Retinoic‐acid receptor β expression in melanocytes

ARTICLE

Auteur(s) : Nelly BOEHM¹, Brigitte SAMAMA¹, Bernard CRIBIER², Cécile ROCHETTE-EGLY³

¹ Institut d'Histologie, Faculté de Médecine, 4 rue Kirschleger 67085, Strasbourg, France
² Clinique Dermatologique, Hôpitaux Universitaires, Strasbourg, France
³ Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch Graffenstaden, France

Article accepted on 2/10/2003

Retinoids, the natural and synthetic analogues of vitamin A, have been used to manage various skin disorders including skin cancers [1] as well as for their chemopreventive effects [2]. Retinoids act through their ability to modulate cell growth, differentiation and apoptosis. The effects of retinoids are mediated by nuclear receptors that belong to the steroid/thyroid hormone nuclear receptor family [3]. Two distinct classes of nuclear proteins have been identified: the retinoic acid receptors (RARs) and retinoid X receptors (RXRs). Each class consists of three subclasses: α, β, γ. RARs and RXRs act as ligand-dependent transcription factors, which bind to cis-acting DNA sequences located in the promoter of RA-target genes. RXRs form either homodimers or heterodimers with RARs but also with other members of the nuclear receptors family [3]. In adult human epidermis and in cultured keratinocytes, mRNAs for RARα and γ and RXRα and β are detected, with predominant expression of RARγ and RXRα [4-6]. As only very few RARβ transcripts or protein levels could be detected in skin or keratinocytes [7], we aimed to re-evaluate RARβ expression by immunohistochemistry of normal skin epidermis and of melanocytic tumours on routinely fixed and paraffin embedded sections.

Materials and methods

Skin specimens

In this study, eighteen paraffin-embedded archival skin biopsies were used. Six corresponded to normal skin; six other biopsies were benign melanocytic tumours and the last six were melanomas. Some tumours also contained normal adjacent tissue. The diagnoses were based on typical histopathologic features and were confirmed independently by two dermatopathologists.

Primary antibodies

The RXRα and RARβ mouse monoclonal antibodies used in this study were Mab 4RX-3A2 [8] and Mab 8β-10B2 [9]. Melanocytes were immunostained using monoclonal antibodies against protein S-100 (Novocastra) and Melan-A/MART-1 antigen (Melan-A; Dako Corporation).

Immunocytochemistry

Five µm thick sections were mounted on chrome alum subbed glass slides. After deparaffinization, antigens were demasked by microwave heating. Endogenous peroxidase was neutralised by H2O2 and aspecific reactions were prevented by incubating sections in blocking buffer (normal horse serum 2%) diluted in PBS-Tween (pH 7.4). Blocking buffer was used for reagent solutions and all washes were performed in PBS-Tween. Sections were incubated with primary antibodies (RARβ and RXRα antibodies:1/10 000; Melan-A: 1/200) in a moist chamber at 4 °C overnight. Then, they were incubated with biotinylated-horse antimouse antibody for one hour at room temperature and for a further 30 minutes in streptavidine-biotine-peroxidase complex (Vectastain ABC Elite; Vector Laboratories). VIP (Vector Laboratories) was used as chromogen, giving a purple reaction. Sections were counterstained with methyl green. Omission of primary antibodies, or their immunoadsorbtion with the corresponding peptide, both resulted in disappearance of the signal.
The number of RARβ positive nuclei in the basal layer of epidermis was compared to the number of Melan-A positive cells on adjacent sections in both normal biopsies and normal epidermis in tumour biopsies.

Results

All signals described for retinoid receptors were localized in the nuclei. In each category of biopsies, similar results were observed.

