Foxn1 regulates lineage progression in cortical and medullary thymic epithelial cells but is dispensable for medullary sublineage divergence - PubMed (original) (raw)
Foxn1 regulates lineage progression in cortical and medullary thymic epithelial cells but is dispensable for medullary sublineage divergence
Craig S Nowell et al. PLoS Genet. 2011 Nov.
Abstract
The forkhead transcription factor Foxn1 is indispensable for thymus development, but the mechanisms by which it mediates thymic epithelial cell (TEC) development are poorly understood. To examine the cellular and molecular basis of Foxn1 function, we generated a novel and revertible hypomorphic allele of Foxn1. By varying levels of its expression, we identified a number of features of the Foxn1 system. Here we show that Foxn1 is a powerful regulator of TEC differentiation that is required at multiple intermediate stages of TE lineage development in the fetal and adult thymus. We find no evidence for a role for Foxn1 in TEC fate-choice. Rather, we show it is required for stable entry into both the cortical and medullary TEC differentiation programmes and subsequently is needed at increasing dosage for progression through successive differentiation states in both cortical and medullary TEC. We further demonstrate regulation by Foxn1 of a suite of genes with diverse roles in thymus development and/or function, suggesting it acts as a master regulator of the core thymic epithelial programme rather than regulating a particular aspect of TEC biology. Overall, our data establish a genetics-based model of cellular hierarchies in the TE lineage and provide mechanistic insight relating titration of a single transcription factor to control of lineage progression. Our novel revertible hypomorph system may be similarly applied to analyzing other regulators of development.
Conflict of interest statement
The authors have declared that no competing interests exist.
Figures
Figure 1. Foxn1R is a hypomorphic allele.
(A) A LoxP flanked cassette containing the 5′ engrailed 2 splice acceptor site (SA), the SV40TAg coding region, an internal ribosome entry site coupled to an enhanced green fluorescent protein/neomycin resistance fusion protein (IRES-EGFPneo), and the CMAZ transcriptional pause was inserted into intron 1b of the Foxn1 locus of mouse ES cells (Figure S1). E, exon. (B) Images show representative pairs of thymic lobes from six week old WT, Foxn1R/+ and Foxn1R/R as indicated. Scale bars represent 1cm. Thymocyte numbers - Foxn1R/R, 5.0×107±2.3×106; Foxn1R/+, 2.6x109±9.5×107; WT littermates 1.9×108 ±8.1×106. WT versus Foxn1R/+ p = 3.7×10−6; WT versus Foxn1R/R p = 2.0×10−5, n = 3 for each genotype. TEC numbers: WT, 1.3×106±8.4×105; Foxn1R/+, 1.3×107±1.8×106; Foxn1R/R, 3.6×105±2.1×105. WT versus Foxn1R/+ p = 0.00054; WT versus Foxn1R/R p = 0.09374. (C) Immunoblots performed on nuclear extracts isolated from whole E12.5 thymus tissue were probed sequentially with anti-Foxn1 (left panel) and anti-alpha-tubulin (loading control; right panel). Arrows indicate bands corresponding to Foxn1 and alpha-tubulin. Genotypes are indicated above each image. WT, wild-type; R/+, Foxn1R/+; R/R, Foxn1R/R. (D) RT-PCR analysis of Foxn1R/R TEC reveals splicing around the inserted cassette. cDNA from whole adult thymus from wild type, Foxn1R/+ or Foxn1R/R thymi was