ArticlePDF AvailableLiterature Review

Disruption in the Inhibitory Architecture of the Cell Minicolumn: Implications for Autisim

Authors:
  • International Group of Neuroscience IGN (Independent)
Article

Disruption in the Inhibitory Architecture of the Cell Minicolumn: Implications for Autisim

Abstract and Figures

The modular arrangement of the neocortex is based on the cell minicolumn: a self-contained ecosystem of neurons and their afferent, efferent, and interneuronal connections. The authors' preliminary studies indicate that minicolumns in the brains of autistic patients are narrower, with an altered internal organization. More specifically, their minicolumns reveal less peripheral neuropil space and increased spacing among their constituent cells. The peripheral neuropil space of the minicolumn is the conduit, among other things, for inhibitory local circuit projections. A defect in these GABAergic fibers may correlate with the increased prevalence of seizures among autistic patients. This article expands on our initial findings by arguing for the specificity of GABAergic inhibition in the neocortex as being focused around its mini- and macrocolumnar organization. The authors conclude that GABAergic interneurons are vital to proper minicolumnar differentiation and signal processing (e.g., filtering capacity of the neocortex), thus providing a putative correlate to autistic symptomatology.
Content may be subject to copyright.
http://nro.sagepub.com
The Neuroscientist
DOI: 10.1177/1073858403253552
2003; 9; 496 Neuroscientist
Manuel F. Casanova, Daniel Buxhoeveden and Juan Gomez
Disruption in the Inhibitory Architecture of the Cell Minicolumn: Implications for Autisim
http://nro.sagepub.com/cgi/content/abstract/9/6/496
The online version of this article can be found at:
Published by:
http://www.sagepublications.com
can be found at:The Neuroscientist Additional services and information for
http://nro.sagepub.com/cgi/alerts Email Alerts:
http://nro.sagepub.com/subscriptions Subscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://nro.sagepub.com/cgi/content/refs/9/6/496
SAGE Journals Online and HighWire Press platforms):
(this article cites 63 articles hosted on the Citations
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
496 THE NEUROSCIENTIST Lateral Inhibition in Autism
Copyright © 2003 Sage Publications
ISSN 1073-8584
Notwithstanding the major classes of neurons, identify-
ing and classifying inhibitory interneurons has proven
very difficult. Lund and Lewis (1993) reported 13 sepa-
rate classes of local circuit neurons in monkey prefrontal
cortex, only 5 of which resembled those found in
humans. The potentially large variety for different cell
types has negated the concept that “the” neuron repre-
sents an element of the brain in the same way that hepa-
tocytes or myocytes characterize their respective organs.
In other parts of the body, specific cells unequivocally
manifest the particular function of that organ. However,
the same cannot be said of the cerebral cortex where
individual neurons seem incapable of producing
thoughts or complex behaviors. Furthermore, the large
variety of neuronal cell types questions the wisdom of
those studies that focus on single-cell pathology in com-
plex behavioral disorders such as autism or schizophre-
nia. Mental function requires a multitude of neuronal
networks working together in temporal synchrony. In
this regard, a focus on the columnar organization of the
neocortex rather than on individual cells may better
show how the mind works under normal conditions and
pathological states.
A system exists when the properties of the whole are
greater than the sum of its functionally related parts—
that is to say, when the examination of its constituent
elements results in unexpected consequences. The pres-
ence of systems within an organ implies a graded cytoar-
chitectural organization. In the primate brain, embryo-
genic systems composed of vertical columns of cortex
are called “modules.In these arrangements, connectiv-
ity within modules is stronger than between them.
Modules embody a second property: their constituent
cells share similar stimulus/response properties. The
smallest neocortical module capable of information pro-
cessing is the minicolumn (Mountcastle 1998).
Strict biological reductionism would require that the
function and differentiation of the minicolumn be com-
pletely explained in terms of its component neurons.
However, minicolumns have emergent properties that
are not proper to the individual neurons (Fig. 1). These
properties are mediated by the relationship among the
neurons and their connections (i.e., cortical afferent,
efferent, and interneuronal). It is noteworthy that signal
processing by minicolumns may become self-reinforcing,
that is, reveal a tendency to magnify small effects under
appropriate conditions. In neurophysiology, this effect is
mediated through positive feedback. This phenomenon
is well known in neurosurgery, which employs subpial
transection rather than cortical ablation for cases in
which focal seizures involve the primary cortices or the
language regions of the brain (e.g., Landau-Kleffner
syndrome). This surgical procedure is intended to tran-
sect those horizontal connections of minicolumns impli-
cated in the kindling of an ictal event. Because affected
Disruption in the Inhibitory Architecture
of the Cell Minicolumn: Implications for Autisim
MANUEL F. CASANOVA, DANIEL BUXHOEVEDEN, and JUAN GOMEZ
The modular arrangement of the neocortex is based on the cell minicolumn: a self-contained ecosystem of
neurons and their afferent, efferent, and interneuronal connections. The authors’ preliminary studies indi-
cate that minicolumns in the brains of autistic patients are narrower, with an altered internal organization.
More specifically, their minicolumns reveal less peripheral neuropil space and increased spacing among
their constituent cells. The peripheral neuropil space of the minicolumn is the conduit, among other things,
for inhibitory local circuit projections. A defect in these GABAergic fibers may correlate with the increased
prevalence of seizures among autistic patients. This article expands on our initial findings by arguing for the
specificity of GABAergic inhibition in the neocortex as being focused around its mini- and macrocolumnar
organization. The authors conclude that GABAergic interneurons are vital to proper minicolumnar differen-
tiation and signal processing (e.g., filtering capacity of the neocortex), thus providing a putative correlate to
autistic symptomatology. NEUROSCIENTIST 9(6):496–507, 2003. DOI: 10.1177/1073858403253552
KEY WORDS: Autistic disorder, Cerebral cortex/pathology, Computer simulation
This study was funded by grants from the Stanley Medical Research
Foundation, the VA Merit Review Board, and the National Institutes of
Mental Health (grants MH61606 and MH62654).
Department of Psychiatry and Department of Neurology, Pathology,
Anatomy and Cell Biology; Medical College of Georgia (MFC).
Department of Psychiatry, Medical College of Georgia (DB, JG).
Address correspondence to: Manuel F. Casanova, Gottfried and
Gisela Kolb Endowed Chair, Department of Psychiatry and
Behavioral Science, University of Louisville, Health Sciences
Center, 500 South
Preston St., Bldg A, Louisville, KY 40292 (e-mail:
m0casa02@louisville.edu).
PROGRESS IN CLINICAL NEUROSCIENCE
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
Volume 9, Number 6, 2003 THE NEUROSCIENTIST 497
patients are not born with seizures but develop them dur-
ing infancy, the minicolumn in this instance exhibits a
perversion of Hebbian learning.
Minicolumns are reiterative elements of the neocortex
(Fig. 2). They are defined anatomically by the position
of bundles: those composed of apical dendrites, myeli-
nated axons, and double-bouquet axons. In Nissl-stained
sections, these components are integrated into single-
cell radial structures spanning layers II through VI (Fig.
3). The total content of cells per minicolumn varies from
60 to 100. Certain areas, such as koniocortex (i.e., stri-
ate cortex), show a two- to threefold increase in num-
bers. The genesis of this structure appears early in gesta-
tion. A series of symmetrical divisions of periventricular
precursor cells first define the total number of mini-
columns. Then, a subsequent wave of asymmetrical divi-
sions determines the total number of cells within the
minicolumns. Although the radial path of minicolumnar
precursor cells is well known, some interneurons may
pursue a longer tangential approach to the cortex
(Corbin and others 2001).
Recent reports have suggested that autism may result
from a minicolumnopathy (Casanova and others 2002b,
2002c). More specifically, minicolumns in the brains of
autistic patients were both more numerous and narrower
than normal. Furthermore, their constituent neurons
were more dispersed than in controls, accounting for a
normal cellular density. Similar results have now been
reported in high-functioning autism or Asperger’s syn-
drome (Casanova and others 2002a). In both autism and
Asperger’s syndrome, narrowing of the minicolumns
was most prominent for the peripheral neuropil com-
partment. Because this compartment encases the
unmyelinated projections of some interneurons,
researchers have postulated a deficit of inhibition in
autism (Casanova and others 2002b, 2002c). This would
provide a pathological correlate to some autistic symp-
tomatology, for example, the expression of seizures and
amelioration of certain behavioral traits with anticonvul-
sants (Casanova, Buxhoeveden, and Brown 2002). This
article will focus on deficits of inhibition as they pertain
to the organization of cortical minicolumns. It will also
suggest how such a deficit may contribute to dysfunction
in autism.
Minicolumns and Macrocolumns
The somas of cortical neurons are not randomly distrib-
uted in space; they are organized in both layers and dif-
ferent-sized columns or modules. Similarly, some den-
dritic and axonal ramifications that begin or end in these
somas are wired in parallel groups of fibers. This repet-
itive pattern of connections and cellular bodies, a puta-
tive “crystalline structure of the cortex, is the key to
understanding the huge processing capacity of the brain
and its modular arrangement (see Box 1).
The modular organization of the brain may be said to
span four hierarchical levels: 1) the individual minicol-
umn, 2) the engagement of multiple minicolumns in
structures less than a single macrocolumn, 3) an entire
macrocolumn, and 4) large-scale networks of macro-
columns. This hierarchical model does not include a
level composed of multiple adjacent macrocolumns. In
barrel field cortex, blocking GABA-mediated synaptic
inhibition causes a lateral spread of excitation for layers
II/III and V that is restricted to the barrel of origin
(Petersen and Sakmann 2001).
1
It thus appears that neu-
rons from a particular macrocolumn do not communi-
cate with neighboring barrels. Rather, outputs from indi-
vidual barrels are preferentially directed toward neurons
of the parent column (Laaris and Keller 2002).
