Gene expression analysis of melanocortin system in vitiligo

Gene expression analysis of melanocortinsystem in vitiligo

Kulli Kingo a,c, Eerik Aunin b,c, Maire Karelson a,Mari-Anne Philips b,c, Ranno Ratsep b,c, Helgi Silm a,Eero Vasar b,c, Ursel Soomets c,d, Sulev Koks b,c,e,*

Journal of Dermatological Science (2007) 48, 113—122

www.intl.elsevierhealth.com/journals/jods

aDepartment of Dermatology and Venerology, University of Tartu, EstoniabDepartment of Physiology, University of Tartu, EstoniacCentre of Molecular and Clinical Medicine, University of Tartu, EstoniadDepartment of Biochemistry, University of Tartu, Estoniae Institute of Veterinary Medicine and Animal Sciences, Estonian University of Life Sciences, Estonia

Received 18 April 2007; received in revised form 15 June 2007; accepted 18 June 2007

KEYWORDSVitiligo;Melanocortin system;TRP1;DCT;mRNA expression

Summary

Background: The melanocortin system in the skin coordinates pigmentation andimmune response and could be implicated in the pathogenesis of vitiligo.Objectives: We aimed to analyze changes in expression of genes involved in skinpigmentation (melanocortin system and enzymes involved in melanin synthesis).Methods: With quantitative RT-PCRwemeasured themRNA expression levels of eightgenes from themelanocortin system and two enzymes involved inmelanogenesis. RNAwas extracted from both lesional and non-lesional skin of vitiligo patients and in non-sun-exposed skin of healthy subjects.Results: POMC (proopiomelanocortin) expression was lower in lesional skin com-pared to non-lesional skin. Expression of melanocortin receptors was increased inunaffected skin of vitiligo patients compared to healthy subjects and decreased inlesional skin compared to uninvolved skin of vitiligo patients, the differences werestatistically significant in the cases of MC1R (melanocortin receptor 1) and MC4R(melanocortin receptor 4). TRP1 and DCT genes were down-regulated in lesional skincompared to non-lesional vitiligo skin or skin of healthy controls and up-regulated inuninvolved vitiligo skin compared to healthy control samples. In non-lesional skin,POMC expression was not elevated, possibly indicating that systemic influences areinvolved in up-regulation of MC receptor genes. Decreased expression of the analyzedgenes in the lesional skin is not surprising, but statistically significant increased

* Corresponding author at: Department of Physiology, University of Tartu, 19 Ravila Street, 50411 Tartu, Estonia. Tel.: +372 7 374 335;fax: +372 7 374 332.

E-mail address: [email protected] (S. Koks).

0923-1811/$30.00 # 2007 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.jdermsci.2007.06.004

114 K. Kingo et al.

expression of studied genes in non-lesional skin from vitiligo patients is not describedpreviously.Conclusion: In our mind, up-regulation of melanocortin system in non-lesional skincould be systemic compensation to restore normal pigmentation in lesions.# 2007 Japanese Society for Investigative Dermatology. Published by Elsevier IrelandLtd. All rights reserved.

Fig. 1 The components of the melanocortin system:POMC, proopiomelanocortin; ASIP, agouti signalling pro-tein; a-MSH, alpha-melanocyte-stimulating hormone; b-MSH, beta-melanocyte-stimulating hormone; g-MSH,gamma-melanocyte-stimulating hormone; ACTH, adreno-corticotropic hormone; AGRP, agouti-related protein;MC1R, melanocortin receptor 1; MC2R, melanocortinreceptor 2; MC3R, melanocortin receptor 3; MC4R, mel-anocortin receptor 4; MC5R, melanocortin receptor 5.

1. Introduction

Vitiligo is an acquired cutaneous disorder that pre-sents with gradual skin depigmentation produced bythe deterioration of melanocyte functions. Severalhypotheses have proposed to explain the dysfunc-tion and/or loss of melanocytes in epidermis ofvitiligo patients [1,2]. These include an autoimmunemechanism, an auto-cytotoxic mechanism and anabnormality in melanocytes or in surrounding kera-tinocytes-producing factors necessary for the survi-val and function of melanocytes [3—5]. Till now thepathogenesis of vitiligo remains partially under-stood and probably involves various combinationsof diverse mechanisms.

The melanocortin system is an evolutionarily con-served regulatorymodule, operating as a coordinatorand executor of responses to stress. The melanocor-tin system consists of the melanocortin peptides a-,b- and g-melanocyte-stimulating hormone (a-MSH,b-MSH and g-MSH) and adrenocorticotropic hormone(ACTH), that are post-translational products of theproopiomelanocortin (POMC) prohormone gene; theendogenous melanocortin receptor antagonists:agouti (ASIP) and agouti-related protein (AGRP);the family of fivemelanocortin receptors (MCRs) thatexert the biological effects of the POMC peptides [6].The components of the melanocortin system andtheir relationships are shown in Fig. 1. Describedsystem is mainly expressed in the brain but also inmany peripheral tissues including the skin.

