Neuroscience Vol. 42, No. 2, pp. 483-495, 1991

0306-4522/91 $3.00+ 0.00 Pergamon Press plc © 1991 IBRO

Printed in Great Britain

ANGIOTENSIN CONVERTING ENZYME IN THE MONKEY (MACACA FASCICULARIS) BRAIN VISUALIZED BY IN VITRO AUTORADIOGRAPHY S. Y. CHAI,*~"M. J. McKINLEY,:~ G. PAXINOS~jand F. A. O. MENDELSOHN* *University of Melbourne, Department of Medicine, Austin Hospital, Heidelberg, Victoria 3084, Australia :~Howard Florey Institute of Experimental Physiology and Medicine, Parkville, Victoria 3052, Australia §School of Psychology, University of New South Wales, Kensington, N.S.W. 2033, Australia Abstract--Anglotensin converting enzyme is localized in the monkey (Macaca fascicularis) brain by in vitro autoradiography using the radiolabelled inhibitor, [~25I]351A. This radioligand binds with high affinity and specificity to monkey cortical sections. Specific inhibitors of converting enzyme, lisinopril and perindoprilat compete for the radioligand binding to monkey cortex sections with inhibitory constants of 10 nM. High concentrations of anglotensin converting enzyme occur in most components of the basal ganglia including the caudate nucleus, putamen, the internal and external globus pallidus, nucleus accumbens, ventral pallidum and the reticular part of the substantia nigra. The distribution of converting enzyme in the caudate nucleus and putamen is heterogeneous, with prominent patches of higher activity. The patches in the caudate nucleus correspond closely with the acetylcholinesterase-poor striosomes. In the hypothalamus, very high levels of angiotensin converting enzyme occur in the median eminence and the pituitary stalk and high concentrations occur in the supraoptic and suprachiasmatic nuclei. Moderate, diffuse binding is observed in the median preoptic nucleus, the medial preoptic area, and in the anterior, lateral, ventromedial, posterior and arcuate nuclei. In the circumventricular organs, the subcommissural and subfornical organs exhibit high levels of angiotensin converting enzyme. The organum vasculosum of the lamina terminalis and the pineal body display moderate enzyme activities whereas the area postrema is devoid of labelling. The interpeduncular nucleus and, in the hippocampal formation, the molecular layer of the dentate gyrus are also intensely labelled. High levels of angiotensin converting enzyme activity are also detected throughout the cerebral cortex with laminations of higher activity corresponding to cell dense layers of the cortex. Layered binding is also present in the cerebellar cortex, with the most intense labelling in the molecular layer. Moderate concentrations of converting enzyme also occur in the paraventricular, medial habenula, lateral habenula and central median nuclei of the thalamus, the amygdala, the central gray, the locus coeruleus, the parabrachial nucleus and dorsal tegmental nucleus. The dorsal vagal complex, inferior olivary nucleus and the caudal subnucleus of the spinal trigeminal nucleus all display high levels of binding. Moderate, diffuse labelling is found throughout the reticular region and is also present in the gracile and cuneate nuclei. Although the overall distribution of angiotensin converting enzyme in the monkey brain resembles that in the rat, there are some striking differences. These include the high levels of binding throughout the monkey cerebral cortex and in the interpeduncular and suprachiasmatic nuclei.

Angiotensin converting enzyme (ACE; EC 3.4.15.1) is responsible for the formation of the biologically active octapeptide angiotensin II (Ang II) from angiotensin I, but cleaves other peptide substrates including bradykinin, [Met]- and [Leu]-enkephalin, substance P, neurotensin, beta-endorphin, dynorphin and luteinizing hormone releasing hormone (LHRH). 2s A C E distribution has been well described in the rat central nervous system by biochemical assays after microdissection, 2v immunohistochemical4,s and autoradiographic 5'2°'29studies. High concentrations of the

"['To whom correspondence should be addressed. Abbreviations: ACE, angiotensin converting enzyme; ACHE, acetylcholinesterase; Ang II, angiotensin II; EDTA, ethylenediaminetetra-acetate; LHRH, luteinizing hormone releasing hormone. 483

enzyme occur in the choroid plexus, forebrain circumventricular organs, basal ganglia and the neurosecretory hypothalamic nuclei. In some regions of the rat brain, the distribution of A C E correlated closely with the distribution of Ang II-like immunoreaetivity17 and Ang II receptors. 21 These include the forebrain circumventricular organs and the dorsal vagal complex. However, the function of ACE, particularly in brain regions where high levels of enzyme activity are not paralleled by Ang II-like immunoreactivity or Ang II receptors, such as in the basal ganglia and hippocampus, is still undetermined. The existence of this enzyme in the central nervous system of other m a m m a l i a n species, particularly the primates, is not well documented. Recently, we used the technique of in vitro autoradiography to localize ACE in the h u m a n medulla oblongata l" and basal

S.Y. CHAIet al.

