A pharmacological study of the angiotensin receptor and tachyphylaxis in smooth muscle

Gen. Pharmac., 1976, Vol. 7, pp. 177 to 183. Pergamon Press. Printed in Great Britain

A PHARMACOLOGICAL STUDY OF THE ANGIOTENSIN RECEPTOR A N D TACHYPHYLAXIS IN SMOOTH MUSCLE JOHN M. STEWART, RICHARD J. FREER, LEONIDES REZENDE, CLARA PE1NAand GARY R. MATSUEDA

Department of Biochemistry, University of Colorado School of Medicine, Denver, CO, U.S.A. (Receired 18 December 1975) Abstract--The interaction of angiotensin with its receptor has been studied on the basis of the tachyphylaxis shown by the rat uterus towards angiotensin II when pH and Ca2+ concentration are below physiological levels, t4C-Angiotensin binding and 45Ca2÷-uptake investigations suggest tachyphylaxis to be due to increased binding at low pH and Ca-' + concentration. Studies with alkylating (affinity labeled) angiotensin derivatives containing the N-mustard chlorambucil suggest a "Chami+re type" inhibition at the Ca-binding site of the receptor and an irreversible inhibition at an anionic site. Angiotensin inhibitors containing chlorambucil do not alkylate tissue but are competitive inhibitors suggesting that the aromatic side chain in angiotensin may induce conformational changes in the receptor. The results obtained lead to a logical model for the angiotensin receptor allowing for normal activation by the hormone as well as for production of tachyphylaxis.

THE FIRSTstep in the physiological action of angiontensin II (Ang) and other peptide hormones is presumed to be binding of the peptide to receptors on the membrane of the muscle cells. The chemical nature of these receptors, the details of the interaction between the hormone and the receptor, and the mechanism by which this combination stimulates the response of the cell have been the subject of much discussion and intense research for years. Recently several investigators have attempted to isolate the receptor substance for Ang (Devunck et al., 1974) and other hormones (De Robertis, 1975; Schimke, 1973) from cell membrane-preparations. While these studies have yielded much information there still is need for a study of the receptors in situ, for only in this way can the details of the agonist-receptor interaction be elucidated and only by studying intact tissue is there a hope for understanding the mechanism whereby the stimulus-contraction coupling is effected. Earlier investigations (Lin & Goodfriend, 1970; Glossman et al., 1974) have contributed much to our present understanding of Ang receptors. We have combined data from several lines of investigation to assemble a model of Ang-receptor interaction. Significant contributions to this study came from investigation of the phenomenon of tachyphylaxis to Ang in isolated smooth muscle, and we propose a

(ClCH~ CH2)2N

(CH2)3C02H

Fig. 1. The structure of chlorambucil.

A s p - A r g - V a I - T y r - rl e - H is - P r o - Phe Angiofensin

II

Chl - Arg - V a I - T y r - I le - H i s - P r o - P h e [C,hl =] - a n g i o f e n s i n

II

A ~ - Pro - P r o - Gly - Phe - S e r - Pro - Phe -Arg Brodykinin

Fig. 2. Structures of peptides used.

hypothesis for a mechanism of tachyphylaxis production. Our initial study of angiotensin receptors involved the use of Ang derivatives containing the nitrogen mustard alkylating agent chlorambucil (Chl) (Fig. 1). Such derivatives can serve as probes to determine the location of nucleophilic sites on the receptor. If peptide-receptor combination holds the alkylating moiety of the chlorambucil residue adjacent to a nucleophilic site on the receptor, the ensuing reaction should lead to formation of a stable covalent bond between the peptide and the receptor, with consequent permanent inhibition of the receptor. Of a series of fragments of Ang containing chlorambucil, we found (Paiva et al., 1972) that one, [CHll]-Ang (Fig. 2) was a specific, irreversible inhibitor of the response of isolated guinea pig ileum to Ang (Fig. 3). Retreatment of the ileum with [ChP]-Ang or use of a larger dose of the inhibitor abolished completely the response to

177

178

JOHN M. STEWART, RICHARD J. FREER, LEONIDES REZENDE, CLARA PENA AND GARY R. MATSUEDA

I00 F Angiotensin

~,.~=~--------.o

60 4O

'°I

20

i0 -io

E IOOF

"~ o

60

,~

40

I

10-9

lO-a

I

I

i0 -7

10.6

10-5

t0 -?

