Amygdalostriatal Projections in Reptiles: a Tract-Tracing Study in the Lizard Podarcis Hispanica

THE JOURNAL OF COMPARATIVE NEUROLOGY 479:287–308 (2004)

Amygdalostriatal Projections in Reptiles: A Tract-Tracing Study in the Lizard Podarcis hispanica

AMPARO NOVEJARQUE,1 ENRIQUE LANUZA,2 AND FERNANDO MARTI´NEZ-GARCI´A1* 1Departament de Biologia Funcional i Antropologia Fı´sica, Facultat de Cie`ncies Biolo`giques, Universitat de Vale`ncia, ES-46100 Vale`ncia, Spain 2Departament de Biologia Cellular, Facultat de Cie`ncies Biolo`giques, Universitat de Vale`ncia, ES-46100 Vale`ncia, Spain

ABSTRACT Whereas the lacertilian anterior dorsal ventricular ridge contains unimodal sensory areas, its posterior part (PDVR) is an associative center that projects to the hypothalamus, thus being comparable to the amygdaloid formation. To further understand the organization of the reptilian cerebral hemispheres, we have used anterograde and retrograde tracing techniques to study the projections from the PDVR and adjoining areas (dorsolateral amygdala, DLA; deep lateral cortex, dLC; nucleus sphericus, NS) to the striatum in the lizard Podarcis hispanica. This information is complemented with a detailed description of the organization of the basal telencephalon of Podarcis. The caudal aspect of the dorsal ventricular ridge projects nontopographically mainly (but not exclusively) to the ventral striatum. The NS projects bilaterally (with strong ipsilateral dominance) to the nucleus accumbens, thus recalling the posteromedial cortical amygdala of mammals. The PDVR (especially its lateral aspect) and the dLC project massively to a continuum of structures connecting the striatoamygdaloid transition area (SAT) and the nucleus accumbens (rostrally), the projection arising from the dLC being probably bilateral. Finally, the DLA projects massively and bilaterally to both the ventral and dorsal striatum, including the SAT. Our findings lend further support to the view that the PDVR and neighboring structures constitute the reptilian basolateral amygdala and indicate that an emotional brain was already present in the ancestral amniote. These results are important to understand the comparative significance of the caudal aspect of the amniote cerebral hemispheres, and specifically challenge current views on the nature of the avian caudal neostriatum. J. Comp. Neurol. 479:287–308, 2004. © 2004 Wiley-Liss, Inc.

Indexing terms: striatum; reptiles; comparative neuroanatomy; emotional brain

The telencephalon of reptiles, like the one of birds, dis- Verkley, 1978; Ulinski, 1983; Bruce and Butler, 1984b). As plays a subventricular structure just dorsal to the stria- a consequence, the DVR was compared with the sensory tum that protrudes into the ventricle, thus known as dorsal ventricular ridge (DVR; see Ulinski, 1983). The comparative significance of this structure, i.e., its counterpart in the mammalian brain, has been the center of a Grant sponsor: Spanish MCyT/FEDER; Grant number: BFI2001-3535; Grant sponsor: Universitat de Vale`ncia, Programa Cinc Segles. long debate (see Lohman and Smeets, 1990; Karten, 1991, *Correspondence to: Fernando Martıńez-Garcıá, Departament de Biolo- 1997; Butler, 1994; Striedter, 1997; Kaas and Reiner, gia Funcional i Antropologia Fı´sica (Unitat de Morfologia Microsco`pica), 1999). The use of tract tracing methods in reptiles and Facultat de Cie`ncies Biolo`giques, Universitat de Vale`ncia, C. Dr. Moliner, birds demonstrated that the DVR is the target of ascend- 50, ES-46100, Burjassot, Vale`ncia, Spain. E-mail: fernando.mtnez- [email protected] ing projections from at least three different dorsal tha- Received 20 February 2004; Revised 17 June 2004; Accepted 14 July lamic nuclei usually considered the putative relays of vi- 2004 sual, auditory, and somatosensory information (Karten, DOI 10.1002/cne.20309 1969; Karten et al., 1973; Lohman and van Woerden- Published online in Wiley InterScience (www.interscience.wiley.com).

© 2004 WILEY-LISS, INC. 288 A. NOVEJARQUE ET AL. neocortex of mammals or to some neuronal populations The alternative views by Bruce and Neary (1995b) and within it (Karten, 1991; Powers and Reiner, 1993; Butler, Striedter (1997) can be reconciled using hodological data 1994)., in lizards that indicate the presence of two functionally In reptiles, the DVR gives rise to descending projections distinct anteroposterior divisions in the DVR (Lanuza et to the underlying dorsal striatum (Hoogland, 1977; Ulin- al., 1998). The anterior DVR (ADVR), apparently contains ski, 1978; Voneida and Sligar, 1979; Gonza´lez et al., 1990), three unimodal sensory areas (auditory, medial; somato- which are commonly seen as the link between the telen- sensory, intermediate; visual, lateral; see Ulinski, 1983), cephalic sensory centers and the motor system, thus me- whereas the posterior DVR (PDVR) seems to be an asso- diating sensorimotor integration (see Ulinski, 1983). In ciative structure that receives convergent projections from line with the data on the thalamic afferents, these projec- the three sensory fields of the ADVR (Andreu et al., 1996; tions were interpreted as equivalent to the corticostriatal Lanuza et al, 1998). The intratelencephalic and extrate- circuits of the mammalian brain. lencephalic afferents of the PDVR (Belekhova and Chk- Recent data on the development, topology, connections, heidze, 1991, 1992; Andreu et al, 1996; Lanuza et al., and expression of homeotic genes in the telencephalon of 1998) together with its projections to the hypothalamus mammals and nonmammalian amniotes have seriously (Bruce and Neary, 1995a; Lanuza et al., 1997; Martıńez- challenged this view. First, studies on the expression of Marcos et al., 1999) strongly suggest that the PDVR is homeotic genes during development (Smith-Fernańdez et part of the reptilian basolateral amygdala. al., 1998; Puelles et al., 2000) indicate that the DVR is not In view of this parcellation of the reptilian DVR, the homologous to the neocortex but to parts of the lateral extent and significance of the projections from the DVR to pallium (piriform and/or endopiriform) plus a region just the basal ganglia needs to be reassessed. The aim of the ventral to it and dorsal to the striatum. This area, which present study is to characterize the striatal projections of has been called ventral pallium (Puelles et al., 2000), the caudal aspect of the dorsal ventricular ridge in the Old apparently composes parts of the mammalian claus- World lizard Podarcis hispanica. To do so, we have first troamygdaloid complex, including the ventral endopiri- studied the cyto- and chemoarchitecture of the basal fore- form area and parts of the basolateral amygdala. This brain of Podarcis. Then, we have analyzed the antero- interpretation partially agrees with the hypothesis by grade transport resulting from tracer injections in the Bruce and Neary (Bruce and Neary, 1995b), according to PDVR and adjacent areas, as well as the retrograde label- which the reptilian DVR is comparable to the lateral ing after tracer injections in the basal ganglia. The results amygdala of mammals. An alternative hypothesis, which of these experiments indicate the presence of important also fits partially the data of expression of homeotic genes, projections from the putative basolateral amygdala was already put forward by Holmgren (1925) and has been (PDVR and adjacent structures) to different parts of the revisited recently by Striedter (1997), who proposed that striatum, including the nucleus accumbens, which would the DVR of the sauropsidian forebrain is comparable to be equivalent to the amygdalostriatal projections present the mammalian claustroendopiriform complex. in the mammalian telencephalon. These findings have

Abbreviations

ABC avidin–biotin complex Nac nucleus of the anterior commissure ac anterior commissure Naot nucleus of the accessory olfactory tract Acb nucleus accumbens Nmfb bed nucleus of the medial forebrain bundle AChase acetyl cholinesterase NS nucleus sphericus ADVR anterior dorsal ventricular ridge OT olfactory tubercle AONv ventral anterior olfactory nucleus PAP peroxidase–antiperoxidase method aot accessory olfactory tract PB phosphate buffer BDA biotinylated dextran amine PDVR posterior dorsal ventricular ridge BST bed nucleus of the stria terminalis PHA-L Phaseolus vulgaris leucoagglutinin BSTl bed nucleus of the stria terminalis, lateral PT pallial thickening BSTm bed nucleus of the stria terminalis, medial PVA periventricular area Љ DAB 3,3 -diaminobenzidine RDA tetramethylrhodamine-labeled dextran amine DBN diagonal band nucleus RC retrochiasmatic area DBNh diagonal band nucleus, horizontal limb S septum DBNv diagonal band nucleus, vertical limb SAT striatoamygdaloid transition area DC dorsal cortex SATl striatoamygdaloid transition area, lateral aspect DLA dorsolateral amygdaloid nucleus SATm striatoamygdaloid transition area, medial aspect dLC deep lateral cortex Si nucleus septalis impar DSt dorsal striatum DVR dorsal ventricular ridge sm stria medullaris GP globus pallidus SO supraoptic nucleus HRP horseradish peroxidase SP substance P LA lateral amygdala St striatum proper LC lateral cortex st stria terminalis lfb lateral forebrain bundle TB Tris buffer LHA lateral hypothalamic area TBS Tris-buffered saline lot lateral olfactory tract TH tyrosine hydroxylase MA medial amygdala VAA ventral anterior amygdala MC medial cortex VP ventral pallidum mfb medial forebrain bundle VPA ventral posterior amygdala MPA medial preoptic area zl zona limitans MPO medial preoptic nucleus AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 289

important implications for the understanding of the com- seconds on/off; 3–5 ␮A) applied during 5–15 minutes were parative neuroanatomy and evolution of the telencephalon used to inject all the tracers, and a mild continuous reten- of amniotes. tion current (Ϫ0.9 ␮A) was applied during entrance and withdrawal of the micropipette to avoid diffusion of the tracer. Injections were performed taking the anterior ver- MATERIAL AND METHODS tex of the pineal scale as a reference point for stereotaxic Tract-tracing experiments coordinates. Injections in striatal structures were done using a contralateral approach to ensure that portions of To study the amygdaloid projections to the basal fore- the cortex and/or dorsal ventricular ridge overlying the brain in the lizard Podarcis hispanica, we first injected desired target were unaffected by the injection or by the neuroanatomical tracers by iontophoresis in different pipette track. To do so, the skull was drilled on the me- ϭ parts of the putative amygdaloid formation (n 11; PDVR diodorsal cortex and the micropipette was introduced us- and adjacent structures) and analyzed the resulting an- ing the appropriate angle to get access to the contralateral terograde labeling in the basal forebrain. Then, tracer striatum through the septum. injections were placed into the basal forebrain (n ϭ 10; After a survival time of 7–10 days (HRP experiments) or mainly in the dorsal and ventral striatum) to examine the 10–14 days (PHA-L, BDA, RDA experiments), animals distribution of retrogradely labeled cells in the putative received an overdose of Ketolar and were perfused tran- amygdala to identify the cells of origin of the amygdaloid projections to the basal forebrain. As a control, additional scardially with 2 ml of saline solution (0.9% NaCl), fol- injections were made in different neighboring areas to the lowed by 10–15 ml of fixative (4% paraformaldehyde in 0.1 amygdala and the basal forebrain (ADVR, n ϭ 4; dorsal M PB, pH7.4). Brains were carefully removed from the cortex, n ϭ 8; lateral cortex, n ϭ 3; septum, n ϭ 6). skull and post-fixed for 4–24 hours, in the same fixative In the present work, four different tracers were used: solution at 4°C. Then, they were immersed in 30% sucrose ␮ horseradish peroxidase (HRP; Type VI; Sigma, St. Louis, in PB at 4°C until they sank, and 40- m-thick frontal MO), Phaseolus vulgaris leucoagglutinin (PHA-L; Vector sections obtained with a freezing microtome were collected Laboratories, Burlingame, CA), biotinylated dextran into three matching series. amine (BDA, 10,000 molecular weight; Molecular Probes, In animals injected with HRP, the peroxidase activity Eugene, OR), and tetramethylrhodamine-labeled dextran was revealed by using 3,3Ј-diaminobenzidine (DAB, Sig- amine (RDA, 10,000 molecular weight, lysine fixable; ma; 0.02% in 0.05 M TB, pH 8.0) as the chromogen and Fluoro-Ruby; Molecular Probes). To minimize the number 0.01% H2O2. In most cases, nickel salts (0.4% nickel am- of animals, every lizard usually received two tracer injec- monium sulphate) were added as an enhancing agent of tions. Therefore, each injection received a code consisting the reaction product (DAB-Ni). of a letter indicative of the tracer (H for horseradish per- To visualize PHA-L, endogenous peroxidase activity oxidase, P for PHA-L, B for BDA and R for RDA) followed was first inhibited by incubating the sections for 30 min-

