Structure-activity correlations of morphinelike analgetics based on

a centrally mediated opioid effect on rumen motility in conscious goats and sheep ... Quantum chemical and physicochemical influences on structure...
2 downloads 0 Views 582KB Size
STKGCTUHE-ACTIVITY

RELATIONS O F XIORPHISC ASALGETICS

Journal of Medicinal Cheniistry, 1970, T’oi. 13, S o . 5 801

Structure-Activity Correlations of Morphine-like Analgetics Based on Efficiencies Following Intravenous and Intraventricular Application E B E R H A R D ICUTTER,

Thonnae Iiesearch Laboratories, 795 Biberach an der IZiss, West Gerntuny *ALBERT

HERZ,HANS-JORG TESCHEJIACHER, A N D RSINER HESS

Max Planck Institzit fur Psychiatrie, 8 Xunich 83, lt-est Gelmany Received Jlarch 30, 1970 Analgetic efficiency of morphine-like drugs following intraventricular application is found to be a good approximation for their “receptor activities.” Evidence is preiented indicating that ths quotient log (CLYentrICLY) parallels the lipophilicity of the compounds and is a measure for their tendency to penetrate the blood-brain barrier. Penetration ability can be represented extrathermodynamically by log P.

Structure-activity correlations have been performed recently using several mathematical models. 1-3 Of special interest is the method of Hansch and coworkers using computerized regression analysis to find those physicochemical parameters of importance which determine the change in biological activities when substituents are changed in a set of congeneric drugs. According to Hansch3 eq 1 can be used as a first approximation: log l / C

=

a X log A

+ b X log k, + c

(1)

C is the molar drug concentration necessary to cause a constant equivalent biological response (e.g., LDbo, EDboetc.). A is the probability of the molecules reaching their sites of action in a given time interval and k , is the rate-determining critical step at the receptor site, which is responsible for the visible biological response. The k , values of a set of congeneric drugs should parallel their act’ivities in the receptor compartment4 (“receptor activities.”) Following these lines, even with gross structural differences in a set of congeneric drugs, quite good correlations can be obtained.b,6 However one has to keep in mind that the scope of eq 1 has certain limits. Correlation studies with highly stereospecific acting drugs’J have shown, that in this case only minor structural changes can be performed on the basic structure of a set of congeneric drugs, otherwise it is impossible to account for the corresponding changes in the physicochemical properties in a quantitative manner. In addition the structural changes must be perfwmed in a part of the basic molecular entity where a similar mode of binding to the receptor for a11 congeners can be expected. One of the main rwsons for this difficulty is that one has to assume a more or less flexible nature for the receptor. Therefore the st’eric interactions between flexible receptor and flexible drug can be very complex. This aspect n-as especially pointed out by Portogheseg in the case of morphine-like analgetics. These findings are supported by a paper of Casy and (1) S. M. Free, J r . , a n d J. Ti. Wilson, J . M e d . Chem., ‘7, 395 (1964). (2) K . BoEek, J. Kopechj., M. Krivucovd, a n d D. Vlachovb, Ezperentiu, 20, 667 (1964). (3) C. Hansch a n d T. Fujita, J . Amrr. Chem. Soc., 88, 1616 (19fi3). ( 4 ) J. *V. v a n Rossum, Adaan. Drug Res., 5, 193 (1966), ( 5 ) C. Hansoh, Farm. Ed. Sci., 2S, 2!X3 (1968). ( 6 ) C . Hansch, A . R.Steward, S. SI. .\nderson, a n d I). Iled, Chem.. 8, 609 (1965).

Parulkar’o on the different mode of binding of morphine-like analgetics of different configurations t o the analgetic-receptors. Unfortunately these authors had to base their structure-activity discussions regarding drug-receptor interactions on biological data obtained following peripheral application. However, these data are complex in nature. The structural change on the drug has as a result not only a change in the drug-receptor interactions but also changes in the rate of penetration through the blood-brain barrierl1-l3 and changes in the rates of metabolism and eliminati~n.’~Therefore Portoghese had to restrict his structure-activity discussion to the comparison of the effects of similar substituents on different sets of “analgesiophoric dmgs.”15,16 To overcome these difficulties we have measured the analgetic activity following intraventricular application. These activities should be better suited to represent drug-receptor interactions than activities obtained following intravenous application. I n addition one would expect to obtain quite useful information regarding the factors governing penetration through the blood-brain barrier in comparing analgetic activity following intravenous and intraventricular application, respectively. The analgetic activities of the compounds were determined in rabbits by means of the tooth-pulp test. A detailed description of the technical procedure for the intraventricular application and a more qualitative discussion of the penetration phenomenon is described elsewhere.” I t is the purpose of this paper to show that a properly chosen pharmacological test procedure can reveal quite useful structure-activity correlations i n spite of the complications which are involved with structure-activity studies of highly stereospecific acting drugs. 1;igure 1 shows the chemical structures of tlie (IO) .\. F. C a a y a n d .\. 1’. Parulkar. ibid., 12, 178 (lY69). (11) M. W.Bradburg a n d H. Dal-son, in “Absorption a n d Distrihiition

