You are on page 1of 8

A NEW CONCEPT

OF NARCOTIC
ANALGESICS
ACTION

September 1965

609

A New Concept on the Mode of Interaction of Narcotic Analgesics with Receptors


PHILIPS. PORTOGHESE
Department of Pharmaceutical Chemistry, College of Pharmacy, Cniuersity of Minnesota, Minneapolis, Minnesota 55455
Received March 18, 1965
Existing data on the relationship between molecular geometry and narcotic analgesic activity has been interpreted in terms of differing modes of analgesic-receptor interactions. A method for detecting similarities or
differences in molecular binding modes has been accomplished by comparing the variation of activity in two or
more different series of compounds when identical changes of the N-substituent are made. A parallel change in
activity is suggestive of similar binding modes while a nonparallel relationship is indicative of dissimilar interactions. This method has been extended quantitatively to demonstrate the existence of a linear free-energy relationship when binding modes are similar. The value of the slopes from such a relationship are approximately
unity. This suggests that, when compounds in two different series interact with the receptors in a similar fashion,
identical substituents contribute to the binding by a similar mechanism. The concept or" three-point interaction
arid possible pitfalls inherent in its applicability to the interaction of analgesics with receptors is discussed.

Although it is generally agreed that narcotic analgesics exert their action by interacting with specific
receptors, there is little known about the nature of such
receptors. One approach to t'he investigation of receptor t'opography is the correlat'ion of absolute stereochemistry with analgesic potency. In this connection,
the configuration of a variety of optically active structures has been determined.l-7 A rather perplexing
aspect of this problem is that the more active
enantiomers of analgesic molecules, having degrees of
structural similarity and a common asymmetric
center, are not all stereochemically related.
Another phenomenon which has not been adequately
rationalized has a bearing on the pharmacophoric conformation(s) of mobile analgesics. It appears that at
least two conformations of the phenylpiperidine moiety
contained in various compounds are capable of exerting
high analgesic potency. For example, the conformationally restricted compound I 8 is as pot'ent as morphine even though the aromatic ring is fixed in the equatorial position. This is in contrast to the phenyl-

HO

N-Me

piperidine moiety of morphine, where the aroinat'ic


group is constrained in t'he axial conformation. Since
I is a c h e , it is conceivable that other phenylpiperidine
compounds exert their action in the equatorial conformation when serious nonbonded interactions preclude
a significant amount of axial conformer. For example,
the change in free energy on going from the equatorial
(IIe) to axial conformer (IIa) in the pot'ent analgesic,
~)roniedol,gis in excess of 7 kcal./mole. Since it
seems unlikely that such a large energy barrier can
(1) A . H. Beckett in "Progress in Drug Research," Vol. 1, E. Juoker, Ed.,
Birkhauser Verlag, Basel, 1959, p. 455.
(2) A. H. Beckett, G . Kirk, and R. Thomas, J. Chem. Soc., 1386 (1962).
(3) &I. Nakazaki, I. Mite, and N. Toshioka, Bull. Chem. SOC.Japan, 36,
161 (1963).
(4) H. R. Sullivan, J. R. Beck, and A. Pohland, J. Org. Chem., 28, 2381
(1963).
(5) A. Rheiner, Jr., and A. Brossi, Ezperientia, 2 0 , 488 (1964).
(6) P. S. Portoghese and D. L. Larson, J . Pharm. Sci., 63, 302 (1964).
(7) P. S. Portoghese. J . .\fed. Chem., 3, 147 (1965).
(8) N . B. Eddy, Chem. I n d . (London), 1462 (1959).
(9) I. N. Naaarov, N. S. Prostakov, and N . I. Shvetsov, J . Gen. Chem.

CSYI1, 24, 2TY8 (1956).

be surmounted as a consequence of an analgesicreceptor interaction, IIe possibly could be a pharniacophoric conformation.

