b Department of Neuroimaging Sciences, Institute of Psychiatry, King’s College London, is a functionally integrated system interconnected by short- and.
132 KB – 14 Pages
PAGE – 1 ============
Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737Contents lists available at ScienceDirect Neuroscience and Biobehavioral Reviews journal h om epa ge: www.elsevier.com/locate/neubiorevReviewA revised limbic system model for memory, emotion and behaviourMarco Catania,, Flavio Dell’Acquaa,b,c, Michel Thiebaut de Schottena,d,aNatbrainlab, Department of Forensic and Neurodevelopmental Sciences, Institute of Psychiatry, King’s College London, UKbDepartment of Neuroimaging Sciences, Institute of Psychiatry, King’s College London, UKcNIHR Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust and Institute of Psychiatry, King’s CollegeLondon, UKdUMR S 975; CNRS UMR 7225, Centre de Recherche de l’Institut du Cerveau et de la Moelle épinière, Groupe Hospitalier Pitié-Salpêtrière, 75013 Paris,Francea r t i c l e i n f oArticle history:Received 27 November 2012Received in revised form 15 May 2013Accepted 1 July 2013Keywords: Limbic systemTractography White matter connectionsBrain networksEmotion Memory Amnesia Dementia Antisocial behaviourSchizophrenia Depression Bipolar disorderObsessive–compulsive disorderAutism spectrum disordera b s t r a c tEmotion, memories and behaviour emerge from the coordinated activities of regions connected by thelimbic system. Here, we propose an update of the limbic model based on the seminal work of Papez,Yakovlev and MacLean. In the revised model we identify three distinct but partially overlapping networks:(i) the Hippocampal-diencephalic and parahippocampal-retrosplenial network dedicated to memory andspatial orientation; (ii) The temporo-amygdala-orbitofrontal network for the integration of visceral sen-sation and emotion with semantic memory and behaviour; (iii) the default-mode network involved inautobiographical memories and introspective self-directed thinking. The three networks share corticalnodes that are emerging as principal hubs in connectomic analysis. This revised network model of thelimbic system reconciles recent functional imaging findings with anatomical accounts of clinical disorderscommonly associated with limbic pathology.© 2013 Elsevier Ltd. All rights reserved.Contents1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17252. Anatomy of the limbic system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17262.1. Fornix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17262.2. Mammillo-thalamic tract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17262.3. Anterior thalamic projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17272.4. Cingulum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17272.5. Uncinate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17283. Functional anatomy of the limbic system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17294. Limbic syndromes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17304.1. Hippocampal-diencephalic and parahippocampal-retrosplenial syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17304.2. Temporal-amygdala-orbitofrontal syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17314.3. Default network syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17335. Conclusions and future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1734Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1734References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1734Corresponding authors at: Natbrainlab, Department of Forensic and Neurodevelopmental Sciences, Institute of Psychiatry, 16 De Crespigny Park, London SE5 8AF, UK.E-mail addresses: m.catani@iop.kcl.ac.uk (M. Catani), michel.thiebaut@gmail.com (M. Thiebaut de Schotten).0149-7634/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.neubiorev.2013.07.001
PAGE – 2 ============
M. Catani et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737 17251. IntroductionThe limbic system is a group of interconnected cortical andsubcortical structures dedicated to linking visceral states and emo-tion to cognition and behaviour (Mesulam, 2000). The use of theterm ‘limbic’ has changed over time. Initially introduced by ThomasWillis (1664) to designate a cortical border encircling the brain-stem (limbus, Latin for ‘border’) (Fig. 1) the term has been usedin more recent times to indicate a progressively increasing num-ber of regions dedicated to a wide range of functions (Marshalland Magoun, 1998; Mega et al., 1997). Paul Broca (1878) held theview that ‘le grand lobe limbique’ was mainly an olfactory