E Volutionary Divergence of the Reptilian and the Mammalian Brains

Brain Research Reviews 39 (2002) 141–153 www.elsevier.com/locate/brainresrev

Review E volutionary divergence of the reptilian and the mammalian brains: considerations on connectivity and development Francisco Aboitiza,* , Juan Montiel a , Daniver Morales b , Miguel Concha a,c

aPrograma de Morfologıa´´, Instituto de Ciencias Biomedicas, Facultad de Medicina, Universidad de Chile, Santiago, Chile bDevelopmental Neurobiology Laboratory, The Rockefeller University, New York, NY, USA cDepartment of Anatomy and Developmental Biology, University College, London, UK Accepted 27 June 2002

Abstract

The isocortex is a distinctive feature of the mammalian brain, with no clear counterpart in other amniotes. There have been long controversies regarding possible homologues of this structure in reptiles and birds. The brains of the latter are characterized by the presence of a structure termed dorsal ventricular ridge (DVR), which receives ascending auditory and visual projections, and has been postulated to be homologous to parts of the mammalian isocortex (i.e., the auditory and the extrastriate visual cortices). Dissenting views, now supported by molecular evidence, claim that the DVR originates from a region termed ventral pallium, while the isocortex may arise mostly from the dorsal pallium (in mammals, the ventral pallium relates to the claustroamygdaloid complex). Although it is possible that in mammals the embryonic ventral pallium contributes cells to the developing isocortex, there is no evidence yet supporting this alternative. The possibility is raised that the expansion of the cerebral cortex in the origin of mammals was product of a generalized dorsalizing inﬂuence in pallial development, at the expense of growth in ventral pallial regions. Importantly, the evidence suggests that organization of sensory projections is signiﬁcantly different between mammals and sauropsids. In reptiles and birds, some sensory pathways project to the ventral pallium and others project to the dorsal pallium, while in mammals sensory projections end mainly in the dorsal pallium. We suggest a scenario for the origin of the mammalian isocortex which relies on the development of associative circuits between the olfactory, the dorsal and the hippocampal cortices in the earliest mammals.  2002 Elsevier Science B.V. All rights reserved.

Theme: Other systems of the CNS

Topic: Comparative neuroanatomy

Keywords: Amygdala; Dorsal cortex; Dorsal ventricular ridge; Homology; Isocortex; Pallium; Regulatory genes; Ventral pallium

Contents

1 . Introduction. the problem of isocortical origins ...... 142 2 . The pallium of amniotes...... 143 3 . Connectional and neurochemical comparisons: the recapitulation hypothesis...... 143 4 . Conﬂicting connectional evidence: the outgroup hypothesis...... 145 5 . Developmental criteria ...... 145 5 .1. Other components of the DVR ...... 147 5 .2. Dorsoventral gradients and their relation to the IT/VP ...... 147 6 . Different patterns of brain organization in reptiles and mammals ...... 148 7 . A scenario for isocortical origins: olfaction, the hippocampus and the thalamofugal visual system...... 149 7 .1. Fossil brains ...... 149

*Corresponding author. Depto. de Psiquiatria, Facultad de Medicina, Pontiﬁcia Universidad Catolica de Chile, Marcoleta No. 387, 28 Piso, P.O. Box 114-D, Santiago 1, Chile. Fax: 156-2-665-1951. E-mail address: [email protected] (F. Aboitiz).

8 . Final comment ...... 150 Acknowledgements ...... 151 References...... 151

1 . Introduction. the problem of isocortical origins assumption that there are comparable components (homologues) in the different taxa, despite their superficial The mammalian isocortex (or neocortex) is a character dissimilarities. In order to understand the origin of the unique to the brain of mammals in several respects. First, it mammalian isocortex, it is fundamental to determine has undergone an enormous expansion, especially in the which ancestral structure(s) gave rise to it. Although tangential domain [70]. Second, it has a six-layered obviously the ancestor is unavailable to study, there are architecture which differs from the three-layered array of sister taxa (reptiles and birds) whose brains can be simpler telencephalic laminar structures such as the hip- compared with those of mammals in order to infer the pocampal formation, the olfactory cortex and the reptilian characteristics of the common ancestor. Commonly used cortices [3,92]. On the other hand, the telencephalon of criteria for determining neural homology are connectivity reptiles has a small, thin cortex and a prominent periven- [42,43,56], neurochemistry [74,75], topographical location tricular structure named dorsal ventricular ridge (DVR), [1,2,34] and embryonic origins [4,40,66,67,87,90]. Un- which receives many thalamic sensory projections (Fig. 1). fortunately, when dealing with the homologues of the There have been important disagreements as to which isocortex, these different methodologies have led to di- components of the non-mammalian telencephalon can be verging interpretations. compared to the isocortex of mammals. This problem is In this paper we will address the issue of a possible complicated by the intricate topography of the hemispheres correspondence between the reptilian dorsal cortex and the in some vertebrate classes, and by the absence of a unified mammalian isocortex, and will review recent molecular criterion to establish neural homology. Homology is a evidence supporting this interpretation. Then, the general central problem to comparative anatomy, since the organization of the mammalian and reptilian brains will be evolutionary considerations regarding the origin and di- discussed in light of these new findings, and we will versification of any structure are usually based on the propose a scenario for the origin of the mammalian brain.

Fig. 1. Coronal section of the cerebral hemispheres of a reptile and a mammal, indicating in gray the different components of the pallium. The subpallium is shown in white. ADVR, anterior dorsal ventricular ridge; AM, amygdala (only part of which is pallial); CL, claustrum; DCx, dorsal cortex; DMCx, dorsomedial cortex; HIP, hippocampus; ICx, isocortex; LCx, lateral cortex; MCx, medial cortex; OCx, olfactory cortex; PT, pallial thickening (present only in turtles); STR, striatum. Medial is to the left. F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 143

