Monoalkyl sulfates as alkylating agents in water, alkylsulfatase rate enhancements, and the “energy-rich” nature of sulfate half-esters

Abstract

Alkyl sulfate monoesters are involved in cell signaling and structure. Alkyl sulfates are also present in many commercial detergents. Here, we show that monomethyl sulfate acts as an efficient alkylating agent in water, reacting spontaneously with oxygen nucleophiles >100-fold more rapidly than do alkylsulfonium ions, the usual methyl donors in living organisms. These reactions of methyl sulfate, which are much more rapid than its hydrolysis, are insensitive to the nature of the attacking nucleophile, with a Brønsted β_nuc value of −0.01. Experiments at elevated temperatures indicate a rate constant of 2 × 10⁻¹¹ s⁻¹ for the uncatalyzed hydrolysis of methyl sulfate at 25°C (t_1/2 = 1,100 y), corresponding to a rate enhancement of ≈10¹¹-fold by a human alkylsulfatase. Equilibria of formation of methyl sulfate from methanol and sodium hydrogen sulfate indicate a group transfer potential (ΔG′_pH7) of −8.9 kcal/mol for sulfate ester hydrolysis. The magnitude of that value, involving release of the strong acid HSO₄⁻, helps to explain the need for harnessing the free energy of hydrolysis of two ATP molecules in activating sulfate for the biosynthesis of sulfate monoesters. The “energy-rich” nature of monoalkyl sulfate esters, coupled with their marked resistance to hydrolysis, renders them capable of acting as sulfating or alkylating agents under relatively mild conditions. These findings raise the possibility that, under appropriate circumstances, alkyl groups may undergo transfer from alkyl sulfate monoesters to biological target molecules.

Dimethyl sulfate and other diesters of sulfuric acid are powerful alkylating agents, widely used in organic synthesis and chemically induced mutagenesis. Monoalkyl sulfates, in contrast, undergo spontaneous hydrolysis only at high temperatures (1) and appear never to have been shown to act as alkylating agents under any conditions. Many commercial detergents, formulated on the basis of that presumed lack of reactivity, incorporate SDS, also known as sodium lauryl sulfate, as their active ingredient. In living organisms, O-sulfated lipids and polysaccharides, produced through the intermediacy of 3′-phosphoadenosine 5′-phosphosulfate, are involved in maintaining cell structure. Sulfate monoesters also play a central role in recently discovered signaling functions that are sensitive to variations in their detailed structures (for recent reviews, see refs. 2 and 3).

The identification of roles for sulfuric acid monoesters in biological signaling and structure has aroused increasing interest in the chemical properties of these negatively charged molecules, and of the enzymes responsible for their biosynthesis and degradation. To appreciate the effectiveness of any enzyme as a catalyst, it is of interest to compare the rate of an enzyme reaction with the rate of a reaction proceeding in water in the absence of a catalyst. Recent work has shown that hydrolases that act on phosphate monoesters produce rate enhancements exceeding those produced by other enzymes that have been characterized in this way (4). It seemed desirable to obtain similar information about the proficiencies of sulfate ester hydrolases as chemical catalysts, by examining the hydrolysis of monoalkyl sulfate ester anions in neutral solution. We also wished to determine the group transfer potentials of sulfate monoesters, as an aid to understanding their roles in metabolism and the reason for the requirement for harnessing the free energy of hydrolysis of two ATP molecules in the course of activating sulfate for the biosynthesis of sulfate esters in living organisms (5).

The hydrolysis of monomethyl sulfate is catalyzed by acid (6–8) and alkali (8), and proceeds slowly in the neutral pH range (Fig. 1, top pathway) (8). Aryl esters behave somewhat similarly and undergo nucleophilic attack by amines and other nucleophiles at sulfur (Fig. 1, middle pathway) (9). In the present work, we began by determining the heats and entropies of activation associated with monomethyl sulfate hydrolysis in 1 M HCl and 1 M KOH. Monitoring the hydrolysis of monomethyl sulfate in buffered solutions of the sodium salt, we observed initial rates of uncatalyzed hydrolysis that did not vary over most of the pH range near neutrality. We also encountered rapid buffer-catalyzed decomposition of monomethyl sulfate, at rates far exceeding its rate of hydrolysis in unbuffered solutions. Further experiments showed that monoalkyl sulfates serve as efficient alkylating agents in water at moderately elevated temperatures. In the absence of enzymes, these half-esters react spontaneously with oxygen and nitrogen nucleophiles (Fig. 1, bottom pathway) more rapidly than do alkylsulfonium ions such as S-adenosylmethionine, the usual methyl donor in living organisms.

