- Email: [email protected]
The effects of the aminopeptidase inhibitors amastatin and bestatin on angiotensin-evoked neuronal activity in rat brain
Brain Research, 424 (1987) 299-304 Elsevier
299
BRE 12996
The effects of the aminopeptidase inhibitors amastatin and bestatin on angiotensin-evoked neuronal activity in rat brain Joseph W. Harding and Dominik Felix Division of Animal Physiology, Berne (Switzerland) (Accepted 7 April 1987) Key words: Angiotensin II; Angiotensin III; Brain; Iontophoresis; Amastatin; Bestatin; Sarl-angiotensin II
During a recent comparison of iontophoretically applied angiotensin II (All) and angiotensin III (AIII) in the paraventricular nucleus of the rat, we observed that the response latency for AIII was much shorter than that for AII. This suggested that AII may have to be converted to AIII before it becomes active. To test this hypothesis we performed 3 experiments. (1) We examined the effects of bestatin, an adainopeptidase B inhibitor, on the activity of applied AII and AIII. (2) Next, we monitored the effects of amastatin, a specific aminopeptidase A inhibitor, on the action of co-applied AII or AIII. (3) And, finally, we examined the response to the aminopeptidase-resistant analog SarLAII, both applied alone and in combination with AII or AIII. Bestatin, while having no activity of its own, dramatically enhanced the actions of both All and AIII. Amastatin, on the other hand, had little effect on AIII's action and diminished or totally blocked AII-dependent activity. Like bestatin, amastatin had no effect alone. Sarl-AII reduced spontaneous activity of angiotensin-sensitive neurons and inhibited the actions of AII and AIII in a reversible manner. The same cells were also blocked by the recognized angiotensin antagonist Sar l,Iles-AII. In total these results strongly support the notion that AII must be converted to AIII in the brain before it is activated.
INTRODUCTION The brain angiotensin system plays an i m p o r t a n t role in the central control of cardiovascular function and b o d y water balance 9'14'2°'23. Much of the work that has been u n d e r t a k e n to characterize this important n e u r o p e p t i d e system has p r o c e e d e d u n d e r the premise that angiotensin II ( A I I ) , the o c t a p e p t i d e , is the active form of angiotensin in the brain. Nevertheless, a n u m b e r of recent studies, including this one, suggest that the h e p t a p e p t i d e , angiotensin III ( A I I I ) , may be the centrally active form of angiotensin. A l though not without opposition 5, others have already implied that A I I I m a y play an i m p o r t a n t role in regulating aldosterone release from the adrenal cortex 4. Initial support for a pivotal role for A I I I in the b r a i n can be traced to the first radioligand binding characterizations of brain angiotensin receptors, where IleS-AIII was shown to bind m o r e tightly than
A I I 2'3. A n examination of 125I-AII binding in o t h e r angiotensin-sensitive species including gerbil 12,18, m o n k e y t9, and rabbit 29, indicated little or no brain binding for 12SI-AII, while significant 125I-AIII binding could be d e m o n s t r a t e d . In line with the binding data, electrophysiological studies in cat subfornical organ 6 and rat paraventricular nucleus 11 demonstrated that iontophoretically applied A I I I was a more effective stimulator of neuronal activity than was A l l . This latter study also provided new data which suggested not only that A I I I was m o r e potent, but that A I I may have to be c o n v e r t e d to A I I I before it becomes active. In o r d e r to e x p a n d on this observation and to test the hypothesis that A l l undergoes an obligatory conversion to A I I I during activation, we p e r f o r m e d 3 additional experiments in which either the formation of A I I I from A l l or o t h e r angiotensins was blocked or the destruction of A I I I was attenuated. This block-
Correspondence: J.W. Harding. Present address: Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164, U.S.A. 0006-8993/87/$03.50 (~ 1987 Elsevier Science Publishers B.V. (Biomedical Division)
300 ade was carried out directly using the aminopeptidase inhibitors bestatin and amastatin or indirectly through the use of the aminopeptidase-resistant analog Sarl-AII 1°. MATERIALS AND METHODS Experiments were performed on 22 angiotensinsensitive cells from the paraventricular (16) and lateral septal (6) nuclei of adult Wistar-Kyoto (WKY) rats. Thirteen female rats (200-250 g) were used in this study. They were anesthetized with i.p. doses of thiopentane sodium (50 mg/kg). Extracellular action potentials were recorded through the 2 M NaCl-containing barrel of a 5-barrel glass micropipette and counted by a ratemeter. The other channels of the micropipettes contained various combinations of the compounds to be ejected microiontophoretically with appropriate cationic or anionic current: (1) Ile 5All (Sigma), prepared as a 10-3 M solution in distilled water, final pH 3.5; (2) IIeLAIII (Sigma), prepared as a 10-3 M solution in distilled water, final pH 4.5; (3) Sarl-AII (Peninsula Labs), prepared as a 10-3 M solution in distilled water, final pH 4.5; (4) Sarl,IleS-AII (Sigma), prepared as a 10 -3 M solution in distilled water, final pH 4.5; (5) bestatin hydrochloride (Sigma), prepared as a 5 x 10-3 M solution in distilled water, final pH 3.0; and (6) amastatin hydrochloride (Sigma), prepared as a 4 × 10 -3 M solution in distilled water, final pH 7.0. Compensation current was always used to prevent any direct current effects. The recording electrode solution also contained Fast green FCF, allowing the histological determination of the electrode site. RESULTS The first experiment examined the effects of the co-application of the aminopeptidase B inhibitor, bestatin, on the stimulatory actions of AII and AIII. Since bestatin is known to inhibit the action of aminopeptidase B 26, which cleaves basic (B) residues from the N-terminal, while having little effect on aminopeptidase A, it would be expected to preferentially block the cleavage of Arg from the N-terminal of AIII, thus interfering with AIII's major catabolic route 15.17,22. The result should be an enhancement of the stimulatory effect of iontophoretically applied