Our main argument along this paper is that the evidence on the isocortex has a dual origin, part of it deriving from the regulatory gene expression provides a notable example of reptilian dorsal cortex, and the other part deriving from the how molecular developmental biology may help to clarify DVR (especially the ADVR). An alternative hypothesis important issues in the evolution of the nervous system. suggests that the isocortex derives mostly from the reptilian dorsal pallium, and has been called the outgroup hypothesis [1,2,4,61,68,90]. Below, we will summarize 2 . The pallium of amniotes some of the evidence favouring each of the two possi- bilities. The reptilian pallium has a three-layered cortex, consist- ing of a medial and a dorsomedial component (both comparable to the mammalian hippocampal formation), a dorsal cortex and a lateral or olfactory cortex [98] (Fig. 1). 3 . Connectional and neurochemical comparisons: the In birds, there is a medial hippocampus and a lateral recapitulation hypothesis olfactory cortex. The avian dorsal cortex is a relatively complex, multilaminated structure called the Wulst. The Before comparing the sensory systems of reptiles and reptilian dorsal cortex and the Wulst of birds receive visual mammals, it will be useful to brieﬂy discuss some aspects projections as well as some somatosensory input [56]. of thalamic connectivity. Thalamic sensory nuclei have Additionally, in reptiles and birds (together are called been recently classiﬁed in two main classes: lem- sauropsida), many non-olfactory sensory projections termi- nothalamic and collothalamic [16,17]. Lemnothalamic nate in the dorsal ventricular ridge (DVR; Fig. 1), which nuclei receive projections from lemniscal systems which bulges into the lateral ventricle above the basal ganglia do not synapse in the midbrain (see Fig. 2). Examples of [96,97]. The DVR is the most expansive telencephalic lemniscal pathways are the somatosensory pathways that component of reptiles and birds and is a main integratory use the spino-thalamic route, and the visual thalamofugal center in their brains. It consists of an anterior part pathway that goes directly from the retina to the dorsal (ADVR) and a posterior or basal part (PDVR). The ADVR lateral geniculate nucleus. Collothalamic nuclei receive (which in birds corresponds to the hyperstriatum ventrale, sensory pathways that synapse in the midbrain colliculi neostriatum and ectostriatum) receives much of the sensory (optic tectum and torus semicircularis; Fig. 2). The most input, and its output is directed mainly to the corpus prominent of the collicular pathways are the visual tec- striatum and to the PDVR. The latter (corresponding to the tofugal pathway, which originates in the retina and relays archistriatum in birds) has been compared to parts of the in the optic tectum, projecting then to the thalamic nucleus mammalian amygdala and projects to the hypothalamus rotundus of sauropsids (‘R’ in Fig. 2) or the pulvinar [46,47,96]. nucleus of mammals (‘P’ in Fig. 2); and the auditory On the other hand, the pallium of mammals mainly pathway, which synapses in the inferior colliculus/torus consists of a hippocampal formation (archicortex), an semicircularis and in the thalamic medial geniculate body. olfactory cortex (paleocortex) and an isocortex or neocor- Lemnothalamic nuclei project to the dorsal cortex of tex interposed between them. There is also a claus- sauropsids and to dorsomedial aspects of the isocortex of troamygdaloid complex that contains both pallial and mammals (i.e., primary visual and somatosensory cortices), subpallial elements. The isocortex receives most of the while collothalamic nuclei project to the ADVR of saurop- ascending sensory input from the thalamus and projects to sids and to ventrolateral isocortical regions of mammals the hippocampus and to the amygdala, as well as sending (i.e., visual extrastriate and auditory cortices) [16,17]. output to several lower brain centers including the corpus On the basis of similarities in sensory projection sys- striatum, thalamus, several brainstem nuclei and the spinal tems, some investigators [43,56,60,85] proposed that the cord. During development, the isocortex originates at least dorsal cortex of reptiles is homologous to the primary in large part from the dorsal pallium [61,70,71] (see also (striate) visual cortex, and to the somatosensory and motor Ref. [4]). Recent studies indicate that additionally, cells cortices of mammals since they all receive lemnothalamic originating in the embryonic corpus striatum migrate projections (see Fig. 1). Furthermore, as the auditory tangentially in a dorsal direction and become incorporated radiation and the tectofugal visual pathway (both col- into the isocortex and the hippocampus, mostly as lothalamic) end in the avian/reptilian ADVR, and in the GABAergic interneurons [10,12,39]. Tangential migration mammalian auditory and extrastriate visual cortices, the of inhibitory neurons from the subpallium has been DVR was considered to be homologous to the ventrolateral recently described in birds [20], which suggests that this isocortex [41–43,60,85]. The point of this hypothesis is mechanism pre-dates the common ancestor of mammals that the common ancestor to both reptiles and mammals and sauropsids. had a reptilian-like brain, perhaps already with a DVR or Which reptilian structures correspond to the mammalian with a primordial component which developed into a true isocortex? One hypothesis—which has been termed the DVR in sauropsids, while in the line leading to mammals recapitulation hypothesis [41–43,60,61]—postulates that this component was transformed into lateral isocortex [76]. 144 F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153

Fig. 2. Diagrams summarizing some visual projections in reptiles and in mammals, which illustrate the differences between collicular and lemniscal sensory pathways. Visual projections originate in the retina (RET) and project in two separate pathways. The lemniscal pathway (in purple) is directed to the dorsal lateral geniculate nucleus (G) in the lemnothalamus (LT), and from there to the dorsal cortex (DCx) of reptiles and to the primary visual cortex (V1) of mammals. The collicular pathway (in red) projects to the optic tectum (or superior colliculus, OT) and from there to the reptilian anterior dorsal ventricular ridge (ADVR), via the nucleus rotundus (R) in the collothalamus (CT). In mammals, this pathway projects to the extrastriate visual cortex (Est, in green), via the pulvinar nucleus (P). This nucleus has been classically considered to be homologous to (R), but some recent analyses [21,32] consider that it is a derived character of mammals, with no clear homologue in sauropsids. The reptilian dorsal cortex (DCx) has connections with the M/DMCx. In mammals,V1 exerts control over the ESt; all sensory isocortical areas indirectly project to the hippocampus (HIP). The reptilian posterior dorsal ventricular ridge (PDVR) is related to parts of the mammalian amygdalar system and receives projections from the ADVR. In mammals, the collicular pathway projects to the amygdala (AM) after a thalamic relay. DL, VM, dorsolateral and ventromedial components of the ADVR. F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 145

4 . Conflicting connectional evidence: the outgroup (isocortex) either arose de novo in mammals, or existed in hypothesis the common ancestor to reptiles and mammals and were subsequently lost in the dorsal cortex of sauropsids. As mentioned, the outgroup hypothesis implies that the isocortex arises mostly from the dorsal pallium, i.e., is most likely homologue to the reptilian dorsal cortex. First, 5 . Developmental criteria we will briefly review some connectional evidence point- ing to important differences between the reptilian ADVR Another and long used homology criterion is develop- and the mammalian isocortex. For example, some authors mental origin. Many authors have argued that the general- [14,21,32,68,73,105] have recently claimed that in terms of ized morphology and topographic relations are shown most overall connectivity, the ADVR is more similar to some clearly in early developmental stages, which facilitates regions of the claustroamygdalar complex of mammals, cross-species comparisons [30,62,83,90]. This assumption than to the ventrolateral isocortex. Both the ADVR and the is valid in cases in which there is cross-species conservat- mammalian laterobasal amygdala receive collothalamic ism of embryonic processes while adult morphology tends projections, and both project to the ipsilateral cortex, the to diverge. Alternatively, in cases of embryological di- corpus striatum and the subpallial amygdala. On the other versity with adult conservatism, comparison of adult hand, the isocortex projects to many other brain regions structures and relations may provide a more accurate beside these. As a consequence, the thalamic nucleus diagnosis for homology [2,4]. In the case of the amniote rotundus (‘R’ in Fig. 2) of sauropsids has been considered telencephalon, phylogenetic evidence points to a notable by these authors to be more comparable to the intralaminar conservation of early embryonic structure with adult nuclei of mammals (which receive collothalamic input and diversification [2,4,90], which puts weight on embryologi- project to the lateral amygdala) than to the mammalian cal comparisons as a reliable homology criterion. pulvinar (‘P’ in Fig. 2). There are two other important Studies indicate that during embryogenesis, the ADVR differences in connectivity between the ventrolateral iso- develops from the lateral aspect of the pallium, in a cortex and the ADVR, which need to be mentioned here. position deep (in the radial direction) and ventral to the The first is that the primary visual cortex of mammals olfactory cortex [40,91]. On the other hand, most of the projects densely to the extrastriate visual cortex (especially isocortex is considered to originate from the dorsal pallium V2) [58,79], while in reptiles, projections from the dorsal (excepting the above-mentioned tangentially migrating cortex to the ADVR are considered to be much weaker GABAergic cells that originate in the subpallium). This [96,98]. Second, the mammalian isocortex projects re- evidence was initially claimed to support the hypothesis of ciprocally to the entorhinal cortex and from there to the non-homology between the ADVR and the isocortex hippocampus [36,80,100], while in reptiles few if any [1,2,4,90]. Furthermore, studies of expression patterns of connections have been reported from the ADVR to the regulatory homeobox-like genes in the embryonic fore- hippocampus [96–98]. Finally, despite the similarities in brain have revealed a conserved mosaic organization in sensory connectivity between the ventrolateral isocortex which the different compartments develop into specific and the ADVR, it is important to note that sensory brain components in the adult [31,57,65,84]. In the em- pathways may not always end in comparable structures. In bryonic mammalian telencephalon, distinct genetic amphibians, collothalamic pathways terminate mainly in markers for pallial and for subpallial regions have been the corpus striatum, which is clearly not homologue of detected. For instance, the corpus striatum (Fig. 3) arises either the isocortex or the ADVR [16]. Therefore, only from the embryonic lateral and medial ganglionic emi- some sets of connections are similar between the lateral nences, which are located in the lateral subpallium and isocortex and the ADVR; and these that are similar (inputs express the marker genes Dlx1 and Dlx2 [11]. The from collicular pathways) are reported to end mainly in the cerebral cortex arises mostly from the embryonic pallium subpallium of amphibians, which is not homologue of and is characterized by the expression of genes of the Emx, either the ADVR or the isocortex. Otx and Pax families [8,51,64,65,86]. Smith Fernandez´ et Summarizing, the hypothesis of homology between the al. [87] confirmed these findings for amphibians, reptiles lateral isocortex and the DVR (recapitulation hypothesis) and birds, showing that the amphibian dorsal pallium, the assumes that the sensory pathways associated to midbrain reptilian dorsal cortex and the avian Wulst express the colliculi end in the same targets in reptiles and mammals, same pallial markers as the isocortex. These authors also while the gross morphology of these targets (isocortex and identified for the first time an intermediate territory (IT; DVR) has diverged in these two taxa. On the other hand, see Fig. 3) in the equatorial region of the hemisphere, the alternative possibility of non-homology between iso- located according to them between the pallium and the cortex and DVR (outgroup hypothesis) implies that collicu- subpallium of amphibians, reptiles, birds and mammals, lar pathways have changed their targets in the course of which does not express either the Emx1 or Dlx1 markers reptilian and mammalian divergence. If this hypothesis is but is largely positive for the gene Pax6 [87]. More recent correct, collothalamic projections to the dorsal pallium reports [66,67] have confirmed the existence of the IT 146 F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153

