15 Telencephalon: Neocortex

Introduction...... 491 – Dendritic Clusters, Axonal Bundles and Radial Cell Cords As (Possible) Sulcal Pattern ...... 498 Constituents of Neocortical Minicolumns . . 582 Structural and Functional Subdivision – Microcircuitry of Neocortical Columns .... 586 ofNeocortex...... 498 – Neocortical Columns and Modules: – Structural Subdivision 1: Cytoarchitecture . . 498 A Critical Commentary ...... 586 – Structural Subdivision 2: Myeloarchitecture . 506 Comparative Aspects ...... 591 – Structural Subdivision 3: Myelogenesis ....510 Synopsis of Main Neocortical Regions ...... 592 – Structural Subdivision 4: Connectivity .....510 – Introduction...... 592 – Functional Subdivision ...... 516 – Association and Commissural Connections . 592 – Structural and Functional Subdivision: – Functional and Structural Asymmetry Overview ...... 528 of the Two Hemispheres ...... 599 – Structural and Functional Localization – Occipital Lobe ...... 600 in the Neocortex: Current Research – Parietal Lobe ...... 605 and Perspectives ...... 530 – TemporalLobe...... 611 Neocortical Afferents ...... 536 – Limbic Lobe and Paralimbic Belt ...... 617 Neocortical Neurons – FrontalLobe ...... 620 and Their Synaptic Relationships ...... 544 – Insula...... 649 – IntroductoryNote...... 544 – Typical Pyramidal Cells ...... 544 – Atypical Pyramidal Cells ...... 559 – Local Circuit Neurons ...... 560 Introduction Microcircuitry of Neocortex ...... 569 – Introduction...... 569 – Networks of Pyramidal Neuron ...... 570 The neocortex is an ultracomplex, six-layered – Interneuronal Systems ...... 571 structure that develops from the dorsal pallial Neocortical Columns and Modules ...... 575 sector of the telencephalic hemispheres (Figs. – Introduction...... 575 2.24, 2.25, 11.1). All mammals, including – The Investigations of Lorente de Nó: monotremes and marsupials, possess a neocor- Elementary Units and Glomérulos ...... 576 tex, but in reptiles, i.e. the ancestors of mam- – The Columnar Organization mals, only a three-layered neocortical primor- of the Somatosensory Cortex ...... 576 dium is present [509, 511]. The term neocortex – The Columnar Organization refers to its late phylogenetic appearance, in oftheVisualCortex...... 578 comparison to the “palaeocortical” olfactory – The Auditory Cortex ...... 579 cortex and the “archicortical” hippocampal – TheMotorCortex...... 579 – Columnar Patterns Shown by the Cells cortex, both of which are present in all am- of Origin and the Terminal Ramifications niotes [509]. of Cortico-cortical Connections ...... 579 The size of the neocortex varies greatly – Minicolumns and the Radial Unit Hypothesis among the various mammalian groups. In of Cortical Developmen ...... 581 some insectivores, such as the hedgehog, its 492 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.1. Brains of some mammals drawn on the same scale. The neocortex is shown in red 15 Telencephalon: Neocortex 493

Fig. 15.2. The telencephalon of the opossum (A) and human (B). The position and extent of the neocortex (shown in red) and of a number of subcortical centres (marked in A and B by the same hatchings/symbols) are indicated. In the opossum, the neocortex covers the remaining parts of the telencephalic hemispheres like a cap. In humans, the enormously expanded neocortex almost completely surrounds all subcortical centres. amc, amygdaloid complex; aon, anterior olfactory nucleus; olb, olfactory bulb; ot, olfactory tubercle; sep, precommissural septum; str, corpus striatum 494 Section II Structure of Spinal Cord and Brain Parts size does not exceed that of the “older” parts (Fig. 15.4 B). The characterization and determi- of the cortex, but in primates and cetaceans it nation of these layers is generally based on attains remarkable proportions, becoming by Nissl-stained material. Although the study of far the largest centre in the brain (Figs. 15.1, such material, in which the dendritic and axo- 15.2). Stephan and Andy [709] determined the nal processes of the neurons remain largely average index of progression for the neocortex unstained, yields in itself very little insight into (and for many other brain structures) in a the structural and functional organization of number of insectivores, prosimians and simi- the cortex, the results of such cytoarchitectonic ans, including humans. These indices express analyses are important because they provide a how many times larger the neocortex is in a very useful general framework for studies em- particular group of species than in that of a ploying other, more critical techniques. typical basal insectivore of the same size. It Beginning at the surface, the six neocortical was found that in prosimians the neocortex is layers are as follows: (I) lamina molecularis, on average 14.5 times, and in the simians 45.5 (II) lamina granularis externa, (III) lamina times larger than in basal insectivores. In hu- pyramidalis externa, (IV) lamina granularis in- mans, the neocortex appeared to be 156 times terna, (V) lamina pyramidalis interna and (VI) larger than that of basal insectivores. lamina multiformis. As regards the functional differentiation of I. The lamina molecularis or lamina zonalis the neocortex, in many “primitive” mammals contains only very few cell bodies. (marsupials, insectivores and rodents) much of II. The lamina granularis externa is composed the neocortex is occupied by projection areas of small, densely packed cell bodies. The which either receive, via the thalamus, im- name of this layer is misleading because pulses directly related to the various special most of its constituent somata belong to senses, or are concerned with the steering of small pyramidal neurons which, like all motor activity. The sensory fields comprise the typical cortical pyramids, direct their somatosensory cortex, receiving impulses from apices toward the surface. sense organs situated in the skin, the muscles III. The lamina pyramidalis externa is a thick and the joints, then the primary visual cortex layer in which pyramidal somata prevail. and finally the primary auditory cortex. These somata increase progressively in size There is evidence suggesting that already in from superficial to deep. primitive mammals the various projection IV. The lamina granularis interna consists of areas just mentioned are separated from each small, densely packed, pyramidal and non- other by narrow strips of non-projection cor- pyramidal somata. tex (see Cajal [84], p. 830 and figure 531). V. The lamina pyramidalis interna consists What we now see is that these strips expand mainly of medium-sized and large, loosely enormously to form a large area known as the arranged pyramidal somata. parietotemporal association cortex. Another VI. The lamina multiformis is composed of rel- area of association cortex develops in front of atively tightly packed, spindle-shaped so- the motor cortical areas (Fig. 15.3). This region mata. Golgi material reveals that most of is known as the prefrontal association cortex. these somata belong to modified pyramidal All cortical areas maintain afferent and efferent cells. connections with subcortical centres. However, these various association cortical areas are pri- The neocortex is frequently also designated as marily connected with other cortical fields. isocortex [784, 791] or homogenetic cortex.Brod- The neocortex shows a laminated structure mann [70, 71] used the latter term to indicate throughout its extent. Whereas in the piriform that, even though some cortical areas in the adult and hippocampal parts of the mammalian cor- brain may have more or less than six layers, the tex three layers can be distinguished, in the entire neocortex has passed through a basic, six- neocortex six layers are usually recognized layered stage during ontogeny (Fig. 15.4 A). 15 Telencephalon: Neocortex 495

Fig. 15.3. Lateral views of the brains of the hedgehog (A), the prosimian Galago (B) and human (C) to show the evolutionary differentiation of the neocortex. In the hedgehog almost the entire neocortex is occupied by sensory (ss, vis, ac) and motor (m) areas. In the prosimian Galago the sensory cortical areas are separated by an area occupied by association cortex (pa). A second area of association cortex (aa) is found in front of the premotor cortex (pm). In the human the association areas are strongly developed. aa, anterior association area; ac, auditory cortex; i, insular cortex; m, motor cortex; pa, posterior association area; pm, premotor cortex; ss, somatosensory cortex; vis, visual cortex 496 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.4. The laminar pattern of the human neocortex: A in a newborn, diagrammatic; B Nissl preparation, adult; C myelin sheath preparation, adult. After Vogt and Vogt [786] and Rose [637]. Bi, Bo, inner and outer stripes of Baillarger 15 Telencephalon: Neocortex 497

Apart from the neuronal perikarya, several cortical neurons. Many extrinsic afferents also other structural elements in the cortex show a take a radial course after having entered the more or less distinct laminar arrangement. cortex from the deep white matter. Myelin-stained and reduced silver preparations The systematic study of the disposition of reveal the presence of tangentially oriented fi- myelinated fibres in brain structures is known bre concentrations at several levels, and Golgi as myeloarchitectonics. Detailed myeloarchitec- material shows that these fibres and their ter- tonic analyses of the human cerebral cortex minal ramifications may contribute to plexi- have been carried out by C. and O. Vogt [784, form zones, which are likewise tangentially ar- 786, 789–791]. So far as the neocortex is con- ranged. Thus, lamina I contains a dense plexus cerned, the layers distinguished in these anal- of horizontally running extrinsic and intrinsic yses correspond to the cytoarchitectonic layers fibres that contact the apical dendritic bou- of Brodmann and others (Fig. 15.4 B); however, quets of pyramidal neurons situated in the they are designated with arabic, rather than deeper layers. Tangential plexuses of fibres in with roman numerals (Fig. 15.4 C). layer IV and deep in layer V are known as the Lorente de Nó [416] concluded on the basis outer and inner striae of Baillarger. These of the study of extensive Golgi material that striae are probably formed primarily by intrin- the main pattern of connections between corti- sic cortical axons, i.e. axons of local circuit cal neurons is in a vertical direction. He ad- neurons and collateral branches of pyramidal vanced the idea that the cerebral cortex essen- cell axons [340]. The outer stria of Baillarger is tially consists of small, radially arranged sets particularly well developed in the visual cortex, of neurons having a thalamic afferent fibre as where it is referred to as Gennari’s (or Vicq their axis. According to Lorente de Nó, these d’Azyr’s) line (Fig. 5.13 C). column-like elementary units contain all types Although tangential lamination is a promi- of cortical cells and within their confines the nent feature of the neocortex, many of its con- whole process of the transmission of impulses stituent elements show an evident radial orien- from the afferent fibre to the efferent axon is tation. Prominent among these are the pyrami- accomplished. Electrophysiological studies of dal cells. The apical dendritic shafts of these the somaesthetic [494] and the visual cortex ubiquitous elements extend peripherally, and [308] have provided powerful evidence for the many of them reach the most superficial layers presence of functionally discrete radial col- before forming their terminal tufts. The axons umns or modules in these sensory areas. Later, of the pyramidal cells are also radially orient- morphological studies [242] showed that in ed. They emerge from the basis of the cell other areas the cortical modules are organized bodies and descend toward the white matter. around cortico-cortical afferents rather than The pyramidal cell axons acquire a myelin thalamic inputs. The cortical columns will be sheath at a short distance from the soma and further discussed in a later section of the pres- assemble in bundles that increase in size as ent chapter. they descend and as more axons are added. Some quantitative data on the cerebral cor- These bundles are known as radial fasciculi. tex are presented in Table 15.1. The apical dendrites of the pyramidal cells are also arranged in bundles [194, 196, 565, 568]. These axonal and dendritic bundles impose upon the neuronal cell bodies an arrangement in slender radially oriented columns that extend throughout the thickness of the cortex. Not only the main axons of the pyramidal cells, but also their collateral branches often show a radial orientation, and the same holds true for the axonal systems of many intrinsic 498 Section II Structure of Spinal Cord and Brain Parts

Table 15.1. Quantitative data on the human cerebral cortex Volume (both hemispheres) 517 cm3 (males) Pakkenberg and Gundersen [525] 440 cm3 (females) Surface (both hemispheres) 1470–2275 cm2 Blinkov and Glezer [56] Elias and Schwartz [173] Pakkenberg and Gundersen [525] Depth of neocortex 1.5–5 mm von Economo and Koskinas [796] Total number of neurons (both hemispheres) 22.8 ´109 Pakkenberg and Gundersen [525]

Sulcal Pattern Structural and Functional Subdivision of Neocortex

The human cerebral cortex, as that of other large mammals, is strongly folded (Fig. 15.1) Structural Subdivision 1: Cytoarchitecture [31]. Convolutions or gyri are separated by fis- sures or sulci. In humans, almost two thirds of Although the laminar basic pattern is recogniz- the neocortex is hidden away in the depth of able throughout the neocortex, this structure is the sulci [64]. The sulcal pattern of the human not homogeneous. Differences in the relative cerebral hemispheres has been the subject of thickness and cell density of the various layers, numerous studies, as for instance those of Ret- and in the size, shape and arrangement of the zius [615], Bailey and von Bonin [30] and Ono neuronal perikarya, are present (Figs. 15.6, et al. [519]. These studies have shown that the 15.7) and have been used to divide the neocor- overall sulcal pattern on the telencephalic sur- tex into cytoarchitectural areas. Similarly, dif- face is highly characteristic for humans (Fig. ferences in the pattern of myelinated fibres (lo- 15.5 A–C) and that the same holds true for cal development and distinctness of the striae other mammalian species. However, the indi- of Baillarger, length of the radial fasciculi) have vidual sulci show a considerable intersubject been used in myeloarchitectonic parcellations variability. They may vary in position and of the cortex (see below). course (Fig. 15.5 G). Many sulci may show one During the first half of the twentieth cen- or several interruptions (Fig. 15.5 A,F) and tury, several (groups of) investigators, namely some may be doubled over a certain part of Campbell [85], Brodmann [70, 71], von Econo- their trajectory (Fig. 15.5 E). The sulcal pattern mo and Koskinas [796], Bailey and von Bonin not only varies among individuals but also var- [30] and Sarkissov et al. [649], have produced ies between the two hemispheres in the same cytoarchitectonic maps of the human cerebral individual. Some sulci, including the lateral cortex (Table 15.2). In all of these maps, the and central sulci and the parieto/occipital and cortex is parcellated into a number of juxta- cingulate sulci, mark the boundaries between posed areas or fields. Brodmann’s map, which cerebral lobes (Fig. 1.4). is by far the most widely used, is shown in Fig. 15.8. Brodmann distinguished 44 sharply delineated areas, each of which he designated with a different figure. He emphasized that the boundaries between these areas generally do not coincide with the sulci on the cerebral surface. Von Economo and Koskinas [796] and Sarkissov et al. [649] subdivided several of Brodmann’s areas into smaller units; hence, their total number of fields is somewhat larger 15 Telencephalon: Neocortex 499

1 Lateral sulcus (Sylvian fissure) 10 Cingulate sulcus 2 Central sulcus (of Rolando) 11 Parieto-occipital sulcus 3 Precentral sulcus 12 Calcarine sulcus 4 Postcentral sulcus 13 Occipitotemporal sulcus 5 Superior frontal sulcus 14 Collateral sulcus 6 Inferior frontal sulcus 15 Lateral orbital sulcus 7 Intraparietal sulcus 16 Medial orbital sulcus 8 Superior temporal sulcus 17 Arcuate orbital sulcus 9 Inferior temporal sulcus 18 Olfactory sulcus

Fig. 15.5 A–G. Synopsis of cerebral sulci. A–C Lateral, medial and ventral aspect of a cerebral hemisphere, showing the most important sulci. D–F Three variations in the course of the cingulate sulcus, as depicted by Retzius [615]. In D there is a single, continuous cingulate sulcus; in E the sulcus is doubled (10a,b) and in F it is interrupted twice. G Ten variants in the course of the sulci on the orbitofrontal cortex, derived from Ka- nai [358], have been transferred to a standard outline of this region 500 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.6. Areas 17 and 18 and their border zone in the human brain. Reproduced from Vogt and Vogt [786]. The position of the line of Gennari (G) corresponds with that of the cell-poor layer IVb 15 Telencephalon: Neocortex 501

Fig. 15.7. Cytoarchitecture of several neocortical areas, designated with Brodmann’s [70, 71] numbers 502 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.8. Brodmann’s famous map of the cytoarchitecture of the human cerebral cortex. The latest (1914) version is shown here [71]. In the pericentral and retrosplenial regions some (parts of) cytoarchitectonic areas have been transferred to the free surface of the hemisphere. The small auxiliary chart shows the cytoarchitecture of the insula and the upper surface of the superior temporal gyrus 15 Telencephalon: Neocortex 503

Fig. 15.9 A,B. Cytoarchitectonic analysis of the human cerebral cortex, according to Sarkissov et al. [649]. These authors mapped the results of their analysis not only on lateral and medial views (not reproduced here), but also on upper (A) and frontopolar views (B) of the hemispheres. They followed Brodmann’s numbering scheme and introduced self-explanatory symbols for subareas and transitional zones 504 Section II Structure of Spinal Cord and Brain Parts

Table 15.2. Architectonic subdivisions of the human neocortex Author(s) Type of analysis No. of fields Judgement of borders Campbell (1905) [85] Cyto- (myelo-) 14 Sharp Brodmann (1909, 1914) [70, 71] Cyto- 44 a Sharp von Economo and Koskinas (1925) [796] Cyto- 54 Generally not sharp Vogt and Vogt (1919, 1926) [784, 791] Myelo- >200 Very sharp Bailey and von Bonin (1951) [30] Cyto- 8 Generally not sharp Sarkissov et al. (1955) [649] Cyto- 52 Many transition zones a The highest numbered area in Brodmann’s map is area 52; however, because areas numbered 13–16 and 48–51 are lacking the total number of areas in his map amounts to 44

than that of Brodmann. According to von The cortex of fundamental type 1 is distin- Economo and Koskinas, who designated the guished by its lack of distinct granular layers various cytoarchitectonic areas with combina- II and IV, whereas the pyramidal layers III and tions of letters and figures (PA1, TE2 etc.), the V are strongly developed. Reference to Fig. interareal boundaries are generally not sharp. 15.10 B,C shows that this agranular cortex oc- The map of Sarkissov and colleagues (Fig. cupies a region situated directly in front of the 15.9) is patterned after that of Brodmann. central sulcus, which corresponds to the areas Their nomenclature followed Brodmann’s num- 4 and 6 of Brodmann, and that this type of bering scheme. However, they intercalated cortex is also found in the paracentral lobule transition zones between several of the areas on the medial surface of the hemisphere and that Brodmann had distinguished. in the rostral part of the cingulate gyrus. With- Several authors have attempted to allocate in this region, area 4 is characterized by the the various cytoarchitectonic areas to a smaller presence of the giant pyramidal cells of Betz in number of variant types. It is worthy of note its fifth layer (Fig. 15.7 F). Because the agranu- in this context that as early as 1874, Betz [51] lar cortex gives rise to prominent corticobulbar pointed out that the human cerebral cortex is and corticospinal projections, it may be con- divided by the central sulcus of Rolando into sidered the prototype of “motor” cortex. an anterior part in which pyramidal cells pre- The other type of heterotypical cortex (la- dominate and a posterior part where granular belled 5) is characterized by its richness in cells prevail. In the anterior part he described small, granular cells and the strongly devel- the giant pyramidal cells that bear his name. oped layers II and IV. Layers III and V, on the Von Economo [795], who also focussed on the other hand, are poorly developed. Areas be- distribution of pyramidal and granular cells, longing to this type of cortex are found in the recognized five fundamental cytoarchitectural anterior part of the postcentral gyrus, along types in the human neocortex, which he la- the calcarine fissure and in a limited part of belled 1–5 (Fig. 15.10). The cortical types la- the upper surface of the superior temporal belled 2, 3 and 4 contain the six typical neo- gyrus. The type 5 cortex is called granular cor- cortical layers described above, although not tex or koniocortex (in Greek konios = dust). It all of these are developed to the same degree is characteristic of those cortical areas that rein the different types. Von Economo designated ceive the great afferent systems, i.e. the soma- these areas as homotypical, thus contrasting tosensory projection, the acoustic projection them to the heterotypical cortical types 1 and and the optic radiation. A particularly well-dif- 5, in which, at least in the fully developed cor- ferentiated granular cortex is found in the area tex, the six layers cannot be clearly discerned. striata (area 17 of Brodmann). As shown in 15 Telencephalon: Neocortex 505

1 Heterotypical, agranular cortex (motor cortex) 2 Homotypical cortex, frontal type 3 Homotypical cortex, parietal type 4 Homotypical cortex, polar type 5 Heterotypical, granular cortex (primary sensory cortical areas)

Fig. 15.10. The five principal types of neocortex (A), as distinguished by von Economo [795] and their distribution over the medial (B) and lateral surfaces (C) of the cerebral hemispheres 506 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.6, the number of granular cells is ex- regions the analyses of the “classical” cyto- ceedingly high here and the inner granular architectonists, as well as those of later in- layer is clearly subdivided into three sublayers, vestigators, have led to widely diverging re- IVa,b and c. The middle of these, which is rel- sults (Fig. 15.11) [518, 718]. atively poor in cells, coincides with the layer of myelinated fibres known as the line of Gennari In his notable monograph of 1909, Brodmann or Vic d’Azyr. [70] did not confine himself to the human The three types of homotypical cortex (2, 3 brain, but also presented cytoarchitectonic and 4; Fig. 15.10) occupy an intermediate posi- analyses of the cortex of a considerable num- tion between the agranular (1) and granular ber of other mammalian species, among them (5) types. They are named according to their two primates, a prosimian, a carnivore and an main, but not exclusive localization: frontal insectivore. The number of areas identified in type (2), parietal type (3) and polar type (4). these animals was smaller than that in the hu- Types 2 and 3 both contain numerous small man. Thus, he delineated 24 areas in the maca- and medium-sized pyramidal cells, but in type que Cercopithecus, 26 in the marmoset Calli- 3, granular cells are more prominent than in thrix, 14 in the prosimian Lemur and 12 in the type 2 (Fig. 15.7 A,E). The polar type (4) occu- insectivore Erinaceus. In the monkeys studied, pies small areas near the frontal and occipital homologues of the human primary motor poles, hence its name. It is characterized by its areas 4 and 6 and the primary sensory areas 3, thinness and high content of granular cells. In 1, 2 and 17 could be identified, and the same both the parietal (3) and polar (4) types of holds true for several of the human “associa- cortex, the multiform layer (VI) is particularly tion” areas, such as the frontal areas 9 and 12, well developed (Fig. 15.7 E). The homotypical the parietal areas 5 and 7 and the temporal cortical types (2, 3 and 4) occupy much of areas 20–22 (Fig. 15.12). Brodmann’s compara- what Flechsig [191, 192] and many later inves- tive cytoarchitectonic explorations laid the tigators have provisionally designated as asso- foundation for later electrical stimulation stud- ciation cortex. ies as well as for experimental analyses of cor- If we consider the cytoarchitectonic analyses tical fibre connections (see below). of Campbell [85], Brodmann [70, 71], von Econ- omo and Koskinas [796], Bailey and von Bonin [30] and Sarkissov et al. [649] in light of the Structural Subdivision 2: Myeloarchitecture classification of cortical types just discussed, the following conclusions can be drawn: It has already been mentioned that C. and O. 1. The heterotypical agranular and granular Vogt [784, 786, 789–791] subjected the human areas, i.e. the primary motor and primary cerebral cortex to a detailed myeloarchitectonic sensory fields, have been recognized and de- analysis. The myelinated fibres in the cortex lineated in all of these analyses. show two principal orientations, tangential and 2. The homotypical cortex was left largely undi- radial. The tangential fibres tend to form lami- vided by Campbell and by Bailey and von Bo- nae, which in general can be readily identified nin; hence, the number of areas distinguished in conjunction with the corresponding layers by these authors is considerably smaller than observed in Nissl preparations (Fig. 15.4 B,C). that recognized by the others (Table 15.2). The radially oriented fibres are arranged in 3. With regard to the location and subdivision bundles, termed radii, which ascend from and of the three types of homotypical cortex, descend to the subcortical white matter. The the results of Brodmann, von Economo and Vogts noticed that the number and distinctness Koskinas and Sarkissov et al. are in general of the tangential fibre layers show considerable closely comparable. Exceptions should be local differences in the cortex, and that the made, however, for the medial temporal lobe same holds true for the extent to which the ra- and the orbitofrontal cortex. In both of these dii penetrate into the cortex. Focussing on 15 Telencephalon: Neocortex 507

Fig. 15.11 A–G. Cytoarchitectonic parcellations of the human orbitofrontal cortex. Note that the analyses of the seven (groups of) authors have led to widely diverging results. Courtesy of Drs. H. Uylings and G. Raj- kowska. LOS, lateral orbital sulcus; MOS, medial orbital sulcus; OLF, olfactory sulcus; TOS, transverse orbital sulcus. References cited: Brodmann [71], von Economo and Koskinas [796], Beck [44], Sarkissov et al. [649], Petrides and Pandya [577], Hof et al. [301], Öngür et al. [518] 508 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.12. Cytoarchitectonic map of the cerebral cortex of the macaque Cercopithecus, according to Brod- mann [70] 15 Telencephalon: Neocortex 509

Fig. 15.13 A–C. Myeloarchitecture of some cortical areas, according to the Vogts. The numbers have been added according to the Brodmann scheme. Reproduced from Rose [637] 510 Section II Structure of Spinal Cord and Brain Parts these differences, they were able to delineate a view, clusters of late-maturing areas form three large number of areas whose boundaries were association centres in the human brain: a large found to coincide with those of Brodmann’s posterior association centre, occupying much of cytoarchitectonic fields (Fig. 15.13) [787]. They the parietal, occipital and temporal lobes; an emphasized that the areal boundaries are very anterior association centre, expanding in front sharp and omnilaminar, i.e. characterized by of the motor projection areas; and a small in- changes in all layers. In their parcellation of sular association centre (Fig. 15.14). the cortex, the Vogts went much further than Brodmann. Many of the areas distinguished by the latter were subdivided into several subar- Structural Subdivision 4: Connectivity eas, which in turn were claimed to be subdivis- ible into still smaller units, termed campuli Cytoarchitectonic studies form an indispens- [787]. able first step in the analysis of any part of the The Vogts did not investigate the temporal central nervous system. They should logically and occipital lobes; hence, their myeloarchitec- be followed by exploration of the fibre connec- tonic map of the human cortex remained in- tions of the delineated grisea. The fibre con- complete [64]. nections may be used as auxiliary criteria in the detection and homologization of morphological entities, but secondly and more impor- Structural Subdivision 3: Myelogenesis tantly, they may give salient clues as to the functional significance of their targets. Experi- Myelinated fibres also took a central position mental studies on two sets of connections, tha- in the investigations of Flechsig [191–193]. lamocortical and cortico-cortical, have led to However, this author did not study the disposi- very useful functional subdivisions of the neo- tion of the myelinated fibres in the mature cor- cortex. It is important to note that, because hu- tex, but rather the temporal order in which the mans, for obvious reasons, cannot be subjected fibres of the white matter immediately under- to interventional techniques, our knowledge of neath the different parts of the cortex become the fibre connections mentioned is almost en- myelinated during development. Flechsig dis- tirely based on studies in non-human primates, tinguished 45 myelogenetic areas, which he particularly the rhesus macaque. numbered according to their sequences of The thalamic nuclei and their connections myelination during foetal and early postnatal have been extensively discussed in Chap. 8. Let life (Fig. 15.14). He classed these areas in three it suffice to recall here that the four groups of categories, primordial areas (1–16), which specific thalamic nuclei, sensory relay nuclei, show signs of myelination before birth, inter- motor relay nuclei, limbic nuclei and associa- mediate areas (17–36), which become myeli- tion nuclei, project to specific areas or regions nated between birth and the second postnatal of the cortex and together innervate the entire month and terminal areas (37–45), which mye- cortex (Fig. 8.4). linate after the second postnatal month. The sensory relay nuclei, i.e. the lateral pos- Flechsig observed that the prenatally myeli- terior nucleus, the lateral geniculate body and nating fibres belong to the large sensory and the medial geniculate body, project to and motor projections. For this reason he charac- functionally define the primary sensory cor- terized the primordial areas as projection areas. tices. The ventral posterior nucleus receives af- According to Flechsig, the postnatally maturing ferents from the somatosensory pathways and fibres form associative links between different projects to the areas 3, 1 and 2, which collec- parts of the cortex; hence, he designated the tively form the primary somatosensory area, intermediate and terminal areas as association S1. The lateral geniculate body receives affer- areas. He speculated that these areas subserve ents from the retina and projects to the pri- higher cognitive and mental functions. In his mary visual cortex, V1, which corresponds to 15 Telencephalon: Neocortex 511

Fig. 15.14. Myelogenetic subdivision of the human cerebral cortex, according to Flechsig [193]. The numbers of the areas indicate the sequence of myelination during fetal and early postnatal life: 1–16 (cross-hatched), early myelinating, primordial areas; 17–36 (vertical lines), intermediate areas; 37–45 (white), terminal or “association” areas. Reproduced from von Bonin [793] 512 Section II Structure of Spinal Cord and Brain Parts area 17. This area is also known as the striate subsidiary nuclei, each with receiving afferents area, because of the presence of the line of from one or more specific subcortical source Gennari, or the calcarine cortex, because it and sending efferents to a specific cortical area surrounds the calcarine sulcus. The medial or region. However, the distinguishing feature geniculate body, which represents the thalamic of these two cell masses as a whole is that their processing station in the auditory pathway, subcortical afferents are rather weakly devel- sends its efferents to the primary auditory cor- oped and that they receive their principal, tex, A1, which consists of the areas 41 and 42. driving afferents from the cortex. They func- The motor relay nuclei comprise the ventral tion primarily as links in cortico-thalamo-cor- anterior nucleus (VA) and the anterior and tical association paths, hence their name [680, posterior divisions of the ventral lateral nu- 681]. The mediodorsal nucleus projects to the cleus (VLa, VLp). VA is the principal target of prefrontal cortex, which corresponds to Flech- fibres ascending from the pars reticulata of the sig’s anterior association centre (Fig. 15.14). substantia nigra and projects to the frontal eye The cortical output of the pulvinar is to the field (area 8) and adjacent parts of the pre- areas of parieto-temporo-occipital cortex, inter- frontal cortex. VLa is the terminus of afferents calated between the primary somatosensory, from the internal segment of the globus palli- auditory and visual areas, i.e. to Flechsig’s pos- dus and projects to the premotor cortex (area terior association centre. However, these pro- 6). VLp, finally, receives a massive input from jections are less specific than was previously the cerebellar nuclei and sends its efferents to thought. Thus, the mediodorsal nuclei send ax- the primary motor cortex, M1 (area 4). It ons to several cortical regions other than the should be mentioned that the efferent projec- prefrontal, among them the cingulate, insular tions from the three motor relay nuclei, VA, and parietal regions, as well as the premotor VLa and VLp, are focussed on the cortical and motor accessory areas, while the pulvinar areas mentioned but that their terminal fields has additional connections with the frontal eye overlap to some extent. field and with the anterior and posterior cin- The limbic nuclei comprise the anterior nu- gulate, retrosplenial, insular and parahippo- clear complex and the lateral dorsal nucleus. campal areas. These structures are linked by being the tar- The question of whether a cortical area that gets of hippocampal fibres arising from the receives projection fibres from a specific tha- subicular complex, reaching the thalamus via lamic element possesses structural characteris- the fornix, the mamillary nuclei and the ma- tics that permit its morphological delimitation millothalamic tract (Fig. 12.13). They are the was specifically addressed by Rose and Wool- source of thalamic fibres to the cingulate gyrus sey [634] in a study on the structure and tha- (areas 23, 24, 32), including the retrosplenial lamic connections of the cingulate gyrus in areas 29 and 30, this projection extending as rabbit and cat. They found that the cingulate far posteriorly and inferiorly as the presubicu- cortex in these animals can be subdivided into lum and parasubiculum and the entorhinal cor- three cytoarchitectonic areas, and that each of tex (area 28) [712]. The sources of the afferents these areas co-extends the distribution field of to the anterior nuclear complex and the lateral a particular nucleus within the anterior nuclear dorsal nucleus, as well as the targets of their group. Such close correlations between cytoar- efferents, all form part of the limbic system. chitecture and the projection fields of individu- The association nuclei include the mediodor- al thalamic nuclei were later also found in sal nucleus and the pulvinar, two cell masses many other regions of the cerebral cortex [342, that are particularly well developed in pri- 401, 636, 670]. mates. The pulvinar involves more than one Experimental studies on the cortico-cortical half of the whole thalamic volume in humans connections in the rhesus monkey have consid- [550]. Both cell masses actually represent cell erably increased our insights into the functional complexes. They can be subdivided into several organization of the association cortices. The fol- 15 Telencephalon: Neocortex 513 lowing summary of the results of these studies is tex, A1 (areas 41 and 42), which is located principally based on review articles by Pandya in Heschl’s gyrus on the posterior aspect of and colleagues [526, 529, 532, 579], Van Hoesen the temporal plane. The connectivity found [773] and Mesulam [463–465]. The latter made a in the monkey brain suggests that the poste- constructive effort to integrate cytoarchitectural rior part of the superior temporal cortex and clinical findings in humans with experi- (area 22) displays the projections of the mental evidence obtained from the non-human proximal auditory association cortex, where- primate to form a working model of the human as the more anterior part of this gyrus may cerebral cortex (Fig. 15.15). represent the distal auditory association cor- 1. The association areas can be subdivided into tex. two main types: modality-specific (unimo- Unimodal association areas for taste and dal) and high-order (heteromodal). Each vestibular sensation have not been identified primary sensory area is adjoined by a so far. Some areas in the posterior orbitolat- modality-specific sensory association area. eral cortex and anterior insula (13 a, I am) These unimodal sensory association areas may represent unimodal olfactory associa- can be further subdivided into proximal or tion areas (see Figs. 11.7, 11.8). upstream and distal or downstream compo- 3. Mesulam [463, 464] considers the premotor nents. Proximal areas are only one synapse cortex anterior to the primary motor area, away from the corresponding primary sen- M1, as a motor analogue of the modality- sory area, whereas distal areas are at a dis- specific sensory association areas, chiefly tance of two or more synapses from the rel- because it provides the principal cortical in- evant primary area. put to M1. According to Mesulam, this mo- 2. The unimodal sensory association areas to- tor association area (MA) is composed of gether occupy most of the post-Rolandic area 6, including the supplementary motor neocortex. region (M2), posterior area 8 and area 44. It The somatosensory unimodal association has reciprocal connections with the unimo- cortex (SA) is situated in the parietal lobe, dal sensory association areas [526, 532]. directly behind the primary somatosensory 4. The high-order heteromodal association areas 3, 1 and 2, which are collectively de- areas, which are sometimes also designated signated as S1. It is supposed to occupy as polymodal or supramodal areas, receive parts of areas 5 and 7 in the superior parie- their principal cortical afferents from (1) tal lobule and may also include parts of area unimodal areas, particularly their distal 40 in the anterior portion of the inferior pa- zones, (2) other heteromodal areas and (3) rietal lobule. The subdivision of the somato- paralimbic areas. In the primate brain, three sensory association cortex into proximal and heteromodal association areas, parietotem- distal areas in the human remains to be elu- poral, medial temporal and prefrontal, have cidated. been identified. The visual unimodal association cortex (VA) occupies much of the occipital lobe The parietotemporal heteromodal association and extends far anteriorly into the lower parts area is situated at the juncture of the unimodal of the temporal lobe. Its proximal zone con- sensory association cortices, from which it re- sists of the areas 18 and 19 which, together ceives convergent bimodal and trimodal affer- forming the circumstriate belt, surround the ents. This area includes the caudal part of the primary visual of striate cortex (V1). The dis- superior parietal lobule (area 7), most of the tal zone of the unimodal visual association inferior parietal lobule (areas 39 and 40) and a cortex includes areas 20, 21 and 37. forwardly extending strip, formed by lateral The unimodal auditory association cortex temporal cortex within the banks of the superi- (AA) covers the superior temporal gyrus or temporal sulcus at the juncture of areas 21 (area 22). It flanks the primary auditory cor- and 22 (Fig. 15.15). 514 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.15. Subdivision of the human neocortex into functional zones. This subdivision is largely based on experimental studies of cortico-cortical connections in monkeys. The primary somatosensory (S1), visual (V1) and auditory (A1) cortical areas (in dark red) project via short connections to the adjacent unimodal somatosensory (SA), visual (VA ) and auditory (AA) association areas (in medium red). These unimodal sensory association areas in turn project to an elongated parietotemporal heteromodal association area (PTHA, in light red). The unimodal and heteromodal sensory association areas project massively to the prefrontal cortex (PFC,inlight red). These long associational projections are diagrammatically indicated by red-outlined, white fibres. Sequences of short connections successively link the various parts of the PFC with the motor association area (MA,inlight grey) and the primary motor area (M1,indark gray) (based on Mesu- lam [463, 465]). TP, temporopolar cortex 15 Telencephalon: Neocortex 515

The medial temporal heteromodal associa- Feed-back. The great majority (75% or tion area is formed by the perirhinal areas 35 more) of the cortical feed-forward connections and 36, which are interposed between the ento- are reciprocated by descending systems pro- rhinal area 28 and the visual association areas jecting back to areas from which an input was 19, 37 and 20. received. The fibres forming these feed-back The prefrontal heteromodal association area systems originate nearly always from the infra- is situated in front of the motor association granular layers V and VI and terminate largely cortex and includes areas 9 and 10, 45–47 and in layers I and VI [623]. the anterior parts of areas 8, 11 and 12. It re- Parallel Processing. Extensive anatomical and ceives afferents from the unimodal sensory as- physiological studies on the visual system of sociation areas, particularly from their distal the rhesus monkey have shown that different parts and from the parietotemporal and medial features of the visual image remain segregated temporal heteromodal association areas. An in the striate cortex and in their further pro- important output system of the prefrontal het- jection to the extrastriate visual areas (see eromodal association area is formed by se- Chap. 19 for references and details). These ex- quences of short association fibres that succes- trastriate visual areas are organized into two sively link the anterior orbitofrontal cortex, the hierarchically organized and functionally spe- polar and lateral prefrontal areas 9, 10 and 46, cialized processing pathways: an occipitoparie- the motor association cortex and the primary tal pathway or “dorsal stream”, concerned with motor cortex. spatial vision and movement, and an occipito- If we survey the data just discussed, it may temporal pathway or “ventral stream” for ob- be concluded that most of the human neocor- ject identification (Fig. 19.6). Areas along both tex is occupied by association areas of various pathways are organized hierarchically, such kinds and that the boundaries between these that the initial, low-level inputs are trans- areas do not closely correspond to those of the formed into more complex and specific repres- cytoarchitectonic fields, as delineated by Brod- entations through successive stages of process- mann and others. ing. Both pathways project separately to differ- Combining connectional data with the re- ent parts of the prefrontal cortex. There is evi- sults of functional studies has revealed some dence suggesting that parallel processing chan- basic features of the flow of information nels are also present in the central somatosen- through the cortex. These features, which may sory and auditory systems. be designated as hierarchical processing, feed- It should be emphasized that the various back and parallel processing, will now be processing streams do not operate in isolation. briefly discussed. There is substantial intermixing and cross-talk Hierarchical Processing. We have seen that between them at successive levels of process- multisynaptic feed-forward systems can be ing. Association fibres forming lateral connec- traced from the various primary sensory cor- tions between areas at the same processing lev- tices via successive association areas to the el differ from ascending and descending fibres premotor and motor cortices. In light of the in that they terminate in a columnar pattern results of physiological experiments on the vis- involving all cortical layers. It is assumed that ual system [308, 309], it is assumed that the via these interlinking fibres the information sequential processing of information within processed within the individual channels is these feed-forward systems becomes progres- combined into behaviourally relevant percepts. sively more complex and may be characterized Together, the processing streams and their in- on this account as hierarchical [180, 770]. terconnections form distributed hierarchical These feed-forward systems arise largely from networks subserving a distributed hierarchical pyramidal neurons in layer III and terminate processing of information [180, 770]. in and around layer IV of the cortical area they send axons to [623]. 516 Section II Structure of Spinal Cord and Brain Parts

Functional Subdivision fected. He associated this sensory aphasia with damage to the posterior part of the left hemi- The first systematic attempt to localize differ- sphere, in the region where the occipital, parie- ent functions in different regions of the cere- tal and temporal regions meet [811]. It is im- bral cortex was made by Franz Joseph Gall portant to note that Wernicke in his remark- (1758–1828) and his collaborator Johann able 1874 monograph, and in a later textbook Spurzheim (1758–1832), the founders of phre- [812], did not confine himself to attributing nology [187, 256]. Gall and Spurzheim main- the different forms of aphasia to particular tained that the cerebral cortex is composed of areas or centres. He designed circuit diagrams discrete organs or regions that represent differ- for reading, writing and speech and postulated ent mental faculties (“seelische Kräfte”), and that certain disorders of communication could that there are as many such organs as there result from damage to the connections between are mental faculties. They distinguished 35 the various processing centres included in these such faculties, including arithmetic, hope, diagrams. speech, causality, destructiveness and parental During the last decade of the nineteenth love (Fig. 15.16). Gall and Spurzheim, more- century, and the first decade of the twentieth over, suggested that the organs subserving century, several authors, including Hugo Liep- these faculties correspond to prominences of mann (1863–1925) and Joseph Jules Déjerine the overlying skull. On that account, they as- (1849–1917), extended and substantiated Wer- serted to be able to disclose a person’s intellec- nicke’s idea that the interruption of cortico- tual and moral abilities by palpating the cortical fibre connections (“disconnection”) bumps of his/her skull. may cause various kinds of behavioural im- Gall and Spurzheim’s doctrine was strongly pairment. Liepmann [410, 411] published sev- contested by Marie-Jean-Pierre Flourens (1794– eral studies on apraxias, the higher-order dis- 1867), who stated on the basis of extensive ab- orders in the execution of skilled movements. lation experiments that the crucial factor deter- He showed that several sites of damage, includ- mining how an animal organism is affected by ing the corpus callosum and the white matter brain damage is the amount of tissue removed deep to the supramarginal gyrus, could pro- and not its location. In short: only the size and duce apraxia by disconnecting pathways join- not the site of the lesion matters. ing selected sensory, motor and language In spite of its empirical weakness, phrenol- areas. Déjerine [143, 144] analysed the various ogy has had a positive and lasting influence on types of acquired reading disorders (alexias). modern neuroscience because it stimulated the He found, inter alia, that pure alexia, the in- discussion of whether specific functions and ability to read with a preserved ability to qualities of the mind could be localized within write, may be caused by a lesion in the left oc- the convolution of the brain. During the sec- cipital lobe, which disconnects the calcarine ond half of the nineteenth century a number of cortex from the angular gyrus. clinical, anatomical and experimental land- In 1868, John Harlow (1819–1907) reported mark studies appeared that seemed to favour on the case of Phineas Gage, who exhibited localization. profound changes in personality, social con- In 1865, Paul Broca (1824–1880) described duct and emotional stability, following an acci- several patients in whom circumscript lesions dent in which an explosion drove an iron bar at the basis of the third convolution of the left – more than a meter in length and weighing frontal lobe had led to a motor aphasia, i.e. the about 6 kg – through the anterior part of his inability to produce articulated speech [67]. A brain, destroying the prefrontal cortex bilater- few years later, Carl Wernicke (1848–1905) re- ally [278]. Harlow’s case study was followed by ported on a second type of language disorder, several other publications, reporting profound namely the inability to comprehend speech character changes following lesions of the while having speech production relatively unaf- frontal lobes [187]. 15 Telencephalon: Neocortex 517

Fig. 15.16. Gall and Spurzheim’s “phrenological head”, showing 35 mental faculties, residing in specific organs of the cerebral cortex, projected upon the surface of the skull 518 Section II Structure of Spinal Cord and Brain Parts

The landmark experiments in 1870 of Gus- dogs and monkeys lead to blindness [499] and tav Theodor Fritsch (1838–1927) and Eduard during the 1880s, several clinicians identified Hitzig (1838–1907) led to the discovery of a blindness with damage to the occipital cortex motor zone in the cortex. They found that [256]. Salomon Henschen (1847–1930) collect- electrical stimulation of the anterior part of ed over 160 clinical cases of blindness and the cerebral cortex in dogs produces distinct hemianopsia after cortical lesions from the lit- motor reactions on the contralateral side and erature [288]. This vast amount of material led presented evidence suggesting a topographical him to identify the “centre of vision” or “corti- arrangement of muscle groups in the area that cal retina” with the striate cortex [256]. was stimulated [212]. Their experiments were The fact that lesions in the region of the confirmed by David Ferrier (1843–1928) for central sulcus can cause somatosensory distur- various mammals, including monkeys [182]. bances in human patients was recognized in Ferrier also induced lesions in the precentral 1884 by Moses Allen Starr (1854–1932). He no- cortex of monkeys. He noted that the lesions ticed that in the majority of cases of such dis- were associated with paralysis and weakness of turbances, the lesion was posterior to the cen- muscles on the opposite side. The experiments tral sulcus in the postcentral gyrus or the adja- of Fritsch and Hitzig [212] and Ferrier [182] cent parietal region [705]. In the 1890s, Freder- confirmed the inferences on the existence of ick Mott (1853–1926) [493] and Hermann somatotopically arranged motor centres in the Munk [500] reported that, in monkeys, lesions brain made by John Hughlings Jackson (1835– in the region of the central sulcus cause loss of 1911) from observations on patients suffering touch and pressure sensations as well as paral- from a certain type of epileptic seizure, now ysis. Neither Mott nor Munk produced or ana- commonly designated as Jacksonian fits. This lysed lesions confined to the postcentral gyrus. kind of convulsion starts locally in a group of Due to their studies, the prevailing view at the muscles and then spreads in an orderly fashion end of the nineteenth century was that the so- to other muscles [326]. For example, the con- matosensory and motor areas overlap [187]. tractions may begin in the lower part of the During the period just discussed, several in- face and spread to the ipsilateral arm and, in vestigators rejected the idea of cortical local- turn, to the corresponding leg or vice versa. ization. Prominent among them were Charles- During the last quarter of the nineteenth Edouard Brown-Séquard (1817–1894) and century, there was an extensive search for the Friedrich Goltz (1834–1902). Brown-Séquard localization of sensory centres in the cortex [73] rejected Broca’s idea that speech can be (see Finger [187] for a detailed account). As re- localized in the cortex. He presented evidence gards audition, Ferrier [182] and Ferrier and showing that lesions outside Broca’s region can Yeo [183] reported on the effects of lesions of affect speech and he cited patients with dam- the superior temporal gyrus in monkeys. It age to Broca’s area who did not exhibit speech was found that animals with bilateral lesions defects [187]. Brown-Séquard proposed that were totally irresponsive to hearing and that cells subserving speech and other functions animals with unilateral lesions were deaf in the are distributed over many parts of the brain. contralateral ear. In 1891, Charles K. Mills Goltz [248, 249] mistrusted both the results (1845–1931) presented two cases of totally deaf obtained from electrical stimulation of the cor- patients, one with a marked atrophy of both tex and the observations made on animals with superior temporal convolutions, the other with local cortical ablations after short survival severely damaged superior temporal gyri, due times. Goltz performed large and bilateral ab- to successive strokes [471]. These findings lations of the cerebral cortex in dogs and kept were considered to be highly supportive of Fer- these animals alive for a long time. He ob- rier’s experimental findings with monkeys. served that in these animals permanent deficits In 1881, Hermann Munk (1839–1912) re- were limited to higher psychic functions such ported that lesions of the occipital lobes in as intelligence and memory. Like Flourens, 15 Telencephalon: Neocortex 519

Goltz stated that it was the size and not the lo- tary sensory and motor functions, but he em- cation of the lesion that determined the sever- phasized that the answer to the question of ity of its effects on such higher functions what is actually localized remains obscure. He [257]. believed that “Leistungen”, i.e. purposeful ac- In spite of the objections just briefly out- complishments, are not the product of single lined, it may be stated that at the beginning of cortical areas, but rather involve the whole the twentieth century, i.e. at the time that the brain, which functions as a dynamic organiz- architectonists began their mapping studies, ing entity. Von Monakow introduced the con- the idea that at least certain functions can be cept of diaschizis of collaboration, which im- localized to particular areas of the cerebral plies that a breakdown in the widely correlated cortex prevailed. The monograph in which structures of the brain would result in a gross Campbell [85] presented the first cytoarchitec- corresponding defect syndrome instead of the tonic map of the human cerebral cortex was loss of function in the circumscribed area ac- entitled: “Histological studies on the localiza- tually hit by the lesion. As regards central vis- tion of cerebral function”. Brodmann [70, 71] ual disturbances, he maintained that many of and von Economo [795] also emphasized that these are due to interruptions of pathways the cytoarchitectonic fields shown in their rather than to the lesion as such. We will see maps actually represent functional entities. that many of the ideas of von Monakow re- Brodmann stated that the cerebral cortex emerged and became dominant in the second should not be considered as a single organ, half of the twentieth century. but rather a complex of organs, and that each The material of Kleist [373, 374] included, of the different organs (i.e. the separate cytoar- apart from numerous regular clinical cases, chitectonic fields) subserves a particular set of 300 persons who had sustained local brain in- functions. He pointed out that many of these juries during World War I. He summarized his fields coincide with well-known physiological findings concerning the localization of func- centres, specifying that his area 17 co-extends tions in the cerebral cortex in a famous map, with the “clinical visual area of Henschen” and shown here in Fig. 15.17. It will be seen that that his precentral region (area 4) corresponds Kleist subdivided the cortex according to Brod- to the “electromotor zone”. However, Brod- mann (see Fig. 15.8), and that he provided al- mann took issue with the idea of the phrenolo- most all of the cytoarchitectonic areas distin- gists that complex mental functions such as guished by the latter with a functional label. memory, will or fantasy could be localized in The overall functions of the primary sensory circumscribed cortical areas. Rather, he be- and motor areas, which were well known by lieved that these functions result from the con- the end of World War I, were correctly indi- joint activities of a large number of areas dis- cated. However, Kleist went far beyond that by tributed more or less widely over the cortical attributing all sorts of higher cognitive and surface. mental functions and faculties to many other During the twentieth century, the study of areas. Thus, he associated temporal area 21 the effects of brain lesions in human patients with acoustic awareness (“akustische Aufmerk- continued to play a prominent role in the lo- samkeit”), prefrontal area 10 with motor skill calization-anti-localization debate. That the (“motorische Handlungsfolgen”) and orbitofron- outcome of such studies could differ consider- tal area 11 with personal and social ego ably is illustrated by the work of Constantin (“Selbst- und Gemeinschafts-Ich”). It was par- von Monakow (1853–1930) and Karl Kleist ticularly because of this detailed localization of (1879–1960), two clinical neurologists who psychic functions that many of his colleagues both had a vast amount of clinical material at disposed of Kleist’s map as “brain mythology” their disposal. [119]. Uttal [744] recently characterized this Von Monakow [329, 797, 798] admitted that map as a modern manifestation of phrenologi- there was some sort of localization of elemen- cal thinking. 520 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.17 A,B. Localization of functions in the cerebral cortex, according to Kleist [373]. The numbers indicate Brodmann’s cytoarchitectonic areas 15 Telencephalon: Neocortex 521

The most ardent anti-localizationist of the of little value for the planning of experimental twentieth century was, beyond any doubt, Karl work, because the areal subdivisions are in Lashley (1890–1954). He subjected animals, large part anatomically meaningless, and be- mostly rats, to complex maze-learning tests cause individual variation is too great to make and analysed their performance before and the map significant for a single specimen. after damaging various parts of the cerebral They emphasized that the preparation of maps cortex [392, 393]. He found that their perfor- representing functional units required that ob- mance in the maze after operation was entirely jective criteria and standards be developed and independent of the location of the cortical le- that other methods of confirmation be used, sion and only depended on its size. On the ba- particularly the study of fibre connections. sis of his experimental results, Lashley formu- Difficulties with the cytoarchitectonic meth- lated two general principles of brain organiza- od, somewhat similar to those of Lashley and tion, equipotentiality and mass action. With Clark, were experienced by Bailey and von Bo- the term equipotentiality he designated the ap- nin in their cytoarchitectonic studies of the parent capacity of any intact part of a func- cortex of the rhesus macaque [794], chimpan- tional area to carry out the functions lost by zee [29] and human [30]. With regard to the destruction of the whole. He considered it human cortex, they stated that vast areas are likely that this capacity holds only for associa- so closely similar in structure that any attempt tive parts of the cortex and for functions more at subdividing them would be unprofitable, if complex than simple sensory or motor co-ordi- not impossible. It has already been mentioned nation. Lashley pointed out that the equipoten- that the total number of areas distinguished by tiality is not absolute but subject to the princi- Bailey and von Bonin [30] is much lower than ple of mass action, whereby the efficiency of that of other cytoarchitectonists (Table 15.2). performance of a complex function may be re- Braak [64] criticized the results of these au- duced in proportion to the extent of brain in- thors. He stated that they missed or neglected jury within an equipotential area. Overall, obvious cytoarchitectonic boundaries and sus- Lashleys ideas about functional localization pected that they were content with superficial had much in common with those of Flourens observations. and Goltz, but contrary to these authors, he Several authors have attempted to localize accepted the general concepts of specialized functional areas in the cortex by electrical sensory and motor areas in the cortex [187]. stimulation. Vogt and Vogt [784, 785, 787, 788] Lashley not only challenged the functional stimulated the cortex in monkeys (Cercopithe- but also the structural subdivision of the cere- cus) under general anaesthesia, and Foerster bral cortex. He and his collaborator Clark [197] did the same in patients who underwent [394] independently studied the brains of two brain operations under local anaesthesia. They spider monkeys (Ateles geoffroyi), both produc- found that, although the classical motor cortex ing a cytoarchitectonic map. The two maps on stimulation at low strength typically reacted showed little agreement. They then compared with isolated contractions of small muscle their two specimens and found that most of groups, quite characteristic, more complex mo- the differences were not observer-dependent. tor effects could be evoked from many other They noticed that there was considerable varia- cortical areas. Thus, eye and head movements tion in size and appearance of corresponding to the contralateral side could be elicited from areas from brain to brain and that some areas areas 8 and 19, and stimulation of the parietal could be seen in the one brain, but not in the cortex produced synergic flexion of the con- other. Furthermore, they remained unable to tralateral arm and leg (Fig. 15.18 C). Their pro- locate the majority of the neocortical areas rec- cedure of identifying functionally equivalent ognized by other investigators in this monkey areas involved three steps: (1) They plotted the and in the rhesus macaque. Lashley and Clark results of the stimulation experiments in the concluded that standard architectonic maps are Cercopithecus monkeys on a cytoarchitectonic 522 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.18 A–C. Functional mapping by electrical stimulation. A The various effects of electrical stimulation of the cerebral cortex of the monkey Cercopithecus, plotted with particular symbols on a cytoarchitectonic map of the same species. B Comparable stimulation effects, elicited in humans, plotted with the same symbols on a cytoarchitectonic map of the human cortex. C Chart of motor effects elicited by electrical stimulation of the human cerebral cortex. A and B are reproduced from Vogt and Vogt [787]; C is taken from Foer- ster [197] 15 Telencephalon: Neocortex 523 map of the same species, using different sym- tant findings with regard to the functional orga- bols for the various reaction patterns (Fig. nization of the neocortex. Thus, it has been 15.18 A). It appeared that almost all of the shown that most of the post-Rolandic cortex is areas characterized by particular reactions ex- devoted to processing specific aspects of a single actly matched with a particular cytoarchitec- sensory modality and that numerous separate tonic area. (2) Next, they compared the cytoar- areas for processing unimodal sensory informa- chitecture of Cercopithecus monkeys with that tion are present in this vast region. Monkeys of humans and were able to identify and to de- have over 30 cortical areas for processing visual lineate in the latter equivalents of all the struc- information, at least 15 for somatosensory infor- turally and functionally characterized areas in mation, and some 20 for auditory information the monkey brain. (3) Finally, they transferred [353]. Moreover, it has been shown that many the results of the stimulation experiments in of these areas are occupied by orderly represen- humans (Fig. 15.18 C) to their map of the hu- tations or maps of receptor surfaces. man brain, using the same symbols for the Recording the activity from single neurons various reaction patterns as had been used in under particular experimental or behavioural the monkey (Fig. 15.18 B). They concluded that conditions has yielded important clues about many cortical areas in the monkey area are the functional significance of the areas in structurally and functionally equivalent (i.e. which the elements studied are embedded. homologous and analogous) to particular areas Thus, in the visual cortex, there are neurons in the human cortex. It is noteworthy that Vogt that respond selectively to various directions and Vogt [787] proposed to reserve the term of motion, specific colours, the orientation of cortical field (“Rindenfeld”) for areas repre- line segments, the density of visual texture and senting both structural and functional units. many other visual features [313, 536]. In the Electrical stimulation experiments on the extrastriate infratemporal visual area IT, many human cerebral cortex in awake patients un- cells respond only or best to highly complex dergoing brain surgery have also been per- stimuli and some are selective for faces [258]. formed by Penfield and his associates [544, Cells in the anterior premotor cortex play a 545, 548]. These experiments have extended role in the preparation of movements [828], our knowledge concerning the somatotopic or- whereas the activity of neurons in certain parts ganization of the primary somatosensory and of the prefrontal cortex is correlated with de- primary motor areas considerably. They layed-response performance and hence with showed that, in both of these cortical areas, working memory [215]. the different parts of the body are represented Returning to the systems level of functional in territories of very different sizes. Thus, in localization in the neocortex, it should be the somatosensory cortex, the face and hand mentioned that in 1965 Norman Geschwind areas are very large, whereas the trunk and the (1926–1984) published a series of seminal pa- proximal parts of limbs occupy relatively small pers on disconnection syndromes in animals territories (Fig. 15.19 A). In the motor cortex and humans [227]. Referring to the studies of those parts of the body capable of performing Wernicke, Liepmann and Déjerine (see above), the most differentiated and delicate movements he posited that many disorders of the higher have the largest representation (Fig. 15.19 B). functions of the nervous system, such as the Woolsey and collaborators [830–833] have aphasias, apraxias and agnosias, may be con- carefully mapped the somatosensory and mo- sidered disturbances produced by anatomical tor cortical areas in a variety of mammals, disconnection of primary receptive and motor using electrical stimulation and also recording areas from one another. He emphasized the evoked cortical potentials. These studies have importance of long cortico-cortical connec- given results concordant with those in humans. tions emanating from the unimodal sensory Studies using single-cell recordings and elec- association areas and pointed out that discon- trical stimulation of cells have generated impor- nection of cortical areas can be achieved by le- 524 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.19. Diagrams showing the relative proportion of the representation of the different parts of the body in (A) the postcentral primary somatosensory cortex, and (B) the precentral primary motor cortex. From Penfield and Rasmussen [548] 15 Telencephalon: Neocortex 525 sions involving either white matter connections the early-nineteenth-century, phrenology-based or by damage to association areas that consti- maps (Fig. 15.16), as well as to the early-twenti- tute obligatory way stations between the pri- eth-century, clinical evidence-based map of Kle- mary sensory, motor, and limbic regions of the ist (Fig. 15.17). All of these maps show detailed brain. Herewith Geschwind provided a general functions, allocated to distinct territories of the conceptual framework for higher brain distur- cortex. However, an important difference is that bances, based on the neocortical wiring pattern the phrenology maps and the Kleist map, con- (Fig. 15.20). Geschwind’s papers became the trary to neuroimaging maps, are typical “one- manifesto of behavioural neurology and played function–one-brain-area maps”. an important role in the development of the It stands to reason that, in order to become current concept that networks of intercon- scientifically meaningful, the raw products of nected multiple specialized cortical territories brain imaging, i.e. the activity-related blobs form the morphological substrate of higher and patches, have to be placed in the context brain functions [3, 92, 464, 465]. of the structural and functional architecture of It is noteworthy that, according to Catani the brain. As regards morphological architec- and ffytche [92], not only disconnection, but ture, many authors confine themselves to indi- also local hyperconnectivity or combinations cating the location of the spots of activity de- of these two abnormalities may play a role in tected in terms of the gross anatomy of the the pathogenesis of neurological and psychia- cortex, for instance: activity in “the inferior tric disorders. Thus, it has been suggested that, parietal lobule” or in “the posterior part of the in autism, a combination of fronto-frontal hy- superior frontal sulcus”. Many others use Brod- perconnectivity with frontal disconnection mann’s map (Fig. 15.8), as extrapolated by Ta- from other brain regions may be present [115]. lairach and Tournoux [727] onto a standard- Functional neuroimaging techniques such as ized brain atlas, as a frame of reference. Given positron emission tomography (PET) and the considerable interindividual variability of functional magnetic resonance imaging (fMRI) the human brain at both the gross anatomical play a prominent role in localizing functions and the structural levels, neither of these two and functional complexes in the neocortex. approaches is entirely satisfactory [66]. We will Maps constructed by these techniques depict come back to this issue in a later section. neural activity as coloured spots or blobs, Insight into the functional architecture of either on the surface of the brain or on brain the neocortex may be gained by considering sections in various directions. PET and fMRI the following features of neuroimaging data. monitor changes in blood flow or metabolism 1. In most cognitive tasks, two or more corti- in specific regions of the brain while human cal areas are activated [83]. The activated subjects perform various sensory, motor or areas may be considered as nodal points in cognitive tasks. The blood flow signals and the networks underlying the different cogni- metabolic signals reflect changes in neuronal, tive processes being investigated. particularly synaptic, activity [602]. 2. Reviews of functional neuroimaging results Neuroimaging studies have not only con- across different cognitive domains clearly firmed and extended previous knowledge about show that cortical regions, such as the pre- localization of function in the primary sensory frontal cortex and the parietal area 7, are and motor cortical areas, but have also demon- engaged in a wide variety of cognitive de- strated a pervasive form of localization in a wide mands. The most parsimonious explanation variety of cognitive tasks [83]. The activity of of this kind of activation is that they reflect identifying and localizing cortical areas related cognitive processes that are tapped by tasks to specific sensorimotor or behavioural events, in different domains [83]. However, it is also using imaging techniques, is generally denoted conceivable that more refined analyses will as brain mapping. The resultant maps or charts lead to a further functional parcellation of (Fig. 15.21) show a striking resemblance to the areas involved. 526 Section II Structure of Spinal Cord and Brain Parts

Pure alexia

Fig. 15.20. Geschwinds’ disconnection syndromes. Cortico-cortical pathways in the human brain and the disconnection syndromes to which their interruption may lead. Geschwind and his predecessors Liepmann and Déjerine also indicated that damage to sectors of the corpus callosum may also lead to typical disconnection syndromes. These have been omitted from this scheme (based on [92]). ANG+ SM, angular and supramarginal gyri; AU, auditory cortex; B, Broca’s speech area; PREMOT, premotor cortex; SS, somatosensory cortex; VIS, visual cortex; W, Wernicke’s speech area

a Attention [684] l Coherence processes in t Selective auditory attention [413] b Calculation [684] language comprehens- u Evaluation of unpleasant, arousing c Activation arising from a working ion [184] words [435] memory task [66] m Executive function in working v Cross-modal sensory processing d Processing of incorrect arithmetic memory [523] [396] equations [461] n Embarrassment [724] w Error processing [460] e Maternal love [39] o Perception of visual motion: V5 x Categorization of written object de- f Performance of a verbal working [808] scriptions [260] memory task [583] p Written sentence comprehension y Pursuit eye movements [50] g Reading aloud irregular words [109] z Painful visceral sensation [662] [289] q Processing of color patterns [38] a1 Saccades [684] h Meditative state [418] r Verb comprehension [259] b1 Visual stimulation [506] j Finger opposition task [713] s Judgement of the grammaticality of k Visual-spatial orienting [450] sentences [807] 15 Telencephalon: Neocortex 527

Fig. 15.21. Activation loci, found in a number of functional imaging experiments, projected upon the lateral (A) and medial (B) surface of the cerebral hemisphere. For significance of letters, see opposite pape 528 Section II Structure of Spinal Cord and Brain Parts