AIII. If AII is indeed converted to A I I I during activation, then bestatin would be expected to potentiate the action of All as well. Co-application of bestatin with either All or AIII uniformly and dramatically potentiated the stimulatory activity of both angiotensins. There appeared to be two distinct classes of bestatin-sensitive cells. The first type was characterized by an initial insensitivity to both AII and AIII that changed to high sensitivity subsequent to bestatin application (Fig. 1A). The two cells of this type that were examined, both retained some angiotensin sensitivity after the cessation of bestatin application. The large majority of cells that were subjected to bestatin application were initially All- and AIII-sensitive (6 out of 8). In all cases, co-application of bestatin produced a tremendous potentiation of the stimulatory action of both angiotensins (Fig. 1B). In no instance did bestatin exhibit any activity itself (Fig. 1A,B). The second experiment determined the influence of amastatin, an aminopeptidase A inhibitor, on coapplied All and AIII. Because amastatin specifically inhibits the action of aminopeptidase A 1, which cleaves acidic (A) amino acids from the N-terminal, it should block the removal of Asp from All and its conversion to A I I I while having little effect on the degradation of AIII. If our hypothesis is true, the effect of amastatin should be to block the effects of iontophoretically applied All while having minimal influence on AIII. Amastatin co-application (11 cells examined) with All or AIII produced differential effects. With All, amastatin uniformly reduced (Fig. 2B) or totally blocked (10 of 11 cells were totally blocked) the stimulatory effects of All (Fig. 2A,B). In one instance the blockade occurred after a substantial delay (Fig. 2A). Amastatin's effects on the stimulatory action of AIII were unlike those seen with All. In most instances (9 out of 11 cells) amastatin had little discernible effect on AIII-dependent stimulatory activity. The two remaining cells exhibited a reduction in AIII-dependent activity (Fig. 2B). In both cases, higher ejection currents were used (>80 nA). The attenuating effects of amastatin on All stimulation were reversible in all cells examined, with normal responsiveness returning in 5-10 rain. Like bestatin, amastatin alone exhibited no activity. The final experiment examined the activity of ion-
301 I0 f
A ~ut_l
L0
t ji:]=q
m A N G I]110
20
ANG ll'40
20 I
i===!
40
jL--=L~i
10
L
80
jR- =!
20
80
t
BESTATIN 100
[
_
_
~
A N G II15
I
~ 20
10
==~E[_ ANG ] 1 8 0
~ 20 J
t==K 10
._m~l_ 40
BESTATIN 6 0
1 min
m~ L 20
..
1
--
80
I
Fig. 1. Response of rat paraventricular neurons to iontophoretically applied angiotensin II (AII) or angiotensin III (AIII) alone and with co-application of the aminopeptidase B inhibitor bestatin. Experimental details are described in the text. The frequency (f) of response is expressed in spikes/s. A: this cell was initially unresponsive to either AII or AIII at doses used. However, with bestatin co-application, both AII and AIII induced vigorous responses. Although reduced, some of their stimulatory ability remained after the cessation of bestatin application. Bestatin exhibited no activity itself. B: this cell was responsive to both AII and AIII. The cell's sensitivity and responsiveness to AII and AIII were again enhanced by bestatin. After termination of bestatin ejection, AII and AIII responses returned to near initial levels. Again bestatin had no effect alone.
5
ANGTr 70
ANG m 4 0 I
L
AMASTATIN 5 0
I
ANG fl130
1 rain
I
ANG ~1"8 0
J
/%MASTATIN 100
I
Fig. 2. Responses of one rat paraventricular (A) and one septal (B) neuron to iontophoretically applied angiotensin I[ (AH) or angiotensin [ I I ( A [ [ [ ) alone and in combination with the aminopeptidase A inhibitor amastatin. Experimental details are described in the text. The frequency (f) of response is expressed in spikes/s. A: this cell was responsive to both A H and A I I [ . Following 5 min of amastatin application the responsiveness of the cell to A I ! was blocked while the effect of A H I was unaltered. After termination of amastatin application, the responsiveness to A I [ returned. B: this cell exhibits response characteristics similar to cell A except that the response to A I I is not totally blocked and appears to be manifested sooner after the initiation of amastatin application.
302 tophoretically applied Sarl-AII, which is an angiotensin analog that has been modified in such a way as to interfere with aminopeptidase-dependent metabolism ~°. This compound, which is known to be a peripheral 'superagonist '23 and to bind to central sites 16, is therefore only slowly convertible to AIII. If con-
[-5
ANGIT50
A N G fir 10 mln
version to A I I I is necessary for angiotensin activation in the brain, S a r L A I I should be inactive and could be predicted to act as an antagonist. The aminopeptidase-resistant angiotensin analog Sarl-AII possessed no stimulatory activity but instead tended to reduce the spontaneous activity of angiotensin-sensitive cells. Other spontaneous nonangiotensin-sensitive cells were unaffected by Sar IAII. In every angiotensin-sensitive cell examined, Sarl-AII application produced total and long-lasting blockade of both A I I - and AIII-dependent stimulatory activity (Fig. 3). This blockade was completely reversible. In addition, the same angiotensin-sensitive cells that were inhibited by Sarl-AlI could also be blocked by the known angiotensin antagonist Sarl,IleS-All. DISCUSSION
--J
SAR 1 - ANGTr 5 0
t
J SARI,ILE 8-ANGIT50
"~
Fig. 3. Response of rat septal neuron to iontophoretically applied angiotensin II (All) or angiotensin III (AIII) alone or in combination with the aminopeptidase-resistant analog Sar 1All. Experimental details are described in the text. The frequency (J) of response is expressed in spikes/s. This spontaneously active cell was responsive to both AlI and AIII. Sar L AII acted both to reduce the spontaneous activity of the cell and to completely block the stimulatory action of both AII and AIII. After cessation of SarLAII application, the responsiveness to AII and AIII returned. However, the cell's sensitivity to All and AIII could again be blocked by the established angiotensin antagonist Sarl,IleS-AII. Like Sarl-AlI, Sarl,IleS-AII also reduced spontaneous activity.
