ORIGINAL ARTICLE
Coding and Non-Coding Polymorphisms in Alcohol Dehydrogenase Alters Protein Expression and Alcohol-Associated Erythema Lynn K Pershing1, Yuexian Chen1, Ariana N. Tkachuk2, Holly L. Rausch1, Kasia Petelenz-Rubin1, Judy L. Corlett1 and Maurine R. Hobbs2 Ethanol (EtOH), isopropyl alcohol (IPA), and propylene glycol (PG) increase topical drug delivery, but are sometimes associated with erythema. A potential genetic basis for alcohol-associated erythema was investigated as the function of polymorphisms in coding and non-coding regions of class IB alcohol dehydrogenase (ADHIB) and evaluated for altered gene expression in vitro and metabolic activity in vivo via altered skin blood flow (Doppler velocimeter) and erythema (reflectance colorimeter a*) following topical challenge to 5 M EtOH, IPA, PG, and butanol (ButOH). Promoter polymorphisms G-887A and C-739T and exon G143A form eight ADHIB haplotypes with different frequencies in Caucasians vs Asians and exhibit variable gene expression and metabolic activity. Polymorphisms C-739T and G-887A independently alter gene expression, which is further increased by IPA and PG, but not EtOH or ButOH. EtOH and ButOH increase erythema as a function of skin blood flow. IPA increases skin blood flow without erythema and PG increased erythema with decreased skin blood flow, all as a function of ADHIB haplotype. PG-induced erythema was uniquely associated with tumor necrosis factor-a expression. Thus, erythema following alcohol exposure is alcohol type specific, has a pharmacogenetic basis related to ADHIB haplotype and can be functionally evaluated via Doppler velocimetry and reflectance colorimetry in vivo. Journal of Investigative Dermatology (2008) 128, 616–627; doi:10.1038/sj.jid.5701105; published online 29 November 2007
INTRODUCTION Topical glucocorticosteroids are widely used in dermatology practice for the treatment of inflammatory dermatoses including psoriasis, dermatitis, and lichen planus. These products produce a blanching response in human skin, also known as the vasoconstrictor response, which has been used as a surrogate marker to compare potency, bioequivalence, and bioavailability of dermatologic products (Cornell and Stoughton, 1985; Stoughton, 1992; FDA Guidance, 1998). The skin blanching response is highly variable in humans. Dermatologic corticosteroid products often contain alcohols, such as propylene glycol (PG) and/or isopropanol (IPA) to increase drug solubility and enhance delivery of the drug into
1
Department of Dermatology, University of Utah, Salt Lake City, Utah, USA and 2Department of Internal Medicine, University of Utah, Salt Lake City, Utah, USA Correspondence: Professor Lynn K Pershing, Department of Dermatology, University of Utah, 1715 E Laird Ave., Salt Lake City, Utah 84108-1806, USA. E-mail:
[email protected] Abbreviations: ADH, alcohol dehydrogenase; ADHIB, class IB ADH; ButOH, butanol; EtOH, ethanol; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; IPA, isopropyl alcohol; LD, linkage disequilibrium; LDV, laser-Doppler velocimetry; NRE2, negative regulatory element 2; PG, propylene glycol; RT, room temperature; SNP, single-nucleotide polymorphism; TNF-a, tumor necrosis factor-a Received 8 March 2007; revised 20 June 2007; accepted 5 July 2007; published online 29 November 2007
616
Journal of Investigative Dermatology (2008), Volume 128
skin. Preliminary studies (Pershing et al., 1999 abstr.; Rausch et al., 2001 abstr.) demonstrate that commercial topical corticosteroid products containing alcohol(s) can either increase or decrease corticosteroid activity in human skin compared with the same drug concentration in a vehicle without alcohol(s). We hypothesized that alcohol and its intermediary metabolism may play a role in the resulting corticosteroid response. Alcohol dehydrogenase (ADH) is primarily responsible for alcohol(s) metabolism. There are seven ADH genes in the human genome, all located within 380 kb on chromosome 4q21–23 (Yasunami et al., 1990; Edenberg, 2000) with varying affinity and activity for alcohol(s), and a broad spectrum of other substrates including hydroxysteroids, retinol, and hydroxy fatty acids (Jornvall et al., 2000). Class I ADH is composed of three genes, ADHIA (a.k.a. ADH1), ADHIB (a.k.a. ADH2), and ADHIC (a.k.a. ADH3), which differ by single-nucleotide polymorphisms (SNPs; Yasunami et al., 1990). ADHIB and ADHIC are expressed in the differentiated viable epidermis, fibroblasts, endothelial cells (Cheung et al., 1999), as well as the upper third of keratinocytes around the hair shaft, and lymphocytes, but is undetectable in pigmented cells (unpublished personal observations). Class IB ADH (ADHIB) is known to contain a number of SNPs in both coding and non-coding regions of the gene (www.genome.utah.edu/genesnps/) whose frequencies vary as a function of Race (Roychoudhry and Masatoshi, 1988; & 2007 The Society for Investigative Dermatology
LK Pershing et al. ADHIB Haplotype and Skin Erythema
(Osier et al., 2002). The G143A SNP in exon 3 is associated with a nonsynonymous change of the amino acid from arginine to histidine (R47H). This SNP changes the pKa of the encoded protein from 8.5 to 10.0 that is associated with 40- to 100-fold increases in Km and Vmax of EtOH metabolism (Hurley et al., 1994; Edenberg, 2000). Increased alcohol metabolism produces more aldehyde, which causes vasodilation and is the accepted basis for the alcohol flushing response to EtOH in Asians (Wilkin and Fortner, 1986). We hypothesized that polymorphisms in class I ADH may be associated with the erythema observed following application of dermatologic glucocorticosteroid products containing alcohols. This study evaluates 10 SNPs in the coding and non-coding regions of ADHIB and ADHIC, establishes eight ADHIB haplotypes and their functional association with erythema and altered cutaneous blood flow in human skin in vivo following topical alcohol exposure. RESULTS SNPs in class I ADH have variable frequencies in Asians vs Caucasians
Evaluation of 1,000 bp upstream from the initiation site of ADHIB and ADHIC revealed three SNPs in ADHIB and one SNP in ADHIC. These four promoter SNPS, along with four SNPs in the coding region of ADH1B (ID125, OMIM 103720), and two SNPs in the coding region of ADHIC (ID126; OMIM 103730) were evaluated in the enrolled Caucasian (n ¼ 70) and Asian (n ¼ 19) subjects (www.genome. utah.edu/genesnps/). Frequency of SNPs in class I ADH genes are significantly different in Caucasians vs Asians (Table 1). The fast alcohol
metabolizing nonsynonymous G143A SNP discussed above is more frequent in Asians (35% heterozygous and 65% homozygous variant) than Caucasians (11% heterozygous variant and 89% wild type). Other SNPs in exon 3 of ADHIB (C171A and A178T) and in exon 5 (C1108T) were predominately wild type in both populations and therefore removed from haplotyping analysis. In contrast, Asians are primarily wild type (94%) at SNPs C-887A and T-752C in the promoter region of ADHIB, whereas Caucasians demonstrate higher frequency of the SNP (26% heterozygous and B19% homozygous variants) at both locations. Asians are either homozygous (94%) or heterozygote variant (6%) at C-739T in the 50 promoter, whereas Caucasians are primarily heterozygous variant (56%) or wild type (31%), but both populations demonstrate high LD with the G143A SNP in exon 3. SNP frequency in the 50 promoter, exons 6 and 8 in ADHIC are also different between the two populations. SNP frequencies of T-845C and G815A and A1048G in ADHIC are higher in Caucasians (B24, 69, and 69%, respectively) than Asians (6, 18, and 18% respectively). Many SNPs within class I ADH are in high linkage disequilibrium (LD); C-887A and T-752C in ADHIB have a D 0 value of 0.8 in Caucasians and 1.0 in Asians (Figure 1). All three SNPs in ADHIC (T-845C, G815A and A1048G) are in high LD with each other (D0 ¼ 1.0) in both populations. There is also high LD between particular SNPs between ADHIB and ADHIC in both populations. The ADHIB promoter SNP (C-739T) and exon 3 SNP (G143A) are in LD (D0 ¼ 1.0) with ADHIC promoter SNP (T-845C), exon 6 (G815A), and exon 8 (A1048G) SNPs in both Race groups.
