Articulo exam1

7/31/2019 Articulo exam1

1/13

Estradiol-16r-c a r b o x y l i c A c i d E s t e r s a s L o c a l l y A c t iv e E s t r o g e n s

David C. Labaree, Toni Y. Reynolds, and Richard B. Hochberg* ,

Dep ar tm en t of Ob st etr ics an d Gy n ecolog y, an d Com pr eh en si ve Ca n cer Cen ter , Y al e U n iv ers it y S ch ool of M ed ici n e,

N ew H av en , Con n ecti cu t 06 52 0

R eceiv ed Decem ber 7, 20 00

We attempted to design analogues of estradiol to act as locally active estrogens withoutsignificant system ic action. We synth esized a ser ies of 16R-car boxylic acid subst itu ted st eroidsan d th eir esters a nd t ested th eir action in severa l assays of estrogenic action, including estr ogenreceptor (ER) binding, estrogenic potency in Ishikawa cells (hu man endometrial car cinoma),ra t ut erine weight (systemic action), an d m ouse vaginal r eductases (local action). All of theestradiol substituted carboxylic acids (formic, acetic and propionic acids) were devoid ofestrogenic action. To the contrary, many of the esters had marked estrogenic potency in thereceptor an d the Ish ikawa a ssays. The esters of the 16R-formic acid series had the highest ERaffinity with litt le differen ce between t he st ra ight-cha in alcohol esters (from met hyl to n-butyl).However, estr ogenic action in the Ishikawa assa y decreased precipitously with esters longerth an t he eth yl ester. This decrease corr elated well with th e increas ed rat e of estera se hydrolysisof longer est ers as deter mined in incubat ions with r at h epat ic microsomes. The most promisingcandidates, th e methyl, ethyl, a nd fluoroethyl esters of the format e series, were t ested for

system ic an d local action in th e in vivo models. All thr ee, especially th e fluoroeth yl ester, sh oweddivergence between systemic and local estrogenic action. These metabolically labile estrogenswill be extr emely useful for th e ther apeu tic tr eatm ent of th e vaginal dyspareu nia of men opausein women for whom systemic estrogens ar e contr aindicated.

I n t r o d u c t i o n

It is well-recognized tha t phar macologic estrogenadministration (hormone replacement therapy, H RT)can alleviate most, if not al l , of the symptomologyassociated with menopause. These symptoms includebut ar e n ot limited t o bone loss a ssociated with osteo-porosis, heart disease associated with changes in bloodlipids and lipoproteins, hot flashes, and vaginal dys-pareunia.1 However, there are risks associated with

estrogen administra tion in HRT a s well as oral contr a-cept ive use, and they include an associat ion withendometrial cancer, breast cancer, and stroke. Thus,although there are many therapeutic benefits of HRT,there are significant risks.2-4

Estrogen therapy affects a number of organs bothdirectly and indirectly, and some of these outcomes aredeleterious. Consequently, where possible, symptomol-ogy that could be ameliorated by local rather thansystemic administration could limit the adverse sideeffects of estrogen t her apy. One such syndr ome tha t canbe treated directly, caused by estrogen deprivation orestr ogen an ta gonists, is vagina l dyspar eun ia, a commondisorder which affects a large proportion of women,

approximately 40% within 10 years of the onset ofmenopause.5 Dyspareunia is associated with a severephysical and psychological impact, for it is not onlypainful but it can dramatically influence a womens self-image, leading to clinical depression.6-8 While topicalapplication of estr ogens to th e vaginal mucosa ha s beenused to treat the vaginal dyspareunia of menopause,these estr ogens a re a dsorbed into the blood an d resu lt

in significant blood levels of estrogens.9-12 Thus, thisther apy may not be used where systemic estrogens arecontraindicated. Another possible use for local estrogensincludes the topical administration to aging skin. Theskin conta ins ER and it is an estr ogen target organ th atcan respond to estrogen replacement.13

Since t opically applied estrogen is adsorbed int o theblood, its pu rpose is defeated . A potent estr ogen wh oserange is limited to the tissue to which it is applied would

be ideal for the treatment of these disorders. Similartherapeutic agents with locally l imited actions havebeen termed soft drugs,14 compounds that have alimited region of activity due to r apid meta bolic inac-tivation. Ester groups have been used to convey softdrugproperties to biologically active molecules becausehydrolytic enzymes, including esterases, are ubiqui-tously distributed.15 In these drugs the esters ar e theactive a gents wh ile th e hydrolysis pr oducts, th e corr e-sponding carboxylic acids, are inactive. For example,locally active glucocorticoids have been developed asan tiinflam mat ory agent s for the skin. These are carboxyanalogues of steroids that are esterified. The parentcarboxylic acids do not bind to the glucocorticoid recep-

tor an d ar e biologically iner t, while their corr espondingesters bind to the glucocorticoid receptor with highaffinity.15-17 The esters are rapidly hydrolyzed to t hehorm onally ina ctive par ent s ter oidal carboxylic acid byubiquitous est era ses. Consequently, th eir effect is local-ized to the area of the skin to which they are appliedbecause their rapid inactivation prevents systemic ac-tion.18

Similarly, in a study designed to produce affinitychromatographic supports for the purification of t heestr ogen r eceptor (ER) it was found t ha t car boxylic acid

* Corresponding author. Phone: (203)785-4001. Fax: (203)737-4391.E-mail: r ichard.h ochber [email protected].

Department of Obstetrics and Gynecology.

1802 J . M e d . Ch e m . 2001, 44 , 1802-1814

10.1021/jm000523h CCC: $20.00 2001 American Chemical SocietyPublished on Web 04/27/2001


2/13

an alogues of estr adiol (E 2) at C-7R- and C-17R are verypoor ligands if they bind at all, but the methyl esters ofthese s ame ana logues ha ve much impr oved affinity forthe ER.19 It appears from those results th at a chargedcarboxylic acid gr oup in proximity t o th e st eroid ringinterferes with binding to the ER an d tha t mask ing thecharge by esterificat ion r everses th is interferen ce. Con-sequently, we thought i t l ikely that a locally activeestrogen could be designed using a concept similar tothat used for the glucocorticoids. We synthesized afam ily of 16R-carboxylic acid analogues of E 2 of var yingalkyl chain length s tha t were esterified with a num berof alcohols of different size and substituents (Figure 1)and performed a structure-activity st udy in order todetermine the feasibility of producing a locally activeestrogen. The 16R-position of E 2 was chosen for substi-tut ion becaus e it is chemically accessible a nd becau sesome substitutions there are known to be tolerated byt he ER.20 To produce a potent but locally confinedestr ogen, specific alcohols of a r an ge of ster ic bulk, chainlength, and leaving group ability were chosen to con-stru ct ester s of the car boxylic acid conta ining ster oidsin order t o balan ce ER affinity an d estr ogen a ction withrat e of enzymat ic hydrolysis. The car boxylic acids an dth eir ester s were tested for th eir affinity for the ER an dtheir ability to activate an estrogen-inducible gene intissue culture. The results were correlated with therelat ive rate at which the esters are hydrolyzed byhepat ic estera se(s). Finally, candidate est ers were t estedin in vivo bioassa ys for t heir system ic and local a ction.

C h e m i s t r y

The synthesis of the 16R-formyl ester analogues ofestradiol 9-18 i s shown in Scheme 1 and employsmethodology used previously by the Katzenellenbogengroup to prepar e 16R-hydroxymethyl-substituted estra-diol der ivatives.20 Deprotona tion of 3-benzylestrone withNaH in THF followed by a cylation with et hyl forma tegave the enol ether 1, which was protected with E tI a s

the et hyl enol ether 2. Stereoselective reduction of the17-ketone with LiAlH4 gave the 17-hydroxyl compoun d3 that was protected with Ac2O in pyridine to give theacetate 4. Acid hydrolysis of 4 with 10% aqueous HClproduced the 16R-aldehyde 5 as t he only isomer . Oxida-tion with CrO3-H 2SO 421 gave the acid 6 , which wasdeprotected with KOH-MeOH followed by hydrogenol-ysis with 5% Pd-C/H 2 to produce the acid 8, E 16-1,0.