3. Combining neuroimaging data with the re- detailed functional label (Fig. 15.17), and sults of single-cell recordings in non-human that the Vogts sought to identify structural- primates may help determine the functional functional units in the neocortex of monkey specialization of a given area in more detail. and humans, by combining the results of cy- 4. The functional role played by any cortical toarchitectonic analyses with those of elec- area is defined largely by its connections. trical stimulation experiments (Fig. 15.18). Thus, it is important to complement the re- 3. The primary sensory and motor areas are sults of neuroimaging experiments with structurally and functionally distinct. They studies aimed at determining the afferent can be easily identified as heterotypical and efferent connections by which the areas agranular and granular areas (types 1 and 5 (spots of activity) detected are embedded in of von Economo: Fig. 15.10), and as early the cortical network. maturing myelogenetic projection areas (Fig. 15.14). Hodologically, the primary sensory Although techniques have recently become areas are defined by receiving the thalamo- available to globally trace the tracts of the liv- cortical endsegments of the great sensory ing human brain [40, 108, 335, 488, 588], infor- projections, whereas the primary motor cor- mation concerning the connectivity of particu- tex is characterized by its massive cortico- lar areas in the human cortex still must be de- bulbar and corticospinal projections. rived and extrapolated from experimental 4. The remainder of the neocortex is formed studies in monkeys. Nevertheless, we have at by Flechsig’s late-maturing association cor- present a fairly accurate picture of the connec- tex. In the vast region occupied by this cor- tional frameworks involved in a large number tex, the structural differences between the of specific cognitive and behavioural opera- different areas are rather subtle, which ex- tions, such as spatial orientation, object recog- plains why the results of the various cytoar- nition, language [464, 465], attention [553] and chitectonists are less concordant here than decision forming [522]. The neural circuitry in the primary areas. All of the areas form- related to many of these operations is not con- ing part of the association cortex occupy, as fined to the neocortex, but also includes one homotypical areas, an intermediate position or several subcortical centres. The cerebellum in von Economo’s classification system (Fig. is consistently activated in a variety of cogni- 15.10). tive processes, but the nature of its role in 5. Experimental studies on the cortico-cortical these processes is still obscure [83]. connections are and have been of para- mount importance for both the morphological and functional interpretation of the asso- Structural and Functional Subdivision: ciation cortex. However, because such stud- Overview ies cannot be performed on humans, they have to be carried out on laboratory ani- 1. The structure of the neocortex is not homo- mals, preferably monkeys. Hence, in prac- geneous throughout. tice, the acquisition of data on the fibre con- 2. Many authors have attempted to subdivide nections of a given area in the human cortex the neocortex or parts thereof into structur- involves the following four steps: (1) identi- al entities, using cytoarchitectonic and/or fying and locating an area in the human myeloarchitectonic criteria. Most of these cortex; (2) identifying and locating the authors considered the resultant architec- homologue of that human area in the cortex tonic areas or fields not merely as structural of the monkey; (3) establishing the fibre entities, but also as functional units. We connections of that area in the monkey cor- have seen that, on the basis of clinical evi- tex by using experimental tract-tracing tech- dence, Kleist provided nearly all of the areas niques, and (4) extrapolating the results of delineated by Brodmann (Fig. 15.8) with a these experiments to the human brain. A 15 Telencephalon: Neocortex 529

serious limitation of this indirect method is Localizationism claims that each region of the that many areas in the human neocortex neocortex represents an independent organ, have no obvious homologue in the monkey dedicated to a complete and distinct function. (Figs. 15.8, 15.12) [120]. This idea was first put forward by the phrenol- Experimental studies in the monkey have ogists Gall and Spurzheim (Fig. 15.16), culmi- shown that (a) the primary somatosensory, nated in the work of Kleist (Fig. 15.17) and is visual and auditory cortical areas project via still tangible in many current mapping studies short connections to adjacent unimodal as- aimed exclusively at localizing functions. Such sociation areas; (b) these unimodal sensory functional cartography with the aid of neuro- association areas project in turn to two imaging techniques has been characterized by strips of heteromodal sensory association some [508, 601, 744, 745] as “neophrenology”. cortex, situated in the lateral and medial Anti-localizationism or holism holds that parts of the parietal and temporal lobes; (c) higher cognitive and mental functions are dis- the unimodal and heteromodal sensory as- tributed homogeneously throughout the cortex sociation areas project massively to the pre- and require the integrated activity of the entire frontal cortex which, on this account, should cortex. The degree of cognitive or behavioural be considered as a higher association cortex, impairment following a cortical lesion is sim- and (e) sequences of short connections suc- ply proportional to the amount of tissue de- cessively link the various prefrontal areas, stroyed, with localization being entirely irrele- the premotor area and the primary motor vant. The most prominent anti-localizationists area (Fig. 15.15). It is important to note that were Flourens, in the early nineteenth century, the boundaries between the various associa- Goltz, in the late nineteenth century, and Lash- tion areas thus defined do not show a close ley, in the twentieth century. correspondence to those of the cytoarchitec- The presently prevailing concept of (dis)con- tonic fields, as delineated by Brodmann and nectionism has its roots in the clinical work of others. Wernicke, Liepmann, Déjerine and Geschwind 6. A combination of connectional data with the and has been further developed by Mesulam results of microphysiological studies has [463–465], Van Essen [180, 770], Friston [201, shown that the wiring of the neocortex may 211] and many others. It states that, although be characterized as a distributed hierarchi- functional localization and functional speciali- cal network that contains numerous inter- zation are important principles, they do not of- twined processing streams. fer a complete or sufficient explanation of cor- 7. Unit activity recordings, the effect of lesions tical organization. The physiological changes and particularly the results of neuroimaging elicited in the cortex by particular cognitive studies have firmly established functional processes or behaviours should be explained in segregation as a principle of neocortical or- terms of distributed patterns of changing neu- ganization, not only in the primary cortices, ral activity in networks of interconnected, but also in the association regions (Fig. functionally specialized areas. Cognitive and 15.21). This having been said, it should im- mental abilities are not the product of single mediately be added that the precise morpho- and separate cortical regions, but rather result logical identification of the functionally seg- from the functional integration of the elemen- regated domains is a difficult problem and tary processing operations occurring in a much current research is devoted to this smaller or larger number of functional areas. (see below). The neural mechanism of many neurological 8. As regards the functional organization of disorders can be understood as dysfunction the neocortex, there are three different con- within specific neural circuits. cepts or paradigms: localizationism, anti-localizationism or holism and (dis)connec- tionism. 530 Section II Structure of Spinal Cord and Brain Parts

Structural and Functional Localization in the the macaque monkey. A similar architectonic Neocortex: Current Research and Perspectives analysis of the human OMPFC, carried out by Öngür et al. [518], revealed that all of the areas Current research on the subdivision of the recognized in the macaque monkey have coun- neocortex concentrates on the following two terparts in humans. On the basis of these results interrelated questions: (1) what are the true and of studies on the fibre connections of the and fundamental neocortical units, and (2) various areas in the monkey [88, 89], Öngür et how do we determine and visualize the loca- al. concluded that each of the cortical areas dis- tion and extent of these units. tinguished is a module with specific input-out- As regards the first question, we have seen put relations and a unique role in information that Campbell, Brodmann, the Vogts and von processing. They considered it likely that much Economo held that the cytoarchitectonic and of the cortex consists of such discrete structural myeloarchitectonic fields they delineated repre- and functional modules. sent fundamental structural as well as functional Observer-independent procedures in the units. Although there is much respect and ap- structural parcellation of the cortex include preciation for the work of the classical architec- computer-assisted quantification of cortical tonists, it is generally felt that their analyses morphology in terms of (a) cell density, neu- show two serious methodological shortcomings ron size and laminar thickness [605, 659, 660]; and, hence, do not provide a reliable baseline for (b) packing density and laminar distribution further studies on the organization of the cortex. of myelinated fibres [12]; and (c) receptor ar- These shortcomings are: (1) the classical parcel- chitecture, i.e. the density and laminar distri- lations are based on the study of microscopic bution of different receptor types as visualized sections stained according to a single technique, by autoradiography [230, 851]. Areal bound- either the Nissl technique for neuronal perikar- aries are fixed at locations where these features ya or the Weigert technique for myelinated fi- change significantly. bres, and (2) the areal boundaries were estab- Roland and Zilles [626] expect that combin- lished using purely visual inspection of histolog- ing the results of quantitative morphological ical sections and were, hence, influenced by studies that apply observer-independent proce- varying observer-dependent conditions, such dures with data derived from functional neuro- as individual abilities in pattern recognition imaging studies will lead to the detection of [660]. In current architectonic studies these functional cortical fields. They advance the hy- methodological shortcomings are avoided by pothesis that the organization of the cortex is using several complementary histological tech- based on such functional fields, each occupy- niques [747] and/or applying observer-indepen- ing a certain, relatively large territory of the dent procedures [659, 660]. Thus, Carmichael cortex, and postulate that all neurons and sy- and Price [86] presented an architectonic divi- napses within these fields perform a co-opera- sion of the orbital and medial prefrontal cortex tive computation. (OMPFC) of the macaque monkey based on nine A preliminary reconnaissance of the current different cytoarchitectonic, myeloarchitectonic literature on the neocortex, aimed at detecting and immunohistochemical stains. These stains entities comparable to, or at least foreshadow- included Nissl, myelin and acetylcholinesterase, ing the modules of Öngür et al. [518] or the and immunohistochemical stains for parvalbu- functional fields of Roland and Zilles [626], min (PV), calbindin (CB) and a membrane- yields the following results: bound glycoprotein. A cortical area was defined 1. Just as the classical architectonists indicated, as distinct if it was differentiated in at least three the human neocortex can be subdivided into different stains. This analysis subdivided many a number of juxtaposed structural fields. of the areas originally described by Brodmann 2. Recent investigations on the primary sen- [70] and Walker [802] and resulted in the iden- sory and motor areas generally confirm the tification of 22 different areas in the OMPFC of results of the classical architectonists [7, 15 Telencephalon: Neocortex 531

231, 598–600], although two subfields [230] mine and visualize the location and extent of have been detected within the primary mo- structural and functional units – has to do tor cortex and three subfields [489] within with two further shortcomings, or rather limi- the primary auditory cortex. tations, of the work of the classical architecto- 3. Detailed quantitative analyses have con- nists. Firstly, these studies were all based on a firmed the presence of Brodmann’s frontal single or a few brains and, hence, interindivid- areas 9 and 46 [604, 605] and 44 and 45 [6, ual variations in the position and extent of the 8] as distinct architectonic entities. various architectonic areas were completely ne- 4. In many regions of the neocortex, including glected. Secondly, the classical architectonists the orbitofrontal cortex [518, 751], the pari- only analysed the superficially exposed parts of etal cortex [849], the cingulate region [783], the cortex, leaving the sulcal cortices largely the retrosplenial region [376] and the extra- unexplored. This is a serious limitation indeed striate visual region [352, 408, 850], the because, as already mentioned, in humans al- number of architectural areas detected is most two thirds of the cortex are hidden away larger, and sometimes considerably larger in the depths of the sulci. than in the classical studies. As regards the first limitation, the pioneer- 5. Taking these data together, we estimate that ing studies of Filimonov [186] on the visual about 150 juxtaposed structural and poten- areas 17, 18 and 19, and of Kononova [383] on tially functional entities are present in the the frontal areas 44 and 45 have brought to human neocortex. light that these areas show a considerable intersubject variability (Fig. 15.22). Their obser- Comparative architectonic studies [516, 518, vations have been fully confirmed by later 577, 580, 581, 605] have shown that many of studies on the same areas [6–8]. Similar varia- the delineable structural entities in the human bility has also been reported for many other neocortex have distinct counterparts (homolo- cortical areas, including the primary motor gues) in the neocortex of the macaque monkey. [230, 598, 599] and somatosensory cortices The results of experimental studies on the af- [231, 232], the primary auditory cortex [489, ferent and efferent connections of these com- 600], the prefrontal areas 9 and 46 [604, 605] parable areas in the cortex of the monkey can and the orbitofrontal cortex (Fig. 15.23) [751]. be transferred to the human brain and may Interindividual variability appears to be a provide important clues as to the functional general feature of neocortical architectonic areas significance of the entities involved. Using data and it should be added that this microstructural collected in the macaque connectivity database variation is superimposed upon the also consid- CoCoMac [710], Passingham et al. [541] dem- erable macrostructural variation pertaining to onstrated that each cortical area has a unique the overall size and shape of the hemispheres pattern of cortico-cortical connections, a defin- as well as to the sulcal and gyral pattern. It will ing “connectional fingerprint”. They suggested be clear that this variability seriously hampers that the connectional fingerprints underlie the the establishment of spatial relationships be- cell-firing differences observed between areas tween structural and functional domains. during different tasks or task events. Passing- In human brain imaging studies, it is com- ham et al. [541] designate this pattern as a mon practice to determine the location of the “functional fingerprint”. They indicate that activation foci detected by transferring these functional brain imaging will be a useful tool loci to the three-dimensional version of Brod- for detecting such functional fingerprints be- mann’s chart, incorporated in the stereotaxic cause it enables us to compare activations reference system of Talairach and Tournoux across many cortical areas and across a wide [727]. Because the considerable structural range of tasks. variability of the cortex is completely neglected The second central questions in current rein this procedure, it is apt to lead to erroneous search on the neocortex – how do we deter- conclusions [748]. 532 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.22 A,B. Interindividual variability of cytoarchitectonic areas in normal human adults. A Prefrontal cortical areas, among them areas 44 and 45, which together form Broca’s motor speech region; B the visual cortical areas 17, 18 and 19. Reproduced from [748]. References cited: Kononova [383], Filimonov [186] 15 Telencephalon: Neocortex 533

Fig. 15.23. Cytoarchitectonic (sub)areas covering the orbital surface of the frontal lobe in three different brains. Note the considerable interindividual variability (cf. Fig. 15.11) (based on Uylings et al. [751]) 534 Section II Structure of Spinal Cord and Brain Parts 15 Telencephalon: Neocortex 535

Fig. 15.24 A–C. Flat maps of the surface of the left human hemisphere, according to Van Essen [765]. In order to avoid gross distortions, several cuts have been made in the regions corresponding to the medial surface of the hemisphere. The dashed, arrow-headed curves in (A) connect corresponding points. In (A) the sulcal surfaces are shaded and the principal sulci are labelled. In order to facilitate orientation, outlines of the structures visible in a lateral view of the hemisphere are shown in red.In(B) The cytoarchitectonic fields of Brodmann (see Fig. 15.8) have been extended over the sulcal parts of the neocortex. In (C) the results of a detailed architectonic study of the orbital and medial prefrontal cortex [518] and of a number of functional imaging studies on the localizations of visuotopic areas [275, 590, 736, 764] have been superimposed on Brodmann’s subdivision. It is noteworthy that the 30 (sub)areas delimited, together occupy 20% of the neocortical surface. Extrapolating from these data, it may be estimated that the neocortex is composed of some 150 (sub)areas or units. calc s, calcarine sulcus; central s, central sulcus; cing s, cingulate sulcus; collat s, collateral sulcus; fos, fronto-orbital sulcus; ifs, inferior frontal sulcus; inf temp s, inferior temporal sulcus; ips, intraparietal sulcus; occ temp s, occipitotemporal sulcus; olf s, olfactory sulcus; pos, parieto-occipital sulcus; postcentr s, postcentral sulcus; precentr s, precentral sulcus; sup fr s, superior frontal sulcus; sup temp s, superior temporal sulcus; tr orb s, transverse orbital sulcus 536 Section II Structure of Spinal Cord and Brain Parts

In order to cope with the problem of inter- and colleagues [160, 180, 762–764, 766, 767]. subject variability of brain structures, compu- These authors have worked out computerized ter-based, three-dimensional probability refer- procedures by which the entire surface of a ce- ence systems have been developed that incor- rebral hemisphere can be unfolded and flat- porate data on the gross anatomy of a group of tened into a two-dimensional reconstruction or brains and provide an adequate probabilistic flat map. Three such flat maps are shown in framework for both microstructural and func- Fig. 15.24. In the first (Fig. 15.24 A), the sulcal tional studies [452–454, 624, 625, 627]. Spatial surfaces are shaded and selected sulci are la- normalization procedures render it possible to belled. In the second (Fig. 15.24 B), the cytoar- fit structural data derived from individual chitectonic fields, which in Brodmann’s chart brains (or groups of brains) into this reference (Fig. 15.8) cover the superficially exposed parts system and the same holds true for the results of the hemisphere, have been extended over of neuroimaging studies [66]. It is necessary to the sulcal domains. This map should be con- superimpose structural and functional data sidered as provisional and tentative because: within the reference system so that structural- (1) the borders of the sulcal parts of the fields functional relationships can be rigorously have not been determined, but rather esti- tested. This procedure provides the basis for mated, and (2) data concerning the interindivi- making sound statistical statements concerning dual variability of the fields have not been in- the probability that a particular activity focus cluded. In the third flat map (Fig. 15.24 C), the lies within a certain architectonic unit. results of a detailed architectonic study of the It will be appreciated that the generation of orbital and medial prefrontal cortex [268, 518] probability maps of all of the, say 150, struc- and of a number of fMRI studies on the local- tural units of the human neocortex is a for- ization of visuotopic areas [275, 590, 736] have midable task. Two groups of neuroanatomists, been superimposed on Brodmann’s parcella- that of Zilles and Amunts [6–8, 97, 169, 230– tion. Flat maps allow the entire neocortical 232, 489, 490, 599, 600], and that of Uylings surface to be seen in a single view; in the pres- [747, 748, 750, 751] are currently working on ent authors’ opinion, therefore, they are very this mega-project, which will take at least 12 well suited for visualizing the results of such years to complete. Within the framework of studies. If the data concerning the localization this project, the neocortex of 10–15 normal hu- and variability of the approximately 150 neo- man subjects will be analysed, using various cortical architectonic units, once established staining techniques and observer-independent (see above), would be transferred to a surface- procedures. If the final results of this project based atlas, this mega-operation would yield a matched with the results of large series of care- very useful probabilistic framework for the fully carried out functional imaging studies topical specification of fMRI results and other would yield a (probabilistic) functional profile functional data. A probabilistic map of the hu- for all of the architectural units delineated, this man visual cortex is already available [771]. neophrenological achievement would still represent no more than a first step in the analysis of the structural-functional relationships in the neocortex. The next step would be to systema- Neocortical Afferents tically analyse how these various units or centres are integrated in the circuits and processing streams subserving higher cognitive and In the neocortex, the fibres of all extrinsic af- mental functions. ferent systems follow a radial course. Golgi The final problem left by the classical archi- studies and experimental investigations using tectonists, i.e. the representation and charting anterograde degeneration and tracer tech- of the vast neocortical areas hidden in the sul- niques have shown that most of these afferent ci, has been successfully tackled by Van Essen systems, after having entered the cortex from 15 Telencephalon: Neocortex 537 the deep white matter, distribute their terminal spread projections to deep cortical layers (lam- branches preferably in one or more layers. inae V, VI or both). The third class encom- Lorente de Nó [414, 416] provided a detailed passes a number of nuclei that share a pattern description of thalamic afferent fibres in the of dense, widespread projections to lamina I, rodent neocortex. His Golgi material displayed though terminations in other laminae may or two different laminar distributions of terminal may not be present. Because these nuclei pro- arborizations: the “specific” distribution, which jecting to lamina I generally occupy a position is densely aggregated in lamina IV, with or adjacent to the intralaminar nuclei, they were without extension to lamina III (a, b in figure collectively designated by Herkenham [293] as 15.25 A), and the “unspecific” distribution, the paralaminar nuclei. The finding that differ- which is sparsely distributed throughout all ent (groups of) thalamic nuclei project in a cortical layers, but appears predominantly in particular laminar fashion to smaller or larger lamina VI (c in Fig. 15.25 A). The “specific” af- parts of the neocortex has been confirmed by ferents were considered to originate from “spe- a large number of studies on different animals, cific” thalamic sensory relay nuclei, such as although with regard to the projections of cer- the medial and lateral geniculate bodies and tain nuclei different results have been obtained the ventral posterior nucleus, and to terminate by different authors. Thus, Berendse and Groe- in specific sensory cortical fields. The “unspe- newegen [49], who studied the cortical projec- cific” afferents, which were observed to extend tions of the midline and intralaminar thalamic over multiple cortical fields, were believed to nuclei in the rat using PHA-L as an antero- originate from other, as yet undetermined tha- grade tracer, reported that the fibres originat- lamic sources. Frost and Caviness [213] stud- ing from the intralaminar nuclei do not termi- ied the intracortical distributions of the projec- nate exclusively in the deep cortical layers (as tions of a number of different thalamic loci to reported by Herkenham), but also project to the neocortex in the mouse using an axon de- lamina I. Corresponding findings have been re- generation technique. They found that in this ported for the cat and the monkey [209, 364, animal virtually the entire tangential extent of 640, 641]. the neocortex receives a projection from the The cortical projections of some thalamic thalamus. The areas of termination of thalamo- centres have been analysed in detail by placing fugal axons appeared to be segregated into multiple small focal injections of anterograde three tiers: an outer tier in layer I, a middle tracers within their confines or by introducing tier in layer IV and/or III and an inner tier in horseradish peroxidase (HRP) into the cortex layer VI. In most fields, terminating axons be- using a micropipette after recording from indi- longing to the middle or the inner tier or both vidual thalamocortical fibres. Such an analysis were found to extend over some distance into of the ventroanterior-ventrolateral complex of layer V. Herkenham [291–293] examined the the thalamus of the cat revealed, for instance, cortical projections of individual thalamic nu- that fibres originating from the ventrolateral or clei in the rat using tritiated amino acids as the caudal part of this complex are distributed anterograde tracers. He found that the thalamic in laminae I, III and IV of the parietal cortex, nuclei can be grouped into three classes ac- whereas fibres arising from rostral or dorso- cording to the laminar patterns of their corti- medial portions of the complex are almost al- cal projections. The first class includes the tha- ways confined to lamina I [356]. However, the lamic relay nuclei for vision, audition and so- most detailed information of this type present- matic sensibility. Their cortical projections ter- ly available concerns the projection from the minate mainly in lamina IV, lamina III or both. lateral geniculate body to the primary visual The fibres of this class clearly correspond to cortex in primates. In this group, the lateral the “specific” afferents of Lorente de Nó [416]. geniculate nucleus is a laminar structure in The second class includes the intralaminar tha- which separate magnocellular, parvocellular lamic nuclei, which issue sparse but wide- and intercalated zones can be distinguished. In 538 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.25 A–C. Neocortical afferents. A “Specific” (a,b) and “unspecific” afferents (c,d) as observed in Golgi material of the somatosensory cortex of the mouse. Left, same cortical area as observed in Nissl preparations (based on Lorente de Nó [414]). B Cortico-cortical afferent fibre arborization in the auditory cortex of the macaque monkey (redrawn from Szentágothai [721]). C Callosal axons in the somatosensory cortex of the same species, anterogradely labelled by horseradish peroxidase injected in the corpus callosum. Redrawn from Jones [339] 15 Telencephalon: Neocortex 539 the primate primary visual cortex, the laminar ing the nucleus basalis of Meynert, which pattern is very distinct, and lamina IV can be sends numerous cholinergic as well as GABA- divided into four subzones, designated as IVA, ergic fibres to the cortex (Fig. 14.15); (4) the IVB, IVCa and IVCb (Fig. 15.26). Investigations hypothalamus, comprising orexinergic, hista- of many different authors, summarized by Fitz- minergic and melanin-concentrating hormone patrick et al. [189, 190] and Lund [425], re- (MCH)-containing neocortically projecting vealed that there are at least six discrete popu- neuron groups (Figs. 10.5, 10.7 B, 14.15); (5) lations of geniculocortical axons, differing the mesencephalic dorsal raphe and central su- markedly from one another in laminar distri- perior nuclei, which provide the entire neocor- bution and tangential spread (Fig. 15.26): tex with richly ramifying serotoninergic fibres; 1. Coarse fibres originating from the magno- (6) the ventral tegmental area and the substan- cellular layers of the lateral geniculate nu- tia nigra, sending dopaminergic axons mainly, cleus projecting to lamina IVCa with wide- but not exclusively, to the prefrontal cortex; spreading terminal fields that span the en- and (7) the locus coeruleus in the rostral pon- tire depth of the lamina. tine tegmentum, which is the source of nora- 2. Fibres resembling those mentioned above, drenergic fibres spreading over the entire neo- the terminal fields of which are, however, cortex. confined to the upper half of lamina IVCa. The claustral and amygdaloid efferents have 3. Axons originating presumably from both the been discussed in Chap. 13. The neocortical parvocellular and the intercalated layers of projections from the basal forebrain, the hypo- the lateral geniculate nucleus forming small, thalamus and the upper brain stem together dense clusters of terminal branches in lami- constitute the ventral branch of the ascending na IVA and contributing some rising collat- arousal system (Figs. 10.5, 14.15). A detailed erals to the adjacent zone of lamina III. discussion of these various projections is be- 4. Fibres originating from the intercalated yond the scope of the present work. As far as layers of the lateral geniculate nucleus, the the cholinergic, GABAergic and monoaminer- terminal arborizations of which participate gic projections are concerned, it may be stated in the formation of small, dome-shaped for- that, although all of these projections innervate mations in lamina III. the entire cortex, they are not to be considered 5. Fibres from the parvocellular layers forming as “diffuse and non-specific” [538]. Rather, small, dense terminal fields in lamina IVCb. they show a high degree of anatomical specific- 6. Fine fibres likewise originating from the ity, both for particular cortical areas and for parvocellular layers of the lateral geniculate particular laminae within a single cortical area nucleus, extending over large distances hori- (Fig. 15.27 A–C) [198, 264, 331, 467, 492, 651]. zontally, that contribute terminals to upper In some regions, for instance the primary vis- lamina VI and to lamina I (for a detailed ual cortex, different monoamines may show discussion of the visual system the reader is clearly complementary laminar patterns of in- referred to Chap. 19). nervation (Fig. 15.27 C,D). These data strongly suggest that the effects of the cholinergic, GA- Although the thalamus is a major source of BAergic and monoaminergic systems are not subcortical input, it is not the only one. In generalized excitation or inhibition but rather fact, more than ten different extrathalamic subregion-specific enhancement or diminution of cortical structures projecting to the neocortex activity in limited neuronal ensembles during have been identified [734]. In most of these, certain stages of information processing [198]. the nature of the neurotransmitter utilized by Immuno-electron microscopic studies have their constituent neurons has been established. shown that cholinergic [13, 43, 307], dopamin- These structures include: (1) the claustrum ergic [677], noradrenergic [534] and serotonin- (Fig. 13.8); (2) the basolateral amygdaloid nu- ergic axons [141, 535] in the neocortex form clei (Fig. 13.7); (3) the basal forebrain, includ- synaptic contacts with pyramidal as well as 540 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.26. Distribution of thalamic inputs from the lateral geniculate nucleus to the primate primary visual cortex of the macaque monkey. M1, M2, input from magnocellular layers; P1–P3, input from parvocellular layers; I1, I2, input from intercalated layers. The ocular dominance bands are indicated by dashed lines, 400– 500 lmapart.Left, same cortical area as observed in Nissl preparations. Redrawn from Lund [425] 15 Telencephalon: Neocortex 541

Fig. 15.27 A–D. Monoaminergic innervation of three different regions of the neocortex of the squirrel monkey. A–C Noradrenergic (NA) innervation of the dorsolateral prefrontal cortex (areas 9, 10), the primary somatosensory cortex (areas 3, 1, 2) and the primary visual cortex (area 17), respectively. D Serotoninergic (SER) innervation of the primary visual cortex. Bars represent 200 lm. Slightly modified from Morrison and Magistretti [491] 542 Section II Structure of Spinal Cord and Brain Parts with (GABAergic) non-pyramidal cells. Descar- proposed to optimize task performance (ex- ries and colleagues have devoted a long series ploitation). When utility in the task declines, of publications, summarized in [148], to the locus coeruleus neurons exhibit a tonic activity synaptology of cholinergic and monoaminergic mode, associated with disengagement from the axons. They found that, in the neocortex, these current task and a search for alternative behav- axons are provided with varicosities and term- iours (exploration). The locus coeruleus is sup- inals which, although filled with typical “syn- posed to produce these patterns of activity un- aptic” vesicles, quite often lack the membrane der the control of the cingulate and orbitofron- functional complexes that are typical of sy- tal cortices, both of which project strongly to napses. They consider it likely that these non- it. junctional structures are involved in non-syn- From studies of the cognitive context gov- aptic or volume transmission. Evidence in sup- erning the activity of locus coeruleus neurons port of the occurrence of this mode of neuro- in rats and monkeys, Bouret and Sara con- transmission in the neocortex has been pre- cluded that these neurons are activated in situ- sented by Smiley and Goldman-Rakic for both ations requiring interruption of on-going be- the serotoninergic [692] and for the dopamin- haviour and rapid adaptation. This locus coer- ergic projections [691]. uleus activation appeared to occur whenever It is known that the various ascending ex- there is a change in the environmental impera- trathalamic pathways modulate the responsive- tive, such as the appearance of a novel, unex- ness of cortical neurons that process sensory pected sensory event. They suppose that the input, co-ordinate motor output and perform noradrenaline signals from the locus coeruleus higher brain functions, such as mood, atten- facilitate changes in the forebrain networks tion, motivation, cognition, learning and mem- that are mediating specific cognitive functions. ory [198, 264]. In addition to thalamic and extrathalamic The dopaminergic projection is a key modu- subcortical afferents, each particular neocorti- lator of motivational cognitive and motor func- cal area also receives a strong input from other tions [244, 652] and the serotoninergic projec- neocortical areas. These association fibres tion helps regulate wake-sleep cycles and mod- come from either the same or the opposite ulates the sensory gating of behaviourally rele- hemisphere; in the latter case they are called vant cues in the environment [465]. callosal fibres. According to Lorente de Nó The cholinergic projection to the neocortex [414, 416], the association fibres give off collat- is implicated in arousal, sleep-wake cycles, vis- erals in the deep laminae, especially VI, but ual information processing learning, memory their main territory of distribution is in lami- and selective attentional functions [239, 650, nae I–IV, and especially II and III (d in Fig. 835]. 15.25 A). Goldman and Nauta [242] made small The noradrenergic projection to the neocor- focal injections with tritiated amino acids into tex arises, as mentioned, from the locus coeru- various areas of the association cortex of mon- leus in the rostrodorsal pons. Traditionally, this keys. They found that the anterogradely trans- nucleus has been thought to play a role in vigi- ported label accumulated in narrow, 200- to lance and in responsiveness to novel stimuli 300-lm-wide columns in relatively distant re- [21, 199]. gions. On the basis of a study of Golgi material, Recently, Aston-Jones and Cohen [19, 20] Szentágothai [721, 723] concluded that the ar- and Bouret and Sara [59] have developed new borization space and pattern of individual cor- concepts on the function of the locus coeru- tico-cortical axons correspond in size with the leus. Aston-Jones and Cohen point out that columns revealed by isotopic labelling. He ob- during the waking state, the locus coeruleus served that these fibres pass radially through neurons exhibit two modes of activity, phasic the cortex and issue relatively short branches and tonic. Phasic activity is driven by the out- at all levels (Fig. 15.25 B). Cortico-cortical fi- come of task-related decision processes and is bres defining a narrow radial zone of termina- 15 Telencephalon: Neocortex 543 tion have also been observed in the primate cial pyramids situated in lamina II extend not sensorimotor cortex [286, 339]. Such a fibre, only their apical dendrites in lamina I; these labelled by HRP injected in the corpus callo- elements often have basal dendrites which pass sum is depicted in Fig. 15.25 C. The major laterally from the soma and ascend to layer I layers of termination of callosal fibres in the [779]. These elements are most probably primate somatosensory areas are laminae I–IV. strongly excited by the lamina I extrinsic affer- In the motor cortex, they terminate in a com- ents. It is interesting to note that the afferents parable pattern in laminae I–III. Studies on from different thalamic nuclei, which, after cortico-cortical projections based on the trac- having traversed the cortex, spread in lamina ing of individual fibres, like the one cited I, terminate in different subzones of that layer above, are scant. However, an overall picture of [290] and the apical dendritic tufts of the pyr- the mode of termination of these fibres can be amids thus receive stratified input from differ- deduced from axon degeneration studies of ent sources. cortico-cortical projections and from tracer Until recently, it was widely held that the studies in which such fibres are labelled en specific sensory thalamic afferents mainly dis- masse. A detailed survey of this literature, pre- tributed to lamina IV (or its sublayers) of the sented by Innocenti [322], revealed that most primary sensory cortices contact only a single cortico-cortical fibres terminate in layers III category of neurons, i.e. the spiny stellate cells, and IV in primates. and that the input is then processed sequen- Rockland and Pandya [623] reported that, in tially by hierarchically organized chains of the visual cortex of the rhesus monkey, the py- neurons (see e.g. Hubel and Wiesel [308, 310] ramidal neurons of the superficial layers pro- and Eccles [166]). Golgi-electron microscopy ject to the middle layer (principally layer IV) studies combined with degeneration and tracer of their cortical target areas, whereas the deep experiments [814, 815] conclusively showed layer pyramids project outside the middle that the specific sensory thalamic afferents to layers, mainly to layers I and VI. It has already the cortex synapse with pyramidal cells and been mentioned that Felleman and Van Essen with non-pyramidal intrinsic cortical elements. [180] classified the projections to layer IV as According to White [814, 815], this finding feed-forward pathways and those projecting challenges the concept that thalamic input is outside layer IV as feed-back projections. processed by hierarchically organized chains of From the foregoing, it appears that most of neurons and lends support to the notion that the extrinsic cortical afferent systems distribute the function of the cerebral cortex depends themselves in layered arrays, and it may be heavily on parallel processing mechanisms. added that, in the primary sensory cortices, an However, it should be noted that in highly dif- elaborate laminar segregation of thalamic in- ferentiated sensory cortices different groups of puts appears to be reflected in the stratification local circuit neurons may well be involved in visible in the cellular architecture. Morphologi- the parallel processing of sensory information. cal evidence suggests that all extrinsic neocor- We have seen that, in the primary visual cortex tical afferents are excitatory, raising the ques- of primates, the magnocellular zones of the lat- tion of how the intrinsic machinery of this eral geniculate nucleus project to sublamina cortex is driven by these afferents. IVCa, whereas the parvocellular zones project Lamina I contains the apical dendritic bou- to IVCb (Fig. 15.26). These two sublaminae are quets of pyramidal cells of laminae II, III and characterized by the presence of different V. Because in this layer only sparsely scattered classes of spiny stellate cells and by an almost intrinsic neurons are present, it is reasonable complete lack of pyramidal neurons [422–424, to assume that the extrinsic thalamic afferents 446]. Moreover, it has been established that the spreading in it synapse principally with pyra- principal projection of the IVCa spiny stellate midal neurons and thus have a direct access to neurons is to sublamina IVB, whereas the ma- the long-axonal outflow of the cortex. Superfi- jority of IVCb spiny stellate neurons project to 544 Section II Structure of Spinal Cord and Brain Parts sublaminae IIIC and IVA [190, 422]. These Typical Pyramidal Cells findings suggest a continued separation of magnocellularly and parvocellularly derived in- The typical pyramidal cells constitute the larg- formation within the primate primary visual est and most characteristic category of neocor- cortex [425]. tical neurons. In the visual cortex of rat, cat We have seen that the deeper layers of the and monkey, according to Winfield et al. [824], cortex (V and VI) also receive direct thalamic these elements account for about 65% of the inputs. It is known that these layers contain total neuronal population, and in other cortical many pyramidal and pyramid-like cells and areas this percentage may be even higher. A only a relatively small number of intrinsic ele- typical pyramidal neuron is provided with a ments. The axons of many of these deep pyra- radially oriented apical dendrite that forms a midal cells possess recurrent, ascending collat- terminal tuft in laminae I and II, with basal erals and the axons of a certain proportion of dendrites that radiate out from the base of the the local circuit neurons in laminae V and VI soma and with an axon descending downward are also directed to more superficial cortical to leave the cortex (Fig. 15.29). The apical den- layers. There is experimental evidence indicat- drites may give rise to one or more horizontal ing that, in the deeper layers of the primary or obliquely ascending side branches. All of somatosensory cortex of the mouse, thalamo- the dendrites of typical pyramidal cells are cortical afferents contact both pyramidal and more or less densely covered with spines (Fig. non-pyramidal cells [294, 295, 813]. The deep 15.30 A) [561]. Positionally, the apical dendritic pyramidal cells may well receive an “aspecific” tuft with its ramifications in laminae I and II thalamic input via collaterals spreading in lam- is the only feature shared by all typical pyra- ina VI (Fig. 15.25 A (c) and a “specific” thala- midal cells. The somata and their basal dend- mic input via their apical dendrites in lamina ritic systems may vary in position from lamina IV. II to lamina VI, and the length of the apical Finally, the callosal and ipsilateral associa- dendrites varies accordingly (Fig. 15.29). In the tion fibres form nearly all their synapses with following discussion of the typical pyramidal the spiny dendrites of pyramidal cells [322, cells, the afferents impinging upon the various 815]. Because these fibres terminate principally parts of their receptive surface will be consid- in laminae III and IV, it seems likely that the ered first; the relation between the position of superficial pyramidal neurons form their main their somata and the destination of their axons target. will then be dealt with. Finally, attention will be paid to the patterns of distribution of their collaterals. The somata of the pyramidal cells are not Neocortical Neurons under the direct influence of any extrinsic af- and Their Synaptic Relationships ferent system. Rather, they are specifically addressed by one type of local circuit neuron, i.e. the basket cell (Fig. 15.31 A), although several Introductory Note other types of non-pyramidal cells may contribute to some extent to the somatic innerva- The neurons of the neocortex can be classified tion of pyramidal cells [137]. The somata of as belonging to two basic types: pyramidal the basket cells, which, like those of the pyra- cells and non-pyramidal cells. The former can midal cells, vary in distance from the pia and be subdivided into typical and atypical pyra- are concentrated in laminae III and V [177, midal cells (Figs. 15.28–15.30). The elements 436]. Their poorly ramifying dendrites, which belonging in these categories will now be dis- bear few or no spines, radiate in all directions, cussed. with a tendency toward vertical and horizontal trajectories. The axons are either ascending or 15 Telencephalon: Neocortex 545

Fig. 15.28. Pyramidal neurons, as observed in Nissl preparations of the human neocortex (reproduced from [786]). 1, Largest pyramidal cell from the IIIrd layer of area 17; 2, stellate cell from the VIth layer of area 18; 3, large pyramidal cell from the IIIrd layer of area 18; 4, Meynert cell from the Vth layer of area 17; 5, largest pyramidal cell from the IIIrd layer of area 4; 6, Betz cell from the Vth layer of area 4. Note that the Meynert cell and the Betz cell, both of which represent by far the largest cell type in their respective cytoarchitectonic areas, and which, hence, are both designated as “giant cells” [84], differ considerably in absolute size. a, axon 546 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.29. Typical pyramidal neurons in the mammalian neocortex

Fig. 15.30 A–F. Typical and atypical pyramidal neurons in the primate neocortex. A Typical pyramidal cell in the human neocortex [84]; B small pyramidal neuron with poorly developed apical dendrite (arrowheads) and a recurving axon from the striate cortex of the macaque monkey [422]; C a “star pyramid” in layer IV of the somatosensory cortex of the squirrel monkey [336]; D–F, spiny stellate cells in layers IVB, IVA and IVCb, respectively, of the primary visual cortex of the macaque monkey [422, 426] 15 Telencephalon: Neocortex 547

Fig. 15.31 A,B. Intrinsic cortical neurons contacting specific parts of the receptive surface of the pyramidal cells. A Basket cells, whose axonal terminals participate in the formation of pericellular nests surrounding the somata and proximal dendrites of pyramidal cells (partly based on Jones [339]). B Chandelier cells, whose numerous vertical axonal terminal portions specifically contact the axon initial segments of pyramidal cells. Based on Peters [555], Fairén et al. [177], Somogyi et al. [701] and Jones [339] 548 Section II Structure of Spinal Cord and Brain Parts descending and provide four or more horizon- than in the infragranular layers [570, 690]. Be- tal collaterals at various levels. The horizontal cause the pyramidal neurons projecting via the collaterals, which can extend for 1 mm or more corpus callosum to the contralateral hemi- in either direction, issue at intervals short side sphere are situated mainly in laminae II and branches. These side branches terminate in III, it has been suggested that the chandelier pericellular baskets surrounding the somata cells principally influence cortico-cortical cir- and proximal dendrites of pyramidal cells. The cuitry [570, 698]. basket cell terminals contain flattened vesicles, In the hippocampal cortex, the perikarya of and the synapses made with the pyramidal the pyramidal neurons are concentrated in a cells are of the symmetrical variety. They use single layer, and all of these elements extend GABA as a neurotransmitter and exert a their basal dendrites in one and the same powerful inhibitory action on their target ele- plexiform zone, the stratum oriens (Fig. 12.9). ments [343]. There is evidence suggesting that However, in the neocortex, the pyramidal so- the basket cells are major recipients of thalam- mata are situated at different levels and, given ic and commissural inputs [555, 690, 700]. the laminar organization of many extrinsic and Moreover, the somata of the basket cells are intrinsic cortical fibre systems, it is to be ex- contacted by numerous GABAergic terminals, pected that the basal dendritic systems of dif- which may well arise from the axons of other ferent pyramidal neurons are involved in dif- basket cells [287]. ferent synaptic relationships, depending on the Like the somata, the axon hillocks and ini- laminae or sublaminae within which these den- tial axonal segments of the pyramidal cells are dritic systems occur. Thus, it has been estab- also the target of a special category of local cir- lished that the basal dendrites of pyramidal cuit neurons, i.e. the axo-axonic or chandelier cells situated in laminae II and III receive cal- cells (Fig. 15.31 B) [176, 555, 699]. These ele- losal afferents [100] and that lamina III border ments occur in laminae II–V, but are most pyramids extend their basal dendrites into prominent in lamina III. Their dendrites may lamina IV, where they are contacted by term- be grouped in an upper and a lower tuft or inals of thalamocortical fibres [303, 643]. In- spread in all directions. The axon ramifies trinsic cortical axons, i.e. axons of local circuit many times and produces a dense plexus in neurons and collateral branches of pyramidal the vicinity of the parent soma. From this axo- cells, are concentrated in lamina IV and deep nal plexus, numerous (up to 300) vertically in lamina V, where they form the external and oriented arrays of terminals arise that contact internal striae of Baillarger [340, 756]. The ba- the initial axonal segments of pyramidal cells. sal dendrites of certain groups of pyramidal The axo-axonic synapses between the chande- neurons exhibit a specific affinity to the termi- lier cells and the pyramidal neurons are sym- nal ramifications of the axons within these metrical, and the vesicles in the presynaptic striae. In the primate primary visual cortex, axon terminals are flattened [178, 345, 690]. the external stria of Baillarger (also known It has been shown that the axon terminals of there as the line of Gennari) is strongly devel- the chandelier cells contain the inhibitory neu- oped and occupies a discrete sublayer IVB, in rotransmitter GABA; hence, it may be assumed which no thalamic afferents are present (Fig. that these elements inhibit the pyramidal cells 15.26). According to Valverde [755], its axonal [570]. Because the chandelier cells exert this components include horizontal collaterals of action at a strategic site, i.e. the trigger zone descending axons of pyramidal and stellate for the initiation of action potentials, they may cells of sublayer IVA, horizontal branches of be considered to have a decisive influence on axons of neurons located within sublayer IVB the output from the pyramidal neurons. It has and ascending ramifications of spiny stellate been found that the number of axo-axonal sy- cells residing in sublayer IVC. Lund [422] re- napses along the initial segments of pyramidal ported that the basal dendrites of pyramidal cells is much larger in the supragranular layers neurons situated in sublayer IVA turn sharply 15 Telencephalon: Neocortex 549 downward from the soma to enter sublayer belled by intracellular injections with HRP IVB, where they fan out horizontally, markedly have conclusively shown that the axon col- avoiding any ramifications within lamina IVA laterals of these elements make contact pre- (Fig. 15.32E). To the author’s knowledge, the dominantly with the dendrites of other py- question of which neurons contribute axons or ramidal cells [218, 370, 457]. axonal branches to the internal stria of Baillar- 3. The vertically elongate axonal systems of some ger has never been specifically addressed; how- types of cortical local circuit neurons. These ever, the Golgi studies by Cajal [84] and Val- elements, which will be discussed below, in- verde [755, 756] suggest that the axons of both clude spiny stellate cells (Fig. 15.30 D–F) and superficial and deep pyramids issue numerous various types of smooth or sparsely spiny horizontally oriented collaterals into this stria non-pyramidal neurons (Fig. 15.36 (M–Q). and that certain local circuit neurons with ascending axons, situated in lamina VI, also con- The recurrent ascending axons or axonal tribute collaterals to it. In the primate motor branches of pyramidal neurons, as well as the and visual cortices, basal dendrites of pyrami- ascending and descending axons of spiny stel- dal cells form a conspicuous plexus within the late cells, have been observed to assemble in domain of the internal stria. In both of these highly characteristic, radially oriented bundles cortices, large pyramidal neurons occur, [177, 423, 755, 756]. known in the motor cortex as Betz cells and in Turning now to the apical dendritic shafts the visual cortex as Meynert cells (Fig. 15.28 of the typical pyramidal cells, it is important (4, 6). Beyond their extraordinary size, these to note that these conspicuous processes are neurons have several other features in com- well placed to receive input from a variety of mon. Both cells are characterized by the pres- axonal pathways known to terminate within ence of long horizontally oriented basal den- specific cortical layers. There is considerable drites that extend into the internal stria, and variation in apical shaft length, ranging from both further contribute to this stria with addi- essentially no shaft at all (pyramidal cells in tional horizontal dendrites emanating from the the superficial zone of lamina II) to apical den- lateral surface of the soma and even from the drites of 2 mm or more (pyramidal cells in proximal stem of the apical dendrite [63, 64, lamina VI). Moreover, particularly in primates, 84, 654, 657]. The long, coarse collaterals of the apical dendrites of many deep pyramidal the Meynert cells may also contribute to the cells do not extend into the subpial cortical internal stria (Fig. 15.35) [622]. zone, but rather terminate in a variably devel- Whereas the horizontal axonal systems dis- oped terminal tuft deeper within the cortex. cussed so far contact the basal dendrites of py- Given these variations in length and in posi- ramidal neurons situated at one particular lev- tion of their apical dendrites, it is evident that el, the neocortex also contains vast numbers of different pyramidal cells may receive different vertically oriented axonal elements, which po- samples of lamina-specific extracortical and in- tentially contact the basal dendritic systems of tracortical afferents, and these differences are pyramidal elements situated at different levels. still further accentuated by the fact that the These vertical axonal elements, which play a apical dendrites of different pyramidal neurons prominent role in the so-called radial coupling may exhibit different specific affinities to par- of neuronal elements and hence in the colum- ticular afferent systems. The evidence for such nar functional organization of the cortex, may specific affinities includes (1) the presence of be categorized as follows: distinct, lamina-specific differences in the den- 1. Thalamocortical and cortico-cortical associa- sity of spines along the apical dendrites, (2) tion fibres (Fig. 15.25). the presence of lamina-specific side branches 2. Axons and recurrent collaterals of pyramidal on the apical dendrites and (3) direct proof neurons (Figs. 15.30 B,C, 15.32 D–F). Ultra- that apical dendritic segments of different py- structural analyses of pyramidal neurons la- ramidal cells passing through a particular layer 550 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.32 A–G. Features of pyramidal neurons. A-F Semidiagrammatic representations of pyramidal neurons in the primary visual cortex of the rhesus monkey. The apical dendrites of the elements A–C issue their side branches and terminal tufts in particular (sub)layers; the density of spines along these processes show distinct lamina-specific differences. The elements D–F, which are situated at different levels, contribute axon collaterals to particular levels. G shows two pyramidal neurons, whose apical dendrites receive multiple asymmetrical synapses (open circles) from the axons of a bipolar cell. G also shows that smooth or sparsely spiny, non-pyramidal cells (ss) form symmetrical synapses (filled circles) with pyramidal (py) and bipolar cells (b) and that thalamic afferents (th) contact pyramidal, bipolar and smooth or sparsely spiny, non-pyramidal cells forming asymmetrical synapses (A–C are redrawn from Lund [423]; D–F are redrawn from Lund and Boothe [426]; G is based on Peters [554]) 15 Telencephalon: Neocortex 551 may receive highly different numbers of sy- randomly distributed along the apical shafts. napses from the afferents concentrated in that However, systematic studies of large numbers layer. These three aspects will now be briefly of apical dendrites have shown that in several discussed. cortical areas these processes are given off at se- 1. The presence of distinct, lamina-specific dif- lective levels, i.e. as they pass through particular ferences in the density of spines along the apical layers. Thus, Lorente de Nó [415] observed that, dendrites. Spines are present in abundance on in the entorhinal cortex, the ascending shafts of the dendrites of pyramidal neurons, where they the pyramids of the deeper laminae always issue function as the primary postsynaptic structures some side branches in lamina III, but give off of the cell [178]. Quantitative analyses have none at all during their passage through lamina shown that these spines are not evenly distrib- II. Lund and Boothe [426] observed that the api- uted along the dendrites. Characteristically, the cal dendrites of a particular population of lam- most proximal portions of the dendrites ema- ina VI pyramidal neurons in the primate visual nating from the pyramidal somata are devoid cortex issue side branches selectively in lamina of spines. From these initial segments onward, IVCa, whereas the apical dendrites of another the concentration of spines increases gradually, lamina VI pyramidal population in this cortex attains a maximum some 80 lm away from the give off side branches at the border of lamina soma and then declines again distally [240, V and IVCb and split up into their terminal den- 561, 753]. Superimposed over this general pat- dritic tufts in lamina IVA (Fig. 15.32 B,C). Ac- tern of spine distribution, the apical dendrites cording to Lund [425], these various dendritic of pyramidal neurons may display distinct, lam- patterns suggest that deeper pyramidal neurons ina-specific changes in spine density. For exam- may receive thalamic input within laminae IVCa ple, Lund [422] found that in the primary visual and IVA. cortex of the monkey, the apical dendrites of py- 3. Direct proof that apical dendritic segments ramidal cells lying in laminae V and VI show of different pyramidal cells passing through a par- markedly fewer spines in lamina IVCb than in ticular layer may receive highly different numbers the densely spiny portion of the same shafts in of synapses from the afferents concentrated in laminae V and VI. The number of spines on that layer. The thalamocortical connectivity of these apical dendrites in general increases again pyramidal neurons has been the subject of a se- in laminae above IVCa, but not to the level ries of quantitative studies by White and collab- shown on the proximal portion of the shaft orators [294–297, 817, 818, 820]. Hersch and (Fig. 15.32A,B). According to Lund [422], this White [294, 295] reported that, in layer IV of reduction in spines on the apical dendritic shaft the primary somatosensory cortex of the mouse, as it passes through laminae IVCb is true of the the proportion of thalamocortical synapses re- great majority of the pyramidal cells of laminae ceived by apical dendrites belonging to various V and VI, including the giant pyramidal cells of sizes of Golgi-impregnated lamina V and lamina Meynert. It is known that lamina IVCb receives VI pyramidal neurons ranged from 1.3% to afferents specifically from the parvocellular 14.6% of all the asymmetrical synapses formed layers of the lateral geniculate nucleus. Thus, onto these dendrites. In subsequent work in the findings of Lund [422] might suggest that, the same cortex, the number of thalamocortical in the monkey, the deep pyramids in the visual synapses in lamina IVon the apical dendrites be- cortex are not the primary targets of these par- longing to deep pyramids that project from the ticular afferents. primary somatosensory cortex to either the ven- 2. The presence of lamina-specific side trobasal nucleus of the thalamus [296, 818], the branches on the apical dendrites. The apical den- primary motor cortex [820] or the ipsilateral drites of many pyramidal neurons issue small striatum [297] was determined. It appeared that numbers (some three to eight) of oblique or hor- these three populations of pyramidal neurons izontally oriented side branches (Fig. 15.30 A). have very different thalamocortical connectivity At first sight, these side branches seem to be patterns. Thus only 0.3–0.9% of the total number 552 Section II Structure of Spinal Cord and Brain Parts of apical dendritic synapses in lamina IV of cor- 4. Ascending axons from multipolar or bitufted ticostriatal projection neurons were made with neurons situated in deeper cortical layers. thalamic afferents. For the cortico-cortical and These elements, which have smooth or corticothalamic elements examined, these values sparsely spinous dendrites, are known as were 1–7% and 7–20%, respectively. Martinotti cells (Fig. 15.36 Q). Large elements So far, I have dealt with tangentially orient- of this type are found in laminae V and VI, ed afferent systems establishing synaptic con- but they are also present in the more superfi- tacts with particular segments of the apical cial cortical layers [177]. Marin-Padilla [438] dendrites of pyramidal neurons. However, reported that the morphology of the axonal there is also evidence for the presence of ra- termination of Martinotti neurons resembles dially oriented afferents that make repeated quite closely the arborization of the apical contacts with these processes. Thus, Scheibel dendritic tufts of pyramidal cells and consid- and Scheibel [656] mentioned that “non-specif- ered it likely that Martinotti cells form dual ic” cortical afferent fibres originating from the sets with pyramidal neurons of similar corti- brain stem and the medial thalamus break up cal depth. With regard to the function of the into a series of branches that ascend radially Martinotti cells, Marin-Padilla [438] sug- through the cortex, establishing sequences of gested that these elements are inhibitory and axodendritic contacts with the spines of apical that the inhibition takes place specifically be- shafts and terminal branches of pyramidal tween the axonal terminals of a given Marti- neurons. It is also known that the axons of a notti cell and the dendritic tufts of the pyra- certain category of cortical local circuit neu- midal neurons with which it forms a dual set. rons, known as bipolar cells, typically give rise 5. Axons of so-called horizontal neurons, which to vertically oriented branches which parallel are situated within lamina I itself [84]. The the trajectories of clustered pyramidal apical dendrites and axonal branches of these ele- dendrites (Fig. 15.32 G). These branches form ments are entirely confined to lamina I and multiple asymmetrical synapses with the extend parallel to the surface of the cortex spines of the apical dendrites [179, 554]. Apart (Fig. 15.36 D). Some of their axonal branches from “non-specific” cortical afferents and the may attain a considerable length. There is im- axons of bipolar cells, collateral branches of munohistochemical evidence (summarized pyramidal cell axons may also make repeated, by Vogt [779]) indicating that the horizontal climbing fibre-like contacts with the apical neurons are all GABAergic and also contain dendrites of (other) pyramidal cells [218]. the neuropeptide cholecystokinin. The final part of the dendritic system of typical pyramidal neurons to be considered is In summary, it may be stated that the apical the set of dendritic branches emanating from dendritic branches of neocortical pyramidal the tip of the apical dendrite. These so-called neurons receive input from various sources. apical dendritic tufts or terminal dendritic Thalamic afferents and recurrent collaterals of bouquets extend into lamina I, in which, to- pyramidal cells presumably exert an excitatory gether with numerous tangentially running ax- influence on these dendrites, whereas the axo- ons, they constitute a typical plexiform zone. nal endings of the Martinotti cells and the The axonal components observed within this horizontal cells may well exert an inhibitory zone include: influence on them. The horizontal cells have 1. Fibres originating from the intralaminar and been suggested to receive a specific input from midline thalamic nuclei [49, 640]. monoaminergic fibres originating from neu- 2. Monoaminergic fibres arising from the brain rons situated in the brain stem [438], but tha- stem [438]. lamic afferents may also impinge on these in- 3. Recurrent collaterals of the axons of pyrami- trinsic lamina I elements and the monoaminer- dal neurons, principally situated in laminae gic fibres may also directly contact the pyrami- II and III [414, 656, 755]. dal apical dendrites. However, it should be em- 15 Telencephalon: Neocortex 553 phasized that so far none of the synaptic con- jecting to the remaining subcortical sites tend tacts suggested by light microscopy material to occupy an intermediate position. has been verified with the aid of experimental The corticothalamic projections to the “spe- ultrastructural techniques. cific” thalamic relay nuclei arise exclusively Having discussed the afferents making con- from large pyramids in lamina VI. tact with the various parts of the receptive sur- Although most cortical neuronal populations face of the typical pyramidal neurons, I will projecting to a particular cortical or subcorti- now turn to the axons of these elements. It has cal target show a distinct laminar specificity, it already been mentioned that these processes is not uncommon to find some degree of overall leave the cortex and pass either to other ip- lap in the boundaries demarcating different silateral or contralateral cortical regions or to populations of projection neurons. This raises one or several subcortical centres. The latter the question of the extent to which projections may include the striatum, i.e. the nucleus cau- to particular targets of cortical efferent neu- datus and the putamen, the various “specific” rons are made up by collaterals of axons pro- and “non-specific” thalamic nuclei, the nucleus jecting to other centres. In this context, it may ruber, the superior colliculus or tectum mesen- be recalled that Cajal [84] noted in his Golgi cephali, the pontine nuclei, the medulla oblon- studies of the brains of rodents that numerous gata and the spinal cord. subcortical centres are supplied by collaterals Retrograde tracing studies have shown that of corticofugal fibres. Thus, he observed that, the cell bodies of pyramidal neurons projecting during their descent through the internal cap- to particular cortical or subcortical targets are sule, such corticofugal fibres issued numerous preferentially located in particular cortical collaterals to the striatum or thalamus and that layers or sublayers (Fig. 15.33). The following pyramidal tract neurons in the brain stem gave summary of the laminar relationships of corti- off collateral branches to several centres, in- cal efferent cells is based on reviews by Jones cluding the red nucleus, the pontine nuclei, [337] and White [815], to which the reader is and the dorsal column nuclei. Double-labelling referred for the primary sources. experiments, i.e. experiments in which two dif- Cortico-cortical and callosally projecting fi- ferent and distinctive retrogradely transported bres arise predominantly from pyramidal neu- labels are injected into two different known rons in lamina II and III; however, in rodents terminal fields, have revealed that double-pro- and primates, significant numbers of these fi- jecting neurons do occur in the neocortex. bres have been found to originate from ele- Catsman-Berrevoets and Kuypers [93] reported ments situated in the infragranular layers. As the presence of double-labelled cells in the mo- regards the superficial layers, it has been estab- tor cortex of the monkey after injecting mark- lished that the smaller, more superficially situ- ers into the magnocellular part of the red nu- ated pyramids tend to project to ipsilateral cor- cleus and spinal cord, and Rustioni and Hayes tical areas situated nearby, whereas the larger, [644] found double-labelled cells in the sensory more deeply placed cells tend to project to cortex of the cat after injecting markers into contralateral and to more remote ipsilateral the dorsal column nuclei and spinal cord. cortical areas. However, in these and other comparable ex- Pyramidal neurons situated in lamina V periments, the number of double-labelled cells have been shown to project subcortically to the appeared to be very small, implying that the intralaminar and other “aspecific” thalamic nu- degree of subcortical collateralization of corti- clei, the striatum, the red nucleus, the tectum, cofugal fibres is limited, too [337]. The abun- the medulla oblongata and the spinal cord. The dance of such collaterals observed by Cajal smallest and most superficially situated ele- [84] might well have to do with the fact that ments in this layer project to the striatum, the Golgi material studied by that author was while the largest and most deeply situated cells exclusively derived from the brains of very project to the spinal cord. The elements pro- young animals. 554 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.33. Laminar location of the perikarya of pyramidal cells projecting to other parts of the cerebral cortex and to subcortical centres. aspec thal, aspecific thalamic nuclei; tect, tectum; nrub, nucleus ruber; med obl, medulla oblongata; spec thal, specific thalamic nuclei; spin cord, spinal cord. Based on White [815] 15 Telencephalon: Neocortex 555

The axons of all typical pyramidal neurons V, up to 2.64 mm); Ojima et al. ([515]: cat release a number of collaterals before entering auditory cortex, laminae II, III, 0.7–2.5 mm); the subgriseal white matter (Fig. 15.30 A). McGuire et al. ([457]: macaque visual cortex, These collaterals may ramify within close prox- lamina III, 2 mm); Kisvárday and Eysel ([368]: imity in the parent cell body or may descend, visual cortex, lamina III, up to 2.8 mm); Mel- ascend or travel for shorter or longer horizon- chitzky et al. ([459]: macaque prefrontal cortex, tal distances within the cortex. Early studies of lamina III, 6 mm); and Rockland and Knutson the intracortical distribution of pyramidal cell ([622]: macaque primary visual cortex, lamina axon collaterals were based exclusively on the VI, 8 mm). study of Golgi material, but the real extent of As regards the extent of the long-range col- these processes has only recently been revealed laterals, it has been observed that processes of by experimental studies in which single pyra- this type do not remain within the cytoarchi- midal cells were intracellularly injected with tectonic area in which their parent soma lies, HRP, or in which small extracellular injections but may project to adjacent cortical areas [138, of tracer substances, such as biocytin or bioti- 218]. nylated dextran amine (BDA), are placed in the The number of major collaterals with long trajectory of fibres or collaterals. Given the nu- horizontal trajectories issued by the main ax- merical preponderance of pyramidal neurons, ons of pyramidal neurons varies; most of the there can be no doubt that the intracortical cells examined by DeFelipe et al. [138] had one collateral branches of these neurons together to three long collaterals, whereas the elements constitute the largest single category of axons studied by Ojima et al. [515] issued two to five in the neocortex. The endings of these collat- such processes. eral branches, like those of the main axons, all The long-range primary collaterals are make synapses of the asymmetrical/round ves- usually coarse and well myelinated. 1 They give icle variety and use the excitatory amino acids off thin, unmyelinated, bouton-laden secondary glutamate and aspartate as neurotransmitters. branches that are predominantly oriented per- The local axon collaterals of pyramidal neu- pendicular to the cortical surface. Remarkably, rons may show quite characteristic distribution these secondary branches are emitted in clus- patterns. Thus, Lund and Boothe [426] found ters at more or less regular intervals (Fig. that the axons of pyramidal cells situated in 15.35). Because of the overall vertical orienta- different layers of the primary visual cortex of tion of the individual secondary branches, the macaque monkey selectively issue short, these clusters often have a column-like appear- horizontal, collateral branches in different ance. Secondary branches arising from differ- layers (Fig. 15.32 D,E). ent major collaterals of the same parent axon Apart from local collaterals, the axons of py- in different layers often converge upon the ramidal neurons may also give rise to long, same cluster [138, 370, 515]; the same holds horizontally disposed branches (Figs. 15.34, true for the focussed terminal branches of ma- 15.35). Pyramidal neurons emitting such long- jor collaterals of different pyramidal cells range collaterals have been the subject of nu- [138]. Rockland and Knutson [622] observed merous studies, of which the following may be that the axons of Meynert cells, i.e. large pyra- mentioned (for the sake of brevity, the animals midal elements in the infragranular layers of studied, the cortical areas and laminae in the primary visual cortex, issue a number of which the parent somata were located and an coarse, remarkably long (up to 8.0 mm) collat- indication of the length of the collaterals ob- eral branches, which pass horizontally through served are included in the references): DeFelipe the VIth layer of that cortical area. Each of et al. ([138]: monkey sensory motor cortex, 1 lamina III, up to 6 mm); Kisvárday et al. Myelinated fibres fail to impregnate with the Golgi l stain; this may explain why these long-range collat- ([370]: cat visual cortex, lamina III, 1500 m); erals remained unnoticed in the classical Golgi stud- Gabbott et al. ([218]: cat visual cortex, lamina ies by Cajal [84] and Lorente de Nó [414, 416]. 556 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.34. A pyramidal neuron in layer III of the primary visual cortex of the cat. The element was intracellularly filled with horseradish peroxidase and reconstructed from 80-lm-thick serial sections. The intracortical system of collaterals forms distinct clusters, one near the cell’s dendritic field in layer III (A), another just below the cell in layer V (B), and two at a distance of some 1000 lm from the soma in layers III and V (C,D). Modified from Kisvárday et al. [370] 15 Telencephalon: Neocortex 557

