vulgaris monoclonal autoantibodies isolatedby phage display. J Clin Invest 115:888–99

Posthaus H, Dubois CM, Muller E (2003) Novelinsights into cadherin processing by subtili-sin-like convertases. FEBS Lett 536:203–8

Sekiguchi M, Futei Y, Fujii Y, Iwasaki T, Nishika-wa T, Amagai M (2001) Dominant auto-

immune epitopes recognized by pemphigusantibodies map to the N-terminal adhesiveregion of desmogleins. J Immunol 167:5439–48

Stanley JR, Amagai M (2006) Pemphigus, bullousimpetigo, and the staphylococcal scalded-skin syndrome. New Engl J Med 355:1800–10

Yokouchi M, Adly M, Kuroda K, Hachiya T,Stanley JR, Amagai M et al. (2009)Pathogenic epitopes of autoantibodies inpemphigus reside in the amino-terminaladhesive region of desmogleins whichare unmasked by proteolytic processingof prosequence. J Invest Dermatol 129:2156–66

See related commentary on pg 2096

Non-Neuronal Expression of Transient Receptor PotentialType A1 (TRPA1) in Human SkinJournal of Investigative Dermatology (2009) 129, 2312–2315; doi:10.1038/jid.2009.58; published online 12 March 2009

TO THE EDITORThe temperature-sensitive channels,which belong to the transient receptorpotential (TRP) superfamily, play animportant role in skin biology. Inaddition to being expressed in sensoryneurons, several members of this fa-mily, TRPV1, TRPV3, and TRPV4,which are activated by warm to hottemperatures (442, 434–38, and427–34 1C, respectively), are broadlyexpressed in non-neuronal cells of theskin and are involved in the control ofkeratinocyte differentiation, inflamma-tory skin responses, and hair growth(reviewed in Bıro et al., 2007).

TRPA1 is a distant family member ofthe TRP superfamily channels, which islocalized in a subset of nociceptivesensory neurons, and showed a re-sponse to cold temperature startingnearly at 17 1C, the threshold of noxiouscold for humans (Story et al., 2003).TRPA1 can also be activated by thenumber of pungent natural compounds,environmental irritants, and formalin,as well as by endogenous proalgesicagents (Bandell et al., 2004; McNamaraet al., 2007; Trevisani et al., 2007). Inaddition, TRPA1 is capable of mediat-ing acute and inflammatory pain, atleast in part, through crosstalk with thesignaling pathway induced by theproinflammatory peptide, bradykinin(Bautista et al., 2006). However, theexpression and functions of TRPA1 in

non-neuronal cells in skin remain as yetunknown.

The aim of this study was to explorea role of TRPA1 in skin biology bystudying its expression in distinct cuta-neous cell populations (keratinocytes,fibroblasts, and melanocytes) as well asby assessing whether pharmacologicalactivation of TRPA1 would have effectson gene expression programs in epider-mal keratinocytes. Human scalp skinsamples were obtained from five pa-tients after face-lift surgery, with writtenconsents approved by the InstitutionalReview Board to ensure subject protec-tion and adherence according to theDeclaration of Helsinki Principles. Byreal-time PCR analysis (see Supplemen-tary Material), the TRPA1 mRNA ex-pression was observed in primarycultures of human epidermal keratino-cytes, melanocytes, and fibroblasts(Figure 1a). Relative quantification re-vealed that TRPA1 mRNA levels werehigher in melanocytes than that infibroblasts and keratinocytes. By wes-tern blot analysis, the TRPA1 proteinexpression was also seen in all celltypes examined, and its expressionlevels in fibroblasts were relativelyhigher than those in melanocytes andkeratinocytes (Figure 1b). To determinelocalization of TRPA1 in skin, weperformed immunofluorescence analy-sis. TRPA1 immunoreactivity was de-tected in the basal layer of the

epidermis, in the dermis, and in theepithelium of the hair follicle. Bydouble immunofluorescence, we ob-served colocalization of TRPA1 withthe melanocyte marker, pMel-17, in thedistinct cells of the basal layer of theepidermis, suggesting that TRPA1 isexpressed in the keratinocytes as wellas in the melanocytes (Figure 1c and d;for details, see also SupplementaryText).

