March 1965
HYPOCHOLESTEREJ~IC BASICCARBIKOLS
peutic ratios are compared to those of pentobarbital or phenobarbital. Preliminary structure-activity correlations led us to make the following general statements. (1) The nature of the S-substituent is critical. One-carbon interruption between the aryl moiety and the nitrogen atom as well as alkyl branching of the a-carbon are prerequisites for sustaining hypnotic activity. Lengthening of the side chain to include two carbon atonis, with or without branching ( i e . , 46 and 47), or omission of branching, ci la 45, or direct attachment of the aryl group upon the nitrogen (44),leads to total loss of hypnotic properties. ( 2 ) Differences in hypnotic potency among the various esters are relatively minor. The presence of the ester moiety per se is essential;
223
the corresponding carboxylic acids are totally inactive. A detailed pharniacological study, including test results obtained in other animals, will be published elsewhere. Acknowledgment.--We are indebted to Messrs. T. Van Offenwert and A. Knaeps for the preparation of a number of starting materials and to lfessrs F. Sels and W.Verkest for analytical determinations reported herein. Pharmacological assistance was rendered by A h . F. Leenaerts. We also extend our thanks to Dr. C. van de Westeringh for his participation in fruitful chemical discussions during the course of this work. Financial support of the “Instituut tot Aanmoedinging van het Wetenschappelijk Onderzoek in Nijverheid en Landbouw” is gratefully acknowledged.
Hypocholesteremic Agents. 1II.l Basic Carbinols and Related Compounds JAMES H. SHORT,U R S U L - 4 BIERMACHER, D a N I E L A. DUNNIGAN, G. FREDERICK LAMBERT, DONALD L. J I A R T I N , C A R L w.S O R D E E N , A T D HOX-ARD B. b
7
~
Organic Chemistry Department and Pharmacology Department, Research Division, d bbott Laboratories, S o r t h Chicago, Illinois Received August 21, 1964
A series of 58 basic carbinols and related compounds has been synthesized, mostly by means of the Grignard reaction, and examined for hypocholesteremic activity. One compound, a-[P(2-diethylaminoethoxy)phenyl]cu-phenyl-5-acenaphthenemethano1, proved t o be considerably more potent than triparanol in both rats and
mice.
Much effort in recent years has been expended in the search for chemical agents which will significantly lower the blood cholesterol level of hypercholesteremic individuals. The rationale behind use of such drugs is the belief, not conclusively proven, that formation of atherosclerotic plaques is directly connected to the amount of cholesterol in the b l ~ o d . Although ~~~ a number of drugs are known to possess hypocholesteremic activity, none are entirely satisfactory. I n 1959, triparanol was introduced for this purpose, and both animal and clinical studies indicate it to be both effective and consistent in its a~tivity.5,~It was soon discovered, however, that, as the level of cholesterol is reduced, the level of its biogenetic precursor, desmosterol, is increased and total sterol concentration of the plasma is not reduced as much as determinations of cholesterol would seem to indicatee23 , 6 A further possible disadvantage of triparanol is its lack of potency. A typical dose for human patients is 250 mg. dailye6 For drugs which are given over long periods of time, it might be advantageous to be able to give one effective a t a lower dose. This work, then, was undertaken for two reasons; first, to find a drug similar to triparanol effective at a dose of no more than 50 ing.,’ ( 1 ) Paper I1 RI. Freifelder and H. B. JFright. J . M e d . Chem., 1, 664 (1964). ( 2 ) “New and Nonofficial Drugs, 1962,” J. B. Lippincott Co., Philadelphia, Pa., 1962, p. 616. (3) R. H. Furman and C . a. Robinson, Jr., M e d . CZzn. N . Am., 46, 935
(1961).
(4) C. Moses, Angiology, 13,59 (1962).
(5)
37
W. Hollander and A. Chobanian, B M Q , Boston X e d . Quart., 10,
(1959).
(6) (a) M. Friedman, S. 0. Byers, and R. H. Rosenman. Progr. Cardzoiascular Dzseases, 4, 419 (1962); (b) W. Hollander, A . V. Chobanian, and R. \+. \%ilkins, J . A m . M e d . Assoc., 114, 5 (1960).
day, and, second, one which would give a better reduction in total sterols. Chemically, triparanol (I) is a derivative of 1,1,2triphenylethanol. As such, there are many possible modifications which might lead to interesting struc-
CH, I
ture-activity relationships. In a previous publication,’ we established that the 4-(2-(diethylaminoethoxy)phenyl group may be replaced by a pyridine ring and activity maintained. The most potent compound of that series is l,l-diphenyl-2-(4-pyridyl)ethanol. Its potency is about the same as that of I. Investigation of pyridine derivatives is continued in this paper, and modifications of the diethylaniinoethoxy side chain have been studied extensively. Triarylmethanol homologs have also been investigated, as has replacement of benzene rings with polynuclear ring systems. I n addition, a group of tetrahydrofuran derivatives and some ethylene derivatives has been synthesized. Chemistry.-Four general methods were used to obtain the carbinols (IV, VI, IX, and XI), ethers, and ethylene derivatives described in Table I. ( i )H. B. Wright, D. A. Dunnigan, and 1 , 113
(1964).