The highest level of modularity within the brain is a
network of macrocolumns. These networks are geo-
graphically noncontiguous and widely distributed
throughout the neocortex. Although these networks hold
the promise of unraveling global cognitive processes,
this article will not deal with them. Alternatively, the
lowest level of neocortical modularity is the minicol-
umn. The minicolumn is the unit of anatomy that reiter-
ates itself throughout the neocortex while its function
simultaneously reflects the holistic properties of the
brain: afferent, efferent, and central processing
(Buxhoeveden and Casanova 2002a, 2002b). Reelin, a
glycoprotein produced and secreted by Cajal-Retzius
cells, appears to be involved in the development of these
vertical structures (Nishikawa and others 2002).
Discussions of minicolumns usually focus on their com-
ponent neurons and connections. However, astrocytes
with interlaminar processes may provide an essential
A2EY7C
Fig. 1. Recurrence or reentry is an essential feature within the
organization of minicolumnar assemblies. An increase in the
number of minicolumns is accompanied by a still larger increase
in their connective infrastructure. Those minicolumns that tend
to discharge together will be selected and their synapses
strengthened. Slight modifications in the connectivity of involved
minicolumns can generate novel representations that are not
reducible to inputs of the original minicolumns. It is through the
proliferation and apposition of supernumerary minicolumns that
new cortical areas may be formed in the course of evolution. In
this illustration, the two interconnected minicolumns are not
blank slates but have developed a stimuli history with other, not
necessarily adjacent, minicolumns.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
498 THE NEUROSCIENTIST Lateral Inhibition in Autism
role by maintaining the spatial definition of mini-
columns and isolating them from neighboring units
(Colombo and others 2000; Colombo 2001). Thick,
Nissl-stained sections allow the visualization of mini-
columns as central cellular cores surrounded by cell-
poor areas. Cell-poor areas (the peripheral neuropil
space of the minicolumn) are rich in unmyelinated
(inhibitory) axon fibers, dendritic arborizations, and
synapses (Buxhoeveden and Casanova 2002a, 2002b).
Multiple minicolumns aggregate into hexagonal-like
arrangements called macrocolumns (Fig. 4).
The cells of macrocolumns respond to similar but not
identical features (Tanaka 1997). The subtle variations in
response properties seem to be a function of individual
minicolumns within the larger module. Estimates for the
number of minicolumns occupying a macrocolumn
range between 60 and 80 (Favorov and Kelly 1994a,
1994b; Calvin 1998). Macrocolumns have been the
object of greater scientific inquiry than have the mini-
columns that constitute them. Examples of macro-
columns include barrels in rodent somatosensory cortex,
segregates in cat and monkey somatosensory cortex, and
hypercolumns in visual cortex. Macrocolumns have
been studied by electrophysiological means, immunocy-
A
B
C
D
Fig. 2. Minicolumns during development. (A) Ontogenetic cell columns at 18 weeks of gestation. (B) Ontogenetic cell columns at 28
weeks of gestation. Note onset of lamination with increased lateral and vertical spacing between cells in layer III. (C) Cell columns
(pyramidal cell arrays) in a 4-year-old. (D) Pyramidal cell arrays in a 50-year-old. Layer III is in the center and layer IV at the bottom.
(Nissl-stain, (A) at 200X magnification, (B), (C), and (D) at 100X magnification.)
Box 1: Lashley Revisited
The function of minicolumns lends itself to com-
parisons with logic gates and even to microproces-
sors. In the simpler example, minicolumns as bio-
logical gates have the advantage of small size, low
power consumption, and reliability. The white mat-
ter provides for the connectivity of the logic gate
(minicolumn), making the thalamocortical, associ-
ation, and commissural connections the circuit dia-
grams of the brain. Minicolumns throughout the
neocortex may follow the same anatomical and
physiological template; that is, they may act as a
single logic gate. The fact that all minicolumns may
stem from a primordial template is exemplified by
the unvarying use of layers II and III for association
(e.g., corticocortical integration), layer IV for
reception, and layers V and VI for efferent connec-
tivity. It is therefore noteworthy that combinations
of single logic gates (i.e., NOR) can provide for
higher-level functions capable of implementing any
computer program.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
Volume 9, Number 6, 2003 THE NEUROSCIENTIST 499
tochemistry, and optical imaging. Physiologically,
macrocolumns may exist as “topographic entities due in
large part to the operations of the spiny-stellate cells of
layer IV and the double-bouquet cells of the superficial
layers” (Favorov and Kelly 1994a; Favorov and Kelly
1994b, p 408). Anatomically, dendritic branching and
axonal patches (intrinsic and extrinsic) are used to
define the size of macrocolumns. In somatosensory cor-
tex, the anatomical basis of a macrocolumn is the termi-
nation of thalamic afferents that occur in focal clusters
measuring hundreds of microns in width (Mountcastle
1997). Research has found that axonal terminal patches
typically match the size of dendritic arbors (Elston and
Rosa 2000; Lubke and others 2000). This fact suggests a
predetermined (not coincidental) structure for their
aggregate fields. Thalamic activity is relayed in the ver-
tical direction and limited in the horizontal direction by
an input cluster. This activity engages all neuronal types
across all horizontal lamina. The bundling of afferent
axons also includes association and commissural sys-
tems. Thus, the widths of the terminal fields of callosal
afferents equal those of the thalamic bundles with which
they interdigitate or, under certain circumstance, become
convergent (i.e., pyramidal cells in layer IIIb can both
emit callosal fibers and receive direct input from thala-
mic afferents) (Mountcastle 1997).
Minicolumns and Inhibitory Circuits
The neocortex, with its minicolumnar organization, pre-
sents a specific type of inhibitory circuit. Minicolumns
use at least two basic forms of inhibition: lateral (or sur-
round) and intrinsic. Lateral inhibition sharpens the bor-
ders of minicolumns and increases their definition
(Marin-Padilla 1970; Szentagothai 1978; DeFelipe and
others 1990; Favorov and Kelly 1994a, 1994b; DeFelipe
1999). The primary source for this inhibitory effect may
be derived from axon bundles of double-bouquet cells
Fig. 3. Vertical components in the cortex. The following elements contribute to the anatomical vertical orientation, as well as to phys-
iological vertical bias of the neocortex: (A) Pyramidal cell arrays. Pyramidal cell somas of layers III, V, and VI are vertically oriented
(Schlaug and others 1995). (B) Interneurons. Studies have shown that GABAergic cells tend to be vertically aligned within the cortex
(DeFelipe and Jones 1985; Somogyi and Soltesz 1986). Marin-Padilla (1970) reported a vertical alignment of basket cells in human
motor cortex. (C) Axons. Each minicolumn contains one myelinated efferent bundle extending from layers II/III to layer VI (Peters and
Sethares 1996) and at least one double-bouquet axon bundle (DeFelipe and others 1990; Peters and Sethares 1997). Layer IV stellate
cells send axons vertically to supragranular layers. (D) Dendrites. Layers V/VI pyramidal cells have bundles of vertically oriented den-
drites that extend upward within a column (Peters and Sethares 1991). Dendrites of nonpyramidal cells also have strong vertical ori-
entation (Jin and others 2001; Kozloski and others 2001). (E) Overlay. The previously mentioned components are grouped together
within the confines of a minicolumn.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
500 THE NEUROSCIENTIST Lateral Inhibition in Autism
(DeFelipe and others 1990; Favorov and Kelly 1994a)
and basket cells (Marin-Padilla 1970) (Fig. 5). The
axons of double-bouquet cells arrange themselves in
essentially repeatable patterns varying between 15 and
30 µm wide, depending on the cortical area examined
(DeFelipe and others 1990; DeFelipe 1999), and are
con
sidered by some to be the source of lateral inhibi-
tion of neighboring minicolumns. The synaptic patterns
of
double-
bouquet axons are complex, apparently con-
tacting a variety of cell types including pyramidal, spiny
stellate cells, and nonpyramidal cells. All of the contacts
are dendritic of which approximately 60% are found on
small to medium dendritic shafts (1-2 µm) and the
remainder on distal spines. Among the possible func-
tions that have been postulated for double-bouquet cells
are that 1) they inhibit inhibitory interneurons (disinhi-
bition) within a column so as to allow the unopposed
flow of vertical excitation from spiny stellates within
that minicolumn, and 2) the vertical descent of axons
across the lamina inhibits dendritic terminals belonging
to excitatory cells in neighboring minicolumns.
However, the implication that double-bouquet axon bun-
dles are the source for a powerful surround inhibition
(Favorov and Kelly 1994a) has not been demonstrated.
The study by DeFelipe and others (1986) in monkey
sensory-motor cortex also suggests that basket cells have
several functions regarding columnar organization. A
majority of the horizontal axonal connections remains
within a macro-size column, whereas a small number of
these extend beyond this to encompass two or more
macrocolumns. Most dendritic fields of basket cells are
contained within a more narrow area averaging about
800 by 500 µm and vertically crossing at least three lay-
ers. The long-range (horizontal) myelinated axons cross
the equivalent of two or more macrocolumns, with some
extending as far as 2 mm. However, these long-range con-
nections are said to represent only a minority of the col-
laterals from basket cells. Therefore, although basket cells
receive input that is restricted to one macrocolumn, some
are capable of influencing signals in neighboring units.
The basket cell dendrite appears well suited for unification
of a macrocolumns size unit, whereas that double-bouquet
cell is clearly suited for minicolumn size control.
Intrinsic inhibition, on the other hand, regulates the
flow of information within a minicolumn. One function
of intrinsic inhibition is the transfer of information
between horizontal layers within a narrow vertical corti-
cal field. Already during development, excitatory thala-
mic inputs establish connections with minicolumns.