Fig. 15.35. Diagrammatic representation of a Meynert cell (M) and a lamina III pyramidal cell (P) in the primary visual cortex of the macaque monkey. The Meynert collaterals have (1) relatively small terminal clusters concentrated in lamina VI; (2) large terminal specializations; (3) a small number of terminations per cluster, and (4) a termination-free zone in the vicinity of the soma. On all of these points, these collaterals differ from those of the smaller supragranular pyramidal cells. Modified from Rockland and Knutson [622]. ax, axon 558 Section II Structure of Spinal Cord and Brain Parts these collaterals forms a number of small ter- put relationships. Most pyramidal cells ex- minal clusters, containing a small number of amined appeared to synapse predominantly large terminal boutons (up to 3.0 lm), and with other pyramidal elements. However, each of these collaterals has an initial termina- the collaterals of groups of corticothalamic tion-free zone. All of these features appeared projection neurons in the primary sensory to be markedly different from those of hori- cortex of the mouse were observed to form zontal intrinsic connections in the supragranu- over 90% of their synapses with the smooth lar layers of the primary visual cortex (Fig. dendrites of non-pyramidal neurons [819]. 15.35). It is noteworthy that the basal dendrites Pyramidal neurons whose axon collaterals of Meynert cells show a remarkable asymme- contact the dendrites of pyramidal and non- trical spread (Fig. 15.35) [84]. It has been pro- pyramidal cells in about equal numbers have posed [515] that their asymmetrical dendrites also been found [456, 825]. make Meynert cells sensitive to visual motion, 4. The number of synapses made by the collat- and in particular to motion in one direction. erals of a given pyramidal neuron with one Comparable asymmetry does not characterize other individual pyramidal neuron is pre- the collateral axonal fields of supragranular py- sumably generally very limited [218, 370, ramidal neurons [622]. Li et al. [409] recently 456, 457, 720, 721]. However, the collaterals investigated how Meynert cell collaterals are of one pyramidal cell contact numerous mapped in relation to the functional architec- other pyramidal cells and, conversely, one ture of the primary visual area in macaque pyramidal cell receives the converging input monkeys. They found that such collaterals may of numerous other pyramidal cells. cross over several pairs of ocular dominance 5. Neurons contacted by the collaterals of pyra- columns and contact both left- and right-eye midal neurons have been identified as non- ocular dominance columns. According to Li et spiny bipolar cells [456], non-spiny multipo- al. [409], these findings suggest hat the system lar cells [819] and basket cells [457], and it of Meynert intrinsic collaterals is involved with has also been established that some of the binocular interactions over wide sectors of the dendrites postsynaptic to pyramidal collat- visual field. The main axons of the Meynert erals are immunoreactive to GABA [370]. It cells project to the middle temporal visual area is known that most types of non-spiny or (MT) and to subcortical targets including the sparsely spiny non-pyramidal cells, includ- superior colliculus [778]. ing basket cells, chandelier cells and double In order to gain insight into the way in bouquet cells, use GABA as a neurotransmit- which pyramidal neurons participate in the inter and that these elements are the principal trinsic circuitry of the neocortex, the synaptic source of the GABAergic, symmetrical sy- connections of axon collaterals belonging to napses that impinge upon the somata, proxi- neurons of this type have been examined with mal dendrites and axon initial segments of the electron microscope in several areas and in pyramidal neurons [306]. All in all, it seems different species [128, 172, 218, 365, 370, 456, reasonable to assume that the GABAergic in- 457, 459, 816, 817, 819, 825]. The results of terneurons in the neocortex receive input di- these studies may be summarized as follows: rectly from the axon collaterals of pyramidal 1. Collaterals of pyramidal cells only form sy- neurons and in turn synapse with pyramidal napses of the asymmetrical/round vesicle neurons. These circuits probably provide the variety and, hence, may be considered to ex- morphological substrate for both feed-for- cite their targets. ward and feed-back inhibition of pyramidal 2. Practically all of the synaptic contacts made neurons [306, 815]. by these collaterals are on dendritic spines or dendritic shafts. If we survey the data concerning the typical 3. Different types of pyramidal neurons may pyramidal neurons discussed above, it appears show striking differences in their local out- that these elements show a quite remarkable 15 Telencephalon: Neocortex 559 structural diversity. This diversity may concern where they may be abundant [336, 422, 823]. their size, their laminar position, the branch- Their small, spherical or ellipsoid somata in ing pattern of their dendrites, the density of Nissl preparations often form one or several spines along their apical dendrites, their affini- conspicuous granular zones, which is why lam- ty to particular afferent systems, the cortical inae I–III and V–VI in the sensory cortices are or subcortical target regions to which their often designated as the supragranular and in- main axons project, the distribution of their fragranular layers. The dendrites of these spiny axonal collaterals and their patterns of intra- stellate cells are emitted at several points from cortical synaptic output. Certain structural their soma and are generally confined to the properties are clearly correlated. It has been fourth layer or to the sublayer in which their demonstrated that the somata of pyramidal soma is situated. They may be strongly strati- neurons projecting to a particular target are fied horizontally or may show a radiating or not only located in one and the same layer or even vertically elongated distribution [423]. sublayer, but also show striking similarities The spiny stellate cells are most probably the with regard to dendritic morphology, thalamo- principal, but not the exclusive targets of the cortical connectivity and distribution of their thalamocortical afferents terminating in lamina axon collaterals. It seems likely that all pyrami- IV of the sensory cortical areas [208]. Ahmed dal neurons projecting to a particular target et al. [4] studied the connectivity of spiny stel- are in receipt of similar extracortical and intra- late cells in the visual cortex of the cat. They cortical inputs and that they participate in a presented evidence suggesting that, of the similar way in the intrinsic circuitry of the ce- asymmetrical synapses on these elements, 45% rebral cortex. comes from lamina VI pyramidal cells, 28% from other spiny stellate cells and 6% from thalamic afferents. The source of the remaining Atypical Pyramidal Cells 21% of asymmetrical synapses remained obscure. Spiny stellate cells presumably receive In the mammalian neocortex, neurons occur inhibitory afferents from neurogliaform cells, that lack one or several of the features charac- basket cells and chandelier cells [338, 424, terizing typical pyramidal cells, but that are 815]. nevertheless considered to belong to the pyra- The axon of the spiny stellate cells may as- midal cell group. These atypical pyramids in- cend to more superficial layers or descend to clude: (a) elements in which the apical den- deeper layers. Many of these elements are pro- drites are shortened (Fig. 15.32 B,C), reduced vided with both ascending and descending ax- (Fig. 15.30 B) or absent [171, 756, 775]; (b) onal branches. Short collateral branches estab- “star pyramids”, in which the dendrites radiate lishing synaptic contacts within the sublayer of out from the soma in all directions, rather origin are invariably present (Fig. 15.30 D–F). than forming a basal skirt (Fig. 15.30 C); (c) The axon terminals of spiny stellate cells form sparsely spiny pyramidal cells [149, 277, 315, asymmetrical synapses mainly with dendritic 836, 837]; (d) “pure” projection pyramidal spines [446, 646, 779]. The investigations of neurons, whose axons do not emit any intra- Lund and collaborators [422–424, 426] have cortical collaterals [234]; and (e) intrinsic py- shown that the spiny stellate cells, present in ramidal neurons, whose axons remain within the various sublayers of layer IV of the monkey the cortex and often ascend (Fig. 15.30 B) or primary visual cortex, differ markedly with re- form ascending and descending branches (Fig. spect to their projections to other layers. It 15.30 C) [363, 756]. However, the most remark- seems likely that the short local collaterals of able modified pyramidal neurons are doubtless the spiny stellate cells mainly contact other the so-called spiny stellate cells (Fig. 15.30 D– spiny stellate cells [646] and that their ascend- F). These elements occur exclusively in lamina ing and descending axonal branches establish IV of primary sensory areas of the neocortex, numerous synaptic contacts with superficial 560 Section II Structure of Spinal Cord and Brain Parts and deep pyramidal neurons, but direct evi- 1. They are evidently non-pyramidal, i.e. they dence for the existence of such outputs is lack- have no conical soma and lack a dominant ing [137]. Nevertheless, it may be safely as- apical dendrite. On that account, the group sumed that spiny stellate cells play a crucial is often referred to as non-pyramidal, but role in the radial propagation of the activity this designation is not entirely satisfactory fed by thalamocortical afferents into layer IV because many neurons belonging to the py- of primary sensory cortical areas. ramidal category do not show a pyramidal Although spiny stellate cells represent typical morphology (Fig. 15.30 C–F). local circuit neurons and lack an apical den- 2. Their dendrites bear only few spines or are drite, they nevertheless have to be considered entirely spine-free. This is a very important as modified pyramidal neurons [423, 756]. The distinguishing feature, even though the den- reasons for this interpretation are as follows: drites of some otherwise typical pyramidal 1. The dendrites of pyramidal neurons as well neurons are also aspinous [315]. as those of spiny stellate cells are densely 3. Their somata have both symmetrical and covered with spines and the axon terminals asymmetrical axosomatic synapses, whereas of both cell types form synapses of the pyramidal cell bodies possess only symme- asymmetrical type, mainly with dendritic trical axosomatic synapses [557]. spines [403, 646]. 4. Their axons do not leave the cortex, which 2. In lamina IV of the primary sensory areas is why the cells under consideration are of- containing spiny stellate cells, neurons pro- ten referred to as local circuit neurons. vided with a more or less developed apical However, it should be recalled that many py- dendrite are frequently observed that in all ramidal cells also possess exclusively intra- other respects cannot be distinguished from cortical axons (Fig. 15.30 B–F) and, hence, typical spiny stellate neurons. The “star pyr- also belong to the category of cortical local amids” in the somatosensory cortex of the circuit neurons. squirrel monkey [336] are a good example 5. With a single exception, their axon termi- of such an intermediate form (compare Fig. nals contain flattened vesicles and form 15.30 C with F). symmetrical synapses with their postsynap- 3. Peinado and Katz [543] have presented evi- tic targets, both features suggesting an inhi- dence that during ontogeny lamina IV stel- bitory function [177, 403, 539]. late cells initially extend an apical dendrite 6. An inhibitory function is also suggested by to lamina I and only later lose this process the fact that most of the cells under consid- and develop their mature stellate morphol- eration use GABA as their primary neuro- ogy. transmitter [285, 306, 371, 558, 616]. 7. A certain proportion (25–30%) of the GABAergic cortical neurons also express one Local Circuit Neurons or several neuropeptides. The neuropeptides detected in cortical neurons include sub- In the preceding sections, the neocortical pyra- stance P, vasoactive intestinal polypeptide midal neurons have been discussed. It was (VIP), cholecystokinin (CCK), neuropeptide pointed out that this category not only encom- Y (NPY), somatotropin-release-inhibiting passes true pyramidal cells, but also atypical factor (SRIF), corticotropin-releasing factor elements that have lost one or several of the typ- (CRF) and tachykinin (TK) [304, 350, 440]. ical pyramidal characteristics. Pyramidal neu- Several subpopulations of GABAergic corti- rons account for 60–85% of the total neuronal cal neurons appeared to be definable on the population of the neocortex [240, 556, 589, basis of their immunoreactivity for certain 824]. The remaining 15–40% of neocortical neu- neuropeptides. To give a few examples: in rons include a variety of morphological types the visual cortex of the rat a subpopulation that have the following features in common: of bipolar cells can be labelled with antibod- 15 Telencephalon: Neocortex 561

ies to VIP [559], whereas in the neocortex of many of the morphological types of neocorti- the monkey, subpopulations of chandelier cal interneurons can be subdivided into two or cells and of double bouquet cells have been more biochemically definable subgroups. The shown to be immunoreactive for CRF [406] synaptic relationships of the various types of and TK [140], respectively. Another subpop- neocortical interneurons to be discussed are ulation of double bouquet cells has been semidiagrammatically indicated in Fig. 15.37 B. shown to be immunoreactive for SRIF [133]. Stellate neurons are found in all cortical 8. It has been shown that differential immuno- layers (Fig. 15.36 A,C,D) [416]. The dendrites reactivity for the calcium-binding proteins of these elements radiate from the soma in all parvalbumin (PV), calbindin (CB) and calre- directions and branch infrequently. Their axo- tinin (CR) can be used as a selective marker nal arborization forms a local plexus occupy- for different subpopulations of neocortical ing approximately the same territory as that local circuit neurons [105, 134, 174]. In the covered by the dendrites. Distinctive axonal primate neocortex, among the most charac- terminals are lacking [563] (Fig. 15.36 B). Be- teristic types of neurons immunoreactive for cause of the limited spread of their axonal sys- PV are chandelier cells and large basket tem, the stellate cells are also referred to as lo- cells, whereas for CB they are double bou- cal plexus neurons. quet cells, and for CR, double bouquet and Neurogliaform or spiderweb cells form a spe- bipolar cells [134]. cial class of stellate cells (Fig. 14.36 B). These elements have a small, spherical soma and short In brief, the mammalian neocortex contains a sinuous dendrites. Their axons arborize profu- large population of non-pyramidal, inhibitory sely around the soma, forming a dense feltwork. interneurons with smooth or sparsely spinous Characteristically, small empty spaces appear in dendrites. These elements use GABA as their this feltwork, representing the positions occu- primary neurotransmitter and some also pro- pied by the unstained somata of other neurons duce one or several neuropeptides. The popu- [338]. Neurogliaform or spiderweb cells have lation of neocortical neurons thus outlined is been observed in all layers of the cortex, but morphologically heterogeneous, and numerous they are particularly concentrated in the primate authors, including Lorente de Nó [416], Jones somatosensory and primary visual cortex. The [336], Feldman and Peters [179], Peters and function of the spiderweb cells is unknown. Jones [561], Fairén et al. [177], Lund [424], Jones [338] considered it likely that the elements Lund and Yoshioka [428], Lund and Lewis concentrated in lamina IVof sensory cortices re- [427], Lund et al. [430] and DeFelipe [136] ceive thalamic afferents and synapse mainly have attempted to subdivide this population of with the spiny stellate cells of that layer. Kisvár- neurons into different groups or types, using day et al. [371] observed synaptic contacts be- the size and shape of the somata, the shape of tween the axons of neurogliaform cells and the the dendritic field and the number, distribu- distal dendrites of pyramidal cells. tion, length and branching patterns of individ- There are numerous neurons in the neocor- ual dendrites, the preferred direction of axons tex which, judging from the disposition of and axonal branches and configuration of axo- their dendritic tree, would fall in the category nal terminals as criteria. The following discus- of stellate cells, but which are distinguished sing of the non-pyramidal, smooth or sparsely from the elements in that category by the spinous local circuit neurons in the mamma- course and/or mode of termination of their ax- lian neocortex is based on the publications ci- ons. The chandelier cells and basket cells, ted above, on the excellent characterizations of which have already been discussed in a pre- particular cell types [338, 343, 554, 555, 563, vious section, represent two distinct types of 695], published in the first volume of the Cere- such specialized stellate cells. bral Cortex [560], and on other sources to be Chandelier cells have been thus named be- quoted below. As has already been indicated, cause their profuse axonal plexuses give rise to 562 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.36. Local circuit neurons with smooth or sparsely spiny dendrites in the neocortex. A, Multipolar neuron with axonal arcades in lamina III of the somatosensory cortex of the squirrel monkey (Jones [336]); B, neurogliaform or spiderweb cell from layer IV of the somatosensory cortex of the squirrel monkey (Jones [338]); C, stellate neuron with sparsely spiny dendrites extending from lamina IVC to lamina VI in the primary visual cortex of the macaque (Lund [422]); D, horizontal cell in lamina I of the neocortex of the hedgehog (Valverde and Facal-Valverde [757]); E, stellate cell from lamina IV in rat visual cortex (Peters and Saint Marie [563]); F, G, chandelier cells in laminae III and V of the neocortex of the macaque (Jones [340]); H, J, basket cells in laminae II and V of the neocortex of the macaque (Jones [340]); K, neuron with local, beaded axon in lamina IV of the prefrontal cortex of the macaque (Lund and Lewis [427]); L, bipolar or vertical cas- cade neuron in lamina III of the prefrontal cortex of the macaque (Lund and Lewis [427]); M, neuron with rising axon in lamina IV of the prefrontal cortex of the macaque (Lund and Lewis [427]); N, neuron with axon connecting laminae II and III in the prefrontal cortex of the macaque (Lund and Lewis [427]); O, double bouquet cell in lamina II of the primary visual cortex of the macaque (Werner et al. [810]; P, neuron with rising axon in upper zone of lamina V of the primary auditory cortex of the cat (Fairén et al. [177]); Q, Mar- tinotti cell in lamina V of the visual cortex of the cat (Wahle [801]); R, horizontally oriented neuron in lamina VI of the prefrontal cortex of the macaque (Lund and Lewis [427]) 15 Telencephalon: Neocortex 563 a large number (up to 300) of highly charac- get soma. One basket cell contributes to nu- teristics, vertically oriented “candles”, each merous baskets, and terminal axonal branches consisting of a preterminal axonal branch that of several basket cells contribute to a single, forms a short row of terminal boutons. These pericellular plexus. The somata of basket cells vertical arrays, which are designated as axon are concentrated in laminae II and V. terminal cartridges, synapse with the axon ini- These cells – with their radiating short den- tial segments of pyramidal cells (Figs. 15.31B, drites and their long horizontal axonal 15.36 F,G) [177, 555]. The axon initial segments branches – form a distinct group of cortical lo- of spiny stellate cells may also receive synapses cal circuit neurons, which have been desig- from chandelier cells [424]. nated as large basket cells [177]. Cells of this Chandelier cells occur in layers II–V, but are type have so far only been observed in the cat most common in layer II. Because the pyrami- and monkey [343]. They can express PV, CB, dal neurons projecting to the ipsilateral and NPY, CCK and occasionally CR and SRIF. They contralateral neocortex are mainly situated in never express VIP [440]. layers II and III and because the axon initial In the neocortex of different mammals, both segments of these supragranular pyramids re- “higher” and “lower” small, smooth or sparse- ceive a much richer synaptic supply from the ly spinous intrinsic neurons have been ob- chandelier cells than the infragranular ele- served, whose axons consistently produce mul- ments, it has been suggested that chandelier tiple synaptic contacts on the somata of pyra- cells principally influence cortico-cortical cir- midal cells. Fairén et al. [177] named these ele- cuitry [570, 698]. Chandelier cells typically ex- ments small basket cells. According to their ob- press one or both of the calcium-binding pro- servations, these small basket cells are multi- teins PV and CB [134]. polar elements with rather wide dendritic It is noteworthy that the axon initial seg- fields, sometimes showing a predominant den- ments of pyramidal neurons are not exclusively dritic tuft oriented towards the pial surface. contacted by chandelier cells. Gonchar et al. The axon is primarily descending and typically [250] detected in the visual cortex of rat and produces frequently ramifying and “curvy” col- monkey a population of somatostatin-express- lateral branches. They differ from other basket ing, GABAergic neurons, whose axons inner- cells in that they express VIP [440]. vate somata, dendritic spines and initial seg- A third class of basket cells is formed by the ments of pyramidal neurons. It is also note- so-called nest basket cells. These elements, worthy that, according to DeFelipe [135], loss which were only recently shown to represent a of chandelier cells may represent a key compo- distinct class of soma-targeting cell, occupy an nent in the aetiology of epilepsy. intermediate position between large and small Basket cells are among the largest non-pyra- basket cells [805]. Their primary axonal midal cells in the neocortex (Figs. 15.31 A, branches are long and ramify sparsely, much 15.36 H,J). Their poorly ramifying dendrites like those of the large basket cells, but their radiate in all directions, but vertically oriented clusters of secondary axonal branches more dendrites prevail in some and give these cells a closely resemble those of the small basket cells. bitufted appearance. The axons of the basket The nest basket cells do not typically express cells are either ascending or descending and CR and never express VIP [440]. give rise to four or more horizontal branches Approximately 50% of all inhibitory neocor- at various levels. These collateral branches are tical interneurons are basket cells [440]. myelinated and may reach a length of 1 mm or The classical descriptions of basket cells pre- more. At intervals, they issue short ascending sented by Cajal [84] created the impression or descending terminal branches that contrib- that these elements synapse exclusively with ute to the formation of pericellular baskets the somata and proximal dendrites of pyrami- around pyramidal cell bodies. Each terminal dal neurons. However, in several later studies branch forms a series of synapses with its tar- [207, 369, 697, 700], evidence was presented 564 Section II Structure of Spinal Cord and Brain Parts 15 Telencephalon: Neocortex 565 , chan- Ch , etc., cortical II , I , bipolar cell; Bi , basket cells; Ba , thalamocortical fibres; . B thc , etc., different types of interneurons mentioned in the I2 , ; inhibitory neurons and their terminals by filled profiles in I1 black , spiny stellate cells; SS , horizontal cell of Cajal; and excitatory plus inhibitory elements in A HC , pyramidal neurons; P hown by open profiles in , double bouquet cell; DB , neurogliaform or spiderweb cell; N , cortico-cortical fibre; cc Neocortical circuits showing excitatory elements in , Martinotti cell; M . Modified from Nieuwenhuys [509] Fig. 15.37. layers. Excitatory neurons and their synaptic terminals are s red delier cell; text; 566 Section II Structure of Spinal Cord and Brain Parts suggesting that axon terminals of basket cells rons described in the literature is not possible also synapse with the somata of spiny stellate here. However, some, i.e. the bipolar cells, the cells, the distal dendrites and axons of pyrami- bitufted cells, the double bouquet cells and the dal cells and the somata and dendrites of non- elements with ascending axons known as Mar- pyramidal neurons. Kisvárday et al. [372], tinotti cells, may be briefly commented upon. using biocytin as a label and taking advantage Bipolar cells have a small, spindle-shaped, of the fact that large basket cells are GABAer- vertically oriented cell body, from which one gic and contain PV, demonstrated that in the primary ascending and one primary descend- visual cortex (area 18) of the cat these cells not ing dendrite arises (Fig. 15.32 G (b)) [554]. only synapse with pyramidal neurons, but also These two processes and their ramifications, establish an average of four to six perisomatic which bear few or no spines, extend for long contacts onto other large basket cells. A large distances through the depth of the cortex, pro- basket cell in lamina II was found to synapse ducing a narrow and very elongated dendritic with the somata of 58 other large basket cells, tree. The axons of the bipolar cells usually whereas a large lamina V basket cell appeared arise from one of the primary dendrites and to contact 33 of its fellow cells. From these ob- form a plexus that is also narrow in spread servations, Kisvárday et al. [372] concluded and vertical in orientation [554]. Bipolar cells that large basket cells form an interconnected occur in layers II–VI and typically express VIP network in area 18 of the visual cortex. As- and CR [440]. suming that the GABAergic large basket cells Ultrastructural investigations have shown are inhibitory, they proposed that: (a) a large that there are two different populations of bi- basket cell provided direct perisomatic inhibi- polar neurons, one forming symmetrical sy- tion onto a number of pyramidal cells (at least napses and the other forming asymmetrical sy- 200–300) and in the range of about 50 other napses [106, 177, 559, 562]. The bipolar cells basket cells, (b) the target cells of the directly with axons forming symmetrical synapses re- inhibited basket cells become facilitated via a lease principally GABA, but also express VIP disinhibitory effect and (c) the number of neu- [440, 458, 571]. The axons of these inhibitory rons disinhibited through this process may well elements preferentially synapse with dendritic greatly exceed the number of elements that are shafts. directly inhibited by the large basket cells. A certain proportion of the neocortical bipo- Vertically Oriented Neurons. If we consult lar neurons that form symmetrical synapses the extensive inventories of neocortical neu- can be labelled with antibodies to ChAT, a spe- rons provided by Cajal [84], Lorente de Nó cific marker for cholinergic neurons. The axon [414], Jones [336], Feldman and Peters [179], terminals of these cholinergic bipolar cells Fairén et al. [177], Lund [422, 424], Lund and most commonly synapse with small- to medi- Yoshioka [428], Lund and Lewis [427], Lund et um-sized dendritic shafts and less frequently al. [429, 430], DeFelipe [136] and others, it ap- with apical dendrites and with the somata of pears that numerous smooth or sparsely spi- neurons [307, 540]. nous local circuit neurons in this structure The bipolar cells that form asymmetrical sy- show an overall vertical orientation. This ori- napses are excitatory by releasing only VIP entation may concern their dendritic trees [440]. The axons of these elements give rise to (Fig. 15.36 K), their axonal systems (Fig. vertically oriented branches which parallel the 15.36 M–Q) or both (Fig. 15.36 L). Among the trajectories of clustered apical dendrites of py- neurons with vertically oriented axons, ele- ramidal neurons, forming multiple synapses ments with rising axons (Fig. 15.36 M,P,Q), with spines on these processes (Fig. 15.32 G) descending axons (Fig. 15.36 L,N) and both [562]. However, axonal branches of the excita- descending and rising axons (Fig. 15.36 O) may tory bipolar cells have also been observed to be distinguished. A discussion of all of the contact the shafts of apical dendrites and the types of vertically oriented cortical interneu- somata and dendrites of non-pyramidal cells. 15 Telencephalon: Neocortex 567

Bitufted cells have dendrites arising mainly axon terminals of these cells do not form sy- from the upper and lower poles of the soma, napses with apical dendrites, but rather with forming, as the name implies, two dendritic basal dendrites and with oblique branches of tufts [557]. The dendrites diverge initially, but the apical dendrites of pyramidal neurons and at some distance from the soma often assume with postsynaptic structures belonging to non- a radial orientation. Many neurons of this type pyramidal neurons. Double bouquet cells ex- have local axonal plexuses which partly overlap press CB [139, 146] and can also express VIP with the field occupied by their dendritic trees or CCK [440]. [557]. From these local axonal plexuses, ra- Martinotti cells are multipolar or bitufted dially oriented branches may extend into the neurons with smooth or sparsely spinous den- adjoining cortical layers. Bitufted cells are dendrites. Their distinguishing feature is a long dritic-targeting cells that occur in layers II–VI ascending axon which reaches layer I, where it [702]. They can express NPY, SRIF, VIP, CCK, forms a terminal arborization. The axon CB and CR, but not PV [440]. emerges either from the upper surface of the Among the bitufted cells, elements are found soma or from an ascending dendrite. The ini- whose axons produce tight, radially oriented tial part of the axon gives rise to a number of long plexuses of thin, parallel axonal branches. descending collaterals, which form a local ter- These plexuses are generally designated as minal plexus [84, 177, 442, 643, 754]. “Classi- “bundles” or “horse-tails”. Smooth or sparsely cal” Martinotti cells have a single ascending spinous cortical neurons provided with these axon, but Jones et al. [351] and Wahle [801] highly characteristic long, fascicular axonal described “double bouquet type” Martinotti systems have been commonly referred to as cells whose axon branched into a bundle of double bouquet cells [177, 554, 561, 695]. Cajal two to eight long, ascending collaterals (Fig. [84] used the name cellules à double bouquet 15.36 Q). dendritiques to refer to a number of cell types Martinotti cells occur in all cortical layers ex- with highly different axonal ramification pat- cept layer I, but they have been most frequently terns, including elements with the long radially found in layers V and VI. They use GABA as a oriented arrays of axonal branches discussed neurotransmitter [696] and may additionally above. However, it has become customary to contain a neuropeptide, e.g. TK [351] or soma- designate only smooth or sparsely spinous cells tostatin [801]. Little is known with certainty showing this particular axonal pattern as dou- about the afferent and efferent connections of ble bouquet cells. In the current literature, the Martinotti cells. On the basis of a Golgi even elements with these long vertical axonal study of the visual cortex of the mouse, Ruiz- branches whose dendritic trees are not of a bi- Marcos and Valverde [643] suggested that Marti- tufted appearance are still referred to by this notti cells situated in layer V of the cortex re- name (Fig. 15.36 O) [336, 695]. ceive terminals from both superficial and deep Typical double bouquet cells have only been pyramidal cells, that their local axonal plexuses observed in laminae II and III of the neocortex contact deep pyramidal cells and that their ter- of cats and primates [33, 695]. It has been minal plexuses in layer I impinge upon the api- shown that double bouquet cells form symme- cal dendritic tufts of pyramidal cells. It has been trical synapses containing flat or pleiomorphic observed that the axons of certain Martinotti vesicles with their target structures and use cells, after having reached lamina I, give rise GABA as their principal neurotransmitter [139, to long, horizontal collaterals [414]. According 695]. It was formerly believed that the verti- to Szentágothai [721], these branches may run cally oriented axons of double bouquet cells for several millimetres through lamina I, making mainly synapse with the apical dendrites of synapses with the apical dendrites of numerous pyramidal neurons [84, 103, 721]. However, the pyramidal cells. studies by Somogyi and Cowey [694] and by Horizontal Cells. Local circuit neurons show- DeFelipe et al. [139, 140, 146] showed that the ing an overall horizontal orientation are almost 568 Section II Structure of Spinal Cord and Brain Parts exclusively found in layers I and VI. It is im- synthesize GABA [760] and that their cell portant to note that these layers, although far bodies have both symmetrical and asymmetri- apart in the adult cortex, are both derivatives cal axosomatic synapses [556] indicates that of a single embryonic pallial zone, i.e. the pri- they represent inhibitory local circuit neurons. mordial plexiform layer. During the formation Remaining Inhibitory Interneurons. There of the cortex, immature bipolar cells migrate are many other smooth or sparsely spinous, in- peripherally and together form a compact cor- hibitory neurons in the neocortex that are dif- tical plate, which splits up the primordial ficult to classify or that have not yet been rec- plexiform layer into a superficial and deep ognized as belonging to a particular morpho- zone. The former becomes layer I, and the lat- logical type [136]. In Fig. 15.37 B some of these ter gives rise to the deep zone of layer VI in elements, designated provisionally as I1–I4, are the mature cortex. The intervening layers are included. all derivatives of the cortical plate (Fig. 2.10 H– Element I1, which is situated in layer II, re- K) [437]. ceives excitatory afferents from bipolar cells, Horizontal Cells of Cajal. Bipolar neurons whereas its axon forms symmetrical, presum- occurring in layer I of the cortex are known as ably inhibitory synapses with the apical den- the horizontal cells of Cajal. These elements drites and the somata of superficial pyramidal are provided with one or a few long, smooth neurons [562]. dendrites that pursue a course parallel to the Element I2, which is situated in layer IV, is cortical surface. Their axons, which pass hori- contacted by thalamic afferents and its axonal zontally like the dendrites, may attain a con- endings impinge upon the somata of superfi- siderable length (Fig. 15.36 D). Curiously en- cial pyramidal neurons, as well as on the so- ough, horizontal cells in layer I may have more mata and proximal dendrites of bipolar cells than one axon [84, 757]. The elements under [562]. discussion are GABAergic and additionally Element I3 is a local plexus neuron situated contain the neuropeptide CCK [779]. The ax- in the upper part of layer IV. It receives affer- ons of the horizontal cells probably enter into ents from thalamocortical fibres, from axon synaptic contacts with the apical dendritic collaterals of superficial and deep pyramidal branches of pyramidal neurons [416, 757]. Sev- neurons, and from basket cells. Its ascending eral authors [84, 167, 438, 537, 757] have ex- axon forms synaptic contacts with the apical pressed the opinion that the conspicuous Ca- dendritic shafts, somata, and axon initial seg- jal-Retzius cells found in layer I of the imma- ments of superficial pyramidal neurons [815]. ture cortex undergo morphological changes Element I4, finally, is a local plexus neuron and transform to horizontal cells. situated in the upper part of layer V. It receives Horizontal Cells in Layer VI. The deep zone terminals from thalamocortical fibres, and its of layer VI contains numerous medium-sized axonal endings establish synaptic contacts with horizontal cells (Fig. 15.36 R) [427, 497, 498]. the somata and apical dendritic shafts of deep One or a few long dendrites arise from the pyramidal neurons [815]. ends of their fusiform cell bodies, but some It is noteworthy that defects in the neuro- shorter dendrites may also arise from the transmission of GABAergic neocortical inter- upper and lower surface of the soma. Their ax- neurons may well play a role in the pathophys- ons, which like the principal dendrites often iology of schizophrenia. A detailed discussion pursue a horizontal course, give off varicose of the voluminous literature on this subject is side branches. In Golgi material, the axons of beyond the scope of the present work (see [46] the horizontal cells under discussion can rarely for a review). However, the results of two re- be traced to their final destination, and very cent studies, those of Lewis et al. [407] and of little is known about the afferent and efferent Konopaske [384], may be briefly mentioned. connections of these elements. The fact that Lewis et al. [407] found that a deficiency in the horizontal cells in layer VI most probably signalling by brain-derived neurotropic factor 15 Telencephalon: Neocortex 569 through its receptor, the tyrosine kinase Trk 1. The assembled data are derived from (tropomyosin-related kinase) B, leads to re- many different species and from many different duced GABA synthesis in the PV-containing cortical areas; hence, it can not be expected a subpopulation of inhibitory GABAergic neu- priori that the diagrams present a reliable pic- rons in the dorsolateral prefrontal cortex of in- ture of the microcircuitry of the mammalian dividuals with schizophrenia. They indicate cortex. However, it is worthy of note that the that the resulting alteration in perisomatic in- extensive ontogenetic and cytoarchitectonic hibition of pyramidal neurons contributes to a studies of Brodmann [70] have shown that all diminished capacity for the synchronized neu- neocortical areas in all of the many species he ronal activity that is required for working investigated represent variations on a common memory function. basic plan, which he designated as “der Konopaske et al. [384] presented evidence sechsschichtige tektogenetische Grundtypus”. suggesting that in schizophrenia the neurotrans- Rockel et al. [620] counted the number of neu- mission in another type of neocortical interneuronal cell bodies in a narrow strip (30 lm) ron, the chandelier cell, might also be compro- through the depth of the neocortex in several mised. As has been discussed, the axon term- functional areas (motor, somatosensory, pri- inals of these GABAergic elements form linear mary visual, frontal, parietal and temporal) arrays, termed cartridges, that synapse on the and in many species (mouse, rat, cat, monkey axon initial segments of pyramidal neurons and human). With the exception of the pri- (Fig. 15.31 B). These cartridges are immunoreac- mary visual cortex in primates the same abso- tive for the GABA membrane transporter-1, lute number (about 110) of neurons was found which regulates the duration and efficiency of in all areas and in all species. In the primate GABAergic neurotransmission. Konopaske et primary visual cortex there appeared to be ap- al. [384] reported that the density of GABA proximately 2.5 times more neurons. According membrane transporter-1-immunoreactive car- to Rockel et al. [620] these findings suggest tridges is reduced in schizophrenia, particularly, that the intrinsic structure of the neocortex is but not exclusively, in the prefrontal cortex. basically uniform and that differences in cytoarchitecture and function reflect differences in extrinsic connections. On the basis of a study of a vast array of Golgi material, Szentágothai Microcircuitry of the Neocortex [721] concluded that the various neocortical cell types show only little variation from mouse to human. Moreover, the ratio of differ- Introduction ent cell types has been shown to be essentially the same in the motor and visual cortices of The data concerning the synaptic connections rats and cats [824]. It may be concluded that of neocortical neurons and principal neocorti- the neocortex of different mammalian species cal afferents, discussed in the previous section, is built according to a common basic plan, and were employed to compose two summary dia- that, hence, the structural features assembled grams of the microcircuitry of the neocortex in Fig. 15.37 A,B may well present an image of (Fig. 15.37 A,B). Moreover, the principal results that plan. However, the reservation should be of some recent studies on the connectivity of a made that the spiny stellate cells depicted (SS) certain type of interneuron have been summa- occur only in specialized primary sensory cor- rized in an additional diagram (Fig. 15.38). tices and that excitatory bipolar cells (Bi) have (The symbols used in the text – Ba, Bi etc. – been demonstrated so far only in the primary correspond to those in the figures and are ex- visual cortex of the rat [559, 562] and cat plained in the legend of Fig. 15.37.) Before [177]. commenting on these diagrams, a few caution- 2. If our diagram presents an image of the ary remarks should be made. microcircuitry of the neocortex, it does so 570 Section II Structure of Spinal Cord and Brain Parts merely in a qualitative sense. The total number targets, whereas cortico-cortical fibres arise of neocortical synapses in the neocortex has mainly from superficial pyramidal neurons. been estimated to be 3 ´1014 [104]. Our dia- The very extensive axon collateral systems of gram contains 157 synapses. Because excitatory pyramidal neurons primarily contact other py- (grossly: pyramidal) neurons are by far more ramidal neurons. There is evidence suggesting numerous than inhibitory (grossly: non-pyra- that superficial pyramidal neurons contact other midal) neurons and because the intracortical superficial and deep pyramidal neurons and that axon terminal- and collateral systems of most deep pyramidal neurons impinge on other deep of the former are vastly more extensive than and superficial pyramidal neurons. It is known those of the latter, the excitatory synapses may that, in the primary visual cortex, the axon col- be expected to outnumber the inhibitory sy- lateral systems of pyramidal neurons constitute napses. The validity of these estimates has reciprocal patchy networks, which can be traced been confirmed by the quantitative synaptolog- over distances of up to 7 mm [368]. It has been ical studies of DeFelipe et al. [142]. These suggested that these networks link sites with authors found that in neocortical areas as dif- similar physiological characteristics, such as ferent as the hindlimb area of the rat somato- orientation preference. This would imply that, sensory cortex and the human anterolateral within the primary visual cortex, different inter- temporal cortex, symmetrical synapses formed woven networks of pyramidal axon collaterals 10.7% and 11.5% of the total synapse popula- are present and that the extent of these networks tion, respectively. However, in our diagram would be confined to that cortical area. there are 76 excitatory and 82 inhibitory sy- Another remarkable local network of inter- napses. More specifically, about 10% of the ex- connected pyramidal neurons is found in the citatory synapses (7 out of 76) is formed by prefrontal cortex. The prefrontal cortex (PFC) terminals of cortico-cortical fibres. Because has been identified as the key neocortical re- cortico-cortical projections constitute by far gion supporting working memory [244]. Extra- the largest neocortical input system, this per- cellular recordings during delayed response centage is most probably much higher. tasks have shown that a considerable fraction 3. Several functionally important structural of prefrontal cortical neurons remain active features have been neglected. For example, it is after the cue (e.g. a particular sensory stimulus known that different thalamic nuclei project to or event) and until the task is completed. Such different sets of neocortical (sub)layers and activity, which can persist for several seconds, that the same holds true for particular cell has been proposed as the neural correlate of types within some thalamic nuclei, for instance working memory [804]. It has been recently the lateral geniculate nucleus. In our diagrams established [806] that the PFC of the ferret no distinction has been made between these contains a network of heavily interconnected various projections. Rather, they are taken to- cells with characteristic dual apical dendrites gether as if representing one single thalamo- and a particularly wide and dense system of cortical system. basal dendrites. These “hyper-reciprocally” connected cells appeared to have synaptic and functional properties essential for supporting Networks of Pyramidal Neurons persistent activity in the PFC. A continuous network of excitatory elements The pyramidal neurons (P) are doubtless the potentially involving the entire neocortex and principal neurons in the neocortex (Fig. even extending into the hippocampal region is 15.37 A). They are not only by far the most nu- constituted by the ipsilaterally and contralater- merous cellular elements in that structure, but ally projecting cortico-cortical pyramidal neu- also constitute its sole output system and its rons. The continuity of this network is empha- largest input system. Separate sets of deep py- sized by the fact that the cortico-cortical fibres ramidal neurons project to different subcortical terminate throughout the neocortex mainly in 15 Telencephalon: Neocortex 571 the superficial layers, where the cortically pro- lier cells and other spiny stellate cells and prob- jecting pyramidal neurons are concentrated. ably impinge on basket cells and double bouquet The presence of this strongly developed, ubi- cells (DB). Basket cells receive afferents from quitous network warrants the conclusion that other basket cells and probably from spiny stel- the neocortex communicates first and foremost late cells and double bouquet cells and make ef- with itself. However, the fact that the cortico- ferent contacts with spiny stellate cells and prob- cortical fibres most probably impinge not only ably with smooth stellate elements (e.g. I3). on superficial pyramidal neurons, but also on Excitatory local circuit neurons, i.e. spiny the apical dendrites and terminal dendritic bou- stellate cells and some bipolar cells, are, as far quets of deep pyramidal neurons indicates that as it is known, confined to primary sensory the various subcortical centres to which these areas. Inhibitory local circuit neurons are of elements project are continuously kept informed many different types and occur throughout the about the successive transformation of data oc- neocortex. Some types of local circuit neurons curring along the cortico-cortical processing are lamina-specific, i.e. their somata are situ- streams. This applies in particular to the cau- ated mainly in one or in two adjacent layers date-putamen complex and to the pontine nu- (SS, DB, HC, M). clei, centres which are known to receive projec- In the highly differentiated primary visual tions from almost the entire neocortex. cortex of the rhesus monkey, thalamocortical Thalamocortical fibres (thc) synapse directly fibres carrying particular kinds of visual infor- with pyramidal neurons, although the number mation terminate in sharply defined sublayers of contacts made by such fibres with particular of lamina IV (Fig. 15.26) [189, 190, 425]. The types of pyramidal neurons is subject to con- detailed Golgi studies carried out by Lund and siderable variation [815]. her associates [424, 428, 430] showed that each of these thalamic recipient sublayers, and some other layers as well, contain several types of lo- Interneuronal Systems cal circuit neurons. The dendritic trees of all of these elements are strictly confined to the All types of neocortical local circuit neurons layer or zone in which their soma is situated. have been reported to establish synaptic con- Their axons project to more superficial layers, tacts with pyramidal neurons. Most of these to deeper layers or to both. These interlaminar elements, namely spiny stellate cells (SS), bipo- projections are highly specific, targeting from lar cells (Bi), neurogliaform cells (N), basket one to four laminar divisions, depending on cells (Ba), horizontal cells of Cajal (HC) and the type of neuron. The strictly radial orienta- cells of the provisional types I3 and I4, have tion of all of these projections is presumably been suggested to receive a direct input from connected with the preservation of the detailed the thalamus, and some of these (SS, Bi, Ba, topographical map existing within one of the I3) are also contacted by axon collaterals of py- thalamic recipient sublayers. Some of these ramidal neurons. Martinotti cells (M) are pri- types of local circuit neurons may well be in- marily intercalated between different pyramidal volved in the selective processing and transfer neurons. The nature of the input to one type of one particular type of visual information. of local circuit neuron, the chandelier cell Important clues to the specific functional (Ch), is entirely unknown. roles of the various types of inhibitory inter- Several types of neocortical interneurons are neurons derive from their differential axonal in receipt of afferent contacts from other inter- arborizations (Fig. 15.36) and from the loca- neurons and/or establish efferent contacts with tion of their synaptic connections with pyra- such elements. Prominent among them are the midal neurons [25]. Most of these inhibitory spiny stellate cells and the basket cells. Spiny cell types distribute their synaptic contacts stellate cells are contacted by basket cells and preferentially to selected membrane domains most probably by neurogliaform cells, chande- of pyramidal elements [440, 702, 815]. Thus, 572 Section II Structure of Spinal Cord and Brain Parts chandelier cells (Ch) terminate on the axon During the last decade, important new find- initial segments of pyramidal neurons, whereas ings concerning the connectivity and the regula- basket cells (Ba) target the somata and proxi- tion of the activity of neocortical inhibitory in- mal dendrites of these elements (Fig. 15.31). terneurons have been reported. Thus, it has The contacts of the chandelier and basket cells been shown that (1) groups of inhibitory inter- are optimally localized for controlling the out- neurons are coupled through gap junctions, (2) put and oscillatory synchronization of groups the axons of certain types of interneurons form of pyramidal neurons [206]. Small stellate cells autapses at their own soma or dendrites, and (3) such as those belonging to the provisional certain types of interneurons are specifically or types I1, I2 and I4 may also participate in preferentially contacted by particular extrathala- these specific functions. Neurogliaform cells mic neuromodulatory systems. These phenom- (N) preferentially innervate the proximal and ena, and their (putative) functional significance, mid-range dendritic domains. (This feature of will now be briefly discussed (Fig. 15.38). the neurogliaform cells is not shown in Fig. 1. Electrical coupling of neocortical inhibi- 15.37 B.) The neurogliaform neurons are opti- tory interneurons. Gap junctions or electrical mally positioned to influence dendritic synapses are specialized loci where channels processing and integration of synaptic inputs bridge the plasma membranes of two adjacent [440]. Finally, the preferential innervation of nerve cells. By providing a low-resistance reci- distal dendritic branches and apical tuft re- procal pathway for ions and small organic gions of pyramidal neurons by double bouquet molecules, such electrical connections permit cells (DB), horizontal cells of Cajal (HC) and the direct transmission of electrical signals be- Martinotti cells (M), allows these elements to tween neurons [107, 223, 693]. Gap junctions affect local dendritic integration. have been observed between smooth neurons Double bouquet cells are abundant in the in electronmicroscopic studies of the primate primate neocortex [146]. It has been frequently neocortex [688, 689]. Electrophysiological stud- observed that the dendritic spines postsynaptic ies on slices of rat neocortex established the to the double bouquet cells receive, in addi- existence of a network of electrically coupled tion, an asymmetrical synapse from a putative inhibitory interneurons [222]. These interneu- excitatory terminal [137]. In the human neo- rons, which were functionally characterized as cortex at least 47% of the spines postsynaptic fast-spiking cells, could be identified as PV-ex- to double bouquet cells were found to be in re- pressing basket cells [729]. Fukuda et al. [214] ceipt of such a dual innervation [146]. found that large PV-containing cells in layer II/ Dendritic spines are the principal targets of III of the cat visual cortex form about 60 gap the thalamocortical and cortico-cortical affer- junctions with other cells. Most, though not all ents and of the local axonal arborizations of of these junctions were between proximal den- pyramidal and spiny stellate cells, which are drite sites. Further studies revealed that other known to form asymmetrical, excitatory sy- types of neocortical inhibitory interneurons, napses [128, 137, 815]. However, because the designated as low-threshold-spiking [235], majority of dendritic spines of pyramidal neu- multipolar bursting [54] and late-spiking cells rons are outside layer IV, i.e. the principal area [99] also form type-specific, electrically coupl- of termination of thalamocortical afferents and ed networks. The same holds true for neuro- spiny stellate cell axons, it may be assumed gliaform cells [683]. However, the network that the main source of asymmetrical (excita- formed by the latter cells appeared to be not tory) synapses on spines with dual innervation strictly type-specific. In slices of rat somato- is from intracortical axons (local and cortico- sensory cortex, gap junctions connected ap- cortical). Hence, it is reasonable to assume that proximately 50% of neurogliaform cells, but double bouquet cells control the level of excita- 20% of these elements also formed gap junctions tion in circuits linking pyramidal neurons with with other interneurons, including fast-spiking other pyramidal neurons [146]. basket cells, regular-spiking non-pyramidal cells 15 Telencephalon: Neocortex 573

Fig. 15.38. Neocortical micronetwork. This micronetwork includes some specific afferent fibres from the thalamus (thc), a fibre representing a neuromodulatory system (mod), originating from an extrathalamic subcortical centre, a number of pyramidal neurons (P) and, finally, a set of GABAergic, inhibitory interneurons, represented by two basket cells (Ba). The thalamocortical fibres form excitatory synapses (e) with pyramidal and basket cells. The axons of the basket cells make inhibitory contacts with the somata of the pyramidal cells. When the basket cells are triggered by their thalamocortical afferents, they will exert a strong feed-forward inhibition on the pyramidal cells. The axons of the pyramidal neurons leave the cortex. These processes are provided with collateral branches, which establish excitatory synaptic contacts with other, related pyramidal neurons (not shown) and with the inhibitory interneurons (i.e. the basket cells) from which they receive input. The pyramidal axon collaterals and the basket cells form pathways for feed-back inhibition. The neuromodulatory system selectively addresses the set of inhibitory interneurons (Ba) with which they make some synaptic (s) and numerous non-synaptic (ns) contacts. This system may exert an excitatory or an inhibitory influence, depending on the neurotransmitter used by its afferent fibres and/or by the receptors expressed by their target neurons. The interneurons are interconnected by excitatory, reciprocal electrical synapses (el) as well as by inhibitory chemical synapses (ch). Moreover, the axons of the interneurons (Ba) emit collateral branches that form numerous autaptic contacts (au) with their own soma and proximal dendrites 574 Section II Structure of Spinal Cord and Brain Parts and bitufted cells. It is not known whether the of the present chapter, it was shown that several networks of inhibitory interneurons extend in- neurotransmitter systems, including those con- definitely across the neocortex or whether they taining acetylcholine, serotonin, dopamine and have distinct boundaries [107]. noradrenaline, ascend to the neocortex, and that The cerebral cortex displays synchronized, interneurons are a major target of these systems. rhythmic activity, and a variety of oscillations It was also demonstrated that, although a certain accompany states of sensory perception, motor proportion of the terminals of these systems performance, arousal and sleep [45, 82]. The make classical synapses, many others are not as- diverse networks of inhibitory interneurons sociated with a specialized postsynaptic struc- most probably operate as precision clockworks ture. The neurotransmitter molecules released for the entrainment of the various types of by these non-synaptic terminals diffuse over cortical oscillations. some distance through the extracellular space 2. Autaptic innervation of neocortical inhibi- to act on extrasynaptic receptors. This mode tory interneurons. Autapses are transmitter re- of intercellular communication – known as vol- lease sites made by the axon of a neuron on its ume transmission – enables the various extra- own soma or dendrites. Tamás et al. [728] thalamic neurotransmitter systems to exert a si- studied the synaptic relations of various types multaneous modulatory influence on large num- of neurons in the visual cortex of the cat dur- bers of interneurons. ing intracellular biocytin labelling and corre- Xiang et al. [835] studied the influence of lated light and electron microscopy. They acetylcholine on the excitability of two types found that the axons of two types of inhibitory of inhibitory interneurons in layer V of the rat interneurons, namely basket cells and den- visual cortex, namely fast-spiking cells and drite-targeting cells form significant numbers low-threshold spike cells. They found that ace- (10–20) of synaptic contacts with their own so- tylcholine elicits hyperpolarization in fast-spik- matodendritic surfaces. These autapses ap- ing cells through activation of muscarine re- peared to be domain-specific, i.e. those formed ceptors, resulting in disinhibition of their py- by basket cells concentrated on the perisomatic ramidal cell targets. The low-threshold spike region, whereas those formed by dendrite-tar- cells, on the other hand, appeared to be excited geting cells were located on more distal den- by acetylcholine through the activation of nico- drites. Bacci et al. [23] recorded autaptic activ- tine receptors. Bacci et al. [25] cited evidence ity in neocortical fast-spiking GABAergic inter- suggesting that the dopaminergic, noradrener- neurons. It appeared that in these neurons the gic and serotoninergic systems exert similar autaptic activity has significant inhibitory ef- selective and differential modulations on par- fects on repetitive firing and increased the cur- ticular subgroups of inhibitory interneurons, rent threshold for evoking action potentials. thus providing a substrate for fine control of As noted above, during cortical oscillations, information flow through cortical networks. fast-spiking elements (as well as other types of Yoshimura and Callaway [841, 842] recently inhibitory neurons) probably synchronize the provided some further examples of “fine-scale activity of groups of pyramidal neurons [45, specificity” in the organization of cortical net- 82]. It is believed that inhibitory autaptic works. Studying the mutual functional rela- transmission enables fast-spiking cells “to tionships between pyramidal neurons and sense their own firing and regulate it in phase those between pyramidal neuron and fast-spik- with that of other fast-spiking cells, resulting ing interneurons in layer II/III of the rat visual in synchronous inhibitory neurotransmission cortex, they found that (1) layer II/III pyrami- onto pyramidal neurons” [25]. dal neurons only share common input from 3. Several extrathalamic neurotransmitter sys- layer IV and from within layer II/III in the tems ascend to the neocortex, where they specifi- minority of cases in which they are directly in- cally or preferentially target particular types of terconnected to each other; (2) fast-spiking in- inhibitory interneurons. In a previous section hibitory interneurons connect preferentially to 15 Telencephalon: Neocortex 575 neighbouring pyramidal cells that provide Finally, each of the various cohorts of inhibi- them (via their axon collaterals) with recurrent tory interneurons impinging on a particular py- excitation; and (3) these pairs of reciprocally ramidal network is specifically addressed by one connected fast-spiking interneurons and pyra- or more of the extrathalamic modulatory sys- midal cells share common specific excitatory tems. The influence exerted by these systems de- input. Yoshimura and Callaway [841] consider- pends on the nature of the neurotransmitter ed it likely that these sets of reciprocally con- used by their afferent fibres and on the receptor nected neurons contribute to the synchroniza- types expressed by their target neurons. tion of activity within particular neuronal sub- It is important to note that not only the in- populations. hibitory input, but also the excitatory input to Each neocortical area contains at least sev- pyramidal neurons belonging to the same net- eral types of pyramidal neurons. In a previous work may be specific. Yoshimura et al. [842] section of this chapter, it was shown that these studied the connections to pyramidal neurons types of pyramidal neurons may differ from in layer II/III of the visual cortex of the rat. each other with regard to (1) the laminar posi- They found that reciprocally connected pyra- tion of their perikarya; (2) the laminar spread midal neurons share common excitatory input of their dendritic branches (Fig. 15.32); (3) the from layer IV and within layer II/III. However, laminar density of their dendritic spines, and, adjacent layer II/III neurons that were not con- hence, their laminar afference (Fig. 15.32); (4) nected to each other appeared to share very lit- the destination of their axons (Fig. 15.33); and tle (if any) common excitatory input from (5) the spread and extent of their cortical axon layers IV and II/III. collaterals (Figs. 15.34, 15.35). These data indi- The question arises as to how strict the sep- cate that the various types of pyramidal cells aration is between the various pyramidal net- are differently embedded in the microcircuitry works, including their satellite cohorts of inhi- of the neocortex. Given the fact that pyramidal bitory interneurons? The answer to this ques- neurons of the same type are strongly and re- tion is largely unknown. However, it seems ciprocally connected by axon collaterals [806, likely that the abundant double bouquet cells 841], it may be concluded that each particular with their vertically oriented axonal systems neocortical area contains a number of net- contact pyramidal neurons belonging to differ- works of interconnected, type-specific pyrami- ent networks, and we have seen that the neuro- dal neurons. The number of pyramidal net- gliaform cells form gap junctions with several works present within a given neocortical area other types of inhibitory interneurons. is unknown and the same holds true for the actual extent of these networks. In light of the recent literature reviewed above, it seems likely that the various pyrami- Neocortical Columns and Modules dal elements belonging to a particular network are in receipt of afferents from cohorts of inhibitory interneurons, each cohort contacting a Introduction specific domain of the receptive surface of the pyramidal neurons involved. The inhibitory in- During the last 50 years, physiological and terneurons forming these cohorts are all of the morphological evidence has accumulated indi- same type and are generally reciprocally con- cating that several parts of the neocortex of nected by inhibitory chemical synapses and by many different mammalian species are com- excitatory electrical synapses [729]. posed of radially oriented, column-like units or Thalamic inputs selectively contact and modules. The concept that column-like mod- strongly excite the interneurons belonging to ules represent fundamental units of the mam- particular cohorts, while others receive weaker malian neocortex has gained wide acceptance or no thalamic inputs [235]. in the literature. In the following, a survey will 576 Section II Structure of Spinal Cord and Brain Parts be presented of the findings and considerations portion of the somatosensory cortex to which that led to this concept. This survey will be fol- the large mystacial vibrissae on the snout pro- lowed by a brief critical and cautionary com- ject via relays in the main trigeminal nucleus mentary. and in the thalamus. It appeared that each “glomérulo” or “barrel”, as they came to be called, is the primary cortical representation of The Investigations of Lorente de Nó: one vibrissal follicle. The vibrissae are ar- Elementary Units and Glomérulos ranged in a stereotypical grid-like pattern in which each vibrissa has a unique position. The first to propose a modular structure of the Within the somatosensory cortex, each barrel neocortex was Lorente de Nó [416]. He also occupies a unique position and the topo- claimed that, in small radially oriented cylin- graphical arrangement of these structures ders having a specific thalamocortical fibre as closely corresponds to that of the vibrissae. their axis, all elements of the cortex are repre- The largest barrels in mice are elliptical in sented and that in these elementary units the cross-section, with a greatest diameter of whole process of transmission of impulses 300 lm (Woolsey and van der Loos [834]; van from the afferent fibre to the efferent axon may der Loos and Woolsey [759]). According to van theoretically be accomplished. Lorente de Nó’s der Loos [758], barrels are the visible lamina concept was based on the observations that (a) IV counterparts of cortical columns. the terminal branches of thalamocortical fibres Staiger et al. [704] recently studied the spiny may form discrete cylindrical patches in lami- neurons in layer IV of the somatosensory cor- na IV of the cortex, (b) the efferent system of tex of the rat. It appeared that this layer, in ad- the cortex is formed by radially oriented axons dition to numerous spiny stellate cells (58%), of pyramidal neurons and (c) the axons of also contains star pyramidal cells (25%) as well many cortical local circuit neurons are likewise as pyramidal elements (18%). As regards the radially oriented. spread of the axonal ramifications of these cell The results of earlier investigations of Lor- types, the authors reported that the axonal ar- ente de Nó have played an additional role in bors of the spiny stellate cells are largely con- the development of the ideas concerning co- fined to their related columns, that the pyra- lumnar organization of the neocortex. In 1922 mids show extensive transcolumnar axonal Lorente de Nó [414] described discrete cylin- branches and that the star pyramids occupy an droid aggregations of cells in lamina IV of intermediate position (Fig. 15.39 D). It is note- what he believed to be the auditory cortex of worthy that the columns denoted by the au- the mouse, which he termed “glomérulos” (Fig. thors have no structural boundaries above and 15.39 A). Golgi material revealed that the spe- below layer IV. cific thalamic afferents to this cortex constitute dense patches of terminal ramifications which coincide with the glomérulos (Fig. 15.39 B) and The Columnar Organization that the dendrites of the neurons in lamina IV of the Somatosensory Cortex are largely confined to the glomérulo in which their soma is located (Fig. 15.39 C). The glo- Mountcastle [494–496] analysed the functional mérulos appeared to contain numerous spi- organization of the somatosensory cortex in nous stellate cells. The axons of these local cir- the cat, using microelectrodes to record the ac- cuit neurons were found to descend to the tivity of single cells. During radial penetration, deeper cortical layers, where they issue numer- he encountered cells with similar receptive ous collaterals. Some of these collaterals were fields throughout the depth of the cortex that observed to ascend to lamina III. responded to stimulation of the same type of Some 50 years later, it became clear that cutaneous receptors located at a particular site. Lorente de Nó [414] had actually analysed the Penetrations parallel to the pial surface and 15 Telencephalon: Neocortex 577

Fig. 15.39 A–D. The somatosensory cortex of rodents. A Cytoarchitecture; the fourth layer contains specialized structures consisting of a cylinder of densely packed cells surrounding a region of lower cell density. These structures were designated as glomérulos by Lorente de Nó [414] and as barrels by Woolsey and van der Loos [834]. B Specific somatosensory efferents (b) from the thalamus form patches of a dense terminal feltwork, which correspond to the glomérulos. These patches are separated by zones (Z), in which the axonal feltwork is clearly less dense. The fibres of the feltwork surround the neuronal somata in the fourth cortical layer. C Two glomérulos as observed in Golgi preparations. D The glomérulos or barrels contain three different types of spiny neurons: spiny stellates, star pyramids and pyramids. These cell types differ with regard to the spread of their axonal arbors. A, B and C are reproduced from Lorente de Nó [414]; D is based on Staiger et al. [704] 578 Section II Structure of Spinal Cord and Brain Parts crossing the radial axis of the cortex appeared relative changes in metabolic activity following to pass through blocks of tissue 300–500 lmin peripheral stimulation. size, and neurons with identical properties Microelectrode recordings also revealed the were encountered in each of them. Sharp tran- presence of parallel stripes, about 500 lm wide, sitions were observed from a block with one alternately receiving their input from the ipsilat- set of properties to the adjacent block with dif- eral and the contralateral eye. The morphologi- ferent properties. These findings led Mountcas- cal correlate of these ocular dominance columns tle ([494], p. 430) to hypothesize that “there is could be strikingly visualized using silver im- an elementary unit of organization in the so- pregnation techniques and autoradiographically matic cortex made up of a vertical group of by means of transneural transport of radiola- cells extending through all the cellular layers.” belled amino acids injected into one eye. He termed this unit a “column”. On the basis In the monkey striate cortex, staining for of experiments in which multiple, closely the mitochondrial enzyme cytochrome oxidase spaced penetrations were made, Mountcastle revealed an array of densely staining, peg-like concluded that the individual columns have a structures in laminae II and III, which can be width of maximally 500 lm. He believed that seen lying in register with less densely staining functionally active columns are able to isolate formations in the cortical layers below lamina themselves from their surroundings by exert- IV. These blobs, as they were termed, are ellip- ing an inhibitory action on neurons in their soid in tangential sections, measuring roughly inactive neighbours, a process which he desig- 150 ´200 lm. Their most conspicuous superfi- nated as “pericolumnar inhibition”. cial (i.e. supragranular) parts receive a separate projection from the intercalated layers of the lateral geniculate body (Fig. 15.26). Physiologi- The Columnar Organization cal studies have shown that the cells within the of the Visual Cortex blobs respond selectively to the colour of a stimulus, without regard to its orientation. The detailed physiological and morphological A hypercolumn is considered to represent studies performed by Hubel and Wiesel and the circuitry necessary for the analysis of a collaborators [308, 310–314, 404, 412] on the given, discrete region in the visual field. It is primary visual cortex of the monkey have led conceived to contain a set of nine orientation to the identification of three different kinds of columns, together encompassing a complete column-like structure in that area: orientation cycle of orientation through 1808, an adjoining columns, ocular dominance columns and pair of right and left ocular dominance col- blobs. Sets of these three column-like struc- umns and several blobs. tures are conjectured to be united in larger en- In a previous section, it has been mentioned tities, termed hypercolumns. that the axons of many neocortical pyramidal The orientation columns were detected by neurons give rise to long, horizontally running electrophysiological recordings. During radial collaterals, which issue clusters of terminal penetrations with microelectrodes, cells having branches at regular intervals (Fig. 15.35). Inter- identical axes of orientation (i.e. cells that re- estingly, experimental studies using different sponded strongest to a bar of light in one par- techniques, i.e. retrograde tracing, 2-deoxyglu- ticular orientation) were encountered through- cose autoradiography and cross-correlation out the thickness of the cortex. During tangen- analysis, have produced evidence suggesting tial penetrations, the electrode encountered a that, in the primary visual cortex, these hori- shift in the axis of orientation of the light bar zontal collaterals mediate communication be- of about 108 every 300–100 lm. The morpho- tween functionally related columns, e.g. blobs logical substrate of the functional columns of similar colour opponency and columns of thus detected could be visualized with the aid similar orientation preference (for a review, see of the 2-deoxyglucose technique of mapping Gilbert [237]). 15 Telencephalon: Neocortex 579