484

forebrain and midbrain. 7 In this study, the complete distribution of A C E in the monkey brain is mapped by autoradiography. EXPERIMENTAL

were incubated with acetylthiocholine and ethopropazine in a sodium acetate buffer, pH 5.0, containing 0.1% copper sulphate and 0.1% glycine for 20 h. After incubation, the sections were rinsed in tap water and developed in 1% sodium sulphide for I0 min.

PROCEDURES

Radioligand The radioligand used to label ACE was [125I]351A, a tyrosyl analogue of the potent specific ACE inhibitor, lisinopril. 3° This compound was radioiodinated by the chloramine T method 15and purified on a Sephadex SP-C25 column. The binding of [t2sI]35IA to rat lung2° and caudate putamen 5 membranes and to human medulla oblongata ~a and caudate nucleus7 sections has been characterized and shown to exhibit high specificity and high affinity for ACE. vitro autoradiographic localization of angiotensin converting enzyme Two Macaca fascicularis monkeys were given a lethal In

dose of sodium pentobarbitone, their brains were removed and divided into blocks of 3 x 3 x 2 cm. The tissue blocks were then frozen by immersion into liquid xylene chilled to -40°C and stored at -80°C until used. Twenty-micrometre coronal sections were cut in a cryostat maintained at - 15°C, thaw-mounted onto gelatine-coated slides and dried under reduced pressure for 2 h. The incubation procedure was as previously described. 5,7 Briefly, following a 15 min preincubation in 10 mM sodium phosphate buffer, pH 7.4, containing 150mM sodium chloride and 0.2% bovine serum albumin, the sections were incubated for I h in a fresh aliquot of the same buffer to which had been added 0.3,uCi/ml of [125I]351A (approximately 130pM). The sections were then washed four times in phosphate buffer at 0°C for 1 min, dried under a stream of cool air and exposed to Agfa-Scopix CR-3B X-ray film for four days. Nonspecific binding was determined in parallel incubations to which 1 #M lisinopril or 1 mM EDTA were added. The sections directly adjacent to the ones used for incubations were stained with thionin or Luxol Fast Blue for the anatomical localization of brain nuclei.

Characterization of binding properties The specificity of [~25I]351A binding to monkey cortical sections was determined by competition studies with ACE inhibitors, lisinopril (Merck, Westpoint, PA) and perindoprilat (Servier Neuilly Sur Seine, France), the enkephalinase inhibitor ZincoV (Calbiochem-Behring, Los Angeles, CA), the renin inhibitors SR 42128 (Dr P. Corvol) and RI008 (Bachem, Torrance, CA) and leupeptin (Sigma, St. Louis, MO). For these competition studies, serial sections of monkey cortex were incubated with a range of concentrations, 10 -12 to 10-6M, of the inhibitors, which have been added to the incubation medium containing 130 pM of []25I]351A. The X-ray films were developed in an automatic processor and the autoradiographs generated were analysed with an Eye-Corn model 850 image processor (Spatial Data, Springfield, VA) coupled to a Dec-11/23 LSI computer. The optical densities on the autoradiographs were remapped in terms of radioactivity bound per unit area (d.p.m./mm 2) by calibration curves fitted by the computer using radioactivity standards. These tESI-radioactivity standards were prepared by adding known amounts of radioactivity to 5mmdiameter disks of 20-/zm-thick brain sections mounted on slides. The binding isotherms obtained from each of the competition curves were analysed by an iterative modelfitting computer programme LIGAND. 23

Acetylcholinesterase histochemistry Alternate sections of monkey caudate-putamen were processed for acetylcholinesterase (ACHE) using the modified method of Koelle and Friedenweld) 6 The sections

RESULTS

Characterization of binding properties of [125I]351A The radioligand, [12511351A, displayed high affinity, specific binding to monkey cortical sections. The binding is completely abolished by 1/~M lisinopril or 1 m M E D T A . The ability of the converting enzyme inhibitors lisinopril and perindoprilat, the enkephalinase inhibitor ZincoV, and the renin inhibitors, to compete for the radioligand binding to monkey cortex sections is illustrated in Fig. 1. Lisinopril and perindoprilat competed for the binding of the radioligand with inhibitory constants (K3 of 1 . 1 + 0 . 2 6 x 10 - S M and 1 . 0 + 0 . 3 3 x 10 - 8 M respectively, which are obtained from computer analysis of the competition curves (Fig. 1). The enkephalinase and renin inhibitors are ineffective at competing for the radioligand binding, even at high concentrations.

Autoradiographic localization of angiotensin converting enzyme Non-specific binding determined in the presence of I/~M lisinopril produced no visible image on the X-ray film and therefore the autoradiographs shown in Figs 2-8 represent specific binding only. Basal ganglia. Most components of the basal ganglia contain high densities of A C E (Fig. 3A-D). The caudate nucleus displays a heterogeneous pattern of enzyme distribution (Fig. 2) with prominent patches of higher densities in all but the tail portion.

o

•

.0

-

.

LOG

. [DRUG]

-

.

