Journal of Medicinal ChaisttnJ, 1971, Vol. 14, No. 6 493
POTENTIAL RENININHIBITORS
Lysophosphatidylethanolamine and 2-Desoxylysophosphatidylethanolamine Derivatives. 1. Potential Renin Inhibitors FRANCIS R. PFEIFFER,*SUZANNE C. HOKE,CLARAK. MIAO,RALPH E. TEDESCHI, JOSEPHINE PASTERNAK, RICHARD HAHN,ROBERTW. ERICKSON, HERMAN W. LEVIN, AND JERRY A. WEISBACH CHARLOTTE A. BURTON, Research and Development Division, Smith Kline & French Laboratom'es, Philadelphia, Pennsylvania
19101
Received October 50, 1970
A wide variety of derivatives of phospholipids related to natural phosphatidylethanolamines were prepared and evaluated as potential renin inhibitors in two in vitro assays. Lysophosphatidylethanolaminesand 2-desoxylysophosphatidylethanolamineswere also synthesized and evaluated. None of the new compounds were more active than a phospholipid mixture isolated from hog kidney but two compounds, 6g and 6j, which contain the 1adamantyl moiety, showed comparable activity in both assays. I n this series the lyso and 2-desoxy derivatives were generally active a t comparable concentrations regardless of the hydrophobic group. General procedures for the preparation of ethanolamine derivatives are also described.
A phospholipid isolated from dog,' hog,l rat,2 and human3n4was reported to be a natural precursor of a renin inhibitor. This inhibitor was said to repress the action of renin in vitro,' inhibit the in vitro response to injected renin in bilaterally nephrectomized rats,5 and reduce the blood pressure of rats with acute and chronic renal hypertension by 70-130 mm when dosed daily (from 4 to 12 days) by im (3.5 mg/kg), oral (150 mg per 250-350 g rat), or infusion (5 mg/kg) routes.6 The phospholipid had no effect on the blood pressure of normotensive rats. The "active" inhibitor mas derived from the phospholipid by phospholipase hydrolysis and mas provisionally identified as a lysophosphatidyl amino acid derivative.' A later report2 suggested that the preinhibitor, isolated from different species and different tissue, was a phosphatidylethanolamine with a unique fatty acid complement. Our interest in phospholipids7 and their involvement in enzymatic reactionss led us to investigate the potential for development of a phospholipid renin inhibitor as an effective mediator of the renin-angiotensinaldosterone system. We isolated a phospholipid from hog kidney, according to the published procedure,' which, as reported, inhibited the conversion of renin substrate to angiotensin I in tn-o in vitro assays. This substance was found to be a mixture of many molecular species of phosphatidylethanolamine esterified with a relatively high proportion of polyunsaturated fatty acids. Stearate, oleate, and linoleate were esterified predominately at the primary hydroxyl, and almost half of the ester at the secondary hydroxyl was arachidonate with lesser amounts of long-chain polyunsaturated fatty acids.g For this study, a series of synthetic phospholipids
(1) S. Sen, R. R. Smeby, and F. M. Bumpus, Biochemistry, 6, 1572 (1967). (2) D. H. Osmond, R . R . Smeby, and F. M . Bumpus, J . Lab. Clin.M e d . , 73, 795 (1969). (3) D. Ostrovsky, S. Sen, R. R. Smeby, and F. M. Bumpus, Circ. Res., 21, 497 (1967). (4) D. H. Osmond, L. A . Lewis, R . R. Smeby, and F. M . Bumpus, J . Lab. Clin.Med., 78, 809 (1969). (5) S. Sen, R. R. Smeby, and F. M . Bumpus, Amer. J . Physiol., 214, 337 (1968). (6) R. R. Smeby, S. Sen, a n d F. M . Bumpus, Circ. R e s . , 21, Suppl., 11, 11-129 (1967). (7) F. R . Pfeiffer, S. R. Cohen, and J. A . Weisbach. J . Org. Chem., 34, 2795 (1969). (8) F. R. Pfeiffer,C. K. Miao, and J. A . Weisbach, i b i d . , 36,221 (1970). (9) F. R . Pfeiffer, unpublished results.
and phospholipid-like derivatives were prepared and their renin-inhibitory activity evaluated in two in vitro test systems. It was anticipated that once structureactivity relationships had been investigated, inhibitory phospholipids could be prepared which would be stable to natural hydrolytic enzymes and thus be potentially useful for in vivo studies. Potential inhibitors have been prepared including the phosphatidylethanolamine derivatives 6, lysophosphatidylethanolamines 7, 2-desoxylysophosphatidylethanolamines11 and 13, and other related compounds. Chemistry.-Compounds 6 were prepared either by the condensation of the Ag salt of the phosphate 21° with the I derivatives 1 (Scheme I, method A), or by the phosphorylation of a 1,2-diglyceride 3 with the dichlorophosphate 4' (method B). Successive acylationsll of 3-iodopropane-1,2-diol with the appropriate acid chlorides gave compounds 1 (Table I). Compounds 3 were synthesized by a simplified method which is applicable to the preparation of optically active, unsaturated 1,2-diglyceride~~ (Scheme 11, Tables I1 and 111). The protected intermediates 5a and 5b were not isolated; treatment of 5a at 0" with dry HC11° or reaction of 5b with Zn in AcOH7 afforded the synthetic phosphatidylethanolamines 6 (Table IV) . Enzymatic hydrolysis of 6 with phospholipase A from snake venom12 gave the lysophosphatidylethanolamines 7 which were not purified, but were used directly in the assays. As activity resided in the lyso derivatives 7 (compounds 6 were generally inactive-see Biological Evaluation below) it was of interest to prepare the 2-desoxylysophosphatidylethanolamine derivatives, the esters 11 and ethers 13, which are related to 7 but are devoid of the 2-OH group. Previously, it was shown that other desoxyphospholipids were active in certain test systems. Desoxylysophosphatidyl choline derivatives have anticlotting activity13 and are cleaved by lipase and phospholipase preparations. l4 Desoxylysophosphatidyl glycerols exhibited some activity in the (10) F. J. M . Daeman, Chem. P h y s . Lipids, 1, 476 (1967). (11) G. H. de Haas and L. L. M. van Deenen, R e d . Trav. Chim. Pays-Bas, SO, 951 (1961). (12) C . Long and I . F. Penny, Biochem. J., 66, 382 (1957). (13) H. Kaneko and I. Hara, BULL Soc. Chim. B i d . , 46, 895 (1964); Chem. Abstr., 62,2951h (1965). (14) H. Eibl and 0. Westphal, Justus Liebigs Ann. Chem., 709, 244 (1967), and ref cited therein.