Fig. 1.

Pathways of decomposition of methyl sulfate in the presence of nucleophiles: hydrolysis (top pathway), nucleophilic attack at sulfur (middle pathway), nucleophilic attack at carbon (bottom pathway).

Results

When the hydrolysis of methyl sulfate (0.1 M) was conducted in 1 M HCl, over the temperature range between 40°C and 110°C, extrapolation to 25°C yielded a rate constant of 1.7 × 10⁻⁸ s⁻¹, with ΔH^‡ = 24.6 kcal/mol and TΔS^‡ = −3.9 kcal/mol. Experiments in 1 M KOH, conducted over the range between 40°C and 110°C, yielded an extrapolated rate constant of 8.3 × 10⁻⁸ s⁻¹ at 25°C, with ΔH^‡ = 18.8 kcal/mol and TΔS^‡ = −8.2 kcal/mol.

In solutions buffered in the range between 3 and 10, the appearance of methanol resulting from hydrolysis of sodium methyl sulfate (0.02 M) followed simple first-order kinetics. An Arrhenius plot of initial rate constants, observed over the range between 100°C and 190°C, was linear (Fig. 2). Extrapolation to 25°C yielded a rate constant of 2 × 10⁻¹¹ s⁻¹, with ΔH^‡ = 32 kcal/mol. The value of TΔS^‡ was −1.9 kcal/mol, a near-zero value consistent with unimolecular decomposition through a loose transition state. Similarly small entropies of activation have been reported for the decomposition of both the mono- and dianions of methyl phosphate (4).

Fig. 2.

Log₁₀ of the rate constant (s⁻¹) for hydrolysis of methyl sulfate (0.02 M) in potassium phosphate buffer (0.1 M, pH 6.8, at 25°C), plotted as a function of the reciprocal of absolute temperature in K.

In the presence of increasing concentrations of buffers, much higher rates of decomposition of methyl sulfate were observed. The buffer bases used in these experiments included ammonia, methylamine, dimethylamine and trimethylamine, the inorganic phosphate dianion and trianion, the methyl phosphate dianion, hydroxylamine, thiosulfate, and sulfite. In every case, proton NMR showed that the buffer base had been methylated. In the case of the inorganic phosphate trianion, a second methyl group began to be transferred to methyl phosphate before the concentration of methyl phosphate had reached a maximum. An apparent first-order rate constant for each of these reactions was estimated by multiplying the rate constant observed for decomposition of the substrate by the fraction of that reaction that led to the methylation product. Second-order rate constants for each of these reactions were then calculated by dividing the apparent first-order rate constant for reaction with the buffer by the concentration of buffer base that was a present. These experiments were performed over temperatures spanning a range of at least 70°C, and Arrhenius plots of the results were used to obtain the rate constants and activation parameters in Table 1. That table also includes approximate values observed for the uncatalyzed (s⁻¹) and acid-catalyzed (M⁻¹·s⁻¹) hydrolysis of methyl sulfate anion. In Fig. 3, the rate constants for methyl sulfate reactions with each of these nucleophiles, extrapolated to 25°C, are plotted as a logarithmic function of the pK_a value of the conjugate acid of each nucleophile.

Table 1.