In the skin, the components of the melanocortinsystem are expressed in different cells. POMC mRNAis detected in the keratinocytes and melanocytes ofnormal epidermis and in the pilosebaceous unit [7].Among POMC-derived peptides a-MSH and ACTH arethe most abundant in the skin [8,9]. MC1R, the moststrongly expressed melanocortin receptor in thehuman skin, has its highest expression inmelanocytes[10]. The most active peptide for MC1R is a-MSHfollowed by ACTH, while b-MSH and g-MSH causeweak activation of this receptor [11]. The expressionof MC2R has been detected in the melanocytes andadipocytes and this receptor is activated by ACTH[11,12]. MC4R, which is activated both by a-MSH andACTH, has been reported to be expressed in dermalpapilla cells [11,13]. TheexpressionofMC5Rhasbeenestablished in sebocytes, adipocytes and skin mast

cells [12—14]. The binding of the POMC peptides tothis receptor is similar to that of MC1R. No manifes-tation of MC3R mRNA in human skin has beenreported.

Functional purpose of the melanocortin system inthe skin is to respond to external and internalstresses through local pigmentation, immune, epi-dermal, adnexal and vascular structures to stabilizeskin function and prevent disruption of internalhomeostasis [15,16]. Components of the melanocor-tin system are involved in determining skin and hairphenotypes as well as in different skin inflammatorydisorders and malignancies. An association betweenthe variability of the POMC gene and red hair phe-notype has been demonstrated [17]. The abnormalexpression of POMC peptides has been established insevere atopic dermatitis, psoriasis, scarring alope-cia and inflammatory keloids as well as in the nod-ular type and metastatic melanomas and basal cellcarcinoma [16]. The MC1R gene variants are linkedwith red hair and poor tanning ability [18—21].Additionally, MC1R has been reported as a genetic

Melanocortin system expression in vitiligo 115

Fig. 2 MC1R, MC2R, MC3R, MC4R and MC5R mRNA levels(relative to housekeeping gene HPRT mRNA level) in theskin from healthy controls (HCS).

risk factor for melanoma, basal cell carcinoma andsquamous cell carcinoma [22,17]. ASIP inhibits thebinding of a-MSH to MC1R, with resulting inhibitionof melanogenesis [23,24]. Polymorphism in the ASIPgene at position 8818 was found to be associatedwith dark hair and brown eyes [25]. The role of AGRPin pigmentation has not been verified. As AGRPblocks a-MSH binding to melanocortin receptors,it is possibly implicated in human pigmentation [26].

The purpose of present study was to examine theexpression variations of genes encoding mediatorsof the melanocortin system in non-lesional andlesional skin of vitiligo patients and in skin of healthycontrols with the aim to explore the regulation ofthe cutaneous stress response system in vitiligo. Inprevious studies, a reduction in the level of thePOMC peptide a-MSH has been demonstrated bothin lesional skin and serum of vitiligo patients[27,28]. Moreover, Graham et al. demonstratedusing a-MSH immunoreactive positive melanocytesthat low expression of a-MSH in the lesional skin ofvitiligo patients resulted from decreased expressionof the peptide rather than a reduction in melano-cyte numbers [29]. The relationship between vitiligoand MC1R and ASIP gene polymorphisms has also

Fig. 3 POMC mRNA levels (relative to housekeepinggene HPRT mRNA level) in skin from healthy controls(HCS), non-lesional vitiligo skin (NLS) and lesional vitiligoskin (LS). Bars indicate mean � S.E.M. *p < 0.05.

been studied, but association between variations ofthese genes and susceptibility to vitiligo has notbeen proved [30]. No studies have analyzed geneexpression levels of the melanocortin system invitiligo samples up to the present time.

2. Results

Gene expression levels of POMC, the five melano-cortin receptors (MC1R—MC5R) and endogenousmelanocortin receptor antagonists (ASIP and AGRP)were measured by quantitative reverse transcrip-tase-polymerase chain reaction (QRT-PCR) in punchbiopsies from lesional and non-lesional skin of viti-ligo patients (n = 31) and from non-sun-exposed skinof healthy subjects (n = 24). In addition, levels oftwo genes encoding enzymes concerned with mel-anogenesis — tyrosinase-related protein-1 (TRP1)and dopachrome tautomerase (DCT) — were mea-sured. The presence of mRNA expression of exam-ined genes in skin of healthy controls and vitiligopatients became evident in QRT-PCR with a rela-tively large number (27—36) of amplification cycles.In the samples, MC1R demonstrated the highestexpression (amplification after 27 cycles), whereasthe levels of MC2R, MC3R, MC4R and MC5R mRNAswere low (amplification after 32—36 cycles) (Fig. 2).POMC mRNAwas detected at 28—32 cycles, ASIP andAGRP at 34—36 cycles. TRP1 and DCT mRNAs weredetected at 28—30 cycles. None of the studied geneshad statistically significant gender-related differ-ences in mRNA expression levels.

A difference in POMC mRNA expression level,although marginally satisfying the standard criteriaof statistical significance ( p < 0.05), was evidentbetween lesional and non-lesional skin of vitiligosubjects (Fig. 3). The POMC mRNA expression was1.9 fold lower in involved skin compared with unin-volved skin in the vitiligo group. POMC expressionwas similar in uninvolved vitiligo skin and in skinbiopsies of healthy controls. No significant differ-ences in ASIP expression as well as AGRP expressionwere established when comparing the study groups(data not shown).