6

.0

M

Fig. 1. Competition curves showing displacement of binding of [t2sI]351A to monkey cortical sections by the ACE inhibitors, lisinopril (Q) and perindoprilat ( I ) . The enkephalinase inhibitor ZincoV (I-1), the renin inhibitors (O) and leupeptin ( 0 ) were ineffective in competing for the binding at 1-#M concentrations.

Angiotensin converting enzyme in monkey brain The putamen also shows a similar pattern o f binding but with less prominent patches (Fig. 2). A C E activity is also detected in striatal bridges which traverse the internal capsule, linking the caudate nucleus with the putamen (Fig. 3A-F). The nucleus accumbens displays moderate levels of A C E (Fig. 3A). As with the caudate nucleus and putamen, the internal and external divisions of the globus

pallidus demonstrate high densities of binding (Fig. 3E, F). Ventral to the anterior commissure, the ventral pallidum displays high concentrations o f A C E activity which is continuous with the globus pallidus (Fig. 3B, C). The olfactory tubercle contains low concentrations of enzyme activity whereas the islands o f Calleja magna are devoid of binding (Fig. 3A).

Abbreviations used in figures anterior commissure MS nucleus aecumbens mt Acb acs7 accessory facial nucleus MVe anterodorsal thalamic nucleus Op AD Arc arcuate nucleus anteroventral thalamic nucleus opt AV BA basoanterior amygdaloid nucleus Or BL basolateral amygdaloid nucleus ox BM basomedial amygdaloid nucleus Pa BST bed nucleus of the stria terminalis PAG central nucleus of the amygdala PeRt C fields CA1-4 of Ammon's horn PirCx CAI~ CC corpus callosum PnR ChPI choroid plexus PP CinCx cingulate cortex PrH CI claustrum PrS CLi caudal linear nucleus of the raphe Pu CMn centromedian nucleus of the thalamus Pul CN candate nucleus PVA cuneiform nucleus Py CnF cp cerebral peduncle R csc commissure of the superior colliculus Rad cuneate nucleus Re Cu ca cuneate fasciculus RMg DB diagonal band ROb DPB dorsal parabrachial nucleus RPa DR dorsal raphe nucleus RTg DTg dorsal tegmental nucleus f fornix S Fi fissure SCh gelatinosus layer of the caudal spinal SCO Ge5 trigeminal nucleus sop GPe external division of the giobus pallidus SFO GPi internal division of the giobus pallidus sm Gr gracile nucleus SNr GrCb granular layer of the cerebellum So granular layer of the dentate gyrus SO GrDG ic internal capsule Sol IC inferior colliculus sol icp inferior cerebellar peduncle Sp5 InG intermediate gray layer of the superior sp5 colliculus SuG inferior olivary nucleus IO IP interpeduncular nucleus SuVe L lateral nucleus of the amygdala TemCx LaCb lateral cerebeHar nucleus VL LC locus coeruleus VLG lateral habenula nucleus VMH LHb 11 lateral lemniscus VP lv lateral ventricle VPL MD mediodorsal thalamic nucleus VTA median eminence xscp ME MeCb medial cerebellar nucleus MHb medial habenula nucleus ZI ml medial lemniscus Zo mlf medial longitudinal fasciculus 4 MnPO median preoptic nucleus 4n MnR median raphe nucleus 7 molecular layer of the cerebellum 7n MoCb; MolCb MoDG; MolDG molecular layer of the dentate gyrus 10 MPB medial parabrachlal nucleus 12 ac

485

medial septum mammillothalamic tract medial vestibular nucleus optic nerve layer of the superior eolliculus optic tract oriens layer optic chiasm paraventricular hypothalamic nucleus periaqueductal gray parvocellular reticular nucleus piriform cortex pontine raphe nucleus peripeduncular nucleus prepositus hypoglossal nucleus presubiculum putamen pulvinar paraventricular thalamic nucleus pyramidal tract red nucleus radiatum layer reuniens nucleus raphe magnus nucleus raphe obseurus raphe pallidus nucleus reticulotegmental nucleus of the puns subiculum suprachiasmatic nucleus subcommissural organ superior cerebellar peduncle subfornieal organ stria medullaris of the thalamus reticular part of the substantia nigra supraoptic nucleus superior olivary nucleus nucleus of the solitary tract solitary tract spinal trigeminal nucleus spinal trigeminal tract superficial gray layer of the superior colliculus superior vestibular nucleus temporal cortex ventrolateral thalamic nucleus ventrolateral geniculate nucleus ventromedial hypothalamic nucleus ventral pallidum ventral posterolateral thalamic nucleus ventral tegraental area decussation of the superior cerebellar peduncle zona incerta zonal layer of the superior colliculus trochlear nucleus trochlear nerve facial nucleus facial nerve dorsal motor nucleus of the vagus hypoglossal nucleus

486

S.Y. CrtAt et al.

