Foxn1 Protein Expression in the Developing, Aging, and Regenerating Thymus

Immanuel Rode, Vera C. Martins, Günter Küblbeck, Nicole Maltry, Claudia Tessmer and Hans-Reimer Rodewald
J Immunol December 15, 2015, 195 (12) 5678-5687; DOI: https://doi.org/10.4049/jimmunol.1502010

Abstract

The forkhead box N1 (Foxn1) protein is the key regulator of thymic epithelial cell (TEC) development, yet how Foxn1 functions remains largely unknown. All mature TECs arise from Foxn1-expressing progenitors/immature TECs and it is widely assumed that TECs as a whole are defined by Foxn1 expression. However, data on the Foxn1 protein are virtually lacking. In this study, we developed novel tools to visualize Foxn1 protein expression at single-cell resolution. We generated Foxn1 knock-in mice expressing a C-terminal hemagglutinin-tagged Foxn1 protein, and a cytometry-grade monoclonal anti-Foxn1 Ab. We evaluated Foxn1 expression patterns in TEC subsets and its dynamics during normal thymus development, aging, injury, and regeneration. Upon challenges, upregulation of Foxn1 was a common feature of thymus regeneration, but the timing of Foxn1 expression changed and the responding TEC subsets depended on the type of treatment. Whereas dexamethasone and recombinant human fibroblast growth factor 7 promoted expansion of Foxn1+Ly51+CD80− TECs, castration led to expansion of Foxn1+Ly51−CD80+ TECs. Collectively, Foxn1 expression is highly heterogeneous in the normal thymus, with large fractions of Foxn1low or Foxn1− TECs accumulating with age. Furthermore, Foxn1 expression is responsive to perturbations.

This article is featured in In This Issue, p.5523

Introduction

Inactivating mutations in the forkhead box N1 (Foxn1) gene cause the athymic “nude” phenotype in humans, rats, and mice (1, 2). The 10 Foxn1 exons encode a 648-aa transcription factor of the forkhead box family. The first noncoding exon 1 is alternatively transcribed, and Foxn1 mRNA comes in two isoforms as either exons 1a–9 or exons 1b–9 (3). Within the Foxn1 protein, the evolutionary conserved DNA binding domain and the activation domain have been functionally characterized in vitro (4, 5).

The thymus, the main site of T cell development, arises from a shared parathyroid/thymus anlage from third pharyngeal pouch endoderm. Separation of the common anlage coincides with the arrival of hematopoietic precursor cells from the bone marrow inside the thymus at embryonic day (E)12 (6, 7). Although Foxn1-deficient mutants lack a functional thymus, a Foxn1-deficient thymic anlage nevertheless forms and migrates to its final mediastinal position above the heart (8). Lack of Foxn1 function causes a developmental arrest in thymic epithelial cells (TECs), presumably at an immature stage in a cell-intrinsic manner (1, 2, 9–13). As a consequence, the Foxn1-deficient epithelium remains alymphoid and TECs fail to further differentiate into cortical and medullary subsets.

The discovery of Foxn1 as the gene mutated in nude mice dates back >20 y (1, 3), yet until now it has been virtually impossible to study this important protein. For example, Foxn1 protein expression has not been amenable to flow cytometry on TEC subsets or to immunoprecipitation. Levels of Foxn1 expression have mostly been inferred from analysis of RNA that suggested declining Foxn1 expression to parallel age-dependent thymus involution (4, 5, 14). Introduction of Foxn1 alone appears to be sufficient to reprogram mouse embryonic fibroblasts into functional TECs (6, 7, 15), suggesting that Foxn1 is a determining master regulator of TEC development even beginning from non-TECs.

In this study, we report the generation and application of new robust tools to study Foxn1 protein expression. We analyzed thymic Foxn1 expression throughout most of the embryonic and adult life of a mouse. Additionally, we investigated Foxn1 expression in the absence of normal T cell development, as well as in the course of thymic insults and regeneration. The results show that already in young mice large fractions of bona fide TECs are Foxn1−/low, precluding Foxn1 as a universal TEC marker, and that Foxn1 is a highly dynamic protein in response to thymic injury.

Materials and Methods

Mice and targeting

Foxn1lacZ/lacZ (8), Foxn1-Cre (16), Rosa26-YFP (17), and Rag2−/− (18) mice were described earlier. C57BL/6J mice were either bred at the animal facility of the German Cancer Research Center or purchased from Charles River Laboratories.

Foxn1HA/HA knock-in mice were generated by conventional gene targeting using the construct shown in Supplemental Fig. 1A. All animal experiments were performed in accordance with institutional and government regulations and were approved by the Regierungspräsidium (Karlsruhe, Germany).

Epithelial cell isolation

Thymic stromal cell isolation was done with slight modifications as described earlier (19). In contrast to the published method, minced thymi were directly subjected to digestion without discarding any fractions, and lymphoid cells (anti-CD45 beads, Miltenyi Biotec) and erythrocytes (anti-Ter119 beads, Miltenyi Biotec) were manually depleted using MACS technology. Embryonic thymi were digested in 96-well V-bottom plates, and the resulting single-cell suspensions were directly used for flow cytometry without depletion.

Alveolar epithelial cells were isolated from the lungs as described (20). Small intestinal epithelial cells were isolated by incubating cleaned and finely minced small intestine fragments three times for 20 min each at 37°C in RPMI 1640 (5% FCS, 1.25 mM EDTA) while shaking. Supernatants were filtered (70 and 30 μm) to yield single-cell supensions.

Flow cytometry and counting

For intracellular stainings (Foxp3 staining kit, eBioscience), cells were first incubated with a Live/Dead fixable violet dead cell stain kit (Life Technologies) and then Fc blocked, and surface stainings were directly blocked. Simultaneous staining of YFP and Foxn1 required replacing the eBioscience fixative with 3.7% paraformaldehyde in PBS. Fluorescein-di-β-d-galactopyranoside (FDG) staining was done with the FluoReporter lacZ flow cytometry kit (Life Technologies). Abs used were CD45 (30F11) Brilliant Violet 421/PE-Cy7, Aire (5H12) Alexa Fluor 488, Ki67 (SolA15) PE-Cy7 (eBioscience), CD326 (G8.8) PerCp-Cy5.5/PE-Cy7/Alexa Fluor 647, CD80 (16-10A1) Brilliant Violet 605, rat IgG1, κ (RTK2071) PE-Cy7 (BioLegend), Ly51 (6C3) PE, CD45 (30F11) FITC (BD Biosciences), anti-hemagglutinin (HA; GG8-1F3.3.1) PE (Miltenyi Biotec), and Foxn1 (2/41) Alexa Fluor 647 (available on request).

For TEC counting, Accudrop beads (BD Biosciences) of a fixed concentration were spiked into all samples just before they were acquired on the LSRFortessa.