The m ethyl, ethyl, propyl, and butyl ester s [E16-1,1(9), E16-1,2 (10 ), E16-1,3 (11), and E16-1,4 (13 )] wereprepared by reacting 8 with th e appr opriate alcohol inthe presence of SOCl2; the isopropyl, neopent yl, mono-fluoro-, difluoro- and t rifluoroeth yl ester s [E16-1,3i (12 ),E16-1,5neo (14 ), E16-1,2F 1 (16 ), E16-1,2F 2 (17 ), a nd

E16-1,2F 3 (18 )] were prepared by reacting 8 with theappr opriate a lcohol in th e presen ce of pTsOH. The vinylester, E16-1,2vin (15 ), was prepar ed from 8 through avinyl-exchan ge reaction with vinyl propionate andPdCl2-LiCl as catalyst.22 In the 1H NMR spectra of allthese esters, the signal for H-17R appears at about 3.90 ppm with a coupling const an t (J17R-16) of 7.4-9 Hz,indicating tha t th ese esters have the sam e stereochem-istry at C-16. The magnitu de of this coupling consta nt

is consistent with that seen in related C-16R,17-substituted steroids.23 Proof of the stereochemistry atC-16 was obtained by reduction of E16-1,2 with LiAlH 4,a reagent known not to affect epimerizable asymmetriccenters.24,25 The reduction gives the hydroxymethylsteroid whose 1H NMR spectru m is identical with tha tof th e known 16R-hydroxymeth yl estrad iol 19 , preparedby the l iteratu re procedure.20

The synthesis of the 16R-car boxymeth yl ana logues ofestra diol (22, 23 , 26 ) is shown in Scheme 2. Deproton-ation of 3-benzylestrone with LDA in THF at 0 Cfollowed by alkylation with ethyl bromoacetate at


3/13

decreases the binding. Thus, in the series of methylesters, E16-1,1 ha s a RBA ) 35%; E16-2,1, 5%; an d E 16-3,1, 1%, where RBA is the relative binding a ffinity in

the rat uterine estrogen receptor. Likewise, in the ethylester s eries, E16-1,2 ha s a RBA) 40% and E16-2,2, 5%.This pr ecipitous decline with increasing number of

S c h e m e 1a

a (a) NaH, THF ; ethyl forma te (estrone-3-benzyl ether f 1); (b) K2CO 3, EtI, acetone (1 f 2); (c) LiAlH 4, E t 2O (2 f 3); (d) Ac2O, pyridine(3 f 4); (e) 10% aqueous HCl, THF (4 f 5); (f) Jones oxidat ion (5 f 6); (g) KOH-MeOH, 50 oC (6 f 7); (h) 5% P d-C/H 2, EtOH (7 f 8);(i) ROH, SOCl2 or ROH, pTsOH or vinyl propionate, P dCl2-LiCl, MeOH (8 f 9-18 ).

S c h e m e 2a

a (a) LDA, THF; ethyl bromoacetate (estrone-3-benzyl ether f 20 ); (b) Li(OtBu)3AlH, THF (20 f 21 ); (c) 5% Pd-C/H 2, EtOH (21 f 22 );(d) 5% KOH-MeOH (22 f 23 ); (e) 5% KOH-MeOH (21 f 24 ); (f) MeOH, SOCl2 (24 f 25 ); (g) 5% P d-C/H 2 (25 f 26 ).

1804 J ou rn al of M ed icin al Ch em istry, 2001, V ol. 44, N o. 11 L abaree et al.


4/13

carbon atoms did not occur with the alcohol portion of

the ester. Her e th e length of the a lcohol portion of theester had no effect on ER binding: E16-1,1, E16-1,2,and E16-1,3 all had RBAs of approximately 35-40%,an d E16-1,4 at 28% was n ot significan tly differen t fromthe others. Consequently, the pairs E16-1,2 and E16-2,1, and E16-1,3 and E16-3,1 are all esters with thesame number of carbon and oxygen atoms, but theirbinding to the ER is very different . Noteworth y, all ofthese E16-1 esters were at least as potent as th e natu ralestrogen, estr one (E1) (RBA) 30%). However, th e E16-1esters ma de from bulky alcohols had a markedly de-creased affinity for the receptor: the RBA of theneopentyl ester E16-1,5neo was 7%, and that of theisopropyl ester E16-1,3I was 7%. The fluoroeth yl ester,

E16-1,2F 1, was a good l igand, with an RBA of 35%,comparable to that of E16-1,2. Additional fluorine atomsdecreased the affinity for the ER: E16-1,2F2, RBA )

10%; E16-1,2F 3, RBA)

7%. The vinyl ester, E16-1,2vin,also had a RBA of 6%.A few of the E16-1 esters that had high affinity for

ERR were also tested (by Drs. Paul Shughrue andHeather Harris) for binding to ER. In these experi-men ts bindin g to th e ligan d binding domain (LBD) (seeExperimental Section) of both the human ERR and ERwere measu red in par allel. The binding of the carboxyanalogues to the expressed human ERR (Table 2) issomewha t lower tha n th e binding to ERR of ra t uter inecytosol (Table 1) but shows the s ame tren d: E16-1,1 )E16-1,2. = E16-1,2F1 > E16-1,2F 2 (the RBAs here arenot statistically different). More importantly, none ofthese ester s bind well to ER, on the order of 1-2% of

ERR. The ERR to ER ratio range from 60 for E16-1,1to 90 for E 16-1,2F1. This differen ce in th e affinity of theE 2-16R-alkyl esters for the 2 ERs is not unexpected,because i t ha d pr eviously been reported t hat another16R-substi tuted ER ligand, 16R-iodoestradiol, whichbinds with high affinity to the classical ER, ERR,27 bindsonly poorly to ER.28 Apparently, many substituents atC-16R th at do not affect binding to ERR impede bindingby ER.

Al l of the E 2-16R-alkyl carboxylic acids and theiresters were tested for their est rogenic potency bymeasuring their effect on the induction of alkalinephosphatase (AlkP) in Ishikawa cells (a human en-dometrial adenocarcinoma cell line). We have previously

S c h e m e 3a

a (a) AcO2, pyridine (16R-allyl-3-benzylestradiol f 29); (b) BH 3-THF; Et 3NO, diglyme, 150 oC (29 f 30 ); (c) Jones oxidat ion (30 f 31 );(d) KOH-MeOH, 55 oC (31 f 32 ); (e) MeOH, SOCl2 (32 f 33 ); (f) 5% Pd-C/H 2, EtOH (33 f 34 ); (g) KOH-MeOH, 60 oC (34 f 35 ).

T a b l e 1 . Est rogenic Properties of E2-16R-alkylesters

compdaER

(RBAb)Ishikawa cellAlkP (RSAc)

esterase(RHAd)

E 2 100 100 -E 1 30 ( 11 7 ( 2 -E16-1,0 0 0 -E16-1,1 35 ( 4 10 ( 3 45 ( 10E16-1,2 40 ( 10 11 ( 5 100E16-1,3 34 ( 9 2 ( 0 230 ( 50E16-1,3i 7 ( 4 0.1 ( 0.06 140 ( 10E16-1,4 28 ( 17 1 ( 0 350 ( 50E 16-1,5n eo 14 ( 3 1 ( 0.6 50 ( 5E16-1,2F1 35 ( 3 13 ( 5 420 ( 40E16-1,2F2 10 ( 3 4 ( 1 2350 ( 275E16-1,2F3 7 ( 1 3 (1 8200 ( 1800E 16-1,2vin 6 ( 2 0.4 ( 0.3 20300 ( 5500E16-2,0 0 0 -E16-2,1 5 ( 2 0.5 ( 0.1 340 ( 30E16-2,2 5 ( 1 0.3 ( 0.1 700 ( 120

E16-3,0 0.1 ( 0.1 0 -E16-3,1 1 ( 1 0 910 ( 100

a Abbreviat ions are shown in Figure 1 , with examples asfollows: E16-1,0 is the formic acid analogue of E 2. E16-1,1 is themethyl ester of the formate analogue, etc. 3i is the isopropyl ester,5neo is the neopentyl ester , and 2vin is th e vinyl ester. F is fluorinesubstitution in the 2-position of an ethyl ester, i.e., F 3 is the 2,2,2-trifluoroethyl ester. b RBA is the relative binding affinity in therat uterine estrogen receptor (ER) assay, where E 2 ) 100. c RS Ais the relative stimulatory activity in the induction of alkalinephosphatase (AlkP) activity in the Ishikawa estrogen bioassay,where E 2 ) 100. d RHA is the relative hydrolytic activity in theesterase a ssay with ra t hepa tic microsomes in compar ison to E16-1,2 ) 100. The dash (-) indicat es not done. All values are (SD .

T a b l e 2 . Binding of Selected E 2-16R-Alkylesters t o the LBD ofHuman ERR and ER

compd ERRa E Ra ERR/E R

E 2 100 100 1E16-1,1 19 ( 9 0.3 ( 0.2 62 ( 10E16-1,2 19 ( 5 0.3 ( 0.1 77 ( 12E16-1,2F1 16 ( 2 0.1 ( 0.1 89 ( 3E16-1,F2 13 ( 9 0.2 ( 0.1 67 ( 4

a RBA of the indicated est er compared t o E2. Values are (SD .This assay measures the inhibition of the binding of [3H]E 2 inlysates ofE. coli in which the LDB of human ERR and ER wereseparately expressed. Abbreviations are in Table 1, LBD is theligand binding domain.

E st ra d iol E st ers as L ocal ly A cti ve E st rog en s J ou rn al of M ed ici n al Ch em is tr y, 20 01 , V ol. 44 , N o. 11 1805


5/13

shown that this assay accurately assesses the potencyof a wide variety of estrogenic compounds.29 As can beseen in Table 1, the potencies determined in thisexperiment are different than those measured in the ERassa y. Three of the esters, E 16-1,1, E16-1,2, and E16-1,2F 1, ha d fairly high est rogenic potencies (compar edto E 2) with RS As of 10, 11, an d 13%, respectively. Theywere at least as effective as E 1 (7%) and p robably moreso, although the differences were not significant. Con-

tra riwise, several of the est ers th at h ad r elatively highRBAs in the ER binding a ssay, including E 16-1,3 andE16-1,4, ha d a mu ch lower potency in the Ish ikawa cellassay. Most of the other esters also had considerablylower estr ogenic action tha n would have been pr edictedon the basis of their r eceptor affinity.