Fig. 2. Adjacent coronal sections through the rostral striatum showing [t2~I]351Abinding (A) and AChE staining (B). The arrows mark the positions Of the AChE-poor striosomes. The reticular part of the substantia nigra displays high levels of ACE (Fig. 4A--C). Lower levels of enzyme activity are found in the compact part of the nigra (Fig. 4B). Septum and hypothalamus. The septum and nucleus of the diagonal band display low levels of ACE (Fig. 3A). The medial preoptic area shows moderate enzyme activity with the highest levels in the median preoptic nucleus (Fig. 3B-D). Of the neurosecretory nuclei, the supraoptic nucleus demonstrates high density of binding (Fig. 3E) and the paraventricular nucleus moderate levels which are indistinguishable from the surrounding hypothalamic areas (Fig. 3E). The suprachiasmatic and arcuate nuclei and the median eminence display high levels of ACE

(Fig. 3D-F) whereas the pituitary stalk contains extremely high enzyme activity. In other parts of the hypothalamus, the anterior, lateral, ventromedial and posterior nuclei all display moderate levels of enzyme, whilst the mammillary bodies are conspicuously devoid of binding. Thalamus and epithalamus. The paraventricular and centromedian nuclei of the thalamus contain moderate ACE activity (Fig. 3E, F). The other parts of the thalamus display low but significant levels of binding, particularly in the anteroventral nucleus (Fig. 3E). The medial habenula displays a heterogeneous binding pattern with a high density medial strip and an unlabelled rostrolateral part (Fig. 4A, C). The

Fig. 3. Autoradiographs demonstrating the distribution of ACE in coronal sections through the monkey forebrain progressing from rostral (A) to caudal (F) extent. The arrowheads in A, B and D indicate the location of stria~al bridges and the asterisk in F marks the position of the fibres within the internal capsule which contain high levels of ACE. 487

Fig. 4. Autoradiographs demonstrating the distribution of ACE in coronal sections through midbrain and pons, progressing from rostra1 (A) to caudal (F) extent. 488

the monkey

Fig. 5. Autoradiographs demonstrating the distribution of ACE in coronal sections through the monkey medulla oblongata and cerebellum, progressing from rostral (A) to caudal (H) extent. 489

490

S.Y. CI-IAIet al.

lateral habenula shows moderate enzyme activity (Fig. 4A). The medial dorsal nucleus is largely devoid of binding except for its posterior extent. Amygdala and hippocampal formation. On the whole, the amygdala displays moderate levels of ACE, particularly in the basomedial, basolateral and lateral subnuclei (Fig. 3C-E). The basoanterior and central nuclei contain lower densities of binding. The most striking feature in the hippocampus is the high density of ACE activity in the molecular layer of the dentate gyrus (Fig. 6A, B). In contrast, the granular layer is devoid of enzyme activity (Fig. 6A, B). The polyform layer of the dentate gyrus and the pyramidal layer of Ammon's horn show moderate levels of the enzyme. The entorhinal cortex displays moderate and the subiculum low levels of binding (Fig. 6A, B). Visual system. The dorsal lateral geniculate nucleus is conspicuous by its lack of ACE activity. However, low enzyme activity is present in the ventral lateral geniculate nucleus (Fig. 4B). The zonal/superficial and intermediate gray layers of the superior colliculus display moderate levels of ACE whereas lower levels are found in the optic and intermediate white layers (Fig. 4B, D). The pulvinar displays low densities of the enzyme throughout its extent (Fig. 4C, D). Clearly evident are the oculomotor and trochlear nuclei which are devoid of binding.

Interpeduncular and ventral tegmental area. The interpeduncular nucleus displays high ACE activity with an extreme density ventrally (Fig. 4C). The remaining ventral tegmental area shows moderate levels of binding (Fig. 4A, B). Reticular formation and central gray. The reticular formation of the medulla exhibits moderate to low ACE activity with lower levels in the gigantocellular nucleus. The parvocellular reticular nucleus displays moderate levels of binding. The pedunculopontine tegmental nucleus and all of the lateral tegmentum at the same level show low concentrations of ACE. Moderate enzyme activity is present in the central gray with higher levels in a conspicuous median strip connecting the aqueduct with the posterior commissure (Fig. 4C). Lateral to the central gray, the locus coeruleus displays moderate ACE activity (Fig. 4F). Raphe nuclei. The dorsal, median raphe nuclei (Fig. 4D) and the raphe pallidus (Fig. 5B) show moderate to low levels of binding as did the raphe obscurus (Fig. 5C), magnus (Fig. 4F) and pontine raphe (Fig. 4E). The interfascicular and caudal linear nuclei exhibit some ACE activity (Fig. 4C). Brainstem nuclei associated with respiratory, cardiovascular and other autonomic function. The nucleus of the solitary tract displays moderate enzyme activity with an intensely labelled strip which separates it from the dorsal motor nucleus of the vagus

Fig. 6. Higher magnification of autoradiograph showing ACE distribution in the monkey hippocampus (B) and the adjacent section stained with thionin (A).