These inputs are disseminated via spiny stellate cells
(and apical dendrites of lamina V) radially to supra- and
infragranular layers to establish a vertical organization
based on shared excitation. In addition, thalamocortical
afferents participate in a feedforward pathway to
inhibitory interneurons in layer IV. After an initial volley
of excitatory input, thalamic connections excite inhibito-
ry interneurons, which then fire on short latency. The
axons of layer IV interneurons are anatomically aligned
vertically. This suggests that after a thalamic volley, inhi-
bition flows upward to the supragranular layers, in the
same way as excitation from the stellate and pyramidal
cells. Interestingly, excitatory interneurons in layers IIIb
and IV are responsible for projecting information verti-
cally in narrow bundles 100 to 200 µm wide. Because
this is larger than a typical minicolumn but smaller than
a macrocolumn, it raises the possibility that dynamic
columnar units comprise multiple minicolumns within
the confines of a “parent” macrocolumn (level 2 of the
modular organization, vide supra). Intrinsic inhibition
would be capable of subdividing them, whereas disinhi-
bition would serve to merge them.
AB
A B
Fig. 4. Vertical compartments within minicolumns. (A)
Schematic representation of the major vertical compo-
nents of cell columns viewed from a horizontal perspec-
tive (blue circles: myelinated bundles; green circles:
apical
dendrite bundles; red circles: double-bouquet axon
bundles). The illustration represents a tangential slice
at the upper level of layer IV in striate cortex. Peripheral
neuropil space is labeled “A”; core area is labeled “B.”
The core space comprises 90% of the neurons in a
column. The relationship between core area and the
peripheral neuropil space varies. In human area Tpt,
the core generally occupies between 50% and 60% of
the total area. The three fiber bundles depicted here are
found in each minicolumn and are a regularly spaced,
common feature of the primate neocortex (Peters and
Sethares 1997). Different projections are segregated
into separate parallel channels. This configuration
allows for parallel processing among neighboring mini-
columns. (B) Minicolumn input and output. Schematic
two-dimensional representation of major minicolumn
pathways and circuits. Information is received and sent
in all dimensions of the column: on top via layer I fibers;
at the bottom via the major thalamic, association, and
commissural pathways; and within the column (gray
matter) via horizontal connectivity.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
Volume 9, Number 6, 2003 THE NEUROSCIENTIST 501
The excitation-inhibition interface is vital to neuro-
logical function because it determines the filtering prop-
erties
2
of the tissue and the ultimate fate of an incoming
signal. Inhibition and excitation act not in opposition but
as part of a synchronized whole. Exploiting temporal
stimulus parameters requires precise timing of sensory
pathways (Grothe and Klump 2000). Temporal dynamics
in the brain are generated by propagation delays of neu-
ral pathways, synaptic delays, membrane time constants,
and the interaction of excitatory and inhibitory factors in
cellular compartments and ionic channels. Recently,
GABAergic inputs have been seen as crucial to phase
locking and for precise coincidence detection over a
wide range of stimulus intensities (Grothe and Klump
2000). In motor cortex, intrinsic cortical circuits pro-
mote temporal coordination of cortical modules in the
execution of complex movement patterns (Keller 1993).
In the spinal cord, sensory-motor processing is provided
DB
CH
BC
BD
ADAXON
DB
Barrel Column
ActivatedInhibited
A
B
C
Fig. 5. Inhibitory cells. (A) Double-bouquet cell: the axon bundles
of these cells contact small to medium dendritic shafts and distal
spines of apical and basal dendrites. The double-bouquet cell
sends long vertical axons in tight bundles hundreds of microns
into the cortex, beginning deep in layer II and extending to layer V,
although this varies with cortical area. Chandelier cell: this
interneuron synapses directly onto the axon hillock of pyramidal
cells. Basket cell: this inhibitory cell contacts the dendritic initial
segment and cell soma thereby modulating the input phase to the
pyramidal cell. Budd and Kisvarday (2001) reported peak connec-
tions of clutch cell axons (axon terminals) within a 30 to 45 µm
radius of the cell (i.e., the size of the minicolumn). (B) This figure
postulates how inhibitory interneurons can create smaller inhibito-
ry domains within larger units of function. Red lines indicate exci-
tatory input. In section A, afferents first contact strong inhibitory
interneurons, whereas in section B they converge first onto the
excitatory cells. The result is a biphasic response within a single
macrocolumn. The inhibitory neurons in section B would either be
inhibited or activated after the excitatory phase in section A has
been initiated. Porter and others (2001) showed that in mouse bar-
rel cortex, only a subset of inhibitory interneurons is activated by
thalamocortical inputs and that the afferents preferentially excite
the inhibitory interneurons. Inhibitory interneurons responded
more vigorously at lower thresholds and shorter latencies than
excitatory ones did. In barrel columns, as in minicolumns, inhibito-
ry neurons enhance spatial contrast. In this instance, it occurs
between principal and adjacent whiskers (Shimegi and others
1999). It does so by differentiating small versus large magnitude
responses (Brumberg and others 1996). Basket and chandelier
cells participate in surround inhibition in barrel cortex. (C)
Mountcastle (1997) referred to a “vertical stream of inhibition” pro-
vided by the axon bundles of double-bouquet cells. In this dia-
gram, minicolumn 2 receives a strong stimulus and is immediate-
ly able to inhibit its neighboring columns. This creates the contrast
needed to enhance somatosensory discrimination. The exact role
of the double-bouquet cell is far from understood, and other
interneurons such as basket and chandelier cells may be involved
in the lateral inhibition of minicolumns.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
502 THE NEUROSCIENTIST Lateral Inhibition in Autism
by tightly knit circuits containing a variety of inhibitory
cell types such as the Renshaw cell (recurrent inhibition)
and the Ia inhibitory interneuron (reciprocal inhibition).
The excitatory convergence model of Hubel and
Wiesel has been the dominant paradigm for approxi-
mately 40 years (1962). However, an increased role for
local inhibitory circuits is being proposed for the visual
cortex—in this regard, a model for other cortical areas.
Some recent articles have discussed emerging models
for the functioning of orientation columns in the visual
cortex. These studies propose that intracortical inhibito-
ry systems are heavily implicated for selectivity of ori-
entation (Vidyasagar and others 1996; Sompolinsky and
Shapley 1997). The theory of cross-orientation inhibi-
tion (Somers and others 1995; Vidyasagar and others
1996) suggests that intracortical inhibition, rather than
excitatory input from the thalamus, creates orientation
selectivity.
Based on these facts, we conclude that whereas the
spread of excitation indicates the capacity of the mini-
column to fuse and form bigger units, the spread of inhi-
bition determines the capacity of a module to subdivide
into smaller units. The presence of inhibitory synapses in
key topographical locations indicates the potential of the
module for self-division. The dynamic competition
between inhibition and excitation determines whether a
region of interest exhibits discontinuity (modular limit)
or continuity (modular fusion). In this context, complex
mental pathology can result from a disruption in this
competition. Although Hussman (2001), for instance,
suggested that autism might be caused by an array of
defects in relatively independent systems, some have
argued that it could be reduced to a single dysfunction,
that is, cortical inhibitory defect (Table 1). Among the
reasons given as support for the hypothesis is that
pathology relating to GABA receptors is a common fea-
ture in several suspected etiologies of autism. Both dis-
inhibition of GABAergic influence and excessive stimu-
lation of non-NMDA glutamate receptors generate pathol-
ogy similar to autism. Furthermore, the most prevalent
genetic or environmental factor found among the first
100 cases in the South Carolina autism project is an abnor-
mality of the chromosome 15q that has three GABA
receptor subunit genes. Last, computer models that alter
excitation and inhibition levels provide for representa-
tions of core autistic symptoms. Some of these models
have critically tied the fundamental shaping of the
macrocolumn during development to the correct balance
of excitation and inhibition of individual minicolumns.
Computer Models of Excitation and Inhibition
The model proposed by Favorov and Kelly (1994a,
1994b) closely parallels physiological studies, differenti-
ates between mini- and macrocolumns, and uses the
main types of cells responsible for excitation and inhibi-
tion. In this regard, their study may be more biologically
accurate than others discussed below. Favorov and Kelly
(1994a, 1994b) ran computer models of segregate for-
mation (macrocolumns) in primary somatosensory cor-
tex based on studies in rodent and monkey cortex. They
studied segregates according to the effects of 1) excita-
tory thalamocortical input, 2) thalamic input plus lateral
excitation, 3) the latter two plus the addition of inhibi-
tion, and 4) the effects of varying the synaptic weights of
inhibition and excitation. Their model used three kinds
of cells, those representing thalamic input (spiny stel-
late), lateral excitation (pyramidal cell), and inhibition
(double-bouquet cell). They argued that during perinatal
development, minicolumns play an important role in the
selection of thalamic connections to neighboring mini-
columns. The thalamic connections to individual mini-
columns are shaped by the primary interaction of mini-
columns with those neighbors belonging to the same
segregate. In other words, within-segregate self-
organization
of minicolumns induces an orderly pattern
of afferent connections. Lateral inhibition provides the
minicolumn with diverse receptive fields arranged in a
shuffled, yet orderly manner. Favorov and Kelly (1994a,
1994b) found that in the presence of too much inhibi-
tion, minicolumns within a segregate became highly dis-
ordered. On the other hand, with a diminution of lateral
inhibition, the segregates tended to merge. That is, exci-
tatory lateral connections caused interconnected mini-
columns to act more alike.
In a normal segregate, not all minicolumns are acti-
vated at the same time but respond to different thalamic
inputs. Inhibition provided by GABAergic cells, like the
double-bouquet neurons, accounts for some of the sys-
tem’s plasticity. An example in somatosensory cortex
concerns directional selectivity. A stimulus for a partic-
ular direction will initially activate a single minicolumn
(“x”) that in turn inhibits its immediate neighbors (“y”
and “z”). The neighbor neurons (“y” and “z”) are thus
inhibited from responding when the stimulus arrives at
their own receptive fields. The first minicolumn receives
little inhibition during most of its stimulus period.