The data presented above are consistent with the proposal that A I I is converted to A I I I during activation. Although the results of each experiment alone may not provide unequivocal proof of this hypothesis, we feel that the three experiments together strongly support it. The aminopeptidase inhibitor bestatin, while selective for aminopeptidase B 26, which removes basic residues, is not totally specific and can inhibit other aminopeptidases like aminopeptidase M 24. Therefore, it is not possible to unequivocally attribute the enhancement of AII- and AIII-dependent activity to an accumulation of A I I I alone. Nevertheless, the data are consistent with an obligatory conversion and at the very least demonstrate that the action of A I I I is tremendously potentiated when degradation is suppressed. In accord with these findings is the recent observation that intracerebroventricularly applied bestatin potentiates the pressor 2s and dipsogenic (Wright et al., unpublished observations) action of both A I I and AIII. Interestingly, bestatin given alone can stimulate drinking and raise blood pressure by what appears to be an angiotensin-dependent mechanism, i.e. its effects are blockable by the angiotensin antagonist Sar a,Thr~-AII (Wright, Batt and Harding, unpublished observations). The action of amastatin is again selective but not totally specific 24. Amastatin preferentially inhibits aminopeptidase A, which cleaves acidic amino acids
303 from the N-ierminal 1. The ability of amastatin to differentially block the action of All while having little effect on AIII-dependent activity strongly supports the idea that an AII-to-AIII activation step is necessary. It appears that amastatin's effect must be attributable to inhibition of a selective aminopeptidase, like aminopeptidase A, since inhibition of more nonselective peptidases should also affect AIII's action. One could predict, based on the results of the bestatin experiment, that such an inhibition of A I I I metabolism should greatly enhance its activity - - an observation that could not be made while using amastatin. Although a behavioral and physiological assessment of amastatin's effects has just begun, it is already clear that, as in the electrophysiological studies, amastatin's effects differ from those of bestatin. For example, when given alone, amastatin does not exhibit the potent dipsogenic action of bestatin, suggesting that it does not inhibit the metabolism of active endogenous angiotensins. Furthermore, recent experiments which examined the effects of amastatin pretreatment on the pressor activity of i.c.v.-applied All and A I I I demonstrate a selective decrease in the size of the blood pressure response to All but not A I I I (Sullivan and Wright, unpublished). Based on studies in the periphery, Sarl-AII has been classifed as an angiotensin 'superagonist'Zk The enhanced activity has been directly attributed to its resistance to aminopeptidase action 1°. In this study, however, Sarl-AII acted not as a 'superagonist' but instead as a potent antagonist. One possible explanation for this difference is that the vascular and adrenal angiotensin receptors differ in their specificity for angiotensins. While the C-terminal may be the sole critical determinant for inducing agonist activity at peripheral receptors, it may be that the central receptors possess requirements for both N- and C-terminals. Specifically, the central receptors may require an exposed N-terminal arginine. Since Sarl-AII cannot be effectively converted to A I I I , the active form, its binding to the receptor produces an antagonist effect. Recent physiological studies from our laboratory group have examined the effects of i.c.v, applied Sarl-AII on blood pressure. The results, which at first glance appear contradictory to the iontophoretic study, indicate that Sarl-AII does produce an elevation in blood pressure that is longer lasting than that produced by All or AIII. However, unlike A l l and
AIII, Sarl-AII is effective only as a single dose producing a profound tachyphylaxis and inhibition of further angiotensin stimulation. Because Sarl-angio tensins are effectively degraded in the ventricular space 13, the activity witnessed with Sar~-AII may, in fact, result from the production of active AIII. Its longer-lasting action is most likely a consequence of Sarl-AII's resistance to peptidase action, therefore allowing it to remain in the vicinity of the receptors as preactive material for a longer period. The tachyphylaxis/inhibition could result from the actual binding of Sarl-AII itself to receptors with the delay being due to its very slow binding on rate (Erickson and Harding, unpublished observations). The major objection to the above hypothesis can be drawn from numerous behavioral studies which indicate a greater potency with regard to elevating blood pressure and stimulating drinking for All than for A I I I when injected i.c.v. 7'25. If the central angiotensin receptor does preferentially interact with AIII, then why is All more physiologically active? The solution to this dilemma is provided by two recent studies from our laboratory group. In the first study, the pressor and dipsogenic potencies of these peptides following i.v. application in rats was reexamined 27. The results clearly showed that at low doses A I I I and All had equivalent potencies and only at higher doses, similar to those used in most previous studies, was AII more potent. These data led to the suggestion that A I I I may be more labile than AII in the cerebroventricular space. Thus, at higher doses, All would act as a depot of pre-active material, while at lower doses the enhanced stability of AII would be offset by the higher affinity of AIII. The concept of an angiotensin precursor acting as an effective dipsogen is not new but has in the past focused on the latent activity of renin substrate and angiotensin 18. The idea that All is more degradation-resistant than AIII, thus acting as a more stable precursor, was recently confirmed by additional studies that measured the cerebroventricular metabolism of 125I-angiotensins 13. These studies determined that the degradation of A I I I to inactive fragments occurs at 3 times the rate measured for AII and A I I I and has an amazingly short half-life of 7.7 s in the ventricular space. These facts together suggest that any apparent lack of biological potency observed for A I I I can be traced to its very active metabolism.