The reason for the discrepancy between ER bindingand estrogenic potency became clear when t he E 2-16R-alkyl esters were tested a s substr ates for t he estera se(s)in ra t hepa tic microsomes. It can be seen in Table 1 tha tther e are dra mat ic differen ces in the ra te of the ester asereaction with t he various esters. In general, the longerthe alkyl chain, regardless of whether i t i s in thecarboxylic acid or a lcohol port ion of th e est er function,

the more rapid the hydrolyt ic cleavage. This i s asexpected, for i t has been shown that increasing thelipophilicity of the alcohol portion of an ester leads toan increased r at e of enzymat ic cleavage,30 an d it is cleartha t the furth er removed tha t the ester function is fromthe bulky steroid nucleus, the more accessible it is tothe enzyme. Steric hindrance is also a factor in thehydr olysis of the br an ched chain a lcohols, E16-1,3i an dE16-1,5neo, in which the relative rate of reaction ismarkedly decreased, with RHA ) 140 and 50, respec-tively, compar ed t o E16-1,3, RHA ) 230, where RH A isthe relative hydrolytic activity in the esterase assaywith r at hepat ic microsomes. The fluorine-substitutedesters, E16-1,2F 1, E 16-1,2F 2, and E16-1,2F 3, showed alarge increase in enzymat ic hydr olysis that wa s directlyrelated to the number of f luorine atoms at the 2-posit ion. The rapid rate of hydrolysis of esters of fluorinated alcohols h as been ascribed to the increasedacidity of the leaving group alcohol.31 Likewise, vinylesters have been shown to be excellent substrates foresterases,32 consistent with the very high rate of cleav-age of E16-1,2vin.

In evalua ting th ese compoun ds, binding to the E R is,of course, the most important factor in the determina-tion of estrogenic potential. However, in biologicalsystems, additiona l factors such a s catabolism m ust beweighed. In general , in these experiments, the estro-

genic potency of each compound as determined by thestimulation of AlkP in the Ishikawa cells (RSA) isconsistent with its binding to th e E R (RBA), providedthat i ts susceptibil i ty to esterase cleavage (RHA) isconsider ed. Because t he car boxylic acid ana logues of E2ar e inactive (Table 1), th e ra te of hydrolysis of the est ersis an importa nt factor. This can be seen in th e estrogenicpotency of the straight chain E 216-1 esters, E 216-1,1through E 216-1,4. They all have a pproximat ely the sa meRBA, but their RSA decreases with increasing chainlength, which reflects t heir increa sing enzymat ic cleav-age rat es with increasing chain length. The E16-2 andE16-3 esters h ave low RBAs an d high RHAs, and conse-quently, th e potency of all of th ese compounds is low.

The convergence of binding (RBA) and hydrolysis(RHA) as the determinant of potency (RSA) does nothold a s well for the fluorinated esters. As can be seenin Ta ble 1, E16-1,2F1 and E 16-1,2 ha ve about t he sameRBA for ERR, but since the monofluorinated ester iscleaved at a bout 4 times th e ra te of the n onfluorina tedester , i t might be assumed that E16-1,2 should beconsiderably more potent. However, both esters havethe same RSA. This also appears to be t ru e for thedifluoro and trifluoro ethyl ester s. While both of these

fluorinated esters h ave approximately the same RBAas E16-1,3i, their RHA is at least 17-60 times greater.If these two factors, enzymat ic hydrolysis and ERbinding, are consider ed, then t he isopropyl ester s houldbe the more potent. However, the isopropyl ester isalmost ina ctive, while conversely th e difluoro an d t ri-fluoro ethyl esters, although weak estrogens, are sig-nificantly more active. An explanation for these appar-ently conflicting findings may be that competit ivebinding studies, which are indirect measurements, donot always a ccur ately reflect t he t ru e binding affinity.More likely though, the RBA does not necessarily reflectthe ligan d-indu ced conformat iona l changes of th e r ecep-tor which directly affects the tr an scriptiona l stimulat ion

of estrogen responsive genes.33

Thus, receptor st imula-tion of genes is more complex th an is appar ent fromligand binding. In any case, t he fluorine-substi tutedesters are more potent estrogens than would be pre-dicted on t he ba sis of their RBA and RHA.

Three estrogen esters, th e methyl (E16-1,1), ethyl(E16-1,2), and fluoroethyl (E16-1,2F 1) ester s of E16-1,0,that were most potent in the Ishikawa assay were testedfor their systemic estrogenic activity in the classical invivo assay, u terotrophic st imulation of th e immaturera t .34 In this assay al l of the test compounds wereadministered in sesame oil. Again, E 2 is included forcomparison. As can be seen in Figure 2, 5 ng of E 2produced a statistically significant stimulation in the

F i g u r e 2 . In vivo systemic estrogenic assay (uterotrophicassa y). Immat ur e female ra ts (22 days old) were injected withE16-1,1, E16-1,2, and E16-1,2F 1 as well as E 2 in sesame oilonce daily for 3 days. On the fourth day th e an imals were

killed, and t he ut eri were removed an d weighed. The total doseis shown. Controls were injected with sesame oil. Error bar sa re (SD . n ) 5. *P < 0.05, **P < 0.001 when compared t o th econtrol.



6/13

weight of the uterus. As hypothesized, the systemicpotency of the est ers wa s very low. E16-1,1, the met hylester, produced a small uterotrophic effect (P < 0.05)a t 1 0 g (total dose per animal) in the experimentshown. In two separate experiments, a small effect atthis dose was a lso detected, but it was n ot st atisticallysignificant. This ester produced a reproducible andstatistically significant stimulation at 30-50 g (P 99% pure.

Methyl (3,17-Dihydroxyestra-1,3,5(10)-trien-16r-yl)-formate (9, E16-1,1). A solution of 28 mg (0.088 mmol) ofcarboxylic acid 8 and 9.66 L (0.132 mmol) of SOCl2 in 2 mLof MeOH was st irred a nd h eated a t 40 C for 2.25 h in a 5 mLflask equipped with a reflux condenser. The reaction mixturewas poured into saturated aqueous NaHCO3 (50 mL) andextracted with CH 2Cl2 (3, 50 mL). The combined organicextracts were dried over N a 2SO 4 and concentrated in vacuo,giving a brown oil. Purification of t he residue by f lashchromatography on a 2 16 cm column of silica gel usinghexan es/EtOAc (1:1) as eluen t ga ve 23 mg (79%) of9 as a whitesolid. Purification of this mat erial by H PLC in system H -3,280 nm, gave 17 m g of9 for bioassa y. Data for 9: TLC, T-6,Rf 0.325; 1H N MR (500 MHz, CDCl3, D2O) 0.84 (s, 3H, H -18),3.75 (s, 3H, OCH 3), 3.89 (d, 1H, J ) 7.7 Hz, H-17R), 6.57 (d,1H , J ) 2.8 Hz, H-4), 6.63 (dd, 1H, J ) 8.3, 2.8 Hz, H -2), 7.15(d, 1H, J ) 8.3 Hz, H-1); HRMS (ES) calcd for C20H 30NO4 (M+ NH 4+) m/e 348.2175, found m/e 348.2191; HP LC system, H -3,280 nm, tR ) 14 min, and system H-4, 280 nm, tR ) 8 min,>99% pure.

Ethyl (3,17-Dihydroxyestra-1,3,5(10)-trien-16r-yl)for-mate (10, E16-1,2). Compound 10 was pr epared by est erifi-cation of acid 8 (14 mg, 0.043 mmol) with EtOH as describedfor the preparation of9. Purification of the residue by flashchromatography on a 1 15 cm column of silica gel usinghexanes/EtOAc (1:1) as eluent gave 14 mg (94%) of 10 a s awhite solid. Purification of this material by HPLC (H-3, 280nm , tR ) 11 min) gave 11 mg of10 for bioassa y. Data for 10 :TLC, T-6, Rf 0.46; 1H NMR (500 MHz, CDCl3, D2O) 0.84(s,3H, H -18), 1.31 (t, 3H, J ) 7.1 Hz, OCH 2CH3), 3.88 (d, 1H, J) 8.0 Hz, H-17R), 4.21 (q, 2H, J ) 7.1 Hz, OCH2CH 3), 6.57 (d,1H , J ) 2.8 Hz, H-4), 6.63 (dd, 1H, J ) 8.4, 2.8 Hz, H -2), 7.15



9/13

(d, 1H, J ) 8.4 Hz, H-1); HRMS (ES) calcd for C21H 32NO 4 (M+ NH 4+) m/e 362.2331, found m/e 362.2331. Anal. (C21H 28O4)C, H.

P r o p y l ( 3 ,1 7-Dihydroxyestra-1,3,5(10)-trien-16 r-yl)-format e (11, E16-1,3). Compound 11 was prepared by esteri-fication of acid 8 (66 mg, 0.209 mmol) with n -propanol asdescribed for the preparation of9. Purification of the residueby flash chromatography on a 2 15 cm column of silica gelusing hexanes/EtOAc (2:1) as eluent gave 46 mg (62%) of11as a wh ite solid. Pu rificat ion of 22 mg of th is mat erial by HP LC(H-5) gave 20 m g of11 for bioassay. Data for 11 : TLC, T-2, Rf) 0.41; 1H N MR (500 MHz, CDCl3) 0.85 (s, 3H, H-18), 0.98(t, 3H, J ) 7.6 Hz, CH 3), 3.88 (d, 1H , J ) 7.4 Hz, H-17R), 4.11(m, 2H, OCH 2), 6.57 (d, 1H, J ) 2.5 Hz, H-4), 6.63 (dd, 1H, J) 8.3, 2.5 Hz, H-2), 7.16 (d, 1H, J ) 8.3 Hz, H-1); HRMS (ES)calcd for C22H 34NO4 (M + NH 4+) m/e 376.3488, found m/e376.2493; HPLC system, H-3, 280 nm, tR ) 12 min, and systemH-5, 280 n m, tR ) 9.5 min, >99% pure.

