Mechanistic studies of β-arylsulfotransferase IV

Abstract

Sulfotransferases are an important class of enzymes that catalyze the transfer of a sulfuryl group to a hydroxyl or amine moiety on various molecules including small-molecule drugs, steroids, hormones, carbohydrates, and proteins. They have been implicated in a number of disease states but remain poorly understood, complicating the design of specific, small-molecule inhibitors. A linear free-energy analysis in both the forward and reverse directions indicates that the transfer of a sulfuryl group to an aryl hydroxyl group catalyzed by β-arylsulfotransferase IV likely proceeds by a dissociative (sulfotrioxide-like) mechanism. Values for the Brønsted coefficients (β_nuc and β_lg) are +0.33 and −0.45, giving Leffler α values of 0.19 and 0.61 for the forward and reverse reactions, respectively.

Sulfotransferases (STs) are an important class of enzymes that catalyze the transfer of a sulfuryl group from 3′-phospho-adenosyl-5′-phosphosulfate (PAPS) to a hydroxyl or, less frequently, an amine to give the sulfated product and the cofactor byproduct 3′-phosphoadenosyl-5′-phosphate (PAP) (Fig. 1). They can be divided into two groups: the cytosolic STs and the membrane-associated STs. Cytosolic STs catalyze the sulfonation of xenobiotics for detoxification and hormones for regulation. Membrane-associated STs catalyze the sulfonation of proteins and carbohydrates for processes such as cellular signaling and modulation of receptor binding (1–9). Recent studies have implicated the STs in a number of disease states including entry of the herpes virus (10), entry of HIV (11, 12), chronic inflammation (13), and various forms of cancer (1–9). For this reason there is considerable interest in the design of small-molecule inhibitors that are specific for a particular ST. Although a number of inhibitor designs and syntheses have been published, to date the inhibitors are relatively weak and have always been discovered through semirational methods because of the lack of a complete mechanistic picture (14).

Figure 1.

General reaction catalyzed by the STs.

A number of structural and mechanistic studies have been carried out on STs. Crystal structures have been solved for estrogen ST (15), catecholamine ST (16), dopamine ST (17), hydroxysteroid ST (18), and the ST domain of heparan sulfate N-deacetylase/N-ST-1 (19). Sequence alignment data from these structures as well as other STs show there to be a great deal of homology at the amino acid level for certain residues near the active site. Furthermore, mutagenesis studies carried out on these conserved residues show several of them to be critical to enzyme function. One such study of estrogen ST using a bound vanadate indicates an in-line attack of PAPS by the steroid hydroxyl group (20). Moreover, early studies of β-arylsulfotransferase IV (β-AST-IV) show a random, rapidly equilibrating bi-bi mechanism (21). In the kinase field, the transition state of a protein tyrosine kinase has been illuminated through the use of linear free-energy studies (22). This information then was used to design specific inhibitors of the enzyme by using the transition-state model developed (23).

In the present study, we sought to gain further insight into ST function using β-AST-IV as a model. This enzyme is readily accessible, easy to assay (24, 25), and shows high active-site homology with the other solved STs as well as other STs of interest (Fig. 2). Theoretical studies indicate that sulfuryl transfer can proceed through one of two transition states (Fig. 3) (26, 27). Substantial nucleophile participation can be operative, in which case there is an associative transition state or leaving-group ability that can dominate, leading to a dissociative (sulfotrioxide-like) transition state. In the case of an associative transition state, the charge build-up on the sulfuryl group could be stabilized by the active-site lysines. The dissociative transition state would have more charge build-up on the bridging oxygen of the phosphate group that could be stabilized by the active-site threonines. Studies of uncatalyzed sulfuryl transfer reactions indicate that the latter is likely operative (28–33). In an attempt to elucidate this mechanism, a linear free-energy analysis was performed in both the forward and reverse directions as well as pH-rate profiles on several of the substrates. The data indicate a dissociative (sulfotrioxide-like) transition state. In this case, the distance between the nucleophilic species and the electrophilic sulfur center is expected to be ≤3 Å. If an associative transition state were operative, this distance would be <2 Å. This knowledge of transition-state distance should aid in designing bisubstrate ST inhibitors.

Figure 2.

Sequence alignment of various cytosolic STs. The β-AST-IV primary sequence is aligned with those cytosolic STs for which crystal structures are available. Light-gray highlight, conserved residues near the active sites; dark-gray highlight, residues found to be critical in the active sites.

Figure 3.

Two possible reaction pathways for heparan sulfate N-ST-catalyzed sulfonation. (a) Associative pathway. In this case the majority of charge build-up would be on the sulfate group. The two lysines shown (614 and 833) could act to stabilize charge build-up. (b) Dissociative pathway. The charge build-up on the bridging phosphate oxygen could be stabilized by the indicated threonines (617 and 618).