494
Joitrnal of Medicinal Chemistry, 1971, Vol. 14, No. 6
PFEIFFER, et al.
CHPOCOR
J
TABLE I 1,2-DI.lCYL-3-DESOXY-3-IODO-S?L-GLPCF,ROLS
CHzOCOR
I
IZ'COZCHCHJ xo, Ia
K'l R' Yield, % [a]25~ deg , (c, %Ib Formulac C&(CHi)is CH?=CH(CH? j 8 78d +5.1(1.04) Cs&dOa 11) CII,-CH(CH2 j8 CHa=CH(CHz j, 89 $4.7 (0.90) CZ&~IOI 1 (' 1-Adamantyl CHz=CH(CH,), 63 + 2 . 9 (0.89) Cz5H3dOa 1t l 1 -.\clsmaiityl CI-I3(CH,)7CH=CH(CH,)ij 7" f 0 . 8 (1.02) C3~Hdoa I cH ( C I I);CII=CI-I ~ c: r-r li 1-ildamaiityl 31 f3.7 (0.88) C3&3104 li' I-Adamaiityl 1-Adamant yl 67d f0.7j1.14) Cz&sI01 IC' CII,ICHL),CII=CfI~CI~z)i' CH3(CH2),CH=CH(CHijie 61 + 3 . 1 (0.78) CasHnIOa The I-Rcyl-::-de.-oxy-3-iodo prerursors of l a , IC, Id, and l e were prepd by reported procedures [see G. 11. de Haas and L. L. AI. ~ 8 1 1 lkeiieii, R e d . il'ruc. Chini. P a y s - h s , 80, 951 (1961j] : l-o~tadecanoyl-3-desoxy-3-iodo-sn-glycero1, mp 64-66' (from E t G A f e O H ) ; [ u ] ? ~ I +l.K2'' ) i c 1.0, CHC1,) [lit: nip 65-66"; [03% +1.S" ( e 10, CHC13)]; 1-(l-adamantoyl)-3-desoxy-3-iodo-sn-gly~erol, mp 1 .io" (c 1.14, CHCl;] [Anal. (CI3H1,I0B)C, H, I] : l-(cis-9-oct~adecenoyl)-3-desoxy-3exane-petr ether:: [ a ]2% [ a ] * %S1.G.i' (c 1.1, CIICl,) [lit: mp U . 4 " ; [ o ] * ~ D+1.9" ( e 10, CHC13). * Optical rotations weredetdinCHC13. ' -,ill compd5 \\-ere anal. for C, 13, I. d AI1 compds are oils except la, mp 26-2s" (from MeOHj, and If, mp 108-110" (fromEtzO-MeOH!. I)oiible bond is trans. 1 Double bond is cis. 'l
+
Journal of Medicinal Chemistry, 1971, Vol. 14, No. 6 495
POTENTIAL RENININHIBITORS
TABLE I11 1,8-DI ACYL971-QLY CEROLS
r-
RCOz H
bHzOH No.
R
Yield,
7'
MP, 'C
Recrystn solvent
[ a l z 6 D , deg
( e , %)"
Formulab
3a CHa(CHz)ie 87 Hexane 74-74.56 -2.6(1.0) c39H7605 3b 1-Adamantyl 90 Hexane 98.5-100 - 0 . 4 (1.1) C2&.05 3c 1-Adamantylmethyl 91 Oil -0.7 (0.9) CZ7H4005 3d DihydrophytyP 82 Oil -1.5(1.2) Ci13H8406 J. C. Sowden and H. 0. L. Fischer [ J .Amer. Chem. Soc., 63, a Optical rotations were detd in CHCla. I,All compds anal. for C, H. See footnote c of Table 11. 3244 (1941)] reported mp 74-74.5", [a]D -2.7" (c 6.18, CHC13).
TABLE IV
0-[ ~,~-DIACYL-S~-C~LYCERO-~-PHOSPHORYL] ETHANOLAMINES CHZOCOR
I
R'COzCH LH20P02CHzCH2~H2 AH No.
