HAL Id: hal-00527058
https://hal.archives-ouvertes.fr/hal-00527058
Submitted on 18 Oct 2010
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Contrast gain control and cortical TrkB signaling shape
visual acuity
J Alexander Heimel, M Hadi Saiepour, Sridhara Chakravarthy, Josephine M
Hermans, Christiaan N Levelt
To cite this version:
J Alexander Heimel, M Hadi Saiepour, Sridhara Chakravarthy, Josephine M Hermans, Christiaan N
Levelt. Contrast gain control and cortical TrkB signaling shape visual acuity. Nature Neuroscience,
Nature Publishing Group, 2010, �10.1038/nn.2534�. �hal-00527058�
1
Contrast gain control and cortical TrkB signaling shape visual acuity
J. Alexander Heimel, M. Hadi Saiepour, Sridhara Chakravarthy, Josephine M. Hermans & Christiaan N.
Levelt
Molecular Visual Plasticity Group, Netherlands Institute for Neuroscience, an institute of the Royal Netherlands Academy of Arts and
Sciences, Meibergdreef 47, 1105 BA Amsterdam, The Netherlands. Correspondence should be addressed to J.A.H.
(heimel@nin.knaw.nl)
During development, aging and in amblyopia, visual acuity is much below the limitations set by
the retina. Expression of brain-derived neurotrophic factor (BDNF) in the visual cortex is
reduced in these situations. We have tested the hypothesis that TrkB/BDNF regulates cortical
visual acuity in adult mice. We found that genetically interfering with TrkB/BDNF signaling in
pyramidal cells in the mature visual cortex reduced synaptic strength and resulted in a loss of
neural responses to high spatial frequency stimuli. Responses to low spatial frequency stimuli
were unaffected. This selective loss was not accompanied by a change in receptive field sizes or
plasticity but exclusively by a reduction in apparent contrast. We demonstrate that a
dependence on spatial frequency in the Heeger normalization model explains this selective
effect of contrast reduction on high resolution vision, and argue that it involves contrast gain
control operating within the visual cortex.
Resubmission date: March 5
th
, 2010
2
The limits of visual acuity and contrast sensitivity are set by the eye, but what we perceive is determined by the visual
cortex
1
. In healthy, mature people and animals, the visual acuities of the retina and the cortex are well-matched
2
, but this
match is neither automatic nor unbreakable. Differences between cortical and retinal acuity are most apparent during
development, when cortical acuity continues to rise after retinal development is completed
3
, and during aging, when
behavioral acuity falls even without obvious changes in the eye or the thalamus
4
. Differences also occur as a result of
cortical injury or erroneous development. This is the case with amblyopia, the most prevalent (2-4%) visual impairment
in young people. Amblyopia is a reduced psychophysical acuity in one or both eyes. It is believed to be caused by
deficient processing in the visual cortex
5
, but the mechanisms underlying the dissociations of retinal and cortical acuity
in amblyopia and in the healthy aging and developing brain are unclear. Interestingly, there is a good match between
changes in the cortical expression level of brain-derived neurotrophic factor (BDNF) and changes in visual acuity.
During development, acuity and BDNF levels rise
6
, while both slowly decrease with age
4,7
. This relationship between
BDNF and acuity also holds for experimentally induced amblyopia. BDNF mRNA and protein levels
8,9
and acuity
10
in
the primary visual cortex (V1) responding to a monocularly deprived eye are all below normal. In amblyopic rats
receiving environmental enrichment
11
or antidepressant treatment
12
, increased BDNF expression in the cortex was seen
in parallel to the restoration of visual acuity. Moreover, transgenic mice overexpressing BDNF in the forebrain show a
faster rise of cortical acuity
6
even when reared in darkness
13
. Although there is a wealth of data on the involvement of
BDNF and its main receptor TrkB in neuronal development
14
, synaptic efficacy
15
, -morphology
16
and -plasticity
17,18
, it
has remained unknown how BDNF promotes visual acuity at the coding level and whether BDNF signaling plays a role
in acuity in the mature cortex. For these reasons we studied visual acuity in adult transgenic mice where, after normal
development is completed, cortical TrkB/BDNF signaling is impaired. We found a loss of acuity, caused by a reduction
in apparent contrast. Using a combination of experiments and modeling, we show the involvement of cortical gain
control in the selective loss of responses to visual stimuli with high spatial frequencies and the maintenance of
responses to low spatial frequencies.
RESULTS
Genetic inhibition of TrkB signaling in the adult cortex
To investigate the role of TrkB signaling in cortical acuity in the mature animal we overexpressed a dominant negative
TrkB.T1-EGFP fusion protein
19
in a large proportion of pyramidal cells after the maturation of cortical acuity. This was
achieved by crossing mice carrying a Cre-dependent TrkB.T1-EGFP-transgene under the control of the Thy-1
promoter
16
with G35-3 Cre-recombinase transgenic animals
20
(Supplementary Fig. 1). In G35-3-Cre mice, Cre-
3
recombination is restricted to excitatory neurons in the neocortex, hippocampus and amygdala
20
while the retina
21
,
thalamus and superior colliculus are unaffected. In TrkB.T1-EGFP x G35-3 double transgenic animals, TrkB.T1-EGFP
was expressed in excitatory neurons of the hippocampus and pyramidal neurons of the neocortex (Fig. 1a). Expression
was absent in layer 4 of the neocortex (Fig. 1b), probably due to the lack of Thy-1 promoter activity in this layer. No
transgene expression was detected in the locus coeruleus or basal forebrain neuromodulatory regions. Transgene
expression started around 5 weeks after birth, after the end of the critical period for ocular dominance plasticity
22
and
after maturation of visual acuity
6,10
.