Normal skin

RXRα protein was detected as a strong signal in epithelial cells: epidermis (Fig. 1A, C), sebaceous gland (Fig. 1A, C) and hair follicle (Fig. 1B). In the epidermis, the strongest signal was observed in spinocellular layers, with scattered positive cells in the basal layer (Fig. 1C); in sebaceous glands, the strongest signal was observed in differentiating cells but disappeared in fully differentiated cells before they detached from the gland (Fig. 1A, C); in sudoral ducts, one to two nuclei per cross-section, mainly localised in the basal layer were weakly stained. Endothelial cells and fibroblasts were only occasionally positive. Adipocyte nuclei were more strongly stained.
The expression of RARβ in the epidermis was found to be rather low when compared to that of RXRα. RARβ immunopositivity was localised only in two regions of the epidermis: in the basal layer, round cells with a clear cytoplasm, regularly placed along the basal lamina were stained; in the upper part of the epidermis, some granulous cells were stained (Fig. 1E). In sebaceous glands, only differentiated cells (with picnotic nuclei) were stained (Fig. 1F). In the dermis, scattered cells (fibroblasts, endothelial cells, and muscular cells) were stained.
When comparing Melan-A with RARβ immunoreactivity in the basal layer (Fig. 1D, 1E), a ratio of 1.25 between Melan A- and RARβ positive cells was observed.

Melanocytic tumours

RXRα antibodies did not give any signal in melanocytic tumours, neither in nevi nor in melanomas (Figs. 2C,  3D,  3E). Moreover, keratinocytes localised above melanoma but not benign tumours were less stained than the adjacent normal skin (Fig. 3D,  3E).
In contrast, RARβ antibodies gave a signal in a great number of nevus cells (Fig. 2A, 2B) with the strongest signal in the nests of nevus cells localised in the superficial dermis (Fig. 2D,  2E). However, no immunoreactivity could be detected in melanoma (Fig. 3B).
As shown in Figs 3A, B, C which illustrate a melanoma developed from a nevus (Fig. 3A), nevus cells express RARβ (Fig. 3C), but not melanoma cells (Fig. 3B).

