WO2000021486A2

WO2000021486A2 - Prevention of uv-induced functional vitamin a deficiency through use of topically applied retinoid

Info

Publication number: WO2000021486A2
Application number: PCT/US1999/023591
Authority: WO
Inventors: John J. Voorhees; Gary J. Fisher
Original assignee: The Regents Of The University Of Michigan
Priority date: 1998-10-14
Filing date: 1999-10-12
Publication date: 2000-04-20
Also published as: IL142349A0; WO2000021486A9; CA2344696A1; WO2000021486A3; JP2003524604A; BR9914531A; AU6513399A

Abstract

The vitamin A metabolite all-trans retinoic acid (RA) is critical for normal skin function. Ultraviolet irradiation (UV) markedly reduces the mRNA and protein of the two major nuclear retinoid receptors, RAR-η and RXR-α in human skin in vivo. One half the dose of UV that causes skin reddening was sufficient to reduce retinoid receptor mRNA levels. Maximal reduction of RAR-η and RXR-α proteins occurred between 8 and 16 hours after UV irradiation. With multiple exposures to UV, RXR-α remained decreased, but RAR-η recovered to normal levels. Application of RA 24 hours before UV exposure partially prevented loss of nuclear retinoid receptors. UV irradiation completely prevented RA induction of two retinoid receptor-regulated genes, cellular retinoic acid binding protein-II(CRABP-II) and RA 4-hydroxylase. In contrast, UV irradiation did not affect 1,25-dihydroxyvitamin D3 induction of the vitamin D receptor-regulated gene, vitamin 0 24-hydroxylase, indicating that UV selectively interferes with the retinoid-signaling pathway. These data demonstrate for the first time that UV specially reduces retinoid receptor levels and dramatically suppresses retinoid-responsive gene expression in human skin in vivo. In effect, UV causes a functional vitamin A deficiency that could have deleterious effects on skin function, contributing to skin photoaging and carcinogenesis, which can be ameliorated by application of a retinoid prior to exposure.

Description

PREVENTION OF UV-INDUCED FUNCTIONAL VITAMIN A DEFICIENCY THROUGH USE OF TOPICALLY APPLIED RETINOID

Background

All-trans retinoic acid (RA) is a critical regulator of cell growth and differentiation 5 in developing and adult mammalian skin. Two families of nuclear retinoid receptors, retinoic acid receptors (denoted RARs) and retinoid X receptors (denoted RXRs) mediate the biological effects of RA. Each family of nuclear retinoid receptors is comprised of three subtypes, α, β, and γ, which are ligand-activated transcription factors. RAR-γ and RXR-α function as heterodimers to mediate the pleiotropic biological activities of

10 retinoids by activating transcription of retinoic acid-responsive genes. Nuclear retinoid receptor gene knock-outs and over-expression of dominant negative retinoid receptor transgenes have shown that retinoid receptors are required for normal skin development and function. Simultaneous null mutations of two RARs produce phenotypic changes that closely mimic those observed in vitamin A deficient animals. Thus, loss of retinoid

15 receptors results in a state of functional vitamin A deficiency.

Several lines of evidence indicate that alterations in nuclear retinoid receptors are associated with cellular transformation and tumor formation. RAR transcripts (RAR-α and RAR-γ) are decreased in benign skin tumors in mice and are essentially absent in undifferentiated squamous cell carcinomas in man. Several studies have demonstrated that

20 retinoids are effective for treating or preventing certain kinds of cancers, in both animal and human models. In breast cancer cells, retinoid inhibition of cell growth correlates with RAR-α expression. RA has proven to be effective in inhibiting tumor promoter- induced transformation in cultured cells. RAR-β mRNA selectively lost in premalignant oral lesions was restored by treatment with 13 -cis retinoic acid.

2.5 Skin cancer is the most prevalent form of malignancy in man. The major cause of skin cancer is ultraviolet (UN) irradiation from the sun. UN irradiation is a complete carcinogen, capable of causing cell transformation and promoting tumor formation. UN irradiation causes DΝA damage that can result in mutations as a consequence of imperfect DΝA repair. Such mutations can ultimately lead to cell transformation. UN irradiation

30 also induces several immediate early genes for transcription factors such as AP-1 and ΝF-

KB, the activation of which lead(s) to complex secondary changes in expression of multiple target genes. These genes are involved in arrest of the Gl phase of the cell cycle, DΝA repair mechanisms, and p53-dependent programmed cell death, among other functions. Many UN-inducible genes such as c-jun, c-fos, c-myc, plasminogen activator, coUagenases,

35 and ornithine decarboxylase are involved in tumor progression in a variety of cancers.

