PII S0361-9230(97)00343-2

Neuropeptides in the Cat Amygdala

P. MARCOS,* R. COVENAS,*1 J. A. NARVAEZ,* J. A. AGUIRRE,* G. TRAMU†AND S. GONZALEZ–BARON*

*Universidad de Malaga, Facultad de Medicina, Departamento de Fisiologia,Campus de Teatinos s/n, Malaga, Spain

†Universite de Bordeaux 1, Laboratoire de Neurocytochimie Fonctionnelle, C.N.R.S.;Talence cedex, France

[Received 3 December 1996; Revised 10 June 1997; Accepted 3 September 1997]

ABSTRACT: The distribution of seven neuropeptides was stud-ied in the cat amygdala using an indirect immunoperoxidasetechnique. No labeling was found for luteinizing hormone-re-leasing hormone or b-endorphin (1–27). Sparse a-melanocyte-stimulating hormone-immunoreactive fibers were found in thebasomedial nucleus of the amygdala, whereas a low density offibers containing a-neo-endorphin was observed in the anterioramygdaloid area. Neurotensin was observed in fibers of theanterior amygdaloid area (low density) and both the lateral (lowdensity) and the medial part (moderate density) of the centralnucleus. A low density of fibers containing neurokinin A wasfound in the anterior amygdaloid area, the basolateral nucleus,and the medial part of the central nucleus. A moderate densitywas observed in the basomedial nucleus and in the medial andcortical nuclei. Fibers containing somatostatin-28 (fragment 1–12) were observed in all the amygdaloid nuclei, whereas immu-noreactive cell bodies were found in all the nuclei except in themedial part of the central nucleus and the medial nucleus.Perikarya containing neurokinin A were observed in the latternucleus. The results point to a discrete distribution of peptider-gic fibers in the cat amygdala, as well as the occurrence ofneurons containing neurokinin A and somatostatin-28 (frag-ment 1–12). The distribution of the peptides studied in the cat iscompared with the location of the same peptides in the amyg-dala of other species. The possible diencephalic origin of thepeptidergic fibers is also discussed. © 1998 Elsevier Sci-ence Inc.

KEY WORDS: Immunocytochemistry, Cat, Amygdaloid com-plex, Somatostatin, Neurokinin A, Neurotensin.

INTRODUCTION

The amygdala has been implicated in neuroendocrine, visceral,and pain mechanisms [11] as well as in behavioral mechanisms,such as ingestion, reproduction, defense [21,22], aggression, mem-ory, and learning (see [9] for a review). The amygdala has alsobeen reported to be involved in several neurological disorders,including Alzheimer’s and Huntington’s diseases [9].

Few data are available on the distribution of peptides in the catamygdala. Substance P has been detected in neurons of the medialamygdaloid nucleus [22], and adrenocorticotropin-like immunore-

activity has been observed in fibers of the anterior amygdaloidarea, the medial division of the central nucleus, and the medialnucleus of the cat amygdaloid complex [20]. However, in othermammals (e.g., rat, monkey, humans) several immunocytochem-ical and radioimmunoassay studies of peptides in the amygdalahave been performed [1–4,19,26].

The aim of the present work was to determine the anatomicaldistribution of several peptides belonging to different families inthe cat amygdala. These substances include tachykinins (neuro-kinin A), pro-dynorphin-derived peptides (a-neo-endorphin),pro-opiomelanocortin-derived peptides (b-endorphin,a-melano-cyte-stimulating hormone), inhibitors of growth hormone (soma-tostatin), and releasing factors (luteinizing hormone-releasing hor-mone), as well as neurotensin. Our findings are discussed in lightof previous studies on the distribution of the above-mentionedpeptides in the amygdala of the rat, monkey, and humans.

MATERIALS AND METHODS

Eight male adult cats (2–3 kg body weight), obtained fromcommercial sources (CRIFFA, Barcelona, Spain), were used inthis study. Each animal was kept in a cage under standard condi-tions of light (lights on at 06.00, off at 20.00) and temperature(25°C) and had free access to food and water. The animals re-mained in their cages for 10 days before experiments, which wereconducted following institutional approval.

Four animals were deeply anesthetized with ketamine (40–50mg/kg intraperitoneally), heparinized, and perfused via the ascend-ing aorta with 500 ml of cold 0.9% NaCl. This prerinse wasimmediately followed by fixative: 3 1 of 4%paraformaldehyde in0.15 M phosphate-buffered saline (PBS) (pH 7.2). The brains wereremoved and placed in the same fixative for 12 h at 4°C. After thispostfixation, the brains were cryoprotected by immersion in su-crose baths (10–30%) until they sank. Using a cryostat, 80mm-thick frontal sections were cut, collected in PBS, and kept at 4°C.Six or seven series of adjacent sections were used for immunocy-tochemistry with different antibodies, while the remaining sectionswere stained with the Nissl technique.