However, a stimulus moving in another direction would
first activate the opposing minicolumn “y” with the
same sequence of events, thereby inhibiting “x” and pre-
venting inhibition of “y” during most of the stimulus
Table 1. GABAergic Abnormalities in Autism
Alterations in platelet, plasma, and urine GABA levels
Abnormalities in the gene coding for reelin
a
as well
as in its tissue levels
Abnormalities in the long arm of chromosome 15
(near a cluster of genes coding for GABA receptor
subunits)
A paradoxical effect of benzodiazepines on autistic
individuals
Reduced GABAergic receptor binding in the
hippocampus
Increased incidence of seizures
The table is a summary of evidence from Dhossche and oth-
ers (2002).
a
A glycoprotein involved in the regulation of GABAergic trans-
mission.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
Volume 9, Number 6, 2003 THE NEUROSCIENTIST 503
period. This permits individual minicolumns to respond
differently to different stimuli such as point location,
directional motion detection, and pressure (Box 2).
According to Favorov and Kelly (1994b), thalamic con-
nections to minicolumns are not prescribed according to
some preconceived idea (e.g., the mapping relationship
between skin and cortex) but are selected by the model
itself. The interaction between thalamus and mini-
columns is therefore determined by its history of senso-
ry stimulation.
Gustafsson (1997) has proposed another model based
on a neural circuit theory of autism referred to as “inad-
equate cortical feature maps. This model is based on
Kohonen’s “physiological self-organizing map” with
Hebbian learning,
3
which has a columnar structure as an
essential feature. He proposed that proper self-organization
of feature maps does not take place in autism. Features
are defined as “characteristics of classes of objects that
are useful for distinguishing the objects within a class.
Gustafsson claimed that too much lateral inhibition nar-
rows the range of information for a column. A given col-
umn will be active (i.e., the output of neurons within it
will be simultaneously high) when a particular set of fea-
tures is present in the input. During development,
columnar structures develop and become ordered so that
objects with similar features activate columns that are in
close proximity to each other (Kohonen 1988; Favorov
and Kelly 1994a). The simplest lateral interaction
between neurons is described according to a Mexican
hat function (Fig. 6) in which neighboring neurons coop-
erate and distant neurons compete. This connectivity
gives capacities to detect spatial differences in neural
activity, and the network behaves as a generalized edge
or boundary detector. Edge detectors enhance contrast
and ignore irrelevant details. Defects of this mechanism
at higher levels, according to Gustafsson, cause an insis-
tence on sameness and attention to irrelevant data.
A column does not distribute synaptic weights evenly
among neurons. This allows for variability in the pro-
cessing of the objects that activate the column.
According to Gustafsson (1997), a wide column that
contains many neurons has a wider spread of synaptic
weights than a narrow one has. Narrow columns with
fewer cells facilitate “discrimination, whereas wider
columns facilitate “generalization. Our studies have
shown similar results: narrow columns (discrimination)
in the brains of autistic individuals and broader ones
(generalization) in dyslexia (Casanova and others 2002b,
2002c; Casanova, Buxhoeveden, Cohen, and others
2002). This has suggested “that minicolumns exist with-
in a phenotypic spectrum that intertwines the inhibitory/
excitatory flow of neocortical information with a tweak-
ing of the signal-to-noise ratio relevant to feature extrac-
tion” (Casanova, Buxhoeveden, Cohen, and others 2002,
p 110).
Cohen (1994) created another computer model of
autism using artificial neural networks. The author test-
ed the effects of having too many or too few neurons
offering connections. The study revealed that too many
connections produced excellent discrimination but infe-
rior generalization because of an overemphasis on
details unique to the training set. Although this study
used neurons as opposed to cell columns, it represents a
similar theme. An alteration in the balance between exci-
tation and inhibition may account for some aspect of
autistic behavior. This study suggests a loss of the abili-
ty to correctly oscillate between discrimination (inhibi-
tion) and generalization (excitation). Unfortunately, this
model does not take into account the effects of inhibi-
tion, it is not a self-organizing model (which we prefer),
nor does it speculate where such increases or decreases
in numbers of neurons occur within the modular organi-
zation of the brain, whether in minicolumns or in macro-
columns. It is fair to say that when this study was per-
Fig. 6. Lateral excitation and inhibition are modeled by bell-shaped functions with radial profiles shown (left). Each function is multi-
plied by some positive constant representing its absolute strength, then inhibition is subtracted from excitation to yield the Mexican
hat model. In a module of normal minicolumns (center) with Mexican hat lateral connections, the active column in red has a net exci-
tatory effect on its immediate neighbors shown pink, a net inhibitory effect on more distant minicolumns shown blue, and no signifi-
cant effect on distant minicolumns. In a module in which the minicolumns are narrower, corresponding to a 50% reduction in inhibi-
tion, lateral excitation dominates (right).
Box 2: Video Game Technology
We may think of minicolumns as a space bar in a
video game. In one game, the space bar is used to
fire a cannon at invading aliens. Implementing
another software may cause a different function for
the space bar, for example, rotating a space ship or
causing a character to jump. The potentially limit-
less number by which minicolumns may be used or
combined provides a mathematical explanation to
the vast number of operations entailed in complex
cognitive processes.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
504 THE NEUROSCIENTIST Lateral Inhibition in Autism
formed, no clear evidence existed for cortical abnormal-
ities in autism.
Because it is obvious that either too much excitation
or inhibition can result in a breakdown in the processing
of information, the question is which one of the afore-
mentioned models best represents what is seen in
autism. We agree with Gustafsson (1997) in the use of
the columnar model and the role of inhibition but ques-
tion his conclusions for the following reasons: 1) It fails
to account for the predilection of epilepsy-seizure activ-
ity seen in autism. Seizures are facilitated by a move
away from discreteness toward a synchronization of sig-
nals. If the pathophysiology of autism involved too much
lateral inhibition, it would not explain their increased
susceptibility to seizures. A further difficulty is trying to
correlate the autistic symptomatology of hypersensitivi-
ties and increased discriminatory abilities (which would
require narrow fields of inhibitory control) with seizures
(which suggest a loss of inhibition). If minicolumns lose
their lateral inhibition to the extent that they cannot
maintain their individuality, a merging between them
occurs with a resultant diffusion of the afferent input
(feature extraction) spread among several columns (Fig.
6). 2) Narrow columns by themselves do not produce
abnormal behavior. Humans and nonhuman primates
have much smaller minicolumns in visual cortex com-
pared to those of other mammals. In this instance, the
smaller columns allow for more interconnections and
result in greater complexity required for primate vision
(Peters and Yilmaz 1993). Similarly, the fact that mini-
columns are also smaller in V1 than in other regions of
the human cortex does not imply pathology. Last, the
brains of nonhuman primates contain smaller mini-
columns than those of autistic patients without manifest-
ing autistic-like behavior. The uniqueness of the mini-
column in autism is that they are smaller relative to the
norm for the human condition. This suggests that they
are abnormal, either in their internal configuration, their
interconnections, or both. 3) In addition to lateral mod-
els of inhibition such as portrayed by the Mexican hat
model, critical inhibitory defects can occur within a col-
umn. One example is the temporal balance between
feedforward excitation and inhibition among thalamic
afferents to layer IV stellate cells that regulate the inflow
of thalamic information into the minicolumn. 4) The
Mexican hat model (Fig. 6) used by the author may be
incomplete. The Mexican hat model states that lateral
excitatory connections drive neighbors to develop simi-
lar afferent connections, whereas minicolumns farther
away are driven to develop dissimilar afferent connec-
tions by inhibitory actions.
4
Physiological data (Favorov
and Kelly 1994a, 1994b) show that the receptive fields
of individual minicolumns are not strictly topographical
in organization. An adjacent column may be responsive
to a very different part of the larger receptive field
encountered in the segregate.
5
Although the segregate
will exhibit the same feature extraction, the response
properties of individual minicolumns within that segre-
gate vary. In this regard, a more general function can be
used, for instance, a Gabor function (Pollen and Ronner
1983; Field and Tolhurst 1986).
The Gabor function (Field and Tolhurst 1986) has
been useful in explaining edge detection and contrast
enhancement for the visual system. It seems that these
properties are diminished by a disruption of the normal
balance between excitation and inhibition. In Figure 6,
for instance, we represent the lateral extension of excita-
tion by a bell-shaped function, a cubic spline, which
peaks at the center of a minicolumn. The lateral exten-
sion of inhibition, on the other hand, is represented by
another bell-shaped function with a lower peak but of
greater extent. In this way, the subtraction between exci-
tation and inhibition, estimated for instance in terms of
synaptic density, synaptic weight, or membrane poten-
tial, can represent a good approximation of the Mexican
hat function (Wilson and Bergen 1979). It can be seen
that the alternation of dominance between the distribu-
tion of excitation and inhibition can produce the lateral
inhibition effects and its spatial filtering capacities. The
more fragile factor in this configuration appears to be
the inhibition around the periphery of the minicolumn,
where a deficit of inhibitory fibers in autism has been
suggested (Casanova and others 2002b). The result of
such a defect is the disruption of the flow of information
between minicolumns.
One may conclude from these models that if we rep-
resent the lateral extension of inhibition or excitation by
bell-shaped functions of different peaks and widths, the
net balance between both determines the computational
abilities (filtering properties) of the model. Models
based on the Mexican hat function can explore the gen-
eral properties of the spatial distribution of inhibition
and excitation, but only to a limited extent. The Mexican
hat function implies that interactions between units are
symmetrical and singular, whereas cortical lateral inter-
actions are asymmetric, anisotropic, and usually with
many other columns.
