Integrative and Comparative Biology

SYMPOSIUM Integrative and Comparative Biology Integrative and Comparative Biology, volume 55, number 6, pp. 949 961 doi:10.1093/icb/icv003 Society for Integrative and Comparative Biology Crocodilian Forebrain: Evolution and Development Michael B. Pritz 1 Molecular Neurosciences Department, Krasnow Institute for Advanced Study, George Mason University, 4400 University Drive, MS 2A1, Fairfax, VA 22030, USA From the symposium Integrated Biology of the Crocodilia presented at the annual meeting of the Society for Integrative and Comparative Biology, January 3 7, 2015 at West Palm Beach, Florida. 1 E-mail: michael.pritz@denlabs.com Synopsis Organization and development of the forebrain in crocodilians are reviewed. In juvenile Caiman crocodilus, the following features were examined: identification and classification of dorsal thalamic nuclei and their respective connections with the telencephalon, presence of local circuit neurons in the dorsal thalamic nuclei, telencephalic projections to the dorsal thalamus, and organization of the thalamic reticular nucleus. These results document many similarities between crocodilians and other reptiles and birds. While crocodilians, as well as other sauropsids, demonstrate several features of neural circuitry in common with mammals, certain striking differences in organization of the forebrain are present. These differences are the result of evolution. To explore a basis for these differences, embryos of Alligator misissippiensis were examined to address the following. First, very early development of the brain in Alligator is similar to that of other amniotes. Second, the developmental program for individual vesicles of the brain differs between the secondary prosencephalon, diencephalon, midbrain, and hindbrain in Alligator. This is likely to be the case for other amniotes. Third, initial development of the diencephalon in Alligator is similar to that in other amniotes. In Alligator, alar and basal parts likely follow a different developmental scheme. Introduction In the telencephalon of all adult amniotes, a layered structure of varying complexity is located above the lateral ventricle in both the transverse and sagittal plane and is termed the cortex. Similarly, the telencephalon of all amniote brains contains nonlaminated areas usually surrounded by borders of varying degrees of distinctiveness. These areas are commonly referred to as nuclei and are located internal to the lateral ventricle in all planes of section. In the telencephalon of mammals, most of these latter regions comprise the basal ganglia. However, in reptiles and birds, located between the cortex and basal ganglia and internal to the lateral ventricle, lies a nuclear area seemingly unique to sauropsids and known as the dorsal ventricle ridge (Ulinski 1983). In crocodilians, the dorsal ventricular ridge includes: the dorsolateral area, the intermediolateral area, and the nucleus of the lateral olfactory tract (Crosby 1917). Of these, experimental observations have been limited to the dorsolateral area. A representative transverse section of a crocodilian telencephalon is shown beside that of a comparable, transverse section of a mouse s brain to illustrate these differences (Fig. 1). A variety of sophisticated approaches have been used to unravel the organization of the forebrain (telencephalon and diencephalon) in amniotes. Despite considerable information on a variety of amniotes, comparisons between cortical areas in the telencephalon, as well as homologies between the dorsal ventricular ridge of sauropsids and the telencephalic regions of mammalian brains remain in dispute (Bruce 2007; Butler et al. 2011). Regardless of the ultimate interpretation, examination of crocodilian brains should prove key to understanding these relationships since crocodilians are the reptiles most closely related to birds (Walker 1972; Whetstone and Martin 1979; Hedges 1994). Experimental observations on the forebrain both of adult (juvenile Caiman crocodilus) and developing (embryonic Alligator mississippiensis) crocodilians are summarized. These findings demonstrate both differences and similarities in the organization of Advanced Access publication March 30, 2015 ß The Author 2015. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved. For permissions please email: journals.permissions@oup.com.

950 M. B. Pritz Fig. 1 Cytoarchitecture of telencephalons of Caiman and mice. Transverse Nissl-stained sections through the telencephalon of Caiman (A) and a mouse (B) illustrate similarities and differences. Note the presence of a nuclear area, the dorsal ventricular ridge (DVR), located internal to the lateral ventricle (marked by an asterisk, *) in Caiman (A) and the location of the basal ganglia in these two species. the forebrain between these reptiles and other amniotes. These data provide an overview of present knowledge and point to experimental approaches that should provide information to address some of the unanswered questions. Organization of the forebrain in C. crocodilus Dorsal thalamus: identification of nuclear groups and neural circuitry While a number of characters have been used to define the dorsal thalamus, the most common attribute, and the one most frequently used in mammalian studies, is the connections with the telencephalon (Jones 2007). Accordingly, a similar approach was employed in crocodilians using the following strategies. One series of experiments made large injections of a retrograde tracer into various regions of the telencephalon with the object of injecting all parts of this area. The goal was to identify all thalamic nuclei that projected to the telencephalon. These observations were supplemented by injections of anterograde tracers into specific dorsal thalamic areas. Additional experiments examined the neural circuitry of a variety of sensory systems from the periphery centrally to the telencephalon. These latter experiments focused on specific circuits and specific neuronal aggregates. Using these approaches, 11 dorsal thalamic nuclei that project to the telencephalon in Caiman have been identified (Pritz 2014). The neural circuitry of some of these nuclei is known; for others, many details remain incomplete. Nevertheless, these data suggest that dorsal thalamic nuclei in crocodilians can be grouped into several categories based on the following features: the telencephalic target, the fiberbundle connecting these forebrain structures, and whether thalamic projections to the telencephalon are ipsilateral or bilateral. Using this scheme, six categories of dorsal thalamic nuclei have been recognized based on the available data (Pritz 2014). These groups include the following subdivisions: (1) nuclei that project to the cortex bilaterally and utilize the medial forebrain bundle (dorsolateralis anterior), (2) nuclei that project ipsilaterally to the cortex (diagonalis), (3) nuclei that project to the ipsilateral primordial general cortex (dorsal geniculate), (4) nuclei that project both to the ipsilateral cortex and to the dorsal ventricular ridge (dorsomedialis anterior), (5) nuclei that project to the ipsilateral dorsal ventricular ridge and utilize the lateral forebrain bundle (rotundus, reuniens pars centralis, reuniens pars diffusa, medialis complex posterior, posterocentralis, and the area ventrolateralis), and (6) nuclei that project to the ipsilateral basal ganglia via the lateral forebrain bundle (medialis complex anterior). Of these nuclei mentioned above, only the medialis complex anterior likely has reciprocal connections with the telencephalon (Pritz unpublished observations) Reciprocal connections of other thalamic nuclei that project to the telencephalon in crocodilians have yet to be demonstrated. These observations are summarized in Table 1.

