T cell development is tightly controlled by thymic stromal cells. Alterations in stromal architecture affect T cell maturation and the development of self-tolerance. The monogenic autoimmune syndrome APECED (autoimmune-polyendocrinopathy-candidiasis-ectodermal dystrophy) is characterized by the loss of self-tolerance to multiple organs. Although mutations in the autoimmune regulator (AIRE) gene are responsible for this disease, the function of AIRE is not known. Here we report on the spatial and temporal pattern of murine Aire expression during thymic ontogeny and T cell selection. Early during development, thymic Aire transcription is critically dependent on RelB and occurs in epithelial cells in response to lymphocyte-mediated signals. In adult tissue, Aire expression is confined to the medulla and the corticomedullary junction, where it is modulated by thymocytes undergoing negative selection. Aire may determine thymic stromal organization and with it the induction of self-tolerance.

The thymus is the primary lymphoid organ for the development of T cells of the αβTCR lineage. Lymphoid cells differentiate in direct physical contact with thymic stromal cells, and this interaction is not only critical for thymocyte selection and maturation but is equally essential for the induction of a functionally competent thymic stromal compartment. The thymic primordium of mice arises bilaterally at day 10 of gestation (E10)4 by fusion of the third pouch endoderm with ectoderm from the corresponding branchial clefts. The two tissues are separated by a thin layer of mesenchyme originating from the cephalic neural crest (1). While the relative contribution of each germ layer to the mature thymic epithelial compartment remains to be determined, a small subpopulation of epithelial cells has been characterized that represents a common precursor cell for mature cortical and medullary epithelium (2, 3). At day E11.5, hemopoietic precursor cells seed the thymus anlage although its epithelial cells are not yet competent to fully support T cell development (4). This capacity is only achieved after further maturation when epithelial cells have differentiated into distinct stromal cell types (reviewed in Ref. 5). Finally, the thymic microenvironment is composed of an integrated network of epithelial reticular cells and nonepithelial stromal cells (i.e., fibroblasts, macrophages, and interdigitating reticular cells) each characterized by typical structural, antigenic, and functional features. Notably, the organization of the thymic epithelium differs from most other epithelial organs in the body: rather than forming a sheet of cells positioned on a basement membrane, thymic epithelial cells (TEC) form a three-dimensional meshwork (6).

The space between the diverse thymic stromal cells is occupied by thymocytes at different stages of development as defined by their expression of CD4 and CD8 (7). The most immature population of intrathymic T cell precursors lack CD4, CD8, and CD3 expression and are referred to as triple negative (TN). These cells can be further classified into four independent subpopulations according to their sequential expression of CD44 and CD25: early thymic immigrants are CD44+CD25 (TN I) and develop via a CD44+CD25+ (TN II) and a CD44CD25+ (TN III) stage to immature thymocytes with a CD44CD25 (TN IV) phenotype (8). Subsequently, thymocytes begin to express transiently CD8 before become CD4+CD8+ (double positive, DP) cells, a population that constitutes the majority of thymocytes. At this stage, DP cells express an αβTCR that renders them subject to either positive or negative selection dependent on their Ag specificity. Only a minority (∼3%) of the DP cells are positively selected to generate mature single positive (SP) helper (CD4+CD8) and cytotoxic (CD4CD8+) T cells (9). The developmental progression from immature to mature thymocytes occurs sequentially and in distinct thymic microenvironments (10).

The typical architecture of the complex thymic stroma is critically dependent on intercellular communications. Early in ontogeny, fibroblastoid cells control the differentiation of mesenchymal cells to form a regular epithelial compartment. Depending on cell-cell interactions between developing thymocytes and stromal cells (termed thymic cross-talk), distinct microenvironments are created that allow all steps in T cell maturation to occur. Importantly, the lack of inductive signals from developing T cells prevents the formation of the distinct cortical and medullary microenvironments (11, 12). Consequently, the absence of a normal stromal compartment hinders the orderly maturation of thymocytes and impedes the establishment of self-tolerance (13, 14, 15). Conversely, diverse experimental models of autoimmunity, such as Chagas disease, scleroderma, lupus erythematosus, and insulin-dependent diabetes, have been correlated with an obvious disorganization of thymic stromal architecture (16).

Clonal deletion and the induction of anergy represent two major mechanisms that establish self-tolerance among thymocytes. Immunity to self-Ags is also prevented by the presence of thymus-derived T cells, which play an active role in regulating the autoreactive potential of cells that have neither been clonally deleted nor rendered anergic (17, 18, 19, 20). However, the relative importance of these mechanisms for the establishment of self-tolerance remains to be determined for particular Ags.

The cellular and molecular mechanisms responsible for many forms of autoimmune diseases have yet to be defined due to the fact that these conditions are polygenic and have different environmental triggers responsible for their clinical emergence. In contrast, monogenic autoimmune syndromes provide appropriate models to gain further in-depth insights into the complex molecular processes associated with the loss of self-tolerance. Autoimmune-polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED, also known as autoimmune polyglandular syndrome 1; OMIM 240300) is a rare autosomal recessive disease with no known HLA association. Along with the autoimmune lymphoproliferative syndromes 1 and 2 (21, 22), APECED is recognized as one of three autoimmune disorders known to be caused by a single gene defect.

The gene responsible for APECED has been designated AIRE (autoimmune regulator) and has been mapped to chromosome 21q22.3 (23, 24, 25). Mutations in the coding region of AIRE result in truncated proteins devoid of a normal function. The murine Aire gene has recently been cloned and mapped to chromosome 10, revealing a structural organization and sequence homology highly conserved to its human ortholog (26, 27, 28). Aire encodes a predicted protein of 552 aa that contains structural features that anticipate a role in gene transcription. The protein embodies a proline-rich region, four LXXLL motifs, two plant PHD zinc-finger domains (which are restricted to nuclear proteins including transcriptional coactivators and chromatin-modulating proteins of the polycomb and trithorax groups), and a SAND domain (a sequence present in Sp100, Aire, NucP41/75 and DEAF-1/suppressin) (29, 30, 31). Moreover, the AIRE gene product is localized to distinct spherical nuclear structures, further corroborating its putative function as a regulator of gene transcription (32).

The autoimmune manifestations of APECED encompass hypoparathyroidism, primary adrenocortical failure, and chronic mucocutaneous candidiasis (29). APECED is characterized by lymphocytic infiltrations, the presence of a wide variety of tissue-specific, T cell-dependent autoantibodies, and a yet-unidentified defect in T cell function (29, 33, 34).

To detail its role in thymic function and in T cell tolerance induction, we investigated the temporal and spatial expression of Aire in embryonic and adult mice and in models of positive and negative thymic selection. Here we report that Aire expression is restricted to a distinct subpopulation of TEC, and its expression requires thymocyte stromal cell-cell interactions, a critical prerequisite for the generation of an appropriate thymic architecture. Moreover, Aire expression is dependent on RelB function, and the number of Aire-positive thymic stromal cells correlates with the presence of thymocytes undergoing negative selection.

