phase ii metabolism in human skin: skin explants...

1521-009X/43/1/126–139$25.00 http://dx.doi.org/10.1124/dmd.114.060350DRUG METABOLISM AND DISPOSITION Drug Metab Dispos 43:126–139, January 2015Copyright ª 2014 by The American Society for Pharmacology and Experimental Therapeutics

Phase II Metabolism in Human Skin: Skin Explants Show FullCoverage for Glucuronidation, Sulfation, N-Acetylation, Catechol

Methylation, and Glutathione Conjugation s

Nenad Manevski, Piet Swart, Kamal Kumar Balavenkatraman, Barbara Bertschi, Gian Camenisch,Olivier Kretz, Hilmar Schiller, Markus Walles, Barbara Ling, Reto Wettstein, Dirk J. Schaefer,

Peter Itin, Joanna Ashton-Chess, Francois Pognan, Armin Wolf, and Karine Litherland

Drug Metabolism and Pharmacokinetics (N.M., P.S., G.C., O.K., H.S., M.W., K.L.), Pre-clinical Safety (K.K.B., B.B., F.P., A.W.),and Clinical Sciences and Innovation Translational Medicine (J.A.-C.), Novartis Institutes for BioMedical Research, Novartis Pharma,Basel, Switzerland; and Department of Plastic, Reconstructive, Aesthetic and Hand Surgery (B.L., R.W., D.J.S.), and Department

of Dermatology (P.I.), University Hospital Basel, Basel, Switzerland

Received July 30, 2014; accepted October 21, 2014

ABSTRACT

Although skin is the largest organ of the human body, cutaneous drugmetabolism is often overlooked, and existing experimental modelsare insufficiently validated. This proof-of-concept study investigatedphase II biotransformation of 11 test substrates in fresh full-thicknesshuman skin explants, a model containing all skin cell types. Resultsshow that skin explants have significant capacity for glucuronidation,sulfation, N-acetylation, catechol methylation, and glutathione conju-gation. Novel skin metabolites were identified, including acyl glucu-ronides of indomethacin and diclofenac, glucuronides of 17b-estradiol,N-acetylprocainamide, and methoxy derivatives of 4-nitrocatecholand 2,3-dihydroxynaphthalene. Measured activities for 10 mM sub-strate incubations spanned a 1000-fold: from the highest 4.758pmol·mg skin–1·h–1 for p-toluidine N-acetylation to the lowest 0.006pmol·mg skin–1·h–1 for 17b-estradiol 17-glucuronidation. Interindividual

variability was 1.4- to 13.0-fold, the highest being 4-methylumbelliferoneand diclofenac glucuronidation. Reaction rates were generally linearup to 4 hours, although 24-hour incubations enabled detection ofmetabolites in trace amounts. All reactions were unaffected by theinclusion of cosubstrates, and freezing of the fresh skin led to loss ofglucuronidation activity. The predicted whole-skin intrinsic metabolicclearances were significantly lower compared with correspondingwhole-liver intrinsic clearances, suggesting a relatively limited contri-bution of the skin to the body’s total systemic phase II enzyme-mediated metabolic clearance. Nevertheless, the fresh full-thicknessskin explants represent a suitable model to study cutaneous phase IImetabolism not only in drug elimination but also in toxicity, asformation of acyl glucuronides and sulfate conjugates could play a rolein skin adverse reactions.

Introduction

With 1–2 m2 of surface area (Verbraecken et al., 2006) and 2–4 kg oftotal body weight (Brown et al., 1997), human skin is often exposed totherapeutic drugs, cosmetic ingredients, and environmental xenobiotics,either after topical or transdermal administration. In addition, manysystemically administered drugs distribute into the skin, with local ex-posure approaching (Klimowicz et al., 1988; Borg et al., 1999; Krishnaet al., 2010), or sometimes even exceeding, corresponding plasma levels(Zucchi et al., 2001; Brunner et al., 2002). As a result, cutaneous drugmetabolism is relevant for pharmacotherapy, both from the perspectivesof metabolic elimination and activation of prodrugs, as well as theformation of reactive metabolites that may cause skin toxicity (Sharmaet al., 2013). Despite this potential importance, cutaneous drug me-tabolism is generally overlooked and rarely discussed during drugdevelopment. A better understanding of skin metabolism could also

close the general in vitro–in vivo extrapolation gap for drugs whereextensive extrahepatic metabolic clearance is suspected (Hiraoka et al.,2005; Gundert-Remy et al., 2014).Drug metabolism is often a two-step process. In the first phase,

compounds undergo reactions of oxidoreduction and hydrolysis, intro-ducing or exposing nucleophilic groups; in the second phase, nucle-ophilic groups are conjugated with an endogenous molecule, such asglucuronic acid or glutathione, yielding hydrophilic and easily ex-cretable drug conjugates with often diminished pharmacological ac-tivity. This study focused on activities of cutaneous phase II metabolicenzymes. A literature survey suggested that mRNA expression ofphase II metabolic enzymes is higher than expression of correspondingphase I enzymes in human skin (Luu-The et al., 2009; Hu et al., 2010;van Eijl et al., 2012). Numerous detected cutaneous transferases(Table 1), namely UDP-glucuronosyltransferases (UGTs), sulfotransferases(SULTs), N-acetyltransferases (NATs), catechol-O-methyltransferase(COMT), and glutathione-S-transferases (GSTs), indicate a poten-tially significant role of the human skin in metabolic elimination anddetoxification.

dx.doi.org/10.1124/dmd.114.060350.s This article has supplemental material available at dmd.aspetjournals.org.

ABBREVIATIONS: AcCoA, acetyl coenzyme A; CDNB, 1-chloro-2,4-dinitrobenzene; CLint;skin, skin intrinsic clearance; CLint;liver, liver intrinsicclearance; COMT, catechol-O-methyltransferase; DMSO, dimethylsulfoxide; GST, glutathione-S-transferase; NAT, N-acetyltransferase; MTT,3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; PAPS, 39-phosphoadenosine 59-phosphosulfate; SAM, S-(59-adenosyl)-L-methionine;SULT, sulfotransferase; UDPGA, UDP-a-D-glucuronic acid; UGT, UDP-glucuronosyltransferase.

126

http://dmd.aspetjournals.org/content/suppl/2014/10/21/dmd.114.060350.DC1Supplemental material to this article can be found at:

at ASPE

T Journals on O

ctober 12, 2020dm

d.aspetjournals.orgD

ownloaded from

http://dx.doi.org/10.1124/dmd.114.060350

http://dx.doi.org/10.1124/dmd.114.060350

http://dmd.aspetjournals.org

http://dmd.aspetjournals.org/content/suppl/2014/10/21/dmd.114.060350.DC1

http://dmd.aspetjournals.org/

Despite the reports of gene expression (Table 1), functional data suchas enzymatic activities are largely missing, especially with the freshhuman skin explants. Table 2 summarizes literature reports of phase IImetabolic activities detected in different skin models. Most of the studieswere performed with a single test compound and with experimentalmodels that do not reflect the full complexity of the human skin—forexample, isolated keratinocytes or subcellular fractions. Moreover, sinceskin is a fibrous tissue that is notably more difficult to homogenize thanliver, preparation of skin subcellular fractions may reduce enzyme activitiesbelow the detection limits. Use of three-dimensional reconstructed skin

models and immortalized cell lines (e.g., HaCaT cells) raises concernsof altered gene expression, even if existing studies are mostly en-couraging (Bonifas et al., 2010b; Jackh et al., 2011; Gotz et al., 2012).Skin explants were rarely used in the past and mostly dermatomed tothickness below 500 mm (Ademola et al., 1993; Moss et al., 2000;Goebel et al., 2009; Zalko et al., 2011), thus minimizing the potentialmetabolic contributions of skin dermis. Taken together, difficulties inaccessing fresh human tissue, lack of alternative and simpler modelvalidation against the full-thickness human skin, limited selection oftest substrates, and potential enzyme inactivation in skin subcellular

TABLE 1

Expression of phase II metabolic enzymes in human skin

Literature reports are based on either mRNA expression or detection of the corresponding enzymes at the protein level.

Family Individual Enzymes Detected Reference

UGT 1A6, 1A8, 1A10a, 2A1, 2A3, 2B4, 2B15, 2B17, 2B28 Luu-The et al., 2009; Hu et al., 2010SULT 1A1a, 1A2, 1A3, 1A4, 1B1, 1E1a, 1E2, 2A1, 2B1a Dooley et al., 2000; Higashi et al., 2004; Luu-The et al., 2009;

Hu et al., 2010; Kushida et al., 2011NAT NAT1a, NAT2 Luu-The et al., 2009; Bonifas et al., 2010a; Hu et al., 2010; van

Eijl et al., 2012COMT Single human gene is detected Luu-The et al., 2009; Hu et al., 2010; van Eijl et al., 2012GST aa, ma, Pa, Q, z, V Blacker et al., 1991; Raza et al., 1991; Luu-The et al., 2009; Hu

et al., 2010; van Eijl et al., 2012

aEnzymes with higher reported expression.

TABLE 2

Reported activities of the phase II drug metabolism enzymes in different human skin models

See original publications for further details. Literature reports of phase II enzyme expression are presented in Table 1.