The other four cats were placed under ketamine anesthesia in astereotaxic apparatus for surgery. Body temperature was main-

1 Address for correspondence: Rafael Coven˜as, Universidad de Salamanca, Facultad de Medicina, Departamento de Biologı´a Celular y Patologı´a,Laboratorio de Neuroanatomı´a de los Sistemas Peptide´rgicos, Avda. Campo Charro s/n., 37007-Salamanca, Spain, Tel: 34 23 299400; Fax: 34 23 294549;E-mail: covenas@gugu.usal.eg

Brain Research Bulletin, Vol. 45, No. 3, pp. 261–268, 1998Copyright © 1998 Elsevier Science Inc.Printed in the USA. All rights reserved

0361-9230/98 $19.001 .00

261

tained at 37°C with a feedback-controlled heating pad. A hole wasdrilled in the skull, and a glass micropipete filled with a salinesolution containing colchicine was introduced into the brain. Theseanimals received unilateral injections of colchicine (300mg dilutedin 5 ml of saline solution) in the lateral cerebral ventricle toenhance the immunoreactivity of cell bodies. After a 2-day sur-vival time, the animals were again deeply anesthetized and thensubject to perfusion and tissue processing identical to those de-scribed above.

Antisera

Polyclonal primary antibodies were raised in rabbits againsttheir respective immunogens. These were prepared by coupling thefollowing antigens to a carrier protein (human serum albumin)with glutaraldehyde: the whole synthetic peptide, in the case ofa-neo-endorphin (NEO-END), neurokinin A (NKA), neurotensin(NT), a-melanocyte-stimulating hormone (MSH), and luteinizinghormone-releasing hormone (LH-RH); peptide fragments, as theportion 1–27 of theb-endorphin (END) and the fragment 1–12 ofsomatostatin-28 (SOM). Rabbits were initially immunized withimmunogens emulsified with Freund’s adjuvant and then, at 2-week intervals, were given booster doses of incomplete Freund’sadjuvant. Plasma from rabbits was obtained 10 days after threesuch booster injections, and periodically thereafter. The immuno-logical properties of the primary antibodies have been previouslyreported, and the specificity of the immunostaining was controlledin the present study following the protocols described in previouspapers [5–8,17,18,23].

Immunocytochemistry

Free-floating sections were processed for immunostaining asdescribed previously [5–8,17,18,23]. The dilutions used for theprimary polyclonal antisera were as follows: 1/1,000 in PBS forNEO-END, NKA, NT, and SOM and 1/1,500 in PBS for MSH,LH-RH, and END. In all cases, the secondary antibodies—sheepantirabbit immunoglobulins coupled to horseradish peroxidase—were diluted 1/250 in PBS. Triton X-100 (at 0.2%) was added toPBS in all cases to enhance the penetration of antibodies; a lack ofantibody penetration in the deeper section levels was observedwhen Triton X-100 was not used. Finally, tissue-bound peroxidasewas reacted with 3,39-diaminobenzidine (DAB). To enhance thesensitivity of the reaction, the glucose oxidase-nickel-DABmethod [25] was used in some series of sections to visualizeNEO-END and MSH.

To ensure the validity of our procedure, several controls wereimplemented. Some hypothalamic sections of both cat and rat wereidentically processed at the same time as the amygdaloid sectionsof the cat brain. The immunohistochemical results obtained fromthe control sections, developed by both DAB and glucose-oxidase-nickel DAB procedures, were identical to those obtained previ-ously [5–8,19,23].

Mapping was carried out from the sections using a cameralucida according to the stereotaxic atlas of Jasper and Ajmone-Marsan [14]; the nomenclature of the anatomical structures wasderived from Krettek and Price [15].

The density of the immunoreactive fibers was graded into threecategories: high, moderate, and low. This procedure was carriedout by viewing the sections under light illumination at a constantmagnification with reference to photographs of defined series ofdensities [18]; these photographs were taken from brain sections ofcats treated similarly, and the primary antibody used in the studywas obtained in the same way as those used in the present work.The size of cell bodies was measured with the nucleus in the focalplane using a micrometer grid and verified in several series of

photographs by comparing the neuron diameter to scale barsdetermined previously. Characterization of cell bodies was carriedout according to the criterion of Ljungdahl et al. [16]. Cell profileswith the largest diameter below 15mm were termed small, thosewith a diameter of 15–25mm medium-sized, and those with adiameter above 25mm large. The number of cells measured persection in each animal is indicated in the Results section.

