Impact of Human SULT1E1 Polymorphisms on the Sulfation of 17β-Estradiol, 4-Hydroxytamoxifen, and Diethylstilbestrol by SULT1E1 Allozymes

Abstract

Background and objectives: Previous studies have revealed that sulfation, as mediated by the estrogen-sulfating cytosolic sulfotransferase (SULT) SULT1E1, is involved in the metabolism of 17β-estradiol (E2), 4-hydroxytamoxifen (4OH-tamoxifen), and diethylstilbestrol in humans. It is an interesting question whether the genetic polymorphisms of SULT1E1, the gene that encodes the SULT1E1 enzyme, may impact on the metabolism of E2 and these two drug compounds through sulfation.

Methods: In this study, five missense coding single nucleotide polymorphisms of the SULT1E1 gene were selected to investigate the sulfating activity of the coded SULT1E1 allozymes toward E2, 4OH-tamoxifen, and diethylstilbestrol. Corresponding cDNAs were generated by site-directed mutagenesis, and recombinant SULT1E1 allozymes were bacterially expressed, affinity-purified, and characterized using enzymatic assays.

Results: Purified SULT1E1 allozymes were shown to display differential sulfating activities toward E2, 4OH-tamoxifen, and diethylstilbestrol. Kinetic analysis revealed further distinct K_m (reflecting substrate affinity) and V_max (reflecting catalytic activity) values of the five SULT1E1 allozymes with E2, 4OH-tamoxifen, and diethylstilbestrol as substrates.

Conclusions: Taken together, these findings highlighted the significant differences in E2-, as well as the drug-sulfating activities of SULT1E1 allozymes, which may have implications in the differential metabolism of E2, 4OH-tamoxifen, and diethylstilbestrol in individuals with different SULT1E1 genotypes.

Fig. 1.. Ribbon diagram of the structure of human SULT1E1-17β-estradiol (E2)-3’-phosphoadenosine 5-phosphate (PAP) complex showing the locations of amino acid residues involved in the SULT1E1 cSNPs studied.

The structure of SULT1E1 (Protein Data Bank code: 4JVL), containing the polypeptide from Asp2 (N-terminus) to Glu293 (C-terminus), was depicted using Structure Comparison Analysis in USCF Chimera, a molecular modeling software [37]. E2 and PAP molecules in the structure are shown by bond structures. Loop-1 (Phe80-Asp90), loop-2 (Met144-Ser153), and loop-3 (Pro235-Gly25), shown by magenta color, form a gate for substrate entry [57, 58]. Side chains of the amino acid residues involved in the SULT1E1 cSNPs Ala43, Ala131, Arg186, Pro214 and Asp220 are indicated by bond structures (red color). PAP, (3′-phosphoadenosine 5′-phosphate); E2, (17β-Estradiol).

Fig. 2.. SDS gel electrophoretic pattern of the purified human SULT1E1 allozymes.

SDS-PAGE was performed on a 12% gel, followed by Coomassie blue staining. Samples analyzed in lanes 1–7. Lane 1 corresponds to the migrating positions of protein molecular weight markers co-electrophoresed. Samples analyzed in lanes 2 through 7 correspond to SULT1E1-WT, SULT1E-A43D, SULT1E1-A131P, SULT1E1-R186L, SULT1E1-P214T, and SULT1E1-D220V.

Fig. 3.. Concentration dependence of the sulfation of E2 by human wild-type SULT1E1.

The fitting curve was generated based on Michaelis-Menten kinetics. Data shown represent calculated mean ± standard deviation derived from three experiments.

Fig. 4.. Specific activities of the human SULT1E1 allozymes toward E2.

(a) Using 0.5 μM E2. (b) Using 2 μM E2. c) Using 4 μM E2. Data shown represent mean ± standard deviation derived from three independent determinations. One-way ANOVA was performed followed by Tukey’s post hoc analysis. **** Statistical significant p<0.0001 from SULT1E1-WT.

Fig. 5.. Concentration dependence of the sulfation of 4OH-tamoxifen by human wild-type SULT1E1.

The fitting curve was generated based on Michaelis-Menten kinetics. Data shown represent calculated mean ± standard deviation derived from three experiments.

Fig. 6.. Specific activities of the human SULT1E1 allozymes toward 4OH-tamoxifen.

(a) Using 10 μM 4OH-tamoxifen. (b) Using 50 μM 4OH-tamoxifen. (c) Using 200 μM 4OH-tamoxifen. Data shown represent mean ± standard deviation derived from three independent determinations. One-way ANOVA was performed followed by Tukey’s post hoc analysis. **** Statistical significant p<0.0001 from SULT1E1-WT.

Fig. 7.. Concentration dependence of the sulfation of diethylstilbestrol by human wild-type SULT1E1.

The fitting curve was generated based on Michaelis-Menten kinetics. Data shown represent calculated mean ± standard deviation derived from three experiments.

Fig. 8.. Specific activities of the human SULT1E1 allozymes toward diethylstilbestrol.

(a) Using 0.5 μM diethylstilbestrol. (b) Using 3 μM diethylstilbestrol. c) Using 6 μM diethylstilbestrol. Data shown represent mean ± standard deviation derived from three independent determinations. One-way ANOVA was performed followed by Tukey’s post hoc analysis. **** Statistical significant p<0.0001 from SULT1E1-WT.

Fig. 9.. Hydrophobic interaction and hydrogen bond analyses of the SULT1E1 allozymes.

Atoms interacted with Ala43 (A), Ala131 (B), Arg186 (C), Pro214 (D), and Asp220 (E) are colored by the blue-white-red gradient (left panels; WT). Estimated interaction formed with Asp43 in A43D (A), Pro131 in A131P (B), Leu186 in R186L (C), Thr214 in P214T (D), and Val220 in D220V (E) are colored by the blue-white-red gradient (right panels; substituted). Side-chain conformation of a substituted residue was simulated using the Dunbrack backbone-dependent rotamer library [36]. Hydrophobic and hydrogen bond interactions of the substituted residues were also simulated by Find Clashes/Contacts tool in USCF Chimera software [37]. Top five-ranked rotamers of each substituted residue are modeled using the Dunbrack backbone-dependent rotamer library [36] and interaction was analyzed by Find Clashes/Contacts tool in USCF Chimera software. Hydrogen bonds formed with Cys122 (C) and Arg200 (E) are shown by blue slid lines (left panels). ¹Wild-type human SULT1E1; PAP, (3’-phosphoadenosine 5’-phosphate); E2, (17β-Estradiol).