Sulfation of “Estrogenic” Alkylphenols and 17β-Estradiol by Human Platelet Phenol Sulfotransferases*

We have investigated the ability of alkylphenols to act as substrates and/or inhibitors of phenol sulfotransferase enzymes in human platelet cytosolic fractions. Our results indicate: (i) straight chain alkylphenols do not interact with the monoamine-sulfating phenol sulfotransferase (SULT1A3); (ii) short chain 4-n-alkylphenols (C < 8) are substrates for the phenol-sulfating enzymes (SULT1A1/2), which exhibit two activity maxima against substrates with alkyl chain lengths of C1–2 and C4–5; (iii) long chain 4-n-substituted alkylphenols (C ≥ 8) are poor substrates and act as inhibitors of SULT1A1/2; (iv) human platelets contain two activities, of low and high affinity, capable of sulfating 17β-estradiol, and 4-n-nonylphenol is a partial mixed inhibitor of the low affinity form of this activity. We conclude that by acting either as substrates or inhibitors of SULT1A1/2, alkylphenols may influence the sulfation, and hence the excretion, of estrogens and other phenol sulfotransferase substrates in humans.

Endocrine disrupters are exogenous substances in the environment which can influence endocrine function in humans and other animals (1). A number of these chemicals have estrogenic activity and are thus termed "xenoestrogens." Phytoestrogens are naturally occurring xenoestrogens produced by plants. Man-made xenoestrogens include the alkylphenols, nonylphenol and octylphenol, and bisphenol A; environmental exposure to these compounds has been reported to modify sexual development and reproductive function in amphibians (2,3), crustacea (4,5), and fish (6). In mammals, evidence is less clear, but there is widespread public concern that they may exert similar effects on human reproductive health and be involved in the initiation of some hormone-dependent cancers (7)(8)(9)(10)(11)(12)(13)(14).
17␤-Estradiol and alkylphenols share a common structural motif in the phenolic A ring of 17␤-estradiol and the phenol moiety of alkylphenols (Fig. 1), and it has been suggested that alkylphenols may act as endocrine disrupters by mimicking the activity of 17␤-estradiol at estrogen receptors. Indeed, in a variety of cells transfected with human estrogen receptors, alkylphenols have been shown to bind weakly and provoke modest estrogenic effects (15)(16)(17). Conversely, alkylphenols have been reported to promote estrogenic signaling by inhibiting androgen receptor activation in some tissues (18,19). However, there is also evidence that alkylphenols can disrupt en-docrine-mediated events by inhibiting enzymes involved in the metabolism of sex steroids. Thus exposure of rats to octylphenol during fetal or perinatal development has been shown to decrease the expression of P450 17␣-hydroxylase/C17-20 lyase (20,21), the enzyme system responsible for the transformation of C 21 steroids into C 19 steroids. Hence, exposure to alkylphenols may disrupt the production of both androgens and estrogens at key times during development.
By acting as structural mimetics, alkylphenols may also disrupt the elimination of steroids from the body. Estrogenic hormones are excreted from the body following metabolic conversion to biologically less active water-soluble metabolites via cytochrome P450-mediated hydroxylation, glucuronidation, Omethylation, or sulfation (22), and many of these pathways are also utilized for the metabolism of alkylphenols in fish and mammals (23)(24)(25)(26)(27). Exposure of the invertebrate Daphnia magna to either nonylphenol polyethoxylate or nonylphenol has been shown to disrupt endocrine function by decreasing both the glucuronidation and sulfation of testosterone (28), but it was not demonstrated whether the alkylphenols were substrates for the D. magna testosterone sulfotransferase in this study.
