Microstructure of the neocortex: comparative aspects

Journal of Neurocytology 31, 299–316 (2002) Microstructure of the neocortex: Comparative aspects J AV I E R D E F E L I P E ∗ , L I D I A A L O N S O - N A N C L A R E S and JON I. ARELLANO Instituto Cajal (CSIC), Avenida Dr. Arce, 37, 28002 Madrid, Spain [email protected] Received December 16, 2002; accepted January 8, 2003 Abstract The appearance of the neocortex, its expansion, and its differentiation in mammals, represents one of the principal episodes in the evolution of the vertebrate brain. One of the fundamental questions in neuroscience is what is special about the neocortex of humans and how does it differ from that of other species? It is clear that distinct cortical areas show important differences within both the same and different species, and this has led to some researchers emphasizing the similarities whereas others focus on the differences. In general, despite of the large number of different elements that contribute to neocortical circuits, it is thought that neocortical neurons are organized into multiple, small repeating microcircuits, based around pyramidal cells and their input-output connections. These inputs originate from extrinsic afferent systems, excitatory glutamatergic spiny cells (which include other pyramidal cells and spiny stellate cells), and inhibitory GABAergic interneurons. The problem is that the neuronal elements that make up the basic microcircuit are differentiated into subtypes, some of which are lacking or highly modified in different cortical areas or species. Furthermore, the number of neurons contained in a discrete vertical cylinder of cortical tissue varies across species. Additionally, it has been shown that the neuropil in different cortical areas of the human, rat and mouse has a characteristic layer specific synaptology. These variations most likely reflect functional differences in the specific cortical circuits. The laminar specific similarities between cortical areas and between species, with respect to the percentage, length and density of excitatory and inhibitory synapses, and to the number of synapses per neuron, might be considered as the basic cortical building bricks. In turn, the differences probably indicate the evolutionary adaptation of excitatory and inhibitory circuits to particular functions. Introduction The appearance of the neocortex, along with its expansion and differentiation in mammals represents one of the fundamental events in the evolution of the vertebrate brain. This brain region is where faculties such as speech and thought are localized, those which distinguish humans from other mammals. Thus, understanding its structure and function is one of the prime aims in neuroscience. In general, the neocortex of all species contains a set of elements similar to that of any other part of the brain. That is, two major types of neurons (projection cells and interneurons), glia (astrocytes, oligodendrocytes and microglia), nerve fibers (extrinsic and intrinsic) and blood vessels. Furthermore, the physiological properties, neurotransmitters and neuroactive peptides, receptors and ion channels, and other compounds generally expressed by cortical neurons are not unique to the neocortex, but are found throughout the brain. Thus, one of the fundamental questions in neuroscience is, what neural substrates make a human being human? In other words, what is ∗ To whom correspondence should be addressed. 0300–4864  C 2003 Kluwer Academic Publishers special about the neocortex of humans and how does it differ from that of other species? Santiago Ramón y Cajal, made some of the most important early contributions to the study of the cortical microstructure in different species (DeFelipe & Jones, 1988) and he highlighted the need to answer these questions in his book Recuerdos de mi vida (Cajal, 1917, pp. 345–350): “At that time, the generally accepted idea that the differences between the brain of [non-human]mammals (cat, dog, monkey, etc.) and that of man are only quantitative, seemed to me unlikely and even a little offensive to human dignity. . . but do not articulate language, the capability of abstraction, the ability to create concepts, and, finally, the art of inventing ingenious instruments. . . seem to indicate (even admitting fundamental structural correspondences with the animals) the existence of original resources, of something qualitatively new which justifies the psychological nobility of Homo sapiens?. . . ’’. 300 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O In spite of the many comparative studies performed since the original findings of Cajal, it is somehow surprising that we still don’t have any real answers to this fundamental question. In this presentation we shall mainly deal with some of the structural aspects of the neocortex that are related to our knowledge of cortical microorganization. Fundamental aspects of the intrinsic organization of the neocortex The neocortex is commonly described as a 6-layered structure that is subdivided into cortical areas. These areas are distinguished by their histological and neurochemical characteristics, their connections (afferent/efferent fiber systems) and their functional properties. There are two major groups of cortical neurons: spiny neurons and smooth or sparsely spiny neurons. Spiny neurons include the population of pyramidal and spiny stellate cells that are characterized by the excitatory asymmetric synapses they form (glutamatergic). The majority of cortical neurons are pyramidal cells that are found in layers II–VI and constitute the bulk of the projection cells (commonly they are classified according to their projection site). Most spiny stellate cells are interneurons and they are located in the middle cortical layers. Smooth or sparsely spiny neurons are interneurons and include a large population of neurons that form inhibitory symmetric synapses (GABAergic), as well as a small population that are non-GABAergic. Smooth interneurons are found in layers I–VI and form a morphologically and physiologically heterogeneous group. Furthermore, a number of smooth interneurons have been shown to contain more than one neuroactive substance, commonly a neuroactive peptide. Different types of interneurons can also be recognized by their chemical and electrical synaptic connectivity. Additionally, specific morphological types, with particular molecular attributes and connections, have been shown to present distinct physiological characteristics (for reviews see Peters & Jones, 1984; Jones & Hendry, 1986; White, 1989; DeFelipe & Fariñas, 1992; DeFelipe, 1993, 2002; Nieuwenhuys, 1994; Kawaguchi & Kubota, 1997; Thomson & Deuchars, 1997; Somogyi et al., 1998; Galarreta & Hestrin, 2001; Silberberg et al., 2002; Wang et al., 2002). In general, cortical neurons are thought to be organized into multiple, small repeating microcircuits. However, as pointed out by Silberberg et al. (2002), given the great diversity of anatomical, molecular and physiological types of neurons, and the intricate connectivity, it is evident that a unique, discrete microcircuit cannot exist. However, in spite of the large number of different elements that constitute cortical circuits, it is possible to draw up a common basic microcircuit. The skeleton of this basic microcircuit is formed by a pyramidal cell and its input-output connections (DeFelipe, 2002). That is excitatory inputs (asymmetrical synapses) only arrive at the dendritic arbor and originate from extrinsic afferent systems and spiny cells (which include other pyramidal cells and spiny stellate cells). Inhibitory inputs (symmetrical synapses), which mostly originate from GABAergic interneurons, terminate on the dendrites, soma and axon initial segment. These interneurons are interconnected between themselves with the exception of chandelier cells, which only form synapses with the axon initial segment of pyramidal cells. These microanatomical characteristics have been found in all cortical areas and species examined so far and, therefore, they can be considered as fundamental aspects of cortical organization. It is clear that cortical areas show important differences within the same and different species. The contention is that, since the times of Cajal, some researchers have emphasized the similarities whereas others the differences. For the former group of researchers the histological differences are essentially fortuitous, the functional differences of the various cortical areas are a consequence of the differential connectivity of their afferent and efferent fiber systems. For the second group, the morphological differences between cortical areas are fundamental as are the differences in connectivity. Thus, the working hypothesis posed by Cajal when he started to examine the human cerebral cortex (Cajal, 1899), can still be taken as a current working hypothesis. “. . . if the gray cerebral cortex is an aggregation of organs with diverse functions, each of them must possess a special structure, within a [fundamental] plan whose general lines are appropriate for the whole cortex.’’ The basic uniformity of the neocortex and species-specific variations COLUMNAR ORGANIZATION In 1938, Lorente de Nó introduced the important idea of the “elementary cortical unit of operation.’’ He considered the cortex as consisting of small cylinders composed of vertical chains of neurons that crossed all cortical layers and that had specific afferent fibers as their axis (Lorente de Nó, 1938): “. . . all the elements of the cortex are represented in it, and therefore it may be called an elementary unit, through which, theoretically, the whole process of the transmission of impulses from the afferent fibre to the efferent axon may be accomplished.” This idea formed the basis of the hypothesis of the columnar organization of the cerebral cortex that developed later, mainly after the works of Mountcastle, Hubel and Wiesel (for reviews see Hubel & Wiesel, 1977; Cortical microstructure Mountcastle, 1978, 1997; Jones, 2000a; Buxhoeveden & Casanova, 2002). The columnar organization is mostly related to the migration of neurons from the ventricular and subventricular zones into radial columns during development (Rakic, 1988). This radial migration has been suggested as a mechanism, known as the radial unit hypothesis (Rakic, 1988), by which the neocortex could expand enormously during evolution as a sheet of cells with a basically uniform thickness rather than increasing in size as a globe (reviewed in Rakic, 2002; see also Chenn & Walsh, 2002). NUMBER AND PROPORTION OF NEURONAL TYPES WITHIN A CORTICAL COLUMN In 1980, a highly influential paper by Rockel, Hiorns and Powell emphasized the basic uniformity of the cortex (Rockel et al., 1980). These authors counted the number of neurons in tissue samples (30 µm wide by 25 µm thick) across the whole cortex, from the pial surface to the white matter from different functional areas (motor, somatic sensory, visual, frontal, parietal and temporal) of a variety of species (mouse, rat, cat, monkey and man). They found that the absolute number of neurons was very similar in all areas and in all species (approximately 110), with the exception of the binocular part of area 17 of the monkey and human where there were approximately 2.5 times more neurons. They assessed tissue columns of 30 µm width because, according to the studies of Mountcastle, Hubel and Wiesel, this was approximately the width of the smallest functional column in the neocortex. Thus, the number of neurons within this column would represent a quantitative reflection of the anatomical basis of the functional column. Rockel et al. (1980) also pointed out that the pattern of the intrinsic cortical connections, as revealed by various anatomical methods in the laboratory of Powell, and by other authors (e.g., Gilbert & Kelly, 1975; Lund et al., 1975; Jones & Wise, 1977; for a review, see Jones, 1983), was essentially uniform throughout the neocortex. They also pointed out that the cytoarchitectonic variations between different areas of the same brain and between the same areas in different species were correlated with variations in the relative sizes of the major afferent and efferent fiber pathways of each area. Early electron microscope studies identified ultraestructural characteristics that distinguish the somata of pyramidal and nonpyramidal neurons (Colonnier, 1968; Jones & Powell, 1970). On the basis of these, several authors have reported that the proportions of these two major neuronal types were the same in visual, motor and somatic sensory cortex of the monkey and in the visual and motor areas of the rat and cat (Sloper, 1973; Tömböl, 1974; Sloper et al., 1979; Winfield et al., 1980). All these features led to Rockel et al. (1980) to conclude that 301 “in mammalian evolution the area of neocortex increases in larger brains but the number of and proportions of neuronal types through the depth remains constant, except in area 17 of primates. From these and other findings it is suggested that the intrinsic structure of the neocortex is basically more uniform than has been thought and that differences in cytoarchitecture and function reflect differences in connections.’’ (see also Powell, 1981). However, using other more appropriate quantitative methods (including the disector method), it has not been possible to confirm these conclusions in the rat and cat (Beaulieu & Colonnier, 1989; Beaulieu, 1993; Skoglund et al., 1996), nor that each cortical unit contains a similar number of neurons in a variety of other species (Haug, 1987; Stolzenburg et al., 1989; see Table 1). SPECIES - SPECIFIC VARIATIONS IN CORTICAL MICROSTRUCTURE The majority of the studies of cortical microstructure have been based on the examination of relatively few species (mouse, rat, cat, macaque monkey and human). But, when other less commonly used species are considered, such as, elephants, pigmy shrew, manatees, dolphins, giraffes or apes, it is obvious that their cortical microanatomy is rather different (e.g., Haug, 1987; Reep et al., 1989; Stolzenburg, 1989; Glezer et al., 1988, 1993; Hof et al., 1999, 2000; Preuss & Coleman, 2002). For example, Dexler (1913) described clumps of neuronal cell bodies in layer VI of the cerebral cortex of dugongs and he called these aggregations “basal Rindenkerne’’ (basal cortical nuclei). These cellular aggregates were also been found in the cerebral cortex of manatees, in a region that Reep et al. (1989) called the “cluster cortex’’ to emphasize this feature. This peculiar cytoarchitecture has not been described in any other species, suggesting that it is a unique trait of the sirenia (Reep et al., 1989; see also Johnson et al., 1994). Dramatic differences in the neuron density and cytoarchitectonic organization can be seen in Nisslstained sections from the dorsal parietal cortical region of the human (somatosensory area 1), the giraffe (unchacterized cytoarchitectonic area), and the platypus (somatosensory bill area; Fig. 1). For example, in the human vertical aggregates of neurons can be distinguished that form relatively thin and long minicolumns. The giraffe neocortex is mainly characterized by a thick layer I, the presence of short and thick minicolumns or clusters (“septa’’) separated by regions with few neurons (“hollows’’) in layer V and, above all, the existence of clusters of neurons in layer II. The neocortex of the platypus is characterized by having a high density of neurons, a dense and thin layer II, and short minicolumns in layer VI (see also Elston et al., 1999b). 302 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O Table 1. Data accumulated when considering all cortical layers. Total number of synaptic profiles studied Range (mean ± s.e.m.) of asymmetrical and symmetrical synaptic profiles per 100 µm2 Range (mean ± s.e.m.) of all synaptic profiles per 100 µm2 Mean cross-sectional lengths of all synapses (µm ± s.e.m.) Synaptic density of asymmetrical and symmetrical synaptic profiles per 108 /mm3 (mean ± s.e.m.) Synaptic density of all synapses per 108 /mm3 (mean ± s.e.m.) % of asymmetrical synapses % of symmetrical synapses No. of neurons/mm3 No. of neurons per column of 30 × 25 µm through the depth of the cortex No. of synapses/neuron Human temporal cortex (n = 3) Rat hindlimb Mouse barrel somatosensory y cortex (n = 3) cortex (n = 2) Mouse visual cortex (n = 1)∗ 2195 2523 2054 897 11–18 (14.9 ± 0.7) 9–21 (17.9 ± 2.2) 22–35 (28.2 ± 1.4) 24–38 (29.4 ± 1.8) 20–36 (29.9 ± 0.79) 32–46 (38.9 ± 0.95) 50–75 (63.1 ± 1.96) 40–64 (51.3 ± 2.1) 0.28 ± 0.01 0.29 ± 0.01 0.22 ± 0.01 0.21 ± 0.09 5.42 ± 0.28 6.46 ± 0.26 12.83 ± 0.65 14.46 ± 0.93 10.94 ± 0.34 13.97 ± 0.33 29.31 ± 1.02 25.19 ± 1.16 89 11 24186 48 89 11 54468 75 84 16 120315 109 89 11 29821 18015 21983 – – – ∗ Data from mouse visual cortex were not used for statistical comparisons with data from mouse barrel cortex, nor with the rat and human cortex. All synapses include asymmetrical, symmetrical and uncharacterized synapses. However, in spite of the importance of applying comparative morphology to understand the basic organization of the neocortex (e.g., Petrides & Pandya, 1994), relatively little attention has been paid to these studies. In conclusion, the notion of a basic uniformity in the neocortex, with respect to the density and types of neurons per column is not valid for all species. Therefore, these features are not an essential or general feature of the neocortex. SPECIES - SPECIFIC VARIATIONS IN NEURONAL TYPES Immunocytochemical analyses have demonstrated important differences in the proportions of interneurons between species. For example, in the rat GABAergic cells form 15% of the population in all cortical areas whereas in the primate, they reach 20% in the visual cortex and up to 25% in other cortical areas (Hendry et al., 1987; Beaulieu et al., 1992; Beaulieu, 1993; Micheva & Beaulieu, 1995; Gabbott & Bacon, 1996; del Rĺo & DeFelipe, 1996; Gabbott et al., 1997; Meinecke & Peters, 1997; but see Ren et al., 1992). Interestingly, while Beaulieu (1993) found the same proportion of GABAergic neurons (15%) in the occipital, parietal and frontal cortex of the rat, the numerical density of neurons in the frontal cortex (34,000/mm3 ) was significantly lower than that in the occipital (52,000/mm3 ) and parietal (48,000/mm3 ) regions. The fixed proportion of interneurons, irrespective of the number of neurons, is in keeping with the idea of the uniformity of cortical Fig. 1. A, B, and C—Low-power photomicrographs of 100-µm-thick vibratome sections stained with thionin from area 1 of the primary somatosensory cortex of the human (A), dorsal parietal cortex of the giraffe (unchacterized cytoarchitectonic area; B), and the bill representation of the primary somatosensory cortex of the platypus C). D–I—Higher magnification images of the regions boxed in A, B, and C. In the human (A, D), vertical aggregations of neurons forming relatively thin and long minicolumns can be readily observed (arrow). In the giraffe (B, F, G), layer