Crocodilian forebrain 951 Table 1 Summary of thalamo telencephalic connections in Caiman Thalamic nucleus Telen. Target Ipsi- Contra- FB Reciprocity Dorsolateralis anterior General cx Yes Yes MFB No Dorsomedialis anterior General cx & DVR Yes No?? Dorsal geniculate 1 0 General cx Yes No?? Diagonalis General cx Yes No?? Rotundus DVR Yes No LFB No Reuniens pars centralis DVR Yes No LFB No Reuniens pars diffusa DVR Yes No LFB No Medialis complex posterior DVR Yes No LFB No Area ventrolateralis DVR Yes No LFB? Posterocentralis DVR Yes No LFB? Medialis complex anterior Basal ganglia Yes No LFB Probably Abbreviations: contra-, contralateral; cx, cortex; DVR, dorsal ventricular ridge; FB, forebrain; ipsi, ipsilateral; LFB, lateral forebrain bundle; MFB, medial forebrain bundle; Telen., telencephalon;?, unknown; 1 0, primordial. Certain of the nuclei that comprise one of these categories (see group 5, above) have been studied in greater detail. This analysis revealed additional features shared by three of these nuclei: reuniens pars centralis, rotundus, and medialis complex posterior. These similar features include: neural circuitry and topographic projection to the telencephalon. Each of these nuclei utilizes the lateral forebrain bundle and projects to the ipsilateral dorsal ventricular ridge. Because these fibers pass through the ventrolateral area, axons from each of these thalamic nuclei are likely to synapse on intervening neurons of the basal ganglia although definitive proof would require ultrastructural confirmation. The topography of termination in the telencephalon, as well as the position of efferent axons in the lateral forebrain bundle, reflects the location of each respective nucleus in the dorsal thalamus (Fig. 2). Nucleus reuniens pars centralis, a caudo-medial nucleus, which, in Caiman is fused at the midline, projects to a caudo-medial portion of the dorsal ventricular ridge. Axons from this nucleus travel in a medial portion of the lateral forebrain bundle (Pritz 1974b, 1995). Nucleus rotundus, which is located anterior and lateral to the nucleus reuniens pars centralis, has efferents located in a lateral part of the lateral forebrain bundle that end in an antero-lateral portion of the dorsal ventricular ridge (Pritz 1975, 1995). The medialis complex posterior, which is located between nuclei reuniens pars centralis and rotundus, has fibers that travel in a central portion of the lateral forebrain bundle to terminate in a part of the dorsal ventricular ridge that lies between the projection zones of nucleus reuniens pars centralis and nucleus rotundus (Pritz and Stritzel 1994c; Pritz 1995). Summaries illustrating these fiber paths are available elsewhere (Fig. 9, Pritz [1975] for audition and vision and Fig. 16, Pritz and Stritzel [1994c] for somatosensation from the body surface). These neural circuits and their termination in the dorsal ventricular ridge are not unique to crocodilians but have been described in other reptiles and in birds (Nieuwenhuys et al. 1998; Butler and Hodos 2005; Bruce 2007). However, unlike reptiles and birds, similar circuits in mammals end in the cortex (Jones 2007) rather than in the dorsal ventricular ridge of nonmammalian amniotes, which is organized as a nucleus. Furthermore, each of these three nuclei shares additional similarities in their neural circuitry, beginning at the periphery (Table 2). In this simplified scheme, receptors for audition, vision, and somatosensation from the body surface are connected to bipolar cells, which, in turn, synapse on cells with long axons. Regardless of the modality, these long-axon cells project to the contralateral midbrain (Burns and Goodman 1967; Braford 1973; Repérant 1975; Pritz and Stritzel 1989; Derobert et al. 1999). From this third -order element, neurons terminate bilaterally in a dorsal thalamic target with the densest projection located ipsilaterally (Braford 1972; Pritz 1974a; Pritz and Stritzel 1990a). A similar grouping of neural circuits is also present in other reptiles as well as in birds (Nieuwenhuys et al. 1998; Butler and Hodos 2005; Bruce 2007). Other features shared by this dorsal thalamic category in Caiman are certain patterns of histochemical staining in the dorsal ventricular ridge (Pritz and Northcutt 1977).