304 Taken together, the results of this study support a paramount role for AIII in the brain. Furthermore, they strongly suggest that the activity witnessed with AII is dependent upon its prior conversion to AIII. ACKNOWLEDGEMENTS
This work was supported by Grants TW 01112 and REFERENCES 1 Aoyagi, T., Tobe, H., Kojima, F., Hamada, M., Takeuchi, M. and Umezawa, H., Amastatin, an inhibitor of aminopeptidase A, produced in actinomycetes, J. Antibiot., 31 (1978) 636-638. 2 Bennett Jr., J.P. and Snyder, S.H., Angiotensin II binding to mammalian brain membranes, J. Biol. Chem., 251, 23 (1976) 7423-7430. 3 Bennett, J.P. and Snyder, S.H., Receptor binding interaction of the angiotensin II antagonist, 125I-(Sarcosinel,leucineS)angiotensin II, with mammalian brain and peripheral tissues, Eur. J. Pharmacol., 67 (1980) 11-26. 4 Devynck, M.-A., Pernollet, M.G., Matthews, P.G., Khosla, M.C., Bumpus, F.M. and Myer, P., Specific receptors for des-Aspl-angiotensin II ('angiotensin III') in rat adrenals, Proc. Natl. Acad. Sci. U.S.A., 74 (1977) 4021-4032. 5 Douglas, J.G., Khosla, M.C. and Bumpus, F.M., Efficacy of octa- and heptapeptide antagonists of angiotensin II as inhibitors of angiotensin III binding in the rat adrenal glomerulosa, Endocrinology, 116 (1985) 1598-1602. 6 Felix, D. and Schlegel, W., Angiotensin receptive neurons in the subfornical organ. Structure-activity relations, Brain Research, 149 (1978) 107-116. 7 Fitzsimons, J.T., The effect on drinking of peptide precursors and of shorter chain fragments of angiotensin II injected into the rat's diencephalon, J. Physiol. (London), 214 (1971) 295-303. 8 Fitzsimons, J.T. and Kucharczyk, J., Drinking and haemodynamic changes induced in the dog by intracranial injection of components of the renin-angiotensin system, J. Physiol. (London), 276 (1978) 419-434. 9 Ganten, D., Lang, R.E., Lehmann, E. and Unger, T., Brain angiotensin: on the way to becoming a well-studied neuropeptide system, Biochem. Pharmacol., 33 (1984) 3523-3528. 10 Hall, M.M., Khosla, M.C., Khairallah, P.A. and Bumpus, F.M., Angiotensin analogs: the influence of sarcosine substituted in position 1, J. Pharmacol. Exp. Ther., 188 (1974) 222-228. 11 Harding, J.W., Imboden, H. and Felix, D., Is angiotensin III the centrally active form of angiotensin?, Experientia, 42 (1986) 706. 12 Harding, J.W., Stone, L.P. and Wright, J.W., The distribution of angiotensin II binding sites in rodent brain, Brain Research, 205 (1981) 265-274. 13 Harding, J.W., Yoshida, M.S., Dilts, R.P., Woods, T.M. and Wright, J.W., Cerebroventricular and intravascular metabolism of ~25I-angiotensins in rat, J. Neurochem., 46 (1986) 1292-1297. 14 Johnson, A.K., Neurobiology of the periventricular tissue surrounding the anteroventral third ventricle (AV3V) and its role in behavior, fluid balance, and cardiovascular control. In Smith, Galosy and Weiss (Eds.), Circulation, Neurobiology, and Behavior, Elsevier, Amsterdam. 1982, pp. 277-295.
HL 32063 from the NIH and Grant 831145 from the American Heart Association and its Washington affiliate to J.W.H., as well as Grant 3.627-0.84 from the Swiss National Science Foundation and assistance from the 'Stiftung zur F6rderung der wissenschaftlichen Forschung an der Universit/it Bern' to D.F. We would especially like to thank J. Conger for preparing this manuscript. 15 Kugler, P., On angiotensin-degrading aminopeptidases in the rat kidney. In Beck, Hild, Van Limborgh, Ortman, Pauly, and Schiebler (Eds.), Advances in Anatomy, Embryology, and Cell Biology, 76 (1982) 1-86. 16 Mendelsohn, F.A.O., Quirion, R., Saavedra, J.M., Aguilera, G. and Catt, K.J., Autoradiographic localization of angiotensin receptors in rat brain, Proc. Natl. Acad. Sei. U.S.A., 81 (1984) 1575-1579. 17 Mizutani, S., Akiyama, H., Kurauchi, O., Taira, H., Narita, O. and Tomodo, Y., In vitro degradation of angiotensin II (A-II) by human placental subcellular fractions, pregnancy sera, and purified placental aminopeptidases, Acta Endocrinol., 110 (1985) 135-139. 18 Petersen, E.P., Camara, C.G., Abhold, R.H., Wright, J.W. and Harding, J.W., Characterization of angiotensin binding in the gerbil brain using [lZ5I]angiotensin III as the radioligand, Brain Research, 321 (1984) 225-235. 19 Petersen, E.P., Camara, C.G., Abhold, R.H., Wright, J.W., Harding, J.W., Characterization of angiotensin binding in the African Green monkey, Brain Research, 341 (1985) 139-146. 20 Phillips, M.I., Brain renin-angiotensin and hypertension. In G.P. Guthrie Jr. and T.A. Kotchen (Eds.), Hypertension and the Brain, Futura, Mount Kisco, NY, 1984, pp. 63-81. 21 Regoli, D., Park, W.K. and Rioux, F., Pharmacology of angiotensin, Pharmacol. Rev., 26 (1974) 69-123. 22 Regoli, D., Riniker, D. and Brunner, H.R., The enzymatic degradation of various angiotensin II derivatives by serum, plasma, or kidney homogenates, Biochem. Pharmacol., 12 (1963) 637-646. 23 Reid, I.A., Actions of angiotensin II on the brain: mechanisms and physiological role, Am. J. Physiol., 246 (1984) F533-F543. 24 Rich, D.H., Moon, B.J. and Harbeson, S., Inhibition of aminopeptidases by amastatin and bestatin derivatives. Effect of inhibitor structure on slow-binding processes, J. Med. Chem., 27 (1984) 417-422, 25 Tonnaer, J.A.D.M., Wiegant, V.M., DeJong, M. and DeWied, D., Central effects of angiotensin on drinking and blood pressure: structure-activity relationships, Brain Research, 236 (1982) 417- 428. 26 Umezawa, H., Aoyagi, T., Suda, H., Hamada, M. and Takeuchi, T., Bestatin, an inhibitor of aminopeptidase B, produced by actinomycetes, J. Antibiot., 29 (1976) 97-99. 27 Wright, J.W., Morseth, S.L., Abhold, R.H. and Harding, J.W., Comparison of angiotensin II- and III-induced pressor action and dipsogenicity in rats, Am. J. Physiol., 249 (1985) R514-R521. 28 Wright, J.W., Sullivan, M.J. and Harding, J.W., Dysfunction of central angiotensinergic aminopeptidase activity in spontaneously hypertensive rats, Neurosci. Lett., 61 (1986) 351-356. 29 Wright, J.W., Sullivan, M.J., Petersen, E.P. and Harding, J.W., Brain angiotensin II and Ill dipsogenicity in the rabbit, Brain Research, 358 (1985) 376-379.