Angiotensin converting enzyme in monkey brain (Fig. 5C-E). The dorsal motor nucleus itself is moderately labelled and can be contrasted with the adjacent hypoglossal nucleus which exhibits lower activity (Fig. 5D-F). The nucleus ambiguus is devoid of binding. The medial and lateral parabrachial nuclei exhibit low levels of binding (Fig. 4F). A band of moderate labelling which appears to extend from the dorsal vagal complex ventrally to the lateral reticular nucleus is evident in several of the sections (Fig. 5D-F). A similar binding pattern is observed in the human medulla oblongata. Orofacial motor nuclei. The motor component of the tn.'geminal and the facial nuclei are negative. Somatosensory system. Most nuclei in the system display moderate to low ACE activity, including the gracilis and cuneate nuclei (Fig. 5F-H) and the principal nucleus of the trigeminal. The nucleus of the spinal trigeminal displayed moderate binding rostrally with lower levels at the caudal extent (Fig. 5B-E). In contrast, the gelatinosus subnucleus of the caudal part of the spinal nucleus of the trigeminal is quite strongly labelled (Fig. 5G, H). Auditory system. Only the medial geniculate nucleus displays moderate ACE activity in this system. The cochlear nucleus, the lateral superior olive, the nucleus of the trapezoid body, the nucleus of the lateral lemniscus and the inferior colliculus are all devoid of enzyme activity (Fig. 4B-F).

Vestibular system. The vestibular nuclei are unlabelled except for the low ACE activity in the superior vestibular nucleus, whereas the prepositus hypoglossal nucleus shows moderate enzyme activity (Fig. 5B). Precerebellar nuclei and the red nucleus. Moderate ACE activity is present in all the subdivisions of the inferior olivary nucleus (Fig. 5C-F). The red nucleus displays lower levels of ACE (Fig. 4A) and the pontine nucleus is negative. Cerebellar and cerebral cortices. Striate labelling is evident in the cerebellar cortex; the molecular layer displays intense and the granular layer low ACE activity (Figs 5 and 7A, B). The deep cerebellar nuclei display low levels of labelling except for the hileum of the lateral cerebellar nucleus (Fig. 5B-E). All of the cortical mantle displays high ACE activity throughout its layers with laminations evident especially in the occipital cortex (Fig. 8A, B). The laminar pattern of enzyme distribution is distinguishable between the cellular and fibrous layers of the cortex. The higher concentrations of ACE appear to register mainly with cell-dense layers of the cortex (Fig. 8A, B). Meninges, ventricles and periventricular structures. The choroid plexus shows low levels of ACE and, in some regions of the ependyma, moderate enzyme activity is present (Fig. 5C). High density of labelling

Fig. 7. Higher magnification of autoradiograph showing ACE distribution in the monkey cerebellum (B) with the adjacent thionin-stained section (A). NSC 4 2 / 2 ~

491

S.Y. CHAIet al.

492

Fig. 8. Higher magnification of autoradiograph showing the laminar pattern of ACE distribution in the monkey occipital cortex (B) with the adjacent thionin-stained section (A). is detected in the subcommissural (Fig. 5C, D) and subfornical organs (Fig. 3C) and moderate densities in the organum vasculosum of the lamina terminalis. The pineal gland displays moderate enzyme activity and the area postrema is totally devoid of binding. The pia mater is moderately labelled in some regions, including the area surrounding the internal capsule. Fibre tracts. All fibre tracts are generally negative except for the intense binding observed in the caudal, ventral part of the internal capsule (Fig. 3F). The binding appears to be associated with neuronal fibres although the resolution of the autoradiographs does not allow localization of the binding to the cellular level. Streaks of labelling are also observed in parts of the superior cerebellar peduncle and cerebral peduncle (Fig. 4A). Slight activity is present within the fasciculus retroflexus (Fig. 4A). The solitary tract also displays some activity but it was not clear if the binding was associated with cells within the tract. DISCUSSION

Properties of [175I]351A binding The radioligand, [125I]351A,displayed specific, high affinity binding to monkey brain sections. The specific ACE inhibitors lisinopril and perindoprilat, competed for the binding of the radioligand to monkey cortical sections with inhibitory constants of 10nM. These inhibitors appear to be slightly less

potent in monkey cortex than in post mortem human caudate-putamen or nucleus of the solitary tract sections, which have Ki values of between 1 and 9 nM for the two inhibitors. The enkephalinase and renin inhibitors were ineffective at high concentrations in competing for the radioligand binding.

Localization of angiotensin converting enzyme in the monkey brain The overall pattern of ACE distribution in the monkey brain resembles that of the rat and human, where the most striking feature is the high density of ACE in the basal ganglia. Some important differences between rat and monkey brains are the high densities of binding throughout the cerebral cortex, in the interpeduncular nucleus and the suprachiasmatic nucleus of the monkey brain which are not evident in the rat. Basal ganglia. The basal ganglia of the monkey, similar to the rat and human, s'7 display high levels of ACE activity in the striatum, globus pallidus and reticular part of the substantia nigra. In this study, ACE is also observed in striatal bridges which traversed the internal capsule, joining the caudate nucleus with the putamen and in fascicles in the fibre system which project caudally from the internal capsule to the substantia nigra. A similar pattern is also observed in the human basal ganglia. 7 Although the resolution obtained from this technique does not