Discussion

The study re-evaluates RARβ expression in normal skin. It also demonstrates that both RARβ and RXRα are absent in melanoma cells.
Our results corroborate biochemical results showing that very few RARβ transcripts or protein could be detected in skin or keratinocytes [4, 7, 10]. We show that among skin epithelial structures, only the most superficial keratinocytes of both epidermis and hair follicle and fully differentiated sebaceous cells express RARβ. RARβ expression differs among keratinocyte subtypes: in non-keratinizing squamous epithelia such as oesophageal and cervical epithelia, RARβ mRNA and protein are detected [12-14]. Zou et al., 1999 [14] observed an inverse association between RARβ expression and keratinization markers in squamous carcinoma cells, implicating this RAR isotype in suppression of keratinizing terminal differentiation.
A RARβ immunostaining is also observed in the basal layer. Based on several criteria, this staining would concern melanocytes. An indirect proof comes from the fact that RARβ is expressed in nevus cells, although not all cells are stained in either normal skin or benign tumours. Double staining with anti-S-100 protein and RARβ antibodies would give definitive results; unfortunately, very high S-100 protein immunostaining obscured the RARβ signal.
In contrast, melanoma cells do not express RARβ. Similar results were reported concerning murine melanoma cells lines (S91 and B16 cell lines). It is interesting to note that melanoma cell lines display highly variable responsiveness to RA. Although many of them are RA-resistant, the murine lines described above are RA-sensitive since their treatment with RA or synthetic elective RAR agonists induces differentiation, growth arrest as well as the induction of RARβ expression [15-18].
RARβ expression is decreased in many premalignant and malignant squamous cell carcinomas (lung, head and neck, oesophagus, mammary gland, pancreas, cervix) (for review see 10) and RARβ reexpression usually correlates with growth arrest and differentiation. Therefore, our observation that RARβ is also decreased in melanoma, confirms the notion that this receptor is the target retinoid receptor in most cancers and that it would act as a «tumour suppressor». Adult mice conditionally knocked out for RARβ will give more insights concerning the role of this receptor in melanocytes and epithelial cell growth and differentiation.
We also studied the expression of RXRα in skin. Our results are in accordance with other results showing that this receptor is strongly expressed in skin [19, 20], especially in epithelial structures. RXRα expression in epidermis is in agreement with very recent results from Chapellier et al., 2002 [21] which showed that topical retinoid signal is transduced by RXRα/RARγ heterodimers in suprabasal epidermal keratinocytes, which in turn, stimulate proliferation of basal keratinocytes via a paracrine signal. This is also in agreement with our recent report showing that RXRα expression is mainly present in basal cells of the cyclic vaginal epithelium, during the proliferating stage [22]. Staining was very strong in hair follicle, a region of cell proliferation; in that respect, RXR agonists were reported to stimulate hair growth in vitro [23].
In most tumours (such as lung cancers), RXRα expression is substantially altered or even increased [24]. However, in the present study, we did not observe any expression of RXRα in melanic tumours. This is in agreement with an antiproliferative role for RXRα. Indeed, ablation of RXRα in mouse embryocarcinoma cells (F9 cells) resulted in an increased cell growth and the suppression of RA antiproliferative response [24]. Moreover, temporally controlled RXRα ablation in mouse epidermis resulted in hair follicle degeneration and in keratinocyte hyperproliferation [25].
In fact, our results indicate that melanoma are characterised by a simultaneous decrease in RARβ and an absence of RXRα that would result in deficient RARβ/RXRα heterodimers. Note that a concomitant decrease in RARβ and other RXR isotypes (RXRβ) has been observed in some lung cancers [26]. Consequently, such defects would alter a variety of pathways under the control of RA, including cellular differentiation programs, cell cycle control and the expression of genes that are regulated by RARβ/RXRα heterodimers. Collectively, all these results corroborate the notion that many tumours are characterised by defective RAR/RXR heterodimers.
What is the cause of the loss in RARβ and RXRα observed in melanoma? A number of mechanisms have been proposed such as a defect in the transcription of both receptors. Indeed, in many types of cancer such as lung cancer, the loss of RARβ expression has been correlated to the hypermethylation of the RARβ promoter [27] and/or to a deficient acetylation of histones [28] resulting in both cases, in an aberrant repressive state of chromatin. However, other mechanisms should not be excluded. Indeed, many types of cancers including melanoma and breast cancers have been correlated to an aberrant upregulation of the receptor tyrosine kinases [29, 30] and of the downstream MAP kinases [31, 32]. As the turnover of many nuclear receptors including RARs and RXRs is regulated by their phosphorylation [33, 34], one can hypothesise that the observed loss of RARβ and RXRα might reflect an accelerated turnover due to their aberrant phosphorylation by MAPKs.
All these hypotheses are of great importance from the therapeutical point of view, especially because most melanoma cell lines are RA-resistant [35]. The combination of RA with drugs reversing the aberrant repressive state of chromatin and/or inhibiting the aberrant kinase pathways [36, 37] might help to overcome some RA-resistant melanoma. n

Acknowledgements. We are grateful to Professor P. Chambon for the interest he took in this work, making it possible. We thank Professor E. Grosshans for helpful comments on the manuscript. The technical assistance of P. Boos and E. Varnaison is gratefully acknowledged.

References

1. Orfanos CE, Zouboulis CC, Almond-Roesler B, Geilen CC. Current use and future potential role of retinoids in dermatology. Drugs 1997; 53: 358-88.

2. Levine N, Moon TE, Cartmel B, Bangert JL, Rodney S, Dong Q, Peng YM, Alberts DS. Trial of retinol and isotretinoin in skin cancer prevention: a randomised, double-blind, controlled trial. Southwest Skin Cancer Prevention Study group. Cancer Epidemiol Biomarkers Prev 1997; 6: 957-61.

3. Chambon P. The molecular and genetic dissection of the retinoid signalling pathway. Recent Prog Horm Res 1995; 50: 317-32.

4. Elder JT, Fisher GJ, Zhang QY, Eisen D, Krust A, Kastner P, Chambon P, Voorhees JJ. Retinoic acid receptor gene expression in human skin. J Invest Dermatol 1991; 96: 425-33.

5. Elder JT, Aström A, Petterson U, Tavakkol A, Griffiths CE, Krust A, Kastner P, Chambon P, Voorhees JJ. Retinoid acid receptors and binding proteins in human skin. J Invest Dermatol 1992; 98: 36S-41S.

6. Torma H, Rollman O, Vahlquist A. Detection of mRNA transcripts for retinoic acid, vitamin D3, and thyroid hormone (c-erb-A) nuclear receptors in human skin using reverse transcription and polymerase chain reaction. Acta Derm Venereol 1993; 73: 102-7.