Table 1. SNPs frequency in class I ADH: Caucasian vs Asian Caucasian1 Gene ADHIB
Nt3 C-887A
Aa4 # #
Asian2
*1/*15
*1/*2
*2/*2
*1/*1
*1/*2
*2/*2
0.57
0.24
0.19
0.94
0.06
0.00
0.55
0.27
0.18
0.94
0.06
0.00
0.31
0.56
0.13
0.00
0.06
0.94
ADHIB
T-752 C
ADHIB
C-739 T
ADHIB
G 143 A
R 48 H
0.89
0.11
0.00
0.00
0.35
0.65
ADHIB
C 171 A
N 57 K
1.00
0.00
0.00
1.00
0.00
0.00
ADHIB
A 178 T
T 60 S
0.95
0.05
0.00
1.00
0.00
0.00
ADHIB
C 1108 T
R 370 C
1.00
0.00
0.00
1.00
0.00
0.00
ADHIC
T -845 C
0.76
0.23
0.01
0.94
0.06
0.00
ADHIC
G 815 A
Q 271 R
0.31
0.50
0.19
0.82
0.18
0.00
ADHIC
A 1048 G
I 350 V
0.31
0.50
0.19
0.82
0.18
0.00
ADH, alcohol dehydrogenase; ADHIB, class IB ADH; SNP, single-nucleotide polymorphism. n=63. n=17. 3 nt=nucleotide. 4 aa=amino acid. 5 *allele 1=wild type, allele 2=contains SNP. #wild type (allele *1) is opposite of that listed in NCBI database. 1 2
www.jidonline.org
617
LK Pershing et al.
–887
–752
–752
–739
–739
143
143
–845
–845
815
815
1,048
1,048
0.0
0.1
0.2
0.3
0.4
0.5
0.7
0.8
0.9
IC
IC 1,048
ADH IB IC
815
IB
–845
nt
–887
IB
143
Gene IB
–739
–739
IC
–752
–752
IC
–887
nt
ADH IB IC
1,048
IB
815
IB
–845
IB
143
Gene
–887
ADHIB Haplotype and Skin Erythema
1.0
D′ Figure 1. LD between SNPs in ADHIB and ADHIC. (a) Caucasians, n ¼ 70. (b) Asians, n ¼ 19. Extent of LD assessed by D0 value ranging from highest (red) D0 ¼ 1.0 (100%) to lowest (black), D0 ¼ 0.0 (0%). D0 40.80 (80%) considered to be significant.
Table 2. ADHIB diplotype in Caucasian vs Asian populations Allele 1
Allele 2 Caucasian1 (%)
Asian2 (%)
G
4.3
0.0
21.4
0.0
44.3
0.0
17.1
0.0
ADHIB diplotype
Allele pairs
C-887A
C-739T
G143A
/
C-887A
C-739T
G143A
A
1/1
C
C
G
/
C
C
B C D
5/1
A
C
G
/
C
C
G
5/3
A
C
G
/
A
C
G
3/1
C
T
G
/
C
C
G
3/3
C
T
G
/
C
T
G
5/3
A
C
G
/
C
T
G
5/7
A
C
G
/
A
T
G
E
1/2
C
C
G
/
C
C
A
2.9
0.0
F
5/2
A
C
G
/
C
C
A
4.3
5.9
5/5
A
C
G
/
A
C
A 4.3
88.2
1.4
5.9
G H
3/2
C
T
G
/
C
C
A
3/4
C
T
G
/
C
T
A
5/4
A
C
G
/
C
T
A
ADHIB, class IB alcohol dehydrogenase. n=70, relative frequency of ADHIB haplotype (%). n=17, relative frequency of ADHIB haplotype (%).
1 2
ADHIB SNPs form haplotypes with different frequencies in Asians vs Caucasians
Given the high LD between the ADHIB promoter SNPs C-887A and T-752C, and the high LD between ADHIB and ADHIC, haplotyping was performed on three SNPS of ADHIB only: C-887A, C-739T, and G143A. Combining heterozygote and homozygote variants at each loci into a single haplotype, eight ADHIB haplotypes were identified in Caucasians and three haplotypes in Asians (Table 2). The majority of Caucasians were predominantly haplotype ADHIB C (44.3%), ADHIB B (21.4%), and ADHIB D (17.1%) with less frequency of haplotypes ADHIB A (4.3%), ADHIB F (4.3%), ADHIB G (4.3%), and ADHIB E (1.4%). Asians were 618
Journal of Investigative Dermatology (2008), Volume 128
primarily ADHIB G (88.2%) with lesser amounts equally distributed between ADHIB F (5.9%) and ADHIB H (5.9%) haplotypes. This frequency distribution in ADHIB haplotypes between the two populations is consistent with the known high frequency of G143A SNP among Asians, and low frequency in Caucasians (Table 1). ADHIB F and ADHIB H haplotypes exist at lower frequencies in both populations. ADHIB promoter SNPs are associated with a negative regulatory region
ADHIB is known to have a negative regulatory region within the 1,600 to 468 bp of the 50 promoter (Yu et al., 1994) (Figure 2). Three ADHIB SNPs in the current study population
LK Pershing et al. ADHIB Haplotype and Skin Erythema
–1,600 bp
Net negative regulation –809
–468 bp
Table 3. Sequence consensus of NRE2 in ADHIB vs other genes Gene
bp
ADH1B promoter SNP
11
50 nucleotide sequence 30 (993) TTTT CCCT ATA (982) (993) TTTT CACT ATA (982)
ADHIB 5′
ADHIC 5′
ADH1B promoter SNP
11
(745) TATT CATC TTC (736) (745) TATT CATT TTC (736)
–752 –739
–887
3′
–845
3′
Figure 2. SNPs in the NRE2 of ADHIB and ADHIC 50 promoter. Net negative regulation of ADHIB and ADHIC are associated with the general region 1,600 nt to 809 and 809 to 468 nt in the 50 promoter of ADHIB and ADHIC (Edenberg, 2000). ADHIB contains NRE1 located 336 nt to 350 nt and NRE2 located 743 nt to 732 nt in ADHIB which contain the C-887A, T-752C, and C-739T SNPs. ADHIC also contains a known NRE in the region containing the T-845C SNP.
IFN-b promoter
11
(60) AATT CCTC TGA (50)
IL-8 promoter
11
(1,415) AATT CCTC TGA (1,405)
iNOS promoter
11
(6,749) AATT CCTC AGC (6,739)
Consensus
11
WWTT CMTC WKM
ADHIB, class IB alcohol dehydrogenase; iNOS, inducible nitric oxide synthase; NRE2, negative regulatory element 2; SNP, single-nucleotide polymorphism.
2.0
SNPs in ADHIB coding and non-coding regions influence expression and production
Given that the C-887A and C-739T SNPs lie within a known negative regulatory region, we hypothesized that various ADHIB haplotypes should have altered mRNA and protein production. Lymphocytes isolated from known ADHIB haplotypes A, B, C, D, E, and G were evaluated for ADHIB transcript and protein concentration (Table S2) under basal controlled conditions. Increased ADHIB expression is associated with increased protein production (r ¼ 0.92) (data not shown). The extent of ADHIB production under basal control conditions is ADHIB haplotype dependent (Figure 3). ADHIB A, wild type at both 50 promoter loci and G143 in exon 3 is associated with an intact NRE2 and low expression and protein production (filled bar). The C-887A SNP in ADHIB B or the C-739T SNP in ADHIB C deregulates the negative regulatory region, thereby is associated with increased gene expression and protein production. Unexpectedly, SNPs C887A and C-739T together (ADHIB D) decreased expression and protein production twofold relative to the individual SNPs. ADHIB E containing the G143A SNP in exon 3 with an
1.5 ADHIB/GAPDH
lie within the negative regulatory region: two of those SNPs, T-752C and C-739T are part of a negative regulatory element 2 (NRE2) (Yu et al., 1994), the other C-887A SNP is upstream of the NRE2, but still within the general negative regulatory region. No SNPs were detected in NRE1 in the current population (Edenberg, 2000). The ADHIB promoter regions containing 15 bp around the 887 nt (50 993 to 979 30 ) and 12 bp around the 739 nt (50 745 to 732 30 ) were compared for consensus sequence among other known gene NRE and response elements. Table 3 illustrates that the NRE region encompassing the wild type or C-887A and C-739T SNPs of ADHIB show some similarity among other NREs previously identified in IFN-b (Nourbakhsh and Hauser, 1999), IL-8 (Nourbakhsh et al., 2001), and inducible nitric oxide synthase (Feng et al., 2002).
