
Finally, Michael Halassa (Massachusetts Institute of
Technology) presented data from in vivo recordings from
TRN cells. Using optogenetic techniques it was shown that
when TRN cells are activated they evoke bursting behavior
in ventral posterior medial (VPM) thalamic cells. In awake
behaving animals it was demonstrated that some TRN cell
firings are correlated with ongoing spindle behavior and
some are not. Some TRN neurons showed correlated firing
to behavioral tests, while others did not. The latter two results
taken together provide evidence that TRN neurons have
heterogeneous functions.
Short talks
Following a short break in the rejuvenating sun, Matthew
Diamond (SISSA, Trieste, Italy) moderated the f irst set of
short (15 min) talks. First up was Carsten Pfef fer (University
of California San Diego) whose talk addressed the issue of
inhibition of inhibitory cells within the mouse visual cortex.
Using genetic techniques to label distinct GABAergic
populations along with an adenovir us tagged with channel
rhodopsin, the interconnectivity of these cells was addressed.
Using single-cell polymerase chain reaction (PCR) from the
nuclei of the recorded cells it was possible to identify five
classes of GABAergic neurons in layers 2/3 and 5. In sum,
vasointestinal peptide (VIP)
+
neurons predominantly inhib-
ited somatostatin
+
neurons, parvalbumin
+
neurons inhibited
other parvalbumin
+
neurons and pyramidal cells whereas
somatostatin
+
cells inhibited all classes of cells except
somatostatin
+
neurons. Next, Robin Wagener (University
of Medicine, Goettingen, Germany) showed that in the
severely disorganized cortex of the reeler mouse it was still
possible to have intact thalamocortical inputs. Using mole-
cular markers that differentiate cortical laminae it was shown
that thalamocortical axons target layer 4 fated neurons and
using optogenetics it was possible to show that these synapses
can be activated and lead to columnar-like activation in
the reeler mouse. The short talk session was concluded by
Hong Li (Yale University) who demonstrated that the proper
development of barrel cortex requires thalamocortical neuro-
transmission. A double knockout mouse was constructed
where glutamate receptors VGlut1 and VGlut2 were deleted
and it was shown that somatotopy was disrupted when the
resultant brains were stained with Nissl or the thalamocortical
afferents were analyzed despite the fact that within these
animals the trigeminal and thalamic nuclei develop normally.
Further analysis of these animals showed that layer 4 was
reduced in width and it had fewer stellate cells.
Following lunch Dirk Schubert (Radboud University,
the Netherlands) moderated another session of short talks.
Leading off was Matthew Diamond (SISSA, Trieste, Italy)
who defined the issue of what the rodent is sensing through its
whiskers in a neuroethological context, specifically can it
when living in underground tunnels detect vibrations made
by predators above the ground. To that end, a behavioral
test was devised where different vibrations were presented to
the left vs. right whisker pads which the rat was trained to
discriminate. Rats were able to make discriminations at delays
of up to 8 s (humans could perform a comparable task with
delays up to 12 s). Recordings from anterior motor cor tical
regions (M2) showed that some neurons responded to the
go signal and based on their firing rates it was possible to
predict the behavior of the animal. Matthew Evans
(University of Sheffield, UK) considered issues concerning
how the firing rates of cortical neurons can be interpreted
based on different encoding schemes. A key concept was
the idea that sensory systems are tuned to the state of the
environment they are interacting with. To attack this problem
a four-stage program was proposed of measurement (done by
mechanoreceptors), preprocessing (neuronal tuning curve
distributions), representation (neuronal tuning curve shapes),
and inference (cortical outputs). Following the short talks,
Joshua C. Brumberg (Queens College, CUNY) moderated
the annual data blitz where attendees presented their latest
results in quick secession before the meeting adjourned to the
poster session and finally an alfresco dinner by the sea.
Neurovasculature coupling
The second day of Barrels XXVI began with an introduction
and then moderation by David Kleinfeld on the topic of
neurovascular structure and control of brain metabolism.
Serge Charpak (University Paris V) began the first talk of
the morning with work demonstrating the ability of 2-photon
phosphorescence lifetime microscopy to image the partial
pressure of oxygen (PO
2
) at micron-scale resolution in the
olfactory bulb of rats. He found that PO
2
levels peak in
individual erythrocytes and that PO
2
levels between two
peaks represents the PO
2
levels existent in the neuropil.
He presented in vivo data from chronically imaged rats and
found that sniffing an odor selectively resulted in a PO
2
decrease prior to functional hyperemia in the nerve layer
of the glomeruli. He emphasized that the dip that is seen in
functional imaging techniques prior to hyperemia, such as
those seen in fMRI, is reflected in the micron scale within
capillaries and suggested that measuring the prehyperemia
‘‘dip’’ could be a way to gain higher temporal resolution
in functional imaging studies. The second talk of the day was
given by Anna Devor (University of California San Diego
and the Harvard Medical School) who began by noting the
overarching goal of using model systems to understand the
human brain. Using 2-photon microscopy she presented
data that blood vessel dilation speed is specific to cortical
depth and branch order. Specifically, that deeper cortical
layers dilate faster than superficial layers, emphasizing that
although layer 4 is thought to be the most metabolically
active, it does not have the fastest dilating vessels. It was then
demonstrated that local glutamate application caused dilation
of vessels in layer 1 and that astrocyte calcium signaling was
locked to vascular dilation patterns. Interestingly, astrocyte
activity was shown to peak after maximum vasodilatation,
and thus astrocytes appear too slow to control vasculature
behavior. Further, using an IP
3
R2-knockout mouse, in which
astrocyte signaling is blocked in response to mGluR activity,
it was shown that dilation still occurred. Taken together these
data suggest that astrocytes are not playing a direct role
in regulating vasodilation. She then went on to propose a
hypothesis of local signaling within compartments of neurons
that may selectively secrete prostaglandins. Next, Catherine
Hall (University College London) differentiated the role of
2 P. Chu et al. Somatosens Mot Res, Early Online: 1–6
Somatosens Mot Res Downloaded from informahealthcare.com by Queens College on 02/06/14
For personal use only.