I
EtCOO

Me
IIe

I+

IIa

In view of the absence of a coherent interpretation of


steric factors and other phenomena as related to the
mode of interaction of molecules with analgesic receptors, existing structure-activity relationships have been
reinterpreted in the light of newer developments in the
areas of biological and medicinal chemistry.
The Analgesic-Receptor Interaction.-There
are
three possible modes of interaction of narcotic analgesics with receptors which merit consideration. Different analgesic molecules may exert their action by:
(1) interacting with a single species of receptors, (2)
interacting with two or more common species of receptors, and/or (3) interacting with different species of receptors. Since a variety of structures possessing narcotic analgesic activity has been shown to be antagonized by nalorphineJ8~'0
it is conceivable that the inode
of interaction in certain cases involves either of the first
two possibilities. Quantitative studies" with analgesic antagonists may provide greater insight into this
problem. Recent experiments along these lines have
indicated that nalorphine does not necessarily antagonize different analgesics to the same degree.12 Although
quantitative experiments are presently one of the best
means of investigating possible modes of interaction of
narcotic analgesics with receptors, such studies are open
to question since it is not known with certainty whether
nalorphine is purely a competitive antagonist. Moreover, since the effects of certain highly potent analg e s i c ~ do
' ~ not appear to be completely antagonized by
nalorphine, it seems likely that there may be several
different species of narcotic analgesic receptors. It ap(10) L. A. Woods. Pharmacol. Rev., 8, 175 (1956).
(11) L. Grumbitch and H. I. Chernov, Federatzon Proc., 2 0 , 165 (1961).
(12) (a) B. hI. Cox, Brzt. J . Pharmacol., 22,289 (1964): (b) R. >Veatch.
'I
T K Adler, and E. L Way, J . Pharmacol. Ezptl. Therap 146, 11 (1964).
(13) R W. Bentley, A. L. A. Boura, A . E. Fitzgraeld. D G . Hardy, A.
AIcCoubre), 31. L. Aikman, and R. E. Lister, 'Vature, 206, 102 (1965).

010

PHlLlP

d.

1'0111 0 G l l Y b b

\ 01. h

till

located OII these iiiacroiiiolecules iiiay promote a rearrangement of the receptor substance which contributes
to the binding process.'j The resulting macroinolecular
perturbationl6 ultimately may lead to an analgesic
effect. It is quite possible that several different type5
of complexes could give rise to the pharmacological
cff ec t. If the iiiacroniolecular rearrangement does 110 t
lead to analgesia, then the conipouiid may act as a
competitive antagonist (ie., in rats arid mice). Inactivity can also be caused by a lack of affinityz4for
the receptor due to unfavorable hydrophobic interaction. It is likely that the value of the ED50 is dependent both on the ability of a niolecule to induce a
proper macromolecular perturbation16 arid on the
affinity for t,he receptor.

Configurational Selectivity of Analgesic Receptors.'l'he ttereoselectivity of analgesic receptors toward a


variety of narcotic arialgeiics is amply documented
(Table I). 111 all caies activity is distributed uiiequally in eriaiitiomeric molecules. I t is significant that
the more active enantiomers are not all configurationally related. For example, methadone (1) and thiambutene (8) are related to (R)-alanine while the carbethoxy analog of methadone (3), diampromide (lo),
and a-methadol (4) are of opposite configuration.
The aforementioned cases are quite obvious because of
the inversion in configurational selectivity of the
analgesic receptors. Quite often, however, the selectivity of the receptors, though not inverted, is changed
considerably when groulls in a molecule are altered
(see Table I).
In conipouiids structurally related to methadone,
alter atioii of the ketonic carbonyl gioup causes profound changes in the enantiomeric potency ratio. For
iristance, reduction of methadone to a-niethadol (4)
followed by transforination to the acetate derivative
(6) vaii5es the (R), (S), aiid ( R ) i\oniers of 1, 4, aiid
6, rehl)ecatively, to be more artive. A mere cehaiige in
relative itereocdhemidry rail also iiiflueiice the anti1)odal discriiiiiiiatory power of the receptois. Thus,
the more active enantiomer of a-methadol (4) i q in the
(S) series while the more potent optlcal antipode of
p-niethadol ( 5 ) posseises the (6R)configuration.
T'ariations in the discriniinatory aptitude of analgesic
receptors toward various enantiomers can be caused by
several factors which may be important singly or in
combination. These are as follow: (1) Differences in
the course of binding of stereochemically related compounds to a single species of receptors. Figure 1 illustrates schematically how two different niolecules of
opposite configuration may interact with an analgesic
receptor. Such differences in the binding mode will be
manifested by an inversion of the configurational selectivity of the receptor. Dipole3 conceivably can be sites
which are hydrogen bondirig protori donor5 (X) or acceptors (Y).. i s illu\trated, interaction of methadone
with an analgesic receptor may involve hydrogen bonding of the ketonic carbonyl group by X, whereas with
cy-methanol, OH.. .Tniay occur. Hence, alteration of
a polar group on an analgesic molecule can afford a
compound which may interact with dipolar sites and
hydrophohic areas that differ from t h o w involved i n thc
binding of the unaltered structure.6 ( 2 ) Differences in
(24) E. .J. .\riens, ".\Iolecular Pliarmaculoyy
Inc., Nexv Tork, K. Y.,1864, p. 232.