Materials and Methods

Reagents.

All reagents used were purchased from Sigma–Aldrich except PAPS, which was purchased from Calbiochem. Reagents were used as purchased except where noted. Mutagenesis was done by using the QuikChange site-directed mutagenesis kit from Stratagene. Colorimetric measurements were carried out on a Beckman DU 650 spectrophotometer. Recombinant β-AST-IV was overexpressed and purified as described (24, 25).

Synthesis of Fluorinated Molecules.

All molecules used were synthesized from known procedures as shown in Scheme S1 (34, 35). Generally, all procedures were done in an inert atmosphere in flame-dried glassware. One equivalent of the appropriate 4-fluorophenol was dissolved in anhydrous DMSO, and 3 equivalents of sodium nitrite were added. The solution turned orange in a short time. Reactions were stirred at room temperature overnight. Water and concentrated HCl were added, and the resulting yellow solutions were refluxed for 5 min followed by cooling in an ice-water bath. The resulting solution was extracted three times with ether, and the pooled organic layers were washed with water and dried over magnesium sulfate. After removal of solvent in vacuo, the desired compounds were recrystallized in carbon tetrachloride. Yields ranged from 70% to 95%.

Scheme 1.

Synthesis of fluorinated molecules.

Sulfonation of the resulting p-nitrophenol (PNP) compounds has also been reported (34, 35). Briefly, all procedures were done under anhydrous conditions in an inert atmosphere. N,N-dimethylaniline (3.7 equivalents) was dissolved in carbon disulfide and cooled in an ice-water bath. Chlorosulfonic acid (1.4 equivalents) was added slowly followed by the appropriate fluorinated PNP (1.0 equivalents). The resulting viscous, orange slurry was heated to 35°C for 1 h followed by cooling to room temperature overnight. Four equivalents of potassium hydroxide in a 1.4 M solution then was added followed by two washes with toluene and two washes with ethyl acetate. Water was removed by rotary evaporation, and the resulting compound was purified by recrystallization from water. Yields ranged from 40% to 75%.

General Assay Conditions.

In the forward direction, all assays were conducted in 100 mM Tris, pH 6.8, with 5 mM 2-mercaptoethanol, 1 mM PAPS, and varied concentrations of the phenol in a final volume of 490 μl. The substrate mixture was allowed to equilibrate for 5 min, and the reactions were initiated with 10 μl of β-AST-IV solution to give a final enzyme concentration of 0.34 μM. Reaction progress was monitored at 405 nm for 5 min.†

In the reverse direction, all assays were conducted in 100 mM bis-trispropane, pH 7.6, with 5 mM 2-mercaptoethanol, 1 mM PAP, and varied concentrations of the phenol sulfate in a final volume of 490 μl. The reactions were initiated with 10 μl of β-AST-IV solution to give a final enzyme concentration of 0.34 μM. Reaction progress was monitored at 405 nm for 5 min.

Results and Discussion

pH-Rate Profile for the Forward Reaction.

Theoretical studies implicate a number of residues involved in stabilizing the transition state through proton donation and an active-site histidine that likely acts to remove a proton from the nucleophile (Fig. 3) (26, 27). In an effort to gain insight into the protonation state of the reaction, a series of pH-rate profiles was conducted (Fig. 4). In the forward direction, PNP, 2-fluoro-PNP, and 3-fluoro-PNP were used as substrates. The pH was varied between 4 and 10, but only values shown in Fig. 4a gave reproducible results, which is perhaps due to the instability of the enzyme at low and high pH values. Very little change in the second-order rate constant is seen over this range. It has been postulated that one of the first steps in the reaction is protonation of the leaving phosphate group, the pKa of which is likely in this range (26, 27). However, the acidity of this group would be increased greatly with coordination to the threonines shown in Fig. 3, masking the observation of a transition in this range.

Figure 4.

pH-rate profiles for β-AST-IV-catalyzed sulfuryl transfer in the forward and reverse directions. (a) k_cat/K_m vs. pH for PNP, 2-fluoro-4-nitrophenol (2F-PNP), and 3-fluoro-4-nitrophenol (3F-PNP) in the forward direction. (b) k_cat/K_m vs. pH for the reverse reaction using 4-nitrophenol sulfate. (c) k_cat/K_m for the reverse reaction using 4-methylumbelliferyl sulfate.