I s o p r o p y l ( 3 ,1 7-Dihydroxyestra-1,3,5(10)-trien-16r-yl)formate (12, E16-1,3i). A solution of 44 mg (0.14 mmol)of acid 8 an d 10 m g (0.053 mm ol) of pTsOH in 2-propan ol (20mL) was stirred a nd h eated at 85 C for 18 h. A Dean-S ta r kt r a p f i l le d wi th 4 s ie ve s wa s a dde d a nd he a t ing wa scontinued for 18 h. the reaction mixture was allowed to coolto room temperat ure, poured into satu rated a queous Na HCO3(20 mL), and extracted with CH 2Cl2 (3, 50 mL). The combinedorganic extracts were dried over Na

2SO

4and concentra ted in

vacuo. Pu rificat ion of the residu e by flash chr omatograp hy ona 2 15 cm column of silica gel using hexanes/EtOAc (2:1) aseluent gave 21 mg (43%) of12 as a white s olid. Pur ification ofthis material by HPLC in system H-3, 280 nm, followed bycrystallization from acetone/petroleum eth er, gave 8 mg of12for bioassay. Data for 12 : TLC, T-6, Rf 0.55; 1H NMR (500MHz, CDCl3) 0.84 (s, 3H, H-18), 1.28 (d, 3H, J ) 6.5 Hz,CH 3), 1.282 (d, 3H, J ) 6.5 Hz, CH 3), 3.85 (br d , 1H, J ) 9.3Hz, H-17R), 5.07 (sept, 1H, J ) 6.5 Hz, -CH-), 6.57 (d, 1H, J) 2.7 Hz, H-4), 6.63 (dd, 1H, J ) 8.3, 2.7 Hz, H-2), 7.15 (d,1H , J ) 8.3 Hz, H-1); HRMS (ES) calcd for C22H 34NO4 (M +NH 4+) m/e 376.2488, found m/e 376.2485; HP LC system , H-3,280 nm, tR ) 11 min, and system H-6, 280 nm, tR ) 7 min,>99% pure.

n -Buty l (3,17-Dihydroxyestra-1,3,5(10)-trien-16 r-yl)-

format e (13, E16-1,4). Compound 13 was prepared by esteri-fication of acid 8 (59 mg, 0.19 mmol) with bu tan ol as describedfor the preparat ion of9. Purification of the residue by flashchromatography on a 2 15 cm colum n of silica gel gave 50mg (73%) of 13 as a white solid. Data for 13 : TLC, T-2, Rf0.30; 1H NMR (500 MHz, CDCl3) 0.85 (s, 3H, H-18), 0.96 (t,3H , J ) 7.4 Hz, -CH 3), 3.87 (d, 1H, J ) 8.0 Hz, H-17R), 4.15(m, 2H, OCH 2), 6.57 (d, 1H, J ) 2.7 Hz, H-4), 6.63 (dd, 1H, J) 8.4, 2.7 Hz, H-2), 7.15 (d, 1H, J ) 8.4 Hz, H-1); HRMS (ES)calcd for C23H 36NO4 (M + NH 4+) m/e 390.2644, found m/e390.2647. HPLC system H-3, 280 nm , tR ) 11 min, and systemH-6, 280 n m, tR ) 9 min, >99% pure.

2 ,2 -D i m e t h y l p r o p y l ( 3 ,1 7-Dihydroxyestra-1,3,5(10)-trien-16r-yl)formate (14, E16-1,5ne o). Compound 14 wa sprepa red by est erificat ion of acid 8 (85 mg, 0.27 mm ol) with 2mL of neopentyl a lcohol in 10 mL of benzene a s described for

the preparation of 12 . Purification of the residue by flashchromatography on a 2 16 cm column of silica gel usinghexanes/EtOAc (3:1) gave 92 mg (89%) of14 as a white solid.Data for 14 : TLC, T-5, Rf 0.77; 1H N MR (500 MHz, CDCl3) 0.86 (s, 3H, H -18), 0.97 (s, 9H, CH 3), 3.83 & 3.87 (AB quart et,2H , J ) 10.5 Hz, -CH 2-), 3.90 (d, 1H, J ) 8.1 Hz, H-17R),6.57 (d, 1H, J ) 2.6 Hz, H -4), 6.63 (dd, 1H, J ) 8.5, 2.6 Hz,H-2), 7.15 (d, 1H, J ) 8.5 Hz, H-1); HRMS (ES) calcd forC24H 38NO4 (M + NH 4+) m/e 404.2801, found m/e 404.2809.Anal. (C24H 34O4) C, H.

V i n y l ( 3 ,1 7-Dihydroxyestran-1,3,5(10)-trien-16r-yl)-formate (15, E16-1,2 vin). A solution of 24 mg (0.077 mmol)of acid 8 in 2 mL of vinyl propiona te was s tirr ed as 20 L of a0.1 M solution of PdCl2-LiCl in vinyl propionat e was added.This solution wa s prepa red by combinin g 17 mg (0.1 mm ol) ofPdCl2 and 4.2 mg (0.1 mmol) of LiCl in 1 mL of MeOH with

heating to dissolve, evaporation of the solvent, and resuspen-sion in 1 mL of vinyl propionat e. The reaction mixture wasstir red at 92 C for 4 h, poured into CH 2Cl2 (50 mL), andwashed with saturated a queous Na HCO3 (20 mL) and H 2O(20 mL). The organic layer was dr ied over Na 2SO 4 a ndconcentrated in vacuo. Purification of the residue by flashchromatography on a 2 17 cm column of silica gel usinghexanes/EtOAc (3:1) gave 20 mg (76%) of15 as a white solid.Purification of 7 mg of this material in HPLC system H-7,followed by crystallization from Et 2O/hexane, gave 5 mg of15for bioassay. Data for 15 : TLC, T-6, Rf 0.57; 1H NMR (500MHz, CDCl3) 0.86 (s, 3H, H-18), 3.95 (d, 1H, J ) 7.7 Hz,H-17R), 4.63 (dd, 1H, J ) 6.3, 1.6 Hz, vinyl-H), 4.94 (dd, 1H,J ) 14.0, 1.6 Hz, vinyl-H), 6.57 (d, 1H, J ) 2.8 Hz, H-4), 6.64(dd, 1H, J ) 8.3, 2.8 Hz, H-2), 7.16 (d, 1H, J ) 8.3 Hz, H-1),7.33 (dd, 1H, J ) 14.0, 6.3 Hz, vinyl-H); HRMS (ES) calcd forC21H 30NO4 (M + NH 4+) m/e 360.2175, found m/e 360.2171;HPLC system H-3, 280 nm, tR ) 12.21 min, an d system H-8,280 nm, tR ) 11.5 min, >99% pure.

2-F l u o r o e t h y l ( 3 ,1 7-Dihydroxyestra-1,3,5(10)-trien-16r-yl)formate (16, E16-1,2 F1). Compound 16 was preparedby est erificat ion of acid 8 (16 mg, 0.05 mmol) with 1.5 mL offluoroethanol in 1.5 mL of toluene as descr ibed for thepreparation of12 . Pu rification of th e r esidue by flash chro-matography on a 2 16 cm column of silica gel using hexanes/Et OAc (2:1) as elu ent gave 11 m g (59%) of16 as a wh ite solid.

Pu rificat ion of th is mat erial by HP LC in syst em H-3, followedby crysta llizat ion from Et 2O/petroleum eth er, gave 8 mg of16for bioassay. Data for 16 : TLC, T-6, Rf) 0.375; 1H NMR (500MHz, CDCl3) 0.85 (s, 3H, H-18), 3.91 (d, 1H, J ) 8.2 Hz,H-17R), 4.36-4.44 (m 2H, CH2CH 2F), 4.64 (dt, 2H , J ) 47.4,4.2 Hz, CH 2CH2F), 6.57 (d, 1H, J ) 2.8 Hz, H-4), 6.63 (dd, 1H,J ) 8.3, 2.8 Hz, H-2), 7.16 (d, 1H, J ) 8.3, H-1); HRMS (ES)calcd for C21H 31FNO4 (M + NH 4+) m/e 380.2237, foun d m/e380.2248; HPLC, system H-3, 280 nm, tR ) 13.0 min, andsystem H-8, 280 nm, tR ) 8.5 min, >99% pure.

2,2-Difluoroethyl (3,17-Dihydroxyestra-1,3,5(10)-trien-16r-yl)formate (17, E16-1,2 F2). Compound 17 was preparedby esterification of acid 8 (25 mg, 0.08 mmol) with 2,2-difluoroethanol as described for the preparation of 16 . Puri-fication of this residue by flash chromatography on a 2 17cm colum n of silica gel u sin g hexa nes /EtO Ac (3:1) followed by

hexanes/EtOAc (2:1) as eluent gave 29 mg of 17 as a yellowoil. Purification of this material by HPLC with system H-3gave 23 mg (74%) of17 as a clear colorless oil. Fu rt her H PLCpurification of 6 mg of this material with system H-7, followedby crystallization from Et 2O/hexanes, gave 5 mg of 17 forbioassay. Data for 17 : TLC, T-6, Rf 0.5; 1H NMR (500 MHz,CDCl3) 0.85 (s, 3H, H-18), 3.91 (d, 1H, J ) 8.3 Hz, H-17R),4.32-4.39 (m, 2H, CH2CH F 2), 5.99 (tt, 1H , J ) 55.2, 4.0 Hz,CH 2CHF 2), 6.57 (d, 1H, J ) 2.8 Hz, H-4), 6.64 (dd, 1H, J )8.4, 2.8 Hz, H -2), 7.16 (d, 1H , J ) 8.4 Hz, H-1); HRMS (ES)calcd for C21H 30F 2NO4 (M + NH 4+) m/e 398.2143, foun d m/e398.2148; HPLC system H-3, 280 nm, tR ) 12.8 min, andsystem H-8, 280 nm, tR ) 10.5 min, >99% pure.