"

Vol. 1, Academic Press

Q
(6R)

(6s)

Figure l.--Sn illustration of how different polar groups in


analgesic molecules may came inversion in the configiirational
selectivity of an analgesic receptor. Hydrogel1 bonding protoii
donor a i d acceptor dipoles are denoted by s a i d y, respectively.
The anionic site is represeut,ed by - .

the partitioniiig of coiifiguratioiially related niolecules


011 two or more coniinoii species of receptors. Hence, if
two analgesics are partitioned iii different proportioils
on two or more comnioii receptors which do not have the
same stereoselectivity, an apparelit change in the stereoselectivity of the receptors will result. (3) Binding of
stereochemically related compounds to different species
of receptors having dissimilar configurational selectivity.
I t may be possible to make a distinction between the
first two cases aiid the third possibility by studies with
analgesic antagonists. However, it niay prove difficult
to distinguish between (1) aiid ( 2 ) by such a method.
The situation niay be more complex if there are combinations of case (2) and (3).
Inasmuch as no quantitative studies with analgesic
antagonists have beeii carried out oii compounds such
as a-methadol aiid the basic anilide analgesics [whose
more active eiiantionieters are all stereochemic~allyrelated to (S)-alanine], firm coiiclusioiis rcgardiiig thc
mode of iiiteractioii of these compounds with aiialgesic
receptors cannot presently be made.
It is of int,erest that. inversion in the corifiguratioiial
selectivity of muscarinic rece1)tors toward muscaroiie2j
has been suggested to be caused by biiidiiig of muscarine
and inuscarone with different portions of the receptor.26
Similar behavior has been displayed by a-chymotrypsin
in the hydrolysis of various asymmetric esters. 27,28
These observations have likewise been interpreted as
being due tjo differences in the interaction of substrates
with the active
Variation of N-Substituents as a Means of Comparing Modes of Drug-Receptor Interaction.-It is known
that variation in the substituent attached to the basic
nitrogen of ari a~ialgesiophore~~
can exert a profound
influence 011 activity.3z For example. replacing a methyl
with a ciriiiamyl group in .l-l)heiiyl-.l-propio~ioxypiperidine results in a hundredfold increase in analgesic activ( 2 5 ) P. G. Waser. Pharmacol. Rev., 13, 465 (1961).
(26) B. Bellean and J. Puranen, J . ,Wed. Chem.. 6 , 326 (1963).
(27) G . E. Hein, R. A. IIcGriff, and C. Niemann. J . A m . Chem. S o c . , 82,
1830 (1960).
(28) S . G. Cohen and R . P.Weinstein, ibid.,86, 5326 (1964).
(29) I. l i . JVilson an(l 13. 1'. Crlanger, i/d.,
82, 6422 (1960).
(30) E. S. .\\vad, 11. Seuratti, and H. S. Hartley, .I. B i d . Clrem., 235,
PC35 (1960).
(31) Analyesiopliore is defined as the analgesic molecule less the substituent
on the Iiasic nitrogen.
(32) P. h..Tanssen and N . 15. Eddy, J . 3 f t d . ]'harm, Chem., 2 , 31 (1960).