by a four-figure number indicative of the specimen. As utes in a 1% H2O2 solution in Tris-buffered saline (TBS), shown previously, all four tracers gave consistent antero- pH 7.6. Then the indirect peroxidase–antiperoxidase grade and retrograde transport (Lanuza et al., 1997, 1998; method (PAP; Sternberger, 1979) was applied (goat anti– Martıńez-Marcos et al., 1999). PHA-L, 1:2,000; donkey anti-goat IgG, 1:50; goat PAP, For this study, 42 adult specimens (both sexes) of Po- 1:400; Vector Laboratories), and the resulting peroxidase darcis hispanica (48–54 mm snout-cloaca length) were activity was revealed by using DAB as described above. used. Animals were captured in Vale`ncia, Spain, between For BDA detection, after endogenous peroxidase inhibi- 1988 and 1998 under a license issued by the Conselleria tion, we incubated the sections in the avidin–biotin com- d’Agricultura i Medi Ambient of the Valencian Govern- plex (ABC Elite kit; Vector Laboratories) in TBS with ment. Animals were maintained in terraria with food and 0.3% Triton X-100, either overnight at 4°C or for 2 hours water ad libitum, under natural day/night cycle, at 22– at room temperature. The resulting peroxidase label was 30°C. Throughout the experimental work, animals were revealed with DAB-Ni as described above. For RDA detec- treated according to the guidelines of the European Com- tion, the sections were observed using an epifluorescence munities Council (86/609/EEC) on the treatment of exper- microscope equipped with the filter N2.1 (Leica, Heidel- imental animals. berg, Germany). Selected injections of RDA were treated For surgery, animals were preanesthetized with halo- thane (2-brom-2-chlor-1,1,1-trifluorethan, 99%; Aldrich, for immunodetection of tetramethyl rhodamine using a Steinheim, Germany) and then deeply anesthetized with specific primary antibody (rabbit anti-tetramethyl rhoda- an intramuscular injection of 6.5␮l of Ketolar (ketamine mine, Molecular Probes), diluted 1:4,000 in TBS with 0.3% chlorohydrate 50 mg/ml, Parke-Davis, El Prat de Llobre- Triton X-100, followed by a standard PAP procedure (goat gat, Spain) per gram of body weight. Then they were anti-rabbit IgG, 1:100; rabbit PAP, 1:800; followed by de- placed in a stereotaxic apparatus with a small bird adap- tection of peroxidase activity with DAB). tor (Stoelting, Wood Lane, IL), and a small trephine hole Sections treated for HRP, PHA-L, BDA, or RDA immu- was opened in their skull to gain access to the desired nodetection were rinsed in warm (30–40°C) 0.2% gelatin area. Tracers were injected from stock solutions (HRP, (diluted in TB) and mounted onto clean slides. In most 10% in 0.05 M Tris buffer [TB], pH 8.6; PHA-L, 2.5% in cases, preparations were subsequently counterstained 0.01 M phosphate buffer [PB], pH 7.4; BDA, 10% in 0.01 M with acidic toluidine blue and cover-slipped with Per- PB, pH 7.6; RDA, 10% in 0.01 M PB) by means of iono- mount (Fisher Scientific, Fair Lawn, NJ). For observation phoresis using a current generator (Midgard Precision of RDA fluorescence, sections were collected immediately Current Source, Stoelting) through glass micropipettes after sectioning, mounted as described above, air-dried, with an inner diameter tip of 20–40 ␮m. Positive pulses (7 and cover-slipped with Mowiol (Osborn and Weber, 1982). 290 A. NOVEJARQUE ET AL. Histochemistry and immunohistochemistry Chemoarchitecture of the basal A collection of slides of brain sections of Podarcis telencephalon of Podarcis hispanica stained with different techniques is available in our labo- In mammals, the basal forebrain is composed of differ- ratory (see Font et al., 1995). The acetyl cholinesterase ent structures that include the basal ganglia, nucleus of histochemistry (AChase; Geneser-Jensen and Blackstad, the diagonal band, substantia innominata, and the nu- 1971) and the immunohistochemical detection of tyrosine cleus basalis of Meynert. The basal ganglia are composed hydroxylase (TH) and substance P (SP) resulted particu- of the dorsal (caudate putamen) and ventral striatum larly useful to characterize the chemoarchitecture of the (nucleus accumbens plus olfactory tubercle) as well as the basal forebrain. Briefly, TH immunoreactivity was de- lateral globus pallidus and ventral pallidum. Caudally, tected by using a monoclonal anti-TH (Incstar, Stillwater, some of these structures merge with other divisions of the MN), a biotinylated horse anti-rabbit IgG and the ABC forebrain such as the bed nucleus of the stria terminalis method; SP immunoreactivity was detected with a poly- and the medial or intracapsular globus pallidus. clonal rabbit anti-SP (CRB, Cambridge, UK) and the per- The study by Russchen et al. (1987) was the first one in oxidase antiperoxidase method (for methodological de- trying to identify all these structures in a reptile using a tails, see Font et al., 1995). battery of (immuno)histochemical markers that label more or less specifically some of these nuclei. We have Image acquisition and processing used some of these markers (acetyl cholinesterase, SP and TH), together with Nissl staining, to delineate the compo- Light microscopic images were photographed in a Leitz nents of the basal telencephalon in Podarcis hispanica. DMRB microscope equipped with a Leica DC 300 digital Our results indicate a similar picture to the one described camera (Leica Microsystems, Wetzlar, Germany). Digital in the gecko (Russchen et al., 1987) and allow us to pro- images were imported into Adobe Photoshop (Adobe Sys- pose a more detailed cyto- and chemoarchitecture of the tems, Mountain View, CA) and converted to gray scale reptilian basal telencephalon. images. For the final figures, brightness and contrast were By using classic cytoarchitectonic criteria (cell size, adjusted and resolution was set at 600 dpi. No additional shape, and density; distribution of fibers and fiber tracts; filtering or manipulation of the images was performed. staining properties), several cell groups can be distin- The final figures were composed and labeled with Adobe guished in the basal telencephalon of Podarcis hispanica. Photoshop. The dorsal boundary of the basal ganglia is clear at rostral levels, due to a cell-free zona limitans (zl) that separates them from the ADVR (Fig 1A–AЈ, B–BЈ). Laterally the zl RESULTS ends in the nucleus of the lateral olfactory tract. Caudally, A detailed description of the basal telencephalon of Po- the zl disappears and this reflects the end of the ADVR darcis hispanica, including cytoarchitecture and chemoar- and the appearance of the PDVR (Lanuza et al., 1998; Fig. 1AЉ,BЉ). chitecture, is first provided to facilitate an appropriate At rostral levels, the ventral striatum (nucleus accum- description of the tract-tracing experiments. Subse- bens, Acb; olfactory tubercle, OT; Olmos and Heimer, quently, according to our experimental design, we de- 1999) occupies most of the medial aspect of the basal scribe the anterograde labeling in the striatopallidal com- telencephalon, thus clearly protruding into the septal re- plex after injections in the amygdaloid complex. Finally, gion (Fig. 1A,B). The OT is present only at these levels, the retrograde labeling that resulted from injections of where it occupies the ventral surface of the cerebral hemi- tracers in the basal telencephalon is used to check the spheres and caps ventrally the rostral Acb. More caudally results of the anterograde experiments and to locate in (Fig. 1AЈ,BЈ) the OT disappears and its position is occupied detail the cells of origin of these projections. by the diagonal band nucleus (DBN), whereas the Acb According to Lanuza et al. (1997; 1998), the caudal becomes reduced to a cell group surrounding the ventral aspect of the dorsal ventricular ridge and neighboring sulcus of the lateral ventricle. This is accompanied by the areas constitute the amygdala of Podarcis, which is com- expansion of the other main striatal structure, the stria- posed of a dorsal division (encompassing the posterior tum proper (St or dorsal striatum, DSt), which at this level dorsal ventricular ridge, the dorsolateral amygdala, and occupies most of the mediolateral extent of the dorsal the lateral amygdala); a cortical, olfacto-recipient division basal telencephalon (Fig. 1AЈ,BЈ). (formed by the nucleus sphericus, the nucleus of the ac- The histochemical markers used in the present study cessory olfactory tract, the nucleus of the lateral olfactory reveal further heterogeneities in the caudal levels of the tract, the ventral–anterior amygdala, and the ventral– ventral striatum. Thus, whereas the rostral tip of the posterior amygdala); and a centromedial division (com- ventral striatum looks highly (and quite uniformly) (im- posed of the striatoamygdaloid transition area, the bed muno)reactive for AChase, SP and TH (Fig. 1C–E), the nucleus of the stria terminalis, and the medial amygdala). picture changes dramatically at slightly more caudal sec- In addition, here we consider the existence of a distinct tions (Fig. 1CЈ–EЈ). There, the caudal aspect of the Acb nucleus laterally adjacent to the ADVR and deep to the displays a moderate-to-low AChase (Fig. 1CЈ) and TH (Fig. rostral lateral cortex, which we name deep lateral cortex 1EЈ) (immuno)staining, as well as a very low SP immuno- (dLC) and consider it to be a rostral extension of the reactivity that contrasts with the moderate staining of the amygdala. This structure is rostral to the dorsolateral surrounding structures (Fig. 1DЈ). amygdala (DLA) but, in contrast to it, shows a low cellular Concerning the DSt, its dorsal and ventral parts display density and is negative for acetyl cholinesterase histo- distinct cytoarchitectonic and histochemical features. The chemistry. As we will see, the dLC displays a distinct dorsal aspect of the DSt, which lies adjacent to the zl, pattern of connections with the striatum. shows a relatively high cell density (Fig. 1A–AЈ,B–BЈ) but AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 291

