Tetrahedron Letters,Vo1.23,No.40,pp 4147-4150,1982 oo40-4039/82/404147-04$03.00/O Prmted m Great Brltam 01982 Pergamon Press Ltd.

THE SYNTHESIS OF FURAN-BASED SECOPROSTACYCLINS

John Saunders*, David C. Tlpney and Peter Robins.

(I.C.I. PLC, Pharmaceuticals Drvlslon, Alderley Park, Macclesfleld, Cheshire, SK10 4TG, England)

Summary Furan-based secoprostacycllns (2-6) have been syntheslsed by two routes dlfferlng In the stage at which the-furan ring 1s generated.

In contrast to the classical prostaglandlns (PG), prostacyclln (PG12, 1) has a direct effect on the tuberular reabsorption of sodium In the k1dney.l. Since the Merck group had shown2. useful, PG-like actlvlty may St111 be retalned In very simple prostanold derlvatlves (termed secoprostaglandins), we decided to make sunpler, stable analogues of PG12 to test as novel dluretlc agents. Furan-based secoprostacycllns (2-6) combined these features, the furan ring provides the crItIcal sp2-character at C-6 (FGrnumberlng) and also Incorporates the enol-ether functlonallty Into a more stable environment

In prlnclple, compound 2 could be syntheslsed from the parent furan 1 by lntroductlon of the bottom side-chain by Wlttlg oleflnatlon of the carbonyl group followed by alkylatlon of the resulting furan at the 5-posltlon to give the C2o skeleton. However, reactlon of the anIon of 2 with ethyl bromovalerate gave g4. rather than 14. The problem was solved by lntroductlon of the c-5 chain In reduced form as follows. 5-Bromo-1-pentanol was reacted xnmedlately with an excess of dunethyl t-butylsllyl chloride to give (by GC/MS) a 4 1 mixture of the bromide 10 and the chloride 11 - -* Since only the bromide reacted with

OH

OH Ill (lOIX=Br IlllX Cl

171R=CHJ 18)R-H

(91 (131 (14)R-CH3 (2) R-H

Reagents. 1. Ph3P+(CH2)7CH3Br-,KOBut, PhMe, 90°, furan added at 25', 11. H2, 10% Pd-C, abs. EtOH, 111. n-BuL1, THF, -5", 2.5h; 12 added at -60°, 1". n-BuqN+F-, O", THF, 0.5h, v. C5H5NH+ClCr03-, CH3C02Nc CH2C12, 20°, 3h, Vl. 2N-NaOH, AgN03, H20, 2O", 2h.

4147

4148

the metalated derlvatlve of 2, the mixture was converted to the lodlde 12 prior to use, desllylatlon of the resulting furan gave the alcohol 13 ready for oxlda;on to the acid. This was most conveniently accompllshed zn two steps zth pyrldlnlum chlorochromate III methylene dlchlorlde followed by sliver oxide oxldatlon of the IntermedIate aldehyde. For purlflcatlon the crude acid was lmmedlately esterlfled with dlazomethane, following chromatography and saponlflcatlon of the ester the pure acid 2 was Isolated as Its sodium - salt.

Appllcatlon of this sequence to a furan with a functlonallsed lower side-chain (eg 16, prepared in two steps from I and5the phosphonlum salt x) falled at the metalatlon stage regardless of the proteLtlng group and the ratlo of base to furan that was used. Addltlon of the top side-chain in two stages enabled the synthesis to be completed. The furan 22 was prepared via the usual Wlttlg reactlon between I and 18 (to give 20) followed by hydrogenation and acetylatlon. Formylatlon gave 23 and this reacted selectively at the aldehyde function with the ylld from 24. Hydrogenatlonof the Wlttlg product generated the alcohol 25 which was oxldlsed G give 2. Oleflnatlon of L with the phosph- orane from 19L gave 2land repetltlon of the steps described above afforded 4. - -

O._Ph

(19)

I O_OLAO/

Ph3; MO- Ph I- Ph3; "v"./OvPh Br-

(16) 118) (24)