The BALB/c, C57BL/6, Rag-2−/−, Tgε26, B6.RAG-2−/− I-Ab, and B6.RAG-2−/− I-Abm12 mouse strains were housed at the Animal Facility of the Kantonsspital Basel and the Basel Institute for Immunology, respectively, according to the Institutional Review Boards. The B6.RAG-2−/− I-Ab and B6.RAG-2−/− I-Abm12 mice express the 3BBM74 TCR as a transgene on the RAG-2null background (see Ref. 35).

Thymic tissue was obtained from adult mice and from embryos after timed pregnancies where detection of the vaginal plug was considered as day 0 of gestation (E0). Thymic tissue from RelB−/− mice was kindly provided by Dr. Li Wu (Melbourne, Victoria, Australia) and Dr. Philippe Naquet (Marseille, France). Thymic tissue was embedded in OCT (Tissue-Tec, Miles, Elkhart, IN) for analysis by immunohistology and in situ hybridization (ISH).

Thymocytes were obtained by tissue disruption using frosted glass slides, whereas TEC were prepared by enzymatic digestion of whole thymi as described elsewhere (36, 37). After enrichment of low-density cells by Percoll gradient for the sorting of TEC, cell subsets were stained with a combination of CD3, CD4, CD8, CD11c, and CD45 (all from PharMingen, San Diego, CA), as well as mAbs 29 (36), G8.8 (38), and CDR1 (39), respectively. The different lymphocyte and stromal subpopulations were sorted by flow cytometry (FACStar; Becton Dickinson, Mountain View, CA) to a purity of at least 98%.

Total RNA was isolated from tissues or sorted cells using TRI reagent (Molecular Research Center, Cincinnati, OH). Sorted cells were supplemented with 10 μg of yeast transfer RNA (Life Technologies, Basel, Switzerland) as a carrier. For the generation of cDNA, RNA was treated with RNase-free DNaseI (Roche Molecular Biochemicals, Gipf-Oberfrick, Switzerland) and then reverse transcribed using SuperScript II reverse transcriptase with either oligo(dT) (for tissues) or random hexamers (for sorted cells) as primers (Life Technologies). For PCR, various amounts of cDNA were used with 1× Taq PCR buffer (1.5 mM Mg2+), 0.5 U Taq, 0.2 mM dNTP (all from Sigma, St. Louis, MO), and 0.4 μM of each oligonucleotide. The following oligonucleotide pairs were designed from publicly available data: Aire, 5′-TGC ATA GCA TCC TGG ACG GCT TCC and 5′-CCT GGG CTG GAG ACG CTC TTT GAG; as well as 5′-TCT ACT GAG TGC TGG GAA TGA G, and 5′-CAG GAA GAG AAG GGT GGT GTC (see Fig. 3,B); Gapdh, 5′-ACC ACA GTC CAT GCC ATC AC and 5′-TCC ACC ACC CTG TTG CTG TA; Whn, 5′ATG GAG ACC TTG GGA CTG AC and 5′-TGG CTG AGT GGC ATA GGA GA; pTα, 5′-ATC ACA CTG CTG GTA TAT GGA and 5′-TCA GAG GGG TGG GTA AGA TC; β-actin, 5′-GTC GGC CGC TCT AGG CAC CAA and 5′-CTC TTT GAT GTC ACG CAG GAT TTC. PCR amplification for cDNA obtained from tissues used 25, 26, or 30 cycles, respectively, as indicated for Gapdh, 28 cycles for Whn and pTα, and 30 cycles for Aire. Amplification of cDNA from sorted cells used 30 cycles for Gapdh and β-actin and 35 cycles for Aire. For data shown in Fig. 3 B, the relative amount of the first-strand cDNAs produced from each sorted stromal subpopulation was estimated after amplification of a reference β-actin cDNA fragment. PCR products were separated on a 1.7% agarose gel, visualized by staining with SYBR Gold (Molecular Probes, Eugene, OR), and images were analyzed using the QuantityOne gel-doc system (Bio-Rad, Richmond, CA).

FIGURE 3.

Aire expression is restricted to a subpopulation of TEC. A, ISH of adult thymic tissue localizes Aire expression to cells at the corticomedullary junction and in the medulla. Counter-staining with Methylene Green (original magnification, ×10). B, RT-PCR analysis for Aire expression by distinct subpopulations of thymic stroma cells: MHCII+ CD11c+ dendritic cells and Ag 29-positive stromal cells that are either CD11c (medullary epithelial cells) or CD11c+ (dendritic cells). C, Immunohistology for 29 Ag expression in thymic tissue of Tgε26 and RAG-2null mutant mice (original magnification, ×40).

FIGURE 3.

Aire expression is restricted to a subpopulation of TEC. A, ISH of adult thymic tissue localizes Aire expression to cells at the corticomedullary junction and in the medulla. Counter-staining with Methylene Green (original magnification, ×10). B, RT-PCR analysis for Aire expression by distinct subpopulations of thymic stroma cells: MHCII+ CD11c+ dendritic cells and Ag 29-positive stromal cells that are either CD11c (medullary epithelial cells) or CD11c+ (dendritic cells). C, Immunohistology for 29 Ag expression in thymic tissue of Tgε26 and RAG-2null mutant mice (original magnification, ×40).

Close modal

Tissues from 6-wk-old C57BL/6 mice were collected and snap frozen in liquid nitrogen. Total RNA was isolated using TRI reagent, and mRNA was isolated from total RNA using the Oligotex kit (Qiagen). For each sample, 1 μg of mRNA was electrophoresed in a 1% agarose-formaldehyde gel, transferred to Hybond N+ membrane (Amersham Pharmacia Biotech, Uppsala, Sweden) by overnight capillary blotting in 10× SSC. The membrane was probed with [α32-P]-labeled Aire (54–630 bp; accession no. AJ132243) or Gapdh (566–1017 bp; accession no. M32599) cDNA fragments. Hybridization was performed for 2 h at 65°C using the QuickHyb buffer (Stratagene, La Jolla, CA) according to the manufacturer’s instructions. The blot was exposed overnight using a PhosphoImager Screen (Amersham Pharmacia Biotech), and results were analyzed by ImageQuant software (Bio-Rad).

