Elsevier

Brain Research Bulletin

Volume 57, Issues 3–4, February–March 2002, Pages 499-503
Brain Research Bulletin

Spatial memory and hippocampal pallium through vertebrate evolution: insights from reptiles and teleost fish

https://doi.org/10.1016/S0361-9230(01)00682-7Get rights and content

Abstract

The forebrain of vertebrates shows great morphological variation and specialized adaptations. However, an increasing amount of neuroanatomical and functional data reveal that the evolution of the vertebrate forebrain could have been more conservative than previously realized. For example, the pallial region of the teleost telencephalon contains subdivisions presumably homologous with various pallial areas in amniotes, including possibly a homologue of the medial pallium or hippocampus. In mammals and birds, the hippocampus is critical for encoding complex spatial information to form map-like cognitive representations of the environment. Here, we present data showing that the pallial areas of reptiles and fish, previously proposed as homologous to the hippocampus of mammals and birds on an anatomical basis, are similarly involved in spatial memory and navigation by map-like or relational representations of the allocentric space. These data suggest that early in vertebrate evolution, the medial pallium of an ancestral fish group that gave rise to the extant vertebrates became specialized for processing and encoding complex spatial information, and that this functional trait has been retained through the evolution of each independent vertebrate lineage.

Introduction

The forebrain of vertebrates shows an impressive range of morphological variation and specialized adaptations 9, 28, 31. For example, whereas in amniotes (i.e., mammals, birds, and reptiles) and other non-actinopterygian groups the telencephalon develops by a process of evagination, in the actinopterygian fish (for instance, teleost fish), the telencephalon undergoes a process of eversion or outward bending during the embryonic development [26, 27, 32; see Fig. 1]. These different developmental processes produce notable morphological variation, mainly two solid telencephalic hemispheres separated by an unique ventricle in the actinopterygian radiation that contrasts with the hemispheres with internal ventricles in non-actinopterygians. However, an increasing amount of neuroanatomical data reveal that the evolution of the vertebrate forebrain could have been more conservative than previously realized. Thus, the pallial region of the actinopterygian telencephalon, contains subdivisions presumably homologous with various pallial areas in amniotes. It includes a possible homologue of the medial pallium or hippocampus 5, 8, 9, 28, 29, 31, although the cell masses derived from the embryonic alar prosencephalic plate, which form the medial pallium in non-actinopterygians, occupy a lateral position in actinopterygians (Fig. 1).

In this context, our research has focused on the brain mechanisms that underlie spatial cognition in vertebrates, with a comparative and multidisciplinary approach. Given that an impressive amount of data demonstrate that the hippocampus of mammals and birds is critical for map-like or relational memory representations of the allocentric space 4, 7, 11, 12, 34, 43, we investigated whether the reptilian medial cortex and the teleost lateral telencephalic pallium, both proposed to be homologous to the mammalian hippocampus, also play a similar role in spatial cognition.

In this review, we present data supporting the notion that mammals, birds, reptiles, and teleost fish share a number of similar basic spatial cognition mechanisms, in particular, that all of these vertebrate groups have place memory capabilities, based on map-like or relational memory representations of the allocentric space; like the hippocampus of mammals and birds, the reptilian medial cortex is critical for cognitive mapping capabilities; and the teleost lateral telencephalic pallium, the presumed homologue of the amniote hippocampus, is selectively involved in spatial cognition.

Considerable experimental effort has been devoted to the analysis of spatial learning and memory mechanisms in different vertebrate species. From this analysis, it became clear that the spatial capabilities of mammals and birds are remarkably similar. For example, besides orienting themselves on the basis of discrete cue representations and other nonrelational mechanisms, every mammalian and avian species studied use hippocampal-dependent, cognitive mapping strategies to navigate to a goal by means of encoding its spatial relationships with a number of landmarks in a map-like, allocentric representation that provides a stable frame of reference called place memory 1, 34, 47. In a number of thorough behavioral studies devoted specifically to investigate the spatial strategies used by reptiles and fish to solve spatial problems, we have shown the presence of complex spatial learning and memory capabilities in these vertebrate groups that closely parallel those described in mammals and birds. The most noteworthy result of these studies is that turtles and teleost fish are able to use cognitive mapping strategies to solve spatial problems 19, 22, 24, 39, 48. Thus, both turtles and goldfish trained in a variety of standard place learning procedures are able to reach the goal even when novel start positions are used and they have to adopt shortcuts and novel routes in the absence of local cues 22, 24, 39. Moreover, turtles and goldfish can use a number of widely distributed visual cues to solve spatial tasks by means of encoding the whole spatial arrangement, such that none of those cues is essential by itself to locate the goal 22, 24, 39.