The degree of penetration of antibodies into the sections waschecked by focusing the different deep levels of the sections undera light microscope using a 403 lens. In all cases, immunoreactiv-ity was found in all the deep section levels.

RESULTS

Figures 1 and 2 show the distribution of the neuropeptides(SOM, NT, MSH, NEO-END, and NKA) studied in the cat amyg-dala based on the results obtained from both control and col-chicine-treated cats. Colchicine treatment was necessary to visu-alize NKA-positive cell bodies. However, SOM-like-immu-noreactive (ir) perikarya were observed without this treatment,although colchicine did increase the number of neurons containingSOM. The results were quantitatively the same with the DAB andglucose-oxidase-nickel DAB developing procedures. However, theintensity of staining was considerably increased when the glucoseoxidase-DAB method was employed in the cases in which theDAB method had yielded extremely weak labeling (NEO-END,MSH). The amygdala was devoid of LH-RH and END immuno-reactivity.

Immunohistochemical staining was always restricted to neuro-nal cell bodies and/or fibers. Glial cells were consistently devoid oflabeling.

NEO-END

A low density of NEO-END–like-immunoreactive (NEO-END-like-ir) fibers was found throughout the caudorostral exten-sion of the anterior amygdaloid area (Fig. 3A). These fibers werethin, varicose, short (10–15mm), and arborized, not showing anytopographical prevalence.

MSH

Sparse MSH-like-ir fibers were observed in the basomedialnucleus of the amygdala (Fig. 3B). At caudorostral levels (exceptat A 14.0, where we did not observe MSH-containing fibers) suchfibers were varicose, relatively thin, long (95–120mm) andstraight. In the caudalmost levels (from A 9.5 to A 11.5), MSH-like-ir fibers were distributed homogeneously throughout the sur-face of the basomedial nucleus, being restricted to its ventral partfrom levels A 12.0 to A 13.5.

NT

A low density of NT-like-ir nerve processes was detected in theanterior amygdaloid area (Fig. 3C). Thick puncta were present inits dorsalmost region, while some NT-like-ir thin fibers running inthe mediolateral direction were observed in the ventral region ofthis nucleus. In general, these thin fibers were short (10–60mm),varicose, and straight, although some relatively long fibers (90–150 mm) were also detected apparently surrounding nonimmuno-reactive cell bodies.

At all caudorostral levels the lateral part of the central nucleus(Fig. 3D) displayed a low density of NT-like-ir thin, varicosefibers. In general, these were long (90–140mm) and scatteredacross the surface of the nucleus. In addition, a few puncta con-taining NT were also present. The medial part of the centralnucleus (Fig. 3E) also consistently showed moderately dense NT-

262 MARCOS ET AL.

like-ir fibers along its caudorostral extension; positive puncta werevery abundant around nonimmunoreactive cell bodies. A few short(20–60mm), varicose, and arborizing NT-containing fibers werealso detected in the central nucleus, running in the lateromedialdirection.

NKA

The anterior amygdaloid area displayed a low density of short(10–50 mm), thick, nonvaricose fibers weakly NKA-like-ir

throughout its caudorostral extension; these were principally lo-cated in the dorsal-most region of the nucleus.

A low density of NKA-positive puncta was detected in both theanterior and posterior divisions of the basolateral nucleus, mainlydistributed at the periphery of those nuclei. A few short (10–50mm), thin NKA-like-ir fibers were also observed in both divisionsof the basolateral nucleus.

Similar types of NKA-containing nerve elements were sparselydistributed in the medial part of the central nucleus, although morehomogeneously than in the basolateral nucleus.

Both the anterior (Fig. 3F) and posterior cortical nuclei of theamygdala displayed a moderate density of NKA-like-ir puncta,

FIG. 1. Distribution of NEO-END- and MSH-like-ir fibers in frontal planesof the cat amygdala from caudal (A) to rostral levels (E) corresponding tothe Jasper and Ajmone–Marsan stereotaxic atlas [14]. In this and thefollowing figures the anterior level (in mm) with respect to the zerostereotaxic point of each section is indicated at the lower right. Immuno-reactive fibers are represented by varicose lines. AAA, anterior amygdaloidarea; BL, basolateral nucleus of the amygdala; Bla, anterior division of thebasolateral nucleus of the amygdala; Blp, posterior division of the baso-lateral nucleus of the amygdala; BM, basomedial nucleus of the amygdala;Cel, lateral part of the central nucleus of the amygdala; Cem, medial part ofthe central nucleus of the amygdala; Coa, anterior cortical nucleus of theamygdala; Cop, posterior cortical nucleus of the amygdala; D, dorsal; La,Lateral nucleus of the amygdala; M, medial; Me, medial nucleus of theamygdala.

