C-3- and C-4-Substituted Bicyclic Coumarin Sulfamates as Potent Steroid Sulfatase Inhibitors

Synthetic routes to potent bicyclic nonsteroidal sulfamate-based active-site-directed inhibitors of the enzyme steroid sulfatase (STS), an emerging target in the treatment of postmenopausal hormone-dependent diseases, including breast cancer, are described. Sulfamate analogs 9–27 and 28–46 of the core in vivo active two-ring coumarin template, modified at the 4- and 3-positions, respectively, were synthesized to expand structure–activity relationships. α-Alkylacetoacetates were used to synthesize coumarin sulfamate derivatives with 3-position modifications, and the bicyclic ring of other parent coumarins was primarily constructed via the Pechmann synthesis of hydroxyl coumarins. Compounds were examined for STS inhibition in intact MCF-7 breast cancer cells and in placental microsomes. Low nanomolar potency STS inhibitors were achieved, and some were found to inhibit the enzyme in MCF-7 cells ca. 100–500 more potently than the parent 4-methylcoumarin-7-O-sulfamate 3, with the best compounds close in potency to the tricyclic clinical drug Irosustat. 3-Hexyl-4-methylcoumarin-7-O-sulfamate 29 and 3-benzyl-4-methylcoumarin-7-O-sulfamate 41 were particularly effective inhibitors with IC50 values of 0.68 and 1 nM in intact MCF-7 cells and 8 and 32 nM for placental microsomal STS, respectively. They were docked into the STS active site for comparison with estrone 3-O-sulfamate and Irosustat, showing their sulfamate group close to the catalytic hydrated formylglycine residue and their pendant group lying between the hydrophobic sidechains of L103, F178, and F488. Such highly potent STS inhibitors expand the structure–activity relationship for these coumarin sulfamate-based agents that possess therapeutic potential and may be worthy of further development.


■ INTRODUCTION
Breast cancer is a major health threat to women of all age groups and a prime contributor to cancer deaths in women. About two-thirds of cases when first diagnosed are classified as hormone-dependent (ER + ), in which to grow and develop the tumors need estrogens, which act via the estrogen receptor (ER). 1 Endocrine therapy administered by the oral route is an effective form of treatment for this type of cancer. 2 Although newer targeted agents such as mTOR and CDK4/6 inhibitors, e.g., everolimus and palbociclib, are now gaining recognition in treatment, they are expensive and are administered in conjunction with endocrine therapy. 3,4 Currently, the firstline treatment for patients with hormone-dependent breast cancer (HDBC) includes either a selective estrogen-receptor modulator, such as tamoxifen, which blocks the action of estrogens at the ER or an ER downregulator (SERD), 5 or a "third-generation" aromatase inhibitor (AI) such as letrozole, anastrozole, and exemestane. This strategy leads to a reduction in the biosynthesis of estrogens and has been found to be superior to tamoxifen alone. 6 Also, in one study, anastrozole has shown significant preventative activity in high-risk postmenopausal women with undiagnosed breast cancer. 7 However, resistance will inevitably occur and blocking the action of estrogens at the ER and inhibiting the aromatase enzyme are not the only strategies available for endocrine therapy. There is now evidence that inhibition of steroid sulfatase (STS), 8 the enzyme that converts the biologically inactive estrone sulfate to estrone, as well as dehydroepiandrosterone sulfate to dehydroepiandrosterone, may render significant estrogen deprivation in patients treated with an STS inhibitor. This strategy works in an intracrine fashion because in the postmenopausal setting tumor cells can convert the large reservoir of circulating estrone sulfates, imported through organic anion transporters, to active estrogen in situ.