The Auditory Cortex Columnar Patterns Shown by the Cells of Origin and the Terminal Ramifications Evidence of a columnar organization of the of Cortico-cortical Connections primary auditory cortex has also been presented. During microelectrode penetrations Anatomical evidence for the existence of verti- normal to the cortical surface, cells with the cal columns or bands in the cerebral cortex same characteristic frequency were encoun- has been provided by studies on the origin tered throughout the entire depth of the cortex and termination of cortico-cortical connec- [1]. Binaural interaction bands have also been tions. Thus, Jones et al. [347], using an antero- described, exhibiting either binaural summa- grade tracer, demonstrated that both commis- tion or inhibition. Experiments in which sin- sural and ipsilateral cortico-cortical fibres aris- gle-unit mapping was combined with autora- ing and terminating in the somatic sensory diographic tract tracing revealed that, in the cortex of monkeys terminate in distinct verti- primary auditory cortex of the cat, clusters of cally oriented columns with a variable width of summation responses coincide with bands re- 500–800 lm. Each of these columns appeared ceiving a heavy innervation from the contralat- to be separated from its neighbours by a gap eral primary auditory cortex, whereas suppres- of approximately 500 lm. Each column was sion responses were recorded in regions of considered to represent the terminals of a bun- sparse contralateral innervation [75]. dle of cortico-cortical axons emanating from cells at the centre of each of the loci in which the tracer was injected. Using a retrograde The Motor Cortex tracer, Jones et al. [347] also demonstrated that the cells of origin of the cortico-cortical pro- Within the motor area of the neocortex, there jections investigated, found predominantly in are small loci where stimulation with weak lamina III, are also aggregated in vertically ori- currents produces movements executed by a ented clusters (Fig. 15.40). Goldman and Nauta single muscle. It has been demonstrated that [242] injected an anterograde tracer into three the loci for producing contraction of a given cytoarchitectonically distinct regions (areas 4, muscle extend perpendicularly from the sur- 9 and 12) of the frontal lobe of rhesus mon- face to the depth of the cortex. These radial ar- keys. Neurons in these various regions are rays of stimulation points have been called cor- known to project both contralaterally and ipsi- tical motor columns. The diameter of these col- laterally to cytoarchitectonically diverse areas umns is approximately 1 mm in cross-section within the frontal, temporal and parietal lobes. [16]. Using two penetrating microelectrodes, They found that all of these cortico-cortical one for stimulation and the other for record- projections terminate in vertically oriented col- ing, it was demonstrated that stimulation of a umns, 300–700 lm wide, which alternate in given column produces a pericolumnar zone of regular sequence with zones of comparable inhibition [18]. Whereas the afferent input ar- width that are free of such terminals. Interest- riving at the somatosensory columns is modal- ingly, the pattern and dimensions of cortical ity specific, the cortical motor columns receive columns in the prefrontal cortex, defined by polymodal input from sensors in the target cortical afferent inputs, appeared to be strik- muscle itself, from the joint to which the mus- ingly similar in the rat, squirrel monkey and cle inserts and from the skin overlaying the rhesus monkey, although the brains of these muscle, i.e. sensors that are stimulated when animals differ considerably in size [76, 325]. In the target muscle contracts [17]. order to determine the pattern of termination of two converging cortico-cortical systems in the same animal, Goldman-Rakic and Schwartz [246] followed a double anterograde labelling strategy. Using rhesus monkeys, they im- 580 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.40. The general arrangement of cortico-cortical fibres, shown diagrammatically in a lisencephalic brain. The fibres interconnect radial columns with a diameter of 200–300 lm. Ipsilateral connections originate mainly from pyramidal neurons in layer III (cells shown at left in outlines), while commissural connections (shown in solid black) derive from layers III, V and VI. Reproduced from Szentágothai [721]. TH, thalamus 15 Telencephalon: Neocortex 581 planted HRP pellets in area 7 of the parietal HRP and tritiated amino acids were made in lobe in one hemisphere and injected a mixture two topographically matched regions of the of tritiated amino acids in area 9 of the frontal frontal lobe in each hemisphere. The cortical lobe of the other hemisphere. It appeared that, projections from both of these homotopic pairs in the prefrontal cortex, contralateral callosal of areas appeared to converge in common co- fibre columns interdigitate with ipsilateral as- lumnar territories. sociational fibre columns. Goldman-Rakic and Schwartz [246] also investigated the spatial organization of the populations of cells in the Minicolumns and the Radial Unit Hypothesis prefrontal cortex projecting to the parietal as- of Cortical Development sociational cortex and to the contralateral prefrontal cortex using HRP or fluorescent dyes as In 1979, Mountcastle [495] introduced what he retrograde tracers. They demonstrated that believed to be “the basic modular unit of the these two populations are inversely related in neocortex” under the name minicolumn.He their relative densities over portions of the characterized this unit as a narrow chain of prefrontal cortex examined. Moreover, in the neurons extending radially across the cellular HRP material, the high-density patches of ret- layers II–VI. Referring to the ontogenetic stud- rogradely labelled neurons appeared to coin- ies of Rakic (see below), he indicated that the cide with the afferent fibre columns revealed minicolumn is produced by the iterative divi- by anterograde transport. In a later study (Se- sion of a small cluster of progenitor cells, and lemon and Goldman-Rakic [669]), terminal la- that the resultant neuroblasts migrate to their belling originating from prefrontal and parietal destination in the cortex along radial glial injections in the same hemisphere was investi- cells. According to Mountcastle, a minicolumn gated in a large number of areas of conver- occupies a radial cylinder of cortical space gence in both the ipsilateral and the contralat- with a diameter of about 30 lm. On the basis eral hemisphere. In most of the target areas, of cell counts in the neocortex of various prefrontal and parietal terminal fields formed mammals, carried out by Rockel et al. [619], an array of interdigitating columns, but in Mountcastle estimated that each minicolumn some other areas prefrontal and parietal pro- contains about 110 neurons, except for the stri- jections converged on the same column or ate cortex where the number is about 2.5 times cluster of adjacent columns but terminated larger. He posited that minicolumns contain all within different laminae. For instance, in the the major cortical neural cell phenotypes, in- depths of the superior temporal sulcus, pre- terconnected in the vertical dimension [496]. frontal terminals occupied laminae I, III and V, According to Mountcastle, the cortical col- with parietal terminals filling complementary umns, which he sometimes also designated as laminae IV and VI. macrocolumns [495], are to be considered as From the experiments of Goldman-Rakic aggregations of several hundred minicolumns and Schwartz [246] and Selemon and Gold- bound together by short-range horizontal con- man-Rakic [669], it appears that projections nections [496]. from two different (heterotopic) cortical re- The studies of Rakic [606–610], on the onto- gions (in the same or opposite hemispheres) geny of the neocortex in the rhesus monkey, remain segregated within common cortical tar- which strongly influenced the development of gets, either by terminating in separate alternat- Mountcastle’s minicolumn concept, may be ing columns or by concentrating in comple- summarized as follows: The matrix zone of the mentary laminae within the same column. The embryonic pallium is divided by glial septa relationship between the termination zones of into columns of precursor or stem cells termed projections from paired homotopic areas was “proliferative units”. The postmitotic cells pro- studied by McGuire et al. [457] in the cerebral duced by these proliferative units find their cortex of the rhesus monkey. Injections with way to the primordial cortex by following the 582 Section II Structure of Spinal Cord and Brain Parts shafts of elongated, radially oriented glial cells, necessarily the size of the simplest anatomical which stretch across the embryonic pallial wall. unit. In fact, their findings imply that, regard- Eventually, all postmitotic cells generated in a less of their areal size, modules will contain single proliferative unit form a morphological- constant cell numbers (Swindale [719]). ly identifiable stack of neurons in the cortex. These entities, which may be designated as “ontogenetic” or “embryonic” columns, form, Dendritic Clusters, Axonal Bundles and according to Rakic, the fundamental building Radial Cell Cords As (Possible) Constituents blocks in the developing neocortex [608]. Each of Neocortical Minicolumns proliferative unit is supposed to produce multiple neuronal phenotypes. Rakic [608] advanced A number of examples of repeating microar- the idea that the proliferative units in the ma- rays of cortical elements could be interpreted trix zone of the embryonic pallium constitute as conforming to a minicolumnar radial pat- a proto-map of prospective cytoarchitectonic tern. These include the clusters of apical den- areas. Each of these areas is composed of a drites of pyramidal neurons, the bundles of large number of ontogenetic columns, which myelinated axons and the column-like cell become the basic processing units in the cere- soma arrays running orthogonal to the hori- bral cortex. According to Rakic [608], the rela- zontal laminae. tively constant size of these units in different The existence of clusters or bundles of den- mammalian species suggests that, during evo- drites in the neocortex was discovered inde- lution, the cortex has expanded by the addi- pendently and almost synchronously by Peters tion of such radial units rather than by their and Walsh [568] in rat somatosensory cortex enlargement. This postulate was designated as and by Fleischhauer et al. [196] in the rabbit the “radial unit hypothesis” [608, 610]. and cat sensorimotor cortex. Since then, simi- Mountcastle’s estimate that a minicolumn lar bundles of dendrites have been revealed in contains about 110 cells was, as he indicates, a number of other neocortical areas in mice, based on the work of Rockel et al. [619]. In rats, rabbits and cats [140, 194, 195, 569, 826]. this publication and in a later, more extensive In each of these areas there appeared to be a study [620], the results of a quantitative com- similar basic organization: the apical dendrites parative analysis of the neocortex were re- of groups of layer V pyramidal neurons form ported. The number of neuronal cell bodies clusters, which during their ascent are joined were counted in a narrow strip, 30 lm wide, by the apical dendrites of pyramidal cells situ- through the depth of the neocortex in several ated in the more superficial layers. different functional areas (somatic sensory, vis- Radially oriented bundles of myelinated fi- ual, motor, frontal, parietal and temporal) in a bres occur in all parts of the neocortex (Figs. number of different species (mouse, rat, cat, 15.4 C, 15.13). These bundles are known as ra- monkey and human). These counts gave re- dii or as the radiations of Meynert [85]. As al- markably constant values of 110 ±10 cells in all ready mentioned in a previous section of this areas and in all species studied. The only ex- chapter, local differences in the size and pe- ception to this similarity appeared to be the bi- ripheral extent of these bundles play a promi- nocular part of the visual cortex in a number nent role in the myeloarchitectonic parcellation of primate brains, in which approximately 2.5 of the cortex. times as many neurons were found. Rockel et Column-like arrays of cell bodies can be ob- al. [620] remarked that they chose a width of served in many parts of the neocortex. They 30 lm for counting the cortical perikarya be- are particularly conspicuous in the human foe- cause, according to Hubel and Wiesel [312] tal cortex and in the mature temporal cortex of and others, this is approximately the width of humans and other primates [341]. Brodmann the simplest functional column in the neocor- [70], who beautifully depicted these cell arrays tex. However, they pointed out that this is not (Fig. 15.41), indicated that there is a correla- 15 Telencephalon: Neocortex 583

Fig. 15.41. Radial arrays of cell bodies in the primary motor area 4 (A) and the visual areas 17 and 18 (B)of 8-month-old human fetuses. Reproduced from Brodmann [70]. G, location of line of Gennari 584 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.42 A,B. Arrangement of pyramidal neurons and double bouquet cells in 23-lm-wide radial modules of the striate cortex of the macaque monkey. Neuronal somata and dendrites are shown in red,axonsin black. A The arrangement of the apical dendrites of pyramidal cells in a module; for clarity, only one half of the neurons present are shown. The somata of spiny stellate cells are represented by open circles, while those of inhibitory, GABAergic neurons are represented by circles filled with small dots. The total number of spiny (excitatory) and GABAergic (inhibitory) neurons within each layer are given on the right. B A representation of three modules to show the arrangement of the dendrites and axons of the pyramidal cells. Each module contains one double bouquet cell (DBC). The axonal system of the double bouquet cell runs alongside the clustered apical dendrites of pyramidal neurons in layers IV and II/III and then alongside the bundle of myelinated axons. Modified from Peters and Sethares [566, 567] 15 Telencephalon: Neocortex 585 tion between them and the bundles of myeli- are regularly arranged, and their average cen- nated fibres, and several later authors, among tre-to-centre spacing is similar to that of the them Fleischhauer [195] and Buxhoeveden and dendritic clusters. In the deeper layers of the Casanova [81], arrived at a similar conclusion. cortex, each myelinated axon bundle can be However, von Economo and Koskinas [796], paired with a nearby cluster of apical dendrites who thoroughly investigated this relation, stat- (Fig. 15.42 B) [566]. Peters and Sethares envi- ed that in certain cortical regions, in which the sioned that the pyramidal cell modules repre- neuronal perikarya were clearly arranged in sent the basic functional units in the cortex, vertical columns, a corresponding pattern of and that in essence they are equivalent to radial fibre bundles was absent. It is important Mountcastle’s minicolumns. to note that neither in the foetal nor in the Column-like domains of spontaneously co- adult stage do the arrays of cell bodies extend active pyramid-like neurons have been ob- uninterruptedly through all cortical layers. served in slices of neonatal rat cortex [844]. The phenomenon of apical dendrite bund- The neurons within each domain appeared to ling led Peters and Sethares [564–566] to the be coupled by gap junctions. It has been sug- concept that the neocortex is composed of py- gested [341] that these domains are compar- ramidal cell modules. In accordance with ear- able to the pyramidal cell modules. However, lier studies on rodents and cats, these authors their diameter (50–120 lm) is much larger found that the apical dendrites of the pyrami- than that of the modules, whereas their num- dal cells in layer V of the primary visual cortex ber of constituent cells (35 ±17) is smaller. of the rhesus monkey cluster together in dis- The investigations of Vercelli et al. [776] tinct and regularly spaced groups as they as- have shown that pyramidal cells that share an cend through the various sublayers of lamina apical dendritic bundle may project to specific IV and pass into superficial layers II/III. The targets. Studying the rat visual cortex, they apical dendrites of the pyramidal neurons in found that neurons projecting to the ipsilateral those layers are then added to the clusters of or contralateral cortex form bundles together layer V apical dendrites so that the clusters of and with neurons projecting to the striatum, apical dendrites gradually increase in size as but not with those projecting to the lateral they ascend. Upon reaching layer I, the apical geniculate body or the superior colliculus or dendrites split up into their terminal tufts (Fig. through the cerebral peduncle. Bundles com- 15.42 A). When fully formed, the clusters composed of the apical dendrites of neurons pro- prise the dendrites of some 50 pyramidal cells. jecting to the striatum and through the cere- The apical dendrites of the pyramidal neurons bral peduncle were also observed. The layer VI in layer VI do not add to the clusters, but neurons projecting to the lateral geniculate rather ascend independently to form their ter- body formed separate dendritic bundles. Ver- minal rufts in layer IV (Fig. 15.42 A). celli et al. [776] concluded that the neocortex On the basis of these findings, Peters and is organized in minicolumns of output neu- Sethares [566] proposed that the primary vis- rons. They considered it likely that the neurons ual cortex is composed of pyramidal cell mod- in an output minicolumn are closely related. ules. This concept is shown diagrammatically Before leaving the pyramidal cell modules, it in Fig. 15.42 B, where three such modules are may be mentioned that Peters and Sethares depicted. The average diameters of the mod- [567] related one type of interneuron, namely ules are equal to the mean centre-to-centre the double bouquet cells, to these entities. spacing of the clusters of apical dendrites, Double bouquet cells are abundant through- namely about 23 lm. out the primate neocortex [140, 146]. The cell The primary visual cortex of the rhesus bodies of these elements are situated in layer monkey also contains prominent vertical bun- II and upper layer III. Their axons form tightly dles of myelinated fibres. These bundles, which packed bundles of collaterals that descend to are largely composed of pyramidal cell axons, layer V, where they disperse in horizontal 586 Section II Structure of Spinal Cord and Brain Parts branches. These vertically oriented double and Casanova equate these radial arrays with bouquet axonal systems or horse tails are Mountcastle’s minicolumns. equally spaced [140, 146]. Peters and Sethares [567] found that the average centre-to-centre spacing of the horse tails in the primary visual Microcircuitry of Neocortical Columns cortex of the rhesus monkey is about 23 lm, a value similar to that of the centre-to-centre Two noted neuroscientists, Szentágothai and spacing of the clusters of apical dendrites Eccles, attempted to present the microcircuitry forming the axes of the pyramidal cell modules of the neocortical columns. Szentágothai’s and that of myelinated axon bundles that con- [721–723] opinion was that “the basic unit of tain the efferent axons of these neurons. It ap- cortical architecture – the true cortical col- peared that each pyramidal cell module is as- umns – is anatomically based not on the mode sociated with one horse tail axon bundle (Fig. of termination of the specific sensory afferents, 15.42 B). Within the superficial layers II/III, the as one might logically assume, but on that of double bouquet cell axons run alongside the cortico-cortical afferents”. The size of these co- apical dendritic clusters, while in layer IVC lumnar units of cortico-cortical afferent termi- they are closely associated with the bundles of nations, according to Szentágothai, is remark- myelinated axons. It is important to note that ably constant, measuring around 200–300 lm the double bouquet axonal branches do not sy- in diameter (Fig. 15.40). On the basis of light napse with the apical dendrites but rather with and electron microscopy data, Szentágothai in- basal dendrites and with side branches of the dicated how the various types of presumed ex- apical dendrites of pyramidal neurons and citatory and inhibitory interneurons are con- with postsynaptic structures belonging to local nected with the cortical afferent system and circuit neurons [1, 136]. The branches of apical with the cortical output neurons, i.e. the pyra- and basal dendrites of pyramidal neurons ex- midal cells. He emphasized that the internal tend horizontally for hundred of micra; hence, connectivity of each columnar unit is predomi- it is clear that the influence exerted by the nantly vertical, in a way that would favour a double bouquet cells does not remain confined subdivision of each column into narrower mi- to narrow cylindroid modules, but rather is crocolumns, and that the space of many of dispersed over pyramidal neurons with apical these microcolumns is occupied by the dend- dendrites located in many different apical den- ritic and/or axonal ramifications of individual dritic bundles [341]. interneurons (Fig. 15.43). Eccles’ [166] sketch The third major radial configurations in the of the neuronal structure of the cortical mod- neocortex, i.e. the arrays of neuronal perikarya ule conforms to that of Szentágothai and is as observed in Nissl preparations, deserve little largely based on the work of the latter. comment. On the basis of a study of the development of the human auditory cortex, Krmpo- tic-Nemanic [387] suggested that there is a di- Neocortical Columns and Modules: rect continuum between the ontogenetic radial A Critical Commentary columns as described by Rakic and adult cell columns. These cellular columns are the focus In the preceding pages, the evidence leading to of two recent reviews by Buxhoeveden and Ca- the concept that the mammalian neocortex is sanova [80, 81]. The authors indicate that the composed of repetitive radially oriented colum- visibility of these structures in Nissl prepara- nar units has been reviewed. Although this tions depends on the linear arrangement of py- concept has dominated the neurobiological lit- ramidal cells and the existence of cell-poor, erature on the cortex for the last 50 years, it is neuropil zones surrounding them. In layers II, by no means generally accepted. Numerous IV and VI of the adult neocortex, the radial ar- authors, including Towe [738], Valverde [756], rays are usually one cell wide. Buxhoeveden Swindale [719], Purves et al. [597], Jones [341] 15 Telencephalon: Neocortex 587

Fig. 15.43. Some features of the organization of a neocortical module or column, “defined” by a cortico-cortical (either association or callosal) afferent, ascending in its centre, as drawn by Szentágothai [721]. The module consists of a tall cylinder, 300 lm in diameter, extending through the full thickness of the cortex. The flat cylinder of the same diameter corresponds to the termination space of a specific thalamocortical afferent. This afferent fibre forms synaptic contacts with two different types of spiny stellate cells, SS1 and SS2,aswell as with a neurogliaform cell: ngf; SS1 has both an ascending and a descending arborization; SS2 has only an ascending arborization, whereas ngf has only a descending arborization. The ascending arborization of SS1 synapses with a typical double bouquet cell (db), whose bifurcating axon forms a long radial slender arborization. It should be mentioned that Szentágothai considered all of the neuronal elements depicted to be excitatory in nature. However, it is now known that neurogliaform cells and double bouquet cells are GABAergic, inhibitory interneurons. Reproduced from Szentágothai [721] 588 Section II Structure of Spinal Cord and Brain Parts and Horton and Adams [305], have challenged clude all cortical layers. He designated these the validity of this concept. Some of the main units as “elementary units of cortical opera- objections may be summarized as follows: tion”, because he believed that within them 1. The original concept of Rakic, according “the whole process of the transmission of im- to which all neurons forming a radial unit, pulses from the afferent fibre to the efferent both pyramidal and non-pyramidal, stem from axon may be accomplished”. Lorente de Nó a small cluster of progenitor cells in the pallial did not specify the pattern of connectivity neuroepilium, is untenable. As has been dis- within these units, but during the last decades, cussed in Chap. 2, many, if not most, of the several attempts have been made to character- neocortical non-pyramidal cells are produced ize the basic neocortical microcircuit (see Fig. in the subpallial matrix and migrate tangen- 15.44) [136, 157, 679]. However, in none of tially into the developing pallium (Fig. 2.25). these attempts has the question been specifical- 2. The thesis that columnar structures form ly addressed as to whether the elements in- the basic structural and functional units of the volved in these microcircuits are confined to neocortex has so far not been substantiated by small, cylindrically shaped spaces. The pyrami- direct evidence concerning the specific neuro- dal cell modules described by Peters and nal composition and the specific mode of op- Sethares (Fig. 15.42) [565–567] could be the eration of these units. The statement that the starting point for this type of research. Such a columns contain all cell phenotypes [76, 474] module would comply with Lorente de Nó’s has not been proven and is even unprovable elementary unit concept if it could be shown because an indeterminate number of cortical that: (a) all axons of the constituent pyramidal neurons have not been characterized in detail neurons pass to the same target(s); (b) all con- as yet. Moreover, recent histochemical and phys- stituent pyramid neurons are strongly and re- iological studies have shown that the diversity of ciprocally connected by axon collaterals; (c) all neocortical interneuron types is much larger sets of interneurons impinging on the various than was previously thought [136, 507]. somatodendritic compartments of the pyrami- It is important to note that the models pre- dal cells are situated within the confines of the sented by Szentágothai [721, 723] and Eccles module; and (d) all of its afferents terminate [166] are not based on analyses of the local within the 23-lm-wide cylindrical space of the connectivity within particular morphological module. As regards (c), we have seen that the units; rather, findings on cortical neurons and their connections derived from different cortical areas in different species were assembled Fig. 15.44 A–C. Neocortical microcircuits; excitatory and placed within the spatial framework of a elements in black, inhibitory elements in red. A Ba- cylinder with a diameter of 300 lm. Hence, sic neocortical circuit for sensory areas, modified these models should be considered as theoreti- from Shepherd [679]. The diagram includes elements representing superficial and deep pyramidal cal constructs, showing at best how the local cells (P1, P2), spiny stellate cells (SS) and inhibitory circuitry within columns of cortical tissue, de- interneurons (I1, I2). Afferent inputs (AFFs) excite fined by streams of cortico-cortical afferents, pyramidal cells (1, 2, 5, 6) as well as inhibitory cells could be organized. Mountcastle’s [495] defini- (1, 4). In addition, there is feed-forward excitation tion, stating that “a cortical column is a com- (AFF3) through spiny stellate cells. The axons of the P cells have collaterals (ac) that recurrently and lat- plex processing and distributing unit that links erally excite the P cells (ac1,2) as well as inhibitory a number of inputs to several outputs”, does cells (ac3) that provide for feed-back and lateral in- not contain any information concerning the hibition. B Canonical neocortical circuit, based on specific functional design of these columns. the studies of Douglas et al. [157], as represented by 3. According to the original concept of Lor- Shepherd [679]. P1 and P2 represent the superficial and deep populations of (excitatory) pyramidal ente de Nó [416], the neocortex is built up of cells. Element I represents the population of inhibi- cylindrically shaped units, composed of verti- tory, GABAergic neurons. Neurons within each pop- cal arrays of interconnected neurons that in- ulation form connections with other members of 15 Telencephalon: Neocortex 589

that population. All three populations receive direct activation from the thalamus, but because the thalamic input provides only about 10% of the excitatory input, 90% of the excitation stems from intracortical connections between P cells. The excitation of nearby P cells is provided by axon collaterals of these elements. These axon collaterals also excite the interneuron-pool to provide feed-back and lateral inhibition. Modified from Shepherd [679]. C Skeleton of the basic neocortical microcircuit, according to DeFelipe [136]. The diagram shows the synaptic inputs and main patterns of local connections of the pyramidal cell. Excitatory inputs to this element arrive exclusively to its dendritic arbor and originate from extrinsic afferent systems and spiny cells (which include other pyramidal neurons and spiny stellate cells). Inhibitory inputs, which originate from GABAergic interneurons, terminate on the proximal dendrites, soma and axon initial segment. These interneurons are interconnected between one another, with the exception of the axo-axonic chandelier cells. Reproduced from DeFelipe [136] 590 Section II Structure of Spinal Cord and Brain Parts double bouquet cells conform to the modules 5. Although column-like entities of a given (Fig. 15.42 B) but that their axons most prob- type are readily apparent in the neocortex of ably do not exclusively contact the dendrites some species, they are often not detectable in belonging to the pyramidal cells situated with- other, sometimes closely related animals. The in “their own” module. As regards (d), the cor- following examples of such inconsistencies are tical afferents are not of an appropriate size to quoted from Purves et al. [597] and Horton match an individual pyramidal cell module. and Adams [305], to whom the reader is re- The terminal arbors of thalamocortical and ferred for primary sources and details: cortico-cortical afferents are generally 200 lm (a) Ocular dominance columns are present or more in diameter (Figs. 15.40, 15.43). This in Old World monkeys, but reportedly, absent does not exclude, of course, that these afferents in some New World monkeys. participate in the formation of larger process- (b) Remarkably, it was found in squirrel ing units (see below). An important general monkeys that some members of this species question to be addressed in the near future is: have ocular dominance columns in only part what is the relationship (if any) between the of the visual cortex, leaving other regions of cylindrically shaped units proposed by Lorente binocular cortex bereft of columns. de Nó, Mountcastle and others and the neocor- (c) Blobs in the primary visual cortex of the tical functional networks, discussed in the pre- rhesus monkey are reportedly concerned with vious section of the present chapter. processing information about colour. However, The various column-like cortical entities these structures are also present in nocturnal have been distinguished on the basis of highly primates with cone-poor retinas (and, there- different morphological and/or physiological fore, poor colour vision). Moreover, blobs are properties, e.g. similarity in physiological re- absent in some non-primate species with cone- sponse properties, similarity in motor re- rich retinas and excellent colour vision, such sponses following stimulation, spatial segrega- as the tree shrew and the ground squirrel. tion of sets of thalamocortical afferents, high (d) Barrels are well developed in the mouse, concentrations of the enzyme cytochrome oxi- rat, squirrel, porcupine and walrus, but are ab- dase, concentrations of spiny stellate cells in sent in other species that have prominent facial lamina IV and the mode of termination of whiskers such as the dog, cat, raccoon and tree bundles of cortico-cortical fibres. It is unlikely shrew. Furthermore, barrels occur in the guin- that all of these entities are derivatives of one ea pig, which hardly uses its whiskers, and in and the same basic cortical module, especially the chinchilla, another cavimorph that has no as they may differ considerably in size. It can- whisking behaviour at all. not be excluded that, as Szentágothai [722] 6. The barrel cortex of rodents and the pri- surmised, the units defined by cortico-cortical mate visual cortex are frequently cited as ar- afferents represent the “true cortical columns”. chetypical examples of columnar structures. With regard to these structures, however, it However, Jones ([341], p. 5021) has pointed should be emphasized that (a) their presence out that these two structures “represent end- has so far not been established in monotremes, points in the evolution of the two orders, one marsupials or insectivores, (b) their intrinsic of which noses and whisks its way around its structure is unknown and (c) their functional environment and the other of which extracts significance remains to be elucidated [305]. an extraordinary richness of visual detail from It is also important to note that the vertically its environment”. aligned rows of cell bodies observed in Nissl preparations of many cortical areas, which are If we survey the data discussed above, the fol- often equated to cortical minicolumns, never ex- lowing conclusions seem to be warranted: tend uninterruptedly from layers II to VI and 1. During the last 50 years, numerous different that most of the column-like formations de- types of classes of column-like entities have scribed have no definite boundaries [621]. been detected in different parts of the neo- 15 Telencephalon: Neocortex 591

cortex of many different mammalian species contacts with the apical dendritic extensions with the aid of a variety of physiological of pyramidal neurons. The sources of these and anatomical techniques. There is little fibres include the olfactory bulb, the thala- evidence of a systematic relation between mus and other cortical regions. The fibre these classes of “columns”. composition of the superficial layer varies 2. There is very little evidence in favour of the from region to region. Thus, in the prepiri- concepts that (a) the entire mammalian neo- form cortex, secondary olfactory fibres oc- cortex is composed of column-like entities, cupy a superficial position, whereas cortico- (b) all of these entities represent variations cortical fibres form a deeper zone. In the on one and the same theme, (c) all of these sensory part of the neocortex of primitive entities essentially have the same structure mammals, numerous thalamocortical fibres and (d) they all essentially subserve the reach lamina I, but in the hippocampus and same function. in neocortical association area, cortico-cortical fibres prevail in this layer. In lamina I of all cortices, fibres from different sources tend to be arranged in different sublaminae. Comparative Aspects 5. In neocortical sensory regions, thalamic afferents terminate massively in lamina IV. In the primary visual cortex of prosimians and If we compare the neocortex with the “older” primates, different laminae of the lateral parts of the mammalian cortex, i.e. the prepiri- geniculate nucleus project to different sub- form or olfactory cortex (Figs. 11.5, 11.6) and layers of lamina IV. the hippocampal cortex (Figs. 12.6–12.9), the 6. Inhibitory interneurons, using GABA as following similarities and differences can be their neurotransmitter, occur in all cortices. observed: All inhibitory interneurons in all cortices 1. Pyramidal neurons occupy a central position impinge directly with some or all of their in the circuitry of all of these cortices. These terminals on pyramidal neurons. In all cor- elements are provided with two, spatially tices, separate sets of inhibitory interneu- separated, dendritic systems; an apical den- rons terminate mainly or exclusively on one dritic system, expanding in the most super- of the three receptive domains of the pyra- ficial cortical layer; and a basal dendritic midal cells, i.e. the dendritic compartment, system, the laminar position of which may the soma and the axon initial segment. vary considerably. 7. Some inhibitory interneurons receive their 2. In all cortices, the pyramidal neurons are principal input from extrinsic afferents and excitatory in nature and potentially consti- mediate feed-forward inhibition of pyramidal tute a continuous network extending neurons, e.g. the horizontal cells in lamina I throughout the cortex. Nearby connections of the piriform cortex and in the neocortex. in this network are provided by axon collat- Other inhibitory interneurons receive their erals of pyramidal neurons, whereas longer main input via axon collaterals of pyramidal intracortical connections consist of the main neurons, thus forming part of feed-back loops axons or axonal branches of more remote (e.g. the large multipolar cells in lamina III of pyramidal neurons. In the neocortex, each the piriform cortex) and still others, receiving pyramidal neuron receives by far its largest excitatory inputs from extrinsic afferents and input from other pyramidal neurons. from axon collaterals of pyramidal cells, are 3. Pyramidal neurons constitute the output involved in both feed-forward and feed-back system of all of these cortices. inhibition of pyramidal neurons, e.g. basket 4. In all cortices, the most superficial layer cells in the hippocampus and the neocortex. contains numerous tangentially running af- 8. Inhibitory interneurons, exerting their influ- ferent fibres, making excitatory synaptic ence on other inhibitory interneurons, and 592 Section II Structure of Spinal Cord and Brain Parts

thus a disinhibition of pyramidal neurons, hence, has little to do with brain function. are present in all cortices. The elements in- However, it appears to be possible to subdivide volved in such inter-interneuronal inhibi- all of them into a number of functional do- tory circuits may be of the same type, e.g. mains. Throughout the description, the still the large multipolar cells in the piriform widely used area parcellation of Brodmann cortex and the basket cells in the neocor- (Fig. 15.8) will be used as a reference system. tex, or of different types, e.g. the so-called The discussion of the various lobes is prefaced L/M cells and the basket cells in the hippo- with some remarks on the neocortical associa- campus and the basket cells and double tion and commissural connection and on the bouquet cells in the neocortex. functional and structural asymmetry of the 9. In the neocortex as well as in the hippo- two cerebral hemispheres. campus groups of inhibitory interneurons have been observed that are interconnected via chemical synapses and gap junctions Association and Commissural Connections and also innervate themselves via autapses [24, 101, 362]. Association and commissural fibres, which in- 10. Excitatory interneurons have so far only terconnect different neocortical areas, make up been demonstrated in the neocortex. These most of the white matter of the cerebral hemi- elements, the spiny stellate cells, represent sphere. Commissural fibres cross in the corpus transformed pyramidal neurons. They are callosum and in the caudal part of the anterior abundant in primary sensory cortices, commissure. Association systems can be rather where they play a prominent role in the ra- arbitrarily subdivided into short and long asso- dial propagation of the activity fed by tha- ciation fibres. Short association fibres may re- lamocortical afferents into lamina IV of main within the grey matter of the cortex or pass these cortices. through the superficial white matter between neighbouring cortical areas as U fibres. Long association systems are located in deeper parts of the white matter, lateral and medial to the coro- Synopsis of Main Neocortical Regions na radiata and the internal capsule. In the human brain, long association systems are mainly known from gross dissection [145, 421], Introduction although imaging techniques have recently become available, rendering it possible to visualize In this section an overview of the structure these bundles in the living human brain [40, and principal connections of the main neocor- 108, 335, 488, 588]. The most important long as- tical regions will be presented. We will confine sociation systems are named and depicted in ourselves here mainly to cortico-cortical pro- Figs. 15.45 and 15.46. It can be seen that these jections. Neocortical afferents have been dis- systems interconnect cortical regions in differ- cussed in a previous section of the present ent lobes within the same hemisphere. chapter and efferents of the motor cortex will The superior occipitofrontal (or subcallosal) be dealt with in Chap. 21. For discussion of fasciculus is situated dorsolateral to the caudate the relations of the neocortex with the basal nucleus, immediately underneath the most me- ganglia and cerebellum, we refer the reader to dial part of the radiation of the corpus callo- Chaps. 14 and 20, respectively. In this overview sum. Its fibres connect the occipital and tem- the classical subdivision of the hemisphere into poral regions with the frontal lobe. the occipital, parietal, temporal, limbic, frontal The superior longitudinal fasciculus is located and insular lobes will be followed (Fig. 1.4). It along the laterosuperior border of the putamen. is important to note that this subdivision was It is separated from the superior occipitofrontal based originally on the bones of the skull and, fasciculus by the fibres of the most proximal 15 Telencephalon: Neocortex 593 part of the corona radiata. It forms a large arcu- tion of the present chapter (Structural Subdivi- ate bundle that interconnects the frontal lobe sion 4: Connectivity), it may be good to repeat with the three post-Rolandic lobes. Its posterior some general principles applying to the organi- part splits up into a brachium posterius, which zation of the connections (Figs. 15.15, 15.46, fans out into the parietal, occipital and posterior 15.47). It should be emphasized that our temporal lobes, and a strongly curved brachium knowledge of the cortico-cortical connections anterius, which is connected with the anterior is almost entirely based on experimental stud- part of the temporal lobe. ies in non-human primates [463–465, 526, 529, The inferior occipitofrontal fasciculus and 532, 579, 773]. the uncinate fasciculus form part of a single fi- 1. Cascades of short association fibres inter- bre complex. The compact, intermediate parts connect the primary, modality-specific areas of these bundles are situated directly under- of the cortex, which receive their sensory in- neath the claustrum and pass via the limen input from the thalamus, with the modality- sulae from the frontal to the temporal lobe. specific parasensory association areas. The The inferior occipitofrontal fasciculus consists somatosensory association cortex is found of fibres that connect the lateral parts of the in the parietal lobe, the visual association frontal lobe with the inferior temporal and me- areas in the occipital lobe and the temporal dial and lateral occipitotemporal gyri and with lobe below the superior temporal sulcus, the occipital lobe. The uncinate fasciculus con- and the auditory association areas are in the nects the inferior frontal gyrus and the orbital superior temporal gyrus and the temporal surface of the frontal lobe with anterior por- operculum. These short connections will be tions of the temporal lobe. reviewed according to their modality. The inferior longitudinal fasciculus follows 2. Modality-specific areas of parasensory asso- the lateral wall of the posterior and inferior ciation cortex are connected with multimo- horns of the lateral ventricle. It connects the dal sensory areas located at their borders. occipital lobe with most parts of the temporal These areas constitute a belt extending from lobe. Tusa and Ungerleider [739] provided ex- the junction of the occipital and parietal perimental evidence indicating that the inferior lobes, through the caudal part of the superi- longitudinal fasciculus in the rhesus monkey or temporal gyrus, into the superior tempo- consists of a series of U fibres that sequentially ral sulcus. Long association systems connect connect adjacent regions in the striate, prestri- the modality-specific parasensory associa- ate and inferior temporal cortex. On that action cortex and the multimodal areas in the count they proposed replacing the term “infe- occipital, temporal and parietal lobes with rior longitudinal fasciculus” by the term “occi- the premotor and prefrontal cortex of the pitotemporal projection system”. frontal lobe. Short association fibres inter- The cingulum, finally, is a large system of connect the prefrontal cortex, the premotor shorter and longer association fibres, which is area and the motor cortex. Short association situated in the white matter of the cingulate fibres interconnect the motor cortex and the gyrus, following this convolution throughout primary somatosensory cortex. its entire length (Fig. 12.13). Anteriorly, it ex- 3. Connections from the parasensory and mul- tends around the genu of the corpus callosum timodal association cortices and the pre- to the subcallosal area. Posteriorly, it arches frontal cortex to limbic structures pass by around the splenium of the corpus callosum way of the cingulum to the medial part of and continues downward and forward within the temporal lobe. Other fibres originating the parahippocampal gyrus. From there, its from parasensory association cortices reach constituent fibres fan out in the adjoining limbic structures via the insula. parts of the medial temporal lobe. 4. Most association connections are reciprocal. Although the neocortical association systems 5. Connections from the primary sensory areas have already been dealt with in a previous sec- to their neighbouring association areas usual- 594 Section II Structure of Spinal Cord and Brain Parts

1 Superior occipitofrontal fasciculus 2 Site of corona radiata 3 Superior longitudinal fasciculus 4 Superior longitudinal fasciculus, brachium posterius 5 Superior longitudinal fasciculus, brachium anterius 6 Outline of insula 7 Inferior occipitofrontal fasciculus 8 Inferior longitudinal fasciculus 9 Site of anterior commissure 10 Uncinate fasciculus

Fig. 15.45. Long association bundles of the right cerebral hemisphere in a lateral view (1/1´). Part of the superior longitudinal fascicle has been removed to show the superior occipitofrontal fascicle 15 Telencephalon: Neocortex 595

B Broca’s speech region M1 primary motor area MT Middle temporal visual association area S1 primary sensory area W Wernicke’s speech region

Fig. 15.46. Long association connections of the neocortex of the left hemisphere. The multimodal association cortex in the superior temporal sulcus and the parieto-occipital sulcus is dotted 596 Section II Structure of Spinal Cord and Brain Parts

ly originate from the supragranular layers and The corpus callosum is divided into a terminate in and around layer IV. This type of curved rostral part, the genu, a large middle connection is therefore considered to transfer part, the body or truncus, and a thickened cau- information in a forward direction. The re- dal part, the splenium. A tapering rostrum con- verse (feed-back) connections originate in nects the posteriorly recurved part of the genu the infragranular layers and terminate in with the lamina terminalis (Figs. 3.7, 15.48). layers I and VI [623]. The laminar analysis The fibres of the corpus callosum, which fan of association connections may therefore re- out into the white matter of both hemispheres, veal the direction of information transfer. form the radiation of the corpus callosum (Figs. 5.4–5.10). The extension of these fibres Fibres interconnecting neocortical regions in in the most frontal parts of the hemisphere is the two hemispheres pass through two com- called forceps minor, whilst the similar but missural systems, the anterior commissure and wider sweep of fibres toward the occipital the corpus callosum (Fig. 3.7). Both of these poles is called forceps major (Figs. 5.29, 15.48). commissures develop in the commissural plate, The fibres destined for the basal parts of the i.e. the thickened, most rostral part of the lam- temporal and occipital lobes form a thin fibre ina terminalis (Fig. 2.6). plate, the tapetum, which accompanies the dor- The anterior commissure is throughout most solateral wall of the inferior horn of the lateral of its extent a compact, transversely oriented ventricle (Fig. 15.48). bundle (Fig. 12.4). It comprises a small ante- Commissural connections are homotopic or rior limb and a much larger posterior limb heterotopic. Homotopic fibres interconnect cor- (Fig. 15.48). The anterior limb interconnects responding cortical areas in both hemispheres. olfactory structures, such as the anterior olfac- The fibres interconnecting the frontal lobes tory nuclei and the primary olfactory cortices, pass in the anterior half of the corpus callo- on the two sides (Fig. 11.7). The posterior sum, the others pass in the posterior half, with limb, which passes laterally and somewhat the parietal fibres anterior to those from the backward through the most inferior parts of temporal lobe, and the occipital lobe fibres the lentiform nucleus (Figs. 5.5, 5.6, 5.21–5.24), most posteriorly, in the splenium [156, 533]. interconnects the anterior portions of the mid- Experimental studies in rhesus monkeys dle and inferior temporal gyri. [126, 361] have shown that there are consider- A corpus callosum is lacking in monotremes able regional variations as regards the distribu- and marsupials [511]. In eutherians, the size of tion of callosal afferents. Thus, commissural fi- the corpus callosum parallels that of the neocor- bres are absent from most of the primary vis- tex. Thus, it is small in primitive mammals such ual cortex (area 17) and the same holds true as the hedgehog (Fig. 12.1 D), larger in prosi- for the areas representing the distal parts of mians (Fig. 12.1 E), and attains its maximal size the limbs in both the primary somatosensory in humans (Fig. 12.1 C). During ontogeny, the and the primary motor cortices. Most of the human corpus callosum manifests itself initially association areas, on the other hand, are as a small, compact bundle in the most dorsal strongly interconnected by callosal fibres. part of the commissural plate (Fig. 2.6 A). Dur- Heterotopic commissural connections connect ing further development, it thickens, expands a cortical area with non-corresponding areas backwardly (Fig. 2.6 B–D) and ultimately covers in the contralateral hemisphere. The pattern of the entire diencephalon (Fig. 7.1). heterotopic connections of a particular cortical In the human adult, the corpus callosum is area often mimics its association connections a broad plate of fibres that interconnects the with other areas in the ipsilateral hemisphere. neocortical portions of the two hemispheres. It It is sometimes possible to distinguish between forms the floor of the longitudinal fissure and forward and backward types of heterotopic the roof of the lateral ventricles (Figs. 5.4– connections on the basis of their laminar ori- 5.11). gin and termination [126]. 15 Telencephalon: Neocortex 597

B Broca’s speech region p.h. parahippocampal gyrus pl.t. planum temporale W Wernicke’s speech region

Fig. 15.47 A,B. Short association connections of the cerebral cortex. A Lateral view. The frontal and part of the parietal operculum has been removed; the temporal operculum is retracted. B Medial view. The primary and secondary sensory areas (S1 and S2), the primary and secondary motor areas (M1 and M2), the primary visual and auditory cortices, the paralimbic association cortex and the insula, and the limbic cortex are indicated with different shadings. The multimodal cortex in the superior temporal sulcus and the parieto-occipital sulcus is dotted 598 Section II Structure of Spinal Cord and Brain Parts

1 Olfactory bulb 8 Body of caudate nucleus 2 Forceps minor 9 Body of corpus callosum 3 Genu of corpus callosum 10 Tail of caudate nucleus 4 Head of caudate nucleus 11 Tapetum 5 Rostrum of corpus callosum 12 Splenium of corpus callosum 6 Anterior commissure, anterior limb 13 Radiation of corpus callosum 7 Anterior commissure, posterior limb 14 Forceps major

Fig. 15.48. Commissural connections of the telencephalon as seen from the basal side of the brain (1/1´) 15 Telencephalon: Neocortex 599

Association and commissural connections asymmetry began in the 1860s and 1870s often originate from and terminate in strips, with the discovery of Broca [67] and Wer- which in turn are separated from each other nicke [811] that language functions are lat- by strips lacking these particular connections. eralized to the left hemisphere. The notion This pattern of alternating positive and nega- of a dominant hemisphere was first pre- tive strips has a periodicity of 200–1000 lm sented by Hughlings Jackson on the basis of [243, 319, 332, 349]. In the somatosensory and the frequent occurrence of motor (and other auditory cortices, the strips extend perpendic- kinds of) aphasia in right-handed persons ular to the borders of the somatotopic or the when they suffer a hemiplegia of the right isofrequency maps, but a periodicity in the limbs, while a left-sided hemiplegia in such origin and termination of cortico-cortical con- persons usually is not accompanied by nections has also been observed for visual as- aphasia [69]. sociation areas and for the multimodal, frontal 2. Studies on the laterality of motor control. and paralimbic association cortices. Cells of Such studies have shown that in all races origin and their homotopic terminations are more than 90% of the population is natu- located in the same strips. It is not certain rally more skilled with the right hand than whether the ipsilateral association and com- with the left [112, 114]. As is well known, missural pathways are complementary with re- the right hand is controlled by the left spect to the strips [76, 246, 337, 344, 349, 667]. hemisphere. Association and commissural connections 3. Studies in which one of the cerebral hemi- are involved in many higher functions of the spheres is temporarily inactivated. In the so- nervous system. Damage to the primary sen- called Wada test [799], a barbiturate is in- sory or parasensory association area may rejected into the right or left internal carotid sult in perceptual deficits. In contrast, the in- artery, by which the ipsilateral hemisphere terruption of commissural connections or the is inactivated for a short while. The test association connections between the unimodal forms part of the presurgical assessment of area and multimodal or paralimbic association hemispheric language dominance in patients areas may lead to disconnection syndromes. suffering from intractable epilepsy. In the These syndromes have been extensively studied dominant hemisphere, such an injection in experimental animals and in humans [227, transiently blocks speech. Results with this 462, 703]. test have shown that 92–99% of dextral individuals are left hemisphere-dominant for language, while the pattern in non-dextral Functional and Structural Asymmetry individuals includes leftward but also bilat- of the Two Hemispheres eral or even rightward dominance for language [417]. Functional asymmetry of the two hemispheres is 4. Studies in patients in whom the two cerebral a salient feature of human brain organization hemispheres are surgically separated by and cognition. This phenomenon is also termed transection of the corpus callosum and the “hemispheric specialization”, “functional latera- anterior and hippocampal commissures. Op- lization” and “cerebral dominance”. These terms erations of this type are carried out to con- all refer to the fact that the right and left cerebral trol intractable seizures [799]. Such “split- hemispheres have different roles in mediating brain” individuals offer a unique opportuni- behaviour and higher mental processes [226]. ty to test the functions of each hemisphere Evidence for functional asymmetry of the independently of the other [224, 225]. cerebral hemispheres is mainly derived from 5. Neuroimaging studies, which render it pos- studies of the following types. sible to identify and localize changes in me- 1. Clinicopathological studies. The modern era tabolic activity in the brain, correlated with of neuroscientific investigation into cerebral cognitive and mental tasks, have been 600 Section II Structure of Spinal Cord and Brain Parts

widely applied to the study of functional in the volume of cytoarchitectonically defined asymmetries. Moreover, MRI scans can be areas. Thus, Galaburda et al. [221] measured used for in vivo morphometry of brain an auditory association area, designated as Tpt structures [417, 708]. in four brains. Distinct leftward asymmetries were seen in this area, whereby in one case the Overall, the studies listed above have shown left area Tpt was 620% larger than the right. that in right-handed persons the left hemi- Moreover, a perfect rank-order correlation was sphere is mainly concerned with verbal and found between the magnitude of asymmetry of linguistic functions, mathematical skills and area Tpt and the degree of planum temporale analytical thinking, whereas the right hemi- asymmetry. sphere is primarily involved in spatial relation- It is generally accepted that areas 44 and 45 ships, musical and artistic functions and in the of Brodmann in the inferior part of the frontal recognition and expression of emotions. lobe constitute the cytoarchitectonic correlates Left-handed persons form a heterogeneous of Broca’s speech region. Amunts et al. [6] de- group. Only 15% of left-handers show the ex- lineated these two areas in 10 human brains. pected right-hemisphere dominance; a full They found a left-larger-than-right asymmetry 70% of them is left-hemisphere dominant, of area 44 in all brains studied. while the remaining 15% has bilateral language Histological asymmetries have also been abilities [226]. found in areas related to language. Scheibel et Crow and Mitchell [123, 473] concluded al. [655] reported differences in the complexity from lesion studies in stroke patients and func- of the dendritic trees of pyramidal cells be- tional imaging studies of healthy people that tween the speech areas in the left hemisphere some language functions, including discourse and their counterparts on the right side, and planning/comprehension, understanding hu- Hayes and Lewis [281] found that the pyrami- mour, sarcasm and metaphors, are mediated dal neurons in layer III of Broca’s area 45 are by the right hemisphere rather than the left. It consistently larger on the left than on the right may be added that, according to these authors, side. Finally, it may be mentioned that evi- the language deficits in patients with schizo- dence, recently summarized by Hutsler and Ga- phrenia can be interpreted as abnormalities of luske [318], indicates that the width of individ- lateralization. ual columns in the posterior language-oriented Possible morphological correlates of the areas is greater in the left hemisphere than in functional asymmetries just discussed have the right. been described by numerous authors. Ge- schwind and Levitsky [229] reported left–right asymmetries in the temporal speech region. Occipital Lobe They measured the size of the planum temporale, i.e. the posterior region of the superior The occipital lobe occupies the most posterior surface of the temporal lobe, in 100 brains and portion of the hemisphere on its lateral, medial found it to be larger on the left than on the and basal surfaces (Figs. 1.4, 15.49). On the lat- right in 65, with no asymmetry in 24 and with eral or convex surface of the hemisphere, the a reversed pattern in 11 brains. Their observa- anterior boundary is formed by a somewhat ir- tions have been confirmed by several other in- regular parieto-occipital line which, starting vestigators, among them Witelson and Pallie from the superior end of the parieto-occipital [829], Wada et al. [800], Rubens [642] and sulcus, partly follows the anterior occipital sul- Steinmetz [708]. Habib et al. [271] demon- cus (Fig. 3.2). On the medial surface, the ante- strated that the leftward volume asymmetry of rior boundary of the occipital lobe is marked the planum temporale is related to the degree by the parieto-occipital sulcus (Fig. 3.6). An of right-handedness. Asymmetries in the pla- arbitrary, transversely oriented occipitotempor- num temporale have also been demonstrated al line, cutting through the medial and lateral 15 Telencephalon: Neocortex 601

Fig. 15.49 A–D. The occipital lobe. Subdivision (numbers) according to Brodmann, lateral (A) and medial (B) views. The occipital lobe is shown in red. Functional areas adopted from Tootell et al. [737], lateral (C) and medial (D) views. FFA, fusiform face area; LO, lateral occipital area; MST, middle superior temporal visual area; MT, middle temporal visual area; STP, superior temporal polysensory area; V1, V2, VP, etc., visual areas 602 Section II Structure of Spinal Cord and Brain Parts occipitotemporal gyri, forms the anterior nation of fMRI, PET and lesion studies, in- boundary of the occipital lobe on the basal cludes Brodmann’s areas 7, 18–21, 37 and 39 side (Fig. 3.5). On the medial aspect of the (Fig. 15.50) [161]. By this estimate, roughly a hemisphere, the deep calcarine sulcus sepa- third of human neocortex is associated with rates the cuneus from the medial occipitotem- visual processing [771]. poral gyrus. Anteriorly, the calcarine sulcus Our present-day knowledge of the central joins the parieto-occipital sulcus (Fig. 3.6). visual system is largely based on experimental The occipital lobe is occupied by three con- studies in non-human primates, particularly centrically arranged cytoarchitectonic areas, the rhesus monkey. These studies have shown the areas 17, 18 and 19 of Brodmann (Figs. that there is a far-reaching “division of labour” 15.8, 15.49 A,B). in the analysis and processing of visual infor- Area 17 surrounds the calcarine sulcus. It is mation. Many cortical areas outside the pri- also known as the area striata because it is mary visual cortex appear to be specialized for characterized by the presence of a dense layer the processing of specific aspects of vision, of myelinated fibres, known as the line of Gen- such as form, motion or colour. nari. This line is clearly visible in unstained, The extrastriate cortex of the macaque mon- macroscopical sections (Figs. 5.13, 5.14, 5.30, key has been shown to contain a large number 5.31). Histologically, the area striata is typified of visual areas defined by three or more of the by its richness of small, granular cells (“konio- following criteria: (i) presence of retinotopic cortex”). The internal granular layer is subdi- representation of the contralateral visual hemi- vided into three sublayers, IVa–c (Figs. 15.6, field or of part of it; (ii) functional specializa- 15.41 B). The transition zone between layers V tion of neurons; (iii) specific effects of lesions; and VI contains the large cells of Meynert (iv) specific connectivity pattern and (v) spe- (Figs. 15.28 (4), 15.35). cific cyto-, myelo- and/or chemoarchitecture Areas 18 and 19 show a homotypical lami- [850]. More than 30 different areas have been nation pattern. The borderline between areas identified in the extrastriate visual cortex of 17 and 18 is very distinct. The line of Gennari the rhesus monkey [180, 408, 761, 770, 771]. stops abruptly and the various sublayers of Some of these are labelled with a capital “V”, layer IV merge into a single internal granular standing for visual area, followed by a number: layer (Figs. 15.6, 15.41 B). The anterior borders V2, V3 etc. Others are designated with combi- of areas 18 and 19, on the other hand, are hard nations of letters, indicating their topographi- to establish. The borders of areas 17 and 18 cal position. Mainly on the basis of functional show a considerable intersubject variability imaging studies, putative equivalents of 8–10 (Fig. 15.22 B) [7, 186]. of the functional areas detected in the extra- Area 17, which receives its principal afferent striate cortex of the rhesus monkey have been projection from the lateral geniculate body, identified in the human brain (Figs. 15.24 C, represents the primary visual cortex (V1). The 15.49 C,D). areas 18 and 19, which surround area 17, are As regards V2, Zilles [849] indicates that collectively designated as the parastriate belt. this area corresponds to Brodmann’s area 18, The areas 18 and 19, which are strongly vis- although Kaas [352] claims that V2 is much ually responsive, receive their principal affer- narrower than area 18. V2 is strongly activated ents either directly or indirectly from the stri- during shape discrimination [267]. ate area and are, hence, also characterized as Area 19 of Brodmann has been shown to extrastriate visual areas. However, the total ex- harbour a considerable number of functional trastriate visual cortex is not confined to the areas within its confines, including V3, VP, occipital lobe, but rather extends anteriorly V3A, V4v, V8 and V5/MT. V3, which adjoins into the parietal and temporal lobes. The over- the anterior border of the upper part of V2, all extent of extrastriate visual cortex (or visual contributes to dynamic form perception. Many association cortex), as estimated from a combi- of its neurons are both orientation- and mo- 15 Telencephalon: Neocortex 603

Fig. 15.50. Principal visual projections. The total brain area dominated by visual projections is shown in light red; visual centres in darker red. DVPS, dorsal visual processing stream; FF, frontal eye field; ILF, inferior longitudinal fasciculus; PEF, parietal eye field; PF, prefrontal eye field; SF, supplementary eye field; VVPS, ventral visual processing stream; 7b, 20, etc., areas according to Brodmann 604 Section II Structure of Spinal Cord and Brain Parts tion selective [849]. VP, the ventroposterior we confine the discussion to a brief survey of visual area, is situated anteriorly to the basal the principal connections of the various com- part of V2. It has been argued that this area is ponents of the occipital lobe (Fig. 15.50). not a separate functional entity, but rather rep- Efferent fibres from the lateral geniculate resents the ventral part of V3 [847]. V3A occu- body pass to the primary visual cortex (V1) pies the most superior part of the occipital and to the middle temporal visual area (MT) lobe, extending over both the lateral and the [687]. In V1, the fibres from the magnocellular, medial surface of the hemisphere. V3A, V7 parvocellular and intercalated layers of the lat- and adjoining areas in the parietal lobe form a eral geniculate body terminate in different complex that is activated by saccades and at- layers, particularly in the various sublayers of tention [124]. LO, the lateral occipital area, is layer IV (Fig. 15.26). The remaining visual cen- situated anteriorly to V3. Human imaging tres in the occipital lobe receive their thalamic studies have shown that this area is involved in afferents principally from the pulvinar. Addi- shape processing. It responds strongly to tional afferents to the occipital lobe arise from images of objects, but not to scrambled control the amygdala (Fig. 13.7) [205] and the claus- stimuli [241]. V8, which is sometimes desig- trum (Fig. 13.8). The afferents from the thala- nated as V4 [737], borders laterally to LO and mus and the claustrum are reciprocated by ef- extends basally over the posterior parts of the ferent fibres. Several visual cortical centres, in- lateral and medial occipitotemporal gyri. Func- cluding V1–V3A, MT and MST, project to the tional imaging studies have shown that this area pontine nuclei [210]. is involved in colour perception [60, 266, 275, Short association fibres interconnect area 846]. Lesions in this region cause central colour 17(V1) with many extrastriate visual areas, in- blindness (achromatopsia). As its name indi- cluding V2, V3, V3A, V4, V5/MT and MST [5, cates, the fusiform face area (FFA) is located in 161, 742, 768, 769]. the fusiform (or lateral occipitotemporal) gyrus, Analysis of the connections of the various directly in front of V8. Functional imaging stud- extrastriate visual areas has led to the concept ies and single unit electrophysiology from non- that these areas are organized into two path- human primates have shown that this area is se- ways: an occipitoparietal pathway or “dorsal lectively responsive to faces [241, 658, 737]. Le- stream” and an occipitotemporal pathway or sions in FFA lead to prosopagnosia, a selective “ventral stream” [32, 180, 181, 623, 686, 706, difficulty identifying faces. V5 or MT, the middle 735, 741, 743]. As discussed in a previous sec- temporal visual area, is located in the area where tion of the present chapter (Structural Subdivi- the inferior temporal sulcus meets the anterior sion 4: Connectivity), the sequential processing occipital sulcus. This area, which is common of information in the areas along both path- to all primates studied so far [352], is morpho- ways is progressively more complex and may, logically characterized by its dense myelination hence, be characterized as hierarchical. [761]. Functional imaging studies have shown The dorsal, occipitoparietal processing stream that this area is strongly activated by moving vi- is engaged in the perception of relative spatial sual stimuli [266, 317, 846]. Bilateral lesions that location of objects as well as visual guidance include this area cause a severe impairment in of movements towards objects (“where” sys- detecting the movements of objects, known as tem). It includes V1, V2, V3, V3A, V4, V5/MT, cortical akinetopsia. MT is surrounded by sev- MST and STP (superior temporal polymodal) eral other specialized areas which process higher (i.e. a polysensory area situated in the dorsal order aspects of motion perception and motor bank of the superior temporal sulcus, and the planning [653]. One of these, the middle superi- so-called parietal eye field [32, 241, 686]. STP or temporal visual area, MST, is situated ante- responds to visual, auditory and somatosen- riorly to MT [241]. sory stimuli. It interacts with the inferior tem- For a detailed discussion of the visual sys- poral cortex. In the parietal lobe of the rhesus tem, the reader is referred to Chap. 19. Here monkey, the territory surrounding the intra- 15 Telencephalon: Neocortex 605 parietal sulcus contains a series of at least five Parietal Lobe different areas, all of which are concerned with highly specialized visual functions (see next The parietal lobe is bounded on the lateral sur- section). These areas are collectively desig- face anteriorly by the central sulcus, posterior- nated as the parietal eye field. Neuroimaging ly by the parieto-occipital line, and inferiorly studies suggest that all of the areas included in by the posterior ramus of the lateral sulcus this field have a functional equivalent (ana- and an arbitrary parietotemporal line. The lat- logue) in the human brain [124]. The parietal ter passes horizontally through the border area eye field functions as a visual-motor interface. of the supramarginal and angular gyri, on the It is reciprocally connected with V1, V2 and one hand, and the superior and middle tem- V3 through direct and indirect pathways pass- poral gyri on the other (Figs. 3.2, 15.51). On ing through MT, MST and V3A. The frontal the medial surface, the boundaries are: the lobe contains three fields which give rise to parieto-occipital sulcus, the subparietal sulcus eye movements when stimulated, namely the and a vertical line connecting the superior lim- frontal eyefield, situated in the most posterior it of the central sulcus with the cingulate sul- part of the middle frontal gyrus, the prefrontal cus (Fig. 3.6). eye field, located more anteriorly in area 46, The parietal lobe can be subdivided into and the supplementary eye field, situated dor- four parts, the postcentral gyrus, the superior sally, in the most rostral part of area 6. These parietal lobule, the inferior parietal lobule and frontal eye fields are heavily and reciprocally the parietal operculum. The postcentral gyrus connected with the parietal eye field. Moreover, runs parallel to the precentral gyrus and is sit- most of the occipital extrastriate areas, interca- uated between the central and postcentral sulci. lated in the dorsal processing stream, also pro- The horizontally oriented intraparietal sulcus ject directly to the frontal and prefrontal eye divides the region posterior to the postcentral fields [434, 579, 653]. sulcus into superior and inferior parietal lobu- The ventral occipitotemporal processing les. The latter consists of two gyri, the supra- stream is concerned with pattern discrimina- marginal gyrus, arching around the upturned tion and the visual identification of objects end of the lateral sulcus, and the angular (“what” system). Areas V1, V2, LO, V4/V8 and gyrus, which surrounds the ascending, termi- the entire inferotemporal cortex, composed of nal part of the superior temporal sulcus (Fig. areas 37, 20 and 21, are included in the ventral 3.2). The parietal operculum is not visible on stream. Most of the fibres, interconnecting the the free surface of the hemisphere. It is sit- various areas participating in this stream, pass uated deep to the posterior part of the lateral forwardly in the inferior longitudinal fasciculus sulcus and connects the postcentral gyrus and (Fig. 15.45) [579]. The separation of the dorsal the anterior part of the supramarginal gyrus and ventral streams is not absolute. Thus, it is with the insula. In Fig. 15.51 A,C, the parietal known that in the macaque monkey, the poste- operculum is brought to the external surface of rior part of the inferotemporal cortex is reci- the hemisphere by rotation. procally connected with the parietal and fron- Brodmann [70, 71] divided the parietal lobe tal eye fields [155]. into nine cytoarchitectonic areas: 1, 2, 3, 5, 7, Interruption of the dorsal and ventral pro- 31, 39, 40 and 43 (Figs. 15.8, 15.51 A,B). Areas cessing streams leads to distinct disconnection 3, 1 and 2 occupy the postcentral gyrus as syndromes. Thus, judgements of spatial rela- three narrow strips. They extend for some dis- tions are impaired by lesions in the posterior tance over the medial surface of the hemi- parietal region, whereas object recognition is sphere into the anterior part of the superior selectively impaired by lesions in inferior parietal lobule. Area 3 was later subdivided occipital and temporal areas [743]. into a small transitional zone 3 a, situated in or near the fundus of the central sulcus [388], and a larger zone 3 b. The areas 5 and 7 occu- 606 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.51 A–D. The parietal lobe. Subdivision according to Brodmann, lateral (A) and medial views (B). The parietal lobe is shown in red. Functional areas, lateral (C) and medial views (D). In A and C, the parietal operculum is brought to view by rotation. AIP, anterior intraparietal area; ANG, angular gyrus; CIP, caudal intraparietal area; LIP, lateral intraparietal area; MIP, medial intraparietal area; OP1–4, parietal opercular areas, according to Eickhoff et al. [169, 170]; par operc, parietal operculum; PIVC, parieto-insular vestibular cortex; PV, parietal ventral area; SPPC, superior polymodal parietal cortex; SSAC, somatosensory association cortex; SUMG, supramarginal gyrus; S1, primary somatosensory cortex; S2, second somatosensory area; VIP, ventral intraparietal area; W, Wernicke’s region; 1, 2, 3, etc., areas according to Brodmann; 7a,b, subdivisions of area 7; 2V, 3a, vestibular cortical areas 15 Telencephalon: Neocortex 607 py most of the superior temporal lobule. On The parietal operculum contains two somato- the medial side of the hemisphere, area 7 ex- sensory areas, the parietal ventral area and the tends over the precuneus. Areas 39 and 40 second somatosensory area. The parietal ventral roughly coincide with the angular and supra- area (PV) is located immediately adjacent to the marginal gyri, respectively. Area 31 is situated S1 cortex. It is posteriorly followed by the sec- on the medial side of the hemisphere, where it ond somatosensory area (S2) (Fig. 15.51 C) adjoins the posterior part of the cingulate [153]. PV and S2 roughly correspond to the cy- gyrus. It forms part of the limbic system and toarchitectonic areas OP4 and OP1, respectively will not be further considered here. The cy- [169]. Areas PV and S2 both contain a somatoto- toarchitecture of the parietal operculum was pic representation of the contralateral body sur- recently analysed by Eickhoff et al. [169, 170]. face. These two representations form mirror re- Four cytoarchitectonic areas, termed OP1–4, versals of each other along their common bor- were identified, of which OP4 and possibly der. The somatosensory system will be exten- OP3 correspond approximately with Brod- sively discussed in Chap. 16. Here we confine mann’s area 43. the discussion to mentioning that the three parts Cytoarchitectonically, four of the five funda- of S1 are reciprocally connected with each other mental cortex types, as distinguished by von and with the various parts of the superior pari- Economo [795], are represented in the parietal etal lobule [346, 348, 389, 586, 781]. S2 is densely lobe. Area 3 a, with its strongly developed pyra- interconnected with the primary sensory area midal layers and inconspicuous granular layers, 3b, PV and the inferior parietal lobule, whereas closely resembles the adjacent primary motor PV is interconnected with area 3 b, the superior cortex (type 1). Area 3b, which has a strongly and inferior parietal lobules, the premotor cor- developed internal granular layer IV and poorly tex, the frontal eye field and the medial auditory developed pyramidal layers III and V, must be belt areas [154]. Some of these connections are classified as a typical koniocortex (type 5). diagrammatically indicated in Fig. 15.52 A. The remaining parietal areas are all homotypi- The vestibular cortex includes an elongated cal, belonging either to types 2 or 3 (Fig. 15.10). area located in the inferior part of somatosen- In the following discussion of the connec- sory area 3 a, an area 2v situated around the tional and functional relations of the parietal anterior tip of the intraparietal sulcus and the lobe, this structure will be provisionally di- parieto-insular vestibular cortex (Figs. 15.51 C, vided into four parts, the somatosensory and 15.72). The different vestibular cortical areas vestibular cortices, and the superior, interme- are interconnected and the PIVC occupies a diate and inferior polymodal parietal cortices central position in this network [265]. The ves- (Fig. 15.51 C,D). tibular system will be discussed in Chap. 17. The somatosensory cortex occupies the post- The superior polymodal parietal cortex occu- central gyrus and part of the parietal opercu- pies the superior parietal lobule, except for its lum. The cytoarchitectonic areas 3b, 1 and 2, most anterior part (area 5), which forms the which cover the postcentral gyrus, are collec- (unimodal) somatosensory association cortex. tively known as the primary somatosensory There is evidence suggesting that this polymo- cortex, S1. (Area 3 a is usually included in the dal region is important for hand–eye coordina- motor cortex.) The contralateral half of the tion [333, 334, 733]. It is reciprocally connected body surface is somatotopically mapped onto with the prestriate cortex, via the dorsal proces- the S1 region of each hemisphere (Figs. sing stream, somatosensory area 5 and the dor- 15.18 C, 15.19 A). The lower limb is represented sal part of area 8, which contains the smooth eye in the upper part, the face in the lower part, movement-related frontal eye field [400]. The and the trunk and upper limb in the inter- superior parietal lobule also projects to the dor- mediate part of S1. Each of the three primary sal part of the premotor cortex (area 6) and to sensory areas contain a complete body map the supplementary motor area on the medial [354] (Fig. 16.5 F). surface of the frontal lobe [444, 579]. 608 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.52 A,B. Principal connections of the parietal lobe. Connections of the superior and opercular (A) and inferior parts (B) of the parietal lobe. A, angular gyrus; DVPS, dorsal visual processing stream; FF, frontal eye field; F2vr, F4, F5, specialized areas within the premotor cortex; MST, middle superior temporal visual area; MT, middle temporal visual area; PHC, parahippocampal cortex; PM, premotor cortex; PMd,v, dorsal and ventral subdivisions of premotor cortex; S, supramarginal gyrus; SMA, supplementary motor area; SPL,m, medial part of superior parietal lobule; STP, superior temporal polysensory area; 20, 21, etc., areas according to Brodmann 15 Telencephalon: Neocortex 609

