Electronic structure, antimalarial activity, and phototoxicity of selected

quine, w'hich has a 4-amino moiety, is not metabolically attacked at the 2 position but the quinolinemethanol are. Derivatives of quinolinemethanol of...
0 downloads 0 Views 393KB Size
1s

Electronic Structure, Antimalarial Activity, and Phototoxicity of Selected Quinolinemethanol Derivatives and Analogs1

The electronic structures of eighteen derivatives and analogs of c~iiiiioliiiemetharioiwere calculated by the Iliickel molecular orbital method. The results were searched for possible cori,elationi among the eleclronic~ iiidicaes and phot,otoxicity and ’or antimalarial a d v i . X parameter, l i l H O \ l O E:LI:\IO~ is introduced as :I measure of molecular “electroiiegativity.” The 2 posi xi was fouiid to he the most positive in d l the molecule-; ralculated except those which have a 4-amino group. This observation msy be relnt,eti to the fact that chloroquine, which has a 4-amino moiety, is nut metabolicdly attacked at. the 2 po4tioti but the quitiolinemethanok are.

+

Derivatives of quinolinemethanol of type I are reported to tic both active against Plasnzodiur~ berghei iri rnicc :md phototoxic.? The 2-phenyl group is helieved to contribute to the high activity by prcventing oxidation t o c:trbostyril.2 The activity dccreases markedly and phototoxicity is not observed, however, when an > S H moiety replaces >CHOH (derivativeq of type 11) oi’ if the “phenyl group is removed (derivatives of type III) . 4 Several laboratories, including ours, have a5 their target a molecdar modification of quinolinemethltnol which would preserve the high activity of the ”phenyl derivatives and reduce the phototoxicity. RCHOH

I

Methods All of ihe calculations were in the simple ITuckel approxiriiaiI O I I . ~ . ~‘I’he sernrempiric*alparameters8 are t hose rrcorrirrirtltieti 1)y Streitwirser“nd

nie given in Table I.

R

RNH

I

I

Results and Discussion R = RICHOH or R’NH

I

I1

I11

Chloroquine (Y11, Figure 1) is also :L derivative of quinoline and serves as an interesting compound for comparison. It undergoes metabolic attack on the :mino side chain5 rather than at the 2 position as is found in quinolinemethanols of type 111. &o, since 2-phenyl analogs of chloroquine of type I1 are not phototoxic,* the phototoxicity of the quinolinemethunols of type I seems to depend on both the presence of a phenyl group at the 2 position and a methanol moiety nt the 4 position. R’ith the idea that the electronic structures of these molecules are at least partially responsible for the activities and phototoxicities, Huckel molecular orbital (HJIO) calculations were performed on variations of the quinoline ring system. Thc aim is to use the results to assist in the rational design of molecules having maximum antimalarial activity and minimum phototoxicity . (1) This research is being supported by the U. S. -4rmy Medical Research and Development Command ( D 4-49-lY3-hID-2779), the National Science Foundation ( G B - i 3 8 3 ) ,and a grant from Eli Lilly and Company. Thls paper is Contribution No. 441 from the Army Research Program on hlalaria. Computer facilities n e r e provided through Grant HE-09495 from the National Institutes of Health 11, ( 2 ) D. W . Boykin, J r , A R Patel, and It. E. Lutz, .I. .lied. Ch~m., 273 (1968). ( 3 ) R. M. Pinder and A. Burger, tbtd., 11, 267 (1968). (4) W. E. Rothe and D. P. Jacobus, zbzd., 11, 366 (1968). ( 5 ) R. T. Williams, “Detoxication Mechanisms,” 2nd ed, John Wiles and Sons. Inc., New York, N. Y., 1959, p 652.

