Until recently, the STRIATUM was seen rather as a funnel for widespread neocortical afferents, being a smaller structure than the cortical mantle projecting to it. The striatum appeared to funnel its outflow through the even smaller GLOBUS PALLIDUS to the smaller still motor nuclei of the THALAMUS (ventroanterior and ventrolateral nuclei [VA/ VL]) and then to more restricted areas of the FRONTAL CORTEX, including the SUPPLEMENTARY MOTOR AREA. Its functions were seen as predominantly motor and it comprised the major component of what was known as the EXTRAPYRAMIDAL MOTOR SYSTEM. Neuroanatomical data have dramatically altered this view. The fundamental principle of striatal circuitry, namely cortex-to-striatum-to-pallidum-to-thalamus-to-cortex is still appropriate, although striatopallidal outflow also influences a variety of BRAINSTEM structures, such as the PEDUNCULOPONTINE TEGMENTAL NUCLEUS. However, rather than appearing as a funnel, it is now accepted that there exist a series of more or less segregated cortico-striato-pallidothalamo-cortical loops. The discrete cortical origins and topography of these loops indicates separable functions. For example, the following loops are generally agreed upon: motor loop, oculomotor loop, so-called cognitive loops (at least two) and affective loops (at least two).
Important general points in the organization of these loops are that: (1) the cortical origins are more widespread than the cortical areas of termination of the thalamic projection; (2) quite specific striatal, pallidal and thalamic domains are associated with particular cortical inputs for each loop—hence the notion of segregated loops. For example, the motor loop, which is best understood, originates in the sensory-motor cortex and anterior parietal cortex; it projects in a somatotopic way to the PUTAMEN which in turn projects to discrete portions of the internal pallidal segment/SUBSTANTIA NIGRA pars reticulata and also the external pallidal segment/SUBTHALAMIC NUCLEUS subsidiary loop, all of which retain a somatotopic organization. The outflow from the internal pallidal segment/substantia nigra pars reticulata reaches the ventral anterior and ventral lateral thalamus, again with somatotopicity preserved and this thalamic nucleus projects back primarily to the supplementary motor area and also to the primary SOMATOSENSORY CORTEX, which is of course itself somatotopically organized. The remarkable feature of this loop circuitry is its precise compartmentalization at each node, leading to the notion of segregated loops. The oculomotor loop involves the FRONTAL EYE FIELDS and supplementary frontal eye fields; the affective loops involve the anterior cingulate cortex and medial orbitofrontal cortex; the cognitive loops involve the dorsolateral and lateral orbital parts of the PREFRONTAL CORTEX.
These loops are primarily dependent upon projections to discrete domains of the dorsal striatum, especially the CAUDATE NUCLEUS which in turn, for each loop, projects to specific sub-regions of the globus pallidus and ventral anterior or medial dorsal thalamus. A similar organization has been demonstrated for the ventral striatum, where the cortical origins involve the basolateral AMYGDALA, HIPPOCAMPUS, ANTERIOR CINGULATE CORTEX and PRELIMBIC CORTEX, the striatal node is mainly the NUCLEUS ACCUMBENS and the re-entrant cortical circuitry involves serial relays through the ventral pallidum and DORSOMEDIAL NUCLEUS OF THE THALAMUS, the latter projecting specifically to the prefrontal cortex.
Within this general scheme, there is also considerable fine detail in the organization of the striatum and its outputs and this has helped clarify the way that hypokinetic states arise following the loss of DOPAMINE from the striatum, arising either experimentally or in PARKINSON’S DISEASE. The outflow of the dorsal striatum is now known to be organized in the form of direct and indirect striatal pathways. The direct path consists of projections from the MEDIUM SPINY NEURONS in the striatum to the globus pallidus internal segment (GPi) and the substantia nigra pars reticulata (SNr): these are in essence separate parts of the same structure. They in turn project to the thalamus, SUPERIOR COLLICULUS and pedunculopontine tegmental nucleus. The indirect path consists of projections from striatal medium spiny neurons to the external segment of the globus pallidus (GPe), thence to the subthalamic nucleus (STN) and to the GPi/ SNr. Striatal and pallidal neurons are GABAERGIC (that is, they are inhibitory output neurons); the subthalamic nucleus neurons are GLUTAMATERGIC (and therefore excitatory). Thalamic neurons projecting to the cortex and also cortical projections to the striatum are both glutamatergic. It can be seen then that activation of the striatum via excitatory cortical afferents will increase activity in the direct path, so inhibiting GPi/SNr and disinhibiting the excitatory thalamic drive to the cortex. Increasing activity in the indirect path, on the other hand, will disinhibit the STN, increasing the excitatory drive to GPi/SNr, thereby increasing the inhibitory input to the thalamus and so opposing activity in the direct path. However, the NIGROSTRIATAL DOPAMINE SYSTEM powerfully modulates this relationship; dopamine release tends to increase the activity of the direct path, via dopamine D1 receptors (see D1–D5 DOPAMINE RECEPTORS) localized to this population of striatal neurons, but simultaneously also tends to decrease the activity of the indirect path via dopamine D2 receptors localized to this population of striatal neurons. This is the basis of the so-called ‘disinhibitory’ principle of communication through striatopallidal circuitry and the way that cortical drive to the striatum and its interaction with dopamine reinforces cortical activity subserving movement.
Clearly, it is inappropriate to regard the striatum simply as a motor structure. The striatum has cognitive and affective functions as well, which are dependent upon the specific organization of the segregated corticostriatal loops. While we understand a great deal about the functioning of the motor loop, we know as yet much less about the functions of the other corticostriatal loops, nor how these apparently segregated loops might interact with each other.
BARRY J.EVERITT
This is the complete article, containing 916 words
(approx. 3 pages at 300 words per page).
