Variation in AKR1C3, which encodes the neuroactive steroid synthetic enzyme 3α-HSD type 2 (17β-HSD type 5), moderates the subjective effects of alcohol

Abstract

Rationale: Animal models suggest that neuroactive steroids contribute to alcohol's acute effects. We previously reported that a common nonsynonymous polymorphism, AKR1C3 2 in the gene encoding the enzyme 3α-HSD2/17β-HSD5, and a synonymous single nucleotide polymorphism (SNP), rs248793, in SRD5A1, which encodes 5α-reductase, were associated with alcohol dependence (AD).

Objectives: The aim of the study was to investigate whether these polymorphisms moderate subjective effects of alcohol in humans and whether AKR1C3 2 affects neuroactive steroid synthesis.

Methods: Sixty-five Caucasian men (34 lighter and 31 heavier drinkers; mean age 26.2 years) participated in a double-blind laboratory study where they consumed drinks containing no ethanol or 0.8 g/kg of ethanol. Breath alcohol, heart rate (HR), and self-reported alcohol effects were measured at 40-min intervals, and genotype was examined as a moderator of alcohol's effects. Levels of the neuroactive steroid 5α-androstane-3α,17β-diol and its precursors, 3α,5α-androsterone and dihydrotestosterone, were measured at study entry using GC/MS.

Results: Initially, carriers of the AD-protective AKR1C3 2 G allele had higher levels of 5α-androstane-3α,17β-diol relative to the precursor 3α,5α-androsterone than C allele homozygotes. AKR1C3 2 G allele carriers exhibited greater increases in heart rate and stimulant and sedative effects of alcohol than C allele homozygotes. The genotype effects on sedation were observed only in heavier drinkers. The only effect of the SRD5A1 SNP was to moderate HR. There were no interactive effects of the two SNPs.

Conclusions: The observed effects of variation in a gene encoding a neuroactive steroid biosynthetic enzyme on the rate of 17β-reduction of androsterone relative to androstanediol and on alcohol's sedative effects may help to explain the association of AKR1C3 2 with AD.

Fig. 1. Change in heart rate as a function of alcohol and AKR1C3*2 genotype

Left panel: Heart rate change from baseline following the placebo beverage. Right panel: Heart rate change from baseline following alcohol. Heart rate was measured approximately every 40 min and displayed as a change from baseline. C-allele homozygotes had a smaller alcohol-induced increase in heart rate than G-allele carriers. Data are displayed as mean ± S.E.M. of change in heart rate.

Fig. 2. Biphasic Alcohol Effects Scale (BAES) stimulation as a function of alcohol and AKR1C3*2 genotype

Placebo beverage (left panel) and Alcohol (right panel) laboratory sessions. CC genotype subjects had lower self-reported levels of stimulation following alcohol than G-allele carriers. Data are displayed as mean ± S.E.M. of BAES stimulation score (sum of 7 items).

Fig. 3. Biphasic Alcohol Effects Scale (BAES) sedation as a function of alcohol and AKR1C3*2 genotype

Placebo (left panel) and Alcohol (right panel) laboratory sessions. CC genotype subjects had lower self-reported levels of sedation following alcohol than G-allele carriers. Data are displayed as mean ± S.E.M. of BAES sedation score (sum of 7 items).

Fig. 4. Change in heart rate as a function of alcohol and SRD5A1 rs248793 genotype

Left panel: Heart rate change from baseline following the placebo beverage. Right panel: Heart rate change from baseline following alcohol. AD-associated risk allele (G) homozygotes had a greater alcohol induced-increase in HR than minor C-allele carriers. Data are displayed as mean ± S.E.M. of change in heart rate.

Fig. 5. AKR1C3*2 genotype moderation of acute sedative effects of alcohol are seen in heavier but not lighter drinkers

BAES sedation scores following alcohol administration for 34 lighter drinkers (left panel) and 31 heavier drinkers (right panel). CC genotype subjects had lower levels of sedation following alcohol only in the heavier drinker group. Data are displayed as mean ± S.E.M. of BAES sedation score (sum of 7 items).

Fig. 6. Interaction of AKR1C3*2 genotype with androstanediol precursors

Study baseline plasma androsterone, DHT and androstanediol levels were measured using GC/MS. There was a significant main effect of androsterone level and interaction of androsterone with genotype in predicting androstanediol levels, panel A. The simple slopes for the GLM parameter estimates for each of the three genotypes are shown in panel B.