299
BRE 12996
The effects of the aminopeptidase inhibitors amastatin and bestatin on angiotensin-evoked neuronal activity in rat brain Joseph W. Harding and Dominik Felix Division of Animal Physiology, Berne (Switzerland) (Accepted 7 April 1987) Key words: Angiotensin II; Angiotensin III; Brain; Iontophoresis; Amastatin; Bestatin; Sarl-angiotensin II
During a recent comparison of iontophoretically applied angiotensin II (All) and angiotensin III (AIII) in the paraventricular nucleus of the rat, we observed that the response latency for AIII was much shorter than that for AII. This suggested that AII may have to be converted to AIII before it becomes active. To test this hypothesis we performed 3 experiments. (1) We examined the effects of bestatin, an adainopeptidase B inhibitor, on the activity of applied AII and AIII. (2) Next, we monitored the effects of amastatin, a specific aminopeptidase A inhibitor, on the action of co-applied AII or AIII. (3) And, finally, we examined the response to the aminopeptidase-resistant analog SarLAII, both applied alone and in combination with AII or AIII. Bestatin, while having no activity of its own, dramatically enhanced the actions of both All and AIII. Amastatin, on the other hand, had little effect on AIII's action and diminished or totally blocked AII-dependent activity. Like bestatin, amastatin had no effect alone. Sarl-AII reduced spontaneous activity of angiotensin-sensitive neurons and inhibited the actions of AII and AIII in a reversible manner. The same cells were also blocked by the recognized angiotensin antagonist Sar l,Iles-AII. In total these results strongly support the notion that AII must be converted to AIII in the brain before it is activated.
INTRODUCTION The brain angiotensin system plays an i m p o r t a n t role in the central control of cardiovascular function and b o d y water balance 9'14'2°'23. Much of the work that has been u n d e r t a k e n to characterize this important n e u r o p e p t i d e system has p r o c e e d e d u n d e r the premise that angiotensin II ( A I I ) , the o c t a p e p t i d e , is the active form of angiotensin in the brain. Nevertheless, a n u m b e r of recent studies, including this one, suggest that the h e p t a p e p t i d e , angiotensin III ( A I I I ) , may be the centrally active form of angiotensin. A l though not without opposition 5, others have already implied that A I I I m a y play an i m p o r t a n t role in regulating aldosterone release from the adrenal cortex 4. Initial support for a pivotal role for A I I I in the b r a i n can be traced to the first radioligand binding characterizations of brain angiotensin receptors, where IleS-AIII was shown to bind m o r e tightly than
A I I 2'3. A n examination of 125I-AII binding in o t h e r angiotensin-sensitive species including gerbil 12,18, m o n k e y t9, and rabbit 29, indicated little or no brain binding for 12SI-AII, while significant 125I-AIII binding could be d e m o n s t r a t e d . In line with the binding data, electrophysiological studies in cat subfornical organ 6 and rat paraventricular nucleus 11 demonstrated that iontophoretically applied A I I I was a more effective stimulator of neuronal activity than was A l l . This latter study also provided new data which suggested not only that A I I I was m o r e potent, but that A I I may have to be c o n v e r t e d to A I I I before it becomes active. In o r d e r to e x p a n d on this observation and to test the hypothesis that A l l undergoes an obligatory conversion to A I I I during activation, we p e r f o r m e d 3 additional experiments in which either the formation of A I I I from A l l or o t h e r angiotensins was blocked or the destruction of A I I I was attenuated. This block-
Correspondence: J.W. Harding. Present address: Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164, U.S.A. 0006-8993/87/$03.50 (~ 1987 Elsevier Science Publishers B.V. (Biomedical Division)
300 ade was carried out directly using the aminopeptidase inhibitors bestatin and amastatin or indirectly through the use of the aminopeptidase-resistant analog Sarl-AII 1°. MATERIALS AND METHODS Experiments were performed on 22 angiotensinsensitive cells from the paraventricular (16) and lateral septal (6) nuclei of adult Wistar-Kyoto (WKY) rats. Thirteen female rats (200-250 g) were used in this study. They were anesthetized with i.p. doses of thiopentane sodium (50 mg/kg). Extracellular action potentials were recorded through the 2 M NaCl-containing barrel of a 5-barrel glass micropipette and counted by a ratemeter. The other channels of the micropipettes contained various combinations of the compounds to be ejected microiontophoretically with appropriate cationic or anionic current: (1) Ile 5All (Sigma), prepared as a 10-3 M solution in distilled water, final pH 3.5; (2) IIeLAIII (Sigma), prepared as a 10-3 M solution in distilled water, final pH 4.5; (3) Sarl-AII (Peninsula Labs), prepared as a 10-3 M solution in distilled water, final pH 4.5; (4) Sarl,IleS-AII (Sigma), prepared as a 10 -3 M solution in distilled water, final pH 4.5; (5) bestatin hydrochloride (Sigma), prepared as a 5 x 10-3 M solution in distilled water, final pH 3.0; and (6) amastatin hydrochloride (Sigma), prepared as a 4 × 10 -3 M solution in distilled water, final pH 7.0. Compensation current was always used to prevent any direct current effects. The recording electrode solution also contained Fast green FCF, allowing the histological determination of the electrode site. RESULTS The first experiment examined the effects of the co-application of the aminopeptidase B inhibitor, bestatin, on the stimulatory actions of AII and AIII. Since bestatin is known to inhibit the action of aminopeptidase B 26, which cleaves basic (B) residues from the N-terminal, while having little effect on aminopeptidase A, it would be expected to preferentially block the cleavage of Arg from the N-terminal of AIII, thus interfering with AIII's major catabolic route 15.17,22. The result should be an enhancement of the stimulatory effect of iontophoretically applied