Hisfomine

20

g

10-8

'::r Brodykinin

o --i

::L f < (< i0 -to

10-9 Concentration,

,

I

10-8

10-7 M

Fig. 3. Dose-response curves for angiotensin II, histamine, and bradykinin on guinea pig ileum before ( H ) and after (0 ©) treatment for 15 min with 10-s M [Chll]angiotensin II. Ang, still without affecting the response to other agonists. It is clear from these results that Ang receptors are unique, and that blockade of them does not interfere with the response of the tissue to bradykinin or histamine. Since it had earlier been postulated (Paiva & Paiva, 1966) that there is on the Ang receptor an anionic site which combines with the guanidinium group in the arginine residue at position 2 of Ang (Fig. 4), a logical hypothesis is that the group being alkylated by [ChlI]-Ang is this anionic group, perhaps a carboxylate ion (Fig. 5). A practical limitation of [Chll]-Ang as an inhibitor is that it possesses significant agonist action. This is due to the fact that the N-terminal aspartic acid residue is not necessary in the peptide molecule for good agonist activity (Page & Bumpus, 1973). In an attempt to overcome this undesirable feature, we introduced the chlorambucil residue into two known Ang inhibitors: we synthesized [Chl 1, Ilea]-Ang and [ChP, N-methylphenylalanineS]-Ang.These substitutions in the 8-position of Ang yield potent inhibitors (Khosla et al., 1972; Pefia et al., 1974). Much to our surprise these two chlorambucil derivatives were both good competitive inhibitors, but gave no evidence of alkylating the tissue to produce irreversible inhibition. One other chlorambucil peptide produced significant inhibition of the response of ileum to Ang. This was Chl-Pro-Phe-Arg, in which chlorambucil is

attached to the C-terminal tripeptide fragment of bradykinin (Fig. 2). This derivative did not affect the response of the tissue to bradykinin (Freer & Stewart, 1972a), but did produce permanent inhibition of the response of ileum to angiotensin. This inhibition was different from that produced by [Chll]-Ang, since in this case the maximum contraction could still be obtained (Fig. 6) after treatment of the tissue with the inhibitor. In this case use of larger doses of the inhibitor or retreatment of the tissue did not cause any change in the dose-response curve. Clearly the tissue was already maximally alkylated, but the inhibition was very different from that produced by [Chll] Ang. One possible explanation for this type of result (Rocha e Silva, 1969) is that in this case the alkylating agent was alkylating a nucleophilic site adjacent to the receptor, rather than directly upon it (Fig. 7). In this situation, high agonist concentrations could still give effective occupation of the receptor and produce the maximal response, but when the agonist was washed out, the inhibitor could return to its position occluding the receptor and give continued inhibition. At about this time we observed that tachyphylaxis to Ang was very severe in isolated rat uterus (in low calcium medium) if the experiments were done at a pH below the physiological range. It occurred to us that protonation of the imidazole of the histidine residue of Ang might be the cause of the tachyphylaxis, since the onset of tachyphylaxis occurred in the region where that group should be titrated. To test this hypothesis a series of analogs with substitutions in the 6-position was tested (Freer & Stewart, 1973). Asp NH II

Arg

NH-C-NH 3 +

®

Vol Tyr - - - - @ O H

Ar

Ile

Ar 8 G

[--'@--] "'-V~ ¢o 2+

V Pro 3

Phe CO2(~)

Ar

®

Fig. 4. Proposed normal angiotensin-receptor interaction. The receptor area is indicated by the large vertical rectangle. The anionic site to the right of the receptor is normally occupied by Ca 2+.