Reactions of nucleophiles with methyl sulfate

B	pKa (BH+)*	k25°C, M−1·s−1	ΔG‡25°C, kcal/mol	ΔH‡, kcal/mol	TΔS‡25°C, kcal/mol
Nucleophile
NH3	9.24	5 × 10−8	27.3	16.5	−10.8
CH3NH2	10.62	1.4 × 10−6	25.4	16.1	−9.3
(CH3)2NH	10.64	1.6 × 10−6	25.3	16.1	−9.2
(CH3)3N	9.76	1.3 × 10−5	24.0	18.0	−6.0
HPO4−2	6.58	8 × 10−9	28.4	21.5	−6.9
PO4−3	12.8	2.4 × 10−7	26.4	18.4	−8.0
CH3OPO4−2	6.31	5 × 10−10	30.0	23.8	−6.2
NH2OH	5.96	8 × 10−7	25.7	18.2	−7.5
SO3−2	7.0	4 × 10−6	24.7	17.4	−7.3
S2O3−2	1.72	1.8 × 10−6	25.2	20.1	−4.1
OH−	15.7	8 × 10−8	27.0	18.8	−8.2
Hydrolysis
H2O (H+)		1.7 × 10−8	28.5	24.6	−3.9
H2O		2.2 × 10−11	31.9	30.0	−1.9

*See ref. 12.

Fig. 3.

ΔG^‡ for the reaction of nucleophiles (B) with the methyl sulfate anion (Table 1), plotted as a function of the pK_a value of the conjugate acid (BH) of the attacking nucleophile.

To compare the reactivity of monomethyl sulfate with the reactivities of other methylating agents in water, dimethylamine was used as a common acceptor. Table 2 shows that methyl sulfate surpasses both tetramethylammonium and trimethylsulfonium ions in its reactivity as a methylating agent. Table 3 shows a comparison of the reactivities of methyl, ethyl, and isopropyl sulfates, using the dianion of inorganic phosphate as a common acceptor. At 25°C, the reactivities of these methylating agents fall in the ratio 1:0.04:0.02. The relative magnitude of these values is consistent with the steric effects expected for S_N2 displacement (10).

Table 2.

Kinetics of methyl transfer to dimethylamine

Methyl donor	k25°C, M−1·s−1	ΔG‡25°C, kcal/mol	ΔH‡, kcal/mol	TΔS‡25°C, kcal/mol
Me2NH2+*	4 × 10−13	+34.4	+25.9	−8.5
Me4N+*	2 × 10−12	+33.6	+30.1	−3.5
Me3S+*	1.5 × 10−8	+28.1	+22.2	−5.9
MeSO4−	2 × 10−6	+25.3	+16.1	−9.2

*See ref. 21.

Table 3.

Kinetics of methyl transfer to PO₄⁻²

Methyl donor	k25°C, M−1·s−1	ΔG‡25°C, kcal/mol	ΔH‡, kcal/mol	TΔS‡25°C, kcal/mol
MeSO4−	8 × 10−9	+28.4	+21.5	−6.9
EtSO4−	3.6 × 10−10	+30.2	+23.7	−6.5
iPrSO4−	1.8 × 10−10	+30.6	+28.4	−2.2

To determine the equilibrium constant for hydrolysis of methyl sulfate, methanol (0.1 M) was incubated with varying concentrations of sulfuric acid (1–4 M) at 100°C, for 15 h, a period of incubation at least eight times longer than the half-time for approach to equilibrium in 1 M HCl at that temperature. At 100°C, and at temperatures ranging up to 150°C, the apparent equilibrium constant for the reaction shown in Fig. 4 remained constant with a value of 0.027 ± 0.002 M⁻¹, based on an assumed water activity of unity. Accordingly, we assume that the value of the equilibrium constant at 25°C is very similar. The resulting ΔG value for hydrolysis (−2.13 kcal/mol), corrected for the free energy of ionization of HSO₄⁻ (pK_a = 2.0) (11) at pH 7 (−6.8 kcal/mol), corresponds to a group transfer potential (ΔG′_pH7) (12) of −8.9 kcal/mol for sulfate ester hydrolysis at pH 7.

Fig. 4.

Equilibrium of hydrolysis of the methyl sulfate anion at pH 2, a pH value corresponding to the first pK_a value of methyl sulfate and the second pK_a value of H₂SO₄.

Discussion

The literature discloses no direct precedent for the present alkyl transfer reactions. However, both monoalkyl (13) and monoaryl (14) sulfates have been shown to undergo C-O cleavage in alkali, and Fig. 3 shows that the OH⁻ ion falls on the same line as other nucleophiles. The near-zero value of the slope of that line (β_nuc = 0.01) indicates that the present reactions occur at similar rates, implying that there is little if any development of positive charge on the nucleophile in the transition state for these reactions.

