Biology Reference
In-Depth Information
There is an opinion that binding is performed by cortico-cortical connections in the “global
workspace”, which consists of the sensory upstream unimodal, downstream unimodal,
heteromodal and limbic neocortical zones (Baars, 2002; Mesulam, 1998). We assume that C-
BG-Th-C loops and dopamine favours the binding and its attentional enhancement due to
following reasons. Release of dopamine lasts during 100-200 ms. Therefore it might support
asynchronous selection and simultaneous conjunction of neuronal patterns representing
different features of visual stimulus in numerous cortical areas. Interdepending changes in all
stages of processing in diverse C-BG-Th-C loops is promoted by existence of not only
reciprocal but also non-reciprocal connections between dopaminergic cells and striatal loci
that belongs to different cortico-BG circuits (Haber, 2003). Based on the non-reciprocal
projections dopaminergic cells, which influence processing in the higher-order C-BG-Th-C
loops, could also influence activity reorganization in the lower-order C-BG-Th-C loops. In
addition, dopaminergic neurons from both VTA and SNc are projected as into the NAcc, as
into the dorso-lateral striatum (Lynd-Balta and Haber, 1994) (Fig. 1). Thus divergent
dopaminergic projections can simultaneously promote processing in C-BG-Th-C loops that
analyze diverse attributes of visual stimulus even though the attention was not especially
attracted to all attributes.
I
NTERPRETATION OF
S
OME
A
TTENTIONAL
E
FFECTS
BY
P
ROPOSED
M
ECHANISM
Elaboration of a new attentional model is reasonable, if it allows explain some
experimentally revealed effects which mechanisms were unclear from point of view of
commonly accepted mechanism of attention, that is based on feedback connections from
higher to lower cortical areas and then to reticular thalamic nucleus. For example, this
mechanism cannot explain data denoting the disinhibition as a mechanism of attentional
strengthening of visual cortical responses (Mehta et al., 2000) since disinhibition requires a
chain of inhibitory cortical interneurons but their number is very small (less than 5%). Except
for that, it is unclear how targets for disinhibition or inhibition could be chosen by attention
taking into account significant convergence and divergence of interconnections between
interneurons and pyramidal cells. In our model, attentional strengthening of visual cortical
responses is in principle the result of disinhibition of thalamic cells that increases excitation
of neocortical neurons, and initial neuronal response itself determines the choice of cells
which activity must be increased.
From common point of view it is unclear why the attention directed on a certain attribute
of a stimulus strengthens responsivity of neurons preferring this attribute, and suppresses
reactions of neurons for which other attributes are preferable (Martinez-Trujillo and Treue,
2004). It was also obscured, why responses to ignored stimuli are attenuated (O'Connor et al.,
2002; Treue and Maunsell, 1999). There is an opinion that various mechanisms underlie these
effects (Hillyard et al., 1998). In contrast, in our model, the unified mechanism underlies both
these effects. If stimulus evokes dopamine release and thus can attract attention, dopamine-
dependent synaptic modifications promote disinhibition of the thalamus and subsequent
increase of neocortical responses, while the lack of dopamine and therefore the absence of
attention leads to rise of thalamic inhibition by the BG and subsequent suppression of
