Specific expression of lacZ and cre recombinase in fetal thymic epithelial cells by multiplex gene targeting at the Foxn1 locus

Abstract

Background: Thymic epithelial cells (TECs) promote thymocyte maturation and are required for the early stages of thymocyte development and for positive selection. However, investigation of the mechanisms by which TECs perform these functions has been inhibited by the lack of genetic tools. Since the Foxn1 gene is expressed in all presumptive TECs from the early stages of thymus organogenesis and broadly in the adult thymus, it is an ideal locus for driving gene expression in differentiating and mature TECs.

Results: We generated two knock-in alleles of Foxn1 by inserting IRES-Cre or IRES-lacZ cassettes into the 3' UTR of the Foxn1 locus. We simultaneously electroporated the two targeting vectors to generate the two independent alleles in the same experiment, demonstrating the feasibility of multiplex gene targeting at this locus. Our analysis shows that the knockin alleles drive expression of Cre or lacZ in all TECs in the fetal thymus. Furthermore, the knockin alleles express Cre or lacZ in a Foxn1-like pattern without disrupting Foxn1 function as determined by phenotype analysis of Foxn1 knockin/Foxn1 null compound heterozygotes.

Conclusion: These data show that multiplex gene targeting into the 3' UTR of the Foxn1 locus is an efficient method to express any gene of interest in TECs from the earliest stage of thymus organogenesis. The resulting alleles will make possible new molecular and genetic studies of TEC differentiation and function. We also discuss evidence indicating that gene targeting into the 3' UTR is a technique that may be broadly applicable for the generation of genetically neutral driver strains.

Figure 1

Multiplex gene targeting at the Foxn1 locus. A. Map of the wild-type (wt) Foxn1 locus and the targeting vectors used for gene targeting. Locations of the 5' and 3' flanking probes (fp) are shown. Exons 7–9 are shown as boxes, coding regions are black and UTR sequences are grey. IRES-Cre or IRES-lacZ cassettes including the floxed or frt neo positive selection cassette were inserted into a unique XhoI site (X) in the 3'UTR in exon 9. The targeting vector homology was flanked by two TKgenes for negative selection. B. Southern blot analysis of ES cell clones. Genomic DNA was digested with EcoRV, and probes with the 5' flanking probe. The wild-type 14.8 kb band is present in all samples. Correctly targeted cell lines with bands corresponding to the two different targeted alleles are present in lanes 3–6. Lane two shows an abberant clone. C. PCR analysis of the lacZ alleles used for genotypeing. The larger, 600 bp band is the wild-type allele, the 300 bp band is the lacZ (z) allele. Marker (M) shown is the Promega 1 kb ladder. V, EcoRV; E, EcoRI; B, BamHI; K, KpnI; S, SspI; X, XhoI.

Figure 2

Foxn1^ex9lacZ and Foxn1^ex9cre expression from E10.5 to newborn. Cre expression was assayed using the R26R reporter allele. Panels C, F, I, L show Cre negative control embryos with no expression. Black arrows in (A-I) point to the third pharyngeal pouch. White arrows in (J-L) indicate the Foxn1-negative parathyroid region of the primordium. For sagittal sections, ventral is left and anterior is up. For transverse sections, dorsal is up. A, B, Lateral views of E10.5 Foxn1^+/ex9lacZ (A) and Foxn1^+/ex9cre;R26R^+/- (B) whole embryos, and D, E, coronal sections through E10.5 Foxn1^+/ex9lacZ (D) and Foxn1^+/ex9cre;R26R^+/- (E) embryos showing no expression in the third pharyngeal pouch. G, H, Lateral views of E11.5 Foxn1^+/ex9lacZ (G) and Foxn1^+/ex9cre;R26R^+/- (H) embryos showing expression only in the developing thymus primordium. J, K, Sagittal sections through E11.5 Foxn1^+/ex9lacZ embryo (J) and E11.5 Foxn1^+/ex9cre;R26R^+/- embryo (K) showing expression restricted to the ventral thymus domain of the 3^rd pouch-derived organ primordium. M, Sagittal section through E12.5 Foxn1^+/IRESlacZ thymus primordium. N, Transverse section through E12.5 Foxn1^+/ex9cre;R26R^+/- thymus primordium. O, Transverse section through E12.5 thymus showing in situ hybridization for Foxn1 mRNA. P, Section through E16.5 Foxn1^+/ex9lacZ thymus. Q, R, Sections through E16.5 (Q) and newborn (R) Foxn1^+/ex9cre;R26R^+/- thymi showing more intense staining in the prospective medullary regions. c = cortex, m = medulla.

Figure 3

Analysis of reporter expression in neonatal TECs after Cre recombination. Five-day old mice carrying both the Foxn1^ex9cre and R26YFP reporter genes were analyzed for the expression of YFP in TEC populations by FACS. Left panel shows the percentage of YFP positive cells in the CD45^-MHC Class II⁺ population. Right panel shows the percentage of YFP positive cells in the CD45^-UEA1⁺ population.

Figure 4

Effect of neo^r deletion on Cre expression. Ventral views of two E12.5 Foxn1^+/ex9cre;R26R^+/- embryos, x-gal stained to show Cre activity. A, Prior to deletion of the neo^r cassette, extensive non-specific staining was seen throughout the embryo. B, Thymus-specific expression after removal of the neo^r cassette. The heads have been removed from both embryos to allow visualization of the thymic primordia. Arrows indicate one of the thymic lobes. h = heart.

Figure 5

Phenotypic analysis of mice carrying Foxn1^ex9lacZ and Foxn1^ex9cre alleles. A, Semi-quantitative RT-PCR of Foxn1 RNA in thymi from heterozygous and homozygous Foxn1^ex9cre or Foxn1^ex9lacZ newborn mice. Both the 586 bp Foxn1 full-length product and the 124 bp exon 2–4 spliced product are first seen at 24 cycles in all genotypes. B-E, Keratin 5 and keratin 8 expression in the E16.5 thymus. Transverse sections through E16.5 thymic lobes from (B) wild-type, (C) Foxn1^+/ex9cre, (D) Foxn1^+/ex9lacZ and (E) Foxn1^{ex9lacZ/ex9lacZ} embryos. Keratin 8 is shown in green and marks cortical epithelial cells, while keratin 5 marks medullary epithelial cells and is shown in red. F, FACS analysis of thymocytes from Foxn1^ex9lacZ and Foxn1^ex9cre mice. Thymocytes from newborn mice were stained with anti-CD4-APC, anti-CD8-FITC, anti-CD44-PE, and anti-CD25-Biotin followed by Streptavidin-PerCP. The upper panels show expression of CD4 and CD8 on total thymocytes. The lower panels show expression of CD44 and CD25 on the gated CD4^- CD8^- subpopulation.