Fig. 3. The cerebral hemispheres of reptiles, birds and mammals (only one hemisphere is shown; medial is to the left), indicating the pallium (dark grey), which during development expresses the marker genes Emx1/2, Otx1/2, Pax6 and Tbr1, and gives rise to the Wulst (W) and hyperstriatum ventrale (HV) of birds, and to cortical structures in reptiles and mammals. Light grey indicates the intermediate territory or ventral pallium (IT/VP), which is largely negative for the genes Emx1/2 and Otx1/2, but positive for Pax6 and Tbr1. The IT/VP gives rise to the anterior dorsal ventricular ridge (ADVR) of reptiles, to the neostriatum (N) of birds (which corresponds to a large part of the ADVR), and to the laterobasal amygdala and ventral claustrum (AM) of mammals. The subpallium (white) expresses Dlx-type genes during embryogenesis and gives rise to the corpus striatum (STR) among other structures. In each vertebrate class, the projection sites of the auditory pathway (a), and the collicular (vc) and lemniscal (vl) visual pathways are indicated. DCx, dorsal cortex; ICx, isocortex, OCx, olfactory cortex. Based on Smith Fernandez´ et al. [87] and Puelles et al. [66].

(which has been named ventral pallium, VP, by these zones, disappearing from the neuroepithelial surface [87]. authors), extending its definition by showing that, like Somewhat in agreement with this, Swanson [94] has other pallial regions, the IT/VP also expresses the regula- proposed that the claustral complex of mammals (includ- tory marker gene Tbr1. ing the basolateral amygdala, the claustrum, the endo- Parts of the olfactory cortex, the olfactory bulbs, the piriform nucleus and isocortical layers 6b-7) is an early- basolateral amygdalar complex, and the ventral claustrum produced component, embryologically related to the sub- (among other regions) of mammals derive from the IT/VP plate or a subplate-like structure (see also Ref. [45]). If [67,87]. On the other hand, part of the lateral cortex, the these interpretations are true, most late-produced com- olfactory bulb and an important part of the reptilian ADVR ponents of the ADVR might not have a strict homologue in and of the avian neostriatum develop from the IT/VP (Fig. the mammalian isocortex. 3). This molecular evidence has been considered to This evidence poses a serious challenge to the recapitu- strongly support the early suggestions of non-homology lation hypothesis of homology between the ventrolateral between the isocortex and the ADVR (outgroup hypoth- isocortex and the ADVR [41–43], since it requires that in esis) [1,2,4], and proposals of homology between the mammals the IT/VP contributes cells to isocortical de- reptilian ADVR and parts of the mammalian claus- velopment. However, the topographic location of the IT/ troamygdalar complex [14,34,90]. It is of interest to note VP is such that the lateral (olfactory) cortex is interposed that Holmgren [34] originally named the sauropsidian between it and the isocortex. Thus, a massive tangential DVR and the mammalian claustroamygdaloid regions as migration of neurons from the IT/VP to the dorsal the ‘hypopallium’, which definitely agrees with the molec- pallium, producing the visual extrastriate and the auditory ular findings. cortices would be needed if the recapitulation hypothesis is Smith Fernandez´ et al. [87] argued that in reptiles and correct. The evidence that many isocortical GABAergic birds, the IT/VP remains as a distinct neuroepithelial zone cells originate in the subpallial corpus striatum and migrate until late development, period in which it gives rise to dorsally into the isocortex [10], raises the possibility that most of the ADVR. On the contrary, in mammals this some cells from the IT/VP also migrate to the dorsal territory was described as producing only the above pallium. Just like cells from the ganglionic eminences that mentioned early-generated components. In later embryonic migrate tangentially into the isocortex keep expressing stages, the mammalian IT/VP is considered to be obliter- their Dlx1 subpallial marker [10,12], if cells from the ated between the Emx1-positive and the Dlx1/2-positive IT/VP migrate tangentially into the isocortex they might F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 147 maintain their molecular identity. This would make the 5 .2. Dorsoventral gradients and their relation to the IT/ auditory and visual extrastriate cortices largely negative for VP Emx1. However, recent studies indicate that Emx1 can be considered as a marker of practically all pyramidal cells in In this context, it is important to note that the regulation the cerebral cortex, and of most glutamate-containing of distinct transcription factors may produce overall effects neurons in dissociated cortical cultures [19], something in the dorsoventral patterning of the hemispheres. For that would be unexpected if the ventrolateral isocortex example, in Nkx2.1 loss of function mutants, a dorsaliza- (including auditory and visual extrastriate cortices) were tion of the basal ganglia is observed, in which the striatum largely negative to this gene. Up to this point, no evidence is enlarged at the expense of the development of the more has been reported of a massive migration from the IT/VP ventral pallidum [93]. On the other hand, in the Pax6 into the developing isocortex. mutant, a ventralization of the developing basal ganglia is observed in which the medial ganglionic eminence ex- 5 .1. Other components of the DVR pands into the territory of the more dorsally located lateral ganglionic eminence, which results in underdevelopment The DVR is quite a complex structure which has many of the striatum [89]. In this mutant, there is also a dorsal intriguing components. We will discuss two of these, the displacement of the limits of expression for marker genes hyperstriatum ventrale of birds, and the posterior DVR, such as Emx1, Tbr1, Shh, Dlx1, Nkx2.1 and others, which has been sometimes termed the ‘reptilian producing malformations in the claustrum, endopiriform amygdala’. The avian hyperstriatum ventrale (a structure nucleus, insular cortex and piriform/lateral cortex. Finally, located ventral to the Wulst, which has been considered to the isocortex is also severely disrupted, presumably due to belong to the avian DVR), expresses Emx1 during de- failure of differentiation and migration of cortical velopment (Fig. 3), suggesting that it may derive from the progenitors, especially late-produced neurons in the ven- dorsal or lateral pallium and not from the IT/VP [66,87]. trolateral pallium [28,89]. Another observation is that the In subsequent articles, Puelles et al. [67] and Guirado et al. Pax6 2/2 mutant develops a dorsal ventricular ridge-like [32] subdivide the reptilian ADVR into dorsolateral and structure in the lateral pallium [18], which according to the ventromedial moieties (Fig. 2). These components would authors may consist of cells that failed to migrate to the correspond to the hyperstriatum (which is Emx1 positive) lateral isocortex and other regions. However, this DVR- and the neostriatum (Emx1 negative) of birds, respectively. like structure may derive of subpallial components and of These authors further argue that the hyperstriatum/dorsola- components originating in the dorsal, lateral and ventral teral component of the avian/reptilian ADVR (Emx1 pallium (all positive to Pax6 and dependent on this gene positive) might correspond to the mammalian dorsolateral for migration). If this is the case, it may not be strictly claustrum, the dorsal endopiriform nucleus and the ba- comparable to the sauropsidian DVR, which originates somedial amygdala (and perhaps some elements of the mostly from the ventral pallium. lateral isocortex). On the other hand, the neostriatum/ The above evidence may suggest that in the evolution of ventromedial component of the ADVR (Emx1 negative) the vertebrate pallium, expansion of the domains of corresponds to the ventromedial claustrum, ventral endo- expression of regulatory genes, produced by overall dor- piriform nucleus and