1. Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP, Collège de France, BP 163, 67404 Illkirch Cedex, France
2. Institute of Molecular & Cellular Biosciences, University of Tokyo, Yayoi, Bunkyo-ku , Tokyo 113, Japan
3. CREST, Japan Science and Technology , Kawaguchi, Saitama 332, Japan
4. These authors contributed equally to this work

Correspondence to: Pierre Chambon¹ Correspondence and requests for materials should be addressed to P.C. (e-mail: Email: chambon@igbmc.u-strasbg.fr).

To ablate RXR in epidermis, we engineered mice carrying LoxP-site-containing (floxed) RXR^L2 alleles (Fig. 1a) and used the K5–Cre–ER^T transgenic line in which tamoxifen (Tam) efficiently induces Cre-mediated recombination in basal layer keratinocytes¹⁹. K5–Cre–ER^T(tg/tg) /RXR^L2/L2 mice mated with RXR ^+/- (Fig. 1a; ref. 16) or RXR^L2/+ mice yielded 'pro-mutant' mice hemizygous (tg/0) for K5–Cre–ER^T and carrying either one RXR^L2 and one RXR null (-) allele (K5–Cre–ER ^T(tg/0)/RXR^L2/–genotype) or two L2 alleles (K5–Cre–ER^T(tg/0)/RXR ^L2/L2 genotype). At 14 weeks old, the pro-mutant mice were treated with Tam (5 days, 1 mg per day), and then retreated 2, 4 and 6 weeks later. Six weeks after the first Tam treatment (AFT), 80% of RXR ^L2 alleles were converted into RXR^L- alleles in the epidermis of mice carrying one or two floxed alleles ( Fig. 1b). By 12 weeks AFT, almost all RXR^L2 alleles had been converted (Fig. 1b). As expected¹⁹, no RXR disruption occurred in vehicle (oil)-treated mice (data not shown) and Cre-mediated excision of RXR exon 4 was restricted to epidermis and some epithelia in which the K5 promoter is also active (for example, tongue, salivary gland, oesophagus; Fig. 1c).

Figure 1: Tamoxifen-induced RXR null mutation in adult mouse epidermis mediated by Cre–ER^T.

a, Diagram of the wild-type RXR genomic locus (+), the floxed RXR L2 allele, the RXR L- allele obtained after Cre-mediated excision of exon 4 (encoding the DNA-binding domain), and the RXR null allele (-)¹⁶. Black boxes indicate exons (E2–E4). Restriction enzyme sites and probe X4 location are indicated. BamHI fragments are in kilobases (kb). B, BamHI; C, ClaI; E, EcoRI; H, HindIII, S, SpeI; X, XbaI. Arrowheads in L2 and L- alleles indicate LoxP sites. b, Tamoxifen (Tam)-induced generation of K5–Cre–ER ^T-mediated RXR^L- alleles illustrated by Southern blot analysis of epidermal DNA isolated 6 (lanes 1–3) and 12 (lanes 4–6) weeks after the first Tam (1 mg) injection series (AFT). All mice were K5–Cre–ER^T(tg/0) and the RXR genotypes are indicated. BamHI-digested DNA fragments corresponding to RXR (+), L2, L- and (-) alleles are displayed. c, Tissue-specificity of Cre-ER^T-mediated RXR disruption. WT (+), L2 and L- alleles were identified by PCR on DNA extracted from various organs of K5–Cre–ER ^T(tg/0)/RXR^L2/+ mice, 12 weeks AFT. d, Tamoxifen-induced generation of RXR null alleles in adult mouse epidermis using K14–Cre–ER^T2(tg/0)or K14–Cre–ER ^T2(0/0) mice (designated (tg/0) and (0/0), respectively). PCR analysis of genomic DNA from epidermis (E) and dermis (D), isolated two weeks after injection of either Tam (0.1 mg) (+) or vehicle (-). Mouse genotypes are indicated and PCR fragments corresponding to RXR (+), L2 and L- alleles are displayed.

