LOUISF. FIESERAND ARTHURP. RICHARDSON
3156
Vol. 70
(a) THE CHEMICAL LABORATORY OF HARVARD UNIVERSITY AND (b) THE MEDICAL SCHOOLOF THE UNIVERSITY OF TENNESSEE AND THE SQUIBB INSTITUTE FOR MEDICAL RESEARCH]
[CONTRIBUTION FROM
Naphthoquinone Antimalarials. 11. Correlation of Structure and Activity Against P. Zophurae in Ducks1 BY LOUISF. FIESER~ AND ARTHURP.RICHARDS ON^
The assays for suppressive activity against P. lophurae in ducks discussed in this paper were conducted under the direction of one of us, Richardson, and are reported in the CMR monograph2 (F-1 procedure). The other author is responsible for the analysis of the assay data by a scheme that departs from the practice generally followed by pharmacologists in the field. Evaluation of Drug Activity.-The accepted method of estimating drug activity in terms of the quinine equivalent (Q) as described by Blanchard and Marshall2consists in either matching the dose of quinine and of the drug under assay which cause the first sharp drop in parasitemia relative to that of the untreated controls, or else matching the doses which cause a drop to roughly analogous levels. That this scheme may lead to considerable variability in the interpretation of the assays can be seen from an examination of the illustrative example cited by Blanchard and Marshall and reproduced in Fig. 1 in the form of two dose-response curves. The doses described as producing the first “definite decrease” in parasitemia for quinine bisulfate and for Drug No. 760 are 30 and 75 mg., indicated on the curves as points a and a’, whence Q = 30/75 X 59.12Yo3= 0.24. But the choice of dosage is arbitrary, and equally “sharp Q=0.23 h
C
-30
t‘
ranges of 30 to 40 mg., and of 50 to 90 mg., respectively Thus the assays might have indicated quinine equivalents varying, in the extreme case, from 0.20 to 0.47. The other pair of matched doses, b and b’, fortuitously produce responses that are more nearly analogous, but they still are not exactly the same. The variability and uncertainty can be eliminated by the simple expedient of reading’from dose response curves the dosages corresponding to any arbitrarily selected level of reduction in parasite count; for example points c and c’ a t the 95% level. Such E D (Effective Dose) values are objectively derived from an unambiguous definition of terms and they also have the advantage of increased reliability resulting from the graphical averaging of assays a t several dosage levels. In December, 1944, the accumulated data from 46 standardization assays of quinine bisulfate were analyzed by this method, and the average E D values (for quinine bisulfate = Q) and average deviations found were as follows EDosQ = 17.35 * 2.86 (16.5 %) mg./kg./dose = 11.38 EDT~Q
The values calculated for the 70% level are thus somewhat less reproducible than the ED95 values and, as will be shown presently, are less Xenerally derivible from &e naphthoquinone assays. Thus the ED95 value, the dose (given three times a day) that is required to produce a 95% reduction in parasitemia as compared to the controls, has been employed throughout this work. The E D Bvalue ~ derived from an assay of a given drug is corrected (to ED9bc)for any deviation from the standard response noted in a parallel assay of quinine bisulfate as follows
90 110 130 150 170 190 Dose, mg./kg. day. Fig. 1.-Methods of determining quinine equivalents. 50
70
&ODs’’ would have been observed had the operator kmployed for his second doses amounts 6f the standard drug and of No. 760 anywhere in the (1) See Paper I , Ref. 1, for acknowledgments to CAIR and the Rockefeller Foundation. (2) Wiselogle, “Survey of Antimalarial Drugs,” 1940. (3) The percentage of quinine base in the bisulfate.
* 2.40 (21.1 %) mg./kg./dose
210
ED95c = E D QX~ 17.35/ED,rQ
The quinine equivalents (free base) cited in Paper I were calculated from the formula 10.26 (59.12 % of 17.35)/ED~5~ Q In the nineteen control assays of quinine bisulfate completed since the adoption of the standard value EDg6Q = 17.35 (done chiefly a t New Brunswick rather than Memphis), the average doses
Oct., 1948
CORRELATION OF STRUCTURE AND ACTIVITY OF ANTIMALARIALS
are somewhat higher: EDg6Q = 20.12 * 2.67 and ED7oQ = 12.56 * 1.73. For the total 65 assays the average ratio is ED,,Q/EDe6Q = 0.65. In 177 individual assays of naphthoquinones (some repeats, some late results not averaged) where the results can be analyzed from dose-response curves there are 144 cases where ED,o and ED96 values can both be calculated and the over-all average ratio is 0.50 == 0.23. In 50 instances an ED95 value is available where the figure is not, and in only 13 cases is the reverse situation encountered. Tabulation.--The compounds listed in this paper are all 3-alkyl or aryl derivatives of 2hydroxy-1,4-naphthoquinoneor closely related substances, and each of the eighteen series into which they have been classified includes substances that possess significant activity. Some 48 other miscellaneous naphthoquinones, including a number from the Hooker collection, together with the ten quinones described in Paper XIV, have been assayed and found inactive and are not listed in the accompanying Tables. Each of these Tables except the last one lists the members of the series concerned in the order of the number of carbon atoms in the side chain, whether these atoms are in uninterrupted combination or are separated by oxygen, nitrogen, or sulfur atoms. Since the 0-acyl and 0-alkyl derivatives of the strongly acidic hydroxynaphthoquinones, as well as the halonaphthoquinones that may be considered as the acid chlorides of these substances, are readily convertible into the free hydroxy compounds by mild hydrolysis, these substances are all treated as derivatives of the hydroxyquinones and are entered immediately after the parent compounds. Substances with aliphatic and alicyclic side chains are listed in Tables I-X, those with unsaturated substituents of mixed types in Table X I , and those with aralkyl and aryl groups in Tables XII-XIV.
