Abstract
Sulfation is an important reaction in the biotransformation of steroid hormones, neurotransmitters, drugs, and other xenobiotics, yet little is known about the effects of organic solvents on sulfotransferase (SULT) activities in vitro. Initial experiments found that surprisingly low levels of solvent had dramatic effects on sulfotransferase activity. Consequently, we evaluated the effects of five commonly used solvents (methanol, ethanol, acetonitrile, dimethyl sulfoxide, and dimethyl formamide) on activities of cDNA-expressed sulfotransferase isozymes 1A1 (4-nitrophenol sulfation), 1A3 (dopamine sulfation), 1E1 (ethynylestradiol sulfation), and 2A1 (dehydroepiandrosterone sulfation). In addition, 1-hydroxypyrene was used as a general fluorescent probe for all four sulfotransferase isoforms examined. When substrates were present at their respective isoform-specific Km values, methanol and ethanol (0.4%, v/v) generally had less effect than acetonitrile, dimethyl sulfoxide, and dimethyl formamide on sulfotransferase activities. Acetonitrile, a commonly used solvent in cytochrome P450 studies, inhibited SULT1A1 activities (∼40%) at 0.4% (v/v), but activated SULT1E1-mediated 1-hydroxypyrene sulfation ∼2.6-fold. Assuming a two-site kinetic model, studies revealed that solvent affected Vmax1,Vmax2, and the Ki value of 1-hydroxypyrene sulfation mediated by SULT1E1. In contrast, the Km value was not affected, suggesting that solvent may potentially alter binding interactions of the second substrate molecule, but not the first. Additional experiments with expressed SULT1A1, supplemented with control protein, revealed that the inhibitory effect of solvent (0.4%, v/v) was reduced to <15% for all solvents examined. Thus, it is recommended that ethanol is used as the preferred solvent vehicle and that incubations with expressed enzyme contain >12 μg/ml total protein.
Sulfate conjugation is an important pathway in the detoxication of xenobiotics as well as in the modulation of endogenous compounds such as thyroid hormones, steroids, and neurotransmitters (Coughtrie et al., 1998). This reaction is catalyzed by sulfotransferases (SULTs1), a family of cytosolic enzymes (Nagata and Yamazoe, 2000). Sulfotransferases facilitate the transfer of sulfate from a cosubstrate, 3′-phosphoadenosine-5′-phosphosulfate (PAPS), to an acceptor molecule forming a sulfate conjugate. In human, at least 11 members of SULT have been identified and divided into several gene families based on their amino acid sequence and substrate specificity. Members of SULT1 family (divided into four subfamilies: 1A, 1B, 1C, and 1E) sulfate primarily phenols and catechols, whereas SULT2 family members (from subfamilies 2A and 2B) catalyze sulfation of steroids. SULT4A1, the only member of the SULT4 family, is an orphan SULT selectively expressed in brain tissues with unknown function (Falany et al., 2000; Walther et al., 2000).
Traditionally, the major focus of drug metabolism has been directed toward cytochrome P450, a family of enzymes that catalyze the oxidation of a variety of endogenous and exogenous compounds. However, there has been a growing interest in the role that phase II enzymes, such as sulfotransferases and UDP-glucuronyl transferases, may play in drug metabolism (Meloche et al., 2002; Prueksaritanont et al., 2002). Consequently, it is important that in vitro conditions for phase II enzymes are selected such that accurate kinetic values can be obtained for extrapolation to the in vivo condition. Many substrates for sulfotransferase are by nature lipophilic compounds (e.g., steroids) and thus the use of organic solvent as a vehicle is often one condition necessitated for in vitro studies. Although several reports have dealt with the effects of organic solvents on cytochrome P450 activities (Chauret et al., 1998; Busby et al., 1999; Tang et al., 2000) and UDP-glucuronyl transferase activities (Dehal et al., 2002), similar analysis has not yet been conducted for sulfotransferases. The effect of solvent on sulfotransferase activity in hepatocyte cultures has been reported, but no isoform selective probes were used in these studies (Easterbrook et al., 2001).