These data indicate that turtles and goldfish, like mammals and birds, are able to use place strategies based on map-like or relational memory representations of the allocentric space. Because in mammals and birds the hippocampus is critical for cognitive mapping strategies, the question is: Are the cognitive mapping capabilities observed in reptiles and teleost fish also based on homologous neural mechanisms?

Considerable experimental evidence shows that the hippocampus of mammals and birds is critical for encoding the environmental information in map-like or relational memory representations for spatial navigation 4, 7, 34, 43. In these vertebrate groups, damage to the hippocampal formation produces selective impairments in spatial tasks that require the encoding of reciprocal relationships among environmental features (place learning), but not in tasks requiring the subject to approach a single cue or requiring nonspatial discriminations 3, 14, 15, 25, 33, 35, 37, 44. Similarly, lesions to the medial cortex (MC) in turtles produce dramatic and selective impairments in place learning and memory. For example, MC lesions produce a significant memory deficit in turtles trained in a place procedure in a dry-maze analogue, as evidenced by their post-surgery failure to navigate to the goal 6, 17. In addition, MC lesions impair turtles to use place strategies in standard arm-maze tasks, as indicated by their performance in transfer and probe trials. Thus, although MC turtles, like sham animals, learn to reach the goal during training trials, the MC turtles fail to navigate to the goal from novel start locations (transfer trials) and when the visual cues in the proximity of the goal are excluded (probe trials), indicating that these animals are unable to use a cognitive mapping strategy [17]. In addition, like hippocampus lesioned mammals and birds, MC-lesioned turtles present a clear spatial reversal learning deficit 17, 23. Furthermore, MC lesions in turtles, like hippocampal lesions in mammals and birds, do not impair (or even facilitate) the use of guidance and other nonrelational strategies to reach the goal 6, 17, 23.

It is worthy to note that these results showing a striking functional similarity between the reptilian medial cortex and the hippocampus of mammals and birds have considerable comparative value. Two characters can only be considered homologous in two or more taxa if they can be traced back to the presumptive common ancestor of those taxa 10, 46, 52. The modern reptiles, however, including chelonians, cannot be viewed as the ancestral stock from which living birds and mammals evolved, and in addition, little evidence has remained in the fossil record for characters such as brain and behavior. Thus, any hypothesis of homology concerning spatial memory systems in amniotes must be inferred from the distribution of characters in the amniote extant species on the base of a principle of parsimony 13, 16, 30, 31, 36, 52.

These data revealing the selective involvement of the reptilian medial cortex in spatial cognition suggest that the presence of the hippocampus as a fundamental structure of a spatial memory system could be a primitive character in amniotes that could have been present in the common reptilian ancestor of modern turtles, mammals, and birds that inhabited the earth in the Mesozoic era, about 200 million years ago. Furthermore, a number of functional results suggest the possibility that this trait may have appeared even earlier in the evolution of vertebrates, i.e., experimental evidence showing that the lateral telencephalic pallium of the actinopterygian fish is similarly involved in spatial cognition.

Recently, we have obtained a considerable amount of evidence showing that the teleost fish capability to navigate using map-like or relational spatial memory representations depends on telencephalic structures 6, 18, 20, 41, 42, 49, 50. Telencephalic ablations in goldfish produce dramatic deficits in spatial tasks requiring place memory strategies, but not in tasks requiring egocentric strategies or cue guidance 6, 18, 20, 21, 41, 42. Thus, the effects of complete telencephalic ablation on the performance of fish in spatial tasks closely resemble those following hippocampal damage in mammals, birds, and reptiles.

In order to identify what subregions of the teleost telencephalon are most directly implicated in spatial memory, Vargas et al. [50] evaluated the possible spatial learning-related changes in neuronal protein synthesis by means of a silver stain (AgNOR) with high affinity for the argyrophilic proteins associated with the nucleolar organizer regions (NORs) of the neurons [38]. With this objective, the transcriptive activity was studied in goldfish trained in a spatial learning task by measuring the NOR relative area of the neurons of the lateral and the medial pallium. The analysis showed that the NOR area of the neurons in the lateral pallium of goldfish trained in the spatial learning task increased significantly relative to the NOR area in the lateral pallium of the animals trained in a control procedure (Fig. 2). In addition, the increase in protein synthesis was restricted to the lateral pallium, as no between-group differences were observed in the neurons of the medial pallium. Furthermore, and just as important, the increase in the transcriptive activity of the neurons in the lateral pallium was dependent on spatial learning, as no increase was observed in the animals trained in a cue learning procedure [51]. These results are notable, because as a consequence of the eversion process that characterizes telencephalic development in actinopterygians, the telencephalic area that occupies a topological position equivalent to the amniote medial pallium or hippocampus is the lateral pallium 5, 8, 28, 29, 31.