Ab generation and labeling

Full-length, native, recombinant mouse Foxn1 was used to immunize mice in collaboration with the mAb Facility at the German Cancer Research Center. One purified clone (2/41, mouse IgG2b) was labeled for all further experiments using the Alexa Fluor 647 Ab labeling kit (Life Technologies).

Dexamethasone treatment

Seven- to eight-week-old Foxn1HA/HA mice received a single i.p. injection of either 20 mg/kg dexamethasone (DEX; Sigma-Aldrich) dissolved in DMSO (Sigma-Aldrich) or an equal volume of DMSO.

Recombinant human fibroblast growth factor 7 injection

Six-day-old C57BL/6J mice were i.p. injected with either 5 mg/kg recombinant human fibroblast growth factor 7 (rhFGF7; Kepivance [palifermin] from Sobi) or an equal amount of solvent (PBS) for 3 consecutive days. Three (i.e., postnatal day 9) and six days later (i.e., postnatal day 12), mice were analyzed by flow cytometry.

Surgical castration

For surgical castration, 9- to 11-wk-old C57BL/6J mice were anesthetized and a small ventral abdominal midline incision was made to reveal the testes. These were removed along with surrounding fatty tissue. Sham castration was performed using the same surgical procedure, but without removal of the testes.

Histology

Embryonic day 14.5 embryos, used for thymus staining in Fig. 1A, were fixed overnight (4% paraformaldehyde, 4°C), and 1-wk-old thymus lobes (Supplemental Fig. 2B) from C57BL/6J and Foxn1HA/HA mice were fixed for 1.5 h. Fixation was followed by cryoprotection in 20% sucrose/PBS (w/v). After OCT embedding in −80°C isopentane, embryonic sections were stained with anti-ΔNp63 (Poly6190, BioLegend) and anti-HA (3F10, Roche). Anti-ΔNp63 was detected with anti-rabbit Alexa Fluor 488 (A-11008, Thermo Fisher Scientific), and anti-HA was detected with anti-rat Alexa Fluor 546 (A-11081, Thermo Fisher Scientific). Endogenous biotin in thymus sections was blocked (E-21390, Thermo Fisher Scientific) before incubation with anti–Foxn1Bio (2/41) and anti-HA (3F10, Roche). The biotinylated (EZ-Link-kit 21327, Thermo Fisher Scientific) Foxn1 Ab was detected with anti–biotin-Cy3 (Jackson ImmunoResearch Laboratories), and anti-HA was detected with anti-rat Alexa Fluor 488 (A-11006, Thermo Fisher Scientific).

Southern blotting

Nonradioactive Southern blotting was performed on SspI/SpeI-digested tail DNA of Foxn1HA/HA knock-in mice using a biotinylated Foxn1 probe (Supplemental Fig. 1A).

Expression systems

Foxn1 and fragments of various lengths were cloned into pEGFP-C1 and electroporated (Gene Pulser Xcell, Bio-Rad) into Cos-1 cells for transient expression.

For immunization, Foxn1 cDNA was cloned into pET-22b(+) (Novagen). Expression of the protein in BL21-CodonPlus (DE3)–RIL bacteria (Agilent Technologies) was induced with isopropyl β-d-thiogalactopyranoside (Sigma-Aldrich).

X-gal staining

Newborn thymi were fixed for 15 min with 2% paraformaldehyde on ice, transferred into X-gal staining solution, and incubated at 37°C with gentle agitation. The staining reaction was stopped by repeated washes with unsupplemented PBS. Staining solution (in unsupplemented PBS) included 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 2 mM MgCl2, 0.01% sodium deoxycholate, 0.02% Nonidet P-40, and 1 mg/ml X-gal.

Results

New tools to measure Foxn1 expression at the single-cell level

Most information on Foxn1 expression is based on RNA data or knock-in/transgenic reporter alleles (8, 14, 21). However, Foxn1 protein and reporter levels may differ due to transcriptional variations, posttranslational modifications, or differences in protein stability comparing the reporter and the native Foxn1 protein. To overcome these disadvantages, we followed two different experimental approaches. First, we engineered a mouse carrying a tagged version of Foxn1, and second, we generated a monoclonal, flow cytometry grade Ab. Because modifications of the Foxn1 locus had previously resulted in hypomorphic alleles (22), we decided to attach the small HA tag of 9 aa to the C terminus of Foxn1 by gene knock-in in embryonic stem cells (Supplemental Fig. 1A, 1B). Homozygous Foxn1HA/HA mice were born with fur and thymus, indicating that the mutant allele encodes a functional Foxn1 protein. We compared total thymus cell numbers and phenotypic stages of T cell development in Foxn1HA/HA to Foxn1+/+ C57BL/6 mice (Supplemental Fig. 1C–E). All parameters were comparable, except for an increase in cell numbers at 6 wk of age in early backcrosses of Foxn1HA/HA mice compared with C57BL/6 controls (Supplemental Fig. 1C). Reanalysis of 10th generation backcrosses of Foxn1HA/HA mice revealed comparable cell numbers also in 6- to 8-wk-old mice (Supplemental Fig. 1C, 1E, red symbols). We next analyzed whether the HA-tagged Foxn1 protein showed a nuclear localization in TECs. Double staining against Foxn1-HA and the epithelial cell–specific transcription factor ΔNp63 revealed colocalization within the nuclei of TECs (Fig. 1A). Collectively, these data show that HA-tagged Foxn1 correctly localizes to the nucleus and functions well in lieu of the normal protein.

FIGURE 1.
FIGURE 1.

New Foxn1 tools are TEC specific. (A) E 14.5 thymus sections of either Foxn1+/+ (upper panel) or Foxn1HA/HA (lower panel) mice were stained with Abs against HA (red, left) and ΔNp63 (green, middle) and counterstained with DAPI (blue, merge on the right; scale bars, 20 μm). One representative out of two independent experiments is shown. (B) Newborn thymic stroma (live, CD45−) of either Foxn1+/+ (upper panel) or Foxn1HA/HA mice (lower panel) was intracellularly stained with Abs against HA (left), Foxn1 (middle), or a combination of both (right). Note that the epitope of the Foxn1 Ab is not affected by the presence of the C-terminal HA tag (middle). One representative out of three experiments is shown. (C) Newborn thymic stroma (live, CD45−) of Foxn1+/+ (left), Foxn1+/lacZ (middle), or homozygous nude Foxn1lacZ/lacZ mice (right) was stained with either FDG (upper panel) or anti-Foxn1 (lower panel). Note that the live, CD45−EpCAM− population shows no signs of Foxn1 staining. One representative out of three independent experiments is shown.