Sulfation has a major role in regulating the active concentrations of a variety of biologically important molecules, including steroids, catecholamines, and peptides, and it is also important in the detoxification of many xenobiotics (29,30). In humans, the balance of sulfation and desulfation plays an important role in modulating the activity and transport of steroid hormones: the "inactive" sulfated forms of many steroids, particularly estrone and dehydroepiandrosterone, are found in the circulation at concentrations 10 -30-fold higher than the unconjugated steroids. Furthermore, sulfation appears to prolong the half-life of these compounds in the circulation. Hence steroid sulfates represent an important depot of potentially "active" steroids following desulfation by steroid sulfatases (31)(32)(33). Hydrolysis of estrone sulfate by steroid sulfatase is the major source of plasma estrogens in men and postmenopausal women, and the activity of steroid sulfatase is elevated in breast tumors (33). In fact, there is a strong inverse correlation between the level of expression of steroid sulfatase within breast tumors and disease-free survival times (34). In many breast tumors, estrogen sulfotransferase activity is much lower than in normal breast tissue, and loss of this "inactivating" pathway may explain why tumor cells are extremely sensitive to the mitogenic effects of estradiol and estrone sulfate (35,36).
Sulfation reactions require PAPS 1 as a sulfate donor and are catalyzed by members of the SULT family. These enzymes are widely expressed in human tissues and have been classified into two broad subtypes: the phenol sulfotransferases (SULT1) and steroid sulfotransferases (SULT2) (29,30). The human SULT1 family is made up of at least six homodimeric enzymes with amino acid identities ranging from 47 to 96%. SULT1A3, sometimes called monoamine-sulfating phenol sulfotransferase, exhibits a strong substrate preference for catecholamines such as dopamine. Two closely related but kinetically distinct isoforms, SULT1A1 and SULT1A2, also known as phenol-sulfating phenol sulfotransferases, show a distinct preference for phenols such as 4-nitrophenol (30,37,38). On the basis of their structural similarity to these sulfotransferase substrates ( Fig. 1), it seems likely that alkylphenols may be substrates or inhibitors of either SULT1A1/2 and/or SULT1A3. Furthermore, SULT1A1/2 can sulfate 17␤-estradiol and other estrogens in human liver extracts and is the major route of estrogen sulfation in some breast cancer cell lines (39 -42). Hence, if alkylphenols do influence SULT1A1/2 activity, they might also be expected to interfere with estrogen sulfation in these, and possibly other, cancer cells.
Preparation of Platelet Cytosol Fractions from Platelet Concentrates-Apheresis packs of platelet concentrates (stored in acid citrate buffer at 22°C) were obtained within 5 days of donation from the National Blood Service, Birmingham, U. K. Red blood cell and lymphocyte contamination of the platelet concentrates was Ͻ Ͻ 0.1% (v/v). Platelets were collected by centrifugation at 8,000 ϫ g for 5 min at 4°C. The pellet was washed three times with 10 mM TES-buffered saline containing 4 mM EDTA (pH 7.0 at 4°C) and finally resuspended in one-eighth of the original pack volume of 10 mM TES storage buffer containing 0.25 M sucrose and 2 mM 2-mercaptoethanol (pH 7.0 at 37°C). The platelets were disrupted by ultrasonication for 3 ϫ 10-s bursts at 4°C and then centrifuged at 100,000 ϫ g for 60 min. The supernatant was stored in 1-ml portions at Ϫ20°C before use.
Measurement of Sulfotransferase Activity-Sulfotransferase activity was measured in the platelet cytosol fractions by assessing incorporation of 35 S into dopamine (SULT1A3) and 4-nitrophenol (SULT1A1/2), essentially as described in Refs. 43 and 44. These compounds are routinely used as standard substrates to assess SULT1A activities in many laboratories. To simulate in vivo conditions, TES buffer was used to adjust the pH of the assays to 7.0 at 37°C (45). It is probable that total platelet phenol-sulfating activity is a measure of SULT1A1 and SULT1A2 activity. Hence, we have adopted the nomenclature SULT1A1/2 in the text where there may be involvement of the SULT1A2 isoform. However, the concentration of 4-nitrophenol (6.7 mM) used in most of our experiments was well below the reported K m of SULT1A2 for this substrate (41); we therefore believe that under these conditions, the majority of measured phenol-sulfating phenol sulfotransferase activity was SULT1A1. Alkylphenols were dissolved and diluted in dimethyl sulfoxide, and the final concentration of the vehicle was 1% (v/v) in all incubations, including controls. Prior to use, the platelet homogenate was diluted with TES storage buffer so that the amount of PAPS consumed in the reaction was about 10% for screening enzyme activity and less than 5% for the determination of apparent kinetic constants. Platelet protein concentration was measured by the method of Bradford (46) using bovine serum albumin as standard.