952 M. B. Pritz Fig. 2 Thalamo telencephalic topography in Caiman. Horizontal sections projected onto a single-dimension of the diencephalon (A) and dorsal ventricular ridge (B) are shown. Note that the topography of similarly coded thalamic areas is preserved in the dorsal ventricular ridge. Because these areas are projected onto a single dimension in the dorsal ventricular ridge, areas of seeming overlap (open areas in B) are actually separate. Abbreviations: Dla, nucleus dorsolateralis anterior, MCp, medialis complex posterior; OT, optic tract; Rc, nucleus reuniens pars centralis; Rd, nucleus reuniens pars diffusa; Rt, nucleus rotundus; TRT, tectoreuniens tract; c, caudal; m, medial; IIIv, third ventricle. This figure was re-drawn from Fig. 3 in Pritz (1995). Comparable information on specific circuits in other categories of thalamo telencephalic projections in crocodilians is fragmentary. These gaps in knowledge point to areas where further morphological information (see Table 1) is needed. These additional data will likely reveal similarities yet to be described and/or differences in the organization of pathways in the forebrain. Local circuits neurons in the dorsal thalamus In the dorsal thalamus of amniotes, two types of neurons are present: local circuit neurons (also called interneurons) and relay cells. Axons of local circuit neurons remain within their region of origin whereas axons of relay (projection) cells terminate outside of this area (Jones 2007). With the exception of the dorsal geniculate nucleus (Pritz and Stritzel 1994b), these previously identified dorsal thalamic nuclei in crocodilians lack local circuit neurons and contain only projection cells. These observations were based on two types of experiments. One approach used massive injections of a tracer into various telencephalic targets and subsequent counting of retrogradely labeled neurons in specific dorsal thalamic nuclei. Only rarely was an unlabeled neuron observed in the following dorsal thalamic nuclei: rotundus (Pritz and Stritzel 1986), reuniens pars centralis (Pritz and Stritzel 1986), and dorsolateralis anterior (Pritz and Stritzel 1987). The other technique independently confirmed these latter findings utilizing a different approach: immunocytochemistry. Based on the observations that local circuit neurons are immunoreactive to antibodies to gamma amino butyric acid (GABA) or to glutamic acid decarboxylase (GAD), neurons in these dorsal thalamic nuclei were examined using this methodology. With the exception of the dorsal geniculate nucleus, the following dorsal thalamic nuclei were unlabeled: dorsolateralis anterior, dorsomedialis anterior, reuniens pars centralis, reuniens pars diffusa, rotundus, diagonalis, posterocentralis, and medialis complex posterior (Pritz and Stritzel 1988; 1994a). Lack of GABA/ GAD immunoreactive neurons in certain dorsal thalamic nuclei is not unique to crocodilians but has been described in turtles (Belekhova et al. 1991) and chameleons (Bennis et al. 1991) and pigeons (Domenici et al. 1988; Granda and Crossland 1989). However, some GABA immunoreactive cells have been noted surrounding the borders of certain dorsal thalamic nuclei in turtles (Belekhova et al. 1991) and a few such immunoreactive neurons have been observed in certain dorsal thalamic nuclei in pigeons (Veenman and Reiner 1994). Similar to crocodilians (Pritz and Stritzel 1994b), GABA immunoreactive neurons are present in the dorsal geniculate of turtles (Belekhova et al. 1991; Rio et al. 1992; Kenigfest et al. 1995) and chameleons (Bennis et al. 1991) and in its avian homolog in pigeons (Domenici et al. 1988; Granda and Crossland 1989; Veenman and Reiner 1994; Miceli et al. 2008). In mammals, the percentage of local

Crocodilian forebrain 953 Table 2 Generalized pattern of synaptic elements of sensory systems that synapse in the midbrain of Caiman crocodilus a Sensory modality Synaptic element Audition Vision Body surface somatosensation Receptors Hair cells Rods & cones Somatosensory receptors Bipolar cells Spiral ganglion Retinal bipolar cells Dorsal root ganglion Long axon cells Cochlear nuclei Retinal ganglion cells Dorsal column nucleus Midbrain Torus semicircularis cn Optic tectum Intercollicular area Thalamus N. reuniens pc N. rotundus Medialis complex posterior Telencephalon DVR DVR DVR a Modified from Pritz and Stritzel (1994c). Abbreviations: cn, central nucleus; DVR, dorsal ventricular ridge; N., nucleus; pc, pars centralis. circuit neurons in individual dorsal thalamic nuclei varies. In some small-brained species, GABA immunoreactive neurons are either absent or sparsely present in some dorsal thalamic nuclei while being present in other dorsal thalamic nuclei in the same species. On the other hand, large-brained mammals have local circuit neurons present throughout dorsal thalamic nuclei in varying percentages (Jones 2007). Telencephalic projections to the dorsal thalamus Another characteristic feature of organization of the forebrain in mammals is reciprocal connections between thalamic nuclei and their respective areas of cortical projection (Jones 2007). To date, this feature does not appear to be present in nuclei that project to the anterior dorsal ventricular ridge in crocodilians (Pritz 2014). Limited observations suggest that this may be a feature of the medialis complex anterior (Pritz unpublished observations). As data are incomplete, other dorsal thalamic nuclei may possess this property. Rather than having reciprocal connections with the non-cortical telencephalon directly, telencephalic efferents arise from the basal ganglia in crocodilians (Brauth and Kitt 1980; Brauth 1988). Similar neural circuits are shared by other reptiles and birds (Hoogland 1977; Voneida and Sligar 1979; Russchen and Yonker 1988; Reiner et al. 1998). In turtles (Hall et al. 1977; Ulinski 1986; Kenigfest et al. 1998), reciprocal connections between the dorsal geniculate nucleus and the cortex have been documented. A similar feature also has