.Fig. 1. Frontal sections of the rostral (left column), intermediate blue in B. C–C؆: Sections treated for acetylcholinesterase histochemistry -middle column), and intermediocaudal (right column) right cerebral D–E؆: Sections immunolabeled for substance P (D) and tyrosine hydrox) hemisphere of Podarcis hispanica, illustrating the cytoarchitecture (A,B) ylase (E). Asterisks in CЈ–EЈ indicate the position of a characteristic cell and chemoarchitecture (C–E) of the striatopallidal complex. A–B؆: Semi- group within the ventral aspect of the St (see the Results section). For schematic camera lucida drawings frontal sections stained with toluidine abbreviations, see list. Scale bar ϭ 250 ␮minEЉ (applies to A–EЉ). 292 A. NOVEJARQUE ET AL. a low AChase reactivity (Fig. 1C–CЈ) and very faint SP erate TH-immunoreactive innervation (Fig. 1EЈ–EЉ). The (Fig. 1D–DЈ) and TH (Fig. 1E–EЈ) immunoreactivity. In medial pallidum is composed of two cell groups: the DBN contrast, the ventral aspect of the DSt displays a high and the nucleus of the medial forebrain bundle (Nmfb), reactivity for AChase (Fig. 1C–CЈ); a moderate, patchy SP which are rostral to the nucleus of the anterior commis- immunoreactivity (Fig. 1D–DЈ); and a dense TH- sure (not shown). Histochemically, these structures dis- immunoreactive innervation (Fig. 1E–EЈ). The lateral bor- play a neuropile moderately reactive for AChase histochem- der of the ventral aspect of the DSt shows at these levels istry (Fig. 1CЈ), especially prominent at caudal levels (Fig. a characteristic strong (immuno)reactivity for all three 1CЉ), as well as numerous AChase-reactive cells (only visible markers used in this work (asterisk in 1CЈ–EЈ). at higher magnification, not shown). Substance P immuno- More caudally, where the zl disappears, the nucleus of histochemistry renders a moderate to high labeling of the the accessory olfactory tract (Naot) deepens into the cere- Nmfb, whereas the cell layer of the DBN remains virtually bral hemisphere, increases its size and displays a promi- unstained (Fig. 1DЈ–DЉ). Finally, the DBN/Nmfb displays nent AChase-reactive spot that occupies its medial border scarce (if present) TH-immunoreactive fibers (Fig. 1EЈ–EЉ). Љ Љ (Fig. 1A –C ). At these levels, a dense-celled structure Anterograde transport from injections in occupies most of the striatal territory. This cell group corresponds to the striatoamygdaloid transition area de- the reptilian pallial amygdala fined in the gecko by Russchen and Jonker (1988). On the A total of 11 injections were located in the PDVR and basis of their differential innervation by calcitonin-gene adjacent areas. The resulting anterograde labeling (Fig. 2) related peptide (CGRP) -immunoreactive fibers, Martıńez- is described below. Garcıá et al. (2002b) considered a medial (SATm) and Anterograde labeling in the basal forebrain after lateral divisions (SATl) within the striatoamygdaloid injections in the PDVR. Eight injections were aimed at transition area (SAT), the latter displaying a denser the PDVR, four of which were restricted to this structure. CGRPergic innervation fibers. The SATm seems a caudal In the case illustrated (P9228; Fig. 3), the tracer was continuation of the Acb, with which it shares most (immu- entirely restricted to the central part of the PDVR (Figs. no)histochemical features (Fig. 1), although the SATm 2A, 3E). Three main bundles of labeled fibers were ob- displays a slightly higher AChase reactivity. On the other served to leave the injection site. One of them reached a hand, the SATl seems a lateral continuation of the dorsal series of pallial structures at different rostrocaudal levels, aspect of the DSt, and it is just medial to the enlarged namely the Naot (Fig. 3B–E), the lateral aspect of the Naot (Fig. 1AЉ–BЉ). As compared with the SATm, the SATl PDVR, and the whole LA (Fig. 3D,E) as well as, more shows a lower cell density (Fig. 1BЉ) and a somewhat rostrally, the dLC (Fig. 3B,C). The second bundle of la- fainter (immuno)reactivity for AChase (Fig. 1CЉ) and TH beled fibers coursed ventromedially up to the BST. This (Fig. 1EЉ). The SATm is continuous caudally with a struc- bundle of labeled fibers continued rostrally to innervate ture crossed by the stria terminalis (st) as it leaves the the SAT (Figs. 2AЈ, 3C,D) and the Acb (especially its PDVR (Lanuza et al., 1998), which is named, therefore, central part, Figs. 2AЉ, 3A,B). A few labeled fibers could be the bed nucleus of the stria terminalis (Fig. 2B; BST). observed in the VP (Figs. 2AЉ, 3A,B). In the pallidum of Podarcis hispanica, three compart- The third bundle of labeled fibers left the injection site ments can be distinguished attaining to cytoarchitectonic, and crossed the midline trough the anterior commissure histochemical, and hodological features: the globus palli- (ac). In the contralateral hemisphere, labeled fibers were dus, the ventral pallidum, and the medial pallidum. The located in the BST with a similar distribution to that lateral pallidum or globus pallidus (GP), which is enriched observed in the ipsilateral hemisphere (Fig. 3E). More- in large cells with abundant Nissl substance, occupies the over, a few scattered labeled fibers were seen in the medial ventrolateral border of the telencephalic hemispheres at part of the SAT and in the center of the Acb (Fig. 3A–C). mid-to-caudal levels (Fig. 1AЈ–AЉ,BЈ–BЉ), where it sur- Anterograde labeling in the basal forebrain after rounds ventrolaterally the lateral forebrain bundle (lfb). injections in the DLA. Six tracer injections affected the The GP is quite heterogeneous from a histochemical point DLA. The BDA injection site illustrated (case B9506; Figs. of view. Thus its anterior portion looks uniformly pale in 2B, 4D) completely involved the DLA but also encom- preparations reacted for AChase histochemistry (Fig. 1CЈ) passed the lateral aspect of the caudal dorsal cortex (DC) or TH immunostaining (Fig. 1EЈ), but SP immunoreactiv- and the dorsal aspect of the caudal lateral cortex (LC). In ity reveals a deep, dorsal, strongly positive field plus a addition, some tracer might have diffused into the lateral ventral superficial part virtually devoid of labeling (Fig. edge of the PDVR. Several small bundles of labeled fibers 1DЈ). On the other hand the caudal GP shows a mild, left the injection site coursing medially and sharply patchy reactivity for AChase (Fig. 1CЉ), an apparently turned ventral to reach the st and the ac (Fig. 4D). A laminar pattern of SP immunoreactivity (Fig. 1DЉ) and a densely labeled terminal field was observed bilaterally in moderate patchy innervation by TH-immunoreactive fi- the whole SAT (Figs. 2BЈ, 4C) and BST (Fig. 4C). In the bers (Fig. 1EЉ). contralateral SAT, labeling was less profuse and showed a More medially within the basal telencephalon, just ven- more lateral location relative to the ipsilateral SAT. In tral to the Acb and SATm, and medial to the lfb, there is both hemispheres, several labeled fibers coursed rostrally a group of big, sparse cells that apparently constitutes the to innervate structures in the basal telencephalon. Dense ventral pallidum (VP) of Podarcis. At rostral levels (Fig. anterograde labeling was found throughout the whole exten- 1B) the VP is seen as a narrow, cell-poor area ventral to sion of the Acb (Figs. 2BЉ, 4A,B), especially in its central and the Acb and lateral to the OT, moderately reactive for dorsal parts and around the ventral sulcus of the lateral AChase (Fig. 1C) and SP (Fig. 1D), and light TH immu- ventricle. A large number of labeled fibers was seen bilater- noreactivity (Fig. 1E). More caudally, the VP displays a ally in the OT (Fig. 4A,B). Thus, the ventral striatum dis- faint reactivity for AChase (Fig. 1CЈ–CЉ), weak (but het- played bilateral (symmetric) labeling with ipsilateral pre- erogeneous) SP immunoreactivity (Fig. 1DЈ–DЉ), and mod- dominance. The VP also showed quite a dense anterograde AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 293

Fig. 2. Anterograde labeling in the continuum nucleus accumbens/ Anterogradely labeled fibers are also quite dense in the dorsal stria- striatoamygdaloid transition area after tracer injections in the PDVR tum (DSt) (BЉ). C–C؆: On the other hand, tracer injections centered in -and adjoining structures. A–A؆: Injections restricted to the PDVR (A, the deep lateral cortex (C, case B9510) also result in a dense antero case P9228) give rise to anterograde labeling in the medial (SATm) grade labeling of the medial and lateral SAT (CЈ), but at more rostral and lateral (SATl) striatoamygdaloid transition area (AЈ), and in the levels, labeling is restricted to the Acb and VP (CЉ). In all three nucleus accumbens (Acb, AЉ). B–B؆: On the contrary, tracer injections injections, labeled fibers can be seen leaving the main terminal fields encompassing the dorsolateral amygdala (B, case B9506) render a in striatal territories to reach the ventral pallidum. For abbreviations, denser and more extensive anterograde labeling in the medial and see list. Scale bars ϭ 250 ␮m in C (applies to A–C), CЈ (applies to lateral SAT (BЈ), as well as in the Acb and ventral pallidum (BЉ). AЈ–CЈ), BЉ (applies to AЉ–CЉ). labeling that was also observed in the contralateral hemi- the overlying cortex (dorsal, lateral, and medial areas). sphere (Figs. 2BЉ, 4B). The pattern of labeling (see Martıńez-Garcıá et al., 1993) Anterograde labeling was very profuse within the DSt was very similar to the one described above. Moreover, (Fig. 4A–C), where labeled fibers were present bilaterally control injections in the dorsal cortex overlying the DLA in the whole DSt (Figs. 2BЉ, 4B), reaching also the Naot. This rendered scarce labeling in the Acb/SAT. This, together labeling was not homogeneously distributed. For instance, a with the results of retrograde tracing (see below), indi- distinct stripe of labeled fibers was observed just ventral to cates that most of the fiber labeling observed in the basal the zona limitans in both hemispheres (not shown). At cau- forebrain after injections encompassing the DLA was due dal levels of the ipsilateral hemisphere, a few labeled fibers to anterograde transport from DLA neurones. extend from striatal regions to the GP (Fig. 4C). It is interesting to note that, although injections of BDA In two cases (H8819 and H8820), a small fragment of or HRP in the PDVR reliably rendered thalamic retro- crystallized HRP was placed into the DLA after removing grade labeling (see Lanuza et al., 1998), after injections in 294 A. NOVEJARQUE ET AL.

Fig. 3. A–E: Semischematic camera lucida drawings of frontal Most of the fibers course rostrally to innervate the ipsilateral contin- sections through the brain of a lizard that received a PHA-L injection uum Acb–SAT. The brain on top shows the rostrocaudal level of each (striped area) restricted to the PDVR (injection P9228), showing the section. For abbreviations, see list. Scale bar ϭ 500 ␮m in E (applies resulting anterograde labeling (lines and dots) in the telencephalon. to A–E). the DLA, the thalamus was devoid of retrogradely labeled Anterograde labeling in the basal forebrain after cells. This finding is indicative of a distinct pattern of injections in the dLC. In one case (B9510), the injection afferents and efferents of the DLA compared with neigh- site was rostral to the DLA, thus being centered in the boring areas such as the PDVR that displays substantial dLC. This injection also involved cells belonging to the afferents from the thalamus (Lanuza et al., 1998). overlying LC and DC (Figs. 2C, 5D). Labeled fibers leaving AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 295 the injection site could be followed up to a dense field of and the BST (Fig. 5F). This fiber system extended ros- fiber and terminal labeling in the whole SAT, the caudal trally up to the Acb and the VP, where a dense field of edge of the DSt and portions of the GP (Figs. 2CЈ, 5D,E) terminal labeling was present (Figs. 2CЉ, 5B–D). The anterograde labeling in the Acb was mainly located in its most caudal part and in its ventral half. In the contralateral hemisphere, very few labeled fibers were observed in the BST and SAT. In addition, a few fibers were observed in the contralateral Acb and VP, with a similar distribution to that described for the ipsilateral labeling (Fig. 5C). The anterograde labeling found in the contralateral hemisphere is due to fibers that cross the middle line throughout the ac. Although some fibers could also be seen in the habenular commissure, they are likely to be connecting the ipsi- and contralateral LC (Martıńez-Garcıá et al., 1986). Anterograde labeling in the basal forebrain after injections in the NS. Four injections involved different portions of the NS. In the case illustrated (H9331 Fig. 6), the injection was located in the centromedial aspect of the nucleus. Numerous labeled fibers entered the anterior commissure and gave rise to abundant anterograde and retrograde labeling in the contralateral NS (particularly in its medial part, Fig. 6G). Intense anterograde labeling was also observed bilaterally in the medial amygdala (MA; especially in its rostral parts; Fig. 6E,F), the BST (Fig. 6F) and the Naot (Fig. 6D–F). Dense anterograde labeling was present ipsilaterally in the ventral anterior (VAA) and ventral posterior amygdala (VPA; Fig. 6D–F), with a few fibers present in the contralateral VAA (Fig. 6D). In the ventromedial telencephalon, sparse anterograde labeling was observed in the ventral striatum and ventral pallidum. Thus, a few labeled fibers were visible in the OT, as well as in the ventromedial aspect of the Acb (Fig. 6A–C) and SATm (Fig. 6D). On the other hand, the anterograde labeling in the VP (Fig. 6A–C) was especially dense at rostral levels. Labeling in the ventral striatum and ventral pallidum was bilateral with strong ipsilateral dominance. The DSt was devoid of fiber labeling except for its most rostral level, where a few fibers appeared in the dorsolateral DSt coursing from the aot to the ventral striatum (Fig. 6A). A large injection involving most of the NS (H8810) displayed the same pattern of anterograde labeling. Remarkably, the rostral parts of the ipsilateral Acb displayed a relatively dense network of labeled fibers and terminals, whereas no labeling was observed in the contralateral ventral and dorsal striatum. Retrograde labeling in the reptilian pallial amygdala after injections in the ventral and dorsal striatum A total of 10 injections were located in the ventral and dorsal striatum. The resulting retrograde labeling after these injections is described below.

Fig. 4. A–D: Camera-lucida drawings of the anterograde labeling (lines and dots) observed in the cerebral hemispheres of a lizard that received a BDA injection (striped area) encompassing the DLA (case B9506). Labeling is bilateral, with a rough symmetrical distribution in the basal forebrain. Labeled ﬁbers are mainly found in the ventral and dorsal striatum, including both divisions of the SAT. The approximate level of each section is indicated in the brain on top. For abbreviations, see list. Scale bar ϭ 500 ␮m in D (applies to A–D). 296 A. NOVEJARQUE ET AL.

