You are not authenticated, login.
text: sort by
tags: modified
type: chronology
hide / / print
ref: Parent-1995.01 tags: basal ganglia anatomy review STN GPe DBS date: 02-22-2012 15:48 gmt revision:17 [16] [15] [14] [13] [12] [11] [head]

PMID-7711769[0] Functional anatomy of the basal ganglia. I. The cortico-basal ganglia-thalamo-cortical loop.

  • Pallidal and nigral neurons have wide dendritic arborizations at right angles to the unbranched incoming striatal axons, leading to (hypothetically) a confulence of information from distinct functional striatal territories on many neurons and to extreme reception convergence [242].
    • This pattern suggests that projections arising from very small areas of the cortex may extend through very large regions of the striatum, particularly along the rostrocaudal plane.
    • Individual striatal neurons receive relatively few synapses from restricted cortical areas; this makes it difficult to conceive how the cortico-striatal projection system could convey information in a highly specific manner; specificity does not exist at a cellular level.
  • Cortex to striatum:
    • Virtually all cortical functional areas contribute, at varying degrees, to the cortico-striatal projection, inputs from the sensorimotor cortex being particularly extensive and those from the visual cortex much less so.
    • Cortico-sriatal projection originates from neurons located in both supragranular (layers I-III) and infragranular (V,VI) cortical layers.
    • Cortical neurons project ipsilaterally or contralaterally, but not usually bilaterally.
    • Cortical cells arborize on restricted, topologically defined domains in the striatum.
    • Restricted cortical regions project to parasagitally elongated domains in the caudate nucleus.
      • this seems to be a general feature. see B and C below.
      • Reminds me of the cerebellum.
    • non-adjacent cortical areas (prefrontal and pareital cortices)project to adjacent striatal territories.
    • The association, sensorimotor, and limbic cortical areas project in a segregated manner onto threes distinct striatal regions referred to as the associative, sensorimotor, and limbic striatal territories.
    • In this view, cortical information is not directly transposed at striatal level, but is integrated and transformed into strict associative, sensorimotor, and limbic functional modalities.
  • Convergence and divergence:
    • There is a vast reduction in the number of neurons from the cortex to the striatum.
    • This has led many to infer overlap or convergence.
    • Actual projection is patchy -- divisions of striosomes and extrastriosomal matrix -- with the individual axons sending out further sub-patches.
      • This degree of segregation breaks down for sensorimotor territory.
    • cortico-striatal neurons in infragranular layers project principally to striosomes while those in supragranular layers send their axons to the matrix. things are tightly organized.
  • The output cells of the matrix are grouped in clusters in relation to the different projection systems that lead from the striatum to the GPe and GPi. These are called 'matrisomes'.
    • These might be a way of bringing into proximity different cortical signals so they can be recombined in novel ways.
    • That said, there was substantial topographical overlap of the frontal eye field and the supplementary eye field, and though these are closely interdigitated they do not mix.
  • Medium spiny neurons:
    • The primary projection neurons of the striatum.
    • GABA. Plus substance P, enkephalin, dynorphin and neurotensin. (!)
      • The coexistence of GABA with a given peptide in a spiny neuron is in correlation with it's target site.
      • At that time they didn't know what the peptides did.
    • Axon emits several collaterals:
      • Local axonal arborizations restricted tot he dendritic domain of its cell of origin or a nearby cell -- inluding an 'autonapse' or of nearby projection neurons.
      • Less common axonal arborization goes far beyond and often does not overlap the dendritic domain of the cell of origin.
    • Projected to by the cortex, thalamus, and the SNc.
    • Usually silent, except with cortical / thalamic input.
  • Interneurons in the striatum are non-spiny.
    • Less than 2% (of entire striatal population, not just interneurons) them are huge, cholinergic cells.
      • These form symmetric synapses on virtually all parts of MSN.
    • Medium, 1% of population, have short axons and are GABA ergic.
    • Second medium, nitrous oxide signaling interneurons.
    • SNc efferents synapes ontot the base of the spines, but only on MSN that have cortical afferents.
    • Thalamic input synapse onto morphologically distinct type of MSN.
    • Destruction of the dopaminergic nicgro-striatal pathway results in a decrease in levels of mRNA for substance P and increase in mRNA for enkephalin.
  • Striatal MSN projections:
    • Relatively discrete in cats and monkeys; highly collateralized in rats, where many neurons project to GPe, GPi, SN, or some pair.
  • Fibers from the associative territory massively invade the whole extent of SNr, without clear territorial demarcation.
    • Meanwhile, inputs from the limbic striatal territory appears to be widely distributed in the substantia nigra & VTA.
  • Most authors think that the distinction between the GPi and SNr is artificial -- they are split by the internal capsule.
    • However, GPi is mostly sensorimotor, while SNr is associative.
  • Projections from striatum to pallidus * SNr very organized and layered.
    • Pictures. read the paper. words do not do this justice.
    • For example, injections of anterograde tracers in various sectors of the striatum produce elongated, longitudinally oriented terminal fields that cover nearly the entire rostrocaudal extent of the substantia nigra.
    • "The dorsal climbing fibers and the corresponding wooly fibers from replicable modular units whose boundaries do not respect the limit between SNc and SNr compartments. ... They are distrinuted along the rostrocaudal extent of the substantia nigra according to a remarkably precise and constant sequence.
  • As in [1]: striatal and subthalamic terminals converge onto the same pallidal neurons within these regions of overlap, possibly in register with those from the striatum.
    • The striato-pallidal fibers and striato-nigral fibers arborize at least twice in the target structures, suggesting there are multiple copies of the same information to distinct subsets of pallidal/nigral populations.
      • Meanwhile, GPi/SNr axons are highly collateralized and not strictly confined to disctinct subnuclei.
      • That is, output is both convergent and divergent.
      • There are several multi-laminar models of the SNr [54] or the globus pallidus [243].
  • Regarding information funneling due to the very large dendritic fields of pallidal neurons:
    • anterograde double-labeling experiments in the squirrel monkey clearly indicate that neighboring striatal cell populations do not have overlapping terminal fields in the GP or SN.
      • Axons from adjacent striatal cell populations produce two sets of terminal fields that interdigitate but never mix.
      • cortical information is conveyed and integrated along multiple, segregated channels.
  • Output of GPi/SNr = VA, VL thalamus, both ipsi and contralateral.
    • Lesser: pedunculopontine tegmental nucleus & centromedian thalamus, superior colliculus.
    • Highly collateralized output.
    • Lamellar distribution of cells that share similar functional characteristics.
    • Synapse almost exclusively on thalamic projection neurons.
    • Centromedian nucleus: no projection to the cortex; rather projects to the striatum, hence is involved in regulation.
    • Pedunculopontine nucleus: mostly re-afferent back to the BG!
      • innervation of the SNc, subthalamic nucleus, and the pallidum. [95,149,186-188,202,207,215,263,277].
      • Acetylcholine output.
      • Deep cerebellar nuclei project to the pedunculopontine nuclei in primates.
  • GPe: efferent fibers from large terminal boutons that make synapses mostly of the symmetrical type with proximal dendrites and soma of GPi/SNr neurons. These GABA synapses may be of ultimate importance in regulating activity.
    • Also projects to the reticulothalamic region, which supplies GABA synapses to the rest of the thalamus, hence GPe can disinhibit most of the thalamus. Such complexity.