Recently, we reported that UN irradiation up-regulates AP-1 activity and AP-1 -driven matrix-degrading metalloproteinase genes in human skin in vivo. Increased AP-1 activity and matrix metalloproteinase gene expression might be inhibited by pretreatment of skin with RA. In the course of these studies, we investigated the effect of UV radiation on retinoid receptor expression and retinoid signaling. We found that UV irradiation causes rapid and major reduction of retinoid receptors and an almost total loss of RA-responsive gene expression in human skin in vivo. In essence, UV causes a functional vitamin A deficiency in human skin. This functional vitamin A deficiency begins at UV doses that do not cause even minimal sunburn, and can be substantially prevented by treatment of skin with RA prior to UV exposure.

SUMMARY AND OBJECTS OF THE INVENTION In light of the foregoing, it would be useful to prevent UV-induced functional vitamin A deficiency. As described in the experiments below, UV irradiation of human skin causes reductions in RAR-γ protein and mRNA and RXR-α protein and mRNA, and so it would also be useful to inhibit the loss/reduction in these proteins and the mRNA coding therefor. As such, we have found that topical pretreatment with a retinoid prior to UV exposure inhibits the reductions in RAR-γ and RXR-α proteins caused by UV radiation. Retinoid receptors, especially RXR, are necessary for human skin to respond to vitamin D (e.g., 1,25-dihydroxyvitamin D₃). Accordingly, the present method includes inhibiting the loss of at least one type of retinoid receptor after exposure of human skin to UV radiation by topically pretreating the skin later to be exposed to the UV radiation with a retinoid, preferably in amounts of 0.01% to 5%, more preferably 0.05% to 1%. By preventing the loss of retinoid receptors, this invention also includes preventing functional vitamin A deficiency in human skin because of UV exposure, which method comprises pretreating the skin to be exposed to UV radiation with a retinoid, preferably in amounts of 0.01% to 5%, more preferably 0.05% to 1%. Pretreatment with the retinoid (or a compatible mixture of retinoids) preferably occurs at least 8, more preferably 12, and most preferably about 24 hours prior to exposure to UV radiation.

FIGURE LEGENDS

Fig. 1. Ultraviolet irradiation reduces nuclear retinoid receptor proteins and mRNAs in adult human skin in vivo. (A) Time course of reductions of RAR-γ protein

(open bars) and RXR-α protein (hatched bars) in human skin in vivo following exposure to 2MED UV. Inset shows representative Western blots for RAR-γ and RXR-α Data are presented as fold change relative to non-irradiated control skin, and are expressed as means ± SEM, N=4-6. *p<0.05 versus control. (B) Time course of reductions of RAR-γ mRNA (open bars) and RXR-α mRNA (hatched bars) in human skin in vivo following exposure to 2MED UV. Inset shows representative Northern blots for RAR-γ RXR-α and 36B4 (internal control). Retinoid receptor hybridization signals were normalized to those of 36B4. Data are presented as fold change of normalized values relative to non-irradiated control skin, and expressed as means ± SEM, N=7-12. *p<0.05 versus control. (C) UV dose dependence for reduction of RAR-γ mRNA (open bars) and RXR-α mRNA (hatched bars) in human skin in vivo. Skin was obtained for analyses eight hours following exposure to the indicated doses of UV. Inset shows representative Northern blots for RAR-γ RXR-α and 36B4 (internal control). Retinoid receptor hybridization signals were normalized to those of 36B4. Data are presented as fold change of normalized values relative to non-irradiated control skin, and expressed as means ± SEM, N=7-12. *p<0.05 versus control.

Fig. 2. Repeated exposure to UV irradiation causes sustained reduction of RXR-α but not RAR-γ in human skin in vivo. (A) Northern analyses of RAR-γ mRNA (open bars) and RXR-α mRNA (hatched bars) were performed on total RNA extracted from UV-irradiated and non-irradiated human skin. Separate skin sites received one, two, or three exposures to IMED UV at 24-hour intervals. Skin was obtained for analyses eight hours following the last UV exposure. Inset shows representative Northern blots for RAR- γ RXR-α and 36B4 (internal control). Retinoid receptor hybridization signals were normalized to those of 36B4. Data are presented as fold change of normalized values relative to non-irradiated control (CTRL) skin, and are expressed as means ± SEM, N=4 *p<0.05 versus control. (B) RAR-γ protein (open bars) and RXR-α protein (hatched bars) levels were determined by Western blot. Skin sites received one or three exposures to IMED UV at 24-hour intervals. Separate skin was obtained for analyses eight (second and fourth pairs of bars) or 56 hours (third pair of bars) after the last UV exposure. Inset shows representative Western blots for RAR-γ and RXR-α. Data are presented as fold change of protein levels relative to protein level in non-irradiated control (CTRL) skin, and expressed as means ± SEM, N=5-9. *p<0.05 versus control.