2,2,2-Trifluoroethyl (3,17-Dihydroxyestra-1,3,5(10)-trien-16r-yl)formate (18, E16-1,2 F3). Compound 18 wa s

prepa red by est erificat ion of acid 8 (25 mg, 0.079 mm ol) with2,2,2-trifluoroethanol as described for the preparation of16 .Purification of the residue by flash chromatography on a 2 17 cm column of silica gel using hexanes/EtOAc (4:1) as eluentgave 19 mg of 18 . HPLC purification of this material withsystem H-3 gave 17 mg (54%) of 18 as a slightly yellow oil.Further HPLC purif ication of 3 mg of this mater ial withsystem H-8, followed by crystallization from Et 2O/hexanes,gave 2 mg of18 as an amorphous solid. Data for 18 : TLC,T-6, Rf 0.58; 1H N MR (500 MHz, CDCl3) 0.86 (s, 3H, H-18),3.92 (d, 1H, J ) 8.3 Hz, H -17R), 4.52-4.58 (m, 2H, CH2CF 3),6.57 (d, 1H, J ) 3.0 Hz, H -4), 6.64 (dd, 1H, J ) 8.4, 3.0 Hz,H-2), 7.15 (d, 1H, J ) 8.4 Hz, H-1); HRMS (ES) calcd forC21H 29F 3NO4 (M + NH 4+) m/e 416.2049, foun d m/e 416.2051;HPLC system H-3, 280 nm, tR ) 11.6 min, and system H-8,280 nm, tR ) 13.5 min, >99% pure.



10/13

16r-Hydrox yme thyle stra-1,3,5(10)-trien-3,17-diol (19).A solution of 5 mg (0.015 mmol) of ethyl ester 10 a nd 5 m g(0.13 mm ol) of LiAlH 4 in 1 mL of anh ydrous THF was stirredat 0 C for 1 h under N 2. The reaction was quenched with 1mL of EtOAc, poured int o satur ated a queous Na-K tar tarate(5 mL), and extracted with EtOAc (3, 5 mL). The combinedorganic extracts were dried over Na 2SO 4 and concentra ted invacuo giving 3.5 mg (75%) of19 20 as a white solid: TLC, T-5,Rf 0.51; 1H NMR (500 MHz, DMSO-d6 + D2O) 0.68 (s, 3H,H-18), 3.15 (d, 1H, J ) 8.1 Hz, H-17R), 3.28 (dd, 1H , J ) 10.3,7.1 Hz, CH2OH), 3.51 (dd, 1H, J ) 10.3, 4.1 Hz, CH2OH), 6.42(d, 1H, J ) 2.6 Hz, H -2), 6.49 (dd, 1H, J ) 8.4, 2.6 Hz, H-2),7.02 (d, 1H, J) 8.4 Hz, H-1); HRMS (ES) calcd for C 19H 26O3Na(M + Na+) m/e 325.1780, foun d m/e 325.1782.

Eth yl (3-Benz yloxy -17-oxoest ra-1,3,5(10)-trien -16r-yl)-a c e t a t e ( 2 0 ) . A solution of 2.196 g (6.09 mmol) of estronebenzyl ether in 20 mL of anhydrous THF was added in oneportion a t 0 C to a s olution of 5.79 mm ol of LDA (2.89 mL ofa 2 M solution in heptane, THF, ethylbenzene) in THF (10mL). The resulting mixture was cooled to -4 5 C a n d asolution of 3.05 g (2.0 mL, 18.3 m mol) of ethyl br omoacetat ein 8 mL of THF was added d ropwise over 5 min . The reactionwas stir red at 99% pure.

3,17-Dihydroxyestra-1,3,5(10)-trien-16r-ylacetic Acid(23, E16-2,0). Compound 23 was prepared by saponificationof eth yl ester 22 (43 mg, 0.12 mmol) as described for 7, giving40 mg (100%). Furth er pu rification of 6 mg of this ma terialby HPLC with system H-1, followed by acid/base extraction,gave 5 m g of23 for bioassay. Data for 23 : 1H NMR (500 MHz,

DMSO-d6) 0.70 (s, 3H, H-18), 3.12 (d, 1H, J) 7.8 Hz, H-17R),6.42 (d, 1H, J ) 2.5 Hz, H -4), 6.49 (dd, 1H, J ) 8.7, 2.5 Hz,H-2), 7.03 (d, 1H, J ) 8.7 Hz, H -1), 8.96 (s, 1H, OH ); HRMS(ES) calcd for C20H 30NO4 (M + NH 4+) m/e 348.2175, found m/e348.2188; HPLC system H-1, 280 nm, tR ) 10 min, and systemH-10, 280 nm , tR ) 12 min, > 99% pure.

3-Benzyloxy-17-hydroxyestra-1,3,5(10)-trien-16r-yl-acetic Acid (24). Compound 24 was prepared by saponifica-tion of ester 21 (107 mg, 0.238 mm ol) as d escribed for 7, giving60 m g (60%) of24 as a white solid. This material was used

without furt her pu rification in the n ext step. TLC, T-5, Rf 0.36.Methyl (3-Benzyloxy-17-hydroxyestra-1,3,5(10)-trien-16r-yl)acetate (25). Compound 25 was prepared by esteri-fication of cru de 24 (60 mg) with MeOH as described for thepreparat ion of9. Purification of the residue by flash chroma-tography on a 2 15 cm column of silica gel using CHCl 3/Et OAc (5:0.15) gave 57 mg (92%) of25 as a white solid. Datafor 25 : TLC, T-5, Rf 0.8; 1H N MR (500 MHz, CDCl3 + D2O) 0.85 (s, 3H, H-18), 3.37 (d, 1H, J ) 7.3 Hz, H-17R), 3.71 (s,3H, OCH 3), 5.04 (s, 2H, benzylic), 6.72 (d, 1H, J) 2.8 Hz, H-4),6.79 (dd, 1H, J ) 8.5, 2.8 Hz, H-2), 7.21 (d, J ) 8.5 Hz, H-1),7.31-7.44 (m, 5H, Ar-H); HRMS (ES) calcd for C28H 34O4Na(M + Na+) m/e 457.2355, foun d 457.2342.

Me t h y l 3 ,1 7-Dihydroxyestra-1,3,5(10)-trien-16r-yl-acetate (26, E16-2,1). Compound 26 was prepared by hydro-genolysis of 25 (55 mg, 0.126 mmol) as described for the

preparat ion of8. Purification of the residue by flash chroma-tography on a 2 17 cm column of silica using hexanes/EtOAc(2:1) as eluent gave 33 mg of 26 . HPLC purification in sixportions with system H-3 gave 28 mg (64%) of 26 as a whitesolid. Data for 26 : TLC, T-6, Rf 0.456; 1H NMR (500 MHz,CDCl3) 0.86 (s, 3H, H-18), 3.38 (d, 1H, J ) 7.3 Hz, H-17R),3.71 (s, 3H, OCH 3), 6.56 (d, 1H, J ) 2.5 Hz, H-4), 6.63 (dd,1H , J) 8.4, 2.5 Hz, H -2), 7.16 (d, 1H, J) 8.4 Hz, H-1); HRMS(ES) calcd for C21H 32NO4 (M + NH 4+) m/e 345.2066, found m/e345.2082; HPLC system H-3, 280 nm, tR ) 11 min, and systemH-11, 280 nm , tR ) 18 min, >99% pure.

3-Benzyloxy-17-hydroxyestra-1,3,5(10)-trien-16-yl-a c e t a l d e h y d e ( 2 7 ) . A solution of 10 mg (0.022 m mol) of21in anh ydrous toluene (200 L) was stirr ed at -60 C as 0.0669mmol of Dibal (44 L of a 1.5 M solution in t oluene) was ad ded.The reaction was stir red at -60 C for 2 h, quenched withMeOH (2 mL), poured into H 2O (5 mL), and extracted withEt OAc (3, 5 mL). Combined organ ic extra cts were dr ied overNa 2SO 4 an d concentr at ed in vacuo. Pu rificat ion of the residueby flash chromatography on a 2 17 cm column of silica gelusin g hexan es/EtOAc (1:1) as eluent gave 5 m g (59%) of27 a sa clear colorless oil. Data for 27 : TLC, T-6, Rf 0.3; 1H NMR(500 MHz, CDCl3) 0.86 (s, 3H, H -18), 3.33 (d, 1H , J ) 7.3Hz, H-17R), 5.04 (s, 3H, benzylic), 6.72 (d, 1H, J ) 2.5 Hz,H-4), 6.79 (dd, 1H, J ) 8.8, 2.5 Hz, H-2), 7.21 (d, 1H, J ) 8.8Hz, H -1), 7.31-7.44 (m, 5H , Ar-H), 9.83 (s, 1H, CHO); HRMS(ES) calcd for C27H 36NO3 (M + NH 4+) m/e 422.2695, found m/e422.2681.