In the present study, five commonly used solvents, methanol, ethanol, acetonitrile, dimethyl sulfoxide, and dimethyl formamide, were evaluated for their effects on cDNA-expressed SULT activities. Assays were performed on known substrates as well as a general fluorescent probe, 1-hydroxypyrene.
Materials and Methods
Chemicals. PAPS, 1-hydroxypyrene, formic acid, ammonium formate, 4-nitrophenol, 4-nitrophenol sulfate, dopamine, dehydroepiandrosterone (DHEA), sodium EDTA, and sodium 1-decanesulfonic acid were purchased from Sigma-Aldrich (St. Louis, MO). [3H]Dopamine, [3H]ethynylestradiol, and [3H]DHEA were obtained from PerkinElmer Life Sciences (Boston, MA). All solvents and other reagents were of analytical grade or higher.
Human Tissue Cytosol and cDNA-Expressed Sulfotransferases. Cytosol prepared from human liver and jejunum were purchased from XenoTech, LLC (Lenexa, KS) and Tissue Transformation Technologies (Edison, NJ), respectively. Recombinant human sulfotransferase prepared from insect cells, infected with baculovirus containing human SULT cDNA, was obtained in-house (SULTs 1A1 and 1A3; Merck Research Laboratories, West Point, PA) or from PanVera Corp. (SULTs 1E1 and 2A1; Madison, WI).
Incubations. Incubation mixtures (0.5 ml) contained 100 mM sodium phosphate (pH 7.4), 1 mM magnesium chloride, 0.19 to 80 μg of cytosolic protein, 0 to 10 μl of solvent, 50 μM PAPS, and probe substrates. 4-Nitrophenol (1 μM), dopamine (10 μM), ethynylestradiol (6 nM), and DHEA (2 μM) were used as marker substrates for SULTs 1A1, 1A3, 1E1, and 2A1, respectively. 1-Hydroxypyrene was used at the concentration of 10 nM, 1 μM, 20 nM, and 2 μM for SULTs 1A1, 1A3, 1E1, and 2A1, respectively. Concentration of the marker substrates was chosen based on the Km values reported in the literature (Falany et al., 1989; Veronese et al., 1994) or determined in-house (Table 1) (Schrag et al., 2002). For ethynylestradiol and DHEA, stock solutions in organic solvent (by the vendor) were pipetted into clean tubes and then dried under nitrogen. The dried substrate was resuspended in the incubation mixture, followed by sonication for 1 min. The recovery of the substrate in the incubation mixture was determined to be >95%. The amount of cytosolic protein used for 1-hydroxypyrene sulfation in expressed SULTs 1A1, 1A3, 1E1, and 2A1 was 0.4, 1.38, 0.19, and 0.9 μg, respectively. In addition, the activity of SULTs 1A1 (4-nitrophenol), 1A3 (dopamine), 1E1 (ethynylestradiol), and 2A1 (DHEA) was determined using 1.6, 20, 0.19, and 6.75 μg of protein, respectively. Sulfation of 4-nitrophenol and 1-hydroxypyrene was performed using 10 and 2 μg of human liver cytosol, whereas 80 and 6 μg were used in human jejunum preparations, respectively. The reaction was initiated with the addition of PAPS after a 2-min preincubation at 37°C. Acetonitrile (0.1 ml) was added to stop the reaction at 5 to 30 min. The samples were centrifuged at 2000g for 10 min, and the supernatant fractions were analyzed by HPLC. Standard curves of the sulfation product were run at the beginning and end of each HPLC analytical run to confirm metabolite stability. For the kinetic studies of 1-hydroxypyrene sulfation, 0.2 to 1.4 μg of cytosolic protein were incubated with 4 nM to 20 μM substrate for 8 min. Substrate consumption was <8% in all incubations performed. The formation rate of the sulfate product was linear with respect to incubation time as well as the amount of cytosolic protein used. To estimate the amount of 1-hydroxypyrene sulfate, a standard curve was prepared by incubating a known concentration of 1-hydroxypyrene in the presence of PAPS and SULT1A1 and letting the reaction go to completion. The resulting yield of 1-hydroxypyrene sulfate was >97% in all cases.