Fig. 3. Pretreatment with topical RA reduces loss of RAR-γ and RXR-α in UV- irradiated human skin in vivo. Adult human skin was treated on two sites with vehicle (VEH) and on two sites with 0.1% all-trans retinoic acid (RA). UV (2MED) was administered to skin 24 hours after application of vehicle or RA. Skin biopsies were obtained eight or sixteen hours following exposure to UV. RAR-γ (open bars) and RXR-α (hatched bars) protein levels were determined by Western blot analyses. Insets show representative Western blots of RAR-γ and RXR-α proteins. Data are presented as fold change in retinoid receptor protein level relative to control vehicle-treated, non- irradiated skin (CTRL), and expressed as means ± SEM, N=4-6. *p<0.05 for comparisons of RAR-γ and RXR-α between RA and VEH.

Fig. 4. UV irradiation abolishes responsiveness to all-trans retinoic acid (RA), but not 1 ,25-dihydroxyvitamin D₃ in human skin in vivo. (A) UV irradiation blocks RA induction of CRABP-II mRNA in human skin in vivo. UV (2MED) was administered to adult human skin two hours before application of RA or vehicle (VEH). Skin was obtained for Northern analyses eight hours following application of RA or vehicle. Inset shows representative Northern blots for CRABP-II mRNA and 36B4 (internal control). CRABP-II hybridization signals were normalized to those of 36B4. Data are presented as fold change of normalized values relative to non-irradiated vehicle-treated skin, and expressed as means ± SEM, N=6. *p<0.05 versus RA treatment. (B) UV irradiation blocks induction of RA 4-hydroxylase (RA 4-OHase) mRNA by RA in human skin in vivo. UV (2MED) was administered to adult human skin six hours before application of RA or vehicle. Skin was obtained for Northern analyses 24 hours following application of RA or vehicle. Inset shows representative Northern blots for RA 4-OHase mRNA and 36B4 (internal control). RA 4-hydroxylase hybridization signals were normalized to those of 36B4. Data are presented as fold change of normalized values relative to non-irradiated, vehicle-treated skin, and expressed as means ± SEM, N=5. *p<0.05 versus RA treatment. (C) UV irradiation does not block induction of vitamin D 24-hydroxylase (Vit. D 24-

OHase) mRNA by 1,25-dihydroxyvitamin D₃ (Vit. D) in human skin in vivo. UV (2MED) was administered to adult human skin six hours before application of 1,25- dihydroxyvitamin D₃ or vehicle. Skin was obtained for Northern analyses 24 hours following application of 1,25-dihydroxyvitamin D₃ or vehicle. Inset shows representative Northern blots for 24-OHase mRNA and 36B4 (internal control). The 36B4 hybridization was used to normalize the retinoid receptor hybridization signals. Data are presented as fold change of normalized values relative to non-irradiated, vehicle-treated skin, and expressed as means ± SEM, N=6. *p<0.05 versus, 1,25-dihydroxyvitamin D₃ treatment.

EXPERIMENTAL RESULTS Unless indicated otherwise, experiments were conducted in vivo with human volunteers.

Ultraviolet irradiation reduces retinoid receptor mRNA and protein levels in human skin in vivo. Initially we investigated the effect of UV irradiation on RAR-γ and RXR-α proteins, the major nuclear retinoid receptors in human skin. Nuclear extracts from UV-irradiated and non-irradiated human skin biopsies were analyzed for nuclear retinoid receptor protein levels by Western blot. Significant reductions in RAR-γ and RXR-α proteins were observed following UV irradiation (Fig. 1 A). Reductions in both RAR-γ and RXR-α proteins were detected as early as one and two hours post UV exposure, respectively, and remained reduced for at least 24 hours (Fig. 1 A). Eight hours after UV exposure, RXR-α protein was reduced over 70%, and RAR-γ protein was reduced over 80% in UV-irradiated skin, compared to non-irradiated skin (Fig. 1 A). At 16 hours post UV, RXR-α protein was further reduced (85%), while RAR-γ protein recovered 20%, to one-half of its original level. By 48 hours after UV irradiation, RXR-α and RAR-γ protein levels returned to values found in non-irradiated skin (Fig. 1 A).

