoprecipitated γδ TCRs from freshly isolated (PBMC) or PHA- and IL-2–expanded PBMCs (expan.) of patient B (Pt B) and healthy donors (N1 and N2). The open and closed arrows indicate two different glycosylation forms of CD3δ in the γδ TCR. This experiment was repeated three times. (B) Western blotting of the CD3ε, CD3δ, and CD247 chains contained in immunoprecipitated TCRαβ from freshly isolated (PBMC) or PHA- and IL-2–expanded PBMCs (expan.) of patient B (Pt B) and healthy donors (N1 and N2). (C) Western blotting of the CD3ε and CD247 chains contained in immunoprecipitated TCRs from purified naive and memory T cells from two healthy donors.
We next examined whether this CD247 defect might be attributable to the predominantly naive phenotype of the cells from patient B. When comparing the amount of ζ (CD247) protein relative to the amount of CD3ε protein in the TCR complexes precipitated from naive and memory T cells of healthy donors, we found no apparent differences (Fig. 4C). Thus, reduction of CD247 within the TCR complexes of naive T cells from patient B could be a consequence of the thymoma. Indeed, we found that the TCR complexes of thymocytes isolated from an additional three thymoma patients contained less CD247 than did thymocytes from three normal thymuses (Fig. 5A). Moreover, the thymomas contained more thymocytes expressing the TCRγδ than the normal thymi, as determined by flow cytometry (Fig. 5B).
Sequencing of the CD247 gene from patient B
Following isolation of genomic DNA from purified γδ T cells from patient B, the eight exons with and without 10 bp intronic sequence of the CD247 gene (encoding the ζ-chain [CD247]) were sequenced. No mutations were detected.
Discussion
Thymoma patients can develop clinically overt immune deficiency, but the underlying pathophysiology is largely unknown (3). A relative increase in circulating naive T cells in thymoma patients has been described, but the mechanism and clinical significance remain unknown (4). In this study, we address these questions based on observations in a patient with thymoma, chronic leishmaniasis, and an expansion of rare γδ T cell subsets in the peripheral blood in the presence of normal numbers of circulating lymphocytes as well as CD4 and CD8 T cells. Clinically, the diagnosis of immunodeficiency in this patient was based on cutaneous anergy and the fulminant clinical presentation of leishmaniasis with multiple relapses in the absence of immunosuppressive therapy. Histomorphological evidence for immunodeficiency was that this patient’s leishmania-infected macrophages completely lacked signs of epithelioid cell transformation in all biopsy specimens and at autopsy. Because such activated macrophages typically appear in Th cell–induced granulomas, this absence of epithelioid cell transformation in the infected lymph nodes of our patient documents a Th cell defect. Despite his chronic infection, most γδ and αβ T cells of this patient displayed a naive phenotype (Fig. 1A, 1B) and Gaussian spectratypes (Fig. 1C, Supplemental Fig. 3), which suggested a generalized defect in T cell activation. The low NK cell numbers fluctuating with disease activity could be explained as a reaction to leishmaniasis (27). The clinical observation that the patient only suffered from relapses of visceral leishmaniasis, but did not develop persistent or recurrent infections with other pathogens, might indicate impairment of his immunosurveillance against only certain intracellular microbes. He presumably became infected with Leishmania in Sicily or while traveling to Africa as a salesman. Nevertheless, it is possible that had he been exposed to other intracellular pathogens such as Mycobacteria spp., Salmonella spp., T. gondii, Cryptococcus neoformans, or Histoplasma capsulatum, he might have also developed chronic clinically significant infections with these microorganisms. A lack of exposure is supported by the fact that the patient had no Abs against T. gondii and, therefore, belonged to the ∼60–70% seronegative persons in his age group that do not harbor persistent Toxoplasma parasites.