structurecommon to all mammalian brains, although he argued that its func-tions were not limited to olfaction (Fig. 2). After Broca’s publicationthe accumulation of experimental evidence from ablation studiesin animals broadened the role of the limbic structures to includeother aspects of behaviour such as controlling social interactionsand behaviour (Brown and Schäfer, 1888), consolidating memories(Bechterew, 1900), and forming emotions (Cannon, 1927).Anatomical and physiological advancements in the field ledChristfield Jakob (1906) (Fig. 3) and James Papez (1937) (Fig. 4)to formulate the first unified network model for linking action andperception to emotion. According to Papez emotion arises eitherfrom cognitive activity entering the circuit through the hippocam-pus or from visceral and somatic perceptions entering the circuitthrough the hypothalamus. In the case of emotion arising from cog-nitive activity, for example, ‘incitations of cortical origin would passfirst to the hippocampal formation and then down by way of the fornixto the mammillary body. From this they would pass upward throughthe mammillo-thalamic tract, or the fasciculus of Vicq d’Azyr, to theanterior nuclei of the thalamus and thence by the medial thalamocor-tical radiation [or anterior thalamic projections] to the cortex of thegyrus cinguli [. . .] The cortex of the cingular gyrus may be looked onas the receptive region for the experiencing of emotion as the result ofimpulses coming from the hypothalamic region [or the hippocampalformation][. . .] Radiation of the emotive process from the gyrus cingulito other regions in the cerebral cortex would add emotional colouringto psychic processes occurring elsewhere (Papez, 1937)’A decade later, Paul Yakovlev (1948), proposed that theorbitofrontal cortex, insula, amygdala, and anterior temporal lobeform a network underlying emotion and motivation (Fig. 5). In twoseminal papers published in 1949 and 1952, Paul. MacLean crys-tallised previous works by incorporating both Papez and Yakovlevview into a model of the limbic system that has remained almostunchanged since (MacLean, 1949, 1952). MacLean concluded thatthe limbic cortex, together with the limbic subcortical structures,is a functionally integrated system interconnected by short- andlong-range fibre bundles (Fig. 6).The development of tracing methods for studying long axonalpathways added details to the anatomical model of the limbicFig. 1. The limbic system described for the first time by Thomas Willis (1664) toindicate cortical regions located around the brainstem.system (Crosby et al., 1962). These methods allowed, for example,the description of long and short connections of the cingulatecortex in animals. Further, the combination of anatomical methodswith experimental procedures was used to demonstrate a directlink between specific limbic structures and behavioural response(e.g. amygdala and aggressive response). Unfortunately, axonaltracing could not be applied to human anatomy for the study ofthe biological underpinnings of those abilities that characterisehuman mind (e.g. emotions; empathy). Also animal models werenot suitable for studying anatomical differences in psychiatricconditions such as autism and schizophrenia.In the 1990s the use of functional neuroimaging methods (e.g.PET, fMRI) and later diffusion tractography offered the possibilityof studying the functional anatomy of the limbic system in the liv-ing human brain. A major finding that emerged initially from PETstudies and later confirmed with fMRI was the identification of a‘default network’, consisting of a set of regions that activate underresting-state condition and deactivate during task-related func-tions (Buckner et al., 2008; Raichle et al., 2001; Raichle and Snyder,2007; Shulman et al., 1997) (Fig. 7). The most medial regionsof the default network correspond to the most dorsal portionof the Papez circuit and are interconnected through the dorsalcingulum.Diffusion imaging is an advanced MRI technique based on opti-mised pulse sequences, which permits the quantification of thediffusion characteristics of water molecules inside biological tis-sues (Le Bihan and Breton, 1985). Given that cerebral white mattercontains axons, and that water molecules diffuse more freely alongaxons than across them (Moseley et al., 1990), it is possible toobtain in vivo estimates of white matter fibre orientation by mea-suring the diffusivity of water molecules along different directions(Basser et al., 1994). By following the orientation of the waterFig. 2. Paul Broca (1878) identified the limbic system as mainly an olfactory structure of the mammalian brain.