Northern analyses of RNA extracted from UV-irradiated and non-irradiated human skin biopsies revealed that UV (2MED) also markedly reduced both RAR-γ and RXR-α mRNA levels (Fig. IB). Significant reduction of RXR-α and RAR-γ transcripts were observed as early as two and four hours post UV, respectively, and were maximally reduced at eight hours post UV (Fig. IB). Reduction of RXR-α mRNA paralleled reduction in RXR-α protein. Reduction of RAR-γ mRNA lagged behind reduction of RAR-γ protein, suggesting that reduced RAR-γ gene expression does not fully account for the initial reduction of RAR-γ protein following UV irradiation. Levels of RXR-α and

RAR-γ mRNA were decreased approximately 85% and 75%, respectively, relative to non- irradiated skin at eight hours post UV (Fig. IB). RAR-γ mRNA returned to the non- irradiated control level by 16 hours post UV, whereas RXR-α mRNA remained 60% reduced for at least 24 hours following exposure to UV (Fig. IB). UV irradiation reduced both RAR-γ and RXR-α mRNA levels in a dosedependent manner (Fig. 1C). Significant reductions in mRNA for both retinoid receptors were evident in human skin exposed to one-half the amount of UV that causes skin reddening (t^'.e., 0.5 MED). Thus, retinoid receptor mRNA can be reduced by doses of UV that do not cause any noticeable skin reddening. In addition to RAR-γ and RXR-α human skin also expresses detectable RAR-α protein. UV irradiation reduced RAR-α protein in a manner similar to that described above for RAR-γ and RXR-α (data not shown). We could not reliably detect RAR-β, RXR-β, or RXR-γ proteins in either untreated or UV-irradiated human skin.

Repeated exposure to UV has divergent effects on RAR-γ and RXR-α mRNA and protein levels.

We next investigated the effect of multiple UV exposures on RAR-γ and RXR-α mRNA levels. For these studies, human volunteer's skin received one, two, or three UV exposures (1 MED) at 24-hour intervals, and was analyzed eight hours following the last UV exposure. Northern analyses of RNA extracted from UV-irradiated and non-irradiated human skin revealed that multiple UV exposures affected RXR-α to a greater degree than

RAR-γ (Fig. 2A). As expected, both RAR-γ and RXR-α mRNA were decreased eight hours after the first exposure to UV. Eight hours after the second exposure, RAR-γ mRNA exhibited a slight recovery relative to the level observed after the first UV exposure, while RXR-α mRNA was further decreased, to about 25% of its level in non- irradiated skin. Eight hours following the third UV exposure, RAR-γ mRNA had recovered to approximately 85% of the level in non-irradiated skin, while RXR-α mRNA remained reduced at 35% of control levels (Fig. 2A). We next examined the effect of multiple UV exposures on RAR-γ and RXR-α protein levels (Fig. 2B). Skin was exposed three times to 1 MED UV at 24-hour intervals and analyzed eight hours after the last exposure (a total of 56 hours after the first exposure). For comparison, separate sites on each subject were also irradiated once with 1 MED UV, and skin was analyzed eight hours and 56 hours after irradiation. As expected,

RAR-γ and RXR-α protein levels were reduced eight hours following a single UV exposure (Fig. 2B). By 56 hours after a single UV exposure, both RAR-γ and RXR-α proteins had returned to levels comparable to those in non-irradiated skin (Fig. 2B). However, following three 1 MED UV exposures, RXR-α protein remained significantly reduced, compared to its level in non-irradiated skin. In contrast, RAR-γ protein returned to the original level seen in non-irradiated skin, following three UV exposures. These data indicate that daily UV exposures result in sustained reduction of RXR-α mRNA and protein in human skin.

Pretreatment with topical RA partially blocks UV-induced loss of RAR-γ and RXR-α proteins in human skin.

We next investigated the effect of topical application of RA on UV-induced reduction of retinoid receptors. For these studies, skin was pretreated with RA or its vehicle 24 hours prior to UV irradiation. Skin biopsies were obtained from irradiated and adjacent non-irradiated skin eight or 16 hours after irradiation. RA treatment alone had no effect on RAR-γ or RXR-α protein levels (data not shown). At eight hours post UV,

Western blot analyses revealed that both RAR-γ and RXR-α proteins in vehicle-treated skin were markedly reduced (approximately 80% and 70% reduced, respectively), compared to their levels in non-irradiated skin (Fig. 3). UV irradiation of RA-pretreated skin also caused reductions in RAR-γ and RXR-α proteins at eight hours post UV. However, these reductions were less than in vehicle-treated skin (Fig. 3). This protective effect of RA was more pronounced at 16 hours post UV. RA pretreatment completely blocked loss of RAR-γ and reduced loss of RXR-α by more than 30% some 16 hours after UV irradiation (Fig. 3). RA pretreatment had no effect on UV-induced reductions in RAR-γ and RXR-α mRNA (data not shown). UV irradiation causes functional vitamin A deficiency in human skin in vivo.