Sense and antisense digoxigenin-labeled (Roche Molecular Biochemicals) cRNA probes were generated by in vitro transcription using cDNA specific for Aire, Whn, and Rag-1 as templates. ISH was performed as previously reported (40). In short, frozen sections (15 μm) were cut from thymic tissue embedded into OCT, air dried at room temperature (RT; 20 min), fixed in 4% paraformaldehyde (10 min), washed tree times with PBS, and finally acetylated (10 min). Prehybridization was performed overnight at RT with hybridization buffer (50% formamide, 5× SSC, 1× Denhardt’s, 100 μg/ml yeast transfer RNA, 100 μg/ml salmon sperm DNA) using a chamber humidified with 5× SSC. The hybridization mixture was prepared by adding 200 ng of digoxigenin-labeled probe per milliliter of hybridization buffer, which was first heated (85°C, 5 min) and subsequently chilled (4°C). The hybridization mixture was spread over the sections, which were then covered with siliconized coverslips and sealed with rubber cement (Stanford, Bellwood, IL). Slides were placed in Quadriperm dishes (Heraeus Instruments, Hanau, Germany), and hybridization was performed overnight at 68°C in a 5× SSC humidified chamber. Slides were washed initially in 5× SSC at 72°C and subsequently in 0.2× SSC at 70–72°C (60 min) followed by 0.2× SSC at RT (5 min). Next, slides were rinsed with 0.1 M maleic acid, 0.15 M NaCl buffer (pH 7.5), blocked for 1 h in 1% blocking reagent (Roche Molecular Biochemicals) diluted in maleic acid buffer, and incubated for 1 h with alkaline phosphatase-conjugated anti-digoxigenin Ab (Roche Molecular Biochemicals) diluted in blocking solution (1:2500). Slides were washed for 1 × 10 min and 2 × 30 min in equilibrated maleic acid buffer (100 mM Tris-HCl, 100 mM NaCl, 5 mM MgCl2, pH 9.5). For color development, slides were placed upside-down in a Quadriperm dish containing nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl-phosphate substrate (Roche Molecular Biochemicals; nitroblue tetrazolium, 112.5 μl; 5-bromo-4-chloro-3-indolyl-phosphate, 87.5 μl; diluted in equilibrated maleic acid buffer). The reaction was performed in the dark at RT and stopped by transferring the slides into PBS containing 1 mM EDTA (10 min). The tissue was counterstained with Methylene Green (0.05% in PBS), and slides were coverslipped using Kaiser’s gelatin (Merck, Darmstadt, Germany). Positive signals are represented by dots of strong purple color, while unspecific signals are detected as small and diffuse dots of weaker intensity. These latter signals are also present in the absence of a specific probe (data not shown).

Frozen thymic sections (5 μm) were fixed with acetone (Merck) at RT for 15 min and washed in PBS. Endogenous peroxidase activity was blocked by a 15-min incubation with 3% H2O2, 0.1% NaN3 in PBS. The sections were blocked with normal goat serum (10% in PBS) for 30 min. Individual thymic sections were incubated with the appropriate dilution of primary Ab (30 min, RT), washed three times in PBS, and incubated with HRP-conjugated goat anti-rat Ig for 30 min. After additional washing in PBS, the substrate amino-ethyl-carbazole (Sigma) was applied to the sections. The tissues were washed in PBS, counterstained with Mayer’s Hemalaun (Merck), mounted with Crystal/Mount (Biomeda, Foster City, CA), and coverslipped.

To detail the expression pattern of Aire in different murine tissues, a Northern blot analysis was performed. As shown in Fig. 1 A, specific mRNA was detected mainly in thymus, and to a lesser extent in spleen and lymph nodes. At least three distinct bands were consistently observed in each of these organs, suggesting the existence of multiple Aire transcripts.

FIGURE 1.

Expression of Aire in different murine tissues. A, Northern blot analysis of mRNA from thymus (Th), liver (Li), brain (Br), spleen (Sp), bone marrow (BM), lymph node (LN), stomach (St), skin (Sk), and kidney (K) analyzed with a probe specific for Aire and Gapdh, respectively. The markers on the right represent the 28S (upper band) and the 18S (lower band) ribosomal RNAs. B, RT-PCR analysis of mRNA from CDR1+G8.8+ TEC and from isolated thymocyte subpopulations. Thymic tissue was either gently digested for the retrieval of stroma cells or mechanically separated for the isolation of thymocytes using frosted glass slides. Cell suspensions were stained and sorted by flow cytometry as outlined in Materials and Methods. Thymocytes were defined by their characteristic forward/side scatter and enriched for the phenotypic subpopulations of CD3CD4CD8 (TN), CD4+CD8+ (DP), or either mature CD4+CD8 (SP4) or CD4CD8+ (SP8) thymocytes. Total thymic RNA not subjected to reverse transcription was used as a negative control (designated RNA). C, Expression of Aire, Whn, pTα, and Gapdh was assessed by RT-PCR using thymic tissues from embryos at different developmental ages (days post conception, p.c.) and from adult mice.

FIGURE 1.

Expression of Aire in different murine tissues. A, Northern blot analysis of mRNA from thymus (Th), liver (Li), brain (Br), spleen (Sp), bone marrow (BM), lymph node (LN), stomach (St), skin (Sk), and kidney (K) analyzed with a probe specific for Aire and Gapdh, respectively. The markers on the right represent the 28S (upper band) and the 18S (lower band) ribosomal RNAs. B, RT-PCR analysis of mRNA from CDR1+G8.8+ TEC and from isolated thymocyte subpopulations. Thymic tissue was either gently digested for the retrieval of stroma cells or mechanically separated for the isolation of thymocytes using frosted glass slides. Cell suspensions were stained and sorted by flow cytometry as outlined in Materials and Methods. Thymocytes were defined by their characteristic forward/side scatter and enriched for the phenotypic subpopulations of CD3CD4CD8 (TN), CD4+CD8+ (DP), or either mature CD4+CD8 (SP4) or CD4CD8+ (SP8) thymocytes. Total thymic RNA not subjected to reverse transcription was used as a negative control (designated RNA). C, Expression of Aire, Whn, pTα, and Gapdh was assessed by RT-PCR using thymic tissues from embryos at different developmental ages (days post conception, p.c.) and from adult mice.

Close modal

To ascribe Aire expression to a distinct thymic cell lineage, RT-PCR was performed on cDNA from freshly isolated subpopulations of thymocytes, TEC and thymic dendritic cells, respectively. While thymocytes at all maturational stages failed to express Aire, analysis from cDNA obtained from freshly isolated TEC resulted in amplification of Aire transcripts (Fig. 1,B). The use of Aire-specific primers with cDNA from MHC class II+ CD11c+ thymic dendritic cells resulted only in a very faint amplification product (see below and Fig. 3 B). Thus, stromal cells but not lymphoid cells (at any stage of their intrathymic development) are responsible for Aire expression in adult thymic tissue.

Next we determined the pattern of Aire expression during thymic ontogeny using RT-PCR. For this purpose, RNA was isolated from thymic tissue of C57BL/6 embryos taken at distinct gestational ages and analyzed for Aire, Whn, and pTα transcripts. The latter two transcripts were used to monitor the two dominant cell types present during early thymic ontogeny, i.e., epithelial cells and immature thymocytes. Whn is a transcription factor typically expressed in all thymic epithelial cells and critical for their growth and differentiation (41, 42). In contrast, pTα is a T lymphoid-specific cell-surface molecule expressed exclusively in early thymocytes, where it is essential for the maturational transition of a CD44CD25+ (TN III) to a CD44CD25 (TN IV) phenotype (43, 44).

Aire-specific cDNA was detected only after E14, while Whn- and pTα-specific transcripts were already apparent at E12 and E13, respectively. By day 14, commitment to thymocytes of the TCR αβ lineage has already occurred as demonstrated by the transcription of the pTα gene (Fig. 1,C and Ref. 45). Moreover, thymic epithelial subpopulations with a cortical and medullary phenotype can be distinguished at this point in time (36). Abundant Aire expression was demonstrated at E16, a time during thymic development when the formation of DP thymocytes has been initiated but TCR-mediated thymic selection has not yet begun. The relative abundance of Aire transcripts decreased somewhat after day 16 but remained detectable throughout fetal and postnatal life (Fig. 1 C). The observed changes in Whn and Aire expression during thymic development are likely accounted for by shifts in the relative frequency of Aire-positive epithelial cells. Taken together, these results demonstrate that Aire expression in thymic stromal cells emerges relatively late during ontogeny but is sustained into adulthood.