7. Fisher GJ, Voorhees JJ. Molecular mechanisms of retinoid actions in skin. FASEB J 1996; 10: 1002-13.

8. Rochette-Egly C, Gaub MP, Lutz Y, Ali S, Scheuer I, Chambon P. Retinoic acid receptor-beta: immunodetection and phosphorylation on tyrosine residues. Mol Endocrinol 1992; 12: 2197-206.

9. Rochette-Egly C, Lutz Y, Pfister V, Heyberger S, Scheuer I, Chambon P, Gaub MP. Detection of retinoid X receptors using specific monoclonal and polyclonal antibodies. Biochem Biophys Res Commun 1994; 204: 525-36.

10. Xu XC, Wong WYL, Goldberg L, Baer SC, Wolf JE, Ramsdell WM, Alberts DS, Lippman SM, Lotan R. Progressive decreases in nuclear retinoid receptors during skin squamous carcinogenesis. Cancer Res 2001; 61: 4306-10.

11. Crowe DL, Hu L, Gudas LJ, Rheinwald JG. Variable expression of retinoic acid receptor (RAR beta) mRNA in human oral and epidermal keratinocytes; relation to keratin 19 expression and keratinization potential. Differentiation 1991; 48: 199-208.

12. Geisen C, Denk C, Gremm B, Baust C, Karger A, Bollag W, Schwarz E. High-level expression of the retinoic acid receptor beta gene in normal cells of the uterine cervix is regulated by the retinoic acid receptor alpha and is abnormally down-regulated in cervical carcinoma cells. Cancer Res 1997; 57: 1460-7.

13. Zhang W, Rashid A, Wu H, Xu XC. Differential expression of retinoic acid receptors and p53 protein in normal, premalignant, and malignant oesophageal tissues. J Cancer Res Clin Oncol 2001; 127: 237-42.

14. Zou CP, Hong WK, Lotan R. Expression of retinoic acid receptor beta is associated with inhibition of keratinization in human head and neck squamous carcinoma cells. Differentiation 1999; 64: 123-32.

15. Clifford JL, Petkovich M, Chambon P, Lotan R. Modulation by retinoids of mRNA levels for nuclear retinoic acid receptors in murine melanoma cells. Mol Endocrinol 1990; 4: 1546-55.

16. Spanjaard RA, Sugawara A, Ikeda M, Chin WW. Evidence that retinoid X receptors mediate retinoid-dependent transcriptional activation of the retinoic acid receptor beta gene in S91 melanoma cells. J Biol Chem 1995; 270: 17429-36.

17. Xiao Y, Desai D, Quick TC, Niles RM. Control of retinoic acid receptor expression in mouse melan by cyclic AMP. J Cell Physiol 1996; 167: 413-21.

18. Desai SH, Boskovic G, Eastham L, Dawson M, Niles RM. Effect of Receptor-Selective Retinoids on Growth and Differentiation Pathways in Mouse Melanoma Cells. Biochem Pharmacol 2000; 59: 1265-75.

19. Reichrath J, Mnssinger T, Kerber A, Rochette-Egly C, Chambon P, Bahmer FA, Baum HP. In situ detection of retinoid-X receptor expression in normal and psoriatic human skin. Br J Dermatol 1995; 133: 168-75.

20. Reichrath J, Mittmann M, Kamradt J, Muller SM. Expression of retinoid-X receptors (-α,-β,-γ) and retinoic acid receptors (-α,-β,-γ) in normal human skin: an immunohistological evaluation. Histochem J 1997; 29: 127-33.

21. Chapellier B, Mark M, Messaddeq N, Calléja C, Warot X, Brocard J, Gérard C, Li M, Metzger D, Ghyselinck NB, Chambon P. Physiological and retinoid-induced proliferations of epidermis basal keratinocytes are differently controlled. EMBO J 2002; 21: 3402-13.

22. Boehm N, Château D, Rochette-Egly C. Retinoid receptors in rat vaginal and uterine epithelia: changes with ovarian steroids. Mol Cell Endocrinol 1997; 132: 101-8.