Epidermal keratinocytes serve as firstline of defense that protects organismfrom environmental stressors, includingcold temperature and chemical irri-tants, which are capable of activatingTRPA1. To explore a possible func-tional role for TRPA1 in the epidermis,primary normal human epidermal ker-atinocytes were treated with the phar-macological TRPA1 agonist, icilin(10 mM; 24 hours) (Werkheiser et al.,2006; Doerner et al., 2007). Compara-tive analysis of global gene expressionprofiles in keratinocytes treated withicilin and vehicle control was per-formed using Agilent microarray tech-nology (Santa Clara, CA) and real-timePCR (Supplementary Text).

Microarray analysis of the icilin-treated and control keratinocytesshowed two-fold or higher changes inthe expression of 241 genes encodingthe adhesion/extracellular matrix mole-cules, in cell cycle/apoptosis andcytoskeleton/cell motility markers, andin molecules involved in the control ofcell differentiation, metabolism, signal-ing, and transcription (Figure 2a and b,

Abbreviations: GDF15, growth differentiation factor 15; HSP, heat shock protein; TRP, transient receptorpotential

2312 Journal of Investigative Dermatology (2009), Volume 129

R Atoyan et al.TRPA1 Expression in Human Skin

Table S1). Genes for further validationby quantitative real-time PCR wereselected on the basis of their changesin expression levels compared with thecontrols, as well as on the basis of theirfunctional significance for the epider-mal structure and functions (Table S2).

The most prominent differenceswere seen in the expression of genesinvolved in the control of keratinocyteproliferation and differentiation, includ-ing the selected members of the trans-forming growth factor-b superfamily.Specifically, we detected about four-to five-fold increase in the expression ofBMP7 and growth differentiation factor15 (GDF15) mRNA in icilin-treatedkeratinocytes compared with the con-trol (Figure 2c). The role of BMP7 andGDF15 in the keratinocyte biology wasnot entirely clarified yet; however, itwas recently suggested that GDF15can be involved in keratinocyte differ-entiation. In particular, knockdown ofthe endogenous GDF15 by siRNAmediated cell growth and inhibitedthe expression of the epidermal differ-entiation markers (Ichikawa et al.,2007).

Furthermore, microarray and real-time PCR analyses of icilin-treatedkeratinocytes revealed marked changesin the expression of genes encoding the

selected members of the heat shockprotein (HSP) family. The HSP27mRNA expression was substantiallyincreased, whereas the expression ofHSP90 mRNA was decreased in kerati-nocytes treated with 10 uM icilin, com-pared with the control (Figure 2d). It isknown that HSP27 expressing in thesuprabasal layers in the human epider-mis is involved in the regulation ofdifferentiation-associated genes actingas a chaperone of cornification (Trau-tinger et al., 1995; Hell-Pourmojibet al., 2002; Jonak et al., 2002). Similarto HSP27, HSP90 is also constitutivelyexpressed in human keratinocytes (Ed-wards et al., 1991) However, HSP90 isgenerally increased in proliferatingcells and is implicated in the controlof cell cycle (reviewed in Helmbrechtet al., 2000). Therefore, alterations inthe expression levels of selected HSPsin response to icilin exposure furthersuggest that icilin treatment couldmodulate proliferation/differentiationprogram in the keratinocytes.

Together with changes in the expres-sion of genes that encode regulatoryfactors stimulating keratinocyte differ-entiation, the expression of severalcyclins and cyclin-dependent kinasesinvolved in the control of cell cycledownregulated, whereas the expression

of cyclin-dependent kinase inhibitorswere upregulated in icilin-treatedcells, compared with the control. Forinstance, the expression of cdc2 tran-script decreased approximately 2.5-foldin icilin-treated keratinocytes comparedwith the vehicle control. In contrast, theexpression of cyclin-dependent kinaseinhibitor p21 increased approximatelyfour-fold in the keratinocytes culturedwith icilin than that in the control(Figure 2e).