FIG. 2. Distribution of SOM-, NKA-, and NT-like-ir structures in frontalplanes of the cat amygdala from caudal (A) to rostral levels (E) corre-sponding to the Jasper and Ajmone–Marsan stereotaxic atlas [14]. Immu-noreactive fibers and puncta are represented, respectively, by continuousand/or varicose lines and small dots, whereas the localization of perikaryais indicated by dots. In the case of NKA, each dot represents five stainedcell bodies. For the SOM-like-ir neurons, the quantitative correspondencebetween dots and cells is established as follows: in the cortical nucleus,1:10, whereas at A 13.5 correspondences are 1:3 in the lateral nucleus, and1:4 in the basolateral nucleus. In the remaining levels, for all the amyg-daloid nuclei, the correspondence is approximately 1:1. Abbreviations areused as in Fig. 1.

NEUROPEPTIDES IN THE CAT AMYGDALA 263

homogeneously distributed throughout these nuclei. A few short(15–50 mm) and varicose fibers containing NKA were also de-tected in both nuclei.

The basomedial and medial nuclei showed a low density ofNKA-like-ir fibers caudally. From A 12.0 to more rostral levels, adensity increasing from low to moderate was detected in thesenuclei. In the basomedial nucleus, distinct positive fibers wereseen: in its ventral part NKA-like-ir puncta were the most commontype of positive nerve structures, while in its dorsal region long

(90–100mm), thin and varicose fibers were observed running inthe dorsoventral direction. Fibers in the medial nucleus weredistributed more homogeneously than in the basomedial nucleus.In general, these NKA-like-ir fibers were thin, very short (10–35mm), and straight. In this nucleus, five NKA-positive cell bodiesper section were detected at those levels in which the density offibers was moderate. These neurons were only observed aftercolchicine treatment. The majority of such cell bodies containingNKA were round or elongated, small (12mm in diameter), and

FIG. 3. Immunoreactive fibers, stained for different neuropeptides, observed in several nuclei of the catamygdaloid complex. (A) Immunoreactive fibers containing NEO-END in the anterior amygdaloid area. (B)MSH-like-ir fibers in the basomedial nucleus of the amygdala. (C) NT-like-ir fibers in the anterior amygdaloidarea. (D) Immunoreactive fibers containing NT in the lateral part of the central nucleus of the amygdala. (E)NT-like-ir fibers in the medial part of the central nucleus of the amygdala. (F) NKA-like-ir fibers in the anteriorcortical nucleus of the amygdala. Scale bars: 50mm.

264 MARCOS ET AL.

characterized by one or two stained processes of limited extent(Fig. 5A, and C), although one or two medium-sized perikaryawere also observed (Fig. 5B).

SOM

All the amygdaloid nuclei showed SOM-like-ir fibers. Exceptin the medial nucleus and the medial part of the central nucleus,SOM-containing cell bodies were detected in the amygdaloid

complex. Cell bodies were visualized without colchicine treat-ment, although injections of colchicine revealed a greater numberof SOM-positive neurons.

Throughout its caudorostral extension, the anterior amygdaloidarea (Fig. 4A) displayed a moderate density of homogeneouslydistributed SOM-like-ir nerve elements. Several types of SOM-immunostained nerve processes were observed: puncta; thin, long(90–115mm), varicose, arborizing fibers; and thick, short (15–40

FIG. 4. SOM-like-ir fibers and perikarya in the cat amygdala. (A.) Immunoreactive fibers and cell bodies(arrows) in the anterior amygdaloid area. (B.) SOM-like-ir fibers and perikarya (arrows) in the posteriordivision of the basolateral nucleus of the amygdala. (C.) Immunoreactive fibers and cell bodies (arrows)containing SOM in the lateral nucleus of the amygdala. (D.) Immunoreactive fibers in the medial part of thecentral nucleus of the amygdala. (E.) Cell bodies (arrows) and fibers containing SOM in the lateral part of thecentral nucleus of the amygdala. (F.) SOM-like-ir fibers and perikarya (arrows) in the basomedial nucleus ofthe amygdala. Scale bars: 50mm.