AIII. If AII is indeed converted to A I I I during activation, then bestatin would be expected to potentiate the action of All as well. Co-application of bestatin with either All or AIII uniformly and dramatically potentiated the stimulatory activity of both angiotensins. There appeared to be two distinct classes of bestatin-sensitive cells. The first type was characterized by an initial insensitivity to both AII and AIII that changed to high sensitivity subsequent to bestatin application (Fig. 1A). The two cells of this type that were examined, both retained some angiotensin sensitivity after the cessation of bestatin application. The large majority of cells that were subjected to bestatin application were initially All- and AIII-sensitive (6 out of 8). In all cases, co-application of bestatin produced a tremendous potentiation of the stimulatory action of both angiotensins (Fig. 1B). In no instance did bestatin exhibit any activity itself (Fig. 1A,B). The second experiment determined the influence of amastatin, an aminopeptidase A inhibitor, on coapplied All and AIII. Because amastatin specifically inhibits the action of aminopeptidase A 1, which cleaves acidic (A) amino acids from the N-terminal, it should block the removal of Asp from All and its conversion to A I I I while having little effect on the degradation of AIII. If our hypothesis is true, the effect of amastatin should be to block the effects of iontophoretically applied All while having minimal influence on AIII. Amastatin co-application (11 cells examined) with All or AIII produced differential effects. With All, amastatin uniformly reduced (Fig. 2B) or totally blocked (10 of 11 cells were totally blocked) the stimulatory effects of All (Fig. 2A,B). In one instance the blockade occurred after a substantial delay (Fig. 2A). Amastatin's effects on the stimulatory action of AIII were unlike those seen with All. In most instances (9 out of 11 cells) amastatin had little discernible effect on AIII-dependent stimulatory activity. The two remaining cells exhibited a reduction in AIII-dependent activity (Fig. 2B). In both cases, higher ejection currents were used (>80 nA). The attenuating effects of amastatin on All stimulation were reversible in all cells examined, with normal responsiveness returning in 5-10 rain. Like bestatin, amastatin alone exhibited no activity. The final experiment examined the activity of ion-
301 I0 f
A ~ut_l
L0
t ji:]=q
m A N G I]110
20
ANG ll'40
20 I
i===!
40
jL--=L~i
10
L
80
jR- =!
20
80
t
BESTATIN 100
[
_
_
~
A N G II15
I
~ 20
10
==~E[_ ANG ] 1 8 0
~ 20 J
t==K 10
._m~l_ 40
BESTATIN 6 0
1 min
m~ L 20
..
1
--
80
I
Fig. 1. Response of rat paraventricular neurons to iontophoretically applied angiotensin II (AII) or angiotensin III (AIII) alone and with co-application of the aminopeptidase B inhibitor bestatin. Experimental details are described in the text. The frequency (f) of response is expressed in spikes/s. A: this cell was initially unresponsive to either AII or AIII at doses used. However, with bestatin co-application, both AII and AIII induced vigorous responses. Although reduced, some of their stimulatory ability remained after the cessation of bestatin application. Bestatin exhibited no activity itself. B: this cell was responsive to both AII and AIII. The cell's sensitivity and responsiveness to AII and AIII were again enhanced by bestatin. After termination of bestatin ejection, AII and AIII responses returned to near initial levels. Again bestatin had no effect alone.
5
ANGTr 70
ANG m 4 0 I
L
AMASTATIN 5 0
I
ANG fl130
1 rain
I
ANG ~1"8 0
J
/%MASTATIN 100
I
Fig. 2. Responses of one rat paraventricular (A) and one septal (B) neuron to iontophoretically applied angiotensin I[ (AH) or angiotensin [ I I ( A [ [ [ ) alone and in combination with the aminopeptidase A inhibitor amastatin. Experimental details are described in the text. The frequency (f) of response is expressed in spikes/s. A: this cell was responsive to both A H and A I I [ . Following 5 min of amastatin application the responsiveness of the cell to A I ! was blocked while the effect of A H I was unaltered. After termination of amastatin application, the responsiveness to A I [ returned. B: this cell exhibits response characteristics similar to cell A except that the response to A I I is not totally blocked and appears to be manifested sooner after the initiation of amastatin application.
302 tophoretically applied Sarl-AII, which is an angiotensin analog that has been modified in such a way as to interfere with aminopeptidase-dependent metabolism ~°. This compound, which is known to be a peripheral 'superagonist '23 and to bind to central sites 16, is therefore only slowly convertible to AIII. If con-
[-5
ANGIT50
A N G fir 10 mln
version to A I I I is necessary for angiotensin activation in the brain, S a r L A I I should be inactive and could be predicted to act as an antagonist. The aminopeptidase-resistant angiotensin analog Sarl-AII possessed no stimulatory activity but instead tended to reduce the spontaneous activity of angiotensin-sensitive cells. Other spontaneous nonangiotensin-sensitive cells were unaffected by Sar IAII. In every angiotensin-sensitive cell examined, Sarl-AII application produced total and long-lasting blockade of both A I I - and AIII-dependent stimulatory activity (Fig. 3). This blockade was completely reversible. In addition, the same angiotensin-sensitive cells that were inhibited by Sarl-AlI could also be blocked by the known angiotensin antagonist Sarl,IleS-All. DISCUSSION
--J
SAR 1 - ANGTr 5 0
t
J SARI,ILE 8-ANGIT50
"~
Fig. 3. Response of rat septal neuron to iontophoretically applied angiotensin II (All) or angiotensin III (AIII) alone or in combination with the aminopeptidase-resistant analog Sar 1All. Experimental details are described in the text. The frequency (J) of response is expressed in spikes/s. This spontaneously active cell was responsive to both AlI and AIII. Sar L AII acted both to reduce the spontaneous activity of the cell and to completely block the stimulatory action of both AII and AIII. After cessation of SarLAII application, the responsiveness to AII and AIII returned. However, the cell's sensitivity to All and AIII could again be blocked by the established angiotensin antagonist Sarl,IleS-AII. Like Sarl-AlI, Sarl,IleS-AII also reduced spontaneous activity.