Angiotensin receptor and tachyphylaxis in smooth muscle

,

CHzCl'~l N3N-~NH~/~_OH }Ar i I~N

/ Ar'(~I

179

Ar

"

~

4

l

Ca2+

Ar

< Fig. 5. Postulated alkylation of angiotensin receptor by [ChlZ]-angiotensin II. Production of tachyphylaxis was found to correlate well with the presence of a basic group in the side chain of the amino acid in this position (Fig. 8). A dramatic demonstration is the case of 6-pyrazolylalanine-Ang, which is isosteric with Ang. The pyrazole ring has a pK of 2.2, and can thus never become protonated throughout the range being tested. In contrast, the p K of the imidazole of Ang is 6.4. It has long been established that the imidazole ring is important for the activity of Ang. Our experiments with analogs suggests that the imidazole binds to the receptor because of its aromatic character and because it possesses a slight nucleophilic character. In contrast, when the imidazole is protonated (or when some other residue bearing a positive charge is present) the imidazole combines with the receptor in an abnormal way. This suggests that perhaps there is on the tissue an anionic site near the imidazolecombining site, which can cause this abnormal inter-

NH II NH-C-NH~

AF

®®o2c

Fig. 7. Postulated mode of alkylation of angiotensin receptor by Chl-Pro--Phe-Arg. action leading to tachyphylaxis. The. resulting ionic bond would be expected to be much more stable than the normal aromatic-aromatic interaction. A further characteristic of Ang tachyphylaxis in rat uterus is that it is significant only when the calcium concentration of the medium is reduced below the physiological level (Freer & Stewart, 1972b). At physiological Ca 2 ÷ concentration, tachyphylaxis does not occur even at low pH. When uterus is made tachyphylactic at tow pH and low Ca 2 ÷ concentration, recovery can be accelerated by raising the pH or by increasing the Ca 2÷ concentration. These data sug140~ileS_Ang ~

I00

®

Iles-Orr~-ANG ]I

Vol5 - pzo6- ANG Tr

--

,oo

E 80--

~ s~ "6

60--

'~

1

612

IE

40-:r

2C

20--

i0 -ll

Z

i0 -I0

/,

10-9

Angiotensin Tr concentration~

,

I

I0-s

10 -7

M

Fig. 6. Dose-response curves for angiotensin II on guinea pig ileum before ( H ) , immediately after (O O), and 3 hr after (x x) treatment with Chl-Pro-Phe-Arg.

O~

2

5

4

2 3 Dose NOI

4

I

2

I

3

I

4

Fig. 8. Effect on rat uterus of successive doses of angiotensin II (left), [IleS,Orn6]-angiotensin II (center) and [VaP,Pza6]-angiotensin II (right) at pH 8.9 (¢ O), 8.0 (O-----O/. 6.8 (x x ) and 6.4 (A A). Doses were of the sarnc size, and were given at regular 8 min intervals. Pza is pyrazolylalanine.

180

JOHN M. STEWART, RICHARD J. FREER, LEONIDES REZENDE, CLARA PENA AND GARY R, MATSUEDA

140 Q

•

120

g~

I00 o.

8~ 'BE

80

+. E

60

40

20

0

I

I

I

2

5

4

Dose No,

Fig. 9. Response of rat uterus to repeated doses of angiotensin before and after treatment with Chl-Pro-Phe-Arg. All experiments were done at pH 6.5. The interval between doses was 8 rain. (~ - 0 ) before treatment, 2.0 mM Ca 2 + ; (O O) before treatment, 0.5 mM Ca 2+ ; ( x x ) after treatment, 0.5 mM Ca z+. gest that the anionic site responsible for tachyphylaxis production may be normally saturated with Ca 2 + at physiological concentration of this ion. Turning again to the inhibition produced by C h l P r o - P h e - A r g , integration of these data suggests as a corollary that the nucleophilic site being alkylated by this peptide may be the anionic site normally occupied by calcium with which the protonated imidazole combines to produce tachyphylaxis. If this is the case, alkylation of the tissue with C h l - P r o - P h e - A r g should block tachyphylaxis. As is shown in Fig. 9, this was indeed the case. This observation lends considerable support to the overall hypothesis (Stewart 1973; Stewart and Freer, 1974). This mechanism for tachyphylaxis requires that under tachyphylactogenic conditions Ang should be