[0] Parent A, Hazrati LN, Functional anatomy of the basal ganglia. I. The cortico-basal ganglia-thalamo-cortical loop.Brain Res Brain Res Rev 20:1, 91-127 (1995 Jan)
[1] Parent A, Hazrati LN, Functional anatomy of the basal ganglia. II. The place of subthalamic nucleus and external pallidum in basal ganglia circuitry.Brain Res Brain Res Rev 20:1, 128-54 (1995 Jan)

hide / / print
ref: Parent-1995.01 tags: basal ganglia anatomy review STN DBS date: 02-22-2012 14:40 gmt revision:15 [14] [13] [12] [11] [10] [9] [head]

PMID-7711765[0] Functional anatomy of the basal ganglia. II. The place of subthalamic nucleus and external pallidum in basal ganglia circuitry.

  • 5 'sideways control structures' :
    • subthalamic nucleus (glutamate) STN
    • pars compacta of the substantia nigra (dopamine) SNpc
    • centromedian / parafasicular thalamic complex (glutamate) CM/Pf
    • dorsal raphe nucleus (serotonin)
    • pedunculopontine tegmental nucleus. (glutamate and acetylcholine) PPN
  • STN exitatory on the GPi and SNr. Which are basically the same thing.
  • Largest target is the GPe, to which it is reciprocally connected.
  • STN lesions produce ballism, violent, involuntary, wild, flinging movements usually limited to the side of the body contralateral to the lesion. Symptoms gradually resolve.
  • STN densely packed with soma, dendrites, and long axons.
    • But no (or few) interneurons.
  • Projects to:
    • GPe & GPi, SN, striatum, cerebral cortex, substantia innominata, pedunculopontine tegmental nucleus and the mesencephalic and pontine reticular formation.
    • These projections are topologically organized. Lateral -> dorsal pallidium, medial -> ventral pallidium (GPv).
    • Projections are often collaterals to GPe, GPi, and SNr in rodents; in primates, subsytems are separate.
    • Dorsolateral STN = sensorimotor, ventromedial = 'association'
  • STN projections lie parallel to GP neurons, arranged in lamina along the rostral-caudal axis.
    • These, like in the striatum, are arranged perpendicular to the afferent fibers.
    • Subthalamic and striatal neurons converge upon the same pallidal neurons.
    • "Subthalamic axons arborize throughout large caudorostral portions of the pallidum and appear to influence in a rather uniform manner large subpopulations of pallidal neurons in both pallidal segments."
  • Above: gray cells = pallidal neurons.
    • Suggests that STN cells can excite a rather large / diffuse population of pallidal cells, whereas striatum exerts a more specific inhibitory action.
  • STN neurons project somewhat diffusely and less topographically to SNr, with 'patchy' regions, very similar to other striatal-nigral projections.
    • Still, 90% of synapses in SN are GABA-ergic, < 10% are glutamatergic, so afferents from STN is not too large.
  • electrophysiological studies in the rat have suggested that efferent projections of the subthalamic nucleus control the inhibition of movement by setting the physiological conditions of pallidal and nigral neurons to the appropriate level prior the arrival of striatal signals.
  • STN projection to striatum diffuse, weak, unbranched and 'en passant'.
  • Afferent projections:
    • direct projection from the cerebral cortex. Might be collaterals from the pyramidal tract.
      • In rodents: 40% from the prefrontal cortex, 15% from the ACC, 9% M1.
    • In primates: Mostly M1, somatotopic organization (page 9), monosynaptic.
      • also S1, somatotopic, respond to sensory stimuli.
      • Dorsolateral sector of the subthalamic nucleus appears to be more involved in skeletomotor behavior, whereas the ventromedial sector appears more concerned with occulomotor and associative aspects of behavior [107].
  • Electrical stimulation of the cortex results in the STN a short-latency EPSP (monosynaptic) followed by brief inhibition IPSP (from the GP), then further EPSP.
  • Electrical stimulation of the STN does not elicit movements; stimulation within microzones of the striatum does.
  • more is known about the role of STN in eye movements through the SNr than skeletal motor control.