subjected to RT-PCR using primers specific for Foxn1 exon1a (forward) and exon 3 (reverse); Foxn1 exon1a (forward) and SV40Tag (reverse), or ß actin as shown. Lanes: m, markers; 1, wild type cDNA; 2, wild type no RT control; 3, Foxn1R/+ cDNA; 4, Foxn1R/+ no RT control; 5, Foxn1R/R cDNA; 6, Foxn1R/R no RT control. Bands in lanes 4 and 6 of the ß actin panel are non-specific. Band specificity was confirmed by sequencing. (E) Reversion of the Foxn1R allele in Foxn1R/+ZP3Cre/+ mice. Foxn1R/+ males were mated with ZP3Cre/Cre females to generate Foxn1 R/+: ZP3 Cre/+ mice. Foxn1R/+:ZP3Cre/+ females were then crossed with C57BL/6 males and the resulting offspring analyzed. Two mice (lanes 1 and 2) were identified by PCR as carrying both the Foxn1R allele (using primers specific for the targeted allele) and the ZP3Cre transgene and were subsequently analyzed by Southern blotting of Nde1 digested genomic DNA using a probe flanking the 3′ end of the homology arms. In both cases only a 7.5kb band was present indicating complete excision of the targeted insertion. Lane 3 shows the 8.5kb and 7.5kb bands corresponding to the targeted and WT alleles respectively in a Cre negative Foxn1R/+ littermate. (F-I) Morphological and phenotypic analysis of 6 week-old Foxn1R/+ thymi. (F) Representative H&E staining showing overtly normal architecture in Foxn1R/+ thymi. Magnification, x10. (G, I) Plots show flow cytometric analysis of 6 week old Foxn1R/+ or WT thymi with markers shown. Plots in I) show data after gating against CD45+ and on EpCam+ cells. (H) Immunohistochemical analysis showing normal staining patterns with markers specific for medullary (K14, Aire) and cortical (CDR1) TEC and for the TEC maturation marker MHC Class II (MHCII). Epithelial cells are identified by co-staining with anti-pancytokeratin (PanK) in lower and upper panels. Right panels show co-stains of left panels, right AIRE panel shows detail from left panel. Scale bars 150 µm except right Aire panel which is 11.9 µm. Data shown in F-I are representative of at least three independent experiments. Thymocyte (G) and TEC (H,I) subset distribution is indistinguishable between Foxn1R/+ and WT thymi.
Figure 2. Functional deficit in Foxn1R/R TEC.
Morphological and phenotypic analysis of 6 week old Foxn1R/R thymi. (A) H&E staining showing frequent large medullary cysts in Foxn1R/R thymi. (B) Immunostaining with anti-cytokeratin 5 (K5) and anti-cytokeratin 8 (K8) showing overtly normal cortical and medullary compartments in Foxn1R/R thymi; note the normal morphology of cortical and medullary TEC. (C-E) Flow cytometric analysis of 6 week old wild-type or Foxn1R/R thymocytes with (C) anti-CD4-PE, anti-CD8-FITC, (D) anti-CD44-APC and anti-CD25-PE after gating on Lin- cells (Lin = anti-CD3, anti-CD4, anti-CD8, anti-CD11b, anti-CD11c, Ter119, NK1.1, Gr-1), or (E) B220 after gating on DN1 cells in (D). (E) red lines show Foxn1R/R; blue lines, WT. (F) Flow cytometric analysis of E13.5 or E15.5 fetal thymic primordia with anti-CD45-APC Cy7; anti-CD25-PE, and anti-CD44-APC. PN, postnatal; WT, wild type. (E) shows pooled data from three 6 week old mice, n = 1. All other data are representative of at least three separate analyses.
Figure 3. Reduced Foxn1 expression results in impaired development of both medullary and cortical TEC in postnatal Foxn1R/R thymi.