As TRPA1 is implicated in mediationof inflammation induced by variousirritants and endogenous proalgesicagents, we have also explored whetherthe activation of TRPA1 in the epider-mal keratinocytes would affect theexpression of proinflammatory cyto-kines, including IL-1a and IL-1b, whichare known to be key contributors toskin inflammation. Treatment of epider-mal keratinocytes with 10 mM icilinresulted in the increase in the mRNAexpression of both IL-1 isoforms, com-pared with that of the control (Figure2f). Furthermore, IL-1a protein levelswere elevated in icilin-treated keratino-cytes compared with those in untreatedcells as determined by western blotanalysis (Figure 2g). These findingssuggest that TRPA1 might directly beinvolved in the promotion of cutaneous

Melanocytes Keratinocytes Fibroblasts

Mel

anoc

ytes

Ker

atin

ocyt

es

Fib

robl

asts

1,000

100

10

1

0.1Rel

ativ

e m

RN

A e

xpre

ssio

n

TRPA1

β-Actin

Figure 1. TRPA1 expression in distinct cell populations of human skin. Expressions of the TRPA1 mRNA and protein in primary cell cultures of epidermal

keratinocytes, dermal fibroblasts, and melanocytes isolated from the human skin. (a) Real-time PCR analysis of the TRPA1 expression. (b) Western blot analysis

of 75 kDa TRPA1 protein. (c) TRPA1 immunofluorescence in the human scalp skin sample: epidermis (arrowhead), dermis (arrow), and hair follicle (asterisk). (d)

Double immunofluorescence of TRPA1 (red color) and gp100/pMel-17 (green color); co-localization is shown by arrows. Bars¼100 mm (c) and 50 mm (d).

www.jidonline.org 2313

R Atoyan et al.TRPA1 Expression in Human Skin

inflammation by stimulating the expres-sion of keratinocyte-derived cytokines.

In summary, we provide the firstevidence that TRPA1 is broadly ex-pressed in the skin and may directly beinvolved in the regulation of keratino-cyte differentiation as well as of inflam-matory responses. In addition, thesedata suggest that epidermal keratino-cytes might act in concert with sensoryneurons to perceive the thermal envir-onment and to activate adaptive skinresponses to environmental stressors.

CONFLICT OF INTERESTThe authors state no conflict of interest.

Ruzanna Atoyan1, Doug Shander2 andNatalia V. Botchkareva3

1Department of Dermatology, BostonUniversity, Boston, Massachusetts, USA;2Trichoresearch, Gaitherburg, Maryland, USA

and 3Centre for Skin Sciences, School of LifeSciences, University of Bradford, Bradford, UKE-mail: n.botchkareva@bradford.ac.uk

SUPPLEMENTARY MATERIAL

Supplementary Materials and Methods

Table S1. Genes differentially expressed in theicilin-treated keratinocytes versus control.

Table S2. List of PCR primers.

REFERENCES

Bandell M, Story GM, Hwang SW, Viswanath V,Eid SR, Petrus MJ et al. (2004) Noxious coldion channel TRPA1 is activated by pungentcompounds and bradykinin. Neuron 41:849–57

Bautista DM, Jordt SE, Nikai T, Tsuruda PR,Read AJ, Poblete J et al. (2006) TRPA1mediates the inflammatory actions of envir-onmental irritants and proalgesic agents. Cell124:1269–82

Bıro T, Toth BI, Marincsak R, Dobrosi N, Geczy T,Paus R (2007) TRP channels as novel playersin the pathogenesis and therapy of itch.Biochim Biophys Acta 1772:1004–21

Doerner J, Gisselmann G, Hatt H, Wetzel CH(2007) Transient receptor potential channelA1 is directly gated by calcium ions. J BiolChem 282:13180–9

Edwards MJ, Marks R, Dykes PJ, Merrett VR, MorganHE, O’Donovan MR (1991) Heat shockproteins in cultured human keratinocytes andfibroblasts. J Invest Dermatol 96:392–6