I I

I 0

Ill

C,Hd CH, j4
CsHsCH=CHCHs

IO

.;

.I1
(I

:inalgesic activity relative to meperidine: a value of 10 signifies the cornporiiid is IO t,imes1110reput.ent,than the reference cwnpound.
L'nless otherwise specified cornpounds were administered subcutaneously. h I<. 1%.Thorp and E. Walton, J . Chem. Soc., 559 (1945).
B. Elpern, L. N . Gardner, and L. Grumbach, J . Am. Chem. SOC.,79, 1961 (1937). B. Elpern, 1s'. Wett,erau, P. Carbateas, and L.
Grumbach, ibid., 80, 4916 (1958). e A4nalgesicactivit,y relative to morphine in the sitme sense as in a. f S . B. Eddy, H. Besendorf,
I,. B. Mellett and I,. A. Woods "Progress in Drug Re:tnd B. Pellmorit, Bull. Narcotics, C,. S.D e p f . Social Affairs, 10, 23 (1958).
T,J. H. h g e r and E. L. May, J . O i g . Chem., 25, 984 (1960).
search," Vol. 5 , E. Jucker, Ed., Birkhauser Yerlag, Basel, 1963, p. 157.
Hot plat,e method using mice. 3' Rat t,ail radiant, heat method. t Adrnitiist,eredintraperitoneally.
Q

ity. Lln identical substituent c~haiigoi r i the niorphiIic1


series results in loss of activity."~ The fact that K-ciiiiinniyliioi.morphirie is not ai1 antagonist 3 3 suggests thai
i t \ iriactivity iii:tyJ iri part, be due iiiaiiily to lack oi
affiiiity for the receptors.
plausible tlxplantttion foi
t liih noriyarallel relationship is t h:Lt the ciririaniyl substituents attached to the aforementioned analgesio1)horcs are in different physiocheniical environnient*
on tlir i e c q t o r s . This nould iiieiiii that the binding
niodc of the arialgesiophore 111 phenylpiperidine coni1)oriiid.: i5 different from that in the morphine series.
OF greutcr +pifirailre is the report that S-allyliioriiie1)eridiiie (T:hlc TI, 1. 13 = allyl), unlike nalorp l i i i i c ' (4, I< = :dlyl) n IiicnEi behaves ns :t iiiorphiii(~
:mtngoiiiit :~iidi q virliially iiinctivc ( L e . . in mice or
rat-). I\ iwt a11 aiititgoriiht j4 mid po s.:es a r t ivity conil):~r:ih[(~
to iiiq)(>ridinc. Thi\ yuggest * 1 hat tlic S;dIyl dciivxtivv- of 1 :urd 4 (Tnhlc 11) are coiiiplexiiig
wit 11 t h c :~iialge.:ic receiitor. 1))- diwmilar niodcs.
Siriw thc :iffinity of iialorpliiiie i o i . t hv receptors has
iiof ~ W I Idilriiiiiihrd, t h e great variance in the pharnlacwlogic4:rl cffwt niay be reflective of' a difference in bindiiig niodc. SuraIi dificrt3iicey 111 tlic mode of interactioii
r<:iri bo :11 tiihuted t o u i y of ilie factors which have been
d i * c ~ u i wiIi i coiiiier.t ioii with tlic inversion of configurnf ioii:iI + 1 ( ~ * 1 ivity of niinlge*ic~rerel)iori.

.ill,llougll ;ti1 :IIIaIgehlo~~IlorcI11 Oil(' \rl'l('h of ('01111)oiirids m:iy iiiterart n-jtli aualgesic. receptors in :i ni:tiii i v r which I \ di->iniilar to :I differe111 analgesiophorc.
cwitaiiiiiig :ill idcntical S-substituent, i t is rea*oiiablv

u n i r th:tt, regardlehs of the binding Iilode, :ill


;iii:ilgesics (iiicnludiiig compct itive aiitagonists) :ire i i i v o l v d in ionic. bonding with an identicma1 anionic 4 l v
I Iic ailionic - i t ( > niay bc eiiviiaged as :t i)i\-otal 1 ) o i r i l