R
R'
Method'
Yield? %
Mp, OCc
204, CHa(CHn)ie B 48 CHFCH (CHz)s A 51 206-208 210-212 CHFCH (CHz)s A 38 197-199 CHz=CH(CHz)s A 30 CH~(CHZ)~CH=CH(CHZ)~' A 42 Wax 1-Adamantyl A 33 Waxk I-Adamantyl A 41 207-2 10 B 36 24 196-198 6h C H ~ ( C H ~ ) ~ C H = C H ( C H ZC)H~ I~( C H Z ) ~ C H = C H ( C H Z ) ~ ~A Dihydrophytyl B 25 Waxm 6i Dihydrophytyl' 1-Adamantylmethyl 207-208 1-Adamantylmethyl B 29 6j a The symbols used in this column are explained in the text. Yields are based on chromatogd products. c All compds were crystd Optical rotations detd in CHC13. or pptd from CHCls-MeZCO, except 6g (from CHCls-petr ether) and 6 j (2-butanone-MeCN), e All compds were anal. for C, H, N and were dried a t 40" (1 mm) for 18 hr. f Reported mp 173-175', [ c x ] ~f6.0" ~ D [E. Baer, J. Maurukas, and M. Russell, J . Amer. Chem. SOC.,74, 162 (1952)l. 0 Hemihydrate. h Dihydrate. Hydrate. 2 Sesquihydrate. Purified by thick-layer chromatography on Brinkmann silica gel G F 254 (1 mm) plates using 65: 25: 4 CHC13-MeOH-H20 as the eluent. * See footnote c of Table 11. Purified by column chromatogr on DEAE-cellulose according to the method of G. Rouser, G. Kritchevsky, D. Heller, and E. Lieber, J . Amer. Oil Chem. SOC.,40,426 (1963). Double bond is trans. 0 Double bond is cis. 6a 6b 6c 6d 6e 6f 6g
CHdCHz)ie CHa(CHz)is CHFCH (CHp)S 1-Adamantyl 1-Adamantyl CHa(CHz)7CH=CH(CHz)70 1-Adamantyl
SCHEMEI1
CHzOH Ho&
I
RCOCl
CHzOCOR
I
RCOnCH I
-3
Zn
AcOH
An excess of propane-1,3-diol was acylated with an acid chloride or treated with the appropriate mesylate in DMSO to give, after chromatographic purification, 10 and 12, respectively (Table V). Compounds 10 TABLE V
Wassermann assay.I5 Several l-acyldesoxylysolecithins caused lysis of red cells.I6 CHzOX &HZ
I
CHzOY 10, X = COR; Y = H 11, X = COR; Y = POzCHzCHzNHz I
OH 12, X = CHZR; Y = H 13, X = CHzR; Y = POaCHzCHzNH2 I
bH
hTONOBCYL-
(AND ALKYL)PROPANE-1,3-DIOLS ROCH2CHzCHzOH Yield,
No.
Ra CO(CHZ)&H~* CO (CHz)?CH=CH (CHz)7CHa' CO(1-adamanty1)c CHZ(CH~)I~CH~~ CHz (CH2)7CH=CH (CHz),CH3f
%
Formula'
10a 79 CZlH4203 10b 71 C4iH& 1Oc 74 CltHZZO3 129, 55 C21H4402 12b 52 C21H4202 1 2 ~ CHz(CHz)7(CH=CHCHz)n(CHz)aCHa' 64 CziHIoOz 12d CHzCHs(1-adamantyl) 56 C15H2602 All compds are oils except where indicated. Mp 49-51' Alp 45-47' (from EGO-MeOH). c M p 39-41' (petr ether). (EtzO-MeOH). e All compds were anal. for C, H. f Double bond is cis. (1
(HO)zPOzCH&HzNHCOzCHzCCl~
were phosphorylated with 4 and converted to 11 with Zn in AcOH. Phosphorylation of the ethers 12 with 4 or 14? in the presence of trichlor~acetonitrile~~ (meth(15) K.Inoue and S.Najima, Chem. P h y s . Lipids, 1,360 (1967). (16) F. C.Reman, R . A. Demel, J. deGier, L. L. M . van Deenen, H. Eibl, and 0. Westphal, ibid., 8,221 (1969).
(17) F. Cramer, W.Rittersdorf, and W. Bohm, Justus Liebigs Ann. Chem., 664, 180 (1962).
PFEIFFER, et a l .
496 Journal of Medicinal Chemistry, 2971, Vol. 14, .ITo. G
TABLE VI 0-[1-*4CTL (OR
-.\LKYL)
-%-DESOXYLYSOGLYCER0-3-PHOSPHORYL] I~'TH.\.NOLhMISI;S
CHaOR
1
CHa ~
CH20PO,CH&HaNHa
I
OH No.