Moreover, STS inhibition also decreases levels of androstenediol, an estrogenic androgen, 9 levels of which are unaffected by AI inhbition. Chronic AI treatment leads to compensatory increases in both STS and 17β-HSD1 levels. 10 Moreover, the contribution of both STS and organic anion transporters to AI resistance was recently established and could be overcome by an STS inhibitor. 11 These ideas have led to STS inhibitors reaching phase II clinical trials for several indications in oncology including breast cancer and endometrial cancer and a phase I trial in prostate cancer. 12 The first STS inhibitor discovered with a remarkable potency was the steroidal sulfamate ester estrone-3-Osulfamate (EMATE, 1, Figure 1). 13 This agent is orally active and inhibits STS in an irreversible manner. However, EMATE was subsequently shown to be highly estrogenic in rats and this undesirable property effectively precluded its further development for use in the treatment of HDBC, although the estradiol variant (E2MATE, PGL2001, 1a) has nevertheless proceeded to clinical trials in the hormone-dependent nononcology setting of endometriosis. 12a,b,14 In an attempt to search for a nonestrogenic alternative to 1 with a comparable or even superior STS inhibitory profile, many structurally diverse inhibitors that contain the pharmacophore for irreversible inhibition of STS, i.e., an aryl sulfamate ester, have been developed, 8,12,15 leading to the clinical inhibitor Irosustat 2 ( Figure 1). 16 Initial work focused on designing A/B ring mimics of 1 such as derivatives of indanone, tetralone, and tetrahydronaphthol, 15 and this yielded a series of bicyclic coumarin sulfamates 17 (3−8, Figure 2) that were promising leads, showed a significant improvement over the first lead nonsteroidal candidate 5,6,7,8-tetrahydronaphthalene 7-O-sulfamate, and, more significantly, possessed in vivo activity. A main lead was the two-ring 4-methylcoumarin-7-O-sulfamate (3) that was orally active in vivo and, like EMATE, was a highly potent time-and concentrationdependent STS inhibitor, but importantly with no rodent estrogenic activity. 18,19 The related 3,4-dimethylcoumarin-7-Osulfamate (6) inhibited STS in MCF-7 cells with IC 50 = 30 nM, and a series of derived tricyclic compounds was subsequently synthesized that proved even more potent. 16,17 Further development of this series of nonsteroidal inhibitors led to the discovery of the tricyclic coumarin sulfamate (2) (Irosustat, STX64, 667COUMATE, BN83495, Figure 1) which has proven to be the most successful STS inhibitor to date. 12, 16 Recently, other examples of both mono- 20,21 and two-ring sulfamate-based STS inhibitors 22 have been published, but these compounds are generally still of relatively modest inhibitory activity. Irosustat was the first STS inhibitor to enter clinical trials for postmenopausal patients with advanced HDBC and has shown encouraging results. 12, 23−26 Irosustat has just completed CRUK sponsored phase II trials in both early breast cancer and in advanced breast cancer in combination with an AI, with positive indications of efficacy. 27,28 However, despite the significant progress made in developing irreversible inhibitors of STS and although 1a and 2 have reached clinical trials, their mechanism of action remains unresolved. The crystal structure of human STS has been solved, 29,30 and several hypotheses have been postulated to suggest how a sulfamate-based STS inhibitor might inhibit STS irreversibly. The currently favored hypothesis is a transfer of the sulfamoyl group (or as sulfonylamine) to a hydrated or unhydrated STS active site formylglycine residue, and this leads to inactivation of the active site machinery. 12b Although multiple mechanisms have been proposed for this, e.g. ref.
12e, see ref. 12b for a full up-to-date discussion.
Because of the unique role that a coumarin ring system plays in the design of potent STS inhibitors, we further expand here the bicyclic coumarin sulfamate series exemplified by 3−8 to give derivatives that bear various substituents at the 3-and/or 4-position(s), following on from preliminary encouraging data. 17 The inhibitory activities of most new candidates against STS were evaluated in MCF-7 cells and in placental microsomes. This study, in conjunction with other studies carried out on 2, 16,17 provides a more comprehensive structure−activity relationship (SAR) for coumarin sulfamates and has also produced highly active inhibitors of picomolar potency in vitro. The activity of two of the best bicyclic compounds is supported by molecular modeling and by comparing binding poses with those of the benchmark compounds EMATE and Irosustat.