Autism and Inhibition
Based on the descriptions given thus far, one may rea-
sonably suspect a disruption of the normal balance
between excitation and inhibition in the columnar organ-
ization of autistic patients. Computer modeling suggests
that such an imbalance biases the information-processing
system toward more signal or discrimination. This bias
may explain some of the more publicized features of the
autistic condition: highly focused savant skills (e.g., cal-
endar calculators), hypersensitivities of all sensory
modalities (e.g., flickering of fluorescent lights), and
some eccentricities (e.g., eating the same food, wearing
the same clothes). In this regard, a series of noteworthy
studies report that both children and adults with autism
were superior to a control group in their ability to dis-
criminate novel, highly similar stimuli (Plaisted and oth-
ers 1998). Autistic children also have superior ability to
discriminate display items in visual search tasks
(O’Riordan and Plaisted 2001; O’Riordan and others
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
Volume 9, Number 6, 2003 THE NEUROSCIENTIST 505
2001). O’Riordan (2000) stated that enhanced discrimi-
nation in autism results from low-level perceptual pro-
cessing of incoming stimuli, or what is called the bottom-
up approach. This supports a model in which altered pro-
cessing of sensory input occurs in the neocortex (e.g., a
deficiency in the dampening of filtered information). If
this were the case, the enhancement of improper dis-
crimination would likely operate at the level of the mini-
or macrocolumn.
Another salient feature of autism is that by puberty,
one-third of affected individuals will have suffered at
least two unprovoked seizures (Volkmar and Nelson
1995). Interneurons both help define minicolumnar
organization and play a crucial role in epilepsy. DeFelipe
(1999) proposed that a genetically determined decrease
in the presence of a class of inhibitory interneurons
(chandelier cells) can make the brains of individuals
more susceptible to seizure disorders. He argued that the
considerable variability in human brain size and num-
bers of GABA interneurons suggest significant dissimi-
larities of internal connectivity. Furthermore, networks
of inhibitory interneurons acting as GABA-gated pace-
makers are critically involved in gamma oscillations
(Bragin and others 1995; Traub and others 1996; Grothe
and Klump 2000; Bartos and others 2002; Porjesz and
others 2002). Abnormalities in these mechanisms have
been associated with binding problems (the coactivation
of neural assemblies), which may be present in both
autism and schizophrenia (Shergill and others 2000;
Grice and others 2001; Brock and others 2002; Lee and
others 2003).
The genesis of the minicolumn is defined early in ges-
tation by a series of divisions of primordial cells lining
the anterodorsal aspect of the embryonic ventricles. The
first phase of symmetrical divisions defines the total
number of minicolumns. In primate species, both human
and nonhuman, this process usually transpires during the
first 40 days of gestation. The increased number of mini-
columns reported in autism (Casanova and others 2002b,
2002c) therefore suggests a disruption during the earlier
stages of gestation. The proposed timing correlates well
with the observation of a high incidence of pervasive
developmental disorders in children with prenatal expo-
sure to thalidomide (Rodier and others 1996, 1997) and
the concurrence of structural and functional brainstem
abnormalities in autistic children (Hashimoto and others
1992). This early insult may well interfere with some of
the unfolding capabilities of the developing brain of an
autistic patient.
A normal neonate does not come to the world as a
blank slate. Rather, it demonstrates a number of behav-
iors and potential capabilities that appear to be hard
wired during gestation. Among these capabilities are the
neonate’s preference for the mother’s complex vocaliza-
tions, a temperament style, the capacity for object recog-
nition, an internal grammatical structure, an ability to
parse elements of spoken sounds, and even taste prefer-
ence or aversion. An additional process that begins in
utero is attachment in which the baby co-constructs with
his or her mother an inkling of who he or she is going to
be. This effort, if successful, provides scaffolding for
future social constructs that will affect both early and
adult relationships. Correlational data support the role of
lack of attachment in the development of avoidant
behavior, problems at school, and a proclivity toward
mood disorders. Interference with an internal working
model of relationships (i.e., the initial parental model)
may provide for disturbances in proximity, reciprocity,
and commitment when dealing with other people. This
will in turn require a greater investment from other peo-
ple for an affective bond to develop. Furthermore, a
minicolumnar abnormality may translate difficulties in
the integration of information (e.g., accommodation,
assimilation) into a delay in language acquisition. In all,
minicolumnar abnormalities may incapacitate a patient
as a social being by distorting elements of the child’s
biopsychological experience. Williams James’s famous
portrayal of the internal world of infants as “one great
blooming, buzzing confusion” may better serve to
describe the potential shimmering kaleidoscope of per-
ceptual abnormalities caused by minicolumnopathies.
Notes
1. Nonetheless, cortical plasticity is present in rodent representa-
tional maps and may be due to the extension of a small fraction of thal-
amocortical input beyond the parent barrel (Arnold and others 2001).
2. Filtering refers to the capacity to accept or reject frequency
changes from a given stimulus.
3. The rules for Hebbian learning are the following: 1) the neuron
responding most strongly to an input pattern has its synapses modified
such that it becomes more sensitive to that input, and 2) any two cells
or systems of cells that are repeatedly active at the same time will tend
to become associated.
4. Recent studies using high spatiotemporal resolution BOLD
fMRI have reported signal in the submillimeter range that may be com-
patible with macrocolumnar activation following a Mexican hat func-
tion (Kim and others 2000). These studies offer hope for discerning the
spatial resolution of the macrocolumn under different challenges and
behavioral states.
5. From the “normal” Kohonen (1988) maps, it is well known that
even though there is a strong similarity property in these maps, there
will still be “ravines” (Kohonen’s wording) where neighboring neurons
have distinctly different (synaptic) weight vectors.
References
Arnold PB, Cheng X Li, Waters RS. 2001. Thalamocortical arbors
extend beyond single cortical barrels: an in vivo intracellular trac-
ing study in rat. Exp Brain Res 136:152–68.
Bartos M, Vida I, Frotscher M, Meyer A, Monyer H, Geiger JR, and
others. 2002. Fast synaptic inhibition promotes synchronized
gamma oscillations in hippocampal interneuron networks. Proc
Natl Acad Sci U S A 99(20):13222–7.
Bragin A, Jando G, Nadasdy Z, Hetke J, Wise K, Buzsaki G. 1995.
Gamma (40–100 Hz) oscillation in the hippocampus of the behav-
ing rat. J Neurosci 15(1 Pt 1):47–60.
Brock J, Brown CC, Boucher J, Rippon G. 2002. The temporal binding
deficit hypothesis of autism. Dev Psychopath 14(2):209–24.
Brumberg JC, Pinto DJ, Simons DJ. 1996. Spatial gradients and
inhibitory summation in the rat whisker barrel system. J
Neurophysiol 76(1):130–40.
Budd LML, Kisvarday ZF. 2001. Local lateral connectivity of inhibito-
ry clutch cells in layer 4 of cat visual cortex (area 17). Exp Brain
Res 140:245–50.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
506 THE NEUROSCIENTIST Lateral Inhibition in Autism
Buxhoeveden D, Casanova MF. 2002a. The minicolumnar hypothesis
in neurosciences. Brain 125(5):935–51.
Buxhoeveden D, Casanova MF. 2002b. The minicolumn and evolution
of the brain: a review. Brain Beh Evol 60(3):125–51.
Calvin WH. 1998. Competing for consciousness: how subconscious
thoughts cook on the back burner [online]. Available at:
http://WilliamCalvin.com/1990s/1998JConscStudies.htm.
Casanova MF, Buxhoeveden DP, Brown C. 2002. Clinical and macro-
scopic correlates of minicolumnar pathology in autism. J Child
Neurol 17:692–5.
Casanova MF, Buxhoeveden D, Cohen M, Switala A, Roy E. 2002. The
neuropathology of dyslexia. Ann Neurol 52:108–10.
Casanova MF, Buxhoeveden D, Switala A, Roy E. 2002a. Asperger’s
syndrome and cortical neuropathology. J Child Neurol
17(12):142–5.
Casanova MF, Buxhoeveden D, Switala A, Roy E. 2002b.
Minicolumnar pathology in autism. Neurol 58:428–32.
Casanova MF, Buxhoeveden D, Switala A, Roy E. 2002c. Neuronal
density and architecture (gray level index) in the brains of autistic
patients. J Child Neurol 17(7):515–21.
Cohen IL. 1994. An artificial neural network analogue of learning in
autism. Biol Psych 36:5–20.
Colombo JA. 2001. A columnar-supporting mode of astroglial archi-
tecture in the cerebral cortex of adult primates? Neurobiol
9(1):1–11.
Colombo JA, Fuchs E, Hartig W, Marotte LR, Puissant V. 2000.
“Rodent-like” and “primate-like” types of astroglial architecture in
the adult cerebral cortex of mammals: a comparative study. Anat
Embyol 201:111–20.
Corbin J, Nery S, Fishell G. 2001. Telencephalic cells take a tangent:
non0radial migrations in the mammalian forebrain. Nature
Neurosci Suppl 4:1177–82.
DeFelipe J. 1999. Chandelier cells and epilepsy. Brain 122:1807–22.
DeFelipe J, Hendry S, Hashikawa T, Molinari M, Jones EG. 1990. A
microcolumnar structure of monkey cerebral cortex revealed by
immunocytochemical studies of double bouquet cell axons.
Neurosci 37(3):655–73.
DeFelipe J, Hendry SH, Jones EG. 1986. A correlative electron micro-
scopic study of basket cells and large GABAergic neurons in the
monkey sensory-motor cortex. Neuroscience 17(4):991–1009.
DeFelipe J, Jones EG. 1985. Vertical organization of y-aminobutyric
acid-accumulating intrinsic neuronal systems in monkey cerebral
cortex. J Neurosci 5(12):3246–60.
Dhossche D, Applegate H, Araham A, Maertens P, Bland L, Bencsath
A, and others. 2002. Elevated plasma gamma-aminobutyric acid
(GABA) levels in autistic youngsters: stimulus for a GABA
hypothesis of autism. Med Sci Monit 8(8):1–6.