Angiotensin converting enzyme in monkey brain allow localization of the enzyme at the cellular level, from the streaky pattern of binding it appears that in the primate basal ganglia, ACE may be associated with neuronal fibres within the internal capsule. It has been shown that: (a) in the rat, the enzyme appears to be transported along the striatonigral neurons ~where selective excitotoxin lesion of the striatal neurons depleted ACE in the pallidum, entopeduncular nucleus and substantia nigra 6 and (b) in human, where in Huntington's disease brains which were marked by degeneration of striatal neurons, ACE was decreased in the pallidum and the nigra. ~b In the rat, these basal ganglia structures do not contain significant quantities of Ang II-like immunoreactive neuronal elements t7 and therefore ACE is postulated to be the processing enzyme of some other neuropeptides including substance P, the enkephalins, neurotensin, beta-endorphin and dynorphin, all of which are found in moderate to high concentrations in the rat basal ganglia. 1~ It is not known whether the monkey basal ganglia, which also contain high densities of substance P- and enkephalin-immunoreactive elements, TM have Ang II. However, high densities of binding sites for Ang III have been reported in striatal membranes of the African green monkey. 24 ACE distribution in the primate striatum is heterogeneous, where patches of higher enzyme activity are surrounded by diffuse lower levels. The patches of higher activity in the monkey caudate nucleus correlate with AChE-poor striosomes.12 ACE distribution in the human striatum also observes a striosomal pattern. 7 The striosomal distribution of other neurochemical markers has also been described. These include substance P, enkephalin (which are concentrated in the striosomes) and tyrosine hydroxylase, glutamate decarboxylase, cholineacetyltransferase and somatostatin (which occurs in higher densities in the matrix compartment). 9-13 The nucleus accumbens and ventral pallidum of the rat, 5 human 7 and monkey are moderately labelled and the olfactory tubercle, subthalamic nucleus and compact part of the substantia nigra contain low to undetectable levels of the enzyme. Septum and hypothalamus. In contrast to the rat, the monkey septum and nucleus of the diagonal band are not labelled. Moderate diffuse binding is present in both the monkey and rat medial preoptic area. One consistent feature of ACE distribution in the primate and rodent hypothalami is the moderate to high levels of ACE in the supraoptic and paraventricular nuclei and the median eminence. 5,7 Similarly, in most species studied, high densities of Ang II receptors also occur in these structures. 18ag,2L22These results, together with the finding of dense Ang II-immunoreactive cell bodies and fibres in these structures in the rat, ~7 support the hypothesis of the local production of Ang II in the hypothalamic neurosecretory nuclei to regulate pituitary hormone secretion.

493

One of the most striking differences between the primate and rodent hypothalami is the extremely high density of ACE in the monkey suprachiasmatic nucleus; the rat nucleus contains only low levels of the enzyme. 5 In the other parts of the hypothalamus in the rat, human and monkey, moderate levels of binding are observed except for the mammillary nuclei, which are labelled in the rat 5 but not the primates. 7 Thalamus and epithalamus. The most significant binding in the rat and monkey thalami occurs in the midline nuclear groups, particularly the paraventricular and medial habenula nuclei. The centromedian nucleus of the monkey also contains moderate levels of ACE. Amygdala and hippocampus. The distribution of ACE in the amygdala and hippocampus is similar in all three species: moderate ACE levels are present in the basomedial and basolateral nuclei of the amygdala and a moderate to dense concentration of binding is found in the molecular layer of the hippocampal dentate gyrus. 5,7 The granular layer of the dentate gyrus and the different cellular layers of Ammon's horn all contain low to undetectable levels of the enzyme. Although ACE is present in significant levels in these structures of all three species, its role has yet to be elucidated. Midbrain. In the midbrain, high levels of ACE are found in the interpeduncular nucleus of all three species whereas the remaining ventral tegmental area contains only low levels of binding. 5'7 The ventral part of the lateral geniculate nucleus is labelled in the monkey, in contrast to the rat, where binding is present in the dorsal subnucleus. 5 The superior colliculus of the rat 5 and human 7 displayed low levels of binding whereas in the monkey moderate ACE activity is found in the superficial and intermediate gray layers (present study). The central gray of the rat and human contains moderate concentrations of ACE 5'7 with higher levels in the monkey. The medial geniculate nucleus of the monkey is moderately labelled. Brainstem. The nucleus of the solitary tract and dorsal motor nucleus of the vagus of all three species exhibit moderate densities of binding. 1a,5 The parabrachial nucleus and locus coeruleus of the monkey contain higher levels of the enzyme than in the rat and human. 5 The ambiguus nucleus is labelled in the rat 5 but not in the monkey and human, la In the rat, many of these structures are richly innervated with Ang II-immunoreactive fibres 17and contain Ang II receptors, 21 and are thought to be sites where Ang II may elicit effects on central autonomic function. Cerebral and cerebellar cortices. The most striking feature of ACE distribution in the monkey brain is the moderate to high concentration of the enzyme distributed throughout the cortex, which shows a laminar pattern. Cell-dense layers of the cortex appear to be more intensely labelled. In contrast, in the rat cortex, binding is only confined to cerebral