Statistically significant differences in MC1Rexpression and MC4R expression between healthycontrols and vitiligo patients were established. Spe-cifically,MC1RmRNAexpressionwas 1.6 foldhigher innon-lesional skin of vitiligo patients when comparedto healthy subjects (p < 0.01; Fig. 4(a)). Likewise,the MC4R expression level in vitiligo non-lesional skinwas 1.9 fold higher than in healthy skin (p < 0.01;Fig. 4(b)). In lesional skin, 2.1 fold decrease of MC1R(p < 0.0001; Fig. 4(a)) and 2.5 fold decrease of MC4R(p < 0.01; Fig. 4(b)) expressions were detected

116 K. Kingo et al.

Fig. 4 MC1R (a) and MC4R (b) mRNA levels (relative tohousekeeping gene HPRT mRNA level) in skin from healthycontrols (HCS), non-lesional vitiligo skin (NLS) and lesionalvitiligo skin (LS). Bars indicate mean � S.E.M. **p < 0.01compared to healthy control; ++p < 0.01, +++p < 0.0001compared to non-lesional skin.

Fig. 5 TRP1 (a) and DCT (b) mRNA levels (relative tohousekeeping gene HPRT mRNA level) in skin from healthycontrols (HCS), non-lesional vitiligo skin (NLS) and lesionalvitiligo skin (LS). Bars indicate mean � S.E.M. *p < 0.05,***p < 0.0001 compared to healthy control; +++p < 0.0001compared to non-lesional skin.

when compared to non-lesional skin samples frompatients with vitiligo. The mRNA expression levels ofthe three othermelanocortin receptors (MC2R, MC3RandMC5R)were also increased in unaffected skin anddecreased in lesional skin of vitiligo patients; how-ever, the differenceswere not statistically significant(data not shown).

Suppression of TRP1 and DCT expressions inlesional skin compared to non-lesional vitiligo skinand healthy controls was established. Specifically,6.8 fold decrease of TRP1 mRNA expression ininvolved skin compared with skin of healthy subjects( p < 0.0001; Fig. 5(a)) and 19.7 fold decrease ofTRP1 in lesional skin compared to uninvolved skin inthe vitiligo group ( p < 0.0001; Fig. 5(a)) wasdetected. The DCT mRNA expression was 7.6 foldlower in involved skin compared with skin of healthycontrols ( p < 0.0001; Fig. 5(b)) and 12.9 fold lowerin lesional skin compared to uninvolved skin in thevitiligo group (p < 0.0001; Fig. 5(b)). Contrarily, theTRP1 mRNA expression was 2.9 fold higher in non-lesional skin of vitiligo patients compared to healthycontrols (p < 0.05; Fig. 5(a)). The DCT expressionlevel in non-lesional skin of vitiligo patients showed

also a clear tendency of increase when compared tohealthy subjects while not reaching the level ofstatistical significance ( p = 0.14).

No differences in the expressions of genes of themelanocortin system and genes involved in melaninsynthesis were detected between subgroups of viti-ligo based on the extent of involvement and pro-gression of the disease.

Possible interactions between the expressions ofthe studied genes were examined by Spearman rankcorrelation. In addition, the correlations betweenthe expressions of genes of the melanocortin systemand tyrosinase (TYR), essential enzyme in melaninsynthesis, were examined. In our previous study, astatistically significant decrease in TYR mRNAexpression was observed in lesional skin comparedto non-lesional skin of vitiligo patients and in skin ofhealthy subjects (both p < 0.0001) [31]. TYRexpres-sion was elevated in uninvolved vitiligo skin com-pared to skin samples from controls, but thisdifference was not statistically significant. The dataof the correlations between the expressions of thegenes of the melanocortin system in the controlgroup and patient group are presented in Table 1.

Melanocortin system expression in vitiligo 117

MC4R

(HCS/NLS/L

S)MC5R

(HCS/NLS/L

S)

0.57

**/0

.22/

0.27

0.41

/�0.12

/0.26

0.29

/0.05/

0.17

0.21

/�0.35

/0.29

�0.21

/0.11/

0.29

�0.35

/0.28/�0.01

0.01

/0.15/

0.45

*0.13

/�0.21

/0.24

0.60

* /0.54

**/0

.60**

0.41

/0.20/

0.28

0.79

*** /0.79

*** /0.83

***

0.37

/0.26/

0.76

***

**1/

1/1

0.36

/0.51**/0

.62***

0.36

/0.51**/0

.62***

1/1/

1

Variables of the mRNA expression of different com-ponents of the melanocortin system were positivelyrelated. Among the expressions of TYR, TRP1 andDCT mRNAs, a strong positive correlation equally inskin of healthy controls and in skin of vitiligopatients was observed (r > 0.70; p < 0.0001;Fig. 6(a—c)). Further, positive correlation was notedbetween the levels of MC1R and TRP1 in skin ofhealthy controls (r = 0.47; 95% CI 0.001—0.77;p < 0.05; Fig. 7(a)). Contrarily, no statistically sig-nificant correlation between variables of MC1RmRNA and expressions of TYR, TRP1 and DCT mRNAswere observed in non-lesional and lesional vitiligoskin (Fig. 7(b and c)).