Diluted homogenate (20 l) was incubated with 6.7 M PAPS (containing 6 -90 nCi of [ 35 S]PAPS) in a final volume of 150 l of 20 mM TES buffer containing 1% dimethyl sulfoxide. Standard substrates (4-nitrophenol for SULT1A1/2 and dopamine for SULT1A3, both at a saturating concentration of 6.7 M), SULT1A1/2 inhibitor (6.7 M DCNP), and alkylphenols (0.1-100 M) were added as appropriate. In some experiments, 17␤-estradiol was included at final concentrations ranging from 0.1-200 M. There is no known inhibitor that is specific for SULT1A3; however 6.7 M DCNP inhibits the SULT1A1/2-catalyzed sulfation of 6.7 M 4-nitrophenol by Ͼ 95% and has no significant effect on the sulfation of 6.7 M dopamine by SULT1A3. Therefore, SULT1A1/2 activity was calculated by subtracting sulfation in the presence of DCNP (SULT1A3 activity) from total sulfation in the absence of DCNP. Reactions were started by the addition of platelet cytosol fraction, followed immediately by PAPS, and incubated at 37°C for 40 min. The tubes were then transferred to ice and the reaction stopped by the addition of 0.1 M barium acetate (200 l). Any protein, unreacted PAPS, and free sulfate were precipitated by two additions of 0.1 M barium hydroxide (200 l) followed by 0.1 M zinc sulfate (200 l). The tubes were vortexed and centrifuged at 11,500 ϫ g for 3 min after each addition and the radioactivity measured in 500 l of the final supernatant. The total radioactivity added to each assay was determined in triplicate tubes containing radiolabeled PAPS and solvents only. The radioactivity recovered in each tube was corrected for radioactive decay and the results expressed relative to the activity against 6.7 M of the appropriate standard substrate to allow for variations in enzyme activity between platelet samples. Because many of the sulfotransferases are subject to substrate inhibition (discussed later), we undertook a number of preliminary experiments prior to the determination of the kinetic constants to ensure that the enzyme reactions were linear with respect to assay time over the desired concentration range. Kinetic parameters were estimated by nonlinear regressional analysis using the Enzfitter software package.

Alkylphenols Are Not Substrates for SULT1A3 in Platelet
Cytosol-Sulfotransferases accept a wide range of compounds as substrates, reflecting their role in xenobiotic detoxification. In view of the structural similarities between alkylphenols and known sulfotransferase substrates (Fig. 1), it seemed likely that these xenoestrogens may be able to influence the sulfation of other compounds by competing for multiple sulfotransferase isoforms or for the available pool of PAPS. To determine whether the alkylphenols are substrates for SULT1A3, a range of these compounds (both straight and branched chain) were incubated with platelet cytosol (Ϯ 6.7 M DCNP to inhibit SULT1A1/2 activity). Fig. 2, a and b, shows that, at a concentration of 6.7 M, none of the straight or branched chain 4-alkylphenols tested was a significant substrate for SULT1A3 because sulfation was consistently Ͻ 10% of that observed with a similar concentration of dopamine. These results concur with a recent study that reported that 4-methylphenol acted as a substrate for recombinant SULT1A1 but not for recombinant SULT1A3 (47). We have also examined 35 S incorporation into a limited number of commercially available alkylphenols (all at a final concentration of 6.7 M, Ϯ DCNP) with straight or branched alkyl chains on the 2-or 3-position of the phenol. In the presence of DCNP, 35 S was incorporated into 2-methylphenol, 2-ethylphenol, 2-n-propylphenol, 3-methylphenol, and 3-ethylphenol at approximately 10 -25% of the rate of 35 S incorporation into dopamine (Fig. 2b). At present we cannot distinguish whether these compounds were acting as poor substrates for SULT1A3 or were overcoming the DCNP inhibition and hence acting as weak substrates for SULT1A1/2. Fig. 3, a  and b (open bars), shows that total SULT1A3 activity against coincubations containing 6.7 M alkylphenols and 6.7 M dopamine was similar to that determined with dopamine as the sole substrate; the results in Fig. 2 clearly indicate that dopa- Human platelet cytosol fractions were incubated with alkylphenols or standard substrates at a final concentration of 6.7 M as described under "Experimental Procedures." SULT1A3 activity was determined in the presence of 6.7 M DCNP (open symbols/bars), and SULT1A1/2 activity was calculated as total sulfation in the absence of DCNP less that in the presence of the inhibitor (closed symbols/bars). Sulfation of the alkylphenols is expressed relative to sulfation of appropriate standard substrates (4-nitrophenol for SULT1A1/2 and dopamine for SULT1A3) in parallel incubations. Data in panel a are means Ϯ S.E. for 9-12 determinations from three separate platelet preparations. Data in panel b are means Ϯ S.E. from triplicate determinations from a single platelet preparation. mine sulfate is likely to be the major product in these coincubations.