3157
Compounds with side chains containing oxygen, halogen and nitrogen, respectively, appear in the next three Tables, and Table XVIII lists compounds closely related to substances of established activity or conceivably convertible into such substances in the organism. The previously known compounds that have been assayed are identified by reference notes; the reference letters appear as superscripts to the M-numbers. New compounds that were prepared for special studies or for characterization but not assayed are not listed here but are described (and coded) in the synthesis papers. TABLE I1 ISOALKYL SERIES Side chain
h.I
EDoz
ED960
264" (175) C4 1706b Inactive c 6 1523' 50, 71 72,64 Acetate 1707 Weak Propionate 2215 >80 (77%) Hydroq. diacetate 1708 Feeble Methyl ether 1720 Inactive Hydroq. triacetate 1721 Inactive Chloride 1923 Inactive Ca 1711 70-80 c7 1929 17.7 16.1 CS 287 19, 7, 18 16.4 G 2284 7.9,10.5 8.1,9.3 Cto 300 11 9.8 c11 2287 15 13.3 Fieser, Hartwell and Seligman, THISJOURNAL, 58, 1223 (1936). *Hooker, ibid., 58, 1168 (1936). 'Monti, Gnzs. chim. ital., 45, 52 (1915).
CS
Whenever the assay data permit, the antimalarial activity against P. lophurae in ducks is reported in terms of ED96 and ED96=values, or of the ED96 value alone in case a standardization assay of quinine bisulfate was not conducted. A substance is reported as inactive if i t caused no reduction in the parasite count a t the highest dosage tried TABLE I (usually 100 mg./kg./dose three times per day). n-ALKYL SERIES In case the highest dose tried produced a response Side chain nd EDor EDm that was either feeble or somewhat short of a 95y0 C, 1690" Inactive reduction in parasitemia, the entry under EDq6reports the highest dose tried, indicates that this c 3 263b >150 (21%) was not great enough to produce a 9sE0 effect, and C4 1709* 105 182 C5 17 LOb 67 116 includes in parentheses a figure recording the percentage reduction in parasitemia produced by the C6 268* 76 c7 26Ob 40 compound a t the dose given. This permits the comparison of weakly active compounds. In a C8 2'7 1 23,40 36.5 co 2275 8.3, 11.2 7.7,9.7 few instances the entry indicates that the lowest dose tried destroyed more than 957* of the paraClO 273 19.22 20.4 sites. Unless otherwise noted, the compounds CI1 1926 26, 40 37.1 were all administered orally; a citation is made in CIP 1928 25'30 27.8 C18 1924 48 those few instances of intramuscular administration. The activity graphs are based solely upon CI, 2347 (54%) C16 1714 44 43 data referring to oral administration; uncorrected CIS 2348 Inactive ED values are used when nothing better is availc 1 7 2256 Inactive able. Definition of Series.-The graphs of Figs. a Anderson and Xewrnan, J . Biol. Chem., 103, 197,405 2 and 3 show that in most series of quinones (1933). Hooker, THISJOURNAL, 58, 1163 (1936).
LOUISF. FIESER AND ,ARTHUR P. RICHARDSON
3158
Vol. 70
TABLE I11 hfETHYL-n-ALKYL SERIES Side chain
r
a-CH3 8-CH3 a-CH;+ o-CH, @-CHI Sodium salt o-CHa P-CHs
a
Structure
M
-CHMeCH2CH2CHs -CH2CHMeCH2CH3 -CHMeCoHll-n -CHMeC6Hl3-n -CHzCHMeCsHl1-n
1908 1910 279 280 284
--CHMeC,H15-n --CH2CHMeC~H13-n
3 14 285
Intramuscularly Sodium salt, orally intramusc. Acetate, intramusc. Propionate, intramusc. Hydroq. triacet., intramusc. Methyl ether B-CHa -CH2CHMeC7H15-n y-CH3 -CH2CH2CHMeCBH11-n a-CH3 -CHMeCpHlp-n KH.3 -CH2CHMeCQH19-n Synthesis by R. H. Brown to be reported later.
