|
Background Information
In
subjects with schizophrenia, audiologic, electrophysiologic, and structural
imaging studies indicate abnormalities in auditory sensory processing localized
to the auditory cortex of the superior temporal gyrus. Gray matter volume
reductions in this region appear to be specific to schizophrenia and are present
prior to neuroleptic treatment. These volume reductions are correlated with electrophysiologic measures of auditory sensory processing, and with the
cardinal symptoms of thought disorder and auditory hallucinations. In non-human
primates, cortical processing of auditory input occurs both in parallel and in
series through hierarchically organized auditory core, lateral belt, and
parabelt cortices of the superior temporal gyrus.
Feedforward projections (see
schematic) from
auditory core to lateral belt, and from lateral belt to parabelt, arise
predominantly from layer 3 pyramidal cells and terminate in layers 3c/4 of the
target region. Reciprocal feedback projections arise from both layers 3 and 5,
and terminate predominantly in layer 1. Feedforward projections provide for
rapid transfer of sensory information, facilitating pre-attentive behavioral
responses. In contrast, attentional or top-down mechanisms, which provide for
contextual interpretation of sensory input, are likely to be dependent on
feedback connections. The deficit in precision of early auditory processing,
observed in some subjects with schizophrenia, has been described as
pre-attentive, suggesting a preferential vulnerability of feedforward circuits.
Similarly, our initial studies of auditory cortex in subjects with schizophrenia
have found reductions in pyramidal cell size in a laminar pattern indicative of
alterations of feedforward, but not feedback, circuits
(see
Sweet et al 2003 and
Sweet et al 2004).
Proper interpretation of these findings has been
limited by the lack of data regarding the correspondence between human auditory
association cortex and auditory regions defined in non-human primates, and by
the absence of unbiased methods for determination of pyramidal cell numbers and
size within specific cortical layers. We have been addressing both these issues,
in collaboration
 with Dr. David Lewis at the University of Pittsburgh and Dr. Hans Jørgen
Gundersen of the University of Aarhus in Denmark. In non-human primate, auditory
core, lateral belt, and parabelt regions can be
distinguished by differing cytoarchitecture and intensity of staining for the calcium binding
protein, parvalbumin, and for the enzyme, acetylcholinesterase. We recently
reviewed the defining criteria for auditory core, lateral belt, and parabelt
using these methods in monkeys, and identified an additional division of the parabelt into internal and external subregions, as shown in the figure at right.
 We also evaluated the performance of these criteria
in
parcellating human auditory cortex, and found they could be applied with high
precision. (Follow links for examples of the
cytoarchitecture,
parvalbumin and
acetylcholinesterase
staining for the human auditory
cortical regions). A summary diagram of the locations of the auditory core,
lateral belt, and parabelt in the left superior temporal gyrus of humans, and
the location of the adjacent non-auditory Tpt, is
shown at
right.
Our current
findings in
subjects with schizophrenia have identified reduced
pyramidal cell somal volume in
in layer 3
of auditory cortex. Whether reduced mean somal
volume results from a smaller number of large neurons, an excess of small
neurons, or a diminution in size of neurons without a change in total neuron
number is not known. Determining this requires concurrent determination of
pyramidal cell size and number. Unbiased stereologic methods exist for
quantification of these features in cortex, but
not within individual layers. Because neuronal connectivity and function differs
across layers in the cerebral cortex, for quantitative studies to be fully
informative it is essential that layer specific information not be lost. We have
modified an unbiased stereologic method, the Fixed Axis Vertical Rotator (FAVeR),
for application to auditory cortex.
 The
image at right demonstrates our approach. The cortex of the superior temporal
gyrus is is cut into narrow slabs of uniform thickness. The slabs are separated
into two uniform random series. One (on the left) is sectioned for mapping the
region of interest (e.g. parabelt shown in yellow). From these maps, the
regional boundaries are identified in the 2nd series of slabs (on the right).
The region of interest is cut perpendicular to the pial surface from slabs in
the 2nd series (as marked on a representative slab) to yield several uniform
small blocks. This process is repeated for all slabs. A systematic random
sub-sample of all blocks is selected. Each block is randomly rotated around an
axis perpendicular to the pial surface prior to sectioning. A central section,
cut from the blocks, is used in vertical nucleator estimation of somal volume
and cell number. The cortical layers and a region of interest in layer 3c
(shaded), are indicated.
Additional studies to quantify laminar changes in pre- and postsynaptic elements
in subjects with schizophrenia are ongoing. For example,
axon terminals can be labeled with an antibody to synaptophysin, and their
number can then be determined. Similarly,
dendritic
spines can be labeled with antibody to spinophilin and quantified (shown
below).
In contrast to schizophrenia, relatively less
is known about regional localization of cortical abnormalities that correlate
with psychosis in Alzheimer Disease. We have hypothesized that positive
psychotic symptoms, delusions and hallucinations, whether occurring in
schizophrenia or Alzheimer Disease, are likely to share abnormalities of select
brain circuitry. Studies from our laboratory and others suggest that the prefrontal
and temporal cortex may be particularly affected in Alzheimer Disease with
psychosis. It is likely that these different disorders
will have differing pathologic processes in the vulnerable cortical circuits,
though some overlapping mechanisms are possible. Unlike schizophrenia, Alzheimer
Disease is characterized by prominent neuropathologic features, senile plaques
and neurofibrillary tangles. We have not found, however, that these pathologic
lesions are increased in selected brain regions in those individuals who also
develop psychosis. In contrast, brain concentrations of a marker of neuron
integrity (NAA), which are reduced in subjects with schizophrenia, are also
reduced in subjects with psychosis during Alzheimer Disease (see
Sweet et al). Currently our lab is focused on understanding the
genetic mechanisms
underlying psychosis risk in Alzheimer Disease. We are not conducting
studies of post-mortem Alzheimer disease tissue at this time.
|