Because of the limitations in the forward direction, two pH-rate profiles were also conducted in the reverse direction by using p-nitrophenolsulfate (Fig. 4b) and 4-methylumbelliferyl sulfate (Fig. 4c) as the sulfuryl donors and PAP as the acceptor. An initial decrease in rate was seen after pH 5.0 with p-nitrophenolsulfate followed by a plateau from pH 6.2 to pH 9.2 and then a large drop in rate. A similar pattern was seen for 4-methylumbelliferyl sulfate except that measurements below pH 5.0 were possible, showing an optimal pH at 4.8 with a large drop at lower pH values. In the reverse direction, acid catalysis would be necessary to protonate the leaving aromatic alcohol, because the sulfate substrate used in the study has a relatively high pKa value for the leaving group. If histidine is acting as an acid to protonate the leaving aromatic alkoxide, microscopic reversibility would indicate this to be the base in the forward direction, although most of the phenol substrates used in the forward reaction are ionized under these conditions. The pKa of histidine (5.5–7.0) (36) does fall in the range of the transition seen in Fig. 4 b and c.

Mutagenesis of Active-Site Histidine.

Past mutagenesis studies of estrogen ST (20) have shown that the active-site histidine, conserved in all cytosolic ST, is necessary for catalytic activity. In an effort to gain understanding of the role of conserved His-104 in β-AST-IV, this residue was mutated to alanine and serine, and measurements of kinetic parameters were attempted. For both of the mutants the activity was below 6.0 × 10⁻⁷ μmol/min, the limit of detection. Attempts to measure activity were made by using PNP, the fluorinated molecules, umbelliferone, and imidazole, but all failed to recover activity. These data agree with those of the literature (20), but similar to other cases, it could be a different problem.

Effective Charges for the Forward Reaction.

A number of papers have been published regarding uncatalyzed sulfuryl transfer reactions (28–33). From the data for the uncatalyzed reaction, it is possible to analyze the charge build-up (effective charge) in the transition state and determine Leffler α parameters given by β_nuc/β_EQ, where β_EQ is determined by the difference of effective charges for the reactants and products (Fig. 5), and β_nuc is the Brønsted nucleophile coefficient. Fig. 5a shows the values for β_EQ and β_nuc from the literature giving a Leffler's α of 0.13 (0.23/1.74) (32). This low value of Leffler's α indicates little change in the effective charge on the nucleophile, demonstrative of little nucleophilic participation (37, 38). Fig. 5b shows the expected effective charge on the phenolic oxygen using the β_nuc value determined whether the nucleophile is the deprotonated phenol. Fig. 5c shows the effective charge on the phenolic oxygen of the transition state if the nucleophile is the neutral phenol. By using the β_EQ values shown in Fig. 5 b and c from the literature, the Leffler parameters are calculated to be 0.19 and 0.44, respectively.‡ Although there is precedent in the literature for substantial difference between uncatalyzed and enzyme-catalyzed effective charges (37, 38), making it impossible to rule out neutral phenol as the nucleophile, it seems unlikely that this is the case. A Leffler α parameter of 0.19 in Fig. 5b indicates very little O—S bond formation in the transition state. Furthermore, the effective charge on the phenolic oxygen is −0.67, in close agreement with the literature value (32). It should be noted that the leaving group used in the uncatalyzed studies was different from the current leaving group, and this could be contributing to the differences we have observed.

Figure 5.

Effective charge for the sulfuryl transfer reaction in the forward direction. (a) Expected value of β_nuc for the uncatalyzed reaction taken from the literature. In this case, LG represents substituted pyridines. (b and c) Expected charges in the transition state for a deprotonated and a protonated nucleophile using β_nuc values determined in this study, respectively.

Linear Free-Energy Analysis of the Forward Reaction.

To gain a better understanding of the transition-state structure of catalyzed sulfuryl transfer, a number of fluorinated PNPs with a wide range of pKa values were synthesized (Table 1). Fluorine substitution was chosen because of its electron-withdrawing ability while maintaining a van der Waal's radius similar to hydrogen (1.35 and 1.20 Å, respectively) (39, 40). By using a saturating amount of PAPS, the concentration of the various phenol nucleophiles was varied to determine the apparent second-order rate constants (k_cat/K_m). Under the reaction conditions, each of the phenols may not be in the fully deprotonated state, and thus the k_cat/K_m values shown for PNP and 2-fluoro-4-nitrophenol are adjusted to reflect the amount of unprotonated and neutral species by using Eq. 1 (41, 42),

1

Determination of the rate constants for the tri- and tetrafluoro species, the two most acidic species,§ was not possible. Apparent second-order rate constants were plotted as a function of the pKa of the nucleophile (Fig. 6a). The slope of this line gives β_nuc, which can be used to calculate the degree of charge build-up in the transition state and subsequently the amount of bond formation in the transition state (see above). It is noted that most of the phenol substrates used in the study are ionized almost completely under this condition. A reasonable linear correlation was seen, and β_nuc was determined to be 0.33. This value is slightly higher than the value reported in the literature for the uncatalyzed reaction (32). The calculated effective charge on the phenolic oxygen in the transition state is −0.77 for the uncatalyzed reaction and −0.67 for the catalyzed reaction. This discrepancy could be from the binding energy of the positively charged active site, which would lessen the amount of negative-charge build-up observed (37, 38, 43, 44). It is necessary in this type of analysis that the chemical step (dissociation of the sulfuryl group and attack by the nucleophile) is rate-limiting. Previous kinetic studies indicate this to be likely (21).