Substrate (concentration, mM) Skin Model Observed Activity Reference

Glucuronidation4-Methylumbelliferone Skin EpiDerm microsomes, keratinocytes 1.3–2.0 nmol·min–1·mg–1 Gotz et al., 20124-Methylumbelliferone Microsomes from EpiDerm and Phenion FT ;0.1–2.0 nmol·min–1·mg–1 Jackh et al., 2011Bisphenol A (multiple) Explants (500 mm) Vmax = 16.8 nmol·h–1 Zalko et al., 2011Triclosan Explants (230 mm) ;5% conversion (24 h) Moss et al., 20003-Indolylacetic acid Explants (500 mm, postmortem), microsomes 0.5–2.5 nmol·min–1·mg–1 (microsomes) Ademola et al., 1993Androstanediol Tissue homogenate 0.07–0.40% conversion Rittmaster et al., 1993Testosterone and

dihydroxytestosteroneTissue pieces (1 mm) 0.01–1.40% conversion Lobo et al., 1987

p-Aminophenol,p-nitrophenol

Keratinocytes, fibroblasts 5–10 nmol·h–1·mg–1 Rugstad and Dybing, 1975

Sulfation12-Hydroxynevirapine

(1 mM)Dermatomed skin homogenate (postmortem) Detected (qualitative) Sharma et al., 2013

17b-Estradiol Keratinocytes 0.11–0.58 pmol·min–1·mg–1 Kushida et al., 2011Bisphenol A Explants (500 mm) Vmax = 0.30 nmol·h–1 Zalko et al., 201117b-Estradiol, estrone,

didehydroepiandrosterone,butylparaben

Cytosol from dermatomed skin, keratinocytes 0.05–1.06 pmol·min–1·mg–1 (steroids);butylparaben was detected

Prusakiewicz et al., 2007

Triclosan Explants (230 mm) ;3% conversion (24 h) Moss et al., 2000Minoxidil Cytosol Anderson et al., 1998

N-Acetylationp-Toluidine Skin microsomes, EpiDerm, keratinocytes 0.63–3.03 nmol·min–1·mg–1 Gotz et al., 2012p-Aminobenzoic acid S9 fraction from EpiDerm and Phenion FT 11.2–17.0 nmol·min–1·mg–1 Jackh et al., 2011p-Phenylenediamine HaCaT cells Up to 49.7 nmol·min–1·mg–1 Bonifas et al., 2010ap-Aminobenzoic acid Keratinocytes and HaCaT cells 8.0–44.5 nmol·min–1·mg–1 Bonifas et al., 2010b4-Amino-2-hydroxytoluene,

p-aminobenzoic acid,p-aminophenol

Explants (200–400 mm), HaCaT cells 1.04–1.12 nmol/h per 106 cells Goebel et al., 2009

p-Phenylenediamine Cytosol from skin explants, keratinocytes 0.41–3.68 nmol·min–1·mg–1 Kawakubo et al., 2000Catechol methylation

Adrenaline Explant homogenate Detected Bamshad et al., 1964, 1970Glutathione conjugation

1-Chloro-2,4-dinitrobenzene Cytosol and microsomes from skin explants,EpiDerm

1–60 nmol·min–1·mg–1 (dependingon the model)

Gotz et al., 2012

1-Chloro-2,4-dinitrobenzene HaCaT cell, keratinocytes, melanocytes 45–261 nmol·min–1·mg–1 Zhang et al., 20021-Chloro-2,4-dinitrobenzene,

benzo[a]pyrene 4,5-oxide,styrene 7,8-oxide

Cytosol from skin explants 3.9–25.8 nmol·min–1·mg–1 Raza et al., 1991

Phase II Drug Metabolism in Skin Explants 127

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


fractions all contribute to limited knowledge about phase II metabolismin the human skin.To address these challenges, we investigated activities of UGTs,

SULTs, NATs, COMT, and GSTs in fresh full-thickness human skinexplants. This experimental model was selected based on the presenceof all skin cell types, such as keratinocytes, Langerhans cells, Merkelcells, and melanocytes in the epidermis, as well as fibroblasts andmacrophages in the dermis (Lebonvallet et al., 2010). Therefore, as anadvantage over many simpler experimental models, full-thickness skinexplants show enzymatic activities of both skin epidermis and dermis.With an aim of preserving skin’s maximal metabolic capacity, specialattention was paid to freshness and viability of the explant culture. Toproduce robust results and shed light on interindividual variability, 11substrates of different physicochemical properties (Fig. 1) were in-dividually assayed in a minimum of three different skin donors. To gainbetter understanding of the skin’s relative contribution to the body’s totalsystemic phase II enzyme-mediated metabolic clearance, metabolicactivities measured in skin explants were scaled up to apparent whole-skin intrinsic clearances (CLint;skin) and compared with correspondingwhole-liver intrinsic clearances (CLint;liver). The results presented heresignificantly deepen our knowledge about cutaneous phase II metabo-lism, and open new research directions toward understanding not onlymetabolic elimination but also the toxicological impact of new drugs andcosmetic products.

Materials and Methods

Reagents and Chemicals

Belzer UW cold storage solution (organ preservation medium) was obtainedfrom Bridge to Life (Columbia, SC). Cytotoxicity detection kit plus, a lactatedehydrogenase release assay, was acquired from Roche (Basel, Switzerland).Penicillin-streptomycin-glutamine (10,000 U·ml–1 penicillin, 10,000 mg·ml–1

streptomycin, 29.2 mg·ml–1 glutamine; used as 100-fold dilution) was purchasedfrom Life Technologies (Carlsbad, CA). Williams’ E medium (1.8 mM Ca2+, noglutamine), dimethylsulfoxide (DMSO, $99.7%), insulin from bovine pancreas($27 USP units·mg–1), hydrocortisone (suitable for cell culture), formic acid($98%), 4-methylumbelliferone ($98%), 4-methylumbelliferone-b-D-glucuronidehydrate ($98%), 4-nitrophenyl b-D-glucuronide ($98%), 17b-estradiol ($98%),17b-estradiol-17-b-D-glucuronide sodium salt ($98%), 17b-estradiol-3-b-D-glucuronide sodium salt ($98%), 17b-estradiol-3-sulfate sodium salt ($93%),minoxidil ($99%), minoxidil sulfate ($98%), p-toluidine (99.7%), 49-methylacetanilide(99%), 2,3-dihydroxynaphthalene ($98.0%), 3-methoxy-2-naphthol (97%),4-nitrocatechol (97%), 4-nitroguaiacol (97%), 1-chloro-2,4-dinitrobenzene (CDNB;$99%), procainamide hydrochloride ($99%), N-acetylprocainamide hydrochlo-ride ($99%), UDP-a-D-glucuronic acid trisodium salt (UDPGA; 98–100%),adenosine 39-phosphate 59-phosphosulfate lithium salt hydrate (PAPS; $60%),S-(59-adenosyl)-L-methionine chloride dihydrochloride (SAM; $75%), D-saccharicacid 1,4-lactone monohydrate (D-saccharolactone; $98%), acetyl coenzymeA sodium salt (AcCoA; $93%), and in vitro toxicology assay kit [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) based] were pur-chased from Sigma-Aldrich (Buchs, Switzerland). Indomethacin, indomethacinacyl-b-D-glucuronide, diclofenac sodium salt, diclofenac acyl-b-D-glucuronide,

Fig. 1. Chemical structures of the compounds tested inthis study. Location of metabolic reaction is indicatedwith a dotted circle or arrow. Observed phase II bio-transformation is indicated below the substrate’s name.Physicochemical properties of the compounds are presentedin Table 5. Phase II metabolic activities measured in thefresh full-thickness human skin explants are presented inFig. 3 and Table 6.

128 Manevski et al.

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


triclosan O-sulfate sodium salt, triclosan O-b-D-glucuronide sodium salt, S-(2,4-dinitrophenyl)-glutathione, and 4-hydroxycarbazeran were acquired from TorontoResearch Chemicals (Toronto, ON, Canada). Triclosan was purchased fromMolekula (Munich, Germany). Organic solvents of liquid chromatography–massspectrometry or higher purity grade were used in this study. Stock solutions of thetest substrates were prepared in DMSO (10 mM) and then stored at –20�C untiluse. Stock solution of cosubstrates (UDPGA, PAPS, SAM, and AcCoA) andD-saccharolactone were prepared in purified water (50 mM) and used immediatelyin the assay.

Materials for Tissue Explant Culture

Stericup filter units (0.22 mm) and receiver flasks were purchased fromMillipore (Billerica, MA). Sterile biopsy explant tools, 4 mm in diameter, wereordered from either Acuderm (Fort Lauderdale, FL) or Stiefel (GlaxoSmithK-line company, Research Triangle Park, NC). Falcon 24-well plates werepurchased from Corning (Tewksbury, MA).

Skin Explants Culture

Healthy full-thickness human skin was obtained under informed consent fromindividuals undergoing plastic or reconstructive surgery at the Institute for Plastic,Reconstructive, Aesthetic and Hand Surgery, Basel University Hospital (Basel,Switzerland). The study protocol was approved by the local ethics committee ofBasel, Switzerland, and the study was performed in accordance with theDeclaration of Helsinki (1964 and subsequent revisions). In total, skin samplesfrom 17 individuals were collected (donors’ demographic information is presentedin Table 3). Skin donors were predominantly female Caucasians (87.5%), althoughtwo male subjects were also included in the study (Table 3). Donor age rangedfrom 35 to 73 years, with a median age of 58 years. Depending on the donor, skinsamples were taken from different anatomic regions, namely abdomen (8 donors),breast (3 donors), inguinal area (groin; 2 donors), arm (2 donors), thigh (1 donor),and eyelids (1 donor). To preserve viability and metabolic capacity of the tissue,excised human skin was immediately submerged in the Belzer UW organpreservation solution supplemented with penicillin (100 U·ml–1) and streptomycin(100 mg·ml–1), a specialized medium commonly used for solid organ trans-plantation (Sollinger et al., 1989). Skin was transported to our laboratories at 4�C,generally within 30–90 minutes from the time of surgical excision.

After arrival, skin tissue was handled within a laminar flow hood andprocessed using sterile dissection tools. Adipose tissue and hypodermis wereremoved and, depending on the anatomic region and individual sample, dermiswas trimmed with curved surgical scissors to achieve maximal overall skinthickness of approximately 3 mm. Skin thickness varied based on anatomicregion and individual characteristics of the donor (,1 mm for eyelids, 1–2 mmfor breast and inguinal, 2–5 mm for abdomen and thigh skin). Cylindrical skin

explants, 4 mm in diameter (12.5 mm2), were prepared with a sterile biopsy tooland immediately placed in prewarmed Williams’ E medium supplemented withinsulin (10 mg·ml–1), hydrocortisone (10 ng·ml–1), and penicillin-streptomycin-glutamine (100 U·ml–1 of penicillin; 100 mg·ml–1 of streptomycin; 2 mML-glutamine) (Lu et al., 2007). Medium pH value was 7.4, culturing temperaturewas 37�C, incubator humidity was 90%, and CO2 content was 5%. Skin explantviability after 24 hours of culture was confirmed by histology (hematoxylin andeosin staining), lactate dehydrogenase release, and MTT viability assays.Viability assays were performed according to the manufacturer’s protocols. Skinexplant preparation procedure, composition and volume of the culturing medium,calcium concentration, and the use of antibiotics were optimized prior tometabolic experiments (Litherland et al., manuscript in preparation).