When using dimethylamine as a common acceptor, methyl sulfate is quite reactive compared with other methyl donors, surpassing the trimethylsulfonium ion by a factor of >100 (Table 3). That finding is of biological interest in view of the fact that the sulfonium ion S-adenosylmethionine is the usual methyl donor in biosynthetic reactions. The ease with which nucleophiles attack the methyl group of methyl sulfate raises the possibility that natural sulfate monoesters act, at least occasionally, as alkylating agents in cell walls, cell membranes, and other biological settings. Moderately elevated temperatures are required for the present model reactions to proceed at an observable rate in dilute solution in water. But that requirement might be offset, in some biological settings, by juxtaposition of donors and acceptors in a position conducive to reaction, in such a way as to permit alkyl transfer at ordinary temperatures. Accordingly, these findings open the possibility that oxygen or nitrogen nucleophiles may sometimes react with sulfated alcohol substituents in biological molecules, forming ethers or secondary amines with the release of inorganic sulfate.

The present results raise some concern, at least at first glance, that the dodecyl group of SDS (the active ingredient of many commercial detergents) might react with oxygen or nitrogen nucleophiles with potentially harmful consequences. The rate of reaction of alkyl sulfates with dimethylamine is reduced by increasing substitution at the alkyl group (Table 2), and Kurz (8) has shown that SDS undergoes spontaneous hydrolysis at 11% of the rate observed for methyl sulfate at 90°C. That lowered reactivity appears to reduce the likelihood that sulfate monoesters present in detergents transfer alkyl groups to biological targets to any great extent during their usual time of exposure.

To appreciate the rate enhancement produced by monoalkyl sulfate ester hydrolases, it is of interest to compare their k_cat values with the rate constant for spontaneous hydrolysis of monomethyl sulfate, 2 × 10⁻¹¹ s⁻¹ at 25°C. Of those few alkyl sulfate hydrolases whose kinetic properties have been reported, the best-characterized appears to be a recombinant human N-acetylglucosamine-6-sulfatase (EC 3.1.6.4), with a reported specific activity of 2.7 μmol·min⁻¹·mg⁻¹ (15). Based on a reported subunit molecular mass of 58 kDa, k_cat can be estimated as 2.6 s⁻¹. That turnover number would indicate a rate enhancement (k_cat/k_non) of 1.4 × 10¹¹-fold, a value that is unremarkable compared with the rate enhancements that have been established for other hydrolytic enzymes, which range from 10⁷-fold to 10²¹-fold (16). Based on a K_m value of 1.2 × 10⁻⁵ M for a trisaccharide substrate (15), the value of this enzyme's nominal affinity for the altered substrate in the transition state, k_non/(k_cat/K_m) can be estimated as 9 × 10⁻¹⁷ M.

Sulfatases are unusual among hydrolytic enzymes in that they employ the covalent hydrate of a formylglycine residue (formed posttranslationally from cysteine) as a nucleophile (17), so that substrate hydrolysis proceeds through a sulfuryl-enzyme intermediate. The intervention of a covalently bound intermediate precludes estimation of “transition state affinity” in the usual sense of that term, because there is no uncatalyzed reaction with which it would be appropriate to compare the value of k_cat/K_m for the enzyme reaction (18). If one could measure the rate of sulfuryl transfer to a gem-diol resembling the covalent hydrate of a formylglycine residue, it would be possible in principle to estimate an equilibrium constant for “transition state interchange” (19).