1.0
0.5
0.0 A
B
C D ADHIB haplotype
E
G
Figure 3. Alcohols alter ADHIB production as a function of ADHIB haplotype. Data represent the mean of duplicate evaluations of ADHIB protein concentration normalized to GAPDH in cultured lymphocytes from subjects of six different ADHIB haplotypes. Solid bar ¼ control, shaded bar ¼ EtOH, stripped bar ¼ IPA, hatched bar ¼ PG. ADHIB haplotype represents C-887A, C-739T, G143A, where 0 ¼ wild type, 1 ¼ heterozygote or homozygote variant. Raw optical densities at 260 nm of ADHIB and GAPDH expression for ADHIB haplotypes is provided in Table S2.
intact negative regulatory region in the 50 promoter was associated with mRNA and protein concentrations similar to ADHIB A. Combining the G143A SNP in exon 3 with the 50 promoter C-739T SNP (ADHIB G) was associated with both increased mRNA and protein concentration threefold greater than a SNP in exon 3 only (ADHIB E) and twofold greater than C-739T SNP only (ADHIB C). ADH mRNA and protein production in ADHIB F and ADHIB H were not evaluated due to unavailability of lymphocytes. Thus, combinations of SNPs in the negative regulatory region of the 50 promoter and the coding region of ADHIB are associated with modulation of gene expression and protein production. Alcohol(s) further modulate ADHIB protein concentration as a function of ADHIB haplotype. ADHIB produced in cultured lymphocytes exposed to EtOH (diagonal stripe) was www.jidonline.org
619
LK Pershing et al.
80 70 60 50 40 30 20 10 0 –10 –20
80 70 60 50 40 30 20 10 0 –10 –20
Millivolts
Millivolts
ADHIB Haplotype and Skin Erythema
B
C D E F ADHIB haplotype
G
H
6
6
5
5
4
4
3
3
a*
a*
A
2
2
1
1
0
0
–1
–1
A
B
C D E F ADHIB haplotype
G
H
A
B
C D E F ADHIB haplotype
A
B
C D E F ADHIB haplotype
G
G
H
H
Figure 4. Cutaneous blood flow and erythema is associated with alcohol type and ADHIB haplotype. Cutaneous blood flow (millivolts). (a) Linear chain alcohols: EtOH ¼ stippled bar, ButOH ¼ filled bar. (b) Branched alcohols: IPA ¼ stripped bar, PG ¼ hatched bar. Baseline-adjusted, control site-corrected cutaneous blood flow (millivolts) quantified by Doppler velocimeter 10 minutes after residual alcohol removal for ADHIB haplotypes (n): A (1), B (8), C (19), D (6), E (1), F (2), G (9), H (3). Mean7SEM. Erythema (a*). (c) Linear chain alcohols: EtOH ¼ stippled bar, ButOH ¼ filled bar. (d) Branched chain alcohols: IPA ¼ stripped bar, PG ¼ hatched bar. Baseline-adjusted, control site-corrected a*value by reflectance colorimeter using L*a*b* uniform color space 10 minutes after residual alcohol removal for ADHIB haplotypes (n): A (1), B (8), C (19), D (6), E (1), F (2), G (9), H (3). Mean7SEM.
not altered from their respective basal (filled bar) concentration in all the evaluated haplotypes (Figure 3, Table S2). In vitro exposure to the branched alcohol, IPA (gray bar), increased protein production in ADHIB A, C, and D, but not ADHIB B, E, or G haplotypes. The branched diol, PG (hatched bar), increased protein concentration associated with ADHIB B and C containing the G-887A and C-739T SNPs, respectively, had no effect on ADHIB A, D, and E, and was associated with inhibited protein production twofold from basal conditions in ADHIB G, containing the C-739T and G143A SNPs. Only ADHIB C containing the C-739T SNP was associated with increased protein production by both IPA and PG. Thus, ADHIB protein production can be altered as a function of both alcohol type and ADHIB haplotype. ADHIB haploptypes are associated with skin blood flow and erythema responses
Given that topical alcohol(s) exposure produces changes in cutaneous blood flow in humans (Wilkin and Fortner, 1986) and topical/transdermal drug products containing these alcohols have been documented to produce an erythema skin response (Fisher, 1995; Schachner et al., 2005; Reitamo et al., 2004; Reygagne et al., 2005; Rozenbaum et al., 1996) or allergic reactions (Lutz and El-Azhary, 1997; Bauman and Kerkel, 1999), we hypothesized that some of these skin changes are associated with ADHIB haplotype. Subjects previously haplotyped for ADHIB were evaluated noninvasively for erythema (a*) and cutaneous blood flow activity in volar forearm skin following topical exposure to 5 M aqueous alcohol solutions 620
Journal of Investigative Dermatology (2008), Volume 128
of the linear chain alcohols: EtOH and butanol (ButOH), as well as the branched chain alcohol, IPA and the diol, PG. Cutaneous blood flow (millivolts) and erythema (a*) following both EtOH and ButOH topical exposure was associated with ADHIB haplotype (Figure 4a and c, respectively). ButOH increased mean skin blood flow and erythema two- to seven-fold greater than EtOH in 7/8 ADHIB haplotypes. Skin blood flow (millivolts) following ButOH exposure was increased to the greatest extent in ADHIB E, F, G, and H containing the G143 SNP in exon 3. Increased skin blood flow associated with ADHIB haplotypes containing the G143A SNP is consistent with the well-known 40- to 100fold increase in alcohol metabolism to aldehyde and the flushing response in Asians. EtOH and ButOH had little effect on cutaneous blood flow in ADHIB B, C, and D containing no SNP in exon 3. IPA (gray bar) increased skin blood flow in ADHIB A and E, which are wild type for both the C-887 and C-739 loci in the 50 promoter (Figure 4b) with and without the G143A SNP in exon 3, respectively. IPA increased erythema (a*) to a greater extent than PG (hatched bar) in all ADHIB haplotypes except ADHIB A (Figure 4d). PG decreased cutaneous blood flow (millivolts) in 6/8 ADHIB haplotypes. Decreased erythema (a*) and skin blood flow following PG exposure was only measured in ADHIB F, G, and H, which contain the G143A SNP with either C-887A or C-739T or both SNPs. ADHIB A (containing no SNPs in the 50 promoter or exon 3) was unique among the haplotypes in producing some erythema (a*) to both linear and branched alcohols. Thus,
LK Pershing et al. ADHIB Haplotype and Skin Erythema
6
6
5
5
4
4
3
3
a*
a*
alcohols alter skin blood flow, with or without erythema as a function of alcohol type, and is associated with ADHIB haplotype. Increased erythema (a*) in volar forearm skin in vivo moderately correlated with increased cutaneous blood flow (millivolts) following exposure to EtOH (r ¼ 0.60) (Figure 5a). ButOH produced larger changes in skin blood flow and erythema (a*) (r ¼ 0.76) and was superior in its ability to differentiate between ADHIB haplotypes. The branched alcohol, IPA (gray bar), and the diol, PG (hatched bar) generally decreased cutaneous blood flow (millivolts o0) and did not correlate with small increases in erythema (a*)
2 1
1
0 –1 –20 –10 0
2 0
10 20 30 40 50 60 Millivolts
–1 –20 –10 0
10 20 30 40 50 60 Millivolts
12 10 8 6 4 2 0 –2 –4 –6 –2.0 –1.5 –1.0 –0.5
0.0 a*
0.5
1.0
1.5
2.0
12 10 8 6 4 2 0 –2 –4 –6 –2.0 –1.5 –1.0 –0.5 Blanching
Erythema
Expression
Blanching
Expression
Expression
Figure 5. Erythema is produced with and without alterations in cutaneous blood flow. (a) Linear chain alcohols: EtOH (closed circles) r ¼ 0.87, ButOH (open circles), r ¼ 0.92 of one representative subject from each of six different ADHIB haplotypes A, B, C, D, E, and G. (b) Branched chain alcohols: IPA (open square) r ¼ 0.84, PG (closed triangle) r ¼ 0.53 from one representative subject from each of six ADHIB haplotypes A, B, C, D, E, G. Erythema measured as baseline-adjusted, control site-corrected a* value by reflectance colorimeter. Skin blood flow measured as baseline-adjusted, control site-corrected millivolts. Both parameters represented as mean7SEM response 10 minutes after residual alcohol removal.
