Bb12j09.qxd

Current Biology, Vol. 12, 000–000, May, 2002, 2002 Elsevier Science Ltd. All rights reserved.
PII S0960-9822(02)000-X
Cerebellar Cortex: Computation by
Dispatch
Extrasynaptic Inhibition?
Erik De Schutter
GABA concentration in glomeruli can activate some
GABA receptors.
Several lines of evidence support the view that
In the cerebellar cortex, inhibitory inputs to granule
spillover transmission also makes a contribution. First,
cells exhibit prominent tonic and spillover compo-
two different kinds of synaptic current can be
nents resulting from the activation of extrasynaptic
recorded in a granule cell following Golgi cell stimula-
receptors. A recent study shows how extrasynaptic
tion: fast responses with time courses similar to spon-
inhibition affects information flow through cerebel-
taneous events, and much slower responses with a
lar cortex.
slow decay [4]. Second, granule cells in the cerebel-
lum and cochlear nucleus are the only cells that
express the α subunit of the GABA receptor [5].
The cerebellar cortex contains more granule cells than
Receptors containing the α δ subunit combination
there are neurons in the rest of the brain, but compar-
have a 50-fold higher affinity for GABA than other
atively little is known about their function compared to
GABA receptors, and they do not desensitize upon
the more striking Purkinje cells. Cerebellar granule
prolonged presence of agonist [6]. The δ subunit is
cells are, however, a popular preparation for pharma-
found exclusively in extrasynaptic locations on the
cological studies of receptors and channels. A long
dendrites and somata of granule cells [7]. The tonic
series of studies of granule cell receptors for the
and evoked slow currents recorded from granule cells
inhibitory neurotransmitter γ-amino butyric acid
both have the pharmacological profile of a receptor
(GABA) has now led to some fundamental insights into
containing α and δ subunits, being furosemide-
how a characteristic anatomical specialization of the
sensitive but diazepam- and neurosteroid-insensitive
cerebellar cortex, the glomerulus, may work [1].
[1, 4]. Taken together, these observations strongly
Glomeruli are formed around the large axonal termi-
suggest that the slow evoked currents reflect the
nals of glutamatergic mossy fiber afferents (Figure 1).
activation of extrasynaptic α β
δ receptors by GABA
Each terminal is contacted by dendrites from 50–60
molecules that have ‘spilled over’ from activated
distinct granule cells. The glomeruli also contain
synapses on other granule cells in the same glomeru-
GABAergic synapses that inhibitory Golgi cells make
lus. The diffusion boundaries caused by the glomeru-
with the granule cells, and glutamatergic contacts
lar sheath may further promote extrasynaptic
between the mossy fibers and Golgi cells. The struc-
interaction between granule cells.
ture has a radius of about 2.5 µm and is enwrapped by
It was always assumed that the extrasynaptic
glial sheathing. Granule cells have one to eight den-
activation of GABA receptors might have an impor-
drites, each participating in a different glomerulus.
tant role in cerebellar processing, but until recently
The first indication that GABAergic inhibition of
there has been little experimental evidence for such a
granule cells has unusual properties came from the
notion. In mature animals, where most of the sponta-
discovery that it has a strong tonic component [2,3].
neous GABA
current is tonic, blocking all GABA
This tonic GABA current is much larger than that
receptors with bicuculline leads to increased
evoked by spontaneous events and can be completely
responses of granule cells to current injection [3].
blocked by the GABA
receptor antagonist bicu-
Much progress has now been made by using
culline; its fraction of the total GABA current increases
furosemide to specifically block GABA receptors that
from 5% in young to 99% in mature rats [3]. An attrac-
have an α subunit in cerebellar slices from adult
tive hypothesis is that this phenomenon is caused by
animals [1]. Using this approach, it is estimated that
spillover of GABA molecules between neighboring
97% of the charge evoked during Golgi cell stimula-
Golgi-to-granule cell synapses of the same glomeru-
tion flows through extrasynaptic GABA
receptors!
lus. The initiating event would still be action potential-
This fraction includes the tonic current, which
evoked GABA release, but because the release
accounts for 75% of the charge transfer. These
happens at relatively distant synapses, diffusion and
measurements were, however, made at 29° and tonic
the summation of multiple events will result in delays
inhibition may make a much smaller contribution at
which will act to filter out the synaptic transients. If the
body temperature [8]. Specifically blocking extrasy-
tonic current is indeed caused by GABA spillover, then
naptic GABA
receptors with furosemide causes a
one would expected it to be blocked by tetrodotoxin
leftward shift of the firing curve of granule cells, as
or low external calcium. These manipulations partially
expected for the removal of a shunting inhibition [9],
block the tonic current in young animals and not at all
but it has no effect on the excitability of other neurons
in adult ones [1–3], so it is assumed that the ambient
in the cerebellar cortex.