High resolution image and legend (64K)

Interestingly, hair loss (alopecia) was observed 6–7 weeks AFT in the ventral region of pro-mutant mice, but not in oil-treated pro-mutant mice or in Tam-treated K5–Cre–ER^T(tg/0)/RXR ^L2/+ 'control' littermates (data not shown). At 12–16 weeks AFT, large regions of ventral skin and smaller regions of dorsal skin were hairless (Fig. 2a, b; and data not shown). Cysts became visible under the skin surface and these enlarged and spread all over the body with time (Fig. 2c; and data not shown). With increasing age (> 20 weeks AFT), minor focal lesions appeared on hairless dorsal skin, on chins and behind ears ( Fig. 2d; and data not shown). These were not caused by fights and were formed of crusts on top of hyperproliferative epidermis and inflammatory dermis (see below).

Figure 2: Abnormalities generated by Tam-induced disruption of RXR in skin of adult mouse mediated by K5–Cre–ER^T and K14–Cre–ER^T2.

a, b, A female K5–Cre–ER^T(tg/0)/RXR ^L2/- 'mutant' (mt) mouse, and a female K5–Cre–ER ^T(tg/0)/RXR^L2/+ 'control' (ct) mouse, 16 weeks AFT (1 mg Tam per injection). a, Ventral view. b, Dorsal view. c, Higher magnification of the ventral region of the K5–Cre–ER ^T(tg/0)/RXR^L2/- mouse. Arrow, one of the cysts. d, Dorsal view of a female K5–Cre–ER^T(tg/0)/RXR ^L2/- mouse, 28 weeks AFT. Arrow, minor skin lesion. e–h, Histological analysis. 2-m sections of ventral skin 16 weeks AFT, taken from 'control' (e, g) and 'mutant' ( f, h) mice. hf, hair follicles; u, utriculi; dc, dermal cysts; arrowheads (h), Langerhans cells whose number is increased several-fold in mutant epidermis. i, j, Keratin 6 (K6) immunohistochemistry on 'control' (i) and 'mutant' (j) skin sections (16 weeks AFT). Red, staining of the K6 antibody; cyan, DAPI staining. Arrow (e–j), the dermal–epidermal junction. k, l, Skin appearance of a female K14–Cre–ER^T2(tg/0)/RXR ^L2/L2 'mutant' mouse. k, High magnification of the ventral region, 16 weeks after Tam treatment (0.1 mg per injection). White arrow, a cyst; black arrow, a melanosome-containing utriculus. l, Dorsal view of the same mutant. Arrow, a skin lesion. Scale bars: e, f , 60 m; g, h, 12 m; i, j, 25 m.

High resolution image and legend (157K)

At 16 weeks AFT, histological analysis of ventral and dorsal hairless regions showed hair follicle degeneration, resulting in utriculi and dermal cysts^{20, 21} (Fig. 2 compare e with f; and data not shown). Interfollicular epidermis was hyperplastic with increased incorporation of 5-bromodeoxyuridine (BrdU) and expression of the Ki67 proliferation marker (data not shown). Dermal cellularity was increased and capillaries were dilated (Fig. 2 compare e, g with f, h; and data not shown) underneath the thickened epidermis, reflecting an inflammatory reaction (data not shown). Keratin 6 (K6), which is usually expressed only in hair follicle outer root sheath (ORS), was also expressed in hyperproliferative interfollicular epidermis (Fig. 2i, j), indicating abnormal keratinocyte terminal differentiation²². All abnormalities were less severe, and/or appeared later in males than in females.