16r-Allyl-3-benzyloxyestra-1,3,5(10)-trien-17 -ol (28). Asolut ion of 5 mg (0.013 mm ol) of27 in anhydrous toluene (100L), pyridine (1 L), and THF (33 L) was stirred at -78 C

as 0.0156 mm ol of Tebbe reagent (31 L of a 0.5 M solut ion intoluene) was added by syringe. The reaction was stirred at-78 C for 2 h, at 40 C for 2 h, an d th en a t 0 C for 1 h . Thereaction was quenched with 15% NaOH (25 L), allowed tos t i r f or 0 . 5 h , wa r m e d to r oom te m pe r a tur e , a nd pa sse dthrough a 1 in. plug of Celite. The filter was washed withEtOAc and th e filtrat e was concentra ted in vacuo. Purificat ionof the r e s idue on a 1 17 cm column of silica gel usinghexa nes/E tOAc (2:1) followed by hexa nes /EtO Ac (4:1) gave 1.3mg (25%) of28 26 as a clear colorless oil and 1.8 mg (34%) ofrecovered 27 . Data for 28: TLC, T-6, Rf 0.76; 1H NMR (500MHz, CDCl3) 0.83 (s, 3H, H-18), 2.86 (m, 2H, H-6), 3.33 (d,1H , J ) 7.4 Hz, H-17R), 5.03-5.12 (m, 2H, dCH 2), 5.04 (s,2H, benzylic), 5.85-5.93 (m, 1H, -CH)CH 2), 6.72 d, 1H, J )2.6 Hz, H -4), 6.79 (dd, 1H, J ) 8.7, 2.6 Hz, H-2), 7.21 (d, 1H,J ) 8.7 Hz, H-1), 7.31-7.44 (m, 5H, Ar-H).



11/13

16r-Allyl-3-benzyloxyestra-1,3,5(10)-trien-17-yl Ace-tate (29). A solution of 1.874 g (4.65 mmol) of 28 26 and 5.2mL (55.1 mL) of acetic anhydride in 10.4 m L of anh ydrouspyridine was stir red at room temperature for 16.5 h. Thereaction was poured into H 2O (300 mL) and extracted withCH 2Cl2 (3, 200 mL). Combined organic extracts were driedover Na 2SO 4 and concentrated in vacuo. Purification of theresidue by flash chromatography on a 3 20 cm column ofsilica gel u sing CH 2Cl2 as eluent gave 1.73 g (84%) of29 a s awhit e solid. Dat a for 29 : TLC, T-7, Rf 0.48; 1H N MR (500 MHz,CDCl3) 0.84 (s, 3H, H-18), 2.07 (s, 3H, Ac), 2.84 (m, 2H, H-6),

4.64 (d, 1H, J ) 7.3 Hz, H-17R), 4.99-5.07 (m, 2H, dCH 2),5.04 (s, 3H, benzylic), 5.78 (m, 1H , -CH)), 6.72 (d, 1H, J )2.0 Hz, H-4), 6.78 (dd, 1H, J ) 8.6, 2.0 Hz, H-2), 7.19 (d, 1H,J ) 8.6 Hz, H-1), 7.31-7.44 (m, 5H, Ar-H).

3-Benzyloxy-16r-(3-hydroxypropyl)estra-1,3,5(10)-trien-17-yl Acetate (30). A solution of 629 mg (1.41 mmol) of29in anhydrous diglyme (21 mL) was stirred at 0 C as 1.49 mmolof borane-THF (1.49 mL of a 1 M solution in THF) was added.The rea ction was s tirr ed at 0 C for 0.5 h, allowed to warm toroom tempera tur e, and stirred for 2 h. To this was added 660mg (5.96 mmol) of trimet hylamine oxide, and t he r eaction wa sstirred a nd heated at 150 C for 2 h, cooled to room t emper-atu re, poured int o H2O (150 mL), and extracted with CH 2Cl2(3, 100 mL). The combined organic extracts were washedwith 10% sodium metabisulfite (70 mL), H 2O (70 mL), driedover Na 2SO 4, an d concentra ted in vacuo. Purification of theresidue by flash chromatography on a 3 22 cm column ofsilica gel us ing hexa nes/Et OAc (1.5:1) as elu ent gave 530 mg(81%) of30 as a white solid. Data for 30 : TLC, T-6, Rf 0.34;1H NMR (500 MHz, CDCl3) 0.83 (s, 3H, H -18), 2.09 (s, 3H,OAc), 2.85 (m, 2H , H -6), 3.85 (t, 2H, J ) 5.7 Hz, -CH 2O-),4.63 (d, 1H, J ) 7.8 Hz, H-17R), 5.04 (s, 2H, ben zylic), 6.72 (d,1H , J ) 2.6 Hz, H-4), 6.78 (dd, 1H, J ) 8.2, 2.6 Hz, H -2), 7.19(d, 1H, J ) 2.6 Hz, H-1), 7.31-7.44 (m, 5H, Ar-H); HRMS(ES) calcd for C30H 38O4Na (M + Na+) m/e 485.2668, found m/e485.2679.

3-(3-Benzyloxy-17-acetoxyestra-1,3,5(10)-trien-16 r-yl)-p r o p a n o i c A c i d ( 3 1 ). Compound 31 was prepared by CrO 3oxidation of30 (530 mg, 1.14 mmol) as described for 6, giving474 mg of31 as a wh ite foam: TLC, T-5, Rf 0.64. This mat erialwas used without further purification in the next step. HRMS(ES) calcd for C30H 36O5Na (M + Na+) m/e 499.2460, found m/e499.2449.

3-(3-Benzyloxy-17-hydroxyestra-1,3,5(10)-trien-16r-y l ) p ro p a n o i c A c i d ( 3 2 ). Compound 32 was prepared bysaponificat ion of cru de 31 (474 mg) as described for 7, giving436 mg of32 as a wh ite foam: TLC, T-5, Rf 0.54. This mat erialwas used without further purification in the next step.

Methyl 3-(3-Benzyloxy-17-hydroxyestra-1,3,5(10)-trien-16r-y l ) p ro p a n o a t e ( 3 3) . Compound 33 was prepared byesterification of crude 32 (436 mg) with MeOH as describedfor the preparat ion of9. Purification of the residue by flashchromatography on a 3 21 cm column of silica gel usinghexan es/EtOAc (2:1) as eluen t gave 268 mg (52%, th ree steps )of33 as a white solid. Data for 33 : TLC, T-5, Rf 0.76; 1H NMR(500 MHz, CDCl3) 0.81 (s, 3H, H -18), 2.84 (m, 2H, H -6), 3.31(d, 1H, J ) 7.3 Hz, H-17R), 3.69 (s, 3H, OCH 3), 5.04 (s, 2H,

benzylic), 6.72 (d, 1H, J)

2.8 Hz, H-4), 6.79 (dd, 1H, J)

8.6,2.8 Hz, H-2), 7.21 (d, 1H, J ) 8.6 Hz, H -1), 7.31-7.44 (m, 5H,Ar-H); HRMS (ES) calcd for C29H 36O4Na (M + Na+) m/e471.2511, foun d m/e 471.2513.

Methyl 3-(3,17-Dihydroxyestra-1,3,5(10)-trien-16r-yl)-propanoate (34, E16-3,1). Compound 34 was prepared byhydrogenolysis of33 (37 mg, 0.083 mmol) as described for thepreparat ion of8. Purification of the residue by flash chroma-tography on a 1 20 cm column of silica gel using hexanes/Et OAc (2:1) as elu ent gave 25 m g (84%) of34 . Pu rificat ion of20 mg of this ma terial by HPLC with system H-3 gave 18 mgof 33 for bioassay. Data for 33 : TLC, T-6, Rf 0.45; 1H NMR(500 MHz, CDCl3) 0.81 (s, 3H, H-18), 3.31 (d, 1H, J ) 7.4Hz, H-17R), 3.69 (s, 3H, OCH 3), 6.57 (d, 1H, J ) 2.6 Hz, H-4),6.63 (dd, 1H, J ) 8.4, 2.6 Hz, H-2), 7.16 (d, 1H, J ) 8.4 Hz,H-1); HRMS (ES) calcd for C22H 34NO4 (M + NH 4+) m/e

359.2222, found m/e 359.2231; HPLC system H-3, 280 nm, tR) 14 min, and system H-12, 280 nm), tR ) 14.5 m in, >99 %pure.

3-(3,17-Dihydroxyestra-1,3,5(10)-trien-16r-yl)propan-i o c A c i d ( 3 5, E 1 6 -3 ,0 ) . Compound 35 wa s pr e pa r e d bysaponificat ion of34 (35 mg, 0.097 mmol) as described for 7.Purification of the residue by HPLC in 10 portions with systemH-13 ga ve 20 mg (59%) of35 as a wh ite solid. Dat a for 35 : 1HNMR (500 MHz, DMSO-d6) 0.68 (s, 3H, H-18), 3.08 (d, 1H,J ) 7.2 Hz, H-17R), 6.42 (d, 1H, J ) 2.1 Hz, H -4), 6.49 (dd,1H , J ) 8.2, 2.1 Hz, H-2), 7.03 (d, 1H, J ) 8.2 Hz, H-1), 8.96(s, 1H, OH ); HRMS (ES) calcd for C 21H 32NO4 (M + NH 4+) m/e362.2332, foun d m/e 362.2344; HPLC system H-14, 280 n m,tR ) 12 min, and system H-13, 280 nm, tR ) 23 min, >99 %pure.