■ CHEMISTRY
The compounds synthesized in this work fall into two different series: (i) those with an alkyl group of increasing carbon chain length or other functionalities at the C-4 position of the coumarin ring (A, Figure 3) and (ii) those with an alkyl group of increasing carbon chain length or other functionalities at the C-3 position and with a methyl group at the C-4 position of the coumarin ring (B, Figure 3).

Article
We employed β-keto esters as starting materials for synthesizing coumarins with a substituent at the 3-position, and this also leads to a 4-methyl substituent. Because 4methylcoumarin 7-O-sulfamate (COUMATE) is more active as an STS inhibitor than unsubstituted coumarin 7-Osulfamate, we retained this 4-substituted group, which imbues good activity. 18,19 Thus, to study further the SAR of COUMATE, we ideally needed to keep this 4-methyl group for comparison while we explored different substituents at the 3-position. Apart from coumarin 46a, the bicyclic ring of other parent coumarins was constructed by the Pechmann synthesis of hydroxyl coumarins. For our purposes, this route was preferred because the target structures can be prepared with relative ease by condensing resorcinol with an appropriate βketo ester. The only synthetic hurdle to overcome is the synthesis of the various β-keto esters required because most of them are not available commercially. The 7-hydroxycoumarins synthesized are subsequently sulfamoylated with freshly prepared sulfamoyl chloride to form the corresponding coumarin sulfamates.
The alkanoyl acetate esters required as starting material for the coumarins in the 4-alkyl series (A, Figure 3) were synthesized by treating the inexpensive ethyl potassium malonate with the corresponding acid chloride in the presence of magnesium chloride (MgCl 2 ), triethylamine (Et 3 N), and acetonitrile (CH 3 CN) as the solvent (Scheme 1). 31 This method has the advantage of being relatively safe, clean, economical, and suitable for scaling up with the product produced in high yield and purity, which is free from any unnecessary side products. Rathke and Cowan 32 have shown that the combination of anhydrous MgCl 2 and Et 3 N provides a system with enough basicity for metallating ethyl potassium malonate and that the reactions failed when MgCl 2 was replaced with other metal chlorides such as ZnCl 2 , CuCl 2 , FeCl 3 , TiCl 4 , LiCl, and AlCl 3 . 32 The number of equivalents of reagents used in the reaction determines the yield of the product obtained. For aromatic acid chlorides which have electron-withdrawing substituents such as fluoro, chloro, or nitro groups, 2.1 equiv of potassium ethyl malonate, 2.5 equiv of MgCl 2 , and 2.2 equiv of Et 3 N are optimal, and the yield is around 90%. On the other hand, aliphatic acid chlorides or aromatic acid chlorides containing electron-donating substituents generate side products, which are minimized by employing extra equivalent of Et 3 N to obtain the alkanoyl acetate in high yield.
β-Ketoesters were also prepared efficiently by reacting the corresponding aldehyde with ethyl diazoacetate in the presence of a catalytic amount of tin(II) chloride (SnCl 2 ) (Scheme 1). The mechanism of the reaction is likely to proceed via a betaine intermediate, followed by a preferential migration of the aldehyde hydrogen to the β-carbon i.e., a 1,2-hydride shift producing the required β-keto ester and N 2 as products. The two most noteworthy aspects of this method are its selectivity and the mild conditions involved. The reaction is insensitive to atmosphere, is complete between 1 and 2 h at room temperature, and can be catalyzed by various Lewis acids, such as BF 3 , ZnCl 2 , ZnBr 2 , AlCl 3 , SnCl 2 , GeCl 2 , and SnCl 4 , but the highest yield is obtained with SnCl 2 . 33 Although other common organic solvents can be used, CH 2 Cl 2 is often employed because it gives the best results and can be easily removed.