Alkylphenols with Alkyl Chains of C Ͻ 8 Are Substrates for SULT1A1/2 in Platelet Cytosol-Phenol and a series of 4-nalkylphenols with straight alkyl chains increasing from 1 to 9 carbons in length were tested as substrates for SULT1A1/2. Fig. 2a shows that their ability to act as substrates for SULT1A1/2 was inversely related to the alkyl chain length and hence the hydrophobicity of the compounds. At a concentration of 6.7 M, 4-methylphenol and 4-ethylphenol were sulfated by SULT1A1/2 at a rate similar to that of a saturating concentration of 4-nitrophenol. However, the chain length/sulfation profile was biphasic: the rate of 4-n-propylphenol sulfation was significantly lower than that of either 4-ethyl-(p Ͻ 0.001) or 4-n-butylphenol (p Ͻ 0.01). A possible explanation for this biphasic profile is that SULT1A1 and SULT1A2 may have slightly different preferences for the short chain alkylphenols. Fig. 2b also shows that at a concentration of 6.7 M, 2-methylphenol, 2-ethylphenol, 2-n-propylphenol, 3-methylphenol, and 3-ethylphenol can be sulfated by human platelet SULT1A1/2. Unfortunately, alkylphenols with longer alkyl chains on the 2-or 3-positions were not commercially available so we were unable to investigate whether these compounds were SULT1A1/2 substrates or inhibitors. We also tested a limited number of branched chain alkylphenols as substrates for SULT1A1/2. Fig. 2b shows that their ability to act as substrates decreased as the bulk of the alkyl side chain increased. This is most clearly shown by the 4-butylphenol series, where the rates of sulfation of 6.7 M 4-n-butylphenol, 4-sec-butylphenol, and 4-tert-butylphenol were ϳ 90%, ϳ 65%, and ϳ 28%, respectively, of that of 6.7 M 4-nitrophenol.
To examine these phenomena further we compared the kinetic properties of the sulfation of 4-nitrophenol with those of a series of 4-substituted straight chain alkylphenols. The results (Fig. 4) show that the kinetic profiles for the sulfation of both 4-nitrophenol and two short chain 4-alkylphenols (4-methylphenol and 4-n-amylphenol) were essentially identical: as the substrate concentration was increased, the initial rates rose to a peak value before falling to a lower plateau. All experiments were performed at a saturating concentration of PAPS, so these results cannot be explained by depletion of this vital cosubstrate. Hence it is likely that binding of the substrate affects the initial rate of catalysis. A similar effect has recently been reported with SULT1A1/2 (48) and recombinant human estro- gen sulfotransferase (SULT1E) (49). Zhang and co-workers (49) have examined this phenomenon in some detail using recombinant human SULT1E and have shown that two estrogen molecules bind per SULT1E monomer. They suggest that the partial inhibition occurs after the binding of estrogen to an allosteric regulatory site. We suggest that 4-nitrophenol and short chain 4-alkylphenols may also interact with two common sites on each SULT1A1/2 monomer, thus influencing the initial rate of catalysis.