306 324 305 307 313 328 2357' 329
TABLE IV SERIES DIMETHYL-%-ALKYL Side chain Structure
Type
0 -C(CH$r Cr Ce
Q
CI
-CHC(CHs)a -CHzCHaC(CHa)a -CHMeCHMeCHzCHa -C!HMeCHaCHMen -CHaCHMeCHMea -CHCHMeCHMeCzHr -CHaCHMeCHaCHMea --CHMe(CHa)rCHMez -CHsCHMeCHMeCaHi-n -CHaCHnCHMeCHMeCaHs -CHzCHaCHMeCHzCHMea
M
EDar
1942 1934 2208 309 1939 269 310 270 1944 304 311 283
93 (80) 22.5 (59) >25 (27%) 50 13.9 24 3.3
Sodium salt CU -CHzCHMe(CHz)aCHMea 00 -CHzCHMe(CHz)KHMea -CHzCHGHMe(CHa)aCHMer CII -(CHz)3CHMe(CHz)rCHMez
EDBM
27.6 (53) 12.6 5.6
>SO 30 10.5
27
8.0 1941 333 1933 1974
10 9.0 12,s 14.5
7.7 13.9 12.9
EDai
EDarc
>
50 (64%) 90 (94) 40 6.5 3.4 19 6.2,5.0 3.5,2.2 2.0 2.3 < 1 2.1 2.3 1.6 >lo0 (72%) 4.5 13
20
3.8 34 6.5,6.0 3.3,3.5 1.8 2.5 25 (73%) >go (55%) 20 (38) Inactive (90) 35 5.0 >40 (24%) 3.0,ll.O Inactive Inactive
EDSM
34 8.7 5.2,10.6
CORRELATION OF STRUCTURE AND ACTIVITY OF ANTIMALARIALS
Oct., 1948
methyl, detracts from the 0 physiological potency. Other Iexamples of the same effect are to be found in the results 2o for substituted cyclohexylalkyl derivatives (Table VI). A The Tables include several series in which a cycloalkyl 40 group is separated from the 3 quinonoid nucleus by a methylene chain of increasing length. The cyclohexyl- d alkyl series, which appears representative and has been $ explored more fully than the .3 80 others, is characterized by an 4 activity chart of the usual pattern (Fig. 3) but with a shift in the peak of activity from a Cg to a Clo-C11 side chain. Extrapolation of the
3159
2 4
\
1
\ \ \
\\\
\ \
\ \
\
of this series i t would Possess Fig.2.-Relationship of antimalarial activity to length of side chain in the n-alkyl only feeble activity (EDss70). series. Aciually M-266 distinctly more potent I : E D J8.0) ~ than any of the members side chains. Except for the highly potent of the cyclohexylalkyl series and it is regarded cycloalkyl series, the activities of the peak memas belonging to the special group of compounds bers are all in the range of ED95 4 to 14. The dif(Table X) endowed with particularly high potency associated with the presence of an alicyclic ring TABLE VI1 joined directly to the quinonoid nucleus; even CYCLOPENTYLALKYL SERIES AND MISCELLANEOUS CYCLOthe lower members of this series are unusually ALKYLALKYL COMPOUNDS active, and the higher members include the most -Side chainType Structure M EDps EDnrc potent of all the naphthoquinones examined. CS CY CS Ct
TABLE VI CYCLOHEXY LALKYL SERIES ----Side Type
C7 Cs Co
Cio
Cii C1a a
chainStructure -CKzC6Hil --CHZCHZC~HII -CHz-l(1-Me-cyclohexyl) -CHZCH~:HZ--C~HII Intramuscularly Sodium salt Acetate Propionate" Hydroq. triacetate5 Hydroq. trisulfatc5 Oxime -(CHZ)A-CIHII -CHaCHMeCHzCsHii -CH%CHCHMeCsHi: -(CHn)a(4..Me-CsIIio) -(CHl)scsH11 -CHz-Menthyl -(CH~)YC~H:I
M 1914 1915 364 1916
1975 2203 1970 1972 383 1971 2243 2246 2204 1956 1963 1953
En6 55 22 82 20, 13, 20, 19.2,23.0
EDm
68 24, 22, 19, 18.5,23.2
5.6 6.2 5.0,4.8 5.6.5.0 44 41 9.7 11.0 35 > 40 Inactive 9.4,13.5 10,14 17.4 18 32 39 >50 (99.9%) 8, 16,7 4,10,8 >lo0 (30%) 77
Administered intramuscularly.
Optimum Size of Side Chain.-Table XIX shows the maximum activities observed in several series of quinones having hydrocarbon
-CHz-Cyclopentyl -(CHz)z-Cyclopentyl -(CH2)a-Cyclopentyl -(CHz)L.yclopentyl -CH;-Cpcldictyl 00 -(CHa)r-Cyelopentyl CII Fr. naphthenic acid Ciz -(CHz)a-CoHtr5 - - C H - - C S H ~ ~ - - C ~ H ~(A) ~ (B) Cir -(CHZ)IZCIHP 5
1920 110 35 33.8 2321 17.1 2322 19.1 18.3 16.9 2331 2239 24 22.5 2335 12.2 1968 44 76 2319 < 8 . 7 (98%) < 8 . 0 407 4.9 4.6 408 4.1 3.8 1936 87
5-Perhydrohydrinyl. 4'-Cyclopentylcyclohexyl : (A) (B) m. p . 182".