(Figure 5b). ADHIB F containing the C-887A and G143A SNPs was distinct in being associated with the largest increase in erythema (a*) to IPA with no change in skin blood flow. Thus, erythema (a*) measured in vivo with linear alcohols EtOH and ButOH is the result of increased cutaneous blood flow quantified at the same skin site. PGand IPA-induced erythema (a*) responses, however, were independent of skin blood flow (millivolts). The data suggest that erythema (a*) may be produced to variable extents by different mechanisms following topical exposure to linear vs branched topical alcohols. Erythema following exposure to alcohol(s) is assumed to reflect vasodilation due to metabolism to aldehyde. Erythema is also observed following exposure to alcohols (Latif et al., 2002) in irritation/inflammation responses (Gilder et al., 1993), and is associated with expression of proinflammatory cytokines, such as tumor necrosis factor-a (TNF-a) and IL-1b (Hunziker et al., 1992; Gottlieb, 2001). We hypothesized that the extent of alcohol metabolism is influenced by the amount of ADHIB protein produced and the metabolic activity of the encoded protein, both of which are controlled by the ADHIB haplotype. The relative roles of ADHIB protein produced, its metabolic activity, and TNF-a expression on erythema (a*) following EtOH, IPA, and PG exposure in vivo was evaluated in cultured peripheral blood lymphocytes harvested from ADHIB haplotypes A, B, C, D, E, and G in vitro (Figure 6a, b and c, respectively). Erythema normalized for ADHIB concentration (a*/[ADHIB]) was used as a surrogate marker of protein metabolic activity. Erythema (a*) following exposure to all three alcohols was associated with large increases in ADHIB metabolic activity
12 10 8 6 4 2 0 –2 –4 –6 –2.0 –1.5 –1.0 –0.5 Blanching
0.0 a*
0.5
1.0
1.5
0.0 a*
0.5
1.0
1.5
2.0
Erythema
2.0
Erythema
Figure 6. Erythema (a*) is produced by different mechanisms as a function of alcohol type and ADHIB haplotype. (a) EtOH: ADHIB concentration (ADHIB/GAPDH, filled circle) r ¼ 0.42, ADHIB activity (a*/(ADHIB/GAPDH), open circle) r ¼ 0.89, TNF-a expression (TNF-a/b-actin, þ ) r ¼ 0.58, from a representative subject associated with ADHIB haplotypes A, B, C, D, E, and G. (b) IPA: ADHIB concentration (ADHIB/GAPDH, filled square) r ¼ 0.47, ADHIB activity (a*/(ADHIB/GAPDH), open square) r ¼ 0.85, TNF-a expression (TNF-a/b-actin, þ ) r ¼ 0.47 from a representative subject associated with ADHIB haplotypes A, B, C, D, E, and G. (c) PG: ADHIB concentration (ADHIB/GAPDH), filled triangle) r ¼ 0.72, ADHIB activity (a*/(ADHIB/GAPDH), open triangle) r ¼ 0.92, TNF-a expression (TNF-a/b-actin, þ ) r ¼ 0.94 from a representative subject from ADHIB haplotypes A, B, C, D, E, and G.
www.jidonline.org
621
LK Pershing et al. ADHIB Haplotype and Skin Erythema
(open symbols) and smaller decreases in ADHIB protein concentrations (filled symbols) in vitro. Increasing erythema (a*) with EtOH (Figure 6a) and IPA (Figure 6b) exposure in vivo was independent of TNF-a expression ( þ ) in vitro, supporting alcohol metabolism to aldehyde and its effect on vasodilation as the mechanism of erythema induced by these alcohols. Differences in erythema (a*) following EtOH vs IPA exposure likely reflect differences in substrate binding to ADHIB (Stone et al., 1993). In contrast, PG-induced erythema (a*) was highly correlated with increasing ADHIB metabolic activity (open triangle, r ¼ 0.92) and increasing TNF-a expression in vitro ( þ , r ¼ 0.94) (Figure 6c), supporting alcohol metabolism coupled with cytokine-induced irritation. Thus, different alcohols produced cutaneous erythema (a*) by different mechanisms, but all involving ADHIB concentration and metabolic activity with or without TNF-a as a function of ADHIB haplotype. DISCUSSION Alcohols are often used in topical drug products to increase the active ingredient solubility, thereby increasing the uptake of the desired pharmacological agent into skin. Branched chain alcohols (IPA) and diols (PG) are especially useful in dermatologic formulations of glucocorticosteroids, retinoids, vitamin D analogs, testosterone, and estrogen products. These alcohols are generally assumed to be inactive and therefore have no effect on the active ingredient’s pharmacological activity. Recent studies suggest, however, that these topical ‘‘inactive ingredients’’ are associated with altered glucocorticosteroid (Pershing et al., 1999 abstr.; Rausch et al., 2001 abstr.) and capsaicinoid (Pershing et al., 2006) responses in vivo. SNPs in class I ADH are associated with many diseases
SNPs in the coding region of ADHIB (G143A in exon 3) are associated with functional changes in the encoded protein activity and increased risk for a variety of diseases, including alcoholism (Bosron et al., 1993; Agarwal, 1997; Borras, 2000), cirrhosis (Chao, 1994), myocardial infarction (Hines et al., 2001), fetal alcohol syndrome (Duester, 1991; McCarver et al., 1997), noninsulin-dependent diabetes (Suzuki et al., 2006), esophageal cancer (Chao et al., 2000), Parkinson’s disease (Buervenich et al., 2000), alcohol-associated chronic pancreatitis (Maruyama et al., 1999), and squamous-cell carcinoma (Hori et al., 1997). This report provides early evidence linking class I ADH SNPs with skin erythema. We hypothesized that erythema associated with dermatologic products containing alcohol(s) may have an underlying genetic basis associated with altered alcohol metabolism that are controlled by polymorphisms in the 50 promoter and coding regions ADHIB. ADHIB haplotype frequencies in Caucasians vs Asians
The strong LD of SNPs in exon 3 of ADHIB with those in exons 5 and 6 in ADHIC in the current population agree with previous reports (Chen et al., 1999; Osier et al., 1999). The variable frequency of SNPs in the 50 promoter of ADHIB and ADHIC between Races, yet high LD between these SNPs 622
Journal of Investigative Dermatology (2008), Volume 128
within and between each gene, however, was unexpected. The majority of Asians evaluated were ADHIB G (88.2%), containing the 50 promoter SNP C-739T and exon 3 SNP G143A. In contrast, Caucasians had more ADHIB haplotypes, with the greatest frequency being wild type at nt 143 in exon 3. The most common 50 promoter SNPs was C-739T (ADHIB C, 44.3%), or C-887A (ADHIB B, 21.4%), or both (ADHIB D, 17.1%). Surprisingly, ADHIB A containing no SNPs represented only 4.3% of the current Caucasian population. ADHIB E containing the G143A SNP only was more common in Caucasians (2.9%) than Asians (0.0%). ADHIB F containing G143A and C-887A SNPs occurred in a similar frequency in both Caucasians (4.3%) and Asians (5.9%). ADHIB G containing the G143A and C-739T SNPs, was 20-fold more prevalent in Asians (88.2%) than Caucasians (4.3%). ADHIB H containing all three SNPs (C-887A, C-739T, and G143A) was fourfold more prevalent in Asians (5.9%) than Caucasians (1.4%). ADHIB expression is associated with SNPs in NRE2
SNPs in the 50 promoter located within or near by a known negative regulatory element (NRE2) influence ADHIB transcription. The 15 bp sequence upstream of the previously established region of the NRE2 (Yu et al., 1994) containing the C-887A SNP as well as the 12 bp sequence from 734 to 745 bp within the NRE2 region shares similarities with other published NREs described in genes associated with immune and inflammatory responses (Table 1). SNPs in the NRE2 region of the 50 promoter of ADHIB was associated with increased ADHIB expression above wild type. These data agree with promoter truncation in this region of mouse Adh-1 and derepression of mouse Adh expression (Yu et al., 1994). ADHIB expression correlated with ADHIB production into protein, and differed up to fivefold among ADHIB haplotypes. Mouse strains also differ up to twofold in class I Adh expression (Rex et al., 1987). SNPs in the coding region alone would not be expected to alter ADHIB expression, yet the well-known G143A SNP in exon 3 of ADHIB E was associated with a twofold greater expression than wild type ADHIB A, but twofold less expression than ADHIB B or C, which contain the 50 promoter SNPs (C-887A and C-739T, respectively) without the exon 3 SNP. SNPs at both 50 promoter locations (C-887A and C-739T) as well as the G143A exon 3 SNP (ADHIB H) were associated with decreased ADHIB expression and production. Alcohol-induced alteration of ADHIB protein production was associated with specific ADHIB haplotypes. The linear alcohol, EtOH, produced similar ADHIB expression as untreated basal conditions in all ADHIB haplotypes following exposure to 1.5 mg ml1 concentration (36 mM) and is consistent with low concentration EtOH effects observed in other cell types (Chen et al., 1993; D’Souza et al., 1994; De Vito et al., 2000; Nakatani et al., 2002). In contrast, increased protein production following IPA exposure was greater than after PG exposure as a function of ADHIB haplotype. Only ADHIB C, containing the C-739T SNP was associated with more protein concentration following both IPA and PG. ADHIB G, containing both C-739T and G143A SNPs was
LK Pershing et al. ADHIB Haplotype and Skin Erythema