Figure 1 gives the moieties which were calculated and is purposely arranged so that molecules next to one another differ by only one substitutional or strurtural variation. This facilitates corivcnient and thc most meaningful comparisons of the data since, in the HA10 approximation, relative values in a series of closely related molecules are more significant than the absolute values. Several electronic indices were calculated (bond order, free valence, ctc.), but just thaw which are believed to beiinportant to the interprets‘1t‘ion of the biological data are presented here. Since the methanol moiety arid saturated side chains are not considered as part of the conjugated r-electron systems in the HI\ IO method, quinolinemethanol n ill be considered equivalerit to quinoline (IV), and chloroquine will bc considered equivalent t o 4-ami110-7chloroquinoline (VII). Table I1 gives the net a-electron charge-density distribution‘: at the atomic sites for the moieties in Figure 1. Table 111 gives the energies of the highest occupied and lolvest empty molecular orbitals (HOJIO and LENO) and the sums and differences of thesc energies. The difference of the HOMO arid IJ;-\IO energies correspontls to the lowest a-electronic exit:\(6) 4. Streitwieser, Jr., “SIolecular Orbital Theory for Organic (.’liemixt3, John Wiley and Sons, In?..,Ten. York, N. Y . , 1961. (7) C. A. Coulson and ..1. Rtreitwieaer, Jr.. “Dictionary of x-Electmri C‘alculations,” W. 11. Freeman and Co.. San Francisco, Calif., 1Y65. (8) W.P. Purcell a n d J. A. Singer, J. Chem. Bng. Data. l a , 235 ( I N 7 ) .

ELECTRONIC STRUCTURE

January 1969

OF

QUINOLINEMETHANOLS

19

iV

HN-

HN-

I

IV

: HN-

w

VI

V

: I

xiv

IX

HNm0QNH I

!

8

I

x

$11

--___---___________ HN

I

-NH XI

Figure 1.-Quinoline

XI1 moieties for which the HMO electronic structures were calculated.

tion energy of the molecule. The sum of HOMO and and electron affinity of a molecule in the Huckel apLE110 energies is a parameter which has not been proximation. l1 used previously; it may be considered as a measure Some generalizations may be made about the results of the “electronegativity” of the molecule. The of the calculations. (1) The amino group makes the corresponding theoretical expression for the electromolecule (IV, VIII-XIV) a better electron donor and a negativity of an atom was derived by A\\rullike~~.gworse electron acceptor. I t makes the molecule easier Based on the concept that the energy expended in to excite electronically; one exception to this is the going from the covalent molecule A-B to the ionic 2-amino derivative XIII, which has the same excitation states A+B- and A-B+ is equal if A and B have the energy as the parent compound IV. The amino group same electronegativity, Ilulliken concluded1° that the makes E H O M O ELEMOsmaller. (2) Phenyl group electronegativity of an atom is proportional to the substitution makes the molecule a better electron donor sum of its appropriate valence-state ionization potential and a better electron acceptor. It makes the molecule and electron affinity. The parameter EHoMo easier to excite electronically. The sum of EHoMo I?LEJIO corresponds to the sum of the ionization potential E L E ~becomes ~ O smaller with the single exception of the

+

+

(9) R. S. hlulliken. J. Chem. P k g s . , 2, 782 (1934) (10) R. 5. hlulliken, i b i d . , 46, 497 (1949).

+

(11) W. P. Purcell, J. A. Singer, K. Sundaram, and G. L. Parks in “hIedicinal Chemistry,” A. Burger, Ed., 3rd ed, John Wiley and Sons, Ino.. K e n York, K.Y..in press.