In vitro functional studies confirmed the presence of a T cell activation defect, because this patient’s untouched naive CD4+ T cells showed markedly diminished IFN-γ and IL-2 production following stimulation by PHA or anti-CD3, in comparison with cells from healthy donors (Fig. 2). However, unselected fresh PBMCs from patient B proliferated quite strongly following longer polyclonal stimulations by PHA and IL-2 in vitro (Fig. 3A, 3B). Additionally, his unselected PBMCs showed at least partially effective secondary-type responses to Leishmania Ags in vitro (Fig. 3C). However, one cannot conclude that these memory αβ T cell responses to Leishmania were entirely normal, because no immune competent control patients with previous exposure to Leishmania Ags were available for comparison. In fact, these secondary responses to Leishmania might represent an “old” memory response that was generated when patient B was still thymoma-free. It is also unlikely that the observed functional abnormalities of naive αβ T cells in this patient were due to the naive γδ T cell population, because negatively purified helper αβ T cells displayed defective cytokine production after stimulation in vitro (Fig. 2) and because the strongest proliferative responses to Leishmania Ags were observed with unseparated patient B PBMCs (Fig. 3C). Taken together, these data suggest a mild defect of lymphocyte activation in patient B as the basis for his immunodeficiency, mainly involving lymphokine production. Notably, IFN-γ plays a crucial role in the defense against leishmania and the other intracellular pathogens (28).
In search of a biochemical correlate of this defect, we analyzed the TCR complex of lymphocytes from patient B and found reduced amounts of CD247 in his γδ and αβ T cells (Fig. 4). The CD247 molecule, the subunit that is added last to the TCR during assembly (29), is a critical signal transduction component that is massively phosphorylated upon Ag recognition, and its downregulation results in impairment of TCR signaling (30). Such a defect has been described as a consequence of chronic inflammation in patients with cancers, infections, autoimmune disorders (31–34), including systemic lupus erythematosus (35, 36), or in patients undergoing hemodialysis (37), but not in naive T cells, such as the lymphocytes of patient B. The naive phenotype by itself cannot explain the reduced CD247 chain content in lymphocytes from patient B because naive and memory T cells from healthy donors contained similar amounts of CD247 (Fig 4C), confirming results from earlier studies (38). Besides, the reduction of CD247 in patient B’s lymphocytes was associated with a higher degree of glycosylation of the TCRγδ CD3δ component (25). All TCR alterations in patient B were restored following prolonged T cell activation by PHA-P and IL-2 (Fig. 4A, 4B), which parallels the restoration of CD247 expression in tumor-infiltrating lymphocytes by IL-2 (39). Thus, these abnormalities were most likely acquired rather than genetically determined (40), as also indicated by the lack of mutations in the CD247 (CD3Z) gene of patient B.
A causal relationship of the CD247 defect with thymomas is strongly suggested by the observation that the reduction of CD247 expression was present in thymocytes from other thymoma patients, but not in normal thymocytes (Fig. 5A). Thus, the CD247 abnormality is not just a special characteristic of our patient secondary to the Leishmania infection, but it seems to represent a more general mechanism of thymoma-related immune dysregulation, which may underlie the functional T cell defects in other patients with thymomas as well (7, 8). INF-γ, TNF, myeloid suppressor cells, and decreased levels of l-arginine have been reported to downregulate CD247 (30, 31, 41–43). Thus, the microenvironemnt in some thymomas could contribute to the observed CD247 dysregulation. The accumulation of rare naive γδ T cells in the blood of patient B is very likely attributable to his thymoma. We found that γδ T cells were also increased in thymocytes from other thymoma patients as well (Fig. 5B), whereas others described a preferential development of human γδ T cells in humanized SCID mice following transplantation of human thymoma tissues compared with normal thymic grafts (44). A thymic origin for the peripheral γδ T cells in patient B is directly implied by their naive phenotype, the Gaussian-like (naive) CDR3 lengths distributions, and because they mostly expressed rare variable Vγ and Vδ genes such as Vγ4, Vγ5, and Vγ8, combined with Vδ1 and Vδ3 (detected by mAbs) and Vδ4, Vδ5, and Vδ6 (detected by spectratyping, data not shown). These rare γδ T cell subsets are normally replenished through thymic T cell development, in contrast to the more common Vγ9Vδ2 T cells, which generally expand after activation in the periphery (24, 45).