We next determined the spatial expression of Aire mRNA in fetal thymi using ISH. In E16 tissue, Aire expression was detected in small but distinct aggregates scattered throughout the entire organ (Fig. 2,A). In contrast, ISH using a Whn-specific probe revealed a network of cells that represented the majority of epithelial cells at this stage in thymic development (Fig. 2 B). Similarly, detection of a compact network of TEC was also achieved using Abs specific for cytokeratin 18 (data not shown). Thus, comparison of these two distinct staining patterns indicates that only a subpopulation of TEC express Aire-specific transcripts.

FIGURE 2.

Thymic Aire expression at day 16 of embryonic development is localized to single cells and small aggregates dispersed throughout the organ. A, Thymic tissue from wild-type (a and b), RAGnull (c and d), and Tgε26 (e and f) was analyzed by ISH using an antisense probe specific for Aire (a, c, and e) and Whn (b, d, and f), respectively (original magnification of all panels, ×20). B, RT-PCR for Aire (35 cycles) and Gapdh (26 cycles).

FIGURE 2.

Thymic Aire expression at day 16 of embryonic development is localized to single cells and small aggregates dispersed throughout the organ. A, Thymic tissue from wild-type (a and b), RAGnull (c and d), and Tgε26 (e and f) was analyzed by ISH using an antisense probe specific for Aire (a, c, and e) and Whn (b, d, and f), respectively (original magnification of all panels, ×20). B, RT-PCR for Aire (35 cycles) and Gapdh (26 cycles).

Close modal

Organization and differentiation of thymic stromal cells into distinct microenvironments with a typical cellular architecture are subject to inductive signals provided by thymocytes (6, 11, 46). In particular, the comparison of two mouse strains, Tgε26 and RAGnull mice, has been most informative in revealing the identity of the thymocyte subpopulation critical for the formation of a normal thymic microenvironment: TN II and/or TN III thymocytes provide signals that induce the three-dimensional organization of TEC (11). Evidence for this comes from the fact that Tgε26 mutant mice, which overexpress the human CD3ε chain in high copy number, display a complete arrest in early thymocyte development at the transition of TN I to TN II of intrathymic T cell development, a point in time that corresponds to a developmental stage before E14.5 of normal thymic organogenesis. Consequently, the thymic primordium ceases to develop a mature three-dimensional network of TEC (13). In contrast, thymopoiesis in RAGnull mutant mice is blocked later during development, i.e., at the TN III stage, which relates to E15.5 in the thymic developmental in wild-type mice. The cortical stroma of RAGnull mice reveals a normal cellular composition and a typical architectural organization (13).

To test whether TEC express Aire transcripts before their formation of a three-dimensional architecture, thymi from Tgε26 and RAGnull mice were analyzed using ISH. At E16 of development, the pattern of Aire expression in RAGnull thymi was similar to that seen in age-matched thymi from wild-type mice (Fig. 2,A, c). In contrast, thymic Aire transcripts could not be detected in E16 Tgε26 thymi despite the ample presence of TEC (Fig. 2,A, e and f). This result was confirmed by RT-PCR analysis of fetal Tgε26 and RAG thymic tissue (Fig. 2 B). In Tgε26 thymi, the absence of Aire expression correlates with the lack of TN II/III thymocytes, while the presence of Aire transcripts in RAGnull thymi correlates with the presence of thymocytes at these specific developmental stages. These results are consistent with the idea that TN II/III thymocytes induce the expression of Aire in TEC, which correlates with the formation of a normal thymic architecture.

To document the spatial expression of Aire transcripts in a fully differentiated thymic microenvironment, sections of adult thymi were analyzed by ISH (Fig. 3 A). Aire expression was localized to the medulla and the corticomedullary junction. While ISH within the medulla revealed a punctate pattern of staining, the pattern at the corticomedullary junction suggested the formation of a more continuous arrangement of Aire-positive epithelial cells. In particular, the slender extensions of the medulla that protrude into the cortex displayed several layers of Aire-positive cells, thus forming a concise boundary at the corticomedullary junction.

Several surface markers have been described that are expressed on medullary epithelial cells (36, 39, 47, 48, 49, 50). The Ab designated 29 recognizes a subpopulation of TEC thought to represent stromal cells bearing features of “activated” cells (36, 51). Otherwise characterized by Ia-specific Abs and lectin UEA-1 binding, this subpopulation of 29+ epithelial cells has been implicated in tolerance induction (52, 53, 54). To detect Aire expression in 29+ epithelial cells, stromal cells were isolated by gentle digestion of thymic fragments. Phenotypically distinct subpopulations were separated by flow cytometry using a combination of different markers: CD11c+MHCII+ dendritic cells; CD11c+29+ dendritic cells; and CD11c29+ epithelial cells. cDNAs obtained from these different stromal cell populations were amplified by semiquantitative PCR to detect Aire expression. As seen in Fig. 3 B, specific mRNA was preferentially and strongly expressed in the population of CD11c29+ epithelial cells when compared with all other stromal populations tested. Taken together, 29+ epithelial cells in the medulla and at the corticomedullary junction are to a large extent responsible for Aire expression in adult thymic tissue.

To exclude the possibility that the observed absence of Aire expression in Tgε26 (Fig. 2,A, e) is simply caused by the lack of 29+ epithelial cells, Tgε26 thymi at E16 that lack Aire expression were analyzed by immunohistology for reactivity with either a 29-specific Ab or the UEA-1 lectin (Fig. 3 C and data not shown). Epithelial cells staining positively for either marker could be easily detected in thymi of Tgε26 mice. These results provide further evidence that the lack of inductive signals provided by early thymocytes at the TN II/III phenotype—and not the absence of the responsive epithelial target cells—account for the deficit of Aire expression in Tgε26 thymi.

Medullary epithelial cells mediate negative selection of developing thymocytes (52) and contribute to late thymocyte maturation (55). For its architectural organization, the medulla is dependent on distinct signals provided by postselection thymocytes (12). These signals may likely be mediated by the transcription factor RelB (36) as mice deficient for RelB display a medullary thymic atrophy (56, 57) and aberrant clonal deletion of autoreactive thymocytes (15). Both traits may be the direct consequence of a severe decrease of Ag 29+, UEA-1+ thymic medullary epithelial cells (36).

To test whether Aire expression can be detected in thymi devoid of 29+ epithelial cells, mice homozygous for a null mutation of RelB were analyzed by ISH. The comparison with wild-type thymi demonstrated that Aire expression was completely absent in RelBnull thymi (Fig. 4,A, a and b). Identical results were obtained by use of RT-PCR (Fig. 4 B). Moreover, ISH with a RAG-1-specific probe revealed a dramatic change in the organization of the thymic microenvironment: in lieu of a centrally located medulla, RelBnull thymi displayed multiple small medullary foci dispersed throughout the RAG+ cortex (data not shown). Thus, the absence of Aire expression correlated with a homozygous deficiency for RelB and the disruption of a regular thymic architecture despite the presence of medullary epithelial cells other than 29+ stromal cells.