Fig. 5. A–G: Camera-lucida drawings of seven frontal sections lateral SAT, which extends into the pallidum. The contralateral basal through the forebrain of a lizard that received a BDA injection forebrain shows less dense anterograde labeling. The level of each (striped area) centered in the dLC (case B9510). Anterogradely la- section is shown on a lateral view of the brain of Podarcis (top). For beled ﬁbers (lines and dots) in the ipsilateral hemisphere form a dense abbreviations, see list. Scale bar ϭ 500 ␮m in G (applies to A–G). terminal ﬁeld in the continuum SAT–Acb, including the medial and

Four injections were centered in the Acb. Case B9827 is In the cortical division of the amygdala, retrograde la- described in detail as a representative experiment (Figs. beling was observed bilaterally in the NS (mainly in the 7, 8). This case was a BDA injection encompassing most of mural layer and in its dorsal part, Figs. 7B, 8F,G) and the Acb (Fig. 8) that yielded retrograde labeling in the Naot (Fig. 8D,E). Labeled cells also appeared in the ipsi- dorsal, cortical, and centromedial divisions of the amyg- lateral VAA and VPA (Fig. 8E,F). In the dorsal amygda- dala (Lanuza et al., 1997). loid division, a large number of retrogradely labeled neu- AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 297

Fig. 6. A–G: Semischematic camera-lucida drawings of seven frontal sections through the forebrain of a lizard that received a HRP injection (striped area) in the nucleus sphericus (case H9331), showing the resulting anterograde labeling (lines and dots). The basal forebrain displays scarce ﬁber labeling mainly located in the VP, olfactory tubercle, and the medial aspect of the Acb–SAT. The level of each section is depicted on a lateral view of the brain of Podarcis (top). For abbreviations, see list. Scale bar ϭ 500 ␮m in G (applies to A–G). 298 A. NOVEJARQUE ET AL.

Fig. 7. A: Retrograde labeling in the PDVR and adjoining struc- its lateral border) and in the DLA. In contrast, the dorsal and lateral tures of the brain of Podarcis after an injection of BDA restricted to cortices, the lateral amygdala and the medial PDVR display a very the nucleus accumbens (case B9827). B: The ipsilateral nucleus low density of labeled cells. D: At more rostral levels, a dense group of sphericus displays a high density of retrogradely labeled cells mainly labeled cells is present in the deep lateral cortex (also visible in A), located in the dorsal mural layer. Labeled neurons are also visible in which surrounds the lateral sulcus of the cerebral ventricle and seems the caudal edge of the dorsal cortex. C: At the level of the PDVR to be a rostral extension of the DLA. For abbreviations, see list. Scale retrogradely labeled cells are mainly observed in PDVR (especially in bars ϭ 250 ␮m in B (applies to A,B), D (applies to C,D). rons was observed bilaterally in the DLA (Figs. 7C, 8D–F) In another case (B9807), a larger BDA injection was and ipsilaterally in the PDVR (where cell labeling was espe- centered in the caudal Acb but also involved the rostral cially dense in its most lateral part at caudal levels; Figs. 7C, SAT, the VP, and the ventral portion of the septum. In this 8D,E) and in dLC (Figs. 7D, 8C). Few labeled cells were injection, the amygdala displayed a pattern of retrograde found in the contralateral dLC. The LA was nearly devoid of labeling similar to the one described above. In addition, retrograde labeling (Fig. 7C), thus appearing as an unla- retrograde labeling was found in the ipsilateral LA and beled area separating the Naot and the DLA (Fig. 8D,E). medial PDVR, and contralateral labeling appeared in the Retrogradely labeled cells were also abundant in the MA and SAT (data not shown). centromedial amygdaloid division, located bilaterally In three cases, injections were centered in the DSt, (with a clear ipsilateral predominance) in the lateral BST although they also involved the lateral aspect of the Acb. (Fig. 8D,E) and ipsilaterally in the SAT and medial amyg- In all three cases, the retrograde labeling was similar. dala (MA; Fig. 8F). Injection B9828 (Fig. 9), which is described as a represen- AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 299

Fig. 8. A–G: Retrograde labeling in the telencephalon of a lizard appear relatively free of labeling. Several of these structures show after a tracer injection (case B9827) restricted to the nucleus accum- bilateral labeling. The approximate rostrocaudal level of each section bens (striped area in B). Abundant labeled cells can be seen in the is indicated on a lateral view of the brain of Podarcis. For abbrevia- lateral PDVR and in most of the structures that surround it, such as tions, see list. Scale bar ϭ 500 ␮m in G (applies to A–G). the DLA, dLC, Naot, and NS, whereas the medial PDVR and the LA tative example, was located rostrally in the DSt and par- the anterior–dorsal ventricular ridge but scarce labeling tially involved the VP and the Acb. In the dorsal amygda- in the PDVR (Fig. 9A–D). loid division, retrogradely labeled neurons were located In the cortical amygdaloid division, a few labeled cells ipsilaterally in the dLC and the PDVR (especially at its also appeared in the medial aspect of the Naot (Fig. 9C,D) rostral levels), as well as in the LA (Fig. 9D), and bilater- and in the medial mural layer of the NS (Fig. 9E). Finally, ally in the DLA (with a similar density of cell labeling in in the centromedial amygdaloid division, we found a few both hemispheres; Fig. 9D,E). It is interesting to note that labeled cells in the ipsilateral SAT (mainly in its lateral these injections displayed abundant retrograde labeling in part Fig. 9C) and in the BST (Fig. 9D). Ipsilateral labeling 300 A. NOVEJARQUE ET AL.

Fig. 9. A–E: Camera-lucida drawings of frontal sections through retrograde labeling, whereas the LA ipsilateral to the injection shows the brain of a lizard (case B9828) that received a BDA injection a few labeled neurones. In contrast, the PDVR is virtually devoid of (striped area in A–C) that involved striatum proper plus the lateral labeling. The rostrocaudal level of each section is indicated on a edge of the Acb. Retrogradely labeled cells are visible in the ipsilateral drawing of a lateral view of the brain of Podarcis (top). For abbrevi- ADVR dorsal to the injection site. The DLA shows a dense, bilateral ations, see list. Scale bar ϭ 500 ␮m in E (applies to A–E).