bound more tightly to the receptors than under normal non-tachyphylactogenic conditions. To seek an answer to this question we conducted a series of binding studies in which radioactive angiotensin was used. In contrast to earlier studies of the binding of radioactive angiotensin to tissues, (Lin & Goodfriend, 1970) our experiments were done with angiotensin having a specific activity such that it was possible to do the binding experiments in the same concentration range as was required to produce the physiological response contraction of the tissue. In this way an exact correlation could be drawn between Ang binding and the tissue response. As is shown in Fig.10, Ang binding and tissue contraction occurred over the same concentration range and saturated at similar concentrations. It is interesting that although the thresholds are identical, the biological response curve rises more rapidly than does the binding curve. This type of response may indicate "biological amplification" of the action of Ang, and has been used by some to invoke "spare receptors". When the Ang binding studies were continued with Ang concentrations higher than those shown in Fig. 10, a second saturable binding curve was observed. A similar biphasic binding curve has also been observed in other studies. The effect of Ang concentration, time, Ca 2 + concentration and pH on the binding of Ang to tissues was studied. Figure 11 shows the effect of pH on the binding of Ang to tissue with respect to time, using a fixed (EDso) Ang dose. At this low Ca 2+ concentration, tachyphylaxis is produced at pH 6.8, but not at pH 8.8. It is clear that saturation is reached more rapidly at the high pH, while the amount of Ang bound is greater at the low pH. On the other hand, when a fixed time of 20 min was used for binding and Ang concentration was varied and the data were examined by means of a double reciprocal plot (Fig. 12), it is

300 --

[C 2+

250 -160

8-0

120

6,0

Oo

I E o

g-,g~

200

150 80--

~g

~g

4.0 ~o

z.0 ~

40--

0.1

0.5

I

1.0

I

5.0

I

I0.0

10-3[(14C ) ongiotensin ] , c p m / m l .

Fig. 10. Angiotensin II and rat uterus: comparison of muscle contraction and angiotensin binding at pH 8 . 8 .

/(././-.--

I00

50

I

I

I

I

I

I

5

I0

15

20

25

30

Time,

min

Fig. 11. Time course of binding of a constant dose of angiotensin II to rat uterus as a function of pH.

Angiotensin receptor and tachyphylaxis in smooth muscle

[CoZ+]:0.18mM

-

200

=~ == le0 /~J'/'p/ c ~'

120

,,Q o

40

. pH

8.8 H6.8

I

I

I

I

I

20

40

60

80

I00

IOS/[Angiotensin],

cpm -I x ml

Fig. 12. The effect of pH on binding of angiotensin II to rat uterus. The time allowed for binding was 20 min. clear that the difference in Ang binding at the two pH's is one of rate rather than amount (or total capacity of tissue). We now see that the greater binding observed at low pH in Fig. 11 is a reflection of the higher affinity of Ang for the receptor under these conditions, and that the capacity of the tissues is not changed. We then compared Ang binding and tachyphylaxis production within the same piece of tissue under various conditions (Fig. 13). In this experiment repeated doses of the same amount of Ang were given to the tissue at fixed 8-min intervals, with adequate washing between. Examining the figure we see that

600

-~ 500 ~5.~ 400 g "~ soo

.~_ §

_o o

~

zoo I00,

.....

• ....

"Q'- ~--a

~\

....

-0--. . . .