The α-alkylacetoacetates required for the synthesis of coumarins in the 3-substituted-4-methyl series were prepared conveniently by treating a solution of ethyl acetoacetate in CH 2 Cl 2 with the corresponding alkyl bromide in the presence of potassium carbonate (K 2 CO 3 ) and tetrabutylammonium chloride (Bu 4 NCl) (Scheme 2). 34 Pechmann synthesis of coumarins with a β-keto ester and resorcinol was carried out in the presence of an equimolar mixture of trifluoroacetic acid (CF 3 COOH) and concentrated sulfuric acid (H 2 SO 4 ) warmed from ice-water temperature to room temperature (Schemes 3 and 4). The use of a 1:1 mixture of conc. H 2 SO 4 and conc. CF 3 COOH as the condensing agent for the Pechmann synthesis of coumarins was first described by Hua et al., 35 and, in our hands, such a mixture has been found to be as effective as using conc. H 2 SO 4 alone, which is playing a role as a catalyst. The role of conc. CF 3 COOH in this reaction is not entirely clear, although it might be acting as an organic solvent and lowering the viscosity of the reaction mixture, rendering the stirring process more efficient.
The sulfamoylation reaction was performed by reacting the hydroxyl coumarins with an excess of (∼5 equiv) sulfamoyl Scheme 1. Synthesis of Various Ethyl Alkanoyl Acetates for the Preparation of 4-Alkylcoumarin Sulfamates a a (i) SOCl 2 /tetrahydrofuran (THF), reflux; (ii) (a) MeCN, MgCl 2 , Et 3 N, 10−25°C, 2.5 h, (b) RCOCl, Et 3 N, 0°C, 0.5 h, room temperature (rt), 12 h; (iii) RCHO, SnCl 2 , CH 2 Cl 2 , rt, 3 h. chloride after treating with 1 equiv of NaH as described previously by Woo et al. 36 ■ BIOLOGICAL RESULTS AND DISCUSSION The in vitro inhibition of STS activity by most of the sulfamates synthesized in this work was measured in two assay systems: (i) a preparation of an intact monolayer of MCF-7 cells, which assesses the ability of compounds to cross the cell membrane and inhibit STS under conditions that closely resemble the tissue/physiological situation and (ii) a placental microsomes preparation where a higher concentration of substrate is employed, with which a compound has to compete for binding to the enzyme active site. For the placental microsome STS assay, a saturating substrate concentration of 20 μM was used and inhibitors were tested under initial rate conditions. The MCF-7 STS assay is meant to mimic/reflect  (Tables  1−4). Although time-and concentration-dependence studies have not been carried out to confirm the nature of inhibition for those compounds tested, it is anticipated that they act mechanistically in a similar manner to other aryl sulfamates like 1 and 2, which have been shown to be active-site-directed inhibitors 4-Substituted Compounds. MCF-7. For 4-n-alkyl derivatives 9−15, 17, and 18, all derivatives inhibit STS activity >90% at 1 μM. Based on the % inhibition observed at 0.1 and 0.01 μM, the inhibitory activity of the compounds increases slightly as the chain length of the alkyl group increases, and it peaks at the nonyl (14, 90% at 0.01 μM) and decyl (15, 86% at 0.01 μM) derivatives. This may be attributed to the increase in lipophilicity of the compounds as their alkyl group becomes longer until steric hindrance potentially becomes a limiting factor. A similar observation was reported in studies of (p-Osulfamoyl)-N-alkanoyl tyramines. 37,38 We and others have also noted the existence of a hydrophobic pocket at the end of the steroid binding pocket 39 that might also be accessed by the present compound series to improve binding, although seeking the right compromise between hydrophobicity and steric hindrance is important. 40 For other substituents at the 4position of the coumarin ring, their inhibitory activities vary with the phenethyl derivative 24 (39% at 0.01 μM, IC 50 = 18 nM) and the cyclohexyl derivative 26 (37% at 0.01 μM, IC 50 = 24 nM) being the most active. However, both 24 and 26 are less potent as STS inhibitors than the 4-n-alkyl derivatives. One possibility is that the active site of STS, like many other enzymes with steroids as substrate in general, has limited accommodation for substituents at the C1/C11/C12 edge of the steroid scaffold. Hence, the more flexible aliphatic alkyl chains may be better tolerated by the enzyme active site than the bulkier and more rigid substituents such as phenyl, benzyl, phenethyl, 4-ethylphenyl, and cyclohexyl when these substituents are placed at the 4-position of the coumarin ring system, which mimics the A/B ring of the steroidal STS inhibitor 1.