m. p. 128';
TABLE VI11 DECALYLALKYL SERIES -Side Type
chainStructure Cis -(CHt),-p-Deealyl-cis -(CHz)r-p-Deralyl-frans Cis -(CHs)r-p-Decalgl-cis Isomer A (130") Isomer B (120') -(CHa)r-p-Decalyl-frans Hpdroq. triacetate, intramuscularly
C,,
M 2320 2305 2279 2315 2516 297 2258
393 Oxime -(CHa):-a-Deealyl 2280 -(CH~)r-~-Decal~-l-fvarls 2296
ED98 4.6 7.6.5.7 4,5.5,2 6.5,2.9 3.5.3.2 3.5,9,12 7.5 Inactive 24.0 16.0
ED960 5.5 6.6,5.1 4,4.4,2 5.6,2.5 3.0,2.7 6 , s . 11.5 7.2
24.5 14.5
LOUISF. FIESERAND ARTHURP. RICHARDSON
3160
Vol. 70
is the most potent member assayed and is about twice as active as the C1l homolog. In the other two instances marked as questionable the next higher homolog was not assayed. Influence of Configuration.-The configuration of an alicyclic ring system in the side chain is a factor of considerable importance in the determination of potency, This can be seen from the summary given in Table XX of data in Tables VII, IX and X; the configurations of the first five 100 , , , l O O j , , , , # , , I pairs of isomers have been 5 6 ” 8 9 1 0 1 1 7 8 9 10 11 12 13 14 15 established experimentally; those of the last two pairs 0“ I are assumed on the basis of 10 analogy. The trans-4’-cyclohexylcyclohexyl derivative 20 M-2293 is exceptional not only in absolute potency but 30 in being no less than thirteen times as active as the cis 40 Isoalkyl Series isomer. Dimethyl-@-Alkyl Table VI11 includes results 50 for two pairs of P-decalyl80 alkyl isomer mixtures that are rich, respectively, in cis 70 - 0 1 and trans isomers; of these 0 configurations the former 80 seems to be the more favorable. A cis-8-decalyl group 90 joined :directly to the quinoIOOI , , > , 1001 , , , noid nucleus (Table X ) like4 5 6 7 8 9 1011 4 5 6 7 8 9 1 0 1 1 wise gives rise to higher Fig. 3.---Effect of lengthening the side chain in other series (ED93 plotted against potency than a trans group. Unsaturation.-The asnumber of carbon atoms in the side chain). sav Table XI includes six ferent series seem to fall into distinct groups de- instances permitting Gmparison of ED^^ for unfined by the number and kind of rings present in saturated derivatives with those for the correthe side chain. If no rings are present, maximum sponding saturated compounds. In all but one activity is reached in a Cs-group; if one alicyclic case the introduction of a double bond in the l’-] ring is present the peak is shifted to Cl&; if the 5’-, 6’-, 8‘- or 9’-position of the side chain proside chain contains two alicyclic rings either TABLES joined or fused together the peak is displaced to CYCLOALKYL SERIES C12-C13. A phenyl group seems to produce a still chainmore pronounced shift in the same direction] but -Side Type Structure M EDM EDPM the present data show only that the CISderivative 26 Cs -Cyclopentyl 2326 26
t
I
I
I
#
~
/
f“
1
I
t
I
266 2328 p-Decalyl-trans 2374 CII -CsHloOHp,“ low m. p. 411 -CsHloC~H~,ahigh m. p. 412 cI2 - C ~ H ~ ~ C ~ H ~ ~ * - C ~2327 J -C~H~~GHir*-frans 2293 -CsHlo, CsHIC-cis 401 -CaHlo. C S H‘-frans ~ 400
Ca -Cyclohexyl CIO (3-Decalyl-cis
TABLE IX
-~‘-CPCLOHEXYLCYCLOHEXYLALKYL SERIES chamType Structure hil ED96 Cia - - O H ~ C S H I ~ C ~ H I I - ~ + S 384 10 5 , 9 8 --CHaCsHla CsHii-frons 380 2 7,12 0 C I ~ - ( C I ~ Z ) ~ C S H I ~ C S H I I 2329 - ~ ~ ~ 45 0 2330 7 7 -(CHt)tCsHloCsHit-lrans 2291 16 2 CIS -(CHa)aCaHioCsHit-crs - ( C H ~ ) ~ C ~ H I O C ~ H I I 2292 8 0
I
4
------Side
EDoac
9 7,s 7 2 8,lO 7 7 1 14 6 7 2
4 ’-Cyclopentylcyclohexyl. 4’-Phenylcyclohexyl. a
7.0,9.7
5.5 12.0
7.0 0.6,1,0.4 31.3
8.0
6.9,s.l 5.0
9.0 0.6,1,0.4 29.0 7.4
4’-Cyclohexylcyclohexyl.