lla llb llc
13it 13b
R
CO(CH?)MCH~ CO(CII,),CH=CIl!eH,)7CH," CO(1-adamantyl) CH2(CH?)isCI-I, CH,ICHa),CH=CH(CH,),CH3*
Methoda
Yieldh %
B B B
4'1 :; 4 37
13
40 30
c
c
Recrystn solyent
CHC13-MeOH CHC1a-MEKd CHC13-?VIeaC0 CHC13-?\lepC0 CHCl3-Me,CO
Mp,
oc
202-204 204-206 222-224 224-226 206-208
Formulac
Ci3H48NOsPe Ca3HdOeP Ci6H2sN06P C23HsoNOsPe C?3HasNOjPj
i3j
1Yc CHa(CHa),(Cr-I=Cr~CH,),(CIl~)3CIl3'~ I3 2s CHC1,j-AfEKd 193-196 C23H2JOjPe 13d CH,CH,( 1-adamantyl) B 40 CHCl3-MeCX 208-2 10 Cl,H3aN05P~ a The symbols used in this column are explained in the text. * Yields are based on chromatogd products. All compds were anal. for C, H, N and were dried at 40" (1 mm) for IS hr. N E K = methyl ethyl ketone. e Hemihydrate. I Anal. a3 0.23hydrate by Quarter hydrate by tga; HaO: calcd, 1.23; found, 1.6% Double thermal gravimetric analysis (tga); HgO: calcd, 0.99; found, 0.6. bonds are cis Q
od C) gave 13 after removal of the CC13CHaOC0group enzyme reactions. Renin substrate, in this case a (Table VI). tetradecapeptide,lg n as incubated with renin and the The attempted enzymatic preparation of 0-[l- potential inhibitor. The derived angiotensin I J W ~ (1-adamantoyl) - 3-sn- glycerophosphoryl]ethanolamin~ hydrolyzed in ziitro u i t h a preparation of converting (7b) from the 1,2-di(1-adaniantoyl) derivative 6g ~ a * enzyme to give the octapeptide, angiotensin 11, arid not successful; 6g is inert to the phospholipase A ill Hib-I,eu. The amount of His-Leu released >vas mensnake venom (Crotalus adamanteus). Compound 7b sured spectrophotofluorometrically and compared t (J was readily prepared enzymatically (Scheme 111) appropriate controls. The biological data may be summarized as follows SCHi.\lb, 111 (Table VII). Potential inhibitors were tested at dose.; of 1-5 X JI. Compounds 6g and Sj, which are esterified with 1-adamantoyl moieties, showed good activity in both assays. I n the bioassay 6c and 6d having shorter fatty acyl residues were slightly active. phospholipase A I n contrast nearly all of the lyso derivatives 7, and the 2-desoxylyso derivatives 11and 13exhibited some renininhibitory activity in both assays. These findings are consistent with literature which show that the lyhophosphatidylethanolamines7 are the inhibitor subhtances and the phosphatidylethanolamines 6 are inactive (exceptions noted above) and are precursors of the actual inhibitor 7. The inactivity of l l a in the bioa> could be attributed to insolubility or + incomplete micelle formation. Renin-inhibitory ;ICCHL=CH(CHg)&OOH tivity does n u t seem to depend on the degree of UILsztturation in the acyl groups, nor is an OH at position 2 riecessar?. The specific hydrophobic head of the from 0- [ 1-(1-adamantoyl-2-( 10-undecenoyl)-sn-glycinhibitorb does not seem to be as important as the grobh erophosphoryllethanolamine (6d). hydrophobicity. The ethanolamine moiety seems to be Biological Evaluation.-The abilit,y of t'hese coni;L requirement for inhibition as other phospholipid pounds to inhibit the renin-catalyzed conversion of types were .;hewn to be inactive in a bioassay.' renin substrate to angiotenr;in I \vas examined in 2 test systems. Renin activity ir-as det)ermiriedindirectly by measuring the angiotensin I hydrolysis products, Experimental Section angiotensin I1 and His-Leu. A bioassay and a bioThe compds were routinely checked by ir spectroscopy; meltchemical assay were used. The bioassay is a modificaiiig points were determined in a Thomas-Hoover melting point tion of a reported p r ~ c e d u r e "and ~ involves a two-step apparatus and are corrected; optical rotations were taken on a sequence: (I) i n vitro production of angiotensin I by Perkin-Elmer 141 polarimeter. Tlc was carried out on 0.25-mm the action of hog renin on hog renin subst'rate in the Anal Tech silica gel G F plates20 with the solvent systems: 9CE (9: 1, v/v, cyclohexane-F,t08c), 3CE (3: 1, cyclohexane-EtOAc), presence of a potential inhibitor, and ( 2 ) in vivo bioand CMW (65: 25:4, CHC13-NeOH-H20). The compds were assay of the generated angiotensin I in the anesthedetected by spraying with 40%) H2SOa and heating, with nintized rat1Pindirectly via the pressor response elicited by hydrin, and with a reagent for P.21 Where analyses are indicated the angiotensin I1 formed by the rat's converting enzyme. The biochemical assay also involves two (19) L. T. Skeggs, J r . , K . E. Lentz, J. R . Kahn, and H. Hochstrasser. (18) P. T. Pickens, F. hf. Bumpus, -1.h l . Lloyd, R. R . Smeby, and I . I f . Page. Circ. R e s . , 17,438(1965).
J . E x p . .Wed.. 128, 13 (1968). 120) Analtech, Inc., Wilmington, Del. (21) J . C . Dittmer and R . L. Lester, J . L i p i d Res., 6, 126 (1964).
Journal of Medicinal Chemistry, 1971, Vol. 14, N o . 6 497
POTENTIAL RENININHIBITORS
TABLE VI1 BIOLOQICAL EVALUATION OF POTENTIAL RENININHIBITOR^ -Bioassay-
Compd 6a
6b 6c
6d 6e 6f 6g
6h 6i 6j
78, R
-
(CHz)iscH~
7b, R = 1-adamantyl 70, R = (CHz)rCH=CH(CHz)?CHah 7d, R dihydrophytylC lla
-Biochemicalassay ConcenConcentration Inhibitration Inhibi( X 10-2 tion, ( X 10-3 tion, %* %* M) M) 16 2.67 5 3.08 26 3.65 33 3.69 17 3.02 18 3.02 39 68 1.86 3.80 38 0.93 5 2.69 5 2.49 68 1.76 53 0.88 69 4.15 61 5.39 66 4.18 4.06
45 2.50 1.25
7 3
2.2 1.1
29 15
30 64
llb 110 13a
4.33 5.55
13b 13d
4.36 4.39
49 56
6 (from hog kidney)d 7 (from hog kidney)e
2,671 4.160
5-10 65-95
2.77 1.38 1.33' 2.088
47 22 0 85-90
0 See Experimental Section for d e t d e d descriptions of the assays. * P e r cent change when compared to controls. See footnote c of Table 11. d Isolated from hog kidney acetone powder (Pe ntex Biochemicals Division, Miles Laboratories, Kankakee, Ill. ) by column chromatogr or thick-layer chromatography on silica gel or DEAE-cellulose. e Prepared from the phospholipid isolated from hog kidney by phospholipase A hydrolysis (see Experimental Section). f Assuming an average mol wt of about 750. 0 Assuming an average mol wt of about 470. h Double bond is cis.