laterobasal amygdala of mammals. salizing or ventralizing factors, may have had a key One additional issue concerns the homologies between influence in the origin of new brain architectures such as parts of the mammalian amygdalar complex and the that of mammals. For example, it is possible that in the reptilian PDVR/avian archistriatum. Some authors [54,95] origin of the isocortex, the IT/VP was dwarfed as a divide the amygdala into pallial (cortical and basolateral consequence of some dorsalizing influence that enhanced amygdala) and subpallial (central and medial amygdala) Emx1 expression and was concomitant with the emergence moieties. According to Smith Fernandez´ et al. [87], the and expansion of the isocortex. Nevertheless, only a reptilian PDVR and the avian archistriatum express pallial change in boundaries may not be sufficient for the expan- markers and are comparable to the corticomedial and sion of the mammalian isocortex. Substantial increases in central amygdala of mammals, while Puelles et al. [66] cell proliferation within the territory destined to the dorsal argue that only the posterior archistriatum is pallial. An pallium were probably required to produce the enormous additional interpretation, based on hodological studies, is growth of this structure in early and late mammalian that the reptilian PDVR is homologue to the mammalian evolution. laterobasal amygdala [46]. Dubbeldam [23] claims that the In an argumentation in favor of homology between the avian archistriatum consists of a sensorimotor moiety that mammalian isocortex and the reptilian ADVR, Reiner [76] receives projections from the ADVR, and an ‘amygdalar’ has proposed that the structure ancestral to both ADVR moiety. Perhaps further embryological studies will help and ventrolateral isocortex was either Emx1 negative and clarify the compartmentalization and homologies of the acquired Emx1 expression in the origin of mammals, or it mammalian amygdala and the reptilian PDVR/avian ar- was Emx1 positive and lost Emx1 expression in the chistriatum. evolution of reptiles and birds. In our view, shifts in the 148 F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 boundaries of distinct telencephalic territories could cer- sory information: one receiving mainly olfactory and tainly happen, with the result that cell groups that were lemnothalamic information, which is connected with the initially destined to one compartment become incorporated hippocampus, and the other, receiving mainly col- into another, without need of tangential migration of its lothalamic information (and some lemnothalamic infor- cellular constituents. However, in most reptiles the topog- mation), which is more related to the amygdalar system raphic relation between the IT/VP and the isocortex is (see Fig. 2). such that the lateral cortex is interposed between them In many mammals, the primary visual cortex (with [1,2,66,67], so that a mere shift in gene expression lemnothalamic input) projects heavily into extrastriate boundaries may perhaps not be sufficient to transform the areas (especially V2), and these to higher-order visual IT/VP into an isocortical area. The only exception to this cortices. Since extrastriate and higher-order visual areas situation may be the turtle dorsal cortex, which develops a also receive collicular input (via the pulvinar nucleus), structure called pallial thickening (Fig. 1), located deep to there is an important degree of convergence between the olfactory cortex, and topographically between the lemniscal and collicular pathways (Fig. 2). Both pathways dorsal cortex and the ADVR. Although definitely not an eventually route their information to the hippocampus and IT/VP derivative [87], the pallial thickening deserves amygdala. Although there are some species differences in further embryological studies. It is possible that the pallial hippocampal connectivity, most species show similar thickening is a derived character of turtles, which may be overall patterns of connections [99]. In the rat and in the supported by recent phylogenetic analyses indicating that monkey, nearly all areas in the neocortical mantle (visual, turtles are a rather modified group of diapsid reptiles, and auditory and somatosensory) are bidirectionally connected unrelated to ancestral reptiles [52,77,106]. to the subiculum and entorhinal cortex, which route information into the hippocampus [9,15,48,63,80,102,103]. Likewise, in most mammals studied, visual, auditory and 6 . Different patterns of brain organization in reptiles somatosensory information is transmitted to the mam- and mammals malian amygdala by a series of modality-specific cortico- cortical pathways [55]. These two structures, the hip- As opposed to sauropsids, in which lemniscal and pocampus and the amygdala, process different types of collicular sensory pathways are largely (but not totally) mnemonic information (spatial and emotional, respective- separated in the dorsal cortex and the ADVR, in the ly) [50,53]. Thus, the mammalian hippocampus may mammalian isocortex there is a strong confluence of the receive a much heavier sensory projection than is the case lemniscal and collicular pathways, especially in the case of in reptiles and birds, who may rely more on amygdalar the visual pathways. This suggests different patterns of components (PDVR/archistriatum) than on the former to brain organization in mammals and in reptiles/birds (Fig. process certain types of mnemonic information. For exam- 2). ple, although the hippocampus of sauropsids is known to In reptiles and birds, the collicular pathways are appar- be involved in spatial memory as it is in mammals [78], in ently more important to perception and anatomically more birds not all forms of spatial memory depend on the robust than the lemniscal projections [97,98]. In reptiles, hippocampus. In homing pigeons, hippocampal lesions the dorsal cortex (receiving lemnothalamic projections) is disrupt certain types of spatial learning such as using the strongly connected with the medial/dorsomedial (hip- sun compass directional information, while the capacity to pocampal) and the lateral (olfactory) cortices. The ADVR, learn on the basis of landmark beacons remains spared which receives most collothalamic projections, is con- [29]. nected primarily to the PDVR (comparable to parts of the mammalian amygdalar complex). In turtles and lizards, projections from the dorsal cortex and medial cortex into 7 . A scenario for isocortical origins: olfaction, the the ventromedial (Emx1 negative) ADVR are sparse, but hippocampus and the thalamofugal visual system the pallial thickening (related to the dorsal cortex) and the rostrolateral (somatosensory) dorsal cortex are reciprocally The reviewed evidence provides important insights into connected to the dorsolateral ADVR (Emx1 positive) [32]. the developmental mechanisms leading to the evolutionary Therefore, some integration or convergence of lem- emergence of the isocortex. However, an equally important nothalamic and collothalamic pathways may occur in the question (but more difficult to address) concerns the dorsolateral ADVR (Fig. 2). Both the ventromedial ADVR functional context in which these transformations took and the dorsolateral ADVR project to the PDVR, in which place. In other words, we might ask what are the be- convergence of different sensory modalities may occur, as havioral correlates of the expansion of the dorsal pallium it also happens in the mammalian amygdala. The PDVR and the lemnothalamic visual pathway in early mammals, projects to hypothalamic nuclei, perhaps modulating social and of the IT/VP and the collothalamic visual pathway in and aggressive behavior and visceral responses [46]. sauropsids. We postulate that the divergent evolution of the Summarizing, in the telencephalon of reptiles and birds brains of reptiles and mammals was related to emphasis on two relatively separate systems exist for processing sen- different strategies for processing sensory information. F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 149