I is thick and there are clusters of neurons in layer II (arrows in B). One of these clusters is shown at greater magnification in F. In G the thick minicolumns (arrows) present in layer V are shown and it should be noted that these are separated by relatively large regions with few neurons (“hollows’’) as indicated with asterisks. In the platypus (C, E, H, I), the neocortex is characterized by the high density of neurons (E), the thin and very dense layer II (H, indicated between two lines), and the presence of short and relatively thick minicolumns (arrows) in layer VI (I). Note the high density of neurons in the platypus compared to the human and giraffe. wm, white matter. Scale bar in I: 620 µm for A, B, C; 145 µm for D–G; 100 µm for, H, I. Cortical microstructure 303 304 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O circuits. However, this does not exclude the possibility that the proportion of the different types of interneurons might vary in each cortical area, resulting in regional specialization of inhibitory circuits (see DeFelipe et al., 1999b). The higher proportion of GABA neurons in the primate cortex raises three intriguing but not mutually exclusive possibilities. Firstly, there could be an increase in the number of all types of interneurons already present in non-primate mammals. Alternatively, an increase in the number of certain types of interneurons may have occurred. Thirdly, new types of interneurons may have been added during evolution in the primate cortex. This latter idea was proposed in the times of Cajal, and he indeed thought that unique types of neurons were present in the human cortex (DeFelipe & Jones, 1988). However, it is generally thought that the same types of neurons are found in all species (e.g., Fairén et al., 1984; Tyler et al., 1998). Nevertheless, a number of studies support the theory of evolutionary diversity with respect to the morphology and neurochemical characteristics of both pyramidal cells and interneurons (e.g., Campbell & Morrison, 1989; Lewis & Lund, 1990; Glezer et al., 1993; del Rĺo & DeFelipe, 1997b; Nimchinski et al., 1999; Hof et al., 2000; Elston et al., 1999a, 2001; Preuss & Coleman, 2002; see also Benavides-Piccione & DeFelipe, 2003). For example, the morphology of pyramidal cells differs considerably between different cortical areas within a given species, and between homologous cortical areas of different species. The basal dendritic arbor of these cells displays notable differences in the size, number of bifurcations and spine density between various species of monkeys. Since each dendritic spine is assumed to receive at least one excitatory synapse, these differences indicate species variations in the cortical circuits established by pyramidal cells (see Elston et al., 1999a). Another example is a type of projection neuron that can be distinguished by their spindle-shaped morphology (spindle cells) and that are only found in the cortex of humans and in great apes (Nimchinski et al., 1999). With regards interneurons, a remarkable case of interspecies variation is that of the so-called double bouquet cell, the source of a large number of inhibitory synapses on dendrites within a very narrow column of cortical tissue. These interneurons have been shown to form a widespread and regular microcolumnar structure (DeFelipe et al., 1990; del Rĺo & DeFelipe, 1997a), which is strikingly similar to that shown by bundles of myelinated axons (radial fasciculi) that originate from small vertical aggregates of pyramidal cells (radial minicolumns or pyramidal cell modules). Using immunocytochemical techniques to label double bouquet cells in tangential sections, a high coincidence of bundles of myelinated axons and double bouquet cell axons was found, such that one double bouquet cell is found per pyramidal cell module (del Rĺo & DeFelipe, 1997a). Thus, double bouquet cells seem to constitute a key element of the minicolumnar organization of the cortex (DeFelipe et al., 1990; del Rĺo & DeFelipe, 1997a; Peters & Sethares, 1997; reviewed in DeFelipe, 1997b, 2002; Jones, 2000a). However, while the morphology and distribution of double bouquet cells is similar in the human and macaque neocortex, they are modified or less numerous in the neocortex of other species (e.g., the cat), and may even be absent (e.g., in the mouse and rat). Thus, differences in the morphology, number or distribution of double bouquet cells may represent fundamental differences in cortical microorganization between primates and other species (DeFelipe, 2002). Recently, it has been proposed that changes in the distribution of transcription factors in the telencephalon may generate species-specific programs of GABAergic neurogenesis (Letinic et al., 2002). These authors showed that two lineages of human cortical GABAergic neurons exist. One lineage, that expresses the transcription factors Dlx1/2 and Mash1, makes up 65% of the GABAergic neurons. These cells originate in the ventricular and subventricular zones of the dorsal telencephalon and migrate radially across the intermediate zone to the cerebral cortex. The other lineage, the remaining 35% of GABAergic neurons, express Dlx1/2 but not Mash1, and originate in the ganglionic eminence of the ventral telencephalon from where they migrate tangentially across the intermediate zone to the cerebral cortex. However, in rodents the origin of GABAergic neurons differs since all, or almost all, GABAergic neurons originate in the ganglionic eminence and migrate tangentially to the cerebral cortex (Tan et al., 1998; Anderson et al., 1999; Marin et al., 2001). Thus, we proposed that the double bouquet cells in the human neocortex might originate in the ventricular/subventricular zone from the GABAergic lineage expressing Dlx1/2 and Mash1, and migrate radially along common pathways as the radial units or ontogenetic columns that give rise to the pyramidal cell modules (reviewed in Rakic, 2002). This could represent a mechanism acquired during evolution by which the complementary microcolumnar structures formed by double bouquet cells (and probably, other types of interneurons), and pyramidal cell modules, might be constructed in the primate, but not in the rodent neocortex. Synaptology of the neuropil: Morphology and density of synapses GENERAL CONSIDERATIONS Generally, cortical synapses are divided into two morphological categories: asymmetrical (type 1) and symmetrical (type 2) synapses (Gray, 1959; Colonnier, Cortical microstructure 1968, 1981; Peters, 1987; Peters et al., 1991; Peters & Palay, 1996). These a most easily distinguished by the postsynaptic density, which is prominent in asymmetrical synapses and thin in symmetrical synapses. Differences or similarities in the density of synapses have been reported in various cortical areas and species (see section “Density of synapses’’). However, the number of specimens studied and the extension of cortex examined is frequently relatively small and in many studies, the cortical layers have not been examined systematically. Other considerations, such as the age of the individuals, may also differ considerably between studies, and often this information is not available. This is particularly important since notable decreases in the density of synapses may occur during puberty and throughout adulthood, depending on the cortical area, layer and species examined (reviewed in Rakic et al., 1994; see also Bourgeois et al., 1994; Poe et al., 2001; DeFelipe et al., 1997, 2002). Furthermore, there is no general consensus for counting synapses and, therefore, it is often difficult to compare data obtained in different laboratories (DeFelipe et al., 1999a). Here we have reviewed some of these issues and compared them with data we have obtained regarding the differences in the cortical circuits of distinct species and cortical areas. We have studied the morphology and density of synapses, in relation to their depth in the cortex (from layer I to layer VI), in the adult mouse (3 months old), rat (5 months old) and human (normal cortex from surgically resected brain tissue from three male individuals: 24, 27 and 36 years of age). Some of this material has been used in previous studies (DeFelipe et al., 1997, 2002; Marco & DeFelipe, 1997). We applied the same method to quantify synapses in the thin neuropil (i.e., avoiding the neuronal and glial somata, blood vessels, large dendrites and myelinated axons) of the visual cortex and the somatosensory cortex of the mouse (area 17 and barrel cortex, respectively), the rat hindlimb area, and the anterolateral temporal cortex (T2–T3) of the human (Fig. 2). Furthermore, the neuronal density was also calculated in this material to compare and estimate the number of synapses per neuron in each cortical layer and species. Table 1 data from the mouse visual cortex were not used for statistical comparisons, therefore, unless otherwise specified, the mouse cortex refers to the barrel cortex. For more details of the methods used, see section “methods” on page 312. MORPHOLOGY OF SYNAPSES Types of synapses In most electron microscope studies, the percentage of asymmetrical and symmetrical synapses varied between 80–90% and 20–10%, respectively, in all the cortical areas and species examined (e.g., Blue & Parnavelas, 305 1983; Rakic et al., 1986; Zecevic et al., 1989; Schüz & Palm, 1989; Zecevic & Rakic, 1991; Beaulieu et al., 1992, 1994; Bourgeois et al., 1994; Beaulieu & Colonnier, 1985, 1989; Micheva & Beaulieu, 1996; DeFelipe et al., 1997, 2002; White et al., 1997; Marco & DeFelipe, 1997). Our data confirmed the remarkable consistency in the percentage of asymmetric and symmetric synapses in the human cortex (89% and 11%), rat (89% and 11%) and mouse (84% and 16%) when considering all layers together (Table 2). However, when each layer was analyzed individually, a number of differences were found. For instance, in all layers the percentage of asymmetrical synapses was generally lower in the mouse than in the rat and human, except for layers I, II and IIIa of human and layer I of rat (Table 2). The greatest difference was found in layer IV, where 16% and 14% fewer asymmetrical synapses were seen in the mouse when compared to rat and human, respectively. The maximum difference between layers in the rat and human was only 3%. A given axon terminal forms an asymmetrical or symmetrical synapse with either one (single synapse) or with two or more postsynaptic elements (multiple synapses). The large majority of axon terminals form single synapses with a postsynaptic element in the three species examined (human, 99.7%; rat, 99%; mouse, 97.6%), but notable differences between species were found in certain layers. For example, the greatest proportion of multiple synapses were found in layer IV in each of the three species, the proportions being 0.7% in the human, 1.6% in the rat and 4.5% in the mouse (8 million, 29 million and 173 million multiple synapses per mm3 , respectively). While in layer I, no axon terminals formed multiple synapses in the human cortex, whereas in the rat and mouse there were 0.7% and 0.9%, respectively (12 million and 13 million multiple synapses per mm3 ). CROSS - SECTIONAL LENGTH OF SYNAPTIC JUNCTIONS With few exceptions, the cross-sectional length of synaptic junctions in all cortical areas and species studied so far varies between 0.20–0.40 µm (e.g., Blue & Parnavelas, 1983; Beaulieu & Colonnier, 1985, 1989; Schüz & Palm, 1989; Glezer & Morgane, 1990; Beaulieu et al., 1992, 1994; Huttenlocher & Dabholkar, 1997; White et al., 1997; DeFelipe et al., 1997, 2002; Marco & DeFelipe, 1997). As far as we know, no extensive quantitative electron microscope studies have determined whether interspecies differences might exist between the mean cross-sectional lengths of synaptic junctions in the various cortical layers. Therefore, we have used our material to compare between the human, rat and mouse (Fig. 3 and Table 3). The mean cross-sectional lengths of asymmetrical synapses were significantly shorter in all layers of the mouse cortex when compared to the human and rat (mean length of all layers: 0.30 µm, 0.30 µm 306 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O Fig. 2. Electron micrographs illustrating the thin neuropil of layer IIIa in the human temporal cortex, and of layer II/III of the mouse barrel cortex. Note the higher density of synapses (some of them indicated by arrows) in the mouse cortex. Scale bar: 0.5 µm. Cortical microstructure 307 Table 2. Thickness of layers, number of neurons and of synaptic profiles in the human, rat and mouse cerebral cortex. Number neurons/ mm3 (mean ± s.e.m.) Layer Thickness (µm; mean ± s.e.m.) I II IIIa IIIb IV V VI I–VI 235 ± 13.5 295 ± 10.5 405 ± 16.5 370 ± 14.1 285 ± 10.2 552 ± 34.0 480 ± 26.7 2622 8333 ± 1531 45563 ± 3010 20964 ± 2709 15090 ± 1804 46167 ± 4073 23076 ± 1734 16774 ± 1875 24186 I II–III IV Va Vb VI I–VI 123 ± 4.9 457 ± 9.4 152 ± 7.0 209 ± 10.2 321 ± 8.8 565 ± 11.5 1827 3472 ± 1273 61670 ± 3996 90965 ± 5911 44868 ± 3450 35536 ± 3068 64286 ± 4520 54483 I II–III IV∗ V VI I–VI 69 ± 3.7 235 ± 9.9 208 ± 4.6 248 ± 6.1 451 ± 14.1 1210 18229 ± 2915 137645 ± 6410 181362 ± 6142 77765 ± 6282 122092 ± 7161 120315 Number AS Human 136 147 144 171 136 127 110 971 Rat 158 206 192 201 172 169 1098 Mouse 199 178 145 125 125 772 Number SS Number UC Number all synapses % AS % SS 29 19 19 20 15 14 10 126 176 181 174 186 155 142 84 1098 341 347 337 377 306 283 204 2195 82.4 88.6 88.3 89.5 90.1 90.1 91.7 88.5 17.6 11.4 11.7 10.5 9.9 9.9 8.3 11.5 40 22 17 24 19 10 132 296 241 215 192 184 165 1293 494 469 424 417 375 344 2523 79.8 90.4 91.9 89.3 90.1 94.4 89.3 20.2 9.6 8.1 10.7 9.9 5.6 10.7 35 22 43 23 23 146 264 231 285 191 165 1136 498 431 473 339 313 2054 85.0 89.0 77.1 84.5 84.5 84.1 15.0 11.0 22.9 15.5 15.5 15.9 AS, asymmetrical synaptic profiles; SS, symmetrical synaptic profiles; UC, uncharacterized synaptic profiles. ∗ Values include data from the centers or ”hollows” and the septa of the barrels, according to their relative volumes. and 0.23 µm, in the human, rat and mouse). No significant differences were found between humans and rats except in layer V of the human where asymmetrical synapses were shorter (0.27 µm) compared to layer Va (0.31 µm) and Vb (0.30 µm) of the rat. With regards to symmetrical synapses, they were also generally smaller in the mouse when compared to the human and rat, and longer in the rat than in the human (mean all layers: 0.25 µm, 0.28 µm and 0.21 µm in the human, rat and mouse). Significant differences were found between human, rat, and mouse in all layers, except in layer I, depending on the layer and species (Fig. 3). Furthermore, within a given species, there were significant differences in the size of symmetrical synapses in the human between all layers except I, IV and VI, and in the size of asymmetrical synapses in the mouse between all layers except layer V. In summary, laminar-specific differences in the cross-sectional length of synaptic junctions can be observed between different species and within certain layers of the same species. DENSITY OF SYNAPSES A pioneer in comparative ultrastructural studies of the neocortex, Cragg (1967) observed that there was rela- tively little variation in synaptic density between areas that were so cytoarchitectonically and functionally different as the motor and visual cortex of the mouse and monkey (between 6 and 9 × 108 synapses/mm3 ). From their own data, and that of Cragg and other authors, Colonnier and colleagues argued that the numerical density of synapses was relatively constant throughout the cortical layers, as well between different cortical areas and different species (see O’kusky & Colonnier, 1982; Beaulieu & Colonnier, 1989). Similarly, Schüz and Palm (1989) reported no systematic differences between layers or different cortical regions in the mouse, including areas 6, 8 and 17. A remarkable narrow variability, 15 to 20 synaptic profiles per 100 µm2 , was found in the neuropil of all cortical layers by Rakic and colleagues (considering both asymmetrical and symmetrical synapses) in diverse regions of the rhesus monkey neocortex, including the motor cortex, somatosensory cortex, prefrontal cortex and visual cortex (Rakic et al., 1986; see also Zecevic & Rakic, 1991). Synaptic densities were also found to be similar in the human visual, auditory and prefrontal cortex Huttenlocher and Dabholkar (1997). This uniformity in synaptic density led O’kusky and Colonnier (1982) to propose that it probably reflects the optimal number of synapses, and that it may due to 308 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O Fig. 3. Comparison of the mean cross-sectional lengths (± s.e.m.) and densities of synaptic profiles between human, rat and mouse. Density values obtained in layers II, IIIa and IIIb of human, and layers Va and Vb of rat were recalculated according to the relative thickness of these layers to estimate the representative values of layers II–III and V, respectively. Cortical microstructure 309 Table 3. Measured length (µm) of synaptic profiles in the human, rat and mouse cerebral cortex. Table 4. Number of synapses (× 108 ) per mm3 (Nv) in the neuropil in the human, rat and mouse cerebral cortex. Layer AS I II IIIa IIIb IV V VI I–VI 0.29 ± 0.01 0.31 ± 0.01 0.31 ± 0.01 0.29 ± 0.01 0.29 ± 0.01 0.27 ± 0.01 0.30 ± 0.01 0.30 ± 0.01 I II–III IV Va Vb VI I–VI 0.30 ± 0.01 0.30 ± 0.01 0.29 ± 0.01 0.31 ± 0.01 0.30 ± 0.01 0.29 ± 0.01 0.30 ± 0.01 I II–III IV∗ V VI I–VI 0.24 ± 0.01 0.26 ± 0.01 0.21 ± 0.01 0.24 ± 0.01 0.20 ± 0.01 0.23 ± 0.01 SS Human 0.25 ± 0.01 0.30 ± 0.02 0.29 ± 0.02 0.22 ± 0.01 0.23 ± 0.02 0.20 ± 0.02 0.22 ± 0.02 0.25 ± 0.01 Rat 0.27 ± 0.01 0.28 ± 0.02 0.28 ± 0.02 0.30 ± 0.02 0.30 ± 0.04 0.29 ± 0.02 0.28 ± 0.01 Mouse 0.22 ± 0.01 0.23 ± 0.02 0.20 ± 0.01 0.22 ± 0.02 0.18 ± 0.02 0.21 ± 0.01 UC All synapses Layer Nv AS 0.25 ± 0.01 0.29 ± 0.01 0.27 ± 0.01 0.29 ± 0.01 0.25 ± 0.01 0.27 ± 0.01 0.27 ± 0.01 0.27 ± 0.01 0.27 ± 0.01 0.30 ± 0.01 0.29 ± 0.01 0.29 ± 0.01 0.27 ± 0.01 0.26 ± 0.01 0.28 ± 0.01 0.28 ± 0.01 I II IIIa IIIb IV V VI I–VI 4.78 ± 0.65 4.80 ± 0.60 4.61 ± 0.61 5.75 ± 0.65 4.53 ± 0.40 4.65 ± 0.47 3.56 ± 0.43 4.67 ± 0.21 0.27 ± 0.01 0.30 ± 0.01 0.28 ± 0.01 0.29 ± 0.01 0.29 ± 0.01 0.28 ± 0.01 0.28 ± 0.01 0.28 ± 0.01 0.30 ± 0.01 0.29 ± 0.01 0.30 ± 0.01 0.29 ± 0.01 0.28 ± 0.01 0.29 ± 0.01 I II–III IV Va Vb VI I–VI 5.20 ± 0.48 6.57 ± 0.42 6.40 ± 0.57 6.36 ± 0.50 5.52 ± 0.35 5.72 ± 0.45 5.96 ± 0.19 0.21 ± 0.01 0.24 ± 0.01 0.19 ± 0.01 0.21 ± 0.01 0.20 ± 0.01 0.21 ± 0.01 0.22 ± 0.01 0.25 ± 0.01 0.20 ± 0.01 0.22 ± 0.01 0.20 ± 0.01 0.22 ± 0.01 I II–III IV∗ V VI I–VI 12.72 ± 1.07 10.60 ± 0.77 11.33 ± 0.76 8.16 ± 0.86 9.99 ± 1.09 10.56 ± 0.43 Nv SS Nv UC Human 1.18 ± 0.22 6.95 ± 0.58 0.68 ± 0.17 6.17 ± 0.62 0.65 ± 0.15 6.36 ± 0.51 0.91 ± 0.23 6.28 ± 0.46 0.65 ± 0.16 6.03 ± 0.53 0.72 ± 0.21 5.22 ± 0.51 0.45 ± 0.16 3.07 ± 0.39 0.75 ± 0.07 5.73 ± 0.21 Rat 1.52 ± 0.19 10.77 ± 0.60 0.81 ± 0.16 7.77 ± 0.46 0.62 ± 0.15 7.41 ± 0.63 0.82 ± 0.15 6.40 ± 0.45 0.73 ± 0.23 6.25 ± 0.42 0.35 ± 0.10 5.72 ± 0.60 0.80 ± 0.07 7.39 ± 0.25 Mouse 2.51 ± 0.41 18.95 ± 1.53 1.54 ± 0.25 14.84 ± 0.84 3.54 ± 0.52 23.76 ± 1.05 1.67 ± 0.34 13.81 ± 1.43 2.14 ± 0.71 13.04 ± 1.70 2.27 ± 0.22 16.85 ± 0.72 Nv all synapses 12.67 ± 1.13 11.35 ± 0.97 11.44 ± 0.90 12.73 ± 0.92 11.02 ± 0.74 10.34 ± 0.62 7.00 ± 0.54 10.94 ± 0.34 17.23 ± 0.70 15.02 ± 0.76 14.24 ± 0.96 13.43 ± 0.72 12.29 ± 0.62 11.68 ± 0.74 13.97 ± 0.33 33.73 ± 1.77 26.56 ± 1.32 38.44 ± 1.09 23.28 ± 2.26 24.78 ± 2.55 29.31 ± 1.02 AS. asymmetrical synaptic profiles; SS, symmetrical synaptic profiles; UC, uncharacterized synaptic profiles. ∗ Values were obtained in the centers or “hollows” of the barrels. AS, asymmetrical synaptic profiles; SS, symmetrical synaptic profiles; UC, uncharacterized synaptic profiles. ∗ Values were obtained in the centers or “hollows” of the barrels. some limiting metabolic or structural factor (see also Rakic et al., 1986). However, the methods used by the different authors to estimate synapse number are rather different, including whether or not stereological methods were used. In addition, most comparisons are only qualitative and not based on statistical analysis. For instance, Beaulieu et al. (1992, 1994) found significant differences in the number of synapses per volume between certain layers of both the rat and monkey visual cortex when using the disector method and statistical comparisons. We have found that when considering all types of synapses, the synaptic density considering all layers was lower in the human (1094 million/mm3 ), than in the rat (1397 million/mm3 ) and mouse (2931 million/mm3 ) (Tables 1 and 4). These differences were statistically significant in most layers among the three species (Fig. 3). Moreover, the high density of synapses found in the mouse barrel cortex when compared to the rat and human, was not a peculiarity of this cortical region of the mouse, since the mouse visual cortex also showed a high density (2519 million/mm3 ; Table 1). When asymmetrical and symmetrical synapses were analyzed separately, a significantly higher density was also observed for asymmetrical synapses in all mouse cortical layers when compared the rat and human, and in layers IV and VI of the rat when compared to human. The density of symmetrical synapses was also significantly higher in all layers of the mouse compared the rat and human, but no significant differences were observed between the human and rat. Layer IV demonstrated the greatest differences in symmetrical synapse density between humans and rats when compared to the mouse (65 million/mm3 in the human, 62 million/mm3 in the rat and 354 million/mm3 in the mouse). In conclusion, the density of synapses is not uniform throughout the cortical layers and there are laminar specific differences between different cortical areas in different species. Thus, if an optimal number of synapses for cortical circuits exists, it must be species specific. NUMBER OF SYNAPSES PER NEURON Since the thickness and neuronal density in different layers and species may differ (Table 2, Fig. 4), comparing the density of synapses between species alone is difficult to interpret in terms of connectivity. Therefore, it is common to divide the synaptic density by the corresponding neuronal densities in the same brains 310 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O Fig. 4. Comparison of the means of the laminar thickness (± s.d.) and neuronal densities (± s.e.m.) between human, rat and mouse. The values obtained in layers II, IIIa and IIIb of human, and layers Va and Vb of rat were recalculated according to the relative thickness of these layers to estimate the representative values of layers II–III and V, respectively. to estimate the average number of synapses on each neuron. According to Cragg (1967), this idea originated from the observations of the histologists, Nissl (1898) and Von Economo (1926). Nissl pointed out that in the mole and dog, cortical neurons were more crowded than in man. Based on this observation Von Economo proposed that the greater separation between neurons the richer the fiber plexus will be between them, increasing the opportunity for neuronal interactions. Thus, a wider separation of neurons in humans compared to other species could be taken as an indication of a greater refinement of the connections between neurons. Nevertheless, as dendrites, particularly of pyramidal cells, may cross several layers, the number of synapses per neuron in a given layer is not an accurate estimate of the number of synapses received by neurons in that layer. The most obvious pitfall can be seen layer I, where the number of synapses per neuron is very high (Table 5). However, it is known that most postsynaptic dendrites in layer I, are the apical dendritic tufts of the underlying pyramidal cells. In addition, axon terminals in a given layer may originate from local neurons or from neurons located in other layers or in other cortical areas or subcortical nuclei. Thus, the synapse/neuron ratio should be taken as a useful parameter to compare between cortical areas and species in terms of “general’’ connectivity. Using this approach, Cragg (1967) compared the motor and visual areas of the mouse and macaque monkey, and he found an inverse relationship between neuronal density and the number of synapses per neuron. Furthermore, he found an opposite tendency in the ratio of synapses per neuron between specific areas in both species. In the motor cortex, the density of neurons was lower and the number of synapses per neuron was much greater in the monkey than in the mouse (16.1 × 106 neurons/cm3 ; 60000 synapses/neuron vs. 64.4 × 106 neurons/cm3 ; 13000 synapses/neuron). On the contrary, in the visual cortex there was a greater density of neurons and a lower synapse/neuron ratio in the monkey than in the mouse (110.3 × 106 neurons/cm3 ; 5600 synapses/neuron vs. 92.4× 106 neurons/cm3 ; 7000 synapses/neuron). Unfortunately, as pointed out earlier, comparison with other studies is rather difficult. For example, an inverse relationship between neuronal density and the number of synapses per neuron was also observed for individual layers of the macaque visual cortex by O’Kusky and Colonnier (1982), corroborating the findings of Cragg. However, O’Kusky and Colonnier’s estimates were approximately 60% lower than Cragg (2300 synapses/neuron) in the macaque visual cortex. In any case, O’Kusky and Colonnier suggested that this inverse relationship was due to a limiting factor such that neurons receiving more synapses would have a more complex dendritic arborization, increasing the distance between their cell bodies. In contrast, neurons receiving fewer synapses would have a less complex dendritic arbor, allowing them to be more densely packed. We found here that the principle of the inverse relationship between neuronal density and number of synapses per neuron held true in the human when compared with the rat and mouse. The density of neurons (neurons/mm3 in layers I–VI) was lower in human (24186) than in the rat (54483) and mouse (120315), and the number of synapses per neuron was higher in human (29807) than in rat (18018) and mouse (21133; Table 5; Fig. 5). However, the rat showed a lower density of neurons and synapses/per neuron than mouse and thus, this principle does not appear to be generally applicable. Additionally, when specific layers were compared, this rule was applicable to all layers in the Cortical microstructure 311 Table 5. Number of synapses per neuron in the human, rat and mouse cerebral cortex. Values were obtained by dividing the density of synapses by the density of neurons. Uncharacterized synapses were included in the asymmetrical and symmetrical types, according to the frequency of both types of synapses in each layer. Layer No. AS per neuron I II IIIa IIIb IV V VI I–VI 83883 14886 32573 50266 14379 26806 25482 26096 I II–III IV Va Vb VI I–VI 286409 15992 10493 19484 22588 12545 16127 I II–III IV V VI I–VI 143438 15647 14952 23068 15561 17595 No. SS per neuron Human 19188 2007 4436 6951 1806 3561 2811 3711 Rat 76766 1831 969 2428 2734 756 1891 Mouse 26593 2086 4524 4444 3093 3538 No. all synapses per neuron 103071 16894 37009 57217 16186 30367 28293 29807 363175 17823 11462 21912 25322 13301 18018 170031 17733 19476 27512 18654 21133 AS, asymmetrical synaptic profiles; SS, symmetrical synaptic profiles. rat, but not in the human or mouse (compare Tables 2 and 5). In addition, the laminar distribution of the number of synapses per neuron did not follow a common pattern in the three species. Similarly, there was no common pattern between species or within a given species with respect to the distribution of synapses per neuron when considering asymmetrical and symmetrical synapses (Table 5 and Fig. 5). In human, the ratio of asymmetrical synapses per neuron was higher in all layers when compared to the rat and mouse, except in layer IV of the human which showed a slightly lower synaptic density per neuron than in the mouse (Fig. 5; Table 5). Rat and mouse showed similar ratios in layers II–III, but in the rest of layers the ratio was higher in the mouse than in the rat, particularly in layer IV (42%). For symmetrical synapses, human and mouse displayed very similar values when considering all layers, approximately 50% higher than the rat. However, there were dramatic laminar differences not only between certain layers of the human and mouse when compared to the rat, but also between the human and mouse (Fig. 5 and Table 5). Fig. 5. Graphs showing the number of synapses per neuron in the