To improve the efficiency of Tam-induced Cre-mediated recombination, we engineered K14–Cre–ER^T2 transgenic lines. The K14 promoter is selective for the basal layer of stratified squamous epithelia²³, and Cre–ER^T2 can be induced by a milder Tam treatment (0.1 mg for 5 days)¹⁹. We treated 8–10-week-old K14–Cre–ER^T2(tg/0)/RXR^L2/L2 mice with Tam, together with 'control' littermates of genotype K14–Cre–ER^T2(tg/0)/RXR^L2/+, K14–Cre–ER^T2(0/0)/RXR^L2/+ and K14–Cre–ER^T2(0/0)/RXR^L2/L2. In two weeks, RXR^L2 alleles were fully converted into RXR^L- alleles in epidermis (Fig. 1d, lanes 1 and 7), but not in dermis (Fig. 1d, lanes 2 and 8) of K14–Cre–ER^T2-expressing transgenic mice, demonstrating the higher efficiency of Cre–ER^T2 for mediating the selective somatic mutation of floxed RXR in epidermis after Tam treatment. No L2 to L- allele conversion occurred in 'controls' lacking the K14–Cre–ER^T2 transgene ( Fig. 1d, lanes 5, 6, 11 and 12) or without Tam treatment ( Fig. 1d, lanes 3, 4, 9 and 10). Moreover, 8 weeks after Tam treatment, only the RXR^L- allele was detected in K14–Cre–ER ^T2(tg/0)/RXR^L2/L2 mouse epidermis (data not shown), indicating that RXR was disrupted in most, if not all epidermal stem cells. RXR disruption also occurred in other epithelia in which the K14 promoter is active²⁴ (for example, tongue, oesophagus and stomach, data not shown). Starting 6 weeks after Tam treatment, K14–Cre–ER ^T2(tg/0)/RXR^L2/L2 mice exhibited abnormalities similar to those observed in Tam-treated K5–Cre–ER^T(tg/0) /RXR^L2/L2 mice, that is, marked hair loss with visible cysts and focal skin lesions appearing at later stages ( Fig. 2k and l; and data not shown). The underlying epidermal and dermal histological abnormalities were also similar to those described above for Tam-treated K5–Cre–ER^T(tg/0) /RXR^L2/L2 mice (data not shown).

Thus, disruption of floxed RXR in adult epidermis is achieved faster and with lower Tam doses in K14–Cre–ER^T2 than in K5–Cre–ER^T mice, but the resulting skin abnormalities are similar and, in both cases, more severe in females than in males. Interestingly, these abnormalities are also similar to those exhibited by K14–Cre ^(tg/0)/RXR^L2/L2 or K14–Cre^(tg/0) /RXR^L2/- mice in which floxed RXR alleles are selectively disrupted in the epidermis during fetal development, leading to RXR ablation in epidermal keratinocytes and hair follicle ORS (M.L. et al., manuscript in preparation). Indeed, from three weeks of age, these 'constitutive' epidermis-selective RXR mutants develop progressive alopecia with typical features of degenerated hair follicles, together with utriculi and dermal cysts, which can all be attributed to defects in hair cycles^{20, 21}. Furthermore, these mutants also exhibit interfollicular keratinocyte hyperproliferation, as well as abnormal terminal differentiation (with K6 expression) and increased dermal cellularity associated with a skin inflammatory reaction (M.L. et al., manuscript in preparation).

RXR is expressed at a much lower level than is RXR in mouse skin, and RXR expression is undetectable by polymerase chain reaction after reverse transcription of RNA (RT–PCR; data not shown). Interestingly, we found that the adult skin level of RXR messenger RNA is several fold higher in males than in females. However, the skin of adult male and female RXR^-/- mutants appears normal²⁵ (data not shown) and RXR mRNA levels did not change upon RXR ablation (data not shown). Compound mutants were generated to explore a possible functional redundancy between RXRs.