In the second approach, we immunized mice with full-length recombinant native Foxn1 to produce mouse mAbs. One mAb (clone 2/41) was selected for detailed characterization. This Ab stained Cos-1 cells transiently transfected with Foxn1 cDNA (Supplemental Fig. 2A). As further evidence for the specificity of both reagents, double staining with anti-Foxn1 and anti-HA Abs identified the same cells in Foxn1HA/HA thymus sections (Supplemental Fig. 2B). Epitope mapping with deletion constructs determined that clone 2/41 binds the C terminus of Foxn1 between amino acids 475 and 542 (Supplemental Fig. 2C).

Next, we tested the capacity of Abs against HA or Foxn1 (clone 2/41) to recognize the Foxn1-HA fusion protein or normal Foxn1 protein, respectively, by flow cytometry. Newborn TECs from Foxn1+/+ or Foxn1HA/HA mice were intracellularly stained with Abs against HA, the new Foxn1 Ab (clone 2/41), or a combination of both, and analyzed by flow cytometry. Whereas the anti-HA Ab stained the Foxn1HA/HA TECs exclusively (Fig. 1B, left column), the tag-independent anti-Foxn1 Ab stained TECs of both genotypes (Fig. 1B, middle column). When both Abs were used in combination, Foxn1HA/HA but not Foxn1+/+ TECs were double positive, that is, both Abs stained the same cells (Fig. 1B, right column). To further exclude nonspecific labeling, we used Foxn1LacZ knock-in mice, in which Foxn1 is nonfunctional whereas β-galactosidase is produced under the control of the endogenous Foxn1 promoter (8). According to their allelic composition, TECs of this strain expressed homozygously Foxn1 (Foxn1+/+) or homozygously β-galactosidase (Foxn1LacZ/LacZ) or both heterozygously (Foxn1LacZ/+), which is consistent with the blue X-gal staining of whole thymi from the different genotypes (Supplemental Fig. 2D). Replacing chromogenic X-gal with flow cytometry suitable FDG allowed us to analyze TECs from these genotypes by flow cytometry. Simultaneous Foxn1 and FDG staining of cells was technically incompatible, and hence we stained cells either by FDG (indicating at least one Foxn1LacZ allele) (Fig. 1C, top row) or with anti-Foxn1 Ab (indicating at least one Foxn1+ allele) (Fig. 1C, bottom row) versus the pan-epithelial marker EpCAM. Wild-type Foxn1+/+ TECs stained for Foxn1 only (Fig. 1C, left column), heterozygous Foxn1+/LacZ TECs were positive for FDG and anti-Foxn1 (Fig. 1C, center column), and homozygous nude Foxn1LacZ/LacZ TECs were exclusively positive for FDG (Fig. 1C, right column). Therefore, cells with “TEC identity” could be identified in the Foxn1-deficient thymic rudiment by the active Foxn1 promoter (evident by staining with FDG), but these “wannabe” TECs cannot express Foxn1, and lack of their staining proves the specificity of the monoclonal Foxn1 Ab (clone 2/41). In line with this specificity, the CD45−EpCAM− non-TEC stroma remained unstained (Fig. 1C, bottom row). Moreover, we also stained nonthymic epithelium and found that the tested alveolar (Supplemental Fig. 2E) and intestinal (Supplemental Fig. 2F) epithelial cells were not recognized by the Foxn1 Ab based on flow cytometry.

Foxn1 expression during ontogeny

Specification of pharyngeal epithelium toward TECs occurs independent of Foxn1, but further differentiation into cortical and medullary TECs requires Foxn1 (8). During ontogeny, comprehensive information on Foxn1 expression levels and the distribution of expressing TEC subsets is lacking. Analysis of embryonic development (E14.5–18.5) revealed that ∼94% of CD45−EpCAM+ epithelial cells in the thymus expressed Foxn1 (Fig. 2A). Nonepithelial (CD45−EpCAM−) stromal cells were Foxn1− (Fig. 2A). These high percentages of Foxn1+ TECs remained constant over time in prenatal development (Fig. 2B, 2C), irrespective of TEC subset and proliferative state. The small fraction of Foxn1− TECs was initially mostly limited to Ly51−CD80− cells (termed double-negative TECs in this study) and to Ly51+CD80− cells (termed Ly51+ cortical TECS [cTECs]), but some Ly51−CD80+ cells (termed CD80+ medullary TECs [mTECs]) were also detected (Fig. 2A).

FIGURE 2.
FIGURE 2.

Foxn1 expression remains unaltered from E14.5–18.5. (A) Single thymi of developing C57BL/6J mouse embryos were enzymatically digested and analyzed for Foxn1 expression after gating on live, CD45− events. Foxn1 expression by thymic stromal cells at E14.5 (upper panel) and E18.5 (lower panel) is shown on the left. The indicated frequencies refer to the distribution of Foxn1+ and Foxn1− fractions among EpCAM+ TECs. Expression of Ly51 versus CD80 on Foxn1− (middle) and Foxn1+ (right) TECs (live, CD45−, EpCAM+) is shown. One representative experiment out of two is shown (n = 5 per age and experiment). (B) Frequency of Foxn1+ TECs (live, CD45−, EpCAM+) in individual embryonic thymi from E14.5 to E18.5 (n = 10 per age, mean ± SD). (C) Frequency of Foxn1− TECs (live, CD45−, EpCAM+) in individual thymi from E14.5 to E18.5 (n = 10 per age, mean ± SD). (D) Lineage tracing using Rosa26YFP (left) or Foxn::Cre x Rosa26YFP transgenic littermates (right) at E14.5. Expression of YFP and its correlation with Foxn1 expression by TECs (live, CD45−, EpCAM+) is shown. One representative experiment out of two is shown.

To determine whether Foxn1− cells (Fig. 2C) represent a pre- or post-Foxn1 stage, we lineage-traced these cells using Foxn1::Cre transgenic mice crossed to Rosa26-YFP reporter mice (16, 17). In these mice all cells expressing Foxn1 at any time point in their development and their progeny are YFP+. In E14.5 thymi from Foxn1::Cre transgenic mice, but not Cre− control mice, almost all TECs had a history of Foxn1 expression (Fig. 2D). The vast majority of these cells actively expressed Foxn1 (YFP+Foxn1+). However, ∼5% of these lineage-traced cells were Foxn1 protein-negative (YFP+Foxn1−) (Fig. 2D). Collectively, during embryonic development, TECs almost homogeneously express Foxn1, and consequently this is true for TECs with cTEC and mTEC phenotypes. A very small fraction of cells (∼5%) has gone through a Foxn1 RNA-expressing stage, but fails to express the protein. Hence, these cells are “post-Foxn1 mRNA” cells. Additionally, a minor subset of EpCAM+YFP−Foxn1− cells was also found.