The kinetic data shown in Fig. 4 were fitted to an equation (Equation 1) describing the behavior of an enzyme system in which substrate molecules bind first to the active site to generate product and second to an allosteric site that inhibits catalysis (50).
The kinetic constants V max and K m were obtained from initial rate studies using low substrate concentrations at which inhibitory effects were negligible. For each substrate, V min was set at a value slightly below the rate of product accumulation during the plateau phase (49,50). The equation assumes that the PAPS concentration used in the assay was saturating and that substrate binding was at equilibrium. The best fit K i values for 4-nitrophenol, 4-n-ethylphenol, and 4-n-amylphenol were 5.2 Ϯ 0.8, 12.3 Ϯ 3.2, and 15.1 Ϯ 0.32 M, respectively. Hence the kinetic behavior of 4-nitrophenol and the short chain 4-alkylphenols (Fig. 4) is consistent with the possibility that, as has been suggested for the interaction of estradiol and SULT1E (49), these compounds act both as substrates for SULT1A1/2 and at an allosteric site that inhibits catalysis.
The kinetic parameters for a series of straight chain alkylphenols were determined at low concentrations so that the inhibitory influences of the allosteric regulatory site were negligible (Table I). Although the K m values obtained will be useful for comparing the substrate preferences of SULT1A1/2 for the different alkylphenols, they do not represent a direct measure of substrate affinity. Rather, they will be a compound reflection of several processes in the catalytic cycle and will be highly dependent on the rate constants for catalysis itself and for product release. Therefore, a better indicator of substrate preference is the ratio V max /K m which is known as the specificity constant (51). Because the absolute magnitude of V max for each substrate will vary slightly among platelet preparations, the ratio of the specificity constants for the alkylphenols relative to that of 4-nitrophenol measured in the same experiment can be used to determine the substrate preferences of human platelet SULT1A1/2 (Table I). Of the 4-substituted phenols, 4-methylphenol was the optimal SULT1A1/2 substrate, having a specificity constant slightly greater than that of 4-nitrophenol (as did 2-methylphenol and 2-ethylphenol), which is in agreement with the 35 S incorporation experiments shown in Fig. 2. 4-n-Propylphenol appeared to be the least favorable substrate, having a specificity constant just 5% that of 4-nitrophenol. As the alkyl chain length increased there was a slight recovery in activity relative to 4-nitrophenol; however, the specificity constants for 4-n-butylphenol and 4-n-amylphenol were still only 30% that of 4-nitrophenol. Because longer chain 4-n-alkylphenols were such poor substrates, it was not possible to obtain accurate specificity constants for these compounds.
4-Ethylphenol was a moderately good substrate for SULT1A1/2; the results in Table I and Fig. 3b suggest that it may compete with other substrates for this enzyme. This is of particular interest in view of recent evidence that 4-ethylphenol is a major metabolite of the dietary phytoestrogen genistein in rats and humans (52). The daily exposure of humans to genistein may be several hundredfold greater than that to environmental alkylphenols (53), but it remains to be determined if dietary phytoestrogens can generate sufficient 4-ethylphenol to compromise SULT1A1/2 activity in target tissues.
The conformation of the 4-alkylphenol chain also plays an important role in determining whether the alkylphenols were SULT1A1/2 substrates. This is best illustrated with the 4-butylphenol series (Table II); the specificity constant of 4-n-butylphenol was approximately 27% of that of 4-nitrophenol, whereas the relative specificity constants for 4-sec-butylphenol and 4-tert-butylphenol were ϳ 3% and Ͻ 1%, respectively. For each of these compounds the V max values were similar, but K m increased with increasing chain branching. Therefore, introduction of bulky side groups onto the 4-position may affect substrate binding or orientation within the active site.
We have been able to test the effects of changing the position of the alkyl chain relative to the phenolic hydroxyl group for both methylphenol and ethylphenol. Both of these compounds were substrates for SULT1A1/2 irrespective of the position of the alkyl group on the phenol ring (Fig. 2b). However, the kinetic data in Table I indicate that for both methyl-and ethylphenol, the 2-substituted compounds were the preferred substrates. It has been noted that the catecholestrogens, and 2-hydroxyestradiol in particular, are much better SULT1A1/2 substrates than is 17␤-estradiol (42). Perhaps having a small

FIG. 4. Partial substrate inhibition of SULT1A1/2-mediated sulfation of 4-nitrophenol (panel a) and short chain 4-n-alkylphenols (panel b).