•

////P'-~'°''~

;

d o

18I

under physiological Ca 2+ concentration a transient tachyphylaxis is produced when the pH of the bath is first dropped from 8.4 to 6.8, but the tissue response recovers readily even though the tissue is maintained at the low pH (lower part of figure). Under these conditions, there is no change in the amount of Ang bound to the tissue (upper part of the figure). On the other hand, when the Ca 2+ concentration was low, lowering the pH of the medium produced a powerful and lasting tachyphylaxis which was reversed only by raising the pH of the bath. Under these conditions there was a sharp increase in binding when the bath pH was first lowered. As the bath was mainrained at the low pH, a new plateau of binding was established, slightly elevated over that observed under physiological conditions. When the bath pH was then raised, another increase in binding appeared. The increased binding observed upon lowering the pH can be understood in terms of the increased affinity of Ang for the tissue, while the peak seen on returning the pH to the physiological value can be understood in terms of a more rapid attainment of equilibrium at the high pH, as suggested by Fig. 11. These data thus appeared to be consistent with the hypothesis. If the tachyphylactogenic site on the tissue is indeed a Ca 2+-bindin.g site,_then An~_and Ca 2+ should compete with ea~ch other for this site under tachyphylactogenic conditions. The data presented in Table 1 show that while Ca z÷ concentration did not affect Ang binding at pH 8.8, at the low pH there was indeed an inhibition of Ang binding when the Ca 2÷ concentration of the medium was increased. Experiments were also done in the reverse direction to learn the effect of Ang on uptake of 45Ca 2+ by the tissue. It was found that Ang inhibited Ca 2÷ uptake by the tissue at low pH, but not at high pH. In contrast to our work demonstrating the dependence of tachyphylaxis on calcium concentration and protonation of the histidine imidazole, Paiva et al. (1974) showed that a free s-amino group was necessary in the Ang molecule for tachyphylaxis production. They found that the degree of tachyphylaxis correlated more closely with the degree of protonation of the s-amino group than with protonation of the imidazole. One can accommodate both sets of observations if it is assumed (as originally suggested by V. Nouailhetas) that two positive charges must exist

--

\

Table 1. Inhibition of angiotensin binding by calcium

/

),J

/

-

50

,~ Ca++ Concentration,

Dose pH

2 (control) 8.4

6.8

3

6.8

4 6.8

5 8.4

6

0

8.4

0.18 Ihc-angiotensln

pH 8.8

0.5

]15+_8

2.0

io9 +,.t27

Bound, per cent

Fig. 13. Comparison of angiotensin binding and tachyphylaxis in rat uterus. The same size dose of angiotensin was given repeatedly at 8 min intervals. Calcium concentration was 1.0 mM (dashed lines) or 0.18 mM (solid lines). G.P. 7--2/3--0

pH 6.8

Ioo +- 8

82 + 5

8o + 3

Results a r e expressed as percent of the amount of angiotensin bound at pH 8.8, 0.|8 mMCa++.

182

JOHN M. STEWART, RICHARD J. FREER, LEONIDES REZENDE, CLARA PENA AND GARY R. MATSUEDA

°°2+ lie

iAr ~

"--"

" -

,'At ;I VaJ

N. II

Pro

I

coo e

[,+ )

I

NH~cC-NH-Arg

~ - - - ----,---

Asp

Fig. 14. Postu|ated initial interaction of angiotensin with receptor functional groups (dashed circles) under normal conditions. The complex anionic site at the lower left of the figure is the calcium-specific pore, here occluded.