FIGURE 4.

RelBnull mutant mice lack thymic Aire expression. A, Thymic tissues from RelB-deficient (a) and wild-type mice (b) were analyzed by ISH for Aire expression (original magnification of all panels, ×10). B, RT-PCR for Aire (35 cycles) and Gapdh (30 cycles).

FIGURE 4.

RelBnull mutant mice lack thymic Aire expression. A, Thymic tissues from RelB-deficient (a) and wild-type mice (b) were analyzed by ISH for Aire expression (original magnification of all panels, ×10). B, RT-PCR for Aire (35 cycles) and Gapdh (30 cycles).

Close modal

A normal thymic microenvironment is a prerequisite for generating a repertoire of T cells restricted to self-MHC molecules and tolerant to self-Ags. Therefore, we sought to correlate Aire gene expression with thymocyte selection as Aire has been 1) implicated in the maintenance of tolerance and 2) localized to the anatomical site usually associated with negative selection (58). For this aim, a murine model was analyzed where thymocytes express the 3BBM74 TCR transgene on a RAG-2null background (35). Thymocytes bearing this TCR are positively selected by I-Ab but negatively selected by I-Abm12 (59).

Mice of either the positively selecting or the negatively selecting haplotype were analyzed by ISH for thymic Aire expression. In the positively selecting (I-Ab+) thymi, Aire expression was limited to a few scattered cells located within the sparse thymic medulla (Fig. 5,A). In contrast, abundant Aire expression at the corticomedullary junction and in the extended medulla was seen in thymi undergoing negative selection (I-Abm12+, Fig. 5,B). Under conditions of negative selection, abundant Aire expression may relate to the TCR-mediated negative selection signals leading to apoptotic cell death. Alternatively, Aire expression may result from any form of thymocyte death including death by neglect. To address the relevance of death by neglect for Aire expression, thymic tissue from mice deficient for the expression of both MHC class I and II molecules was analyzed (Fig. 5 C). In these MHCnull mutant mice, almost all thymocytes undergo cell death by neglect due to the absence of TCR ligands and the ensuing lack of TCR-mediated survival signals. Thymic tissue from MHCnull mice revealed Aire expression restricted to a few scattered cells within the medulla. This pattern was comparable to the expression noted in mice with only positive thymic selection. Thus, thymic Aire expression is directly correlated to the presence of apoptotic cell death due to negative selection and appears to be independent of mechanisms involved in cell death by neglect.

FIGURE 5.

Negative thymic selection correlates with abundant expression of Aire. Thymic tissue from TCR-transgenic mice displaying exclusively either positive thymic selection (H-2b, A) or negative selection (H-2bm12, B) were analyzed by ISH and compared with thymic tissue from mice deficient for MHC I and II molecules (C) (original magnification of all panels, ×10).

FIGURE 5.

Negative thymic selection correlates with abundant expression of Aire. Thymic tissue from TCR-transgenic mice displaying exclusively either positive thymic selection (H-2b, A) or negative selection (H-2bm12, B) were analyzed by ISH and compared with thymic tissue from mice deficient for MHC I and II molecules (C) (original magnification of all panels, ×10).

Close modal

We investigated the temporal and spatial pattern of Aire transcription to elucidate its role for regular immune functions. Using Northern blot analysis, abundant Aire expression was localized to the thymus and to a lesser degree to spleen and lymph nodes of mice. Interestingly, different Aire transcripts were detected among mRNA from thymus and secondary lymphoid organs, suggesting the presence of splice variants. However, the pattern of intensity for the three distinct bands did not appear to be unique for each tissue tested (Fig. 1 A). The detection of Aire variants in Northern blots is in keeping with previous PCR analysis demonstrating alternatively spliced transcripts in different tissues. In particular, variants with deletions of exon 6, 8, 10, and/or 11, respectively, have been detected among mRNA from thymic tissue (27, 60). For example, the complete deletion of exon 10 fails to cause a change in the reading frame but results in a shortened protein lacking a substantial part of the proline-rich region; whereas the combined loss of exons 10 and 11 introduces a stop codon in exon 13 and deletes the second PHD zinc-finger, a motif present in a number of chromatin-associated transcriptional regulators (61, 62, 63). It remains to be determined whether all of these transcripts encode functional proteins with distinct capacities.

Previously, little was known concerning the transcriptional regulation of Aire. Our experiments now demonstrate that Aire expression is principally restricted to the subpopulation of Ag 29+ medullary TEC located within the medulla and at the corticomedullary junction. During thymic ontogeny, 29+ stromal cells are first morphologically detected around E14 (36) and may initially represent a population of lineage-committed precursors of medullary epithelial cells. The first Aire-specific transcripts are detected at E14 during thymic ontogeny (Fig. 1 C), in parallel with the appearance of TN II thymocytes. Two days later, a network of 29+ medullary stromal cells is established concomitant with the emergence of more mature thymocytes and the abundant expression of RelB (36, 64). In adult mice, the subpopulation of 29+ TEC represent a network of scattered medullary stromal cells with abundant and reticulated cell processes. These cells share many membrane markers known to be critical for Ag presentation to T cells (51) and have indeed been implicated in negative thymic selection (53, 54).

The cross-talk between developing thymocytes and epithelial cells is critical for the induction of a typical thymic microenvironment (6, 11, 46). However, the molecular nature of the signals that mediate this stromal organization have yet to be defined. Aire is a possible epithelial target molecule for such a pivotal interaction as its expression correlates with a normal stromal organization. The comparison between RAGnull and Tgε26 E16 embryos clearly demonstrates that Aire expression does not occur in a cell-autonomous manner but is induced after provision of activation signals mediated by TN II/III thymocytes. However, the molecular nature of the signals responsible for the transcriptional regulation of Aire remain presently unknown. Thus, Aire constitutes to our knowledge the first epithelial gene product induced by developing early thymocytes and associated with the correct establishment of a regular thymic microenvironment.

RelB belongs to the NF-κB family of transcription factors that are characterized by distinct structural features, interaction with each other and regulation via the IκB inhibitor (65). RelB expression is detected in thymic tissue during embryo development and, after differentiation into distinct stromal compartments, exclusively confined to the medulla (64). In RelB-deficient mice, the outright absence of 29+ thymic stromal cells and the decreased number of thymic dendritic cells has been correlated with an irregular medullary architecture (56, 57) and the loss of efficient negative selection (15, 66). Therefore, our observation of a complete deficiency of Aire transcripts in RelBnull mice may be explained by the absence of 29+ medullary stromal cells. In contrast, it is unlikely that RelB transcription factors constitute a strict requirement for Aire expression, which, in turn, determines the fate (e.g., growth and differentiation) of 29+ epithelial cells because Tgε26 thymi harbor 29+ epithelial cells but lack Aire expression. Therefore, the conclusion can be drawn that Aire expression is not an unconditional requirement for the generation and maintenance of 29+ medullary epithelial cells. But it remains to be formally determined whether Aire transcription is directly RelB dependent. However, the lack of canonical RelB-binding sequences within the first 600 bp immediately 5′ of the Aire start site suggests that Aire transcription is independent of this transcription factor (28).