The intermediate polymodal parietal cortex cortex. There is evidence suggesting that VIP is formed by a number of functional areas, is involved in the perception of self-movements which occupy the banks of the intraparietal and object movements in near extrapersonal sulcus. Recent morphological and physiological space. Premotor subunit F4 is known to be studies in monkeys and neuroimaging studies concerned with the transformation of object in humans have shown that the cortex border- locations into appropriate movements towards ing the intraparietal sulcus harbours multiple, them [445]. highly specialized functional areas, some of The medial intraparietal area (MIP) is lo- which can also be morphologically character- cated in the intermediate part of the medial ized. In general, these areas serve as interfaces bank of the ips. It receives somatosensory and between perceptive and motor systems. They visual afferents and is strongly connected with receive varying combinations of inputs from subunit F2vr, forming part of the dorsal pre- the surrounding visual, somatosensory, vestib- motor area. MIP and F2vr are known to be in- ular and auditory cortices and send strongly volved in the planning, execution and monitor- developed feed-forward projections to particu- ing of reaching movements. lar parts of the premotor cortex, which are re- The lateral intraparietal area (LIP) forms ciprocated by feed-back projections [444]. The part of a network of areas mediating saccades. various intraparietal sulcal areas and their pre- It receives input from several visual areas and motor targets form functional units, each of is interconnected with the frontal eye field (FF) which is dedicated to a particular aspect of and the superior colliculus. In the macaque sensorimotor transformation [445]. The follow- monkey, LIP is found in the lateral wall of the ing survey of the various intraparietal sulcal ips, hence its name. However, comparative areas is principally based on two review arti- functional studies have shown convincingly cles, by Culham and Kanwisher [124] and that the human LIP equivalent is located in the Grefkes and Fink [253], to which the reader is medial ips rather than in the lateral ips. referred for references and details. The approx- The caudal intraparietal area (CIP), finally, imate positions of these areas are indicated in is situated in the medial bank of the posterior Fig. 15.51 C. ips. It receives afferents from several visual The anterior intraparietal area (AIP), which areas, including V3, V3A and V4. Experiments is located on the lateral bank of the anterior in macaque monkeys have shown that neurons intraparietal sulcus (ips), is concerned with in CIP are involved in the analysis of three-di- tactile and visual object processing. It plays a mensional object features and are especially re- crucial role in the crossmodal transfer of ob- sponsive to axis and surface orientations of object information between the sensorimotor and jects in space. Neuroimaging studies have dem- visual systems. The AIP is connected to the onstrated that CIP is also activated during the ventral premotor area, especially to a subunit analysis of surface and pattern orientation in known as F5. The neurons in this subunit dis- humans. charge during specific object-related hand Some of the connections of the intraparietal movements. It seems likely that AIP in combi- cortical areas just discussed are indicated in nation with F5 transforms visual and somato- Fig. 15.52 A. sensory object data into finger movements for Choi et al. [97] recently studied the cytoar- object grasping and manipulation. chitecture of the cortex within the anterior The ventral intraparietal area (VIP) is lo- ventral bank of the human ips. They delineated cated in the fundus of the ips. It receives pro- two areas, which were termed the human injections from several visual areas, especially traparietal area 1 (hIP1) and the human intra- from MT and MST, from somatosensory, audi- parietal area 2 (hIP2). The areas hIP1 and tory and vestibular areas, and from other poly- hIP2 appeared to correspond with the func- modal cortices. It is strongly connected with tionally defined areas VIP and AIP, respec- F4, another subunit of the ventral premotor tively. 610 Section II Structure of Spinal Cord and Brain Parts

The inferior polymodal parietal cortex occu- part of the left superior temporal gyrus. Dam- pies the inferior parietal lobule, composed of age to this area gives rise to receptive or Wer- the angular and supramarginal gyri. As men- nicke’s aphasia, involving deficits in the com- tioned before, the extension of areas 39 and 40 prehension of both spoken and written lan- of Brodmann correspond roughly to those of guage. Long association fibres connect Wer- the angular and supramarginal gyri, respec- nicke’s region, via the anterior limb of the su- tively. perior longitudinal fasciculus, with the premo- There are some differences of opinion con- tor region of the frontal lobe, including Broca’s cerning the interpretation of the structures un- speech region in the inferior frontal gyrus der discussion. Geschwind [227] considered [219, 346, 398, 532]. the human inferior parietal lobule as a “new The following survey of the fibre connec- anatomical structure”. Crosby et al. [122] tions of the angular (area 39, PG) and supra- maintained that areas 39 and 40 have not been marginal gyri (area 40, PF) is principally based recognized in the macaque, and Zilles [849] on the experimental studies of Cavada and designated these areas as “human specific”, Goldman-Rakic [94, 95], Neal et al. [503–505], which means that they have no homologues in Andersen et al. [11], Petrides and Pandya non-human primates. One of the reasons for [579] and Gregoriou et al. [254], which were the interpretation of these authors may be that all carried out in the macaque monkey. Brodmann [70], in his map of the cortex of The angular gyrus is connected with visual Cercopithecus (Fig. 15.12), labelled the entire areas of both the dorsal (MST) and the ventral region posteroinferior to the intraparietal sul- (infratemporal cortex) processing streams, the cus as area 7 and did not delineate areas 39 medial part of the superior parietal lobule, the and 40 in this species. However, it should be intraparietal oculomotor area LIP, the superior noted that Brodmann in his book repeatedly temporal polysensory area STP, and with lim- emphasizes that identical numbers do not indi- bic areas of the posterior cingulate, retrosple- cate absolute homologies on the brain maps. nial and parahippocampal regions. Frontal In discussing the map of Cercopithecus, he in- connections of the angular gyrus are princi- dicates that area 7 of this species should be pally directed to the anterior part of the dorsal considered as an undifferentiated primordium premotor cortex, area 8a and prefrontal areas of all human parietal areas, except for area 5, 45 and 46 (Fig. 15.52 B). There is evidence sug- which is clearly distinguishable in both species. gesting that the angular gyrus may play a role In line with this, Petrides and Pandya [579] in- in the visual guidance of arm movements terpreted the anterior and posterior parts of [254]. the inferior parietal lobule of the rhesus mon- The most prominent connections of the su- key (their areas PF and PG) as homologous to pramarginal gyrus include those to the somato- the human supramarginal and angular gyri, re- sensory areas S1, S2 and area 5, the medial spectively. We fully agree with this interpreta- part of the superior parietal lobule, the vestib- tion because the areas mentioned occupy iden- ular cortex, the anterior cingulate area 24, the tical relative or topological positions in the supplementary motor area, the ventral premo- monkey and the human. Hence, we feel justi- tor cortex, and the adjacent area 44, and pre- fied in extrapolating the results of experimen- frontal areas 45 and 46 (Fig. 15.52 B). Unit ac- tal studies on the fibre connections of the infe- tivity studies in monkeys indicate that the su- rior parietal areas PF and PG in the macaque pramarginal gyrus is involved in the organiza- to the human situation. tion of co-ordinated hand and face movements Wernicke’s speech region extends for an un- [840]. Petrides and Pandya [579] suggested determined distance over the left inferior pari- that the interactions between rostral inferior etal lobule [463, 549]. The core of this region parietal cortex and the ventral frontal region occupies the left temporal plane behind the in the monkey may be necessary for gestural transverse gyrus of Heschl and the posterior communication, which may have preceded the 15 Telencephalon: Neocortex 611 evolution of linguistic communication. An al- but dramatic syndrome of Balint, which is ternative view on the development of brain characterized by (1) simultanagnosia, (2) oculo- mechanisms that gave rise to language has motor apraxia and (3) optic ataxia. Simultanag- been recently presented by Gil-da-Costa et al. nosia is the inability to see all the components of [236]. These authors pointed out that monkeys the visual scene in an integrated way. Oculomo- possess a repertoire of species-specific vocali- tor apraxia or psychic paralysis of gaze is the in- zations that – like human speech – seem to en- ability to voluntarily direct gaze toward a specif- code meaning in arbitrary sound patterns. In- ic part of the visual field. Optic, or visuomotor, terestingly, they found that conspecific calls ac- ataxia is the inability to direct movement of an tivate putative homologues of Wernicke’s and extremity using visual guidance [151]. Broca’s speech regions in the rhesus monkey. In the foregoing, we have divided the posterior parietal cortex provisionally into nine Temporal Lobe functional subregions (Fig. 15.51 C,D). If we survey the data collected on the connections The temporal lobe presents large lateral and ba- and functions of these subregions, it appears sal surfaces and a more limited medial surface. that what has been said about the areas located The lateral surface is dorsally separated from in the banks of the intraparietal sulcus hold the frontal and anterior parietal lobes by the lat- true for the remaining subregions as well. eral sulcus, and more caudally from the posteri- They are all specifically connected with one or or parietal lobe by an arbitrary parietotemporal some sections of premotor cortex, and all are line (Fig. 3.2). The caudal border of the temporal dedicated to the selection and preparation of lobe is formed on the lateral surface by the ante- specific motor patterns, including eye move- rior occipital sulcus and on the basal side by the ments, reaching movement, grasping move- occipitotemporal sulcus and an arbitrary, trans- ments and articulated speech [10, 445, 828]. versely oriented occipitotemporal line (Figs. 1.4, It has already been mentioned that lesions 3.5, 15.53). Ventromedially, the temporal lobe is of the inferior parietal lobule in the left hemi- separated from the limbic lobe by the collateral sphere may lead to Wernicke’s aphasia. When and rhinal sulci (Fig. 3.6). damage spares Wernicke’s speech region, but Two grooves, the superior and inferior tem- involves the adjacent, more superior parts of poral sulci, which roughly parallel the lateral the left inferior parietal lobule, complex com- sulcus, divide the lateral surface of the tempo- binations of anomia (naming disorders), alexia ral lobe into superior, middle and inferior tem- (reading impairment), construction deficits, al- poral gyri. The basal surface of the temporal calculia (difficulty in performing simple arith- lobe is formed by the basal part of the inferior metic calculations), agraphia (loss of writing temporal gyrus and most of the lateral and skills), finger agnosia (inability to recognize, medial occipitotemporal gyri. distinguish and name one’s fingers or those of The superior temporal gyrus includes the other persons) and right-left disorientation temporal operculum because the gyrus extends (inability to name or point to the right and left medially to meet the inferior part of circular sides of objects or body parts) develop. The sulcus around the insula. On the opercular sur- last four symptoms are collectively referred to face of the temporal lobe, which forms the as Gerstmann’s syndrome. floor of the lateral sulcus, there are one or two, Lesions in the inferior parietal lobule of the more or less transversely running convolu- right hemisphere lead to characteristic deficits, tions, the transverse temporal gyri of Heschl. including dressing difficulties, constructional The areas anterior and posterior to these gyri difficulties and multimodal neglect of the left are known as the planum polare and planum hemispace [463]. temporale, respectively (Fig. 18.1 B). Bilateral lesions of the dorsal parts of poste- Brodmann [70, 71] divided the cortex cover- rior parietal lobules may give rise to the rare ing the temporal lobe into eight cytoarchitec- 612 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.53 A–D. The temporal lobe. Subdivision according to Brodmann, lateral (A) and medial (B) views. The temporal lobe is shown in red. Functional areas, lateral (C) and medial (D) views. In A and C, the upper surface of the superior temporal gyrus or temporal operculum is brought to view by rotation. The approximate boundary between the temporal operculum and the lateral surface of the superior temporal gyrus is indicated by a dashed line. A1, primary auditory cortex; AAC, auditory association cortex; ABR, auditory belt region; ACR, auditory core region; APBR, auditory parabelt region; plan pol, planum polare; plan temp, planum temporale, STPC, superior temporal polymodal cortex; TPC, temporopolar cortex; TVAC, temporal visual association cortex; tr gy, transverse gyrus of Heschl; W, Wernicke’s region; 19, 20, etc., areas according to Brodmann 15 Telencephalon: Neocortex 613 tonic areas: 41, 42, 22, 21, 20, 36, 37 and 38 cance of this parcellation remains to be eluci- (Figs. 15.8, 15.53 A,B). Area 41, the primary dated. auditory cortex, is situated on the opercular The auditory association cortex can be di- surface of the temporal lobe. It forms a rather vided into a belt and a parabelt region. The narrow strip which, extending from anterolat- belt region borders on the primary auditory eral to posteromedial, corresponds roughly core, surrounding it anteriorly, laterally and with the (anterior) transverse temporal gyrus posteriorly. Most of it is confined to the tem- of Heschl. Area 42 is located directly posterior poral operculum, but laterally it extends a to area 41 and partly surrounds the latter (Fig. short distance onto the superficial aspect of 15.8 insert). Most of this area is opercular, but the superior temporal gyrus. The belt region its lateral part extends for some distance over consists of several cytoarchitectonic fields, the free lateral surface of the superior temporal among them Brodmann’s area 42 [220]. The gyrus. Areas 22, 21 and 20 were designated by structure of the belt areas is intermediate be- Brodmann as area temporalis superior, -media tween the typical koniocortex of area 41 and and -inferior, respectively, indicating that they the homotypical structure of area 22 and the roughly correspond to the gyri of the same remainder of the temporal lobe. name. Area 20 extends basally over the ante- The belt region is surrounded in turn by an rior part of the lateral occipitotemporal gyrus. extensive parabelt region that covers the re- Area 36 is situated on the medial surface of the maining anterior and posterior parts of the temporal lobe. It is bounded superiorly by the temporal operculum and the lateral surface of rhinal and collateral sulci and inferiorly by the superior temporal gyrus with the exception area 20. Area 37 occupies the posterior part of of its anterior pole. It corresponds largely with the temporal lobe, extending over its lateral Brodmann’s area 22. and basomedial surface. Area 38, finally, covers Short association projections connect the the temporal pole. auditory core with the belt and the latter with In what follows, the temporal lobe will be the parabelt cortex [273, 485, 486, 528, 670]. divided into the following five functional re- Neuroimaging studies on the response to audi- gions: (1) the primary auditory cortex, (2) the tory stimuli show a spread of activity, begin- auditory association cortex, including Wer- ning with activation by simple stimuli of a lim- nicke’s region, (3) the temporal visual associa- ited region of the temporal operculum, and ex- tion cortex, (4) the superior temporal polymo- panding with more complex stimuli to the sur- dal cortex and (5) the temporopolar cortex face of the superior temporal gyrus [478]. (Fig. 15.53 C,D). The auditory system, includ- Commissural connections of the auditory ing the auditory cortices, will be extensively cortex are predominantly homotopical. Lesions discussed in Chap. 18; hence we confine the in the core, belt or parabelt result in degener- discussion here on these subjects to a brief ating fibres in corresponding areas in the con- overview. tralateral hemisphere [272, 486, 670]. The audi- The primary auditory cortex, or auditory tory commissural fibres are located in the pos- core region, corresponds with area 41 of Brod- terior part of the truncus of the corpus callo- mann. It has the typical koniocortical structure sum, together with the commissural fibres of of primary sensory cortices, with a well-devel- the parietal lobe [531]. oped layer IV, consisting of densely packed The auditory association cortex in the poste- granule cells. Morosan et al. [489] recently sub- rior parabelt region grades on the left side into divided the primary auditory cortex into pos- the heteromodal cortex of Wernicke’s speech teromedial, central and anterolateral subareas, region. The precise architectonic borders of which they designated Te1.1, Te1.0 and Te1.2, this region are not known, but it is generally respectively. A similar tripartitioning of the assumed that it includes the posterior portions primary auditory cortex was proposed for the of the planum temporale and superior tempo- monkey (Fig. 18.3 g). The functional signifi- ral gyrus and the most basal parts of the an- 614 Section II Structure of Spinal Cord and Brain Parts gular and supramarginal gyri [26, 398, 465]. sively to visual stimulation and not to stimula- Interruption of the auditory, visual or tactile tion in other sensory modalities [150]. Many projections to Wernicke’s region lead to pure of these neurons respond selectively to various word deafness, pure alexia and tactile aphasia, complex visual features of objects [255, 730]. respectively [227] (Fig. 15.20). As already men- Recordings of field potentials from the surface tioned, damage to Wernicke’s region itself of the brain in patients have shown that an causes receptive or Wernicke’s aphasia. area in the anterobasal temporal lobe is specifi- Experimental studies in rhesus macaques cally involved in the processing of written [274, 576, 579, 631, 632] have shown that the words [512]. auditory association cortex is strongly and re- The inferotemporal cortex, including FFA ciprocally connected with the prefrontal cortex. but excluding MT and MIT, forms part of the Romanski et al. [631, 632] found that these ventral, occipitotemporal visual processing audito-prefrontal connections form two stream, which is concerned with pattern discri- streams, anterior and posterior, connecting dif- mination and visual identification of objects. ferent sectors of the auditory association cor- The inferior longitudinal fascicle, which forms tex with largely different prefrontal regions. the axis of this processing stream, contains, The anterior stream connects the anterior apart from long occipitotemporal projections, belt and parabelt cortex reciprocally with the numerous short fibres that sequentially con- frontal pole (area 10), the anterior part of area nect adjacent regions in the striate, peristriate 46 and a ventral prefrontal region, including and inferotemporal cortices [155, 579, 715, 739, areas 12 and 45. In contrast, the posterior 809] (Fig. 15.45). Efferents from the basal stream mainly interconnects the posterior belt amygdaloid nucleus pass caudally in or close and parabelt cortex with the frontal eye field to the inferior longitudinal fascicle, to termi- (area 8) and the posterior part of area 46 (Fig. nate in the various cortical regions forming 15.54 A). Romanski et al. [632] pointed out that part of the ventral visual processing stream. the regions receiving input from the posterior These amygdalar efferents terminate mainly in belt and parabelt are implicated in spatial pro- the most superficial and deep layers of their cessing, whereas those receiving input from cortical targets and hence resemble cortico- the anterior belt and parabelt are implicated in cortical feed-back fibres [204]. non-spatial, higher functions. They drew a par- Efferents from the inferotemporal cortical allel with the central visual system, where sep- region, particularly its rostral parts, relay ex- arate processing streams, the dorsal “where” tensively processed visual information to: (1) stream and the ventral “what” stream, follow the adjacent superior temporal polymodal re- different courses and terminate ultimately in gion [127]; (2) limbic cortical regions, includ- roughly the same dorsolateral spatial and ven- ing the temporopolar cortex, the perirhinal trolateral object-processing regions of the fron- area 36 and the subiculum [647, 774]; (3) the tal lobe [822]. lateral amygdaloid nucleus [204, 298]; and (4) The temporal visual association cortex repre- the prefrontal cortex. The projection to the lat- sents a rostral extension of the circumstriate ter passes via the uncinate fasciculus and ter- belt, below the superior temporal sulcus. It in- minates mainly in areas 11 and 47/12 [579] cludes areas 20, 21 and 37 of Brodmann, which (Fig. 15.54 B). Commissural fibres from the in- may be collectively designated as the infero- ferotemporal cortex cross in the posterior part temporal visual cortex. The posterior part of of the body of the corpus callosum and in the this region houses several functionally defined anterior commissure [531]. areas, including the middle temporal visual Experimental neuroanatomical studies have area, MT, the middle superior temporal area, shown that the cortex of the upper bank of the MST, and the fusiform face area, FFA (Fig. superior temporal sulcus in the rhesus monkey 15.49 C). In rhesus macaques, neurons in the contains an elongated region that receives con- inferotemporal cortex respond almost exclu- verging input from surrounding auditory, vis- 15 Telencephalon: Neocortex 615

Fig. 15.54A–C. Principal connections of the temporal lobe. Cortical auditory pathways (A); connections of the temporal visual association cortex (B) and connections of the superior temporal polymodal cortex (C). AAPFS, anterior audito-prefrontal processing stream; ABR, auditory belt region; ACR, auditory core region; APBR, auditory parabelt region; FF, frontal eye field; ILF, inferior longitudinal fasciculus; IPL, inferior parietal lobule; LIP,lateral intraparietal area; PAPFS, posterior audito-prefrontal processing stream; SPLm, medial part of superior parietal lobule; STPC, superior temporal polymodal cortex; TPC, temporopolar cortex; TVAC, temporal visual association cortex; UNCF, uncinate fasciculus; VVPS, ventral visual processing stream; 6, 8, 9, etc., areas according to Brodmann 616 Section II Structure of Spinal Cord and Brain Parts ual and somatosensory association areas [346, stream, whereas the inferotemporal cortex be- 670, 675]. Electrophysiological studies have re- longs to the ventral stream. vealed that neurons in this region respond to Somatosensory afferents to the STP cortex auditory, visual or somatosensory stimulation, emanate reportedly from mid-portions of the including some whose response is bi- or tri- inferior parietal lobule and the medial parietal modal [42, 47, 74, 150, 299, 472, 551, 707]. lobe [530, 671]. These findings have led to the conclusion that In addition to the sensory inputs just dis- the sulcal cortex in question is polymodal in cussed, the STP cortex has extensive afferent nature and should therefore be designated as connections with several limbic cortical struc- the superior temporal polymodal or STP cortex tures. As with cortical sensory input, these [150, 346, 670]. projections have a rostral-to-caudal topograph- The elongated strip of STP cortex has been ical organization. Projections from parahippo- divided on cytoarchitectonic and chemoarchi- campal area 35, for example, target area TPO1; tectonic grounds into four rostrocaudally ar- those from entorhinal area 28 target mid-area ranged units, termed TPO1–TPO4 [125, 127, TPO, whereas area TPO4 receives input from 670, 673, 676]. These units are tied together in the insular limbic cortex. Furthermore, areas a sequence of reciprocal connections, compris- TPO2–4 have afferent connections with ante- ing shorter and longer, rostrally directed feed- rior cingulate area 24, posterior cingulate area forward and caudally directed feed-back pro- 23, and retrosplenial areas 29 and 30. Seltzer jections [673]. and Pandya [675], who described the limbic The extrinsic connections of the STP cortex connections just mentioned, pointed out that can be divided into a post-Rolandic and a pre- certain sets of cortical areas projecting to the Rolandic group. The post-Rolandic connections same rostrocaudal segment of the STP cortex relate the STP cortex to auditory, visual and are themselves interconnected. They mention somatosensory association cortices and to sev- as examples that the parahippocampal area 35 eral limbic areas. and rostral superior temporal cortex, which Auditory afferents to the STP cortex origi- target area TPO1, have strong reciprocal con- nate from the superior temporal cortex. They nections and that the same holds true for the show a topical organization in that the caudal, entorhinal area 28 and middle superior tem- middle and rostral zones of the superior temporal cortex, which target areas TPO2–3. poral cortex project to areas TPO4, TPO2–3 Almost all of the sensory and limbic afferent and TPO1, respectively [675]. systems to the various sectors of the STP cor- Visual projections to the STP cortex include tex are reciprocated by efferent systems [37, afferents from (1) the LIP area in the intrapar- 674]. ietal sulcus, which terminate in the TPO4, (2) The pre-Rolandic connections of the STP the caudal inferior parietal lobule, terminating cortex are organized according to the rostral- in areas TPO2–4, (3) the caudal inferotemporal to-caudal differentiation of this multimodal re- cortex, targeting areas TPO3–4, and (4) the gion. Area TPO1 projects to basal (areas 13, rostral inferotemporal cortex terminating in 12, 11 and 14), medial (areas 24, 32 and 9) and area TPO1. The inferotemporal cortex also pro- lateral (areas 10 and 46) sectors of the frontal jects indirectly to area TPO1 via some visual lobe. Areas TPO2–3 project to rostral subdivi- association areas, known as IPa, TEa and sions of the lateral prefrontal cortex, namely TEm, situated in the lateral bank of the superi- areas 46, 9 and 10, whereas area TPO4 projects or temporal sulcus [155, 638, 647, 675, 772]. It to caudal subdivisions (areas 46, 8 and 6) of is noteworthy that the visual projections con- the lateral frontal lobe (Fig. 15.54 C). All of verging on the STP cortex originate from areas these efferent projections, except for those to forming part of both the dorsal and the ventral the basal (orbitofrontal) cortex, are recipro- visual streams. Thus, the LIP area and the infe- cated by afferent systems [672]. Luppino et al. rior parietal lobule belong to the dorsal [432] reported that not only the caudal, but 15 Telencephalon: Neocortex 617 also the rostral part of the STP cortex projects The temporopolar cortex receives afferents to the premotor cortex (area 6). from the adjacent auditory and visual associa- The areal organization of the cortex sur- tion areas and has reciprocal connections with rounding the human superior temporal sulcus numerous paralimbic and limbic areas, includ- has not been explored so far. However, it ing the caudal orbitofrontal cortex, anterior in- seems reasonable to assume that this cortex, sula, entorhinal cortex, and the subicular com- which forms the border region of areas 21 and plex [439, 479] (Fig. 15.56). 22 of Brodmann, contains a polymodal zone It has been demonstrated that single neu- comparable to that in non-human primates rons in the temporopolar cortex of macaque [463]. Recent brain-imaging studies, summa- monkeys exhibit oscillatory activities in re- rized by Zilbovicius et al. [848], have shown sponse to the representation of complex visual that in humans, the superior temporal sulcal stimuli such as photographs of familiar human (STS) cortex is involved in the processing and faces, familiar foods, and familiar non-food integration of complex visual and auditory in- objects [501, 502]. Bilateral lesions of the hu- formation, conducive to understanding of the man temporopolar cortex cause marked retro- mental states and intentions of other individ- grade amnesia, which includes the loss of all uals. The information involved in this social remote memories, but spares anterograde perception includes eye gazes, gestures, facial learning and recent memory [359, 360]. Typi- displays of emotions and voice perception. Zil- cally, patients with these lesions not only lose bovicius et al. [848] also cite brain-imaging re- episodic, autobiographical memories, but also sults, suggesting that in autism, a mental dis- the ability to recognize faces and names of fa- order in which communication deficits are mous persons learned in the remote past [359, most prominent, anatomical and functional ab- 731]. Sewards and Sewards [678] inferred from normalities in the STS cortex are implicated. these lesion data that the neuronal activities The temporopolar cortex, which corresponds providing the awareness of remote (consoli- to area 38 of Brodmann, is commonly designated dated) object recognition occur in the temp- as a paralimbic structure. Morphologically, it oropolar cortex. The fact that patients with ret- lies beyond the limbic lobe, but structurally it rograde recognition memory loss due to le- forms part of the so-called paralimbic belt. As sions of the temporopolar cortex do not lose discussed in Chap. 11 (and in the next section the ability to recognize objects learned in the of the present chapter), the limbic lobe forms recent past, nor the ability to learn to recog- a large arcuate convolution on the medial aspect nize new objects, indicated, according to Sew- of the hemisphere, which extends from the fron- ards and Sewards, that there must be at least tal, via the parietal, into the temporal lobe. As one other cortical area in which neuronal ac- shown by Fig. 12.3, its frontal and temporal ends tivities produce recognition awareness. They are formed by infralimbic area 25 and perirhinal adduce clinical and experimental evidence in- areas 35 and 36, respectively. According to Mesu- dicating that such a centre, providing recogni- lam [463], the gap between the ends of the limbic tion awareness of objects learned in the recent lobe is bridged by three structures, the caudal past, does exist and is located in another para- orbitofrontal cortex (the posterior parts of areas limbic region, the medial orbitofrontal cortex. 12, 13, 14), the anterior insula, and the temporopolar cortex (Fig. 15.55). Together with the cortices covering the various parts of the limbic Limbic Lobe and Paralimbic Belt lobe, these structures form the paralimbic belt. Architectonically, the belt areas provide a con- The limbic lobe, as defined by Broca [68], continuous transition zone between the simple, sists of a large, arciform convolution on the three-layered olfactory and hippocampal cor- medial aspect of the cerebral hemisphere, tices, on the one hand, and the surrounding neo- which surrounds the interhemispheric commis- cortical regions, on the other [463, 479]. sures and the upper brain stem (Fig. 1.4 B). 618 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.55. Brodmann’s cytoarchitectonic map of the medial hemisphere wall. Paralimbic areas are shown in red. The insert represents a small part of the lateral surface of the hemisphere. The lateral sulcus is opened to expose the insula. In combination with the main figure, it illustrates that the caudal orbitofrontal cortex, anterior insula and temporopolar cortex bridge the gap between the prelimbic area 25 and perirhinal areas 35 and 36, thus closing the paralimbic belt. Based on Mesulam [463]. HF, hippocampal formation; Iant, post, anterior and posterior insula; OFc, caudal orbitofrontal cortex; TP, temporopolar cortex 15 Telencephalon: Neocortex 619

29,30

Fig. 15.56. Intrinsic cortical connections of structures forming the paralimbic belt. Particularly strongly developed projections are indicated by heavy lines. HF, hippocampal formation; Iant, anterior insula; TH, TF, medial temporal cortical areas according to von Economo and Koskinas [796]. 12, 13 etc., areas according to Brodmann 620 Section II Structure of Spinal Cord and Brain Parts

The limbic lobe includes the cingulate and Frontal Lobe parahippocampal gyri and also the hippocampal formation. Cytoarchitectonically, the limbic The frontal lobe is strongly developed and lobe encompasses the infralimbic cortex (area comprises approximately one third of the en- 25), the anterior and posterior cingulate cor- tire hemisphere surface. On the lateral, convex tices (areas 24 and 23), the retrosplenial cortex surface it extends from the central sulcus to (areas 26, 29 and 30), the perirhinal cortex the frontal pole. Basolaterally, it is separated (areas 35 and 36), the entorhinal cortex (area from the temporal lobe by the lateral sulcus. 28) and the various parts of the hippocampal On the medial surface, the frontal lobe is sepa- formation (dentate gyrus, Ammon’s horn and rated from the limbic lobe by the cingulate sul- subicular complex). Von Economo and Koski- cus, and from the parietal lobe by an arbitrary nas [796] demonstrated that a region roughly vertical line connecting the superior limit of corresponding to the posterior parts of Brod- the central sulcus with the cingulate sulcus. mann’s areas 35 and 36 contains two distinct The superolateral surface of the frontal lobe separate entities, a superior area TH and an in- is traversed by three sulci, the precentral sul- ferior area TF. As discussed in the previous cus and the superior and inferior frontal sulci. section, the cytoarchitectonic areas just enu- The precentral sulcus runs parallel to the cen- merated constitute, together with the caudal tral sulcus. These two sulci border the precen- orbitofrontal cortex (areas 12, 13, 14 and cau- tral gyrus. The arciform superior and inferior dal part of area 11), the anterior insula and the frontal sulci divide the convex surface in front temporopolar cortex (area 38), the paralimbic of the precentral gyrus into three convolutions, belt (Fig. 15.55). the superior, middle and inferior frontal gyri The intrinsic [55, 96, 152, 324, 377, 397, 439, (Fig. 3.2). Short anterior and ascending 479, 482, 717, 772, 782, 843] and extrinsic con- branches of the lateral sulcus divide the inferi- nections of the various (para)limbic cortical or frontal gyrus into three parts: pars opercu- structures [96, 152, 323, 324, 346, 377, 380, laris, pars triangularis and pars orbitalis (Fig. 397, 441, 480, 527, 579, 639, 716, 772, 782, 843] 3.2). These areas in the dominant hemisphere have been extensively studied in the rhesus (usually the left in right-handed individuals) monkey. The principal results of these studies correspond to the region of Broca [67] associ- are summarized and extrapolated to the hu- ated with the motor aspects of speech. man brain in Figs. 15.56 and 15.57. It will be The basal surface of the frontal lobe overlies seen that the various belt structures (a) are the bony orbit; hence, the cortex in this region strongly and reciprocally interconnected (Fig. is designated as the orbitofrontal cortex. The 15.56); (b) receive extrinsic cortical afferents olfactory bulb and tract lie in a longitudinal from numerous unimodal (7, 19, 20, 21, 22) sulcus near the medial margin of the frontal and polymodal (STSC, 9, 10, 11, 46) sensory lobe, known as the olfactory sulcus (Figs. 3.4, areas (Fig. 15.57 A); and (c) send their extrin- 3.5). The concave area lateral to the olfactory sic cortical efferents mainly to the same asso- bulb and tract bears a number of sulci, which ciation areas (Fig. 15.57 B). For a further dis- together form an H-shaped configuration. cussion of the limbic lobe, the reader is re- These sulci divide this area into four orbital ferred to Chap. 12. The limbic lobe and the gyri, anterior, lateral, posterior and medial. more extensive paralimbic belt form part of a The latter is separated from a fifth orbital large functional entity, designated as the great- gyrus, the gyrus rectus, by the olfactory sulcus er limbic system [510]. This functional system (Figs. 3.4, 3.5, 15.5 C,G). will be treated in the final chapter (Chap. 23) The cytoarchitectonic subdivision of the of the present work. frontal lobe, according to Brodmann [70, 71], is shown in Fig. 15.58 A,B. Among the 14 frontal areas distinguished by that author, there are agranular (4, 6, 24, 25, 32), dysgranular (8, 15 Telencephalon: Neocortex 621

Fig. 15.57 A,B. Extrinsic cortical connections of paralimbic structures. A Afferents; B efferents. STPC, superi- or temporal polymodal cortex. Numbers and other abbreviations as in Fig. 15.56 622 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.58 A–D. The frontal lobe. Subdivision according to Brodmann, lateral (A) and medial (B) views. The frontal lobe is shown in red. Functional areas, lateral (C) and medial (D) views. B, Broca’s speech region; cc, corpus callosum; CAMc,r, caudal and rostral cingulate motor area; DLPFC, dorsolateral prefrontal cortex; F1– F7, subdivisions of motor cortex; FF, frontal eye field; FPC, frontopolar cortex; M1, primary motor cortex; MPFC, medial prefrontal cortex; OPFC, orbital prefrontal cortex; pre-SMA, pre-supplementary motor cortex; SF, supplementary eye field; SMA, supplementary motor area; VLPFC, ventrolateral prefrontal cortex 15 Telencephalon: Neocortex 623

44, 45), as well as granular areas (9, 10, 11, 12, Area 3 a is represented by a narrow strip of 46, 47) (Fig. 15.7 A,F). According to the cytoar- cortex, located in the fundus of the central sul- chitectonic typology of von Economo [795], cus. Its cytoarchitecture is similar to that of areas 4, 6, 24 and 25 are heterotypical and be- area 4, but it receives information from muscle long to type 1, whereas all of the remaining receptors through the thalamus. Its cortical frontal areas are homotypical. Of these, area 46 connections closely resemble those of the pri- and part of area 10 belong to type 3, the orbi- mary motor cortex. tofrontal areas 11, 12 and 47 to type 4 and the The primary motor cortex receives afferents remaining homotypical frontal areas to type 2 from both subcortical and cortical sources. (Fig. 15.10). The subcortical afferents arise principally from The frontal cortex can be divided into two the contralateral dentate nucleus, via the poste- large functional domains, the motor cortex rior part of the ventral lateral thalamic nucleus and the (associative) prefrontal cortex. The and from the ipsilateral globus pallidus, via the motor cortex is situated in front of the central anterior part of the same thalamic nucleus sulcus and extends over the medial surface of (Fig. 14.8). The cortical afferents to the pri- the hemisphere. The prefrontal cortex occupies mary motor cortex originate from several non- the large region that lies rostral to the precen- primary motor areas [164, 475], including the tral motor cortex [581] (Fig. 15.58 C,D). dorsal and ventral premotor areas [481, 711], The motor cortex comprises the primary the supplementary motor area [481] and the motor cortex (M1) and the non-primary motor cingulate motor area [480], and from the pri- cortex. Matelli et al. [443, 445] have subdivided mary somatosensory cortex (S1) and the soma- the somatic motor cortex of the rhesus monkey tosensory association cortex (area 5) [316, 542, into seven areas designated F1–F7. This parcel- 587]. The projection from S1, which arises lation was based on cytoarchitectonic, histo- from areas 1, 2 and 3b, is topographically or- chemical, neurochemical, hodological and ganized. The somatosensory association area is functional data. Modern architectonic analyses primarily concerned with the analysis of pro- and functional evidence, mostly from neuroim- prioceptive information. It seems likely that aging studies, indicate that the organization of the efferents from this association area to the the human motor cortex closely resembles that primary motor cortex provide the latter with of the rhesus monkey. Indeed, almost all of the information on the localization of body parts, structural/functional entities distinguished in necessary for the control of limb movements. the former can also be identified in the latter The primary motor cortex contributes sub- [188, 445, 585]. Hence, the following parcella- stantially to the pyramidal tract. For a detailed tion of the motor cortex applies to both spe- discussion of this fibre system, the reader is cies (Fig. 21.6). referred to Chap. 21 and Fig. 21.9. Here, we Area F1 (or M1) represents the primary motor confine the discussion to mentioning that this cortex. It corresponds to area 4 of Brodmann, tract arises principally from the motor and so- which is characterized by the presence of the matosensory cortical areas surrounding the giant pyramidal cells of Betz (Figs. 15.7 F, 15.28 central sulcus. Its fibres originate from pyrami- (6). Lassek [395] counted approximately 34,000 dal neurons situated in cortical layer V and via Betz cells in area 4 of the human brain. the internal capsule descend to the brain stem Electrical stimulation studies of the primary and spinal cord. The pyramidal fibres originat- motor cortex in both man [197, 544, 545, 548] ing from the precentral gyrus (areas 3a and 4) and experimental animals [261, 784, 785, 830, pass to premotor interneurons and also direct- 831] have revealed the presence of a topo- ly to motoneurons. The direct corticomoto- graphical map within that cortex of the con- neuronal connections are established by coarse tralateral body half, comparable to that found fibres, derived from the giant cells of Betz. in the adjacent primary sensory area (Figs. These direct pyramidal tract connections to 15.18 C, 15.19). motoneurons, especially to those innervating 624 Section II Structure of Spinal Cord and Brain Parts distal extremity muscles, appear to provide the The medial portion of Brodmann’s area 6 is capacity to execute highly fractionated move- occupied by two functional areas, the caudal ments, as typified by independent movements supplementary motor area (SMA) or area F7, of the digits [390]. which is also known as M2, and the rostral It has long been thought that the relation presupplementary motor area (pre-SMA) or between the primary motor cortex and the area F6 [447] (Fig. 15.58 D). The medial wall of skeletal muscular system is fixed and that this the hemisphere contains two further areas re- cortex is exclusively concerned with the execu- lated to motor control. These areas, which may tion of movements and is not responsible for be designated as the caudal and rostral cingu- the design of movement patterns. Both con- late motor areas, CMAc and CMAr, are buried cepts have appeared to be incorrect. Recent in the cingulate sulcus [188, 585]. The CMAc, evidence, summarized by Graziano [252], has which borders on the SMA and the rostral part shown that the connectivity between primary of the primary motor cortex, forms a subfield motor cortex and muscles is not fixed but plas- of Brodmann’s area 24. It contains a population tic, changing constantly on the basis of feed- of gigantopyramidal neurons, similar to those back from the periphery. Moreover, it has been in the adjacent primary motor cortex [62]. The shown that with electrical stimulation trains of CMAr is located roughly at the same rostro- longer duration than used in the classical map- caudal level as the pre-SMA (Fig. 15.58 D). It ping studies cited above (500 ms instead of 50 also forms part of Brodmann’s area 24 but ms), complex movements resembling meaning- probably encroaches upon area 32. CMAr and ful actions, such as putting the hand to the CMAc are also denoted as M3 and M4, respec- mouth, defensive gestures, reaching motions tively [482, 483]. and shaping the hand as if to grasp an object, Classically, the central skeletomotor system can be elicited from the precentral cortex (Fig. was looked upon as strictly hierarchical. The 21.11 C). efferent pyramidal neurons in the primary mo- Area 6 of Brodmann lies immediately ante- tor cortex were collectively designated as the rior to the primary motor cortex. Just like the “upper motoneuron” or “final common path- latter, it extends on the medial wall of the way” for the central control of movements, im- hemisphere (Fig. 15.58 A,B). The portion of pinging either directly or indirectly (i.e. via in- area 6 situated on the lateral surface of the terneurons) on the “lower motoneurons” in the hemisphere corresponds to the premotor cortex brain stem and spinal cord. Projections from (PM), which can be subdivided into a dorsal other motor-related areas, converging upon the (PMD) and a ventral zone (PMV). Both of primary motor cortex, were thought to repre- these zones can be further subdivided into sent a hierarchical level superimposed on that caudal and rostral parts. The caudal part of of the “upper motoneuronal system”. Recent the ventral premotor zone, PMVc or area F4,is experimental neuroanatomical studies, mainly located directly in front of the representation in rhesus monkeys, have shown that the orga- of orofacial movements in the primary motor nization of the central motor system is much cortex. The rostral part of the ventral premotor more complex than indicated by the hierarchi- zone, PMVr or area F5, in the rhesus monkey cal model just sketched. It has been shown that contains two different functional areas, a cau- the motor part of the pyramidal tract, i.e. the dal area F5ab and a more rostrally situated “upper motoneuronal system”, does not origi- area F5c. It seems likely that the most rostro- nate exclusively from the primary motor cor- ventral part of area 6 of non-human primates tex. Rather, this fibre system has appeared to is homologous to Broca’s speech region [445] originate from many of the non-primary motor in humans. The dorsal premotor zone (PMD) areas discussed above, including the PMVc can be structurally and functionally subdivided and PMDc [162, 282], the SMA [163, 283] and into a caudal PMDc or area F2 and a rostral the CMAr and CMAc [162, 163, 283]. Conse- PMDr or area F7 (Fig. 15.58 C). quently, all of these areas have the potential to 15 Telencephalon: Neocortex 625 influence the generation and control of move- specifically and reciprocally connected with a ments independently of the primary motor particular frontal motor area. The various par- cortex. All of the areas giving rise to pyrami- ietal centres and their frontal targets form dal tract fibres also project to the primary mo- functional circuits, each of which is dedicated tor cortex [131, 431, 475, 480, 682, 711]. Thus, to a particular aspect of sensory-motor trans- they influence motor activity through at least formation [445, 828, 847]. These parietofrontal two pathways, corticospinal projections and circuits have been briefly dealt with in the sec- corticocortical projections to the primary motion on the parietal lobe of the present chapter tor cortex. The various non-primary motor and will be further discussed in Chap. 21. areas mentioned, which are all somatotopically The areas forming the rostral group of Ma- organized [282, 283, 480], do not project only telli et al. [445], i.e. the pre-SMA or F6 and the to M1, but they are also strongly and recipro- PMDr or F7, are tightly connected to the pre- cally interconnected [131, 431, 480, 711]. Dum frontal cortex, particularly area 46, and to the and Strick [164] recently studied the digit rep- caudal motor areas [419, 431, 433, 475]. They resentations in M1, PMD, PMV and SMA of ce- are believed to play a role in cognitive aspects bus monkeys. They concluded that the subar- of motor control such as temporal planning of eas of the various clusters concerned with the actions and motivation [445]. generation and control of hand movements The cingulate motor areas do not fit into the form a densely interconnected network, within categorization of Matelli et al. [445]. Just like which a clear hierarchical organization is lack- the premotor areas of the caudal group, CMAr ing. and CMAc are reciprocally connected with the Matelli et al. [445] have pointed out that the primary motor cortex [480, 481] and project various motor areas can be grouped into two directly to the spinal cord [163, 283]. However, major classes. The caudal motor areas F1, F2, unlike the caudal group and conforming with F3, F4 and F5 form the first class. They are the rostral group, CMAr and CMAc receive a typified by their projections to the spinal cord, substantial projection from the prefrontal cor- to F1 and to each other. The rostral areas F6 tex [481]. The two cingulate motor areas are and F7 form the second class. These areas do strongly interconnected [482], and neither pro- not project to the spinal cord, but their des- jects only to the primary motor cortex (M1), cending output terminates in various part of but also to the supplementary motor area (M2) the brain stem. Furthermore, these areas are [480]. All of these connections are somatotopi- not directly connected with F1. According to cally organized. Matelli et al. [445], the two classes also differ The prefrontal afferents to the cingulate mo- with respect to the sources of their principal tor areas comprise a strong projection from “extrinsic” cortical inputs. The areas forming the dorsolateral prefrontal cortex and less sub- the caudal group receive a primary cortical in- stantial projections from the ventrolateral pre- put from the parietal lobe; therefore, they may frontal and caudal orbitofrontal cortices [481]. be characterized as parieto-dependent motor It is important to note that the cingulate motor areas. The rostral areas, on the other hand, re- areas, in addition to the prefrontal and motor- ceive their primary cortical input from the pre- related cortical afferents discussed above, also frontal cortex and may, hence, be characterized receive inputs from diverse and widespread as prefronto-dependent motor areas. The areas limbic cortical areas, including cingulate areas forming the caudal group collectively corre- 24, 23 and 32, retrosplenial areas 29 and 30 spond to Mesulam’s [463] motor association and temporal areas 35, TF and TH [482]. area (Fig. 15.15). As regards the relationship The functional roles of CMAr and CMAc are with the parietal lobe, we have seen that this not well understood. Neuroimaging studies lobe contains a number of centres in which during motor tasks, involving the execution of sensory information from various sources is various arm movements, have shown that processed, and that each of these centres is CMAr is activated in relation to complex tasks, 626 Section II Structure of Spinal Cord and Brain Parts whereas CMAc is activated during simpler most or all of the other eye fields; (2) most of tasks [585]. Morecraft and Van Hoesen [482] the eye fields receive direct input from several suggested that the cingulate motor areas may regions of visual association cortex; (3) in each play an important role in advancing emotional, of these fields electrical stimulation evokes eye motivational and memory-related information movements; (4) surgical lesions or chemical in- generated in the domain of the limbic lobe di- activation of each field produces transient im- rectly to the neocortical motor areas and, pairments of eye movements; and (5) each hence, to the voluntary motor system. field demonstrates increased activity during The non-primary motor cortex also com- eye movement tasks in functional imaging ex- prises the frontal eye field and the supplemen- periments in humans. Lynch and Tian also re- tary eye field which, as their names indicate, port that the saccade subregion of the frontal are both involved in the control of eye move- eye field (FFsac) and the pursuit region of the ments, and Broca’s speech region (Fig. 15.58 C). same field (FFsem) are selectively connected The frontal eye field (FF) is situated at the with distinct subregions in each of the other caudal end of the middle frontal gyrus, in the eye fields. On that account they propose that vicinity of the precentral sulcus [653]. In Brod- there are two parallel cortical oculomotor net- mann’s map, this region is situated within the works, one devoted to the control of saccadic confines of area 6. However, Foerster [197] eye movements and a second devoted to the identified FF with a subfield of area 8, desig- control of pursuit eye movements. nated by Vogt and Vogt [784] as area 8abd The network mediating gaze control encom- (Fig. 15.18 C). According to Rosano et al. [633], passes, apart from the cortical eye fields just FF is predominantly situated just anterior to discussed, numerous subcortical structures, in- area 6, in a transition region between area 6 cluding the thalamus, the basal ganglia, the and area 8 and extending into area 8 proper. cerebellum, the superior colliculus, the para- FF is functionally divisible into a rostral subre- median pontine reticular formation and, as a gion concerned with the generation of saccades matter of course, the various oculomotor nu- (FF sac) and a caudal subregion concerned clei. The structural and functional organization with the generation of smooth eye movements of this network will be discussed in Chap. 19. (FF sem) [434] (Fig. 19.14 A). Broca’s speech region is located in the oper- The supplementary eye field (SF) is located cular and triangular parts of the inferior fron- in the rostrodorsal part of a subfield of Brod- tal gyrus of the dominant (generally the left) mann’s area 6, designated by Vogt and Vogt hemisphere (Fig. 3.2). It is widely accepted that [784] as area 6ab (Fig. 15.58 C) and by Matelli areas 44 and 45 of Brodmann constitute the et al. [445] as a part of PMDr or F7 (Fig. cytoarchitectonic correlates of Broca’s region 15.58 C). [2, 6, 8, 70, 71, 746, 849] (Fig. 15.58 A,C). Both The neocortex contains a number of highly areas contain a thin, inconspicuous inner gran- interconnected regions that make direct contri- ular layer and can, hence, be classified as dys- butions to the initiation and control of volun- granular [6]. The location and extent of areas tary eye movements. In addition to FF and SF, 44 and 45 vary among subjects [383] (Fig. these regions include the medial part of area 7, 15.22 A). A quantitative study of ten different the lateral intraparietal area (LIP), the middle brains [6] revealed that, although the volume superior temporal visual area (MST) and the of area 44 varied considerably among subjects, prefrontal eye field (PF), forming part of area the volume of this area was larger on the left 46 of Brodmann (Figs. 15.50, 15.51, 15.52 A). than on the right side in all ten brains. Area In a recent review, Lynch and Tian [434] sum- 45 did not show such interhemispheric differ- marized our current knowledge of the connec- ences. It seems likely that during evolution, tions and functions of the cortical eye fields areas 44 and 45 have developed from a primor- mentioned above. They point out that: (1) each dium situated in the ventral premotor region of these eye fields is reciprocally connected to in non-human primates [2, 591]. 15 Telencephalon: Neocortex 627

Damage to Broca’s region generally leads to a The anatomical structures and connections motor aphasia. This aphasia, which is also just discussed (and the clinical syndromes at- known as expressive or Broca’s aphasia, consists tached to their damage or interruption) together of a cluster of linguistic features including non- form the classical or Wernicke-Geschwind mod- fluent, effortful speech, impaired repetition, and el of the neural architecture of language [48, 227, relatively preserved comprehension [26]. Non- 228, 811, 812] (Fig. 15.59). During the past de- fluency itself, apart from aspontaneous, sparse cades an enormous body of literature dealing and slow speech, includes reduced or absence with cerebral language organization and its dis- of grammaticality and reduced number of words orders has accumulated. A detailed discussion of per utterance (generally one or a few, expressed this literature is beyond the scope of the present in a “telegraphic” style) [26, 151]. work. Here, we only present a brief survey of Broca’s anterior speech region is strongly con- some of the main results. nected with Wernicke’s posterior speech region 1. The classical model of the neural organiza- and projects to the primary motor cortex. The tion of language was exclusively based on post- fibres interconnecting Broca’s and Wernicke’s re- mortem studies on the brains of patients who gions follow two different routes, dorsal and ven- had suffered from disturbances in comprehend- tral. The fibres following the dorsal route pass ing and/or producing articulated speech. Nu- dorsally on leaving Wernicke’s region, arch merous later studies, based on diverse ap- around the posterior part of the lateral sulcus, proaches, such as intra-operative electrical stim- and then travel rostrally beneath the supramar- ulation of the brain [514, 547, 549], application ginal and somatosensory areas in the parietal op- of radioisotopes for the localization of brain in- erculum, to reach Broca’s region. These fibres farcts [366], structural neuroimaging in patients form part of the so-called arcuate fasciculus, with aphasic disorders (e.g. [385]) and func- i.e. the anterior limb of the superior longitudinal tional neuroimaging in normal subjects during fascicle [145, 398, 532] (Figs. 15.45, 15.46, 15.59). the execution of linguistic tasks (e.g. [200, 451, The fibres following the ventral route pass direct- 593]), have all confirmed that the posterior part ly forward from Wernicke’s region and reach Bro- of the left inferior frontal gyrus and the left par- ca’s region by way of the extreme capsule, passing ietotemporal junction are critically important directly beneath the insular cortex [130], in for the processing of language. However, several which some of them may be synaptically inter- other authors [73, 821] failed to demonstrate any rupted. The efferent fibres from Broca’s region consistent association between aphasic syn- pass to the lower part of area 4, in which the mus- dromes and lesion locations. cles of larynx, tongue and lips are represented. 2. Pure forms of the various aphasias do oc- It has already been mentioned that damage cur, but are very rare because the lesions in- to Wernicke’s region leads to a sensory or re- volved (mostly infarctions in the left middle ceptive aphasia, in which the comprehension of cerebral artery territory) seldom conform to a written and spoken language is severely af- single functional site [65]. fected. Interrupting the connections between 3. It has become clear that, apart from Bro- Wernicke’s and Broca’s regions and between ca’s and Wernicke’s regions, numerous other the fibres passing from Broca’s region to the brain areas participate in speech and language inferior primary motor cortex results in two (Fig. 15.60). other types of aphasia, known as conduction (a) Functional neuroimaging studies [668, aphasia and subcortical motor aphasia. Con- 740] have shown that the intermediate and pos- duction aphasia is characterized by a marked terior parts of the left superior temporal gyrus impairment in repeating spoken language, ac- play a crucial role in storage and retrieval of companied by preserved comprehension [26, linguistic information and that the left middle 48, 151]. Subcortical motor aphasia clinically and inferior temporal gyri are critically in- closely resembles cortical motor aphasia due to volved in lexical and semantic processing [320, damage of Broca’s region [165]. 321, 645]. 628 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.59. The classical or Wernicke-Geschwind model of the circuitry related to the comprehension and production of language. A1, primary auditory area; arcf, arcuate fasciculus; B, Broca’s speech region; M1,primary motor cortex; S1, primary somatosensory cortex; V1, primary visual cortex; W, Wernicke’s speech region 15 Telencephalon: Neocortex 629

Fig. 15.60. Cortical areas involved in the comprehension and production of language. The classic regions of Wernicke (W) and Broca (B) are shown in grey; more recently discovered language-related areas are shown in red. The small window shows part of the insula. a–g, language-related areas discussed in the text; cc,corpus callosum; cg, cingulate gyrus; cs, central sulcus; LIFG, left inferior frontal cortex; ls, lateral sulcus 630 Section II Structure of Spinal Cord and Brain Parts

(b) Electrical stimulation in epileptic pa- goort [276] concluded from fMRI studies on tients [77, 420] and fMRI studies [289, 593] subjects performing verbal control tasks that a have shown that an area situated in the inter- network of areas consisting of the anterior cin- mediate part of the left lateral occipitotemporal gulate cortex and a dorsolateral prefrontal re- gyrus (designated as the “basal temporal lan- gion encompassing parts of Brodmann’s areas 9 guage area”) is involved in word retrieval. and 46 is involved in verbal action planning (c) A study involving computerized lesion and attentional control. reconstruction in a series of aphasia patients (i) Remarkably, the primary motor cortex is [158] revealed that in patients with a disorder also involved in semantic processing. Hauk et in co-ordinating the movements for speech ar- al. [280] showed action words referring to face, ticulation, an area situated in the anterosuperi- arm and leg actions (e.g. to lick, pick or kick) or insula was specifically damaged. The authors in a passive reading test to normal volunteers concluded that this insular area is specialized for under fMRI monitoring. The mere reading of the motor planning of speech. It has been sug- these words appeared to activate loci of the gested [466] that fibres passing from Wernicke’s motor cortex involved in the actual movement to Broca’s region through the insula are synapti- of the tongue, fingers or feet. These findings cally interrupted in the posterior insular cortex. suggest that the brain areas that are used to (d) Functional neuroimaging studies have perform a particular action are also involved shown that the anterior cingulate cortex is in- in the comprehension of the words related to volved in linguistic activities such as selecting that action [132]. verbs to lists of nouns [603] and translation (k) Functional neuroimaging studies have [594]. shown that the right (non-dominant) hemi- (e) The supplementary motor area is consis- sphere contributes substantially to many as- tently activated in functional imaging studies pects of language comprehension and produc- involving speech [159, 593]. tion, including the detection of syntactic er- (f) Neuroimaging studies [58, 185, 572, 573] rors, comprehending contextual and figurative have identified a small region in the left inferi- meaning and prosody (melody, timing and in- or frontal gyrus that becomes active during tonation) [58, 391, 473]. specific linguistic activities, such as processing 4. There is evidence that the cortical do- semantic relationships between words or mains dedicated to language are compartmen- phrases and retrieving semantic information. talized into separate systems for processing dif- This region corresponds roughly to Brod- ferent aspects of language. Ojemann [514] cites mann’s area 47. lesion studies indicating the presence of sepa- (g) Hagoort [276] has recently pointed out rate areas for handling different languages, for that the language-relevant part of the frontal handling different grammatical classes of cortex, apart from the classical Broca region, words or for the naming of specific semantic also includes the inferior prefrontal region cor- categories, such as “animals” or “tools”. Func- responding to area 47 and presumably also the tional neuroimaging research, summarized by antero-inferior part of area 6. He denoted this Bookheimer [58], has shown that within the new entity as left inferior frontal cortex (LIFG). left inferior frontal lobe separate subsystems Hagoort adduces evidence indicating that LIFG are responsible for different aspects of lan- is crucial in unification operations required for guage. It appeared that a rostroventral area, binding single word information received from corresponding to Brodmann’s areas (BAs) 47 memory into larger semantic, syntactic and and 45, contributes to semantic processing; an phonological structures. intermediate area, corresponding to BAs 45 (h) Based on clinical evidence, Damasio and 44, has a role in syntactic processing; [129] reported that certain higher-level aspects whereas a dorsocaudal area including parts of of language formation require an intact pre- BAs 44 and 6 are involved in phonological pro- frontal cortex surrounding Broca’s region. Ha- cessing. Similar subsystems, subserving partic- 15 Telencephalon: Neocortex 631 ular specialized language-related functions Two groups of authors, Petrides and Pandya have also been reported for Wernicke’s region [580, 581] and Price and collaborators [86, [827] and for both Broca’s and Wernicke’s re- 516, 518], modified Brodmann’s original par- gions [300]. cellation of the human prefrontal cortex on the 5. The data reviewed indicate that the neural basis of comparative studies on the brain of network involved in the comprehension and the rhesus monkey and the human. Both production of language is much more extensive groups took Walker’s [802] cytoarchitectonic than was originally envisioned. However, the analysis of the frontal lobe of the rhesus mon- precise wiring of this network remains to be es- key as their starting point. In this analysis tablished. The marked functional specialization (Figs. 15.61 A, 15.62 A, 15.63 A), Walker used detected within language-processing areas sug- Brodmann’s numbering scheme. With regard to gests that several parallel, though interrelated their position, his areas 6, 8, 9, 10, 24, 45 and subnetworks are present. The language-related 46 are directly comparable to the similarly (sub)networks do not function in isolation. numbered areas of Brodmann’s map (Fig. Neuroimaging studies, cited by Bookheimer 15.58 A,B). However, Walker’s area 46 is very [58] and Hagoort [276], have reported increased extensive and occupies a considerable part of activity in Broca’s region during various non- the lateral surface of the frontal lobe (Fig. linguistic tasks. Mesulam [464] indicates that 15.61 A). Contrary to its human counterpart, many cortical nodes participate in the function Walker’s area 24 extends to the basal surface of of more than one cognitive network. Finally, it the brain, and his area 25 largely corresponds should be mentioned that the circuitry related to Brodmann’s area 32 in the human brain to language functions is presumably not con- (Figs. 15.58 B, 15.62 A). Walker designated an fined to the neocortex. Clinical studies, reviewed area occupying the ventrolateral part of the in [26, 48, 391], have shown that local lesions in frontal lobe, corresponding positionally with the basal ganglia, thalamus and even cerebellum parts of Brodmann’s areas 11 and 47, as area may lead to language deficits. 12 (Figs. 15.59 A, 15.63 A). However, Brodmann The prefrontal cortex is distinguished from used number 12 in his latest map of the hu- the motor and premotor areas by (1) the pres- man cortex [71] to specify quite another area, ence of a pronounced internal granular layer, occupying a ventromedial position in the fron- and (2) its strong reciprocal connections with tal lobe (Fig. 15.58 B). Finally, Walker delin- the mediodorsal thalamic nucleus (MD) [402, eated two “new” areas on the basal surface of 635]. On account of these features, the prefrontal the frontal lobe, which he labelled as areas 13 cortex is sometimes denoted as the granular and 14 (Fig. 15.63), neglecting the fact that frontal cortex or as the MD-projection cortex. Brodmann [70] used these numbers to specify However, neither of these two criteria is abso- other (insular) areas in the cortex of non-pri- lute. Thus, the caudobasal part of the prefrontal mate species. cortex is dysgranular rather than granular, and The modifications of Brodmann’s parcella- studies with modern tracing techniques in the tion of the frontal cortex introduced by Walker rhesus monkey have shown that most prefrontal [802] are clearly reflected in the comparative areas that receive MD input are also connected architectonic studies of Petrides and Pandya to other thalamic nuclei, and that, conversely, [578, 580, 581] and Price and collaborators the MD projections do not pass exclusively to [86, 516, 518]. As shown by Figs. 15.61 B,C, the prefrontal cortex [592, 749]. Another impor- 15.62 B–E and 15.63 B,C, both groups use a tant feature of the prefrontal cortex is that it re- mixed Brodmann-Walker nomenclature in their ceives inputs from all unimodal and heteromo- parcellations of the monkey and human cortex dal sensory association areas. On the basis of [748]. The most important deviations from these afferents, the prefrontal cortex may be Brodmann’s subdivision of the frontal cortex, qualified as a high-order heteromodal associa- introduced by the two groups of investigators tion area [463] (Fig. 15.15). mentioned, are the following: 632 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.61 A–C. Cytoarchitectonic maps of the lateral frontal cortex. A Rhesus monkey, according to Walker [802]. B Rhesus monkey, according to Petrides and Pandya [578]. C Human, according to Petrides and Pan- dya [578]. arcuat s, arcuate sulcus; princ s, principal sulcus 15 Telencephalon: Neocortex 633

Fig. 15.62 A–E. Cytoarchitectonic maps of the medial frontal cortex. A Rhesus monkey, according to Walker [802]. B, C Rhesus monkey and human, respectively, according to Petrides and Pandya [578]. D, E Rhesus monkey and human, respectively, according to Öngür and Price [516]. cc, corpus callosum 634 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.63. Cytoarchitectonic maps of the orbitofrontal cortex of the rhesus monkey, according to Walker [802] (A), Petrides and Pandya [578] (B) and Öngür and Price [516] (C). ot, caudal part of the olfactory tract 15 Telencephalon: Neocortex 635