Table

1Sp

earman

rankco

rrelationco

efficient(Spearman

r)forexp

ressionva

luesofmelanoco

rtin

system

POMC(H

CS/NLS/LS)

ASIP(H

CS/NLS/L

S)AGRP(H

CS/NLS/L

S)MC1R

(HCS/NLS/L

S)MC2R

(HCS/NLS/L

S)MC3R

(HCS/NLS/L

S)

POMC

1/1

/1

0.32

/0.42* /0.16

0.05

/0.14/�0.03

0.1/

0.07

/�0.15

0.07

/0.31/

0.54

* /0.28

/0.24

ASIP

0.32/0

.42* /0.16

1/1/

1�0.01

/�0.04

/0.12

0.73

*** /0.27

/0.17

0.28

/�0.05

2/0.23

0.11

/0.15/

0.36

AGRP

0.05/0

.14/�0.02

�0.01

/�0.04/

0.12

1/1/1

�0.06

/�0.09

/0.36

�0.27

/0.02/

0.24

�0.29

/0.27/

0.16

MC1R

0.11/0

.07/�0.15

0.73*** /0.27

/0.17

�0.06

/�0.09

/0.36

1/1/

10.43

/0.29/

0.30

0.02

/0.09/

0.17

MC2R

0.07/

0.31

/0.29

0.28

/�0.05/

0.23

�0.27/

0.02

/0.24

0.43

/0.29/

0.30

1/1/

10.60

* /0.63

*** /0.44

*

MC3R

0.54* /0.28/

0.24

0.11/

0.15

/0.36

�0.29

/0.27/

0.16

0.02

/0.09/

0.23

0.60

* /0.63

*** /0.44

*1/

1/1

MC4R

0.57

**/0.22/

0.27

0.29

/0.05/

0.17

�0.21/

0.11

/0.29

0.01

/0.15/

0.45

*0.60

* /0.54

**/0

.60**

0.79

*** /0.79

*** /0.82

*

MC5R

0.41

/�0.12

/0.26

0.21

/�0.35

/0.29

�0.35/

0.28

/�0.01

0.13

/�0.21

/0.24

0.41

/0.20/

0.28

0.37

/0.26/

0.76

***

*p<

0.05.

**p<

0.01.

***p<

0.001

.

3. Discussion

Thehypothalamic—pituitary—adrenalaxis(HPA)inthebrain is themainmediator of the systemic response tostress.TheskinexpressesanequivalentoftheHPAaxis:acutaneousdefencesystemthatoperatesasacoordi-nator and executor of local responses to stress. Themelanocortin system in the skin is a part of the cuta-neousHPAaxis [12]. In thepresent study,weanalyzedexpression ofmelanocortin system genes in skin sam-ples from patients with vitiligo. Our study samplecontains 25 patients with generalized vitiligo, 1 withsegmental vitiligo, 4 with local form and 1 with uni-versal form. As segmental vitiligo and generalizedvitiligomayhavedifferentpathogenesis,wecomparedmeanexpressionsofstudiedgenesindifferentformsofvitiligo. There was no significant difference in theexpression pattern and we decided to merge theexpressionresultsofdifferent forms (segmental,gen-eralized, localized and universal) of vitiligo into onestudy group.

Genes of themelanocortin system analyzed in thehuman skin in the present study were proopiome-lanocortin (POMC), five melanocortin receptors(MC1R—MC5R), agouti signalling protein (ASIP) andagouti-related protein (AGRP). Our results reaf-firmed the expression of cutaneous stress responsesystem in human skin, whereas expression of MC3RmRNA in human skin has not previously beenreported. Further studies using histochemical meth-ods must be conducted to localize MC3R expressionto distinct cell types in the skin. The expressionprofiles of all five melanocortin receptors wereanalogous between the study groups, a markedlysignificant difference between lesional and non-lesional vitiligo skin ( p < 0.0001) was establishedonly in the case of MC1R. Probably, the expressionlevels of the genes indicate their functional rele-vance. In our samples, MC1R demonstrated the high-est expression, while the rest of the melanocortin

118 K. Kingo et al.

Fig. 6 (a) The relation of the levels of TYR (TYR_HCS), TRP1 (TRP1_HCS) and DCT (DCT_HCS) in skin of healthy subjects.(b) The relation of the levels of TYR (TYR_NLS), TRP1 (TRP1_NLS) and DCT (DCT_NLS) in non-lesional vitiligo skin. (c) Therelation of the levels of TYR (TYR_LS), TRP1 (TRP1_LS) and DCT (DCT_LS) in lesional vitiligo skin.

receptors had relatively low expression levels.Moreover, the correlation analysis presented a mod-erate up to strong positive relation between thelevels of mRNA of the different melanocortin recep-tors. High homology in sequences may be the causeof the observed similar expression profile of themelanocortin receptors [11]. Alternatively, this kindof expression profile of the melanocortin receptorscould be explained by concomitant expression ofthese genes. Concomitant expression of MC1R andMC2R has been reported in human keratinocytes andepidermal melanocytes [32—34]. Concomitantexpression of MC1R and MC4R has been observedin human dermal papilla cells [35].

In the skin, the potential role of the melanocortinsystem in the network of cutaneous inflammationand pigmentation has been verified. The possiblerole of markers of the melanocortin system in thepathogenesis of vitiligo remains unclear. In the pre-sent study, POMC mRNA expression was found to belower in lesional skin compared with non-lesional

skin in the vitiligo group ( p < 0.05). At the sametime, we did not detect difference in the expressionof POMC between non-lesional skin from vitiligopatients and skin from healthy controls. The mRNAexpression levels of the melanocortin receptors,mediating the effects of POMC peptides, weredecreased in lesional skin compared to uninvolvedskin and increased in unaffected skin of vitiligopatients compared to healthy subjects, the differ-ences were statistically significant (p < 0.01) in thecases of MC1R and MC4R. Decreased expression ofMC1R in lesional skin is not surprising, because thisfits the loss of functional melanocytes–—major celltype expressing MC1R. Contrarily, MC4R is expressedonly in dermal papilla cells. Similar down-regulationof MC1R and MC4R in lesional skin may occur due tohigh homology in their sequences, which may pro-vide similar regulatory regions employed in lesionalskin. Increased expression of MC receptors in non-lesional skin of patients is not so well understood.This is not caused by natural UVR exposure, because