In an attempt to shed some light on this problem, we cues (like odors) may be used in the elaboration of spatial suggest below that the mammalian isocortex originated in maps in the mammalian and reptilian hippocampi, and the dorsal pallium by virtue of the relations of the dorsal possibly in mammalian ancestors. Furthermore, more than cortex with the medial/dorsomedial cortex (hippocampus) being strictly involved in spatial memory, the hippocampus and the olfactory cortex (Fig. 2), and their participation in has been recently considered to be crucial for storing hippocampal episodic and spatial memory [1]. Briefly, in episodic memory, that is, the remembrance of behaviorally early mammals spatial memory may have strongly de- relevant events and sequences of events, from which pended on nonspatial clues like olfaction to develop maps spatial maps and other forms of memory emerge [27]. of space. For example, odors may permit to recognize Therefore, these olfactive–hippocampal networks may specific places and routes made by the animal in its daily have participated in rather broad aspects of memory routine. Thus, the circuit connecting olfactory cortex and formation beside purely spatial memory. hippocampus may have been quite important for the We postulate that the lateral, dorsal, and medial cortices behavior of early mammals. With the colonization of were put to use by the first mammals to make relatively diurnal niches after the demise of dinosaurs, the visual elaborate, largely olfactory-based representations of space system perhaps became also important for spatial memory. and other behaviorally relevant items, in which specific In this condition, the dorsal cortex of early mammals, odors labeled particular objects, places and routes. As- originally receiving visual and somatosensory lem- sociative networks between the dorsal cortex and the nothalamic input, may have become progressively in- olfactory system, via the hippocampus, may have become volved in associative networks for hippocampal memory. increasingly important in these animals to develop mul- This may have triggered the expansion of the dorsal cortex tisensorial maps of space. Eventually, the contribution of into an isocortex, which also begun to receive collicular the visual system became undoubtedly necessary in the visual and auditory information to participate in associative elaboration of more precise maps of space. The dorsal networks with other sensory modalities. cortex, receiving visual information from the thalamofugal This proposal is partly based on Lynch’s [50] original visual pathway, may have thus become an important hypothesis of the origin of the isocortex, and assumes that sensory processing system in the early mammalian brain the development of olfaction was a key event in early [1]. Other parts of the isocortex (visual extrastriate and mammalian evolution. It has been repeatedly proposed that auditory) may have originated from an expansion of the in ancestral mammals, olfaction was an important sensory dorsal cortex to accommodate the incoming mesencephalic modality [37,38,44]. Endocasts of mesozoic mammals inputs that became increasingly involved in associative indicate relatively large olfactory bulbs and perhaps an interactions with the primary visual cortex, and with the elevated rhinal fissure [38], suggesting a large olfactory olfactory system through the hippocampus. cortex in relation to the rest of the pallium. Likewise, in The early involvement of the primary visual cortex in small-brained insectivores and in some marsupials, olfac- processing spatial information implies that it developed at tory-related structures occupy a much larger proportion of least a crude spatial representation of the visual field. The the volume of the brain than is the case in larger-brained visual cortex of reptiles has a rather poor visuotopic species with a well-developed isocortex [25,26,88,101]. In organization, due to its tangential synaptic organization the subsequent evolution of early and late mammals, [1,3,4,96,98]. In these animals, visual topographic infor- olfactory and limbic (including hippocampal) structures mation is mostly processed in the optic tectum. The scale together but in large part separately of the isocortex development of a columnar organization of the isocortex, [88], perhaps indicating that the latter has become largely associated with the increase in cellular depth and the independent of limbic components for certain types of acquisition of an inside-out neurogenetic gradient during sensory and motor processing. development allowed the establishment of more sophisti- The olfactory cortex is heavily connected with the cated retinotopic maps in the primary visual cortex [1,3,5– hippocampal cortex in both reptiles and mammals. In 7]. This permitted the latter to drive the activity of other reptiles, there is a circuit interconnecting the medial cortical areas that begun to receive visuotopically orga- (hippocampal) cortex, the dorsal cortex (receiving the nized tectofugal projections. In addition, the auditory lemnothalamic pathways) and the olfactory cortex [50,96]. projection to the cerebral cortex may have benefited from Furthermore, there is evidence that the medial and dorsal the cortical representation of space through the elaboration cortices of reptiles participate in spatial navigation, and of a more sophisticated sound localization system (see Ref. these structures have been recently found to use nonspatial [1], and below). clues in tasks of spatial orientation [22]. Although it has long been considered that the mammalian hippocampus 7 .1. Fossil brains encodes mainly visual spatial memory [13,72], recent findings indicate that it also represents olfactory infor- Endocast information indicates that early mammal-like mation [24,59,104], and that there is an interleaved segre- reptiles (therapsids) had narrow, tubular hemispheres with gation of spatial and nonspatial information along the no signs of telencephalic expansion [35,44,69]. Increase in length of this structure [33]. This suggests that nonspatial brain size, resulting from a generalized growth of the 150 F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153