We next investigated the functional consequences of UV-induced reduction of RAR and RXR proteins on retinoid receptor-dependent gene expression in human skin. The genes for cellular retinoic acid binding protein-II (CRABP-II) and RA 4-hydroxylase contain functional RA response elements in their promoters, and are induced by RA in human skin and skin cells. We therefore employed induction of CRABP-II and RA 4- hydroxylase transcripts by RA as indicators of retinoid receptor function. In these studies, RA treatment induced CRABP-II nine-fold and RA 4-hydroxylase 17-fold in human skin in vivo (Fig. 4A, B). UV alone did not affect either CRABP-II or RA 4-hydroyxlase mRNA levels. However, when 2 MED UV was administered two hours (for CRABP-II) or six hours (for RA 4-hydroxylase) before topical application of RA, induction of CRABP-II and RA 4-hydroxylase transcripts was essentially abolished (Fig. 4A, B). These data indicate that reduction of RAR and RXR proteins by UV results in functional impairment of receptor-dependent retinoid responsiveness in human skin in vivo.

In addition to heterodimerizing with RARs, RXRs heterodimerize with other nuclear receptors, including thyroid hormone receptors, peroxisome proliferator-activated receptor, and vitamin D₃ receptor (VDR). The promoter of the vitamin D 24-hydroxylase gene contains a vitamin D response element that is activated by 1,25-dihydroxyvitamin D₃ in human skin. This induction involves transcriptional activation by VDR/RXR heterodimers. We therefore examined whether UV alters 1,25-dihydroxyvitamin D₃ induction of vitamin D 24-hydroxylase in human skin.

Vehicle-treated human skin expressed low levels of vitamin D 24-hydroxylase mRNA. A single application of 0.05% 1 ,25-dihydroxyvitamin D₃ induced vitamin D 24- hydroxylase mRNA 30-fold (Fig. 4C). Exposure of human skin to UV had no effect on the level of vitamin D 24-hydroxylase mRNA (data not shown). Similarly, UV irradiation six hours before application of 0.05% 1,25-dihydroxyvitamin D₃had no effect on 1,25- dihydroxyvitamin D₃ induction of vitamin D 24-hydroxylase mRNA (34-fold + 4.7) (Fig. 4C). Thus, in contrast to the observed impairment by UV irradiation of RA responsiveness, UV irradiation had no apparent effect on 1,25-dihydroxyvitamin D₃ responsiveness in human skin.

DISCUSSION

The data presented above demonstrate that UV irradiation causes loss of RAR-γ and RXR-α, the predominant nuclear retinoid receptors in human skin. This UV-induced loss of RAR-γ and RXR-α occurred at both the mRNA and protein levels. The deficiency of nuclear retinoid receptors was associated with a near complete loss of RA induction of RA-responsive genes in human skin in vivo. This state of retinoid non-responsiveness can be equated with a state of functional vitamin A deficiency. In view of the demonstrated requirement of retinoids to maintain normal skin cell homeostasis, loss of retinoid responsiveness would be expected to have deleterious consequences for skin function. Reduction of retinoid receptors is therefore a heretofore-unrecognized mechanism by which UV irradiation can damage human skin.

UV-induced loss of nuclear retinoid receptor proteins was rapid, occurring within one to two hours after a single exposure to UV. UV-induced loss of RAR-γ mRNA lagged behind loss of RAR-γ protein, suggesting that the initial reduction of RAR-γ protein may have resulted from accelerated breakdown, rather than reduced synthesis. UV-induced loss of RXR-α mRNA and protein were coincident, making it difficult to assess the relative contribution of reduced mRNA and accelerated breakdown to the initial reduction of RXR-α protein. UV generates reactive oxygen species that damage proteins through chemical oxidation. Oxidized proteins are recognized and degraded by proteasome and other disposal pathways. We have observed that RAR-γ and RXR-α are degraded by the proteasome pathway in cultured human keratinocytes. It is likely, therefore, that the loss of retinoid receptors that occurs in human skin following UN irradiation results from both reduced synthesis and proteasome-mediated degradation. Treatment of skin with RA prior to UV exposure mitigated UV-induced loss of RAR-γ and RXR-α proteins. This protective effect of RA was most evident 16 hours after

UV irradiation, at which time the level of RAR-γ protein was similar to that found in non- irradiated skin. RA pretreatment, however, had no effect on UV-induced decreases in RAR-γ and RXR-α mRΝA. These data suggest that RA may stabilize the retinoid receptors by retarding their breakdown. This could occur as a direct consequence of RA binding to RAR-γ and 9-cis retinoic acid, which is formed from RA in skin, binding to