It is of interest to consider the thermodynamic driving force for alkylation by sulfate monoesters, as well as the free energy changes involved in their biosynthesis. At temperatures ranging from 100°C to 150°C, the apparent equilibrium constant for the formation of the methyl sulfate monoanion from the bisulfate monoanion (Fig. 4), CH₃OH + HSO₄⁻ → CH₃OSO₃⁻ + H₂O, is almost invariant, 0.027 M⁻¹ based on an assumed water activity of unity. Assuming that the same value applies at 25°C, the resulting ΔG value for hydrolysis (−2.13 kcal/mol), corrected for the free energy of ionization of HSO₄⁻ (pK_a = 2.0) (11) at pH 7 (−6.8 kcal/mol), corresponds to a group transfer potential (ΔG′_pH7) (12) of −8.9 kcal/mol for sulfate ester hydrolysis under physiological conditions. That value, which we believe may be typical of sulfate monoesters, is substantially more negative than the group transfer potentials of simple phosphoric acid esters (e.g., glucose 6-phosphate, ΔG′_pH7 = −3.3 kcal/mol) (12) or carboxylic acid esters (e.g., ethyl acetate, ΔG′_pH7 = −4.7 kcal/mol) (12). The much more negative group transfer potential of sulfate monoesters can be attributed, at least in part, to the much greater strength of the acid released by sulfate ester hydrolysis (HSO₄⁻, pK_a = 2.0) (11), when compared with the acids released by the hydrolysis of monoesters of phosphoric (e.g., CH₃HPO₄⁻, pK_a = 6.3) (12) or carboxylic (e.g., CH₃COOH, pK_a = 4.8) (12) acids.

In their early characterization of 3′-phosphoadenosine 5′-phosphosulfate, Robbins and Lipmann (5) drew particular attention to the use of, and apparent need for, two molecules of ATP in the process of generating activated sulfate in the biosynthesis of sulfate esters. Those investigators showed that the sulfate group transfer potential of 3′-phosphoadenosine 5′-phosphosulfate is 10–11 kcal/mol more negative than for the cleavage of ATP to adenosine 5′-phosphate and inorganic phosphate, and amounts to ≈18 kcal/mol (20). The magnitude of the group transfer potential of methyl sulfate (ΔG′_pH7 = −8.9 kcal/mol), indicated by the present findings, helps to explain why, in intermediary metabolism, it is necessary to harness the free energy of hydrolysis of two ATP molecules (trapped in the intermediate 3′-phosphoadenosine 5′-phosphosulfate) to synthesize a single sulfuric acid monoester. The free energy of hydrolysis of a single molecule of ATP (ΔG′_pH7 = ≈7 kcal/mol in the presence of excess Mg⁺²) (12) would not suffice to drive the formation of methyl sulfate (ΔG′_pH7 = −8.9 kcal/mol for hydrolysis) to completion.

We conclude that the “energy-rich” nature of monoalkyl sulfates, coupled with their kinetic resistance to hydrolysis, renders them capable in principle of acting as alkylating agents in neutral aqueous solution. It remains to be seen whether such reactions occur in biological systems.

Materials and Methods

Most experiments were conducted in aqueous solution in sealed quartz tubes (5-mm outside diameter) at elevated temperatures ranging from 50°C to 150°C. Above pH 10, where quartz dissolves at elevated temperatures, reactions were conducted in microcentrifuge tubes (1.5-ml capacity) or Teflon-lined acid digestion bombs (Parr Instrument, no. 276AC). Reaction mixtures were placed in a Thermolyne convection oven, with the temperatures controlled within a range of excursion of 1.5°C, and were removed at intervals for analysis by proton NMR. Aliquots of the reaction mixture were then diluted 50-fold with D₂O that contained pyrazine (0.01 M), added as an integration standard. The disappearance of the sulfate ester and the appearance of products were monitored by using the integrated intensities of the methyl protons, and the concentrations of reactants and products were determined by comparing their intensities with that of the added pyrazine standard.

In the presence of buffers (typically 0.1 M), the hydrolysis of monoalkyl sulfate anions (typically 0.02 M) followed first-order kinetics in the pH range between 3 and 10. In these experiments, the disappearance of the sulfate monoester followed first-order kinetics but became more rapid with increasing buffer concentration. Apparent first-order rate constants were obtained by using the equation k = ln(A_o/A)/t. In buffered solutions, hydrolysis took place in competition with methylation of the buffer base. Product ratios were found to remain constant during each experiment, and apparent first-order rate constants for hydrolysis and for methylation were obtained by dividing the overall rate constant observed for disappearance of the sulfate monoester into two rate constants (one for hydrolysis and one for methylation), using the constant ratio observed between the two sets of products. Second-order rate constants for methylation were obtained by dividing those apparent first-order rate constants by the concentration of the buffer nucleophile.