PAGE – 3 ============
1726 M. Catani et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737Fig. 3. The limbic system as an integrated system of cortical and subcortical structures linked by projection and association tracts was described for the first time by ChristfriedJakob in 1906. Cg, cingulum; Tr, trigonum; C. Call, corpus callosum; N.A.T., anterior thalamic nuclei; Tal., thalamus; Az, bundle of Vicq d’Azyr; CM, mammillary bodies; H,hippocampus; U, uncus; Bo, olfactory bulb; SL, Septum pellucidum.molecules displacement, diffusion imaging tractography recon-structs 3D trajectories of white matter pathways closely resemblingtracts described in post-mortem animal tracing studies (Dauguetet al., 2007; Thiebaut de Schotten et al., 2011a, 2012) and humanbrain dissections (Basser et al., 2000; Catani et al., 2012a; Dell’Acquaand Catani, 2012; Dell’Acqua et al., 2010, 2012; Jones, 2008; Laweset al., 2008; Thiebaut de Schotten et al., 2011b; Forkel et al., 2012).One of the advantages of tractography is the ability to study theinterindividual variability of white matter tracts in the healthy pop-ulation and correlate white matter abnormalities with symptomsseverity in patients with neurological and psychiatric disordersthat involve the limbic system (Catani et al., 2012b; Catani et al.,2013a). In the following paragraphs we will use integrated infor-mation from animal studies and tractography findings in humanto describe in detail the anatomy of the main limbic pathways(Fig. 8).2. Anatomy of the limbic system2.1. FornixThe fornix is mainly a projection tract connecting the hip-pocampus with the mammillary body, the anterior thalamicnuclei, and the hypothalamus; it also has a small commissuralcomponent known as the hippocampal commissure (Aggleton,2008; Crosby et al., 1962; Nieuwenhuys et al., 2008). Fibres arisefrom the hippocampus (subiculum and entorhinal cortex) of eachside, run through the fimbria, and join beneath the splenium ofthe corpus callosum to form the body of the fornix. Other fimbrialfibres continue medially, cross the midline, and project to thecontralateral hippocampus (hippocampal commissure). Most ofthe fibres within the body of the fornix run anteriorly beneath thebody of the corpus callosum towards the anterior commissure.Above the interventricular foramen, the anterior body of the fornixdivides into right and left columns. As each column approachesthe anterior commissure it diverges again into two components.One of these, the posterior columns of the fornix, curve ventrallyin front of the interventricular foramen of Monroe and posteriorto the anterior commissure to enter the mammillary body (post-commissural fornix), adjacent areas of the hypothalamus, andanterior thalamic nucleus. The second component, the anteriorcolumns of the fornix, enter the hypothalamus and project to theseptal region and nucleus accumbens (Aggleton, 2008). The fornixalso contains some afferent fibres to the hippocampus from septaland hypothalamic nuclei (Nieuwenhuys et al., 2008).2.2. Mammillo-thalamic tractThe fibres of the mammillo-thalamic tract (bundle of Vicqd’Azyr) originate from the mammillary bodies and after a very shortFig. 4. The limbic system according to James Papez (Papez, 1937) is an exact duplicate of Jakob’s original drawing. Papez never quoted the work of Jakob and it is possiblethat he didn’t know about his work, which was published in an Argentinean journal with scarce international diffusion (La Semana Médica). Nevertheless the similaritiesbetween the two models are striking. To give credit to the work of Jakob we suggest the use of the eponym Jakob-Papez circuit. a, anterior nucleus; cc, corpus callosum; cn,caudate nucleus; cp, cingulum posterior; d, gyrus dentatus; f, fornix; gc, gyrus cinguli; gh, gyrus hippocampi; gs, gyrus subcallosus; h, hippocampus; m, mammillary body; mt,mammillo-thalamic tract; p, pars optica hypothalami; pr, piriform area; sb, subcallosal bundle; t, tuber cinereum; td, tractus mammillo-tegmentalis; th, tractus hypophyseus;u, uncus.