in the molecule simultaneously for tachyphylaxis to occur. When the possibility of either one of these is eliminated, either by modification of the histidine residue or of the amino terminus, tachyphylaxis cannot occur. It should be pointed out also that the experimental conditions used by Paiva et al., were quite different from those we used. An amalgamation of all these data leads to a logical model for the functioning of the Ang receptor and its normal activation by the hormone, as well as for the production of tachyphylaxis. Figure 14 depicts the Ang molecule, showing those side chain functional groups which are essential for the biological activity. The other functions shown in dashed circles are suggested to exist on the receptor. We may assume that the carboxylate ion of the C-terminal phenylalanine residue of Ang interacts with a cationic site on the receptor initially, followed by aromatic-aromatic interaction of the imidazole with an appropriate receptor site. After this initial binding of the peptide to the receptor, activation of the receptor would involve a concerted action of the two aromatic side chains of the tyrosine and phenylalanine residues with aromatic sites on the receptor surface. This interaction would cause a conformational change in the receptor to bring the appropriate aromatic sites together and allow for firm binding (Fig. 15). This conformational change in the receptor is facilitated by interaction of the arginine side chain with an anionic site on the receptor. Once this receptor conformational change has occurred, Ca 2+ may then enter the muscle cell, either through a Ca 2+-specific pore which is thus opened, or by activation of a calcium pump. In this model, the aromatic side chain of the phenylalanine residue does not participate in initial binding of the peptide to the receptor, but is required for inducing the conformational change necessary for initiating muscle contraction. This defines the sufficient

conditions for production of an inhibitor hormone analog, which is indeed the case with those analogs with substitutions for the°phenylalanine. The production of excellent inhibitors by methylation of the Pro-Phe peptide bond (the N-methylphenylalanine analogs) (Pefia et al., 1974) shows that the aromatic side chain on the phenylalanine residue must not only be present, but must also be in the correct position. Methylation of this peptide bond probably prevents this aromatic group from assuming the conformation necessary for effective interaction with the receptor. Experimental evidence for this hypothesis involves the three 1-chlorambucil analogs mentioned above. We propose that in order for the receptor to be alkylated by [Chll]-Ang, the receptor conformational change associated with the biological response must first take place in order that the susceptible ionic site on the receptor be brought into close juxtaposition to the alkylating agent. Competitive inhibitors by definition do not induce this receptor conformational change, and thus when the chlorambucil is placed in an inhibitor molecule, the anionic site is not brought into range of the alkylating agent. These analogs are thus competitive inhibitors only and do not permanently inactivate the tissue. In an attempt to "find" the nucleophilic site on the receptor without induction of the conformational change, several chlorambucil-Ellea]-angiotensin analogs were synthesized in which the chlorambucil residue was spaced away from the amino end of the Ang molecule by insertion of one or more amino acid residues. Unfortunately, none of these was able to alkylate the receptor. In light of this receptor model we can now examine the phenomenon of tachyphylaxis. If the imidazole is protonated (as a low pH) and the CaZ+-binding site is not occupied (as would be the case at low Ca 2÷ concentration), a different and abnormal folding of the Ang molecule may occur when the peptide binds

lie

(At ;

"-', Va I

CJ +

Pro

~ ~

tcoo ® ~-I-.)

II NH~NH-Arg

," . . . . 7-_'~ . . . . . ~ fl

Asp

Fig. 15. Postulated activation of the angiotensin receptor by the hormone.

Angiotensin receptor and tachyphylaxis in smooth muscle

Tie

Pro

\

r-,.,.,.,.,.,.,.,.,~ Tyr ~ HO,~,,~I \

Vol

.H \

Phe I

"-"

II

/~--

a~ "~

t~-x

Asp-N H3 (~)i~ }

Fig. 16. Postulated angiotensin-receptor interaction in tachyphylaxis. to the receptor. In this case, not only would the receptor conformation necessary for tissue response not occur, but the a m i n o end of the peptide could well be in a very different position, bringing it into juxtaposition to another anionic site on the tissue. In this case, if this amino group were also protonated (Fig: 16) t h e a d d i t i o n a l strong ionic binding indicated might occur, leading to tachyphylaxis. If, as has been suggested, the binding of Ang to its receptors occurs in a sequential manner, starting with the carboxyl end, it is apparent that the final (and presumably ratelimiting) step in the production of tachyphylaxis would be combination of the protonated or-amino group with the tissue, thus leading to the observed excellent correlation of tachyphylaxis and protonation of this group. This model suggests several types of experiments to be done in the future which should further test the validity of the model, and if correct, add much to our knowledge of the functioning of the Ang receptor. Several such experiments are currently in progress in our laboratory.