Placental Microsomes. Apart from the two lower members 3 and 4 and the two higher members 17 and 18, other 4-nalkyl derivatives (5−16) tested show >90% inhibition at 1 μM. However, the SAR between chain length and inhibitory activity is not clear because all derivatives evaluated show the same order of magnitude in regard to % inhibition at 0.01 μM and IC 50 values. Nonetheless, the n-pentyl (10, IC 50 = 40 nM) and n-dodecyl (17, IC 50 = 45 nM) derivatives appear to be the most active inhibitors in this group. For those derivatives that have bulkier substituents at the 4-position, they are significantly less potent than their n-alkyl derivatives with the exception of 23 (benzyl, IC 50 = 64 nM), 24 (phenethyl, IC 50 = 82 nM), and 26 (cyclohexyl, IC 50 = 42 nM), the IC 50 values of which are of the same order of magnitude as those of n-alkyl derivatives. As expected for a cell-based assay, the IC 50 values against STS obtained for compounds in Table 1 are much lower than those obtained from the cell-free placental microsomes assay (Tables 2 and 3). A similar phenomenon was observed in previous work. 41 3-Substituted-4-methyl Compounds. MCF-7. A relatively smaller number of synthesized compounds in this series compared to their 4-substituted relatives were tested for their inhibitory activities. From the results available, 3-alkylated-4methyl compounds 28−33 show >97% inhibition at 0.1 μM, whereas compounds 39−43, which have other substituents at the 3-position, inhibit STS between 29 and 98% at 0.1 μM. Of those five compounds that have IC 50 values determined, 29 is the most potent (0.68 nM), closely followed by 41 and 42 (1 and 1.1 nM, respectively). On comparing 23 (4-benzyl, IC 50 = 75 nM, Table 1) and 24 (4-phenethyl, IC 50 = 18 nM, Table 1) with 41 (3-benzyl-4-methyl, IC 50 = 1 nM) and 42 (3phenethyl-4-methyl, IC 50 = 1.1 nM), there is 1 order of magnitude difference between the potency of the two pairs of compounds. This finding suggests that placing either a benzyl or a phenethyl group at the 3-position of the coumarin ring produces a more potent STS inhibitor. It is anticipated that on binding of 41 and 43 into the active site of STS, the coumarin ring of which is designed to mimic the A/B ring of 1, their substituents at the 3-position extend into the same area where the C/D ring of 1 resides.
Placental Microsomes. On the basis of the IC 50 values available, it appears that compounds with shorter alkyl chains at the 3-position of the coumarin ring (28, n-pentyl, IC 50 = 12 nM and 29, n-hexyl, IC 50 = 32 nM) are more potent STS inhibitors than those with longer alkyl chains (32)(33)(34)37;IC 50 > 300 nM). This contrasts with those compounds in the 4alkylated series ( which are tighter and fall within the same order of magnitude. This finding suggests that a long alkyl chain placed at the 4position of the coumarin ring may interact better with the enzyme active site than its counterpart placed at the 3-position of the coumarin ring. To this effect, the sulfamate group of 4alkylated compounds may be better positioned within the catalytic site of the enzyme for inactivation. For compounds 40−43, the inhibitory activity observed at 0.1 μM starts from 65% for 40 (phenyl) and rises to 94% for 41 (benzyl) before it falls to 91 and 47% for 42 (phenethyl) and 43 (phenylpropyl), respectively. A similar pattern is observed when the IC 50 values of 40 (54 nM), 41 (8 nM), and 42 (33 nM) are compared. In regard to potency, the benzyl group is therefore the optimal substituent for this group of 3substituted-4-methyl coumarin sulfamates. It is possible that the phenyl, phenethyl, and phenylpropyl groups interact less favorably with or are less well accommodated by the enzyme active site (Table 4).
