Decline of FOXN1 gene expression in human thymus correlates with age: possible epigenetic regulation

Abstract

Background: Thymic involution is thought to be an important factor of age related immunodeficiency. Understanding the molecular mechanisms of human thymic senescence may lead to the discovery of novel therapeutic approaches aimed at the reestablishment of central and peripheral T cell repertoire.

Results: As an initial approach, here we report that the decline of human thymic FOXN1 transcription correlates with age, while other genes, DLL1, DLL4 and WNT4, essential for thymopoiesis, are constitutively transcribed. Using a human thymic epithelial cell line (hTEC), we show that FOXN1 expression is refractory to signals that induce FOXN1 transcription in primary 3D culture conditions and by stimulation of the canonical WNT signaling pathway. Blockage of FOXN1 induceability in the hTEC line may be mediated by an epigenetic mechanism, the CpG methylation of the FOXN1 gene.

Conclusion: We showed a suppression of FOXN1 transcription both in cultured human thymic epithelial cells and in the aging thymus. We hypothesize that the underlying mechanism may be associated with changes of the DNA methylation state of the FOXN1 gene.

Fig. 1

DLL1, DLL4, FOXN1 and WNT-4 gene expression in human thymic samples. Gene expression was assessed by real-time quantitative PCR. DLL1 (a), DLL4 (b), FOXN1 (c) and WNT-4 (c). The graphs show mean ± SEM of normalized expression values from each group. Difference between the samples was analyzed through Kruskal-Wallis non-parametric test followed by Dunn’s multiple test. Data was considered statistical significant when p < 0.05

Fig. 2

Histology, thymic epithelial organization and FOXN1 protein expression. Photomicrographs show morphological aspects of thymus samples from (a) five days (“Postnatal” group) and (b) 57 years-old donors (“Adult” group). Paraffin sections were stained with haematoxilin-eosin stain and analyzed by light microscopic examination. (4x, objective magnification). c = cortical region; m = medullary region; a = Adipose tissue; arrow = connective tissue septum. Panels (c) and (d) show cytokeratin staining of “Postnatal” (5 days-old) and “Adult” thymic samples (57 years-old), respectively. Frozen thymus sections were reacted with specific pan anti-cytokeratin polyclonal rabbit antibody followed by Alexa-488 conjugated anti-rabbit Ig antibody (green) before visualization in the fluorescent microscope. (20x, objective magnification). Photomicrographs (e) and (f) show FOXN1 protein expression in thymus cryosections from “Postnatal” (5 days-old) and “Adult” donors (57 years-old), respectively. In order to demonstrate thymic epithelium associated FOXN1 expression, the sections were first reacted with anti-human FOXN1 antibody (red) and subsequently double stained with pan anti-cytokeratin antibody (green). (20x, objective magnification)

Fig. 3

hTEC line, a model of aging thymic epithelium: three-dimensional cell culture in cellulose scaffolds. Conventional hTEC monolayer culture (b) and hTEC 3D culture (c) maintained for four days in cellulose macroporous scaffolds were immunocytochemical staining for cytokeratin with DAPI nuclear counterstaining. (20x, objective magnification). qPCR detection of FOXN1 (d), DLL4 (e) and DLL1 (f) transcripts, in conventional monolayer and 3D cultures. Thymic stroma from 5 days-old sample served as control (“Stroma”). Undetectable level of expression is labelled with “*”

Fig. 4

hTEC line, a model of aging thymic epithelium. WNT signaling pathway induction. Conventional hTEC monolayer cultures stained with anti-β-catenin monoclonal antibody (green) (a–c) and anti-human FOXN1 (red) (d–f) with DAPI nuclear counterstain: non-treated control (a, d), LiCl-treated for 6 h (b, e) and 18 h (c, f). (40x objective magnification). FOXN1 staining on a thymic human sample was used as a positive control (Fig. 4f, insert). In (g) is shown the RT-PCR detection of FOXN1 transcript after treatment with 20 mM of LiCl for 6 and 18 h. Non-treated cultures and TEC treated with 20 mM of NaCl were used as negative controls. Legend – F: FOXN1 amplification product; F (−): reverse transcriptase reaction control without enzyme from FOXN1 amplification product; H: HPRT-1 amplification product; H (−): reverse transcriptase reaction control without enzyme from HPRT-1 amplification product

Fig. 5

FOXN1 genomic context, potential methylation substrate C20 and CpG methylation status on hTEC. Panel (a) shows the genomic localization of C20 region on FOXN1 gene and (b) (upper panel) presents the relative GC content, CpG island prediction, repetitive region (simple), histone modifications and transcriptional factor binding sites on C20 region. Histone modifications and transcriptional factor binding sites were evaluated by ENCODE consortium laboratories using chromatin immunoprecipitation followed by DNA sequencing [39]. The CpG island was predicted using MethPrimer web-based software [68] and the repetitive region was obtained in the RepeatMasker database [74]. Lower panel in (b) shows the C20 nucleotide sequence. Above is the normal genome sequence and below is the bisulfite converted sequence. In red is presented the CpG residues. The arrows (> and <) indicate the forward and reverse specific bisulfate primers used for sequencing. The CpG island is underlined in black and the yellow label marks the repetitive region. CpG methylation status on C20 region was obtained for FOXN1expressing human skin (c) and FOXN1 non-expressing human leukocytes (d) from the ENCODE database on DNA Methylation by Reduced Representation Bisulfite Sequencing data provided by ENCODE/HudsonAlpha laboratory. FOXN1 gene graphic representation and DNA elements regulation were adapted from UCSC genome browser from the human genome assembly hg19 using “ENCODE regulation” track features [39, 67]. The graph in (e) represents the percentage of CpG methylation, for each residue, and in hTEC cell line clone sequences, while the diagrams in (f) show the 13 CpG-s presented as lollipop diagrams, where the methylated residue is shown in black circles, while the non-methylated residue is in open circles. Percent values were obtained after multiple sequence alignment using ClustalW algorithm followed for methylated CpG quantification using BiQ analyzer software