been noted in birds (Adamo 1967; Karten et al. 1973; Miceli et al. 1987). Whether this circuit is present in crocodilians remains to be seen. Organization of the thalamic reticular nucleus One nucleus that is integral to the organization and function of the forebrain in mammals is the thalamic reticular nucleus. Several features characterize this nucleus in mammals. First, the thalamic reticular nucleus projects to all dorsal thalamic nuclei. Second, its neurons are located within the internal capsule, the fiber bundle interconnecting the dorsal thalamus and cerebral cortex. Third, the thalamic reticular nucleus is composed of a homogeneous group of inhibitory neurons that utilize GABA and GAD. These same neurons also contain the calcium-binding protein, parvalbumin. Fourth, dendrites of the thalamic reticular nucleus are oriented perpendicular to the fibers of the internal capsule (Jones 2007). In crocodilians, a thalamic reticular nucleus was determined based on injections of a retrograde tracer into two caudal dorsal thalamic nuclei: rotundus and the medialis complex posterior (Pritz and Stritzel 1990b). Subsequent to these injections, retrogradely labeled cells were located within the fibers of the dorsal peduncle of the lateral forebrain bundle (Pritz and Stritzel 1990b), a fiber tract connecting the dorsal thalamus with the telencephalon. In Caiman, the thalamic reticular nucleus contains at least two groups of neurons based on immunocytochemical properties. One cell-type projects to the dorsal thalamic nuclei, is immunoreactive for parvalbumin (Pritz and Stritzel 1991, 1993), and has dendrites that are oriented parallel to the fibers of the lateral forebrain bundle (Pritz and Stritzel 1991). The other neuronal group contains neurons immunoreactive to GAD and has its processes oriented perpendicular to the axons of the lateral forebrain bundle (Pritz and Stritzel 1990b). In turtles (Kenigfest et al. 2005) and lizards (Diaz et al. 1994), similar dorsal thalamic projections of a thalamic reticular nucleus have been described. In turtles, some neurons in this nucleus that are immunoreactive both to GAD and to parvalbumin project to the dorsal thalamus (Kenigfest et al. 2005). In pigeons, a thalamic reticular nucleus that projects to the dorsal thalamus has been identified experimentally (Benowitz and Karten

954 M. B. Pritz 1976). Its neurons are immunoreactive to GABA (Domenici et al. 1988; Granda and Crossland 1989; Veenman and Reiner 1994). Neural development in A. mississippiensis Despite sharing certain features of similar neural circuitry, the forebrain in adult sauropsids, including crocodilians, appears quite different from that of adult mammals. These differences must have occurred through evolution and development. Evolutionary explanations require further data on features described above with subsequent comparisons with other amniotes. These include characters such as fiber connections and cellular properties including the molecular signature both of areas of the brain and of individual neurons. On the other hand, studies focusing on development should determine which processes have occurred to produce these different adult forebrains. To address some of these latter questions, experiments were undertaken to address three basic questions. First, are the very early development of both the forebrain and other individual regions of the brain in Alligator similar to that of other amniotes? Second, is the later development of each of these primary brain vesicles of the brain similar after each of these areas has undergone segmental divisions but before each respective area has undergone internal cytoarchitectonic differentiation? Third, specifically focusing on the diencephalon, is its initial development similar among amniotes? Without answering these basic questions, further studies to investigate later development would be difficult to interpret accurately. Early development of the brain in Alligator and other amniotes Early in development, the brains of all amniotes, including those of crocodilians, follow a similar plan. The brain begins as a single, hollow vesicle (Nieuwenhuys et al. 1998), which subsequently undergoes a series of subdivisions. Initially, these transformations occur in the transverse plane as this single vesicle is divided into three, with the formation of: forebrain or prosencephalon, midbrain or mesencephalon, and hindbrain or rhombencephalon (Vaage 1969; Vieira et al. 2010). Subsequently, two events occur. One separates the forebrain transversely into an anterior, secondary prosencephalon, and a posterior, diencephalon (Vaage 1969; Puelles et al. 1987). The other divides the entire neural tube longitudinally into a dorsal, alar, and a ventral, basal plate. At this time, a section perpendicular to the long axis of this longitudinal plane contains four parts: roof plate, alar plate, basal plate, and floor plate (Puelles 1995). While these four areas are present in a section perpendicular to this longitudinal axis, the area of a given component varies, depending on which part of the brain s vesicle is examined (Fig. 3). Crocodilians undergo these general developmental transformations in a manner similar to that of other amniotes (Pritz 2008). At this four-vesicle stage in amniotes, the diencephalon becomes further subdivided transversely into segments known as prosomeres. Each prosomere ultimately forms the following divisions observed in the brains of adults: ventral thalamus (prethalamus), dorsal thalamus (thalamus, including the epithalamus), and pretectum (Puelles and Rubenstein 2003). Each diencephalic prosomere contains both an alar and a basal component (Puelles and Rubenstein 2003). At this time in development in Alligator, prior to stage 14.5, each prosomere is cytoarchitectonically undifferentiated (see Pritz [2008] for Alligator and for review of other amniotes). While the prosomeres of the diencephalon remain morphologically homogenous at this time in development in Alligator (before