1. In Brodmann’s map, area 46, which occu- We discussed the results of the architectonic pies a central position on the lateral surface of analysis of Petrides and Pandya and Price and the frontal lobe, is separated from area 8 by collaborators in some detail because they form the ventrocaudal part of area 9 (Fig. 15.58 A). the basis of extensive hodological studies on However, Petrides and Pandya [578] observed the frontal lobe of the rhesus monkey and that that the architecture of the portion of the mid- of a sound extrapolation of the results of these dle frontal gyrus, labelled area 9 in the map of studies to the human brain. In what follows, Brodmann, is more akin to area 46. They also wherever necessary, Brodmann’s areas will be observed that this cortex corresponds in archi- denoted with the letters “BA” and areas derived tecture to the cortex found in the caudal part from the mixed Brodmann-Walker parcellation of the sulcus principalis in the rhesus monkey, as “BWA”. and that has been included in area 46 by Walk- The various cytoarchitectonic areas forming er (Fig. 15.61 A). Petrides and Pandya therefore the prefrontal cortex are strongly and recipro- labelled the portion of area 9 that lies on the cally interconnected [89, 581]. An important middle frontal gyrus as area 9/46 (Fig. output channel from the prefrontal cortex is 15.61 C), to acknowledge its architectonic simi- formed by sequences of short association fibres larity to area 46 and its inclusion with area 9 which, via the premotor cortex, converge upon in Brodmann’s map. the primary motor cortex [419, 711]. 2. In Brodmann’s map, the region rostroven- The fibre connections of the prefrontal cor- tral to area 45 has been labelled area 47 (Fig. tex are multifarious. The afferents from the 15.58 A). Petrides and Pandya [580] observed various association areas are reciprocated by that this region has architectonic features com- efferent projections. It receives cholinergic and parable to those of Walker’s area 12 in the rhe- GABAergic afferents from the basal nucleus of sus monkey (Fig. 15.61 A). They therefore la- Meynert, histaminergic, orexinergic and mela- belled this region area 47/12 (Fig. 15.61 C). nin-concentrating hormone-containing fibres 3. On the medial surface of the human fron- from the hypothalamus, serotoninergic fibres tal lobe, Petrides and Pandya [581] (Fig. from the mesencephalic raphe nuclei, dopa- 15.62 C) as well as Price and collaborators minergic fibres from the ventral tegmental area [516, 518] have replaced area 12 and part of and noradrenergic fibres from the locus coeru- area 11 of Brodmann (Fig. 15.58 B) by area 14 leus. It has reciprocal connections with various of Walker (Fig. 15.63 A). parts of the limbic lobe (Fig. 15.57), the amyg- 4. It is remarkable that in the map of Price dala (Figs. 13.6, 13.7), septum and hypothala- and collaborators, the medial part of Brod- mus, and it projects to the caudate nucleus mann’s area 10 is extraordinarily large and (Fig. 14.1), the periaqueductal grey (PAG) and abuts directly on area 24, splitting area 32 into the pons. Several of these connections will be a dorsal, human (h) and a ventral monkey (m) further considered below. part (Fig. 15.62 E). The prefrontal cortex is critically involved in 5. We have seen that Walker, in his architec- complex brain functions such as orientation tonic analysis of the frontal lobe of the rhesus and attention, decision making on the basis of monkey, has delineated two “new” areas on the current exteroceptive and interoceptive infor- orbital surface of that lobe, which he labelled mation and past experience, planning and se- areas 13 and 14 (Fig. 15.63 A). Petrides and quencing of actions, emotionality and person- Pandya [581] and Price and collaborators [516, ality. Large bilateral lesions of the prefrontal 518] confirmed the presence of these areas in cortex lead to a disorder known as frontal lobe the rhesus monkey (Fig. 15.63 B,C) and also syndrome. The essential features of this syn- identified them in the human (Fig. 15.11 E,G). drome are [151]: (1) diminished capacity to It is important to note that Price and collabo- sustain attention and concentration; (2) lack of rators divided most of the areas they delineated spontaneity and initiative; (3) inhibition and into two to four subareas (Figs. 11.8, 15.11 G). impulsivity of affect, thought and action; (4) 636 Section II Structure of Spinal Cord and Brain Parts inability to plan, organize and execute complex brain functions, such as integrating sensory behaviour; and (5) loss of social decorum. analysis with motor activity, selective attention, Comparable changes have been observed in working memory, planning and reasoning. In psychiatric patients who underwent prefrontal what follows, a brief survey of the principal leucotomy. The principle of this operation is to connections of the lateral prefrontal cortex cut the fibre connections to and from the pre- (LPFC) will be presented, to which some func- frontal region. This was done by inserting a tional notes will be attached. “leucotome” through a burr hole and moving Experimental studies in rhesus monkeys the instrument in a coronal plane (Fig. 15.64). [578, 580, 581] have shown that the various As a rule these operations were performed bi- prefrontal areas are tightly and reciprocally in- laterally. Prefrontal leucotomy was introduced terconnected and that this connectional net- by the Portuguese neurologist Egas Moniz in work extends uninterruptedly over the medial 1936 [476] and was thenceforward strongly ad- prefrontal and orbitofrontal regions [89, 516]. vocated by the Americans Walter J. Freeman The LPFC represents the highest level of and James W. Watts [202, 203]. In the two de- sensorimotor integration. It receives highly cades following its introduction, tens of thou- processed information from multiple sensory sands of severely disturbed psychiatric patients modalities and plays a prominent role in the underwent leucotomies all over the world [330, planning and organization of voluntary goal- 752]. For his discovery, Moniz was awarded the directed behaviour. The sensory information is Nobel Price in 1949. The principal target group conveyed to the LPFC by massive projections of the intervention consisted of chronic, hospi- emanating from post-Rolandic unimodal and talized patients, suffering from depression, an- polymodal sensory association area (Figs. xiety states and obsessive-compulsive disor- 15.15, 15.46). ders. Although Moniz [477] qualified prefrontal A combination of electrophysiological re- leucotomy euphorically as “highly effective and cordings and anatomical tract-tracing in rhe- always safe”, it soon became clear that the in- sus monkeys [245, 513, 631, 632, 822], as well tervention produced serious changes in per- as neuroimaging studies in humans [113, 116, sonality, including apathy, slowness, lack of in- 147, 455], have shown that the LPFC is orga- itiative, carelessness, poor judgement and in- nized in separate dorsal spatial cognition- and hibited behaviour in social situations [803]. ventral object cognition- and pattern cognition From a neuroanatomical point of view, the op- domains. For a discussion of the relevant elec- erations were very crude, with poor control of trophysiological and neuroimaging data, the the actual place of the section, due to the con- reader is referred to the publications cited; siderable interindividual variations of brain here we confine ourselves to mentioning that size and shape, and skull-brain relationships. the visual and auditory projection streams in- Furthermore, there were unintended side ef- volved in the processing of spatial information fects, such as haemorrhages, sometimes found converge upon the dorsal part of the LPFC, far from the site of the section [468, 469] (Fig. whereas visual and auditory streams, con- 15.65). Fortunately, the advent of effective psy- cerned with pattern discrimination, mainly tar- chopharmaceuticals put an end to the era of get the ventral part of the LPFC (Fig. 15.66). prefrontal leucotomy. However, this functional segregation is not ab- The prefrontal cortex can be divided into solute. In both domains units processing both three major regions: lateral, medial and orbital. object identity and location have been found The lateral prefrontal cortex encompasses [613]. BAs 8, 9, 46, 10 and 47 (Fig. 15.58 A). BAs 9 All the pathways that convey sensory infor- and 46 together correspond to BWAs 9, 9/46d, mation from posterior cortical areas to the 9/46v and 46. BA 47 roughly corresponds to LPFC contain reciprocal fibres, projecting back BWAs 47/12 (Fig. 61 C). This cortical region is to areas from which input was received [581, known to participate in numerous higher-order 582]. These feed-back fibres form part of the 15 Telencephalon: Neocortex 637

Fig. 15.64. Drawing showing how the brain is incised during a prefrontal leucotomy. An identical incision is made on the opposite side. Reproduced from Freeman and Watts [202]

Fig. 15.65. Section of frontal lobes in plane of leucotomy incision, 4 months after operation in which the left anterior cerebral artery was injured. Based on a photograph in Freeman and Watts [202] 638 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.66. Schematic diagram showing that visual and auditory projection streams involved in the processing of spatial information, target the dorsal part of the lateral prefrontal cortex (shown in red), whereas projection streams subserving pattern discrimination target the ventral part of the lateral prefrontal cortex (in black). A1, primary auditory cortex; AAPFS, anterior audito-frontal stream; DVPS, dorsal visual processing stream; PAPFS, posterior audito-prefrontal stream; V1, primary visual cortex; VVPS, ventral visual processing stream 15 Telencephalon: Neocortex 639 morphological substrate of attention, i.e. the afferents from the limbic lobe as well as from mechanism that enables us to direct our pro- the mediodorsal, ventral anterior, and the ante- cessing resources to a subset of the available rior and posterior parts of the ventral lateral information [124]. This mechanism allows us thalamic nucleus, and from a large number of to selectively process the information relevant subcortical, extrathalamic structures. to current goals. A combination of the results The afferents from the limbic lobe reach the of neuroimaging studies in humans with data LPFC via two fibre systems: the dorsal and derived from tract-tracing studies in monkeys ventral limbic pathways [579]. The dorsal lim- suggests that, apart from the LPFC, the visual bic pathway originates from the rostral and attentional control system involves several way caudal cingulate cortex (areas 24 and 23) and stations in the dorsal and ventral visual process- the retrosplenial cortex (area 30) and passes ing streams (Fig. 15.67) [552, 553]. Egner and forward in the cingulate bundles to the LPFC. Hirsch [168] recently presented neuroimaging The ventral limbic pathway originates from the evidence suggesting that the LPFC exerts atten- posterior parahippocampal cortex, particularly tional control on post-Rolandic cortical areas by area TF, and reaches the LPFC via the extreme amplifying task-relevant information rather capsule [247, 397]. Both limbic pathways are than by inhibiting distracting stimuli. bidirectional, i.e. they also contain fibres origi- The organization of mental processes, such as nating from the LPC and terminating in the the sequencing of complex goal-directed behav- cingulate, retrosplenial and parahippocampal iours, not only depends on the selective atten- cortices [247, 377, 397]. Neuroimaging studies tional control of incoming sensory information, have shown that various subregions of the cin- but also on the ability to keep the information gulate cortex can be activated by stimuli re- selected available for some time. This ability to lated to simple emotions, such as happiness, transiently maintain and manipulate a limited sadness, anger and fear [780]. It seems likely amount of information to guide thought or be- that the cinguloprefrontal projections may sub- haviour is known as working memory [405]. serve the assigning of emotional valence to the The LPFC plays a key role in working memory. sensory information processed in the LPFC. It Classic experiments on monkeys and chimpan- is well known that the hippocampal formation zees in the 1930s [327, 328] have shown that, fol- and adjacent medial temporal structures repre- lowing bilateral lesions of the LPFC, the animals sent essential components of the memory sys- were still fully able to perform auditory or visual tem (Chap. 12). Because these structures are discrimination tasks or recall spatial orientation tightly connected to all other parts of the lim- of objects, providing tests were made in im- bic lobe (Figs. 12.13, 15.56), it has been sug- mediate memory. However, the animals ap- gested that the limbico-prefrontal pathways peared to be unable to perform such tasks if a may also form part of the network involved in delay of more than a few seconds was introduced the encoding and retrieval of long-term mem- between the stimulus and the response. The re- ory [579, 685]. lation between this temporal integrative func- The LPFC is reciprocally connected with the tion and the LPFC has been substantiated elec- lateral, parvocellular division of the mediodor- trophysiologically [215, 217]. Cells in the LPFC sal thalamic nucleus [712]. As discussed in of the monkey were found to fire persistently Chap. 8, this nucleus belongs to the so-called at high rates while the animals retained an item higher-order thalamic relays, which means that of visual information in short-term memory. Le- it receives its principal or “driver” afferents sion studies in patients [575] and neuroimaging mainly from the cortex and plays a potentially studies in normal human subjects [574] have significant role in cortico-cortical communica- also shown that the LPFC is critically involved tion [681]. in working memory. Fibres passing from the ventral anterior Apart from the projections from the various (VA) and anterior ventral lateral (VLa) thalam- sensory association areas, the LPFC receives ic nuclei to the LPFC represent the final stage 640 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.67. Some centres involved in the processing of visual information. Their feed-forward connections (shown in black) originate from the primary visual cortex and proceed to the lateral prefrontal cortex, via the dorsal and ventral visual processing streams. Red arrows indicate potential feedback connections, forming part of the visual attentional control system. Based on Pessoa et al. [552, 553]. AITA, anterior inferior temporal area; DVPS, dorsal visual processing stream; EVAA, extrastriate visual association area; LIP, lateral intraparietal area; LPC, lateral prefrontal cortex; PITA, posterior inferior temporal area; V1, primary visual cortex; VVPS, ventral visual processing stream 15 Telencephalon: Neocortex 641 of the associative or cognitive loop, forming suggested [611] that the prefronto-cerebellar part of the direct striatal circuit (for details, system may facilitate the skilled execution of see Chap. 14). This loop is composed of: (1) routine cognitive operations. corticostriate fibres which project to the cau- The afferents to the LPFC from extrathalamic date nucleus, (2) striopallidal fibres passing to subcortical centres do not differ from those to the internal segment of the globus pallidus, (3) other neocortical regions. The structure and pallidothalamic fibres terminating in VA and overall functions of these afferent systems have VLa, and (4) the thalamocortical fibres already been discussed in the section “Neocortical affer- mentioned, which pass back to the same corti- ents” of the present chapter. Let it suffice to re- cal region from which the loop originated call that the neurons in several of these subcor- (Figs. 14.1, 14.8, 14.17). tical centres are particularly responsive to novel Classically, the striatum has been regarded and motivationally relevant sensory stimuli and as a part of the motor system, particularly be- that their efferents modulate the excitability of cause disturbances in its functions can lead to the cortical neurons on which they impinge. severe movement disorders. However, it has The LPFC exerts its control on motor behav- gradually become clear that the striatum also iour by cascades of short projections, which participates in cognitive functions. Lesions or reach the caudal premotor areas (F2–F5) and electrical stimulation of the caudate nucleus re- the primary motor cortex (M1, F1), after syn- sult in cognitive deficits, as revealed by delayed aptic interruption in the rostral premotor areas response and delayed alternation tasks [839]. (F6, F7) [419, 433, 445, 475, 578, 580]. The Neuroimaging studies indicate that the stria- caudal premotor and primary motor areas, tum is involved in procedural learning and which send direct projections to the spinal memory processes [524], and patients with ba- cord, are closely involved in movement execu- sal ganglia disorders exhibit impairments in tion, whereas the rostral premotor areas pro- the planning and execution of constructional cess more “cognitive” aspects of motor control, tasks [399] and in the switching from one be- such as sequence generation and motor learn- havioural strategy to another [110, 111]. ing [269]. It has already been mentioned that Just like the striatum, the cerebellum is not the LPFC has a substantial direct projection to only concerned with motor control, but also the cingulate motor areas. participates in cognitive functions. The pre- Based on fMRI evidence, several recent stud- frontal cortex and the contralateral cerebellar ies [238, 378, 379, 612] suggest that the human cortex are interconnected by a complex corti- frontal polar cortex is specifically concerned cocerebellar circuit, comprising a feed-forward with complex cognitive functions. Koechlin et or afferent limb and a feed-back or efferent al. [378] found that this cortical region is se- limb. Topographically organized corticopontine lectively activated when subjects held in mind and pontocerebellar projections form the feed- a main goal while performing concurrent forward limb, whereas the feed-back limb is (sub)goals. Ramnani and Owen [612] con- composed of cerebellar corticonuclear projec- cluded that the frontal polar cortex has a spe- tions, efferents from the dentate nucleus to the cific role in integrating the outcomes of two or posterior part of the ventral lateral thalamic more separate cognitive operations in the pur- nucleus (VLp), and efferents from the latter to suit of a higher behavioural goal. the prefrontal cortex (Fig. 14.8). These connec- The complex of symptoms shown by pa- tions are considered to represent the morpho- tients with damage to the LPFC has been logical substrate for the cerebellar control of termed the dysexecutive syndrome [27]. Pa- cognitive processing. For a survey of the evi- tients with this syndrome are characterized by dence indicating the participation of the cere- diminished insight and judgement, poor plan- bellum in cognitive functions, the reader is re- ning and decision making and impairments of ferred to the section “A brief excursion to the attention, working memory and the temporal cerebellum” in Chap. 14. It has recently been organization of recent events [216, 483]. 642 Section II Structure of Spinal Cord and Brain Parts

The orbital prefrontal cortex (OPFC) has been Fig. 15.68 A–D, these findings are extrapolated subdivided in different ways. Brodmann [70, to the human brain. 71], who did not study this region in detail, sub- As already discussed in Chap. 11, the OPFC divided it into two areas: 11 and 47 (Fig. receives extrinsic sensory afferents from five dif- 15.11 A). Uylings et al. [751], who recently sub- ferent sources. Olfactory, gustatory, autonomic, jected the OPFC to a thorough cytoarchitectonic visual and somatosensory projections have been analysis, took Brodmann’s parcellation as their traced to restricted and different areas in the point of departure. They confirmed the presence caudal and lateral parts of this region [89, 90] of Brodmann’s areas 11 and 47, but subdivided (Fig. 15.68 D). The olfactory projections origi- the latter into three medial and two lateral sub- nate from the primary olfactory cortex (Fig. areas (Fig. 15.23). At present, the subdivisions of 11.7). The gustatory and autonomic afferents the LPFC, presented by Petrides and Pandya stem from different parts of the parvocellular [580, 581] and Price and collaborators [86, section of the ventral posteromedial thalamic 516, 518], which, as discussed before, are both nucleus and reach the caudal OPFC via synaptic derived from Walker’s parcellation of the frontal relays in the insular cortex (see the next section cortex of the rhesus monkey [802], are most of the present chapter). The visual afferents ori- widely used. According to these subdivisions, ginate from the inferior temporal cortex and the the OPFC encompasses four areas, BWAs 11, somatosensory projections arise from the first 13, 14 and 47/12 (Fig. 15.11 E,G). It has already and second somatosensory cortex and from been mentioned that Price and colleagues [516, the parietal association cortex. Anatomical, 518] subdivided each of these four areas into physiological and neuroimaging data [88, 89, several subareas. There is an anterior-posterior 121, 355, 386, 628–630] indicate that the orbital trend in the cytoarchitecture of the OPFC areas in receipt of these sensory inputs and the [751], which manifests itself as a gradual reduc- network in which they are embedded are in- tion in the granularity of layer IV. Hence, the volved in the analysis and integration of food- posterior portions of BWAs 13, 14 and 47/12 related sensations and play an important role are either dysgranular or agranular. Mesulam in the control of feeding. The caudal sectors of [463] included this dysgranular/agranular re- the orbital network provide a basis for conver- gion of the OPFC in what he called the paralim- gence from the various unimodal sensory areas bic belt (Figs. 15.55, 15.56). onto multimodal areas in the centromedial re- Apart from medially extending portions of gion of the OPFC (Fig. 15.68 D). BAs 9 and 10 and BWA 14, the medial prefrontal cortex (MPFC) includes Brodmann’s ante- Fig. 15.68 A–E. The orbital and medial prefrontal cor- rior limbic area 24, prelimbic area 32 and in- tex (OMPFC). Cytoarchitectonic parcellation of the fralimbic area 25 (Figs. 15.58 B, 15.62 C). BAs medial (A) and orbital prefrontal cortex (B), according 24 and 25 are situated within the confines of to Öngür et al. [518]. C Intrinsic connections comprising the medial cortico-cortical network (in black), and the limbic lobe and form part of the paralim- four prominent output channels of the medial prefron- bic belt (Figs. 15.55–15.57). tal cortex (in red), originating from: (1) the infralimbic Because the OPFC and MPFC share many cortex; (2) the prelimbic cortex, (3) the anterior cin- connectional and functional features, they are gulate cortex, and (4) the cingulate motor areas. Based commonly considered to form a single com- on Carmichael and Price [89], Price et al. [595], An et al. [9], Öngür et al. [517] and Morecraft and Van Hoe- plex, the orbital and medial prefrontal cortex sen [482]. D Intrinsic connections comprising the or- (OMPFC). bital cortico-cortical network (in black) and unimodal The detailed experimental hodological stud- sensory inputs to the orbital prefrontal cortex (in red). ies of Price and collaborators [89, 516, 595] in Based on Carmichael and Price [89] and Price et al. the rhesus monkey have shown that the cytoar- [595]. E Connections between the orbital and medial cortico-cortical networks. Based on Carmichael and chitectonic areas of the OMPFC can be allocated Price [89]. cc, corpus callosum; CMA, cingulate motor to two groups or networks, one restricted to the area; facial n, facial nucleus; hypoth, hypothalamus; OPFC and the other confined to the MPFC. In PAG, periaqueductal grey 15 Telencephalon: Neocortex 643 644 Section II Structure of Spinal Cord and Brain Parts

Mention should be made here of a remark- The medial wall areas 25 and 32 project to the able problem with regard to the exact localiza- anterior and ventromedial hypothalamic nuclei tion of the olfactory zone in the OPFC. The or- (AHN, VMH) in the medial hypothalamus. The bitofrontal focus of olfactory activity in hu- orbital areas 13a, 12/47 l and Iai selectively in- mans, as assessed by neuroimaging studies nervate the lateral hypothalamic area (LHA), [251, 845], appeared to be situated consider- particularly its posterior part [517]. The AHN ably more anteriorly (Fig. 15.69 A) than that and VMH are nodal points in networks sub- derived from an extrapolation of the experi- serving motivated or goal-oriented behaviours, mental results of Carmichael et al. [90] in the i.e. behaviours essential in maintaining the in- rhesus monkey (Fig. 15.68 D). dividual and species, such as thermoregulation, It is important to note that the processing of agonistic behaviour and reproduction (see sensory information in the OPFC is not con- Chap. 10). The posterior LHA is involved in fined to the physical and chemical properties autonomic control. of the stimuli but also involves their affective The PAG can be subdivided into four longitu- or emotional significance. Emotions can be dinal zones or columns: dorsal dorsolateral, lat- classed as positive or negative. Positive emo- eral and ventrolateral [34, 91]. This structure is tions are elicited by rewards or positive rein- known to play a prominent role in co-ordinating forcers; negative emotions are evoked by pun- behavioural and autonomic responses to escap- ishments or negative reinforcers. A large meta- able and unescapable stressful situations. Elec- analysis of neuroimaging studies in humans trical or chemical stimulation of its lateral col- [386] has shown that different subregions of umn evokes co-ordinated response strategies, the OPFC play different roles in the processing such as threat display, fight or flight, accompa- of emotional information. It appeared that ac- nied by hypertension, tachycardia and a shift tivity in the medial OPFC is related to the in blood from the viscera to the limb muscles. monitoring, learning and memory of the re- In contrast, stimulation of the ventrolateral ward value of positive reinforcers, whereas lat- PAG column elicits quiescence (“freezing”) and eral OPFC activity is related to the evaluation decreasing of blood pressure and heart rate. Ret- of negative reinforcers (“punishers”) that can rograde and anterograde tracing experiments in lead to a change in behaviour (Fig. 15.69 B). rhesus monkeys [9] have shown that the medial This meta-analysis [386] also yielded evidence prefrontal areas 24, 25 and 32 and the orbital indicating that in the OPFC more complex or areas 13 a, 27/12 l and Iai project to the PAG abstract reinforcers (such as monetary gain and that projections from distinct cortical areas and loss) are represented more anteriorly than terminate primarily in individual PAG columns. less complex reinforcers (such as taste). The projections from areas 25 and 32 terminate The OMPFC projects densely to the hypo- primarily in the dorsolateral columns, bilater- thalamus and the mesencephalic periaqueduc- ally. Fibres from area 24 predominantly end in tal grey (PAG) and moderately to lightly to the lateral column, whereas fibres from orbital many brainstem centres, including the ventral areas 13 a, 47/12 l and Iai terminate mainly in tegmental area, the mesencephalic raphe nu- the ventrolateral columns. It seems likely that clei, the pedunculopontine nucleus, the locus the OMPFC, by way of these projections, influ- coeruleus and the parabrachial nuclei. Almost ences the integrated behavioural and autonomic all of these centres form part of the greater responses mediated by the lateral and ventrolat- limbic system, discussed in Chap. 23. The fi- eral PAG columns. The functional significance of bres projecting to the hypothalamus arise from the cortical input to the dorsolateral column is prelimbic area 32 and infralimbic area 25 on not known. It is, however, important to note that the medial surface of the frontal lobe, and this column receives a strong projection from from the orbital areas 13a, 12/47 l and Iai [86, the ventromedial hypothalamic nucleus [517], 517]. These three orbital areas are strongly and that this nucleus is involved in both sexual connected to the medial network (Fig. 15.68 E). and defensive behaviour (see Chap. 10). 15 Telencephalon: Neocortex 645

Fig. 15.69 A,B. The functions of the human orbitofrontal cortex. A The putative olfactory projection area (in red), as based on a meta-analysis of neuroimaging data, appears to be situated substantially rostrally to that extrapolated from experimental hodological studies in the rhesus monkey (in grey). Based on Gottfried and Zald [251]. B A large meta-analysis of neuroimaging data showed that there is a medial-lateral functional distinction in the orbitofrontal cortex, such that activity in the medial orbitofrontal cortex (black dashed el- lipse) is related to the learning, monitoring and memory of the reward value of positive reinforcers, whereas lateral orbitofrontal cortex activity (red dashed ellipses) is related to the evaluation of negative reinforcers (punishers). Based on Kringelbach [386]. In the figures, temporopolar regions have been removed to allow visualization of the insula and caudal orbitofrontal cortex. The cut surfaces of the “temporal peduncles” are hatched 646 Section II Structure of Spinal Cord and Brain Parts

The projections from the OMPFC to the hy- and (c) the voluntary motor system, via the pothalamus and PAG form part of the limbic cingulate motor areas. motor system, which is also designated as the 4. The two networks are strongly intercon- emotional motor system [302]. Fibres descend- nected. By way of these connections, sensory ing from the central nucleus of the amygdala information screened for emotionally and and the bed nucleus of the stria terminalis also motivationally relevant features is trans- contribute to this system. ferred to the medial network. The complex formed by the strongly interconnected rostral and caudal cingulate motor There is a certain resemblance between the areas M3 and M4, which is also embedded in limbic or emotional-motivational sensory-mo- the medial prefrontal network, projects via the tor transfer system just outlined and the cogni- corticobulbospinal tract to the facial nucleus to-motor transfer system through the lateral and spinal cord. The complex is interconnected prefrontal cortex (LPFC). In the latter system, with other cortical motor areas, such as the highly processed information from sensory as- primary motor cortex, the supplementary and sociation areas in the parietal and temporal presupplementary motor areas and the lateral cortex converges onto the LPFC. The LPFC premotor cortex [279, 481] and receives affer- then produces voluntary motor actions based ents from the lateral prefrontal areas 9 and 46 on this sensory input via sequences of short and from many limbic cortical areas, including connections to premotor and motor areas (Fig. posterior cingulate area 23, retrosplenial areas 15.15). 29 and 30, temporal areas 35, TH and TF, tem- The OMPFC is connected to a large number poropolar area 38 [482], as well as from the of other brain structures, including the ante- amygdala [483]. In light of these connectional rior temporal lobe, the amygdala and other data, Morecraft and Van Hoesen [482] sug- limbic structure, and the basal ganglia. These gested that the cingulate motor areas form a connections will now be briefly discussed. strategic cortical entry point for limbic influ- Experimental hodological studies in rhesus ence on the voluntary motor system. They monkeys [36, 274, 381, 484, 576, 579, 581, 631, pointed out that patients with damage to the 672] have shown that the OMPFC is strongly anterior cingulate cortex are often character- interconnected with various parts of the ante- ized as akinetic or docile, attention impaired, rior temporal lobe. Without going into details, mute and lack emotional tone. it may be mentioned that many orbital areas As indicated in Fig. 15.68 E, all medial pre- receive afferents from the auditory association frontal areas are connected to several orbital cortex in the superior temporal gyrus, the in- areas [89]. ferotemporal visual association cortex, and the The data concerning the wiring of the multimodal sensory association cortex in the OMPFC discussed above may be summarized superior temporal sulcus (Fig. 15.54). as follows: The amygdaloid complex projects to various 1. The cytoarchitectonic (sub)areas forming the parts of the OMPFC [90, 233, 484] (Figs. medial and orbital prefrontal cortices are 15.70 A, 15.71). These amygdalocortical projec- both embedded into a network of cortico-cor- tions originate from the basal nucleus, and to tical connections. a lesser extent from the accessory basal and 2. The orbital network contains a “sensory lateral nuclei, and terminate principally in (a) pole” involved in the analysis and integra- the caudolateral entry zone for food-related in- tion of food-related sensations. formation, (b) a medial orbital zone, which 3. The medial network contains a “motor forms an interface between the orbital and me- pole”, which potentially exerts influence on dial networks, and (c) the medial prefrontal (a) subcortical networks subserving the exe- output areas 24 and 32 (Fig. 15.70 A). Most of cution of integrated motivated behaviours, these projections are reciprocated by projec- (b) centres involved in autonomic control tions from the cortex back to the amygdala. 15 Telencephalon: Neocortex 647

Fig. 15.70 A,B. Limbic inputs to the orbital and medial prefrontal cortex. A Afferents from the amygdaloid complex. B Afferents from the entorhinal cortex. Based on Carmichael and Price [87] 648 Section II Structure of Spinal Cord and Brain Parts

Fig. 15.71. Diagrammatic summary of connections of the accessory basal (AB), basal (B) and lateral (L)nuclei of the amygdaloid complex (AM), orbital and medial prefrontal cortex (OMPFC), magnocellular part of mediodorsal thalamic nucleus (MDm), ventral striatum (VS), ventral pallidum (VP) and mesencephalic ventral tegmental area (VTA) and adjacent pars compacta of substantia nigra (SNc). Partly based on Price et al. [595]. AAC, auditory association cortex; ac, anterior commissure; Cau, caudate nucleus; ic, internal capsule; Put, putamen; Thal, thalamus; VAC , visual association cortex 15 Telencephalon: Neocortex 649

The basolateral nuclei of the amygdala are The ventral striatum projects back to the VTA known to receive highly processed visual, audi- and the latter is also reciprocally connected with tory and somatosensory information and to be the ventral pallidum [269] (Fig. 15.71). involved in evaluating the emotional and social There is physiological [14, 15, 61, 628, 666] significance of this information [41, 584] (see and neuroimaging evidence [375, 664, 665] in- Chap. 13). dicating that the limbic circuit, including its In addition to receiving inputs and sending mesencephalic extension, is involved in the ex- outputs to the amygdala, the OMPFC is also pectation and evaluation of reward and in the connected with many other limbic structures, reward-guided selection of goal-oriented be- including the subiculum, entorhinal area 28, haviours. Dopaminergic neurons in the VTA perirhinal areas 35 and 36, and parahippocam- and adjacent SNc are known to play a key role pal areas TF and TH [87, 382, 484]. The subi- in recognizing and predicting rewards [664, culum, entorhinal and perirhinal areas are al- 665]. Dysregulation in the circuits intercon- most exclusively connected with the medial necting the orbitofrontal cortex, ventral stria- and caudal parts of the orbitofrontal cortex tum and dopaminergic midbrain are associated (Fig. 15.70 B). In contrast, the parahippocam- with several mental health disorders, including pal cortex is primarily connected with the me- depression [448, 449], drug addiction [175, dial prefrontal cortex, especially BAs 24, 25 357, 663, 792] and schizophrenia [284]. and 32. Lesions of the orbitofrontal cortex often in- Apart from the direct projections from lim- duce dramatic changes in personality, as exem- bic structures to the OMPFC, there are also in- plified by the famous case of Phineas Gage direct projections, which are synaptically inter- [278], discussed in a previous section of the rupted in the thalamus. The amygdala, entorhi- present chapter. Patients with such lesions are nal cortex and subiculum all project to the me- irritable, impulsive and disinhibited, with a dial, magnocellular part of the mediodorsal characteristic tendency to tactlessness, vulgar- thalamic nucleus (MDm), which in turn is re- ity and disregard for social and moral princi- ciprocally connected with all parts of the ples. In addition, orbitofrontal patients exhibit OMPFC [614]. a severe disorder of attention, and there are As discussed in Chap. 14, the neocortex, ba- marked abnormalities in the realms of reason- sal ganglia and thalamus are interconnected by ing and decision-making [216, 483]. a number of parallel circuits or loops. One of Medial prefrontal lesions commonly lead to these, designated as the limbic loop, involves apathy, manifesting as a severe reduction in the OMPFC and MDm. It originates from the spontaneity, motivation and general motility. medial BAs 24, 25 and 32 and the orbital BWAs Patients with such lesions also show a flattened 11, 13, 14 and 47/12. All of these areas project affect and reduced interest in their environ- to the ventral striatum, composed of the ac- ment. Large bilateral lesions may lead to aki- cumbens nucleus and adjacent sectors of the netic mutism [216, 483]. caudate nucleus and putamen [270, 839]. The ventral striatum, in turn, projects to the ventral pallidum formed by rostral and ventral exten- Insula sions of the globus pallidus. The ventral pallidum, then, projects to MDm, and connections The insula or island of Reil is buried in the between this thalamic nucleus and the OMPFC depths of the lateral sulcus. It is covered by ad- form the final link of the limbic loop (Figs. jacent parts of the frontal, parietal and tempo- 14.17, 15.71). ral lobe, known as the frontal, frontoparietal Dopaminergic neurons in the mesencephalic and temporal opercula (Figs. 2.23, 3.3). The in- ventral tegmental area (VTA) and adjacent parts sula is shaped like a triangle, the apex of of the substantia nigra, pars compacta (SNc) which is directed anterobasally. The insular project to the OMPFC and ventral striatum. cortex is separated from surrounding opercular 650 Section II Structure of Spinal Cord and Brain Parts cortices by a limiting or circular sulcus. A cen- Table 15.3. Connections of the insula tral sulcus divides the insular surface into a Ia Idg Ig aff eff larger anterosuperior and a smaller posteroinferior portion. The anterior part is covered by Somatosensory some short gyri (gyri breves); the posterior SI (areas 3, 1, 2) ++ + part is incompletely separated into two long Area 5 ++ + gyri (gyri longi) (Fig. 3.3). Area 7 b ++ + + As regards the structure of the insula, Brod- VMPo (thal.) ++ + mann [70] indicated that its cortex can be di- Vestibular vided into an agranular anterior part and a VPS, VPI (thal.) ++ + granular posterior part, and that the boundary Auditory between these areas is formed by a vertical line Audit. assoc. ctx (STP, PA) + ++ + which partly coincides with the central insular Motor sulcus (Fig. 15.8). In the insular cortex of the Medial premotor (area 6) ++ ++ + rhesus macaque and human, Mesulam and High-order association Mufson [466] recognized three concentrically Ant. orbfr. ctx (area 11) ++ ++ + + arranged zones, a rostroventral agranular zone, Prefr. ctx (areas 45, 46) ++ + + an intermediate dysgranular zone and a caudo- Banks sup. temp. sulcus + ++ + + dorsal granular zone (Fig. 15.72 A). The inter- Olfactory mediate zone is termed dysgranular because Olfactory cortex ++ + + + the granule cells in layers II and IV are rather scarce and do not display complete laminar Gustatory differentiation. VPMpc, med (thal.) ++ + The fibre connections and functions of the General viscerosensory insula have been extensively reviewed by Me- VPMpc, lat (thal.) ++ + + sulam and Mufson [466] and Augustine [22]. Limbic Much of what follows has been derived from Entorhinal ctx (area 28) ++ + + + these studies. Perirhinal ctx (areas 35, 36) + + The insula receives afferents from the dorsal Temporopolar ctx (area 38) ++ ++ + + thalamus and is connected with the amygdala Post orbfr. ctx (areas 12, 13) ++ + + + and with a large number of cortical areas (Ta- Cingulate gyrus (areas 23, ++ ++ + + ble 15.3). The thalamic nuclei which project to 24) the insula include the ventromedial posterior Amygdala, corticomedial ++ ++ + + nucleus (VMPo), the ventral posterior superior Amygdala, basolateral ++ ++ + + and ventral posterior inferior nuclei, as well as Based on [22, 79, 466, 596, 648] the parvocellular part of the ventral posteromedial nucleus. The ventromedial posterior nucleus (VMPo), part of a shell surrounding the ventral posteri- which is situated in the posterobasal part of or thalamic complex. These nuclei receive af- the thalamus (Fig. 8.2) [57, 118], is in receipt ferents from the vestibular nuclear complex of nociceptive and thermoreceptive spinotha- and project to several cortical areas, including lamic and trigeminothalamic lamina I neurons an area in the posterosuperior insula and adja- [98, 117]. This nucleus sends a somatotopically cent operculum known as the parietoinsular organized projection to the posterosuperior vestibular cortex, PIVC [79]. This vestibular part of the insular cortex [72, 118]. This pro- cortical area overlaps with the INTC. jection area may be designated as the insular The most medial sector of the parvocellular nociceptive and thermoreceptive cortex INTC. part of the ventral posteromedial thalamic nu- The ventral posterior superior (VPS) and cleus (VMpc,m) is in receipt of gustatory pro- ventral posterior inferior (VPI) nuclei form jections from the most rostral part of the nu- 15 Telencephalon: Neocortex 651

Fig. 15.72 A,B. The insula of the rhesus monkey. A Reconstruction of the insula and surrounding regions. The three types of insular cortex, agranular (Ia), dysgranular (Idg) and granular (Ig) are indicated with different hues of red. The heavy grey line represents the circular insular sulcus. Modified from Mesulam and Mufson [466]. B The approximate positions of functional areas in the insula and the course of some fibre connections. Primary sensory areas are shaded. AA, agranular anterior zone; AI, AII, primary and secondary auditory areas; GI, II, primary and secondary gustatory areas; ILC, insular limbic cortex; INTC, insular nociceptive and thermoreceptive cortex; ISAC, insular somatic association cortex; IVSC, insular viscerosensory cortex; OCI, primary olfactory cortex; PA, postauditory cortex; PV, parietal ventral area; PIVC, parietoinsular vestibular cortex; RI, retroinsular cortex; SII, secondary somatosensory area; STP, superior temporal plane; TP, temporopolar cortex 652 Section II Structure of Spinal Cord and Brain Parts cleus of the solitary tract. This medial sector network”, functioning in the analysis and inte- of the VPMpc projects to the granular antero- gration of food-related sensory information superior part of the insula and the adjacent (Figs. 11.7, 11.8) [89]. portion of the frontal operculum. This area, The insula receives afferents from the pri- which represents the primary gustatory cortex, mary somatosensory cortex SI, the somatosen- GI, projects to a more basally situated dysgra- sory association areas 5 and 7b, the primary nular and agranular insular area which, hence, vestibular areas 3a and 2v (located in the basal may be designated as the secondary gustatory part of the central sulcus and the anterior tip cortex, GII [596]. of the intraparietal sulcus, respectively) and Via a synaptic relay in the external medial auditory association areas situated in the ante- parabrachial nucleus, the intermediate and lat- rior and posterior parts of the temporal plane. eral sectors of VPMpc receive viscerosensory All of these sensory cortical areas project to information from the intermediate and caudal the posterosuperior part of the insula, which parts of the nucleus of the solitary tract and may hence be characterized as the insular so- project to an insular area located directly pos- matic association cortex (ISAC). Some high-or- terior to the gustatory areas. This area, which der association areas, including the anterior is known as the insular viscerosensory cortex, orbitofrontal cortex (area 11), the prefrontal IVSC, shows an organotopic ordering. Physio- cortex (areas 45, 46) and the polymodal sen- logical mapping experiments revealed that neu- sory association cortex occupying the banks of rons responding to gastrointestinal sensations the superior temporal sulcus, are also known are located in its anterior part, whereas neu- to be connected with the insula. rons responding to cardiovascular and respira- Finally, it may be mentioned that a consider- tory afferents are located more posteriorly able number of limbic cortical areas, including [648]. A similar topographical organization is the entorhinal, perirhinal, temporopolar, poste- observed in the postgustatory part of the nu- rior orbitofrontal and cingulate cortices, as cleus of the solitary tract, where axons from well as the amygdaloid complex, are recipro- the gastrointestinal tract end more rostrally cally connected with agranular and dysgranu- than those from cardiovascular and respiratory lar sectors in the anterior and anterobasal structures. Hence, it seems likely that this to- parts of the insula. We designate these areas pographical and viscerotopical ordering is collectively as the insular limbic cortex, ILC.It maintained throughout the central viscerosen- has been suggested [466] that the insular so- sory pathway [648]. matic association cortex and the insular limbic If we survey that data concerning the thalam- cortex represent way stations in a somatolim- ic projections just discussed, it appears that the bic projection (Fig. 15.72 B) and that this pro- superior tier of the insula contains four antero- jection may provide a means for interrelating posteriorly arranged primary sensory areas: events in the extrapersonal world with relevant gustatory, viscerosensory, somatosensory (pain motivational states. and temperature) and vestibular (Fig. 15.72 B). The insular cortex has been implicated in ol- The insula is also in receipt of primary and factory, gustatory, viscerosensory, visceromotor, higher-order sensory projections from other somatosensory and vestibular functions. All of cortical areas. Fibres from the primary olfac- these functions, except for the visceromotor tory (prepiriform and periamygdaloid) cortex ones, can be readily associated with structures project to an agranular anterior zone (AA) of and connections discussed above, which are the insula. This zone also receives afferents diagrammatically summarized in Fig. 15.72 B. from the primary and secondary gustatory cor- Olfaction. Electrical stimulation of the ante- tices and from the primary viscerosensory cor- rior insula in humans may lead to olfactory tex (Fig. 15.72 B). As discussed in Chap. 11, it sensations [546]. participates with the adjacent caudal orbito- Taste. The results of both stimulation and frontal cortex in the formation of an “orbital ablation experiments in monkeys point to the 15 Telencephalon: Neocortex 653 presence of a gustatory centre in the anterior Somatosensory Functions. Physiological stud- insula [28, 714]. The detection of taste-sensi- ies in rhesus monkeys [617, 618, 661] have tive neurons in the same area further docu- shown that neurons in the granular part of the ments the presence of this centre [838]. Verha- insula respond to innocuous cutaneous stimuli. gen et al. [777] recently reported that, apart A PET study [78] revealed that the human in- from numerous taste-sensitive neurons, the sula can be activated by vibrotactile stimula- primary gustatory cortex in the rhesus monkey tion. The involvement of the insula in proto- also contains elements responding to non-taste pathic sensibility is documented by clinical properties of oral stimuli, related to the texture evidence. Thus, Biemond [52] described a pa- (viscosity, grittiness) or temperature of food. tient in whom a lesion involving the insula Neurons responding to combinations of these and SII was associated with a dramatic loss of inputs also appeared to be present. These ob- pain perception, and Birklein et al. [53] re- servations further substantiate the concept that cently reported that isolated insular infarction the anterior insula and the adjacent caudal or- may lead to contralateral elimination of cold, bitofrontal cortex are involved in the analysis cold pain and pinprick perception. and integration of food-related information. Vestibular Functions. Neurons in the parieto- Viscerosensory Functions. Electrical stimula- insular vestibular cortex (PIVC) of the rhesus tion of the insula in humans produces nausea monkey respond to vestibular stimuli [262]. and a variety of gastric and abdominal sensa- However, most of these elements also re- tions [546, 547]. Neuroimaging studies in hu- sponded to somatosensory and visual stimula- mans demonstrated activation of an insular tion and were, hence, classified as polymodal area, just caudal to the gustatory cortex, with vestibular units [263]. visceral stimuli such as air hunger, maximal Finally, it may be mentioned that some re- inspiration and Valsalva manoeuvre [35, 367]. cent imaging studies [725, 726] have shown Visceromotor Control. Electrical stimulation that the volume of the insula is significantly of the insular region in humans may elicit a reduced in schizophrenia. variety of visceromotor phenomena, including vomiting and other alterations of gastrointestinal motility, respiratory arrest, as well as changes in heart rate and blood pressure [521, References 546]. Ischaemic strokes involving the insula are frequently accompanied by atrial fibrillation and electrocardiographic abnormalities [102, 1. Abeles M, Goldstein MH Jr (1970) Functional archi- 470, 520, 732]. These findings indicate that the tecture in cat primary auditory cortex: columnar or- insula plays a role in autonomic regulation. It ganization and organization according to depth. J Neurophysiol 33:172–187 is known that the subgenual part of the cingulate 2. Aboitiz F, Garcia VR (1997) The evolutionary origin gyrus (BA25), which is involved in similar func- of the language areas in the human brain. A neuro- tions, projects to several subcortical autonomic anatomical perspective. Brain Res Rev 25:381–396 control centres, including the perifornical part 3. Absher JR, Benson DF (1993) Disconnection syndromes: an overview of Geschwind’s contributions. of the hypothalamus, the periaqueductal grey Neurology 43:862–867 and the parabrachial nuclei [9, 96, 517]. How- 4. Ahmed B, Anderson JC, Douglas RJ, Martin KAC, ever, such connections have so far not been de- Nelson JC (1994) Polyneuronal innervation of spiny scribed for the insular cortex. It is remarkable stellate neurons in cat visual cortex. J Comp Neurol 341:39–49 that the cardiac dysfunctions mentioned occur 5. Allman JM, Kaas JH (1974) The organization of the much more frequently after right-sided strokes second visual area (VII) in the owl monkey: a sec- than after left-sided ones [102, 470]. This find- ond order transformation of the visual hemifield. ing suggests that the autonomic control of cardi- Brain Res 76:247–265 6. Amunts K, Schleicher A, Burgel U et al (1999) Bro- ac activity is lateralized and is mediated by the ca’s region revisited: cytoarchitecture and intersub- right-sided insular cortex. ject variability. J Comp Neurol 412:319–341 654 Section II Structure of Spinal Cord and Brain Parts

7. Amunts K, Malikovic A, Mohlberg H, Schormann 24. Bacci A, Rudolph U, Huguenard JR, Prince DA T, Zilles K (2000) Brodmann’s areas 17 and 18 (2003) Major differences in inhibitory synaptic brought into stereotaxic space – where and how transmission onto two neocortical interneuron variable? Neuroimage 11:66–84 subclasses. J Neurosci 23:9664–9674 8. Amunts K, Schleicher A, Ditterich A, Zilles K (2003) 25. Bacci A, Huguenard JR, Prince DA (2005) Modula- Broca’s region: cytoarchitectonic asymmetry and tion of neocortical interneurons: extrinsic influ- developmental changes. J Comp Neurol 465:72–89 ences and exercises in self-control. Trends Neuros- 9. An X, Bandler R, Öngür D, Price JL (1998) Pre- ci 28:602–610 frontal cortical projections to longitudinal columns 26. Bachman DL, Albert ML (1991) The cerebral orga- in the midbrain periaqueductal gray in macaque nization of language. In: Peters A, Jones EG (eds) monkeys. J Comp Neurol 401:455–479 Normal and altered states of function. Plenum, 10. Andersen RA, Buneo CA (2002) Intentional maps New York, pp 213–262 (Cerebral cortex, vol 9) in posterior parietal cortex. Annu Rev Neurosci 27. Baddeley AD (1986) Working memory. Oxford 25:189–220 University Press, Oxford 11. Andersen RA, Asanuma C, Essick G, Siegel RM 28. Bagshaw MH, Pribram KH (1953) Cortical organi- (1990) Corticocortical connections of anatomically zation in gustation (Macaca mulatta). J Neurophys- and physiologically defined subdivisions within iol 16:499–508 the inferior parietal lobule. J Comp Neurol 296:65– 29. Bailey P, von Bonin G (1950) The isocortex of the 113 chimpanzee. University of Illinois Press, Urbana 12. Annese J, Pitiot A, Dinov ID, Toga AW (2004) A 30. Bailey P, von Bonin G (1951) The isocortex of myelo-architectonic method for the structural clas- man. University of Illinois Press, Urbana sification of cortical areas. Neuroimage 21:15–26 31. Baillarger JGF (1853) De l’étendue de la surface du 13. Aoki C, Kabak S (1992) Cholinergic terminals in cerveau et de ses rapports avec le developpement the cat visual cortex: ultrastructural basis for inter- de l’intelligence. Ann Med Psychol 5:1–9 action with glutamate-immunoreactive neurons 32. Baizer JS, Ungerleider LG, Desimone R (1991) Or- and other cells. Vis Neurosci 8:177–191 ganization of visual inputs to the inferior temporal 14. Apicella P, Ljungberg T, Scarnati E, Schultz W and posterior parietal cortex in macaques. J Neu- (1991) Responses to reward in monkey dorsal and rosci 11:168–190 ventral striatum. Exp Brain Res 85:491–500 33. Ballesteros Yáñez I, Muñoz A, Contreras J et al 15. Apicella P, Scarnati E, Ljungberg T, Schultz W (2005) Double bouquet cell in the human cerebral (1992) Neuronal activity in monkey striatum re- cortex and a comparison with other mammals. J lated to the expectation of predictable environmen- Comp Neurol 486:344–360 tal events. J Neurophysiol 68:945–960 34. Bandler R, Keay KA (1996) Columnar organization 16. Asanuma H (1975) Recent developments in the in the midbrain periaqueductal gray and the inte- study of the columnar arrangement of neurons gration of emotional expression. Progr Brain Res within the motor cortex. Physiol Rev 55:143–156 107:285–300 17. Asanuma H (1987) Cortical motor columns. In: 35. Banzett RB, Mulnier HE, Murphy K et al (2000) Adelman G (ed) Encyclopedia of neuroscience, vol Breathlessness in humans activates insular cortex. 1. Birkhäuser, Boston, pp 281–282 Neuroreport 11:2117–2120 18. Asanuma H, Rosen I (1973) Spread of mono- and 36. Barbas H, Ghashghaei H, Dombrowski SM, Re- polysynaptic connections within cat’s motor cor- mpel-Clower NL (1999) Medial prefrontal cortices tex. Exp Brain Res 16:507–520 are unified by common connections with superior 19. Aston-Jones G, Cohen JD (2005) Adaptive gain temporal cortices and distinguished by input from and the role of the locus coeruleus-norepinephrine memory-related areas in the rhesus monkey. J system in optimal performance. J Comp Neurol Comp Neurol 410:343–367 493: 99–110 37. Barnes CL, Pandya DN (1992) Efferent cortical 20. Aston-Jones G, Cohen JD (2005) An integrative connections of multimodal cortex of the superior theory of locus coeruleus-norepinephrine function: temporal sulcus in the rhesus monkey. J Comp adaptive gain and optimal performance. Annu Rev Neurol 318:222–244 Neurosci 28:403–450 38. Barrett NA, Large MM, Smith GL et al (2001) Hu- 21. Aston-Jones G, Chiang C, Alexinsky T (1991) Dis- man cortical processing of colour and pattern. charge of noradrenergic locus coeruleus neurons Hum Brain Mapp 13:213–225 in behaving rats and monkeys suggests a role in 39. Bartels A, Zeki S (2004) The neural correlates of vigilance. Prog Brain Res 88:501–520 maternal and romantic love. Neuroimage 21:1155– 22. Augustine JR (1996) Circuitry and functional as- 1166 pects of the insular lobe in primates including hu- 40. Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi mans. Brain Res Rev 22:229–244 A (2000) In vivo fiber tractography using DT-MRI 23. Bacci A, Huguenard JR, Prince DA (2003) Func- data. Magn Reson Med 44:625–632 tional autaptic neurotransmission in fast-spiking 41. Baxter MG, Murray EA (2002) The amygdala and interneurons: a novel form of feedback inhibition reward. Nat Rev Neurosci 3:563–573 in the neocortex. J Neurosci 23:859–866 15 Telencephalon: Neocortex 655

42. Baylis GC, Rolls ET, Leonard CM (1987) Functional the posterior portion of the ventral medial nu- subdivisions of the temporal lobe neocortex. J cleus, in the human thalamus. Brain 123:601–619 Neurosci 7:330–342 58. Bookheimer S (2002) Functional MRI of language: 43. Beaulieu C, Somogyi P (1991) Enrichment of cho- new approaches to understanding the cortical or- linergic synaptic terminals on GABAergic neurons ganization of semantic processing. Annu Rev and coexistence of immunoreactive GABA and Neurosci 25:151–188 choline acetyltransferase in the same synaptic 59. Bouret S, Sara SJ (2005) Network reset: a simpli- terminals in the striate cortex of the cat. J Comp fied overarching theory of locus coeruleus nora- Neurol 304:666–680 drenaline function. Trends Neurosci 28:574–582 44. Beck E (1949) A cytoarchitectural investigation 60. Bouvier SE, Engel SA (2006) Behavioral deficits into the boundaries of cortical areas 13 and 14 in and cortical damage loci in cerebral achromatop- the human brain. J Anat 83:147–157 sia. Cereb Cortex 16:183–191 45. Beierlein M, Gibson JR, Connors BW (2000) A net- 61. Bowman EM, Aigner TG, Richmond BJ (1996) work of electrically coupled interneurons drives Neural signals in the monkey ventral striatum re- synchronized inhibition in neocortex. Nat Neurosci lated to motivation for juice and cocaine rewards. 3:904–910 J Neurophysiol 75:1061–1073 46. Benes FM, Berretta S (2001) GABAergic interneu- 62. Braak H (1976) A primitive gigantopyramidal field rons: implications for understanding schizophrenia buried in the depth of the cingulate sulcus of the and bipolar disorder. Neuropsychopharmacology human brain. Brain Res 109:219–223 25:1–27 63. Braak H (1976) On the striate area of the human 47. Benevento LA, Fallon J, Davis BJ, Rezak M (1977) isocortex. A Golgi- and pigmentarchitectonic Auditory–visual interaction in single cells in the study. J Comp Neurol 166:341–364 cortex of the superior temporal sulcus and the or- 64. Braak H (1980) Architectomics of the human bital frontal cortex of the macaque monkey. Exp telencephalic cortex. Springer, Berlin Heidelberg Neurol 57:849–872 New York 48. Benson DF (1985) Aphasia and related disorders: a 65. Brain R (1961) Speech disorders. Aphasia, apraxia clinical approach. In: Mesulam M-M (ed) Princi- and agnosia. Butterworth, London ples of behavioral neurology. FA Davis Company, 66. Brett M, Johnsrude IS, Owen AM (2002) The prob- Philadelphia, pp 193–238 lem of functional localization in the human brain. 49. Berendse HW, Groenewegen HJ (1991) Restricted Nat Rev Neurosci 3:243–249 cortical termination fields of the midline and in- 67. Broca P (1865) Sur le siège de la faculté du langage tralaminar thalamic nuclei in the rat. Neuroscience articulé. Bull Soc d’Anthropol Paris 6:377–393 42:73–102 68. Broca P (1878) Anatomie comparée des circonvolu- 50. Berman RA, Colby CL, Genovese CR et al (1999) tions cerebrales: le grand lobe limbique et la scis- Cortical networks subserving pursuit and saccadic sure dans la serie des mammiferes. Rev Anthropol eye movements in humans: an FMRI study. Hum (Paris) 2:285–498 Brain Mapp 8:209–225 69. Brodal A (1981) Neurological anatomy in relation 51. Betz W (1874) Anatomischer Nachweis zweier Ge- to clinical medicine, 3rd edn. Oxford University hirnzentralen. Zentralbl Med Wiss 12:578–580 Press, New York, IX, 1053 p 52. Biemond A (1956) The conduction of pain above 70. Brodmann K (1909) Vergleichende Lokalisations- the level of the thalamus opticus. Arch Neurol Psy- lehre der Großhirnrinde in ihren Prinzipien darge- chiatry 75:221–231 stellt auf Grund des Zellenbaues. Barth, Leipzig 53. Birklein F, Rolke R, Muller-Forell W (2005) Iso- 71. Brodmann K (1914) Physiologie des Gehirns. In: lated insular infarction eliminates contralateral Von Bruns P (ed) Neue Deutsche Chirurgie, Vol cold, cold pain, and pinprick perception. Neurol- 11. Enke, Stuttgart, pp 85–426 ogy 65:1381 72. Brooks JC, Zambreanu L, Godinez A, Craig AD, 54. Blatow M, Rozov A, Katona I et al (2003) A novel Tracey I (2005) Somatotopic organisation of the network of multipolar bursting interneurons gen- human insula to painful heat studied with high re- erates theta frequency oscillations in neocortex. solution functional imaging. Neuroimage 27:201– Neuron 38:805–817 209 55. Blatt GJ, Rosene DL (1998) Organization of direct 73. Brown-Séquard C-E (1877) Aphasia as an effect of hippocampal efferent projections to the cerebral brain-disease. Dublin J Med Sci 63:209–225 cortex of the rhesus monkey: projections from 74. Bruce C, Desimone R, Gross CG (1981) Visual CA1, prosubiculum, and subiculum to the temporal properties of neurons in a polysensory area in lobe. J Comp Neurol 392:92–114 superior temporal sulcus of the macaque. J Neuro- 56. Blinkov SM, Glezer I (1968) Das Zentralnervensys- physiol 46:369–384 tem in Zahlen und Tabellen. Fischer, Jena 75. Brugge JF, Reale RA (1985) Auditory cortex. In: 57. Blomqvist A, Zhang ET, Craig AD (2000) Cy- Peters A, Jones EG (eds) Association and auditory toarchitectonic and immunohistochemical charac- cortices. Plenum, New York, pp 229–271 (Cerebral terization of a specific pain and temperature relay, cortex, vol 4) 656 Section II Structure of Spinal Cord and Brain Parts

76. Bugbee NM, Goldman-Rakic PS (1983) Columnar 94. Cavada C, Goldman-Rakic PS (1989) Posterior organization of corticocortical projections in parietal cortex in rhesus monkey: I. Parcellation squirrel and rhesus monkeys: similarity of column of areas based on distinctive limbic and sensory width in species differing in cortical volume. J corticocortical connections. J Comp Neurol Comp Neurol 220:355–364 287:393–421 77. Burnstine TH, Lesser RP, Hart J Jr et al (1990) 95. Cavada C, Goldman-Rakic PS (1989) Posterior Characterization of the basal temporal language parietal cortex in rhesus monkey: II. Evidence area in patients with left temporal lobe epilepsy. for segregated corticocortical networks linking Neurology 40:966–970 sensory and limbic areas with the frontal lobe. J 78. Burton H, Videen TO, Raichle ME (1993) Tactile- Comp Neurol 287:422–445 vibration-activated foci in insular and parietal-op- 96. Chiba T, Kayahara T, Nakano K (2001) Efferent ercular cortex studied with positron emission projections of infralimbic and prelimbic areas of tomography: mapping the second somatosensory the medial prefrontal cortex in the Japanese area in humans. Somatosens Mot Res 10:297–308 monkey, Macaca fuscata. Brain Res 888:83–101 79. Büttner-Ennever JA, Gerrits NM (2004) Vestibular 97. Choi HJ, Zilles K, Mohlberg H et al (2006) Cyto- system. In: Paxinos G, Mai JK (eds) The human architectonic identification and probabilistic nervous system, 2nd edn. Elsevier, Amsterdam, pp mapping of two distinct areas within the anterior 1212–1240 ventral bank of the human intraparietal sulcus. J 80. Buxhoeveden DP, Casanova MF (2002) The mini- Comp Neurol 495:53–69 column and evolution of the brain. Brain Behav 98. Christensen BN, Perl ER (1970) Spinal neurons Evol 60:125–151 specifically excited by noxious or thermal stimu- 81. Buxhoeveden DP, Casanova MF (2002) The minicol- li: marginal zone of the dorsal horn. J Neurophy- umn hypothesis in neuroscience. Brain 125:935–951 siol 33:293–307 82. Buzsaki G, Draguhn A (2004) Neuronal oscillations 99. Chu Z, Galarreta M, Hestrin S (2003) Synaptic in cortical networks. Science 304:1926–1929 interactions of late-spiking neocortical neurons 83. Cabeza R, Nyberg L (2000) Imaging cognition II: in layer 1. J Neurosci 23:96–102 An empirical review of 275 PET and fMRI studies. 100. Cipolloni PB, Peters A (1983) The termination of J Cogn Neurosci 12:1–47 callosal fibres in the auditory cortex of the rat. A 84. Cajal SR (1972) Histologie du système nerveux de combined Golgi–electron microscope and degen- l’homme et des vertébrés. Consejo Superior de In- eration study. J Neurocytol 12:713–726 vestigaciones Cientificas. Tome II. Instituto Ramon 101. Cobb SR, Halasy K, Vida I et al (1997) Synaptic y Cajal, Madrid effects of identified interneurons innervating 85. Campbell AW (1905) Histological studies on the both interneurons and pyramidal cells in the rat localization of cerebral function. Cambridge Uni- hippocampus. Neuroscience 79:629–648 versity Press, Cambridge, MA 102. Colivicchi F, Bassi A, Santini M, Caltagirone C 86. Carmichael ST, Price JL (1994) Architectonic subdi- (2004) Cardiac autonomic derangement and ar- vision of the orbital and medial prefrontal cortex in rhythmias in right-sided stroke with insular in- the macaque monkey. J Comp Neurol 346:366–402 volvement. Stroke 35:2094–2098 87. Carmichael ST, Price JL (1995) Limbic connections 103. Colonnier ML (1966) The structural design of of the orbital and medial prefrontal cortex in ma- the neocortex. In: Eccles JC (ed) Brain and con- caque monkeys. J Comp Neurol 363:615–641 scious experience. Springer, Berlin Heidelberg 88. Carmichael ST, Price JL (1995) Sensory and pre- New York, pp 1–23 motor connections of the orbital and medial pre- 104. Colonnier ML (1981) The electron-microscopic frontal cortex of macaque monkeys. J Comp Neu- analysis of the neuronal organization of the cere- rol 363: 642–664 bral cortex. In: Adelman G, Dennis SG, Schmitt 89. Carmichael ST, Price JL (1996) Connectional net- FO, Worden FG (eds) The organization of the works within the orbital and medial prefrontal cortex cerebral cortex. MIT Press, Cambridge, pp 125– of macaque monkeys. J Comp Neurol 371:179–207 153 90. Carmichael ST, Clugnet MC, Price JL (1994) Cen- 105. Condé F, Lund JS, Jacobowitz DM, Baimbridge tral olfactory connections in the macaque monkey. KG, Lewis DA (1994) Local circuit neurons im- J Comp Neurol 346:403–434 munoreactive for calretinin, calbindin D-28k or 91. Carrive P, Morgan MM (2004) Periaqueductal gray. parvalbumin in monkey prefrontal cortex: distri- In: Paxinos G, Mai JK (eds) The human nervous bution and morphology. J Comp Neurol 341:95– system, 2nd edn. Elsevier, Amsterdam, pp 393–423 116 92. Catani M, ffytche DH (2005) The rises and falls of 106. Connor JR, Peters A (1984) Vasoactive intestinal disconnection syndromes. Brain 128:2224–2239 polypeptide-immunoreactive neurons in rat vi- 93. Catsman-Berrevoets CE, Kuypers HGJM (1981) A sual cortex. Neuroscience 12:1027–1044 search for corticospinal collaterals to thalamus and 107. Connors BW, Long MA (2004) Electrical synapses mesencephalon by means of multiple retrograde in the mammalian brain. Annu Rev Neurosci fluorescent tracers in cat and rat. Brain Res 27:393–418 218:15–33 15 Telencephalon: Neocortex 657

108. Conturo TE, Lori NF, Cull TS et al (1999) Track- 125. Cusick CG (1997) The superior temporal poly- ing neuronal fiber pathways in the living human sensory region in monkeys. In: Rockland KS, brain. Proc Natl Acad Sci USA 96:10422–10427 Kaas JH, Peters A (eds) Extrastriate cortex in pri- 109. Cooke A, Zurif EB, DeVita C et al (2002) Neural mates. Plenum, New York, pp 435–468 (Cerebral basis for sentence comprehension: grammatical cortex, vol 12) and short-term memory components. Hum Brain 126. Cusick CG, Kaas JH (1986) Interhemispheric con- Mapp 15:80–94 nections of cortical sensory and motor represen- 110. Cools AR, Jaspers R, Schwartz M et al (1984) Ba- tations in primates. In: Lepore F, Ptito M, Jasper sal ganglia and switching motor programs. In: HH (eds) Two hemispheres – one brain. Func- McKenzie JS, Kemm RE, Wilcock LN (eds) Basal tions of the corpus callosum. Liss, New York, pp ganglia structure and function. Plenum, New 83–102 York, pp 513–544 127. Cusick CG, Seltzer B, Cola M, Griggs E (1995) 111. Cools AR, Van den Bercken JHL, Horstink MWI, Chemoarchitectonics and corticocortical termina- Van Spaendonck KMP, Berger HJC (1984) Cogni- tions within the superior temporal sulcus of the tive and motor shifting aptitude in Parkinson’s rhesus monkey: evidence for subdivisions of disease. J Neurol Neurosurg Psychiatry 47:443– superior temporal polysensory cortex. J Comp 453 Neurol 360:513–535 112. Corballis MC (2003) From mouth to hand: ges- 128. Czeiger D, White EL (1993) Synapses of extrinsic ture, speech, and the evolution of right-handed- and intrinsic origin made by callosal projection ness. Behav Brain Sci 26:199–208 neurons in mouse visual cortex. J Comp Neurol 113. Corbetta M, Miezin FM, Dobmeyer S, Shulman 330:502–513 GL, Petersen SE (1991) Selective and divided at- 129. Damasio AR (1992) Aphasia. N Engl J Med tention during visual discriminations of shape, 326:531–539 color, and speed: functional anatomy by positron 130. Damasio H, Damasio AR (1980) The anatomical emission tomography. J Neurosci 11:2383–2402 basis of conduction aphasia. Brain 103:337–350 114. Coren S, Porac C (1977) Fifty centuries of right- 131. Dancause N, Barbay S, Frost SB et al (2006) Ipsilat- handedness: the historical record. Science 198: eral connections of the ventral premotor cortex in 631–632 a new world primate. J Comp Neurol 495:374–390 115. Courchesne E, Pierce K (2005) Why the frontal 132. de Lafuente V, Romo R (2004) Language abilities cortex in autism might be talking only to itself: of motor cortex. Neuron 41:178–180 local over-connectivity but long-distance discon- 133. de Lima AD, Morrison JH (1989) Ultrastructural nection. Curr Opin Neurobiol 15:225–230 analysis of somatostatin-immunoreactive neurons 116. Courtney SM, Ungerleider LG, Keil K, Haxby JV and synapses in the temporal and occipital cortex (1996) Object and spatial visual working memory of the macaque monkey. J Comp Neurol 283:212– activate separate neural systems in human cortex. 227 Cereb Cortex 6:39–49 134. DeFelipe J (1997) Types of neurons, synaptic 117. Craig AD, Kniffki K-D (1985) Spinothalamic lum- connections and chemical characteristics of cells bosacral lamina I cells responsive to skin and immunoreactive for calbindin-D28K, parvalbu- muscle stimulation in the cat. J Physiol (Lond) min and calretinin in the neocortex. J Chem 365:197–221 Neuroanat 14:1–19 118. Craig AD, Bushnell MC, Zhang E-T, Blomqvist A 135. DeFelipe J (1999) Chandelier cells and epilepsy. (1994) A thalamic nucleus specific for pain and Brain 122:1807–1822 temperature sensation. Nature 372:770–773 136. DeFelipe J (2002) Cortical interneurons: from Ca- 119. Creutzfeldt OD (1983) Cortex cerebri. Springer, jal to 2001. Prog Brain Res 136:215–238 Berlin Heidelberg New York 137. DeFelipe J, Farinas I (1992) The pyramidal neu- 120. Crick F, Jones E (1993) Backwardness of human ron of the cerebral cortex: morphological and neuroanatomy. Nature 361:109–110 chemical characteristics of the synaptic inputs. 121. Critchley HD, Rolls ET (1996) Hunger and satiety Prog Neurobiol 39:563–607 modify the responses of olfactory and visual neu- 138. DeFelipe J, Conley M, Jones EG (1986) Long- rons in the primate orbitofrontal cortex. J Neuro- range focal collateralization of axons arising from physiol 75:1673–1686 corticocortical cells in monkey sensory-motor 122. Crosby EC, Humphrey T, Lauer EW (1962) Cor- cortex. J Neurosci 6:3749–3766 relative anatomy of the nervous system. MacMil- 139. DeFelipe J, Hendry SHC, Jones EG (1989) Sy- lan, New York napses of double bouquet cells in monkey cere- 123. Crow TJ (1997) Schizophrenia as failure of hemi- bral cortex visualized by calbindin immunoreac- spheric dominance for language. Trends Neurosci tivity. Brain Res 503:49–54 20:339–343 140. DeFelipe J, Hendry SHC, Hashikawa T, Molinari 124. Culham JC, Kanwisher NG (2001) Neuroimaging M, Jones EG (1990) A microcolumnar structure of cognitive functions in human parietal cortex. of monkey cerebral cortex revealed by immuno- Curr Opin Neurobiol 11:157–163 cytochemical studies of double bouquet cell axons. Neuroscience 37:655–673 658 Section II Structure of Spinal Cord and Brain Parts