human, rat and mouse. The values obtained in layers II, IIIa and IIIb of human, and layers Va and Vb of rat were recalculated according to the relative thickness of these layers to estimate the representative values of layers II–III and V, respectively. For example, the most noticeable variation between the three species was found in layer IV, where the mouse showed a considerably higher number of symmetrical synapses/neuron than in the human and in the rat. Another interesting difference was found in layers II–III, where this ratio was by far higher in human than in mouse and rat (55 and 61%, respectively), but the rodents showed similar values. In summary, examination of the synaptology of the human, rat and mouse neuropil revealed dramatic 312 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O differences in the density of synapses between the three species. Regarding, the proportion, length and density of asymmetrical and symmetrical synapses, and the ratios of asymmetrical and symmetrical synapses per neuron, there were also notable laminar specific differences between human, rat and mouse, which did not necessarily affect the same layers. Furthermore, the density of synapses was not inversely correlated with the density of neurons in the three species. Thus, certain general features of the neocortical synaptology are applicable to the human, rat and mouse, but we also detect significant differences and this means that the pattern of synaptic organization is characteristic of each cortical area and species. Functional significance of the species differences in the synaptology of the neuropil The main sources of asymmetrical synapses are the corticortical and thalamocortical axons, and the local axon collaterals of pyramidal cells and spiny stellate cells, which are known to be excitatory. In contrast, the main sources of symmetrical synapses are the inhibitory GABAergic interneurons (Houser et al., 1984; White, 1989; Peters et al., 1991; DeFelipe & Fariñas, 1992; Peters & Palay, 1996; Conti & Weinberg, 1999; Jones, 2000b; Amitai, 2001). Furthermore, synaptic size plays an important role in the functional properties of synapses (Mackenzie et al., 1999; Schikorski & Stevens, 1999; Takumi et al., 1999; Kubota & Kawaguchi, 2000; Lüscher et al., 2000). For example, larger synapses seem to contain a greater number of postsynaptic receptors (Mackenzie et al., 1999) and are associated with a greater number of docked synaptic vesicles (Schikorski & Stevens, 1999). We have shown that in each species, the cortical neuropil has its own characteristic layer-specific synaptology. Thus, differences in the density, proportion and size of excitatory and inhibitory synapses among cortical areas or species probably reflects the functional differences of cortical the circuits involved. Finally, there is approximately a 10% increase in the proportion of GABA interneurons in primates when compared to rats (see section “Species-specific variations in neuronal types’’). This must be considered in conjunction with the fact that certain subtypes of interneurons are lacking or greatly modified in some species. If the species differences in the number of synapses per neuron and, therefore, in the synaptic weights, are also taken into account, then this might serve to emphasize the variability in the design of microcircuits between cortical areas and species. The laminar specific similarities between the human, rat and mouse and other species with respect to the percentage, length and density of asymmetrical and symmetrical synapses, and in the number of synapses per neuron, might be considered as basic bricks of cortical organization. In contrast, the differences probably indicate evolutionary adaptations of excitatory and inhibitory circuits to particular functions. Methods Synaptic density per unit area (NA) was estimated from ten electron microscope samples of neuropil from each layer and from each species, using a correlative light and electron microscope technique (for a detailed description see DeFelipe et al., 1999a). The numerical density of synapses per unit volume of the neuropil was calculated using the formula NV = NA/d where NA is the number of synaptic profiles per unit area and d the average cross-sectional length of synaptic junctions. The cross-sectional lengths of synaptic junctions (synaptic apposition length) of all synaptic profiles were measured in the prints using a magnetic tablet (SummaSketch III) and the Scion Image analysis program (Scion corporation, Frederick, Maryland, USA). Statistical comparisons of the means were carried out using ANOVA. When the overall ANOVA was significant, a post-hoc Bonferroni’s test was applied. All these studies were performed with the aid of the SPSS statistical package (SPSS Science, Chicago, IL, USA). Neuronal density was estimated in all layers of each cortical area where synaptic counting was performed using optical disectors as described by West and Gundersen (1990; see also Williams & Rakic, 1988). Ten optical disectors per layer and per case were performed in 100 µm Nissl stained sections adjacent to those used for synaptic counting. Optical disectors were performed using X100 oil objective on a surface of 2400 µm2 with a depth of 40 µm, rendering a study volume of 96000 µm3 per optical disector. Nucleoli for humans and rats, and nuclei for mice were counted to obtain an estimation of the number of neurons. The thickness of the layers was calculated from ten measurements from each layer and case. The total density of neurons and synapses of each area was calculated from the relative density of each layer as a function of its thickness. Processing the tissue for electron microscopy produced a greater shrinkage of the tissue compared to Nissl stained sections. In order to obtain homogeneous estimations of the density of neurons and synapses, this difference in shrinkage was evaluated by measuring the cortical surface in adjacent sections processed for electron microscopy and Nissl, using Scion Image analysis. The shrinkage of the cortical tissue for the material processed for electron microscopy was as follows: 32.8% in human; 26.4% in rat, and 8.1% in mouse. The number of synapses per neuron in each layer was obtained by dividing the density of synapses by the density of neurons for each layer after correction considering the shrinkage. For interspecies comparisons, the values of neuron and synaptic densities of layers II, IIIa and IIIb of human, and layers Va and Vb of rat were recalculated according to their relative thickness, to obtain values representing layer II–III and V, respectively. Acknowledgments We would like to thank Alberto Muñoz for critically reading the manuscript, and C. Nilesh Patel and Paul Manger for kindly supply brain tissue from the giraffe Cortical microstructure and platypus, respectively. This work was supported by grants from the Spanish Ministry of Science and Technology (DGCYT PM99-0105) and the Comunidad de Madrid (08.5/0027/2001.1). L A-N is supported by fellowship from the Spanish Ministry of Science and Technology (FP2000-4989). References AMITAI, Y. (2001) Thalamocortical synaptic connections: Efficacy, modulation, inhibition and plasticity. Reviews of Neuroscience 12, 159–173. ANDERSON, S., MIONE, M., YUN, K. & RUBENSTEIN, J. L. (1999) Differential origins of neocortical projec- tion and local circuit neurons: Role of Dlx genes in neocortical interneuronogenesis. Cerebral Cortex 9, 646– 654. BEAULIEU, C. (1993) Numerical data on neocortical neurons in adult rat, with special reference to the GABA population. Brain Research 609, 284–292. BEAULIEU, C. & COLONNIER, M. (1985) A laminar analysis of the number of round-asymmetrical and flatsymmetrical synapses on spines, dendritic trunks, and cell bodies in area 17 of the cat. Journal of Comparative Neurology 231, 180–189. BEAULIEU, C. & COLONNIER, M. (1989) Number and size of neurons and synapses in the motor cortex of cats raised in different environmental complexities. Journal of Comparative Neurology 289, 178–187. BEAULIEU, C., CAMPISTRON, G. & CREVIER, C. (1994) Quantitative aspects of the GABA circuitry in the primary visual cortex of the adult rat. Journal of Comparative Neurology 339, 559–572. BEAULIEU, C., KISVARDAY, Z., SOMOGYI, P., CYNADER, M. & COWEY, A. (1992) Quanti- tative distribution of GABA-immunopositive and -immunonegative neurons and synapses in the monkey striate cortex (area 17). Cerebral Cortex 2, 295–309. BENAVIDES-PICCIONE, R. & DEFELIPE, J. (2003) Different populations of tyrosine hydroxylaseimmunoreactive neurons defined by differential expression of nitric oxide synthase in the human temporal cortex. Cerebral Cortex 13, 297–307. BLUE, M. E. & PARNAVELAS, J. G. (1983) The formation and maturation of synapses in the visual cortex of the rat. II. Quantitative analysis. Journal of Neurocytology 12, 697–712. BOURGEOIS, J.-P. & RAKIC, P. (1993) Changes of synaptic density in the primary visual cortex of the macaque monkey from fetal to adult stage. Journal of Neuroscience 13, 2801–2820. BOURGEOIS, J.