The selective implication of the teleost lateral telencephalic pallium in spatial cognition was confirmed in a series of experiments designed to analyze the effects of selective pallial lesions on a number of different spatial memory procedures. Thus, lesions in the lateral pallium of the goldfish disrupt post-surgery performance and spatial reversal learning in different spatial memory tasks 6, 21, 40, 49. Following lateral pallium lesions, goldfish trained in a radial arm-maze placed in a room with numerous widely distributed extramaze visual cues lost the ability to navigate to goal location 40, 49. This deficit to find a familiar place is observed whenever the animals are required to navigate from novel start locations (post-surgery transfer trials), but also when well trained start locations and routes are used (post-surgery training trials). The place memory impairment observed after lateral pallium lesions in goldfish is as severe as that produced by the complete ablation of both telencephalic lobes 18, 40, 42. In contrast, medial and dorsal telencephalic pallium lesions did not produce any observable deficit in the post-surgery performance of goldfish trained in place tasks 18, 40, 41, 49. In addition, the involvement of the teleost lateral pallium in spatial cognition seems to be selective to place learning, because lesions restricted to this area, or complete telencephalic ablation, did not impair turn or cue guidance strategies 6, 18, 40, 41, 42, 49.

In summary, our data indicate that the lateral pallium of teleost fish underlies allocentric spatial memory processes, but it is not essential for spatial orientation based on simple stimulus-response associations. That is, these results demonstrate a striking functional similarity between the hippocampal pallium of amniotes and the lateral pallium of the telencephalon of the teleost fish, which contributes to the identification of the actinopterygian hippocampal pallium and provides compelling evidence for an eversion process in the teleost telencephalon, with considerable preservation of the original topology.

Section snippets

Concluding remarks

Spatial memory has proven to be a powerful tool for the comparative study of animal cognition. In addition, the present results emphasize the value of multidisciplinary approaches, in which behavioral, psychobiological, neurophysiological, neuromorphological, and developmental studies converge to decipher important and still unresolved questions about brain and behavior evolution. Thus, although historically a dominant trend has occurred in comparative brain and behavior research emphasizing

References (52)

  • Broglio, C.; Gómez, Y.; López, J. C.; Rodrı́guez, F.; Salas, C.; Vargas, J. P. Stability of the spatial memory function...
  • N. Burgess et al.

    The hippocampal and parietal foundations of spatial cognition

    (1999)
  • A.B. Butler et al.

    Comparative vertebrate neuroanatomyEvolution and adaptation

    (1996)
  • C.B.G. Campbell et al.

    The Scala naturae revisitedEvolutionary scales and anagenesis in comparative psychology

    J. Comp. Psychol.

    (1991)
  • H. Eichenbaum et al.

    Two functional components of the hippocampal memory system

    Behav. Brain Sci.

    (1994)
  • H. Eichenbaum et al.

    Hippocampal representation in place learning

    J. Neurosci.

    (1990)
  • N. Eldredge et al.

    Phylogenetic patterns and the evolutionary process

    (1980)
  • T. Fremouw et al.

    Impaired place learning and unimpaired cue learning in hippocampal-lesioned pigeons

    Behav. Neurosci.

    (1997)
  • M. Good

    The effects of hippocampal-area parahippocampalis lesions on discrimination learning in the pigeon

    Behav. Brain Res.

    (1987)
  • G.J. Henning

    Phylogenetic systematics

    (1966)
  • J.C. López

    Spatial learning and medial cortex in the turtle Pseudemys scripta.

    PhD. dissertation

    (1999)
  • J.C. López et al.

    Dissociation of place and cue learning by telencephalic ablation in goldfish

    Behav. Neurosci.

    (2000)
  • J.C. López et al.

    Multiple spatial learning strategies in goldfish (Carassius auratus)

    Anim. Cogn.

    (1999)
  • J.C. López et al.

    Pallial lesions impair spatial learning in goldfish

    Eur. J. Neurosci.

    (1998)
  • J.C. López et al.

    Spatial learning in turtles

    Anim. Cogn.

    (2001)
  • López, J. C.; Gómez, Y.; Rodrı́guez, F.; Broglio, C.; Jiménez-Moya, F.; Salas, C. Reversal learning deficit in turtles...
  • Cited by (193)

    • Conserved functions of hypothalamic kisspeptin in vertebrates

      2022, General and Comparative Endocrinology
    View all citing articles on Scopus
    View full text