REFERENCES

DEVYNCK M. A., PERNOLLETM. G., MEYER P., FERMANDJIAN S., FROMAGEOT P. ~ BUMPUS F. M. (1974) Solubilisation of angiostensin II receptors in rabbit aortae membranes. Nature, Lond. 249, 67-69.

183

DE ROBERTISE. (1975) Synaptic receptors. In Isolation and Molecular Biology. Mod. Pharmacol. Toxicol. Ser., Vol. 4, Marcel Dekker, New York. FREER R. J. & STEWARTJ. M. (1972a) Alkylating analogs of peptide hormones. I. Synthesis and properties of p-[N,N-bis(2-chloroethyl)amino] phenylbutyryl derivatives of bradykinin and bradykinin potentiating factor. J. Med. Chem. 15, 1-5. FREER R. J. & STEWARTJ. M. (1972b) Some characteristics of uterine angiotensin receptors. In. StructureZ)tctivity Relationships of Protein and Polypeptide Hormones. (Edited by MARGOUUESM. & GREENWOODF. C.) pp. 480-495. Excerpta Medica, Amsterdam. FREER R. J. & STEWARTJ. M. (1973) Synthesis and pharmacology of position 6 analogs of angiotensin II. J. Med. Chem. 16, 733-735. GLOSSMANNH., BAUKALA. J. & CATT K. J. (1974) Properties of angiotensin II receptors in the bovine and rat adrenal cortex. J. Biol. Chem. 249, 825-834. KHOSLA M. C., LEESE R. A., MALOY W. L., FERREIRAA. T., SMEBYR. R. & BUMPUS F. M. (1972) Synthesis of some analogs of angiotensin II as specific antagonists of the parent hormone. J. Med. Chem. 15, 792-795. LIN S.-Y. & GOODFRIENDT. L. (1970) Angiotensin receptors. Am. J. Physiol. 218, 1319-1328. PAGE I. n. & BUMPUSF. M. (Editors) (1973) Angiotensin. Handb. Exp. Pharmacol Vol. XXXVII. Springer-Verlag, Berlin. PAIVA T. B., JULIANO L., NOUAILHETAS V. L. A. & PAIVA

A. C. M. (1974) The effect of pH on tachyphylaxis to angiotensin peptides in the isolated guinea pig ileum and rat uterus. Eur. J. Pharmac. 25, 191-196. PAIVA T. B. & PAIVA A. C. M. (1966) The inhibition of cationic myotropic drugs by compounds 48/80 and 46/108. Biochem. Pharmac. 15, 1303-1308. PAIVA T. B., PAIVA A. C. M., FREER R. J. •

STEWART

J. M. (1972) Alkylating analogs of peptide hormones. 2. Synthesis and properties of p-[N,N-bis(2-chloroethyl) amino] phenylbutyryl derivatives of angiotensin II. J. Med. Chem. 15, 6-8. PE~IA C., STEWARTJ. M. & GOODERiENDT. L. (1974) A new class of angioten in inhibitors: N-methylphenylalanine analogs. Life Sci. 14, 1331-1336. ROCHA E SILVAM. (1969) A thermodynamic approach to problems of drug interaction. I. The "Charnirre theory". Eur. J. Pharmac. 6, 294-302. SCmMKE R. T. (Editor) (1973) Membranes and hormone action. Fedn Proc. Fedn Am. Socs. exp. Biol. 32, 1833-1858. STEWART J. M. (1973) Tachyphylaxis to angiotensin. In Angiotensin (Edited by PAGE, I. H. & BUMPUS F. M.). Handb. Exptl. Pharmacol. Vol. XXXVII, pp. 170-184. Springer-Verlag, Berlin. STEWARTJ. M. & FREER R. J. (1974) Intra-Science Chem. Rep. 8, 159-163.