Foxn1 expression during adulthood

We examined Foxn1 expression in Foxn1HA/HA mice using our Foxn1 mAB 2/41 between 11 d after birth and up to ∼2 y of age. Flow cytometric analysis using anti-Foxn1 revealed that both expression levels and frequencies were highest in the youngest analyzed mice (Fig. 3A–C), which was verified with an independent dataset obtained by anti-HA staining. Frequencies of Foxn1-expressing cells progressively declined with time (Fig. 3A, 3B). Interestingly, this decline began within the first weeks after birth (Figs. 2A, 3A, 3B) and reached a plateau of ∼50% Foxn1− TECs by 10 wk (Fig. 3A, 3B), which remained constant throughout the observation period. On a per-cell basis, however, Foxn1 downregulation continued throughout life (Fig. 3C). Based on expression of the markers Ly51 and CD80, cells marked by Foxn1 expression always included the major TEC populations (Fig. 3D). In contrast, among the Foxn1− cells, double-negative TECs dominated (Fig. 3E). Taken together, we observed a temporal loss of Foxn1-expressing cells together with a long-lasting, yet slow downregulation of Foxn1 levels in the remaining expressers.

FIGURE 3.
FIGURE 3.

Loss of Foxn1 expression with age depends on lymphocytes. (A) Flow cytometry for Foxn1 expression by thymic stroma (live, CD45−) isolated at the indicated time points. The indicated frequencies refer to the distribution of Foxn1+ and Foxn1− fractions among EpCAM+ TECs. One representative out of three to four experiments is displayed. (B) Summary of Foxn1 expression by TECs. Frequencies of Foxn1+ (▪) and Foxn1− (○) TECs (live, CD45−, EpCAM+, n = 4–5 per age, mean ± SD) are shown. (C) Histogram based on FACS plots shown in (A). Foxn1 stainings of the TEC compartment at various ages are overlaid to visualize differences in expression levels. Phenotype of Foxn1+ (D) and Foxn1− (E) TECs (live, CD45−, EpCAM+) isolated from mice of various ages (as indicated) is shown. One representative out of three to four experiments is displayed. (F) Representative staining depicting Foxn1 expression by thymic stroma (live, CD45−) derived from Rag2-deficient mice of various ages. The indicated frequencies refer to the distribution of Foxn1+ and Foxn1− fractions among EpCAM+ TECs. One representative out of three to four experiments is shown. (G) Summary of the frequencies of Foxn1− (○) and Foxn1+ (▪) TECs from Rag2-deficient mice at various ages (n = 6–8 per age, mean ± SD).

To examine whether Foxn1 expression levels are dependent on normally proceeding T cell development, we analyzed Foxn1 expression in Rag2-deficient mice. In these mutants, T cell development is blocked at the double-negative 3 stage. This state is associated with impaired TEC differentiation, which can be corrected by reconstitution with wild-type thymocytes (23). Intriguingly, in the presence of only immature thymocytes (double-negative 3), a very high fraction of TECs expressed Foxn1 up to 28 wk (Fig. 3F, 3G). Hence, progression beyond the double-negative 3 stage induces loss or downregulation of Foxn1, possibly by inducing TEC maturation.

Foxn1 during thymic regeneration

Loss of Foxn1 during adulthood has been linked to thymic involution (14), an assumption that is supported by the premature thymic degeneration caused by genetic reduction of Foxn1 expression (22, 24). Corroborating these findings, elevated Foxn1 expression has been shown to attenuate thymic involution in multiple settings (24–26). Hence, we were prompted to elucidate Foxn1 expression at the single-cell level in response to thymic insults and/or thymic regeneration. Three conditions are commonly induced in mice to study thymus regeneration, that is, glucocorticoid treatment, growth factor treatment, or castration. Glucocorticoids such as DEX are well known for their multiple detrimental effects on both thymocytes and TECs. In contrast to thymic aging, the DEX effect is transient, and thymocyte as well as TEC numbers recover (27, 28). Conversely, thymus regeneration can be induced by castration of mice, which results in a dramatic, but transient, increase of thymus cellularity (23, 29–33). Finally, the injection of Kepivance (palifermin), a ΔN23KGF form of rhFGF7, has been shown to have regenerating capacity on the thymus (33–35).

DEX injection affected female and male Foxn1HA/HA mice alike by reducing thymocyte and TEC numbers drastically within 2 d. By day 7, there was a clear, albeit not yet complete, recovery of cellularity (Fig. 4A, 4B). After 2 d, Foxn1 expression dropped markedly in response to DEX treatment (Fig. 4C, 4D). In addition to its global TEC diminishing effect, DEX strongly influenced the ratio of Foxn1+CD80+ mTECs to Foxn1+Ly51+ cTECs (Supplemental Fig. 3A). The reduction in Foxn1 expression was temporary, that is, 7 d after injection the frequency of Foxn1-expressing cells had already returned to normal (Fig. 4C, 4D). Regrowth of Ly51−CD80+Foxn1+ TECs did not occur within the 7-d observation period, and hence the altered ratio of Foxn1+CD80+ mTECs to Foxn1+Ly51+ cTECs persisted (Supplemental Fig. 3B). In summary, DEX treatment has a profound effect on Foxn1 expression in the thymus, on proportions of Foxn1-expressing TEC subsets and on their cycling properties. Whereas Foxn1 expression recovers within 7 d, TEC homeostasis does not appear to be equilibrated by this time point.

FIGURE 4.
FIGURE 4.

TEC rebound following DEX treatment. Seven- to 8-wk-old Foxn1HA/HA mice were injected once with DMSO or 20 mg/kg DEX in DMSO and thymic regeneration was analyzed at 2 d (▪, n = 6 per treatment) and 7 d (○, n = 9 per treatment) after injection. Females (red) and males (blue) were analyzed separately. Data represent three independent experiments with two to three mice per group (mean ± SD). Absolute thymocyte (A) and TEC (B) numbers are shown. The frequency of Foxn1-expressing TECs is depicted in (C). (D) Stroma (live, CD45−)-specific expression of Foxn1 at 2 (left) or 7 d after injection (right). The indicated frequencies refer to the distribution of Foxn1+ and Foxn1− fractions among EpCAM+ TECs. One representative out of three experiments is shown. Note that the alterations in EpCAM staining are a result of differently labeled (dye) EpCAM Abs.