r ,

il'igurc. 2 ) . Thi, doe. irot ner ar1ly 1111ply 1lint 360


*urrouiiding t hr :iiiioiiiv sit(. i ivailable for bindirig ;
11 iiiuy be only ;L sniall fractioii of this area. Thca
of an analgesioyhore determine the
1)osition of'biiidirig of :LII X-substituent to a siiigk q p wo 01 iiiorc (+oi~iiiioii
y)ecies, and/or to differeiil
)f i ( ~ e p t o r \ . I t is alho cwiiceivable that the S 111 iii:iy l i v cxpitblc of modifying the hindiiig
tlic aiialge.iopliorc.. I t this is t hc caw, c:irli
iii a ierjchs rontaiiiiiig tlic vinic analgesiopborcb
rent S-subititueiii- \vould exert its actioii VI'U
>lightly different binding in ode^.
When diff ercwt arialgesiopho~esinteract with rweptors by siniilzir modes, t hcii iderit ical K-substituents
should contribute to the pharmacological effect by the
-aiiie nier~1iarii.ni. If each compound in one series of
analgesics i;. iaoriipared with :I member in a second series
wliirh 1135 :Ldiff erriit aimlgesiophore and :in iderit ical
S-yubstil ueiii , l l i r i i it niay be possible to determiric.
\L 1 ~lrer
~ 1t ]I(, 1)iiiding inodes of tho :~nalgesiopliores(attacalml to tlic saiiiv S-sul)s;titii~~titj
are siniilnr or iliiit(8

September 1965

Y E W C O N C E P T OF N A R C O T I C h S L G E S I C S

different. Thus, if identical changes in the K-substituent in tmo series of compounds produces parallel
changes in potency, the mode of binding of the analgesiophores should be similar. A relationship which has
a high degree of nonparallel activity is suggestive of
binding modes which are quite dissimilar.
The activity of identically N-substituted analgesiophores has been compiled in Table 11. I t can be seen
that the meperidine series (1) and identically substituted reversed esters (series 2 and 3) exhibit parallel changes in activity. Comparison of the aforementioned series with identically substituted conipounds related to morphine (4,5, and 6) show no parallel
relationship. On the other hand, derivative, of morphine (4), niorphinan (5), and benzoniorphan (6) do
display parallel changes in potency.
The correlations in Table I1 suggest that the analgesiophores (containing identical N-substituents) in
series 1, 2, and 3 are binding to the receptor by similar
modes and that an analogous situation exists among
the structures related to morphine (series 4, 5 , and 6 ) .
Furthermore, the high degree of nonparallel activity
between phenylpiperidine derivatives (series 1, 2, and
3) and morphine-like compounds (series 4, 5, and 6)
strongly suggests that the binding modes of identically
substiuted compounds in the former series may be quite
different from the latter.
If identically substituted compounds in two different
series are interacting with receptors in a similar manner, then the quantitative contribution of various substituents to the analgesic effect should produce, under
steady state conditions, proportionate variations of
activity in both series. If a point is plotted whose
abscissa is the logarithm of the activity for the appropriately substituted compound in one series and whose
ordinate is the logarithm of the activity in an identically
substituted compound in the second series, the resultant
points should describe a straight line. Such a proportionality is known as a linear free-energy relati0nship.3~
Linearity would be a consequence of similar binding
modes of two different analgesics containing identical
N-substituents and would be due to the fact that the
change in A S = 0 or that variations in A S are proportional to changes in A H . If the modes of interaction
are quite dissimilar, then identical X-substituents
attached to two different analgesiophores would experience different physicochemical environments on
the receptors. Such a situation would give rise to nonproportionate differences in A S and AH which would
be manifested by a nonparallel relationship and a
scatter of points. If the binding modes of two different analgesiophores containing identical substituents
are similar, the slope of the regression should be in the
vicinity of 1. This, of course, is dependent on the
assumption that, prior to the drug-receptor interaction, identical substituents on two different analgesiophores will affect the biodistribution of the compounds in a similar fashion. This assumption is quite
reasonable in view of the successful application of substituent constants36 for the purpose of predicting drug
availability a t the site of action. Thus, parallelisni
or nonparallelism of potency is interpreted as a manifestation of events at the receptor level. When the
(35) H. H. Jaffe, Chem. Rev., 63, 191 (1953).
(36) C. Hanwh and T . Fujita, .I. A m . Chem. SOT.,86, 1618 (1964)