(A-G, Ai-Hi) Immunohistochemical (A-E,G,H) and/or flow cytometric (Ai-Hi, F) analysis of postnatal Foxn1R/R thymi showing anti-CD205, Ly51, CDR1, anti-Aire, anti-K14, anti-MHC Class II (MHCII), anti-Dll4, anti-ß5t, anti-PanK and anti-CD40 staining. (A–E,G,H) Right panels show co-stains of left panels, right Aire panel shows detail from left panel. Scale bars 150 µm except for anti-CD205 and Ly51, which are 75 µm and Aire right panel which is 11.9 µm. (Ai-Hi, F) Plots show data after gating against CD45+ and Ter119+ and for EpCam+ cells. Lower panel in Ei shows data after gating on CD45-EpCamlo cells (i.e. cTEC). Red lines, Foxn1R/R; blue lines, WT (Foxn1R/+ rather than WT is shown for Ly51, and is indistinguishable from WT); grey lines, isotype control. All analyses are on 6 week old thymi except E and Ei which are on postnatal day 10-11 thymi. Data shown are representative of at least three independent experiments, except (F), n = 1. (I) 4 week old Foxn1R/R;ROSA26CreErt2 mice were injected intraperitoneally with 1.5 mg 4-hydroxy tamoxifen (4OHT) and sacrificed 2 days later; bottom panels show un-injected littermate controls. Images show immunostaining with CDR1, anti-pancytokeratin, Ly51, anti-MHC Class II and anti-Aire. Data are representative of at least 3 mice for each condition. (J) 6 week old Foxn1R/R;ROSA26CreErt2 (R/R) thymi were injected with 1.5mg 4OHT or carrier only and sacrificed 2 days later. Controls were provided by uninjected Foxn1R/+;ROSA26CreErt2 (R/+) littermates. Frozen sections were stained with anti-Aire and anti-PanK, as in (I), and the numbers of Aire+ TEC per mm2 of medullary area were determined by counting for three medullary areas from two sections from each of three mice for each condition. Untreated R/R versus R/+, p = 0.022; 4OHT-treated R/R versus R/+, p = 0.28; 4OHT-treated R/R versus untreated R/R, p = 0.025. (K) Mature and immature postnatal cTEC (EpCAM+Ly51+MHC Class IIhi and EpCAM+Ly51+MHC Class IIlo cells respectively) and mTEC (EpCAM+Ly51-MHC Class IIhi and EpCAM+Ly51-MHC Class IIlo cells respectively) were purified from 4-8 week old WT C57BL/6 mice by flow cytometry and the relative Foxn1 mRNA expression levels in each cell population determined by QRT-PCR. Foxn1 is expressed at higher levels in WT cTEC than WT mTEC, and at higher levels in mature than in immature TEC in both compartments. Foxn1 expression levels are shown relative to alpha-tubulin and are normalized to MHC Class IIhi cTEC. Data shown are the mean of two technical replicates and are representative of three independent experiments. Error bars show SD.
Figure 4. Quantitiative analysis of the Foxn1R allele.
(A) Flow cytometric analysis of thymocyte development in neonatal Foxn1R/- mice. Representative plots are shown for anti-CD45, anti-CD4 versus anti-CD8 and anti-CD44 versus anti-CD25, after gating on CD45+ cells. Foxn1R/- thymi fail to support the maturation of immature thymocyte progenitors. Note that the tissue dissected in Foxn1R/- mice contains non-thymic tissue as well as the thymic rudiment due to the difficulty of identifying the thymic rudiment in postnatal Foxn1R/- mice. (B) Immunohistochemical analysis of postnatal Foxn1R/- and Foxn1 -/- thymi showing Plet-1, CDR1, K14, K5, K8 and UEA-1 and PanK staining. Images shown are representative of at least three individual mice for each genotype. Scale bars 47.62 µm for Plet-1/PanK and 150 µm for K5/K8, K14/CDR1 and UEA1/PanK. (C) 2-4 month old Foxn1R/-;ROSA26CreErt2 mice were injected intraperitoneally with 1.5 mg 4-hydroxy tamoxifen (4OHT) or carrier, and sacrificed 7 weeks later. Thymi were analyzed with the markers shown. MHCII, MHC Class II. (D) Relative Foxn1 expression levels were determined by QRT-PCR analysis of E12.5 Plet-1+ TEC using the Sybr-Green method. Foxn1 expression was normalized to the geometric mean of three housekeeping genes. Numbers show mean expression level relative to WT. (A-C) Data shown are representative of at least three independent experiments. (D) WT, Foxn1R/+ and Foxn1R/R, data are representative of at least three independent experiments; Foxn1R/- and Foxn1-/-, n = 1.
Figure 5. Divergence of the mTEC lineage is Foxn1-independent.