Melanocortin system expression in vitiligo 119

Fig. 7 (a) The relation of the levels of MC1R (MC1R_HCS) and TYR (TYR_HCS), TRP1 (TRP1_HCS) and DCT (DCT_HCS) inskin of healthy controls. (b) The relation of the levels of MC1R (MC1R_NLS) and TYR (TYR_NLS), TRP1 (TRP1_NLS) and DCT(DCT_NLS) in non-lesional vitiligo skin. (c) The relation of the levels of MC1R (MC1R_LS) and TYR (TYR_LS), TRP1(TRP1_LS) and DCT (DCT_LS) in lesional vitiligo skin.

skin biopsies were taken from truly non-exposedareas. Also, distribution of skin phototypes betweenstudy groups was balanced. Over-expression of MCreceptors possibly indicates the existence of com-pensation system to restore normal pigmentation inlesions. Lack of difference in case of POMC in thenon-lesional and control skin samples suggest thatthis over-expression is induced by systemic, circu-lating influences rather than local ones.

Tyrosinase (TYR) and tyrosinase-related proteins(TRP1 and DCT) are closely related melanocyte-specific gene products involved in melanogenesisand in the biology of melanocytes. Down-regulationof TYR mRNA expression has been documented inlesional skin of vitiligo patients [36]. Accordingly, inour previous study, a statistically significantdecrease in TYR expression was observed in lesionalskin compared to non-lesional skin of vitiligopatients and healthy control subjects (bothp < 0.0001) [31]. In addition, significantly lowerlevel of TRP1 mRNA expression by vitiligo melano-cytes compared to control normal melanocytes isestablished [37]. No expression of DCT mRNA has

been investigated in vitiligo up to the present time.In the present study, suppression of expressions ofTRP1 and DCT mRNAs, as verification of decreasedmelanin synthesis in vitiligo, was established inlesional skin of vitiligo patients. Current literatureindeed suggests that the lesional area is chara-cterized by the absence of melanocytes or by thepresence of ineffective melanocytes. However, notall melanocytes are destroyed in lesions and thisexplains why these mRNAs are still in detectablelevel.

We found expressional up-regulation of genesinvolved in melanin synthesis in uninvolved skinsamples of vitiligo patients compared to healthycontrol samples. In case of TRP1, this differencewas statistically significant. As we also found over-expression of melanocortin receptors in the samesamples, activation of enzyme genes in non-lesionalskin is a consequence of melanocortin stimulation.This indicates functional relevance of our finding. Itis possible that the suppression of melanin produc-tion activates melanocortin receptors transcription-ally through negative feedback to restore normal

120 K. Kingo et al.

pigmentation. On the other hand, melanocortinshave the widest spectrum of immunomodulatory/anti-inflammatory capacities [38—41]. Increasedexpressions of pro-inflammatory cytokines (IL-2,TNF-a, IL-6 and INF-g) in lesional skin comparedto uninvolved skin of vitiligo patients have beenreported [42,43]. Up-regulation of the melanocortinreceptors in non-lesional skin, established in thepresent study, may attempt to inhibit the produc-tion or action of proinflammatory factors andthereby make an effort to suppress inflammatoryresponses in lesional skin of vitiligo patients. Posi-tive correlation between the levels of MC1R andTRP1 was found in the skin of healthy controls, thiscorrelation did not exist in patients. Described find-ing indicates that significant changes in transcrip-tional regulation of the melanocortin and melaninsynthesis system occur in skin during vitiligo.

In conclusion, the expression of genes of themelanocortin system is altered in vitiligo. Decreasedexpression of the genes of the melanocortin systemand enzymes of melanogenesis in the lesional skindemonstrated in the present study is not surprisingand fits with existing data. On the other hand,statistically significant increased expression innon-lesional skin from vitiligo patients is notdescribed previously. In our mind, this up-regulationcould be systemic compensation to restore normalpigmentation in lesions.

4. Materials and methods

The study protocols and informed consent formswere approved by the Ethical Review Committeeon Human Research of the University of Tartu. Allparticipants signed a written informed consent.

The patients and control subjects in the studywere Caucasians living in Estonia. Unrelated patientswith vitiligo (n = 31; 22 females; 9 males; age range22—75 years) from the Department of Dermatology,the University of Tartu, were included in the study.The mean age of vitiligo onset of the patients was30.0 years and the mean duration of vitiligo was 19.2years. Five patients had a family history of vitiligo.None of the patients included in the study hadreceived specific therapy in the previous 6 months.The clinical signs on which the diagnosis of vitiligowas based onwere characteristic loss of skin pigmen-tation with typical localization and white colour onthe skin lesions under Woods lamp. The type ofvitiligo was based on the extent of involvementand the distribution of pigmentation. Focal vitiligoinvolves depigmentation in a localized, non-derma-tomal distribution (n = 4(F)). Segmental vitiligoencompasses depigmentation of the dermatomal,