Drug Metabolism Assays

Skin explants were incubated in 0.5 ml of Williams’ E medium (pH 7.4)spiked with 10 mM test substrate (final DMSO concentration was 0.1%).Incubations were performed for 15–24 hours, depending on the individual skinsample assayed. Time-course assays were performed in the interval from 1 to24 hours. Enzyme kinetics assay of 4-methylumbelliferone glucuronidation wasperformed with substrate concentrations from 1 to 150 mM and incubation timeof 3 hours. To achieve maximal exposure of skin to probe substrates, skin explantswere freely floating in the medium with epidermis facing the air-liquid interface.After incubation time, both skin explants and corresponding incubation mediawere placed in centrifuge tubes and snap-frozen in liquid nitrogen. Until analysis,samples were stored at –80�C. All incubations were performed in duplicate.

To efficiently extract metabolites from the skin explants, skin tissue wascryofractured with a cryoPREP Impactor (Covaris, Woburn, MA). Skin explantswere placed in TT05XT tissue tubes (Covaris), cooled in liquid nitrogen, and thencryofractured once with cryoPREP Impactor (impact strength 2). Cryofractured skinsamples were transferred to centrifuge tubes filled with lysing matrix D (MPBiomedicals, Santa Ana, CA), 1 ml of 70% acetonitrile was added (v/v), andsamples were further homogenized using a FastPrep instrument (MP Biomedicals;agitation speed was 4.0 m·s–1; three cycles of 20 seconds each). Corresponding skinincubation medium was directly mixed with acetonitrile (30:70 v/v) and vortexed for1 minute. All samples were placed overnight at –20�C, centrifuged at 30,000g for30 minutes, and aliquots of the supernatants were spiked with suitable internalstandard. Samples were evaporated to dryness (in vacuum or under nitrogen stream)and then reconstituted with mobile phase. Negative control samples, probe substrate inblank medium, and skin explants without probe substrate were analyzed in parallel.

Analytics

Extracted metabolites were detected and quantified using a Quattro Ultimatriple-quadruple mass spectrometer equipped with an electrospray ion source

TABLE 3

Demographics of skin donors that participated in the study

Out of 16 donors in total, 87.5% were female and 12.5% were male. See Fig. 3 and Table 6 for the measured phase II metabolic activities.

DonorCode

Gender Ethnicity AgeSkin Anatomic

RegionSmoking Alcohol Substrates Tested

8 Female Caucasian 35 Abdomen Daily Never 4-Methylumbelliferone, 17b-estradiol, indomethacin, minoxidil11 Female Caucasian 59 Arm Not known Not known 4-Methylumbelliferone, 17b-estradiol, indomethacin, minoxidil16 Female Caucasian 46 Breast Never Occasionally 4-Methylumbelliferone, 17b-estradiol, indomethacin, diclofenac,

triclosan, minoxidil17 Female Caucasian 43 Thigh Daily 1–2 days/week 4-Methylumbelliferone, diclofenac, triclosan, minoxidil18 Male Caucasian 66 Eyelids Former Occasionally 4-Methylumbelliferone, diclofenac, triclosan22 Female Caucasian 72 Abdomen Never Daily 4-Methylumbelliferone, minoxidil, p-toluidine, isoniazid, procainamide,

4-nitrocatechol, 2,3-dihydroxynaphthalene23 Female Caucasian 59 Inguinal Daily Occasionally 4-Methylumbelliferone, minoxidil, p-toluidine, isoniazid, procainamide24 Female Caucasian 40 Abdomen Never Never 4-Methylumbelliferone, minoxidil, p-toluidine, isoniazid, procainamide25 Female Caucasian 72 Abdomen Never Occasionally 4-Methylumbelliferone29 Female Not known 57 Breast Never Never 4-Methylumbelliferone, 4-nitrocatechol, 2,3-dihydroxynaphthalene30 Male Caucasian 45 Abdomen Daily Occasionally 4-Methylumbelliferone, 4-nitrocatechol, 2,3-dihydroxynaphthalene31 Female Caucasian 59 Abdomen Never Daily 4-Methylumbelliferone, minoxidil, p-toluidine, 2,3-dihydroxynaphthalene32 Female Caucasian 54 Abdomen Never Occasionally 4-Methylumbelliferone, minoxidil, p-toluidine, 2,3-dihydroxynaphthalene36 Female Caucasian 73 Breast Never Occasionally 1-Chloro-2,4-dinitrobenzene37 Female Caucasian 63 Inguinal Never Occasionally 1-Chloro-2,4-dinitrobenzene38 Female Not known 58 Abdomen Never Never 1-Chloro-2,4-dinitrobenzene


at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


(Waters, Milford, MA), coupled with an Agilent 1100 capillary high-performanceliquid chromatography pump (Agilent, Santa Clara, CA) and a CTC Palautosampler (CTC Analytics, Zwingen, Switzerland). Metabolite quantificationwas performed in multiple-reaction monitoring mode, and resulting chromato-grams were analyzed with MassLynx 4.1 software (Waters). Since 3-methoxy-2-naphthol, a methylated metabolite of 2,3-dihydroxynaphthalene, is highlyfluorescent and poorly ionizable, the analysis was performed using an Agilent1100 high-performance liquid chromatography system equipped with a fluores-cence detector. Fluorescence data were analyzed with ChemStation B.04.02software (Agilent). For all used analytical methods, the limits of detection andquantification were estimated based on signal-to-noise ratios of 3 and 10,respectively. Standard curves were generally prepared in the concentration rangeof 1–1000 nM. Further details of the analytical methods are presented in Table 4.

Data Analysis

Amounts of metabolites quantified were normalized to average skin explantweight (for the individual donor) and total media volume (500 ml). To calculatethe total amount of metabolite formed in the incubation, we summed theamounts quantified in skin explant extract and corresponding medium extract.Due to difficulties in quantitative extraction of proteins from the skin tissue,activities measured in skin explants are presented as picomole of metaboliteformed per milligram of skin tissue (not milligram of protein) and hours ofincubation time (pmol·mg skin–1·h–1). Physicochemical properties of the testsubstrates were calculated using ACD Laboratories (Toronto, ON, Canada) orChemAxon (Budapest, Hungary) software. Figure preparation, statistical tests,and curve fitting were performed with GraphPad Prism 6.02 for Windows(GraphPad Software Inc., San Diego, CA).

Prediction of Skin Clearances and Comparison with Liver Clearances

Intrinsic Clearances Derived from Skin Explants. Applying Michaelis-Menten enzyme kinetics [v ¼ Vmax½S�=Km þ ½S�, where v, Vmax, Km, and[S] represent the metabolic activity of the skin explant (pmol·min–1, averageskin explant weight for individual donor is taken into account), the maximumreaction velocity (pmol·min–1), the affinity constant (mM), and the concentra-tion of substrate at which the activity was measured (mM), respectively] andassuming an absence of unspecific intracellular binding events, the maximalintrinsic metabolic skin clearance (at [S] , , Km, ml·min–1) from a single-concentration metabolism assay in the skin explants can be calculated as follows:

CLint;skin ¼ v

½S�

Since reactions of sulfation, N-acetylation, catechol methylation, andglutathione conjugation showed nonlinearity over 24 hours of incubation inskin explants (see Fig. 4), the CLint;skin values for these reactions were multipliedby a correction factor derived from the time-course assays as follows:

Nonlinearity correction ¼ CL4hint;skinCL24hint;skin

¼ v4h�½S�

v24h=½S�

where CL4hint;skin and CL24hint;skin are the intrinsic skin clearances calculated after 4and 24 hours of incubation, respectively. Accordingly, v4h and v24h are theenzyme activities measured after 4 and 24 hours of incubation [calculatedcorrection factors: glucuronidation 1.00 (linear); catechol methylation 1.61;N-acetylation 1.73; glutathione conjugation 2.75; sulfation 2.83]. The in vitro

TABLE 4

Analytical methods used for separation, detection, and quantification of metabolites formed in the human skin explants

For all methods, eluents A and B were 0.1% formic acid in water and 0.1% formic acid in acetonitrile, respectively. Analytes were separated on the following high-performance liquidchromatography (HPLC) columns: Agilent StableBond C18 (column A: 50 � 1.0 mm; column B: 150 � 1.0 mm; 3.5 mm particle size; Agilent), Phenomenex Luna pentafluorophenyl (column C: 150 �1.0 mm, 3 mm particle size; Phenomenex, Torrance, CA), or Waters SunFire C18 column (column D: 150 � 2.1 mm, 3.5 mm particle size; Waters). Eluent flow rates were 0.1 and 0.3 ml/min for 1.0- and2.1-mm diameter analytical columns, respectively. Column temperature was 40�C and injection volume was 10 ml.

AnalyteHPLCColumn

Gradient Internal Standard Ionization TransitionsCone/

CollisionLOD/LOQ

m/z V nM

4-Methylumbelliferone-b-D-glucuronide

C 0–5 min, 3→95% B; 5–8 min, 95%B; 8–8.1 min, 95→3% B; 8.1–17min, 3% B

p-Nitrophenol-b-D-glucuronide

ESI– 351 . 175 50/14 2.28/7.63

17b-Estradiol-17-b-D-glucuronide

A 0–1 min, 3% B; 1–9 min, 3→95% B;9–17 min, 95% B; 17–17.1 min,95→3% B; 17.1–25 min, 3% B


ESI– 447 . 271 50/30 ,1


447 . 271 50/32 ,1

17b-Estradiol-3-sulfate 352 . 272 50/30 ,1Minoxidil sulfate C 0–5 min, 5→95% B; 5–11 min, 95%

B; 11–11.1 min, 95→5% B;11.1–20 min, 5% B


ESI– 288 . 96; 288 . 97; 288. 105; 288 . 208

60/22 ,1

49-Methylacetanilide B 0–5 min, 5→95% B; 5–14 min, 95%B; 14–14.1 min, 95→5% B; 14.1–22 min, 5% B

4-Hydroxycarbazeran ESI+ 150 . 93; 150 . 108 30/20 ,1

3-Methoxy-2-naphthol D 0–12 min, 15→95% B; 12–16 min,95% B; 16–16.1, 95→15% B;16.1–26 min, 15% B

External quantification Fluorescence lex = 254 nm lem =360 nm

n.a. 1.85/5.48

4-Nitroguaiacol A 0–1 min, 5% B; 1–7 min, 5→95% B;7–12 min, 95% B; 12–12.1 min,95→5% B; 12.1–18 min, 9% B

2,3-Dihydroxynaphthalene

ESI– 168 . 123; 168 . 153 35/17;35/12

4.01/13.36

N-Acetylprocainamide C 0–1 min, 5% B; 1–6 min, 5→95% B;6–12 min, 95% B; 12–12.1 min,95→5% B; 12.1–22 min, 5% B

49-Methylacetanilide ESI+ 278 . 121; 278 . 162;278 . 164; 278 . 205

30/20 ,1

Indomethacin acyl-b-D-glucuronide

A 0–1.5 min, 3% B; 1.5–3 min, 3→95%B; 3–13 min, 95% B; 13–13.1 min,95→3% B; 13.1–20 min, 3% B

p-Nitrophenol-b-D-glucuronide

ESI– 532 . 312; 532 . 356 50/11 ,1

Diclofenac acyl-b-D-glucuronide

A 0–1 min, 5% B; 1–5 min, 5→95% B;5–15 min, 95% B; 15–15.1 min,95→5% B; 15.1–22 min, 5% B

Triclosan O-b-D-glucuronide

ESI– 470 . 193; 470 . 294;472 . 193; 472 . 296

25/8 ,1

Triclosan O-sulfate A 0–1 min, 10% B; 1–5 min, 10→95%B; 5–15 min, 95% B; 15–15.1 min,95→10% B; 15.1–22 min, 10% B


ESI– 367 . 287; 369 . 289 20/13 ,1Triclosan O-b-D-

glucuronide463 . 287; 465 . 289 20/13 ,1

ESI+, electrospray ionisation in positive mode; ESI–, electrospray ionization in negative mode; LOD, lower limit of detection; LOQ, lower limit of quantification; n.a., not applicable.