Oct., 1948
CORRELATION
~ _ _ _ Type
OF
STRUCTURE AND ACTIVITYOF ANTIMALARIALS
3161
TABLE XI UXSATURATED SERIES Side chain Structure
7
M
ED96
EDoao
C*
-CH=C (C H S ) ~ 1685" Inactive cs -CH=CHCH (CH1)z 1537b 135 235 c 7 -CHzCH=C(CHa)l 1684" >200 (87%) -CH=CH-C5H11-n 1538" 40 GX -CHzA2-Cyclohexenyl 527 > 80 (78%) cs -CH=CHCd& 1715" > 110 (15%) cs -(CHz),- A2-Cyclohexenyl 374 22.4,36.0 20 2 , 2 7 . 0 -(CHz)a- Aa-Cyclohexenyl 2333 47.5 41 2 -(CH2)7CHdHz 2353d 12.5 ClO -(CHz)8CH=CHg 289 12.0 11 3 Cl7 -Korchaulmoogryl 1945 < 100 (99%) Hooker, THISJOURNAL, 58,1223 (1936). * Hooker, J . Chest. SOL,69, 1358 (1896). Hooker, ibid., 69, 1355 (1896) Synthesis by R , H. Brown to be reported later. TABLE XI1 (a) W-PHENYLALKYL SERIES AND (b) POLYARYLALKYL
SERIES Side chain Structure M EDos (a) o-Phenylalkyl Series 1737a >I40 (43%) C7 -CHzPh 1946b >lo0 (70%) Ci -(CHz)aPb 65 Ce -(CHz)rPE, 1955b 41 CIP -(CHz)rPb 2286 32.0 Cii -(CHt)aPh 2276 16 0 Cic -(CHz)oPh 2301
Type
ED96c
4'-Phenylcyclohexyl. hydro-9-phenanthryl. 36.5 32 7 13.9
Fieser, THIS JOURNAL, 48, 3201 (1926). Hooker, Mohlau and Klopfer, Ber., 32, 2146 (1899). 0
ibid., 58, 1163 (1936).
TABLEXI11 OTHERARALKYLCOMPOUNDS M 2234 2281 1952 2236 2235 2216 2214 2237 2308 2230 2209 2277 2242 2252 2304 2257 2213 2302 295 2254 2253 396 305 2295
ED96
>
EDm
75 (17%) 65 64 14.5 15.0 > 75 (16%) > 75 (84%) 40 (62%) > 75 (68%) 32.4 36 Inactive 18.5 22.9 > 40 (77%) >loo (22%) 22.5 23.0 > 50 (61%) 34.7 36.0 37.2 35.8 62 56 40.7,38 23.5,55 > 20 (18%) 38.0 37.0 114 95 7.5.8.9 9.0,8.3 30.8 34.0
> l o 0 (7%j (100) 56 20.0 > 80 (9%) > 80 (11%) 85
Cyclohexyl.
50
76
1,2,3,4-Tetra-
TABLE XIV ARYLSERIES Aryl group
(b) Polyarylalkyl Series CII -CH(Phjz 1739 Inactive ci4 -CHzCH(Ph)z 2282 >lo0 (56%) CI, -CHzCH( Ph)CsHEH,-p 2274 Inactive CIS --CH%CH(C6H&Hs-p)n 2314 Inactive - C H Z C H ( P ~ ) C ~ H I ( M ~ ) 2283 Z - ~ , ~ Inactive -(CHz)zCH(Ph)CsH4CHz-p 2297 > 100 (8%) Cio -C(Ph)i 1740" Inactive
Side chain Type Structure C, -(CH~)ZC~H~CHI-P Cia -(CHz)zCHMePh -4CHz)rCsHoMe-P -(CHa)zCdHiMex-2,4 -(CHz)aCsH4Et-P -CH:CHEtPh GI -(CHz)iC1HaMez-2,4 -(CHz)iCaHrMer-2,5 -(CHZ)ICSHIM~~-~,~ -CHtCHMe(CHz)zPh -(CHz)aCsHdEt-p -( CHz)z-5-Hydrindyl Ciz -(CHz)iCtHxMea-2,4,6 -(CHz)sCsHdi-Pr)-p -( CH~)r-@-Tetralyl -(CH:)a-)I-B-Hydrindyl CII - ( C H M ~ ~ H ~ ( ~ - B U ) - P -(CHl)tCiH1Me(i-Pr-)2,5 --fC~)i-,3-Tetr.lyl -CHI) I-a-NaphthYl -(CHz)l-&Naphthyl -CHaCsHii-CaH~"-ciS Cir -CHaCsHii. CsHk'-fYanS Ct4 -(CHz)e,9-Tetralyl
Cia -(CHz)aCsHaPh-P 2290 2359 -(CHz)zCaH4CHaPh)-p -(CHZ)Z-CIH~-CSH~~~-P 2323 Cis - ( C H ; ) ~ C ~ H I ( C H ~ P ~ ) - ~2339 -(CHz)1-2-Fluoryl 2272 c i 7 -(CHz)1-Hydrophenanthryl e 2255 CI-S -(CHz)z-ct-Thienyl 2285
hl
EDBk
EDMC
323" >IO0 (947,) -Ci" 2313 (125) o-Chloro 1932 Inactive m-Chloro 45 78 p-Chloro 1938 1937 o-Bromo > l o 0 (41%) p-Bromo 50 29 1935 2212 P-Iodo > 40 (37%) 2217 +Fluor0 > 60 (56%) 2,4-Dichloro 1973 >lo0 (30%) 1966 > 75 (40%) 2,5-Dichloro 2288b > 100 (18%) p-Methoxy 2298 $-Ethoxy >loo (8%) Inactive 1901b P-SOaK Inactive 1743 fi-SOzNHa Inactive 1925 p-SO2KH-2-Pyridyl 1919 Inactive p-SOzNH C (IjHz)=NH 2317b > 50 (8070) o-CH~ 2300 90 78 WZ-CH, 73 127 P-CH, 1967b 2225 Inactive 2-Methyl-4-chloro 2307 2,4-(CH3'12 > l o 0 (63%) 2278 2-a-Naphthyl > SO (53%) 1960h 20-60(M-2376)
OH ED960 6.3(M-2367)
I can-n
When assayed by this method, M-285 (EDssc 1.8) and M-1916 (EDBs~ 5.6) proved to be 2.5-4 times as active as when given by mouth, whereas M2350, the most potent of the four alcohols, is ten times as active when given by the injection route. It is surprising that the seemingly minor change in structure from M-2350 to the isomer M-2376 results in almost complete loss in activity; apparently a higher degree of structural specificity exists among the alcohols than in the unsubstituted compounds.