The data presented above are consistent with the proposal that A I I is converted to A I I I during activation. Although the results of each experiment alone may not provide unequivocal proof of this hypothesis, we feel that the three experiments together strongly support it. The aminopeptidase inhibitor bestatin, while selective for aminopeptidase B 26, which removes basic residues, is not totally specific and can inhibit other aminopeptidases like aminopeptidase M 24. Therefore, it is not possible to unequivocally attribute the enhancement of AII- and AIII-dependent activity to an accumulation of A I I I alone. Nevertheless, the data are consistent with an obligatory conversion and at the very least demonstrate that the action of A I I I is tremendously potentiated when degradation is suppressed. In accord with these findings is the recent observation that intracerebroventricularly applied bestatin potentiates the pressor 2s and dipsogenic (Wright et al., unpublished observations) action of both A I I and AIII. Interestingly, bestatin given alone can stimulate drinking and raise blood pressure by what appears to be an angiotensin-dependent mechanism, i.e. its effects are blockable by the angiotensin antagonist Sar a,Thr~-AII (Wright, Batt and Harding, unpublished observations). The action of amastatin is again selective but not totally specific 24. Amastatin preferentially inhibits aminopeptidase A, which cleaves acidic amino acids
303 from the N-ierminal 1. The ability of amastatin to differentially block the action of All while having little effect on AIII-dependent activity strongly supports the idea that an AII-to-AIII activation step is necessary. It appears that amastatin's effect must be attributable to inhibition of a selective aminopeptidase, like aminopeptidase A, since inhibition of more nonselective peptidases should also affect AIII's action. One could predict, based on the results of the bestatin experiment, that such an inhibition of A I I I metabolism should greatly enhance its activity - - an observation that could not be made while using amastatin. Although a behavioral and physiological assessment of amastatin's effects has just begun, it is already clear that, as in the electrophysiological studies, amastatin's effects differ from those of bestatin. For example, when given alone, amastatin does not exhibit the potent dipsogenic action of bestatin, suggesting that it does not inhibit the metabolism of active endogenous angiotensins. Furthermore, recent experiments which examined the effects of amastatin pretreatment on the pressor activity of i.c.v.-applied All and A I I I demonstrate a selective decrease in the size of the blood pressure response to All but not A I I I (Sullivan and Wright, unpublished). Based on studies in the periphery, Sarl-AII has been classifed as an angiotensin 'superagonist'Zk The enhanced activity has been directly attributed to its resistance to aminopeptidase action 1°. In this study, however, Sarl-AII acted not as a 'superagonist' but instead as a potent antagonist. One possible explanation for this difference is that the vascular and adrenal angiotensin receptors differ in their specificity for angiotensins. While the C-terminal may be the sole critical determinant for inducing agonist activity at peripheral receptors, it may be that the central receptors possess requirements for both N- and C-terminals. Specifically, the central receptors may require an exposed N-terminal arginine. Since Sarl-AII cannot be effectively converted to A I I I , the active form, its binding to the receptor produces an antagonist effect. Recent physiological studies from our laboratory group have examined the effects of i.c.v, applied Sarl-AII on blood pressure. The results, which at first glance appear contradictory to the iontophoretic study, indicate that Sarl-AII does produce an elevation in blood pressure that is longer lasting than that produced by All or AIII. However, unlike A l l and
AIII, Sarl-AII is effective only as a single dose producing a profound tachyphylaxis and inhibition of further angiotensin stimulation. Because Sarl-angio tensins are effectively degraded in the ventricular space 13, the activity witnessed with Sar~-AII may, in fact, result from the production of active AIII. Its longer-lasting action is most likely a consequence of Sarl-AII's resistance to peptidase action, therefore allowing it to remain in the vicinity of the receptors as preactive material for a longer period. The tachyphylaxis/inhibition could result from the actual binding of Sarl-AII itself to receptors with the delay being due to its very slow binding on rate (Erickson and Harding, unpublished observations). The major objection to the above hypothesis can be drawn from numerous behavioral studies which indicate a greater potency with regard to elevating blood pressure and stimulating drinking for All than for A I I I when injected i.c.v. 7'25. If the central angiotensin receptor does preferentially interact with AIII, then why is All more physiologically active? The solution to this dilemma is provided by two recent studies from our laboratory group. In the first study, the pressor and dipsogenic potencies of these peptides following i.v. application in rats was reexamined 27. The results clearly showed that at low doses A I I I and All had equivalent potencies and only at higher doses, similar to those used in most previous studies, was AII more potent. These data led to the suggestion that A I I I may be more labile than AII in the cerebroventricular space. Thus, at higher doses, All would act as a depot of pre-active material, while at lower doses the enhanced stability of AII would be offset by the higher affinity of AIII. The concept of an angiotensin precursor acting as an effective dipsogen is not new but has in the past focused on the latent activity of renin substrate and angiotensin 18. The idea that All is more degradation-resistant than AIII, thus acting as a more stable precursor, was recently confirmed by additional studies that measured the cerebroventricular metabolism of 125I-angiotensins 13. These studies determined that the degradation of A I I I to inactive fragments occurs at 3 times the rate measured for AII and A I I I and has an amazingly short half-life of 7.7 s in the ventricular space. These facts together suggest that any apparent lack of biological potency observed for A I I I can be traced to its very active metabolism.