PAGE – 4 ============
M. Catani et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737 1727Fig. 5. Yakovlev’s amygdala-orbitofrontal network (Yakovlev and Locke, 1961; Yakovlev, 1948). AF, supralimbic corticocortical afferents to the limbic cortex; AM, anteriormedial nucleus of the limbic thalamus; Area diag., area diagonalis (Filimonov); Area periamgd., area periamygdalaris (Filimonov); Area entorhin., area entorhinalis; AV, anteriorventral nucleus of the limbic thalamus; cng-unc-om, orbitomesial interdigitation of the cingulum and uncinate bundle; cng-unc-tm, temporomesial interdigitation of thecingulum and uncinate bundle; cpf, callosoperforant fibres from the supralimbic cortex; Fascia Dent., fascia dentate of the Ammon’s horn; Fis. Hippoc., parahippocampal fissure;h, hippocampal efferents via fornix brevis; Hb, habenular nuclei; Lam. Zon., lamina zonalis; LD, lateral dorsal nucleus of the limbic thalamus; m, afferent and efferent fibres ofthe median thalamus (midline nuclei); Massa intrmd., massa intermedia; Pf, corticoperforant temporoammonic fibres (direct and crossed); pf, corticoperforant (direct andcrossed) fibres of the cingulum to the hippocampus; pm, afferent and efferent fibres of the paramedianthalamus (limbic nuclei); sbc, subcallosal radiations (lateral, ventraland medial) of the cingulum to the septum and ipsi- and contra-lateral striatum and limbic thalamus; Sept. Ar., septal area; T.TH., taenia thalami; VA, ventral anterior nucleusof the limbic thalamus.course terminate in the anterior and dorsal nuclei of the thalamus.A ventrally directed branch projects from the mammillary bodiesto the tegmental nuclei (mammillo-tegmental tract). According toNauta (1958), the mammillo-tegmental tract, together with otherfibres of the medial forebrain bundle, forms an important circuitbetween medial limbic structures of the midbrain and hypothala-mus to relate visceral perception to emotion and behaviour.2.3. Anterior thalamic projectionsThe anterior thalamic nuclei receive projections from the fornixand mammillo-thalamic tract and connect through the anteriorthalamic projections to the orbitofrontal and anterior cingulate cor-tex. The anterior thalamic projections run in the anterior limb of theinternal capsule.2.4. CingulumThe cingulum contains fibres of different lengths, the longestrunning from the amygdala, uncus, and parahippocampal gyrusto sub-genual areas of the frontal lobe (Crosby et al., 1962;Nieuwenhuys et al., 2008). From the medial temporal lobe, thesefibres reach the occipital lobe and arch almost 180 degrees aroundthe splenium to continue anteriorly within the white matter of thecingulate gyrus. The dorsal and anterior fibres of the cingulum fol-low the shape of the superior aspect of the corpus callosum. Aftercurving around the genu of the corpus callosum, the fibres termi-nate in the subcallosal gyrus and the paraolfactory area (Crosbyet al., 1962). Shorter fibres that join and leave the cingulum alongits length, connect adjacent areas of the medial frontal gyrus,paracentral lobule, precuneus, cuneus, cingulate, lingual, andfusiform gyri (Dejerine, 1895; Nieuwenhuys et al., 2008). The cin-gulum can be divided into an anterior-dorsal component, whichconstitutes most of the white matter of the cingulate gyrus, and aposterior-ventral component running within the parahippocampalgyrus, retrosplenial cingulate gyrus, and posterior precuneus. Pre-liminary data suggest that these subcomponents of the cingulummay have different anatomical features. For example, a higherfractional anisotropy has been found in the left anterior-dorsalsegment of the cingulum compared to right, but reduced fractionalanisotropy has been reported in the left posterior-ventral compo-nent compared to the right (Gong et al., 2005; Park et al., 2004;Fig. 6. MacLean’s 1952 (MacLean, 1949, 1952) proposal for a unitary model of the limbic system consisting of Papez circuit and Yakovlev’s amygdala-orbitofrontal network.
PAGE – 5 ============
1728 M. Catani et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737Fig. 7. One of the first PET studies looking at the functional anatomy of the ‘default networks’ (Shulman et al., 1997). Area 1 corresponds to the posterior cingulatecortex/precuneus and area 9 to the anterior cingulate/medial frontal cortex. Theses two areas are interconnected through the dorsal fibres of the cingulum.Wakana et al., 2007). Notwithstanding this, the volume of thecingulum is bilateral and symmetrical in most subjects (Thiebautde Schotten et al., 2011b).2.5. UncinateThe uncinate fasciculus connects the anterior part of the tempo-ral lobe with the orbital and polar frontal cortex (Fig. 8). The fibresof the uncinate fasciculus originate from the temporal pole, uncus,parahippocampal gyrus, and amygdala, then after a U-turn, enterthe floor of the extreme capsule. Between the insula and the puta-men, the uncinate fasciculus runs inferior to the fronto-occipitalfasciculus before entering the orbital region of the frontal lobe.Here, the uncinate splits into a ventro-lateral branch, which termi-nates in the anterior insula and lateral orbitofrontal cortex, and anantero-medial branch that continues towards the cingulate gyrusand the frontal pole (Crosby et al., 1962; Dejerine, 1895; Klinglerand Gloor, 1960). Whether the uncinate fasciculus is a lateralisedbundle is still debated. An asymmetry of the volume and density offibres of this fasciculus has been reported in a human post-mortemneurohistological study in which the uncinate fasciculus was foundto be asymmetric in 80% of subjects, containing on average 30%more fibres in the right hemisphere compared to the left (Highleyet al., 2002). However, diffusion measurements have shown higherFig. 8. Diagrammatic representation of the limbic system and tractography reconstruction of its main pathways. The colours in both figures correspond to the tracts in thelegend.