Images show representative sections from (A) E13.5 thymic primordia after staining for K5 and K8, Cldn4 and Plet-1. (B,C) E15.5 thymic primordia after staining for CD205 and panK (B) or MTS10, UEA-1 and panK (C). Scale bars 47.62 µm. Genotypes of embryos are as designated. Note that the medullary lineage markers Cldn4, K5 and UEA1 are expressed in Foxn1-/- thymi. Bottom right image in (C) shows negative control (NC) for UEA1 stain (Streptavin-Alexa647). Data shown are representative of at least three independent experiments. (D-F) Plots show staining of (D) wild type thymic primordia from E12.5, E13.5 and E15.5 mice, (E) E13.5 and (F) E15.5 mice of the genotypes shown after staining with anti-Plet-1 (MTS20) and anti-MHC Class II (MHCII). All plots show data obtained after gating on EpCAM+ cells. The data shown for Foxn1-/-, Foxn1+/-, and E13.5 Foxn1R/- mice are representative analyses of individual thymic primordia, all other plots show analyses on pooled samples from several embryos. n≥3 for all genotypes. Relative proportions of EpCam+Plet-1+ cells in Foxn1-/- and Foxn1R/- mice:
E13.5:
Foxn1R/-, 70.8±16.0%; Foxn1-/-, 74.3±7.6%.
E15.5:
Foxn1R/-, 59±7.9%; Foxn1-/-, 73.8±13.3%; Foxn1R/- versus Foxn1-/- p = 0.74 (E13.5), p = 0.06 (E15.5).
Figure 6. Foxn1 is required for exit from the earliest TEC progenitor states and regulates the size of the differentiated TEC compartment during ontogeny.
(A) Absolute numbers of Plet-1+ EpCam+MHC Class II- (MHCII)(open bars) and Plet-1-EpCam+ (black bars) cells per thymus at E13.5 and E15.5 for each genotype. Data shown are representative of at least three independent experiments. Error bars show SD. (B) Cell cycle analysis of Foxn1 allelic variants. Images show flow cytometric analysis of E15.5 WT, Foxn1R/+, Foxn1R/R, Foxn1R/- and Foxn1-/- thymi using anti-BrdU-APC and 7-AAD 1 hour after injection with BrdU. Plots are representative plots for each genotype after gating on EpCam+ cells; pairs of plots show data collected on the same day. (C,D,F) QRT-PCR analysis of E13.5 TEC (EpCam+ cells) or E13.5 thymic mesenchymal cells (PDGFRalpha+ cells) for the genes shown. Data are normalized to alpha-tubulin and shown relative to E13.5 WT EpCam+ cells (C,D) or E13.5 WT PDGFRalpha+ cells (F) and show the means from two (Pax1) or three (Dll4, CCL25, Fgf2) biological replicates for each gene. Data for all genes are representative of at least three biological replicates performed using either the Lightcycler or Fluidigm platforms. Error bars show SD. (C’,D’,E) QRT-PCR analysis of E12.5 Plet1+ TEC for the genes shown. Data are normalized to EVA and shown relative to E12.5 WT Plet1+ cells. Data show mean of three technical replicates from one experiment on RNA from pooled thymi for each genotype. CCL25 and Dll4, data are representative of two independent analyses using different analysis platforms. (G) QRT-PCR analysis of mouse embryonic stem cells after transient transfection with either Foxn1 or GFP. Data show Foxn1, Dll4 and CCL25 expression levels relative to ß actin and show means of two (Foxn1) or three (Dll4, CCL25) biological repeats. Error bars show SD. Dll4, p = 0.009 (Foxn1 over-expression versus GFP over-expression); CCL25 p = 0.038 (Foxn1 over-expression versus GFP over-expression). (H) Cartoon shows regulatory network indicated by data shown.
Figure 7. Foxn1 regulation of TE lineage development.
Model of cellular hierarchies in TEC lineage progression based on the genetic data presented. The dotted lines represent lineage transitions in which the precursor-progeny relationship is currently uncertain.
References
- Gordon J, Wilson VA, Blair NF, Sheridan J, Farley A, et al. Functional evidence for a single endodermal origin for the thymic epithelium. Nat Immunol. 2004;5:546–553. - PubMed
- Bennett AR, Farley A, Blair NF, Gordon J, Sharp L, et al. Identification and characterization of thymic epithelial progenitor cells. Immunity. 2002;16:803–814. - PubMed
- Moore-Scott BA, Opoka R, Lin SC, Kordich JJ, Wells JM. Identification of molecular markers that are expressed in discrete anterior-posterior domains of the endoderm from the gastrula stage to mid-gestation. Dev Dyn. 2007;236:1997–2003. - PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Molecular Biology Databases