Elston GN, Rosa MGP. 2000. Pyramidal cells, patches, and cortical
columns: a comparative study of infragranular neurons in TEO, TE,
and the superior temporal polysensory areas of the macaque mon-
key. J Neurosci 20(RC117):1–5.
Favorov OV, Kelly G. 1994a. Minicolumnar organization within
somatosensory cortical segregates I: development of afferent con-
nections. Cereb Cortex 4:408–27.
Favorov OV, Kelly G. 1994b. Minicolumnar organization within
somatosensory cortical segregates II: emergent functional proper-
ties. Cereb Cortex 4:428–42.
Field DJ, Tolhurst DJ. 1986. The structure and symmetry of simple-cell
receptive-field profiles in the cat’s visual cortex. Proc R Soc Lond
Biol Sci 22;228(1253):379–400.
Grice SJ, Spratling MW, Karmiloff-Smith A, Halit H, Csibra G, de
Haan M, and others. 2001. Disordered visual processing and oscil-
latory brain activity in autism and Williams syndrome.
Neuroreport 12(12):2697–700.
Grothe B, Klump BM. 2000. Temporal processing in sensory systems
[review]. Curr Opin Neurobiol 10(4):467–73.
Gustafsson L. 1997. Inadequate cortical feature maps: a neural circuit
theory of autism. Biol Psych 42:1138–47.
Hashimoto T, Tayama M, Miyazaki M, Sakurama N, Yoshimoto T,
Murakawa, and others. 1992. Reduced brainstem size in children
with autism. Brain Devel 14(2):94–7.
Hubel DH, Wiesel TN. 1962. Receptive fields, binocular interaction
and functional architecture in the cat’s visual cortex. J Physiol
160:106–54.
Hussman JP. 2001. Suppressed GABAergic inhibition as a common
factor in suspected etiologies of autism. J Aut Devel Dis
31(2):247–8.
Jin X, Mathers PH, Szabo G, Katarova Z, Agmon A. 2001. Vertical
bias in dendritic trees of non-pyramidal neocortical neurons
expressing GAD67-GFP in vitro. Cereb Cortex 11:666–78.
Keller A. 1993. Intrinsic synaptic organization of the motor cortex.
Cereb Cortex 3(5):430–41.
Kim D-S, Duong TQ, Kim S-G. 2000. High resolution mapping of iso-
orientation columns by fMRI. Nat Neurosci 3:164–9.
Kohonen T. 1988. The “neural” phonetic typewriter. Computer
21(6):11–22.
Kozloski J, Hamzei-Sichani F, Yuste R. 2001. Stereotyped position of
local synaptic targets in neocortex. Science 293:868–972.
Laaris N, Keller A. 2002. Functional independence of layer IV barrels.
J Neurophysiol 87:1028–34.
Lee KH, Williams LM, Breakspear M, Gordon E. 2003. Synchronous
gamma activity: a review and contribution to an integrative neuro-
science model of schizophrenia. Brain Res 41(1):57–78.
Lubke J, Egger V, Sakmann B, Feldmeyer D. 2000. Columnar organi-
zation of dendrites and axons of single and synaptically coupled
excitatory spiny neurons in layer 4 of the rat barrel cortex. J
Neurosci 20(14):5300–11.
Lund JS, Lewis DA. 1993. Local circuit neurons of developing and
mature macaque prefrontal cortex: Golgi and immunocytochemi-
cal characteristics. J Comp Neurol 8;328(2):282–312.
Marin-Padilla M. 1970. Prenatal and early postnatal ontogenesis of the
human motor cortex: A Golgi study. II. The Basket-pyramidal sys-
tem. Brain Res 23:185–91.
Mountcastle VB. 1997. The columnar organization of the neocortex.
Brain 120:701–22.
Mountcastle VB. 1998. Perceptual neuroscience: the cerebral cortex.
Cambridge (MA): Harvard University Press.
Nishikawa S, Goto S, Hamasaki T, Yamada K, Ushio Y. 2002.
Involvement of reelin and Cajal-Retzius cells in the developmental
formation of vertical columnar structures in the cerebral cortex:
evidence from the study of mouse presubicular cortex. Cereb
Cortex 12:1024–30.
O’Riordan M. 2000. Superior modulation of activation levels of stim-
ulus representations does not underlie superior discrimination in
autism. Cognition 77(2):81–96.
O’Riordan M, Plaisted K. 2001. Enhanced discrimination in autism. Q
J Exp Psychol A 54(4):961–79.
O’Riordan MA, Plaisted KC, Driver J, Baron-Cohen S. 2001. Superior
visual search in autism. J Exp Psychol Hum Percept Perform
27(3):719–30.
Peters A, Sethares C. 1991. Organization of pyramidal neurons in area
17 of monkey visual cortex. J Comp Neurol 306(1):1–23.
Peters A, Sethares C. 1996. Myelinated axons and the pyramidal cell
modules in monkey primary visual cortex. J Comp Neurol
365(2):232–55.
Peters A, Sethares C. 1997. The organization of double bouquet cells
in monkey striate cortex. J Neurocytol 26(12):779–97.
Peters A, Yilmaz E. 1993. Neuronal organization in area 17 of cat visu-
al cortex. Cereb Cortex 3:49–68.
Petersen CC, Sakmann B. 2001. Functionally independent columns of
rat somatosensory barrel cortex revealed with voltage-sensitive dye
imaging. J Neurosci 21:8435–46.
Plaisted K, O’Riordan M, Baron-Cohen S. 1998. Enhanced discrimi-
nation of novel, highly similar stimuli by adults with autism during
a perceptual learning task. J Child Psychol Psych 39(5):765–75.
Pollen DA, Ronner SF. 1983. Visual cortical neurons as localized spa-
tial frequency filters: IEEE transactions on systems, man and
cybernetics 13(5):907–16.
Porjesz B, Almasy L, Edenberg HJ, Wang K, Chorlian DB, Foroud T,
and others. 2002. Linkage disequilibrium between the beta fre-
quency of the human EEG and a GABAA receptor gene locus.
Proc Natl Acad Sci 19;99(6):3729–33.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
Volume 9, Number 6, 2003 THE NEUROSCIENTIST 507
Porter JT, Johnson CK, Agmon A. 2001. Diverse types of interneurons
generate thalamus-evoked feedforward inhibition in the mouse bar-
rel cortex. J Neurosci 21(8):2699–710.
Rodier PM, Ingram JL, Tisdale B, Croog VJ. 1997. Linking etiologies
in humans and animal models: studies of autism [review]. Reprod
Toxicol 11(2-3):417–22.
Rodier PM, Ingram JL, Tisdale B, Nelson S, Romano J. 1996.
Embryological origin for autism: developmental anomalies of the
cranial nerve motor nuclei. J Comp Neurol 370:247–61.
Schlaug G, Schleicher A, Zilles K. 1995. Quantitative analysis of the
columnar arrangement of neurons in the human cingulate cortex. J
Comp Neurol 351:441–52.
Shergill SS, Brammer MJ, Williams SC, Murray RM, McGuire PK.
2000. Mapping auditory hallucinations in schizophrenia using
functional magnetic resonance imaging. Arch Gen Psych
57(11):1033–8.
Shimegi S, Ichikawa T, Akasaki T, Sato H. 1999. Temporal character-
istics of response integration evoked by multiple whisker stimula-
tions in the barrel cortex of rats. J Neurosci 19(22):10164–75.
Somers DC, Nelson SB, Sur M. 1995. An emergent model of orienta-
tion selectivity in cat visual cortical simple cells. J Neurosci
15(8):5448–65.
Somogyi P, Soltesz I. 1986. Immunogold demonstration of GABA in
synaptic terminals of intracellularly recorded, horseradish peroxi-
dase-filled basket cells and clutch cells in the cat’s visual cortex.
Neuroscience 19(4):1051–65.
Sompolinsky H, Shapley R. 1997. New perspectives on the mecha-
nisms for orientation selectivity. Curr Opin Neurobiol
7(4):514–22.
Szentagothai J. 1978. The neuron network of the cerebral cortex: a
functional interpretation. The Ferrier Lecture 1977. Proc R Soc
London Biol Sci 201:219–48.
Tanaka K. 1997. Columnar organization in the inferotemporal cortex.
In: Jones EG, Peters A, editors. Cerebral cortex. Vol. 12. New York:
Plenum. p 469–95.
Traub RD, Whittington MA, Stanford IM, Jefferys JGA. 1996.
Mechanism for generation of long-range synchronous fast oscilla-
tions in the cortex. Nature 17;383(6601):621–4.
Vidyasagar TR, Pei X, Volgushev M. 1996. Multiple mechanisms
underlying the orientation selectivity of visual cortical neurones.
Trends Neurosci 19(7):272–7.
Volkmar FR, Nelson DS. 1995. Seizure disorders in autism. J Am Acad
Child Adolesc Psych 29:127–9.
Wilson HR, Bergen JR. 1979. A four mechanism model for threshold
spatial vision. Vision Res 19(1):19–32.
© 2003 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
at PENNSYLVANIA STATE UNIV on April 14, 2008 http://nro.sagepub.comDownloaded from
... Incorporating elevated E/I ratio in our model resulted in a reduced representation of the Poggendorff illusion. Excitation/inhibition imbalance is often suggested as the core characteristic of the brain in ASD and the potential underlying cause of atypical sensory perception of individuals with ASD in multiple sensory modalities, including atypical response to auditory stimuli (Ida-Eto et al., 2017;Kondo & Lin, 2020;Visser et al., 2013), tactile stimuli (Orefice et al., 2016;Tannan et al., 2008) and visual stimuli (Casanova et al., 2003;Chung & Son, 2020;Robertson et al., 2014;Robertson et al., 2016;Rosenberg et al., 2015;Rubenstein & Merzenich, 2003;Snijders et al., 2013;Spiegel et al., 2019). In fact, Flevaris and Murray (2015) found that orientation-specific suppression in V1 decreases with increasing autistic tendency, which is directly relevant to our results. ...