only by symbols of t,he elements, anal. results obtained for these elements were within &0.4rj, of the theoretical values, and were obtained by the Analytical and Physical Chemistry Section of Smith Kline & French Laboratories. Pet,r ether used had bp 30-60". Rlallinckrodt silica gel (100 mesh) was used for column chromatogr. Nomenclature of the optically active glycerol derivatives follows the recommended rules.22 l-(l-Adamantoyl)-2-oleoyl-3-desoxy-3-iodo-sn-glycero~ (Id) (Table I).-A s o h of 7.0 g (0.035 mole) of 3-desoxy-3-iodo-snglycerolz3 in 25 ml of CHCl, (distd from P208) and 2.65 ml of anhyd pyridine was cooled to 0" and treated dropwise with 5.8 g (0.029 mole) of 1-adamantane carbonyl chloridez4 in 25 ml of anhyd CHC13. The mixt was stirred for 18 hr a t 25', dild with 200 ml of EtZO, and then washed with dil HCI, H20,5% NaHS03, HzO, 5% NaHC03, and HzO. After being dried (Na2S04), the ether was concd and the product, (9.5 g) was chromatogd oii 285 g of Florisil iising EtzO-petr ether mixts. A small amt of forerun contaiued 1,2-di(l-adamantoyl)-3-desoxy-3-iodo-sn-glycerol (If), Rr 0.82 (system 9CE). The major product was eluted with Et20 and the homogeneous cuts were crystd from cyclohexane-petr ether to give 4.8 g (46%) of l-(l-adamaiitoy1)-3desoxy-3-iodo-sn-glycerol1 mp 63-65', Rr 0.25 (9CE). Anal. (Ci4H21103) C, H, I. l-(l-Adamantoyl)-3-desoxy-3-iodo-sn-glycerol(14.05 g, 0.039 mole) was dissolved in anhyd CHC13 (30 ml) and anhyd pyridine (2.4 ml) and cis-9-octadecenoyl chloride% (8.7 g, 0.29 mole) in anhyd CHCll (30 ml) was added dropwise with stirring at, 0" (22) See rules of nomenclature of lipids proposed by the IUPAC-IUR Commission on Biochemical Nomenclature [Biochemistry, 6 , 3287 (1967) 1. (23) E. 13aer and H. 0. L. Fischer, J . Amer. Chem. Soc., 7 0 , 609 (1948). Hydrolysis of the isopropylidene precursor was achieved with 1 N HC1 in MeOH at 25' for 18 hr. (24) H . Stetter and E. Rauscher, Chem. Ber., 93, 1161 (1960). (25) The acid chlorides were prepared with oxalyl chloride [F. H. Mattson and R . A . Volpenhein, J . Lipid R e s . , 3 , 281 (1962)l.
under NO. After stirring for 18 hr a t room temp the reaction was worked-up as described above and the crude product. was chromatogd on Florisil (30 g/g of crude product). The column was developed with petr ether-Et20 mixts to give oily Id, Rf 0.65 (9CE). 1,2-Di(l-adamantoyl)-3-desoxy-3-iodo-sn-glycerol (If).-A soln of 5.72 g (0.0283 mole) of 3-desoxy-3-iodo-sn-glycerol, 30 ml of anhyd CHC13, and 7.5 ml of anhyd pyridine was treated dropwise a t 0' with a soln of 12.2 g (0.0614 mole) of l-adamantane carbonyl chloride, stirred for 18 hr at 25", and worked-up as described in the previous example. The crude pink oil crystd on standing and was recrystd to give 9.98 g of If. 1,2-Di(l-adamantoyl)-sn-glycerol3-(ptp,p-Trichloroethyl carbonate) (9b) (Table II).-1-Adamaiitanecarbonyl chloride (28.2 g, 0.142 mole) in EtOH-free, anhyd CHCh (100 ml) was added dropwise to an ice-cold solii of sn-glycerol 3-(P,&p-trichloroethyl carbonate)s (20.0 g, 0.071 mole), anhyd pyridine (14 ml), and anhyd CHC13 (50 ml). After stirring overnight a t room temp, 800 ml of Et20 was added and the mixt was washed with dil HCI, HzO, 5% NaHC03, and H20. The dried (Na~S04) ethereal solution was concd to give a colorless oil which was cryst'd to afford 34.1 g of 9b, Rr 0.76 (3CE). 1,2-Di(l-adamantoyl)-sn-glycerol(3b) (Table HI).-Compd 9b (23.2 g, 0.039 mole) was dissolved in 150 ml of 857, AcOH and 200 ml of Et20 and 48 g of activated ZnZ6was added (slight exotherm). The suspension was stirred a t 20-25' for 1 hr and filtered and the filter cake was washed with addnl EttO. The filtrate was washed with HoO (3 X 100 ml), 57, NaHC03 until neutral, and HzO again. After drying (Na2Y04)the Et20 was evapd and the oily residue was crystd to yield 14.6 g (goyo) of 3b, mp 96-98', Rf0.31 (3CE). 0- [1,2-Di(l -adamantoyl)-sn-glycero-3-phosphoryl] ethanolamine (6g). Method A (Table IV).