Next, we stimulated mice with rhFGF7. rhFGF7 has been shown to have a direct effect on FgfR2-IIIb+ TECs and to cause an initial thymocyte loss followed by a drastic increase in thymic cellularity (33–35). As a result, the stimulated thymus increases in size, in contrast to the recovery to normal size following DEX treatment (28). Injection of rhFGF7 triggered an initial reduction of thymocyte numbers (Fig. 5A) and, already by day 3, an increase in TEC numbers (Fig. 5B), which is consistent with stimulated TEC proliferation on day 3 (Fig. 5C). The frequency of Foxn1-expressing cells declined transiently on day 3 (Fig. 5D). Akin to the DEX recovery, the ratio of Foxn1+CD80+ mTECs to Foxn1+Ly51+ cTECs was altered (Supplemental Fig. 4A). Among Foxn1− and Foxn1+ TECs, Ly51+ cTECs expanded (Supplemental Fig. 4A). Hence, this early thymic stimulation was accompanied by subset changes remarkably similar to the ones observed late during DEX regeneration. The initially major Ly51−CD80+Foxn1+ compartment of control mice was replaced by Ly51intCD80−Foxn1int TECs, which turned into a Ly51+CD80−Foxn1hi population by day 6 (Fig. 5E, right column, Supplemental Fig. 4A). Note that the level of Foxn1 expression on these TECs exceeded the level of control TECs (Fig. 5E, left column).

FIGURE 5.
FIGURE 5.

rhFGF7 treatment generates CD80−Ly51+Foxn1hi TECs. Six-day-old C57BL/6J mice were injected on three consecutive days with either PBS or 5 mg/kg rhFGF7. Thymic cellularity (A), TEC numbers (B), TEC proliferation (C), and Foxn1 expression by TECs (D) were analyzed 3 d (○, three independent experiments with three mice per treatment) or 6 d (▪, three independent experiments with three mice per treatment) after the first injection (mean ± SD). (E) Flow cytometric analysis of Foxn1 expression by thymic stroma (live, CD45−; left) and the phenotype of Foxn1− (middle) and Foxn1+ (right) TECs (live, CD45−EpCAM+) 6 d after the first injection. The indicated frequencies refer to the distribution of Foxn1+ and Foxn1− fractions among EpCAM+ TECs. One representative out of three experiments is shown.

In our final model of thymus modulation, we castrated mice and followed Foxn1 expression over time. Castration led to an increase in thymocyte numbers as early as 8 d after surgery (Fig. 6A). Frequencies of Foxn1 expression did not show major alterations (Fig. 6B). On a per-cell basis, however, all Foxn1-expressing cells had shifted toward a higher expression level at day 14 posttreatment, similar to that of young mice (Fig. 6C). This effect was observed as early as day 8, when thymocyte numbers were already increased, in one third of the cases, and in 100% of the animals at day 14 after surgery, when thymocyte numbers were even higher. Therefore, upregulation of Foxn1 does not seem to be an absolute prerequisite for higher thymopoietic capacity at day 8 of recovery.

FIGURE 6.
FIGURE 6.

Castration generates CD80+Ly51−Foxn1hi TECs. Nine- to eleven-week-old C57BL/6J untreated males (●), sham-castrated controls (□), and castrated males (▴) were analyzed 3, 8, and 14 d after surgery (n = 7–9 per treatment and time point). Thymocyte numbers are shown (A), as are the frequency (B) and levels of Foxn1 expression (C) by day 14 after surgery TECs (live, CD45−EpCAM+; mean ± SD). (D) Foxn1 levels expressed by thymic stroma (live, CD45−; left), the expression of CD80 versus Ly51 on Foxn1− (middle), and Foxn1+ (right) TECs of sham-castrated (top panels) and castrated males (bottom panels) at 14 d after surgery. The indicated frequencies refer to the distribution of Foxn1+ and Foxn1− fractions among EpCAM+ TECs. One representative out of three experiments is shown.

In marked contrast to the treatments described above, where thymic regeneration led to the appearance of a Ly51+Foxn1+ population (Fig. 5E, Supplemental Fig. 3B), CD80+Foxn1+ TECs represented the most abundant population in response to castration (Fig. 6D). This shift in TEC distribution could be first observed 8 d after surgery, and became more prominent 14 d after treatment when these cells uniformly turned Foxn1hi (Fig. 6D, Supplemental Fig. 4B). Collectively, compared with DEX and rhFGF7 treatment, castration has a unique effect on the thymus by strongly expanding CD80+ Foxn1hi TECs.

Discussion

Reliable detection of the Foxn1 protein has been a notorious hurdle in studies on its roles in thymus organogenesis and thymus epithelium. We generated a C-terminal–tagged Foxn1HA allele and a flow-grade Foxn1 Ab. In contrast to a published hypomorphic Foxn1 allele (22), Foxn1HA/HA mice have a phenotypically and functionally normal thymus.

Although data with anti-Foxn1 serum have been reported (21, 36), commercially available antisera did not work robustly in our hands to detect Foxn1. We therefore, in a second approach, raised a tag-independent mAb against endogenous Foxn1. We have mapped the epitope of this mAb to the activation domain, close to the C terminus of Foxn1. The Ab detects Foxn1 by immunofluorescence, by flow cytometry, and by immunoprecipitation in thymus cell lysates.

Foxn1HA/HA mice and the Foxn1-specific mAb enabled us to assess Foxn1 protein expression during thymus development, in thymi with arrested T cell development, during thymus aging, and after thymic insult, as well as regeneration. During prenatal development and up until about 2 wk of age, a very high percentage of all TECs were Foxn1+. By 4 wk of age, still >70% of TECs expressed Foxn1. An earlier report using rabbit anti-Foxn1 serum quantified numbers of Foxn1+ TECs on cytospins and obtained much lower frequencies (36). The reason for this discrepancy is unclear, but it is possible that the representation of Foxn1+ TEC subsets was perturbed on cytospin preparations of TECs, or that the antiserum/assay was less sensitive. The frequencies we report in the present study were comparable in independent experiments using our anti-HA or anti-Foxn1 mAb.

The early “ubiquitous” expression of Foxn1 in the developing thymus is in line with the reported fate-mapping data (37, 38). We addressed the origin of the few Foxn1− TECs in embryonic mice. In line with observations of others, most TECs displayed a history of Foxn1 expression. Costaining of lineage-traced embryonic TECs with anti-Foxn1 Ab identified some EpCAM+ TECs that do not actively express Foxn1 and had not expressed it in the past. The significance of this population is unclear, but it may be of minor importance: due to their EpCAM expression, they cannot represent the recently proposed EpCAM−Foxn1− stem cells, and second, they are unable to rebuild the thymus after prenatal intoxication of Foxn1+ TECs (38, 39). These cells may originate from the incomplete separation of the parathyroid and the thymus (40), as has been proposed previously for the cervical thymus in a reciprocal manner (21).

