| Summary: | <p>Charting connectivity between neurons in the brain is one of the main goals of circuit
neuroscience. Despite great advances in our understanding of cortical circuit
organisation, comparatively little attention has been devoted to studying how the
intricate cortical connectivity emerges. Current views are that early synaptic
connectivity within the cortex and between thalamus and cortex is largely mediated
via silent synapses. Early patterns of network activity synchronise firing in pre- and
post-synaptic elements of circuits and promote correlated activity, enabling synaptic
plasticity and maturation via unsilencing. However, little is known about which
circuits may be in place to promote such patterns of activity. In my thesis I used
optical stimulation and electrophysiology to clarify circuits involved in the maturation
of the thalamocortical circuit in the mouse barrel cortex.</p>
<p>In Chapter 3, I attempted to use optogenetics to study the early organisation and
plasticity of thalamocortical inputs by delivering ChR65 (channelrhodopsin 65) to the
embryonic thalamus via in utero electroporation. Following a detailed
characterisation of the response properties of ChR65-positive developing neurons to
optical stimulation, I was able to demonstrate that expression levels strong enough to
elicit laser-mediated spiking are only attainable post-P10, after the critical period for
thalamocortical plasticity. I focused on characterising thalamic inputs onto excitatory
neurons in L4, L5b and the subplate (SP), finding a focused barrel organisation in L4,
broader horizontal organisation in L5b, and lack of thalamic inputs to subplate
neurons (SPns) after P10.</p>
<p>SPns have been shown to have a key role in early network oscillations and
thalamocortical plasticity. However, relatively little is known about the cortical
circuits in which they are embedded early in development. In Chapter 4, I
characterised intrinsic electrophysiological and network development in the Lpar1-
GFP population of SPns. These cells show highly mature intrinsic properties already
at P6, and receive robust thalamic input from birth, which starts being retracted after
P5. Using glutamate uncaging to map afferent excitatory inputs onto Lpar1-GFP
SPns, I found that these cells start out receiving most of their cortical excitation from
the SP. Between P2 and P3, this recurrent network loses prominence, and L6 becomes
the major source of input. By P5, the organisation of cortical inputs onto Lpar1-GFP
SPns appears stable.</p>
<p>In the Lpar1-GFP line, it is also possible to find fluorescently-labelled neurons in the
cortical plate. These cells are mostly in L5 and L6, and have atypical, non-pyramidal
morphologies. In Chapter 5 I studied these neurons, concluding that they are Lhx6+
Martinotti cells, with axons extending to L1 and arborizing in L4. Lpar1-GFP
Martinotti cells (Lpar1-GFP MCs) receive early, facilitating thalamic input, which is
maintained into maturity. I mapped early cortical inputs onto these cells using
glutamate uncaging and found that whereas early inputs arise from L4, by P10-P15
the totality of input comes from L5b/6. Conversely, I found, by mapping GABAergic
inputs onto L4 spiny stellate neurons throughout development, that these cells initially
derive most of their inhibition from L5b. L4 GABAergic inputs are upregulated
throughout development, concomitant to a decrease and eventual elimination of L5b
inhibition onto L4. I propose that Lpar1-GFP MCs and L4 excitatory neurons form a
reciprocal connection. Using infraorbital nerve (ION) sectioning, I found that early
GABAergic innervation is decreased in ION-sectioned animals compared to controls,
suggesting GABAergic synapse formation is experience-dependent. However, the
laminar organisation of GABAergic inputs was identical between the two groups. At
P10-P15, laminar organisation differed, with ION-sectioned animals receiving more
input from L5b and less from L4, compared to controls. Total GABAergic input was
unaltered by this age, and I propose that it was compensated by an upregulation of
input from infragranular layers.</p>
<p>The combined results of my experiments using optical and electrical stimulation
approaches to study the early thalamocortical system revealed highly dynamic circuits
involving thalamic relay neurons, Lpar1-GFP expressing subplate and L5b Martinotti
neurons, and L4 spiny stellate cells. In agreement with recent findings in the adult
literature, my results support a view of emerging thalamocortical processing that is
increasingly less centred in L4, pointing to the key influence of early transient circuits
involving subplate neurons and infragranular Martinotti cells. These two pioneer cell
types seem to integrate into cortical and thalamocortical circuits early in development
and provide two di-synaptic pathways linking the thalamus to L4, one glutamatergic
(via Lpar1-GFP SPns) and one GABAergic (via L5b Lpar1-GFP MCs). These
transient pathways could collaborate to regulate early rhythmic patterns of activity in
L4 that are fundamental for maturation of thalamocortical pathways.</p>
|
|---|