The initial rate of 4-nitrophenol (panel a, E), 4- methylphenol (panel b, Ⅺ), or 4-n-amylphenol (panel b, ‚) sulfation was determined as a function of increasing substrate concentration (0.1-100 M). Each data point is expressed as the mean of duplicate determinations from at least three separate experiments and three separate platelet preparations. The data are expressed relative to the initial velocity of 6.7 M 4-nitrophenol sulfation measured in the same experiment. methyl-or ethyl-group adjacent to the phenolic hydroxyl group may serve as a structural mimetic of the catecholestrogen motif.
In most cases, coincubating 6.7 M branched chain alkylphenols with 6.7 M 4-nitrophenol did not significantly alter total 35 S incorporation into sulfated products, regardless of the position of substitution. The highly branched 4-tert-octylphenol had no discernible effect upon total sulfation in the presence of either standard substrate, even at a concentration of 100 M. However, coincubation of 4-nitrophenol with either 2-sec-propylphenol or 2-sec-butylphenol, both poor substrates for SULT1A1/2, caused a significant (Fig. 3b, ϳ 70%, p Ͻ Ͻ 0.001) inhibition of the accumulation of sulfated products. The structure of 2-sec-propylphenol is similar to that of the aspirin metabolite salicylic acid (2-hydroxybenzoic acid), which we have previously shown to be a highly selective inhibitor of 4-nitrophenol sulfation (54). Coincubation of 4-nitrophenol with 2-ethylphenol, 2-n-propylphenol, or 2-tert-butylphenol caused much less inhibition of total sulfation (ϳ 10 -20%), suggesting that although a branched chain is important for inhibition, it may also be necessary for the alkylphenol to adopt a planar configuration to interact with the enzyme(s).

17␤-Estradiol Is Sulfated in Human Platelet Cytosol by Two Separate Sulfotransferase Activities That Are Inhibited by 4-n-
Nonylphenol-It has been reported that human platelets do not express estrogen sulfotransferase activity (55). However, because 17␤-estradiol has been shown to be a SULT1A1/2 substrate in human liver, breast cancer cells, and fetal lung extracts (K m ϳ 2-5 M) (39 -42, 56), it seemed likely that it may be sulfated by a similar route in platelets. Fig. 5a shows that 17␤-estradiol (0.1-200 M) was sulfated in human platelet cytosol in a concentration-dependent fashion. The pattern of 17␤-estradiol sulfation was different from that

TABLE II Effect of chain conformation on the kinetic parameters of SULT1A1/2-mediated sulfation of 4-butylphenol compared with 4-nitrophenol
Enzyme activities were determined in duplicate samples from platelet cytosol preparations from three separate individuals as described under "Experimental Procedures." K m and V max values were calculated using the EnzFitter software program and are given as means Ϯ S.E. For further explanation, "Results and Discussion." seen with 4-nitrophenol and the short chain 4-alkylphenols. Thus the rate of sulfate incorporation increased with substrate concentration between 0.1 and 30 M; above 30 M the rate of sulfate incorporation reached a plateau that was approximately 20% of the maximal rate of 4-nitrophenol sulfation. We interpret these data to indicate that 17␤-estradiol is not able to influence the catalytic function of the enzyme by binding to the putative allosteric site. However, 17␤-estradiol sulfation does not appear to follow strict Michaelis-Menten kinetics. An Eadie-Hofstee transformation of the kinetic data in Fig. 5b shows that human platelets contain two enzyme species capable of sulfating 17␤-estradiol. One species had a relatively low affinity (K m 2.4 Ϯ 0.15 M) of the same order of magnitude as the reported K m of human SULT1A1/2 for 17␤-estradiol (39 -42). The second species had a much higher affinity for 17␤estradiol (K m ϭ 0.043 Ϯ 0.01 M); this value is approximately an order of magnitude greater than the reported K m (5 nM) of recombinant human estrogen sulfotransferase (49). Because the platelet preparations are essentially devoid of contamination with other blood cells it is likely that this "high" affinity activity is a true platelet protein. The identity of this enzyme is unknown at present, and its relatively low activity precluded further investigation in the present study.