Fig. 4. Endocasts of mammal-like reptiles and primitive mammals, indicating the progressive increase in brain size. Note that in Morganucodon, expansion of the posterior part of the hemisphere can be observed. More advanced mammals like Triconodon show a more complete expansion of the hemispheres. Based on Rowe [81,82]. Dates in millions of years ago: Probaignognathus, about 235 MYA; Therioherpeton, about 215 MYA; Morganucodon, about 205 MYA; Triconodon, about 155 MYA; Didelphys, about 65 MYA to present. isocortex, occurs in more recent fossil mammals like DVR. The latter appears to be related to ventral pallial Hadrocodium and Triconodon [49,81,82] (Fig. 4). The full structures of mammals such as the basolateral amygdala enlargement of the brain coincides with the detachment of and/or the endopiriform nucleus (ventral claustrum), while the auditory bones from the mandible to form the mam- the isocortex originates largely from the dorsal pallium. malian middle ear [49,81,82], and perhaps with the de- Connectional evidence indicating similarity of sensory velopment of auditory projections into the isocortex. In input in the reptilian ADVR and the mammalian auditory this sense, expansion of the brain, which may be attributed and extrastriate isocortex is weakened by other hodological to cortical growth and to invasion of the collicular sensory evidence that suggests different interpretations. One im- pathways into the isocortex, was related to the develop- portant assumption held by several workers has been that ment of higher auditory acuity, perhaps in relation to more the modern reptilian brain represents a stage somewhat sophisticated spatial representation of sound sources. Inter- comparable to that of the ancestral mammals. On the other estingly, Morganucodon, a primitive mammaliaform from hand, the developmental evidence suggests to us that the the early Jurassic, and taxonomically intermediate between reptilian and mammalian brains may have diverged very the above forms and smaller-brained, more primitive early in their organization. It is possible that the origin of therapsids, shows only partial expansion of the brain. In the mammalian isocortex was partly a developmental this species, widening of the occipital parts of the hemi- consequence of a generalized dorsalizing tendency which spheres can be observed [81,82] (Fig. 4), perhaps evidenc- led to the expansion of the dorsal cortex, while the ing early expansion of the dorsal cortex. evolution of the sauropsidian pallium may have been dominated by the action of some ventralizing factor, leading to the expansion of the IT/VP. 8 . Final comment The above proposals have been complemented with a scenario describing the origin of the isocortex from the We have reviewed developmental evidence for a sepa- dorsal pallium based on the relations of this structure and rate origin of the mammalian isocortex and the reptilian the hippocampus and olfactory cortex. In our view, the F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 151 main signiﬁcance of these scenarios is that they provide an Dlx-1 and Dlx-2 disrupt the striatal subventricular zone and dif- evolutionary framework which may guide future studies of ferentiation of late born striatal neurons, Neuron 19 (1997) 27–37. [12] S. Anderson, M. Mione, K. Yun, J.L.R. Rubenstein, Differential the embryology and structure of the cerebral cortex. origins of neocortical projection and local circuit neurons: role of Furthermore, it is important to stress the point that the Dlx genes in neocortical interneuronogenesis, Cereb. Cortex 9 developmental and functional approaches to evolutionary (1999) 646–654. problems are complementary rather than alternative; these [13] P.J. Best, A.M. White, Hippocampal cellular activity: a brief history of space, Proc. Natl. Acad. Sci. USA 95 (1998) 2717–2719. address different but equally important questions. Finally, [14] L.L. Bruce, T.J. Neary, The limbic system of tetrapods: a compara- a similar scenario for the evolutionary development of the tive analysis of cortical and amygdalar populations, Brain Behav. reptilian DVR is still needed. In sauropsids, mesencephalic Evol. 46 (1995) 224–234. sensory pathways remain as the main processors of topog- [15] G. Buzsaki,´ The hippocampo-neocortical dialogue, Cereb. Cortex 6 raphically organized exteroceptive information, as it may (1996) 81–92. [16] A.B. Butler, The evolution of the dorsal pallium in the telencephalon be in lower vertebrates. Thus, the advent in early mammals of amniotes: cladistic analysis and a new hypothesis, Brain. Res. of an olfactory-based brain and as a consequence, of a Brain Res. Rev. 19 (1994) 66–101. dominant, topographically organized visual lemnothalamic [17] A.B. Butler, The evolution of the dorsal thalamus of jawed projection, may be considered a fundamental innovation in vertebrates, including mammals: cladistic analysis and a new hypothesis, Brain. Res. Rev. 19 (1994) 29–65. vertebrate brain evolution, which drove the collicular [18] A.B. Butler, Z. Molnar,´ Development and evolution of the collopal- pathways out of their ancestral route to the telencephalon. liumin amniotes: a new hypothesis of ﬁeld homology, Brain Res. Bull. 57 (2002) 475–479. [19] C.H. Chan, L.N. Godinho, D. Thomaidou, S.S. Tan, M. Gulisano, J.G. Parnavelas, Emx1 is a marker for pyramidal neurons in the A cknowledgements cerebral cortex, Cereb. Cortex 11 (2001) 1191–1198. [20] I. Cobos, L. Puelles, S. Martınez,´ The avian telencephalic subpallium originates inhibitory neurons that invade tangentially the Supported by FONDECYT Grant 1970294 and by a gift pallium (dorsal ventricular ridge and cortical areas), Dev. Biol. 239 from EMEC S.A. and ENAEX S.A. to Francisco Aboitiz. (2001) 30–45. We wish to acknowledge Alexander Vargas for interesting [21] J.C. Davila,´ S. Guirado, L. Puelles, Expression of calcium-binding discussions. Claudia Andrade provided secretarial help. proteins in the diencephalon of the lizard Psammodromus algirus,J. Comp. Neurol. 427 (2000) 67–92. [22] L.B. Day, D. Crews, W. Wilczynski, Effects of medial and dorsal cortex lesions on spatial memory in lizards, Behav. Brain Res. 118 R eferences (2001) 27–42. [23] J.L. Dubbeldam, Birds, in: R. Nieuwenhuys, H.J. Ten Donkelaar, C. Nicholson (Eds.), The Central Nervous System of Vertebrates, Vol. [1] F. Aboitiz, The evolutionary origin of the mammalian cerebral 3, Springer, Berlin, 1998, pp. 1525–1636. cortex, Biol. Res. 25 (1992) 41–49. [24] J.A. Dusek, H. Eichenbaum, The hippocampus and transverse [2] F. Aboitiz, Homology in the evolution of the cerebral hemispheres: patterning guided by olfactory cues, Behav. Neurosci. 112 (1998) the case of reptilian dorsal ventricular ridge and its possible 762–771. correspondence with mammalian neocortex, J. Hirnforsch. 4 (1995) [25] B.L. Finlay, R.B. Darlington, Linked regularities in the development 461–472. and evolution of mammalian brains, Science 268 (1995) 1578– [3] F. Aboitiz, Evolution of isocortical organization. A tentative 1584. scenario including roles of reelin, p35/cdk5 and the subplate zone, [26] B. L Finlay, M.N. Hersman, R.B. Darlington, Patterns of vertebrate Cereb. Cortex 9 (1999) 655–661. neurogenesis and the paths of vertebrate evolution, Brain Behav. [4] F. Aboitiz, Comparative development of the mammalian isocortex Evol. 52 (1998) 232–242. and the reptilian dorsal ventricular ridge. Evolutionary considera- [27] N.J. Fortin, K.L. Agster, H.B. Eichenbaum, Critical role of the tions, Cereb.Cortex 9 (1999) 783–791. hippocampus in memory for sequences of events, Nat. Neurosci. 5 [5] F. Aboitiz, The origin of isocortical development, Trends Neurosci. (2002) 458–462. 24 (2001) 202–203. [28] T. Fukuda, H. Kawano, N. Osumi, K. Eto, K. Kawamura, His- [6] F. Aboitiz, D. Morales, J. Montiel, The inverted neurogenetic togenesis of the cerebral cortex in rat fetuses with a mutation in the gradient of the mammalian isocortex. Development and evolution, Pax-6 gene, Dev. Brain Res. 120 (2000) 65–75. Brain Res. Rev. 38 (2001) 129–139. [29] A. Gagliardo, M. Mazzotto, V.P. Bingman, Hippocampal lesion [7] F. Aboitiz, J. Montiel, J. Lopez,´ An hypothesis of the early effects on learning strategies in homing pigeons, Proc. R. Soc. evolution of the development of the isocortex, Brain Res. Bull. 57 London B 263 (1996) 529–534. (2002) 481–483. [30] W. Garstang, The theory of recapitulation: a critical re-statement of [8] D. Acampora,´ A. Simeone, Understanding the roles of Otx1 and the biogenetic law, J. Linn. Soc. (Zool.) London 35 (1922) 81–101. Otx2 in the control of brain morphogenesis, Trends Neurosci. 22 [31] G. Gellon, W. McGinnis, Shaping animal body plans in development (1999) 116–122. and evolution by modulation of Hox expression patterns, BioEssays [9] J.P. Aggleton, S.D. Vann, C.J. Oswald, M. Good, Identifying cortical 20 (1998) 116–125. inputs to the rat hippocampus that subserve allocentric spatial [32] S. Guirado, J.C. Davila,´ M.A. Real, L. Medina, Light and electron processes: a simple problem with a complex answer, Hippocampus microscopic evidence for projections from the thalamic nucleus 10 (2000) 466–474. rotundus to targets in the basal ganglia, the dorsal ventricular ridge, [10] S.A. Anderson, D.D. Eisenstat, L. Shi, J.L.R. Rubenstein, Inter- and the amygdaloid complex in a lizard, J. Comp. Neurol. 424 neuron migration from basal forebrain to neocortex: dependence on (2000) 216–232. Dlx genes, Science 278 (1997) 474–476. [33] R. E Hampson, J.D. Simeral, S.A. Deadwyler, Distribution of spatial [11] S.A. Anderson, M. Qiu, A. Bulfone, D.D. Eisenstat, J. Meneses, R. and nonspatial information in dorsal hippocampus, Nature 402 Pedersen, J.L.R. Rubenstein, Mutations of the homeobox genes (1999) 610–614. 152 F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153