Whereas the embryonic and young thymi contain only a small fraction of Foxn1− cells, the proportion of Foxn1− cells increases between 2 and 10 wk, reaching a plateau of roughly 50% Foxn1− TECs. Based on their Foxn1 expression history, these Foxn1− TECs are post-Foxn1 cells rather than representing a Foxn1− lineage (37, 38). Although the Foxn1+ TECs contain all known mature major TEC subsets as well as few double-negative TECs, the latter fraction dominates the Foxn1− TEC population. Currently, we lack information on the relative location of Foxn1+ and Foxn1− TEC subsets in the thymus. Thus, neither the location nor the function of Foxn1− TECs is known. We speculate that these cells are exhausted or are end-stage cells that nevertheless may have a role as scaffold to support the thymus structure. It is plausible that the detrimental effects of either ablation of Foxn1 expression (by conditional deletion of Foxn1-floxed alleles) or Foxn1-expressing cells (by progressive intoxication of Foxn1-expressing cells) acted on the thymus only via the Foxn1+ TEC population. These data imply that Foxn1− TECs, despite their abundance in the adult thymus, are unable to sustain a functional thymus (24, 38, 41–43).

Reduced Foxn1 expression has been linked to thymus aging and involution. Our detailed kinetics of frequencies and levels of Foxn1 expression reveal a surprisingly early decline in Foxn1 expression at a juvenile age. The ubiquitous requirement for Foxn1 expression in all TECs appears to end at around 2 wk of age, and hence in a phase when the thymus is about to enter steady-state function. In this scenario, lower levels of Foxn1 may set the stage for the later occurring involution; however, this famous process is most likely multifactorial.

The thymus is known for its regenerative capacity in response to thymic insults and thymic stimulation. The observed disappearance of the CD80+Ly51− mTECs in response to DEX (28) and the emergence of a prominent CD80+Ly51− mTEC population after castration (23) are well in line with previous publications. Regeneration after DEX treatment reestablished the frequency of Foxn1-expressing cells without changes in Foxn1 levels, and the regrowth of the thymus in response to DEX was characterized by the appearance of a CD80−Ly51+Foxn1+ population.

The long-lasting rhFGF7-induced regeneration (34) relied on the expansion of a different TEC subset than that involved in the transient regeneration observed after castration (32, 44), although each response was accompanied by an upregulation of Foxn1 protein levels. A predominant cTEC-like Ly51+Foxn1+ population could be seen in response to DEX administration and rhFGF7 treatment, as well as in Rag2-deficient mice lacking mature thymocytes. That the transient castration response involved a CD80+Ly51−Foxn1+ mTEC-like population clearly pointed toward two different regenerative mechanisms. Concomitantly, Fgf7 deficiency does not result in an inability to respond to castration (45). It seems that each thymic injury, and thereby the regenerative outcome, has its own characteristic responding TEC population.

Whether the predominant Ly51+Foxn1+ cTEC-like cells of Rag2-deficient mice, as well as the phenotypically identical populations expanding in response to DEX and rhFGF7, represent bona fide cTECs remains to be shown. It seems more likely that some TEC progenitor population may also display the Ly51+Foxn1+ phenotype, as does the vast majority of TECs in unborn mice, and at least some of these cells have been demonstrated to exhibit progenitor function (46). Moreover, cTEC-specific β5t promoter–driven Cre labels mTECs, once more suggesting an overlap between the phenotypes of TEC progenitors and cTECs (47). The appearance of cTEC markers on mTECs might represent a coincidental, transient upregulation of these markers during mTEC development, or it requires a revised order of TEC maturation, as suggested (48).

In summary, our in-depth analysis of Foxn1 expression in TEC subsets in the developing, aging, and regenerating thymus may provide a useful basis to address molecular functions of this transcription factor in the near future.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Carmen Blum and Susan Schlenner for help with the generation of mouse mutants; the Imaging Core Facility for help with immunofluorescence; the German Cancer Research Center animal facility for expert mouse husbandry; and Andrea Weber, Tabea Arnsperger, and Sven Schäfer for technical assistance.

Footnotes

  • This work was supported in part by Deutsche Forschungsgemeinschaft–Sonderforschungsbereich Grant 873-Project B11, as well as by a grant from the Helmholtz Alliance Preclinical Comprehensive Cancer Center (to H.-R.R.).

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    cTEC
    cortical TEC
    DEX
    dexamethasone
    E
    embryonic day
    FDG
    fluorescein-di-β-d-galactopyranoside
    Foxn1
    forkhead box N1
    HA
    hemagglutinin
    mTEC
    medullary TEC
    rhFGF7
    recombinant human fibroblast growth factor 7
    TEC
    thymic epithelial cell.

  • Received September 9, 2015.
  • Accepted October 13, 2015.