    • Venrtomedial sector of STN receives afferents from the frontal eye fields & supplementary eye fields.
    • SNr is known to exert a tonic GABAergic inhibition on neurons in the superior colliculus.
      • Inibition is suppressed by transient GABA inhibition originating from the caudate nucleus (disinhibition).
    • STN, in comparison, seems to suppress eye movements through the SNr -- perhaps to maintain attention on an object of interest, under control of the cortex (FEF). .
      • CF {169} : activation of the STN drives SNr activity, which inhibits the superior colliculus, allowing maintainance of eye position on an object of interest.
  • GPe projects directly to the STN, GABAergic, strong on proximal dendrites (less soma /distal),
    • Collaterals to both the STN and SNr, and to the greater striatum and entopeduncular nucleus.
    • Strong inhibitory effect on STN firing which appears to be chronic:
      • STN firing should only be elicited by strongly coherent or synchronized arrival of information from multiple extrinsic sources.
    • Recall there are two negations through the Striatum (GABA) & GPe (GABA).
  • The hypothesis behind Huntington's disease & PD:
    • PD: pallido-subthalamic pathway activity is decreased, leading to an increase in excitatory activity of STN on BG output structures -> greater GPi /SNr GABA ergic activity -> greater rigidity.
    • Huntingtons: pallido-subthalamic activity increased (striatal neurons lost), decreased excitation of STN -> less GPi/SNr GABAergic activity on VA/VL.
      • "leaving thalamocortical neurons to respond undiscriminatingly to all sorts of inputs and hence to hyperkinesia". Makes sense.
    • Above, classical direct and indirect pathway.
  • Re direct / indirect pathway: the evidence to support this is weak; inputs from the GPe seem to spare the area containing subthalamic cells projecting to the GPi/SNr.
    • Another way: pallidal control of the subthalamic nucleus in primates is exerted principally upon cells projecting back to the GPe and not upon cells projecting to GPi/SNr.
  • Only the centromedian / parafasicular complex of the thalamus projects to the STN. Important -- it is also an output structure of the BG.
    • These might be collaterals of the thalamo-striatal projection system.
    • Projections are topographic.
    • Respects boundaries: centromedian projects to sensorimotor laterodorsal STN; parafasicular nucleus innervates the associative / limbic portion of this structure. The associative projection is much stronger than the sensorimotor.
    • Glutamate.
  • Direct projections from the SNc; STN projects back to the SNr.
    • Dopamine, excitatory; much more present in rats than primates.
    • Marked increase in metabolism following dopamine agonist treatments.
    • Both D1 and D2 present (at least in rats).
  • Direct projections from the pedunculopontine tegmental nucleus to the STN.
    • Cholinergic.
    • Reciprocal -- relays BG information to the brainstem and spinal cord. Locomotion? cardiovascular changes?
  • Dorsal rahpe nucleus
    • Serotonin, obvi.
  • GPe:
    • Originally thought to project to STN to mediate it's glutamate projections
    • now realized to have many outputs, including to the GPi/SNr.
    • Strong afferents to the reticular thalamic nucleus (with bunched arborizations), GPi/SNr ('massive arborizations'), STN, and less to striatum.
    • Fibers from a small striatal cell group arborize twice in each pallidal segments in a rostrocaudal sequence manner.
    • GPe projections to GPI/SNr cell-to-cell.
      • These two together implies that the two striatal terminal fields in the GPe would effect two rostrally located sets of GPI/SNr cells 1 & 2 that are distinct from those innervated by the striatum more caudally than GPi/SNR cells 3 & 4 (above).
  • In animals at rest, striatal neurons are quiet, whereas SNr and GPi are tonically active.


[0] Parent A, Hazrati LN, Functional anatomy of the basal ganglia. II. The place of subthalamic nucleus and external pallidum in basal ganglia circuitry.Brain Res Brain Res Rev 20:1, 128-54 (1995 Jan)