also appeared in the MA (Fig. 9D; a few cells were seen striatum, from the caudal pole of the DVR with a distinct contralaterally, data not shown). “nontopographical” organization. Our data on the cyto- and chemoarchitecture of the basal telencephalon suggest that the caudal aspect of the DISCUSSION DVR not only projects to the DSt but also to a continuum Classic ideas on the organization of the reptilian telen- of ventral striatal structures (Acb and SAT) and, less cephalon (Ulinski, 1983) consider the striatum to be dom- massively, to pallidal structures. It seems compulsory, inated by a topographical input from the anterior aspect of therefore, to discuss first our data on the organization of the DVR. By using both anterograde and retrograde trac- the basal telencephalon of Podarcis in the context of pre- ing, we demonstrate the presence of a set of additional vious descriptions in other reptilian species and of solid projections to striatal territories, including the ventral anatomical data of mammals and birds. Second, we will AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 301 use our results of anterograde and retrograde tracing ex- In mammals, the Acb displays two major compart- periments and similar data in other reptiles, to define the ments, shell and core, which show differential expression site of origin and termination pattern of the projections of calbindin, SP, AChase, and TH (see Jongen-Relo et al., from the DVR to striatal territories in reptiles. Then, a 1994), as well as different patterns of connections (Zahm, comparative approach using data on the connections and 1999). Our histochemical data fit the view put forward by the expression of homeotic genes during embryonic devel- Guirado et al. (1999) that the reptilian Acb also displays a opment in the caudal pallial territories of different verte- rough compartmentalization consisting in a rostromedial brates allow us to interpret the connections between the portion richer in TH, AChase, and SP, and a caudolateral posterior DVR and the striatum as amygdalostriatal. Fi- part, less reactive for all markers. Guirado et al. (1999) nally, we discuss the possible functions of these projec- suggested that the rostromedial Acb of reptiles recalled tions in behavior and the important role they might have the mammalian shell and the caudolateral part the core. played in the evolution of the forebrain in amniotes. This finding is not contradicted by our connectional data because, like the mammalian basolateral amygdala On the organization of the reptilian basal (Wright et al., 1996), the reptilian caudal DVR projects to telencephalon both divisions of the Acb in a complex way. The reptilian DSt is usually compared with the mam- Our description is based on the general scheme of the malian caudate-putamen. However, whereas the caudate cerebral hemispheres proposed by Swanson and Risold putamen receives its main input from the mediodorsal and (1999), further refined by Lanuza et al. (2002). According temporal isocortex, the DSt of reptiles is the target of to this scheme, the subcortical telencephalon is composed descending projections from the ventropallial ADVR and of striatal and pallidal territories. This division relies on displays only minor afferents from the dorsal pallium developmental analysis (i.e., origin from the lateral and (Hoogland and Vermeulen VanderZee, 1989). This might medial ganglionic eminences), as well as on anatomical explain the important differences between the organiza- and neurochemical data from the adult. tion of the reptilian DSt and the caudate putamen. Nev- Thus, striatal territories are dominated by glutamater- ertheless, both structures share a massive and bilateral gic inputs from the pallium, which are rich in zinc. On its input from deep lateropallial structures (basolateral turn, medium spiny GABAergic striatal neurons project amygdala of mammals, Kelley et al., 1982; our unpub- mainly to pallidal/entopeduncular territories and to the lished results in mice; reptilian DLA, this work), as well as tegmentum. The third feature of the striatal compart- their interconnections with the substantia nigra. In this ments is their dopaminergic afferent from the tegmental respect, one of the main traits of the intrinsic organization nuclei to which they project. Both hodological and neuro- of the mammalian dorsal striatum is the striosome– chemical features, such as the presence of zinc-positive matrix organization. Although AChase and TH reveal a terminals (of probable pallial origin; Pe´rez-Clausell, 1988; dense ventrolateral DSt and a relatively unstained dorsal Smeets et al., 1989) and dopaminergic (TH) innervation DSt, as discussed by Russchen et al. (1987), at present, it arising from the tegmentum, fit our proposal for the de- is unclear whether this reflects a striosome-matrix–like lineation of striatal territories. organization. In mammals, the matrix projects mainly to In this respect, the mammalian accumbens receives an the pallidum and the cell-poor substantia nigra pars re- important input from the subicular cortex (Groenewegen ticulata, whereas striosome cells mainly project to the et al., 1987; Brog et al., 1993). A comparable projection to cell-rich sustantia nigra pars compacta. Detailed studies the accumbens arising from the mediodorsal cortex is on the nigral and pallidal projections of the DSt of reptiles found in several lizards (Hoogland and Vermeulen- might help to clarify this issue. VanderZee, 1989; Guirado et al., 1999; this work). Our Finally, the pallidum receives a massive nonglutamater- work expands the findings by Pe´rez-Santana et al. (1997) gic (GABAergic and peptidergic) input from striatal territo- indicating that the reptilian accumbens receives an addi- ries. The delineation of pallidal territories proposed in the tional and massive input from the caudal aspect of the present work fits these features, because all the pallidal dorsal ventricular ridge. This finding includes bilateral structures described are poor in Zn-positive terminals projections from deep portions of the lateral pallium, (Pe´rez-Clausell, 1988) and receive inputs from striatal terri- namely the DLA, and homolateral afferents from the dLC tories (our unpublished results of anterograde tracing from (lateropallial) and PDVR (ventropallial). As discussed be- tracer injections in the Acb and DSt). According to Swanson low, and according to the hypothesis put forward by Bruce and Risold (1999), we propose the pallidum to be composed of and Neary (1995a,b) and Lanuza et al. (1998) on the at least three compartments: a medial pallidum, namely, the amygdaloid nature of the caudal DVR, these projections Nmfb, a ventral pallidum (VP) and a lateral pallidum (glo- might represent an amygdaloaccumbens pathway similar bus pallidus [GP]). The Nmfb is adjacent to the DBN, with to the well-described projection from the mammalian ba- which it shares a dense input from the lateral septum (Font solateral amygdala to the nucleus accumbens (Wright et et al., 1998) and ascending projections to the mediodorsal al., 1996). These pallial afferents give strong support to hippocampal cortex (Bruce and Butler, 1984a; Martıńez- the view that the Acb is the reptilian homologue of its Garcıá and Olucha, 1988). On the basis of this and of other mammalian homonym. In addition, the projections from histochemical features, Font et al. (1998) proposed that the the nucleus accumbens to the ventral pallidum and its Nmfb and DBN constitute the reptilian counterpart of the interconnections with the dopaminergic cell group in the medial septum-diagonal band complex of the mammalian ventral tegmental area, which are characteristic features brain. The VP is the main target for the descending projec- of the ventral striatum of mammals (see Groenewegen et tions of the Acb (Russchen and Jonker, 1988; Smeets and al., 1999), also define the Acb of reptiles (gecko: Russchen Medina, 1995; our unpublished results in Podarcis) and dis- and Jonker, 1988; Smeets and Medina, 1995; our unpub- plays additional histochemical features typical of pallidal lished results in Podarcis). territories. Thus, it shows abundant AChase (cholinergic; 302 A. NOVEJARQUE ET AL. see Medina et al., 1993) cells that project to the pallium (see anterograde (“olfactostriatum”; Lanuza and Halpern, Lanuza et al., 2002) and SP-positive cells and fibers (Fig. 1). 1997) as well as retrograde tracing (Pe´rez-Santana et al., The globus pallidus constitutes the main target of the de- 1997). Our results in lizards confirm this is a projection scending projection from the DSt (Russchen and Jonker, displayed by most (if not all) the squamate reptiles. Trac- 1988; our unpublished results). Like its mammalian hom- ing of the afferents to the accumbens in turtles also results onym (Halliday et al., 1995), the GP of Podarcis shows a in retrograde labeling of superficial cells in the so-called heterogeneous distribution of SP (Fig. 1), which was already nucleus centralis amygdalae (Fig. 1G in Siemen and noticed by Russchen et al. (1987). Although this heterogene- Kunzle, 1994). Because these cells are just deep to the ity fits the patchy organization of the inputs to the GP from input from the olfactory bulbs (Reiner and Karten, 1985), striatal territories (see Russchen and Jonker, 1988), its func- these data might indicate a chemosensory input to parts of tional significance is still unknown. the ventral striatum. It is widely accepted that turtles One of the main findings of our work pertaining to the display a reduced vomeronasal system (compared with organization of the basal telencephalon is the presence of squamate reptiles), because they lack a nucleus sphericus. conspicuous (although mild) projections from pallial terri- Nonetheless, there is evidence that, at least some turtles tories to pallidal ones (see Fig. 2). This indicates that, in have a functional vomeronasal organ (Hatanaka and Mat- reptiles like in mammals (Naito and Kita, 1994), the pal- suzaki, 1993; Murphy et al., 2001). Therefore, the pres- lidum is not just reached by a massive striatal input but is ence of vomeronasal input to portions of the ventral stri- also the target of a sparse cortical projection. atum could be a common feature of reptiles. Projections of the DVR to the striatum in Anterograde transport of tritiated amino acids in caudal DVR of Tupinambis nigropunctatus (Voneida and Sligar, reptiles 1979) revealed a projection from the posterior DVR to the Tract-tracing experiments using lesion-degeneration or ventral striatum (continuum Acb–SAT). The study of the anterograde transport of tritiated amino acids in different striatal afferents of the gecko (Gonza´lez et al., 1990) reptiles (snakes: Ulinski, 1978; Tupinambis nigropuncta- showed not only a topographical projection from the DVR tus: Hoogland, 1977; Voneida and Sligar, 1979; reviewed to the dorsal striatum but also a projection to the ventral by Ulinski, 1983) suggested that the DVR gives rise to a striatum (see their Fig. 10G–I) arising exclusively from topographically organized projection to the dorsal stria- the caudolateral aspect of the DVR. This projection seems tum. These results were largely confirmed in Gekko gecko to terminate in a continuum of structures connecting the by using modern tracing techniques (Gonza´lez et al., SAT (caudally) with the Acb (rostrally), thus clearly in 1990). accord with our findings in Podarcis. Moreover, a similar In addition, our experiments using both retrograde and projection might also be present in snakes, where injec- anterograde tracing clearly indicate that the caudal as- tions of tracers in the Acb result in retrograde labeling in pect of the DVR and adjoining structures (dorsolateral the caudal part of the DVR (Pe´rez-Santana et al., 1997). In amygdaloid nucleus, deep lateral cortex, and nucleus turtles, the lateral aspect of the basal DVR massively sphericus) give rise to four different projections to striatal projects to the ventral striatum (Siemen and Kunzle, territories. 1994). This finding also accords with the situation in Po- First, our results render evidence of a projection from darcis, where the cells projecting to the Acb are mainly the nucleus sphericus to the ipsilateral (and very sparsely found in the lateral PDVR (Figs. 7C, 8D,E). to the contralateral) nucleus accumbens. Retrograde trac- To our knowledge, the presence of a projection to the ing (see Figs. 7B, 8F) labels this projection much better ventral striatum from the deep lateral cortex has not been than anterograde transport (Fig. 6). This finding might reported before in squamate reptiles. This finding is sur- reflect that this is a convergent pathway, so that small prising because our injections in the Acb consistently ren- injections in the nucleus sphericus result in the antero- dered retrogradely labeled cells in the space between the grade labeling of only a few fibers in the Acb, whereas lateral sulcus of the telencephalic ventricle and the ven- dextran amine injections in the Acb give rise a high num- tral lateral cortex (Figs. 7D, 8C). Conversely, injections of ber of labeled perikarya in the NS. tracers involving the dLC resulted in fiber labeling in the The second projection arises from the PDVR and termi- ipsilateral (but to a lesser extent also in the contralateral) nates massively in a continuum of structures that connect nucleus accumbens. Therefore, we must conclude that the SAT (caudally) with the nucleus accumbens (rostrally; such a projection exists in Podarcis, whereas further stud- Fig. 3A–D). ies are needed to determine whether it is also present in The third projection, arises from the dLC and displays a other reptiles. In turtles, cells projecting to the nucleus similar termination pattern to the one originating in the accumbens are found in or deep to areas h and L (Siemen PDVR. However, the results of retrograde labeling after and Kunzle, 1994), adjacent to the lateral cortex, and in tracer injections in the accumbens suggest that the affer- the (ventral) lateral cortex itself. Whether or not this ent input from the dLC is more extensive than the one corresponds to the dLC of lizards requires further evalu- arising in the PDVR. Moreover, anterograde tracing (Fig. ation. 5B–E) indicates that the projection from the dLC to the Our results confirm and expand previous findings in accumbens is bilateral with clear ipsilateral dominance. Podarcis (Martıńez-Garcıá et al., 1993), indicating the The last projection originates in the DLA and termi- existence of a massive, bilateral projection from the nates bilaterally in the whole striatum (both ventral and DLA to both the dorsal and ventral striatum. A similar dorsal divisions). This finding has been confirmed by using projection can be inferred for the gecko from the results anterograde (Fig. 4) and retrograde transport of tracers by Gonza´lez and collaborators (Gonza´lez et al., 1990), (Figs. 7C, 8D–F). who reported anterogradely labeled fibers in both the The presence of a projection from the nucleus sphericus dorsal and ventral striatum after injections of Phaseo- to the ventral striatum has been reported in snakes using lus vulgaris leucoagglutinin in the caudolateral DVR AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 303