... Together, these insights led us to propose that the presence of excitation/ inhibition imbalance or/and reduced top-down modulation in visual cortices could lead to less susceptibility to the Poggendorff illusion and related visual orientationbased illusions. These two traits, excitation/inhibition imbalance and reduced top-down modulation, are likely present in visual cortices of individuals with ASD (Casanova et al., 2003;Chung & Son, 2020;Flevaris & Murray, 2015;Isler et al., 2010;Kessler et al., 2016;Robertson et al., 2014Robertson et al., , 2016Rosenberg et al., 2015;Rubenstein & Merzenich, 2003;Seymour et al., 2019;Snijders et al., 2013;Spiegel et al., 2019). Therefore, the presence of either one or both of these traits in the visual cortices of individuals with ASD could lead to less susceptibility to the Poggendorff and related illusions that share the same rod-and-frame underlying mechanism, such as Zöllner, Roelof, Tilt and Ponzo illusions (Blakemore et al., 1971;Bridgeman et al., 2018;Clifford, 2014;Prinzmetal & Beck, 2001;Seymour et al., 2018). ...
Article
Full-text available
While atypical sensory perception is reported among individuals with autism spectrum disorder (ASD), the underlying neural mechanisms of autism that give rise to disruptions in sensory perception remain unclear. We developed a neural model with key physiological, functional, and neuroanatomical parameters to investigate mechanisms underlying the range of representations of visual illusions related to orientation perception in typically developed (TD) subjects compared to individuals with ASD. Our results showed that two theorized autistic traits, excitation/inhibition imbalance and weakening of top‐down modulation, could be potential candidates for reduced susceptibility to some visual illusions. Parametric correlation between cortical suppression, balance of excitation/inhibition, feedback from higher visual areas on one hand, and susceptibility to a class of visual illusions related to orientation perception on the other hand, provide the opportunity to investigate the contribution and complex interactions of distinct sensory processing mechanisms in ASD. The novel approach used in this study can be used to link behavioral, functional, and neuropathological studies, estimate and predict perceptual and cognitive heterogeneity in ASD, and form a basis for the development of novel diagnostics and therapeutics.
... In fact, many of the genes associated with ASD are involved in the development of both excitatory and inhibitory neurons [24], formation and maturation of synapses, and synaptic plasticity and function expression [25]. Moreover, irregularities in cortical column integrity-defined as the most basic functional unit of neural organization -can lead to deficient connectivity between distant cortical regions and impaired GA-BAergic regulation [26,27]. ...
... In fact, many of the genes associated with ASD are involved in the development of both excitatory and inhibitory neurons [24], formation and maturation of synapses, and synaptic plasticity and function expression [25]. Moreover, irregularities in cortical column integrity-defined as the most basic functional unit of neural organization-can lead to deficient connectivity between distant cortical regions and impaired GABAergic regulation [26,27]. ...
Article
Full-text available
Transcranial direct current stimulation (tDCS) has been proposed as a promising therapy for rehabilitation of neurodevelopmental disorders. In this review, we discuss studies on the impact of tDCS as a therapy for autism, schizophrenia, and attention deficit/hyperactivity disorder, as well as the tDCS’ mechanism of action, and propose future paths of research to optimize tDCS treatment protocols. The mechanism underlying tDCS effects is the modulation of excitatory and/or inhibitory activity, making it a valuable tool for restoring the excitation/inhibition (E/I) balance which is disrupted in many neurodevelopmental disorders. Clinical studies have shown that tDCS therapy is well-tolerated by patients and seems to ameliorate behavior and cognitive functions. Alterations in early development of neuronal circuits lead to disruptions in brain activity in neurodevelopmental disorders. An increasing amount of research into the effects of tDCS on neuronal activity has provided a foundation for its use as a therapy for behavior and cognitive characteristics of neurodevelopmental disorders. Clinical studies show that tDCS appears to ameliorate behavioral and cognitive outcomes of patients with autism, schizophrenia, and attention deficit/hyperactivity disorder. More research is needed to understand the mechanisms of action of tDCS and to optimize treatment protocols.
... circuits. One of the prominent theories in autism states that core traits of the condition occur secondary to altered development of cortical interneurons and resulting shift in balance between excitation and inhibition in cortical circuits across sensory and higher-order cortical areas (Rubenstein & Merzenich, 2003;Casanova et al., 2003;Gogolla et al., 2009;Robertson et al., 2014Robertson et al., , 2016. Dynamics of visual rivalry and the rate of perceptual reversals are similarly hypothesized to depend on excitation-inhibition dynamics and circuit wiring in the competing clusters of neurons coding for rivalrous percepts (Laing and Chow, 2002;Seely & Chow, 2011;Wilson, 2003;van Loon et al., 2013). ...
Preprint
Extracting common patterns of neural circuit computations in autism spectrum and pinning them as a cause of specific core traits of autism is the first step towards identifying cell- and circuit-level targets for effective clinical intervention. Studies in human subjects with autism have identified functional links and common anatomical substrates between core restricted behavioral repertoire, cognitive rigidity, and over-stability of visual percepts during visual rivalry. To be able to study these processes with single-cell precision and exhaustive neuronal population coverage, we developed the visual bi-stable perception task for mice. Our task is based on plaid patterns consisting of two transparent gratings drifting at an angle of 120 degrees relative to each other. This results in spontaneous reversals of the perception between local component motion (motion of the plaid perceived as two separate moving grating components) and integrated global pattern motion (motion of the plaid perceived as a fused moving texture). This robust paradigm does not depend on the explicit report of the mouse, as the direction of the optokinetic nystagmus (OKN, rapid eye movements driven by either pattern or component motion) is used to infer the dominant percept. Using this paradigm, we found that the rate of perceptual reversals between global and local motion interpretations of the stimulus is reduced in MECP2 duplication mouse model of autism. Moreover, the stability of local motion percepts is greatly increased in MECP2 duplication mice at the expense of global motion percepts. Thus, our model reproduces a subclass of core features in human autism (reduced rate of visual rivalry and atypical perception of visual motion). This further offers a well-controlled approach for dissecting neuronal circuits underlying these core features if combined with mesoscale 2-photon imaging and optogenetic manipulation of neuronal circuits.
... The differential term in Equation (1) (the second part of Equation (1)) tries to separate the target area from the background and emphasizes the sharpening of the boundary to improve the anti-interference ability. In Equation (1), d uv e −d uv has a functional surface with an inverted "Mexican hat", which is a computational model of lateral inhibition in neurobiology [42,43]. It means that the nearest and farthest neighbors have the least inhibitory effect on central neurons. ...
Article
Full-text available
Understanding the relationship between the chemical structure and physicochemical properties of odor molecules and olfactory perception, i.e., the structure–odor relationship, remains a decades-old, challenging task. However, the differences among the molecular structure graphs of different molecules are subtle and complex, and the molecular feature descriptors are numerous, with complex interactions that cause multiple odor perceptions. In this paper, we propose to decompose the features of the molecular structure graph into feature vectors corresponding to each odor perception descriptor to effectively explore higher-order semantic interactions between odor molecules and odor perception descriptors. We propose an olfactory perception prediction model noted as HGAFMN, which utilizes a hypergraph neural network with the olfactory lateral inhibition-inspired attention mechanism to learn the molecular structure feature from the odor molecular structure graph. Furthermore, existing methods cannot effectively extract interactive features in the large number of molecular feature descriptors, which have complex relations. To solve this problem, we add an attentional factorization mechanism to the deep neural network module and obtain a molecular descriptive feature through the deep feature combination based on the attention mechanism. Our proposed HGAFMN has achieved good results in extensive experiments and will help product design and quality assessment in the food, beverage, and fragrance industries.
... 12). An increased excitation by glutamate and reduced GABAergic inhibition [9][10][11] (inter al.) results in an overall excitation. Added to these, we have the long-distance innervation from monoamine-releasing neurons (noradrenaline, serotonin, dopamine), stimulating, inhibiting, or modulating the activity of the neurons they innervate. ...
Article
Full-text available
We suggest a later timeline for full language capabilities in Homo sapiens, placing the emergence of language over 200,000 years after the emergence of our species. The late Paleolithic period saw several significant changes. Homo sapiens became more gracile and gradually lost significant brain volumes. Detailed realistic cave paintings disappeared completely, and iconic/symbolic ones appeared at other sites. This may indicate a shift in perceptual abilities, away from an accurate perception of the present. Language in modern humans interact with vision. One example is the McGurk effect. Studies show that artistic abilities may improve when language-related brain areas are damaged or temporarily knocked out. Language relies on many pre-existing non-linguistic functions. We suggest that an overwhelming flow of perceptual information, vision, in particular, was an obstacle to language, as is sometimes implied in autism with relative language impairment. We systematically review the recent research literature investigating the relationship between language and perception. We see homologues of language-relevant brain functions predating language. Recent findings show brain lateralization for communicative gestures in other primates without language, supporting the idea that a language-ready brain may be overwhelmed by raw perception, thus blocking overt language from evolving. We find support in converging evidence for a change in neural organization away from raw perception, thus pushing the emergence of language closer in time. A recent origin of language makes it possible to investigate the genetic origins of language.
... The ability to discriminate stimuli relies on the ability to separate spatially distinct signals, which has been linked to GABA-mediated lateral inhibition (Puts et al., 2011(Puts et al., , 2017Sapey-Triomphe et al., 2019;Tommerdahl et al., 2010). Thus, elevated thresholds observed in our study could be explained, at least in part, by reduced GABAergic inhibition in early childhood autism, likely due to altered local circuitry (Casanova et al., 2003;Casanova et al., 2002). This Wang Y. et al., 2017). ...