130 Manevski et al.

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


intrinsic clearances can be scaled to in vivo (whole skin) by applying eithera skin surface or skin weight scaling factor. Assuming that average surfaceareas of whole human skin and prepared skin explants are 2 � 106 mm2 and12.5 mm2, respectively, estimated skin surface scaling factor is 160,000-fold.Corresponding skin weight scaling factor ranges from 74,000- to 260,000-fold,depending on the average weight of the skin explants prepared from individualdonors [average skin explant weight ranged from 10 to 35 mg; assumed averagetotal skin weight was 2.6 kg for a 70-kg individual (Brown et al., 1997)].Since prepared skin explants had a well defined surface area (12.5 mm2), buthighly variable weight dependent on skin anatomic region and individualdonor, skin surface scaling factor was used in calculations as a more robustparameter.

Intrinsic Clearances Derived from Liver Fractions. The in vitro intrinsicclearances for the corresponding phase II reactions in human liver fractions(microsomes, cytosol, or S9), CLint;liver (ml·min–1·mg–1), were calculated fromenzyme kinetics parameters reported in the literature (Supplemental Table 1).For reactions following Michaelis-Menten enzyme kinetics, intrinsic clearanceswere calculated as described earlier (CLint;liver ¼ Vmax=Km). In the case of Hill-type reactions, the in vitro clearance of the maximally activated enzyme wascalculated as described elsewhere ½CLint;liver ¼ Vmax

S50ðh2 1Þ

hðh2 1Þ1=h, where h is the Hillslope of the fitted curve and S50 is the concentration of substrate at Vmax/2(Houston and Galetin, 2008)]. In the case of multiple literature reports, averageCLint;liver values were used (see Supplemental Table 1). The intrinsic hepatic invitro clearances were scaled up to a ml·min–1·kg–1 body weight basis asdescribed previously (Camenisch and Umehara, 2012). The following scalingfactors from the literature were applied: 52.5 mg of protein/g of liver for livermicrosomes (Iwatsubo et al., 1997), 80.7 mg of protein/g of liver for cytosol(Houston and Galetin, 2008), 121 mg of protein/g of liver for S9 fraction(Nishimuta et al., 2014), and 25.7 g of liver/kg of body weight for the liverweight (Davies and Morris, 1993).

Calculation of Systemic Organ Clearances. Applying the so-called wellstirred model, and assuming instantaneous and homogenous drug distribution(i.e., no permeability limitation), the systemic in vivo clearances for anyperfused organ in the absence of protein binding can be determined as follows(Ito and Houston, 2004):

CLorgan ¼ Q � CLintQþ CLint

where Q is the organ blood flow [Qliver = 20.7 ml·min–1·kg–1 and Qskin = 4.3 7ml·min–1·kg–1 (Davies and Morris, 1993)]. Systemic organ clearances werecalculated for average body weight of 70 kg.

Results

To investigate phase II metabolism in human skin, full-thicknessskin explants were incubated with 11 test compounds of variousphysicochemical properties (Fig. 1; Table 5). Test compounds wereselected based on either good activity toward the target enzymes—for

example, 4-methylumbelliferone for UGTs (Manevski et al., 2013)and 1-chloro-2,4-dinitrobenzene for GSTs (Habig et al., 1974)—orrelevance to pharmacotherapy (minoxidil, indomethacin, diclofenac,17b-estradiol, triclosan). Many of the test compounds are substratesfor several individual enzymes within the single transferase family: forexample, 4-methylumbelliferone (Manevski et al., 2013), 17b-estradiol(Itäaho et al., 2008), diclofenac (King et al., 2001; Kuehl et al., 2005),and indomethacin (Kuehl et al., 2005) are glucuronidated by multipleUGTs; minoxidil is a substrate for several SULTs (Anderson et al.,1998); and 1-chloro-2,4-dinitrobenzene is metabolized by multiple GSTs(Hayes et al., 2005). Since the stratum corneum may severely limit theskin permeation of topically administered drugs, test compounds weredirectly added to the incubation medium to maximize the tissueexposure. Due to limited availability and small size of the human skinsamples, incubations were mainly performed only with a single substrateconcentration of 10 mM. After 1–24 hours of incubation time, both skinexplant and corresponding medium were analyzed for the presence ofspecific phase II metabolites. Histologic analysis of formalin-fixed paraffin-embedded skin tissue, stained with hematoxylin and eosin, showed thatskin explants kept normal morphology and viability at least for the first24 hours (Fig. 2). All reactions were tested in three different skin donors,with the exception of 4-methylumbelliferone glucuronidation, minoxidilsulfation, p-toluidine N-acetylation, and 2,3-dihydroxynaphthalene meth-ylation, which were assayed in skin explants from 13, 7, 5, and 5 individualdonors, respectively.Phase II Metabolic Activities in Human Skin Explants. Results

show that human skin has a significant potential for glucuronidation,sulfation, N-acetylation, catechol methylation, and glutathione conju-gation (Fig. 3; Table 6). At least one phase II metabolite was detectedfor all different tested compounds, whereas 17b-estradiol and triclosanyielded both glucuronide and sulfate conjugates. To our best knowledge,this is the first report of indomethacin glucuronidation, diclofenacglucuronidation, 17b-estradiol 3- and 17-glucuronidation, procainamideN-acetylation, 4-nitrocatechol methylation, and 2,3-dihydroxynaphtha-lene methylation in the human skin.Average activities for 10 mM incubations varied approximately

1000-fold, ranging from the highest, 4.758 pmol·mg skin–1·h–1 forp-toluidine N-acetylation, to the lowest, 0.006 pmol·mg skin–1·h–1 for17b-estradiol 17-glucuronidation (Fig. 3; Table 6). Importantly, for allphase II reactions, at least one representative compound had fasterturnover: for example, rates higher than 1.0 pmol·mg skin–1·h–1 weremeasured for p-toluidine N-acetylation, 4-methylumbelliferone glucur-onidation, 2,3-dihydroxynaphthalene methylation, and triclosan sulfa-tion (Fig. 3A). The rate of 1-chloro-2,4-dinitrobenzene glutathione

TABLE 5

Physicochemical properties of compounds used in this study

Compound Log Pa Log Da (pH 7.4) pKa/pKba Polar Surface Areab Solvent-Accessible Surface Areab Molecular Weight

Å2

4-Methylumbelliferone 2.43 2.34 pKa 8.00 46.53 230.19 176.1717b-Estradiol 4.15 4.15 pKa 10.27 40.46 436.59 272.38Indomethacin 4.25 0.98 pKa 3.96 68.53 476.23 357.79Diclofenac 4.55 1.44 pKa 4.18 49.33 361.15 296.15Triclosan 5.34 5.20 pKa 7.80 29.46 318.47 289.54Minoxidil 20.41 20.42 pKb 5.54 93.63 291.22 209.25p-Toluidine 1.53 1.53 pKb 5.04 26.02 182.92 107.15Procainamide 1.32 20.56 pKb 9.09 58.36 406.46 235.334-Nitrocatechol 1.68 1.03 pKa 6.87 86.28 198.73 155.112,3-Dihydroxynaphthalene 2.11 2.10 pKa 9.10 40.46 220.46 160.171-Chloro-2,4-dinitrobenzene 2.06 2.06 n.a. 91.64 233.30 202.55

n.a., not applicable.aLog P, Log D (pH 7.4), and pKa or pKb were calculated with ACD Laboratories software.bPolar surface area and solvent-accessible surface area were calculated by ChemAxon software.


at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from

http://dmd.aspetjournals.org/lookup/suppl/doi:10.1124/dmd.114.060350/-/DC1



conjugation was also high within the first hour of the incubation, butslowed down afterward. Significant metabolic turnover was confirmedby calculating the apparent substrate depletion during the experiment(Fig. 3B). For p-toluidine, assuming that N-acetylation is the onlymetabolic pathway, up to 70% of substrate was consumed during24 hours of incubation. On the other hand, reactions of minoxidil sulfation,procainamide N-acetylation, diclofenac glucuronidation, and 17b-estradiolglucuronidation on average consumed less than 1% of substrate in thereaction.Phase II metabolism in the skin explants also exhibited interindivid-

ual variability (Table 6). Variation between donors was the highest for4-methylumbelliferone and diclofenac glucuronidation (up to 13-fold);moderate for 17b-estradiol 3-glucuronidation and 3-sulfation, triclosanglucuronidation, minoxidil sulfation, and 2,3-dihydroxynaphthalenemethylation (3- to 5-fold); and lowest for p-toluidine and procainamideN-acetylation, triclosan sulfation, and 17b-estradiol 17-glucuronidation(1- to 2-fold). Higher interindividual variation observed in4-methylumbelliferone and diclofenac glucuronidation might be due toan eyelid skin sample from a male individual (donor 18). Skin samplefrom this donor glucuronidated 4-methylumbelliferone and diclofenacat a rate that was approximately 3-fold faster than the correspondingaverage value. Since cutaneous glucuronidation of 4-methylumbelliferonewas tested in skin explants from 13 individual skin donors, results alsoenabled insight into the relationship between glucuronidation rate anddonor demographics. However, measured glucuronidation rates werenot statistically significantly correlated with donors’ age, smoking habits,or alcohol intake (data not shown).Time Course of Phase II Metabolite Formation in Human Skin