ACTION

G13

Figure 2.--A schematir illlistration of two different positions


of binding to a receptor. The protonated amine iiitrogeii
arid the square denote-: an S--:uhititilent.
is represented by
The anionic sit,e lies directly beneath

+.

mode of binding is not similar, scattering of points may


make the value of the slope indeterminate. If
several dissimilar binding modes exist for each
of two different analgesiophores having the same
N-substituent, depending on the relative population of the modes, degrees of point scattering may
be observed. As differencesin the modes of interaction
increase, the standard deviation of points which comprise the regression will become quite large and result
ultimately in a nonparallel relationship.
Janssen and Eddy3 have reported on the analgesic
activity of several series (Table 11, series 1, 2, and 3)
of identically K-substituted phenylpiperidine derivatives. Since their data are reported with well defined confidence limits, it seemed that they could be
used to demonstrate the existence of a linear freeenergy relationship. An additional advantage was
that regressions obtained from the data of Janssen could
be compared with those of Eddy.
Although the least-square regressions obtained
from both Janssens and Eddys data (Table 111)
represent a first approximation due to the limited
number of compounds tested, it can be seen that the
values of the slopes and the linear correlation coefficients are fairly close to unity. This is consistent,
in a way which is more exacting than qualitative
examination, with the idea that different analgesiophores containing the same substituent are interacting
with the analgesic receptors in a similar manner. It is
evident that the standard error (S.E.) and the standard
deviation (S.D.) for each of two regressions (which
possess identical analgesiophores) obtained from two
different sources are in good agreement. A plot of the
logarithm of the activity (pM/kg.) for series 1 and
series 2 is shown in Figure 3.
Since S.D. represents a measure of point scatter it
should give an indication of the degree of similarity
in the binding of different analgesiophores to the receptors. If the data in Table I11 are significant, the
substantially higher values of S.D. in regressions de(37) P. A. J. Janssen, Synthetic dnalgesics, P a r t I , Pergamon Press
I n r . , New P o r k , N.Y., 1960, p 83.

September 1965

615

N E W COKCEPT O F XARCOTIC ASilLGESTCS h C T I 0 N

TABLE
IV

RELATED
TO NETHADONE
RELATIYE
ANALGESIC
ACTIVITY" OF STRUCTURES
C(C4H3S)z C ~ H S N C O E ~

(I

112
1

R3

RZ = H ; R3 = COEt
= l k ;113 = COEt

3, RZ = Me; R3
4, R * = 11e; 113

2, 1 1 2

CH2
I

( C6Hj)2kCH2CHNRR1

1,

CH
=
=

CHNRR'

CHKRR~

4Ip

11e

S02Et

mam

6'

n
1.4
0

<ti. 1
< 0 ,'3:3, <o ,250
0.7
0.5t
4,i j,7f.o
3 5
0
CiHs
4.0, l . 9 f * n
0
19,f 8 .5f@
<O.lQ
1
C3sO
7.0, 4.5f.Q
20, 5 . 4 f
6
0.20
5.5
0
CsHio
2.6,J 2.5JsQ
a Analgesic activity relative to meperidine; a value of 10 signifies the compound is 10 times more potent than the reference compound.
Data lwere obtained from ref. 3 i , Table V, p. 63.
0 Unless otherwise specified compounds were administered subcutaneously to rats.
e With the exception of the methyl benzyl analog, all compounds in this series
d Values were calculated from footnote j, Table I.
were inactive a t 25 mg./kg. We thank Dr. W7.Wright, Jr., for providing this information. f Average valne. 0 Mice were employed
Animal species not revealed. 3 Administered intraperitoneally.
as test animals. h Footnote I,Table I.