NEUROPEPTIDES IN THE CAT AMYGDALA 265

mm), varicose fibers. Ten SOM-like-ir small (15mm in diameter)cell bodies per section were also detected in the anterior amygdal-oid area, displaying several morphologies: monopolar, bipolar,round, and elongated. These cell bodies emitted two or threeprocesses of limited extent. In some cases, SOM-positive terminal-like elements, of unknown origin, appeared to be in appositionwith these SOM-like-ir cell bodies.

In both the anterior and posterior cortical amygdaloid nuclei amoderate density of SOM-positive fibers was detected. Thin, var-icose, and arborized fibers were homogeneously distributedthroughout these nuclei with no preferential orientation, even whensuch SOM-like-ir fibers were long (90–115mm). The 10 cellbodies per section containing SOM observed in these nuclei wereround, small (15mm) (Fig. 5E and F) or medium-sized (17mm)(Fig. 5D), and were normally featured by one or two processes.

The medial nucleus of the amygdala showed a low density ofSOM-like-ir fibers throughout its caudorostral extension. The ma-jority of such nerve fibers were thin, varicose, long (90–110mm),and arborized, although some shorter and straight fibers were alsoobserved.

In the basolateral nucleus, a caudorostral increase in the densityof SOM-like-ir fibers was detected. At the periphery of its poste-rior division (Fig. 4B), a low density of varicose, thin, relativelylong (90–100mm) fibers was observed; a few puncta containingSOM were also detected surrounding some non-immunoreactivecell bodies. The anterior division of the basolateral nucleus dis-played a moderate density of SOM-like-ir fibers, similar in appear-ance to those described in the posterior division but distributedmore homogeneously. In addition, some thick, short (15–60mm),varicose, arborizing SOM-containing fibers were also observed in

the anterior division. Small SOM-like-ir neurons (15mm) werepresent at a density of eight cell bodies per section in bothdivisions of the basolateral nucleus, being principally located attheir peripheral regions, and scarce (two per section) medium-sized perikarya (17mm in diameter) were detected in the dorsal-most part of the nucleus. These neurons were round or elongatedand generally monopolar, but some perikarya emitted two shortstained processes.

From caudal to rostral levels, the lateral nucleus (Fig. 4C)displayed a low to moderate increase in the density of SOM-containing nerve elements. In the ventral part of the nucleus, thin,varicose, very short (15–25mm), straight fibers and SOM-like-irpuncta were distributed with a similar density. In the dorsal part ofthe lateral nucleus, SOM-like-ir fibers were predominantly long(90–100mm), very thin, nonvaricose, and oriented in the dorso-ventral direction. This fiber arrangement was observed throughoutthe caudorostral extension of the lateral nucleus. SOM-positivecell bodies were restricted to the ventralmost region of the nucleus.Perikarya (nine per section) were round, monopolar (although afew bipolar neurons were also observed), and small (12mm).

A high density of SOM-like-ir puncta, intermingled with thick,short (15–55mm), nonvaricose immunoreactive fibers was de-tected in the medial part of the central nucleus (Fig. 4D) through-out its caudorostral extension. These fibers did not display anypreferential orientation.

Similar types of SOM-positive nerve elements were detected inthe lateral part of the central nucleus (Fig. 4E), albeit with amoderate density. However, in this nucleus, SOM-stained punctawere very abundant, being more concentrated in the central regionof the nucleus. At the periphery of the central nucleus, five SOM-

FIG. 5. Camera lucida drawings of cell bodies containing NKA (A, B, C) and SOM (D, E, F). Thelocalization of the immunoreactive perikarya in the amygdaloid sections is indicated by letters. (A, B,C): Several morphologies of NKA-like-ir neurons detected in the medial nucleus of the amygdala (seetext for details). (D, E, F): SOM-like-ir neurons observed in the anterior cortical nucleus of the catamygdala. Scale bar: 25mm.

266 MARCOS ET AL.

like-ir cell bodies per section were observed, and two neurons weredetected in the central region of the nucleus. Neurons containingSOM were round or elongated and small (13mm) and exhibiteddendritic processes.

A moderate density of SOM-like-ir fibers, similar to the threetypes described in the basolateral nucleus, was observed through-out the caudorostral extension of the basomedial nucleus of theamygdala (Fig. 4F). These stained elements did not display apreferential orientation and were distributed homogeneously.SOM-containing perikarya (eight cell bodies per section) werepresent in the basomedial nucleus. These cell bodies were round orelongated, were medium-sized (22mm), and exhibited at least twodendritic processes.