PAGE – 6 ============
M. Catani et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737 1729fractional anisotropy in the left uncinate compared to the right inchildren and adolescents (Eluvathingal et al., 2007) but not in adults(Thiebaut de Schotten et al., 2011b).3. Functional anatomy of the limbic systemThe limbic system has always been considered as a complexarrangement of transitional structures situated between a vis-ceral ‘primitive’ subcortical brain and a more evolved cortical one(MacLean, 1952; Yakovlev, 1948). The subcortical limbic struc-tures include the amygdala, mammillary bodies, hypothalamus,some thalamic nuclei (i.e. anterior, intralaminar, and medial dorsalgroups) and the ventral striatum (i.e. nucleus accumbens). The neu-rons and fibres composing the subcortical limbic structures presenta simple arrangement, not dissimilar to other subcortical nucleiof the brainstem that regulate basic metabolism, respiration, andcirculation.The cortical components of the limbic system include areas ofincreasing complexity separated into limbic and paralimbic zones(Mesulam, 2000). At the lower level the corticoid areas of the amyg-daloid complex, substantia innominata, together with septal andolfactory nuclei display an anatomical organisation that lacks con-sistent lamination and dendritic orientation. These structures arein part subcortical and in part situated on the ventral and medialsurfaces of the cerebral hemispheres. The next level of organisationis the allocortex of the olfactory regions and hippocampal complex,where the neurons are well differentiated into layers and their den-drites show an orderly pattern of orientation. The corticoid andallocortical regions are grouped together into the limbic zone ofthe cerebral cortex as distinct from the paralimbic zone. The lat-ter is mainly composed of ‘mesocortex’, whose progressive level ofstructural complexity ranges from a simplified arrangement similarto the allocortex, to the most complex six-layered isocortex.The limbic and paralimbic zones can also be divided into olfac-tocentric and hippocampocentric groups (Fig. 9) (Mega et al., 1997;Mesulam, 2000). Each division is organised around a central coreof allocortex. The olfactocentric division is organised around theprimary olfactory piriform cortex and includes the orbitofrontal,insular and temporopolar region. The hippocampocentric divisionis organised around the hippocampus and includes the parahip-pocampal and cingulate cortex. Both divisions have reciprocalconnections with subcortical limbic structures and surroundingisocortical regions (Fig. 9). The two divisions overlap in the anteriorcingulate cortex.Functionally the paralimbic areas contribute to the activity ofthree distinct networks (Fig. 10). The first network, composed of thehippocampal-diencephalic limbic circuit (connected through thefornix and mammillo-thalamic tract) and the parahippocampal-retrosplenial circuit (ventral cingulum), is dedicated to memoryand spatial orientation, respectively (Aggleton, 2008; Vann et al.,2009). Some structures of this network are particularly vulnerableto damage caused by viral infections (e.g. encephalitis) or alcohol(e.g. Korsakoff’s syndrome) (Fig. 10). Imaging studies have docu-mented altered metabolism and reduced functional activation ofthis network also in age-related neurodegenerative disorders suchas mild cognitive impairment (Minoshima et al., 1997; Nestor et al.,2003) and Alzheimer’s disease (Buckner et al., 2005).The temporo-amygdala-orbitofrontal network (connectedthrough the uncinate fasciculus) is dedicated to the integrationof visceral and emotional states with cognition and behaviour(Mesulam, 2000). In animal studies, disconnection of the unci-nate fasciculus causes impairment of object-reward associationlearning and reduced performances in memory tasks involvingtemporally complex visual information (Gaffan and Wilson, 2008).In humans this network engages in tasks that involve naming,single word comprehension, response inhibition, face processingand monitoring of outcomes (Catani et al., 2013a; Amodio andFrith, 2006). Damage to this network manifests with cognitive andbehavioural symptoms characteristic of temporal lobe epilepsy,mood disorders, traumatic brain injury, psychopathy and neurode-generative dementias, including advanced Alzheimer’s disease andsemantic dementia (Fig. 10).The dorsomedial default-mode network consists of a groupof medial regions whose activity decreases during goal-directedFig. 9. Functional-anatomical separation of the limbic system into olfactocentric (blue) and hippocampocentric (red) divisions. Some regions (e.g. BA 10, 11, 21, 22, 24, 32,36, 47) are connected to both divisions.