of

Drugs,” T. B. Minns, Ed., E. B S. Livingstone Ltd., Edinburgh and London. (12) A. Herz, H. .J. Teschemacher, A . Hofstetter, and H. Kurz, Znt. ./. Neuropharmacol., 4, 207 (1965). (13) R. Hess, H. J. Teschemacher, and A. Hera. JYaunyn Schmiedeherga Arch. E z p . Pathol. Pharmakol., 261, 469 (1968). (14) E . J . hri*ns, Fortachr.-Aaneimitt., 10, 429-529 (1966). (15) P. S. Portoghese. J . Pharm. Sci., 64, 1077 (196,j). (16) 1’. S. Portoglieae. .\, A , hfikliail, and II. J . IiiigferLerg, J . M e d . Chem.. 11, 219 (1968). (17) B. r o n Cube, H . J. Teschemacher. A . Herz, a n d R. Hess, Naunyn Schmiedebergs Arch. E z p . Pathol. Pharniukol., 266, 455 (1970).

'H 'ilO

H

H

R PH INE

DihYDRCMCRPHINE 25

H0'

H NORMORPHINE 12

H HYDROMCRPHONE 49

HO' LEVCRPHANCLE ETCRPH I NE

4

iO?5

/CH3

'H5

CH3

P

CH3

TABLI.:

I,OG P

I

O F XISIMUM ACTIVITY( T o )A N D (EL),,) NECtSS.IRY TO OBT.lIN .i COiYST.INT ANTINOCICEPTIVE ACTION FOLLOU I S G I ~ T R A Y I SA ON U D INTR.\\ S IIICUL.iK *4PPLIC.\TIOX

Y . i L u m O F THI:

T m m D ASALGESICSA N D LIonrlzsw

EFFI,,CTIYI: IIOSES

7;, lor: 1"'

('ompmmd

0.15 1.29 0.2.5 --2,04 1.63 - 3 . (J(i -4.00

EDi,.C

(rd 3.6

(min)b

(log 1/ C ) i , d (mole8 PY)

EDirentrflog ( 1 , ' c ) ~ V e u t P TO

(min)@

(rg)

(moles, f i x )

-lox Olmdh

(C,,.,,,, CidCalcd'

I A ~

2.04 2.8 0.6 2.82 -0.780 -0.650 0.110 38.4 0.82 4,0 6.6 1.58 -0,760 -0.776 Fentanyl 0.016 4.S '72.0 0.6'7 8.5 24.0 1.15 -0.480 -0.670 0,190 Fury1 derivative 1,evorp haiiole 8.0 1680.0 -0.81 42.0 150.0 0.23 -1.040 -1.124 0 084 7.6 3i50,U -1.09 11.0 610.0 -0.30 -0.790 -0.85!1 0,069 IIethadoiie hetubemidolie 4,9 4500.0 -- 1 .26 59, 0 205.0 0.03 -1.340 -1.632 0.292 IIydrumorphot ie s.5 6600.0 - - I .36 123.0 12.4 1.38 -2.740 -2.26'7 0.4i3 I'ethidine (1, 53 2.8 10600. 0 -. 1 .62 2.5 1230.0 -.0.71 -0.920 -0.679 0,241 1)ihydionioiphitie -5.00 12.7 13650.0 -1.68 91 . O 24.0 1.08 -2.760 -3.118 0 ,358 .\lorphilie -5.00 14.4 18600.0 -1.81 105.0 20.5 1.14 -2,950 -3.118 0.168 - 5.00 22.0 1,50000.0 - 2 . T4 41 , 0 51.0 0.i2 -3.470 -3.118 Somiorphirie 0,352 Logarithni.. of the partitioil cw+Tic.ieritsof the compoutids between heptane aiid phosphate buffer, p H 7.4. Moments of maximum activity from the dose-activity ciirveh for iv ED18 (increase of threshold to 10 mA). c Iv threshold dose (E&) in the moment of maxid Logarithm of the reciprocal iv threshold dose ill mmoler. miim avtivity. e JIoments of maximum activity from the dose-a(-tivity uii'veh for iveritr EDlo. Iveiitr threshold dose (EDlo) in the moment of maximum activity. Logarithm of the reciprocal iventr threyhold do;ie in mmoles. Log (Ci,.e,ltr,'Ci,.)values observed. Calculated log (Ci,,,,,,/Ci,) using eq 10. 1.2t oi,phitie