[34] N. Holmgren, Points of view concerning forebrain morphology in striate cortex and extrastriate visual areas in the rat, Exp. Brain Res. higher vertebrates, Acta Zool. 6 (1925) 413–477. 94 (1993) 1–15. [35] J.A. Hopson, Paleoneurology, in: C.C. Gans, R.G. Northcutt, P.S. [59] C.E. Myers, M.A. Gluck, Cortico-hippocampal representations in Ulinski (Eds.), Biology of the Reptilia, Vol. 4, Academic Press, New simultaneous odor discrimination: a computational interpretation of York, 1979, pp. 39–146. Eichenbaum, Mathews and Cohen (1989), Behav. Neurosci. 110 [36] R. Insausti, Comparative anatomy of the entorhinal cortex and (1996) 685–706. hippocampus in mammals, Hippocampus 3 (1993) 19–26. [60] W.H.J. Nauta, H.J. Karten, A general profile of the vertebrate brain, [37] H.J. Jerison, in: Evolution of the Brain and Intelligence, Academic with sidelights on the ancestry of the cerebral cortex, in: F.O. Press, New York, 1973, 482 pp. Schmitt (Ed.), The Neurosciences, Second Study Program, Roc- [38] H.J. Jerison, Fossil evidence on the evolution of the neocortex, in: kefeller University Press, New York, USA, 1970, pp. 7–26. E.G. Jones, A. Peters (Eds.), Cereb. Cortex, Vol. 8A, Plenum, New [61] R.G. Northcutt, J.H. Kaas, The emergence and evolution of mam- York, 1990, pp. 285–309. malian neocortex, Trends Neurosci. 18 (1995) 373–379. [39] D. Jimenez,´´ L.M. Lopez-Mascaraque, F. Valverde, J.A. De Carlos, [62] R.G. Northcutt, The origin of craniates: neural crest, neurogenic Tangential migration in neocortical development, Dev. Biol. 244 placodes, and homeobox genes, Israel J. Zool. (suppl.) 42 (1996) 273–313. (2002) 155–169. [63] S.M. O’Mara, S. Commins, M. Anderson, J. Gigg, The subiculum: a [40] B. Kallen,¨´ On the ontogeny of the reptilian forebrain. Nuclear review of form, physiology and function, Prog. Neurobiol. 64 structures and ventricular sulci, J. Comp. Neurol. 95 (1951) 307– (2001) 129–155. 347. [64] M. Pannese, G. Lupo, B. Kablar, E. Boncinelli, G. Barsacchi, R. [41] H.J. Karten, The ascending auditory pathway in the pigeon Vignali, The Xenopus Emx genes identify presumptive dorsal (Columba livia). II. Telencephalic projections of the nucleus telencephalon and are induced by head organizer signals, Mech. ovoidalis thalami, Brain Res. 11 (1968) 134–153. Dev. 73 (1998) 73–83. [42] H.J. Karten, The organization of the avian telencephalon and some [65] L. Puelles, J.L.R. Rubenstein, Expression patterns of homeobox and speculations on the phylogeny of the amniote telencephalon, Ann. other putative regulatory genes in the embryonic mouse forebrain NY Acad. Sci. 167 (1969) 164–179. suggests neuromeric organization, Trends Neurosci. 16 (1993) 472– [43] H.J. Karten, Evolutionary developmental biology meets the brain: 479. the origins of mammalian neocortex, Proc. Natl. Acad. Sci. USA 94 [66] L. Puelles, E. Kuwana, E. Puelles, J.L.R. Rubenstein, Comparison of (1997) 2800–2804. the mammalian and avian telencephalon from the perspective of [44] T.S. Kemp, in: Mammal-Like Reptiles and the Origin of Mammals, gene expression data, Eur. J. Morphol. 37 (1999) 139–150. Academic Press, 1982, 363 pp. [67] L. Puelles, E. Kuwana, E. Puelles, A. Bulfone, K. Shimamura, J. [45] H. Kunzle, S. Raddtke-Schuller, Cortical connections of the claus- Keleher, S. Smiga, J.L.R. Rubenstein, Pallial and subpallial deriva- trum and subjacent cell groups in the hedgehog tenrec, Anat. tives in the embryonic chick and mouse telencephalon, traced by the Embryol. 203 (2001) 403–415. expression of the genes Dlx-2, Emx-1, Nkx-2.1, Pax-6 and Tbr-1, J. [46] E. Lanuza, M. Belekhova, A. Martınez-Marcos,´ C. Font, F. Mar- Comp. Neurol. 424 (2000) 409–438. tınez-Garcıa,´´ Identification of the reptilian basolateral amygdala:an [68] L. Puelles, Thoughts on the development, structure and evolution of anatomical investigation of the afferents to the posterior dorsal the mammalian and avian telencephalic pallium, Phil. Trans. R. Soc. ventricular ridge of the lizard Podarcis hispanica, Eur. J. Neurosci. London B Biol. Sci. 356 (2001) 1583–1598. 10 (1998) 3517–3534. [69] J. Quiroga, The brain of the mammal-like reptile Probainognathus [47] E. Lanuza, A. Martınez-Marcos,´´´ F. Martınez-Garcıa, What is the jenseni (Therapsida, Cynodontia), A correlative paleo-neurological amygdala? A comparative approach, Trends Neurosci. 22 (1999) approach to the neocortex at the reptile-mammal transition, J. 207. Hirnforsch. 21 (1980) 299–336. [48] P. Lavenex, D.G. Amaral, Hippocampal-neocortical interaction: a [70] P. Rakic, Specification of cerebral cortical areas, Science 241 (1988) hierarchy of associativity, Hippocampus 10 (2000) 420–430. 170–176. [49] Z.X. Luo, A.W. Crompton, A.L. Sun, A new mammalia form from [71] P. Rakic, A small step for the cell, a giant leap for mankind: a the early jurassic and evolution of mammalian characteristics, hypothesis of neocortical expansion during evolution, Trends Neuro- Science 292 (2001) 1535–1540. sci. 18 (1995) 383–388. [50] G. Lynch, in: Synapses, Circuits, and the Beginnings of Memory, [72] J.N.P. Rawlins, Neurobiology. A place for space and smells, Nature MIT Press, Cambridge, MA, 1986, 122 pp. 397 (1999) 561–563. [51] A. Mallamaci, R. Iannone, P. Briata, L. Pintonello, S. Mercurio, E. [73] C. Redies, M. Ast, S. Nakagawa, M. Takeichi, M. Martınez-de-la-´ Boncinelli, G. Corte, EMX2 protein in the developing mouse brain Torre, L. Puelles, Morphologic fate of diencephalic prosomeres and and olfactory area, Mech. Dev. 77 (1998) 165–172. their subdivisions revealed by mapping cadherin expression, J. [52] H. Mannen, S.S.L. Li, Molecular evidence for a clade of turtles, Comp. Neurol. 421 (2000) 481–514. Mol. Phylogenet. Evol. 13 (1999) 144–148. [74] A. Reiner, A comparison of neurotransmitter-specific and neuro- [53] S. Maren, Long-term potentiation in the amygdala: a mechanism for peptide-specific neuronal cell types present in the dorsal cortex of emotional learning and memory, Trends Neurosci. 22 (1999) 561– reptiles with those present in the isocortex of mammals, Brain. 566. Behav. Evol. 38 (1991) 53–91. [54] F. Martınez-Garcıa,´´ A. Martınez-Marcos, ´ E. Lanuza, The pallial [75] A. Reiner, Neurotransmitter organization and connections of turtle amygdala of amniote vertebrates: evolution of the concept, evolution cortex: implications for the evolution of mammalian isocortex, of the structure, Brain Res. Bull. 57 (2002) 463–469. Comp. Biochem. Physiol. Comp. Physiol. 104 (4) (1993) 735–748. [55] A.J. McDonald, Cortical pathways to the mammalian amygdala, [76] A. Reiner, A hypothesis as to the organization of cerebral cortex in Progr. Neurobiol. 55 (1998) 257–332. the common amniote ancestor of modern reptiles and mammals, [56] L. Medina, A. Reiner, Do birds possess homologues of mammalian Novartis Found. Symp. 228 (2000) 83–108. primary visual, somatosensory and motor cortices?, Trends Neuro- [77] O. Rieppel, R.R. Reisz, The origin and early evolution of turtles, sci. 23 (2000) 1–12. Annu. Rev. Ecol. Syst. 30 (1999) 1–22. [57] C.B. Moens, S.P. Cordes, M.W. Giorgianni, G.S. Barsh, C.B. [78] F. Rodrıguez,´´ J.C. Lopez, J.P. Vargas, Y. Gomez, ´ C. Broglio, C. Kimmel, Equivalence in the genetic control of hindbrain segmenta- Salas, Conservation of spatial memory function in the pallial tion in fish and mouse, Development 125 (1998) 381–391. forebrain of reptiles and ray-finned fishes, J. Neurosci. 22 (2002) [58] V.M. Montero, Retinotopy of cortical connections between the 2894–2903. F. Aboitiz et al. / Brain Research Reviews 39 (2002) 141–153 153