(see their Fig. 6). Surprisingly, this projection was not 2000) indicate that the DVR is not homologous to the revealed by their experiments of retrograde labeling of neocortex but to a pallial derivative just ventral to the the afferents to the Acb, whereas their injections of lateral (piriform) cortex and dorsal to the St, the so-called retrograde tracers in the dorsal striatum rendered sub- ventral pallium (Puelles et al., 2000). Moreover, the DVR stantial labeling of cell bodies in the nucleus lateralis of is composed of two rostrocaudal divisions with a different the amygdaloid complex (see their Fig. 10). We consider functional significance. The ADVR apparently consists of the projection from the caudolateral DVR, including the unimodal sensory areas (see Ulinski, 1983), including sev- so-called nucleus lateralis of the amygdaloid complex, to eral visual fields (Guirado et al., 2000; Manger et al., the striatum of the gecko equivalent to the projection 2002), whereas the PDVR appears to be an associative from the dorsolateral nucleus of the amygdala to the structure, because it receives convergent projections from striatum in Podarcis. Were this true, the DLA of Podar- the sensory fields of the ADVR (Andreu et al., 1996). Data cis (and probably of other lacertilian lizards) would be on the intratelencephalic (Belekhova and Chkheidze, equivalent to the nucleus lateralis of the amygdaloid 1991; Andreu et al., 1996; Lanuza et al., 1998) and ex- complex of geckonids, a view that is strongly supported tratelencephalic afferents to the PDVR (Lanuza et al., by the fact that both structures display a dense cholin- 1998) as well as on its projections to the hypothalamus ergic innervation, as revealed in different lizards by (Bruce and Neary, 1995a; Lanuza et al., 1997) strongly choline-acetyl transferase immunocytochemistry suggest that this structure is part of the reptilian basolat- (Hoogland and Vermeulen-VanderZee, 1990; Medina et eral amygdala (Martıńez-Garcıá et al., 2002a). In this al., 1993) and AChase histochemistry (Regidor and framework, the results discussed above would indicate the Poch, 1988; Lanuza et al., 1997; unpublished results in existence in reptiles of important projections from the Tarentola mauritanica). These data reveal important basolateral amygdala to striatal territories that would differences in the cytoarchitecture of the DVR (espe- have their counterpart in the mammalian (and avian) cially of its caudal pole) among different groups of liz- brain. ards, which have hindered our understanding of the In mammals, there are two sets of projections from the common pattern of organization of the reptilian cerebral basolateral (pallial) amygdala to striatopallidal territo- hemispheres. ries, namely a projection to the central extended amygdala Our results indicate that the PDVR and adjoining struc- that also reaches parts of the ventral striatum, and the tures in the telencephalon of lizards give rise to two sets of projection to the dorsal striatum. Some of these amygda- efferent projections, namely to the ventral striatum (this lostriatal projections are considered as part of the intrin- work) and to the ventromedial hypothalamus (Lanuza et sic circuitry of the amygdala because they mainly termi- al., 1997). It is interesting to note that the cell groups nate in the central amygdala and arise from portions of giving rise to these two efferent systems are partially the lateral (Pitka¨nen et al., 1995; Pitka¨nen and Amaral, segregated. Thus, whereas projections to the ventromedial 1998) basolateral (or basal; Savander et al., 1995) and hypothalamus arise from the medial PDVR and LA basomedial nuclei (or basal accessory; Petrovich et al., (Lanuza et al., 1997), projections to the Acb mainly arise 1996; Savander et al., 1996). These projections terminate from the lateral PDVR and DLA (see the Results section). not only in the central nucleus of the amygdala but also in However, tracer injections involving parts of the SAT ren- different portions of the bed nucleus of the stria terminalis dered relatively dense retrograde labeling in the medial (Dong et al., 2001) as well as in additional striatopallidal PDVR and LA (see the Results section). Therefore, the areas of the basal forebrain such as the interstitial nu- medial PDVR and the LA seem to project to both the SAT cleus of the posterior limb of the anterior commissure and and hypothalamus. This observation suggests that the the sublenticular substantia innominata. All these struc- efferent systems to the hypothalamus and SAT are func- tures are included in the concept of (central) extended tionally linked (e.g., expression of fear; see Functional amygdala (Alheid et al., 1995). Implications section). Some of these projections of amygdaloid nuclei to the Comparative remarks: amygdalostriatal extended amygdala run further rostrally, thus targeting the medial shell of the nucleus accumbens. In fact parts of projections in amniotes the basomedial and basolateral nuclei of the amygdala The anterior DVR is the target for ascending projections (Groenewegen et al., 1980; Kelley et al., 1982; Russchen from the dorsal thalamic nuclei relaying auditory (medial and Price, 1984; Brog et al., 1993; Petrovich et al., 1996; thalamus/nucleus reuniens), somatosensory (medial– Wright et al., 1996; our results in mice) do project to the posterior and posterocentral thalamic nuclei), and visual nucleus accumbens, which shows other (mainly histo- information (nucleus rotundus; Karten, 1969; Karten et chemical) features in common with the extended amyg- al., 1973; Lohman and van Woerden-Verkley, 1978; Ulin- dala (Alheid et al., 1995; Zahm, 1998). ski, 1983; Bruce and Butler, 1984b) and also projects to Whether or not the shell of nucleus accumbens consti- the dorsal striatum. The conclusion drawn from these tutes the rostral tip of the extended amygdala, the baso- findings is that the DVR–striatal pathway constitutes the lateral and basomedial amygdaloid nuclei of the mamma- link between the sensory systems and the motor one and, lian amygdala project to a continuum of structures that therefore, is the anatomical substrate of sensorimotor in- includes the central amygdala, the (dorsolateral) bed nu- tegration in the reptilian forebrain. This finding is the cleus of the stria terminalis, and the nucleus accumbens main reason to compare it with the (neo)corticostriatal (McDonald, 1991), as well as diverse additional ventral projection of the mammalian brain. This view is seriously striatopallidal structures. This finding clearly recalls to challenged by recent data on the anatomical organization the situation in reptiles where portions of the pallial and comparative meaning of the reptilian telencephalon. amygdala, including the lateral PDVR, the DLA, and the First, studies on the expression of homeotic genes during dLC project to a continuum of structures within the rep- development (Smith-Fernańdez et al., 1998; Puelles et al., tilian ventral striatum (Acb and SAT). This comparison 304 A. NOVEJARQUE ET AL. strongly supports the view, put forward by Bruce and ADVR, namely that it was homologous to the lateral Neary (1995a) and Lanuza et al. (1997, 1998) that the amygdala of mammals. In this respect, both the mamma- PDVR and adjoining structures in the caudal cerebral lian lateral amygdala and the reptilian ADVR are ventro- hemispheres constitute the reptilian homologue of the pallial (see Martıńez-Garcıá et al., 2002a). However, mammalian basolateral division of the amygdala whereas the ADVR projects to the DSt, the lateral amyg- (Martıńez-Garcıá et al., 2002a). dala shows meager projections to the striatum (see Pit- The second amygdalostriatal pathway of the mamma- ka¨nen, 2000). lian forebrain is an extensive projection to the dorsal In reptiles, the PDVR and adjacent regions (including striatum (caudate putamen) that arises mainly, if not the DLA) receive an important projection from regions of exclusively, from the basolateral nucleus of the amygdala the mediodorsal pallium (Hoogland and Vermeulen- (Kelley et al., 1982; Russchen and Price, 1984; Wright et VanderZee, 1989; Lanuza and Halpern, 1997; Lanuza et al., 1996; our unpublished results in mice). In reptiles, the al., 1998) that also project to the Acb (Hoogland and dorsal striatum receives a massive input from the DLA Vermeulen-VanderZee, 1989; Guirado et al., 1999). Al- (see the Results section). This finding points to the DLA as though the reptilian mediodorsal pallium is usually con- the reptilian homologue of the basolateral nucleus of the sidered homologous to the mammalian hippocampal for- mammalian amygdala. Our data indicate that the projec- mation (sensu lato; Hoogland and Vermeulen-VanderZee, tions of the DLA to the dorsal and ventral striatum, like 1989), its connections with the basolateral amygdala and the ones arising in the basolateral nucleus of mammalian ventral striatum are congruent with a field homology be- amygdala (Kelley et al., 1982; Brog et al., 1993; our un- tween the medial and dorsal pallial territories of mam- published results in mice), have a substantial contralat- mals and reptiles (see Martıńez-Garcıá et al., 2002a). eral component. This homology is further supported by The apparent lateropallial nature of the dLC (as indi- additional histochemical (dense AChase and TH- cated by its name, deep to the lateral cortex) and its immunoreactive innervations) and embryological features position rostral to the DLA, suggest that its mammalian (putative lateropallial nature; Martıńez-Garcıá et al., counterpart has to be found in a region anterior to the 2002a) shared by the reptilian DLA and the basolateral basolateral amygdaloid nucleus. Two structures appear as nucleus of the mammalian amygdala. the most likely candidates for the mammalian counterpart Besides its major outputs to the striatum, the main of the dLC. On the one hand, parts of the endopiriform hodological features of the mammalian basolateral amyg- nucleus apparently project to the accumbens (Brog et al., daloid nucleus are its interconnections with the mediodor- 1993), although this projection has not been fully con- sal thalamus and with the frontotemporal cortex (see firmed by using anterograde tracing (Behan and Haberly, Swanson and Petrovich, 1998; Pitka¨nen, 2000). However, 1999). Alternatively, the dLC might be comparable to the counterparts for these connections are not easy to identify insular/perirhinal cortex. This possibility is supported by in the reptilian brain. Thus, whereas both the reptilian the massive projections these cortical areas display to the ADVR (Lohman and Van Woerden-Verkley, 1978; Bruce accumbens (Brog et al., 1993; our unpublished results in and Butler, 1984b; Guirado et al., 2000) and PDVR mice), bed nucleus of the stria terminalis and central (Lanuza et al., 1998) receive a substantial thalamic affer- amygdala (McDonald et al., 1999; our unpublished results ent, the DLA is not connected with the thalamus (see the in mice). In addition, the dLC of lizards (Martıńez-Garcıá Results section). et al., 2002b) and the insular/perirhinal cortex of mam- On the other hand, the reptilian counterpart of the mals (Yasui et al., 1989) also share a specific afferent rich amygdalofrontotemporal circuitry should be searched in in CGRP. Further research on the hodological, embryolog- the connections of the PDVR/LA/DLA with other portions ical and histochemical features of the mammalian lateral of the pallium. This finding would point to the ADVR as a pallium are needed to test this hypothesis. possible candidate for the reptilian frontotemporal cortex, As put forward by Martıńez-Garcıá et al. (2002a), these because it is interconnected with the caudal aspect of the findings have important implications to reinterpret the DVR (Andreu et al., 1996; Lanuza et al., 1998). Neverthe- organization of the avian telencephalon. In their study of less, two facts seriously challenge this view. First, the “corticostriatal” pathways in the pigeon, Veenman et whereas the reptilian DVR (including its anterior pole) al. (1995) report projections from parts of the archistria- seems to belong to the ventral pallium in view of its profile tum and most of the caudolateral neostriatum (Kro¨ner of expression of homeotic genes during embryonic devel- and Gu¨ ntu¨ rku¨ n, 1999) to what they call the visceral– opment (Smith-Fernańdez et al., 1998), similar data in limbic basal ganglia, a continuum of structures comprised mammals (Puelles et al., 2000) suggest that the prefrontal between the medial lobus paraolfactorius (LPO) and the and temporal cortices include no ventropallial derivatives. BST. This finding clearly recalls the projections from the In addition, in the mammalian brain the basolateral posterior DVR to the Acb–SAT we have described in rep- amygdala and prefrontal cortex are not only intercon- tiles. Although the classic view (Zeier and Karten, 1971) nected directly, but they project in a coherent manner to claims that the avian amygdala corresponds to parts of the the nucleus accumbens (Brog et al., 1993; McDonald, so-called archistriatum, this correspondence strongly sug- 1991). To the contrary, the reptilian ADVR projects not to gests that the avian basolateral amygdala includes not the Acb (Gonza´lez et al., 1990; this work) but only to the only parts of the archistriatum but also parts of the over- DSt. Therefore, the most likely hypothesis posits that the lying neostriatum and lateropallial derivatives (such as ADVR of reptiles is not homologous (not even analogous) the so-called area temporo-parieto-occipitalis, TPO). Sev- to the frontotemporal cortex but to some, as yet not iden- eral additional histochemical and hodological data (see tified, ventropallial territory in the rostral telencephalon Martıńez-Garcıá et al., 2002a) as well as data on the projecting to the dorsal but not the ventral striatum. expression of homeotic genes during late embryonic devel- Bruce and Neary (1995b) proposed an alternative hypoth- opment in mammals, birds (Smith-Fernańdez et al., 1998; esis for the comparative significance of the reptilian Puelles et al., 2000), and reptiles (Smith-Fernańdez et al., AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 305