But how will the altered excitability of granule cells
Theoretical Neurobiology, Born-Bunge Foundation, University
that results from extrasynaptic inhibition affect infor-
of Antwerp, Universiteitsplein 1, B2610 Antwerp, Belgium.
mation transfer in the cerebellar cortex? This is not
E-mail: erik@bbf.uia.ac.be
easy to predict as additional properties of the circuitry
Dispatch
R2
need to be taken into account. An increase in granule
Golgi cell afterhyperpolarization [13], unless one can
cell activity resulting from a blockade of extrasynaptic
assume mechanisms which prolong the effect of each
inhibition would also enhance Golgi cell activity
Golgi cell spike. While the synaptic GABA channels
through excitatory contacts made by parallel fibers
already have relatively slow kinetics, the much slower
[10], which may lead to increased synaptic GABA
spillover mechanism seems well suited to having a
release. Besides its inhibitory effect on granule cells,
function in gain control. As a result of spillover,
this may also activate GABA
receptors on mossy
increased Golgi cell activity — through either direct
fibers. It has been shown that such receptors are
mossy fiber activation or indirect parallel fiber activa-
activated by GABA spillover and reduce evoked
tion — will reduce excitability of all granule cells par-
mossy fiber responses in granule cells at low stimula-
ticipating in a glomerulus. This would provide for
tion frequencies [8]. Multiple effects on Purkinje cells
rather a slow gain control mechanism, which may not
are possible, as they are both directly activated by
prevent fast swings in granule cell activity [13].
increased parallel fiber activity and inhibited by
But how does the stronger tetrodotoxin-resistant
stellate/basket cells which also receive parallel fiber
tonic inhibition fit into this picture? It could provide a
constant baseline, raising the threshold for spike
In their recent study, Hamann et al. [1] measured
initiation in granule cells. But if this is desired, it seems
the effect of blocking extrasynaptic GABA receptors
more straightforward to reduce the intrinsic excitabil-
with furosemide on the input and output elements of
ity of granule cells, as observed in transgenic mice
the pathway. They found that the resulting enhanced
that lack the GABA receptor α and δ subunits and
excitability of the granule cells increases the number
so have no tonic current [18]. A more attractive idea is
of spikes they fire in response to mossy fiber stimula-
that tonic inhibition itself is regulated somehow. This
tion by about 100%. Purkinje cells also increase their
could occur at the receptor side, for example by phos-
firing frequency in response to mossy fiber stimulation
phorylation depending on β subunit expression [19], or
by about 100%. These increases reflect an increase in
by changing the baseline GABA concentration in the
the size of evoked excitatory postsynaptic potentials
glomerulus. Reduced tonic inhibition may explain the
(EPSPs), corresponding to a larger number of co-acti-
evidence for spike-mediated inhibition in vivo [13,17],
vated parallel fiber synapses. In conclusion, blocking
which is difficult to reconcile with the in vitro observa-
extrasynaptic inhibition increases the flow of neural
tion that only 3% of the GABA current is spike medi-
activity through cerebellar cortex — and conversely,
ated [1]. If tonic inhibition is regulated, it could have
the tonic inhibition normally present reduces this flow.
additional effects beyond controlling information flow
What may be the functional impact of the reduction
through the cerebellum. Long-term potentiation (LTP)
in activity transmission caused by extrasynaptic inhi-
of mossy fiber-to-granule cell synapses can only be
bition? Hamann et al. [1] refer to David Marr’s [11]
reliably evoked in vitro when inhibition is blocked [20].
seminal work on cerebellar motor learning to suggest
Any mechanism that reduces tonic inhibition will
that decreasing the number of granule cells activated
enhance LTP of this synapse. If it turns out that
by mossy fiber input increases the storage capacity of
extrasynaptic inhibition can be regulated separately in
the cerebellum. This is actually a simplification of what
each individual glomerulus, there could be very inter-
Marr really wrote — that Golgi cell inhibition should
esting implications for the possible computational
keep “the numbers of active parallel fibres . reason-
functions of the glomerulus.
ably small over quite large variation in the number of
active mossy fibres” [11]. In other words, there should
References
Hamann, M., Rossi, D.J. and Attwell, D. (2002). Tonic and spillover
be a dynamic component to the inhibition of granule
inhibition of granule cells control information flow through cerebel-
cell activity, stronger when many mossy fibers are
lar cortex. Neuron 33, 625–633.
active and weaker when few are firing. This was called
Wall, M.J. and Usowicz, M.M. (1997). Development of action poten-
‘automatic gain control’ by Albus [12]. The anatomy
tial-dependent and independent spontaneous GABA
receptor-
mediated currents in granule cells of postnatal rat cerebellum. Eur.
seems to favor such a role for Golgi cells, as the com-
J. Neurosci. 9, 533–548.
bined direct mossy fiber and indirect parallel excita-
Brickley, S.G., Cull-Candy, S.G. and Farrant, M. (1996). Develop-
tion makes them sensitive both to input to the granular
ment of a tonic form of synaptic inhibition in rat cerebellar granule
cells resulting from persistent activation of GABA receptors. J.
layer and the resulting activity of granule cells.