The molecular properties of the Aire gene product predict a function as a regulator of gene transcription. In this capacity, Aire may regulate the architectural organization of the thymic microenvironment via transcriptional control of downstream target genes. This notion is supported by our data demonstrating the complete absence of Aire transcripts and the aberrant organization of the thymic microenvironment in Tgε26 mutant mice. In these animals, an appropriate corticomedullary differentiation is missing, epithelial cells are organized in an abnormal two-dimensional fashion, and the medullary foci are scattered throughout the entire thymus (11). Importantly, morphological similarities exist between Tgε26- and RelB-deficient mice, as both mutant mice display scattered and poorly separated areas of medullary epithelial cells in lieu of a centrally located medulla (36, 56). Furthermore, in RelBnull mice, the lack of distinct stromal compartments is reflected by the abnormal spatial distribution of areas where TCR gene recombination occurs (data not shown).

Signals transmitted via the TCR/peptide/MHC ligand interaction during later stages of thymocyte development affect the organization of the medullary architecture (46, 67) and determine the selectional fate of developing thymocytes. To correlate Aire expression with thymic selection, RAG-2-deficient mice transgenic for a TCR that is positively selected by I-Ab and negatively selected by I-Abm12 were studied. Both Aire expression and the size of the medullary compartment were strikingly different in the positive (I-Ab) and negative (I-Abm12) selecting backgrounds. In thymi undergoing positive selection, Aire-positive medullary epithelial cells were scattered throughout the organ as either single cells or small densely packed aggregates occasionally located close to the thymic capsule. In contrast, thymi undergoing only negative selection contained a large number of Aire-positive epithelial cells in the medulla and at the corticomedullary junction. Aire expression in epithelial cells is associated with TCR-mediated programmed cell death because there are very few Aire-positive cells in MHCnull thymi, which lack TCR ligands and cause thymocyte death by neglect.

Taken together, the results observed in fetal and adult thymic tissue support the hypothesis that thymic Aire expression is regulated at two distinct steps during thymic ontogeny. First, Aire expression is induced in TEC by the presence of TN II/III thymocytes and confers to the formation of a functional microenvironment capable to effect thymocyte selection. Second, Aire expression by epithelial cells is largely modulated at a later stage of thymic development when DP thymocytes are subjected to selection.

Although the affinity/avidity model of thymic selection predicts that the interaction of the TCR with its peptide/MHC ligand determines the developmental fate of an immature thymocyte, the exact tissue requirements for negative selection are presently a point of discussion. However, ample evidence exists that medullary epithelial cells are capable to delete or anergize self-reactive T cells (52, 54, 68). Here we have demonstrated that Aire expression 1) correlates with the correct structural organization of the thymic microenvironment, 2) is localized to cells and anatomical sites known to effect negative selection, and 3) is modulated by thymocytes undergoing negative selection. Because all of these features are critical for appropriate thymic function, it is conceivable that Aire mutations as observed in APECED patients may affect thymic T cell selection and the formation of self-tolerance.

We thank Philippe Naquet (Marseille, France) and Werner Krenger (Basel, Switzerland) for helpful discussions and critical reading of the manuscript; Sandrine Guérin for cell sorting and PCR analysis of thymic stromal cells; Mathias Merkenschlager (London, U.K.) for thymic stromal cDNA; and Barbara Hausmann (Basel, Switzerland) for technical assistance. The Basel Institute for Immunology was founded and is supported by F. Hoffmann, LaRoche & Company, Basel, Switzerland.

1

This work was supported by grants from the Swiss National Science Foundation (3100-043600.95 and 31-55820.98) and the Helmut Horten Foundation.

4

Abbreviations used in this paper: E, embryonic day; Aire, autoimmune regulator; APECED, autoimmune-polyendocrinopathy-candidiasis-ectodermal dystrophy; DP, double positive, i.e., CD4+CD8+; ISH, in situ hybridization; pTα, pre-TCR α-chain; TEC, thymic epithelial cell; TN, triple negative, i.e., CD3CD4CD8; SP, single positive; RT, room temperature.