1998) lend further support to this hypothesis. In this very tu¨ rku¨ n, 1999) in processing of the reward of incoming moment, when the nomenclature of the avian forebrain is stimuli. being updated (see Reiner et al., 2004), answers are ur- In addition, there is evidence of a role for the avian gently needed to clarify this point. archistriatum in modulation of anxiety- and fear-related The full pattern of amygdalostriatal projections is com- behaviors (Lowndes and Davies, 1996). This modulation is pleted with projections arising from the vomeronasal very likely mediated by the archistriatal projections amygdala and terminating in or near the ventral stria- (Veenman et al., 1995; Davies et al., 1997; Dubbeldam et tum. In squamate reptiles, this pattern is represented by al., 1997; Kro¨ner and Gu¨ ntu¨ rku¨ n, 1999) to the so-called the massive projections from the nucleus sphericus to the lateral BST (Aste et al., 1998), a cell group caudal to the “olfactostriatum” reported in snakes by Lanuza and Halp- LPO/accumbens that displays long descending projections ern (1997), a projection that is also present (although to the hypothalamus (Berk, 1987), parabrachial area much less prominent) in Podarcis (see the Results sec- (Wild et al., 1990), and nucleus of the dorsal vagal motor tion). In rodents, both the medial amygdala and the pos- nucleus (Berk, 1987), reminiscent of those of the mamma- teromedial cortical amygdala (putative homologue to the lian central extended amygdala. Other neurochemical nucleus sphericus; Martıńez-Garcıá et al., 2002a) appar- similarities, such as the presence of projection cells con- ently give rise to a relatively minor (compared with the taining corticotropin-releasing factor (Richard et al., 2004) projections from the basolateral amygdala) but substan- and neurotensin (Atoji et al., 1996) further support the tial projection to the accumbens (Canteras et al., 1992, homology between the avian BST and mammalian central 1995; Gomez and Newman, 1992; Brog et al., 1993), which extended amygdala. mainly terminates in the medial shell (in or beyond the In reptiles, functional data are even scarcer. Neverthe- rostral tip of the extended amygdala). less, lesions of the reptilian homologue to the mammalian central extended amygdala, the SAT, impair the expres- Functional and evolutionary implications sion of fear-related defensive behaviors such as tonic im- mobility (Davies et al., 2002), thus suggesting a role of the As mentioned above, the DVR has been considered as SAT and its afferents (from the PDVR, DLA, and dLC) in the sensory telencephalon of reptiles and birds, due to its the expression of fear and fear-related behaviors. In fact, massive afferents from the sensory dorsal thalamic nuclei electric stimulation and lesions of the “amygdaloid region” (Ulinski, 1983). Nevertheless additional data on the con- (caudal aspect of the DVR) of the telencephalon of croco- nections (see above), histochemistry and development of diles (Keating et al., 1970) results in changes in behavior the DVR indicate that, at least, its caudal part has a clear suggestive of a role for this region in the expression of fear. “amygdaloid” nature. Thus, the projection from the PDVR Considered together, these data on the anatomy, neu- to striatal territories probably accomplish a distinct role in rochemistry, and function of the caudal aspect of the DVR the control of behavior. of reptiles and birds strongly suggest that, through its In mammals, there is evidence that the projections from projections to the striatum and to its caudal continuation, basolateral amygdala to the ventral striatum (nucleus they might control the expression of basic emotional be- accumbens) are involved in stimulus-reward associations haviors, namely reward/attraction (projections to the Acb/ (Cador et al., 1989; Everitt et al., 1989; Parkinson et al., medial LPO) or fear/anxiety/aversion (projections to the 2001). On the other hand, the central extended amygdala BST of birds, SAT of reptiles). This possibility strongly (central amygdala and portions of the bed nucleus of the suggest that the ancestral amniote possessed an amygda- stria terminalis) seems to play a key role in the expression lostriatal system that made up the primordium of the of fear and anxiety (Davis and Shi, 1999). Thus, the pro- emotional brain. In the ancestral amniote, the pallial re- jections from the basolateral amygdala to the continuum gions giving rise to projections to the Acb and central extended amygdala–accumbens might be conceived as the extended amygdala would show an associative nature due neural substrate for elaborating basic emotional re- to convergent afferents from the sensory telencephalon, as sponses to incoming stimuli (Davis, 1994; Ledoux, 2000), it happens in modern reptiles (PDVR, Lanuza et al., 1998) namely reward–attraction (projections to the accumbens) and birds (caudolateral neostriatum, Metzger et al., 1998; or fear–anxiety–aversion (projections to the central ex- Kro¨ner and Gu¨ ntu¨ rku¨ n, 1999). Moreover, this ancestral tended amygdala). basolateral amygdala would also display direct afferents Although functional studies on nonmammalian am- from multimodal nuclei of the posterior thalamus, as ob- niotes are scarce, it is tempting to suggest that the pro- served in both reptiles (afferents from the medial– jections from the posterior aspect of the DVR of birds and posteromedial thalamus; Lanuza et al., 1998) and birds reptiles to the ventral striatum might also have a function (thalamic afferents from the dorsolateral posterior thala- in the emotional evaluation of stimuli and the elaboration mus and periovoidal region; Gamlin and Cohen, 1986; of appropriate responses. In this respect, lesions of the Wild, 1987; Durand et al., 1992). Very likely, in ancestral ventromedial basal ganglia (LPO) in newborn chicken re- amniotes, this thalamic afferent would already have a sult in an impaired learning of stimulus–reward associa- component rich in CGRP, like the one present in birds tions (Izawa et al., 2002), as well as in the expression of (Lanuza et al., 2000) and reptiles (Martıńez-Garcıá et al., impulsive behavior (Izawa et al., 2003). This finding indi- 2002b). cates that the avian LPO, like the mammalian nucleus This pattern of connections with redundant multimodal accumbens/ventral striatum is involved in the manage- sensory afferents and outputs to striatal regions mediat- ment of behavioral economy, by assessment of the ing basic emotional responses constitutes the appropriate reward–effort balance. This role of the LPO in reward neuroanatomical framework for emotional tagging of in- signaling, in turn, would implicate its afferents (e.g., from coming stimuli by their association with previously the caudolateral neostriatum and lateral portions of the attractive-rewarding or aversive (fear-eliciting) ones. This archistriatum; Veenman et al., 1995; Kro¨ner and Gu¨n- circuitry would constitute the basic neural machinery for 306 A. NOVEJARQUE ET AL. the emotional expression and learning, the emotional Davis M, Shi C. 1999. The extended amygdala: are the central nucleus of brain, which was probably present in the telencephalon of the amygdala and the bed nucleus of the stria terminalis differentially the ancestral amniote. These findings, along with numer- involved in fear versus anxiety? Ann N Y Acad Sci 877:281–291. Dong HW, Petrovich GD, Swanson LW. 2001. Topography of projections ous other lines of evidence, support the need to discard the from amygdala to bed nuclei of the stria terminalis. Brain Res Rev mammalian “triune brain” hypothesis of MacLean (1978), 38:192–246. which posited that mammals posses a “reptilian brain” Dubbeldam JL, den Boer-Visser AM, Bout RG. 1997. Organization and fundamentally composed of a striatal core upon which efferent connections of the archistriatum of the mallard, Anas platy- limbic (emotional paleomammalian) and neocortical (neo- rhynchos L.: an anterograde and retrograde tracing study. J Comp mammalian) layers grew. To the contrary, at least the two Neurol 388:632–657. Durand SE, Tepper JM, Cheng MF. 1992. The shell region of the nucleus former components of MacLean’s sequentially additive ovoidalis: a subdivision of the avian auditory thalamus. J Comp Neurol scenario were already present in stem amniotes. 323:495–518. Everitt BJ, Cador M, Robbins TW. 1989. Interactions between the amyg- LITERATURE CITED dala and ventral striatum in stimulus-reward associations: studies using a second-order schedule of sexual reinforcement. Neuroscience Alheid GF, de Olmos JS, Beltramino CA. 1995. Amygdala and extended 30:63–75. amygdala. In: Paxinos G, editor. The rat nervous system. San Diego: Font C, Hoogland PV, Vermeulen-VanderZee E, Pe´rez-Clausell J, Academic Press. p 495–578. Martıńez-Garcıá F. 1995. The septal complex of the telencephalon of Andreu MJ, Da´vila JC, Real MA, Guirado S. 1996. Intrinsic connections in the lizard Podarcis hispanica. I. Chemoarchitectonical organization. the anterior dorsal ventricular ridge of the lizard Psammodromus J Comp Neurol 359:117–130. algirus. J Comp Neurol 372:49–58. Font C, Lanuza E, Martıńez-Marcos A, Hoogland PV, Martıńez-Garcıá F. Aste N, Balthazart J, Absil P, Grossmann R, Mu¨ lhbauer E, Viglietti- 1998. Septal complex of the telencephalon of lizards: III. Efferent Panzica C, Panzica GC. 1998. Anatomical and neurochemical definition connections and general discussion. J Comp Neurol 401:525–548. of the nucleus of the stria terminalis in Japanese quail (Coturnix Gamlin PD, Cohen DH. 1986. A second ascending visual pathway from the japonica). J Comp Neurol 396:141–157. optic tectum to the telencephalon in the pigeon (Columba livia). Atoji Y, Shibata N, Yamamoto Y, Suzuki Y. 1996. Distribution of J Comp Neurol 250:296–310. neurotensin-containing neurons in the central nervous system of the Geneser-Jensen FA, Blackstad TW. 1971. Distribution of acetyl cholines- pigeon and the chicken. J Comp Neurol 375:187–211. terase in the hippocampal region of the guinea pig. I. Entorhinal area, Behan M, Haberly LB. 1999. Intrinsic and efferent connections of the parasubiculum, and presubiculum. Z Zellforsch Mikrosk Anat 114: endopiriform nucleus in rat. J Comp Neurol 408:532–548. 460–481. Belekhova MG, Chkheidze DD. 1991. Afferent and efferent connections of Gomez DM, Newman SW. 1992. Differential projections of the anterior and the amygdaloid complex with the anterior dorsal ventricular ridge of posterior regions of the medial amygdaloid nucleus in the Syrian ham- the telencephalon of the turtle Emys orbicularis: peroxidase study. Zh ster. J Comp Neurol 317:195–218. Evol Biokhi Fiziol 27:757–767. Gonza´lez A, Russchen FT, Lohman AH. 1990. Afferent connections of the Belekhova MG, Chkheidze DD. 1992. Connections of the amygdaloid com- striatum and the nucleus accumbens in the lizard Gekko gecko. Brain plex with basotelencephalic, diencephalic and brainstem structures in Behav Evol 36:39–58. turtles: HRP study. Rusa Zh Evol Biokhi Fiziol 28:614–622. Groenewegen HJ, Becker NE, Lohman AH. 1980. Subcortical afferents of Berk ML. 1987. Projections of the lateral hypothalamus and bed nucleus of the nucleus accumbens septi in the cat, studied with retrograde axonal the stria terminalis to the dorsal vagal complex in the pigeon. J Comp transport of horseradish peroxidase and bisbenzimid. Neuroscience Neurol 260:140–156. 5:1903–1916. Brog JS, Salyapongse A, Deutch AY, Zahm DS. 1993. The patterns of Groenewegen HJ, Vermeulen-VanderZee E, te Kortschot A, Witter MP. afferent innervation of the core and shell in the “accumbens” part of the 1987. Organization of the projections from the subiculum to the ventral rat ventral striatum: immunohistochemical detection of retrogradely striatum in the rat. A study using anterograde transport of Phaseolus transported Fluoro-Gold. J Comp Neurol 338:255–278. vulgaris leucoagglutinin. Neuroscience 23:103–120. Bruce LL, Butler AB. 1984a. Telencephalic connections in lizards. I. Pro- Groenewegen HJ, Wright CI, Beijer AV, Voorn P. 1999. Convergence and jections to cortex. J Comp Neurol 229:585–601. segregation of ventral striatal inputs and outputs. Ann N Y Acad Sci Bruce LL, Butler AB. 1984b. Telencephalic connections in lizards. II. 877:49–63. Projections to anterior dorsal ventricular ridge. J Comp Neurol 229: Guirado S, Da´vila JC, Real MA, Medina L. 1999. Nucleus accumbens in the 602–615. lizard Psammodromus algirus: chemoarchitecture and cortical afferent Bruce LL, Neary TJ. 1995a. Afferent projections to the ventromedial hy- connections. J Comp Neurol 405:15–31. pothalamic nucleus in a lizard, Gekko gecko. Brain Behav Evol 46:14– Guirado S, Da´vila JC, Real MA, Medina L. 2000. Light and electron 29. microscopic evidence for projections from the thalamic nucleus rotun- Bruce LL, Neary TJ. 1995b. The limbic system of tetrapods: a comparative dus to targets in the basal ganglia, the dorsal ventricular ridge, and the analysis of cortical and amygdalar populations. Brain Behav Evol amygdaloid complex in a lizard. J Comp Neurol 424:216–232. 46:224–234. Halliday G, Harding A, Paxinos G. 1995. Serotonin and tachykinin sys- Butler AB. 1994. The evolution of the dorsal pallium in the telencephalon tems. In: Paxinos G, editor. The rat nervous system. San Diego: Aca- of amniotes: cladistic analysis and a new hypothesis. Brain Res Rev demic Press. p 929–974. 19:66–101. Hatanaka T, Matsuzaki O. 1993. Odor responses of the vomeronasal sys- Cador M, Robbins TW, Everitt BJ. 1989. Involvement of the amygdala in tem in Reeve’s turtle, Geoclemys reevesii. Brain Behav Evol 41:183– stimulus-reward associations: interaction with the ventral striatum. 186. Neuroscience 30:77–86. Holmgren N. 1925. Points of view concerning forebrain morphology in Canteras NS, Simerly RB, Swanson LW. 1992. Connections of the posterior higher vertebrates. Acta Zool Stockh 6:415–477. nucleus of the amygdala. J Comp Neurol 324:143–179. Hoogland PV. 1977. Efferent connections of the striatum in Tupinambis Canteras NS, Simerly RB, Swanson LW. 1995. Organization of projections nigropunctatus. J Morphol 152:229–246. from the medial nucleus of the amygdala: a PHAL study in the rat. Hoogland PV, Vermeulen-VanderZee E. 1989. Efferent connections of the J Comp Neurol 360:213–245. dorsal cortex of the lizard Gekko gecko studied with Phaseolus Davies DC, Csillag A, Sze´kely AD, Kabai P. 1997. Efferent connections of vulgaris-leucoagglutinin. J Comp Neurol 285:289–303. the domestic chick archistriatum: a Phaseolus lectin anterograde trac- Hoogland PV, Vermeulen-VanderZee E. 1990. Distribution of choline ing study. J Comp Neurol 389:679–693. acetyltransferase immunoreactivity in the telencephalon of the lizard Davies DC, Martinez-Garcia F, Lanuza E, Novejarque A. 2002. Striato- Gekko gecko. Brain Behav Evol 36:378–390. amygdaloid transition area lesions reduce the duration of tonic immo- Izawa E, Zachar G, Aoki N, Koga K, Matsushima T. 2002. Lesions of the bility in the lizard Podarcis hispanica. Brain Res Bull 57:537–541. ventro-medial basal ganglia impair the reinforcement but not the recall Davis M. 1994. The role of the amygdala in emotional learning. Int Rev of memorized color discrimination in domestic chicks. Behav Brain Res Neurobiol 36:225–266. 136:405–414. AMYGDALOSTRIATAL PROJECTIONS IN LIZARDS 307