RXR-α. This mechanism is analogous to that reported for stabilization of the vitamin D receptor by its ligand, 1,25-dihydroxyvitamin D₃. (See Arbour, Ν.C., Prahl, J.M., and DeLuca, H.F., "Stabilization of the vitamin D receptor in rat osteosarcoma cells through the action of 1,25-dihydroxyvitamin D₃," Mol Endocrinol 93, 13007-13012 (1993).) RAR-γ and RXR-α mRNA began to recover between 8 and 16 hours following a single UV exposure. By 16 hours after UV irradiation, RAR-γ mRNA was fully recovered, while RXR-α mRNA remained 50% reduced. This slower recovery of RXR-α mRNA likely contributed to the reduced level of RXR-α protein observed 16 hours after UV irradiation. By 48 hours following a single UV exposure, both RAR-γ and RXR-α had fully recovered. Interestingly, eight hours following three successive UV exposures at

24-hour intervals, RAR-γ mRNA and protein had recovered to their normal levels, whereas RXR-α mRNA and protein remained significantly reduced. Thus, RAR-γ appears to become refractory to the effects of UV irradiation, whereas RXR-α does not. The reason for this difference is not known. The ratio of RXR-α to RAR-γ is approximately five to one in normal human skin (Fisher, G.J. et al., "Immunological identification and functional quantitation of retinoic acid and retinoid X receptor proteins in human skin," J Biol Chem 269, 20629-20635 (1994)). Following multiple UV exposures, this ratio becomes approximately one to one due to reduced RXR-α levels. RAR-γ must heterodimerize with RXR-α to function in skin. RXR-α, however, is a shared partner with several other nuclear receptors in skin, including vitamin D receptors, thyroid hormone receptors, and peroxisome proliferator-activated receptors. It is likely, therefore, that retinoid responsiveness will be limited in skin chronically exposed to UV irradiation by reduced levels of RXR-α, even though RAR-γ levels may not be significantly reduced. In contrast to the observed loss of retinoid responsiveness in UV-irradiated human skin, UV irradiation did not alter the ability of human skin to respond to 1,25-dihydroxyvitamin D₃. This finding indicates that UV selectively interferes with retinoid signaling. The vitamin D receptor in human skin binds to DR3 response elements exclusively as a heterodimer with RXR. The vitamin D 24-hydroxylase gene contains DR3 response elements in its promoter, which confer VDR/RXR-mediated induction by 1,25- dihydroxyvitamin D₃. In human skin, the RXR ligand 9-cis RA synergistically augments 1,25-dihydroxyvitamin D₃ induction of vitamin D 24-hydroxylase, indicating that, in human skin, RXR functions in the vitamin D response. Human skin contains approximately twenty times more RXR-α than vitamin D receptors (Li et al. , submitted manuscript).

Therefore, in spite of RXR-α being significantly reduced in UV-irradiated skin, its level still exceeds that of vitamin D receptors. Under these conditions, heterodimerization of vitamin D receptors with remaining RXR-α would be unimpaired in UV-irradiated skin. This finding likely explains the observed lack of effect of UV irradiation on vitamin D receptor-mediated gene expression in human skin.

In addition to regulating expression of genes that contain retinoic acid response elements, retinoid receptors alter the activity of other families of transcription factors. Most notably, retinoid receptors can interfere with the activity of transcription factor AP-1. UV irradiation activates AP-1 in human skin in vivo, and this activity follows a time course similar to that described above for UV-induced loss of retinoid receptors. Thus,

UV simultaneously reduces retinoid receptors and induces AP-1. AP-1 induction in human skin occurs through increased expression of c-Jun and activation of c-Jun N-terminal kinase (JNK) activity. However, since retinoid receptors can negatively regulate AP-1, retinoid receptor loss following UV irradiation could further enhance AP-1 activation in UV-irradiated human skin.

UV irradiation from the sun is the primary causative agent for premature skin aging (photoaging) and skin cancer, which is the most prevalent form of human malignancy. AP- 1 activity is a critical mediator of photoaging in humans, and is required for tumor progression in the mouse model of skin carcinogenesis. Retinoic acid antagonizes AP-1 action in a human photoaging model, and retinoid receptors are reduced during skin tumor progression, and in approximately 90% of skin tumors in humans. We discovered that a single UV exposure caused reversible reductions in RAR-γ and RXR-α while daily UV exposures resulted in a sustained reduction of RXR-α but not RAR-γ. UV exposures that were too little to cause any skin reddening (0.5MED) resulted, nevertheless, in statistically significant reductions of RAR-γ and RXR-α. Reduced retinoid receptors result in a state of functional vitamin A deficiency that, in conjunction with activated AP-1, could promote photoaging and tumor development in skin. Use of RA may prevent, at least in part, loss of retinoid receptors and thereby act to mitigate the participation of AP-1 in photoaging and skin carcinogenesis.