HPLC Analysis. HPLC analyses were conducted with an Agilent 1100 LC system (Agilent Technologies, Palo Alto, CA), equipped with an auto-injector, a fluorescence detector, and a diode array detector. For tritiated samples, column eluent was analyzed by a Packard (500TR, PerkinElmer Life Sciences) or IN/US (β-ram 2, IN/US Systems, Inc., Tampa, FL) online radiochemical detector equipped with a liquid flow cell. The flow of scintillation cocktail (Packard Ultima-Flo M) was maintained at 3 ml/min. Separation of parent substrate and its sulfate conjugate was achieved by the following methods.
4-Nitrophenol sulfation. 4-Nitrophenol and its sulfate were separated using a method developed by Duescher and Elfarra (1993) with modifications. Analyte was eluted through a Zorbax XDB-C18 column (4.6 mm × 25 cm, 5 μm; Agilent Technologies) at a constant flow of 1 ml/min with 30% (v/v) acetonitrile in water containing 0.1% trifluoroacetic acid. Column effluent was monitored by UV absorbance at 300 nm.
Dopamine sulfation. An HPLC method using sodium decanesulfonate as the ion-pairing agent was adapted to separate dopamine and its sulfate (Hollenbach et al., 1998). Separation was performed on a Spherisorb ODS-2 column (4.6 mm × 25 cm, 5 μm; Thomson Instrument Company, Oceanside, CA). The mobile phase consisted of 0.15 M sodium phosphate (pH 4.0) with 10 mM sodium decanesulfonate and 0.1 mM sodium EDTA in water (solvent A) and methanol (solvent B) at a constant flow of 1 ml/min. The solvent gradient initiated at 5% B for 5 min and then increased linearly to 80% B in 5 min. After isocratic flow (80% B) for 5 min, the gradient returned to 5% B in 0.5 min. The column was re-equilibrated at initial conditions for 7.5 min before injection of the next sample.
Ethynylestradiol sulfation. Ethynylestradiol and its sulfate (3 position) were separated using a Zorbax SB-C8 column (4.6 mm × 7.5 cm, 3.5 μm). The mobile phase consisted of 25 mM ammonium formate, pH 3.0 (solvent A) and 0.1% formic acid in acetonitrile (solvent B). The solvent gradient started from 30% B at 1 ml/min and increased linearly to 50% B in 8 min. The gradient returned to 30% B in 0.2 min, followed by re-equilibration for 4.8 min.
DHEA sulfation. An HPLC method similar to that described above for ethynylestradiol was used with minor modifications. A Metachem Inertsil ODS-3 column (2.1 mm × 5 cm, 5 μm; Ansys Technologies Inc., Lake Forrest, CA) was used with a solvent gradient that increased from 10% B to 80% B in 8 min, and then returned back to 10% B in 0.2 min. The column was re-equilibrated for 4.8 min before the next injection.
1-Hydroxypyrene sulfation. The method established for ethynylestradiol was slightly modified for 1-hydroxypyrene sulfation. The solvent gradient increased from 45% B to 80% B in 0.5 min. After holding at 80% B for 4 min, the gradient returned to 45% B in 0.5 min. The column was then re-equilibrated at initial conditions for 2 min before injection of the next sample. Fluorescence detection was performed with excitation and emission wavelengths set at 340 nm and 390 nm, respectively.