stage 14.5), other major vesicles of the brain are not. The secondary prosencephalon shows distinctly different patterns of cortical layering at stage 11 depending on the location in the cerebral hemisphere (Fig. 4). On the other hand, at stage 13, the alar midbrain contains a layered structure destined to become the optic tectum, whereas an internal structure, the torus semicircularis, exhibits a nuclear organization (Fig. 5). At this time in the development of Alligator embryos, at stage 11, the hindbrain has already begun differentiation (Pritz 1999). While formation of the hindbrain in Alligator is similar to that observed in other species (Pritz 1999), its pattern of development differs from that of the secondary prosencephalon, diencephalon, and midbrain. Early development of the hindbrain follows a two segment rule in which odd-numbered rhombomeres share particular cellular and molecular characteristics alternate with evennumbered segments that display different molecular and cellular properties (Lumsden 2004). These observations suggest that differentiation of the secondary prosencephalon, diencephalon, midbrain, and hindbrain follows a different time-course and, most likely, a different, developmental plan for each of these vesicles (Pritz 2010b). Furthermore, in the diencephalon of Alligator when individual prosomeres remain undifferentiated, basal portions differ from alar parts in the expression of Pax6 (Pritz and Ruan 2009) and in the orientation of

Crocodilian forebrain 955 Fig. 3 Early development of the brain in vertebrates. Schematic, lateral views of an idealized vertebrate brain are shown at: one-vesicle, three-vesicle, and four-vesicle stages. A section perpendicular to the long axis of the brain is shown at the four-vesicle stage for the: secondary prosencephalon, diencephalon, midbrain, and hindbrain. Abbreviations: ap, alar plate; bp, basal plate; fp, floor plate; rp, roof plate; 28, secondary. fiber tracts (Pritz 2010a). In Alligator, at least, this suggests that alar and basal components in the diencephalon may also follow a different developmental program at this time when prosomeres are homogeneous in appearance and have yet to undergo internal cytoarchitectonic differentiation. If a similar pattern also is present in other vesicles of the brain, one potential consequence is that evolution could act in a seemingly independent fashion on each of the alar and basal parts of the secondary prosencephalon, diencephalon, midbrain, and hindbrain. Such features are unlikely to be unique to crocodilians but common to other amniotes. Thus, the responses to the three developmental questions posed previously are as follows. First, very early development of the brain in crocodilians

956 M. B. Pritz Fig. 4 Laminar organization of the cerebral cortex in an Alligator embryo. Differences in cortical lamination are illustrated in an horizontal section of the left hemisphere of a stage-11 Alligator embryo stained with peanut agglutinin at low (A) and higher (B D) magnification to show variation between anterior (B), lateral (C), and medial (D) cortical areas. Abbreviations: m, medial; r, rostral, s, stage. is similar to that observed in other amniotes. Second, development of individual regions of the brain in Alligator embryos differs. This is likely to be the case for other amniotes. Third, initial development of the diencephalon in Alligator is similar to that in other amniotes, although alar and basal parts in Alligator are likely to follow a different developmental scheme. Later development of the brain What about subsequent development of the brain after individual segments have become internally differentiated? Although many details are incomplete, several generalized features are likely common to a process that transforms an embryonic brain into its adult morphology (Fig. 6). Initially, an uncommitted cell will become either a neuron or a glial cell. While glia are clearly important, the following discussion focuses on neurons. This uncommitted cell undergoes mitoses and proliferation before exiting the cell cycle and acquiring a phenotype as either a neuron or a glial cell. Neurons migrate in a variety of ways, radially, tangentially, and over long distances, before ultimately becoming either a relay (projection) or local circuit neuron. These two types of neurons align themselves in only one of two ways: either in layers or as a nucleus surrounded by borders. These processes are influenced by expression of transcription factors and signaling molecules as well as by local environmental factors. Although variation among vertebrates occurs, in general, these events take place before birth. Early in vertebrate development, a continuous layer of cells between the ventricular border and the pial surface of the neural tube stretches from the hindbrain rostrally to the secondary prosencephalon (Senn 1970). In certain regions of the brain, this laminated pattern persists, while in others, neuronal aggregates lose their layered appearance and become organized as nuclei. In the diencephalon of amniotes, a nuclear pattern is seen. However, this is not the morphology observed in other vertebrates. For example, in amphibians and fish a primitive layering pattern persists into adulthood rather than being organized as nuclei (Nieuwenhuys et al. 1998; Butler and Hodos 2005). In the telencephalon, a layered morphology dominates in mammals. In this class, the cortex occupies a much greater extent of tissue than does the comparatively thin rim of layered neurons seen in reptiles and birds (Nieuwenhuys et al. 1998; Butler and Hodos 2005). Other features of the forebrain s organization in amniotes display similarities as well as differences. Some of these characters have been