PAGE – 8 ============
M. Catani et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1724–1737 1731to the mammillo-thalamic tract is the best predictor of the severityof the memory deficit (von Cramon et al., 1985). In patients with col-loid cysts of the third ventricle, the surgical removal of the benigntumour can damage the fornix and result in anterograde amnesia,although it is seemingly not as severe as that seen in diencephalicpatients (Aggleton, 2008).Another form of hippocampocentric memory dysfunction isassociated with lesions to the posterior parahippocampal cortex,retrosplenial cingulate cortex, and posterior precuneus (Valensteinet al., 1987). These patients, in addition to memory deficits, showdifficulties in spatial orientation due to the inability to derivedirectional information from landmark cues in familiar and newenvironments (Vann et al., 2009). Reduced metabolism of the ret-rosplenial cortex has also been reported in patients with mildcognitive impairment (Nestor et al., 2003) and early Alzheimer’sdisease (Minoshima et al., 1997). More recently, a combined cor-tical morphometry and diffusion imaging study found reducedcortical thickness and white matter abnormalities of these regions(Acosta-Cabronero et al., 2010). Compared to surrounding areas,the parahippocampal, posterior cingulate, and precuneus regionsalso have a faster rate of atrophy in pre-symptomatic Alzheimer’sdisease patients (autosomal dominant mutation carriers) (Scahillet al., 2002). Reduced fractional anisotropy has also been found inthe cingulum, hippocampus and the posterior corpus callosum ofcognitively intact subjects with increased genetic risk of dementia(APOE 4 carriers) (Persson et al., 2006).Preliminary evidence suggests that diffusion changes in neu-rodegenerative disorders are likely to reflect severity of underlyingwhite matter pathology. Xie et al. (2005) reported a significantpositive correlation between reduced fractional anisotropy val-ues, atrophy of the hippocampus and decline in the mini-mentalstate examination scores in patients with Alzheimer’s disease. Ina transgenic mouse model over-expressing beta-amyloid precur-sor protein, the diffusivity parameters were significantly correlatedwith the severity of Alzheimer’s disease-like pathology in the whitematter (Song et al., 2004). In humans, Englund et al. (2004) con-ducted a parallel post-mortem neuropathological examination andfractional anisotropy quantification of two brains with dementiaand reported that the degree of white matter pathology correlatedsignificantly with gradually lower fractional anisotropy valuessampled in fifteen regions of interest. Overall, these studies sug-gest that reduced fractional anisotropy in Alzheimer’s disease mayreflect white matter axonal degeneration and myelin loss followingneuronal degeneration of cortical neurons.Damage to the limbic white matter tracts such as the fornix(Concha et al., 2005) and the uncinate fasciculus (Diehl et al.,2008) is also reported in patients with unilateral temporal lobeepilepsy. This damage is diffuse and often extends contralaterallyfrom the side of the suspected seizure. In temporal lobe epilepsypatients with mesial hippocampal sclerosis, the decreased frac-tional anisotropy of the fornix and the associated memory deficitsare correlated with reduced axonal diameter and myelin contentof the fornix fibres (Concha et al., 2010). The diffusion changes inthe left uncinate fasciculus also correlate with the severity of thedeficits in delayed recall (Diehl et al., 2008). It is noteworthy tospecify that pre-operative tractography assessment of the lateral-isation pattern of the temporal tracts can help to predict namingdeficits after the operation in patients with temporal lobe epilepsyundergoing surgery (more left lateralised patients showed worsepostoperative deficits) (Powell et al., 2008).4.2. Temporal-amygdala-orbitofrontal syndromesThe clinical profile of neurodegenerative disorders variesaccording to the network affected by the illness. In advancedAlzheimer’s disease, for example, the extension of the disease to theolfactory (orbitofrontal-amygdala) division is