Explants. To determine the linearity of phase II metabolism in humanskin explants, the time course of five representative reactions were tested,namely 4-methylumbelliferone glucuronidation, minoxidil sulfation,p-toluidine N-acetylation, 2,3-dihydroxynaphthalene methylation, and1-chloro-2,4-dinitrobenzene glutathione conjugation (Fig. 4). Glucuroni-dation rate was linear over 24 hours of incubation (Fig. 4A), but reactionrates of minoxidil sulfation (Fig. 4B), p-toluidine N-acetylation (Fig. 4C),

and 2,3-dihydroxynaphthalene methylation (Fig. 4D) started to decreaseafter about 4 hours. A similar trend was observed for 1-chloro-2,4-dinitrobenzene glutathione conjugation (Fig. 4E), where reaction de-celerated after the first hour of incubation. These results indicate thatactivity rates for 24-hour incubations, expressed as pmol·mg skin–1·h–1,are underestimated for N-acetylation, catechol methylation, sulfation, andglutathione conjugation. Since most reactions were linear up to 4 hours,incubations within this time window are more suitable for experiments inwhich linearity is important—for example, enzyme kinetic assays.Distribution of formed phase II metabolites differed between the

skin explant and corresponding incubation medium over time (Fig. 4).Within initial time points (1–2 hours), compared with the surroundingmedium, larger quantities of metabolites were observed in the skinexplant. In contrast, at later time points (4–24 hours), most of themetabolites were found in the incubation medium. Considering thatmany phase II metabolites are hydrophilic and polar compounds,unlikely to cross the cell membrane passively, these results suggest thathuman skin actively excretes the newly formed phase II metabolites.Leakage from damaged cells into the medium over assay time, another

Fig. 2. Histology of formalin-fixed paraffin-embedded full-thickness skin explantsimmediately after skin arrival (A) and after 24 hours of culture in Williams’ Emedium (B). Paraffin slides were stained with hematoxylin and eosin.

Fig. 3. (A) Activity rates of phase II metabolic reactions in fresh full-thickness humanskin explants (log10 scale). Test substrates were assayed for 15–24 hours, dependingon the individual skin samples. Reactions were performed in duplicate. (B) Apparentsubstrate depletion during 24 hours of incubation (linear scale). Black circles representindividual skin donors, whereas horizontal lines are the average values. Substrateconcentration in incubations was 10 mM. For (B), data are reconstructed based onmetabolite formation, assuming that substrate is consumed only in the indicatedmetabolic reaction. See Table 6 for additional details. 4-MU, 4-methylumbelliferone.

132 Manevski et al.

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


hypothesis leading to predominantly extracellular localization ofmetabolites, was unsupported by the unaltered microscopic mor-phology of the skin explants during the 24-hour period of incubation(Fig. 2).Inclusion of Cosubstrates in Human Skin Explant Culture. In an

effort to optimize incubation conditions for the maximal enzymeactivities, we tested the effects of cosubstrate inclusion in the mediumon activities of phase II metabolic reactions in human skin explants(Fig. 5). Addition of 1 mM UDPGA, PAPS, AcCoA, and SAM did notsignificantly enhance the reaction rates of glucuronidation, sulfation,N-acetylation, and catechol methylation, respectively. Similarly, theinclusion of 1 mM D-saccharolactone, an inhibitor of b-glucuronidase(Oleson and Court, 2008), did not affect the glucuronidation rate of4-methylumbelliferone. Since conjugation with glutathione may alsoproceed as a nonenzymatic reaction (Satoh, 1995), the assays were notperformed in a medium supplemented with glutathione.Influence of Substrate Physicochemical Properties on Human

Skin Phase II Metabolic Rates. To investigate whether physicochem-ical properties of the substrates influence phase II metabolic activities inhuman skin, we plotted log D (pH 7.4), molecular weight, solvent-accessible surface area, and polar surface area against average measuredmetabolic rates (Fig. 6). Activities were generally higher for compoundswith smaller molecular weight (Fig. 6B; one-phase decay fit; r2 = 0.65),solvent-accessible surface area below 300 Å2 (Fig. 6C; one-phase decayfit; r2 = 0.52), and polar surface area below 50 Å2 (Fig. 6D; one-phasedecay fit; r2 = 0.65). Activity correlations with substrates’ log D at pH7.4 (Fig. 6A) and pKa or pKb (not shown) were inconclusive, perhapsdue to the limited number of substrates tested.Effects of Rapid Skin Freezing on Glucuronidation Activity in

Human Skin Explants. To examine whether a freeze-thaw cycleinfluences the metabolic activity of human skin samples, we snap frozefull-thickness human skin in liquid nitrogen, kept it frozen for 7 days at–80�C, thawed at room temperature, and then tested 4-methylumbelli-ferone glucuronidation. An MTT viability assay was also performed inparallel (Fig. 7). In contrast to activities found in the fresh skin explantsfrom the same donor (Fig. 3), results showed that a single freeze-thawcycle diminished the 4-methylumbelliferone glucuronidation below thedetection limit (Fig. 7A). Viability assay with MTT dye showed thatremaining skin viability was below 10% of the control fresh sample(Fig. 7B).

Skin and Liver Metabolism Comparison. To estimate the extentof skin metabolism and to compare it to the liver, measured activitiesin human skin and corresponding reactions in human liver were scaledup to in vivo intrinsic clearances as well as to the corresponding organclearances (see Materials and Methods).In-depth analysis revealed that the liver intrinsic clearances

(CLint;liver) for all tested compounds and reactions were significantlyhigher than the corresponding predicted skin values (CLint;skin) (Fig.8A). The ratios between CLint;liver and CLint;skin ranged from 11-fold(17b-estradiol-3-sulfation) to more than 160,000-fold (diclofenacglucuronidation) (Fig. 8B; Table 7). These large differences inintrinsic clearances evidently also translate into organ clearances.However, since liver organ clearances for a number of reactions arelimited by the hepatic blood flow, whereas all skin organ clearanceswere below the skin blood flow, differences between liver and skinorgan clearances are somewhat smaller (from 11- to 20,000-fold).

Discussion

Human skin is often exposed to therapeutic drugs and cosmeticingredients, either after topical administration or due to distributionfrom the systemic circulation. Recent gene expression studies identifieda number of transferases in the human skin (Table 1), enzymes that areimportant for metabolic elimination and detoxification. Although usefulfor approximating the protein levels, studies of mRNA expression areoften limited by poor correlation between measured mRNA levels andactual protein levels (Izukawa et al., 2009; Maier et al., 2009; Oda et al.,2012; Vogel and Marcotte, 2012). Proteomic profiling may solve thisproblem, but a recent large study failed to detect UGTs and NATs inwhole human skin (van Eijl et al., 2012), despite previous reports oftheir activities (Table 2). Cutaneous enzyme activities were rarely mea-sured in the past, usually with a single substrate or in an experimentalmodel that may poorly reflect the metabolic activity of the in vivohuman skin—for example, skin microsomes or reconstructed three-dimensional skin models (Table 2). Skin explants were also usedpreviously (Moss et al., 2000; Goebel et al., 2009; Zalko et al., 2011),but dermatomed to 230–500 mM, effectively minimizing the possiblecontribution of skin dermis to drug biotransformation.To contribute to a better understanding of phase II metabolism in

human skin, we investigated the biotransformation of 11 test substrates

TABLE 6

Metabolic rates measured in human skin explant model

Fresh full-thickness skin explants, 4 mm in diameter (surface area 12.5 mm2, weight 10–35 mg), were incubated for 15–24 hours (depending on the individual skin sample) with 500 ml of Williams’E medium containing 10 mM test substrate. Incubations were performed in duplicate. Metabolites were detected and quantified in both skin explant and corresponding medium. To calculate the activityrate for each donor, we divided the total amount of formed metabolite by the donor’s average skin explant weight and specific incubation time (pmol·mg skin–1·h–1).

Metabolic Reaction No. Donors Mean Activity Rate S.D. Coefficient of Variation Activity Range (Min–Max) Fold Difference (Max/Min)

pmol·mg skin–1·h–1 % pmol·mg skin–1·h–1

4-Methylumbelliferone glucuronidation 13 2.892 1.913 66.14 0.722–8.217 11.3717b-Estradiol 3-sulfation 3 0.180 0.136 75.74 0.077–0.334 3.3417b-Estradiol 3-glucuronidation 3 0.054 0.034 62.17 0.018–0.085 4.7217b-Estradiol 17-glucuronidation 3 0.006 0.001 18.25 0.005–0.007 1.4017b-Estradiol total rate 0.240Indomethacin glucuronidation 3 0.274 0.091 33.12 0.208–0.377 1.81Diclofenac glucuronidation 3 0.055 0.066 120.71 0.010–0.130 13.00Triclosan glucuronidation 3 1.950 1.387 71.11 0.905–3.524 3.89Triclosan sulfation 3 1.628 0.249 15.28 1.367–1.863 1.36Triclosan total rate 3.578Minoxidil sulfation 7 0.013 0.007 55.27 0.007–0.026 3.71p-Toluidine N-acetylation 5 4.758 0.745 15.66 4.055–5.794 1.43Procainamide N-acetylation 3 0.036 0.008 23.17 0.029–0.045 1.554-Nitrocatechol methylation 3 1.146 0.421 36.72 0.662–1.429 2.162,3-Dihydroxynaphthalene methylation 5 1.778 0.787 44.27 0.710–2.567 3.611-Chloro-2,4-dinitrobenzene glutathione conjugation 3 0.127 0.062 48.83 0.059–0.183 3.10


at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


in fresh full-thickness skin explants (Fig. 1). Since it contains allrelevant cell types and has preserved skin morphology (Lebonvalletet al., 2010), we selected this model and performed a proof-of-conceptstudy to detect the activities of the five major phase II metabolismenzymes: UGTs, SULTs, NATs, COMT, and GSTs. Since phase Ireactions may also significantly contribute to cutaneous biotransforma-tion, activities of cytochrome P450s and other non–cytochrome P450phase I enzymes in human skin explants need to be addressed ina separate study. To investigate interindividual variability, assays wereperformed with 17 skin donors in total, and each individual substratewas tested in a minimum of three donors. Combined with specific andsensitive bioanalytics (Table 4), this study offered a unique advantage indetecting and quantifying novel metabolites formed in human skin. Onthe other hand, due to limited availability of fresh human skin, as wellas the potential differences between extra- and intracellular concen-trations of test compounds in skin explants (generally unsuitable forenzyme kinetic assays), incubations were limited to a single substrateconcentration of 10 mM.Results reveal that human skin explants metabolized all tested