1
Bockman, D. E., M. L. Kirby.
1984
. Dependence of thymus development on derivatives of the neural crest.
Science
223
:
498
2
Röpke, C., P. Van Soest, P. P. Platenburg, W. Van Ewijk.
1995
. A common stem cell for murine cortical and medullary thymic epithelial cells?.
Dev. Immunol.
4
:
149
3
Blackburn, C. C., C. L. Augustine, R. Li, R. P. Harvey, M. A. Malin, R. L. Boyd, J. F. Miller, G. Morahan.
1996
. The nu gene acts cell-autonomously and is required for differentiation of thymic epithelial progenitors.
Proc. Natl. Acad. Sci. USA
93
:
5742
4
Amagai, T., M. Itoi, Y. Kondo.
1995
. Limited development capacity of the earliest embryonic murine thymus.
Eur. J. Immunol.
25
:
757
5
von Gaudecker, B..
1991
. Functional histology of the human thymus.
Anat. Embryol.
183
:
1
6
van Ewijk, W., B. Wang, G. Holländer, H. Kawamoto, E. Spanopoulou, M. Itoi, T. Amagai, Y. F. Jiang, W. T. V. Germeraad, W. F. Chen, Y. Katsura.
1999
. Thymic microenvironments, 3-D versus 2-D?.
Semin. Immunol.
11
:
57
7
Holländer, G., S. J. Burakoff.
1997
. Thymic T cell development. J. L. M. Ferrara, and J. Deeg, and S. J. Burakoff, eds.
Graft Versus Host Disease
1
Marcel Dekker, New York.
8
Godfrey, D. I., J. Kennedy, T. Suda, A. Zlotnik.
1993
. A developmental pathway involving four phenotypically and functionally distinct subsets of CD3CD4CD8 triple-negative adult mouse thymocytes defined by CD44 and CD25 expression.
J. Immunol.
150
:
4244
9
Sebzda, E., S. Mariathasan, T. Ohteki, R. Jones, M. F. Bachmann, P. S. Ohashi.
1999
. Selection of the T cell repertoire.
Annu. Rev. Immunol.
17
:
829
10
Boyd, R. L., C. L. Tucek, D. I. Godfrey, D. J. Izon, T. J. Wilson, N. J. Davidson, A. G. Bean, H. M. Ladyman, M. A. Ritter, P. Hugo.
1993
. The thymic microenvironment.
Immunol. Today
14
:
445
11
Holländer, G. A., B. Wang, A. Nichogiannopoulou, P. P. Platenburg, W. van Ewijk, S. J. Burakoff, J.-C. Gutierrez-Ramos, C. Terhorst.
1995
. Developmental control point in induction of thymic cortex regulated by a subpopulation of prothymocytes.
Nature
373
:
350
12
Shores, E. W., W. van Ewijk, A. Singer.
1991
. Reorganisation and restoration of thymic medullary epithelial cells in T cell receptor-negative SCID mice: evidence that receptor-bearing lymphocytes influence maturation of the thymic microenvironment.
Eur. J. Immunol.
21
:
1657
13
Holländer, G. A., S. Simpson, E. Mitzgouchi, A. Nichogiannopoulou, J. She, J.-C. Gutierrez-Ramos, A. K. Bhan, S. J. Burakoff, B. Wang, C. Terhorst.
1995
. Severe colitis in mice with aberrant thymic selection.
Immunity
3
:
27
14
van Ewijk, W..
1991
. T-cell differentiation is influenced by thymic microenvironments.
Annu. Rev. Immunol.
9
:
591
15
Laufer, T. M., J. DeKoning, J. S. Markowitz, D. Lo, L. H. Glimcher.
1996
. Unopposed positive selection and autoreactivity in mice expressing class II MHC only on thymic cortex.
Nature
383
:
81
16
Takeoka, Y., S. Y. Chen, R. L. Boyd, K. Tsuneyama, N. Taguchi, S. Morita, H. Yago, S. Suehiro, A. A. Ansari, L. D. Shultz, M. E. Gershwin.
1997
. A comparative analysis of the murine thymic microenvironment in normal, autoimmune, and immunodeficiency states.
Dev. Immunol.
5
:
79
17
Saoudi, A., B. Seddon, V. Heath, D. Fowell, D. Mason.
1996
. The physiological role of regulatory T cells in the prevention of autoimmunity: the function of the thymus in the generation of the regulatory T cell subset.
Immunol. Rev.
149
:
195
18
Groux, H., A. O’Garra, M. Bigler, M. Rouleau, S. Antonenko, J. E. de Vries, M. G. Roncarolo.
1997
. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis.
Nature
389
:
737
19
Thornton, A. M., E. M. Shevach.
1998
. CD4+CD25+ immunoregulatory T cells suppress polyclonal T cell activation in vitro by inhibiting interleukin 2 production.
J. Exp. Med.
188
:
287
20
Seddon, B., D. Mason.
1999
. Regulatory T cells in the control of autoimmunity: the essential role of transforming growth factor β and interleukin 4 in the prevention of autoimmune thyroiditis in rats by peripheral CD4+CD45RC cells and CD4+CD8 thymocytes.
J. Exp. Med.
189
:
279
21
Straus, S. E., M. Sneller, M. J. Lenardo, J. M. Puck, W. Strober.
1999
. An inherited disorder of lymphocyte apoptosis: the autoimmune lymphoproliferative syndrome.
Ann. Intern. Med.
130
:
591
22
Wang, J., L. Zheng, A. Lobito, F. K. Chan, J. Dale, M. Sneller, X. Yao, J. M. Puck, S. E. Straus, M. J. Lenardo.
1999
. Inherited human caspase 10 mutations underlie defective lymphocyte and dendritic cell apoptosis in autoimmune lymphoproliferative syndrome type II.
Cell
98
:
47
23
Aaltonen, J., P. Bjorses, L. Sandkuijl, J. Perheentupa, L. Peltonen.
1994
. An autosomal locus causing autoimmune disease: autoimmune polyglandular disease type I assigned to chromosome 21.
Nat. Genet.
8
:
83
24
Nagamine, K., P. Peterson, H. S. Scott, J. Kudoh, S. Minoshima, M. Heino, K. J. Krohn, M. D. Lalioti, P. E. Mullis, S. E. Antonarakis, et al
1997
. Positional cloning of the APECED gene.
Nat. Genet.
17
:
393
25
The Finnish-German APECED Consortium. 1997. An autoimmune disease, APECED, caused by mutations in a novel gene featuring two PHD-type zinc-finger domains. Nat. Genet. 17:399.
26
Wang, C. Y., J. D. Shi, A. Davoodi-Semiromi, J. X. She.
1999
. Cloning of Aire, the mouse homologue of the autoimmune regulator (AIRE) gene responsible for autoimmune polyglandular syndrome type 1 (ASP1).
Genomics
55
:
322
27
Blechschmidt, K., M. Schweiger, K. Wertz, R. Poulson, H. M. Christensen, A. Rosenthal, H. Lehrach, M. L. Yaspo.
1999
. The mouse Aire gene: comparative genomic sequencing, gene organization, and expression.
Genome Res.
9
:
158
28
Mittaz, L., C. Rossier, M. Heino, P. Peterson, K. J. Krohn, A. Gos, M. A. Morris, J. Kudoh, N. Shimizu, S. E. Antonarakis, H. S. Scott.
1999
. Isolation and characterization of the mouse Aire gene.
Biochem. Biophys. Res. Commun.
255
:
483
29
Björses, P., J. Aaltonen, N. Horelli-Kuitunen, M. L. Yaspo, L. Peltonen.
1998
. Gene defect behind APECED: a new clue to autoimmunity.
Hum. Mol. Genet.
7
:
1547
30
Gibson, T. J., C. Ramu, C. Gemund, R. Aasland.
1998
. The APECED polyglandular autoimmune syndrome protein, AIRE-1, contains the SAND domain and is probably a transcription factor.
Trends Biochem. Sci.
23
:
242
31
Aasland, R., T. J. Gibson, A. F. Stewart.
1995
. The PHD finger: implications for chromatin-mediated transcriptional regulation.
Trends Biochem. Sci.
20
:
56
32
Björses, P., M. Pelto-Huikko, J. Kaukonen, J. Aaltonen, L. Peltonen, I. Ulmanen.
1999
. Localization of the APECED protein in distinct nuclear structures.
Hum. Mol. Genet.
8
:
259
33
Chilgren, R. A., P. G. Quie, H. J. Meuwissen, R. Hong.
1967
. Chronic mucocutaneous candidiasis, deficiency of delayed hypersensitivity, and selective local antibody defect.
Lancet
2
:
688
34
Fidel, P. L., Jr, J. D. Sobel.