n-Pr
Allyl

n-Pr
Allyl

)Q

lic OH group, the mode of binding of phenolic and nonphenolic compounds may sometimes be quite dissimilar.
Data on phenolic and nonphenolic analgesics (Table
Y) show that replacing the methyl with a phenethyl
substituent causes an enhancement of analgesic activity
in phenolic morphinans and benzoniorphans, whereas
ronipounds containing no hydroxyl group exhibit a decrease in potency when an identicalchange in substituent
is made. Since phenolic and nonphenolic compounds
do not exhibit parallel changes in activity, it is probable
that the modes of binding are riot the same. Inasmuch as the aforementioned compounds are conformationally immobile, it appears that although niolecular shape is an important criterion, it is not the only
factor which determines the mode of binding. The
same factors which have been discussed in connection
with the variation in the enantiomeric potency ratio of
various analgesics (Table I) niay also be of importance
here. Interestingly, the azabicyclo [3.3.1Inonane derivatives (Table V) do show parallel changes in activity.
TABLEV
~ L N A L G E S I CACTIVITY
O F P H E K O L I C AXD

XONPHENOLIC
COMPOUNDS
Structure

%
' N-R

R2

H
H
OH
OH
OH
OH

Me
CHzCHsCsHS
Me
CHaCH2C6Hj
Me
CH2CH2C6Hs

Me

OH

CH?CH2CsHs

R'

Rz

EDaP

H
H
Me
J\Te
AIe
,Ire

H
H
Ale
Me

18.7
31.9
8 9
0 4
2 6
0 1

18.4

3 99

H
H

1 ) o v (rng./kg.) ndniiriistered siitmitarieoiisly to mice.

This could be coincidental or it may be indicative of


very similar binding modes. hdditional data would be
needed in order to distinguish between these alternatives.
The Concept of Three-Point Interaction as Related
to Analgesic Receptors.-The useful concept of three
point
has been employed in explaining
the antipodal selectivity of both enzymes and drug
receptors. This type of analysis, however, should be
applied with caution as it repie5ents an oversimplified
version of a relationship which can become quite coinplex. Thus, it has been obierved that the configurational selectivity of niu~carini(~
r e c e p t ~ r s and
~ ~ ~of~ *CY~ h y i i i o t r y p s i i i ~is~ -dependent
~~
on the characteristics
of polar and nonpolar groups which are a part of the
drug or substrate. Such pheiioniena have been explained in terms of disiinlilttr interaction with the
receptor or active site. I t is possible that an analogous situation exists for analgesic receptors, in that
different sites on the same receptors have the potential
for binding some common features of structurally dissimilar analgesiophores. Furthermore, if more than one
species of receptors having dissimilar steric requirements are interacting with analgesic molecules, the observed configurational selectivity will represent an
average value rather than a single type of analgesicreceptor interaction. Such differences in the mode of
binding of optically active narcotic analgesics are
usually reflected by variations in the enantiomeric
potency ratio. I n certain cases (Table I, 3, 4, 9, and 10)
this may involve an apparent inversion in the configurational selectivity of the receptors. This means that
stereochemical data cannot be used as the sole criterion
in determining whether binding modes are similar, although it may be employed in the detection of dissimilar modes of interaction. As has been pointed out previously in this communication, an zdentical absolute
stereochemistry of the more active analgesic enantiomers
does not necessarily indicate that these structures are interacting with analgesic receptors by similar inodes unless
variation of the N-substituent produces at least a roughly
qualitative parallel change in activity. lloreover, the
data presented in Tables I1 and IV suggest that it
may be misleading to assume, for example, that
methadone, meperidine, and morphine all exert their
nrtioii bv interacting with recq)tori in ver.y qiniilar

pliariiiacophoric conforiiiatioii-. 111 c~oiiil):irii~g1lit,


niode of binding of various narcotic :iii:ilge>ic<>
o t i c inusi
consider not only the geometric dkpo\iiion of 1)li:ir11ii~cophoric 1noietit.s but also take i i i t o :iwourit t h t , K J ~ P
played by other groups which are riot cieriiicd e
for activity. If quite a differctil patteiii of co
t ions to drug -receptor binding :ire ni:idc hy wriou- miifigural ioually i-rlnt(4 :tiialgSc~sic.t i i o l w i i l r - , I h ( J ( s o 1 I[alii-

-ioii-

r:ri
iloi

( h a i r 11 t rani t h(>threc-point

intrrnctiori conc*cpt f o
voiihgurat ioiial select irvily of i rwl)tor- i i i : i \
iiwr-wi,ily I I P \-:ihd

ioiializr

Acknowledgment. ?he author gratefully acknowledges thc wpport of this project by Public Health
Serviw (;rant TI3 05192, from the Satiorial T r i q t itiitr
(rf Stw r~)hgic.*i
1 I )iw