asymmetric distribution (n = 1(F)). Generalized viti-ligo is characterized by bilateral, symmetric loss ofpigmentation of the torso, face, neck, or extensorsurfaces of the hands and legs (n = 25; 17(F); 8(M)).Universal vitiligo occurs as depigmentation of theentire body surface area (n = 1(M)). The stage ofvitiligo was based on the interval of manifestationof new areas of depigmentation or enlargement ofthe area of depigmentation. The patients weredivided into two subgroups based on the stage ofprogression of the disorder: patientswith progressivevitiligo (active vitiligo, in which new areas of depig-mentation or enlargement of depigmentation wereobserved during the previous 3months; n = 22; 16(F);6(M)) and patients with stable vitiligo (inactive viti-ligo, inwhich no newdepigmentation or enlargementofdepigmentationhadbeenobserved formore than3months; n = 9; 6(F); 3(M)). The control group con-sisted of healthy volunteers (n = 24; 17 female; 7male; age range 21—67 years) free from the positivefamily history of vitiligo and other chronic derma-toses. Control subjects were recruited from healthcare personnel, medical students and patients pre-sent at the dermatological outpatient clinic witheither facial teleangiectasis or skin tags.

Two skin biopsies (Ø 4 mm) were obtained fromeach patient with vitiligo: one from the central partof involved skin and another from non-sun-exposeduninvolved skin. One skin biopsy (Ø 4 mm) from non-sun-exposed skin was taken from healthy controlsubjects. The non-sun-exposed skin was defined asthe skin never exposed to UVR previously and defi-nitely not exposed to natural UVR in the last 12months. Biopsies from uninvolved skin of vitiligopatients and healthy controls were taken from thelower abdomen. All probands had skin phototype II(9 controls, 13 patients) or III (15 controls, 18patients), Fitzpatrick classification [44]. Biopsieswere instantaneously snap-frozen in liquid nitrogenand stored at �80 8C until further use.

Total RNA was isolated from tissues using RNeasyFibrous Tissue Mini Kit (QIAGEN Sciences, MD, USA)following the protocol suggested by the manufac-turer. For tissue homogenization, Ultra-Turrax T8homogenizer (IKA Labortechnik, Germany) wasused. RNA was dissolved in RNase free water andstored until further use at �80 8C. For each RT-PCRreaction, 500 ng of total RNA was converted intocDNA. The reverse transcription reactions wereperformed with a reverse transcriptase (SuperScriptIII; Invitrogen Corp., Carlsbad, CA, USA) andpoly(T18) oligonucleotides in accordance with themanufacturer’s instruction. The reaction mixtureswere incubated at 65 8C for 5 min, at 0 8C for 1 min,at 50 8C for 90 min, at 70 8C for 5 min and finallystored at �80 8C.

Melanocortin system expression in vitiligo 121

Gene expression levels were detected in the ABIPrism 7900HT Sequence Detection System (AppliedBiosystems, Foster City, CA, USA). Reactions werecarried out in 10 ml reaction volumes in four repli-cates.

The expression levels of AGRP, ASIP, the melano-cortin receptor genes and TRP1 and DCT weredetected applying TaqMan-QRT-PCR method usingTaqMan Universal PCR Master Mix (Applied Biosys-tems, Foster City, CA, USA). For the detection ofthe expression levels of AGRP, ASIP, MC3R, MC4R,MC5R, TRP1 and DCT, we used TaqMan Assay-On-Demand FAM-labelled MGB-probe gene expressionassay mixes (20X, Applied Biosystems, Foster City,CA, USA). The assay mixes used were Hs00361403_g1(AGRP), Hs00181770_m1 (ASIP), Hs00252036_s1(MC3R), Hs00271877_s1 (MC4R), Hs00271882_s1(MC5R), Hs00167051_m1 (TRP1) and Hs00157244_m1 (DCT). For the detection of the expression levelsof MC1R and MC2R, we used gene-specific primers(MC1R: forward 50-TGCGGCTGCATCTTCAAG-30, reve-rse 50-TGATGGCATTGCAGATGATGA-30; MC2R: for-ward 50-CTCGATCCCACACCAGGAA-30, reverse 50-TGTGATGGCCCCTTTCATGT-30) and MGB-labelledprobe (MC1R: FAM-TTCAACCTCTTTCTCGCC-NFQ;MC2R: FAM-TCTCCACCCTCCCCAGA-NFQ). For thedetection of HPRT-1 (hypoxanthine phosphoribosyl-transferase-1) expression level, gene-specific pri-mers (HPRT-1 exon 6, 50-GACTTTGCTTTCCTTGGT-CAGG-30; HPRT-1 exon 7, 50-AGTCTGGCTTATA-TCCAACACTTCG-30; final concentrations 300 nM)and VIC-TAMRA-labelled probe (VIC-50-TTTCACCAG-CAAGCTTGCGACCTTGA-30-TAMRA; final concentra-tion 200 nM) were used.

The expression level of POMC was detected usingqPCR Core Kit for SYBR Green I (Eurogentec, Seraing,Belgium) and gene-specific primers (forward 50-CTACGGCGGTTTCATGACCT-30, reverse 50-CCCTCAC-TCGCCCTTCTTG-30, final concentrations 100 nM).

For quantification of mRNA, we used comparativeCt method (DCt value), where the amount of targettranscript was normalized according to the level ofendogenous reference HPRT-1 (hypoxanthine phos-phoribosyl-transferase-1). Adjustment to normaldistribution was tested by the Kolmogorov—Smirnovtest. The distribution of measurements of geneexpressions by the applied method did not followa Gaussian distribution. Mann—Whitney U-test andKruskal—Wallis test were used to test for differencesbetween the groups using GraphPad Prism 4 soft-ware (GraphPad Software, San Diego, CA, USA).Correlation analysis was used to investigate rela-tions between two parameters of one group. Formeasure of correlation, the Spearman rank correla-tion was applied. For all tests, a p value < 0.05 wasconsidered significant.