141. DeFelipe J, Hendry SHC, Hashikawa T, Jones EG 158. Dronkers NF (1996) A new brain region for coor- (1991) Synaptic relationships of serotonin-immu- dinating speech articulation. Nature 384:159–161 noreactive terminal baskets on GABA neurons in 159. Dronkers NF (2001) Language, neural basis of. the cat auditory cortex. Cereb Cortex 1:117–133 In: Wilson RA, Keil FC (eds) The MIT encyclope- 142. DeFelipe J, Alonso-Nanclares L, Arellano JI dia of the cognitive sciences. MIT Press, Cam- (2002) Microstructure of the neocortex: compara- bridge, pp 448–451 tive aspects. J Neurocytol 31:299–316 160. Drury HA, Van Essen DC, Anderson CH et al 143. Déjerine JJ (1891) Sur un cas de cécité verbale (1996) Computerized mappings of the cerebral avec agraphie, suivi d’autopsie. CR Hebdoma- cortex: a multiresolution flattening method and a daire Séances Mém Soc Biol 43:187–201 surface-based coordinate system. J Cogn Neurosci 144. Déjerine JJ (1892) Contribution à l’étude anato- 8:1–28 mopathologique et clinique des différentes varié- 161. Drury HA, Van Essen DC, Corbetta M, Snyder tés de cécité verbale. C R Hebdomadaire Séances AZ (1999) Surface-based analyses of the human Mém Soc Biol 44:61–90 cerebral cortex. In: Toga AW (ed) Brain warping. 145. Déjerine JJ (1895) Anatomie des centres nerveux. Academic Press, New York, pp 337–363 Rueff et Cie, Paris 162. Dum RP, Strick PL (1991) The origin of corti- 146. del Rio MR, DeFelipe J (1995) A light and elec- cospinal projections from the premotor areas in tron microscopic study of calbindin D-28k im- the frontal lobe. J Neurosci 11:667–689 munoreactive double bouquet cells in the human 163. Dum RP, Strick PL (1996) Spinal cord termina- temporal cortex. Brain Res 690:133–140 tions of the medial wall motor areas in macaque 147. Demb JB, Desmond JE, Wagner AD et al (1995) monkeys. J Neurosci 16:6513–6525 Semantic encoding and retrieval in the left inferi- 164. Dum RP, Strick PL (2005) Frontal lobe inputs to or prefrontal cortex: a functional MRI study of the digit representations of the motor areas on task difficulty and process specificity. J Neurosci the lateral surface of the hemisphere. J Neurosci 15:5870–5878 25:1375–1386 148. Descarries L, Mechawar N (2000) Ultrastructural 165. Duus P (1976) Neurologisch-topische Diagnostik. evidence for diffuse transmission by monoamine Thieme, Stuttgart and acetylcholine neurons of the central nervous 166. Eccles JC (1984) The cerebral neocortex: a theory system. Prog Brain Res 125:27–47 of its operation. In: Jones EG, Peters A (eds) 149. Deschênes M, Labelle A, Landry P (1979) Morpho- Functional properties of cortical cells. Plenum, logical characterization of slow and fast pyramidal New York, pp 1–36 (Cerebral cortex, vol 2) tract cells in the cat. Brain Res 178:251–274 167. Edmunds SM, Parnavelas JG (1982) Retzius-Cajal 150. Desimone R, Gross CG (1979) Visual areas in the cells: an ultrastructural study in the developing temporal cortex of the macaque. Brain Res 178: visual cortex of the rat. J Neurocytol 11:427–446 363–380 168. Egner T, Hirsch J (2005) Cognitive control mech- 151. Devinsky O (1992) Behavioral neurology. Hodder anisms resolve conflict through cortical amplifi- and Staughton, London cation of task-relevant information. Nat Neurosci 152. Ding SL, Van Hoesen G, Rockland KS (2000) In- 8:1784–1790 ferior parietal lobule projections to the presubi- 169. Eickhoff SB, Schleicher A, Zilles K, Amunts K culum and neighboring ventromedial temporal (2006) The human parietal operculum. I. Cytoar- cortical areas. J Comp Neurol 425:510–530 chitectonic mapping of subdivisions. Cereb Cor- 153. Disbrow E, Roberts T, Krubitzer L (2000) Soma- tex 16:254–267 totopic organization of cortical fields in the lat- 170. Eickhoff SB, Amunts K, Mohlberg H, Zilles K eral sulcus of Homo sapiens: evidence for SII and (2006) The human parietal operculum. II. Stereo- PV. J Comp Neurol 418:1–21 taxic maps and correlation with functional imag- 154. Disbrow E, Litinas E, Recanzone GH, Padberg J, ing results. Cereb Cortex 16:268–279 Krubitzer L (2003) Cortical connections of the 171. Einstein G, Fitzpatrick D (1991) Distribution and second somatosensory area and the parietal ven- morphology of area 17 neurons that project to tral area in macaque monkeys. J Comp Neurol the cat’s extrastriate cortex. J Comp Neurol 462:382–399 303:132–149 155. Distler C, Boussaoud D, Desimone R, Ungerleider 172. Elhanany E, White EL (1990) Intrinsic circuitry: LG (1993) Cortical connections of inferior tem- synapses involving the local axon collaterals of poral area TEO in macaque monkeys. J Comp corticocortical projection neurons in the mouse Neurol 334:125–150 primary somatosensory cortex. J Comp Neurol 156. Dougherty RF, Ben Shachar M, Bammer R, 291:43–54 Brewer AA, Wandell BA (2005) Functional orga- 173. Elias H, Schwartz D (1969) Surface areas of the nization of human occipital-callosal fiber tracts. cerebral cortex of mammals determined by Proc Natl Acad Sci USA 102:7350–7355 stereological methods. Science 166:111–113 157. Douglas R, Martin KAC, Whitteridge D (1989) A 174. Elston GN, Gonzalez-Albo MC (2003) Parvalbu- canonical microcircuit for neocortex. Neural min-, calbindin-, and calretinin-immunoreactive Comput 1:480–488 neurons in the prefrontal cortex of the owl mon- 15 Telencephalon: Neocortex 659

key (Aotus trivirgatus): a standardized quantita- 191. Flechsig P (1896) Gehirn und Seele. Veit & Co., tive comparison with sensory and motor areas. Leipzig Brain Behav Evol 62:19–30 192. Flechsig P (1901) Developmental (myelogenetic) 175. Everitt BJ, Robbins TW (2005) Neural systems of localisation of the cerebral cortex in the human reinforcement for drug addiction: from actions to subject. Lancet 2:1027–1029 habits to compulsion. Nat Neurosci 8:1481–1489 193. Flechsig P (1920) Anatomie des menschlichen 176. Fairén A, Valverde F (1980) A specialized type of Gehirns und Rückenmarks auf myelogenetischer neuron in the visual cortex of cat: a Golgi and Grundlage. Thieme, Leipzig electron microscope study of chandelier cells. J 194. Fleischhauer K (1974) On different patterns of Comp Neurol 194:761–779 dendritic bundling in the cerebral cortex of the 177. Fairén A, DeFelipe J, Regidor J (1984) Nonpyra- cat. Z Anat Entwicklungsgesch 143:115–126 midal neurons: general account. In: Peters A, 195. Fleischhauer K (1978) Cortical architectonics: the Jones EG (eds) Cellular components of the cere- last 50 years and some problems of today. In: bral cortex. Plenum, New York, pp 201–254 (Cer- Brazier MAB, Petsche H (eds) Architectonics of ebral cortex, vol 1) the cerebral cortex. Raven, New York, pp 99–117 178. Feldman ML (1984) Morphology of the neocorti- 196. Fleischhauer K, Petsche H, Wittkowski W (1972) cal pyramidal neuron. In: Peters A, Jones EG Vertical bundles of dendrites in the neocortex. Z (eds) Cellular components of the cerebral cortex. Anat Entwicklungsgesch 136:213–223 Plenum, New York, pp 123–200 (Cerebral cortex, 197. Foerster O (1936) Motorische Felder und Bahnen. vol 1) In: Bumke O, Foerster O (eds) Handbuch der 179. Feldman ML, Peters A (1978) The forms of non- Neurologie, Bd 6. Springer, Berlin, Heidelberg, pyramidal neurons in the visual cortex of the rat. pp 1–357 J Comp Neurol 179:761–793 198. Foote SL (1987) Extrathalamic modulation of cor- 180. Felleman DJ, Van Essen DC (1991) Distributed tical function. Annu Rev Neurosci 10:67–95 hierarchical processing in the primate cerebral 199. Foote SL, Aston-Jones G, Bloom FE (1980) Im- cortex. Cereb Cortex 1:1–47 pulse activity of locus coeruleus neurons in 181. Felleman DJ, Burkhalter A, Van Essen DC (1997) awake rats and monkeys is a function of sensory Cortical connections of areas V3 and VP of ma- stimulation and arousal. Proc Natl Acad Sci USA caque monkey extrastriate visual cortex. J Comp 77:3033–3037 Neurol 379:21–47 200. Frackowiak RS (1994) Functional mapping of 182. Ferrier D (1876) The functions of the brain. verbal memory and language. Trends Neurosci Smith, Elder & Comp., London 17:109–115 183. Ferrier D, Yeo G (1885) A record of experiments 201. Frackowiak RS, Friston KJ (1994) Functional on the effects of lesions of different regions of neuroanatomy of the human brain: positron the cerebral hemispheres. Phil Trans R Soc Lond emission tomography – a new neuroanatomical B 175:479–564 technique. J Anat 184:211–225 184. Ferstl EC, von Cramon DY (2002) What does the 202. Freeman W, Watts JW (1942) Psychosurgery: in- frontomedian cortex contribute to language pro- telligence, emotion and social behavior following cessing: coherence or theory of mind? Neuro- prefrontal lobotomy for mental disorders. Tho- image 17:1599–1612 mas, Springfield 185. Fiez JA (1997) Phonology, semantics, and the role 203. Freeman W, Watts JW (1950) Psychosurgery in of the left inferior prefrontal cortex. Hum Brain the treatment of mental disorders and intractable Mapp 5:79–83 pain, 2nd edn. Thomas, Springfield 186. Filimonov IN (1932) Über die Variabilität der 204. Freese JL, Amaral DG (2005) The organization of Grosshirn Rindenstruktur. Mitt. 2: Regio occipi- projections from the amygdala to visual cortical talis beim erwachsenen Menschen. J Psychol areas TE and V1 in the macaque monkey. J Neurol 44:1–96 Comp Neurol 486:295–317 187. Finger S (1994) Origins of neuroscience. Oxford 205. Freese JL, Amaral DG (2006) Synaptic organiza- University Press, New York tion of projections from the amygdala to visual 188. Fink GR, Frackowiak RS, Pietrzyk U, Passingham cortical areas TE and V1 in the macaque mon- RE (1997) Multiple nonprimary motor areas in key. J Comp Neurol 496:655–667 the human cortex. J Neurophysiol 77:2164–2174 206. Freund TF (2003) Interneuron Diversity series: 189. Fitzpatrick D, Itoh K, Diamond IT (1983) The la- rhythm and mood in perisomatic inhibition. minar organization of the lateral geniculate body Trends Neurosci 26:489–495 and the striate cortex in the squirrel monkey 207. Freund TF, Martin KAC, Smith AD, Somogyi P (Saimiri sciureus). J Neurosci 3:673–702 (1983) Glutamate decarboxylase-immunoreactive 190. Fitzpatrick D, Lund JS, Blasdel GG (1985) Intrin- terminals of Golgi-impregnated axoaxonic cells sic connections of macaque striate cortex: affer- and of presumed basket cells in synaptic contact ent and efferent connections of lamina 4C. J Neu- with pyramidal neurons of the cat’s visual cortex. rosci 5:3329–3349 J Comp Neurol 221:263–278 660 Section II Structure of Spinal Cord and Brain Parts

208. Freund TF, Martin KAC, Soltesz I, Somogyi P, 226. Gazzaniga MS, Hutsler JJ, Keil FC (2001) Hemi- Whitteridge D (1989) Arborisation pattern and spheric specialization. In: Wilson RA, Keil FC postsynaptic targets of physiologically identified (eds) The MIT encyclopedia of the cognitive thalamocortical afferents in striate cortex of the sciences. MIT Press, Cambridge, MA, pp 369–372 macaque monkey. J Comp Neurol 289:315–336 227. Geschwind N (1965) Disconnexion syndromes in 209. Friedman DP, Bachevalier J, Ungerleider LG, animals and man. Brain 88:237–294, 585–644 Mishkin M (1987) Widespread projections to 228. Geschwind N (1970) The organization of lan- layer 1 of primate cortex. Soc Neurosci Abstr guage and the brain. Science 170:940–944 13:251 229. Geschwind N, Levitsky W (1968) Human brain: 210. Fries W (1990) Pontine projection from striate left-right asymmetries in temporal speech region. and prestriate visual cortex in the macaque mon- Science 161:186–187 key: an anterograde study. Vis Neurosci 4:205– 230. Geyer S, Ledberg A, Schleicher A et al (1996) 216 Two different areas within the primary motor 211. Friston K (2002) Beyond phrenology: what can cortex of man. Nature 382:805–807 neuroimaging tell us about distributed circuitry? 231. Geyer S, Schleicher A, Zilles K (1999) Areas 3a, Annu Rev Neurosci 25:221–250 3b, and 1 of human primary somatosensory cor- 212. Fritsch G, Hitzig E (1870) Über die elektrische tex. Neuroimage 10:63–83 Erregbarkeit des Grosshirns. Arch Anat Physiol 232. Geyer S, Schormann T, Mohlberg H, Zilles K Wiss Med 37:300–332 (2000) Areas 3a, 3b, and 1 of human primary so- 213. Frost DO, Caviness VS Jr (1980) Radial organiza- matosensory cortex. Part 2. Spatial normalization tion of thalamic projections to the neocortex in to standard anatomical space. Neuroimage 11: the mouse. J Comp Neurol 194:369–393 684–696 214. Fukuda T, Kosaka T, Singer W, Galuske RA 233. Ghashghaei HT, Barbas H (2002) Pathways for (2006) Gap junctions among dendrites of cortical emotion: interactions of prefrontal and anterior GABAergic neurons establish a dense and wide- temporal pathways in the amygdala of the rhesus spread intercolumnar network. J Neurosci monkey. Neuroscience 115:1261–1279 26:3434–3443 234. Ghosh S, Fyffe RE, Porter R (1988) Morphology 215. Fuster JM (1973) Unit activity in prefrontal cor- of neurons in area 4 gamma of the cat’s cortex tex during delayed-response performance: neuro- studied with intracellular injection of HRP [cor- nal correlates of transient memory. J Neurophy- rected and issued with original paging in J Comp siol 36:61–78 Neurol 1988 Nov 8; 277(2)]. J Comp Neurol 216. Fuster JM (2001) The prefrontal cortex – an up- 277:290–312 date: time is of the essence. Neuron 30:319–333 235. Gibson JR, Beierlein M, Connors BW (1999) Two 217. Fuster JM, Alexander GE (1971) Neuron activity networks of electrically coupled inhibitory neu- related to short-term memory. Science 173:652– rons in neocortex. Nature 402:75–79 654 236. Gil-da-Costa R, Martin A, Lopes MA et al (2006) 218. Gabbott PLA, Martin KAC, Whitteridge D (1987) Species-specific calls activate homologs of Broca’s Connections between pyramidal neurons in layer and Wernicke’s areas in the macaque. Nat Neuro- 5 of cat visual cortex (area 17). J Comp Neurol sci 9:1064–1070 259:364–381 237. Gilbert CD (1992) Horizontal integration and 219. Galaburda AM, Pandya DH (1983) The intrinsic cortical dynamics. Neuron 9:1–13 architectonic and connectional organization of 238. Gilbert SJ, Spengler S, Simons JS, Frith CD, Bur- the superior temporal region of the rhesus mon- gess PW (2006) Differential functions of lateral key. J Comp Neurol 221:169–184 and medial rostral prefrontal cortex (area 10) re- 220. Galaburda AM, Sanides F (1980) Cytoarchitec- vealed by brain-behavior associations. Cereb Cor- tonic organisation of the human auditory cortex. tex 16:1783–1789 J Comp Neurol 190:597–610 239. Giovannini MG, Bartolini L, Kopf SR, Pepeu G 221. Galaburda AM, LeMay M, Kemper TL, Geschwind (1998) Acetylcholine release from the frontal cor- N (1978) Right-left asymmetries in the brain. tex during exploratory activity. Brain Res 784: Science 199:852–856 218–227 222. Galarreta M, Hestrin S (1999) A network of fast- 240. Globus A, Scheibel AB (1967) Pattern and field spiking cells in the neocortex connected by elec- in cortical structure: the rabbit. J Comp Neurol trical synapses. Nature 402:72–75 131:155–172 223. Galarreta M, Hestrin S (2001) Electrical synapses 241. Goebel R, Muckli L, Kim D-S (2004) Visual sys- between GABA-releasing interneurons. Nat Rev tem. In: Paxinos G, Mai JK (eds) The human ner- Neurosci 2:425–433 vous system, 2nd edn. Elsevier, Amsterdam, pp 224. Gazzaniga MS (1989) Organization of the human 1280–1305 brain. Science 245:947–952 242. Goldman PS, Nauta WJH (1977) Columnar distri- 225. Gazzaniga MS (2005) Forty-five years of split- bution of cortico-cortical fibers in the frontal as- brain research and still going strong. Nat Rev sociation, limbic and motor cortex of the devel- Neurosci 6:653–659 oping rhesus monkey. Brain Res 122:393–413 15 Telencephalon: Neocortex 661

243. Goldman-Rakic PS (1984) Modular organization 260. Grossman M, Smith EE, Koenig P et al (2002) of prefrontal cortex. Trends Neurosci 7:419–424 The neural basis for categorization in semantic 244. Goldman-Rakic PS (1995) Cellular basis of work- memory. Neuroimage 17:1549–1561 ing memory. Neuron 14:477–485 261. Grünbaum ASF, Sherrington CS (1903) Observa- 245. Goldman-Rakic PS (1996) The prefrontal land- tions on the physiology of the cerebral cortex of scape: implications of functional architecture for the higher apes. Proc R Soc Lond B Biol Sci understanding human mentation and the central 69:206–209 executive. Philos Trans R Soc Lond B Biol Sci 262. Grüsser OJ, Pause M, Schreiter U (1990) Locali- 351:1445–1453 zation and responses of neurones in the parieto- 246. Goldman-Rakic PS, Schwartz ML (1982) Interdi- insular vestibular cortex of awake monkeys (Ma- gitation of contralateral and ipsilateral columnar caca fascicularis). J Physiol 430:537–557 projections to frontal association cortex in pri- 263. Grüsser OJ, Pause M, Schreiter U (1990) Vestibu- mates. Science 216:755–757 lar neurones in the parieto-insular cortex of 247. Goldman-Rakic PS, Selemon LD, Schwartz ML monkeys (Macaca fascicularis): visual and neck (1984) Dual pathways connecting the dorsolateral receptor responses. J Physiol 430:559–583 prefrontal cortex with the hippocampal formation 264. Gu Q (2002) Neuromodulatory transmitter sys- and parahippocampal cortex in the rhesus mon- tems in the cortex and their role in cortical key. Neuroscience 12:719–743 plasticity. Neuroscience 111:815–835 248. Goltz F (1881) Über die Verrichtungen des Gross- 265. Guldin WO, Grüsser OJ (1998) Is there a vestibu- hirns. Archiv für die gesamte Neurologie und lar cortex? Trends Neurosci 21:254–259 Psychiatrie 26:1–49 266. Gulyás B (1997) Functional organization of hu- 249. Goltz F (1888) Über die Verrichtungen des Gross- man visual cortical areas. In: Jones EG, Peters A hirns. Pflügers Archiv für die gesamte Physiolo- (eds) Extrastriate cortex in primates. Plenum, gie 42:419–467. In: Von Bonin G (1960) (ed) New York, pp 743–775 (Cerebral cortex, vol 12) Some papers on the cerebral cortex. Translated 267. Gulyás B, Roland PE (1994) Processing and ana- as, “On the functions of the hemispheres”. Tho- lysis of form, colour and binocular disparity in mas, Springfield, pp 118–158 the human brain: functional anatomy by positron 250. Gonchar Y, Turney S, Price JL, Burkhalter A emission tomography. Eur J Neurosci 6:1811– (2002) Axo-axonic synapses formed by somatos- 1828 tatin-expressing GABAergic neurons in rat and 268. Gusnard DA, Ollinger JM, Schulman GL et al monkey visual cortex. J Comp Neurol 443:1–14 (2003) Persistence and brain circuitry. Proc Natl 251. Gottfried JA, Zald DH (2005) On the scent of hu- Acad Sci USA 100:3479–3484 man olfactory orbitofrontal cortex: meta-analysis 269. Haber SN, Johnson Gdowski M (2004) The basal and comparison to non-human primates. Brain ganglia. In: Paxinos G, Mai JK (eds) The human Res Rev 50:287–304 nervous system, 2nd edn. Elsevier, Amsterdam, 252. Graziano M (2006) The organization of behavior- pp 676–738 al repertoire in motor cortex. Annu Rev Neurosci 270. Haber SN, Kunishio K, Mizobuchi M, Lynd-Balta 29:105–134 E (1995) The orbital and medial prefrontal circuit 253. Grefkes C, Fink GR (2005) The functional organi- through the primate basal ganglia. J Neurosci zation of the intraparietal sulcus in humans and 15:4851–4867 monkeys. J Anat 207:3–17 271. Habib M, Robichon F, Levrier O, Khalil R, Sala- 254. Gregoriou GG, Borra E, Matelli M, Luppino G mon G (1995) Diverging asymmetries of tem- (2006) Architectonic organization of the inferior poro-parietal cortical areas: a reappraisal of Ge- parietal convexity of the macaque monkey. J schwind/Galaburda theory. Brain Lang 48:238– Comp Neurol 496:422–451 258 255. Gross CG (1994) How inferior temporal cortex 272. Hackett TA, Stepniewska I, Kaas JH (1998) Subdi- became a visual area. Cereb Cortex 4:455–469 visions of auditory cortex and ipsilateral cortical 256. Gross CG (1997) From Imhotep to Hubel and connections of the parabelt auditory cortex in Wiesel; the story of visual cortex. In: Jones EG, macaque monkeys. J Comp Neurol 394:475–495 Peters A (eds) Cerebral cortex, vol 12. Plenum, 273. Hackett TA, Stepniewska I, Kaas JH (1999) Callo- New York, pp 1–58 sal connections of the parabelt auditory cortex in 257. Gross CG (1999) History of localization of func- macaque monkeys. Eur J Neurosci 11:856–866 tion in the cerebral cortex. In: Wilson RA, Keil 274. Hackett TA, Stepniewska I, Kaas JH (1999) Pre- FC (eds) The MIT encyclopedia of the cognitive frontal connections of the parabelt auditory cor- sciences. MIT Press, Cambridge, pp 203–204 tex in macaque monkeys. Brain Res 817:45–58 258. Gross CG, Bender DB, Rocha-Miranda CE (1969) 275. Hadjikhani N, Liu AK, Dale AM, Cavanagh P, Visual receptive fields of neurons in inferotempo- Tootell RB (1998) Retinotopy and color sensitiv- ral cortex of the monkey. Science 166:1303–1306 ity in human visual cortical area V8. Nat Neuros- 259. Grossman M, Koenig P, DeVita C et al (2002) ci 1:235–241 Neural representation of verb meaning: an fMRI 276. Hagoort P (2005) On Broca, brain, and binding: study. Hum Brain Mapp 15:124–134 a new framework. Trends Cogn Sci 9:416–423 662 Section II Structure of Spinal Cord and Brain Parts

277. Hamada I, Sakai M, Kubota K (1981) Morpholo- connectivity. Plenum, New York, pp 403–446 gical differences between fast and slow pyramidal (Cerebral cortex, vol 5) tract neurons in the monkey motor cortex as re- 294. Hersch SM, White EL (1981) Quantification of vealed by intracellular injection of horseradish synapses formed with apical dendrites of Golgi- peroxidase by pressure. Neurosci Lett 22:233–238 impregnated pyramidal cells: variability in thala- 278. Harlow JM (1868) Recovery from the passage of mocortical inputs, but consistency in the ratios an iron bar through the head. Bull Massachusetts of asymmetrical to symmetrical synapses. Neuro- Med Soc 2:3–20 science 6:1043–1051 279. Hatanaka N, Tokuno H, Hamada I et al (2003) 295. Hersch SM, White EL (1981) Thalamocortical sy- Thalamocortical and intracortical connections of napses involving identified neurons in mouse monkey cingulate motor areas. J Comp Neurol primary somatosensory cortex: a terminal degen- 462:121–138 eration and golgi/EM study. J Comp Neurol 280. Hauk O, Johnsrude I, Pulvermuller F (2004) So- 195:253–263 matotopic representation of action words in hu- 296. Hersch SM, White EL (1981) Thalamocortical sy- man motor and premotor cortex. Neuron 41:301– napses with corticothalamic projection neurons 307 in mouse SmI cortex: electron microscopic de- 281. Hayes TL, Lewis DA (1993) Hemispheric differ- monstration of a monosynaptic feedback loop. ences in layer III pyramidal neurons of the ante- Neurosci Lett 24:207–210 rior language area. Arch Neurol 50:501–505 297. Hersch SM, White EL (1982) A quantitative study 282. He SQ, Dum RP, Strick PL (1993) Topographic of the thalamocortical and other synapses in organization of corticospinal projections from layer IV of pyramidal cells projecting from mouse the frontal lobe: motor areas on the lateral sur- SmI cortex to the caudate-putamen nucleus. J face of the hemisphere. J Neurosci 13:952–980 Comp Neurol 211:217–225 283. He SQ, Dum RP, Strick PL (1995) Topographic 298. Herzog AG, Van Hoesen GW (1976) Temporal organization of corticospinal projections from neocortical afferent connections to the amygdala the frontal lobe: motor areas on the medial sur- in the rhesus monkey. Brain Res 15:57–69 face of the hemisphere. J Neurosci 15:3284–3306 299. Hikosaka K, Iwai E, Saito H, Tanaka K (1988) 284. Heimer L (2000) Basal forebrain in the context of Polysensory properties of neurons in the anterior schizophrenia. Brain Res Rev 31:205–235 bank of the caudal superior temporal sulcus of 285. Hendry SHC, Jones EG (1981) Sizes and distribu- the macaque monkey. J Neurophysiol 60:1615– tions of intrinsic neurons incorporating tritiated 1637 GABA in monkey sensory-motor cortex. J Neuro- 300. Hoen M, Pachot-Clouard M, Segebarth C, Domi- sci 1:390–408 ney PF (2006) When Broca experiences the Janus 286. Hendry SHC, Jones EG (1983) The organization syndrome: an ER-fMRI study comparing sentence of pyramidal and non-pyramidal cell dendrites comprehension and cognitive sequence process- in relation to thalamic afferent terminations in ing. Cortex 42:605–623 the monkey somatic sensory cortex. J Neurocytol 301. Hof PR, Mufson EJ, Morrison JH (1995) Human 12:277–298 orbitofrontal cortex: cytoarchitecture and quanti- 287. Hendry SHC, Houser CR, Jones EG, Vaughn JE tative immunohistochemical parcellation. J Comp (1983) Synaptic organization of immunocyto- Neurol 359:48–68 chemically identified GABA neurons in the mon- 302. Holstege G, Mouton LJ, Gerrits NM (2004) Emo- key sensory-motor cortex. J Neurocytol 12:639– tional motor system. In: Paxinos G, Mai JK (eds) 660 The human nervous system, 2nd edn. Elsevier, 288. Henschen SE (1893) On the visual path and cen- Amsterdam, pp 1306–1324 tre. Brain 16:170–180 303. Hornung JP, Garey LJ (1981) The thalamic pro- 289. Herbster AN, Mintun MA, Nebes RD, Becker JT jection to cat visual cortex: ultrastructure of neu- (1997) Regional cerebral blood flow during word rons identified by Golgi impregnation or retro- and nonword reading. Hum Brain Mapp 5:84–92 grade horseradish peroxidase transport. Neu- 290. Herkenham M (1979) The afferent and efferent roscience 6:1053–1068 connections of the ventromedial thalamic nucleus 304. Hornung JP, De Tribolet N, Tork I (1992) Mor- in the rat. J Comp Neurol 183:487–517 phology and distribution of neuropeptide-con- 291. Herkenham M (1980) Laminar organization of taining neurons in human cerebral cortex. thalamic projections to the rat neocortex. Science Neuroscience 51:363–375 207:532–535 305. Horton JC, Adams DL (2005) The cortical col- 292. Herkenham M (1980) The laminar organization umn: a structure without a function. Philos Trans of thalamic projections to neocortex. Trends Neu- R Soc Lond B Biol Sci 360:837–862 rosci 3:17–18 306. Houser CR, Vaughn JE, Hendry SHC, Jones EG, 293. Herkenham M (1986) New perspectives on the Peters A (1984) GABA neurons in the cerebral organization and evolution of nonspecific thala- cortex. In: Jones EG, Peters A (eds) Functional mocortical projections. In: Jones EG, Peters A properties of cortical cells. Plenum, New York, pp (eds) Sensory-motor areas and aspects of cortical 63–89 (Cerebral cortex, vol 2) 15 Telencephalon: Neocortex 663

307. Houser CR, Crawford GD, Salvaterra PM, Vaughn 324. Insausti R, Amaral DG, Cowan WM (1987) The JE (1985) Immunocytochemical localization of entorhinal cortex of the monkey: II. Cortical af- choline acetyltransferase in rat cerebral cortex: a ferents. J Comp Neurol 264:356–395 study of cholinergic neurons and synapses. J 325. Isseroff A, Schwartz ML, Dekker JJ, Goldman- Comp Neurol 234:17–34 Rakic PS (1984) Columnar organization of callo- 308. Hubel DH, Wiesel TN (1962) Receptive fields, bi- sal and associational projections from rat frontal nocular interaction and functional architecture cortex. Brain Res 293:213–223 in the cat’s visual cortex. J Physiol 160:106–154 326. Jackson JH (1931) Selected writings of John 309. Hubel DH, Wiesel TN (1965) Receptive fields and Hughlings Jackson, 2 vols. Hodder & Stoughton, functional architecture in two non-striate visual London areas (18 and 19) of the cat. J Neurophysiol 327. Jacobsen CF (1935) Function of frontal associa- 18:229–289 tion areas in primates. Arch Neurol Psychiatry 310. Hubel DH, Wiesel TN (1968) Receptive fields and 33:549–569 functional architecture of monkey striate cortex. 328. Jacobsen CF, Wolfe JB, Jackson TA (1935) An ex- J Physiol 195:215–243 perimental analysis of the function of the frontal 311. Hubel DH, Wiesel TN (1969) Anatomical demon- association areas in primates. J Nerv Ment Dis stration of columns in the monkey striate cortex. 52:1–14 Nature (London) 221:747–750 329. Janzen R (1972) Die letzten 100 Jahre Lokalisa- 312. Hubel DH, Wiesel TN (1972) Laminar and co- tionsforschung an der Grosshirnrinde. Z Neurol lumnar distribution of geniculo-cortical fibers in 202:75–79 the macaque monkey. J Comp Neurol 146:421– 330. Jasper HH (1995) A historical perspective: the 443 rise and fall of prefrontal lobotomy. In: Jasper 313. Hubel DH, Wiesel TN (1977) Functional architec- HH, Riggio S, Goldman-Rakic PS (eds) Epilepsy ture of macaque monkey visual cortex. Proc R and the functional anatomy of the frontal lobe. Soc Lond Biol 198:1–59 Raven, New York, pp 97–114 314. Hubel DH, Wiesel TN, Stryker MP (1978) Anatom- 331. Javoy-Agid F, Scatton B, Ruberg M et al (1989) ical demonstration of orientation columns in ma- Distribution of monoaminergic, cholinergic, and caque monkey. J Comp Neurol 177:361–380 GABAergic markers in the human cerebral cor- 315. Hübener M, Schwarz C, Bolz J (1990) Morpholo- tex. Neuroscience 29:251–259 gical types of projection neurons in layer 5 of cat 332. Jenny AB (1979) Commissural projections of the visual cortex. J Comp Neurol 301:655–674 cortical hand motor area in monkey. J Comp 316. Huerta MF, Pons TP (1990) Primary motor cor- Neurol 188:137–146 tex receives input from area 3a in macaques. 333. Johnson PB, Ferraina S, Caminiti R (1993) Corti- Brain Res 537:367–371 cal networks for visual reaching. Exp Brain Res 317. Huk AC, Heeger DJ (2002) Pattern-motion re- 97:361–365 sponses in human visual cortex. Nat Neurosci 334. Johnson PB, Ferraina S, Bianchi L, Caminiti R 5:72–75 (1996) Cortical networks for visual reaching: phys- 318. Hutsler J, Galuske RA (2003) Hemispheric asym- iological and anatomical organization of frontal metries in cerebral cortical networks. Trends and parietal lobe arm regions. Cereb Cortex Neurosci 26:429–435 6:102–119 319. Imig TJ, Brugge JF (1978) Sources of termination 335. Jones DK, Simmons A, Williams SC, Horsfield of callosal axons related to binaural and fre- MA (1999) Non-invasive assessment of axonal fi- quency maps in primary auditory cortex of the ber connectivity in the human brain via diffusion cat. J Comp Neurol 183:637–660 tensor MRI. Magn Reson Med 42:37–41 320. Indefrey P, Levelt WJM (2000) The neural corre- 336. Jones EG (1975) Varieties and distribution of lates of language production. In: Gazzaniga MS non-pyramidal cells in the somatic sensory cor- (ed) The new cognitive neurosciences. MIT Press, tex of the squirrel monkey. J Comp Neurol Cambridge, pp 845–865 160:205–267 321. Indefrey P, Levelt WJM (2004) The spatial and 337. Jones EG (1984) Laminar distribution of cortical temporal signatures of word production compo- efferent cells. In: Peters A, Jones EG (eds) Cellu- nents. Cognition 92:101–144 lar components of the cerebral cortex. Plenum, 322. Innocenti GM (1986) General organization of cal- New York, pp 521–553 (Cerebral cortex, vol 1) losal connections in the cerebral cortex. In: Jones 338. Jones EG (1984) Neurogliaform or spiderweb EG, Peters A (eds) Sensory-motor areas and as- cells. In: Peters A, Jones EG (eds) Cellular com- pects of cortical connectivity. Plenum, New York, ponents of the cerebral cortex. Plenum, New pp 291–354 (Cerebral cortex, vol 5) York, pp 409–418 (Cerebral cortex, vol 1) 323. Insausti R, Muñoz M (2001) Cortical projections 339. Jones EG (1986) Connectivity of the primate sen- of the non-entorhinal hippocampal formation in sory-motor cortex. In: Jones EG, Peters A (eds) the cynomolgus monkey (Macaca fascicularis). Sensory-motor areas and aspects of cortical con- Eur J Neurosci 14:435–451 nectivity. Plenum, New York, pp 113–184 (Cere- bral cortex, vol 5) 664 Section II Structure of Spinal Cord and Brain Parts

340. Jones EG (1987) Cerebral cortex. In: Adelman G ventrolateral complex of the cat thalamus. Neuro- (ed) Encyclopedia of neuroscience, vol 1. Bir- sci Lett 117:280–284 khäuser, Boston, pp 209–211 357. Kalivas PW, Volkow N, Seamans J (2005) Unman- 341. Jones EG (2000) Microcolumns in the cerebral ageable motivation in addiction: a pathology in cortex. Proc Natl Acad Sci USA 97:5019–5021 prefrontal-accumbens glutamate transmission. 342. Jones EG, Burton H (1976) Areal differences in Neuron 45:647–650 the laminar distribution of thalamic afferents in 358. Kanai T (1938) Über die Furchen und Windun- cortical fields of the insular, parietal and tempor- gen der Orbitalfläche des Stirnhirns bei Japa- al regions of primates. J Comp Neurol 168:197– nern. Okajima’s Fol anat japon 18:229–306 248 359. Kapur N, Ellison D, Smith MP, McLellan DL, Bur- 343. Jones EG, Hendry SHC (1984) Basket cells. In: rows EH (1992) Focal retrograde amnesia follow- Peters A, Jones EG (eds) Cellular components of ing bilateral temporal lobe pathology. A neurop- the cerebral cortex. Plenum, New York, pp 309– sychological and magnetic resonance study. Brain 336 (Cerebral cortex, vol 1) 115:73–85 344. Jones EG, Leavitt RY (1974) Retrograde axonal 360. Kapur N, Ellison D, Parkin AJ et al (1994) Bilat- transport and the demonstration of non-specific eral temporal lobe pathology with sparing of projections to the cerebral cortex and striatum medial temporal lobe structures: lesion profile from thalamic intralaminar nuclei in the rat, cat and pattern of memory disorder. Neuropsycholo- and monkey. J Comp Neurol 154:349–378 gia 32:23–38 345. Jones EG, Powell TP (1969) Synapses on the axon 361. Karol EA, Pandya DN (1971) The distribution of hillocks and initial segments of pyramidal cell the corpus callosum in the Rhesus monkey. Brain axons in the cerebral cortex. J Cell Sci 5:495–507 94:471–486 346. Jones EG, Powell TPS (1970) An anatomical study 362. Katsumaru H, Kosaka T, Heizmann CW, Hama K of converging sensory pathways within the cere- (1988) Gap junctions on GABAergic neurons bral cortex of the monkey. Brain 93:793–824 containing the calcium-binding protein parvalbu- 347. Jones EG, Burton H, Porter R (1975) Commissur- min in the rat hippocampus (CA1 region). Exp al and cortico-cortical “columns” in the somatic Brain Res 72:363–370 sensory cortex of primates. Science 190:572–574 363. Katz LC (1987) Local circuitry of identified pro- 348. Jones EG, Coulter JD, Hendry SHC (1978) Intra- jection neurons in cat visual cortex brain slices. J cortical connectivity of architectonic fields in the Neurosci 7:1223–1249 somatic sensory, motor and parietal cortex of 364. Kaufman EFS, Rosenquist AC (1985) Efferent monkeys. J Comp Neurol 181:291–348 projections of the thalamic intralaminar nuclei in 349. Jones EG, Coulter JD, Wise SP (1979) Commissural the cat. Brain Res 335:257–279 columns in the sensory motor cortex of monkeys. J 365. Keller A, Asanuma H (1993) Synaptic relation- Comp Neurol 188:113–136 ships involving local axon collaterals of pyrami- 350. Jones EG, Hendry SHC, DeFelipe J (1987) GABA- dal neurons in the cat motor cortex. J Comp peptide neurons of the primate cerebral cortex: a Neurol 336:229–242 limited cell class. In: Jones EG, Peters A (eds) 366. Kertesz A, Lesk D, McCabe P (1977) Isotope Further aspects of cortical function. Raven, New localization of infarcts in aphasia. Arch Neurol York, pp 237–266 (Cerebral cortex, vol 6) 34:590–601 351. Jones EG, DeFelipe J, Hendry SHC, Maggio JE 367. King AB, Menon RS, Hachinski V, Cechetto DF (1988) A study of tachykinin-immunoreactive (1999) Human forebrain activation by visceral neurons in monkey cerebral cortex. J Neurosci stimuli. J Comp Neurol 413:572–582 8:1206–1224 368. Kisvárday ZF, Eysel UT (1992) Cellular organiza- 352. Kaas JH (1997) Theories of visual cortex organi- tion of reciprocal patchy networks in layer III of zation in primates. In: Jones EG, Peters A (eds) cat visual cortex (area 17). Neuroscience 46:275– Extrastriate cortex in primates. Plenum, New 286 York, pp 91–125 (Cerebral cortex, vol 12) 369. Kisvárday ZF, Martin KA, Whitteridge D, Somo- 353. Kaas JH (2002) Neocortex. In: Ramachandran VS gyi P (1985) Synaptic connections of intracellu- (ed) Encyclopedia of the human brain, vol 3. larly filled clutch cells: a type of small basket cell Academic Press, Amsterdam, pp 291–303 in the visual cortex of the cat. J Comp Neurol 354. Kaas JH (2004) Somatosensory system. In: Paxi- 241:111–137 nos G, Mai JK (eds) The human nervous system, 370. Kisvárday ZF, Martin KA, Freund TF et al (1986) 2nd edn. Elsevier, Amsterdam, pp 1059–1092 Synaptic targets of HRP-filled layer III pyramidal 355. Kadohisa M, Rolls ET, Verhagen JV (2004) Orbi- cells in the cat striate cortex. Exp Brain Res tofrontal cortex: neuronal representation of oral 64:541–552 temperature and capsaicin in addition to taste 371. Kisvárday ZF, Gulyas A, Beroukas D et al (1990) and texture. Neuroscience 127:207–221 Synapses, axonal and dendritic patterns of 356. Kakei S, Shinoda Y (1990) Parietal projection of GABA-immunoreactive neurons in human cere- thalamocortical fibers from the ventroanterior- bral cortex. Brain 113:793–812 15 Telencephalon: Neocortex 665

372. Kisvárday ZF, Beaulieu C, Eysel UT (1993) Net- monkeys: contributions to the cortical pheno- work of GABAergic large basket cells in cat visual type. J Comp Neurol 471:97–111 cortex (area 18): implication for lateral disinhibi- 389. Künzle H (1978) Cortico-cortical efferents of pri- tion. J Comp Neurol 327:398–415 mary motor and somatosensory regions of the 373. Kleist K (1934) Gehirnpathologie. Johann Am- cerebral cortex in Macaca fascicularis. Neurosci- brosius Barth, Leipzig ence 3:25–39 374. Kleist K (1959) Die Lokalisation im Grosshirn 390. Kuypers HGJM (1987) Pyramidal tract. In: Adel- und ihre Entwicklung. Psychiatr Neurol 137:289– man G (ed) Encyclopedia of neuroscience, vol 2. 309 Birkhäuser, Boston, pp 1018–1020 375. Knutson B, Adams CM, Fong GW, Hommer D 391. Lalouschek W (1997) Brain and language: beyond (2001) Anticipation of increasing monetary re- the left cerebral hemisphere. Contributions of ward selectively recruits nucleus accumbens. J subcortical structures, cerebellum, and right Neurosci 21:RC159(1–5) hemisphere to language. Evolution and Cognition 376. Kobayashi Y, Amaral DG (2000) Macaque mon- 3:133–147 key retrosplenial cortex: I. three-dimensional and 392. Lashley KS (1929) Brain mechanisms and intelli- cytoarchitectonic organization. J Comp Neurol gence. University of Chicago Press, Chicago 426:339–365 393. Lashley KS (1933) Integrative functions of the 377. Kobayashi Y, Amaral DG (2003) Macaque mon- cerebral cortex. Physiol Rev 13:1–42 key retrosplenial cortex: II. Cortical afferents. J 394. Lashley KS, Clark G (1946) The cytoarchitecture Comp Neurol 466:48–79 of the cerebral cortex of Ateles: A critical exami- 378. Koechlin E, Basso G, Pietrini P, Panzer S, Graf- nation of architectonic studies. J Comp Neurol man J (1999) The role of the anterior prefrontal 85:223–305 cortex in human cognition. Nature 399:148–151 395. Lassek AM (1940) The human pyramidal tract. II. 379. Koechlin E, Ody C, Kouneiher F (2003) The ar- A numerical investigation of the Betz cells of the chitecture of cognitive control in the human premotor area. Arch Neurol Psychiatry 44:718–724 frontal cortex. Science 302:1181–1185 396. Laurienti PJ, Wallace MT, Maldjian JA et al 380. Köhler C, Shipley MT, Srebro B, Harkmark W (2003) Cross-modal sensory processing in the (1978) Some retrohippocampal afferents to the anterior cingulate and medial prefrontal cortices. entorhinal cortex. Cells of origin as studied by Hum Brain Mapp 19:213–223 the HRP method in the rat and mouse. Neurosci 397. Lavenex P, Suzuki WA, Amaral DG (2002) Peri- Lett 10:115–120 rhinal and parahippocampal cortices of the ma- 381. Kondo H, Saleem KS, Price JL (2003) Differential caque monkey: projections to the neocortex. J connections of the temporal pole with the orbital Comp Neurol 447:394–420 and medial prefrontal networks in macaque mon- 398. Lee Seldon H (1985) The anatomy of speech per- keys. J Comp Neurol 465:499–523 ception. Human auditory cortex. In: Peters A, 382. Kondo H, Saleem KS, Price JL (2005) Differential Jones EO (eds) Cerebral cortex, vol 4. Association connections of the perirhinal and parahippocam- and auditory cortices. Plenum, New York, pp pal cortex with the orbital and medial prefrontal 273–327 networks in macaque monkeys. J Comp Neurol 399. Lees AJ, Smith E (1983) Cognitive deficits in the 493:479–509 early stages of Parkinson’s disease. Brain 383. Kononova EP (1935) [The variability of structure 106:257–270 of the cortex of the brain: inferior frontal gyrus 400. Leichnetz GR (2001) Connections of the medial of adult man] in Russian. Brain Res Inst Publ posterior parietal cortex (area 7m) in the mon- 1:49–118 key. Anat Rec 263:215–236 384. Konopaske GT, Sweet RA, Wu Q, Sampson A, Le- 401. Leonard CM (1969) The prefrontal cortex of the wis DA (2006) Regional specificity of chandelier rat. I. Cortical projections of the mediodorsal nu- neuron axon terminal alterations in schizophre- cleus. II. Efferent connections. Brain Res 12:321– nia. Neuroscience 138:189–196 343 385. Kreisler A, Godefroy O, Delmaire C et al (2000) 402. Leonard CM (1972) The connections of the dor- The anatomy of aphasia revisited. Neurology 54: somedial nuclei. Brain Behav Evol 6:524–541 1117–1123 403. LeVay S (1973) Synaptic patterns in the visual 386. Kringelbach ML (2005) The human orbitofrontal cortex of the cat and monkey. Electron micro- cortex: linking reward to hedonic experience. Nat scopy of Golgi preparations. J Comp Neurol Rev Neurosci 6:691–702 150:53–85 387. Krmpotic-Nemanic J, Kostovic I, Nemanic D 404. LeVay S, Hubel DH, Wiesel TN (1975) The pat- (1984) Prenatal and perinatal development of ra- tern of ocular dominance columns in macaque dial cell columns in the human auditory cortex. visual cortex revealed by a reduced silver stain. J Acta Otolaryngol 97:489–495 Comp Neurol 159:559–576 388. Krubitzer L, Huffman KJ, Disbrow E, Recanzone 405. Lewis DA, Gonzalez-Burgos G (2006) Pathophy- G (2004) Organization of area 3a in macaque siologically based treatment interventions in schizophrenia. Nat Med 12:1016–1022 666 Section II Structure of Spinal Cord and Brain Parts

406. Lewis DA, Lund JS (1990) Heterogeneity of chan- 424. Lund JS (1987) Local circuit neurons of macaque delier neurons in monkey neocortex: corticotro- monkey striate cortex: I. Neurons of laminae 4C pin-releasing factor- and parvalbumin-immuno- and 5A. J Comp Neurol 257:60–92 reactive populations. J Comp Neurol 293:599–615 425. Lund JS (1988) Anatomical organization of maca- 407. Lewis DA, Hashimoto T, Volk DW (2005) Cortical que monkey striate visual cortex. Annu Rev Neu- inhibitory neurons and schizophrenia. Nat Rev rosci 11:253–288 Neurosci 6:312–324 426. Lund JS, Boothe RG (1975) Interlaminar connec- 408. Lewis JW, Van Essen DC (2000) Mapping of ar- tions and pyramidal neuron organisation in the chitectonic subdivisions in the macaque monkey, visual cortex. J Comp Neurol 159:305–334 with emphasis on parieto-occipital cortex. J 427. Lund JS, Lewis DA (1993) Local circuit neurons Comp Neurol 428:79–111 of developing and mature macaque prefrontal 409. Li H, Fukuda M, Tanifuji M, Rockland KS (2003) cortex: Golgi and immunocytochemical charac- Intrinsic collaterals of layer 6 Meynert cells and teristics. J Comp Neurol 328:282–312 functional columns in primate V1. Neuroscience 428. Lund JS, Yoshioka T (1991) Local circuit neurons 120:1061–1069 of macaque monkey striate cortex: III. Neurons 410. Liepmann H (1900) Das Krankheitsbild der of laminae 4B, 4A, and 3B. J Comp Neurol Apraxia (motorischen Asymbolie) auf Grund 311:234–258 eines Falles von einseitiger Apraxie. Monatsschr 429. Lund JS, Henry GH, MacQueen CL, Harvey AR Psychiatr Neurol 8:15–44, 102–132, 182–197 (1979) Anatomical organization of the primary 411. Liepmann H, Maas O (1907) Fall von linksseitiger visual cortex (area 17) of the cat. A comparison Agraphie und Apraxie bei rechtsseitiger Läh- with area 17 of the macaque monkey. J Comp mung. Z Psychiatr Neurol 10:214–227 Neurol 184:599–618 412. Livingstone MS, Hubel DH (1984) Anatomy and 430. Lund JS, Hawken MJ, Parker AJ (1988) Local cir- physiology of a color system in the primate vi- cuit neurons of macaque monkey striate cortex: sual cortex. J Neurosci 4:309–356 II. Neurons of laminae 5B and 6. J Comp Neurol 413. Loose R, Kaufmann C, Auer DP, Lange KW 276:1–29 (2003) Human prefrontal and sensory cortical ac- 431. Luppino G, Matelli M, Camarda R, Rizzolatti G tivity during divided attention tasks. Hum Brain (1993) Corticocortical connections of area F3 Mapp 18:249–259 (SMA-proper) and area F6 (pre-SMA) in the ma- 414. Lorente de Nó R (1922) La corteza cereral del ra- caque monkey. J Comp Neurol 338:114–140 tón. Trabajos Cajal Madrid 20:41–80 432. Luppino G, Calzavara R, Rozzi S, Matelli M 415. Lorente de Nó R (1934) Studies on the structure (2001) Projections from the superior temporal of the cerebral cortex. I. The area entorhinalis. J sulcus to the agranular frontal cortex in the ma- Psychol Neurol 45:381–439 caque. Eur J Neurosci 14:1035–1040 416. Lorente de Nó R (1938) Architectonics and struc- 433. Luppino G, Rozzi S, Calzavara R, Matelli M ture of the cerebral cortex. In: Fulton JF (ed) (2003) Prefrontal and agranular cingulate projec- Physiology of the nervous system. Oxford Univer- tions to the dorsal premotor areas F2 and F7 in sity Press, London, pp 291–330 the macaque monkey. Eur J Neurosci 17:559–578 417. Loring DW, Meador KJ, Lee GP et al (1990) Cere- 434. Lynch JC, Tian JR (2005) Cortico-cortical net- bral language lateralization: evidence from intra- works and cortico-subcortical loops for the high- carotid amobarbital testing. Neuropsychologia er control of eye movements. Prog Brain Res 151: 28:831–838 461–501 418. Lou HC, Kjaer TW, Friberg L et al (1999) A 15O- 435. Maddock RJ, Garrett AS, Buonocore MH (2003) H2O PET study of meditation and the resting Posterior cingulate cortex activation by emotional state of normal consciousness. Hum Brain Mapp words: fMRI evidence from a valence decision 7:98–105 task. Hum Brain Mapp 18:30–41 419. Lu MT, Preston JB, Strick PL (1994) Interconnec- 436. Marin-Padilla M (1969) Origin of the pericellular tions between the prefrontal cortex and the pre- baskets of the pyramidal cells of the human mo- motor areas in the frontal lobe. J Comp Neurol tor cortex: a Golgi study. Brain Res 14:633–646 341:375–392 437. Marin-Padilla M (1978) Dual origin of the mam- 420. Luders H, Lesser RP, Hahn J et al (1991) Basal malian neocortex and evolution of the cortical temporal language area. Brain 114:743–754 plate. Anat Embryol (Berl) 152:109–126 421. Ludwig E, Klingler J (1956) Atlas cerebri humani. 438. Marin-Padilla M (1984) Neurons of layer I. A devel- Karger, Basel opment analysis. In: Peters A, Jones EG (eds) Cel- 422. Lund JS (1973) Organization of neurons in the lular components of the cerebral cortex. Plenum, visual cortex, area 17, of the monkey (Macaca New York, pp 447–478 (Cerebral cortex, vol 1) mulatta). J Comp Neurol 147:455–496 439. Markowitsch HJ, Emmans D, Irle E, Streicher M, 423. Lund JS (1984) Spiny stellate neurons. In: Peters Preilowski B (1985) Cortical and subcortical af- A, Jones EG (eds) Cellular components of the ferent connections of the primate’s temporal pole: cerebral cortex. Plenum, New York, pp 255–308 a study of rhesus monkeys, squirrel monkeys and (Cerebral cortex, vol 1) marmosets. J Comp Neurol 242:425–458 15 Telencephalon: Neocortex 667

440. Markram H, Toledo-Rodriguez M, Wang Y et al 455. McCarthy G, Puce A, Constable RT et al (1996) (2004) Interneurons of the neocortical inhibitory Activation of human prefrontal cortex during system. Nat Rev Neurosci 5:793–807 spatial and nonspatial working memory tasks 441. Martin-Elkins CL, Horel JA (1992) Cortical affer- measured by functional MRI. Cereb Cortex ents to behaviorally defined regions of the infer- 6:600–611 ior temporal and parahippocampal gyri as de- 456. McGuire BA, Hornung JP, Gilbert CD, Wiesel TN monstrated by WGA-HRP. J Comp Neurol (1984) Patterns of synaptic input to layer 4 of cat 321:177–192 striate cortex. J Neurosci 4:3021–3033 442. Martinotti C (1890) Beitrag zum Studium der 457. McGuire PK, Bates JF, Goldman-Rakic PS (1991) Hirnrinde und dem Centralursprung der Nerven. Interhemispheric integration: I. Symmetry and Int Monatsschr Anat Physiol 7:69–90 convergence of the corticocortical connections of 443. Matelli M, Luppino G, Rizzolatti G (1991) Archi- the left and the right principal sulcus (PS) and tecture of superior and mesial area 6 and the ad- the left and the right supplementary motor area jacent cingulate cortex in the macaque monkey. J (SMA) in the rhesus monkey. Cereb Cortex Comp Neurol 311:445–462 1:390–407 444. Matelli M, Govoni P, Galletti C, Kutz DF, Luppino 458. Meinecke DL, Peters A (1987) GABA immunor- G (1998) Superior area 6 afferents from the supe- eactive neurons in rat visual cortex. J Comp Neu- rior parietal lobule in the macaque monkey. J rol 261:388–404 Comp Neurol 402:327–352 459. Melchitzky DS, Sesack SR, Pucak ML, Lewis DA 445. Matelli M, Luppino G, Geyer S, Zilles K (2004) (1998) Synaptic targets of pyramidal neurons pro- Motor Cortex. In: Paxinos G, Mai JK (eds) The viding intrinsic horizontal connections in monkey human nervous system, 2nd edn. Elsevier, Am- prefrontal cortex. J Comp Neurol 390:211–224 sterdam, pp 973–996 460. Menon V, Adleman NE, White CD, Glover GH, 446. Mates SL, Lund JS (1983) Neuronal composition Reiss AL (2001) Error-related brain activation and development in lamina 4C of monkey striate during a Go/NoGo response inhibition task. Hum cortex. J Comp Neurol 221:60–90 Brain Mapp 12:131–143 447. Matsuzaka Y, Aizawa H, Tanji J (1992) A motor 461. Menon V, Mackenzie K, Rivera SM, Reiss AL area rostral to the supplementary motor area (2002) Prefrontal cortex involvement in processing (presupplementary motor area) in the monkey: incorrect arithmetic equations: evidence from neuronal activity during a learned motor task. J event-related fMRI. Hum Brain Mapp 16:119–130 Neurophysiol 68:653–662 462. Mesulam MM (1983) The functional anatomy 448. Mayberg HS, Liotti M, Brannan SK et al (1999) and hemispheric specialization for directed atten- Reciprocal limbic-cortical function and negative tion. The role of the parietal lobe and its connec- mood: converging PET findings in depression tivity. Trends Neurosci 6:384–387 and normal sadness. Am J Psychiatry 156:675– 463. Mesulam MM (1985) Principles of behavioral 682 neurology. Davis, Philadelphia 449. Mayberg HS, Brannan SK, Tekell JL et al (2000) 464. Mesulam MM (1998) From sensation to cogni- Regional metabolic effects of fluoxetine in major tion. Brain 121:1013–1052 depression: serial changes and relationship to 465. Mesulam MM (2002) Brain anatomy and net- clinical response. Biol Psychiatry 48:830–843 works. In: Ramachandran VS (ed) Encyclopedia 450. Mayer AR, Dorflinger JM, Rao SM, Seidenberg M of the human brain, vol 1. Academic Press, Am- (2004) Neural networks underlying endogenous sterdam, pp 469–480 and exogenous visual-spatial orienting. Neuro- 466. Mesulam MM, Mufson EJ (1985) The insula of image 23:534–541 Reil in man and monkey. In: Peters A, Jones EG 451. Mazoyer BM, Tzourio N, Frak V et al (1993) The (eds) Association and auditory cortices. Plenum, cortical representation of speech. J Cogn Neuros- New York, pp 179–226 (Cerebral cortex, vol 4) ci 5:467–479 467. Mesulam MM, Hersh LB, Mash DC, Geula C 452. Mazziotta J, Toga A, Evans A et al (2000) A prob- (1992) Differential cholinergic innervation within abilistic approach for mapping the human brain: functional subdivisions of the human cerebral the International Consortium for Brain Mapping cortex: a choline acetyltransferase study. J Comp (ICBM). In: Toga AW, Mazziotta J (eds) Brain Neurol 318:316–328 mapping: The systems. Academic Press, San Die- 468. Meyer A, Beck E (1945) Neuropathological prob- go, pp 141–156 lems arising from prefrontal leucotomy. J Ment 453. Mazziotta J, Toga A, Evans A et al (2001) A prob- Sci 91:411–422 abilistic atlas and reference system for the human 469. Meyer A, Beck E (1954) Prefrontal leucotomy and brain: International Consortium for Brain Map- related operations: anatomical aspects of success ping (ICBM). Philos Trans R Soc Lond B Biol Sci and failure. Oliver and Boyd, Edinburgh/London 356:1293–1322 470. Meyer S, Strittmatter M, Fischer C, Georg T, 454. Mazziotta J, Toga A, Evans A et al (2001) A four- Schmitz B (2004) Lateralization in autonomic dimensional probabilistic atlas of the human dysfunction in ischemic stroke involving the in- brain. J Am Med Inform Assoc 8:401–430 sular cortex. Neuroreport 15:357–361 668 Section II Structure of Spinal Cord and Brain Parts

471. Mills CK (1891) On the localization of the audi- the olfactory bulb bilaterally. Exp NeuroI 64:295– tory centre. Brain 14:465–472 305 472. Mistlin AJ, Perrett DI (1990) Visual and somato- 488. Mori S, Crain BJ, Chacko VP, van Zijl PC (1999) sensory processing in the macaque temporal cor- Three-dimensional tracking of axonal projections tex: the role of ‘expectation’. Exp Brain Res 82: in the brain by magnetic resonance imaging. 437–450 Ann Neurol 45:265–269 473. Mitchell RLC, Crow TJ (2005) Right hemisphere 489. Morosan P, Rademacher J, Schleicher A et al language functions and schizophrenia: the forgot- (2001) Human primary auditory cortex: cytoar- ten hemisphere? Brain 128:963–978 chitectonic subdivisions and mapping into a spa- 474. Mitrofanis J, deFonseka R (2001) Organisation of tial reference system. Neuroimage 13:684–701 connections between the zona incerta and the in- 490. Morosan P, Schleicher A, Amunts K, Zilles K terposed nucleus. Anat Embryol 204:153–159 (2005) Multimodal architectonic mapping of hu- 475. Miyachi S, Lu X, Inoue S et al (2005) Organiza- man superior temporal gyrus. Anat Embryol 210: tion of multisynaptic inputs from prefrontal cor- 401–406 tex to primary motor cortex as revealed by retro- 491. Morrison JH, Magistretti PJ (1983) Monoamines grade transneuronal transport of rabies virus. J and peptides in cerebral cortex. Trends Neurosci Neurosci 25:2547–2556 6:146–151 476. Moniz E (1936) Essai d’un traitement chirurgical 492. Morrison JH, Foote SL, O’Connor D, Bloom FE de certaine psychoses. Bull Acad Med 115:385– (1982) Laminar, tangential and regional organiza- 392 tion of the noradrenergic innervation of monkey 477. Moniz E (1937) Prefrontal leucotomy in the treat- cortex: dopamine-beta-hydroxylase immunohisto- ment of mental disorders. Am J Psychiatry chemistry. Brain Res Bull 9:309–319 93:1379–1385 493. Mott FW (1894) The sensory functions of the 478. Moore JK, Linthicum FH (2004) Auditory system. central convolutions of the cerebral cortex. J In: Paxinos G, Mai JK (eds) The human nervous Physiol 15:464–487 system, 2nd edn. Elsevier, Amsterdam, pp 1241– 494. Mountcastle VB (1957) Modality and topographic 1279 properties of single neurons of cat’s somatic sen- 479. Moran MA, Mufson EJ, Mesulam MM (1987) sory cortex. J Neurophysiol 20:408–434 Neural inputs into the temporopolar cortex of the 495. Mountcastle VB (1979) An organizing principle rhesus monkey. J Comp Neurol 256:88–103 for cerebral function: the unit module and dis- 480. Morecraft RJ, Van Hoesen GW (1992) Cingulate tributed system. In: Schmitt FO, Worden FG input to the primary and supplementary motor (eds) The neurosciences fourth study program. cortices in the rhesus monkey: evidence for so- MIT Press, Cambridge, pp 21–42 matotopy in areas 24c and 23c. J Comp Neurol 496. Mountcastle VB (1997) The columnar organiza- 322:471–489 tion of the neocortex. Brain 120:701–722 481. Morecraft RJ, Van Hoesen GW (1993) Frontal 497. Mrzljak L, Uylings HBM, Kostovic I, Van Eden granular cortex input to the cingulate (M3), sup- CG (1988) Prenatal development of neurons in plementary (M2) and primary (M1) motor cor- the human prefrontal cortex: I. A qualitative Gol- tices in the rhesus monkey. J Comp Neurol gi study. J Comp Neurol 271:355–386 337:669–689 498. Mrzljak L, Uylings HBM, Van Eden CG, Judás M 482. Morecraft RJ, Van Hoesen GW (1998) Conver- (1990) Neuronal development in human prefron- gence of limbic input to the cingulate motor cor- tal cortex in prenatal and postnatal stages. Prog tex in the rhesus monkey. Brain Res Bull 45:209– Brain Res 85:185–222 232 499. Munk H (1881) Über die Funktionen der Gross- 483. Morecraft RJ, Yeterian EH (2002) Prefrontal cor- hirnrinde. 3te Mitteilung, pp 28–53. A. Hirsch- tex. In: Ramachandran VS (ed) Encyclopedia of wald, Berlin. In: Von Bonin G (1960) (ed) Some the human brain, vol 4. Academic Press, Amster- papers on the cerebral cortex. Translated as “On dam, pp 11–26 the functions of the cortex”. Charles C Thomas, 484. Morecraft RJ, Geula C, Mesulam MM (1992) Springfield, IL, pp 97–117 Cytoarchitecture and neural afferents of orbito- 500. Munk H (1892) Über die Fühlsphären des Gross- frontal cortex in the brain of the monkey. J Comp hirns. Sitzungsberichte der Kaiserlichen Akade- Neurol 323:341–358 mie der Wissenschaft 101:679–723 485. Morel A, Kaas JH (1992) Subdivisions and con- 501. Nakamura K, Mikami A, Kubota K (1991) Unique nections of auditory cortex in owl monkeys. J oscillatory activity related to visual processing in Comp Neurol 318:27–63 the temporal pole of monkeys. Neurosci Res 486. Morel A, Garraghty PE, Kaas JH (1993) Tonoto- 12:293–299 pic organization, architectonic fields, and con- 502. Nakamura K, Mikami A, Kubota K (1992) Oscil- nections of auditory cortex in macaque monkeys. latory neuronal activity related to visual short- J Comp Neurol 335:437–459 term memory in monkey temporal pole. Neu- 487. Mori K, Satou M, Takagi SF (1979) Axonal pro- roreport 3:117–120 jection of anterior olfactory nuclear neurons to 15 Telencephalon: Neocortex 669