Retinoids in general are useful for preventing the functional vitamin D deficiency in human skin caused by exposure to UV radiation. Retinoids, besides vitamin A acid (retinoic acid (RA)), include retinol (vitamin A) and natural and synthetic analogs of vitamin A, vitamin A aldehyde (retinal), including all-trows, 9-cis, and 13-cts retinoic acid), etretinate, and others as described in EP-A2-0379367, US 4,887,805, and US 4,888,342

(the disclosures of which are all incorporated herein by reference). Various synthetic retinoids and compounds having retinoid activity are expected to be useful in this invention, to the extent that they exhibit retinoid activity in vivo, and such are described in various patents assigned on their face to Allergan Inc., such as in the following U.S.

Patents, numbered: 5,514,825; 5,698,700; 5,696,162; 5,688,957; 5,677,451; 5,677,323;

5,677,320; 5,675,033; 5,675,024; 5,672,710; 5,688,175; 5,663,367; 5,663,357; 5,663,347;

5,648,514; 5,648,503; 5,618,943; 5,618,931; 5,618,836; 5,605,915; 5,602,130. Still other compounds described as having retinoid activity are described in other U.S. Patents, numbered: 5,648,563; 5,648,385; 5,618,839; 5,559,248; 5,616,712; 5,616,597;

5,602,135; 5,599,819; 5,556,996; 5,534,516; 5,516,904; 5,498,755; 5,470,999; 5,468,879;

5,455,265; 5,451,605; 5,343,173; 5,426,118; 5,414,007; 5,407,937; 5,399,586; 5,399,561;

5,391,753; and the like, the disclosures of all of the foregoing and following patents and literature references hereby incorporated herein by reference.

METHODS

Procurement of human skin biopsies. Keratome biopsies were obtained from healthy adult human volunteers as previously described. (See Fisher, G. J. et al, "Cellular, immunologic and biochemical characterization of topical retinoic acid-treated human skin," J Invest Dermatol 96, 699-707 (1991).) All procedures involving human subjects were approved by the University of Michigan Institutional Review Board prior to initiation of the study, and all subjects provided written informed consent.

UV irradiation procedure. Adult Caucasians (approximately equal numbers of males and females) with mild to moderate pigmentation and no current skin disease or history of skin disease were irradiated with an Ultralite Panelite lamp containing four

F36T12 ERE-VHO UVB tubes. A Kodacel TA4011/407 filter was mounted 4 cm in front of the tubes to remove wavelengths below 290 nm (UVC). Irradiation intensity was monitored using an IL443 phototherapy radiometer and a SED240/UVB/W photodetector (International Light, Newbury, MA). Spectralradiometry was performed using an Optronic Laboratories OL 754 system. Total irradiance (290-800 nm) was 1.49 x 10^'3 w/cm² at a distance 17 inches from the source. Power output distribution was 47% UVB (290-320 nm), 18% UVA2 (320-340 nm), 9% UVA1 (340-400 nm), and 26% visible and near infrared (400-800 nm). The minimal erythema dose (MED) for each subject was determined 24 hours after irradiation. One MED for all subjects ranged from 30-50 mJ/cm².

In vivo treatments with RA and 1,25-dihydroxyvitamin D₃. RA and its vehicle (70% ethanol, 30% polyethyleneglycol, 0.05% BHT) were applied to skin under occlusion,

24 hours prior to UV treatment. For measurement of CRABP-II mRNA, RA was applied to skin two hours after exposure to UV, and skin was obtained eight hours after treatment with RA. For measurement of RA 4-hydroxylase mRNA RA was applied six hours after UV irradiation, and skin was obtained 24 hours after treatment with RA. 1,25- dihydroxyvitamin D₃ was applied to skin six hours after UV irradiation, and skin was obtained 24 hours later for measurement of vitamin D 24-hydroxylase mRNA.

RNA isolation and Northern blot analysis. For isolation of total RNA, skin biopsies were immediately frozen in liquid nitrogen and stored at -80^° C until subsequent isolation of RNA. Frozen skin was ground into a fine powder in liquid nitrogen, and total RNA was isolated as previously described. RNA samples (40 μg total RNN determined by OD 260nm absorption spectrophotometry) were electrophoretically size-fractionated on 1.2% formaldehyde-agarose gels, transferred to nylon membranes (Schleicher & Schuell, Keene, NH) and UV cross-linked. Blots were hybridized against [³²P]labeled RAR-γ, RXR-α, CRABP-II, RA 4-hydroxylase, and vitamin D 24-hydroxylase cDNA probes, as described. Each blot was stripped and reprobed with [³²P]labeled 36B4 cDNA to normalize RNA loading. Hybridization band intensities were quantified with a STORM Phosphorlmager (Molecular Dynamics, Sunnyvale, CA). Results were expressed as fold changes relative to vehicle-treated control, or non-irradiated control.