C om p e t i t i ve B i n d i n g t o t h e E s t ro g e n R e c e p t o r E Rra n d E R. Binding affinities relative to E 2 were performed inincubations with th e ER (ERR39) in r at ut erine cytosol. FemaleSprague-Dawley rats were castra ted an d sacrificed 24 h later.The uterus was removed, homogenized in ice-cold TEGDMobuffer (10 mM Tris, 1.5 mM Na 2-EDTA, 10% (v/v) glycerol,1.0 mM dithiothreitol, 25 mM sodium molybdate, pH 7.4 at 4C), and centrifuged at 105000 g f or 45 m in a t 4 C . Thesuper na ta nt (cytosol) was frozen on dry ice an d stored at -80C until a ssay. For a ssay, the cytosol was defrosted, diluted,and incubated with 1 nm [3H]E 2 in th e presence and absence

of nonradioactive E 2, estrone (E 1), or th e E2-carboxy analoguesover a range of concentrations from 10-12 to 10-6 M. Incuba-tions were carried out on ice overnight, and bound radioactivitywas separated from free by adsorption with dextran-coatedcharcoal an d qua ntified by counting.27 Relative binding affinity(RBA) was det ermin ed by ana lysis of the displacement curvesby the curve-fitting program Prism. The results shown in Table1 are from at least three separate experiments performed induplicate. A subset of E 2-16R-alkyl esters was also comparedby Drs. Paul Shughr ue and H eather H arris for binding to theLB D of hum a n ERR (M250-V595 )40 a n d h u m a n E R (M214-Q530 ).41 The assa y was performed in competition with [3H]E 2in lysates ofEscherichia coli in which the LBDs are expressedas descr ibed, with the exception that the incubation wasperformed overnight at 0-2 C.28 The results, as the RBAscompa red t o that of E 2 an d th e rat io of th e RBAs of ERR/E R,are shown in Table 2.

E s t r o g e n i c P o t e n c y i n I s h i k a w a C e l l s . The estrogenicpotency of the E 2 analogues was determined in a estrogenbioassay, the induction of AlkP in human endometrial adeno-carcinoma cells (Ishika wa) grown in 96-well microtiter plat esas we have previously described.29 In sh ort, the cells ar e grownin ph enol red free medium with estrogen depleted (char coalstripped) bovine serum in the presence or absence of varyingamount s of th e steroids, across a dose range of at least 6 ordersof magnitude. E 2 a nd E 1 were in cluded for compa rison. After3 da ys , the c e l l s a r e wa she d, f r oz e n, tha we d, a nd the nincubated with 5 mM p-nitrophenyl phospha te, a chromogenicsubstrate for the AlkP enzyme, at pH 9.8. To ensure linearenzymatic analysis, the plates a re m onitored k inetically forthe production of p-nitrophenol a t 405 nm. The relativestimu latory activity (RSA) represen ts t he r atio of 1/EC50 of th e

steroid analogue t o that of E 2

100, using the curve fittingprogram Prism t o determine the E C50 . Each compound wasana lyzed in at least thr ee separate experiments performed induplicate.

In Vivo Estrogen Bioassays: Uterine Weight. Systemicestr ogenic potency was determ ined by a uterotr ophic assa y inimmature rats as described.34 Female Sprague-Dawley rat s,22 days old, were injected subcutaneously daily for 3 days withan injection volume of 0.1 mL of the 16R-alkyl esters, for atotal dose of between 1 a nd 300 g, or of E 2, for a total dose ofbetween 0.001 and 0.1 g, in sesame oil. Control animalsreceived sesame oil . On the fourth day, the animals werekilled, and the uteri were removed, dissected, blotted, andweighed. Ea ch compound was assayed in three separateexperiments with n ) 5. A typical experiment is shown inFigure 2.



12/13

I n Vi v o E s t r o g e n B i o a s s a y s : V a g in a l R e d u c t a s e s . Th eestr ogenic action of locally applied E 2-16R-alkyl esters on thevagina was deter mined by measur ing the indu ction of vaginalreductases.35 Female CD-1 mice were ovariectomized a nd 1week later were instilled with the E 2-16R-alkyl esters or E 2 in10 L of 25% propylene glycol in salin e. In some experiment s,as indicated, the method was modif ied by dissolving andinjecting the estrogens in 10 L of sesame oil . The nextmorning 0.5 mg of 2,3,5-triphenyltetrazolium chloride in 20L of saline was instilled in the vagina. Thirty minu tes later

the animals were killed, and the vaginas removed, washedthoroughly with saline, and then blotted on f i l ter paper .Each vagina was placed in a 12 7 5 m m t es t t u b e a n dextracted for 1 h with ethanol/te trachloroethylene (3:1).Afterwar d, the solvent wa s removed and th e formazan pr oductin the or ga nic e xt r a ct wa s qua nt ified a t 500 nm . Ea chcompound was assayed on at least three separate occasionswith at least five replicates each t ime. A typical experimentis shown in Figure 3.

E s t e r a s e . Esterase activity was measured in rat hepaticmicrosomes essen tially usin g th e conditions described.42 Liverobtained from Sprague-Dawley rat s was washed with phos-pha te-buffered saline, h omogenized in 3 volumes of cold 0.25M sucrose, and centrifuged at 700g for 10 min and then at10000g for 20 min. The resulting supernatant was centrifuged

at 105000g for 60 min. The pellet was suspended in 0.1 Mphosphate buffer (pH 7.4) and washed by centrifugation at105000g for 60 min. The washed pellet was suspended in 0.1M Tris-HCl (pH 8.0) at a concent rat ion of13 mg of protein/mL and frozen at -80 C. For assay, the pellets were thawedand diluted with the same buffer . The incubation mixtureconsisted of the microsomal enzyme preparation, 0.28 mg ofprotein/mL, 50 M E 2-16R-alkyl esters, added in 10 L ofethan ol, all in a final volume of 1 mL of pH 8.0 Tris bu ffer.Since the ra tes of reaction are widely different for the var iousesters , the incubat ion times were varied accordingly to obta inlinear kinetics. At several appropriate time points, 100 Laliquots were withdra wn an d th e reaction was quenched with33 L of CH 3CN, followed by 33 L of a solution of THFcontaining 1 g of the interna l stan dard, 6-ketoestradiol. Thequenched aliquot was centrifuged for several minutes on a

benchtop centr ifuge, a nd 80 L o f t h e s u p e r n a t a n t w a sanalyzed for the esterase h ydrolysis product [the corre-sponding E 2-16R-carboxylic acid: E16-1,0 (8), E16-2,0 (23 ),E16-3,0 (35 )] by reversed-phase HPLC with system H-10 forE 216-1 an d E 216-2 esters and with system H-15 for E 216-3esters. The E 2-16R-carboxyl products (tR for E16-1,0 (8) ) 7min and for E16-2,0 (23 ) ) 9 min, E16-3,0 (35 ) ) 6.5 min)and t he intern al stan dard, 6-ketoestradiol (tR 7.5 min in systemH-10, and 5 min in system H-15), were quantified at 280 nmon the H PLC UV detector. The UV absorbance was convert edto moles of product by comparison to standard curves andcorrected for recovery of the internal standard, 6-ketoestradiol.The velocity of the reaction for each ester, in nmol product/min/mL, was then normalized to t he ester, E16-1,2 (10 ) andis shown in Table 1. as r elative hydr olytic activity (RHA). The

enzymat ic velocity for th e hydr olysis of E16-1,2 (10 ) was 0.9( 0.2 (SD) nmol product/min/mL over the various experiments.Since all of the esters could not be t ested simultan eously, ineach case we compared the rate of hydrolysis of the testcom pound to tha t of E16-1 ,2 (10 ) run concurrently. Allcompounds were tested in triplicate in th ree separat e experi-ments.

The various E 2-16R-alkyl esters, including th e reactive E16-1,2vin (15 ) , were stable under the conditions used in theesterase a ssay. Each of the substrates was incubated withheat -denatu red (1 h at 80 C) enzyme at 37 C for periods thatexceeded the incubation t imes of th e enzyme a ssay. Onlyinsignifican t a mount s of car boxyl products wer e form ed fromany of these esters during the incubations with denaturedenzyme.

R e f e r e n c e s

(1) Sh lipak, M. G.; Simon, J . A.; Vittinghoff, E.; Lin, F.; Barrett -Connor, E .; Knopp, R. H.; Levy, R. I.; H ulley, S. B. Est rogenand Progestin, Lipoprotein(a), and the Risk of Recurr ent Coro-nar y Heart Disease Events After Menopause. J A M A 2000, 28 3,1845-1852.

(2) Beral, V.; Banks, E.; Reeves, G.; Appleby, P. Use of HRT andthe Su bsequent Risk of Cancer. J Epidemiol. Biostat. 1999, 4,191-210.

(3) Bank s, E.; Beral, V. Hormone Replacement Thera py for Second-ary Prevention of Coronary Heart Disease. J A M A 1999, 28 1,794-797.

(4) Beral, V.; Hermon, C.; Kay, C.; Hanna ford, P .; Darby, S.; Reeves,G. Mortality Associated With Oral Contraceptive Use: 25 YearFollow Up of Cohort of 46 000 Women From Royal College ofGeneral Practitioners Oral Contraception Study. B MJ 1999,318 , 96-100.

(5) Sarr el, P. M. Sexuality and Menopause. Obstet. Gynecol. 1990,75 , 26S-30S.