503. Neal JW, Pearson RCA, Powell TPS (1986) The 521. Oppenheimer SM, Gelb A, Girvin JP, Hachinski organization of the corticocortical projection of VC (1992) Cardiovascular effects of human insu- area 5 upon area 7 in the parietal lobe of the lar cortex stimulation. Neurology 42:1727–1732 monkey. Brain Res 381:164–167 522. Opris I, Bruce CJ (2005) Neural circuitry of judg- 504. Neal JW, Pearson RCA, Powell TPS (1990) The ment and decision mechanisms. Brain Res Rev ipsilateral cortico-cortical connections of area 7b, 48:509–526 PF, in the parietal and temporal lobes of the 523. Osaka N, Osaka M, Kondo H et al (2004) The monkey. Brain Res 524:119–132 neural basis of executive function in working 505. Neal JW, Pearson RCA, Powell TPS (1990) The memory: an fMRI study based on individual dif- ipsilateral corticocortical connections of area 7 ferences. Neuroimage 21:623–631 with the frontal lobe in the monkey. Brain Res 524. Packard MG, Knowlton BJ (2002) Learning and 509:31–40 memory functions of the Basal Ganglia. Annu 506. Nebel K, Stude P, Wiese H et al (2005) Sparse Rev Neurosci 25:563–593 imaging and continuous event-related fMRI in 525. Pakkenberg B, Gundersen HJG (1997) Neocortical the visual domain: a systematic 2190N compari- neuron number in humans: effect of sex and age. son. Hum Brain Mapp 24:130–143 J Comp Neurol 384:312–320 507. Nelson S (2002) Cortical microcircuits: diverse or 526. Pandya DN (1987) Association cortex. In: Adel- canonical? Neuron 36:19–27 man G (ed) Encyclopedia of neuroscience, vol I. 508. Nichols MJ, Newsome WT (1999) The neurobiol- Birkhäuser, Boston, pp 80–83 ogy of cognition. Nature 402:C35–C38 527. Pandya DN, Kuypers HGJM (1969) Cortico-corti- 509. Nieuwenhuys R (1994) The neocortex. An over- cal connections in the rhesus monkey. Brain Res view of its evolutionary development, structural 13:13–36 organization and synaptology. Anat Embryol 528. Pandya DN, Sanides F (1973) Architectonic par- (Berl) 190:307–337 cellation of the temporal operculum in rhesus 510. Nieuwenhuys R (1996) The greater limbic system, monkey, and its projection pattern. Z Anat Ent- the emotional motor system and the brain. Prog wicklungsgesch 139:127–161 Brain Res 107:551–580 529. Pandya DN, Seltzer B (1982) Association areas of 511. Nieuwenhuys R (1998) Telencephalon. In: Nieu- the cerebral cortex. Trends Neurosci 5:386–390 wenhuys R, Ten Donkelaar HJ, Nicholson C (eds) 530. Pandya DN, Seltzer B (1982) Intrinsic connec- The central nervous system of vertebrates, vol tions and architectonics of posterior parietal cor- III. Springer, Berlin Heidelberg New York, pp tex in the rhesus monkey. J Comp Neurol 1871–2008 204:196–210 512. Nobre AC, Allison T, McCarthy G (1994) Word 531. Pandya DN, Seltzer B (1986) The topography of recognition in the human inferior temporal lobe. commissural fibers. In: Lepore F, Ptito M, Jasper Nature 372:260–263 HH (eds) Two hemispheres – one brain. Functions 513. O’Scalaidhe SP, Wilson FA, Goldman-Rakic PS of the corpus callosum. Liss, New York, pp 47–73 (1997) Areal segregation of face-processing neu- 532. Pandya DN, Yeterian EH (1985) Architecture and rons in prefrontal cortex. Science 278:1135–1138 connections of cortical association areas. In: Pe- 514. Ojemann GA (1991) Cortical organization of lan- ters A, Jones EG (eds) Cerebral cortex, vol 4. guage. J Neurosci 11:2281–2287 Association and auditory cortices. Plenum, New 515. Ojima H, Honda CN, Jones EG (1991) Patterns of York, pp 3–61 axon collateralization of identified supragranular 533. Pandya DN, Karol EA, Heilbronn D (1971) The pyramidal neurons in the cat auditory cortex. topographical distribution of interhemispheric Cereb Cortex 1:80–94 projections in the corpus callosum of the rhesus 516. Öngür D, Price JL (2000) The organization of monkey. Brain Res 32:31–43 networks within the orbital and medial prefrontal 534. Papadopoulos GC, Parnavelas JG, Buijs RM cortex of rats, monkeys and humans. Cereb Cor- (1987) Monoaminergic fibers form conventional tex 10:206–219 synapses in the cerebral cortex. Neurosci Lett 517. Öngür D, An X, Price JL (1998) Prefrontal corti- 76:275–279 cal projections to the hypothalamus in macaque 535. Papadopoulos GC, Parnavelas JG, Buijs RM monkeys. J Comp Neurol 401:480–505 (1987) Light and electron microscopic immuno- 518. Öngür D, Ferry AT, Price JL (2003) Architectonic cytochemical analysis of the serotonin innerva- subdivision of the human orbital and medial pre- tion of the rat visual cortex. J Neurocytol 16:883– frontal cortex. J Comp Neurol 460:425–449 892 519. Ono M, Kubik S, Abernathy CD (1990) Atlas of 536. Parker AJ, Newsome WT (1998) Sense and the the cerebral sulci. Thieme, Stuttgart single neuron: probing the physiology of percep- 520. Oppenheimer SM (1994) Neurogenic cardiac ef- tion. Annu Rev Neurosci 21:227–277 fects of cerebrovascular disease. Curr Opin Neu- 537. Parnavelas JG, Edmunds SM (1983) Further evi- rol 7:20–24 dence that Retzius-Cajal cells transform to non- pyramidal neurons in the developing rat visual cortex. J Neurocytol 12:863–871 670 Section II Structure of Spinal Cord and Brain Parts

538. Parnavelas JG, Papadopoulos GC (1989) The 556. Peters A (1985) The visual cortex of the rat. In: monoaminergic innervation of the cerebral cor- Peters A, Jones EG (eds) Visual cortex. Plenum, tex is not diffuse and nonspecific. Trends Neu- New York, pp 19–80 (Cerebral cortex, vol 3) rosci 12:315–319 557. Peters A (1987) Cortical neurons. In: Adelman G 539. Parnavelas JG, Sullivan K, Lieberman AR, Web- (ed) Encyclopedia of neuroscience, vol 1. Bir- ster KE (1977) Neurons and their synaptic orga- khäuser, Boston, pp 282–284 nization in the visual cortex of the rat. Electron 558. Peters A, Fairen A (1978) Smooth and sparsely- microscopy of Golgi preparations. Cell Tissue Res spined stellate cells in the visual cortex of the rat: 183:499–517 a study using a combined Golgi-electron micro- 540. Parnavelas JG, Kelly W, Franke E, Eckenstein F scopic technique. J Comp Neurol 181:129–171 (1986) Cholinergic neurons and fibres in the rat 559. Peters A, Harriman KM (1988) Enigmatic bipolar visual cortex. J Neurocytol 15:329–336 cell of rat visual cortex. J Comp Neurol 267:409– 541. Passingham RE, Stephan KE, Kotter R (2002) The 432 anatomical basis of functional localization in the 560. Peters A, Jones EG (1984) Cellular components of cortex. Nat Rev Neurosci 3:606–616 the cerebral cortex. Plenum, New York 542. Pavlides C, Miyashita E, Asanuma H (1993) Pro- 561. Peters A, Jones EG (1984) Classification of corti- jection from the sensory to the motor cortex is cal neurons. In: Peters A, Jones EG (eds) Cellular important in learning motor skills in the monkey. components of the cerebral cortex. Plenum, New J Neurophysiol 70:733–741 York, pp 107–122 (Cerebral cortex, vol 1) 543. Peinado A, Katz LC (1990) Development of corti- 562. Peters A, Kimerer LM (1981) Bipolar neurons in cal spiny stellate cells: retraction of a transient rat visual cortex: a combined Golgi-electron mi- apical dendrite. Soc Neurosci Abstr 16:1127 croscope study. J Neurocytol 10:921–946 544. Penfield W (1958) The excitable cortex in con- 563. Peters A, Saint Marie PL (1984) Smooth and scious man. Liverpool Press, Liverpool sparsely spinous nonpyramidal cells forming lo- 545. Penfield W, Boldrey E (1937) Somatic motor and cal axonal plexures. In: Peters A, Jones EG (eds) sensory representation in the cerebral cortex of Cellular components of the cerebral cortex. Ple- man as studied by electrical stimulation. Brain num, New York, pp 419–445 (Cerebral cortex, vol 60:389–443 1) 546. Penfield W, Faulk ME Jr (1955) The insula; further 564. Peters A, Sethares C (1991) Layer IVA of rhesus observations on its function. Brain 78:445–470 monkey primary visual cortex. Cereb Cortex 547. Penfield W, Jasper H (1954) Epilepsy and the 1:445–462 functional anatomy of the human brain. Little, 565. Peters A, Sethares C (1991) Organization of pyra- Brown, Boston midal neurons in area 17 of monkey visual cor- 548. Penfield W, Rasmussen AJ (1950) Cerebral cortex tex. J Comp Neurol 306:1–23 in man. A clinical study of localization of func- 566. Peters A, Sethares C (1996) Myelinated axons and tion. MacMillan, New York the pyramidal cell modules in monkey primary 549. Penfield W, Roberts L (1959) Speech and brain visual cortex. J Comp Neurol 365:232–255 mechanisms. Princeton University Press, Prince- 567. Peters A, Sethares C (1997) The organization of ton double bouquet cells in monkey striate cortex. J 550. Percheron G (2004) Thalamus. In: Paxinos G, Neurocytol 26:779–797 Mai JK (eds) The human nervous system, 2nd 568. Peters A, Walsh TM (1972) A study of the organi- edn. Elsevier, Amsterdam, pp 592–675 zation of apical dendrites in the somatic sensory 551. Perrett DI, Harries MH, Bevan R et al (1989) cortex of the rat. J Comp Neurol 144:253–268 Frameworks of analysis for the neural representa- 569. Peters A, Yilmaz E (1993) Neuronal organization tion of animate objects and actions. J Exp Biol in area 17 of cat visual cortex. Cereb Cortex 146:87–113 3:49–68 552. Pessoa L, Kastner S, Ungerleider LG (2002) At- 570. Peters A, Proskauer CC, Ribak CE (1982) Chan- tentional control of the processing of neural and delier cells in rat visual cortex. J Comp Neurol emotional stimuli. Cogn Brain Res 15:31–45 206: 397–416 553. Pessoa L, Kastner S, Ungerleider LG (2003) Neu- 571. Peters A, Meinecke DL, Karamanlidis AN (1987) roimaging studies of attention: from modulation Vasoactive intestinal polypeptide immunoreactive of sensory processing to top-down control. J neurons in the primary visual cortex of the cat. J Neurosci 23:3990–3998 Neurocytol 16:23–38 554. Peters A (1984) Bipolar cells. In: Peters A, Jones 572. Petersen SE, Fox PT, Posner MI, Mintun M, EG (eds) Cellular components of the cerebral cor- Raichle ME (1988) Positron emission tomo- tex. Plenum, New York, pp 381–408 (Cerebral graphic studies of the cortical anatomy of single- cortex, vol 1) word processing. Nature 331:585–589 555. Peters A (1984) Chandelier cells. In: Peters A, 573. Petersen SE, Fox PT, Posner MI, Mintun M, Jones EG (eds) Cellular components of the cere- Raichle ME (1989) Positron emission tomo- bral cortex. Plenum, New York, pp 361–380 (Cer- graphic studies of the cortical anatomy of single- ebral cortex, vol 1) word processing. J Cogn Neurosci 1:153–170 15 Telencephalon: Neocortex 671

574. Petrides M (2000) Mapping prefrontal cortical the tracking of brain white matter fascicles. Neu- systems for the control of cognition. In: Toga roimage 12:184–195 AW, Mazziotta JC (eds) Brain mapping: the sys- 589. Powell TPS (1981) Certain aspects of the intrinsic tems. Academic Press, San Diego, pp 159–176 organization of the cerebral cortex. In: Pompeia- 575. Petrides M, Milner B (1982) Deficits on subject- no O, Ajmone Marsan C (eds) Brain mechanisms ordered tasks after frontal- and temporal-lobe le- and perceptual awareness. Raven, New York, pp sions in man. Neuropsychologia 20:249–262 1–19 576. Petrides M, Pandya DN (1988) Association fiber 590. Press WA, Brewer AA, Dougherty RF, Wade AR, pathways to the frontal cortex from the superior Wandell BA (2001) Visual areas and spatial sum- temporal region in the rhesus monkey. J Comp mation in human visual cortex. Vision Res Neurol 273:52–66 41:1321–1332 577. Petrides M, Pandya DN (1994) Comparative ar- 591. Preuss TM (1995) The argument from humans to chitectonic analysis of the human and the maca- animals in cognitive neuroscience. In: Gazzaniga que frontal cortex. In: Boller F, Grafman J (eds) MS (ed) The cognitive neurosciences. MIT Press, Handbook of neuropsychology, vol 9. Elsevier, Boston, pp 1227–1241 Amsterdam, pp 17–58 592. Preuss TM, Goldman-Rakic PS (1991) Myelo- 578. Petrides M, Pandya DN (1999) Dorsolateral pre- and cytoarchitecture of the granular frontal cor- frontal cortex: comparative cytoarchitectonic tex and surrounding regions in the strepsirhine analysis in the human and the macaque brain primate Galago and the anthropoid primate Ma- and corticocortical connection patterns. Eur J caca. J Comp Neurol 310:429–474 Neurosci 11:1011–1036 593. Price CJ (2000) The anatomy of language: contri- 579. Petrides M, Pandya DN (2002) Association path- butions from functional neuroimaging. J Anat ways of the prefrontal cortex and functional ob- 197:335–359 servations. In: Stuss DT, Knight RT (eds) Princi- 594. Price CJ, Green DW, von Studnitz R (1999) A ples of frontal lobe function. Oxford University functional imaging study of translation and lan- Press, pp 31–50 guage switching. Brain 122:2221–2235 580. Petrides M, Pandya DN (2002) Comparative cyto- 595. Price JL, Carmichael ST, Drevets WC (1996) Net- architectonic analysis of the human and the ma- works related to the orbital and medial prefrontal caque ventrolateral prefrontal cortex and cortico- cortex; a substrate for emotional behavior? Prog cortical connection patterns in the monkey. Eur J Brain Res 107:523–536 Neurosci 16:291–310 596. Pritchard TC, Norgren R (2004) Gustatory sys- 581. Petrides M, Pandya DN (2004) The frontal cortex. tem. In: Paxinos G, Mai JK (eds) The human ner- In: Paxinos G, Mai JK (eds) The human nervous vous system, 2nd end. Elsevier, Amsterdam, pp system, 2nd edn. Elsevier, Amsterdam, pp 950– 1171–1196 972 597. Purves D, Riddle DR, LaMantia AS (1992) Iter- 582. Petrides M, Pandya DN (2006) Efferent associa- ated patterns of brain circuitry (or how the cor- tion pathways originating in the caudal prefrontal tex gets its spots). Trends Neurosci 15:362–368 cortex in the macaque monkey. J Comp Neurol 598. Rademacher J, Caviness VS Jr, Steinmetz H, Gala- 498:227–251 burda AM (1993) Topographical variation of the 583. Petrides M, Alivisatos B, Meyer E, Evans AC human primary cortices: implications for neuro- (1993) Functional activation of the human frontal imaging, brain mapping, and neurobiology. Cer- cortex during the performance of verbal working eb Cortex 3:313–329 memory tasks. Proc Natl Acad Sci USA 90:878– 599. Rademacher J, Burgel U, Geyer S et al (2001) 882 Variability and asymmetry in the human precen- 584. Phelps EA, LeDoux JE (2005) Contributions of tral motor system. A cytoarchitectonic and the amygdala to emotion processing: from animal myeloarchitectonic brain mapping study. Brain models to human behavior. Neuron 48:175–187 124: 2232–2258 585. Picard N, Strick PL (1996) Motor areas of the 600. Rademacher J, Morosan P, Schormann T et al medial wall: a review of their location and func- (2001) Probabilistic mapping and volume mea- tional activation. Cereb Cortex 6:342–353 surement of human primary auditory cortex. 586. Pons TP, Kaas JH (1986) Corticocortical connec- Neuroimage 13:669–683 tions of area 2 of somatosensory cortex in maca- 601. Raichle ME (1999) Modern phrenology: maps of que monkeys: a correlative anatomical and elec- human cortical function. Ann N Y Acad Sci 882: trophysiological study. J Comp Neurol 248:313– 107–118 336 602. Raichle ME (2003) Functional brain imaging and 587. Porter LL, Sakamoto K (1988) Organization and human brain function. J Neurosci 23:3959–3962 synaptic relationships of the projection from the 603. Raichle ME, Fiez JA, Videen TO et al (1994) primary sensory to the primary motor cortex in Practice-related changes in human brain func- the cat. J Comp Neurol 271:387–396 tional anatomy during nonmotor learning. Cereb 588. Poupon C, Clark CA, Frouin V et al (2000) Regu- Cortex 4:8–26 larization of diffusion-based direction maps for 672 Section II Structure of Spinal Cord and Brain Parts

604. Rajkowska G, Goldman-Rakic PS (1995) Cytoar- 621. Rockland KS, Ichinohe N (2004) Some thoughts on chitectonic definition of prefrontal areas in the cortical minicolumns. Exp Brain Res 158:265–277 normal human cortex: I. Remapping of areas 9 622. Rockland KS, Knutson T (2001) Axon collaterals and 46 using quantitative criteria. Cereb Cortex of Meynert cells diverge over large portions of 5:307–322 area V1 in the macaque monkey. J Comp Neurol 605. Rajkowska G, Goldman-Rakic PS (1995) Cy- 441:134–147 toarchitectonic definition of prefrontal areas in 623. Rockland KS, Pandya DN (1979) Laminar origins the normal human cortex: II. Variability in loca- and terminations of cortical connections of the tions of areas 9 and 46 and relationship to the occipital lobe in the rhesus monkey. Brain Res Talairach Coordinate System. Cereb Cortex 179:3–20 5:323–337 624. Roland PE, Zilles K (1994) Brain atlases – a new 606. Rakic P (1971) Neuron-glia relationship during research tool. Trends Neurosci 17:458–467 granule cell migration in developing cerebellar 625. Roland PE, Zilles K (1996) The developing Euro- cortex. A Golgi and electronmicroscopic study in pean computerized human brain database for all Macacus rhesus. J Comp Neurol 141:283–312 imaging modalities. Neuroimage 4:S39–S47 607. Rakic P (1972) Mode of cell migration to the su- 626. Roland PE, Zilles K (1998) Structural divisions perficial layers of fetal monkey neocortex. J and functional fields in the human cerebral cor- Comp Neurol 145:61–84 tex. Brain Res Rev 26:87–105 608. Rakic P (1988) Specification of cerebral cortical 627. Roland PE, Geyer S, Amunts K et al (1997) Cytoar- areas. Science 241:170–176 chitectural maps of the human brain in standard 609. Rakic P (1990) Principles of neural cell migra- anatomical space. Hum Brain Mapp 5:222–227 tion. Experientia 46:883–891 628. Rolls ET (2000) The orbitofrontal cortex and re- 610. Rakic P (1995) A small step for the cell, a giant ward. Cereb Cortex 10:284–294 leap for mankind: a hypothesis of neocortical ex- 629. Rolls ET, Baylis LL (1994) Gustatory, olfactory, pansion during evolution. Trends Neurosci 18: and visual convergence within the primate orbi- 383–388 tofrontal cortex. J Neurosci 14:5437–5452 611. Ramnani N (2006) The primate cortico-cerebellar 630. Rolls ET, Critchley HD, Mason R, Wakeman EA system: anatomy and function. Nat Rev Neurosci (1996) Orbitofrontal cortex neurons: role in olfac- 7:511–522 tory and visual association learning. J Neuro- 612. Ramnani N, Owen AM (2004) Anterior prefrontal physiol 75:1970–1981 cortex: insights into function from anatomy and 631. Romanski LM, Bates JF, Goldman-Rakic PS neuroimaging. Nat Rev Neurosci 5:184–194 (1999) Auditory belt and parabelt projections to 613. Rao SC, Rainer G, Miller EK (1997) Integration the prefrontal cortex in the rhesus monkey. J of what and where in the primate prefrontal cor- Comp Neurol 403:141–157 tex. Science 276:821–824 632. Romanski LM, Tian B, Fritz J et al (1999) Dual 614. Ray JP, Price JL (1993) The organization of pro- streams of auditory afferents target multiple do- jections from the mediodorsal nucleus of the tha- mains in the primate prefrontal cortex. Nat lamus to orbital and medial prefrontal cortex in Neurosci 2:1131–1136 macaque monkeys. J Comp Neurol 337:1–31 633. Rosano C, Sweeney JA, Melchitzky DS, Lewis DA 615. Retzius G (1896) Das Menschenhirn. Studien in (2003) The human precentral sulcus: chemoarchi- der makroskopischen Morphologie. Norstedt and tecture of a region corresponding to the frontal Söner, Stockholm eye fields. Brain Res 972:16–30 616. Ribak CE (1978) Aspinous and sparsely-spinous 634. Rose JE, Woolsey C (1948) Structure and rela- stellate neurons in the visual cortex of rats con- tions of limbic cortex and anterior thalamic nu- tain glutamic acid decarboxylase. J Neurocytol clei in rabbit and cat. J Comp Neurol 89:79–347 7:461–478 635. Rose JE, Woolsey CN (1948) The orbitofrontal 617. Robinson CJ, Burton H (1980) Organization of cortex and its connections with the mediodorsal somatosensory receptive fields in cortical areas nucleus in rabbit, sheep and cat. Res Publ Assoc 7b, retroinsula, postauditory and granular insula Nerv Ment Dis 27:210–232 of M. fascicularis. J Comp Neurol 192:69–92 636. Rose JE, Woolsey CN (1949) The relations of tha- 618. Robinson CJ, Burton H (1980) Somatic submod- lamic connections, cellular structure and evoc- ality distribution within the second somatosen- able electrical activity in the auditory region of sory (SII), 7b, retroinsular, postauditory, and the cat. J Comp Neurol 91:441–466 granular insular cortical areas of M. fascicularis. 637. Rose M (1935) Cytoarchitektonik und Myeloarch- J Comp Neurol 192:93–108 itektonik der Groszhirnrinde. In: Bumke O, 619. Rockel AJ, Hiorns RW, Powell TP (1974) Num- Foerster O (eds) Handbuch der Neurologie Band bers of neurons through full depth of neocortex. I: Anatomie. Springer, Berlin, pp 589–778 J Anat 118:371 638. Rosene DL, Pandya DN (1983) Architectonics 620. Rockel AJ, Hiorns RW, Powell TP (1980) The ba- and connections of the posterior parahippocam- sic uniformity in structure of the neocortex. pal gyrus in the rhesus monkey. Soc Neurosci Brain 103:221–244 Abstr 9:222 15 Telencephalon: Neocortex 673

639. Rosene DL, Van Hoesen GW (1977) Hippocampal 653. Schall JD (1997) Visuomotor areas of the frontal efferents reach widespread areas of the cerebral lobe. In: Jones EG, Peters A (eds) Extrastriate cortex and amygdala in the rhesus monkey. Sci- cortex in primates. Plenum, New York, pp 527– ence 198:315–317 638 (Cerebral cortex, vol 12) 640. Royce GJ (1987) Recent research on the centro- 654. Scheibel AB (1979) Development of axonal and median and parafascicular nuclei. In: Carpenter dendritic neuropil as a function of evolving be- MB, Jayaraman A (eds) The basal ganglia II: havior. In: Schmidt FO, Worden FG (eds) The Structure and function – current concepts. Ple- neurosciences fourth study program. MIT Press, num, New York, pp 293–319 (Advances in beha- Cambridge, p VII+1185 vioral biology, vol 32) 655. Scheibel AB, Paul LA, Fried I et al (1985) 641. Royce GJ, Mourey RJ (1985) Efferent connections Dendritic organization of the anterior speech of the centromedian and parafascicular thalamic area. Exp Neurol 87:109–117 nuclei: an autoradiographic investigation in the 656. Scheibel ME, Scheibel AB (1970) Elementary pro- cat. J Comp Neurol 235:277–300 cesses in selected thalamic and cortical subsys- 642. Rubens AB (1977) Anatomical asymmetries of tems: the structural substrates. In: Schmitt FO human cerebral cortex. In: Harnad S, Doty RW, (ed) The neurosciences second study program. Goldstein L, Jaynes J, Krauthamer G (eds) Rockefeller University Press, New York, pp 443– Lateralization in the nervous system. Academic 457 Press, New York, pp 503–517 657. Scheibel ME, Scheibel AB (1978) The dendritic 643. Ruiz-Marcos A, Valverde F (1970) Dynamic ar- structure of the human Betz cells. In: Brazier chitecture of the visual cortex. Brain Res 19:25– MAB, Petsche H (eds) Architectonics of the cere- 39 bral cortex. Raven, New York, pp 43–57 644. Rustioni A, Hayes NL (1981) Corticospinal tract 658. Schiltz C, Sorger B, Caldara R et al (2006) Im- collaterals to the dorsal column nuclei of cats. An paired face discrimination in acquired prosopag- anatomical single and double retrograde tracer nosia is associated with abnormal response to in- study. Exp Brain Res 43:237–245 dividual faces in the right middle fusiform gyrus. 645. Saffran E, Sholl A (1999) Clues to the functional Cereb Cortex 16:574–586 and neural architecture of word meaning. In: 659. Schleicher A, Amunts K, Geyer S, Morosan P, Brown CM, Hagoort P (eds) The neurocognition Zilles K (1999) Observer-independent method for of language. Oxford University Press, Oxford, pp microstructural parcellation of cerebral cortex: A 241–273 quantitative approach to cytoarchitectonics. 646. Saint Marie RL, Peters A (1985) The morphology Neuroimage 9:165–177 and synaptic connections of spiny stellate neu- 660. Schleicher A, Amunts K, Geyer S et al (2000) A rons in monkey visual cortex (area 17): a Golgi- stereological approach to human cortical archi- electron microscopic study. J Comp Neurol tecture: identification and delineation of cortical 233:213–235 areas. J Chem Neuroanat 20:31–47 647. Saleem KS, Suzuki W, Tanaka K, Hashikawa T 661. Schneider RJ, Friedman DP, Mishkin M (1993) A (2000) Connections between anterior inferotem- modality-specific somatosensory area within the poral cortex and superior temporal sulcus re- insula of the rhesus monkey. Brain Res 621:116– gions in the macaque monkey. J Neurosci 120 20:5083–5101 662. Schnitzler A, Seitz RJ, Freund HJ (2000) The so- 648. Saper CB (2002) The central autonomic nervous matosensory system. In: Toga AW, Mazziotta JC system: conscious visceral perception and auton- (eds) Brain mapping, the systems. Academic omic pattern generation. Annu Rev Neurosci Press, San Diego, pp 291–329 25:433–469 663. Schoenbaum G, Roesch MR, Stalnaker TA (2006) 649. Sarkissov SA, Filimonoff IN, Kononowa EP, Preo- Orbitofrontal cortex, decision-making and drug braschenskaja IS, Kukuew LA (1955) Atlas of the addiction. Trends Neurosci 29:116–124 cytoarchitectonics of the human cerebral cortex. 664. Schultz W (1998) Predictive reward signal of do- Medgiz, Moscow pamine neurons. J Neurophysiol 80:1–27 650. Sarter M, Bruno JP (2000) Cortical cholinergic 665. Schultz W (2002) Getting formal with dopamine inputs mediating arousal, attentional processing and reward. Neuron 36:241–263 and dreaming: differential afferent regulation of 666. Schultz W, Tremblay L, Hollerman JR (2000) Re- the basal forebrain by telencephalic and brain- ward processing in primate orbitofrontal cortex stem afferents. Neuroscience 95:933–952 and basal ganglia. Cereb Cortex 10:272–284 651. Sarter M, Bruno JP (2002) The neglected consti- 667. Schwarz ML, Goldman-Rakic PS (1984) Callosal tuent of the basal forebrain corticopetal projec- and intrahemispheric connectivity of the prefron- tion system: GABAergic projections. Eur J Neu- tal association cortex in Rhesus monkey: relation rosci 15:1867–1873 between intraparietal and principal sulcus cortex. 652. Sawaguchi T, Goldman-Rakic PS (1991) D1 dopa- J Comp Neurol 226:403–420 mine receptors in prefrontal cortex: involvement in working memory. Science 251:947–950 674 Section II Structure of Spinal Cord and Brain Parts

668. Scott SK, Johnsrude IS (2003) The neuroanatomi- gliaform cells and various interneuron types in cal and functional organization of speech percep- the neocortex. J Neurosci 25:6278–6285 tion. Trends Neurosci 26:100–107 684. Simon O, Kherif F, Flandin G et al (2004) Auto- 669. Selemon LD, Goldman-Rakic PS (1988) Common matized clustering and functional geometry of cortical and subcortical targets of the dorsolat- human parietofrontal networks for language, eral prefrontal and posterior parietal cortices in space, and number. Neuroimage 23:1192–1202 the rhesus monkey: evidence for a distributed 685. Simons JS, Spiers HJ (2003) Prefrontal and med- neural network subserving spatially guided beha- ial temporal lobe interactions in long-term mem- vior. J Neurosci 8:4049–4068 ory. Nat Rev Neurosci 4:637–648 670. Seltzer B, Pandya DN (1978) Afferent cortical 686. Sincich LC, Horton JC (2003) Independent pro- connections and architectonics of the superior jection streams from macaque striate cortex to temporal sulcus and surrounding cortex in the the second visual area and middle temporal area. rhesus monkey. Brain Res 149:1–24 J Neurosci 23:5684–5692 671. Seltzer B, Pandya DN (1984) Further observa- 687. Sincich LC, Park KF, Wohlgemuth MJ, Horton JC tions on parieto-temporal connections in the rhe- (2004) Bypassing V1: a direct geniculate input to sus monkey. Exp Brain Res 55:301–312 area MT. Nat Neurosci 7:1123–1128 672. Seltzer B, Pandya DN (1989) Frontal lobe connec- 688. Sloper JJ (1972) Gap junctions between dendrites tions of the superior temporal sulcus in the rhe- in the primate neocortex. Brain Res 44:641–646 sus monkey. J Comp Neurol 281:97–113 689. Sloper JJ, Powell TP (1978) Gap junctions be- 673. Seltzer B, Pandya DN (1989) Intrinsic connec- tween dendrites and somata of neurons in the tions and architectonics of the superior temporal primate sensori-motor cortex. Proc R Soc Lond B sulcus in the rhesus monkey. J Comp Neurol Biol Sci 203:39–47 290:451–471 690. Sloper JJ, Powell TPS (1979) An experimental 674. Seltzer B, Pandya DN (1991) Post-rolandic corti- electron microscopic study of afferent connec- cal projections of the superior temporal sulcus in tions to the primate motor and somatic sensory the rhesus monkey. J Comp Neurol 312:625–640 cortices. Philos Trans R Soc Lond B Biol Sci 675. Seltzer B, Pandya DN (1994) Parietal, temporal, 285:199–226 and occipital projections to cortex of the superior 691. Smiley JF, Goldman-Rakic PS (1996) Serotonergic temporal sulcus in the rhesus monkey: a retro- axons in monkey prefrontal cerebral cortex sy- grade tracer study. J Comp Neurol 343:445–463 napse predominantly on interneurons as demon- 676. Seltzer B, Cola MG, Gutierrez C et al (1996) strated by serial section electron microscopy. J Overlapping and nonoverlapping cortical projec- Comp Neurol 367:431–443 tions to cortex of the superior temporal sulcus in 692. Smiley JF, Levey AI, Ciliax BJ, Goldman-Rakic PS the rhesus monkey: double anterograde tracer (1994) D1 dopamine receptor immunoreactivity studies. J Comp Neurol 370:173–190 in human and monkey cerebral cortex: predomi- 677. Sesack SR, Snyder CL, Lewis DA (1995) Axon nant and extrasynaptic localization in dendritic terminals immunolabeled for dopamine or tyro- spines. Proc Natl Acad Sci USA 91:5720–5724 sine hydroxylase synapse on GABA-immunoreac- 693. Söhl G, Maxeiner S, Willecke K (2005) Expres- tive dendrites in rat and monkey cortex. J Comp sion and functions of neuronal gap junctions. Neurol 363:264–280 Nat Rev Neurosci 6:191–200 678. Sewards TV, Sewards MA (2002) On the neural 694. Somogyi P, Cowey A (1981) Combined Golgi and correlates of object recognition awareness: rela- electron microscopic study on the synapses tionship to computational activities and activities formed by double bouquet cells in the visual cor- mediating perceptual awareness. Conscious Cogn tex of the cat and monkey. J Comp Neurol 195: 11:51–77 547–566 679. Shepherd GM (2004) Introduction to synaptic 695. Somogyi P, Cowey A (1984) Double bouquet circuits. In: Shepherd GM (ed) The synaptic or- cells. In: Peters A, Jones EG (eds) Cellular com- ganization of the brain, 5th edition. Oxford Uni- ponents of the cerebral cortex. Plenum, New versity Press, Oxford, pp 1–38 York, pp 337–360 (Cerebral cortex, vol 1) 680. Sherman SM, Guillery RW (2001) Exploring the 696. Somogyi P, Hodgson AJ (1985) Antisera to gam- thalamus. Academic Press, San Diego ma-aminobutyric acid. III. Demonstration of 681. Sherman SM, Guillery RW (2002) The role of the GABA in Golgi-impregnated neurons and in con- thalamus in the flow of information to the cortex. ventional electron microscopic sections of cat Phil Trans R Soc Lond B 357:1695–1708 striate cortex. J Histochem Cytochem 33:249–257 682. Shimazu H, Maier MA, Cerri G, Kirkwood PA, 697. Somogyi P, Soltesz I (1986) Immunogold demon- Lemon RN (2004) Macaque ventral premotor cor- stration of GABA in synaptic terminals of intra- tex exerts powerful facilitation of motor cortex cellularly recorded, horseradish peroxidase-filled outputs to upper limb motoneurons. J Neurosci basket cells and clutch cells in the cat’s visual 24:1200–1211 cortex. Neuroscience 19:1051–1065 683. Simon A, Oláh S, Molnar G, Szabadics J, Tamás 698. Somogyi P, Hodgson AJ, Smith AD (1979) An G (2005) Gap-junctional coupling between neuro- approach to tracing neuron networks in the cere- 15 Telencephalon: Neocortex 675

bral cortex and basal ganglia. Combination of 713. Stokking R, Zuiderveld KJ, Viergever MA (2001) Golgi staining, retrograde transport of horserad- Integrated volume visualization of functional im- ish peroxidase and anterograde degeneration of age data and anatomical surfaces using normal synaptic boutons in the same material. Neu- fusion. Hum Brain Mapp 12:203–218 roscience 4:1805–1852 714. Sudakov K, MacLean PD, Reeves A, Marino R 699. Somogyi P, Freund TF, Cowey A (1982) The axo- (1971) Unit study of exteroceptive inputs to axonic interneuron in the cerebral cortex of the claustrocortex in awake, sitting, squirrel monkey. rat, cat and monkey. Neuroscience 7:2577–2607 Brain Res 28:19–34 700. Somogyi P, Kisvárday ZF, Martin KA, Whitter- 715. Suzuki W, Saleem KS, Tanaka K (2000) Divergent idge D (1983) Synaptic connections of morpholo- backward projections from the anterior part of gically identified and physiologically character- the inferotemporal cortex (area TE) in the maca- ized large basket cells in the striate cortex of cat. que. J Comp Neurol 422:206–228 Neuroscience 10:261–294 716. Suzuki WA, Amaral DG (1994) Perirhinal and 701. Somogyi P, Freund TF, Hodgson AJ et al (1985) parahippocampal cortices of the macaque mon- Identified axo-axonic cells are immunoreactive key: cortical afferents. J Comp Neurol 350:497– for GABA in the hippocampus and visual cortex 533 of the cat. Brain Res 332:143–149 717. Suzuki WA, Amaral DG (1994) Topographic orga- 702. Somogyi P, Tamás G, Lujan R, Buhl EH (1998) nization of the reciprocal connections between Salient features of synaptic organisation in the the monkey entorhinal cortex and the perirhinal cerebral cortex. Brain Res Rev 26:113–135 and parahippocampal cortices. J Neurosci 14: 703. Sperry R (1986) Consciousness, personal identity, 1856–1877 and the divided brain. In: Lepore F, Ptito M, Jas- 718. Suzuki WA, Amaral DG (2003) Where are the per HH (eds) Two hemispheres – one brain: perirhinal and parahippocampal cortices? A his- functions of the corpus callosum. Liss, New York, torical overview of the nomenclature and bound- pp 3–20 aries applied to the primate medial temporal 704. Staiger JF, Flagmeyer I, Schubert D et al (2004) lobe. Neuroscience 120:893–906 Functional diversity of layer IV spiny neurons in 719. Swindale NV (1990) Is the cerebral cortex modu- rat somatosensory cortex: quantitative morphol- lar? Trends Neurosci 13:487–492 ogy of electrophysiologically characterized and 720. Szentágothai J (1975) The ‘module-concept’ in biocytin labeled cells. Cereb Cortex 14:690–701 cerebral cortex architecture. Brain Res 95:475– 705. Starr MA (1884) Cortical lesions of the brain: A 496 collection and analysis of the American cases of 721. Szentágothai J (1978) The Ferrier Lecture, 1977. localized cerebral disease. Am J Med Sci 88:114– The neuron network of the cerebral cortex: a 141 functional interpretation. Proc R Soc Lond B Biol 706. Steele GE, Weller RE, Cusick CG (1991) Cortical Sci 201:219–248 connections of the caudal subdivision of the dor- 722. Szentágothai J (1979) Local neuron circuits of the solateral area (V4) in monkeys. J Comp Neurol neocortex. In: Schmitt FO, Worden FG (eds) The 306:495–520 neurosciences fourth study program. MIT Press, 707. Stein BE, Meredith MA, Wallace MT (1993) The Cambridge, pp 399–415 visually responsive neuron and beyond: multisen- 723. Szentágothai J (1983) The modular architectonic sory integration in cat and monkey. Prog Brain principle of neural centers. Rev Physiol Biochem Res 95:79–90 Pharmacol 98:11–61 708. Steinmetz H (1996) Structure, functional and 724. Takahashi H, Yahata N, Koeda M et al (2004) cerebral asymmetry: in vivo morphometry of the Brain activation associated with evaluative pro- planum temporale. Neurosci Biobehav Rev 20: cesses of guilt and embarrassment: an fMRI 587–591 study. Neuroimage 23:967–974 709. Stephan H, Andy OJ (1964) Quantitative compar- 725. Takahashi T, Suzuki M, Hagino H et al (2004) Bi- isons of brain structures from insectivores to pri- lateral volume reduction of the insular cortex in mates. Am Zool 136:59–74 patients with schizophrenia: a volumetric MRI 710. Stephan KE, Kamper L, Bozkurt A et al (2001) study. Psychiatry Res 131:185–194 Advanced database methodology for the Col- 726. Takahashi T, Suzuki M, Zhou SY et al (2005) Vo- lation of Connectivity data on the Macaque brain lumetric MRI study of the short and long insular (CoCoMac). Philos Trans R Soc Lond B Biol Sci cortices in schizophrenia spectrum disorders. 356:1159–1186 Psychiatry Res 138:209–220 711. Stepniewska I, Preuss TM, Kaas JH (2006) Ipsilat- 727. Talairach J, Tournoux P (1988) Co-Planar Stereo- eral cortical connections of dorsal and ventral taxic Atlas of the Human Brain. 3-D propor- premotor areas in New World owl monkeys. J tioned system: an approach to cerebral imaging. Comp Neurol 495:691–708 Thieme, Stuttgart 712. Steriade M, Jones EG, McCormick DA (1997) 728. Tamás G, Buhl EH, Somogyi P (1997) Massive Thalamus, vol I: Organisation and function. Else- autaptic self-innervation of GABAergic neurons vier, Amsterdam in cat visual cortex. J Neurosci 17:6352–6364 676 Section II Structure of Spinal Cord and Brain Parts

729. Tamás G, Buhl EH, Lorincz A, Somogyi P (2000) 746. Uylings HBM, Malofeeva LI, Bogolepova IN, Proximally targeted GABAergic synapses and gap Amunts K, Zilles K (1999) Broca’s language area junctions synchronize cortical interneurons. Nat from a neuroanatomical and developmental per- Neurosci 3:366–371 spectives. In: Brown CM, Hagoort P (eds) The 730. Tanaka K (1996) Inferotemporal cortex and ob- neurocognition of language. Oxford University ject vision. Annu Rev Neurosci 19:109–139 Press, Oxford, pp 319–336 731. Tanaka Y, Miyazawa Y, Hashimoto R, Nakano I, 747. Uylings HBM, Zilles K, Rajkowska G (1999) Opti- Obayashi T (1999) Postencephalitic focal retro- mal staining methods for delineation of cortical grade amnesia after bilateral anterior temporal areas and neuron counts in human brains. Neu- lobe damage. Neurology 53:344–350 roimage 9:439–445 732. Tatschl C, Stollberger C, Matz K et al (2006) Insu- 748. Uylings HBM, Sanz-Arigita E, de Vos K et al lar involvement is associated with QT prolonga- (2000) The importance of a human 3D database tion: ECG abnormalities in patients with acute and atlas for studies of prefrontal and thalamic stroke. Cerebrovasc Dis 21:47–53 functions. Prog Brain Res 126:357–368 733. Thier P, Andersen RA (1998) Electrical micro- 749. Uylings HBM, Groenewegen HJ, Kolb B (2003) stimulation distinguishes distinct saccade-related Do rats have a prefrontal cortex? Behav Brain Res areas in the posterior parietal cortex. J Neuro- 146:3–17 physiol 80:1713–1735 750. Uylings HBM, Rajkowska G, Sanz-Arigita E, 734. Tigges J, Tigges M (1985) Subcortical sources of Amunts K, Zilles K (2005) Consequences of large direct projections to visual cortex. In: Peters A, interindividual variability for human brain at- Jones EG (eds) Visual cortex. Plenum, New York, lases: converging macroscopical imaging and pp 351–378 (Cerebral cortex, vol 3) microscopical neuroanatomy. Anat Embryol 735. Tigges J, Tigges M, Anschel S et al (1981) Areal (Berl) 210:423–431 and laminar distribution of neurons intercon- 751. Uylings HBM, Sanz-Arigita EJ, de Vos K et al necting the central visual cortical areas 17, 18, 19 (2007) 3-D cytoarchitectonic parcellation of hu- and MT in squirrel monkey (Saimiri). J Comp man orbitofrontal cortex. Correlation with post- Neurol 202:539–560 mortem MRI. [submitted] 736. Tootell RB, Hadjikhani N (2001) Where is ‘dorsal 752. Valenstein ES (1990) The prefrontal area and psy- V4’ in human visual cortex? Retinotopic, topo- chosurgery. Prog Brain Res 85:539–553 graphic and functional evidence. Cereb Cortex 753. Valverde F (1967) Apical dendritic spines of the 11:298–311 visual cortex and light deprivation in the mouse. 737. Tootell RB, Hadjikhani NK, Mendola JD, Marrett Exp Brain Res 3:337–352 S, Dale AM (1998) From retinotopy to recogni- 754. Valverde F (1976) Aspects of cortical organization tion: fMRI in human visual cortex. Trends Cogn related to the geometry of neurons with intra- Sci 2:174–183 cortical axons. J Neurocytol 5:509–529 738. Towe AL (1975) Notes on the hypothesis of co- 755. Valverde F (1985) The organizing principles of lumnar organization in somatosensory cerebral the primary visual cortex in the monkey. In: Pe- cortex. Brain Behav Evol 11:16–47 ters A, Jones EG (eds) Visual cortex. Plenum, 739. Tusa RJ, Ungerleider LG (1985) The inferior long- New York, pp 207–257 (Cerebral cortex, vol 3) itudinal fasciculus: a reexamination in humans 756. Valverde F (1986) Intrinsic neocortical organiza- and monkeys. Ann Neurol 18:583–591 tion: some comparative aspects. Neuroscience 740. Ullman MT (2001) A neurocognitive perspective 18:1–23 on language: the declarative/procedural model. 757. Valverde F, Facal-Valverde MV (1986) Neocortical Nat Rev Neurosci 2:717–726 layers I and II of the hedgehog (Erinaceus euro- 741. Ungerleider LG, Desimone R (1986) Cortical con- paeus). I. Intrinsic organization. Anat Embryol nections of visual area MT in the macaque. J (Berl) 173:413–430 Comp Neurol 248:190–223 758. Van der Loos H (1979) The development of topo- 742. Ungerleider LG, Mishkin M (1979) The striate logical equivalencies in the brain. In: Meisami A, projection zone in the superior temporal sulcus Brazier MAB (eds) Neuronal growth and differ- of Macaca mulatta: location and topographic or- entiation. Raven, New York, pp 331–336 ganization. J Comp Neurol 188:347–367 759. Van der Loos H, Woolsey TA (1973) Somatosen- 743. Ungerleider LG, Mishkin M (1982) Two cortical sory cortex: structural alterations following early visual systems. In: Ingle et al (eds) Analysis of injury to sense organs. Science 179:395–398 visual behavior. MIT Press, Cambridge, pp 549– 760. Van Eden CG, Mrzljak L, Voorn P, Uylings HB 586 (1989) Prenatal development of GABA-ergic neu- 744. Uttal WR (2001) The new phrenology: the limits rons in the neocortex of the rat. J Comp Neurol of localizing cognitive processes in the brain. 289:213–227 MIT Press, Cambridge 761. Van Essen DC (1985) Functional organization of 745. Uttal WR (2002) Précis of the new phrenology: primate visual cortex. In: Peters A, Jones EG the limits of localizing cognitive processes in the (eds) Visual cortex. Plenum, New York, pp 259– brain. Brain and Mind 3:221–228 329 (Cerebral cortex, vol 3) 15 Telencephalon: Neocortex 677

762. Van Essen DC (2002) Windows on the brain: the in area 17 projecting to the extrastriate areas MT emerging role of atlases and databases in neuro- and 19DM and to the superior colliculus in the science. Curr Opin Neurobiol 12:574–579 monkey Callithrix jacchus. J Comp Neurol 763. Van Essen DC (2004) Surface-based approaches 362:233–255 to spatial localization and registration in primate 779. Vogt BA (1991) The role of layer I in cortical cerebral cortex. Neuroimage 23 [Suppl 1]:S97–107 function. In: Peters A, Jones EG (eds) Normal 764. Van Essen DC (2005) A Population-Average, and altered states of function. Plenum, New York, Landmark- and Surface-based (PALS) atlas of hu- pp 49–80 (Cerebral cortex, vol 9) man cerebral cortex. Neuroimage 28:635–662 780. Vogt BA (2005) Pain and emotion interactions in 765. Van Essen DC (2006) SumsDB website: http:// subregions of the cingulate gyrus. Nat Rev sumsdb.wustl.edu:8081/sums/index.jsp Neurosci 6:533–544 766. Van Essen DC, Drury HA (1997) Structural and 781. Vogt BA, Pandya DH (1978) Cortico-cortical con- functional analyses of human cerebral cortex nections of somatic sensory cortex (areas 3, 1 using a surface-based atlas. J Neurosci 17:7079– and 2) in the rhesus monkey. J Comp Neurol 7102 177:179–192 767. Van Essen DC, Maunsell JHR (1980) Two-dimen- 782. Vogt BA, Pandya DN (1987) Cingulate cortex of sional maps of the cerebral cortex. J Comp Neu- the rhesus monkey: II. Cortical afferents. J Comp rol 191:255–281 Neurol 262:271–289 768. Van Essen DC, Maunsell JHR, Bixby JL (1981) 783. Vogt BA, Hof PR, Vogt LJ (2004) Cingulate gyrus. The middle temporal visual area in the macaque: In: Paxinos G (ed) The human nervous system, myelo-architecture, connections, functional prop- 2nd edn. Elsevier, Amsterdam, pp 915–949 erties and topographic organization. J Comp 784. Vogt C, Vogt O (1919) Allgemeinere Ergebnisse Neurol 199:293–326 unserer Hirnforschung. J Psychol Neurol 25:279– 769. Van Essen DC, Newsome WT, Maunsell JHR, Bix- 461 by JL (1986) The projections from striate cortex 785. Vogt C, Vogt O (1926) Die vergleichend architek- (V1) to areas V2 and V3 in the macaque mon- tonische und die vergleichend physiologische Fel- key: asymmetries, areal boundaries, and patchy derung der Grosshirnrinde. Naturwissenschaften connections. J Comp Neurol 244:451–481 14:1191–1194 770. Van Essen DC, Anderson CH, Felleman DJ (1992) 786. Vogt C, Vogt O (1942) Morphologische Gestaltun- Information processing in the primate visual sys- gen unter normalen und pathogenen Bedingun- tem: an integrated systems perspective. Science gen. J Psychol Neurol 50:162–524 255:419–423 787. Vogt C, Vogt O (1954) Gestaltung der topistischen 771. Van Essen DC, Lewis JW, Drury HA et al (2001) Hirnforschung und ihre Förderung durch den Mapping visual cortex in monkeys and humans Hirnbau und seine Anomalien. J Hirnforsch 1:1–46 using surface-based atlases. Vision Res 41:1359– 788. Vogt C, Vogt O (1956) Weitere Ausführungen 1378 zum Arbeitsprogramm des Hirnforschungsinsti- 772. Van Hoesen GW (1982) The parahippocampal tutes in Neustadt/Schwarzwald. J Hirnforsch gyrus. New observations regarding its cortical 2:404–427 connections in the monkey. Trends Neurosci 789. Vogt O (1910) Die myeloarchitektonische Felde- 5:345–350 rung des menschlichen Stirnhirns. J Psychol 773. Van Hoesen GW (1993) The modern concept of Neurol 15:221–232 association cortex. Curr Opin Neurobiol 3:150– 790. Vogt O (1911) Die Myeloarchitektonik des Isocor- 154 tex parietalis. J Psychol Neurol 18:379–390 774. Van Hoesen GW, Rosene DL, Mesulam M-M 791. Vogt O (1927) Architektonik der menschlichen (1979) Subicular input from temporal cortex in Hirnrinde. Allg Z Psychiatr 86:247–274 the rhesus monkey. Science 205:608–610 792. Volkow ND, Fowler JS, Wang GJ (2004) The ad- 775. Vercelli AE, Assal F, Innocenti GM (1992) Emer- dicted human brain viewed in the light of ima- gence of callosally projecting neurons with stel- ging studies: brain circuits and treatment strate- late morphology in the visual cortex of the kit- gies. Neuropharmacology 47 Suppl 1:3–13 ten. Exp Brain Res 90:346–358 793. Von Bonin G (1950) Essay on the cerebral cortex. 776. Vercelli AE, Garbossa D, Curtetti R, Innocenti Charles C Thomas, Springfield, IL GM (2004) Somatodendritic minicolumns of out- 794. Von Bonin G, Bailey P (1947) The neocortex of put neurons in the rat visual cortex. Eur J Neu- Macaca mulatta. University of Illinois Press, Ur- rosci 20:495–502 bana 777. Verhagen JV, Kadohisa M, Rolls ET (2004) Pri- 795. Von Economo C (1927) Zellaufbau der Grosz- mate insular/opercular taste cortex: neuronal rep- hirnrinde des Menschen. Springer, Berlin resentations of the viscosity, fat texture, gritti- 796. Von Economo C, Koskinas GN (1925) Die Cyto- ness, temperature, and taste of foods. J Neuro- architektonik der Hirnrinde des erwachsenen physiol 92:1685–1699 Menschen. Springer, Berlin 778. Vogt Weisenhorn DM, Illing RB, Spatz WB 797. Von Monakow C (1911) Lokalisation der Hirn- (1995) Morphology and connections of neurons funktionen. J Psychol Neurol 17:185–200 678 Section II Structure of Spinal Cord and Brain Parts

798. Von Monakow C (1914) Die Lokalisation im 815. White EL (1989) Cortical circuits. Synaptic orga- Grosshirn und der Abbau der Funktion durch nization of the cerebral cortex: Structure, func- Kortikale Herde. J.F. Bergmann, Wiesbaden tion and theory. Birkhäuser, Boston 799. Wada J, Rasmussen T (1960) Intracarotid injection 816. White EL, Czeiger D (1991) Synapses made by of sodium amytal for the lateralisation of cerebral axons of callosal projection neurons in mouse so- speech dominance. J Neurosurg 17:226–282 matosensory cortex: emphasis on intrinsic con- 800. Wada JA, Clarke R, Hamm A (1975) Cerebral nections. J Comp Neurol 303:233–244 hemispheric asymmetry in humans. Cortical 817. White EL, Hersch SM (1981) Thalamocortical sy- speech zones in 100 adults and 100 infant brains. napses of pyramidal cells which project from Arch Neurol 32:239–246 SmI to MsI cortex in the mouse. J Comp Neurol 801. Wahle P (1993) Differential regulation of sub- 198: 167–181 stance P and somatostatin in Martinotti cells of 818. White EL, Hersch SM (1982) A quantitative study the developing cat visual cortex. J Comp Neurol of thalamocortical and other synapses involving 329:519–538 the apical dendrites of corticothalamic projection 802. Walker AE (1940) A cytoarchitectural study of cells in mouse SmI cortex. J Neurocytol 11:137– the prefrontal area of the macaque monkey. J 157 Comp Neurol 73:59–86 819. White EL, Keller A (1987) Intrinsic circuitry in- 803. Walker AE (1944) Psychosurgery: collective re- volving the local axon collaterals of corticothal- view. Int Abstr Surg 78:1–11 amic projection cells in mouse SmI cortex. J 804. Wang XJ (2001) Synaptic reverberation underly- Comp Neurol 262:13–26 ing mnemonic persistent activity. Trends Neuros- 820. White EL, Hersch SM, Belford GR (1982) Quanti- ci 24:455–463 tative studies of thalamocortical synapses with 805. Wang Y, Gupta A, Toledo-Rodriguez M, Wu CZ, labeled pyramidal cells in mouse SmI cortex. Soc Markram H (2002) Anatomical, physiological, Neurosci Abstr 8:853 molecular and circuit properties of nest basket 821. Willmes K, Poeck K (1993) To what extent can cells in the developing somatosensory cortex. aphasic syndromes be localized? Brain 116:1527– Cereb Cortex 12:395–410 1540 806. Wang Y, Markram H, Goodman PH et al (2006) 822. Wilson FA, Scalaidhe SP, Goldman-Rakic PS Heterogeneity in the pyramidal network of the (1993) Dissociation of object and spatial process- medial prefrontal cortex. Nat Neurosci 9:534–542 ing domains in primate prefrontal cortex. Science 807. Wartenburger I, Heekeren HR, Burchert F et al 260:1955–1958 (2004) Neural correlates of syntactic transforma- 823. Winer JA (1982) The stellate neurons in layer IV tions. Hum Brain Mapp 22:72–81 of primary auditory cortex (AI) of the cat: a 808. Watson JD, Myers R, Frackowiak RS et al (1993) study of columnar organization. Soc Neurosci Area V5 of the human brain: evidence from a Abstr 8:1020 combined study using positron emission tomog- 824. Winfield DA, Gatter KC, Powell TPS (1980) An raphy and magnetic resonance imaging. Cereb electron microscopic study of the types and pro- Cortex 3:79–94 portions of neurons in the cortex of the motor 809. Webster MJ, Ungerleider LG, Bachevalier J (1991) and visual areas of the cat and rat. Brain Connections of inferior temporal areas TE and 103:245–258 TEO with medial temporal-lobe structures in in- 825. Winfield DA, Brooke RN, Sloper JJ, Powell TP fant and adult monkeys. J Neurosci 11:1095–1116 (1981) A combined Golgi-electron microscopic 810. Werner L, Winkelmann E, Koglin A, Neser J, Ro- study of the synapses made by the proximal axon dewohl H (1989) A Golgi deimpregnation study and recurrent collaterals of a pyramidal cell in of neurons in the rhesus monkey visual cortex the somatic sensory cortex of the monkey. (areas 17 and 18). Anat Embryol (Berl) 180:583– Neuroscience 6:1217–1230 597 826. Winkelmann E, Brauer K, Berger U (1975) Zur 811. Wernicke C (1874) Der aphasische Symptomen- columnaren Organisation von Pyramidenzellen komplex. Cohen und Weigert, Breslau im visuellen Cortex der Albinoratte. Z Mikrosk 812. Wernicke C (1881) Lehrbuch der Gehirnkran- Anat Forsch 89:239–256 kheiten. Fischer, Kassel 827. Wise RJ, Scott SK, Blank SC et al (2001) Separate 813. White EL (1978) Identified neurons in mouse neural subsystems within ‘Wernicke’s area’. Brain Sml cortex which are postsynaptic to thalamo- 124:83–95 cortical axon terminals: a combined Golgi-elec- 828. Wise SP, Boussaoud D, Johnson PB, Caminiti R tron microscopic and degeneration study. J Comp (1997) Premotor and parietal cortex: corticocorti- Neurol 181:627–661 cal connectivity and combinatorial computations. 814. White EL (1986) Termination of thalamic affer- Annu Rev Neurosci 20:25–42 ents in the cerebral cortex. In: Jones EG, Peters A 829. Witelson SF, Pallie W (1973) Left hemisphere (eds) Sensory-motor areas and aspects of cortical specialization for language in the newborn. connectivity. Plenum, New York, pp 271–289 Neuroanatomical evidence of asymmetry. Brain (Cerebral cortex, vol 5) 96:641–646 15 Telencephalon: Neocortex 679

830. Woolsey CN (1958) Organization of somatic sen- coding face-hand coordination in Japanese maca- sory and motor areas of the cerebral cortex. In: ques. Somatosens Mot Res 20:115–125 Harlow HF, Woolsey CN (eds) Biological and bio- 841. Yoshimura Y, Callaway EM (2005) Fine-scale spe- chemical bases of behavior. University of Wiscon- cificity of cortical networks depends on inhibi- sin Press, Madison, pp 63–81 tory cell type and connectivity. Nat Neurosci 831. Woolsey CN (1964) Cortical localization as de- 8:1552–1559 fined by evoked potential and electrical stimula- 842. Yoshimura Y, Dantzker JL, Callaway EM (2005) tion studies. In: Schaltenbrand G, Woolsey CN Excitatory cortical neurons form fine-scale func- (eds) Cerebral localization and organization. Uni- tional networks. Nature 433:868–873 versity of Wisconsin Press, Madison, pp 17–32 843. Yukie M (2000) Connections between the medial 832. Woolsey CN, Marshall WH, Bard P (1942) temporal cortex and the CA1 subfield of the hip- Representation of cutaneous tactile sensibility in pocampal formation in the Japanese monkey the cerebral cortex of the monkey as indicated by (Macaca fuscata). J Comp Neurol 423:282–298 evoked potentials. Bull Johns Hopkins Hosp 844. Yuste R, Peinado A, Katz LC (1992) Neuronal do- 70:399–441 mains in developing neocortex. Science 257:665– 833. Woolsey CN, Marshall WH, Bard P (1943) Note 669 on the organization of the tactile sensory area of 845. Zald DH, Pardo JV (2000) Functional neuroimag- the cerebral cortex of the chimpanzee. J Neuro- ing of the olfactory system in humans. Int J Psy- physiol 6:287–291 chophysiol 36:165–181 834. Woolsey TA, van der Loos H (1970) The structural 846. Zeki S (1997) The color and motion systems as organization of layer IV in the somatosensory re- guides to conscious visual perception. In: Jones gion (SI) of mouse cerebral cortex. Brain Res EG, Peters A (eds) Extrastriate cortex in pri- 17:205–242 mates. Plenum, New York, pp 777–809 (Cerebral 835. Xiang Z, Huguenard JR, Prince DA (1998) Choli- cortex, vol 12) nergic switching within neocortical inhibitory 847. Zeki S (2003) Improbable areas in the visual networks. Science 281:985–988 brain. Trends Neurosci 26:23–26 836. Yamamoto T, Samejima A, Oka H (1987) Mor- 848. Zilbovicius M, Meresse I, Chabane N et al (2006) phological features of layer V pyramidal neurons Autism, the superior temporal sulcus and social in the cat parietal cortex: an intracellular HRP perception. Trends Neurosci 29:359–366 study. J Comp Neurol 265:380–390 849. Zilles K (2004) Architecture of the human cere- 837. Yamamoto T, Samejima A, Oka H (1987) Mor- bral cortex. In: Paxinos G, Mai JK (eds) The hu- phology of layer V pyramidal neurons in the cat man nervous system, 2nd edn. Elsevier, Amster- somatosensory cortex: an intracellular HRP dam, pp 997–1055 study. Brain Res 437:369–374 850. Zilles K, Clarke S (1997) Architecture, connectiv- 838. Yaxley S, Rolls ET, Sienkiewicz ZJ (1990) Gusta- ity, and transmitter receptors of human extrastri- tory responses of single neurons in the insula of ate visual cortex. In: Jones EG, Peters A (eds) Ex- the macaque monkey. J Neurophysiol 63:689–700 trastriate cortex in primates. Plenum, New York, 839. Yeterian EH, Pandya DN (1991) Prefrontostriatal pp 673–742 (Cerebral cortex, vol 12) connections in relation to cortical architectonic 851. Zilles K, Schleicher A, Palomero-Gallagher N, organization in rhesus monkeys. J Comp Neurol Amunts K (2002) Quantitative analysis of cyto- 312:43–67 and receptorarchitecture of the human brain. In: 840. Yokochi H, Tanaka M, Kumashiro M, Iriki A Toga AW, Mazziotta JC (eds) Brain mapping: The (2003) Inferior parietal somatosensory neurons methods, 2nd edn. Academic Press, San Diego, pp 573–602