Selective deficits in reversal learning after neostriatum caudolaterale lesions in pigeons: Possible behavioral equivalencies to the mammalian prefrontal system

Abstract

The neostriatum caudolaterale (NCL) of birds is thought to be equivalent to the mammalian prefrontal cortex (PFC) due to its dense dopaminergic innervation, its associative structure, and its importance for cognitive tasks which are known to be affected in mammals with prefrontal lesions. The aim of the present study was to analyze the functional importance of the NCL and its main thalamic afferent structure, the n. dorsolateralis posterior thalami (DLP), in reversal and go/no-go tasks, two behavioral procedures which are often used to assess mammalian prefrontal functions. Using a multiple regression analysis in which structure-specific lesion extents are correlated with different postoperative behavioral measures, the specific contribution of the relevant structures were differentiated from the neighbouring areas CDL (area corticoidea dorsolateralis) and NC (neostriatum caudale). The analyses showed a highly significant contribution of the NCL to reversal but not to go/no-go or to visual discrimination performance, while all other structures under analysis had no impact on any behavioral measure. These results underline the specific contribution of the pigeons’ NCL on a subset of cognitive tasks which are known to be affected by prefrontal lesions in mammals.

Introduction

Comparative neuroanatomical studies have demonstrated important similarities in the chemoarchitecture and connectivity of avian and mammalian forebrains [44]. However, these investigations mostly concentrated on sensory and motor pathways, with only limited emphasis on associative structures. Recently, Ivan Divac and colleagues proposed an area in the caudolateral part of the avian forebrain to be comparable to the mammalian prefrontal cortex (PFC) 10, 35. One hallmark of the PFC is the dense dopaminergic innervation 47, 50. Indeed several studies could show that the neostriatum caudolaterale (NCL), a semilunar area in the caudalmost subventricular part of the avian forebrain, receives a dense dopaminergic innervation comparable to the PFC 10, 12, 34, 48, 51, and displays a high density of dopaminergic D1-receptors [42]. The pattern of intratelencephalic afferents of the NCL also resembles that of the PFC and demonstrates a multisensory convergence 31, 34, 37. Additionally, behavioral studies could demonstrate that lesions of the caudolateral forebrain caused deficits in spatial working memory tasks like delayed alternation 14, 18, 35, 36, and sequential procedures like go/no-go [18], but left performance in various visual discriminations intact 14, 35, 36.

Lesion studies are always faced with the problem that the lesioned area generally also includes, at least to a certain extent, neighbouring structures. This is especially problematic in studying the NCL, since every approach inevitably also damages parts of the overlying area corticoidea dorsolateralis (CDL) and the neighbouring neostriatum caudale (NC). Given the dense reciprocal connections of the CDL with the hippocampal formation [5]and the multimodal nature of the NC [37]it is conceivable that a part of the cognitive deficits after NCL-lesions is due to lesioning these neighbouring structures. One elegant way to deal with this problem is to use multiple-regression analyses based on quantitative reconstructions of structure-specific lesion extents which are correlated with different postoperative behavioral measures [21]. Using this method would allow the statistical assessment of the possibly separate contributions of neighbouring structures to different behavioral tasks.

The aim of the present study was to use multiple-regression analyses to examine the functional importance of the pigeon’s NCL and its thalamic afferent structure, the n. dorsolateralis posterior thalami (DLP) [48], in further tasks which are generally associated with prefrontal functions. If it could be shown that these structures are specifically involved in the procedures examined, the idea of an equivalence of NCL and PFC could be substantiated. Four different behavioral tasks were employed. The first was a reversal procedure since deficits in this task are often observed in mammals after prefrontal 6, 40or hippocampal [3]lesions while lesions of the pigeon’s caudal neostriatum are not associated with reversal deficits [32]. Go/no-go was used as a second task, since it is also often observed to be deficient after lesions of the ventromedial prefrontal cortex 25, 41. Performance attenuations in go/no-go tasks were not observed after ablations of the complete caudal neostriatum in pigeons [39]although deficits were seen after extensive NCL-lesions [18]. Finally a grain–grit and a simultaneous pattern discrimination task were employed to control two very different procedures for possible sensory deficits.