substrates (Fig. 3; Table 6), indicating a potential for metabolicelimination and detoxification of diverse chemical structures. Activitiesmeasured for the same phase II reaction varied widely depending on thetest substrate. As an example, glucuronidation rate was significantly

higher for 4-methylumbelliferone compared with 17b-estradiol. Similartrends were observed for N-acetylation of p-toluidine (high) andprocainamide (low), as well as for the sulfation of triclosan (high) andminoxidil (low). Rates of catechol methylation, for both 4-nitrocatecholand 2,3-dihydroxynaphthalene, and glutathione conjugation of 1-chloro-2,4-dinitrobenzene were generally high, especially during the first fewhours of the incubation (Figs. 3 and 4). Taken together, even if relativeranking of cutaneous reaction rates differ at different concentrations oftest substrates, results suggest that considerable turnover may be ex-pected under favorable conditions. Analysis of the substrates’ phy-sicochemical properties revealed that higher activity is generally favoredfor smaller and less polar substrates (Fig. 6), possibly because thesecompounds are more likely to penetrate the skin cells. Although thisanalysis may indicate potential trends, cutaneous metabolism andpermeation data for more compounds are needed to generate predictivemodels—for example, quantitative structure-activity relationships.Interindividual variation of enzyme activities between skin donors

was from 1.4- to 13.0-fold, depending on the test substrate, beinghighest for the 4-methylumbelliferone and diclofenac glucuronidation(Table 6). Considering that skin donors were of different ages, ethnicities,genders, and lifestyles, and that skin samples from different anatomicregions were used (Table 3), the observed cutaneous interindividualvariability was moderately low, especially compared with variability

Fig. 4. Time-course of 4-methylumbelliferone glucur-onidation (A), minoxidil sulfation (B), p-toluidineN-acetylation (C), 2,3-dihydroxynaphthalene methylation(D), and 1-chloro-2,4-dinitrobenzene glutathione conju-gation (E) in fresh full-thickness human skin explants.Reactions were performed in duplicate. Incubation timewas from 1 to 24 hours. Results are presented as themean value and S.D. In addition to total metaboliteformed (circles), each panel also presents metabolitequantified in either skin explants (squares) or corre-sponding incubation medium (triangles).

134 Manevski et al.

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


reported in hepatic glucuronidation (Court, 2010), N-acetylation (Simet al., 2008), or glutathione conjugation (Temellini et al., 1995). Thisfinding may reflect one of the skin’s principal physiologic roles asa robust barrier against xenobiotics. The sole exception is given by themale eyelid skin sample, which exhibited 3-fold above-average glu-curonidation rates. Since a large majority of the skin donors were female(88%) and only a single eyelid skin sample was available, further

research is needed to investigate whether differences in gender oranatomic region play a major role in cutaneous drug metabolism.Glucuronidation rate was linear during 24 hours of incubation with

skin explants, but rates of sulfation, N-acetylation, catechol methyl-ation, and glutathione conjugation generally decelerated after 4 hours(Fig. 4), suggesting that an enzyme kinetic assay should be performedwith a shorter incubation time. Deviation from reaction rate linearity

Fig. 5. Influence of incubation conditions on4-methylumbelliferone glucuronidation (A), minoxidilsulfation (B), p-toluidine N-acetylation (C), and2,3-dihydroxynaphthalene methylation (D). Incubationtime was 24 hours. Reactions were performed in duplicate.Results are presented as the mean value and S.D. Dif-ferences between incubation conditions were statisticallyinsignificant. SL, D-saccharolactone.

Fig. 6. Influence of log D (pH 7.4 (A), molecularweight (B), solvent-accessible surface area (C), andpolar surface area (D) on phase II metabolic activities infresh full-thickness human skin explants. Calculatedphysicochemical properties of the test substrates arepresented in Table 5.


at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


may generally arise from the depletion of substrate and cosubstrate,inhibition by the reaction products, or both, especially in the case ofhigh-turnover reactions. Inclusion of cosubstrates for phase II enzymes,namely UDPGA, PAPS, AcCoA, and SAM, did not significantlymodify the reaction rates (Fig. 5), however, presumably because thesehighly polar molecules cannot easily penetrate the skin cells. On theother hand, accumulation of metabolites in the medium during the time-course experiments (Fig. 4), not the skin explants, suggested that phaseII metabolites are actively excreted from the skin, diminishing theproduct inhibition within the cells. This hypothesis is supported bya recent finding that human skin expresses a number of ATP-bindingcassette efflux transporters, particularly multidrug resistance–associatedprotein 1 (MRP1/ABCC1) (Osman-Ponchet et al., 2014).Rapid freezing of the human skin before incubations had a strongly

negative effect on both metabolic activity, as measured by the4-methylumbelliferone glucuronidation assay, and general skin explantviability (Fig. 7). Even if some previous studies reported activities inmicrosomes prepared from frozen (Gotz et al., 2012) or cadaver humanskin (Ademola et al., 1993), the results presented here suggest that usingfresh human skin is of high importance for drug metabolism studies. Ifthe frozen full-thickness skin samples are used as an alternative to fresh

tissue, further studies are needed to define the optimal and fullycontrolled freezing conditions, similar to ongoing efforts in preparingcryopreserved liver and kidney tissue slices [Fahy et al. (2013) andreferences therein].Although few phase II biotransformation reactions were reported

in the skin models previously (Table 2), such as in the case of4-methylumbelliferone (Jackh et al., 2011; Gotz et al., 2012), minoxidil(Anderson et al., 1998), triclosan (Moss et al., 2000), and p-toluidine

Fig. 7. Influence of rapid skin freezing on 4-methylumbellifrone glucuronidation(A) and skin viability in MTT assay (B). Incubation times for metabolism and MTTassays were 24 and 2 hours, respectively. Reactions were performed in duplicate.Results are presented as the mean value and S.D. To prepare explants with skin afterfreezing, full-thickness skin from the same donor was snap-frozen in liquid N2, keptat –80�C for 7 days, and then thawed at room temperature. LOQ, lower limit ofquantification; N.A., non-quantifiable amount of analyte.

Fig. 8. (A) Comparison between intrinsic organ clearances predicted for human skin(CLint;skin) and liver (CLint;liver). (B) Fold difference between predicted CLint;liver andCLint;skin. Results are presented as the mean value 6 propagated S.D. error. CLint;liverwas calculated based on enzyme kinetics parameters available from the literature(Supplemental Table 1). 4-MU, 4-methylumbelliferone.

136 Manevski et al.

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from



(Gotz et al., 2012), this study detected several novel metabolites. Asan example, our study was the first to identify N-acetylprocainamide,as well as the methoxy derivatives of 4-nitrocatechol and 2,3-dihydroxynaphthalene. Interestingly, we also found acyl glucuronidesof indomethacin and diclofenac (Fig. 3; Table 6), two nonsteroidalanti-inflammatory drugs that are commonly administered both topicallyand orally. Because acyl glucuronides may be chemically reactive(Regan et al., 2010; Boelsterli and Ramirez-Alcantara, 2011; Dickinson,2011), this finding has potential implications for skin toxicity, such asskin rash, an adverse effect observed for numerous therapeutic drugs(Mayorga et al., 2009; Chanprapaph et al., 2014). A more recent studyalso reported that the formation of 12-hydroxynevirapine sulfate inducedskin rash in rats (Sharma et al., 2013). The present results revealed thatminoxidil, triclosan, and 17b-estradiol are sulfated in human skinexplants (Fig. 3; Table 6). Appearance of acyl glucuronides and sulfatesin the human skin explants suggests that even low metabolic rates maybe relevant for pharmacotherapy and cosmetic research, if the formedmetabolites are implicated in adverse drug reactions.Since previous skin metabolism studies used different experimental

models (e.g., split-thickness explants, cells, or subcellular fractions),a wide range of test substrate concentrations [e.g., 500 mM of4-methylumbelliferone in Jäckh et al. (2011) versus 10 mM in thepresent study], or even different administration routes [topicaltriclosan in Moss et al. (2000) versus media inclusion in the presentstudy], direct comparison of the measured activity rates is challenging.If substrate depletion only is taken into account (i.e., conversion ratepercentage), our results on triclosan glucuronidation and sulfationcorrelate well with the values published previously (Moss et al., 2000).Comparison of the skin and liver organ clearances implies that skin

contributes only marginally to total systemic phase II enzyme-mediated metabolic elimination in humans. Considering the significantlower intrinsic metabolic activities in skin and the fact that human skinflow is approximately 5-fold lower than the corresponding liver bloodflow [300 vs. 1450 ml·min–1 for a 70-kg individual (Davies and Morris,1993)], this is not really surprising. However, it has to be noted herethat, for the current assessment, a series of assumptions were madewhich underlie the high degree of uncertainty which accompanies thisapproach. For example, our efforts might be hampered by 1) lack of anappropriate physiologically based skin model, 2) possible permeability

limitation of intrinsic clearance in skin explants, 3) high inherentvariability of skin blood flow (Petrofsky, 2012), 4) uncertainty of thedrug’s unbound fraction in skin, and 5) almost complete lack of suitablein vivo skin metabolism data for model validation. Besides that, currentassessment exclusively focuses on the systemic elimination effect ofskin metabolism. However, a cutaneous metabolic first-pass effect maysignificantly limit the bioavailability of topically and transdermallyapplied drugs (Zhang et al., 2009). However, the magnitude of theskin’s first-pass effect, besides the specific enzymatic activities, willalso depend on numerous factors such as the applied dose, ad-ministration surface area, skin thickness, drug partitioning, drugdiffusion and permeability in the skin, potential active transportprocesses, local skin blood flow, and even metabolite diffusioncoefficient in the skin (Boderke et al., 2000). As a result of theaforementioned uncertainties, accurate predictions of both skin’stotal metabolic clearance and cutaneous first-pass effect need to bestudied more in depth. The authors hope that the results in skinexplants presented here will stimulate further research in the field ofcutaneous drug metabolism, especially focusing on in vitro–in vivocorrelations, first-pass effect, phase I metabolic reactions, andcross-validation of the available skin models.In conclusion, results demonstrate that human skin has a significant

potential for phase II metabolism through reactions of glucuronidation,sulfation, N-acetylation, catechol methylation, and glutathione conju-gation. Even if the skin’s capacity for phase II biotransformationappears low compared with the liver, cutaneous metabolism remainsa relevant process for topical and transdermal drug administration,cosmetics, the study of adverse drug reactions in the skin, as well asfor all drugs that are administrated systemically but targeted for dermalindications. Moreover, full-thickness human skin explants are an ap-propriate experimental model to study these reactions, especially ifidentification of novel skin metabolites is needed. Although the primaryrole of phase II reactions is metabolic elimination and detoxification,formation of acyl glucuronides and sulfates may contribute to the skintoxicity. As shown by several examples, reaction rates may be high,depending on the chemical structure of the drug and expression levelsof the individual metabolic enzymes. These findings suggest thatskin metabolism should be considered during drug and cosmeticsdevelopment.