Physiol. 497, 753–759.
But at the physiological level, it is more difficult to
Rossi, D.J. and Hamann, M. (1998). Spillover-mediated transmission
reconcile the properties of Golgi cells with a gain
at inhibitory synapses promoted by high affinity alpha6 subunit
GABA receptors and glomerular geometry. Neuron 20, 783–795.
control function [13]. Recordings in vivo show that
Pirker, S., Schwarzer, C., Wieselthaler, A., Sieghart, W. and Sperk,
Golgi cells typically fire a few accurately timed spikes
G. (2000). GABA(A) receptors: immunocytochemical distribution of
in response to natural stimulation [10], followed by a
13 subunits in the adult rat brain. Neuroscience 101, 815–850.
long pause due to afterhyperpolarization [14]. More-
Saxena, N.C. and Macdonald, R.L. (1996). Properties of putative
cerebellar gamma-aminobutyric acid A receptor isoforms. Mol.
over, spike-evoked inhibition by Golgi cells in vivo is
Pharmacol. 49, 567–579.
strong enough to cause the theoretically predicted
Nusser, Z., Sieghart, W. and Somogyi, P. (1998). Segregation of dif-
[15] synchronization of Golgi cell activity along the
ferent GABA receptors to synaptic and extrasynaptic membranes
of cerebellar granule cells. J. Neurosci. 18, 1693–1703.
parallel fiber beam [16,17]. The Golgi cell firing pattern
Mitchell, S.J. and Silver, R.A. (2000). GABA spillover from single
may not be suitable for gain control, as it would
inhibitory axons suppresses low-frequency excitatory transmission
promote rebound firing by granule cells during the
at the cerebellar glomerulus. J. Neurosci. 20, 8651–8658.
Current Biology
R3
Holt, G.R. and Koch, C. (1997). Shunting inhibition does not have a
divisive effect on firing rates. Neural Comput. 9, 1001–1013.
Vos, B.P., Volny-Luraghi, A. and De Schutter, E. (1999). Cerebellar
Golgi cells in the rat: receptive fields and timing of responses to
facial stimulation. Eur. J. Neurosci. 11, 2621–2634.
Marr, D.A. (1969). A theory of cerebellar cortex. J. Physiol. 202,
437–470.
Albus, J.S. (1971). A theory of cerebellar function. Math. Biosci. 10,
25–61.
De Schutter, E., Vos, B.P. and Maex, R. (2000). The function of cere-
bellar Golgi cells revisited. Prog. Brain Res. 124, 81–93.
Dieudonné, S. (1998). Submillisecond kinetics and low efficacy of
parallel fibre-Golgi cell synaptic currents in the rat cerebellum. J.
Physiol. 510, 845–866.
Maex, R. and De Schutter, E. (1998). Synchronization of Golgi and
granule cell firing in a detailed network model of the cerebellar
granule cell layer. J. Neurophysiol. 80, 2521–2537.
Maex, R., Vos, B.P. and De Schutter, E. (2000). Weak common par-
allel fibre synapses explain the loose synchrony observed between
rat cerebellar Golgi cells. J. Physiol. 523, 175–192.
Vos, B.P., Maex, R., Volny-Luraghi, A. and De Schutter, E. (1999).
Parallel fibers synchronize spontaneous activity in cerebellar Golgi
cells. J. Neurosci. 19, 1–5.
Brickley, S.G., Revilla, V., Cull-Candy, S.G., Wisden, W. and Farrant,
M. (2001). Adaptive regulation of neuronal excitability by a voltage-
independent potassium conductance. Nature 409, 88–92.
Cherubini, E. and Conti, F. (2001). Generating diversity at GABAer-
gic synapses. Trends Neurosci. 24, 155–162.
D’Angelo, E., Rossi, P., Armano, S. and Taglietti, V. (1999). Evidence
for NMDA and mGlu receptor-dependent long-term potentiation of
mossy fiber-granule cell transmission in rat cerebellum. J. Neuro-
physiol. 81, 277–287.
Figure 1. Synaptic elements in a glomerulus of the cerebellar
cortex.
Excitatory synapses are represented by triangles and inhibitory
ones by circles.

Source: http://www.tnb.ua.ac.be/publications/full/12015139/DeSchutterCurrBiol02.pdf