Logo Logo
Help
Contact
Switch language to German
Structure and dynamics of the corticothalamic driver pathway in the mouse whisker system
Structure and dynamics of the corticothalamic driver pathway in the mouse whisker system
To generate a sensory percept of the environment, the brain needs to analyze and integrate spatially and temporally distributed sensory signals. Consequently, sensation on a neuronal basis is a distributed, non-linear and dynamic process. Following sensory receptor activation the signal travels through many brain regions wherein the pathway is split, loops back onto itself and joins together with others. At each step, neurons dynamically transform and filter the signal. To understand how the brain arrives at a sensory percept, it is therefore essential to determine the neuronal connectivity along the processing chain, the stimulus specificity of responses as well as the input-output transformations at each station. An interesting model system for investigating these dynamical processes is the rodent whisker system. Rodents can solve highly complicated tasks with their whiskers alone, distributed receptors at the follicles require spatial integration and rhythmic movements suggest temporal processing components. The posterior group nucleus of the thalamus (PO) is in a key position of the whisker sensory system. In addition to being part of the ascending paralemniscal pathway it is mainly driven by somatosensory barrel cortex (BC) and projects to many cortical and subcortical areas. Due to its poor excitability by whisker deflections, its function is unclear. The origin of the corticothalamic drive onto PO neurons are ‘thick-tufted’ layer 5B cortical neurons, which have large synaptic terminals in thalamus. One of those synapses alone has a strong influence on postsynaptic target neurons – a very unusual property for cortical synapses. Here, using quantitative anatomy, in vivo electrophysiology and optogenetics I characterize the organization and input-output computations along the BC L5B to PO pathway. Using a dual anterograde tract tracing approach and large scale anatomical reconstructions we demonstrate that BC L5B synaptic boutons divide PO in 4 subregions with different projection parameters. The lateral area (POm lateral) receives most boutons with the highest density. Additionally, L5B neurons innervate two inhibitory nuclei in thalamus and midbrain that both inhibit PO. In all 6 regions we report map specific projections, with different map orientations, showing that somatotopic projections are the rule in these cortico-subcortical projections. Next we investigated the L5B to POm action potential transfer efficacy during spontaneous slow oscillations in anesthesia. Using pharmacology and cell-type specific optogenetics we show that cortical activity is necessary and L5B activation is sufficient to evoke large excitatory postsynaptic potentials (EPSPs) in POm, typical for L5B inputs. Simultaneous cortical local field potential and L5B as well as POm juxtasomal recordings demonstrate that the gain of action potential transmission is high following periods of relative cortical silence, but dynamically decreases during periods of higher cortical activity. Isolation of individual EPSPs allowed us to determine the frequency dependent adaptation of the L5B to POm synapse in vivo. We determined that approximately half of the recorded POm neurons follow a simple rule of EPSP adaptation, suggesting that the subthreshold activity in these neurons originates from a single active L5B input. Using two independent modeling approaches, we determined that on average POm neurons receive 2-3 functional inputs from BC L5B. Finally we investigated how whisker deflection signals reach POm. We found that POm neurons fall into two groups. Approximately one third of the recorded neurons were activated at a relatively short latency by large EPSPs and fired action potentials following whisker stimulation. All neurons had long latency sub- and suprathreshold responses, due to Up-state initiation by the whisker stimulation. POm whisker responses were entirely dependent on cortex and were blocked by optogenetic cortical inactivation. Taken together we quantified the anatomical and physiological properties of the L5B to POm projection. The connection is sparse, parallel, strong and the dominant input for POm spontaneous activity as well as whisker evoked responses. Its gain is dynamically regulated and depends on cortical activity states.
Not available
Sumser, Anton
2016
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Sumser, Anton (2016): Structure and dynamics of the corticothalamic driver pathway in the mouse whisker system. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)
[img]
Preview
PDF
Sumser_Anton.pdf

16MB

Abstract

To generate a sensory percept of the environment, the brain needs to analyze and integrate spatially and temporally distributed sensory signals. Consequently, sensation on a neuronal basis is a distributed, non-linear and dynamic process. Following sensory receptor activation the signal travels through many brain regions wherein the pathway is split, loops back onto itself and joins together with others. At each step, neurons dynamically transform and filter the signal. To understand how the brain arrives at a sensory percept, it is therefore essential to determine the neuronal connectivity along the processing chain, the stimulus specificity of responses as well as the input-output transformations at each station. An interesting model system for investigating these dynamical processes is the rodent whisker system. Rodents can solve highly complicated tasks with their whiskers alone, distributed receptors at the follicles require spatial integration and rhythmic movements suggest temporal processing components. The posterior group nucleus of the thalamus (PO) is in a key position of the whisker sensory system. In addition to being part of the ascending paralemniscal pathway it is mainly driven by somatosensory barrel cortex (BC) and projects to many cortical and subcortical areas. Due to its poor excitability by whisker deflections, its function is unclear. The origin of the corticothalamic drive onto PO neurons are ‘thick-tufted’ layer 5B cortical neurons, which have large synaptic terminals in thalamus. One of those synapses alone has a strong influence on postsynaptic target neurons – a very unusual property for cortical synapses. Here, using quantitative anatomy, in vivo electrophysiology and optogenetics I characterize the organization and input-output computations along the BC L5B to PO pathway. Using a dual anterograde tract tracing approach and large scale anatomical reconstructions we demonstrate that BC L5B synaptic boutons divide PO in 4 subregions with different projection parameters. The lateral area (POm lateral) receives most boutons with the highest density. Additionally, L5B neurons innervate two inhibitory nuclei in thalamus and midbrain that both inhibit PO. In all 6 regions we report map specific projections, with different map orientations, showing that somatotopic projections are the rule in these cortico-subcortical projections. Next we investigated the L5B to POm action potential transfer efficacy during spontaneous slow oscillations in anesthesia. Using pharmacology and cell-type specific optogenetics we show that cortical activity is necessary and L5B activation is sufficient to evoke large excitatory postsynaptic potentials (EPSPs) in POm, typical for L5B inputs. Simultaneous cortical local field potential and L5B as well as POm juxtasomal recordings demonstrate that the gain of action potential transmission is high following periods of relative cortical silence, but dynamically decreases during periods of higher cortical activity. Isolation of individual EPSPs allowed us to determine the frequency dependent adaptation of the L5B to POm synapse in vivo. We determined that approximately half of the recorded POm neurons follow a simple rule of EPSP adaptation, suggesting that the subthreshold activity in these neurons originates from a single active L5B input. Using two independent modeling approaches, we determined that on average POm neurons receive 2-3 functional inputs from BC L5B. Finally we investigated how whisker deflection signals reach POm. We found that POm neurons fall into two groups. Approximately one third of the recorded neurons were activated at a relatively short latency by large EPSPs and fired action potentials following whisker stimulation. All neurons had long latency sub- and suprathreshold responses, due to Up-state initiation by the whisker stimulation. POm whisker responses were entirely dependent on cortex and were blocked by optogenetic cortical inactivation. Taken together we quantified the anatomical and physiological properties of the L5B to POm projection. The connection is sparse, parallel, strong and the dominant input for POm spontaneous activity as well as whisker evoked responses. Its gain is dynamically regulated and depends on cortical activity states.