TABLE 7

Predicted human skin and liver organ clearances for the studied phase II metabolic reactions

Data are sorted by descending order of CLint;skinvalues. See Materials and Methods and Supplemental Table 1 for further details. Results are presented as thecalculated average values 6 propagated S.D.

Predicted Clearances

Intrinsic Organ Clearances Systemic Organ Clearances

CLint;skin CLint;liver CLskin CLliver

Metabolic Reaction ml·min–1

p-Toluidine N-acetylation 53.6 6 6.4 n.a. 45.4 6 4.6 n.a.Triclosan sulfation 30.7 6 4.7 1638 27.8 6 3.9 7694-Methylumbelliferone glucuronidation 19.3 6 12.7 38,255 6 33,100 17.7 6 10.7 1368 6 682,3-Dihydroxynaphthalene methylation 19.1 6 8.4 n.a. 17.8 6 6.5 n.a.Triclosan glucuronidation 13.0 6 9.2 652 12.3 6 8.4 4504-Nitrocatechol methylation 12.3 6 4.5 n.a. 11.8 6 4.2 n.a.17b-Estradiol 3-sulfation 3.39 6 2.57 37.1 3.34 6 2.50 36.31-Chloro-2,4-dinitrobenzene glutathione conjugation 2.34 6 1.14 99,285 6 56,038 2.32 6 1.12 1429Indomethacin glucuronidation 1.83 6 0.61 2010 6 1166 1.81 6 0.60 805 6 217Diclofenac glucuronidation 0.364 6 0.440 58,791 6 54,473 0.363 6 0.438 1292 6 22817b-Estradiol 3-glucuronidation 0.363 6 0.225 2195 6 2029 0.362 6 0.225 765 6 290Minoxidil sulfation 0.338 6 0.186 32.6 0.337 6 0.185 31.9Procainamide N-acetylation 0.419 6 0.097 28.0 0.418 6 0.097 27.417b-Estradiol 17-glucuronidation 0.041 6 0.007 2487 6 2230 0.041 6 0.007 808 6 288

CLliver, human liver organ clearance; CLskin, human skin organ clearance; n.a., literature data were not available.


at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from



Acknowledgments

The authors thank Marie-Catherine Stutz, Karine Bigot, Arno Doelemeyer,and Armelle Grevot for their help and support with the skin histology. Theauthors thank Bertrand-Luc Birlinger, Judith Streckfuss, Maxime Garnier, andArnold Demailly for providing reagents and technical assistance. The authorsalso thank Kenichi Umehara for his support and discussion regarding skinclearance evaluation. In addition, the authors thank the postdoctoral office atthe Novartis Institutes for BioMedical Research, Novartis Pharma.

Authorship ContributionsParticipated in research design:Manevski, Litherland, Swart, Balavenkatraman,

Ashton-Chess.Conducted experiments: Manevski, Bertschi, Ling.Contributed new reagents or analytic tools: Manevski, Bertschi.Performed data analysis: Manevski.Wrote or contributed to the writing of the manuscript:Manevski, Litherland,

Swart, Camenisch, Walles, Pognan, Schiller, Balavenkatraman, Ling, Kretz,Wettstein, Itin, Schaefer, Wolf.

References

Ademola JI, Wester RC, and Maibach HI (1993) Metabolism of 3-indolylacetic acid duringpercutaneous absorption in human skin. J Pharm Sci 82:150–154.

Anderson RJ, Kudlacek PE, and Clemens DL (1998) Sulfation of minoxidil by multiple humancytosolic sulfotransferases. Chem Biol Interact 109:53–67.

Bamshad J, Lerner AB, and McGuire JS (1964) Catechol O-Methyl Transferase in Skin. J InvestDermatol 43:111–113.

Bamshad J, Serot DI, and Szoko EW (1970) Catechol-O-methyl transferase (COMT) in skin ofpatients with psoriasis. J Invest Dermatol 55:147–148.

Blacker KL, Olson E, Vessey DA, and Boyer TD (1991) Characterization of glutathioneS-transferase in cultured human keratinocytes. J Invest Dermatol 97:442–446.

Boderke P, Schittkowski K, Wolf M, and Merkle HP (2000) Modeling of diffusion and con-current metabolism in cutaneous tissue. J Theor Biol 204:393–407.

Boelsterli UA and Ramirez-Alcantara V (2011) NSAID acyl glucuronides and enteropathy. CurrDrug Metab 12:245–252.

Bonifas J, Hennen J, Dierolf D, Kalmes M, and Blömeke B (2010b) Evaluation of cytochromeP450 1 (CYP1) and N-acetyltransferase 1 (NAT1) activities in HaCaT cells: implications forthe development of in vitro techniques for predictive testing of contact sensitizers. Toxicol InVitro 24:973–980.

Bonifas J, Scheitza S, Clemens J, and Blömeke B (2010a) Characterization of N-acetyltransferase1 activity in human keratinocytes and modulation by para-phenylenediamine. J Pharmacol ExpTher 334:318–326.

Borg N, Götharson E, Benfeldt E, Groth L, and Ståhle L (1999) Distribution to the skin ofpenciclovir after oral famciclovir administration in healthy volunteers: comparison of thesuction blister technique and cutaneous microdialysis. Acta Derm Venereol 79:274–277.

Brown RP, Delp MD, Lindstedt SL, Rhomberg LR, and Beliles RP (1997) Physiological pa-rameter values for physiologically based pharmacokinetic models. Toxicol Ind Health 13:407–484.

Brunner M, Stabeta H, Möller JG, Schrolnberger C, Erovic B, Hollenstein U, Zeitlinger M,Eichler HG, and Müller M (2002) Target site concentrations of ciprofloxacin after singleintravenous and oral doses. Antimicrob Agents Chemother 46:3724–3730.

Camenisch G and Umehara K (2012) Predicting human hepatic clearance from in vitro drugmetabolism and transport data: a scientific and pharmaceutical perspective for assessing drug-drug interactions. Biopharm Drug Dispos 33:179–194.

Chanprapaph K, Vachiramon V, and Rattanakaemakorn P (2014) Epidermal growth factor re-ceptor inhibitors: a review of cutaneous adverse events and management. Dermatol Res Pract2014:734249.

Court MH (2010) Interindividual variability in hepatic drug glucuronidation: studies into the roleof age, sex, enzyme inducers, and genetic polymorphism using the human liver bank asa model system. Drug Metab Rev 42:209–224.

Davies B and Morris T (1993) Physiological parameters in laboratory animals and humans.Pharm Res 10:1093–1095.

Dickinson RG (2011) Iso-glucuronides. Curr Drug Metab 12:222–228.Dooley TP, Haldeman-Cahill R, Joiner J, and Wilborn TW (2000) Expression profiling of humansulfotransferase and sulfatase gene superfamilies in epithelial tissues and cultured cells. Bio-chem Biophys Res Commun 277:236–245.

Fahy GM, Guan N, de Graaf IA, Tan Y, Griffin L, and Groothuis GM (2013) Cryopreservation ofprecision-cut tissue slices. Xenobiotica 43:113–132.

Goebel C, Hewitt NJ, Kunze G, Wenker M, Hein DW, Beck H, and Skare J (2009) Skinmetabolism of aminophenols: human keratinocytes as a suitable in vitro model to qualitativelypredict the dermal transformation of 4-amino-2-hydroxytoluene in vivo. Toxicol Appl Phar-macol 235:114–123.

Götz C, Pfeiffer R, Tigges J, Ruwiedel K, Hübenthal U, Merk HF, Krutmann J, Edwards RJ, AbelJ, and Pease C, et al. (2012) Xenobiotic metabolism capacities of human skin in comparisonwith a 3D-epidermis model and keratinocyte-based cell culture as in vitro alternatives forchemical testing: phase II enzymes. Exp Dermatol 21:364–369.

Gundert-Remy U, Bernauer U, Blömeke B, Döring B, Fabian E, Goebel C, Hessel S, Jäckh C,Lampen A, and Oesch F, et al. (2014) Extrahepatic metabolism at the body’s internal-externalinterfaces. Drug Metab Rev 46:291–324.

Habig WH, Pabst MJ, and Jakoby WB (1974) Glutathione S-transferases. The first enzymatic stepin mercapturic acid formation. J Biol Chem 249:7130–7139.

Hayes JD, Flanagan JU, and Jowsey IR (2005) Glutathione transferases. Annu Rev PharmacolToxicol 45:51–88.

Higashi Y, Fuda H, Yanai H, Lee Y, Fukushige T, Kanzaki T, and Strott CA (2004) Expression ofcholesterol sulfotransferase (SULT2B1b) in human skin and primary cultures of human epi-dermal keratinocytes. J Invest Dermatol 122:1207–1213.

Hiraoka H, Yamamoto K, Miyoshi S, Morita T, Nakamura K, Kadoi Y, Kunimoto F,and Horiuchi R (2005) Kidneys contribute to the extrahepatic clearance of propofol in humans,but not lungs and brain. Br J Clin Pharmacol 60:176–182.

Houston JB and Galetin A (2008) Methods for predicting in vivo pharmacokinetics using datafrom in vitro assays. Curr Drug Metab 9:940–951.

Hu T, Khambatta ZS, Hayden PJ, Bolmarcich J, Binder RL, Robinson MK, Carr GJ, Tiesman JP,Jarrold BB, and Osborne R, et al. (2010) Xenobiotic metabolism gene expression in theEpiDermin vitro 3D human epidermis model compared to human skin. Toxicol In Vitro 24:1450–1463.