-A solti of 1.4 g (0.0028 mole) of I f in 200 ml of aiihyd C6H6 was concd to 100 ml using a Dean-Stark adapter. The soln was cooled slightly and 1.24 g (0.003 mole) of silver tert-bntyl(N-terl-butyloxycarbonyl-2aminoethy1)phosphate'O (previously dried a t 40" in uacuo for several hours) was added. The reaction flask was protected from light, and the mixture was refluxed for 2.5 hr. The warm sohi was filtered through Supercel to remove AgI, and the filtrate was washed with 57h NaHC03 and 1 1 2 0 . The dried (hIgSO4) organic layer was coiicd to give the colorless, oily 5a (11 = I:' = 1-adamantoyl). This wits dissolved in dry Et20 (200 ml) and treated with a stream of dry HC1 a t 0" for 2 hr. After evapn of the solvent at 30" t,he residue was dissolved iii 30 ml of 4: 2: 1, v/v, Etz0-EtOH-H20 and percolated throiigh a column packed with 30 g of Amberlite I R 45 (OH-) resin rwiiig about 300 ml of addnl solvent. The crude 6g was obtained on evapri of the solvent at 30"; it was chromatogd on 250 g of a mixture of 2: 1, w/w, silica gel and Supercel. The column was developed with 10-507, bIeOH in CHCl3. Homogeneous fractions with R r 0.66 (ChIW) were crystd. 0-[I,2-Di(l -adamantylacetyl )-sn-glycero-3-phosphoryl]ethanolamine (6j). Method B (Table IV).-1,2-l)i(l-adamaiitylacetyl)-sn-glycerol (3c) (12.8 g, 0.029 mole, freshly prepd from the carbonate 9c arid used without purification) was dissolved in 50 ml of anhyd CHCI3 and 11.5 ml of anhyd pyridine and added dropwise to an ice-cold solii of 15.25 g (0.043 mole) of dichloro(1\'-p,p,p-trichloroethoxycarbo1iyl-2-amiiioethyl)phosphate(4)' in 50 nil of anhyd CHCl3. The s o h was stirred at room temp for 18 hr, dild with 600 nil of EttO, and wmhed with cold dil HCl and brine. The org layer was concd to give crride 5b (It = 13' = 1-adamantylmethyleiie). This was dissolved in 25 ml of 907; AcOH arid 50 ml of Et20 and treated with 30 g of activated Zn. The suspension was stirred a t 20-25" for 3-4 hr or until tlc (system CMW) indicated complete conversioii to the phosphatidylethanolamine Sj. To the suspension was added solid NaHC03 (60 g), H20 (6 ml), and CHC13 (100 ml). After stirring for 20-30 min, CHCI, (1 1.) and excess NanSOa were added and the stirring was continued for another hour. The solvents were filtered and concd, and the residue was chromatogd on 2 kg of 2 : l silica gel-Supercel with 57, RIeOH in CHC13 as the solvent. From the column were obtained a minor product, Rt 0.85 (ChlW), formed from excess 4, mid the major product 6j, Rf 0.69 (CRIW), which was crystd. l-cis-9-Octadecenoylpropane-1,3-diol (lob) (Table V).-A soln of 5.05 g (0.066 mole) of propane-l,3-diol (dried over Linde 411 molecular sieves), 20 ml of anhyd CHC13, and 1.6 ml of anhyd (26) E. Baer and D. Ruchnea, J . Bioi. Chem., 230, 447 (1958).
498 Journal of Medicinal Chemistry, 1971, Val. 14, No. 6
PFEIFFER, et al.
pyridine was cooled to 0" and a soln of cis-9-octadecenoyl chloride (5.0 g, 0.017 mole) in CHC18 (25 ml) was added dropwise under N2. The soin was stirred for 18 hr at 25O, dild with Et20 (150 ml), and washed with dil IIC1, HzO, NaHCOa, and H20. Chromatog of the dried EtsO concii on Florisil (175 g) with petr ether-EtzO mixtures gave4.14 g of oily lob, Rf 0.35 (3CE). 3-[2-(l-Adamantyl)ethoxy] -1-propanol (12d) (Table V).--.i soln of 12.6 g (0.07 mole) of 2-(l-adamantyl)ethanolz7in 70 ml of anhyd pyridine was cooled i o 0" and treated dropwise with 12.0 g (0.105 mole) of NeSOzC1. The mixt was stirred a t 0' for 2 hr and a t 25" for 4 hr, dild with 500 in1 of EbzO, and washed rapidly with cold H20, iced dil HC1, H20, cold 5% NaHC03, and H20. The dried (Na2S04) ethereal s o h was coiicd to 18 g (100%) of crude mesylate, RI 0.44 (3CE), which was w e d without further purification. NaH (11.39 g of a 6OCC mineral oil dispersion, 0.286 mole) in anhyd DMSO (200 ml) wai. heated at 65-70' for 1 hr under NZand cooled t o 25" and a solii of propane-1,3-diol (21.7 g, 0.286 mole) in DMSO (25 mi) was added rapidly. The mixt v a s theii stirred a t 25" for 0.5 hr and n, soln of 12.3 g (0.048 mole of 241adamantyljethanol-1-mesylatein 25 ml of IIlISO was added at :t rapid rate. The mixt, was stirred an addnl 2 hr. Ice H20 (1.5 1.) was used T O quench the reaction and the product wad estd into EtOAc, washed with brine, dried (KazSO4j, and coned to ail orange oil. This was chromatographed o n 540 g of Florisil, first with petr ether-EtzO mixtures and then with E t 2 0 to give 6.37 gof colorless, oily 12d, Rf 0.23 (3CE). 0- [3-(2-(l-Adamanty1)ethoxy)-l-propylphosphor~-l] ethanolprocediire for this preparatioii amine (13d) (Table VI).-The m s entirely similar to that for 6j (method B). From 3.55 g (0.015 mole) of 4 arid 4 ml (Jf pyridine i i i CHCI,, there wai obtained one major P-posit,ive component, XI 0.58 ( C N W ) ; tbis was treated with 18 g of activated Zri iii 40 mi of Et20 and 20 nil of goc'; AcOH. The usiial work-up gave the criide product ~ h i c l i was chromatogd 0 1 1 SO0 g of 2 : 1 silica gel-Supercel Tvith 4 : 1. v:'v, CHCI~-lIeOK (4 l.), ::: 1 CHC13-31eOII (6 1.