When the benzyl group of 41 is replaced by a cyclohexylmethyl group to give 44, a reduction in potency is observed (at 0.1 μM, 94% for 41 vs 74% for 44). The same pattern, but to a greater extent, is observed when the phenethyl group is replaced by a cyclohexylethyl group as shown by the 91% inhibition of the STS observed for 42 at 0.1 μM compared to the 37% inhibition for 45 at the same concentration. This finding suggests that the more rigid and electron-rich phenyl group may interact better with the enzyme active site (such as through π-interactions with neighboring amino acids) than the more flexible aliphatic cyclohexyl group. The best inhibitors are illustrated in Figure 4, with an attempt to illustrate the mimicry of the steroidal C and D rings and also with some comparative activities shown in Table 5.

■ MOLECULAR MODELING
Docking studies were conducted to explore potential interactions between the substituted bicyclic coumarin derivatives and the STS active site, in a similar fashion to those carried out for STX64/Irosustat and related series members. 16 They show that the two most active compounds 29 and 41 are placed in a very similar fashion to the irreversible STS inhibitor Irosustat, with the sulfamoyl group in close proximity and opposite to the catalytic FGly 75 ( Figure 5), suggesting that a putative sulfamoyl group transfer could also readily occur that might lead to similar irreversible inhibition (although note that no experiments were conducted to explore the reversibility/irreversibility of 29 and 41 against STS). Residue V486 on one side and residues L103 and V177 on the other sandwich the bicyclic ring system. Both compounds possibly form a hydrogen bond (N···O = ∼3.2 Å) from their chromen-2-one oxygen to the NH of G100 in the same manner as Irosustat ( Figure 5). These more potent compounds have   fairly small hydrophobic pendant groups attached to the 3position of the chromen-2-one ring. These hydrophobic moieties lie between the hydrophobic sidechains of L103, F178, and F488. Those compounds with larger pendant groups may be less active due to the hydrophobic nature of the group making the compound less soluble. Alternatively, because STS is a membrane-bound protein and any substrate or inhibitor has to pass through the membrane to access the active site, it may be that larger hydrophobic tails result in the inhibitor failing to fully transit through the membrane: the hydrophobic tail stays, preferentially, embedded in the membrane. On examining the data in Table 5, where the best compounds are benchmarked against the steroidal EMATE and the nonsteroidal Irosustat and, more particularly, against the known two-ring coumarin sulfamate COUMATE, it is readily apparent that highly potent compounds have been designed through the targeted 3-and 4-substitutions undertaken in this work. Some of these (29, 32, 41, and 42) have a potency approaching the clinical drug Irosustat in the more definitive intact MCF-7 cell assay, and of these, 29 is highly significant with a similar picomolar IC 50 . Compound 41 is perhaps of the widest interest with an IC 50 of 1 nM but also with an inhibitory activity better than that of Irosustat in the more challenging placental microsomal STS assay. It has an attractive 3-benzyl substituent that, as for the highly active homolog 42, could potentially be substituted to further refine activity. Moreover, 41 and 42 are structurally distinct from the fully saturated C-ring surrogate of Irosustat and, as more versatile compounds, could form the basis of an attractive series for further optimization and eventual preclinical development. In any case, if we sensibly take COUMATE 3 for comparison, the best compounds are gratifyingly some 100−500 times more potent in the MCF-7 assay.