Crocodilian forebrain 957 Fig. 5 Organization of the alar midbrain in an Alligator embryo. A sagittal section of a stage-13 Alligator brain stained for cresyl violet is shown. Low-power magnification of an enclosed area of the optic tectum (A) is illustrated at higher magnification (B). At this early developmental stage, the optic tectum is layered. On the other hand, the torus semicircularis (*) is organized as a nucleus (A). Abbreviations: c, caudal; d, dorsal; s, stage. discussed previously. While the programs to explain these observations are far from being understood, the following two examples suggest developmental approaches directed at explaining differences in the morphology of adults between mammals and sauropsids. Formation of thalamo telencephalic fiber tracts in amniotes In mammals, the paths of thalamo-cortical axons are influenced by corridor cells (Bielle et al. 2011; Molnár et al. 2012) as well as by several factors expressed in the surrounding forebrain (Braisted et al. 2000, 2009; Lopez-Bendito et al. 2006; Uziel et al. 2006; Molnár et al. 2012; Garel and Lopez-Bendito 2014). Identification of these factors and their expression at different times during development suggests an explanation for the trajectory, course, and locus of termination in mice as opposed to chicks (Bielle et al. 2011) and turtles (Bielle et al. 2011; Tosa et al. 2015). In chicks and turtles, these studies have focused mainly on the course of axons ending in the dorsal ventricular ridge (Bielle et al. 2011; Tosa et al. 2015). This has provided an explanation as to why thalamic efferents in mammals form the internal capsule to end in the cortex while in sauropsids, these fibers enter the lateral forebrain bundle to terminate in the dorsal ventricular ridge. In crocodilians, and most likely in other reptiles and in birds, although not the major fiber tract, several additional paths to the telencephalon have been identified (Pritz 2014). How these other paths relate to these data remains to be determined but may explain the evolution of thalamo telencephalic connections in addition to possibly sculpting the cytoarchitecture of the forebrain in sauropsids. Formation of the cortex in amniotes In mammals, the neocortex develops in an inside-out fashion (Angevine and Sidman 1961), contains a subventricular zone, and possesses a prominent and well-organized cortical plate, which forms in an inside outside manner (Molnár et al. 2006). On the other hand, the comparable region in sauropsids, the dorsal pallium, develops in an outside inside fashion (Tsai et al. 1981; Goffinet et al. 1986), has a rudimentary subventricular zone (if present at all) (Goffinet 1983; Martinez-Cerdeno et al. 2006; Cheung et al. 2007), and has a cortical plate that is rudimentary in turtles although well-developed in lizards (Goffinet 1983). A variety of explanations have been advanced to account for some of these differences in cortical complexity and layering. One possibility is the migratory behavior of local circuit neurons originating in the medial ganglionic eminence to enter the cortical plate. In mice, the cortical plate is permissive to the migration of local circuit neurons through this structure but not in chicks or turtles (Tanaka et al. 2011). The above examples of developmental processes that sculpt the forebrain provide approaches to understand not only development of the forebrain in amniotes but also its evolution. In regards to the diencephalon of crocodilians, the time-period when homogeneous prosomeres become internally subdivided represents the developmental epoch when these regions begin to acquire their respective adult morphologies. How these changes occur will require further information not only on a variety of morphological features but also on developmental

958 M. B. Pritz Fig. 6 Development of cortical and nuclear structures of the brain. The influence of local environmental factors and molecular events over time is shown in this schematic beginning with an undifferentiated cell and ending with transformation into the cortex or into a nucleus. Although the time course for this scheme of development will vary among vertebrates, for the most part, these events will occur before birth. Abbreviation: LCN, local circuit neuron. processes to explain how forebrains of amniotes evolved. Acknowledgments C. Brown helped with Figs. 3 and 6. J. Murphy assisted in the preparation of Figs. 4 and 5. Dr R.M. Elsey and the Louisiana Department of Wildlife and Fisheries provided Alligator eggs. Funding Partial funding for participation in this symposium was provided for by the Society for Integrative and Comparative Biology. Some of the data described in this report were supported by the National Institutes of Health [NS 20120] and by a biomedical research grant from Indiana University School of Medicine. References Adamo NJ. 1967. Connections of efferent fibers from hyperstriatal areas in chicken, raven, and African lovebird. J Comp Neurol 131:337 55. Angevine JB Jr, Sidman RL. 1961. Autoradiographic study of cell migration during histogenesis of cerebral cortex in the mouse. Nature 192:766 8. Belekhova MG, Kratskin IL, Reperan Z, P err Z, Veselkin NP, Kenigfest NB, Tumanova NL, Chkheidze DD. 1991. Localization of GABA-immunoreactive elements in the thalamus of the turtle Emys orbicularis. J Evol Biochem Physiol 27:507 17. Bennis M, Calas A, Geffard M, Gamrani H. 1991. Distribution of GABA immunoreactive systems in the forebrain and midbrain of the chameleon. Br Res Bull 26:891 8. Benowitz LI, Karten HJ. 1976. Organization of the tectofugal visual pathway in the pigeon: a retrograde transport study. J Comp Neurol 167:503 20. Bielle F, Marcos-Mondejar P, Keita M, Mailhes C, Verney C, Nguyen Ba-Charvet K, Tessier-Lavigne M, Lopez- Bendito G, Garel S. 2011. Slit2 activity in the migration