As expected, oil-treated K5–Cre–ER^T(tg/0)/RXR ^L2/-/RXR^-/- and K5–Cre–ER ^T(tg/0)/RXR^L2/L2/RXR^-/-mice did not exhibit skin abnormalities, but 4 weeks after treatment with Tam they began to lose their hair. Large skin regions were hairless 16–18 weeks AFT, and epidermal flaking on the hairless trunk, chin and ears was much more conspicuous than on single RXR mutants (compare Fig. 3a; with Fig. 2d; and data not shown). Crusted skin lesions and ulcers lacking epidermis, not seen in RXR single mutants, were also frequently observed on these RXR/ double mutants at 14–16 weeks AFT, particularly on the hairless trunk skin, behind the ears and around the mouth (Fig. 3a; and data not shown). But wound repair was not overtly impaired in these mutants when skin biopsies were taken from lesion-free hairless regions (data not shown). Histology of hairless skin showed disappearance of hair follicles and presence of utriculi and dermal cysts (Fig. 3b). The epidermis was highly hyperplastic and hyperkeratinized (compare Fig. 3b, c with Fig. 2e, g; Fig. 2f, h), abnormal K6 expression was observed throughout epidermis (Fig. 3d), and an inflammatory reaction with increased dermal cellularity was also observed (Fig. 3b; and data not shown).

Figure 3: Comparison of skin abnormalities exhibited by a Tam-treated K5–Cre–ER ^T(tg/10) RXR^L2/L2/RXR^-/- mouse and a VDR-null mouse.

The figure corresponds to a K5–Cre–ER^T(tg/0)/RXR ^L2/L2/RXR^-/- mouse, 18 weeks AFT (1 mg Tam per injection) (a–d) and to a 14-week-old VDR^-/- mouse (e–h). b, c, f, g, Histological analysis of 2-m dorsal skin sections. d, h, Keratin K6 immunohistochemistry on skin sections. u, utriculi; dc, dermal cysts. Arrows: a, skin lesions; b–h, the dermal–epidermal junction. Scale bar: b, f, 60 m; c, g, 12 m; d, h, 25 m.

High resolution image and legend (116K)

The K5–Cre–ER^T(tg/0)/RXR^L2/L2 /RXR^-/-/RXR^-/- triple mutants treated with Tam did not reveal any further role of RXR in adult skin (data not shown). Thus, RXR can partially compensate for a loss of RXR function. Also, in accordance with the larger amount of RXR in adult male skin, the functional redundancy was more pronounced in males than in females as RXR/RXR double mutant males and females were similarly affected (data not shown) unlike the single mutants (see above).

Collectively, our results demonstrate the effectiveness of Cre–ER ^T recombinases in generating cell-specific temporally controlled targeted somatic mutations in adult mouse tissues. We also show that RXR, whose knockout is lethal in utero^{16, 17}, is important postnatally in processes controlling hair cycling and the proliferation and differentiation of epidermal keratinocytes, even though expression of a dominant-negative RXR mutant in epidermal suprabasal layers has no effect on skin development and maintenance¹⁵.

Our study also reveals a functional redundancy between RXR and RXR, although RXR function is clearly dominant. The molecular events underlying the generation of alopecia and keratinocyte abnormalities in epidermis of mice lacking these receptors are unknown. However, a number of nuclear receptors (for example, RARs, TRs, VDR, PPARs, LXRs) form heterodimers with RXR, and numerous in vitro studies using cultured cells and a few in vivo targeted-mutagenesis studies^{10, 11, 12, 26} show that these heterodimers act as signal transducers in different signalling pathways. Interestingly, VDR is also expressed in hair follicle ORS²⁷ and VDR knock-out mice develop progressive secondary alopecia, indicating that VDR is important in hair cycling^{2, 3, 4}. Alopecia developed by 14-week-old VDR ^-/- mice appears similar externally to that developed by K5–Cre–ER ^T(tg/0)/RXR^L2/L2/RXR^-/- mice 18 weeks after Tam treatment, although the skin of VDR^-/- mice was free of the lesions seen on RXR-ablated epidermis (compare Fig. 3a with e). At the histological level, similar utriculi and dermal cysts were observed (compare Fig. 3b with f), but no keratinocyte hyperproliferation was observed in the epidermis of VDR^-/- mice and keratinocyte differentiation was normal (as revealed by K6 expression (compare Fig. 3c and d with g and h; and data not shown). Thus, alopecia generated by selective RXR ablation in adult mouse keratinocytes may reflect a major role of RXR/VDR heterodimers in hair cycling. Further keratinocyte-specific targeted somatic mutagenesis is necessary to investigate whether other signalling pathways involving nuclear receptors that heterodimerize with RXRs (notably RARs¹, LXRs⁷ and PPARs^{8, 28}) are implicated in the generation of the other skin abnormalities resulting from keratinocyte-selective RXR ablation.