Hell-Pourmojib M, Neuner P, Fischer H, Rezaie S,Kindas-Mugge I, Knobler R et al. (2002)Differential expression of a novel gene inresponse to hsp27 and cell differentiation inhuman keratinocytes. J Invest Dermatol119:154–9

Helmbrecht K, Zeise E, Rensing L (2000) Chaperonesin cell cycle regulation and mitogenic signaltransduction: a review. Cell Prolif 33:341–65

Ichikawa T, Suenaga Y, Koda T, Ozaki T, Nakaga-wara A (2007) TAp63-dependent induction ofgrowth differentiation factor 15 (GDF15) playsa critical role in the regulation of keratinocytedifferentiation. Oncogene 27:409–20

Jonak C, Klosner G, Kokesch C, Fodinger D,Honigsmann H, Trautinger F (2002) Subcor-neal colocalization of the small heat shock

Con

trol

BMP7 GDF15 HSP27 HSP90 CDC2

Transcription

Signaling

MetabolismUnderexpressed genes

Translation

Transcription

Signaling

Apoptosis/cell cycle27%

15%

14%11%

10%

5%

5% 4% 4% 3% 2%

Overexpressed genes

Metabolism

Cytoskeleton

Proteolysis

Growth factor/hormone

Transport

Unknown

Chaperone

23%

20%

11%

1%2%2%2%3%4%5%

7%

10%

10%

Apoptosis/cell cycle

Unknown

Cytoskeleton

Immune response

Adhesion

Chaperone

Extracellular matrix

Growth factor/hormone

Proteolysis

Transport

12

10

8

6

4

2

0

5

4

3

2

1

0Rel

ativ

e m

RN

A e

xpre

ssio

n

Rel

ativ

e m

RN

A e

xpre

ssio

n

p210

1

2

3

4

5

Rel

ativ

e m

RN

A e

xpre

ssio

n

IL-1-α IL-1-βControl lcilin

2.5

2

1.5

1

0.5

0

Rel

ativ

e m

RN

A e

xpre

ssio

n

Icili

n

IL-1α

β-Actin

Figure 2. Alterations of gene expression in the epidermal keratinocytes in response to the TRPA1 agonist treatment. Primary human epidermal keratinocytes

were treated using the 10mM pharmacological TRPA1 agonist, icilin, for 24 hours. Agilent microarray and real-time PCR analyses of global gene expression

profiles in the control and icilin-treated keratinocytes. (a and b) Diagrams showing the ontology of overexpressed (a) and underexpressed (b) genes with two-fold

or higher changes in expression after icilin treatment, as determined by microarray analysis. (c–f) Real-time PCR for the transcript of selected genes with altered

expression levels in icilin-treated and control keratinocytes. (g) Western blot analysis of 17 kDa IL-1a protein in the control and icilin-treated keratinocytes.

2314 Journal of Investigative Dermatology (2009), Volume 129

R Atoyan et al.TRPA1 Expression in Human Skin

protein, hsp27, with keratins and proteins ofthe cornified cell envelope. Br J Dermatol147:13–9

McNamara CR, Mandel-Brehm J, Bautista DM,Siemens J, Deranian KL, Zhao M et al.(2007) TRPA1 mediates formalin-inducedpain. Proc Natl Acad Sci USA 104:13525–30

Story GM, Peier AM, Reeve AJ, Eid SR, Mosbacher J,Hricik TR et al. (2003) ANKTM1, a TRP-like

channel expressed in nociceptive neurons, isactivated by cold temperatures. Cell 112:819–29

Trautinger F, Kindas-Mugge I, Dekrout B, KnoblerRM, Metze D (1995) Expression of the 27-kDa heat shock protein in human epidermisand in epidermal neoplasms: an immunohis-tological study. Br J Dermatol 133:194–202

Trevisani M, Siemens J, Materazzi S, BautistaDM, Nassini R, Campi B et al. (2007)

4-Hydroxynonenal, an endogenous alde-hyde, causes pain and neurogenic inflamma-tion through activation of the irritantreceptor TRPA1. Proc Natl Acad Sci USA104:13519–24