Oct., 1948
CORRELATION OF STRUCTURE AND L ~ C T I V I T YO F -4NTIMALARIALS
TABLE XVI SIDECHAINS HALOGENATED +-Side Type c 4
CS
CT
CS
C,
0 0
Cli C1r e
chainStructure
-
M 2247 1309' 2201 1738 2202 2366 1962 2289 1968 2358 2332 2260 2271 2244 2373 2310 2346 2365 340 2299 2340 2341 2344 2362
EDis Inactive >loo (8%) > 25 (6%) 104 52 > 60 (54%) >IO0 (57%) 18.8 >IO0 (27%) 55 24.6 20.2 19.3 60 28 70 (120) 38 >lo0 (24%) 38.0 30 > 50 (86%) 10.2 13 0
TABLE XVIII INACTIVE RELATED COMPOUNDS ED#*
47.5
18.1 29.8 45 19.5 18.6 57 25 63
32.1
9.4
Hooker, J . Chem. SOC.,61, fill (1892). TABLEXVII NITROGEN-CONTAINING SIDECHAINS
Side chain Type Structure CI -CHzN(CH:)l -CH:NHCHiCHxOH CI -CK%NHC4Hs-n -CH~N(CZHS)a -Morpholinomethyl -(CHz)rCONHz Co -CHlMHCrH11-% -CHsNH-5- (5-Me-dioxanyl) -Piperidinom ethyl C; -CHzMH-Cyclohexyl -2-Me-piperidinomethyl -4-Me-piperidmomethyl -(CHl)rCH (CHs)CONHz -CHzCsH4>.102-9 Cs --CH&HCI-I?CIHI Ci- -(CH~)IINHVHCI -CHzNHCIoHzi-n -(CHL) ieCN Cir -(CH,)loN(CzHr): Cn -CH[CoH4PJ(CHa)i-91i
M 336 360 353 339 345 2232 346 355 344 347 361 362 2240 1954 363 341 354 335 379 1943
3163
EDBS Envac >lo0 (16%) >lo0 (8%) 130 Inactwe Inactive Inactive 2 1 Q O (92%) Inactive 38 25 >100 (63%) 170 160 42 40 Inactive > 75 (60%) Inactive >lo0 (24%) Inactive >lo0 (61%) Inactive > l o 0 (8%)
A second series of interest is that of the p-phenoxyphenylalkyl derivatives, exemplified by the most active member, M-2309, with the side chain (CH2)8CsH40CsH5;this substance seems to be better absorbed than the alcohols, for it is only twiee as potent when given intramuscularly (EDgtc7.2) as by oral administration. The next lower homolog in this series is only half as active as M-2309 and the tetramethylene and pentamethylene homologs are only feebly active. It is very odd, therefore, that potency about half that of M-2309appears again in the higher homolog with no less than nine methylene groups. A partial accounting for this phenomenon is to be found in Paper XXII. Structural specificity is again evident from the fact that an o,o'-bridge linking together the two
Compound
M
1,4-Naphthoquinones : 2-Isoamyl 1743 2-Isoamyl-3-methyl 1732 2-Isoamyl-3-amino 1949 2-Hydroxy-3-isoamyl-6-methyl 2226 2-Hydroxy-3-isoamyl-7-methyl 2229 2-b-Cy~l0hc~ylb~tyl 2318 2,6-Dihydroxy-3-~-cyclohexylpropyl 2207 2-Hydroxy-3-tetrahydrogeranyl-6-bromo 1957 2-OH-3-y-cyclohexylpropyl-5,6,7,8-tetrahydro2248 2-Hydroxy-3-y-cyclohexylpropyl-7,8-benz 2324 2-Isoamyl-2,3-oxido 1900 2-Cyclohexyl-2,3-oxido 2351 2-Hydroxy-3-anilino 1904" 2-Hydroxy-3-p-chloroanilino 1905 2-Isoamylnaphthoresorcinol 348b 2-Phenylnaphthoresorcinol 337" Alkannin 267d Perezone 293d Embelin 1735" Rapanone 1736'
r";I'CqcHoH
1 1
R = -CHtCH(CHt)1 1947 & H ~ R R = -nortetrahydrogeranyl 1948'
iJ,co i
I C02H
0 226F
0
O -
F
0
1
R=
1
1724 256 ~ ! ~ ~ ~ C l - o 272 -CsH,CI-P 276 --C~HJ'T(CHI)~-P 2245
0 0 "Zincke and Wiegand, Ann., 286, 76 (1895). bSoliman and West, J. Chem. SOC.,53 (1944). CVolhard, Ann., 296, 14 (1897). Mayer and Cook, "The Chemistry of Natural Coloring Matters," Reinhold Publishing Corp., New York, N. Y., 1943. .Synthetic, Fieser and JOURNAL, 70,71 (1948). I A dosage of Chamberlin, THIS 100 mg./kg. produced a 32% reduction in parasitemia. EDss < 100 (99%).
