| dc.description.abstracteng | The barrel cortex is the part of the rodent neocortex where the tactile information from the
whiskers is processed. In the barrel cortex, each whisker is represented by a single cortical
column isomorphic to the whiskers on the snout. Since the cortical column is considered to be
the basic unit of all computations, the barrel cortex became a fruitful model for the study of
columnar microcircuits in the mammalian neocortex. The majority of cortical neurons are
excitatory (≈85%) and release the neurotransmitter glutamate while only roughly 15% are
inhibitory and release γ-amino butyric acid (GABA). These cell types are intricately integrated
into the local microcircuit to maintain the balance between excitation and inhibition. The
somatostatin- (SST) expressing cells are one subpopulation of GABAergic interneurons (IN)
that are effectively controlling excitatory cells. They are targeted by other GABAergic IN,
namely parvalbumin- (PV) and vasoactive intestinal polypeptide- (VIP) expressing cells,
forming a circuit motif, which is known as disinhibition. By targeting SST cells, PV and VIP cells
are able to reduce the inhibitory drive on excitatory cells. Although disinhibitory circuits are
extensively studied in modern neuroscience, little is known about the disinhibitory circuitry of
L5 SST cells in mouse barrel cortex. L5 is, according to the canonical microcircuit, the main
output layer of the barrel cortex. Furthermore, it is unknown whether individual presynaptic PV
or VIP cells are able to modulate the output of postsynaptic SST cells. My aim is to extend our
knowledge of the disinhibitory circuitry of SST cells and to analyse the functionality of the PV
to SST and the VIP to SST circuit motifs. Therefore, this study is divided into two chapters.
In the first chapter, three novel disinhibitory circuits in mouse barrel cortex targeting
L5 SST cells are described. Intralaminar whole cell patch clamp recordings revealed that both
PV and VIP cells reliably target L5 SST cells (connectivity rates: ≈25% for PV to SST; ≈23%
for VIP to SST). Translaminar recordings demonstrated that also L2/3 VIP cells are connected
to L5 SST cells with a similar connection probability as the intralaminar connection (≈30%).
However, L2/3 PV cells were not frequently connected to L5 SST cell as only one out of 48
analysed pairs (≈2.1%) showed a connection. PV and VIP connections displayed cell type-
specific differences in unitary synaptic properties and short-term synaptic plasticity. PV
connections were larger in amplitude and shorter in latency. Moreover, PV connections were
depressing at all tested frequencies (1, 8, and 40 Hz) whereas VIP connections were
facilitating but only at high-frequency stimulation (40 Hz). In addition, we found reciprocally
connected pairs in all three analysed connections. This opens a window of opportunity for SST
cells to also inhibit their own inhibitors, adding another level of complexity to the dis/inhibitory
circuitry in mouse barrel cortex.
In the second chapter, we analysed the functionality of the PV to SST and VIP to SST
cell circuits in L2/3 of mouse barrel cortex. While inducing action potential firing in postsynaptic
SST cells, we analysed the effect of simultaneous presynaptic PV and VIP cell supratreshold
stimulation. The known or hypothesized differences in unitary synaptic properties and
subcellular distribution of synapses let us assume differences in action potential modulation by
PV or VIP cells on postsynaptic SST target cells. However, stimulation of individual PV and of
individual VIP cells was able to significantly reduce action potential firing in postsynaptic SST
cells. Within both connections, we saw a large variety of effect strengths. When comparing
effect strength of PV to SST and VIP to SST cell connections there was no significant
difference in spike loss or inter-spike interval increase. When normalizing effect strength for
the number of presynaptic action potentials, the VIP effect was significantly stronger than of
PV cells. A short stimulation of PV and VIP cells, when appropriately timed, was able to
significantly delay firing of postsynaptic SST cells. However, again there was no significant
difference in effect strength between PV to SST and VIP to SST cell connections.
Morphological reconstructions and subsequent putative contact site analysis did not reveal
significant differences between both groups. In line with other observations, putative contact
site location did not correlate with spike loss in reconstructed pairs. Our findings question the
concept of input versus output control by specific inhibitory cell types.
In summary, we discovered novel intralaminar disinhibitory circuit motifs of L5 SST,
homologous to their supragranular counterparts. By investigating translaminar circuit motifs,
we found a VIP cell-specific circuit from L2/3 to L5 SST cells, which displayed similar unitary
synaptic properties and short-term plasticity as the intralaminar L5 to L5 circuit. Furthermore,
we were able to show that, against current concepts, activation of individual PV and VIP cells
was able to decrease the output of postsynaptic SST cells. | de |