304 Taken together, the results of this study support a paramount role for AIII in the brain. Furthermore, they strongly suggest that the activity witnessed with AII is dependent upon its prior conversion to AIII. ACKNOWLEDGEMENTS
This work was supported by Grants TW 01112 and REFERENCES 1 Aoyagi, T., Tobe, H., Kojima, F., Hamada, M., Takeuchi, M. and Umezawa, H., Amastatin, an inhibitor of aminopeptidase A, produced in actinomycetes, J. Antibiot., 31 (1978) 636-638. 2 Bennett Jr., J.P. and Snyder, S.H., Angiotensin II binding to mammalian brain membranes, J. Biol. Chem., 251, 23 (1976) 7423-7430. 3 Bennett, J.P. and Snyder, S.H., Receptor binding interaction of the angiotensin II antagonist, 125I-(Sarcosinel,leucineS)angiotensin II, with mammalian brain and peripheral tissues, Eur. J. Pharmacol., 67 (1980) 11-26. 4 Devynck, M.-A., Pernollet, M.G., Matthews, P.G., Khosla, M.C., Bumpus, F.M. and Myer, P., Specific receptors for des-Aspl-angiotensin II ('angiotensin III') in rat adrenals, Proc. Natl. Acad. Sci. U.S.A., 74 (1977) 4021-4032. 5 Douglas, J.G., Khosla, M.C. and Bumpus, F.M., Efficacy of octa- and heptapeptide antagonists of angiotensin II as inhibitors of angiotensin III binding in the rat adrenal glomerulosa, Endocrinology, 116 (1985) 1598-1602. 6 Felix, D. and Schlegel, W., Angiotensin receptive neurons in the subfornical organ. Structure-activity relations, Brain Research, 149 (1978) 107-116. 7 Fitzsimons, J.T., The effect on drinking of peptide precursors and of shorter chain fragments of angiotensin II injected into the rat's diencephalon, J. Physiol. (London), 214 (1971) 295-303. 8 Fitzsimons, J.T. and Kucharczyk, J., Drinking and haemodynamic changes induced in the dog by intracranial injection of components of the renin-angiotensin system, J. Physiol. (London), 276 (1978) 419-434. 9 Ganten, D., Lang, R.E., Lehmann, E. and Unger, T., Brain angiotensin: on the way to becoming a well-studied neuropeptide system, Biochem. Pharmacol., 33 (1984) 3523-3528. 10 Hall, M.M., Khosla, M.C., Khairallah, P.A. and Bumpus, F.M., Angiotensin analogs: the influence of sarcosine substituted in position 1, J. Pharmacol. Exp. Ther., 188 (1974) 222-228. 11 Harding, J.W., Imboden, H. and Felix, D., Is angiotensin III the centrally active form of angiotensin?, Experientia, 42 (1986) 706. 12 Harding, J.W., Stone, L.P. and Wright, J.W., The distribution of angiotensin II binding sites in rodent brain, Brain Research, 205 (1981) 265-274. 13 Harding, J.W., Yoshida, M.S., Dilts, R.P., Woods, T.M. and Wright, J.W., Cerebroventricular and intravascular metabolism of ~25I-angiotensins in rat, J. Neurochem., 46 (1986) 1292-1297. 14 Johnson, A.K., Neurobiology of the periventricular tissue surrounding the anteroventral third ventricle (AV3V) and its role in behavior, fluid balance, and cardiovascular control. In Smith, Galosy and Weiss (Eds.), Circulation, Neurobiology, and Behavior, Elsevier, Amsterdam. 1982, pp. 277-295.
HL 32063 from the NIH and Grant 831145 from the American Heart Association and its Washington affiliate to J.W.H., as well as Grant 3.627-0.84 from the Swiss National Science Foundation and assistance from the 'Stiftung zur F6rderung der wissenschaftlichen Forschung an der Universit/it Bern' to D.F. We would especially like to thank J. Conger for preparing this manuscript. 15 Kugler, P., On angiotensin-degrading aminopeptidases in the rat kidney. In Beck, Hild, Van Limborgh, Ortman, Pauly, and Schiebler (Eds.), Advances in Anatomy, Embryology, and Cell Biology, 76 (1982) 1-86. 16 Mendelsohn, F.A.O., Quirion, R., Saavedra, J.M., Aguilera, G. and Catt, K.J., Autoradiographic localization of angiotensin receptors in rat brain, Proc. Natl. Acad. Sei. U.S.A., 81 (1984) 1575-1579. 17 Mizutani, S., Akiyama, H., Kurauchi, O., Taira, H., Narita, O. and Tomodo, Y., In vitro degradation of angiotensin II (A-II) by human placental subcellular fractions, pregnancy sera, and purified placental aminopeptidases, Acta Endocrinol., 110 (1985) 135-139. 18 Petersen, E.P., Camara, C.G., Abhold, R.H., Wright, J.W. and Harding, J.W., Characterization of angiotensin binding in the gerbil brain using [lZ5I]angiotensin III as the radioligand, Brain Research, 321 (1984) 225-235. 19 Petersen, E.P., Camara, C.G., Abhold, R.H., Wright, J.W., Harding, J.W., Characterization of angiotensin binding in the African Green monkey, Brain Research, 341 (1985) 139-146. 20 Phillips, M.I., Brain renin-angiotensin and hypertension. In G.P. Guthrie Jr. and T.A. Kotchen (Eds.), Hypertension and the Brain, Futura, Mount Kisco, NY, 1984, pp. 63-81. 21 Regoli, D., Park, W.K. and Rioux, F., Pharmacology of angiotensin, Pharmacol. Rev., 26 (1974) 69-123. 22 Regoli, D., Riniker, D. and Brunner, H.R., The enzymatic degradation of various angiotensin II derivatives by serum, plasma, or kidney homogenates, Biochem. Pharmacol., 12 (1963) 637-646. 23 Reid, I.A., Actions of angiotensin II on the brain: mechanisms and physiological role, Am. J. Physiol., 246 (1984) F533-F543. 24 Rich, D.H., Moon, B.J. and Harbeson, S., Inhibition of aminopeptidases by amastatin and bestatin derivatives. Effect of inhibitor structure on slow-binding processes, J. Med. Chem., 27 (1984) 417-422, 25 Tonnaer, J.A.D.M., Wiegant, V.M., DeJong, M. and DeWied, D., Central effects of angiotensin on drinking and blood pressure: structure-activity relationships, Brain Research, 236 (1982) 417- 428. 26 Umezawa, H., Aoyagi, T., Suda, H., Hamada, M. and Takeuchi, T., Bestatin, an inhibitor of aminopeptidase B, produced by actinomycetes, J. Antibiot., 29 (1976) 97-99. 27 Wright, J.W., Morseth, S.L., Abhold, R.H. and Harding, J.W., Comparison of angiotensin II- and III-induced pressor action and dipsogenicity in rats, Am. J. Physiol., 249 (1985) R514-R521. 28 Wright, J.W., Sullivan, M.J. and Harding, J.W., Dysfunction of central angiotensinergic aminopeptidase activity in spontaneously hypertensive rats, Neurosci. Lett., 61 (1986) 351-356. 29 Wright, J.W., Sullivan, M.J., Petersen, E.P. and Harding, J.W., Brain angiotensin II and Ill dipsogenicity in the rabbit, Brain Research, 358 (1985) 376-379.