Werkheiser JL, Rawls SM, Cowan A (2006) Icilinevokes a dose- and time-dependent increasein glutamate within the dorsal striatum ofrats. Amino Acids 30:307–9

Putting Iatrogenic Risk in Perspective: Basal Cell Cancerin PUVA Patients and AustraliansJournal of Investigative Dermatology (2009) 129, 2315–2316; doi:10.1038/jid.2009.36; published online 5 March 2009

TO THE EDITOROwing to the lack of comparablepopulation-based incidence data, aclinical perspective for assessing basalcell cancer (BCC) incidence in patientsundergoing potentially carcinogenictherapies such as psoralens plus UVA(PUVA) has been largely lacking

(Nijsten and Stern, 2003). The recentprospective cohort study of BCC inci-dence in a subtropical Australian po-pulation provides a clinical, populationcontext for assessing the BCC risk inNorth American PUVA-treated patients(Richmond-Sinclair et al., 2008). Thesedata provide additional evidence that

exposure to cutaneous carcinogens inyounger persons results in greater in-creases of the risk of BCC than compar-able exposure in older persons(Gallagher et al, 1995).

I used methods comparable withthose in Richmond-Sinclair’s study toquantify histologically diagnosed BCCincidence in the PUVA cohort during a10-year period (January 1990 to 31December 1999). Of 1,380 patientsoriginally enrolled, 1,021 were stillparticipating in 1990, 14 years afterfirst exposure to PUVA. A total of 692(67%) patients were followed for theentire decade. Of the remaining, 254had died and 88 were lost to follow-up.Complete data (up to death) wereavailable for 92%.

The PUVA cohort was older than theNambour cohort (43 vs 24% X60 years,w2 Po0.001); a higher proportion weremales (64 vs 56%, w2 Po0.001); fewerpatients were fair-skinned (30% skintypes 1 and 2 vs 55% fair-complectedin the Nambour cohort (w2 Po0.001)).At enrollment, only 10% of the PUVAcohort lived in subtropical areas com-parable with that of the Nambour cohort(below 301 latitude), and most (72%)lived at or above 401 North latitude.Although older, the PUVA cohort had alower percentage of patients with ahistory of BCC before the decade studied(11 vs 18%, w2 Po0.001).

The age-adjusted incidence of bothcohorts of persons developing one ormore BCCs in a decade was nearlyidentical (Table 1). PUVA patients who

Table 1. Age-specific incidence rates of BCC (per 100,000 person-years)1990–1999 for PUVA cohort and IRR with 95% CI compared with thosefor Queensland (Australia)1

PUVA cohortPUVA compared with

Nambour2

Total

patients

Patient-

years Incidence 95% CI IRR 95% CI

Men, age in years

20–39 5 229 2,183 703–4,588 2.17 0.63–7.50

40–59 53 2,306 2,298 1,721–3,006 1.29 0.87–1.91

60+ 90 2,733 3,293 2,648–4,048 0.83 0.63–1.11

Weighted average3 148 5,268 2,425 2,027–2,889 1.35 1.04–1.75

Women, age in years

20–39 7 444 1,577 631–3,248 4.24 0.88–20.42

40–59 23 1,599 1,484 941–2,228 0.86 0.53–1.38

60+ 24 1,253 1,915 802–1,864 0.71 0.45–1.11

Weighted average3 54 3,247 1,606 1,196–2,100 1.26 0.90–1.78

Total 202 9,515 1,892 1,625–2,189 1.32 1.07–1.62

BCC, basal cell cancer; 95% CI, 95% confidence interval; IRR, incidence rate ratio; PUVA, psoralensplus UVA.1Richmond-Sinclair et al. (2008).2IRR of the PUVA cohort compared with that of Queensland (Richmond-Sinclair et al., 2008).3Age standardized to the world population (Ahmad et al., 2001).

Abbreviations: BCC, basal cell carcinoma; PUVA, psoralens plus UVA

www.jidonline.org 2315

RS SternBCC Iatrogenic Risk in Perspective