associated with less protein after PG exposure. Thus, alcohol type differentially influences protein production in association with SNPs in the NRE of 50 promoter and G143A in exon 3 of various ADHIB haplotypes. These data suggest that alcohol metabolites influence ADHIB transcription. ADHIB haplotypes are associated with human skin blood flow and erythema
Alcohol-induced changes in human skin blood flow and erythema (a*) is alcohol type dependent and associated with ADHIB haplotype. ADHIB haplotypes containing the G143A SNP (ADHIB E, F, G, and H) are associated with greater increases in skin blood flow and a coincident increased erythema (a*) following exposure to linear alcohols, EtOH and ButOH, than branched alcohols, IPA or PG. These data are consistent with increased alcohol metabolism to aldehyde, causing increased skin blood flow and the resulting erythema (a*) in vivo. In contrast, SNPs in the 50 promoter of ADHIB without the exon 3 G143A SNP (ADHIB B, C, and D haplotypes) are associated with minimal changes in blood flow or erythema (a*) to all the alcohols evaluated. Only ADHIB A, containing no SNPs, was associated with increased erythema (a*) to EtOH, ButOH, and IPA. The better discrimination of alcohol-associated changes in erythema (a*) measured by the colorimeter compared to the laserDoppler velocimeter changes in skin blood flow have also been observed in patch testing with mild chemical irritants (Lahti et al., 1993), allergic chemicals (Gawkrodger et al., 1991), and UV light exposure (Serup and Agner, 1990). The branched alcohol, IPA, altered cutaneous blood flow (millivolts) to a similar extent as EtOH, but with less associated erythema (a*). ADHIB F containing the C-887A and G143A SNPs was unique among ADHIB haplotypes, in being associated with large changes in erythema (a*) without changes in skin blood flow. In contrast, PG-induced erythema (a*) was generally independent of skin blood flow (millivolts). Reflectance colorimetry-detected erythema (a*) without laser-Doppler velocimetry (LDV)-detected cutaneous blood flow changes could reflect time-dependent changes in alcohol metabolism. Different types of irritants can cause variable changes in the two parameters as a function of time. Cutaneous blood flow changes are induced early and dissipate quickly, whereas erythema evolves over longer time periods and remains at the site longer (Wahlberg, 1984). Differential permeability of linear vs branched alcohol(s) into skin could also alter the changes in erythema without alteration in skin blood flow. Linear alcohols are lipophilic and permeate the skin as a function of carbon number. Permeability coefficient (Kp, cm hour1) of the three carbon propanol (1.4 cm hour1) is less than four carbon ButOH (2.5 cm hour1), six carbon hexanol (13 cm hour1), or nine carbon nonanol (60 cm hour1) (Blank et al., 1967). Shortening the carbon chain length and introduction of additional polar groups lowers the permeability constant of branched alcohols drastically (e.g., ButOH (2–4 cm hour1) vs 2,3butane diol (0.05 cm hour1) (Scheuplein and Blank, 1971). Thus, decreased permeability of IPA and PG may effectively lower their bioavailability to vascular endothelium over the
same exposure time period, thereby producing less concentration-dependent changes on cutaneous blood flow compared to EtOH or ButOH. Erythema without skin blood flow changes following topical exposure to linear vs branched alcohols could also reflect different alcohol substrate binding to the ADHIB protein, thereby altering their rate and extent of metabolism to aldehyde and the resulting vascular responses. ADHIB protein is known to efficiently metabolize primary alcohols (Burnell and Bosron, 1989). ADHIB containing the G143A SNP alters amino acid 48 (formerly described as amino acid 47) from arginine to histadine and is associated with a 20-fold increase in Km and a 45-fold increase in Vmax (Bosron and Li, 1986). Increasing chain length of linear alcohols produces similar Km and Vmax values (Stone et al., 1989). In contrast, increasing chain length of branched alcohols decreases Km and the efficiency of the catalytic activity (Vmax/Km) by ADHIB (Stone et al., 1989). Thus, increased binding and metabolism of linear alcohols produces greater increases in cutaneous blood flow than branched alcohols. Differences in cutaneous erythema with and without associated changes in blood flow could also be influenced by depth of light penetration by reflectance colorimetry vs laser-Doppler velocimetry. Laser-Doppler velocimetry detects changes in skin blood flow at the pre-arteriole depth in skin as evidenced by its sensitivity in detecting vasoconstriction by phenylephrine in vivo (Pershing et al., 1989), but not skin blanching (unpublished personal observations) or septal nasal blood flow following glucocorticosteroid nasal sprays (Cervin et al., 2001). These data are consistent with mediation of IPA and PG-induced skin responses at capillaries located superficially to pre-arterioles and thus would not be detected by LDV. PG-induced erythema (a*) correlated with increased in vitro TNF-a expression but not blood flow. TNF-a plays a pivotal role in inflammatory processes, upregulating cytokine expression in immune responses (Gottlieb, 2001). TNF-a increases vascular permeability and angiogenesis (Diaz et al., 2000) by inducing vascular endothelial growth factor. These vascular changes, however, are not associated with laserDoppler velocimeter-detected changes in cutaneous blood flow (personal observation). In contrast to PG, IPA increased cutaneous blood flow and erythema (a*) independent of TNF-a expression. IPA may therefore be associated with upregulation of another proinflammatory cytokine or other unknown mechanism. Detection of erythema (a*) independent of cutaneous blood flow changes further suggests that PG- and IPA-induced erythema is located superficial to pre-arteriole changes in blood flow detected by laser-Doppler velocimeter. Increased cutaneous blood flow (millivolts) in conjunction with erythema (a*) reflects rapid alcohol metabolism to aldehyde in vivo and is associated with ADHIB E, F, G, and H, which contain the G143A SNP. Decreased reflectance colorimeter a* values (skin blanching) are associated with slow alcohol metabolizing ADHIB B, C, and D and reflect increased TNF-a expression to the parent alcohol. Erythema and increased cutaneous blood flow following topical EtOH, ButOH, and IPA challenge in ADHIB A, containing no SNPs, www.jidonline.org
623
LK Pershing et al. ADHIB Haplotype and Skin Erythema
suggests that some alcohols may have direct effects on additional proinflammatory genes other than TNF-a. Conclusions
SNPs in the 50 promoter and coding region (G143A) of ADHIB form ADHIB haplotypes that are associated with altered gene expression and translation into protein. Eight ADHIB haplotypes exist in Caucasians, with the highest frequencies (44.3, 21.4, and 17.1%), containing the 50 promoter C-739T SNP (ADHIB C), the C-887A SNP (ADHIB B), or both (ADHIB D), respectively. Only three ADHIB haplotypes were identified in the evaluated Asian population with the highest frequency (88.2%) containing the 50 promoter SNP C-739T and exon 3 SNP G143A (ADHIB G). ADHIB E, F, G, and H containing the coding region SNP G143A, are highly associated with increased cutaneous blood flow and erythema in vivo following exposure to linear alcohols, EtOH and ButOH, but not branched alcohols, IPA or PG. The extent of skin blood flow alteration and erythema was modulated by the presence of SNPs in the 50 promoter, which regulates the amount of protein produced. Thus, SNPs in both non-coding and coding regions of ADHIB influence expression and activity of the protein. Furthermore, these ADHIB haplotypes can be differentiated noninvasively by their ability to metabolize linear vs branched alcohols in human skin in vivo by LDV and reflectance colorimetry. Erythema with branched alcohols in some ADHIB haplotypes was independent of changes in cutaneous blood flow, supporting the concept that erythema, like the skin blanching response, is associated with increased capillary permeability that is superficial to the pre-arteriole blood flow changes manifested by EtOH and ButOH detected by LDV. MATERIALS AND METHODS Human subjects The University of Utah medical ethical committee approved all described studies. The human studies were conducted according to the Declaration of Helsinki Principles. Eighty-seven male or female subjects, aged 18–65 years, were enrolled in the described studies after providing written informed consent. Enrolled subjects were from various racial/ethnic groups: 17 Asians (7 males, 10 females), 70 Caucasians (22 males, 48 females). Subjects were remunerated for peripheral blood sample donation and participation in the skin blanching response studies.
in-house prepared 15 mM MgCl2 PCR buffer. Four-hundred nanogram samples of DNA were used in the assay. The resulting PCR products were confirmed for purity by gel electrophoresis before submission to the Institutional DNA Sequencing Core Facility for direct sequencing using capillary gel electrophoresis.