1994
. The role of cell-mediated immunity in candidiasis.
Trends Microbiol.
2
:
202
35
Backstrom, B. T., U. Müller, B. Hausmann, E. Palmer.
1998
. Positive selection through a motif in the αβ T cell receptor.
Science
281
:
835
36
Naspetti, M., M. Aurrand-Lions, J. DeKoning, M. Malissen, F. Galland, D. Lo, P. Naquet.
1997
. Thymocytes and RelB-dependent medullary epithelial cells provide growth-promoting and organization signals, respectively, to thymic medullary stromal cells.
Eur. J. Immunol.
27
:
1392
37
Wu, L., C. L. Li, K. Shortman.
1996
. Thymic dendritic cell precursors: relationship to the T lymphocyte lineage and phenotype of the dendritic cell progeny.
J. Exp. Med.
184
:
903
38
Farr, A., A. Nelson, J. Truex, S. Hosier.
1991
. Epithelial heterogeneity in the murine thymus: a cell surface glycoprotein expressed by subcapsular and medullary epithelium.
J. Histochem. Cytochem.
39
:
645
39
Rouse, R., L. M. Bolin, J. R. Bender, B. A. Kyewski.
1988
. Monoclonal antibodies reactive with subsets of mouse and human thymic epithelial cells.
J. Histochem. Cytochem.
36
:
1511
40
Schären-Wiemers, N., A. Gerfin-Moser.
1993
. A single protocol to detect transcripts of various types and expression levels in neural tissue and cultured cells: in situ hybridisation using digoxigenin-labelled cRNA probes.
Histochemistry
100
:
431
41
Nehls, M., B. Kyewski, M. Messerle, R. Waldschutz, K. Schuddekopf, A. J. Smith, T. Boehm.
1996
. Two genetically separable steps in the differentiation of thymic epithelium.
Science
272
:
886
42
Schlake, T., M. Schorpp, M. Nehls, T. Boehm.
1997
. The nude gene encodes a sequence-specific DNA binding protein with homologs in organisms that lack an anticipatory immune system.
Proc. Natl. Acad. Sci. USA
94
:
3842
43
Fehling, H. J., A. Krotkova, C. Saint-Ruf, and H. von Boehmer. 1995. Crucial role of the pre-T-cell receptor α gene in development of αβ but not γδ T cells. [Published erratum appears in 1995 Nature 378:419.] Nature 375:795.
44
Saint-Ruf, C., K. Ungewiss, M. Groettrup, L. Bruno, H. J. Fehling, H. von Boehmer.
1994
. Analysis and expression of a cloned pre-T cell receptor gene.
Science
266
:
1208
45
von Boehmer, H., I. Aifantis, J. Feinberg, O. Lechner, C. Saint-Ruf, U. Walter, J. Bier, and A. O. 1999. Pleiotropic changes controlled by the T-cell receptor. Curr. Opin. Immunol. 11:135.
46
Shores, E. W., W. Van Ewijk, A. Singer.
1994
. Maturation of medullary thymic epithelium requires thymocytes expressing fully assembled CD3-TCR complexes.
Int. Immunol.
6
:
1393
47
Kampinga, J., P. Nieuwenhuis, B. Roser, R. Aspinall.
1990
. Differences in turnover between thymic medullary dendritic cells and a subset of cortical macrophages.
J. Immunol.
145
:
1659
48
Hirokawa, K., M. Utsuyama, T. Sado.
1989
. Immunohistological analysis of immigration of thymocyte-precursors into the thymus: evidence for immigration of peripheral T cells into the thymic medulla.
Cell. Immunol.
119
:
160
49
Van Vliet, E., M. Melis, W. Van Ewijk.
1984
. Monoclonal antibodies to stromal cell types of the mouse thymus.
Eur. J. Immunol.
14
:
524
50
Surh, C. D., E. K. Gao, H. Kosaka, D. Lo, C. Ahn, D. B. Murphy, L. Karlsson, P. Peterson, J. Sprent.
1992
. Two subsets of epithelial cells in the thymic medulla.
J. Exp. Med.
176
:
495
51
Naquet, P., M. Naspetti, R. Boyd.
1999
. Development, organization and function of the thymic medulla in normal immunodeficient or autoimmune mice.
Semin. Immunol.
47
:
1147
52
Hoffmann, M. W., J. Allison, J. F. A. P. Miller.
1992
. Tolerance induction by thymic medullary epithelium.
Proc. Natl. Acad. Sci. USA
89
:
2526
53
Degermann, S., C. D. Surh, L. H. Glimcher, J. Sprent, D. Lo.
1994
. B7 expression on thymic medullary epithelium correlates with epithlium mediated deletion of Vβ5+ thymocytes.
J. Immunol.
152
:
3254
54
Burkly, L. C., S. Degermann, J. Longley, J. Hagman, R. L. Brinster, D. Lo, R. A. Flavell.
1993
. Clonal deletion of Vβ5+ T cells by transgenic I-E restricted to thymic medullary epithelium.
J. Immunol.
151
:
3954
55
Scollay, R., D. I. Godfrey.
1995
. Thymic emigration: conveyer belts or lucky dips?.
Immunol. Today
16
:
268
56
Weih, F., D. Carrasco, S. K. Durham, D. S. Barton, C. A. Rizzo, R. P. Ryseck, S. A. Lira, R. Bravo.
1995
. Multiorgan inflammation and hematopoietic abnormalities in mice with a targeted disruption of RelB, a member of the NF-κB/Rel family.
Cell
80
:
331
57
Burkly, L., C. Hession, L. Ogata, C. Reilly, L. A. Marconi, D. Olson, R. Tizard, R. Cate, D. Lo.
1995
. Expression of relB is required for the development of thymic medulla and dendritic cells.
Nature
373
:
531
58
Wack, A., H. M. Ladyman, O. Williams, K. Roderick, M. A. Ritter, D. Kioussis.
1996
. Direct visualization of thymocyte apoptosis in neglect, acute and steady-state negative selection.
Int. Immunol.
8
:
1537
59
DiGiusto, D. L., E. Palmer.
1994
. An analysis of sequence variation in the β chain framework and complementarity determining regions of an allo-reactive T cell receptor.
Mol. Immunol.
31
:
693
60
Ruan, Q. G., C. Y. Wang, J. D. Shi, J. X. She.
1999
. Expression and alternative splicing of the mouse autoimmune regulator gene (Aire).
J Autoimmun.
13
:
307
61
Kim, S. S., Y. M. Chen, E. O’Leary, R. Witzgall, M. Vidal, J. V. Bonventre.
1996
. A novel member of the RING finger family, KRIP-1, associates with the KRAB-A transcriptional repressor domain of zinc finger proteins.
Proc. Natl. Acad. Sci. USA
93
:
15299
62
Thenot, S., C. Henriquet, H. Rochefort, V. Cavailles.
1997
. Differential interaction of nuclear receptors with the putative human transcriptional coactivator hTIF1.
J. Biol. Chem.
272
:
12062
63
Ge, Q., D. S. Nilasena, C. A. O’Brien, M. B. Frank, I. N. Targoff.
1995
. Molecular analysis of a major antigenic region of the 240-kD protein of Mi-2 autoantigen.
J. Clin. Invest.
96
:
1730
64
Carrasco, D., R.-P. Ryseck, R. Bravo.
1993
. Expression of relB transcripts during lymphoid organ development: specific expression in dendritic antigen-presenting cells.
Development
118
:
1221
65
Baldwin, A. S., Jr.
1996
. The NF-κB and IκB proteins: new discoveries and insights.
Annu. Rev. Immunol.
14
:
649
66
Wu, L., A. D’Amico, K. D. Winkel, M. Suter, D. Lo, K. Shortman.
1998
. RelB is essential for the development of myeloid-related CD8α- dendritic cells but not of lymphoid-related CD8α+ dendritic cells.
Immunity
9
:
839
67
Brabb, T., E. S. Huseby, T. M. Morgan, D. B. Sant’Angelo, J. Kirchner, A. G. Farr, J. Goverman.
1997
. Thymic stromal organization is regulated by the specificity of T cell receptor/major histocompatibility complex interactions.
Eur. J. Immunol.
27
:
136
68
Hoffmann, M. W., W. R. Heath, D. Ruschmeyer, J. F. Miller.
1995
. Deletion of high-avidity T cells by thymic epithelium.
Proc. Natl. Acad. Sci. USA
92
:
9851