Acknowledgements

This study was financially supported by the target-based funding from the Estonian Ministry of Educa-tion Grant No. SF0180043s07, by University of TartuResearch Grant PARFS 05901, by the EstonianScience Foundation Grants No. 6576, 5712 and5688 and by the Centre of Molecular and ClinicalMedicine Grant VARMC-TIPP.

References

[1] Tobin DJ, Swanson NN, PittelkowMR, Peters EM, SchallreuterKU. Melanocytes are not absent in lesional skin of longduration vitiligo. J Pathol 2000;191:407—16.

[2] Le Poole IC, Das PK, van den Wijngaard RM, Bos JD, Wester-hof W. Review of the etiopathomechanism of vitiligo: aconvergence theory. Exp Dermatol 1993;2:145—53.

[3] Ongenae K, Van Geel N, Naeyaert JM. Evidence for an auto-immune pathogenesis of vitiligo. Pigm Cell Res 2003;16:90—100.

[4] Schallreuter KU, Moore J, Wood JM, Beazley WD, Gaze DC,Tobin DJ, et al. In vivo and in vitro evidence for hydrogenperoxide (H2O2) accumulation in the epidermis of patientswith vitiligo and its successful removal by a UVB-activatedpseudocatalase. J Investig Dermatol Symp Proc 1999;4:91—6.

[5] Moretti S, Spallanzani A, Amato L, Hautmann G, Gallerani I,Fabiani M, et al. New insights into the pathogenesis ofvitiligo: imbalance of epidermal cytokines at sites of lesions.Pigm Cell Res 2002;15:87—92.

[6] Gantz I, Fong TM. The melanocortin system. Am J PhysiolEndocrinol Metab 2003;284:E468—74.

[7] KonoM, Nagata H, Umemura S, Kawana S, Osamura RY. In situexpression of corticotropin-releasing hormone (CRH) andproopiomelanocortin (POMC) genes in human skin. FASEB J2001;15:2297—9.

[8] Wakamatsu K, Graham A, Cook D, Thody AJ. Characterisa-tion of ACTH peptides in human skin and their activation ofthe melanocortin-1 receptor. Pigm Cell Res 1997;10:288—97.

[9] Tsatmali M, Yukitake J, Thody AJ. ACTH1-17 is a more potentagonist at the human MC1 receptor than alpha-MSH. Cell MolBiol (Noisy-le-grand) 1999;45:1029—34.

[10] Roberts DW, Newton RA, Beaumont KA, Helen Leonard J,Sturm RA. Quantitative analysis of MC1R gene expression inhuman skin cell cultures. Pigm Cell Res 2006;19:76—89.

[11] Getting SJ. Targeting melanocortin receptors as potentialnovel therapeutics. Pharmacol Ther 2006;111:1—15.

[12] Slominski A, Wortsman J, Luger T, Paus R, Solomon S.Corticotropin releasing hormone and proopiomelanocortininvolvement in the cutaneous response to stress. Physiol Rev2000;80:979—1020.

[13] Bohm M, Schiller M, Luger TA. Non-pigmentary actions ofalpha-melanocyte-stimulating hormone–—lessons from thecutaneous melanocortin system. Cell Mol Biol (Noisy-le-grand) 2006;52:61—8.

[14] Zhang L, Li WH, Anthonavage M, Eisinger M. Melanocortin-5receptor: a marker of human sebocyte differentiation. Pep-tides 2006;27:413—20.

[15] Tsatmali M, Ancans J, Thody AJ. Melanocyte function and itscontrol by melanocortin peptides. J Histochem Cytochem2002;50:125—33.

122 K. Kingo et al.

[16] Slominski A, Wortsman J, Mazurkiewicz JE, Matsuoka L,Dietrich J, Lawrence K, et al. Detection of proopiomelano-cortin-derived antigens in normal and pathologic humanskin. J Lab Clin Med 1993;122:658—66.

[17] Bastiaens MT, ter Huurne JA, Kielich C, Gruis NA,WestendorpRG, Vermeer BJ, et al. Melanocortin-1 receptor gene var-iants determine the risk of nonmelanoma skin cancer inde-pendently of fair skin and red hair. Am J Hum Genet2001;68:884—94.

[18] Valverde P, Healy E, Jackson I, Rees JL, Thody AJ. Variants ofthe melanocyte-stimulating hormone receptor gene areassociated with red hair and fair skin in humans. Nat Genet1995;11:328—30.

[19] Krude H, Biebermann H, Luck W, Horn R, Brabant G, GrutersA. Severe early-onset obesity, adrenal insufficiency and redhair pigmentation caused by POMCmutations in humans. NatGenet 1998;19:155—7.

[20] Healy E, Flannagan N, Ray A, Todd C, Jackson IJ, MatthewsJN, et al. Melanocortin-1-receptor gene and sun sensitivityin individuals without red hair. Lancet 2000;355:1072—3.

[21] Flanagan N, Healy E, Ray A, Philips S, Todd C, Jackson IJ,et al. Pleiotropic effects of the melanocortin 1 receptor(MC1R) gene on human pigmentation. Hum Mol Genet2000;9:2531—7.