Itäaho K, Mackenzie PI, Ikushiro S, Miners JO, and Finel M (2008) The configuration of the 17-hydroxy group variably influences the glucuronidation of beta-estradiol and epiestradiol byhuman UDP-glucuronosyltransferases. Drug Metab Dispos 36:2307–2315.

Ito K and Houston JB (2004) Comparison of the use of liver models for predicting drug clearanceusing in vitro kinetic data from hepatic microsomes and isolated hepatocytes. Pharm Res 21:785–792.

Iwatsubo T, Hirota N, Ooie T, Suzuki H, Shimada N, Chiba K, Ishizaki T, Green CE, Tyson CA,and Sugiyama Y (1997) Prediction of in vivo drug metabolism in the human liver from in vitrometabolism data. Pharmacol Ther 73:147–171.

Izukawa T, Nakajima M, Fujiwara R, Yamanaka H, Fukami T, Takamiya M, Aoki Y, Ikushiro S,Sakaki T, and Yokoi T (2009) Quantitative analysis of UDP-glucuronosyltransferase (UGT)1A and UGT2B expression levels in human livers. Drug Metab Dispos 37:1759–1768.

Jäckh C, Blatz V, Fabian E, Guth K, van Ravenzwaay B, Reisinger K, and Landsiedel R (2011)Characterization of enzyme activities of Cytochrome P450 enzymes, Flavin-dependent mon-ooxygenases, N-acetyltransferases and UDP-glucuronyltransferases in human reconstructedepidermis and full-thickness skin models. Toxicol In Vitro 25:1209–1214.

Kawakubo Y, Merk HF, Masaoudi TA, Sieben S, and Blömeke B (2000) N-Acetylation ofparaphenylenediamine in human skin and keratinocytes. J Pharmacol Exp Ther 292:150–155.

King C, Tang W, Ngui J, Tephly T, and Braun M (2001) Characterization of rat and human UDP-glucuronosyltransferases responsible for the in vitro glucuronidation of diclofenac. Toxicol Sci61:49–53.

Klimowicz A, Nowak A, and Kadyków M (1988) Plasma and skin blister fluid concentrations oftrimethoprim following its oral administration. Eur J Clin Pharmacol 34:377–380.

Krishna G, Beresford E, Ma L, Vickery D, Martinho M, Yu X, Komjathy S, and Tavakkol A(2010) Skin concentrations and pharmacokinetics of posaconazole after oral administration.Antimicrob Agents Chemother 54:1807–1810.

Kuehl GE, Lampe JW, Potter JD, and Bigler J (2005) Glucuronidation of nonsteroidal anti-inflammatory drugs: identifying the enzymes responsible in human liver microsomes. DrugMetab Dispos 33:1027–1035.

Kushida A, Hattori K, Yamaguchi N, Kobayashi T, Date A, and Tamura H (2011) Sulfation ofestradiol in human epidermal keratinocyte. Biol Pharm Bull 34:1147–1151.

Lebonvallet N, Jeanmaire C, Danoux L, Sibille P, Pauly G, and Misery L (2010) The evolutionand use of skin explants: potential and limitations for dermatological research. Eur J Dermatol20:671–684.

Lobo RA, Paul WL, Gentzschein E, Serafini PC, Catalino JA, Paulson RJ, and Horton R (1987)Production of 3 alpha-androstanediol glucuronide in human genital skin. J Clin EndocrinolMetab 65:711–714.

Lu Z, Hasse S, Bodo E, Rose C, Funk W, and Paus R (2007) Towards the development ofa simplified long-term organ culture method for human scalp skin and its appendages underserum-free conditions. Exp Dermatol 16:37–44.

Luu-The V, Duche D, Ferraris C, Meunier JR, Leclaire J, and Labrie F (2009) Expression profilesof phases 1 and 2 metabolizing enzymes in human skin and the reconstructed skin modelsEpiskin and full thickness model from Episkin. J Steroid Biochem Mol Biol 116:178–186.

Maier T, Güell M, and Serrano L (2009) Correlation of mRNA and protein in complex biologicalsamples. FEBS Lett 583:3966–3973.

Manevski N, Troberg J, Svaluto-Moreolo P, Dziedzic K, Yli-Kauhaluoma J, and Finel M (2013)Albumin stimulates the activity of the human UDP-glucuronosyltransferases 1A7, 1A8, 1A10,2A1 and 2B15, but the effects are enzyme and substrate dependent. PLoS ONE 8:e54767.

Mayorga C, Torres MJ, Fernandez J, Canto G, and Blanca M (2009) Cutaneous symptoms in drugallergy: what have we learnt? Curr Opin Allergy Clin Immunol 9:431–436.

Moss T, Howes D, and Williams FM (2000) Percutaneous penetration and dermal metabolism oftriclosan (2,4, 49-trichloro-29-hydroxydiphenyl ether). Food Chem Toxicol 38:361–370.

Nishimuta H, Houston JB, and Galetin A (2014) Hepatic, intestinal, renal, and plasma hydrolysisof prodrugs in human, cynomolgus monkey, dog, and rat: implications for in vitro-in vivoextrapolation of clearance of prodrugs. Drug Metab Dispos 42:1522–1531.

Oda S, Nakajima M, Hatakeyama M, Fukami T, and Yokoi T (2012) Preparation of a specificmonoclonal antibody against human UDP-glucuronosyltransferase (UGT) 1A9 and evaluationof UGT1A9 protein levels in human tissues. Drug Metab Dispos 40:1620–1627.

Oleson L and Court MH (2008) Effect of the beta-glucuronidase inhibitor saccharolactone onglucuronidation by human tissue microsomes and recombinant UDP-glucuronosyltransferases.J Pharm Pharmacol 60:1175–1182.

Osman-Ponchet H, Boulai A, Kouidhi M, Sevin K, Alriquet M, Gaborit A, Bertino B, Comby P,and Ruty B (2014) Characterization of ABC transporters in human skin. Drug Metabol DrugInteract 29:91–100.

Petrofsky JS (2012) Resting blood flow in the skin: does it exist, and what is the influence oftemperature, aging, and diabetes? J Diabetes Sci Tech 6:674–685.

Prusakiewicz JJ, Harville HM, Zhang Y, Ackermann C, and Voorman RL (2007) Parabens inhibithuman skin estrogen sulfotransferase activity: possible link to paraben estrogenic effects.Toxicology 232:248–256.

Raza H, Awasthi YC, Zaim MT, Eckert RL, and Mukhtar H (1991) Glutathione S-transferases inhuman and rodent skin: multiple forms and species-specific expression. J Invest Dermatol 96:463–467.

Regan SL, Maggs JL, Hammond TG, Lambert C, Williams DP, and Park BK (2010) Acylglucuronides: the good, the bad and the ugly. Biopharm Drug Dispos 31:367–395.

Rittmaster RS, Zwicker H, Thompson DL, Konok G, and Norman RW (1993) Androstanediolglucuronide production in human liver, prostate, and skin. Evidence for the importance of theliver in 5 alpha-reduced androgen metabolism. J Clin Endocrinol Metab 76:977–982.

138 Manevski et al.

at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from


Rugstad HE and Dybing E (1975) Glucuronidation in cultures of human skin epithelial cells. EurJ Clin Invest 5:133–137.

Satoh K (1995) The high non-enzymatic conjugation rates of some glutathione S-transferase(GST) substrates at high glutathione concentrations. Carcinogenesis 16:869–874.

Sharma AM, Novalen M, Tanino T, and Uetrecht JP (2013) 12-OH-nevirapine sulfate, formed inthe skin, is responsible for nevirapine-induced skin rash. Chem Res Toxicol 26:817–827.

Sim E, Lack N, Wang CJ, Long H, Westwood I, Fullam E, and Kawamura A (2008) ArylamineN-acetyltransferases: structural and functional implications of polymorphisms. Toxicology 254:170–183.

Sollinger HW, VernonWB, D’Alessandro AM, Kalayoglu M, Stratta RJ, and Belzer FO (1989) Combinedliver and pancreas procurement with Belzer-UW solution. Surgery 106:685–690, discussion 690–691.

Temellini A, Castiglioni M, Giuliani L, Mussi A, Giulianotti PC, Pietrabissa A, Angeletti CA,Mosca F, and Pacifici GM (1995) Glutathione conjugation with 1-chloro-2,4-dinitrobenzene(CDNB): interindividual variability in human liver, lung, kidney and intestine. Int J ClinPharmacol Ther 33:498–503.

van Eijl S, Zhu Z, Cupitt J, Gierula M, Götz C, Fritsche E, and Edwards RJ (2012) Elucidation ofxenobiotic metabolism pathways in human skin and human skin models by proteomic pro-filing. PLoS ONE 7:e41721.

Verbraecken J, Van de Heyning P, De Backer W, and Van Gaal L (2006) Body surface area innormal-weight, overweight, and obese adults. A comparison study. Metabolism 55:515–524.

Vogel C and Marcotte EM (2012) Insights into the regulation of protein abundance from pro-teomic and transcriptomic analyses. Nat Rev Genet 13:227–232.

Zalko D, Jacques C, Duplan H, Bruel S, and Perdu E (2011) Viable skin efficiently absorbs andmetabolizes bisphenol A. Chemosphere 82:424–430.

Zhang Y, Gonzalez V, and Xu MJ (2002) Expression and regulation of glutathione S-transferaseP1-1 in cultured human epidermal cells. J Dermatol Sci 30:205–214.

Zhang Q, Grice JE, Wang G, and Roberts MS (2009) Cutaneous metabolism in transdermal drugdelivery. Curr Drug Metab 10:227–235.

Zucchi A, Raho E, Marconi B, Nicoli S, Santini M, Allegra F, Colombo P, Bettini R, and Santi P(2001) Plasma and skin concentration of 5-methoxypsoralen in psoriatic patients after oraladministration. J Invest Dermatol 117:379–382.

Address correspondence to: Dr. Karine Litherland, Novartis Institutes forBioMedical Research, Translational Sciences, Drug Metabolism and Pharmaco-kinetics, Integrated Drug Disposition, Fabrikstrasse 14-1.02.7, Novartis PharmaAG, CH-4002 Basel, Switzerland. E-mail: [email protected]


at ASPE

T Journals on O

ctober 12, 2020dm


ownloaded from

mailto:[email protected]


phase ii metabolism in human skin: skin explants...

Documents