494

S.Y. CHA1 et al.

blood vessels: Recent evidence suggests that a series of A C E inhibitors may improve cognitive and learning processes, 2 possibly via inhibition of A C E in the cortex and hippocampus. Our finding of high levels of A C E in the monkey cortex and dentate gyrus of the hippocampus lends support to this hypothesis. A C E in the cerebellum of both the monkey and the rat assume a similar pattern of labelling with moderate levels of the enzyme in the molecular layer. Interestingly, the human cerebellar cortex contains only low levels of ACE. Ventricles and periventricular structures. The choroid plexus of the rat contains the highest concentration of A C E in the central nervous system. 5'27 However, in primates, only a low level of the enzyme is detected. The role of A C E in the rat choroid plexus is not known despite its extremely high density. Higher resolution localization revealed that, in the rat, A C E is associated with the microvilli of epithelial cells lining the surface of the ependyma. 26 In the rat, the forebrain circumventricular organs, including the subfornical organ, organum vasculosum of the lamina terminalis and the median eminence, display extremely high levels of ACE. 5'27 These are important sites where Ang II may elicit central actions on drinking, blood pressure regulation and vasopressin releaseY These structures contain intense Ang II-like immunoreactivity j7 and Ang II receptors. 2~ Since the circumventricular organs have

a deficient blood-brain barrier, A C E in these sites may act either on circulating or on centrallyproduced Ang I. In the monkey, high levels of the enzyme are found in the median eminence and subfornical organ, with lower levels in the organum vasculosum of the lamina terminalis. Therefore, the proposed angiotensin system in these rat circumventricular organs may have a parallel in the primates. However, the presence of Ang II-immunoreactivity in the monkey circumventricular organs has not been reported, although in most other species studied, the circumventricular organs contain high densities of Ang II receptors) 8'19'2j'22

CONCLUSION A C E is widely distributed in the monkey brain. Although the overall pattern of A C E distribution resembled that in rat brain there were some striking differences. This widespread distribution of A C E suggests multiple roles for the enzyme in the processing of different neuropeptides. Acknowledgements--We are grateful to Dr C. Sweet, Merck

Institute for Research, West Point, PA for the gift of 351A. This work was supported by grants from the National Health and Medical Research Council of Australia, National Heart Foundation of Australia and the Austin Hospital Medical Research Foundation.

REFERENCES

1. Aiso M., Potter W. Z. and Saavedra J. M. (1988) Axonal transport of angiotensin converting enzyme in the rat striatonigral pathway. Brain Res. 447, 195-199. la. Allen A. M., Chai S. Y., Clevers J., McKinley M. L., Paxinos G. and Mendelsohn F. A. O. (1988) Localization and characterization of angiotensin II receptor binding and angiotensin converting enzyme in the human medulla oblongata. J. comp. NeuroL 269, 249-264. lb. Arregni A., Bennett J. P., Bird E. D., Yamamura H. I., lversen L. L. and Snyder S. M. (1977) Huntington's chorea: selective depletion of activity of angiotensin converting enzyme in the corpus striatum. Ann. Neurol. 2, 294-298. 2. Barnes J. M., Barnes N. M., Costall B., Coughlan J., Horovitz Z. P., Kelly M. E., Naylor R. J. and Tomkins D. M. (1989) ACE inhibition and cognition. In Current Advances in ACE Inhibition (eds MacGregor G. M. and Severs P. S.), pp. 159-171. Churchill Livingstone, U.K. 3. Beach T. G. and McGeer E. G. (1984) The distribution of substance P in the primate basal ganglia: an immunohistochemical study of baboon and human brain. Neuroscienee 13, 29-52. 4. Brownfield M. S., Reid I. A., Ganten D. and Ganong W. F. (1982) Differential distribution of immunoreactive angiotensin and angiotensin converting enzyme in rat brain. Neuroscience 7, 1759-1769. 5. Chai S. Y., Mendelsohn F. A. O. and Paxinos G. (1987) Angiotensin converting enzyme in rat brain visualized by quantitative in vitro autoradiography. Neuroseience 20, 615-627. 6. Chai S. Y., Christie M. J., Beart P. M. and Mendelsohn F. A. O. (1987) Effects of nigral dopaminergic lesions and striatal exeitotoxin lesions on brain converting enzyme. Neurochem. Int. 10, 101-107. 7. Chai S. Y., MeKenzie J. S., McKinley M. J. and Mendelsohn F. A. O. (1990) Angiotensin converting enzyme in the human basal forebrain and midbrain visualized by in vitro autoradiography. J. eomp. Neurol. 291, 179-194. 8. Defendini R., Zimmerman E. A., Weare J. A., Albenc-Gelas F. and Erdos E. G. (1983) Angiotensin-converting enzyme in epithelial and neuroepithelial ceils. Neuroendoerinology 37, 32-40. 9. Gerfen C. R. (1985) The neostriatal mosaic: compartmental organization of projections from the striatum to the substantia nigra in the rat. J. comp. Neurol. 236, 454-476. 10. Gerfen C. R. (1987) The neostriatal mosaic: compartmental organization of mesostriatal system. In The Basal Ganglia. H--Structure and Function, Current Concepts (eds Carpenter M. B. and Jayaraman A.). Plenum Press, New York. 11. Graybiel A. M. (1984) Neurochemically specified subsystems in the basal ganglia. In Function of the BasalGanglia (eds Evered D. and O'Connor M.), Ciba Foundation Symposium 107, pp. 114-144. Pitman, London. 12. Graybiel A. M., Ragsdale C. W. and Moon-Edley S. (1979) Compartments in the striatum of the cat observed by retrograde cell-labelling. Expl Brain Res. 34, 189-195. 13. Graybiel A. M. and Chesselet M.-F. (1984) Compartmental distribution of striatal cell bodies expressing [Met]enkephalin-like immunoreactivity. Proc. natn. Acad. Sci. U.S.A. 81, 7980-7984. 14. Haber S. and Elde R. (1981) Correlation between met-enkephalin and substance P immunoreactivity in the primate globus pallidus. Neuroscience 6, 1291-1297.