(6) Sarrel, P. M. Sexuality. In The Menopause; Studd, J., Whithead,M. I., Eds.; Blackwell Scientific Publications: London, England,1988; pp 65-75 .

(7) Punnonen, R.; Lovgren, T.; Kouvonen, I. Demonstr ation ofEstr ogen Receptors in t he Skin. J. Endocrinol. Invest. 1980, 3,217-221.

(8) Uzuka, M.; Nakajima, K.; Ohta, S.; Mori, Y. The Mechanism ofEstrogen-Induced Increase in Hyaluronic Acid Biosynthesis,With Special Reference to Est rogen Receptor in t he Mouse Skin.

B ioch im . B ioph ys. Act a 1980, 627, 199-206.(9) Schiff, I.; Tulchinsky, D.; Ryan, K. J. Vaginal Absorption of

Estrone and 17-Estradiol. Fertility Sterility 1977, 28 , 1063-1066.

(10) Rigg, L. A.; Hermann, H.; Yen, S. SC. Absorption of EstrogensFrom Vaginal Creams. N. En g l. J . M ed . 1978, 298 , 195-197.

(11) Mart in, P. L.; Yen, S. SC.; Burn ier, A. M.; Herman n, H. SystemicAbsorption and Sustained Effects of Vaginal Est rogen Cr eams.

J AM A 1979, 242 , 2699-2700.(12) Schiff, I.; Tulchinsky, D.; Ryan, K. J .; Kadner, S.; Levitz, M.

Plasma Estr iol and It s Conjugates Following Oral and VaginalAdministration of Estriol to Postmenopausal Women: Correla-tions With Gonadotropin Levels. Am. J. Obstet. Gynecol. 1980,138 , 1137-1141.

(13) Ha sselquist, M. B.; Goldberg, N.; Schr oeter, A.; Spelsberg, T. C.Isolation and Characterization of the Estrogen Receptor inHuman Skin. J. Clin. Endocrinol. Metab. 1980, 50 , 76-82.

(14) Bodor, N. Designing Safer Drugs Based on the Soft DrugApproach. Trends Pharm. Sci. 1982, 3, 53-56 .

(15) Graffner-Nordberg, M.; Sjodin, K.; Tunek, A.; Hallberg, A.Synthesis and Enzymatic Hydrolysis of E sters , ConstitutingSimple Models of Soft Dr ugs. Chem Pharm Bull (Tokyo). 1998,46, 591-601.

(16) Laur ent , H.; Gerha rds, E.; Wiechert , R. New Biologically ActivePregnan-21-Oic Acid Esters. J Steroid Biochem. 1975, 6, 185-192.

(17) Druzgala, P .; Hochha us, G.; Bodor, N. Soft Dr ugs--10. BlanchingActivity and Receptor Binding Affinity of a New Type ofGlucocorticoid: Loteprednol Etabonate. J Steroid Biochem Mol

B iol. 1991, 38 , 149-154.(18) Lee, H. J .; Soliman, M. R. I. Ant iinflamma tory Ster oids Without

Pituitary Adrenal Suppression. S cience 1982, 215 , 989-991.(19) Bucourt, R.; Vignau, M.; Torelli, V. New Biospecific Adsorbent s

for the Purification of Estradiol Receptor. J. Biol. Chem. 1978,253 , 8221-8228.

(20) Fevig, T. L.; Mao, M. K.; Katzenellenbogen, J. A. EstrogenReceptor Binding Tolerance of 16R-Substituted Estradiol De-rivatives. Steroids 1988, 51 , 471-498.

(21) Rasmusson, G. H.; Arth, G. E. Selective Oxidations of HydroxySteroids. In Organic Reactions in Steroid Chemistry.; Fried, J.,Edwar ds, J., Eds.; Van N ostran d Reinhold Co.: New York, 1972;pp 222

(22) Bjorkquist, D. W.; Bush, R. D.; Ezra, F. S.; Keough, T. Cyclo-polymerizat ion and Regioselective Synt hesis of Vinyl Itaconates.

J . Org . Ch em . 1986, 51 , 3196-3201.(23) Schneider, G.; Hackler, L.; Sohar, P. Preparation of 16R-

Hydroxymethyl-3-Methoxyestra-1,3,5(10)-Trien-17R-ol and Sol-volyis Investigations. Liebigs Ann. Chem . 1988, 679-683.

(24) Noyce, D. S.; Denney, D. B. Steric Effects and Stereochemistryof Lithium Aluminum Hydride Reduction. J . Am. Ch e m. S o c .1950, 72 , 5743-5745.

(25) Vaughan, W. R.; Perry, R., Jr. The Configuration of Isocamphe-nilanol. J . Am . Ch em. S o c. 1952, 74 , 5355-5356.

(26) Fevig, T. L.; Katzenellenbogen, J . A. A Short, StereoselectiveRoute to 16R-(Substituted-Alkyl) Estradiol Derivatives. J. Org.Chem. 1987, 52 , 247-251.

(27) Hochberg, R. B.; Rosner, W. The Inter action of 16R-[125I]Iodo-estradiol With Estr ogen Receptor and Ot her Binding Pr oteins.Proc. Natl. Acad. Sci. U.S.A. 1980, 77, 328-332.



13/13

(28) Shughrue, P. J.; Lane, M. V.; Merchenthaler, I. BiologicallyActive Est rogen Receptor-Beta: Evidence Fr om in Vivo Auto-radiographic Studies With Estr ogen Receptor Alpha-KnockoutMice. Endocrinology 1999, 14 0, 2613-2620.

(29) Littlefield, B. A.; Gu rpide, E .; Mar kiewicz, L.; McKinley, B.;Hochberg, R. B. A Simple a nd Sensitive Microtiter PlateEstrogen Bioassay Based on Stimulation of Alkaline Ph os-phata se in Ishikawa Cells: Estr ogenic Action of5 AdrenalSteroids. Endocrinology 1990, 12 7, 2757-2762.

(30) Pannatier, A.; Testa, B.; Etter, J . Enzymatic Hydrolysis byMouse Skin Homogenates: Structure-Metabolism Relationshipsof Par a-Nitrobenzoate Esters. I n t . J . Ph a r m. 1981, 8, 167-174.

(31) Bart on, P.; Laws, A. P.; Page, M. I. Stru cture-Activity Relation-

ships in the Esterase-Catalysed Hydrolysis and Transesterifi-cation of Est ers an d Lactones. J Chem S oc Perkin Trans 2 1994 ,., 2021-2029.

(32) For ro, E.; Lun dell, K.; Fu lop, F .; Kaner va, L. Pr epa ra tion of th eSter eoisomer s of 2-Cyanocycloalkan ols by Lipase-Cata lysed Acyl-ation. Tetrahedron: Asymm etry 1997, 8, 3095-3099.

(33) Kat zenellenbogen, J. A.; OMalley, B. W.; Katzenellenbogen, B.S. Tripartite Steroid Hormone Receptor Pharmacology: Interac-tion With Multiple Effector Sites As a Basis for the Cell- andPr omoter-Specific Action of These H ormones. Mol. En docrinol.1996, 10 , 119-131.

(34) Emmens, C. W. Estrogens. In Methods in Hormone Research ;Dorfman, R. I., Ed.; Academic Press Inc.: New York, 1962; pp59-111.

(35) Mar tin , L. The Use of 2-3-5-Triphenyltetrazolium Chloride inth e Biological Assay of Oestr ogens. J. End ocrinol. 1960, 20, 187-197.

(36) Lund -Pero, M.; Jeppson, B.; Arn eklo-Nobin, B.; Sjogren, H .-O.;Holmgren, K.; Pero, R. W. Nonspecific Steroidal EsteraseActivity and Distribution in Human and Other MammalianTissues. Clin. Chim. Acta 1994, 22 4, 9-20.

(37) Still , C. W.; Kahn, M.; Mitra, A. Rapid ChromatographicTechnique for Prepar ative Separat ions With Moderate Resolu-tion. J. Org. Chem. 1978, 43 , 2923-2925.

(38) Fr ie d J a nd Edwa rds J AE d s. Org an ic R eact ion s in S ter iodChemistry; Van Nostra nd Reinhold Co.: New York, 1972; pp 314

(39) Kuiper, G. G.; Carlsson, B.; Gran dien, K.; Enm ark , E.; Haggblad,J.; Nilsson, S.; Gustafsson, J. A. Comparison of the LigandBinding Specificity and Transcript Tissue Distribution of Es-

trogen Receptors Alpha and Beta. Endocrinology 1997, 13 8,863-870.(40) Green, S.; Walter, P.; Kuma r, V.; Krust , A.; Borner t, J . M.; Argos,

P.; Chambon, P. Human Oestrogen Receptor CDNA: Sequence,Expr ession and H omology to V-Erb-A. Na t u r e 1986, 32 0, 134-139.

(41) Ogawa, S.; Inoue, S.; Wata nabe, T.; Hir oi, H.; Orim o, A.; Hosoi,T.; Ouchi, Y.; Muramat su, M. The Complete Pr imary St ructur eof Human Estrogen Receptor Beta (HER Beta) and Its Het-erodimerization With ER Alpha in Vivo and in Vitro. Biochem.

B ioph ys. R es. Com m u n. 1998, 24 3, 122-126.(42) Schottler, C.; Krisch, K. Hydrolysis of Steroid Hormone Esters

by an Un specific Carboxylester ase Fr om Pig Liver Microsomes.B ioch em . Ph ar m acol. 1974, 23 , 2867-2875.

JM000523H


Articulo exam1

Documents