Haplotyping methods Haplotypes were established using the enrolled subjects with Quickstart (http://archimedes.well.ox.ac.uk/pise/quickstart.html and Phase v.2.02 (http://archimedes.well.ox.ac.uk/pise/PHASE-simple.html) web available computer software (Stephans et al., 2001; Stephans and Donnelly, 2003). Only those SNPs having a frequency above 5% in the study population were used to establish the haplotype and evaluate LD. SNPs demonstrating high LD within ADHIB or between ADHIC and ADHIB (D0 40.8) were removed from the haplotype analysis. Thus, three SNPs in ADHIB (C-887A, T-739C, and G143A were used to establish ADHIB haplotypes.
ADHIB western assay ADHIB protein was isolated from the lymphocyte cultures as per instructions using NE-PER nuclear and cytoplasmic extraction reagent (no. 78833, Pierce Biotechnology, Rockford, IL). Twenty micrograms of protein samples were separated by 4–12% Bis-Tris gel (Invitrogen, Carlsbad, CA) electrophoresis at 180 V for 1 hour at room temperature (RT). The proteins in the gel were transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA) using 32 V at 41C for 1.5 hours. Nonspecific binding was blocked using 5% powdered milk dissolved in TBS buffer for 2 hours. Membranes were then incubated with 1:2,000 primary rabbit anti-human ADHIB antibody at 41C overnight, (kindly supplied by Dr Hoog, Karolinska Institute, Stockholm, Sweden). Membranes were washed at RT in 1 TBS-T six times, 10 minutes each, before incubation with the 1:2,000 secondary goat anti-rabbit IgG horseradish peroxidaselinked antibody (no. 7074, Cell Signaling Technology, Beverly, MA) at RT for 2 hours. Membranes were washed again with 1 TBS-T buffer six times 10 minutes each. Membrane protein concentrations were quantified after soaking the membrane in SuperSignal West Pico chemiluminescent substrate (Pierce, Rockford, IL) for 5 min at RT followed by a 1 minute RT exposure in an imaging instrument (Image Station 2000R, Eastman Kodak Company Scientific Imaging Systems, New Haven, CT). All data were normalized to glyceraldehyde3-phosphate dehydrogenase (GAPDH) using a 1:4,000 monoclonal primary mouse anti-human GAPDH antibody (no. MMS-5805-200, Covance, Berkeley, CA) and 1:2,000 secondary anti-mouse IgG (A-3682 Fab-specific, Sigma, St Louis, Mo). Raw data are presented in Table S1.
SNP identification Genomic DNA was isolated from peripheral blood samples by the General Clinical Research Center (GCRC) at the University of Utah using a DNA purification kit (Gentra Systems, Minneapolis, MN). DNA samples were stored at 201C until use. Seven SNPs in ADHIB or ADH2 (OMIM 103720) and three SNPs in class IC ADH (ADHIC or ADH3; OMIM 103730), respectively, were evaluated in genomic DNA from the enrolled subjects. Areas of gene(s) containing the 10 known SNPs (www.genome.utah.edu/genesnps/) were amplified using PCR (Techne, Burlington, NJ) with forward and reverse flanking primers (Integrated DNA Technologies Inc., Coralville, IA) designed in-house (Table S1). PCR products were generated using BIOLASETMDNA polymerase kit (Bioline, Randolph, MA) with 624
Journal of Investigative Dermatology (2008), Volume 128
Gene expression assay To evaluate the association of one coding and two non-coding SNPs alone and/or in combination with ADHIB protein production, cultured lymphocytes were isolated from peripheral blood of seven enrolled subjects previously assessed for ADH haplotype. All subjects were wild type for aldehyde dehydrogenase 2 (ALDH2). Lymphocytes were Epstein–Barr virus immortalized by the Institutional GCRC (Pelloquin et al., 1986). Following transformation, lymphocytes were stored at 801C until cultured. Lymphocytes were cultured with 15% charcoal-filtered, dextran-treated fetal bovine serum (HycloneTM, Ogden, UT) in 85% RPMI þ glutamine at 371C under 5% CO2 in sterile 75 cm2 plastic
LK Pershing et al. ADHIB Haplotype and Skin Erythema
flasks (FalconTM, BD Sciences, San Jose, CA) to 80% confluence (105 cells/ml). Six milliliters of the cell culture were then transferred into each well of a sterile, flat bottom six-well tissue culture plate (Sarstedt, Newton, NC) and grown to 80% confluence. Twenty-four hours before experimentation, fetal bovine serum was removed from the culture media. Cultured lymphocytes were exposed to various alcohol treatments in duplicate overnight (18 hours): untreated control (phosphate-buffered saline), EtOH (9 mg), IPA (6.5 mg), and PG (7.6 mg). Alcohol doses were calculated based on anticipated alcohol concentration achieved in the epidermis from in vitro diffusion parameters. In vitro cell viability was verified to be greater than 99% following exposure to all alcohol(s) evaluated by trypan blue dye exclusion assay. Total protein was isolated from the various lymphocyte samples using NE-PER nuclear and cytoplasmic extraction reagent (no. 78833, Pierce Biotechnology, Rockford, IL). RNA was isolated from the lymphocyte cultures using TRIzol (Invitrogen), respectively, and submitted to reverse transcription PCR and PCR assays. PCR primers for the 50 promoter and exon 3 SNPs in ADHIB (Table 1) were designed in-house and used as purchased at 10 pM per reaction (Integrated DNA Technologies Inc., Coralville, IA). Two micrograms of RNA and reverse transcriptase (M-MLV, Invitrogen) was used to generate ADHIB complementary DNA (Techne, Burlington, NJ). The resulting complementary DNA was stored at 201C until use. complementary DNA was amplified for the gene of interest by PCR. PCR products were purified using MinEluteTM (Qiagen Sciences, Valencia, CA) before quantification by 260 nm UV spectrophotometry. All data were normalized to b-actin.