associated with clin-ical manifestations such as semantic deficits, language difficulties,personality changes and other behavioural symptoms (e.g. aggres-sion, disinhibition, etc.), which are not present if the pathology islimited to the hippocampocentric division. Alternatively, the earlystages of the temporal variant of the fronto-temporal dementiaand the semantic variant of primary progressive aphasia (Agostaet al., 2010; Borroni et al., 2007; Catani et al., 2013b) involve theolfactory division first leading to olfactory-gustatory-visceral dys-functions and semantic deficits. As the disease progresses, damageto the hippocampocentric division occur and involve other cogni-tive domains such as memory and spatial orientation.In some patients with temporal lobe epilepsy the behaviouralsymptoms resemble those commonly observed in the Klüver–Bucysyndrome. In the 1930s, Klüver and Bucy conducted a series ofexperiments in rhesus monkeys that consisted of bilateral surgicalremoval of the anterior temporal lobe, which include the amygdalaand temporal pole (Klüver and Bucy, 1939). After the operationthe animals showed a strong tendency to examine objects orally(hyperorality), an irresistible impulse to touch (hypermetamor-phosis), loss of normal anger and fear responses, increased sexualactivity, and inability to recognise visually presented objects. Thesystematic experimental series conducted by Klüver and Bucy,although originally described by Brown and Schäfer in 1888, helpedto understand the functions of the anterior temporal lobe andbehavioural deficits associated with limbic damage in humans. Thefirst Klüver–Bucy syndrome in humans was described in a patientwho received bilateral temporal resection (Terzian and Ore, 1955).In recent times, this is a condition that clinicians observe in patientswith herpes or paraneoplastic encephalitis, tumours, or traumaticbrain injury involving the anterior temporal and orbitofrontal cor-tex (Hayman et al., 1998; Zappala et al., 2012).In children with temporal lobe epilepsy, single-photon emissioncomputed tomography reveals hypoperfusion of the basal gan-glia and the adjacent frontal and temporal limbic regions. Mostof the patients recover after the acute phase, but those withabnormal diffusivity of the temporal and frontal white mattertracts exhibit long-term mental retardation, epilepsy, and persis-tent oral tendency (Maruyama et al., 2009). Some temporal lobeepilepsy patients present with Geschwind’s syndrome, a char-acteristic change in personality consisting of unusual tendenciesto write extensively and in a meticulous manner (hypergraphia),excessive and circumstantial verbal output, deepened cognitiveand emotional responses (e.g. excessive moral concerns), viscos-ity of thought, altered sexuality (usually lack of interest), andhyperreligiosity (Waxman and Geschwind, 1974). The emergenceof psychotic symptoms in temporal lobe epilepsy is associatedwith white matter changes extending to the frontal pathways(Flugel et al., 2006). Behavioural symptoms in epileptic patientscan respond to surgery. Mitchell et al. (1954) described a case oftemporal lobe epilepsy with fetish behaviour. The patient reportedhighly pleasurable ‘thought satisfaction’ derived from looking at asafety-pin and sought seclusion in a lavatory to indulge it. Unfor-tunately, the fetish object also triggered severe seizures, whichrequired surgical treatment. Relief not only of the epilepsy but alsoof the fetishism followed the temporal lobectomy.Psychopathic personality disorder (psychopathy) is charac-terised by features of emotional detachment and antisocial traits(Patrick et al., 1993), and is strongly associated with criminalbehaviour and recidivism (Hare et al., 1999). Approximately 0.75%of male population meets the criteria for psychopathy with enor-mous costs for the society (Blair et al., 2005). It has been estimated,for example, that 15% of the prison population are psychopaths andthey commit approximately 50% more criminal offences than non-psychopathic criminals (Hart and Hare, 1997). Since the report ofthe case of Phineas Gage (Harlow, 1848), who displayed ‘acquired
132 KB – 14 Pages