[22] Kennedy C, ter Huurne J, Berkhout M, Gruis N, Bastiaens M,Bergman W, et al. Melanocortin 1 receptor (MC1R) genevariants are associated with an increased risk for cutaneousmelanoma which is largely independent of skin type and haircolor. J Invest Dermatol 2001;117:294—300.

[23] Suzuki I, Tada A, Ollmann MM, Barsh GS, Im S, Lamoreux ML,et al. Agouti signaling protein inhibits melanogenesis andthe response of human melanocytes to alpha-melanotropin.J Invest Dermatol 1997;108:838—42.

[24] Voisey J, Kelly G, Van Daal A. Agouti signal protein regulationin human melanoma cells. Pigm Cell Res 2003;16:65—71.

[25] Kanetsky PA, Swoyer J, Panossian S, Holmes R, Guerry D,Rebbeck TR. A polymorphism in the agouti signaling proteingene is associated with human pigmentation. Am J HumGenet 2002;70:770—5.

[26] Yang YK, Thompson DA, Dickinson CJ, Wilken J, Barsh GS,Kent SB, et al. Characterization of agouti-related proteinbinding to melanocortin receptors. Mol Endocrinol 1999;13:148—55.

[27] Thody AJ, Ridley K, Penny RJ, Chalmers R, Fisher C, ShusterS. MSH peptides are present in mammalian skin. Peptides1983;4:813—6.

[28] Pichler R, Sfetsos K, Badics B, Gutenbrunner S, Aubock J.Vitiligo patients present lower plasma levels of alpha-mel-anotropin immunoreactivities. Neuropeptides 2006;40:177—83.

[29] Graham A, Westerhof W, Thody AJ. The expression of alpha-MSH by melanocytes is reduced in vitiligo. Ann NY Acad Sci1999;885:470—3.

[30] Na GY, Lee KH, Kim MK, Lee SJ, Kim DW, Kim JC. Polymorph-isms in the melanocortin-1 receptor (MC1R) and agoutisignaling protein (ASIP) genes in Korean vitiligo patients.Pigm Cell Res 2003;16:383—7.

[31] Kingo K, Philips MA, Aunin E, Luuk H, Karelson M, Ratsep R,et al. MYG1, novel melanocyte-related gene, has elevatedexpression in vitiligo. J Dermatol Sci 2006;44:119—22.

[32] Curry JL, Pinto W, Nickoloff BJ, Slominski AT. Human kera-tinocytes express functional alpha-MSH (MC1-R) receptors.In Vitro Cell Dev Biol Anim 2001;37:234—6.

[33] Moustafa M, Szabo M, Ghanem GE, Morandini R, Kemp EH,MacNeil S, et al. Inhibition of tumor necrosis factor-alphastimulated NFkappaB/p65 in human keratinocytes by alpha-melanocyte stimulating hormone and adrenocorticotropichormone peptides. J Invest Dermatol 2002;119:1244—53.

[34] Suzuki I, Cone RD, Im S, Nordlund J, Abdel-Malek ZA. Bindingof melanotropic hormones to the melanocortin receptorMC1R on human melanocytes stimulates proliferation andmelanogenesis. Endocrinology 1996;137:1627—33.

[35] Bohm M, Luger TA, Tobin DJ, Garcia-Borron JC. Melanocortinreceptor ligands: new horizons for skin biology and clinicaldermatology. J Invest Dermatol 2006;126:1966—75.

[36] Machado Filho CD, Almeida FA, Proto RS, Landman G. Viti-ligo: analysis of grafting versus curettage alone, using mel-anocyte morphology and reverse transcriptase polymerasechain reaction for tyrosinase mRNA. Sao Paulo Med J2005;123:187—91.

[37] Jimbow K, Chen H, Park JS, Thomas PD. Increased sensitivityof melanocytes to oxidative stress and abnormal expressionof tyrosinase-related protein in vitiligo. Br J Dermatol2001;144:55—65.

[38] Luger TA, Scholzen TE, Brzoska T, Bohm M. New insights intothe functions of alpha-MSH and related peptides in theimmune system. Ann NY Acad Sci 2003;994:133—40.

[39] Wintzen M, Gilchrest BA. Proopiomelanocortin, its derivedpeptides, and the skin. J Invest Dermatol 1996;106:3—10.

[40] Teofoli P, Motoki K, Lotti TM, Uitto J, Mauviel A. Propiome-lanocortin (POMC) gene expression by normal skin and keloidfibroblasts in culture: modulation by cytokines. Exp Derma-tol 1997;6:111—5.

[41] Wintzen M, Yaar M, Burbach JP, Gilchrest BA. Proopiomela-nocortin gene product regulation in keratinocytes. J InvestDermatol 1996;106:673—8.

[42] Caixia T, Hongwen F, Xiran L. Levels of soluble interleukin-2receptor in the sera and skin tissue fluids of patients withvitiligo. J Dermatol Sci 1999;21:59—62.

[43] Grimes PE, Morris R, Avaniss-Aghajani E, Soriano T, Meraz M,Metzger A. Topical tacrolimus therapy for vitiligo: thera-peutic responses and skin messenger RNA expression ofproinflammatory cytokines. J Am Acad Dermatol 2004;51:52—61.

[44] Fitzpatrick TB, Miyamoto M, Ishikawa K. The evolution ofconcepts of melanin biology. Arch Dermatol 1967;96:305—23.