Impairment of T cell function with thymoma is commonly described as Good’s syndrome, together with hypogammaglobulinemia and reduced or absent B cells (3, 46). However, our patient with hypergammaglobulinemia and normal numbers of B cells clearly did not have Good’s syndrome, which typically manifests itself by susceptibility to encapsulated bacteria, as well as opportunistic viral and fungal infections. Extrapolating from the literature, an isolated T cell immunodeficiency appears to be even more common in thymoma patients (6–8) than classical Good’s syndrome (46). Our findings extend the observation that patients with thymoma have increased percentages of naive T cells (4, 47), and they also suggest that the diminished presence of CD247 in the TCR complex in thymoma patients may lead to T cell immunodeficiency and opportunistic infections with intracellular pathogens, such as Cryptococcus (8), tuberculosis (48), and salmonellosis (49). Similar to our patient B, the clinical presentation in each of these case reports was dominated by a single infection.
How can the expansion of naive γδ T cells, the presence of functionally impaired naive αβ T cells, and the CD247 defect be reconciled? As shown above, the reduced CD247 content is not typical for normal naive T cells, but it appears to be related to T cell development in patients with thymomas. This CD247 abnormality may lead to the accumulation of naive T cells due to the inability of these cells to reach the activation threshold and attain the memory state under physiological conditions of stimulation. The naive phenotype in patient B persisted for years despite multiple reactivations of leishmaniasis, which is compatible with the observations of increased circulating naive T cells in thymoma patients and in patients with paraneoplastic myasthenia gravis (4, 47). Strong TCR signals during intrathymic T cell maturation at the double-negative stage are thought to promote γδ T cell differentiation (50, 51), whereas excessive TCR stimulation of mature T cells induces the CD247 defect in patients with chronic immune activation (30, 31). Thus, the neoplastic thymic epithelium might disturb the T cell development causing both an increased output of γδ T cells and reduced amounts of CD247 in the TCRs of maturing αβ and γδ T cells through excessive stimulation of developing T lymphocytes. This would explain the reversibility of the CD247 defect in our patient after prolonged T cell stimulation in vitro, as has already been observed in the case of tumor-infiltrating lymphocytes (39). Additionally, it could explain the unusual association of higher numbers of naive T cells in the periphery (4, 52) with the development of thymoma-related autoimmune conditions such as myasthenia gravis (47) and giant cell myocarditis (1, 2). A possible mechanism could be the failure of the negative selection, as suggested by observations in cases of a genetically determined T cell hyporesponsiveness due to the PTPN22gain-of-function+1858T(+) genotypes (53). In summary, a CD247 defect, probably induced in the neoplastic thymic microenvironment, leads to the accumulation of naive, hyporesponsive γδ and αβ T cells in thymoma patient resulting in a novel type of acquired T cell immunodeficiency and immune dysregulation of clinical significance.
Acknowledgements
We thank Joachim Clos for the gifts of L. donovani Ags, Lorenzo and Alessandro Moretta, Daniela Wesch, and Dieter Kabelitz as well as Marc Bonneville for many anti-TCRγδ Abs. We also thank Daniela Bukatz and Sabine Glatzel for expert technical assistance, Marie Follo for carefully reading the manuscript, Annette Schmitt-Gräff for helpful discussions, Alexandros Spyridonidis for taking care of the patient and clinical support of our work, and Uwe Mäder and Clemens Kreutz for advice with statistical analysis.
Footnotes
This work was supported by Deutsche Forschungsgemeinschaft Grants CRC620-B6, CRC620-Z2, CRC643-A5, and EXC294 (to the BIOSS Centre for Biological Signalling Studies), José Carreras Leukämie Stiftung Grant DJCLS R 05/01, and by European Union Seventh Framework Programme 2007–2013 (SYBILLA project).
The online version of this article contains supplemental material.
References
Disclosures
The authors have no financial conflicts of interest.