References

1. Fisher, G. J. & Voorhees, J. J. Molecular mechanisms of retinoid actions in skin. FASEB J. 10, 1002– 1013 (1996). | PubMed | ISI | ChemPort |
2. Yoshizawa, T. et al. Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nature Genet. 16, 391–396 (1997). | Article |
3. Li, Y. C. et al. Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc. Natl Acad. Sci. USA 94, 9831–9835 (1997).
4. Li, Y. C. et al. Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139, 4391–4396 (1998). | Article | PubMed | ISI | ChemPort |
5. Kömüves, L. G. et al. Ligands and activators of nuclear hormone receptors regulate epidermal differentiation during fetal rat skin development. J. Invest. Dermatol. 111, 429–433 (1998).
6. Hanley, K. et al. Activators of the nuclear hormone receptors PPAR and FXR accelerate the development of the fetal epidermal permeability barrier. J. Clin. Invest. 100, 705–712 (1997). | PubMed | ISI | ChemPort |
7. Hanley, K. et al. Oxysterols induce differentiation in human keratinocytes and increase Ap-1-dependent involucrin transcription. J. Invest. Dermatol. 114, 545–553 ( 2000). | Article | PubMed | ISI | ChemPort |
8. Hanley, K. et al. Farnesol stimulates differentiation in epidermal keratinocytes via PPAR. J. Biol. Chem. 275, 11484 –11491 (2000). | Article | PubMed | ISI | ChemPort |
9. Mangelsdorf, D. J. et al. The nuclear receptor superfamily: the second decade. Cell 83, 835–839 ( 1995). | Article | PubMed | ISI | ChemPort |
10. Chambon, P. A decade of molecular biology of retinoic acid receptors. FASEB J. 10, 940–954 ( 1996).
11. Kastner, P. , Mark, M. & Chambon, P. Nonsteroid nuclear receptors: what are genetic studies telling us about their role in real life? Cell 83, 859–869 (1995). | Article | PubMed | ISI | ChemPort |
12. Mascrez, B. et al. The RXR ligand-dependent activation function 2 (AF-2) is important for mouse development. Development 125 , 4691–4707 (1998).  | PubMed | ISI | ChemPort |
13. Saitou, M. et al. Inhibition of skin development by targeted expression of a dominant-negative retinoic acid receptor. Nature 374 , 159–162 (1995). | Article | PubMed | ISI | ChemPort |
14. Imakado, S. et al. Targeting expression of a dominant-negative retinoic acid receptor mutant in the epidermis of transgenic mice results in loss of barrier function. Genes Dev. 9, 317– 329 (1995). | PubMed | ISI | ChemPort |
15. Feng, X. et al. Suprabasal expression of a dominant-negative RXR mutant in transgenic mouse epidermis impairs regulation of gene transcription and basal keratinocyte proliferation by RAR-selective retinoids. Genes Dev. 11, 59–71 ( 1997).
16. Kastner, P. et al. Genetic analysis of RXR developmental function: convergence of RXR and RAR signalling pathways in heart and eye morphogenesis. Cell 78, 987–1003 ( 1994). | Article | PubMed | ISI | ChemPort |
17. Sucov, H. M. et al. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007–1018 (1994). | PubMed | ISI | ChemPort |
18. Metzger, D. & Chambon, P. Site- and time-specific gene targeting in the mouse. Methods (in the press).
19. Indra, A. K. et al. Temporally-controlled site-specific mutagenesis in the basal layer of the epidermis: comparison of the recombinase activity of the tamoxifen-inducible Cre-ER^T and Cre-ER^T2 recombinases. Nucleic Acids Res. 27, 4324–4327 (1999). | Article | PubMed | ISI | ChemPort |