DISCUSSION

We have described for the first time the distribution of NKA-like-ir fibers and perikarya in the mammalian amygdala. More-over, the present work reports the distribution of MSH-, NT-,SOM-, NEO-END-like-ir fibers and cell bodies in the cat amyg-dala using an immunoperoxidase technique. A considerable diver-sity in the immunohistochemical distribution of the above-men-tioned peptides has been reported in the amygdaloid complex ofthe cat.

Comparison of the Distribution of Neuropeptides in theMammalian Amygdala

NEO-END. Using radioimmunoassay techniques, NEO-ENDhas been detected in the following amygdaloid nuclei of the rat:basomedial, basolateral, central, cortical, lateral, and medial (see[19]). In the cat, by contrast, no NEO-END-like-ir structures wereobserved in these nuclei. However, in the present study the anterioramygdaloid area was found to contain NEO-END-like-ir fibers,not detected in this nucleus in the rat (see [19]). In addition,NEO-END-like-ir cell bodies have been detected in the centralnucleus of the amygdala in rats using immunofluorescence tech-niques [19]. In the cat, on the other hand, we did not detectNEO-END-like-ir neurons in the amygdaloid complex.

MSH.Comparing our results with previous studies carried outon the distribution of MSH in the rat amygdala (see [19]), thispeptide displays a more widespread distribution in the rat than inthe cat amygdala. Thus, in both species MSH-like-ir fibers havebeen found in the basomedial nucleus. However, MSH has alsobeen detected in the basolateral, central, cortical, and medial nucleiof the rat amygdala, whereas no MSH-like-ir structures wereobserved in these nuclei in the cat.

NT.The distribution of NT-like-ir structures in the amygdala ofboth rat (see [19]) and humans [3] is more widespread than thatobserved in the cat. In all three species, NT-containing fibers havebeen found in the central nucleus of the amygdala and the anterioramygdaloid area. However, in both the rat and humans NT-like-irstructures have been found in the basal, cortical, lateral, and medialnuclei, while no NT-positive elements were detected in theseregions in the cat.

NKA. Our study reports the distribution of NKA-like-ir fibersand cell bodies in the mammalian amygdala (anterior amygdaloidarea; basomedial, basolateral, cortical, and medial nuclei; medialpart of the central nucleus of the amygdala) for the first time,because no previous study has been carried out on the distributionof fibers and perikarya containing NKA in the amygdala of the rat,monkey, or humans.

In comparison with a previous study on the cellular localizationof substance P and NKA-encoding pre-pro-tachykinin (PPT)mRNA in the female rat brain [13], PPT cell bodies seem to bemore widely distributed in the rat amygdala than in the same

region of the cat. Thus, PPT neurons in the rat and NKA-like-irperikarya in the cat were observed in the medial nucleus of theamygdala. However, in the rat, Harlan et al. [13] observed PPTneurons in the anterior amygdaloid area and in the central, cortical,and lateral nuclei of the amygdala, whereas no NKA-like-irperikarya were found in these nuclei in the cat in the present study.These discrepancies could be due to technical parameters and/orspecies/sex differences.

SOM. In general, the distribution of SOM-like-ir structures inthe amygdala of the cat, the monkey, and humans is quite similar[4,10]. Moreover, in the rat amygdala the highest levels of SOMwere detected using radioimmunoassay [2]. However, some minordifferences merit comment. For example, in the cat no SOM-like-ircell bodies were found in the medial part of the central nucleus orin the medial nucleus of the amygdala, whereas in both nucleiSOM-containing perikarya have been found in humans [4].

LH-RH. No LH-RH-like-ir structures were observed in the catamygdala in the present study. However, in the rat, LH-RH hasbeen detected in the central, cortical, medial, and lateral nuclei ofthe amygdala (see [19]).

END. The cat amygdala was not found to contain END-like-irstructures. However, in the rat and in humansb-endorphin (1–31)has been detected in the amygdala using radioimmunoassay tech-niques (see [12,19]).

In general, in the present study the neuropeptides studied dis-played a more restricted distribution in the amygdala of the catthan that described in the rat and humans in previous studies[3,4,19]. In both cats and rats, the distribution pattern of NEO-END [19]. MSH [19], or NT [19] was different using similarprocedures (pretreatment with colchicine, immunocytochemicalmethods). Moreover, the distribution of NT-like-ir [3] or SOM-positive [4] structures in the amygdala of humans has been foundto be more widespread than in the cat, even when the latter waspretreated with colchicine.