Topical alcohol challenge test Fifty-two subjects of various ADHIB haplotypes were evaluated for alcohol-associated changes in cutaneous blood flow and erythema. The randomization schedule for alcohol application for each subject was determined in Excel spreadsheet using the available randomization method (MS Office OSX, Redmond, WA). In this way, all alcohols were evaluated at a 5 M aqueous alcohol concentration simultaneously, in the same subject, in the same study group, to minimize temporal variability. Topical exposure to alcohol(s) has been previously shown to produce greater cutaneous blood flow increases by Doppler velocimetry in Asian than Caucasian subjects (Wilkin and Fortner, 1986). Increased millivolt output from a treated skin site is associated with increased blood flow reflecting increased alcohol metabolism to aldehyde and the resulting vasodilation. Cutaneous blood flow was quantified using a laser-Doppler velocimeter (Med Pacific CR2000, Seattle, WA). Increased alcohol metabolism following a 5 minutes exposure to various 5 M aqueous alcohol solutions is associated with a ‘‘flushing response’’ (Wilkin and Fortner, 1986) or erythema that can be objectively quantified with a reflectance colorimeter (Minolta Chroma meter CR200 (Ramsey, NJ) as an increase in a* scale of the L*a*b* uniform color space (Prins et al., 1998; Pershing et al., 2006). We hypothesized that the extent of the flushing response was associated with ADHIB haplotype. Thus, flushing (erythema) measured as an increase in a* scale of the L*a*b* uniform color space by a reflectance colorimeter (Minolta Chroma Meter CR200, Ramsey, NJ) was independent of the blood flow changes. Changes of 1.0 a* unit by the instrument method is approximately equal to a þ 1 unit change by visual assessment (personal observation, data not shown). Both colorimeter-assessed
skin color (a*) and LDV-assessed blood flow (millivolts) are surrogate bioengineering markers for skin erythema and are variable in the human volar forearm as a function of the test site location and time of day (unpublished results). Thus, all skin data was baselineadjusted and control site-corrected for both parameters. Four alcohols (Spectrum Chemical Manufacturing Corp, Gardena, CA) were topically evaluated on human forearms for their ability to change cutaneous blood flow (LDV millivolts) and skin color (a*): linear chain alcohols, dehydrated EtOH and 1-butanol (ButOH), a branched chain three carbon alcohol, 2-propanol or IPA and the three carbon diol, 1,2-propane diol or PG. Each 5 M aqueous alcohol solution was freshly prepared in double distilled water before use in the assay. The IQ chamber patch testing system (Chemotechnique Diagnostics AB, Malmo, Sweden) was used for the alcohol exposures. Ten 9 9 mm chambers contained in a stiff additive-free polyethylene plastic are supplied on a tape backing, thus allowing evaluation of the four alcohols and untreated sites in duplicate. Specific alcohol(s) and untreated placement on the volar forearm was randomized for each subject using Excel (Microsoft Office 2000). The distance between individual chambers is 12 mm in the row and 20 mm between the rows. The patch testing system was placed a minimum of 3 cm above the wrist fold and 3 cm below the antecubital fossa to obviate large differences in cutaneous blood flow on the volar forearms. The plastic containing 10 chambers per system was used to demarcate each individual test site on the volar forearm. One hundred twenty-five microliters of each 5 M alcohol solution or distilled water (untreated control) was pipetted onto the absorbant pad within each test chamber immediately before placement on previously demarcated locations on the volar forearm for 5 minutes. (Note: a 70% alcohol wipe contains 11.6 M IPA.) Skin test sites were exposed to the alcohol solution(s) for 5 minutes. Alcohol exposure was terminated by removing the chambers and blotting the exposed skin site with Kimwipe tissue (Kimberly-Clark Corp, Roswen, GA). One hour was required per subject to complete the challenge test. Skin color (erythema a*) and cutaneous blood flow (millivolts) were quantified at each skin site before alcohol(s) exposure, and again immediately after residual alcohol removal and again 10 minutes later. Data are presented as the mean response of each associated ADHIB haplotype. CONFLICT OF INTEREST The authors state no conflict of interest.
ACKNOWLEDGMENTS We express our deep thanks to all of the participants who faithfully participated in these studies. This investigation was supported by the Public Health Services research grant numbers M01-RR00064 and C06-RR11234 from the National Center for Research Resources.
SUPPLEMENTARY MATERIAL Table S1. PCR primers used to detect 10 SNPs in class I ADH genes. Table S2. ADHIB western blot assays.
REFERENCES Agarwal DP (1997) Molecular genetic aspects of alcohol metabolism and alcoholism. Pharmacopsychiatry 30:79–84 Bauman L, Kerkel F (1999) Topical glucocorticosteroids. In: Fitzpatrick T (ed). Dermatology in General Medicine. New York: McGraw-Hill, 2713–7
www.jidonline.org
625
LK Pershing et al. ADHIB Haplotype and Skin Erythema
Blank IH, Scheuplein RJ, MacFarlane DJ (1967) Mechanism of percutanaeous absorption. III. The effect of temperature on the transport of nonelectrolytes across the skin. J Invest Dermatol 49:582–9
Gilder FJ, Hodgkinson S, Murray RM (1993) ADH and ALDH genotype profiles in Caucasians with alcohol-related problems and controls. Addiction 88:383–8
Borras E (2000) Genetic polymorphism of alcohol dehydrogenase in Europeans: the ADH2*2 allele decreases the risk of alcoholism and is associated with ADH3*1. Hepatology 31:984–9
Gottlieb AB (2001) Psoriasis: immunopathology and immunomodulation. Dermatol Clin 19:649–57
Bosron WF, Ehrig T, Li T-K (1993) Genetic factors in alcohol metabolism and alcoholism. Semin Liver Dis 13:126–35 Bosron WF, Li T-K (1986) Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases and their relationship to alcohol metabolism and alcoholism. Hepatology 6:502–10 Buervenich S, Sydow O, Carmine A, Zhang Z, Anvret M, Olson L (2000) Alcohol dehydrogenase alleles in Parkinson’s disease. Mov Disord 15:813–8 Burnell JC, Bosron WF (1989) Genetic polymorphism of human liver alcohol dehydrogenase and kinetic properties of the isoenzymes. In: Crow KE and Batt RD (eds). Human Metabolism of Alcohol. Boca Raton: CRC Press, 65–75 Cervin A, Akerlund A, Greiff L, Andersson M (2001) The effect of intranasal budesonide spray on mucosal blood flow measured with laser Doppler flowmetry. Rhinology 39:13–6 Chao YC (1994) Polymorphism of alcohol and aldehyde dehydrogenase genes and alcoholic cirrhosis in Chinese patients. Hepatology 19:36–6 Chao YC, Wang SJ, Chu HC, Chang WK, Hsieh TY (2000) Chinese alcoholic patients with esophageal cancer are generically different from alcoholics with acute pancreatitis and liver cirrhosis. Am J Gastroenterol 95:2958–64 Chen CC, Lu RB, Chen YC, Wang MF, Chang YC, Li TK et al. (1999) Interaction between the functional polymorphisms of the alcoholmetabolism genes in protection against alcoholism. Am J Hum Genet 65:795–807
Hori H, Kawano T, Endo M, Yuasa Y (1997) Genetic polymorphisms of tobacco- and alcohol-related metabolizing enzymes and human esophageal squamous cell carcinoma susceptibility. J Clin Gastroenterol 25:568–75 Hunziker T, Brand CU, Kapp A, Waelti ER, Braatheen LR (1992) Increased levels of inflammatory cytokines in human skin lymph derived from sodium lauryl-sulphate-induced contact dermatitis. Br J Dermatol 127:254–7 Hurley TD, Bosron WF, Stone CL, Amzel LM (1994) Structures of three human beta alcohol dehydrogenase variants. Correlations with their functional differences. J Mol Biol 10:415–29 Jornvall H, Hoog J-O, Persson B, Pares X (2000) Pharmacogenetics of the alcohol dehydrogenase system. Pharmacology 61:184–91 Lahti A, Kopola H, Harila A, Myllyla R, Hannuksela M (1993) Assessment of skin erythema by eye, laser Doppler flowmeter, spectroradiometer, two channel erythema meter and Minolta chroma meter. Arch Dermatol Res 285:278–82 Latif O, Peterson JD, Waltenbaugh C (2002) Alcohol-mediated polarization of type 1 and type 2 immune responses. Front Biosci 1:a135–47 Lutz M, El-Azhary R (1997) Allergic contact dermatitis due to topical application of topical corticosteroids: review and clinical applications. Mayo Clinic Proceed 72:1141–4