3.2 4.5

differs markedly from one compound to the other. I n order to investigate the influerice of lipophilicity on the moment of maximum activity me have derived eq 2 from the data in Table I. log tt

=

=

+

-0.2111 log P 0.933 (2) 0.322 I ? = 0.77.5 1' = O.SS0

=

11 s

Keeping in mind that 7'" i y a complex function of the rate of penetration to aiid away from the receptor site the correlation as depicted in eq 2 is reasonably good. Of the variance in the log (To) values 77% can be explained by a linear regression with log P. The negative slope of this equation indicates that with an increase in the lipophilicity of the compounds the moment of maximum activity takes place earlier. Two examples illustrate that this effect is quite pronounced. The highly lipophilic compound pethidine reaches its niomeiit of maximum activity 2.5 min after injection. 0 1 1 the other hand, morphine, a very polar compound, needs 103 min to exert its maximum analgetic effect. Thew rebults are in accord with the assumption that in the case of intraventricular application there also existb a lipophilic barrier for the polar drugs between the locus of application (ventricular TI all) :ind the receptor regiori. I n the cahe of the w r y lipophilic analgetics thebe compounds penetrate very rapidly to the site arid therefore the monieiit of maximum activity takes place very t>arl\. The main reasoii for this phenomenon seemb to bc their rapid elimination from the receptor region ULU the capillarj- syhtem to other parts of the body, because the lipophilic drugs can penetrate the blood-brain barrier rapidly from both directions The situation is different in the case of the polar drugs. The very polar analgetics need more than 1 hr to accumulate in the reoeptor region. 111addition they can not penetrate the blood-brain b:rrrier rapidly i n order to be distributed by the bloodstream. Therefore, these drugs product! a long-lasting analgetic effect following intraventricular application. I n order to eliminate the influence of the rate of penetration from the ventricular wall t o the reccytor region 011 the aiialgctic activity of the &ugh we have determined the threshold doses to produce a measurable aritiriociceptive activity a t the moment of max-

imum drug coricentratiori in the receptor region. These moments of maximum drug concentrations necessary to reach the threshold to produce an analgetic effect" (EDIlentrvalues in Table I) are identical with the T O entr values in Table I. One could expect that these E D values, a t least a t a first approximation, are independent of the lipophilicity of the compounds and therefore should be suitable to represent the relative activity of the drugs a t their sites. A regression study coiifirms these conclusions.

+ 0.777

t1

=

log (1, C)l\entr= -0.030 log P 11 s = 1.033 1.2 = 0.007

I

=

o.os33

(3)

I n eq 3 log P is expected to represent log A (penetrution) extrathermodynamically. (l/C)l\ entr is a measure for the efficiency of a compound following intraventricular application. Addition of a term in (log P ) *gives no improvement of the correlation. The value of this equation indicates t h a t only 0.7% of the variance in the log (l/C)lventrvalues could be explained by x linear regression with log P . This argues strongly against any important dependence of the log ( l/C)lTentr values on this parameter. This conclusion is supported by the very small blope of eq 3 . In determining the threshold doses in the moments of maximum drug colicentration in the receptor region we indeed have eliminated the influence of the penetration phenomeriori on the EDllelltrvalues. Therefore, one can conclude t h a t thebe ED,, values should parallel the "rcccptor activities" of the corresponding analgetich. I n tcrms of mathematics eq 4 is the result of these conbideratioiih. ?n2

log (1 'C),lentr = b

x

log

11,

(receptor compartment)

+d

(4)

From mother point of view the veri low r 2 value of eq 3 is interesting. It is evident that in this case of highly stereospecific acting drugs the relative receptor activities of this set of analgetics are not log P controlled and that otlicr fnctorb govern the affinity of the drugs to their sites. This result corresponds nicely with other correlation studiess where hydrophobic con-

Journal oj Xedicinal Chemistry, 1970, 1‘01. i j, .Yo. 6 805

parallels very well with the lipophilic character of the four analgeticis. Of special interest is the good agreement of the slopes of eq 9 and 11. This reflects a close parallelism between log ( C b r a i n l C p l a s m a ) and log ( C l v e n t r l Ci,.) . This is additional support for the earlier findings that log ( C l r e n t r / C i v ) is a measure of the different capability of the analgetics to penetrate the blood-brain barrier. TABLE I1 I~.WIO.~CITVELY L.IRI:LI:DAS.\LG~;TICS\\ ITH PHI.SIC.\L Cossraxm log (Cbrain/

(Cbrain?