Table 1.

Kinetic and pK_a data for fluoronitrophenol compounds

^*Determined by using a TitraLab 90 and by manual titration. 

^† R = H; pH 6.8. 

^‡ R = H; pH 6.8. 

^§ R = SO₃; pH 7.6. 

Figure 6.

Linear free-energy analysis in the forward and reverse directions. (Left) The log (k_cat/K_m) is plotted as a function of pKa for the forward reaction. (Right) The same plot for the reverse reaction.

Effective Charges for the Reverse Reaction.

Analysis of the effective charges for the reverse reaction (Fig. 7) has also been conducted. Fig. 8a shows the values of β_EQ and β_lg taken from the literature (32), where β_lg is the Brønsted leaving-group coefficient. From this, a Leffler parameter of 0.87 is calculated. Two possible scenarios using the β_lg value determined in the present study are shown in Fig. 8 b and c. For the case shown in Fig. 8b, the Leffler parameter is 0.26 and for Fig. 8c, the Leffler parameter is 0.61. The former case seems unlikely, because it would indicate very little bond cleavage in the transition state as well as a positive effective charge on the phenolic oxygen. Therefore it seems probable that there is proton donation from one of the active-site residues, likely histidine. It is assumed that the species in the transition state has been stabilized by a proton from His-104.

Figure 7.

Reverse reaction catalyzed by β-AST-IV.

Figure 8.

Effective charge of sulfuryl transfer reaction in the reverse direction. (a) Expected value of β_lg for the uncatalyzed reaction from the literature. Nuc, substituted pyridines. (b and c) Effective charges calculated from β_lg determined for an unprotonated and a protonated transition state, respectively.

Linear Free-Energy Analysis for the Reverse Reaction.

By using the sulfated forms of the fluorinated PNPs listed in Table 1, the apparent second-order rate constants for the reverse reaction (shown in Fig. 7) were measured. Values for the three most acidic compounds were not determined, because they decomposed rapidly under the conditions used. The rate constants determined were plotted as a function of the pKa of the leaving phenol (Fig. 6b) to give a line with a slope equal to the Brønsted leaving-group coefficient (β_lg). The value determined in this study (−0.45) is substantially more positive than the value reported in the literature (−1.51) for the uncatalyzed reaction (32). This discrepancy can be explained through partial protonation of the leaving group in the active site giving a net positive partial charge to the phenolic oxygen.‡

Conclusion

In the present study, we examined the transition state of β-AST-IV-catalyzed sulfuryl transfer using linear free-energy analyses in both the forward and reverse directions. We also investigated the protonation state of active-site residues and substrate with a series of pH-rate profiles. The value obtained for β_nuc indicates little participation of the nucleophile in the transition state. The value obtained for β_lg is substantially lower than the value reported in the literature for the uncatalyzed reaction but seems likely if there is substantial protonation of the leaving group (perhaps through the active-site histidine acting as an acid in the reverse direction) in the transition state. All these data, coupled with past research, point toward a dissociative mechanism with only a small amount of nucleophile participation and substantial departure of the leaving group in the transition state (Fig. 9). We propose the forward reaction as follows: a sulfotrioxide-like high-energy species is largely generated followed by an in-line attack of the nucleophile on the sulfuryl group to form a late-transition state. The active-site His-104 may act as a base to remove the proton from the nucleophile and to ensure proper orientation of the nucleophile. Consistent with the microscopic reversibility principle and the above mechanistic studies, a sulfotrioxide-like species is also generated in the reverse reaction, and the leaving oxygen is likely being protonated by His-104. The nature of the proton transfer in both directions is dictated by the pKa of the nucleophile or the leaving group. Further investigation of the mechanism may be carried out in the future by structural studies and designed probes.

Figure 9.

Proposed mechanism indicating the transition-state structure, the nucleophile, and the participating histidine for the forward reaction.

Abbreviations

ST

sulfotransferase

PAPS

3′-phospho-adenosyl-5′-phosphosulfate

PAP

3′-phosphoadenosyl-5′-phosphate

β-AST-IV

β-arylsulfotransferase IV

PNP

p-nitrophenol