[79] M.G.P. Rosa, L.A. Krubitzer, The evolution of visual cortex: where [93] L. Sussel, O. Marın,´ S. Kimura, J.L. Rubenstein, Loss of Nkx 2.1 is V2?, Trends Neurosci. 22 (1999) 242–248. homeobox gene function results in a ventral to dorsal molecular [80] D.L. Rosene, G.W. Van Hoesen, The hippocampal formation of the respecification within the basal telencephalon: evidence for a primate brain, in: E.G. Jones, A. Peters (Eds.), Cerebral Cortex, Vol. transformation of the pallidum into striatum, Development 126 6, Plenum, New York, 1987, pp. 345–350. (1999) 3359–3370. [81] T. Rowe, Coevolution of the mammalian middle ear and neocortex, [94] L.W. Swanson, Cerebral hemisphere regulation for motivated be- Science 273 (1996) 651–654. havior, Brain Res. 886 (2000) 113–164. [82] T. Rowe, Brain heterochrony and origin of the mammalian middle [95] L.W. Swanson, G.D. Petrovich, What is the amygdala?, Trends ear, Mem. Calif. Acad. Sci. 20 (1996) 71–95. Neurosci. 21 (1998) 323–331. [83] E.S. Russel, in: Form and Function: A Contribution to the History of [96] H.J. Ten Donkelaar, Reptiles, in: R. Nieuwenhuys, H.J. Ten Animal Morphology, University of Chicago Press, Chicago, IL, Donkelaar, C. Nicholson (Eds.), The Central Nervous System of 1916, 383 pp.; Reprinted 1982. Vertebrates, Vol. 2, Springer, Berlin, 1998, pp. 1316–1524. [84] H.C. Seo, B.O. Sætre, B. Havik,û S. Ellingse, A. Fjose, The zebrafish [97] P.S. Ulinski, in: Dorsal Ventricular Ridge: A Treatise on Brain Pax3 and Pax7 homologues are highly conserved, encode multiple Organization in Reptiles and Birds, Wiley, New York, 1983, 284 pp. isoforms and show dynamic segment-like expression in the develop- [98] P.S. Ulinski, The cerebral cortex of reptiles, in: E.G. Jones, A. Peters ing brain, Mech. Dev. 70 (1998) 49–63. (Eds.), Cerebral Cortex, Vol. 8A, Plenum, New York, 1990, pp. [85] T. Shimizu, H.J. Karten, Multiple origins of neocortex: contributions 217–265. of the dorsal ventricular ridge, in: H.P. Zeigler, J.H. Bischof (Eds.), [99] T. Van Groen, I. Kadish, J.M. Wyss, Species differences in the Vision, Brain and Behavior in Birds, MIT Press, Cambridge, MA, projections from the entorhinal cortex to the hippocampus, Brain 1993, pp. 75–86. Res. Bull. 57 (2002) 553–556. [86] A. Simeone, M. Gulisano, D. Acampora, A. Stornaiuolo, M. [100] G.W. Van Hoesen, The parahippocampal gyrus. New observations Rambaldi, E. Boncinelli, Two vertebrate homeobox genes related to regarding its cortical connections in the monkey, Trends Neurosci. the Drosophila empty spiracles gene are expressed in the embryonic 10 (1982) 345–350. cerebral cortex, EMBO J. 11 (1992) 2541–2550. [101] J. Voogd, R. Nieuwenhuys, P.A.M. Van Dongen, Mammals, in: R. [87] A. Smith Fernandez,´´ C. Pieau, J. Reperant, E. Boncinelli, M. Nieuwenhuys, H.J. Ten Donkelaar, C. Nicholson (Eds.), The Wassef, Expression of the Emx-1 and Dlx-1 homeobox genes define Central Nervous System of Vertebrates, Vol. 3, Springer, Berlin, three molecularly distinct domains in the telencephalon of mouse, 1998, pp. 1638–2097. chick, turtle and frog embryos: implications for the evolution of [102] M.P. Witter, Organization of the entorhinal-hippocampal system: a telencephalic subdivisions in amniotes, Development 125 (1998) review of current anatomical data, Hippocampus 3 (1993) 33–44. 2099–2111. [103] M.P. Witter, P.A. Naber, T. Van Haeften, W.C. Machielsen, S.A. [88] H. Stephan, Evolutionary trends in limbic structures, Neurosci. Rombouts, F. Barkhof, P. Scheltens, F.H. Lopes da Silva, Cortico- Biobehav. Rev. 7 (1983) 367–374. hippocampal communication by way of parallel parahippocampal- [89] A. Stoykova, D. Treichel, M. Hallonet, P. Gruss, Pax 6 modulates subicular pathways, Hippocampus 10 (2000) 398–410. the dorsoventral patterning of the mammalian telencephalon, J. [104] E.R. Wood, P.A. Dudchenko, H. Eichenbaum, The global record of Neurosci. 20 (2000) 8042–8050. memory in hippocampal neuronal activity, Nature 397 (1999) [90] G.F. Striedter, The telencephalon of tetrapods in evolution, Brain 613–616. Behav. Evol. 49 (1997) 179–213. [105] M.S. Yoon, L. Puelles, C. Redies, Formation of cadherin-express- [91] G.F. Striedter, A. Marchant, S. Beydler, The ‘neostriatum’ develops ing brain nuclei in diencephalic and alar plate divisions, J. Comp. as part of the lateral pallium in birds, J. Neurosci. 18 (1998) Neurol. 421 (2000) 461–480. 5839–5849. [106] R. Zardoya, A. Meyer, The evolutionary position of turtles revised, [92] H. Super,` E. Soriano, H.B.M. Uylings, The functions of the preplate Naturwissenschaften 88 (2001) 193–200. in development and evolution of the neocortex and hippocampus, Brain Res. Brain Res. Rev. 27 (1998) 40–64.