We have examined the effects of 4-n-nonylphenol (1-20 M) on the sulfation of 17␤-estradiol in human platelets by the "low" affinity activity, which we presume to be SULT1A1/2. Despite its low affinity for 17␤-estradiol, this activity may be of significance in estrogen-dependent mammary tumors, where the local concentration of the hormone has been reported to approach 1 M in some cases (57). The transformations of the kinetic data describing the effects of 4-n-nonylphenol on estradiol sulfation by SULT1A1/2 are shown in Fig. 6; they demonstrate that 4-n-nonylphenol acts as a partial mixed inhibitor of SULT1A1/2-mediated 17␤-estradiol sulfation. This behavior is similar to that of several other known inhibitors of SULT1A1/2 activity including vanillin (58) and quercetin (59). We suggest that the most likely explanation of this complex enzymatic behavior is that 4-n-nonylphenol can bind to free SULT1A1/2 yielding an enzyme-inhibitor complex (dissociation constant K i ϭ 2.8 M) and also bind to the SULT1A1/2-17␤-estradiol complex giving an unreactive enzyme-substrate/inhibitor complex (dissociation KЈ i ϭ 5.4 M). Thus 4-n-nonylphenol may both competitively antagonize the binding of 17␤-estradiol to the active site and also interact noncompetitively with the allosteric regulatory site to decrease sulfation. Further work using purified recombinant SULT1A1/2 will be needed to establish the exact nature of the inhibition of 17␤-estradiol sulfation by 4-n-nonylphenol and other alkylphenols.
Agents that inhibit SULT1A1/2 activity, such as 4-n-nonyl-  (panel b). The initial rate of 17␤-estradiol sulfation was determined as a function of increasing substrate concentration (0.1-100 M). Each data point is the mean of duplicate determinations from at least three independent experiments using separate platelet preparations. All data are expressed relative to the initial velocity of 6.7 M 4-nitrophenol sulfation measured in the same experiments. phenol, may be able to increase local estradiol concentrations in those tissues in which SULT1A1/2 is the only route of steroid sulfation. This may be the case in hormone-dependent tumors of the breast. For example, the MCF-7 hormone-dependent breast cancer cell line has a very active estrone sulfatase pathway (34 -36), but no significant cytosolic estrogen or hydroxysteroid sulfotransferase activity can be detected (39 -41), and SULT1A1/2 is the only relevant pathway for steroid sulfation (42). If a similar situation exists in human mammary tumors, alkylphenols may be able to interfere with the "detoxification" of high levels of 17␤-estradiol in these tissues. This may be an important clinical factor because reduced estrogen and dehydroepiandrosterone sulfation in tumor homogenates from breast cancer patients is associated with a poor responsiveness to endocrine or ablative therapy and a grim clinical prognosis (60,61).
In conclusion, it appears that a number of alkylphenols may be sulfated by SULT1A1/2 and/or act as inhibitors of these enzymes. The sulfation of short chain alkylphenols probably represents a detoxification pathway, but in organisms subject to high levels of exposure to these compounds, extensive alkylphenol sulfation may reduce the sulfation and elimination of endogenous substrates such as estrogens by competing for the available pool of PAPS. Longer chain and branched chain alkylphenols, which are now widespread in the environment, are potent inhibitors of SULT1A1/2 and would be expected to reduce the sulfation of endogenous estrogens, especially in tissues such as mammary tumors where this pathway represents a major route of steroid sulfation and elimination (42). It is also possible that alkylphenol sulfates may themselves present a toxic threat to the organism by influencing the activity of a variety of endogenous sulfatases (62). The ability of alkylphenols to reduce estrogen sulfation may lead to a localized accumulation of free estrogens in tissues, thereby provoking the "environmental estrogenic" effects with which these compounds have become associated.