■ CONCLUSIONS
Synthetic routes to two-ring coumarin 7-O-sulfamate derivatives possessing 3-and 4-modified substitutions were devised, generally using an α-alkylacetoacetate strategy and the Pechman hydroxycoumain synthesis. 15,17 Compounds were shown to inhibit, often highly potently, the emerging clinical drug target STS 8 now validated for hormone-dependent diseases 12 using an intact MCF-7 cell assay and an assay against placental microsomal STS activity. The best compounds were benchmarked for activity against the steroidal sulfamate drug EMATE, 13 the nonsteroidal Irosustat, 12,16 and the known two-ring parent coumarin sulfamate COU-MATE. 18,19 Through the targeted 3-and 4-substitution strategy undertaken, highly potent compounds were designed. In intact MCF-7 cells, compounds 29, 32, 41, and 42 had a potency approaching Irosustat with 29 having an IC 50 of 680 pM. 41 had a similar IC 50 of 1 nM but was also better than Irosustat against placental microsomal STS. With COUMATE 3 taken as the most relevant comparative structural benchmark for non-tricyclic derivatives, the best compounds were ca. 100−500 times more potent in the MCF-7 assay. Both 41 and 42 possess motifs structurally distinct from the fully saturated cyclic C-ring of Irosustat with attractive pendant 3-benzyl and 3-phenethyl substituents, respectively, that could potentially be further optimized through aromatic substitution. Compounds 29 and 41 were modeled into STS in comparison to benchmarks and dock well into the active site, placing the aryl sulfamate moiety opposite the catalytic FGly, as for Irosustat 16 and with their pendant side chains occupying a hydrophobic pocket noted previously. 39 The expectation is that, in a similar fashion to Irosustat and EMATE, such compounds will act as irreversible inhibitors by transfer of their sulfamoyl group to the STS enzyme, 12b,13b although this has not been formally explored here. Thus, the versatile 3-benzyl-4methyl-and 3-phenethyl-4-methyl-derivatives 41 and 42, respectively, and possibly also the 3-n-hexyl-4-methyl-derivative 29 from this study are potent STS inhibitors and could represent new leads for potential preclinical development.

■ EXPERIMENTAL SECTION
In Vitro Steroid Sulfatase Assay. STS inhibitory assays were performed essentially as previously described. 13b, 43 The ability of the compounds synthesized to inhibit E1S was tested in vitro using MCF-7 cells and a placental microsomal preparation from a sulfatase-positive human placenta from a normal term pregnancy and compared with that of EMATE. For the placental microsome STS assay, a saturating substrate concentration of 20 μM was used and inhibitors were tested under initial rate conditions. For the MCF-7 STS assay, a physiological concentration of 3 nM E1S was used. Thus, for the placental microsome assay: Ph 99  Chemicals and Analyses. All reagents were purchased commercially either from Aldrich Chemicals Co. (Gillingham, Dorset, U.K.) or Lancaster synthesis (Morecambe, Lancashire, U.K.). All organic solvents used were of general purpose or analytical grade and were obtained from Fisons Plc. (Loughborough, U.K.) and stored over 4 Å molecular sieves. Anhydrous dimethylformamide (DMF) used for all sulfamoylation reactions was purchased from Aldrich and was stored under a positive pressure of N 2 after use. Sulfamoyl chloride was prepared by adapting a method originally reported by Appel and Berger 44 and was stored as a standard solution in purified sulfur-free dry toluene. 36 Thin-layer chromatography (TLC) was carried out using precoated plates (Merck TLC aluminum sheets silica gel 60 F254, art. no. 5554). Product(s) and starting material were detected by treating plates with a methanolic solution of phosphomolybdic acid followed by heating or simply by viewing directly under UV light. Flash column chromatography was carried out by gradient elution (solvents used are indicated in the text) on wet-packed silica gel (Sorbsil C60). IR spectra were recorded using a PerkinElmer 782 spectrophotometer with peak positions expressed in cm −1 . 1 H and 13 C NMR spectra were recorded using either a Jeol Delta 270 MHz or Varian Mercury VX 400 MHz spectrometer. Chemical shifts (δ) are reported in parts per million (ppm) using an internal standard of tetramethylsilane. Coupling constants (J) are quoted to the nearest 0.1 Hz. Mass spectra were acquired at the Mass Spectrometry Service Centre, Bath and FAB mass spectra used m-nitrobenzyl alcohol as matrix. Elemental analyses were carried out by the Microanalysis Service, Bath. Melting points are uncorrected and were determined using a Reichert-Jung Thermo Galen Kofler block. High-performance  (10) (42), in comparison with EMATE (1), Irosustat (2), and COUMATE (3). Solid lines denote similarity to steroid C and D rings.