References

  1. ↵
    1. Nehls M.,
    2. D. Pfeifer,
    3. M. Schorpp,
    4. H. Hedrich,
    5. T. Boehm
    . 1994. New member of the winged-helix protein family disrupted in mouse and rat nude mutations. Nature 372: 103–107.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Frank J.,
    2. C. Pignata,
    3. A. A. Panteleyev,
    4. D. M. Prowse,
    5. H. Baden,
    6. L. Weiner,
    7. L. Gaetaniello,
    8. W. Ahmad,
    9. N. Pozzi,
    10. P. B. Cserhalmi-Friedman,
    11. et al
    . 1999. Exposing the human nude phenotype. Nature 398: 473–474.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Schorpp M.,
    2. M. Hofmann,
    3. T. N. Dear,
    4. T. Boehm
    . 1997. Characterization of mouse and human nude genes. Immunogenetics 46: 509–515.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Schlake T.,
    2. M. Schorpp,
    3. M. Nehls,
    4. 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: 3842–3847.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Schüddekopf K.,
    2. M. Schorpp,
    3. T. Boehm
    . 1996. The whn transcription factor encoded by the nude locus contains an evolutionarily conserved and functionally indispensable activation domain. Proc. Natl. Acad. Sci. USA 93: 9661–9664.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Condie N. R. M. B. G.,
    2. B. G. Condie
    . 2010. Transcriptional regulation of thymus organogenesis and thymic epithelial cell differentiation. Prog. Mol. Biol. Transl. Sci. 92: 103–120.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Itoi M.,
    2. H. Kawamoto,
    3. Y. Katsura,
    4. T. Amagai
    . 2001. Two distinct steps of immigration of hematopoietic progenitors into the early thymus anlage. Int. Immunol. 13: 1203–1211.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Nehls M.,
    2. B. Kyewski,
    3. M. Messerle,
    4. R. Waldschütz,
    5. K. Schüddekopf,
    6. A. J. Smith,
    7. T. Boehm
    . 1996. Two genetically separable steps in the differentiation of thymic epithelium. Science 272: 886–889.
    OpenUrlAbstract
  9. ↵
    1. Bleul C. C.,
    2. T. Corbeaux,
    3. A. Reuter,
    4. P. Fisch,
    5. J. S. Mönting,
    6. T. Boehm
    . 2006. Formation of a functional thymus initiated by a postnatal epithelial progenitor cell. Nature 441: 992–996.
    OpenUrlCrossRefPubMed
    1. Bleul C. C.,
    2. T. Boehm
    . 2005. BMP signaling is required for normal thymus development. J. Immunol. 175: 5213–5221.
    OpenUrlAbstract/FREE Full Text
    1. Blackburn C. C.,
    2. C. L. Augustine,
    3. R. Li,
    4. R. P. Harvey,
    5. M. A. Malin,
    6. R. L. Boyd,
    7. J. F. Miller,
    8. G. Morahan
    . 1996. The nu gene acts cell-autonomously and is required for differentiation of thymic epithelial progenitors. Proc. Natl. Acad. Sci. USA 93: 5742–5746.
    OpenUrlAbstract/FREE Full Text
    1. Müller S. M.,
    2. G. Terszowski,
    3. C. Blum,
    4. C. Haller,
    5. V. Anquez,
    6. S. Kuschert,
    7. P. Carmeliet,
    8. H. G. Augustin,
    9. H.-R. Rodewald
    . 2005. Gene targeting of VEGF-A in thymus epithelium disrupts thymus blood vessel architecture. Proc. Natl. Acad. Sci. USA 102: 10587–10592.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Jenkinson W. E.,
    2. A. Bacon,
    3. A. J. White,
    4. G. Anderson,
    5. E. J. Jenkinson
    . 2008. An epithelial progenitor pool regulates thymus growth. J. Immunol. 181: 6101–6108.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Ortman C. L.,
    2. K. A. Dittmar,
    3. P. L. Witte,
    4. P. T. Le
    . 2002. Molecular characterization of the mouse involuted thymus: aberrations in expression of transcription regulators in thymocyte and epithelial compartments. Int. Immunol. 14: 813–822.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Bredenkamp N.,
    2. S. Ulyanchenko,
    3. K. E. O’Neill,
    4. N. R. Manley,
    5. H. J. Vaidya,
    6. C. C. Blackburn
    . 2014. An organized and functional thymus generated from FOXN1-reprogrammed fibroblasts. Nat. Cell Biol. 16: 902–908.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Soza-Ried C.,
    2. C. C. Bleul,
    3. M. Schorpp,
    4. T. Boehm
    . 2008. Maintenance of thymic epithelial phenotype requires extrinsic signals in mouse and zebrafish. J. Immunol. 181: 5272–5277.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Srinivas S.,
    2. T. Watanabe,
    3. C. S. Lin,
    4. C. M. William,
    5. Y. Tanabe,
    6. T. M. Jessell,
    7. F. Costantini
    . 2001. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev. Biol. 1: 4.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Shinkai Y.,
    2. G. Rathbun,
    3. K. P. Lam,
    4. E. M. Oltz,
    5. V. Stewart,
    6. M. Mendelsohn,
    7. J. Charron,
    8. M. Datta,
    9. F. Young,
    10. A. M. Stall,
    11. et al
    . 1992. RAG-2-deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68: 855–867.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Rode I.,
    2. T. Boehm
    . 2012. Regenerative capacity of adult cortical thymic epithelial cells. Proc. Natl. Acad. Sci. USA 109: 3463–3468.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Gereke M.,
    2. A. Autengruber,
    3. L. Gröbe,
    4. A. Jeron,
    5. D. Bruder,
    6. S. Stegemann-Koniszewski
    . 2012. Flow cytometric isolation of primary murine type II alveolar epithelial cells for functional and molecular studies. J. Vis. Exp. 2012: e4322.
    OpenUrl
  21. ↵
    1. Terszowski G.,
    2. S. M. Müller,
    3. C. C. Bleul,
    4. C. Blum,
    5. R. Schirmbeck,
    6. J. Reimann,
    7. L. D. Pasquier,
    8. T. Amagai,
    9. T. Boehm,
    10. H. R. Rodewald
    . 2006. Evidence for a functional second thymus in mice. Science 312: 284–287.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Chen L.,
    2. S. Xiao,
    3. N. R. Manley
    . 2009. Foxn1 is required to maintain the postnatal thymic microenvironment in a dosage-sensitive manner. Blood 113: 567–574.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Gray D. H. D.,
    2. N. Seach,
    3. T. Ueno,
    4. M. K. Milton,
    5. A. Liston,
    6. A. M. Lew,
    7. C. C. Goodnow,
    8. R. L. Boyd
    . 2006. Developmental kinetics, turnover, and stimulatory capacity of thymic epithelial cells. Blood 108: 3777–3785.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Sun L.,
    2. J. Guo,
    3. R. Brown,
    4. T. Amagai,
    5. Y. Zhao,
    6. D.-M. Su
    . 2010. Declining expression of a single epithelial cell-autonomous gene accelerates age-related thymic involution. Aging Cell 9: 347–357.
    OpenUrlCrossRefPubMed
    1. Zook E. C.,
    2. P. A. Krishack,
    3. S. Zhang,
    4. N. J. Zeleznik-Le,
    5. A. B. Firulli,
    6. P. L. Witte,
    7. P. T. Le
    . 2011. Overexpression of Foxn1 attenuates age-associated thymic involution and prevents the expansion of peripheral CD4 memory T cells. Blood 118: 5723–5731.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Bredenkamp N.,
    2. C. S. Nowell,
    3. C. C. Blackburn
    . 2014. Regeneration of the aged thymus by a single transcription factor. Development 141: 1627–1637.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Kong F.-K.,
    2. C.-L. H. Chen,
    3. M. D. Cooper