Chen GJ, Huang DS, Watzl B, Watson RR (1993) Ethanol modulation of tumor necrosis factor and gamma interferon production by murine splenocytes and macrophages. Life Sci 52:1319–26
Maruyama K, Takahashi H, Matsushita S, Nakano M, Harada H, Otsuki M et al. (1999) Genotypes of alcohol-metabolizing enzymes in relation to alcoholic chronic pancreatitis in Japan. Alcohol Clin Exp Res 23:85S–91S
Cheung C, Smith CK, Hoog J-O, Hotchkiss SA (1999) Expression and localization of human alcohol and aldehyde dehydrogenase enzymes in skin. Biochem Biophys Res Commun 261:100–7
McCarver DG, Thomason HR, Martier SS, Sokol RJ, Li T (1997) Alcohol dehydrogenase-2*3 allele protects against alcohol-related birth defects among African Americans. J Pharmacol Exp Ther 283: 1095–101
Cornell R, Stoughton R (1985) Correlation of the vasoconstrictor assay and clinical activity in psoriasis. Arch Dermatol 121:63–7 D’Souza NB, Nelson S, Summer WR, Deaciuc IV (1994) Expression of tumor necrosis factor-alpha and interleukin-6 cell-surface receptors of the alveolar macrophage in alcohol-treated rats. Alcohol Clin Exp Res 18:1430–5 De Vito WJ, Stone S, Mori K, Shamgochian M (2000) Ethanol inhibits prolactin- and tumor necrosis factor-alpha-, but not gamma interferoninduced expression of intercellular adhesion molecule-1 in human astrocytoma cells. J Cell Biochem 77:455–64 Diaz BV, Lenoir MC, Ladoux A, Frelin C, Demarchez M, Michel S (2000) Regulation of vascular endothelial growth factor expression in human keratinocytes by retinoids. J Invest Dermatol 275:642–50 Duester G (1991) A hypothetical mechanism for fetal alcohol syndrome involving ethanol inhibition of retinoic acid synthesis at the alcohol dehydrogenase step. Alcohol Clin Exp Res 15:568–72 Edenberg HJ (2000) Regulation of the mammalian alcohol dehydrogenase genes. Prog Nucl Acid Res 64:295–341 Feng X, Guo Z, Nourbakhsh M, Hauser H, Ganster R, Shao L et al. (2002) Identification of a negative response element in the human inducible nitric-oxide synthase (hiNOS) promoter: the role of NF-kappa B repressing factor (NRF) in basal repression of the hiNOS gene. Proc Natl Acad Sci USA 99:14212–7 Fisher D (1995) Adverse effects of topical corticosteroid use. West J Med 162:1141–4 Food and Drug Administration (1998) Guidance for Industry: topical Dermatologic Drug Products NDAs and ANDAs – In Vivo Bioavailability, Bioequivalence, In Vitro Release, and Associated Studies Gawkrodger DJ, McDonaugh AJ, Wright AL (1991) Quantification of allergic and irritant patch test reactions using laser-Doppler flowmetry and erythema index. Contact Dermatitis 24:172–7
626
Hines LM, Stampfer MJ, Ma J, Gaziano JM, Ridker PM, Hankinson SE et al. (2001) Genetic variation in alcohol dehydrogenase and the beneficial effect of moderate alcohol consumption on myocardial infarction. N Engl J Med 344:549–55
Journal of Investigative Dermatology (2008), Volume 128
Nakatani Y, Fukui H, Kitazawa T, Fujimoto M, Yamao J, Uemura M (2002) Effect of alcohol on the secretion of tumor necrosis factor-alpha by macrophages in the presence of rat serum. Alcohol Clin Exp Res 26:81S–5S Nourbakhsh M, Hauser H (1999) Constitutive silencing of IFN-b promoter is mediated by NRF (NF-kb-repressing factor), a nuclear inhibitor of NF-kb. EMBO J 18:6415–25 Nourbakhsh M, Kalble S, Dorries A, Hauser H, Resch L, Kracht M (2001) The NFkb repressing factor is involved in basal repression and interleukin (IL-1) induced activation of IL-8 transcription by binding to a conserved NF-kb-flanking sequence element. J Biol Chem 276:4501–8 Osier M, Pakstis AJ, Kidd JR, Lee JK, Yin SJ, Ko HC et al. (1999) Linkage disequilibrium at the ADH2 and ADH3 loci and risk of alcoholism. Am J Hum Genet 64:1147–57 Osier MJ, Pakstis AH, Soodyall H, Comas D, Goldman D, Odunsi A et al. (2002) A global perspective on generic variation at the adh genes reveals unusual patterns of linkage disequilibrium and diversity. Am J Hum Gen 71:84–99 Pelloquin F, Lamelin JP, Lenoir GM (1986) Human B lymphocytes immortalization by Epstein–Barr Virus in the presence of cyclosporin A. In vitro Cell Dev Biol 22:689–94 Pershing LK, Huether S, Conklin RL, Krueger GG (1989) Cutaneous blood flow and percutaneous absorption: a quantitative analysis using a laser Doppler velocimeter and blood flow computer. J Invest Dermatol 92:355–9 Pershing LK, Petelenz K, Corlett JL (1999) Altered topical corticosteroid Emax parameters in Asian vs. Caucasian skin with propylene glycol. AAPS Pharm Res 1(S1) (abstr) Pershing LK, Reilly CA, Corlett JL, Crouch DJ (2006) Assessment of pepper spray product potency in Asian and Caucasian forearm skin using
LK Pershing et al. ADHIB Haplotype and Skin Erythema
transepidermal water loss, skin temperature and reflectance colorimetry. J Appl Toxicol 26:88–97 Prins M, Swinkels OQ, Kolkman EG, Wuis EW, Hekester YA, van der Valk PG (1998) Skin irritation by dithranol cream. A blind study to assess the role of the cream formulation. Acta Derm Venereol 78:262–5 Rausch HL, Pershing LK, Hobbs MR (2001) Altered Asian skin blanching response to dermatologic corticosteroids containing alcohol is associated with an alcohol dehydrogenase 2 polymorphism. J Invest Dermatol 117:520 (abstr.) Reitamo S, Harper J, Bos JD, Cambazard F, Bukijnzeel-Koomen C, Valk P et al. (2004) 0.03% Tacrolimus ointment applied once or twice daily is more efficacious than 1% hydrocortisone acetate in children with moderate to severe atopic dermatitis: results of a randomized doubleblind controlled trial. Br J Dermatol 150:554–62 Rex DK, Bosron WF, Dwulet F, Li T-K (1987) Purification and characterization of the Danish (Skive) variatnof mouse liver alcohol dehydrogenase. Biochem Genetic 25:111–21 Reygagne P, Mrowietz U, Decroix J, de Waard-van der Speck FB, Acebes LO, Figueiredo A et al. (2005) Clobetasol propionate shampoo 0.05% and calcipotriol solution 0.005%: a randomized comparison of efficacy and safety in subjects with scalp psoriasis. J Dermatol Treat 16:31–6 Roychoudhry AK, Masatoshi N (1988) Human Polymorphic Genes World Distribution. Oxford University Press: New York, USA. pp 416 Rozenbaum M, Birkhauser M, deNooyer C, Lambotte R, Pornel B, Schneider H et al. (1996) Comparison of two estradiol transdermal systems (Oesclims50 and estraderm TTSs 50). II. Local skin tolerability. Maturitas 25:175–85 Schachner LA, Lamerson C, Sheehan MP, Boguniewicz M, Mosser J, Raimer S et al. (2005) Tacrolimus ointment 0.03% is safe and effective for the treatment of mild to moderate atopic dermatitis in pediatric patients: results from a randomized, double-blind, vehicle-controlled study. Pediatrics 116:e334–42 Scheuplein RJ, Blank IH (1971) Permeability of the skin. Phys Rev 51:702–747
Serup J, Agner T (1990) Colorimetric quantification of erythema – a comparison of two colorimeters (Lange Micro Color) and Minolta Chroma Meter CR-200) with a clinical sorting scheme and laser-Doppler flowmetry. Clin Exp Dermatol 15:267–72 Stephans M, Donnelly P (2003) A comparison of Bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet 73:1162–9 Stephans M, Smith N, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 68:978–89 Stone CL, Bosron WF, Dunn MF (1993) Amino acid substitutions at position 47 of human beta 1 beta1 and beta 2 beta 2 alcohol dehydrogenases affect hybrid transfer and coenzyme dissociation rate constants. J Biol Chem 268:892–9 Stone CL, Li T-K, Bosron WF (1989) Stereospecific oxidation of secondary alcohols by human alcohol dehydrogenases. J Biol Chem 264:11112–6 Stoughton R (1992) The vasoconstrictor assay in bioequivalance testing: practical concerns and recent developments. Int J Dermatol 31:26–8 Suzuki Y, Ando F, Ohsawa I, Shimokata J, Ohta S (2006) Association of alcohol dehydrogenase 2*1 allele with liver damage and insulin concentration in the Japanese. J Hum Gen 51:31–7 Wahlberg JE (1984) Erythema-inducing effects of solvents following epicutaneous administration to man – studied by laser Doppler flowmetry. Scand J Work Environ Health 10:159–62 Wilkin J, Fortner GF (1986) Cutaneous vascular sensitivity to lower aliphatic alcohols and aldehydes in Orientals. Alcohol Clin Exp Res 9:522–5 Yasunami M, Kikuchi K, Sarapata D, Yoshia A (1990) The human class I alcohol dehydrogenase gene cluster: three genes are tandemly organized in a 80-kb-long segment of the genome. Genomics 7:152–8 Yu M, Jerome RE, Edenberg HJ (1994) A negative regulatory element upstream from the mouse Adh-1 gene can down-regulate a heterologous promoter. Gene 141:249–54
www.jidonline.org
627