ACS Omega
Article liquid chromatography (HPLC) was performed using a Waters 660E instrument equipped with an autosampler and photo diode array detector. A Waters Radialpak column (RP18, 8 mm × 100 mm) was used. The conditions of elution and analytical data are as indicated for each compound analyzed.
Molecular Modeling. Schrodinger software (running under Maestro 9.0) was used to build and minimize all of the ligands. The ALS75 residue in PDB crystal structure 1P49 (human placental estrone/dehydroepiandrosterone sulfatase) was mutated to the gem-diol form using the Schrodinger software editing tools. Minimization of the resulting structure, with the position of the backbone atoms fixed, allowed the atoms of the gem-diol and surrounding side chains to adopt low-energy conformations. Ligands were docked into the rigid protein using GOLD. A 10 Å sphere centered on the ALS75 sulfate was defined as the binding site. The GOLDScore fitness function was used to score the docked poses (25 for each ligand).
General Methods for the Synthesis of Ethyl 3-Oxoalkanoates for the Preparation of 4-Alkylcoumarin Sulfamates. Method A. 31 To ethyl potassium malonate (2.1 equiv) in MeCN (100 mL/5 g of acid chloride) at 10−15°C and under N 2 was added Et 3 N (3.2 equiv), followed by MgCl 2 (2.5 equiv). The mixture was stirred at 20−25°C for 2.5 h and then at 0°C for 0.5 h before the corresponding acid chloride (1 equiv) was added dropwise during 25 min. The mixture was further treated with Et 3 N (5 mL) and stirred overnight at 20°C . The evaporation residue was dissolved in toluene and reconcentrated. More toluene was added, stirred, and cooled to 10−15°C before aq HCl (1 M, 50 mL) was added cautiously while keeping the temperature <25°C. The organic layer was washed with 1 M aq HCl (50 mL) and water. Drying, evaporation, and distillation or chromatography (CHCl 3 or CHCl 3 /acetone, 10:1) gave the corresponding ethyl αalkanoylacetate.
Method B. 33 To anhydrous SnCl 2 (0.1 equiv) was added CH 2 Cl 2 (∼100 mL/5 g of aldehyde), followed by ethyl diazoacetate (1.05 equiv). The reaction was initiated by adding a few drops of the corresponding aldehyde in CH 2 Cl 2 . When N 2 evolution began, the remaining solution of aldehyde (1 equiv) was added dropwise over 30 min. After the evolution of N 2 had stopped (∼1−3 h), the mixture was washed with brine (50 mL) and extracted twice (Et 2 O). Drying, evaporation, and chromatography (CHCl 3 or CHCl 3 /acetone, 10:1) or distillation gave the corresponding ethyl alkanoylacetate.
General Method for the Synthesis of 3-or 4-Alkyl-7hydroxycoumarins. 17 Resorcinol (1 equiv) was dissolved in the corresponding hot β-keto ester (1 equiv). The resulting syrup was cooled to 0°C and treated dropwise with a mixture of CF 3 COOH (2 equiv) and conc. H 2 SO 4 (2 equiv) while keeping the temperature <10°C. After stirring for 3 h at room temperature, the mixture was cautiously quenched with icewater. The brightly colored gluey mass formed was stirred for further 1 h. The bright yellow/brown precipitate resulted was

ACS Omega
Article