Crocodilian forebrain 959 of guidepost neurons shapes thalamic projections during development and evolution. Neuron 69:1085 98. Braford MR Jr. 1972. Ascending efferent tectal projections in the South American spectacled caiman. Anat Rec 172:275 6. Braford MR Jr. 1973. Retinal projections in Caiman crocodilus. Am Zool 13:1345. Braisted JE, Catalano SM, Stimac R, Kennedy TE, Tessier- Lavigne M, Shatz CJ, O Leary DD. 2000. Netrin-1 promotes thalamic axon growth and is required for proper development of the thalamocortical projection. J Neurosci 20: 5792 801. Braisted JE, Ringstedt T, O Leary DD. 2009. Slits are chemorepellents endogenous to hypothalamus and steer thalamocortical axons into ventral telencephalon. Cereb Cortex 19(Suppl. 1):i144 51. Brauth SE. 1988. The organization and projections of the paleostriatal complex in Caiman crocodilus. In. In: Schwerdtfeger WK, Smeets WJAJ, editors. The forebrain of reptiles: current concepts of structure and function. New York: Karger. p. 60 76. Brauth SE, Kitt CA. 1980. The paleostriatal system of Caiman crocodilus. J Comp Neurol 189:437 65. Bruce LL. 2007. Evolution of the nervous system in reptiles. In: Bullock TH, Rubenstein JLR, Kaas JH, editors. Evolution of nervous systems: a comprehensive reference. Vol. II: The evolution of nervous systems in non-mammalian vertebrates. New York: Oxford University Press. p. 125 56. Burns AH, Goodman DC. 1967. Retinofugal projections of Caiman sklerops. Exp Neurol 18:105 15. Butler AB, Hodos W. 2005. Comparative vertebrate neuroanatomy: evolution and adaptation. Hoboken (NJ): John Wiley & Sons. Butler AB, Reiner A, Karten HJ. 2011. Evolution of the amniote pallium and the origins of mammalian neocortex. Ann NY Acad Sci 1225:14 27. Cheung AF, Pollen AA, Tavare A, DeProto J, Molnár Z. 2007. Comparative aspects of cortical neurogenesis in vertebrates. J Anat 211:164 76. Crosby EC. 1917. The forebrain of Alligator mississippiensis. J Comp Neurol 27:325 402. Derobert Y, Médina M, Rio JP, Ward R, Repérant J, Marchand M-J, Miceli D. 1999. Retinal projections in two crocodilian species, Caiman crocodilus and Crocodylus niloticus. Anat Embryol (Berl) 200:175 91. Diaz C, Yanes C, Trujillo CM, Puelles L. 1994. The lacertidian reticular thalamic nucleus projects topographically upon the dorsal thalamus: experimental study in Gallotia galloti. J Comp Neurol 343:193 208. Domenici L, Waldvogel HJ, Matute C, Streit P. 1988. Distribution of GABA-like immunoreactivity in the pigeon brain. Neuroscience 25:931 50. Garel S, Lopez-Bendito G. 2014. Inputs from the thalamocortical system on axon pathfinding mechanisms. Curr Opin Neurobiol 27:143 50. Goffinet AM. 1983. The embryonic development of the cortical plate in reptiles: a comparative study in Emys orbicularis and Lacerta agilis. J Comp Neurol 215:437 52. Goffinet AM, Daumerie C, Langerwerf B, Pieau C. 1986. Neurogenesis in reptilian cortical structures: 3 H-thymidine autoradiographic analysis. J Comp Neurol 243:106 16. Granda RH, Crossland WJ. 1989. GABA-like immunoreactivity of neurons in the chicken diencephalon and mesencephalon. J Comp Neurol 287:455 69. Hall JA, Foster RE, Ebner FF, Hall WC. 1977. Visual cortex in a reptile, the turtle (Pseudemys scripta and Chrysemys picta). Brain Res 130:197 216. Hedges SB. 1994. Molecular evidence for the origin of birds. Proc Natl Acad Sci USA 91:2621 4. Hoogland PV. 1977. Efferent connections of the striatum in Tupinambis nigropunctatus. J Morphol 152:229 46. Jones EG. 2007. The thalamus. New York: Cambridge University Press. Karten HJ, Hodos W, Nauta WJ, Revzin AM. 1973. Neural connections of the visual wulst of the avian telencephalon. Experimental studies in the pigeon (Columba livia) and owl (Speotyto cunicularia). J Comp Neurol 150:253 78. Kenigfest N, Belekhova M, Repérant J, Rio JP, Ward R, Vesselkin N. 2005. The turtle thalamic anterior entopeduncular nucleus shares connectional and neurochemical characteristics with the mammalian thalamic reticular nucleus. J Chem Neuroanat 30:129 43. Kenigfest NB, Repérant J, Rio JP, Belekhova MG, Tumanova NL, Ward R, Vesselkin NP, Herbin M, Chkeidze DD, Ozirskaya EV. 1995. Fine structure of the dorsal lateral geniculate nucleus of the turtle, Emys orbicularis: a Golgi, combined HRP tracing and GABA immunocytochemical study. J Comp Neurol 356: 595 614. Kenigfest NB, Repérant J, Rio JP, Belekhova MG, Ward R, Vesselkin NP, Miceli D, Herbin M. 1998. Retinal and cortical afferents to the dorsal lateral geniculate nucleus of the turtle, Emys orbicularis: a combined axonal tracing, glutamate, and GABA immunocytochemical electron microscopic study. J Comp Neurol 391:470 90. Lopez-Bendito G, Cautinat A, Sanchez JA, Bielle F, Flames N, Garratt AN, Talmage DA, Role LW, Charnay P, Marín O, et al. 2006. Tangential neuronal migration controls axon guidance: a role for neuregulin-1 in thalamocortical axon navigation. Cell 125:127 42. Lumsden A. 2004. Segmentation and compartition in the early avian hindbrain. Mech Dev 121:1081 8. Martinez-Cerdeno V, Noctor SC, Kriegstein AR. 2006. The role of intermediate progenitor cells in the evolutionary expansion of the cerebral cortex. Cereb Cortex 16(Suppl. 1): i152 61. Miceli D, Repérant J, Villalobos J, Dionne L. 1987. Extratelencephalic projections of the avian visual Wulst. A quantitative autoradiographic study in the pigeon Columbia livia. J Hirnforsch 28:45 57. Miceli D, Repérant J, Ward R, Rio JP, Jay B, Medina M, Kenigfest NB. 2008. Fine structure of the visual dorsolateral anterior thalamic nucleus of the pigeon (Columba livia): a hodological and GABA-immunocytochemical study. J Comp Neurol 507:1351 78. Molnár Z, Garel S, Lopez-Bendito G, Maness P, Price DJ. 2012. Mechanisms controlling the guidance of thalamocortical axons through the embryonic forebrain. Eur J Neurosci 35:1573 85.