Angiotensin converting enzyme in monkey brain

495

15. Hunter W. M. and Greenwood F. C. (1962) Preparation of iodine-131 labelled growth hormone of high specific activity. Nature 194, 495-496. 16. Koelle G. B. and Friedenwald J. S. (1949) A histochemical method for localizing cholinesterase activity. Proc. Soc. exp. Biol. Med. 70, 617~22. 17. Lind R. W., Swanson L. W. and Ganten D. (1985) Organization of angiotensin II immunoreactive cells and fibers in the rat central nervous system. Neuroendocrinology 40, 2~4. 18. McKinley M. J., Allen A. M., Clevers J., Denton D. A. and Mendelsohn F. A. O. (1986) Autoradiographic localization of angiotensin receptors in the sheep brain. Brain Res. 375, 373-376. 19. McKinley M. J., Allen A. M., Clevers J., Paxinos G. and Mendelsohn F. A. O. (1987) Angiotensin receptor binding in human hypothalamus: autoradiographic localization. Brain Res. 4211, 375-379. 20. Mendelsohn F. A. O. (1984) Localization of angiotensin converting enzyme in rat forebrain and other tissues by in vitro autoradiography using 125I-labelled MK351A. Clin. exp. Pharm. Physiol. I1, 431-436. 21. Mendelsohn F. A. O., Quirion R., Saavedra J. M., Aguilera G. and Catt K. J. (1984) Autoradiographic localization of angiotensin II receptors in the brain. Proc. natn. Acad. Sci. U.S.A. 81, 1575-1579. 22. Mendelsohn F. A. O., Allen A. M., Clevers J., Denton D. A., Tarjan E. and McKinley M. J. (1988) Localization of angiotensin II receptor binding in rabbit brain by in vitro autoradiography. J. comp. Neurol. 270, 372-384. 23. Munson P. J. and Rodbard D. (1980) LIGAND: a versatile computerized approach for characterization of ligand-binding systems. Analyt. Biochem. 107, 220-239. 24. Petersen E. P., Abhold R. H., Camara C. G., Wright J. W. and Harding J. W. (1985) Characterization of angiotensin binding in the African green monkey. Brain Res. 341, 139-146. 25. Phillips M. I. (1987) Functions of angiotensin in the central nervous system. A. Rev. Physiol. 49, 413-435. 26. Pickel V. M., Chan J. and Ganten D. (1986) Dual peroxidase and colloidal gold-labelling study of angiotensin converting enzyme and angiotensin-like immunoreactivity in the rat subfornical organ. J. Neurosci. 6, 2457-2469. 27. Saavedra J. M., Fernandez-Pardal J. and Chevillard C. (1982) Angiotensin converting enzyme in discrete areas of the rat forebrain and pituitary gland. Brain Res. 245, 317-325. 28. Skidgel R. A., Defendini R. and Erdos E. G. (1987) Angiotensin I converting enzyme and its role in neuropeptide metabolism. In Neuropeptides and Their Peptidases (ed. Turner A. J.), pp. 165-182. Ellis Horwood, U.K. 29. Strittmatter S. M., Lo M. M. S., Javitch J. A. and Snyder S. H. (1984) Autoradiographic localization of angiotensin-converting enzyme in rat brain with [3H]captopril: location to a striatonigral pathway. Proc. natn. Acad. Sci. U.S.A. 81, 1599-1603. 30. Sweet C. S., Gross D. M., Blaine E. H., Ulm E. H. and Stone C. A. (1981) A new class of converting enzyme inhibitors. In New Trends in Arterial Hypertension (eds Worcel M. et aL), INSERM Symposium No. 17, pp. 263-277. Elsevier/North-Holland Biomedical Press. (Accepted 31 October 1990)