20. Sundberg, J. P. & King, L. E. Mouse models for the study of human hair loss. Dermatol. Clin. 14, 619–632 (1996). | Article | PubMed | ISI | ChemPort |
21. Panteleyev, A. A. , Paus, R. , Ahmad, W. , Sundberg, J. P. & Christiano, A. M. Molecular and functional aspects of the hairless (hr) gene in laboratory rodents and humans. Exp. Dermatol. 7, 249–267 (1998).  | PubMed | ISI | ChemPort |
22. Porter, R. M. , Reichelt, J. , Lunny, D. P. , Magin, T. M. & Lane, B. The relationship between hyperproliferation and epidermal thickening in a mouse model for BCIE. J. Invest. Dermatol. 110, 951–957 ( 1998).
23. Vassar, R. , Rosenberg, M. , Ross, S. , Tyner, A. & Fuchs, E. Tissue-specific and differentiation-specific expression of a human K14 keratin gene in transgenic mice. Proc. Natl Acad. Sci. USA 86, 1563–1567 ( 1989).
24. Wang, X. , Zinkel, S. , Polonsky, K. & Fuchs, E. Transgenic studies with a keratin promoter-driven growth hormone transgene: prospects for gene therapy. Proc. Natl Acad. Sci. USA 94, 219 –226 (1997). | Article | PubMed | ChemPort |
25. Kastner, P. et al. Abnormal spermatogenesis in RXR mutant mice. Genes Dev. 10, 80–92 ( 1996). | PubMed | ISI | ChemPort |
26. Wendling, O. , Chambon, P. & Mark, M. Retinoid X receptors are essential for early mouse development and placentogenesis. Proc. Natl Acad. Sci. USA 96, 547–551 (1999). | Article | PubMed | ChemPort |
27. Reichrath, J. et al. Hair follicle expression of 1,25-dihydroxyvitamin D3 receptors during the murine hair cycle. Br. J. Dermatol. 131, 477–482 (1994). | PubMed | ISI | ChemPort |
28. Braissant, O. , Foufelle, F. , Scotto, C. , Dauça, M. & Wahli, W. Differential expression of peroxisome proliferator-activated receptors (PPARs): tissue distribution of PPAR-,-, and- in the adult rat. Endocrinology 137, 354 –366 (1996). | Article | PubMed | ISI | ChemPort |
29. Metzger, D. et al. Conditional site-specific recombination in mammalian cells using a ligand-dependent chimeric Cre recombinase. Proc. Natl Acad. Sci. USA 92, 6991–6995 ( 1995).
30. Brocard, J. et al. Spatio-temporally controlled site-specific somatic mutagenesis in the mouse. Proc. Natl Acad. Sci. USA 94, 14559–14563 (1997). | Article | PubMed | ChemPort |

Acknowledgements

We thank S. Werner for the human K14 promoter; H. Chiba and P. Kastner for RXR^L2/+, RXR^+/- and RXR ^+/-mice; J. M. Bornert, S. Bronner, N. Chartoire, M. Duval, C. Gérard, R. Lorentz and J. L. Vonesch for technical help; M. LeMeur, R. Matyas and the animal facility staff for animal care; M. Mark for histological analysis; the secretariat for typing the manuscript and the illustration staff for preparing the figures; and all the members of the laboratory for helpful discussions. This work was supported by funds from the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale, the Collège de France, the Hôpital Universitaire de Strasbourg, the Association pour la Recherche sur le Cancer, the Fondation pour la Recherche Médicale, the Human Frontier Science Program, the Ministère de l'Éducation Nationale de la Recherche et de la Technologie and the European Economic Community. M.L. was supported by fellowships from the Association pour la Recherche sur le Cancer and the Fondation pour la Recherche Médicale, A.K.I. by a fellowship from the Université Louis Pasteur (Strasbourg), and J.B. and X.W. by fellowships from the Ministère de l'Education Nationale, de la Recherche et de la Technologie and from the Fondation pour la Recherche Médicale.