-P., GOLDMAN-RAKIC, P. S. & RAKIC, P. (1994) Synaptogenesis in the prefrontal cortex of rhe- sus monkey. Cerebral Cortex 4, 78–96. BUXHOEVEDEN, D. P. & CASANOVA, M. F. (2002) The minicolumn hypothesis in neuroscience. Brain 125, 935–951. CAJAL, S. R. (1899) Estudios sobre la corteza cerebral humana II: Estructura de la corteza motriz del hombre y mamı́feros superiores. Revista Trimestral Microgáfica 4, 117–200. Translated in DEFELIPE, J. & JONES, E. G. 313 (1988). Cajal on the Cerebral Cortex. New York: Oxford University Press. CAJAL, S. R. (1917) Recuerdos de mi vida, Vol. 2: Historia de mi labor cientı́fica. Madrid: Moya. CAMPBELL, M. J. & MORRISON, J. H. (1989) Monoclonal antibody to neurofilament protein (SMI-32) labels a subpopulation of pyramidal neurons in the human and monkey neocortex. Journal of Comparative Neurology 282, 191–205. CHENN, A. & WALSH, C. A. (2002) Regulation of cerebral cortical size by control of cell cycle exit in neural precursors. Science 297, 365–369. COLONNIER, M. (1968) Synaptic patterns on different cell types in the different laminae of the cat visual cortex. An electron microscope study. Brain Research 9, 268– 287. COLONNIER, M. (1981) The electron-microscopic analysis of the neuronal organization of the cerebral cortex. In The Organization of the Cerebral Cortex (edited by SCHMITT, F. O., WORDEN, F. G., ADELMAN, G. & DENNIS, S. G. ) pp. 125–152. Cambridge: MIT Press. CONTI, F. & WEINBERG, R. J. (1999) Shaping excita- tion at glutamatergic synapses. Trends in Neuroscience 22, 451–458. CRAGG, B. G. (1967) The density of synapses and neurones in the motor and visual areas of the cerebral cortex. Journal of Anatomy 101, 639–654. DEFELIPE, J. (1993) Neocortical neuronal diversity: Chemical heterogeneity revealed by co-localization studies of classic neurotransmitters, neuropeptides, calcium binding proteins and cell surface molecules. Cerebral Cortex 3, 273–289. DEFELIPE, J. (1997) Microcircuits in the brain. In Lecture Notes in Computer Science, Vol. 1240 (edited by MIRA, J., MORENO-D ÍAZ, R. & CABESTANY, J. ) pp. 1–14. Berlin: Springer. DEFELIPE, J. (2002) Cortical interneurons: From Cajal to 2001. In Changing Views of Cajal’s Neuron (edited by AZMITIA, E., DEFELIPE, J., JONES, E. G., RAKIC, P. & RIBAK, C. ) Progress in Brain Research 136, 215–238. DEFELIPE, J. & JONES, E. G. (1988) Cajal on the Cerebral Cortex. New York: Oxford University Press. DEFELIPE, J. & FARI ÑAS, I. (1992) The pyramidal neuron of the cerebral cortex: Morphological and chemical characteristics of the synaptic inputs. Progress in Neurobiology 39, 563–607. DEFELIPE, J., HENDRY, S. H. C., HASHIKAWA, T., MOLINARI, M. & JONES, E. G. (1990) A microcolum- nar structure of monkey cerebral cortex revealed by immunocytochemical studies of double bouquet cell axons. Neuroscience 37, 655–673. DEFELIPE, J., MARCO, P., FAIR ÉN, A., & JONES, E. G. (1997) Inhibitory synaptogenesis in mouse somatosensory cortex. Cerebral Cortex 7, 619–634. DEFELIPE, J., MARCO, P., BUSTURIA, I. & MERCHANPEREZ, A. (1999a) Estimation of the number of synapses in the cerebral cortex: Methodological considerations. Cerebral Cortex 9, 722–732. DEFELIPE, J., GONZ ÁLEZ-ALBO, M. C., DEL R ÍO, M. R. & ELSTON, G. N. (1999b) Distribution and patterns of connectivity of interneurons containing calbindin, calretinin and parvalbumin in visual areas of the 314 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O occipital and temporal lobes of the macaque monkey. Journal of Comparative Neurology 412, 515–526. DEFELIPE, J., ARELLANO, J. I., MERCH ÁN-P ÉREZ, A., GONZ ÁLEZ-ALBO, M. C., WALTON, K. & LLIN ÁS, R. (2002) Spaceflight induces changes in the synaptic circuitry of the postnatal developing neocortex. Cerebral Cortex 12, 883–891. DEL R ÍO, M. R. & DEFELIPE, J. (1996) Colocalization of calbindin D-28k, calretinin and GABA immunoreactivities in neurons of the human temporal cortex. Journal of Comparative Neurology 369, 472–482. DEL R ÍO, M. R. & DEFELIPE, J. (1997a) Double bouquet cell axons in the human temporal neocortex: Relationship to bundles of myelinated axons and colocalization of calretinin and calbindin D-28k immunoreactivities. Journal of Chemical Neuroanatomy 13, 243–251. DEL R ÍO, M. R. & DEFELIPE, J. (1997b) Colocalization of parvalbumin and calbindin d-28k in neurons including chandelier cells of the human temporal neocortex. Journal of Chemical Neuroanatomy 12, 165–173. DEXLER, H. (1913) Das Hirn von Halicore dugong Erxl. Gegenbaurs Morphologisches Jahrbuch 45, 97–190. ELSTON, G. N., TWEEDALE, R. & ROSA, M. G. P. (1999a) Cortical integration in the visual system of the macaque monkey: Large scale morphological differences of pyramidal neurones in the occipital, parietal and temporal lobes. Proceedings of the Royal Society of London Series B 266, 1367–1374. ELSTON, G. N., MANGER, P. R. & PETTIGREW, J. D. (1999b) Morphology of pyramidal neurones in cy- tochrome oxidase modules of the S-I bill representation of the platypus. Brain Behavior and Evolution 53, 87– 101. ELSTON, G. N., BENAVIDES-PICCIONE, R. & DEFELIPE, J. (2001) The pyramidal cell in cogni- GLEZER, I. I. & MORGANE, P. J. (1990) Ultrastructure of synapses and Golgi analysis of neurons in neocortex of the lateral gyrus (visual cortex) of the dolphin and pilot whale. Brain Research Bulletin 24, 401–427. GLEZER, I. I., HOF, P. R., LERANTH, C. & MORGANE, P. J. (1993) Calcium-binding protein-containing neu- ronal populations in mammalian visual cortex: A comparative study in whales, insectivores, bats, rodents, and primates. Cerebral Cortex 3, 249–272. GRAY, E. G. (1959) Axo-somatic and axo-dendritic synapses of the cerebral cortex: An electron microscopic study. Journal of Anatomy 93, 420–433. HAUG, H. (1987) Brain sizes, surfaces, and neuronal sizes of the cortex cerebri: A stereological investigation of man and his variability and a comparison with some mammals (primates, whales, marsupials, insectivores, and one elephant). The American Journal of Anatomy 180, 126–142. HENDRY, S. H. C., SCHWARK, H. D., JONES, E. G. & YAN, J. (1987) Numbers and proportions of GABA- immunoreactive neurons in different areas of monkey cerebral cortex. Journal of Neuroscience 7, 1503–1519. HOF, P. R., GLEZER, I. I., CONDE, F., FLAGG, R. A., RUBIN, M. B., NIMCHINSKY, E. A. & VOGT WEISENHORN, D. M. (1999) Cellular distribution of the calcium-binding proteins parvalbumin, calbindin, and calretinin in the neocortex of mammals: Phylogenetic and developmental patterns. Journal of Chemical Neuroanatomy 16, 77–116. HOF, P. R., GLEZER, I. I., NIMCHINSKY, E. A. & ERWIN, J. M. (2000) Neurochemical and cellular spe- cializations in the mammalian neocortex reflect phylogenetic relationships: Evidence from primates, cetaceans, and artiodactyls. Brain, Behavior and Evolution 55, 300–310. tion: A comparative study in man and monkey. Journal of Neuroscience 21, RC163, 1–5. FAIR ÉN, A., DEFELIPE, J. & REGIDOR, J. (1984) Nonpyramidal neurons. General account. In Cellular Components of the Cerebral Cortex, Vol. 1, Cerebral Cortex (edited by PETERS, A. & JONES, E. G. ) pp. 201–253. New York: Plenum Press. GABBOTT, P. L. A. & BACON, S. J. (1996) Local circuit neurons in the medial prefrontal cortex (areas 24a,b,c, 25 and 32) in the monkey: II. Quantitative areal and laminar distributions. Journal of Comparative Neurology 364, 609–636. cerebral cortex. Functional Properties of Cortical Cells. In Cerebral Cortex (edited by JONES, E. G. & PETERS, A. ) pp. 63–89. New York: Plenum Press. HUBEL, D. H. & WIESEL, T. N. (1977) Functional architecture of macaque monkey cortex. Proccedings of the Royal Society of London Series B 198, 1–59. HUTTENLOCHER, P. R. & DABHOLKAR, A. S. (1997) Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology 387, 167– 178. GABBOTT, P. L. A., DICKIE, B. G. M., VAID, R. R., HEADLAM, A. J. N. & BACON, S. J. (1997) Local- JOHNSON, J. I., KIRSCH, J. A., REEP, R. L. & SWITZER, R. C. 3RD (1994) Phylogeny through brain traits: More circuit neurones in the medial prefrontal cortex (areas 25, 32 and 24b) in the rat: Morphology and quantitative distribution. Journal of Comparative Neurology 377, 465–499. GALARRETA, M. & HESTRIN, S. (2001) Electrical synapses between GABA-releasing interneurons. Nature Reviews Neuroscience 2, 425–433. GILBERT, G. D. & KELLY, J. P. (1975) The projection of cells in different layers of the cat’s visual cortex. Journal of Comparative Neurology 163, 81–106. GLEZER, I. I., JACOBS, M. S. & MORGANE, P. J. (1988) Implications of the “initial brain’’ concept for brain evolution in Cetacea. The Behavioral and Brain Sciences 11, 75–116. characters for the analysis of mammalian evolution. Brain, Behavior and Evolution 43, 319–347. JONES, E. G. (1983) The columnar basis of cortical circuitry. In The Clinical Neurosciences (edited by WILLIS, W. D. ) pp. 357–383. New York: Churchill Livingstone. JONES, E. G. (2000a) Microcolumns in the cerebral cortex. Proceedings of the National Academy of Sciences of USA 97, 5019–5021. JONES, E. G. (2000b) Cortical and subcortical contributions to activity-dependent plasticity in primate somatosensory cortex. Annual Review of Neuroscience 23, 1–37. JONES, E. G. & POWELL, T. P. S. (1970) Electron microscopy of the somatic sensory cortex of the cat. I. Cell HOUSER, C. R., VAUGHN, J. E., HENDRY, S. H. C., JONES, E. G. & PETERS, A. (1984) GABA neurons in Cortical microstructure 315 types and synaptic organization. Philosophical Transactions of the Royal Society of London Biological Sciences 257, 1–11. JONES, E. G. & HENDRY, S. H. C. (1986) Co-localization of GABA and neuropeptides in neocortical neurons. Trends in Neuroscience 10, 71–76. JONES, E. G. & WISE, S. P. (1977). Size, laminar and columnar distribution of efferent cells in the sensory-motor cortex of monkeys. Journal of Comparative Neurology 175, 391–438. KAWAGUCHI, Y. & KUBOTA, Y. (1997) GABAergic cell subtypes and their synaptic connections in rat frontal cortex. Cerebral Cortex 7, 476–486. KUBOTA, Y. & KAWAGUCHI, Y. (2000) Dependence of GABAergic