23. Billoni N, Gautier B, Mahe YF, Bernard BA. Expression of retinoid nuclear receptor superfamily members in human hair follicles and its implication in hair growth. Acta Derm Venereol 1997; 77: 350-5.

24. Clifford JL, Chiba H, Sobieszczuk D, Chambon P. RXRalpha–null F9 embryonal carcinoma cells are resistant to the differentiation, anti-proliferative and apoptotic effects of retinoids. EMBO J 1996; 15: 4142-55.

25. Li M, Chiba H, Warot X, Messaddeq N, Gérard C, Chambon P, Metzger D. RXRα ablation in skin keratinocytes results in alopecia and epidermal alterations. Development 2001; 128: 675-88.

26. Moldvay J Grignon Y, Vignaud JM, Rochette-Egly C, Martinet Y, Martinet N. Down-regulation of retinoids receptors in malignant mesothelioma. Eur Resp J 2001; 18: 17s (Abstract n° 237).

27. Arapshian A, Kuppumbatti YS, Mira-y-Lopez R. Methylation of conserved CpG sites neighbouring the beta retinoic acid response element may mediate retinoic acid receptor beta gene silencing in MCF-7 breast cancer cells. Oncogene 2000; 19: 4066-70.

28. Suh YA, Lee HY, Virmani A, Wong J, Mann KK, Miller WH, Gazdar A, Kurie JM. Loss of retinoic acid receptor beta gene expression is linked to aberrant histone H3 acetylation in lung cancer cell lines. Cancer Res 2002; 62: 3945-9.

29. Salh B, Marotta A, Wagey R, Sayed M, Pelech S. Dysregulation of phosphatidylinositol 3-kinase and downstream effectors in human breast cancer. Int J Cancer 2002; 98: 148-54.

30. Tari AM, Lim SJ, Hung MC, Esteva FJ, Lopez-Berestein G. Her2/neu induces all- trans retinoic acid (ATRA) resistance in breast cancer cells. Oncogene 2002; 21: 5224-3

31. Oh AS, Lorant LA, Holloway JN, Miller DL, Kern FG, El-Ashry D. Hyperactivation of MAPK induces loss of ERα expression in breast cancer cells. Mol Endocrinol 2001; 15: 1344-59.

32. Govindarajan B, Bai X, Cohen C, Zhong H, Kilroy S, Louis G, Moses M, Arbiser JL. Malignant transformation of melanocytes to melanoma by constitutive activation of mitogen-activated protein kinase kinase (MAPKK) signalling. J Biol Chem 2003; 278: 9790-7.

33. Gianni M, Bauer A, Garattini E, Chambon P, Rochette-Egly C. Phosphorylation by p38MAPK and recruitment of SUG-1 are required for RA-induced RAR degradation and transactivation. EMBO J 2002; 21: 3760-9.

34. Gianni M, Tarrade A, Nigro EA, Garattini E, Rochette-Egly C. The AF-1 and AF-2 domains of RARγ2 and RXRα cooperate for triggering the transactivation and the degradation of RARγ2/RXRα heterodimers. J Biol Chem 2003; 278: 34458-66.

35. Schadendorf D, Worm M, Jurkovsky K, Dippel E, Reichert U, Czarnetzki BM. Effects of various synthetic retinoids on proliferation and immunophenotype of human melanoma cells in vitro. Recent Results Cancer Res 1995; 139: 183-93.

36. Gianni M, Kopf E, Bastien J, Oulad-Abdelghani M, Garattini E, Chambon P, Rochette-Egly C. Down-regulation of the phosphatidylinositol 3-kinase/Akt pathway is involved in retinoic acid-induced phosphorylation, degradation and trascriptional activity of retinoic acid receptor γ2. J Biol Chem 2002; 277: 24859-62.

37. Chang F, Lee JT, Navolanic PM, Steelman LS, Shelton JG, Blalock WL, Franklin RA, McCubrey JA. Involvement of P13K/Akt pathway in cell cycle progression, apoptosis, and neoplastic transformation: a target for cancer chemotherapy. Leukemia 2003; 17: 590-603.