Izawa E, Zachar G, Yanagihara S, Matsushima T. 2003. Localized lesion of tions of the amygdaloid formation of the lizard Podarcis hispanica. caudal part of lobus parolfactorius caused impulsive choice in the Brain Behav Evol 41:156–162. domestic chick: evolutionarily conserved function of ventral striatum. Martıńez-Garcıá F, Martıńez-Marcos A, Lanuza E. 2002a. The pallial J Neurosci 23:1894–1902. amygdala of amniote vertebrates: evolution of the concept, evolution of Jongen-Relo AL, Voorn P, Groenewegen HJ. 1994. Immunohistochemical the structure. Brain Res Bull 57:463–469. characterization of the shell and core territories of the nucleus accum- Martıńez-Garcıá F, Novejarque A, Landete JM, Moncho-Bogani J, Lanuza bens in the rat. Eur J Neurosci 6:1255–1264. E. 2002b. Distribution of calcitonin gene-related peptide-like immuno- Kaas JH, Reiner A. 1999. The neocortex comes together. Nature 399:418– reactivity in the brain of the lizard Podarcis hispanica. J Comp Neurol 419. 447:99–113. Karten H. 1969. The organization of the avian telencephalon and some Martıńez-Marcos A, Lanuza E, Font C, Martıńez-Garcıá F. 1999. Afferents speculations on the phylogeny of the amniote telencephalon. Ann N Y to the red nucleus in the lizard Podarcis hispanica: putative pathways Acad Sci 167:164–180. for visuomotor integration. J Comp Neurol 411:35–55. Karten HJ. 1991. Homology and evolutionary origins of the neocortex. McDonald AJ. 1991. Topographical organization of amygdaloid projections Brain Behav Evol 38:264–272. to the caudatoputamen, nucleus accumbens, and related striatal-like areas of the rat brain. Neuroscience 44:15–33. Karten HJ. 1997. Evolutionary development biology meets the brain: the origins o mammalian cortex. Proc Natl Acad SciUSA9:2800–2804. McDonald AJ, Shammah-Lagnado SJ, Shi C, Davis M. 1999. Cortical afferents to the extended amygdala. Ann N Y Acad Sci 877:309–338. Karten HJ, Hodos W, Nauta WJH, Revzin AM. 1973. Neural connections of the “Visual Wulst” of the avian telencephalon. Experimental studies in Medina L, Smeets WJAJ, Hoogland PV, Puelles L. 1993. Distribution of the pigeon (Columba livia) and owl (Speotyto cunicularia). J Comp choline acetyltransferase immunoreactivity in the brain of the lizard Neurol 150:253–278. Gallotia galloti. J Comp Neurol 331:261–285. Keating EG, Kormann LA, Horel JA. 1970. The behavioral effects of stim- Metzger M, Jiang S, Braun K. 1998. Organization of the dorsocaudal ulating and ablating the reptilian amygdala (Caiman sklerops). Physiol neostriatal complex: a retrograde and anterograde tracing study in the Behav 5:55–59. domestic chick with special emphasis on pathways relevant to imprint- ing. J Comp Neurol 395:380–404. Kelley AE, Domesick VB, Nauta WJH. 1982. The amygdalostriatal projection in the rat - an anatomical study by anterograde and retrograde Murphy FA, Tucker K, Fadool DA. 2001. Sexual dimorphism and develop- tracing methods. Neuroscience 7:615–630. mental expression of signal-transduction machinery in the vomeronasal organ. J Comp Neurol 432:61–74. Kro¨ner S, Gu¨ ntu¨ rku¨ n O. 1999. Afferent and efferent connections of the caudolateral neostriatum in the pigeon (Columba livia): a retro- and Naito A, Kita H. 1994. The cortico-pallidal projection in the rat: an antero- anterograde pathway tracing study. J Comp Neurol 407:228–260. grade tracing study with biotinylated dextran amine. Brain Res. 653: 251–257. Lanuza E, Halpern M. 1997. Afferent and efferent connections of the nucleus sphericus in the snake Thamnophis sirtalis: convergence of Olmos JS, Heimer L. 1999. The concepts of the ventral striatopallidal olfactory and vomeronasal information in the lateral cortex and the system and extended amygdala. Ann N Y Acad Sci 877:1–32. amygdala. J Comp Neurol 385:627–640. Osborn M, Weber K. 1982. Immunofluorescence and immunocytochemical procedures with affinity purified antibodies: tubulin-containing struc- Lanuza E, Font C, Martıńez-Marcos A, Martıńez-Garcıá F. 1997. tures. Methods Cell Biol 24:97–132. Amygdalo-hypothalamic projections in the lizard Podarcis hispanica:a combined anterograde and retrograde tracing study. J Comp Neurol Parkinson JA, Crofts HS, McGuigan M, Tomic DL, Everitt BJ, Roberts AC. 384:537–555. 2001. The role of the primate amygdala in conditioned reinforcement. J Neurosci 21:7770–7780. Lanuza E, Belekhova M, Martıńez-Marcos A, Font C, Martıńez-Garcıá F. 1998. Identification of the reptilian basolateral amygdala: an anatom- Pe´rez-Clausell J. 1988. Organization of zinc-containing terminal fields in ical investigation of the afferents to the posterior dorsal ventricular the brain of the lizard Podarcis hispanica: a histochemical study. ridge of the lizard Podarcis hispanica. Eur J Neurosci 10:3517–3534. J Comp Neurol 267:153–171. Lanuza E, Davies DC, Landete JM, Novejarque A, Martıńez-Garcıá F. Pe´rez-Santana L, Marıń O, Smeets WJAJ. 1997. Afferent connections of 2000. Distribution of CGRP-like immunoreactivity in the chick and the nucleus accumbens of the snake, Elaphe guttata, studied by means quail brain. J Comp Neurol 421:515–532. of in vitro and in vivo tracing techniques in combination with TH immunocytochemistry. Neurosci Lett 225:101–104. Lanuza E, Novejarque A, Moncho-Bogani J, Hernańdez A, Martıńez- Garcıá F. 2002. Understanding the basic circuitry of the cerebral hemi- Petrovich GD, Risold PY, Swanson LW. 1996. Organization of projections spheres: the case of lizards and its implications in the evolution of the from the basomedial nucleus of the amygdala: a PHAL study in the rat. telencephalon. Brain Res Bull 57:471–473. J Comp Neurol 374:387–420. Ledoux JE. 2000. Emotion circuits in the brain. Annu Rev Neurosci 23: Pitka¨nen A. 2000. Connectivity of the rat amygdaloid complex. In: Aggle- 155–184. ton JP, editor. The amygdala. A functional analysis. Oxford: Oxford University Press. p 31–115. Lohman AHM, Smeets JAJ. 1990. The dorsal ventricular ridge and cortex Pitka¨nen A, Amaral DG. 1998. Organization of the intrinsic connections of of reptiles in historical and phylogenetic perspective. In: Finley BL, the monkey amygdaloid complex: projections originating in the lateral Innocenti G, Scheich H, editors. The neocortex: ontogeny and phylog- nucleus. J Comp Neurol 398:431–458. eny. New York: Plenum Press. p 59–74. Pitka¨nen A, Stefanacci L, Farb CR, Go GG, Ledoux JE, Amaral DG. 1995. Lohman AHM, van Woerden-Verkley I. 1978. Ascending connections to the Intrinsic connections of the rat amygdaloid complex: projections origi- forebrain in the Tegu lizard. J Comp Neurol 182:555–594. nating in the lateral nucleus. J Comp Neurol 356:288–310. Lowndes M, Davies DC. 1996. The effect of archistriatal lesions on “open Powers AS, Reiner A. 1993. The distribution of cholinergic neurons in the field” and fear/avoidance behavior in the domestic chick. Behav Brain central nervous system of turtles. Brain Behav Evol 41:326–345. Res 72:25–32. Puelles L, Kuwana E, Puelles E, Bulfone A, Shimamura K, Keleher J, MacLean PD. 1978. Why brain research on lizards? In: Greenberg N, Smiga S, Rubenstein JL. 2000. Pallial and subpallial derivatives in the MacLean PD, editors. Behavior and neurology of lizards. An interdis- embryonic chick and mouse telencephalon, traced by the expression of ciplinary colloquium. Rockville, MD: National Institute of Mental the genes Dlx-2, Emx-1, Nkx- 2.1, Pax-6, and Tbr-1. J Comp Neurol Health. p 1–10. 424:409–438. Manger PR, Slutsky DA, Molnar Z. 2002. Visual subdivisions of the dorsal Regidor J, Poch L. 1988. Histochemical analysis of the lizard dorsal cortex: ventricular ridge of the iguana (Iguana iguana) as determined by an acetylcholinesterase, cytochrome oxidase and NADPH-diaphorase electrophysiologic mapping. J Comp Neurol 453:226–246. study. In: Schwerdtfeger WK, Smeets WJAJ, editors. The forebrain of Martıńez-Garcıá F, Olucha FE. 1988. Afferent projections to the Timm- reptiles. Current concepts of structure and function. Basel: Karger. p positive cortical areas of the telencephalon of lizards. In: Schwerdtfeger 77–84. WK, Smeets WJAJ, editors. The forebrain of reptiles. Current concepts Reiner A, Karten HJ. 1985. Comparison of olfactory bulb projections in of structure and function. Basel: Karger. p 30–40. pigeons and turtles. Brain Behav Evol 27:11–27. Martıńez-Garcıá F, Amiguet M, Olucha F, Lo´pez-Garcıá C. 1986. Connec- Reiner A, Perkel DJ, Bruce LL, Butler AB, Csillag A, Kuenzel W, Medina tions of the lateral cortex in the lizard Podarcis hispanica. Neurosci L, Paxinos G, Shimizu T, Striedter G, Wild M, Ball GF, Durand S, Lett 63:39–44. Guturkun O, Lee DW, Mello CV, Powers A, White SA, Hough G, Martıńez-Garcıá F, Olucha FE, Teruel V, Lorente MJ. 1993. Fiber connec- Kubikova L, Smulders TV, Wada K, Dugas-Ford J, Husband S, 308 A. NOVEJARQUE ET AL.

Yamamoto K, Yu J, Siang C, Jarvis ED. 2004. Revised nomenclature In: Sternberger LA, editor. Inmunocytochemistry. New York: John for avian telencephalon and some related brainstem nuclei. J Comp Wiley and Sons. p 104–169. Neurol 473:377–414. Striedter GF. 1997. The telencephalon of tetrapods in evolution. Brain Richard S, Martıńez-Garcıá F, Lanuza E, Davies DC. 2004. Distribution of Behav Evol 49:179–213. corticotropin-releasing factor-immunoreactive neurons in the central Swanson LW, Petrovich GD. 1998. What is the amygdala? Trends Neurosci nervous system of the domestic chicken and Japanese quail. J Comp 21:323–331. Neurol 469:559–580. Swanson LW, Risold PY. 1999. On the basic architecture of the septal Russchen FT, Jonker AJ. 1988. Efferent connections of the striatum and region. In: Numan R, editor. The behavioral neuroscience of the septal the nucleus accumbens in the lizard Gekko gecko. J Comp Neurol region. New York: Springer. p 1–14. 276:61–80. Ulinski PS. 1978. Structure of anterior dorsal ventricular ridge in snakes. Russchen FT, Price JL. 1984. Amygdalostriatal projections in the rat. J Morphol 148:1–22. Topographical organization and fiber morphology shown using the Ulinski PS. 1983. Dorsal ventricular ridge. A treatise on forebrain organi- lectin PHA-L as an anterograde tracer. Neurosci Lett 47:15–22. zation in reptiles and birds. New York: John Wiley and Sons. Russchen FT, Smeets WJAJ, Hoogland PV. 1987. Histochemical identifi- Veenman CL, Wild JM, Reiner A. 1995. Organization of the avian cortico- cation of pallidal and striatal structures in the lizard Gekko gecko: striatal projection system: a retrograde and anterograde pathway trac- evidence for compartmentalization. J Comp Neurol 256:329–341. ing study in pigeons. J Comp Neurol 354:87–126. Savander V, Go CG, Ledoux JE, Pitka¨nen A. 1995. Intrinsic connections of Voneida TJ, Sligar CM. 1979. Efferent projections of the dorsal ventricular the rat amygdaloid complex: projections originating in the basal nu- ridge and the striatum in the tegu lizard, Tupinambis nigropunctatus. cleus. J Comp Neurol 361:345–368. J Comp Neurol 186:43–64. Savander V, Go G-G, Ledoux JE, Pitka¨nen A. 1996. Intrinsec connections Wild JM. 1987. The avian somatosensory system: connections of regions of of the rat amygdaloid complex: projections originating in the accessory body representation in the forebrain of the pigeon. Brain Res 412:205– basal nucleus. J Comp Neurol 374:291–313. 223. Siemen M, Kunzle H. 1994. Connections of the basal telencephalic areas c Wild JM, Arends JJ, Zeigler HP. 1990. Projections of the parabrachial and d in the turtle brain. Anat Embryol (Berl) 189:339–359. nucleus in the pigeon (Columba livia). J Comp Neurol 293:499–523. Smeets WJ, Medina L. 1995. The efferent connections of the nucleus Wright CI, Beijer AVJ, Groenewegen HJ. 1996. Basal amygdaloid complex accumbens in the lizard Gekko gecko. A combined tract-tracing/ afferents to the rat nucleus accumbens are compartmentally organized. transmitter-immunohistochemical study. Anat Embryol (Berl) 191:73– J Neurosci 16:1877–1893. 81. Yasui Y, Saper CB, Cechetto DF. 1989. Calcitonin gene-related peptide Smeets WJ, Pe´rez-Clausell J, Geneser FA. 1989. The distribution of zinc in immunoreactivity in the visceral sensory cortex, thalamus, and related the forebrain and midbrain of the lizard Gekko gecko. A histochemical pathways in the rat. J Comp Neurol 290:487–501. study. Anat Embryol (Berl) 180:45–56. Zahm DS. 1998. Is the caudomedial shell of the nucleus accumbens part of Smith-Fernańdez A, Pieau C, Reperant J, Boncinelli E, Wassef M. 1998. the extended amygdala? A consideration of connections. Crit Rev Neu- Expression of the Emx-1 and Dlx-1 homeobox genes define three mo- robiol 12:245–265. lecularly distinct domains in the telencephalon of mouse, chick, turtle Zahm DS. 1999. Functional-anatomical implications of the nucleus accum- and frog embryos: implications for the evolution of telencephalic sub- bens core and shell subterritories. Ann N Y Acad Sci 877:113–128. divisions in amniotes. Development 125:2099–2111. Zeier H, Karten HJ. 1971. The archistriatum of the pigeon: organization of Sternberger LA. 1979. The unlabeled antibody peroxidase (PAP) method. afferent and efferent connections. Brain Res 31:313–326.