The methodological procedure applied in this study reveals arich variety in staining patterns within the cat amygdala for thedifferent neuropeptides examined. This wide variety of labelingpatterns, in addition to the many functions in which the amygdalahas been implicated, highlights the difficulty of assigning partic-ular functional roles to the neuropeptides studied. Additionally, thelocalization of several neuropeptides in the same nucleus, repeat-edly observed in the present work, makes a description of theirphysiological functions even more difficult, owing to their possiblecoexistence in the same structure and/or intricate interactionsamong them.

Possible Neuropeptide-Containing Pathways in the CatAmygdala

The origin of NT-, NKA-, and SOM-like-ir fibers in the amyg-daloid complex of the cat, as well as whether the cell bodiescontaining NKA and SOM are local or projecting neurons remainto be elucidated. According to the present morphological data, itwould appear that the medial nucleus and the medial part of thecentral nucleus receive afferents containing SOM, because in thesenuclei SOM-like-ir fibers, but no immunoreactive cell bodies, werefound. Moreover, the anterior amygdaloid area and the centralnucleus of the amygdala might receive neurotensinergic as well asNKA-positive afferents, because in these nuclei immunoreactivityfor both NT and NKA was observed only in fibers.

These data, in agreement with previous studies on the afferentconnections of the central nucleus of the cat amygdala [24] and thedistribution of NKA-, NT-, and SOM-like-ir structures in the catdiencephalon [7,8,23], allow us to speculate about the type ofneuropeptide possibly involved in the inputs to the amygdala.

NEUROPEPTIDES IN THE CAT AMYGDALA 267

Perikarya located in the central medial nucleus of the cat thalamussend afferents to the central nucleus of the amygdala [24]. Thispathway could be neurotensinergic, because in the cat medialdorsal nucleus we observed a high density of NT-like-ir cell bodiesand a low density of immunoreactive fibers [8]. Moreover, anNKA-containing pathway to the central nucleus of the amygdala[24] could derive from the lateral habenular nucleus, in which ahigh density of NKA-like-ir cell bodies and a low density of fiberswere found [23]. Finally, the central amygdaloid nucleus couldreceive afferents containing SOM from the anterior hypothalamusand the subparafascicular nucleus, because in these diencephalicnuclei we found a high density of SOM-like-ir cell bodies, but noimmunoreactive fibers [7]. This is in agreement with results show-ing that the central nucleus of the cat amygdala receives affer-ents from the anterior hypothalamus and the subparafascicular nu-cleus [24].

ACKNOWLEDGEMENTS

P. Marcos is supported by the D. G.I.C.Y.T., Spain. The authors wishto thank N. Skinner and Prof. Cooper R. Mackin for revision of the Englishtext. This work has been supported by the D. G.I.C.Y.T. (PB93/0992 andPB 96/1467), Spain.

REFERENCES

1. Amaral, D. G.; Avendan˜o, C.; Benoit, R. Distribution of somatostatin-like immunoreactivity in the monkey amygdala. J. Comp. Neurol.284:294–313; 1989.

2. Benoit, R.; Ling, N.; Bakhit, Ch.; Morrison, J. H.; Alford, B.; Guille-min, R. Somatostatin-28 (1–12)-like immunoreactivity in the rat. En-docrinology 111:2149–2151; 1982.

3. Benzing, W. C.; Mufson, E. J.; Jennes, L.; Stopa, E. G.; Armstrong,D. M. Distribution of neurotensin immunoreactivity within the humanamygdaloid complex: A comparison with acetylcholinesterase- andNissl-stained tissue sections. J. Comp. Neurol. 317:283–297; 1992.

4. Bouras, C.; Magistretti, P. J.; Morrison, J. H.; Constantinidis, J. Animmunohistochemical study of pro-somatostatin-derived peptides inthe human brain. Neuroscience 22:781–800; 1987.

5. Covenas, R.; de Leo´n, M.; Narvaez, J. A.; Tramu, G.; Aguirre, J. A.;Gonzalez–Baron, S. An immunohistochemical mapping ofb-endor-phin (1–27) in the cat diencephalon. Neuropeptides 30:261–271; 1996.

6. Covenas, R.; de Leo´n, M.; Narvaez, J. A.; Tramu, G.; Aguirre, J. A.;Gonzalez–Baron, S. Mapping ofa-melanocyte-stimulating hormone-like immunoreactivity in the cat diencephalon. Peptides 17:845–852;1996.

7. de Leon, M.; Covenas, R.; Narva´ez, J. A.; Tramu, G.; Aguirre, J. A.;Gonzalez–Baron, S. Somatostatin-28 (1–12)-like immunoreactivity inthe cat diencephalon. Neuropeptides 19:107–117; 1991.