Stereoc.heniical Factors Related LO the PoLenc) of hticholinergic


Psychotominietic Drugs

A groiip of glycolate esters of heterocyclic arniiio alcohols have h e n examined froni the point of view of stereuchemistry in a n effort t o correlate pharmacological action with molecular configuration. The examination rras
confined to the properties of the heterocyclic amino gronp and was conducted with respect to the psychotomimetic
properties of the drugs. An important factor for psychotonlimetic potency was the availability of the electron
pair on nitrogen, which, presumably, must combine with an electrophilic center of the biological receptor site.
Conformational factors were important insofar as they contributed t o 1,3-diaxial interference by axial hydrogel1
atoms. Intramolecular int,eractions involving the noitbonded electron pair on nitrogen %erebelieved to diminish potency.

Coiisiderable interest has been focussed on the structurt.-nctivity relationships of wbstituted glycolate
(Stem of heterocyclic amino alcohols which possess
ni:irl.ced psychotogenic ( i . e . , psychotomimetic) propcrtir4.l ? Irevious studies were devoted largely to the
piperidyl arid pyrrolidyl esters, with eriiphasis on variation in the acid moiety of the eqters. Since it was
known that variations in the type of heterocyclic amino
alcohol resulted in striking differences in psychotomiinetic effectiveness, an opportuiiit y was at hand to
examine the possible role of the amine moiety iu the
pliarniacological action of this series of drugs.
R
R

CH i
1V

CHI
I11

CHI
I1

decreaiing potenry

VI

VI1

decreasing potency

Pharmacology.- The drug. iiicluded in I his -tidy


wcre evaluated for their aiiticholinergic poieiicy arid
for their effevt 0 1 1 thow parameter, of behavior which
are belie\-ed io tie related t o the psyrhotomimetic.
action of the agrnt:, in humani. X ni1caIiirmiciit of
hyperactivity in rntc has been found to be :isiinple :iiiti
effective iiieans of evaluating the over-a11 action of t h r
drugs upoii the central nervous sys1eni.l ? .Z inore
direct iiieaxiireiiieiii of the drugs disturbaiicc of highrr
central nervou function can bc made by iiie:iiis of the
-win1 maze and peek tests.3 *-\description of these
arid their reliability 111 predicting the psychotonuiiietic action of t h e glycolate esters i r i liumaris ha5
been dehcribed in detail elwvhere. l , ? Briefly Ihtx
i\pim tilaze consist5 of A niazr which caoritains water i r i
which a mouse i:, (mipelled to swim i n order to escape'
ilro\viiing; t h e iiiore cmtrally nctivc drugs produce :I
grwt cr nuiiiber of errors in perforniaric*e. The perk
i e i t \\ as developed by Iiosnian4 :IS a iiiearis of quarit ipecu1i:tr head-bobhirig :ti
load-swiyiiig inovcaassociatrd with t h r c e ~ ly active glycvlatt.
nce of this test 111
and although thr 4gri
determining the behavioral nberrat ions of the drugs i:,
not loo well understood, i i ha5 proved to 1 ) extremely
~
cffecative in predic+itg p~ychotorninic.tii. potency iii
huiiiarls. The ability of the drugs to block the acetylc.1ioliiie-induc.ed cmitraction of the isolated rabbit
ileum was takcit its :i measure of anticholinergic pu1 Pll(J-

R=c.g., benzilate
.

- -. - -

G. Almod arid J. II. 13iel Intern. Km. Seurobiol., 4, 217 (1962).


( 2 ) I,. G. Abood, A. Ostfeld, and J. >I. Riel, A r c h . intern. piiarmacodyn.
120, 186 (19591.
(1) I..

I:+)

SI. 1.; liosman, Proc. Suc. Gzptl. Biol. .\fed., 115, 728 (1964).
iEid., in press.
1 1 . C. (hang and J. H. Gaddurn, J . Phg.~iut., 79, 255 (1933).

( 4 ) 11. E. Kosman,
f,!)