    . 2002. Reversible disruption of thymic function by steroid treatment. J. Immunol. 168: 6500–6505.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Fletcher A. L.,
    2. T. E. Lowen,
    3. S. Sakkal,
    4. J. J. Reiseger,
    5. M. V. Hammett,
    6. N. Seach,
    7. H. S. Scott,
    8. R. L. Boyd,
    9. A. P. Chidgey
    . 2009. Ablation and regeneration of tolerance-inducing medullary thymic epithelial cells after cyclosporine, cyclophosphamide, and dexamethasone treatment. J. Immunol. 183: 823–831.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Sutherland J. S.,
    2. G. L. Goldberg,
    3. M. V. Hammett,
    4. A. P. Uldrich,
    5. S. P. Berzins,
    6. T. S. Heng,
    7. B. R. Blazar,
    8. J. L. Millar,
    9. M. A. Malin,
    10. A. P. Chidgey,
    11. R. L. Boyd
    . 2005. Activation of thymic regeneration in mice and humans following androgen blockade. J. Immunol. 175: 2741–2753.
    OpenUrlAbstract/FREE Full Text
    1. Roden A. C.,
    2. M. T. Moser,
    3. S. D. Tri,
    4. M. Mercader,
    5. S. M. Kuntz,
    6. H. Dong,
    7. A. A. Hurwitz,
    8. D. J. McKean,
    9. E. Celis,
    10. B. C. Leibovich,
    11. et al
    . 2004. Augmentation of T cell levels and responses induced by androgen deprivation. J. Immunol. 173: 6098–6108.
    OpenUrlAbstract/FREE Full Text
    1. Olsen N. J.,
    2. S. M. Viselli,
    3. K. Shults,
    4. G. Stelzer,
    5. W. J. Kovacs
    . 1994. Induction of immature thymocyte proliferation after castration of normal male mice. Endocrinology 134: 107–113.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Goldberg G. L.,
    2. J. A. Dudakov,
    3. J. J. Reiseger,
    4. N. Seach,
    5. T. Ueno,
    6. K. Vlahos,
    7. M. V. Hammett,
    8. L. F. Young,
    9. T. S. P. Heng,
    10. R. L. Boyd,
    11. A. P. Chidgey
    . 2010. Sex steroid ablation enhances immune reconstitution following cytotoxic antineoplastic therapy in young mice. J. Immunol. 184: 6014–6024.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Kelly R. M.,
    2. S. L. Highfill,
    3. A. Panoskaltsis-Mortari,
    4. P. A. Taylor,
    5. R. L. Boyd,
    6. G. A. Holländer,
    7. B. R. Blazar
    . 2008. Keratinocyte growth factor and androgen blockade work in concert to protect against conditioning regimen-induced thymic epithelial damage and enhance T-cell reconstitution after murine bone marrow transplantation. Blood 111: 5734–5744.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Rossi S. W.,
    2. L. T. Jeker,
    3. T. Ueno,
    4. S. Kuse,
    5. M. P. Keller,
    6. S. Zuklys,
    7. A. V. Gudkov,
    8. Y. Takahama,
    9. W. Krenger,
    10. B. R. Blazar,
    11. G. A. Holländer
    . 2007. Keratinocyte growth factor (KGF) enhances postnatal T-cell development via enhancements in proliferation and function of thymic epithelial cells. Blood 109: 3803–3811.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Erickson M.,
    2. S. Morkowski,
    3. S. Lehar,
    4. G. Gillard,
    5. C. Beers,
    6. J. Dooley,
    7. J. S. Rubin,
    8. A. Rudensky,
    9. A. G. Farr
    . 2002. Regulation of thymic epithelium by keratinocyte growth factor. Blood 100: 3269–3278.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Itoi M.,
    2. N. Tsukamoto,
    3. T. Amagai
    . 2007. Expression of Dll4 and CCL25 in Foxn1-negative epithelial cells in the post-natal thymus. Int. Immunol. 19: 127–132.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Liston A.,
    2. A. G. Farr,
    3. Z. Chen,
    4. C. Benoist,
    5. D. Mathis,
    6. N. R. Manley,
    7. A. Y. Rudensky
    . 2007. Lack of Foxp3 function and expression in the thymic epithelium. J. Exp. Med. 204: 475–480.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Corbeaux T.,
    2. I. Hess,
    3. J. B. Swann,
    4. B. Kanzler,
    5. A. Haas-Assenbaum,
    6. T. Boehm
    . 2010. Thymopoiesis in mice depends on a Foxn1-positive thymic epithelial cell lineage. Proc. Natl. Acad. Sci. USA 107: 16613–16618.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Ucar A.,
    2. O. Ucar,
    3. P. Klug,
    4. S. Matt,
    5. F. Brunk,
    6. T. G. Hofmann,
    7. B. Kyewski
    . 2014. Adult thymus contains FoxN1− epithelial stem cells that are bipotent for medullary and cortical thymic epithelial lineages. Immunity 41: 257–269.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Dooley J.,
    2. M. Erickson,
    3. A. G. Farr
    . 2009. Lessons from thymic epithelial heterogeneity: FoxN1 and tissue-restricted gene expression by extrathymic, endodermally derived epithelium. J. Immunol. 183: 5042–5049.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Guo J.,
    2. Y. Feng,
    3. P. Barnes,
    4. F.-F. Huang,
    5. S. Idell,
    6. D.-M. Su,
    7. H. Shams
    . 2012. Deletion of FoxN1 in the thymic medullary epithelium reduces peripheral T cell responses to infection and mimics changes of aging. PLoS One 7: e34681.
    OpenUrlCrossRefPubMed
    1. Cheng L.,
    2. J. Guo,
    3. L. Sun,
    4. J. Fu,
    5. P. F. Barnes,
    6. D. Metzger,
    7. P. Chambon,
    8. R. G. Oshima,
    9. T. Amagai,
    10. D. M. Su
    . 2010. Postnatal tissue-specific disruption of transcription factor FoxN1 triggers acute thymic atrophy. J. Biol. Chem. 285: 5836–5847.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Guo J.,
    2. M. Rahman,
    3. L. Cheng,
    4. S. Zhang,
    5. A. Tvinnereim,
    6. D.-M. Su
    . 2011. Morphogenesis and maintenance of the 3D thymic medulla and prevention of nude skin phenotype require FoxN1 in pre- and post-natal K14 epithelium. J. Mol. Med. 89: 263–277.
    OpenUrlCrossRefPubMed
  44. ↵
    1. Griffith A. V.,
    2. M. Fallahi,
    3. T. Venables,
    4. H. T. Petrie
    . 2012. Persistent degenerative changes in thymic organ function revealed by an inducible model of organ regrowth. Aging Cell 11: 169–177.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Goldberg G. L.,
    2. O. Alpdogan,
    3. S. J. Muriglan,
    4. M. V. Hammett,
    5. M. K. Milton,
    6. J. M. Eng,
    7. V. M. Hubbard,
    8. A. Kochman,
    9. L. M. Willis,
    10. A. S. Greenberg,
    11. et al
    . 2007. Enhanced immune reconstitution by sex steroid ablation following allogeneic hemopoietic stem cell transplantation. J. Immunol. 178: 7473–7484.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Baik S.,
    2. E. J. Jenkinson,
    3. P. J. L. Lane,
    4. G. Anderson,
    5. W. E. Jenkinson
    . 2013. Generation of both cortical and Aire+ medullary thymic epithelial compartments from CD205+ progenitors. Eur. J. Immunol. 43: 589–594.
    OpenUrlCrossRefPubMed
  47. ↵
    1. Ohigashi I.,
    2. S. Zuklys,
    3. M. Sakata,
    4. C. E. Mayer,
    5. S. Zhanybekova,
    6. S. Murata,
    7. K. Tanaka,
    8. G. A. Holländer,
    9. Y. Takahama
    . 2013. Aire-expressing thymic medullary epithelial cells originate from β5t-expressing progenitor cells. Proc. Natl. Acad. Sci. USA 110: 9885–9890.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Alves N. L.,
    2. Y. Takahama,
    3. I. Ohigashi,
    4. A. R. Ribeiro,
    5. S. Baik,
    6. G. Anderson,
    7. W. E. Jenkinson
    . 2014. Serial progression of cortical and medullary thymic epithelial microenvironments. Eur. J. Immunol. 44: 16–22.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section