ANTIMALARIAL KETOGUAKTLHYDHAZONES

January 1969

phenyl in the 2 position to block metabolic attack, and (3) an XH moiety in the 5 or 8 position to give a low EHOh1O E L E M o or low phototoxicity. Derivatives which might be worthy of synthesis and evaluation include those of type XXII and XXIII where R‘ and

+

21

R2 are the comnionly employed moieties in antimalarial drug design such as

RTHOH RZCHOH

Such syntheses are in progress. Acknowledgment.-The RINH XXIII

XXII

authors thank Dr. David

P. Jacobus for fruitful and interesting discussions and for providing data relevant to these studies.

Antimalarial Activity of Guanylhydrazone Salts of Aromatic Ketones. I. Primary Search for Active Substituent Patterns’

JEFFERSON R. DOAMARAL, ERWISJ. BLANZ, JR.,AND E’REUERIC A. FREXCH Jlount ZLon Hospitul and Jf edztul Center, Chemotherapy Research Laboratory, Palo Alto, Culifoi rLia LJ4dOd Received June 17, 1968 Thirty two guanylhydrazones of aromatic ketones were synthesized and teated in the primary aiitiinalarial scieen in mice infected with Plasmodzum berghei. Biological data have been received on 30 compounds. S i n e compounds showed significant activity and seven of these gave a high percentage of ciireb. The biological results and structure-activity correlation are discussed as well as drug design arid the syiithetic problems involved. Activity of some of these compounds on L1210 leukemia in mice is described.

In 1962 a small group of substituted benzophenone guanylhydrazones (I) was synthesized specifically to study their action on L1210 leukemia in m i ~ e . ~Sev,~ eral of these compounds displayed significant activity.

I However, the dose-response relations arid therapeutic indices were so poor that the project was abandoned. In 1965 and 1966 we sent 17 of these compounds (Table I) to the Walter Reed Army Institute of Research for testing in the primary antimalarial screen (Plasmodium berghei in mice). One compound, 4-fluoro-i’-trifluoromethylbenzophenone guanylhydrazone hydrochloride (15), was quite active and caused some cures. As a consequence a contract was activated to pursue this lead. Referring to structure I, one can introduce one or more substituents in either or both rings and insert various substituents in place of hydrogen on the aminoguanidine moiety. Considering the limited biological data available (Table 111) a simple series was generated wherein the guanidine group was unchanged, a 4-CF3 (1) This investigation n as conducted under Contract DA-49-193-MD3016 from the U S. Army Research and Development Command. This is Contribution No. 313 to the Army Research Program on blalans. (2) F. A. French E J Blanz, Jr , and C. C. Cheng. Proc. Am. Assoc Cancer Kea 4 , 20 (1903) ( 3 ) T l i i v uork i i a b suyuorted 1,) Grant C.1-03287 f r o m tlie h+tio~i+l C unoar Inatitutu.

group was maintained on one ring, and a variety of substituents mere placed on the 4’ position. These groups were chosen with the usual considerations. A range of electronic effects and influences on solubilitydistribution behavior could be studied relatively easily. While this work was in progress information was received on the moderate activity of 7 and the high activity of 17. Consequently, attention was focused on bromo, iodo, and additional trifluoromethyl substituents. The resultant series of compounds is shown in Table 11. Biological Data and Correlations.-An inspection of the data in Table I11 and a consideration of ancillary toxicity data make it apparent that there are two correlations involved in the structure-activity relationships: a toxicity correlation and an antimalarial activity correlation. Any activity which might conceivably have been displayed by compounds substituted only in one ring with fluoro, chloro, or trifluoromethgl is masked by high host toxicity. The same limitation applies nhen both rings bear fluoro or chloro substituents. For monosubstitution the toxicity decreased in the order: CFs > I > F > Br. When bromo and/or trifluoromethyl substituents are present in both rings toxicity is quite low. The only monosubstituted compounds showing antimalarial activity in the primary screen were the 4bromo and particularly the 4-iodo derivatives. In the new series of 4-trifluoromethyl-i’-halo derivatives the order of activity cannot be stated precisely at this time. However, incomplete advance biological data show that both the 4’-chloro and 4’-iodo derivatives are active. Curiously enough, the 4,4’-dibromo derivative is inactive but it has the expected low toxicity. In contrast, both the ~,~’-ditrifluoron~ethyl a i d the (4-homo-