Article
Full-text available
We assessed different aspects of tactile perception in young children (3–6 years) with autism. Autistic and neurotypical children completed vibrotactile tasks assessing reaction time, amplitude discrimination (sequential and simultaneous) and temporal discrimination (temporal order judgment and duration discrimination). Autistic children had elevated and more variable reaction times, suggesting slower perceptual-motor processing speed and/or greater distractibility. Children with autism also showed higher amplitude discrimination and temporal order judgement thresholds compared to neurotypical children. Tactile perceptual metrics did not associate with social or tactile sensitivities measured by parent-reports. Altered tactile behavioral responses appear in early childhood, can be quantified but appear dissociated from sensitivity. This implies these measures are complementary, but not necessarily related, phenomena of atypical tactile perception in autism.
... For instance, genetic variants involving synaptic genes are emerging as recurrent causative factors in the etiology of ASDs (Südhof, 2008;Zoghbi and Bear, 2012;Salpietro et al., 2019). Moreover, different authors speculate on the role of the unbalance between excitatory and inhibitory circuits in the pathogenesis of these disorders (Casanova et al., 2003;Rubenstein, 2010). Recently, genomic data and gene networks analysis suggested a common cause for ASDs during the embryonic development of the cerebral cortex (Parikshak et al., 2013;Willsey et al.,2013). ...
Book
Full-text available
Appropriate systems for modeling human organs and pathologies represent a constant need in all branches of biomedical research and practice. Ttranslation of preclinical results into effective treatments remains poor, highlighting the need for accurate tissue-specific human emulation systems. In this context, great expectations are accompanying the recent developments on organoids. Organoids are three dimensional tissue structures assembled from pluripotent or adult organ-specific stem cells. Their structure allows establishment of complex cell-cell interactions and gradients of oxygen, nutrients and soluble signals that generate tissue-specific heterogeneous cell types. Organoids are increasingly used to model and understand human organ-specific physiology in both normal and pathological conditions. As such, they emerge as powerful tools for personalised and precision medicine.
... Furthermore, at the neuronal level, the normal performance of cortical circuits in mammals depends on preserving the balance between excitatory and inhibitory (GABAergic) synaptic activity (E/I balance). It has been hypothesized that autism can caused by E/I immbalance in neural circuits that mediate language and social behaviors (Casanova et al., 2003;Rubenstein and Merzenich, 2003;Uzunova et al., 2016). Here, some of the studies that have suggested E/I imbalance as an important underlying pathophysiology for cortical dysfunction in ASD and the possibility of asymmetric changes in inhibitory neurons in these cortical areas in ASD will be discussed. ...
Article
Full-text available
Lateralized specialization of the two cerebral hemispheres is a fundamental structural hallmark of the human brain and underlies many cognitive functions and behavioral abilities. In typical developing individuals the influence of handedness on performance of various sensory modalities and the cortical processing has been well recognized. Increasing evidence suggests that several neurodevelopmental and psychiatric disorders such as bipolar disorder, schizophrenia, and autism spectrum disorders (ASD) are associated with abnormal patterns of cerebral lateralization. Individuals with ASD exhibit abnormal structural and functional lateralization of circuits subserving motor, auditory, somatosensory, visual face processing, and language-related functions. Furthermore, a high prevalence of atypical handedness has been reported in ASD individuals. While the hemispheric dominance is also related to functions other than handedness, there is a clear relationship between handedness and language-related cortical dominance. This minireview summarizes these recent findings on asymmetry in somatosensory and auditory cortical structures associated with language processing in ASD. I will also discuss the importance of cortical dominance and interhemispheric disruption of balance between excitatory and inhibitory synapses as pathophysiological mechanisms in ASD.
... The neurodevelopmental hypothesis found a further resonance in anatomopathological studies showing abnormal brain cytoarchitecture in ASD, characterized by an atypical laminar organization in the prefrontal and temporal cortices (Stoner et al., 2014) and an augmented presence of narrower cell minicolumns. The ensuing altered internal organization could impact the inhibitory local circuit projections, thus affecting the filtering capacity during information processing (Casanova et al., 2003). Moreover, advances in genetics have highlighted a mutation in ASD individuals of some of the genes involved in the glutamatergic synapses' development and functional regulation (Bonnet-Brilhault et al., 2016;Bourgeron, 2015;Laumonnier et al., 2006). ...
Article
In this review, we discuss the topic of intersubjectivity and its disruption in Autism Spectrum Disorder, both from a developmental and neuroscientific perspective. We present evidence of the impact of an early neurodevelopmental disorder on the blooming of intersubjective abilities. We propose the existence of a basic, pre-reflective social cognition that originates from the sensorimotor networks and is anchored to their teleological organization. We present neurophysiological evidence of its involvement in the direct understanding of three core elements of action: its category, its goal, and its dynamic contour. Finally, we discuss to which extent the disrupted development of brain neuroarchitecture and connectivity in Autism could also possibly impact on basic social cognition, thus prevent- ing fluid, preverbal alignment with the others.
Article
Autism spectrum disorders (ASD) are a group of neurodevelopmental disorders that develop in early life due to interaction between several genetic and environmental factors and lead to alterations in brain function and structure. During the last decades, several mechanisms have been placed to explain the pathogenesis of autism. Unfortunately, these are reported in several studies and reviews which make it difficult to follow by the reader. In addition, some recent molecular mechanisms related to ASD have been unrevealed. This paper revises and highlights the major common molecular mechanisms responsible for the clinical symptoms seen in people with ASD, including the roles of common genetic factors and disorders, neuroinflammation, GABAergic signaling, and alterations in Ca⁺² signaling. Besides, it covers the major molecular mechanisms and signaling pathways involved in initiating the epileptic seizure, including the alterations in the GABAergic and glutamate signaling, vitamin and mineral deficiency, disorders of metabolism, and autoimmunity. Finally, this review also discusses sleep disorder patterns and the molecular mechanisms underlying them.
Article
Full-text available
The cellular generation and spatial distribution of gamma frequency (40-100 Hz) activity was examined in the hippocampus of the awake rat. Field potentials and unit activity were recorded by multiple site silicon probes (5- and 16-site shanks) and wire electrode arrays. Gamma waves were highly coherent along the long axis of the dentate hilus, but average coherence decreased rapidly in the CA3 and CA1 directions. Analysis of short epochs revealed large fluctuations in coherence values between the dentate and CA1 gamma waves. Current source density analysis revealed large sinks and sources in the dentate gyrus with spatial distribution similar to the dipoles evoked by stimulation of the perforant path. The frequency changes of gamma and theta waves positively correlated (40-100 Hz and 5-10 Hz, respectively). Putative interneurons in the dentate gyrus discharged at gamma frequency and were phase-locked to the ascending part of the gamma waves recorded from the hilus. Following bilateral lesion of the entorhinal cortex the power and frequency of hilar gamma activity significantly decreased or disappeared. Instead, a large amplitude but slower gamma pattern (25-50 Hz) emerged in the CA3-CA1 network. We suggest that gamma oscillation emerges from an interaction between intrinsic oscillatory properties of interneurons and the network properties of the dentate gyrus. We also hypothesize that under physiological conditions the hilar gamma oscillation may be entrained by the entorhinal rhythm and that gamma oscillation in the CA3-CA1 circuitry is suppressed by either the hilar region or the entorhinal cortex.
Article
Full-text available
Although it is possible that the breadth of impairments in autism are caused by multiple defects in relatively independent systems, autism may instead reflect dysfunction in a single factor shared in common by many systems. Specifically, the severe disruptions observed in autism may be linked to suppression of GABAergic inhibition, resulting in excessive stimulation of glutamate-specialized neurons and loss of sensory gating. This view recognizes the possibility of multiple etiologies in autism. That is, there may exist a spectrum of genetic and environmental factors which impair inhibitory tone in a manner that expresses itself as autistic pathology. Keywords: GABA, autism, excitatory/inhibitory balance, E/I
Article
Children with a diagnosis of autism and normally developing children, matched for age and general ability, were tested on a series of visual search tasks in 2 separate experiments. The children with autism performed better than the normally developing children on difficult visual search tasks. This result occurred regardless of whether the target was uniquely defined by a single feature or a conjunction of features, as long as ceiling effects did not mask the difference. Superior visual search performance in autism can be seen as analogous to other reports of enhanced unique item detection in autism. Unique item detection in autism is discussed in the light of mechanisms proposed to be involved in normal visual search performance.
Chapter
A key function of the primate brain is recognition of objects from their visual images. The recognition process is flexible, tolerating marked changes in images due to changes in illumination, viewing angle, and pose of the object. Moreover, the primate visual system processes images of novel objects based on previous visual experience of similar objects. Generalization may be an intrinsic property of the primate visual system. In this review the neural organization essential for these flexible aspects of visual object recognition in TE of the inferotemporal cortex is discussed.
Article
The modular organization of nervous systems is a widely documented principle of design for both vertebrate and invertebrate brains of which the columnar organization of the neocortex is an example. The classical cytoarchitectural areas of the neocortex are composed of smaller units, local neural circuits repeated iteratively within each area. Modules may vary in cell type and number, in internal and external connectivity, and in mode of neuronal processing between different large entities; within any single large entity they have a basic similarity of internal design and operation. Modules are most commonly grouped into entities by sets of dominating external connections. This unifying factor is most obvious for the heterotypical sensory and motor areas of the neocortex. Columnar defining factors in homotypical areas are generated, in part, within the cortex itself. The set of all modules composing such an entity may be fractionated into different modular subsets by different extrinsic connections. Linkages between them and subsets in other large entities form distributed systems. The neighborhood relations between connected subsets of modules in different entities result in nested distributed systems that serve distributed functions. A cortical area defined in classical cytoarchitectural terms may belong to more than one and sometimes to several distributed systems. Columns in cytoarchitectural areas located at some distance from one another, but with some common properties, may be linked by long-range, intracortical connections.