phenyl groups of M-2309 destroys the activity (at least as judged from oral assays). The few CleCZZ methoxy, phenoxy, and P-phenylphenoxy compounds thus far examined have not exhibited high potency. Halogenated Side Chains (Table XVI).The introduction of an w-bromo substituent into the side chain of 2-hydroxy-3-decyl-l,4-naphthoquinone practically abolishes the activity, but substitution in a larger alkyl group has not been investigated. An aromatically bound halo snb-
LOUISF. FIESER AND ARTHURP. RICHARDSON
3164
TABLE XIX PEAKACTIVITIES Number of rings
Series
n-Alkyl iso-Alkyl P-Methyl-n-alkyl Cyclohexyla1kyl Cyclopentylalkyl ~-Decalylalkyl-cis P-Decalylalkyl-trans 4’-Cyclohexylcyclohexylalkyl-cis 4’-Cyclohexylcyclohexylalkyl-truns Cycloalky1 w-Phenylalkyl
Side chain having maximum activity
G
0 0
C O
0
ce
1 1 2 2
ClO-CIl C,O(?) Cis CII
2
CIS
EDw of most active member
8.7 8.7 4.4 10.5-11.4 12.2 3.5 5.8 9.2
6.7 2 e12 0.67 2 Cld ?) 1(Bz) CIS(?) 13.9
TABLE XX ACTIVITIESOF GEOMETRICAL ISOMERS Side Chain :
-(CH2)n
R
n
Cyclohexy1 C yclohexyl C yclohevyl Cyclohexyl Phenyl Cyclopentyl Cyclopentyl
0
4 1 3 - R EDos cis/EDos trans
13.4 1.3 5.8 1.9 3.9 1.7 1.2
1 2 3 0 0
1
stituent in an aralkyl side chain seems to have the effect of enhancing activity. Thus the p-chlorophenylethyl derivative M-2289 is slightly more potent than M-1916, whereas the unsubstituted parent compound is practically devoid of activity. 0
0 M-2289 (EDg5o 18.1)
0
0 M-2344 (ED956 9.4)
The C3 and C11 aralkyl homologs of M-2289 show decreased activity, but the C16 derivative, M-2344, exhibits twice the potency of the Cg-derivative. This resurgence of activity in a higher homolog constitutes another example of the phenomenon mentioned in the preceding section. Although the data of Table XVI on the relative effect of ortho and para halo-substitution are not fully consistent, the burden of evidence is that the para location is the more favorable. The assays of the a t most feebly active 2-hydroxy-3-aryl-l,4naphthoquinones (Table XIV) tend to show that the order of preference is: para > ortho > meta. The assay results include two comparisons that indicate that a fluoro substituent in the benzene ring. is not as favorable for activitv as a chloro or br&o substituent. One comp&ison indicates that an iodo substituted aralkyl derivative is
Vol. 70
slightly less active (on the weight basis) than the chloro and bromo compounds. A comparison of the relative potencies of chloro and bromo derivatives is best considered in conjunction with the relative antirespiratory activities, and discussion of this point is therefore deferred to Paper XXII. I t is noteworthy that hydrogenation of the aromatic ring of the m-trifluoromethylphenylethyl derivative results in more than three-fold increase in activity. Nitrogen-Containing Side Chains (Table XVII) .-The present results show that activity is completely lost on the introduction into a C S - C ~alkyl ~ side chain of a terminal group of the type -NH2, -N(CzHc)z, -CN, or -CONHI. Compounds with larger side chains containing the same substituents remain to be investigated. Surprisingly, the only nitrogen-containing quinones found to possess even moderate activity are Mannich bases containing only six or seven carbon atoms in the side chain. The most active member, M-344, is a bright red, water-soluble substance having the character of a dipolar ion. 0
M-344 (EDgse 25)
Variations in the Naphthoquinone Nucleus (Table XVIII).-The results of such trials as have been made tend to show that little variation can be made from the standard nucleus of a 2hydroxy- 1,Cnaphthoquinone without practically complete loss of activity. Thus substitution of halogen or hydroxyl a t the 6-position in the aromatic ring destroys the potency exhibited by the parent compounds. The introduction of a methyl group a t the 6- or 7-position in hydrolapachol results in deactivation, whereas methyl substitution in the side chain enhances the potency. The effect of these and other groups in all four available positions and the possibility of offsetting the effect of a substituent by extending the alkyl side chain is the subject of a further investigation. The acidic quinonoid hydroxyl group appears to be essential to activity. Quinones related to hydrolapachol, hi-1916 or M-1971 but having in place of hydroxyl any one of the following substituents are completely inactive: H, CH3, NHz, C1, SH . The activity of hf-1916 is completely destroyed on hydrogenation of the aromatic ring. The resulting substance can be regarded as a substituted benzoquinone, and its lack of activity would suggest that the only hope of discovering active benzoquinone analogs would be through some adjustment in the size of the side chain. A preliminary assav of a thiophene isolog of M-1916 indicates thac the substake possess& definite activity but
Oct., 1948
DIENE SYNTHESIS O F l,$-NAPHTHOQUINONE ANTIMALARIALS
does not define the range of potency. A benzolog of the phenanthrenequinone series is completely inactive. Transformation to Active Compounds in the Organism.-The observation that several precursors of vitamin K1 or of 2-methyl-l,4-naphthoquinone are apparently converted efficiently into these antihemorrhagic substances in the animal body4 suggested trial of comparable precursors of the naphthoquinone antimalarials. It appears, however, that the duck organism has relatively little power for effecting such biological transformations of orally administered materials. Thus even simple esters and hydroquinone triacetates of such active compounds as M-1916 showed a t most feeble activity in oral assays. When given intramuscularly M-1916 propionate proved (4) Fieser, Tishler and Sampson, J . B i d . Chcm., 187, 659 (1941).