C. Biermacher, J . .lied. Chem.,
~
~
:r z:
Y
:Ii
-
c
March 1965
HYPOCHOLESTEREMIC BASICCARBINOLS
n
-
n h
m
s
h
2 7.
v
w
*
m
5;
O
N 3
O
R
"
225
Compd.
lt2
H
I >
-
li
!I 11)
11 H 4-CH, 3-CH3 2,5-(CH, ) L 3,5-(CH, !i FE H
11 I2
IT
:3
1 )
Ci
I
il
1:i 14 1.i I (i
17 1s 1!1
20 21 a 'I'he hydroxy ketoues required fur the preparatioii of the ketoiies iii this iatile were 11l)taiiidfr(Jii1 ~ ~ ~ i i i i i i e rsiriirces ~ ~ i a l (A1dric.h desc~rihedi n (lie literatiire, or :ire tlesc.1 Cheriiioal Co., Inc., The Dow Chemical Co., or Tennessee Eastiiiaii Co. )! --ere prepared in the Experimental section. This compound has been reported kiy G . 1 3 I'aco :tiid C!, The required hydroxy ketone was prepared by the Fries reacstioii as described by It, f ' h e ) i ? . Soc.., 77, 2522 (1955). The required hj-droxy ketone \\-as prepared by L Buiidesmann, and F. Wieners, Ann., 447, 162 (19261. The required hydrrisy ke by F. Benington, R. D orin, L. C. Clark, Jr., and R.P. Fox, J . Org. Chciil., 23, ethoxy)-l-acet'oriaphth . The required hydrosy ketone was prepared by the Cheiii., 135, 49 (1932). The required hydroxy ketone was prepared by refluxirig ~ - h ~ t ~ r c ~ s g . b ~ i i z o ~arid ) ~ i emilfury1 l i ~ ~ i i echloride in benzene for 4 hr. t o give a product melting at, 178-180'. 11.Neiicki and E. Stoeber [[A,,., 30, 1772 (189711 reported 1n.p. 176". I'rcparation of this coinpound is described in U. S.Patent 2,914,562 (R.E.Allen, F. 1'. Pdopoli, F;. L. Schumanri, arid 11. G. \'nri Cariipeii. %Jr.:to the JJ7rri,S. AIerrell Co., Nov. 24, 1959 i . Q
iiig ethylene derivative.
R' I\'
A l ttempts to prepart' L- I I I ether were uniformly unsuccessful, but lr was f o r n i d snioothly and iii good yield wheii tetrahydrofuran was used as tlici solvent. Preparation of t hv (;rignard reagelit in this manner was discoverc,d indepcndcii t 11, b y Lrdniccr arid co-workers.' . h o t lier advantag(, of tetrahydrofuran a b a solvent for Crigiiaid reactioris 1s the increased solubility of the Grignard coiiiplesc- i i i bilk solveiit as conipared with that of cthcr.
3 fethod X involves the reaction of p-chlorobenzylriiagnesiuin chloride (11)with a ketone (HI), generally a derivative of acetophenone or benzophenone, in whicli the dialkylaminoallioxy group is part of the lietoiic nioiety. The Grignard reagent used in iiic>thod €3 is 4-
I.
1
OM
IS
T h reaction of various Grignard reagents (T-11) with -1-(2-diethyla1ninoethoxy)berizophenoiie [ H I , R ' = phenyl; R 2 = I-OCH2CH2S(CzHe)s]coinprises method C. llethod D invol~csthe reaction of t h
eR'
OH
i H 5 C 2 p C H 2 C H 2 0 ~ + R' VI
(.'-diethylaminoethoxy)phe~iyli~iagnesiuiiibromide (V). The ketone I11 did not necessarily carry a basic suhs t h e n t . In some cases, the carbinols spoiita~ieously lost water during the work-up yielding the correspond-
K'
X
XI
HYPOCHOLESTEREJIIC BASICCARBIXOLS
March 1965
sodio derivative of a methylpyridine (X), or a niethylpyrazine, with 4-inethylbenzophenorie (111, R1 = C6H5; R2 = A-CH,), to give the corresponding carbinol XI. This method was not restricted to 2- and 4picolines, but was also successful using 3-picoline. Formation of sodium salts of 2- and 4-picolines has been long known, but it was not until 1951 that Brown and Illurpheyg observed such a reaction with 3-picoline. We have, however. found, as would be expected, that the latter reacts less readily with sodamide than does its isomers. Both 2- and 4-picoline react with sodainide to form the respective sodium salts at room temperature with evolution of heat. The formation of the sodium salt of 3-picoline does not occur at room temperature, but it is necessary to heat the reactants on the steam bath for 1 hr. This difference in reactivity allow one to form a monosodium salt from 3,4-lutidine.10 Several 7-chlorobutyrophenones were allowed to react with 4-(2-diethylaniinoethoxy)phenylmagnesiuni bromide. The products isolated were not the ychlorobutylcarbinols. The carbinols spontaneously lost HC1, and the substances actually isolated are pre-
1
TABLEI11 RESULTS OF ROUTINE SCREENIKG FOR HTPOCHOLESTEREMIC ACTIVITYIN NICE C~mpd.~
2 3
4 6 7
8 10 12
13 14 21 22 23
OH I--\
24 25
OCHjCHjN(C2H , ) 2
XI1
suined to be 1,l-disubstituted tetrahydrofurans (XII). The various ketones required for preparation of the carbinols and other compounds in Table I, which have riot been described previously in the literature, are described in Table 11. I