Determination of Kinetic Parameters. With the exception of 1-hydroxypyrene sulfation in SULT2A1, data from the kinetic studies were fitted into the equation derived from a two-substrate model by Zhang et al. (1998).
It is important to note that the Km value for the binding of the second substrate is denoted as Ki in this equation. Data from SULT2A1-mediated 1-hydroxypyrene sulfation were fitted into the Michaelis and Menten equation. Curve-fitting was performed using SigmaPlot (SPSS Inc., Chicago, IL).
Results and Discussion
In the present study, probe substrates were used at concentrations corresponding to isoform-specific apparent Km values. Dopamine, ethynylestradiol, and DHEA sulfates were detected using [3H]-labeled material, whereas the sulfate of 1-hydroxypyrene was detected with fluorescence. 1-Hydroxypyrene was a substrate for all the sulfotransferase isoform tested. Thus, it was developed as a sensitive, convenient, and universal probe to monitor sulfotransferase activity.
Kinetic studies of 1-hydroxypyrene sulfation by different SULT isozymes revealed that substrate inhibition, a phenomenon commonly observed in sulfate conjugation (Sundaram et al., 1989), was detected in reactions catalyzed by SULTs 1A1, 1A3, and 1E1, but not SULT2A1 (Fig. 1). Since the consumption of 1-hydroxypyrene was <8% in all incubations performed, it is unlikely that the formation and accumulation of 3′, 5′-diphosphoadenosine caused the observed inhibition. Based upon a previous report in the literature by Zhang et al. (1998), a two-site model was selected to describe the kinetic data observed. The use of this model is further supported by the recent observation of two 4-nitrophenol molecules within the active site of crystalized SULT1A1 (Gamage et al., 2003). Moreover, it was also observed that the corresponding SULT1A1-mediated sulfation of 4-nitrophenol yielded substrate inhibition kinetics, similar to that observed for sulfation of 1-hydroxypyrene. The apparent Km, Ki, and Vmax values for 1-hydroxpyrene sulfation are reported in Table 1. It is interesting to note that the Km value of the first binding site ranged from 8 nM to 3 μM, indicating that 1-hydroxypyrene has relatively high affinity for all four SULT isoforms tested.
The effect of different solvents on cDNA-expressed SULT activities was shown in Fig. 2. In general, the alcohols, methanol and ethanol, had relatively less inhibitory effects on SULT activities than acetonitrile, dimethyl sulfoxide, and dimethyl formamide. Statistical analysis (analysis of variance test) revealed that ethanol had the least inhibitory effect on SULT1A1-mediated 1-hydroxypyrene sulfation among the five solvents tested (p < 0.001). In contrast, dimethyl formamide inhibition of 1-hydroxypyrene sulfation catalyzed by SULT1A3 (p < 0.001) and SULT2A1 (p < 0.05) was more significant than the other test solvents. Acetonitrile, the preferred solvent in cytochrome P450-mediated reactions (Chauret et al., 1998; Busby et al., 1999), significantly inhibited SULT1A1 activity by ∼40%, but activated SULT1E1-mediated 1-hydroxypyrene sulfation ∼2.6-fold (p < 0.001).
Ethanol was chosen as a probe solvent to study the concentration-dependent inhibition of sulfotransferase activities (Fig. 3). In all cases, the inhibition of sulfotransferase activity increased with ethanol concentration. Generally, 0.1 and 0.5% (v/v) ethanol inhibited SULT activities up to 10 and 20%, respectively. Ethanol content at 2% (v/v) was sufficient to inhibit ≥40% of the SULT1A1 activity and ≥60% of the SULT1E1 activity, whereas ≤25% activity was inhibited for SULT1A3 and SULT2A1, indicating SULTs 1A1 and 1E1 are more susceptible to ethanol than other SULT isoforms. In addition, 1-hydroxypyrene sulfation appeared to be slightly more sensitive to ethanol, especially at 1 to 2% (v/v), as compared with other known substrates.