Compound

Cplaama)

C,I”

lloiphiiie 1)ihydromorphine Feiit any1 Etorphitie

0.046 0.053 10.58 8.69

-1.34 -1.28 1.02 0.94

log P

-5.00 -5.00 1.29 0.15

The parallelism between the intraventricular activity and the receptor activity leads to some qualitative conclusions regarding the drug-receptor interactions. Figure 1 makes possible the comparison of the chemical structures of the analgetics with their relative intrinsic activities. It is interesting to note that the polar compounds like hydromorphine, dihydromorphine, morphine, and normorphine have a greater activity a t the receptor than the synthetic derivatives methadone, Betobemidone, and pethidine. Polar functions, like hydroxy, ether, or keto groups seem to be favorable for

specific drug-receptor interactions. This is in accord with the findings of Porthogeseg that the introduction of an OH group in an analgetically active molecule can enhance analgetic activity and simultaneously change the mode of binding. Therefore, the high activity of the nonpolar synthetic analgetics follox-ing intravenous application can easily be explained by a good penetration of these compounds through the blood-brain barrier to the reaction site and seems not to be due to especially favorable drug-receptor interactions. I n etorphine, the most active compound in this set of analgetics, both favorable properties are combined within one molecule. The results of this work underline the importance of the passive penetration of drugs through the bloodbrain barrier. S o criterion %as found which would be in accord with a special importance of an active transport2’ through this barrier for the studied analgetics. I n addition, the statistical analysis of the pharmacological results has shown, that the efficiencies of the drugs following intraventricular application parallel largely their receptor activities. Acknowledgments.-This u ork was supported by the World Health Organization. The authors are also indebted to Professor C. Hansch of Pomona College, Claremont, Calif., for several heipful cQmmentson this xvork. E. Kutter wishes to express his gratitude to Professor H. llachleidt for supporting this kind of research. (21) J. T. Scrafani a n d C . C . Hug, Ezochem. Phormacol., 17, 1557 (1968).

Optical Isomers of Miscellaneous Strong Analgetics EVEHETTE

L.;\1.4Y’a

A S D J I I K I O TAKED.4Ib

1,uboralo~yof Chcnizstry, .\’atzonal Inslitulc o j a l r t h r i l t ~und JlduboItc Uisrascs, ,Yatzonal Instztutes of Health, Bethtsda, Jlargland 20014

Rtceived A p r d 10, 1970 Oplical isomers of a-3,9-diethy1-2’-methoxy-(la), a-2,.i-dimethyl-D-ethyl-2’-hydro.uy(Id ), atid 2’-hyd!n-;y2-methyl-6,7-benzomorphans (IC) arid of .i-(m-hydroxyphenyl)-Z-methylmorphan(2) have been prepared arid compared with parent racemates in analgetic activity, physical dependence capacity, and antsgonistic behavior. Racemate IC, (+)-lc, (-)-IC and ( - )-2 have morphine-like analgetic and nalorphine-like antagonistic action

I n continuation of our rosearch designed to effect fnvorable separation of morphine-like effects by optical resolution,2 the antipodec of a-.5,9-diethy1-2’-methoxy(la), a-2,5-dimethyl-9-ethyl-2’-hydroxy-(ld), and 2’hydroxy- (IC) 2-methy1-6,7-benzomorphans and of 5(m-hydrosyphenyl)-2-methylmorphan(2) have been prcp:wed. Compounds (+)- and (-)-la were obtained by CH-N, methylation of (+)- and (-)-lb. Optical resolution of IC,Id, and 2 was effected with (+)-3-bromoS-camphorsulfonic acid ammonium salts,3 d-10-camphorsulfonic acid, and cl-mandelic acid, respectively. (1) (a) T o v h o m inquiries should be addressed: (b) Visiting Fellow from Tokyo. 1968-196Y. ( 2 ) (a) J. €1. .Leer, A. E. Jacobson, and E. L. l f a y , J . .Wed. Chem.. 12, 288 ( 1 9 6 9 ) : ( b ) E. L. M a y a n d S . E. E d d y , ibid.. 9, 581 (1966). ( 3 ) F r o m Aldricii Chemical Company a s d-a-bromocrtmphor-rr-sulfonic acid.

Rd la.R=Me: Rl=RL=Et b, R = H; R, = R,= Et C, R = R i = R , = H d, R = H ; R , = Me; R, = Et

Hd 2

Pharmacology. In Table I are given analgetic activities as determined in the mouse hot plate method14 and physical dependence capacities and antagonistic (4) (a) N . B.E d d y a n d D. Leimbach. J . Pharmncol. E x p . T h e r . , 107, 385 (1953); (b) A. E. Jacobson a n d E. L. M a y . J. M e d . Chem., 8, 563 (1965).