960 M. B. Pritz Molnár Z, Metin C, Stoykova A, Tarabykin V, Price DJ, Francis F, Meyer G, Dehay C, Kennedy H. 2006. Comparative aspects of cerebral cortical development. Eur J Neurosci 23:921 34. Nieuwenhuys R, Donkelaar HJ, Nicholson C. 1998. The central nervous system of vertebrates. New York: Springer. Pritz MB. 1974a. Ascending connections of a midbrain auditory area in a crocodile, Caiman crocodilus. J Comp Neurol 153:179 97. Pritz MB. 1974b. Ascending connections of a thalamic auditory area in a crocodile, Caiman crocodilus. J Comp Neurol 153:199 213. Pritz MB. 1975. Anatomical identification of a telencephalic visual area in crocodiles: ascending connections of nucleus rotundus in Caiman crocodilus. J Comp Neurol 164:323 38. Pritz MB. 1995. The thalamus of reptiles and mammals: similarities and differences. Brain Behav Evol 46:197 208. Pritz MB. 1999. Rhombomere development in a reptilian embryo. J Comp Neurol 411:317 26. Pritz MB. 2008. Early diencephalon development in Alligator. Brain Behav Evol 71:15 31. Pritz MB. 2010a. Forebrain and midbrain fiber tract formation during early development in Alligator embryos. Brain Res 1313:34 44. Pritz MB. 2010b. Do early vertebrate brain subdivisions develop in similar or different ways? Brain Behav Evol 75:312 3. Pritz MB. 2014. Dorsal thalamic nuclei in Caiman crocodilus. Neurosci Lett 581:57 62. Pritz MB, Northcutt RG. 1977. Succinate dehydrogenase activity in the telencephalon of crocodiles correlates with the projection areas of sensory thalamic nuclei. Brain Res 124:357 60. Pritz MB, Ruan YW. 2009. PAX6 immunoreactivity in the diencephalon and midbrain of Alligator during early development. Brain Behav Evol 73:1 15. Pritz MB, Stritzel ME. 1986. Percentage of relay and intrinsic neurons in two sensory thalamic nuclei projecting to the non-cortical telencephalon in reptiles Caiman crocodilus. Brain Res 376:169 74. Pritz MB, Stritzel ME. 1987. Percentage of intrinsic and relay cells in a thalamic nucleus projecting to general cortex in reptiles, Caiman crocodilus. Brain Res 409:146 50. Pritz MB, Stritzel ME. 1988. Thalamic nuclei that project to reptilian telencephalon lack GABA and GAD immunoreactive neurons and puncta. Brain Res 457:154 9. Pritz MB, Stritzel ME. 1989. Reptilian somatosensory midbrain: identification based on input from the spinal cord and dorsal column nucleus. Brain Behav Evol 33:1 14. Pritz MB, Stritzel ME. 1990a. Thalamic projections from a midbrain somatosensory area in a reptile, Caiman crocodilus. Brain Behav Evol 36:1 13. Pritz MB, Stritzel ME. 1990b. A different type of vertebrate thalamic organization. Brain Res 525:330 4. Pritz MB, Stritzel ME. 1991. Calcium binding protein immunoreactivity in a reptilian thalamic reticular nucleus. Brain Res 554:325 8. Pritz MB, Stritzel ME. 1993. Neuronal subpopulations in a reptilian thalamic reticular nucleus. Neuroreport 4:791 4. Pritz MB, Stritzel ME. 1994a. Glutamic acid decarboxylase immunoreactivity in some dorsal thalamic nuclei in Crocodilia. Neurosci Lett 165:109 12. Pritz MB, Stritzel ME. 1994b. Morphological and GAD immunocytochemical properties of the dorsal lateral geniculate nucleus in a reptile. Brain Res Bull 33:723 6. Pritz MB, Stritzel ME. 1994c. Anatomical identification of a telencephalic somatosensory area in a reptile, Caiman crocodilus. Brain Behav Evol 43:107 27. Puelles L. 1995. A segmental morphological paradigm for understanding vertebrate forebrains. Brain Behav Evol 46:319 37. Puelles L, Amat JA, Martinez-de-la-Torre M. 1987. Segmentrelated, mosaic neurogenetic pattern in the forebrain and mesencephalon of early chick embryos: I. Topography of AChE-positive neuroblasts up to stage HH18. J Comp Neurol 266:247 68. Puelles L, Rubenstein JL. 2003. Forebrain gene expression domains and the evolving prosomeric model. Trends Neurosci 26:469 76. Reiner A, Medina L, Veenman CL. 1998. Structural and functional evolution of the basal ganglia in vertebrates. Brain Res Brain Res Rev 28:235 85. Repérant J. 1975. [Retinal projections in Caiman sclerops. Autoradiographic study]. C R Acad Sci D 280:2881 4. Rio JP, Réperant J, Ward R, Miceli D, Medina M. 1992. Evidence of GABA-immunopositive neurons in the dorsal part of the lateral geniculate nucleus of reptiles: morphological correlates with interneurons. Neuroscience 47:395 407. Russchen FT, Jonker AJ. 1988. Efferent connections of the striatum and the nucleus accumbens in the lizard Gekko gecko. J Comp Neurol 276:61 80. Senn DG. 1970. The stratification in the reptilian central nervous system. Acta Anat (Basel) 75:521 52. Tanaka DH, Oiwa R, Sasaki E, Nakajima K. 2011. Changes in cortical interneuron migration contribute to the evolution of the neocortex. Proc Natl Acad Sci USA 108:8015 20. Tosa Y, Hirao A, Matsubara I, Kawaguchi M, Fukui M, Kuratani S, Murakami Y. 2015. Development of the thalamo-dorsal ventricular ridge tract in the Chinese soft-shelled turtle, Pelodiscus sinensis. Dev Growth Differ 57:40 57. Tsai HM, Garber BB, Larramendi LM. 1981. 3 H-thymidine autoradiographic analysis of telencephalic histogenesis in the chick embryo: I. Neuronal birthdates of telencephalic compartments in situ. J Comp Neurol 198:275 92. Ulinski PS. 1983. Dorsal ventricular ridge: a treatise on forebrain organization in reptiles and birds. New York: J Wiley & Sons. Ulinski PS. 1986. Organization of corticogeniculate projections in the turtle, Pseudemys scripta. J Comp Neurol 254:529 42. Uziel D, Garcez P, Lent R, Peuckert C, Niehage R, Weth F, Bolz J. 2006. Connecting thalamus and cortex: the role of ephrins. Anat Rec A Discov Mol Cell Evol Biol 288:135 42. Vaage S. 1969. The segmentation of the primitive neural tube in chick embryos (Gallus domesticus). A morphological, histochemical and autoradiographical investigation. Ergeb Anat Entwicklungsgesch 41:3 87. Veenman CL, Reiner A. 1994. The distribution of GABAcontaining perikarya, fibers, and terminals in the forebrain

Crocodilian forebrain 961 and midbrain of pigeons, with particular reference to the basal ganglia and its projection targets. J Comp Neurol 339:209 50. Vieira C, Pombero A, Garcia-Lopez R, Gimeno L, Echevarria D, Martinez S. 2010. Molecular mechanisms controlling brain development: an overview of neuroepithelial secondary organizers. Int J Dev Biol 54:7 20. Voneida TJ, Sligar CM. 1979. Efferent projections of the dorsal ventricular ridge and the striatum in the Tegu lizard, Tupinambis nigropunctatus. J Comp Neurol 186:43 64. Walker AD. 1972. New light on the origin of birds and crocodiles. Nature 237:257 63. Whetstone KN, Martin LD. 1979. New look at the origin of birds and crocodiles. Nature 279:234 6.