3165
to be almost as active as the free hydroxy compound but the triacetate was only feebly active. In view of these findings i t is hardly surprising that orally administered oxides and naphthoresorcinols showed no activity. Two of the compounds listed in Table XVIII are ketols that are readily transformed into hydroxyalkylnaphthoquinones by oxidation with copper sulfate and alkali (Paper XII) and the one of higher molecular weight showed signs of a t least weak activity; this type of precursor is of interest by virtue of greatly increased water solubility.
Summary The results of this investigation are summarized in Paper I of the series. CAMBRIDGE,
MASSACHUSETTS
NEWBRUNSWICK, NEWJERSEY RECEIVED MAY13, 1947
[CONTRIBUTION FROM THE CHEMICAL LABORATORY OF HARVARD UNIVERSITY]
Naphthoquinone Antimalarials. 111. Diene Synthesis of 1,4-Naphthoq~inones~ BY LOUISF. FIESER
A reinvestigation of the reactions involved in the not fully redissolve. One solution is to add a diDiels-Alder synthesis of naphthoquinones and in chromate-sulfuric acid mixture to the acetic acid the hydroxylation of the products has led to the solution of the adduct a t 65-70’ ; the processes of development of shorter routes to several of the isomerization and oxidation then proceed concursimple naphthoquinone intermediates that have rently and the sparingly soluble component I11 been required for the synthetic program and has never reaches a concentration high enough to also provided a practical method of preparing cer- cause tar formation. A still better laboratory tain hydroxyalbylnaphthoquinones that are dif- procedure is based upon the finding that nitrous acid in acetic acid solution5 is a specific reagent for ficultly accessible by peroxide alkylation. In the procedure2 heretofore usually followed oxidation to the dihydronaphthoquinone stage the addition of butadiene to a benzoquinone is (V) and no further.6 The reaction proceeds very conducted in benzene, ligroin or alcohol at about readily without tar formation, and when chromic 70’ in a pressure vessel, the low-melting and highly acid is subsequently added a smooth conversion soluble adduct (11) is isolated, isomerized to 111, to the naphthoquinone is realized. and this is oxidized with chromic acid to the The preparation of pure a-naphthoquinone by naphthoquinone. It has now been found that the either procedure is far simpler than by the previDiels-Alder addition can be conducted efficiently ous best method from a-naphthol’ and the reacwithout the use of pressure equipment and a t tions can be applied on any scale. The compound room temperature in glacial acetic acid, a solvent thus becomes a practical intermediate. One use well adapted to use in the subsequent steps of is in the preparation of 2-chloro- and 2,3-dichloroisomerization and oxidation. The isolation of the 1,4-naphthoquinone. Another application affords adduct is unnecessary, for isomerization to I11 can a better route to 2-hydroxy-1,4-naphthoquinone be conducted by brief heating following the addi- (VIII) than that from Boron fluoride tion to the solution of a mineral acid and stannous (best as etherate) has been found a superior catachlorideae4; the parent substance separates in lyst for the Thiele reaction9 with acetic anhydride, colorless needles of high purity in excellent yield. and by a simple procedure pure 1,2,4-triacetoxyThe direct oxidation of 5,8-dihydro-1,4-naphtho- naphthalene (VII) can be prepared in quantity. hydroquinone (111, R = H) with chromic acid to A nearly quantitative conversion to pure 2-hya-naphthoquinone a t first presented the difficulty (5) Trial of this reagent was suggested b 7 an observation by Fieser that unless an excessive volume of acetic acid is and Peters, THISJOURNAL, 13,4083,4090 61931). (6) A previous search for an oxidizing agent better than chromic employed an intermediate tar separates and does acid led to the discovery of the alkylating action of lead tetraacetate; (1) See Paper I, Ref. 1, for acknowledgments to CMR and the Rockefeller Foundation. (2) Diels and Alder, Bcr., 62, 2382 (1929). (3) Fieser, Tishlcr and Wendler, THISJOURNAL, 62, 2861 (1940). (4) Fiescr and Chang, i b i d . , 64, 2043 (1942).
Fieser and Chang.4 (7) Fieser, “Organic Syntkeses,” Coll. Vol. I , John Wiley and Sons, Inc.. New York, N. Y.,1941, p. 383. (8) Fieser and Martin, “Organic Syntheses,” 21, 56 (1941). (9) Thiele and Winter, Ann., 311, 347 (1900).