8. de Leon, M.; Covenas, R.; Narva´ez, J. A.; Tramu, G.; Aguirre, J. A.;Gonzalez–Baron, S. Neurotensin-like immunoreactivity in the dien-cephalon of the adult male cat. Peptides 12:257–264; 1991.

9. De Olmos, J. Amygdala. In: Paxinos, G., ed. The human nervoussystem. San Diego: Academic Press; 1990:583–710.

10. Desjardins, C.; Parent, A. Distribution of somatostatin-immunoreac-tivity in the forebrain of the squirrel monkey: Basal ganglia andamygdala. Neuroscience 47:115–133; 1992.

11. Fox, R. J.; Sorenson, C. A. Bilateral lesions of the amygdala attenuateanalgesia induced by diverse environmental changes. Brain Res. 648:215–221; 1994.

12. Gramsch, Ch.; Ho¨llt, V.; Meharein, P.; Pasi, A.; Herz, A. Regionaldistribution of methionine-enkephalin- andb-endorphin-like immuno-reactivity in the human brain and pituitary. Brain Res. 171:261–270;1979.

13. Harlan, R. E.; Garcı´a, M. M.; Krause, J. E. Cellular localization ofsubstance P- and neurokinin A-encoding pre-protachykinin m-RNA inthe female rat brain. J. Comp. Neurol. 287:179–212; 1989.

14. Jasper, H. H.; Ajmone–Marsan, C. A. A stereotaxic atlas of thediencephalon of the cat. Ottawa: National Research Council of Can-ada; 1966.

15. Krettek, J. E.; Price, J. L. A description of the amygdaloid complex inthe rat and cat with observations on intra-amygdaloid axonal connec-tions. J. Comp. Neurol. 178:255–280; 1978.

16. Ljungdahl, A.; Ho¨kfelt, T.; Nilsson, G. Distribution of substance P-likeimmunoreactivity in the central nervous system of the rat. I—Cellbodies and nerve terminals. Neuroscience 3:861–943; 1978.

17. Marcos, P.; Coven˜as, R.; de Leo´n, M.; Narvaez, J. A.; Tramu, G.;Aguirre, J. A.; Gonza´lez–Baron, S. Neurokinin A-like immunoreac-tivity in the cat brainstem. Neuropeptides 25:105–114; 1993.

18. Marcos, P.; Coven˜as, R.; Narva´ez, J. A.; Tramu, G.; Aguirre, J. A.;Gonzalez–Baron, S. Alpha-neo-endorphin-like immunoreactivity inthe cat brainstem. Peptides 14:1263–1269; 1993.

19. Palkovits, M. Neuropeptides in the brain. In: Martini, L.; Ganong,W. F., eds. Frontiers in neuroendocrinology, vol. 10. New York:Raven Press; 1988:1–44.

20. Rao, J. K.; Hu, H.; Prasad, C.; Jarayaman, A. The distribution patternof adrenocorticotropin-like immunoreactivity in the cat central ner-vous system. Neurosci. Lett. 71:48–52; 1986.

21. Shaikh, M. B.; Siegel, A. Neuroanatomical and neurochemical mech-anisms underlying amygdaloid control of defensive rage behavior inthe cat. Braz. J. Med. Biol. Res. 27:2759–2779; 1994.

22. Shaikh, M. B.; Steinberg, A.; Siegel, A. Evidence that substance P isutilized in medial amygdaloid facilitation of defensive rage behavior inthe cat. Brain Res. 625:283–294; 1993.

23. Velasco, A.; de Leo´n, M.; Covenas, R.; Marcos, P.; Narva´ez, J. A.;Tramu, G.; Aguirre, J. A.; Gonza´lez–Baron, S. Distribution of neuro-kinin A in the cat diencephalon: An immunocytochemical study. BrainRes. Bull. 31:279–285; 1993.

24. Volz, H.-P.; Rehbein, G.; Triepel, J.; Knuepfer, M. M.; Stumpf, H.;Stock, G. Afferent connections of the nucleus centralis amygdalae. Ahorseradish peroxidase study and literature survey. Anat. Embryol.181:177–194; 1990.

25. Shu, S.; Ju, G.; Fan, L. The glucose oxidase-DAB-nickel method inperoxidase histochemistry of the nervous system. Neurosci. Lett. 85:169–171; 1988.

26. Zaphiropoulos, A.; Charnay, Y.; Vallet, P.; Constantinidis, J.; Bouras,C. Immunohistochemical distribution of corticotropin-like intermedi-ate lobe peptide (CLIP) immunoreactivity in the human brain. BrainRes. Bull. 26:99–111; 1991.

268 MARCOS ET AL.