synaptic areas on the interneuron type and target size. Journal of Neuroscience 20, 375–386. LETINIC, K., ZONCU, R. & RAKIC, P. (2002) Origin of GABAergic neurons in the human neocortex. Nature 417, 645–649. LEWIS, D. A. & LUND, J. S. (1990) Heterogeneity of chandelier neurons in monkey neocortex: Corticotropinreleasing factor- and parvalbumin-immunoreactive populations. Journal of Comparative Neurology 293, 599– 615. LORENTE DE N Ó, R. (1938) Architectonics and structure of the cerebral cortex. In Physiology of the Nervous System (edited by FULTON, J. F. ) pp. 291–330. New York: Oxford University Press. LUND, J. S., LUND, R. D., HENDRICKSON, A. E., BUNT, A. H. & FUCHS, A. F. (1975) The origin of efferent pathways from the primary visual cortex, area 17, of the macaque monkey as shown by retrograde transport of horseradish peroxidase. Journal of Comparative Neurology 164, 287–304. L ÜSCHER, C., NICOLL, R. A., MALENKA, R. C. & MULLER, D. (2000) Synaptic plasticity and dynamic modulation of the postsynaptic membrane. Nature Neuroscience 3, 545–550. MACKENZIE, P. J., KENNER, G. S., PRANGE, O., SHAYAN, H., UMEMIYA, M. & MURPHY, T. H. (1999) Ultrastructural correlates of quantal synaptic function at single CNS synapses. Journal of Neuroscience 19, RC13, 1–7. MARCO, P. & DEFELIPE, J. (1997) Altered synaptic circuitry in the human temporal epileptogenic neocortex. Experimental Brain Research 114, 1–10. MARIN, O., YARON, A., BAGRI, A., LAVIGNE, M. & RUBENSTEIN, J. TESSIERL. (2001) Sorting of striatal and cortical interneurons regulated by semaphorin-neuropilin interactions. Science 293, 872–875. MICHEVA, K. D. & BEAULIEU, C. (1995) Postnatal development of GABA neurons in the rat somatosensory barrel cortex: A quantitative study. European Journal of Neuroscience 7, 419–430. MICHEVA, K. D. & BEAULIEU, C. (1996) Quantitative aspects of synaptogenesis in the rat barrel field cortex with special reference to the GABA circuitry. Journal of Comparative Neurology 373, 340–354. MOUNTCASTLE, V. B. (1978) An organizing principle for cerebral function: The unit module and the distributed system. In The Mindful Brain (edited by MOUNTCASTLE, V. B. & EDELMAN, G. M. ) pp. 7–50. Cambridge: MIT Press. MOUNTCASTLE, V. B. (1997) The columnar organization of the neocortex. Brain 120, 701–722. NIEUWENHUYS, R. (1994) The neocortex. An overview of its evolutionary development, structural organization and synaptology. Anatomy and Embryology 190, 307– 337. NIMCHINSKY, E. A., GILISSEN, E., ALLMAN, J. M., PERL, D. P., ERWIN, J. M. & HOF, P. R. (1999) A neuronal morphologic type unique to humans and great apes. Proceedings of the National Academy of Sciences of USA 96, 5268–5273. O’KUSKY, J. & COLONNIER, M. (1982) A laminar analysis of the number of neurons, glia, and synapses in the visual cortex (area 17) of adult macaque monkeys. Journal of Comparative Neurology 210, 278–290. PETERS, A. (1987) Synaptic specificity in the cerebral cortex. In Synaptic Function (edited by EDELMAN, G. M., GALL, W. E. & COWAN, W. M. ) pp. 373–397. New York: John Wiley. PETERS, A. & JONES, E. G. (1984) Cerebral Cortex Vol. 1: Cellular Components of the Cerebral Cortex. New York: Plenum Press. PETERS, A. & PALAY, S. L. (1996) The morphology of synapses. Journal of Neurocytology 25, 687–700. PETERS, A., PALAY, S. L. & WEBSTER, H. DE F. (1991) The Fine Structure of the Nervous System. Neurons and their Supporting Cells. New York: Oxford University Press. PETERS, A. & SETHARES, C. (1997) The organization of double bouquet cells in monkey striate cortex. Journal of Neurocytology 26, 779–797. PETRIDES, M. & PANDYA, D. N. (1994) Comparative architectonic analysis of the human and the macaque frontal cortex. Handbook of Neurophysiology 9, 17–58. POE, B. H., LINVILLE, C. & BRUNSO-BECHTOLD, J. (2001) Age-related decline of presumptive inhibitory synapses in the sensorimotor cortex as revealed by the physical dissector. Journal of Comparative Neurology 439, 65–72. POWELL, T. P. S. (1981) Ceratin aspects of the intrinsic organization of the cerebral cortex. In Brain Mechanisms and Perceptual Awareness (edited by POMPEIANO, O. & AJMONE MARSAN, C. ) pp. 1–19. New York: Raven Press. PREUSS, T. M. & COLEMAN, G. Q. (2002) Human-specific organization of primary visual cortex: Alternating compartments of dense Cat 301 and calbindin immunoreactivity in layer 4A. Cerebral Cortex 12, 671–691. RAKIC, P. (1988) Specification of cerebral cortical areas. Science 241, 170–176. RAKIC, P. (2002) Evolving concepts of cortical radial and areal specification. In Changing Views of Cajal’s Neuron (edited by AZMITIA, E., DEFELIPE, J., JONES, E. G., RAKIC, P. & RIBAK, C. ) Progress in Brain Research 136, 265–280. RAKIC, P., BOURGEOIS, J.-P., ECKENHOFF, M. F., ZECEVIC, N., & GOLDMAN-RAKIC, P. (1986) Con- current overproduction of synapses in diverse regions of the primate cerebral cortex. Science 232, 232–235. RAKIC, P., BOURGEOIS, J.-P. & GOLDMAN-RAKIC, P. S. (1994) Synaptic development of the cerebral cortex: 316 D E F E L I P E , A L O N S O - N A N C L A R E S and A R E L L A N O Implications for learning, memory, and mental illness. Progress in Brain Research 104, 227–243. TAN, S. S., KALLONIATIS, M., STURM, K., TAM, P. P., REESE, B. E. & FAULKNER-JONES, B. (1998) Sepa- REN, J. Q., AIKA, Y., HEIZMANN, C. W. & KOSAKA, T. (1992) Quantitative analysis of neurons and glial cells rate progenitors for radial and tangential cell dispersion during development of the cerebral neocortex. Neuron 21, 295–304. THOMSON, A. M. & DEUCHARS, J. (1997) Synaptic interactions in neocortical local circuits: Dual intracellular recordings in vitro. Cerebral Cortex 7, 510–522. T ÖMB ÖL, T. (1974) An electron microscopic study of the neurons of the visual cortex. Journal of Neurocytology 3, 525–531. in the rat somatosensory cortex, with special reference to GABAergic neurons and parvalbumin-containing neurons. Experimental Brain Research 92, 1–14. REEP, R. L., JOHNSON, J. I., SWITZER, R. C. & WELKER, W. I. (1989) Manatee cerebral cortex: Cy- toarchitecture of the frontal region in Trichechus Manatus Latirostris. Brain, Behavior and Evolution 34, 365–386. ROCKEL, A. J., HIORNS, R. W. & POWELL, T. P. S. (1980) The basic uniformity in structure of the neocortex. Brain 103, 221–244. SCHIKORSKI, T. & STEVENS, C. F. (1999) Quantitative fine-structural analysis of olfactory cortical synapses. Proceedings of the National Academy of Sciences of USA 96, 4107–4112. SCH ÜZ, A. & PALM, G. (1989) Density of neurons and synapses in the cerebral cortex of the mouse. Journal of Comparative Neurology 286, 442–455. SILBERBERG, G., GUPTA, A. & MARKRAM, H. (2002) Stereotypy in neocortical microcircuits. Trends in Neuroscience 25, 227–230. SKOGLUND, T. S., PASCHER, R. & BERTHOLD, C. H. (1996) Heterogeneity in the columnar number of neurons in different neocortical areas in the rat. Neuroscience Letters 208, 97–100. SLOPER, J. J. (1973) An electron microscope study of the neurons of the primate motor and somatic sensory cortices. Journal of Neurocytology 2, 351–359. SLOPER, J. J., HIORNS, R. W. & POWELL, T. P. S. (1979) A qualitative and quantitative electron microscopic study of the neurons in the primate motor and somatic sensory cortices. Philosophical Transactions of the Royal Society of London Series B 285, 141–171. SOMOGYI, P., TAM ÁS, G., LUJAN, R. & BUHL, E. H. (1998) Salient features of synaptic organization in the cerebral cortex. Brain Research Reviews 26, 113–135. STOLZENBURG, J. U., REICHENBACH, A. & NEUMANN, M. (1989) Size and density of glial and neuronal cells within the cerebral neocortex of various insectivorian species. Glia 2, 78–84. TAKUMI, Y., RAM ÍREZ-LE ÓN, V., LAAKE, P., RINVIK, E. & OTTERSEN, O. P. (1999) Different modes of ex- pression of AMPA and NMDA receptors in hippocampal synapses. Nature Neuroscience 2, 618–624. TYLER, C. J., DUNLOP, S. A., LUND, R. D., HARMAN, A. M., DANN, J. F., BEAZLEY, L. D. & LUND, J. S. (1998) Anatomical comparison of the macaque and marsupial visual cortex: Common features that may reflect retention of essential cortical elements. Journal of Comparative Neurology 400, 449–468. WANG, Y., GUPTA, A., TOLEDO-RODRIGUEZ, M., WU, C. Z. & MARKRAM, H. (2002) Anatomical, physiolog- ical, molecular and circuit properties of nest basket cells in the developing somatosensory cortex. Cerebral Cortex 12, 395–410. WEST, M. J. & GUNDERSEN, H. J. (1990) Unbiased stereological estimation of the number of neurons in the human hippocampus. Journal of Comparative Neurology 296, 1– 22. WHITE, E. L. (1989) Cortical Circuits: Synaptic Organization of the Cerebral Cortex. Structure, Function and Theory. Boston: Birkhäuser. WHITE, E. L., WEINFELD, L. & LEV, D. L. (1997) A survey of morphogenesis during the early postnatal period in PMBSF barrels of mouse SmI cortex with emphasis on barrel D4. Somatosensory & Motor Research 14, 34–55. WILLIAMS, R. W. & RAKIC, P. (1988) Three-dimensional counting: An accurate and direct method to estimate numbers of cells in sectioned material. Journal of Comparative Neurology 278, 344–352. WINFIELD, D. A., GATTER, K. C. & POWELL, T. P. S. (1980) A electron microscopic study of the types and proportions of neurons in the cortex of the motor and visual areas of the cat and rat. Brain 103, 245–258. ZECEVIC, N. & RAKIC, P. (1991) Synaptogenesis in monkey somatosensory cortex. Cerebral Cortex 1, 510–523. ZECEVIC, N., BOURGEOIS, J.-P. & RAKIC, P. (1989) Changes in synaptic density in motor cortex of rhesus monkey during fetal and postnatal life. Developmental Brain Research 50, 11–32.