~1, nlld 2: 1 CHCl3-J1eOII (6 1. ). T h e homogeneous fractioiis from the 3: 1 and 2 : I eluates were cr>-std to give white 13d, Ii't 0.33 (ChIW ). 0- [3-(cis-9-Octadecenyloxy )-1-propylphosphoryl] ethanolamine (13b). Method C (Table VI).-A .solti of 10.0 g (0.031 mole) of 3-(cis-!)-octadeceiiyioxy)-l-~~r~~p;~iioi (12b), 10.5 of the dicyclohexylammoniiim salt of ~ ~ - ~ , ~ , ~ - t r i c h l o r o e t l i o s ~ - c a r h o i i ~ 1 - 2 amiiioethylphosphoric acid (14),' 100 nil of anhyd pyridine, anti 43 g of CC13CS17 was heated on :t steam bath for 4 hr urider :I positive head of Sz. The dark reactioii mist was cooled, dild w t h 400 ml of H20, and waihed with E t 2 0 16 X 100 nil). The aq layer was concd in zucuo t o a t)rowii ,syrup which w;i> dissolved i t i 2: 1 , v'v, THF-HzO (100 nil) and +tirred with 30-40 g of Amberlite IR 120 (1-1-) for 13 miti. After the resiii wa.: filtered : ~ n d1 he filtrate WBS coricd in zyicuo, the residue was dissolvcd in 100 iril of Et20 and 50 nil of 9Or; i\cOH :iiid treated with 20 g of Z n and ivorked-up as in method B. The (.rude 13b wa.: a guni (S g) which wa3 chroniatogd oil SO0 g of 2 : 1 silica gel-Siipercel using !):2, v v, CHCl3-1\IeOH. The homogenon. ciit. were erystd t o give buff-c>olored13b, IZi 0.56 iClIJV), Bioassay Procedure.-Iii 1 nil of et hyleiie glycol i v t i s dirwlvecl 6 mg of the potential inhihilor, theii 1 nil of nti aq s o h eoiitaiiiitig 12 iimoles of hog reriin Yubstrate28 was added, followed by 0.5 ml of (I.$):; saline s o h coiitaiiiiiig 1 i i i i i t of hog r e ~ i i i i . ~()tie ~ 1 0 t.TO drllpa CJf it 5 s o h of clii,~opropy!phoaphorofliiorid:~~ e (DFP) in i-PrOH was iiroduced and the mist wnh ndjusied t i , pH F with about a drop of 1 .l/ XaH&'04 ~olii. The total vol was atmiit 2 . 5 nil giviiig an iiihitiitor ( ~ ~ i i c of i i 2 [rig nil. Thimist was incubated a t 37' fOl. 1 hr! theii placed iii a boiliiig f I I 0 hath for 10 niiii, chilled iti :I solid C'O1-lIesCO bath, and finall>lyophilized. The residiie w:is recoiistiiiited by the addri of 1 1111 of IIaO aiid the aiigiotensiti I that w a , ~formed waq measured hj- its pressor effect o n the aiiehthetized rat prepn. This w a y iimilar t o the procedure de,-cribed hy Pickens, et d . 1 8 .%dull, male, Charles River rat?, weighing 160-210 g, were uied. lIeaii arterial blood pre-yiire was ineii-wed from the carotid artery 11)meniii of a Sailhorn trarisdiicer (\rode1 267 AC) arid recorded uii :I niiiltichatitiel Saii1)orti recorder ilIodel 964). llriig soliis were giveii by nie:iii- of :icaiinulated jugnlar or femoral vein. The prepii wai. :illowed t o eyiiilitiratc for 1 hr hefore proceeding with
55
..
. .
. ~
(Pi) H . Stetter and E . Rauscller, Chem. R e r . , 93,2034 (1960). (28) Pentex 13iocliemicals Dirision, Miles Laboratories, liankakee, Ill. 129) Sutritional niocliemicals Corporation, Clel-eland. Ohio.
430) Cibs Corporation, Summit, S . .J. ,311 Scliwar;. '1Iann. Orangeburg, ? 1. i. d 2 ) Pipwe Chemical Cuinpanv. Rockford. 111.
FOLICACIDANTAGONISTS.4
Journal of Medicinal Chemistry, 1971, Vol. 14, No. 6 499
Griffin, Robert G. Pendleton, James F. Kerwin, Allen Misher, and Harry Green for helpful discussions. We are grateful to Miss Roberta Carraine for technical
assistance with the bioassay, Miss Margaret Carroll and staff for elemental analyses, and Mr. Walter Hamill and staff for optical rotations.
Potential Folic Acid Antagonists. 4. Synthesis and Dihydrofolate Reductase Inhibitory Activities of 2,4,6-Triamino-5-arylazopyrimidines S. S. CHATTERJEE, D. R. GARRISON, R . KAPROVE, J. F. MORAN, A. M. TRIGGLE, D. J. TRIGGLE,*AND A. WAYNE Department of Biochemistry, Schools of Medicine, Dentistry, and Pharmacy, Department of Biochemical Pharmacology, School of Pharmacy, and The Center for Theoretical Biology, State University of New York at Buffalo, Buffalo, iVew York 14214 Received January 14, 1971 A series of 49 2,4,6-triamino-5-arylazopyrimidines have been synthesized and examined for their inhibitory activity toward chicken liver dihydrofolate reductase. This activity was comparatively independent of the substituent character of the &aryl group; however, para substituents (almost without exception) reduced activity, meta substituents maintained activity, and small o-alkyl substituents increased activity. The most active compound in the series was found to be 2,4,6-triamino-.i-o-ethylphenylazopyrirnidine.