The acetonitrile-mediated activation of 1-hydroxypyrene sulfation (SULT1E1) was examined with additional kinetic studies. As acetonitrile content was increased from 0 to 0.4% (v/v), the apparent first and second Vmax values, determined by fitting data into a two-substrate binding model, increased 58 and 328%, respectively (Fig. 4 and Table 2). However, both Vmax values dropped >50% when the acetonitrile content was further increased to 2 from 0.4% (v/v). Interestingly, the apparent Km value remained at ∼21 nM in spite of the changes in Vmax values. For comparison, the effect of ethanol on 1-hydroxypyrene sulfation was also studied. In agreement with the results shown in Fig. 2C, SULT1E1 activity was inhibited by ethanol in a concentration-dependent manner with a resulting ∼75% drop in both of the Vmax values in the presence of 2% (v/v) ethanol. However, the apparent Km value was not significantly affected by ethanol at concentration up to 2% (v/v), similar to that observed in acetonitrile. In the context of a two-site model, these results suggest that the impact of solvent lies on binding of the second substrate molecule (Ki) and its effects on productive catalysis (Vmax). It is also noteworthy that effects on Vmax could be due to a number of factors such as protein conformational change or altered affinity/binding of PAPS.
In addition to experiments with expressed enzyme, the effect of solvent was investigated in tissue cytosol. Substrates for SULT1A1, 4-nitrophenol and 1-hydroxypyrene, were chosen for this study because expressed SULT1A1 was relatively more sensitive to solvents as compared with other SULT isoforms (Fig. 1). In general, SULT1A1 activity in cytosolic preparation of human liver (Fig. 5A) and jejunum (Fig. 5B) was relatively more resistant to solvents [up to 0.4% (v/v)] than that determined for recombinant SULT1A1. To determine whether sulfotransferase is stabilized by the higher level of protein in the tissue preparations (5- to 50-fold higher, relative to expressed enzyme), the solvent effect on 4-nitrophenol and 1-hydroxypyrene sulfation in expressed SULT1A1 was further examined. Incubations containing expressed SULT1A1 were supplemented with cytosolic protein from control insect cells (12 μg/ml final concentration). The inhibitory effect of solvent [0.4% (v/v) methanol, ethanol, acetonitrile, dimethyl sulfoxide, or dimethyl formamide] on SULT1A1 activity was reduced to <15% of control values in the presence of supplemented protein (data not shown), indicating that the addition of control cytosolic protein can counterbalance the effect of solvents. These results are important because, in many cases, the specific activity of sulfotransferase in an expressed system is substantially greater than that found in native tissue cytosol. As a consequence, to ensure linear product formation conditions, expressed enzyme generally is diluted and the incubation time shortened to reduce substrate consumption. In such cases, it would therefore be prudent to supplement incubations with control cytosol to minimize solvent effects that would otherwise be more profound at lower protein concentrations. The degree to which this will be an issue will depend on the rate of substrate turnover for any given substrate and the sensitivity of the analytical assay.
The present study demonstrates that solvent can have substantial effects, either in the form of activation or inhibition, on cDNA-expressed human sulfotransferase activities in vitro. When the use of solvent and low protein concentrations are unavoidable, ethanol at ≤0.5% (v/v) is recommended. It is also recommended that incubations maintain a final incubation protein concentration > 12 μg/ml to avoid the more profound effects of solvent that are observed at lower protein concentrations.
Acknowledgments
We thank Dr. Tom Rushmore for cloning and expressing the human SULTs 1A1 and 1A3 genes.
Footnotes
-
↵1 Abbreviations used are: SULT, sulfotransferase; PAPS, 3′-phosphoadenosine-5′-phosphosulfate; DHEA, dehydroepiandrosterone; HPLC, high-performance liquid chromatography.
- Received May 5, 2003.
- Accepted August 21, 2003.
- The American Society for Pharmacology and Experimental Therapeutics