17) (ZU)R=H

(Zl)R = n-C5Hll

OH C02H

(3)R=H

li R (4)R = n-C5H11

OH

Reagents. 1 18 or l9, n-BuLI, THF, -10’ to 25’, 0.5h, furan added at -lo”, II. H2, 10% Pd-C, 111.&O, Et3N, 4-(CH3)2NC5H4N, 25’, 3h, I”. POC13, DMF, 5O, lh, then

Na2C03, 1. 24, n-BuLl, THF, -lo”, furan added at -55’,v1.H2, 10% Pd-C, - VII. oxldatlon as above. -

Clearly this route could also be applied to the preparation of 5 and 5 from the aldehyde 8. However, we also wlshed to exploit the well-documented7 furan synthesis based on the acid- or base-catalysed ring closure of cl,Y-ketoalkynes. The IntermedIate $_

4149

required for the crucial cycllsatlon reactlon was obtalned by sequential alkylatlon of

t-butyl acetoacetate first with l-bromo -2-heptyne and then with the lodoacetate E2, the

dlalkylated material 30 was the malor product. Reverse addltlon of the side-chains using

the more reactive propargylllc bromide in the second step avolded this problem Exposure

of 28 to a trace of acid gave the furan 32 III 65% yield and this was hydrolysed to 2. If ze reactlon was Interrupted after on5 two hours, the IntermedIate 31 could be Isolated.

Substltutlon of l-bromo-Z-heptyne by methyl 7-bromoheptynoate8. In the xkylatlon of 11 followed by acid treatment of the product 2 afforded the ester 22, saponlflcatlon

completed the synthesis of g9.

I.i “K

0

(28)R-CH3 131lR CH3

129lR CO$H3

0

WIR-CH3 l51R=CH3

(33)R CO$t$ (6iR C02H

IMI

Reagents. 1. NaH, THF, O", 1-lodo-4-acetoxynonane (E), then reflux, lbh,

11 NaH, THF, 5', propargylllc bromide added at 5" then reflux, 3h, 111. PhMe,

z20, p-TsOH, reflux, lbh, 1~. 2N-NaOH, MeOY, 25", 48h.

References and Notes.

1. P.M. Bolger, G.M. Elsner, P.W. Ramwell, L.M. Slotkoff and E.J. Corey, Nature -9 1978, 271,

467.

2 J.B. Blcklng, C.M. Robb, R.L. Smith, E.J. Cragoe, F.A. Kuehl and L.R. Mandel, J.Medlcln.

m., 1977, 20, 35.

3 K.C. Nlcolaou, W.E. Barnette and R.L. Magolda, J.Amer.Chem.Soc , 1979, 101, 766.

4. All new compounds gave spectra (IR, NMR, MS) consistent with the asslgned structure and satisfactory accurate mass measurement or combustion analysis

5 Slmllar results were obtalned when either benzyl or tetrahydropyranyl was used as protecting group.

6. l-chloro-3-octanol, required in the synthesis of 19, was prepared by two alternatlve routes The first Involved sequential alkylatlon of 1,3-dlthlane with 1-bromopentane and the tetrahydropyranyl ether of 2-bromoethanol followed by desulphurlsatlon, chlorlnatlon and borohydrlde reduction of the ketone. This same IntermedIate was also avaILable from acroleln by addltlon of hydrogen chloride followed by pentylatlon of

4150

the aldehyde with pentyl magnesium bromide.

7. e.g. K.E. Schulte, J. Relsch and A. Mock, Arch.Pharm., 1962, 295, 627. During the course of this part of our programme we became aware of the accldental synthesis of a close analogue of 1 by the Merck group - J.Medlcln.Chem., 1978, 1, 1011.

8 E.J. Corey and H.S. Sachdev, J.Amer.Chem.Soc., 1973, 95, 8483.

9 All compounds were tested for their ablllty to lnhlblt the aggregation of human platelets Induced by ADP in platelet-rich plasma (Dr. M. Johnson) and as dluretlc agents (Dr S.T. Kau) which Involved oral doslng to salt-loaded male Wlstar rats. No compound was sufflclently active to merit further study.

(Received In UK 26 July 1982)