Previous studies in this l a b ~ r a t o r y l - ~and elsewhere4-' have revealed that 5-arylazopyrimidines exhibit significant inhibitory activity towards folatedihydrofolate reductase. I n continuation of our studies of the structural requirements for this activity we report the synthesis and inhibitory activities of a series of 2,4,6-triamino-5-arylazopyrimidines.
Experimental Section8 Synthetic Procedure.-2,4,6-Triamino-5-arylazopyrimidines (Table I ) were prepd by the general method described by Timmis and his coworkers.' The diazotized amine (0.1 mole) free from H K 0 2 was added to a vigorously stirred soln of 2,4,6-triaminopyrimidine (0.1 mole) in H20 (350 ml) a t 0' with sufficient SaOAc to maintain the pH at 6-7. When the addn was complete, the mixt was stirred at 0-5" for 10 hr and then a t 10' for 2-3 days. The product was collected and washed well with H20 and airdried: crude yields ranged from 70 to 100%. The compds were recrystd from i-PrOH which caused considerable losses but had the advantage of giving analytically pure specimens with the minimum of manipulation. Enzyme Procedure.-Dihydrofolate reductase was partially purified from chicken liver according to Kaufman and Gardiner9 their procedure was followed to step I\', yielding a prepn with an increase in specific activity of 125- to 150-fold over that of the starting supernatant. The standard assay mixt contained 5 X 10-6 X T P N H and 6 X 10-5 N dihydrofolatelo in 50 m.11 phosphate buffer at pH
*
T o whom correspondence should be addressed a t the Department of Biochemical Pharmacology, School of Pharmacy, State University of Kew York a t Buffalo, Buffalo, N. Y. 14214. (1) J. Hampshire, P . Hebborn, A . 31. Triggle, and D. J. Triggle, J . M e d . Chem., 8 , 745 (196.5); ibid., 11, 583 (1968). (2) hl. Chadwick, J. Hampshire, P. Hebborn, .i. 11. Triggle, and D. J. Triggle, i h i d . , 9 , 874 (1966). (3) J. Hampshire, P. Hebborn, A . M.Triggle, and D. J. Triggle, J . Pharm. Sci., 66, 453 (1966). (4) G. M. Timmis, D . G . I. Felton, H . 0. J. Collier, a n d P . J. Huskinson. J . Pharm. Pharmacol., 9, 46 (1957). ( 5 ) E. J. Modest. H . K.Schlein, a n d G. E. Foley, ihid., 9, 68 (1957). (6) (a) K . Tanaka, K . Kaaiwara, Y. Aramaki, E . Omura, T . hraki, J. Watanabe, &Kamashina, I. T. Sugawa, Y . Sanno, Y. Sugino, Y. Ando, and K. h a c , Gann., 47, 401 (1956); (b) M. Kamashina, Chem. Pharm. Bull., 7 , 17 (1959)- and immediately preceding papers. ( 7 ) P. Roy-Burman and D. Sen, Biochem. Pharmacol., 13, 1437 (1964). ( 8 ) Melting points were recorded on a Thomas-Kofler hot stage and are corrected. -2nalyaes were performed by Dr. A . E. Bernhardt and, where indicated only by symbols of the elements, were within 0.4% of the theoretical values. (9) B. T. Kaufman and R. Gardiner, J . B i d . Chem., 241, 1319 (1966). (10) S. J. Futterman, i h i d . , 118, 1031 (1957).
7.9. The velocity was detd spectrophotometrically a t 340 nM using a cell of 1-cm path length. Inhibitors were added as aq solns if sol, or if insol in H20, in DMSO to give a final concn of 3.370 DMSO. When DMSO was employed the controls contained an equal amt of the solvent. Experiments with 10 watersol compds selected at random showed that this concn of DMSO had little effect on the extent of inhibition produced. Initial velocities were detd over the first 2 min of the reaction in the presence of a t least 6 inhibitor concns which were chosen to give a range of inhibition from 20 to 80%. From these data the [I]/ [ S ] ~ . Svalues listed in Table I were determined. The [I]/[S] ratios for 50% inhibition provide a convenient numerical manner for denoting changes in the affinity of an antagonist' with structuresincewhen [SI> 5Km, [I]/[S]o.fi = K,/Km.
Results and Discussion
It is quite apparent from the data of Table I that many of the substituents have but modest effects on the inhibitory activity of the parent compound, 2,4,6triamino-5-arylazopyrimidine (1). I n general, the introduction of para substituents (4,7,12,15,17,21,23, 26,28,31,34) reduces activity relative to the parent compound: this effect is most marked with hydrocarbon substituents (4,7,9,12,15,17)although there is no obvious correlation with substituent size. I n contrast, introduction of the more polar p-C1, p-Br, p-I, p-IIeO, and p-COzEt substituents produces a smaller reduction in activity; 2,4,6-triaminod-p-iodophenylazopyrirnidine (26) appears to be the sole example in which a para substituent increases activity above that of the parent unsubstituted compound (1). Introduction of meta substituents (3,6,14,18,20,22, 25,27,30,33,35) has only modest effects on the inhibitory activity. I n general, activity is maintained or increased slightly and this is most marked with m-I (25) which shows a 4-fold increase. These effects of meta substituents appear to be relatively independent of substituent character. The effect of nonpolar ortho substituents reaches a maximum a t o-Et ( 5 ) which shows a 5.5-fold increase in activity: further increases in size of the substituent does not produce any further increase in activity. Other substituents (Cl, CFB, M e O 4 , CF3) do not increase activity and it is interesting that the iodo sub-