Synaptic efficacy is reduced by overexpression of TrkB.T1
Before investigating the effects of impaired TrkB signaling on visual processing in vivo, we assessed its effects at the
synapse level. BDNF signaling is known to modulate synaptic transmission
15,18
, and overexpression of TrkB.T1 inhibits
BDNF-induced enhancement of excitatory transmission
23
. We therefore wanted to know whether interfering with TrkB
signaling in pyramidal neurons of adult V1 altered synaptic transmission. We measured the response of pyramidal cells
to electrical stimulation in slices of adult visual cortex of wild-type and TrkB.T1-EGFP-expressing mice. Because there
was no expression of the transgene in layer 4 and TrkB/BDNF signaling has been implicated in both pre- and
postsynaptic modulation of synaptic strength, we first studied the intralaminar connections of layer 2/3 to layer 2/3
neurons, which are the most abundant synapses in layer 2/3
24
. We recorded intracellular responses in layer 2/3
pyramidal neurons to extracellular stimulation by an electrode displaced 200 μm horizontally from the recording
electrode in the same layer (Fig. 1c-d). These evoked responses were lower in TrkB.T1-EGFP-expressing animals than
in wild-type mice (response to 200 μA, T1: 0.97±0.07 nA, 13 cells; wt: 1.9±0.2 nA, 6 cells; p=0.003, t-test, Fig. 1e)
confirming a reduction in synaptic strength. Spike initiation thresholds were not different (T1: −35.1±0.9 mV; wt:
−34.2±0.5 mV; p=0.5, t-test) nor was paired-pulse facilitation, a predominantly presynaptic phenomenon (T1: 1.18±
0.04; wt: 1.19± 0.03; p=0.9, t-test). Although the unchanged paired-pulse ratio did not fully exclude a presynaptic cause
of the reduced synaptic strength, it did suggest a primarily post-synaptic phenotype. If so, we would expect a change in
layer 4 to 2/3 connections as well. Therefore, we also recorded the local field potential (LFP) in layer 2/3 in response to
extracellular stimulation by an electrode positioned in layer 4 (Fig. 1f-g). Evoked LFP responses are an indication of
combined synaptic activity in the recording electrode's vicinity. In TrkB.T1-EGFP-expressing animals, responses were
well below those in wild-type mice (response to 260 μA, T1: 1.50±0.14 mV, 14 slices, 6 mice; wt: 1.93±0.12 mV, 12
slices, 5 mice; p=0.03, t-test, Fig. 1h). The paired-pulse ratio was again unchanged (T1: 0.89±0.04; wt: 0.98±0.09;
p=0.33, t-test). Together these results strongly suggest that in the TrkB.T1-EGFP expressing mice, synaptic transmission
4
to layer 2/3 neurons was reduced through a postsynaptic mechanism.
Evoked LFPs are dominated by excitatory synaptic transmission, but contain the combination of excitatory and
inhibitory post-synaptic potentials. As chronic changes in BDNF signaling have been shown to positively correlate with
the amount of perisomatic inhibition
6,25
by parvalbumin-expressing interneurons we also determined whether the
TrkB.T1-EGFP transgenic animals showed changes in inhibitory inputs using immunohistochemistry. We examined the
parvalbumin-positive boutons around pyramidal cell bodies, which provide the major source of inhibition in the cortex.
This revealed that perisomatic inhibitory synapses were decreased in TrkB.T1-EGFP animals (Supplementary Fig. 2),
because both bouton number (T1: 2707 puncta in 418 cells, 3 mice; wt: 9048 puncta in 874 cells, 3 mice; p<0.001, t-
test) and average bouton diameter were reduced (T1: 0.623±0.008 μm; wt: 0.715±0.005 μm; p<0.001, t-test). This
finding suggests that the observed decrease in the evoked responses in layer 2/3 neurons was caused by an even larger
decrease in excitatory transmission which was partially compensated for by reduced inhibitory input.
Acuity loss after inhibition of TrkB signaling
We next addressed the question whether the changes in synaptic transmission in adult TrkB.T1-EGFP-expressing mice
were accompanied by a reduction in visual acuity by using optical imaging of intrinsic signal. Example responses to
high contrast phase-reversing sinusoidal gratings of a wild-type mouse are shown in Figures 2a-b. Acuity was defined
as the null response point of a threshold-linear curve fitted to the data (Fig. 2b). TrkB.T1-EGFP-expressing mice had a
strongly reduced acuity (T1: 0.40±0.05 cycles per degree, N=7; wt: 0.54±0.03 cpd, N=18; p=0.01, t-test, Fig. 2c).
Closer inspection of the average response curves in Figure 2d shows that responses to the lowest tested spatial
frequency (0.1 cpd) were not reduced (p=0.6, t-test). Because during development, increasing BDNF levels not only
induce the increase in visual acuity but also the onset of the critical period
6
, we tested the possibility that TrkB.T1-
EGFP expression in the adult visual cortex would affect ocular dominance plasticity. However, no difference in ocular
dominance plasticity was observed between wild-type and TrkB.T1-EGFP-expressing mice (Supplementary Fig. 3).
Acuity loss is caused by a reduction of apparent contrast
There are three functional mechanisms through which the excitatory and inhibitory synaptic changes could cause
reduced acuity, illustrated in Supplementary Figure 4. The first mechanism is the enlargement of receptive field
centers of neurons in the visual cortex. Sampling input from a larger retinal area would reduce the response to a high
spatial frequency grating, because both dark and light areas will fall within single ON-/OFF-subfields of a neuron. The