Nuclear extract preparations from human skin. Fresh skin biopsies were placed in 0.25% trypsin, 0.1% EDTA for 30 min. at 37°C. Trypsinization was stopped by adding

Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum, and cells were released from the tissue by scraping. Released cells were pipetted up and down several times to form a single cell suspension. Cell suspensions were passed through a nylon filter to remove residual tissue. Cells (mainly keratinocytes) were pelleted by brief centrifugation at 4,000 rpm in a table-top centrifuge at 4°C. Nuclear extracts were prepared as previously described. (Fisher, G.J. et al., "Molecular basis of sun-induced premature skin ageing and retinoid antagonism," Nature 379, 335-339 (1996); Xiao, J.H., Durand, B., Chambon, P., and Voorhees, J.J., "Endogenous retinoic acid receptor RAR-γ X receptor RXR-α heterodimers are the major functional forms regulating retinoid- responsive elements in adult human keratinocytes," J Biol Chem 270, 3001-3011(1995).)

Extracts were aliquoted and stored at -80°C for future use. Protein determinations were carried out according to the method of Bradford, with BSA as standard (see Bradford, M., "A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding," AnalBiochem 72, 248-251(1976)).

Expression of retinoid receptor proteins. Human RAR-α, RAR-β, RAR-γ, RXR-α, and RXR-β were transiently over-expressed in Hela cells using Superfect transfection reagent (Qiagen, Chatsworth, CA) according to the manufacturer's instructions. Twenty-four hours after transfection, culture media were removed, cells were scraped into Eppendorf tubes, and lysed in buffer containing 50 mM Tris-HCl (pH 8.0), 5 mM EDTA, 150 mM NaCl, 0.5% NP-40, and 1 mM DTT. Cells were sonicated for 30 sec. on ice, using a probe sonicator. After sonication, the homogenate was centrifuged at 15,000 x g for 20 min. to remove particulate material. The supernatant was collected as the source of recombinant retinoid receptors, which were used as standard for Western analyses, as described in the text.

Western blot analysis. Equal amounts (50μg) of nuclear proteins from human skin were denatured by boiling for ten min. in SDS sample buffer. Samples were separated by SDS polyacrylamide gel electrophoresis, using 10% acrylamide gels. After electrophoresis, the separated proteins were transferred to Immobilon-P membranes (Millipore Co., Bedford, MA). Membranes were blocked with 5% nonfat dry milk in PBS containing 0.05% Tween-20 (PBST) for at least one hour. After the blocking reaction, membranes were incubated with subtype-specific anti-retinoid receptor antibodies (diluted 1 : 1000 in PBST) (Santa Cruz Biotechnology, Santa Cruz, CA). The receptor-antibody complexes were detected by enhanced chemiluminescence (ECL) or by enhanced chemifluorescence (ECF) (ECL and ECF Western blotting systems available from Amersham Corp., Arlington Heights, IL). Bands were quantified by laser densitometry or by STORM Phosphorlmager. Statistics. Comparisons among treatment groups were made with the paired t test. Multiple pairwise comparisons were made with the Tukey Studentized Range test. All/? values are two-tailed, and significant when ≤0.05.

The foregoing description is meant to be illustrative and not limiting. Various changes, modifications, and additions may become apparent to the skilled artisan upon a perusal of this specification, and such are meant to be within the scope and spirit of the invention as defined by the claims.

Claims

What is claimed is:

1. A method for inhibiting the reduction in at least one of RXR and RAR proteins caused by irradiation of human skin with UV light, by providing a retinoid in a carrier suitable for topical administration to human skin, and applying the retinoid to human skin prior to irradiation with the UV light in an amount effective to inhibit the UV-mediated degradation of at least one of RAR and RXR.

2. The method of claim 1, wherein reduction in RXR is prevented.

3. The method of claim 1, wherein reduction in RXR-α is prevented.

4. The method of claim 1, wherein the retinoid is selected from the group consisting of retinoic acid and retinol.

5. The method of claim 1, wherein the UV light is from sunlight.

6. A method for inhibiting functional vitamin D deficiency characterized by reduced retinoid receptors because of exposure of human skin to UV radiation, which method comprises pretreating the skin later to be exposed to UV radiation with a retinoid in an amount effective to inhibit the UV-mediated degradation of at least one of RAR and

RXR.

7. The method of claim 6, wherein reduction in RXR is prevented.

8. The method of claim 6, wherein reduction in RXR-α is prevented.

9. The method of claim 6, wherein the retinoid is selected from the group consisting of retinoic acid and retinol.

10. The method of claim 6, wherein the UV light is from sunlight.