CA1 network model: interneuron contributions to epileptic deficits (Shuman et al 2020)

 Download zip file   Auto-launch 
Help downloading and running models
Accession:256311
Temporal lobe epilepsy causes significant cognitive deficits in both humans and rodents, yet the specific circuit mechanisms underlying these deficits remain unknown. There are profound and selective interneuron death and axonal reorganization within the hippocampus of both humans and animal models of temporal lobe epilepsy. To assess the specific contribution of these mechanisms on spatial coding, we developed a biophysically constrained network model of the CA1 region that consists of different subtypes of interneurons. More specifically, our network consists of 150 cells, 130 excitatory pyramidal cells and 20 interneurons (Fig. 1A). To simulate place cell formation in the network model, we generated grid cell and place cell inputs from the Entorhinal Cortex (ECLIII) and CA3 regions, respectively, activated in a realistic manner as observed when an animal transverses a linear track. Realistic place fields emerged in a subpopulation of pyramidal cells (40-50%), in which similar EC and CA3 grid cell inputs converged onto distal/proximal apical and basal dendrites. The tuning properties of these cells are very similar to the ones observed experimentally in awake, behaving animals To examine the role of interneuron death and axonal reorganization in the formation and/or tuning properties of place fields we selectively varied the contribution of each interneuron type and desynchronized the two excitatory inputs. We found that desynchronized inputs were critical in reproducing the experimental data, namely the profound reduction in place cell numbers, stability and information content. These results demonstrate that the desynchronized firing of hippocampal neuronal populations contributes to poor spatial processing in epileptic mice, during behavior. Given the lack of experimental data on the selective contributions of interneuron death and axonal reorganization in spatial memory, our model findings predict the mechanistic effects of these alterations at the cellular and network levels.
Reference:
1 . Shuman T, Aharoni D, Cai DJ, Lee CR, Chavlis S, Page-Harley L, Vetere LM, Feng Y, Yang CY, Mollinedo-Gajate I, Chen L, Pennington ZT, Taxidis J, Flores SE, Cheng K, Javaherian M, Kaba CC, Rao N, La-Vu M, Pandi I, Shtrahman M, Bakhurin KI, Masmanidis SC, Khakh BS, Poirazi P, Silva AJ, Golshani P (2020) Breakdown of spatial coding and interneuron synchronization in epileptic mice. Nat Neurosci 23:229-238 [PubMed]
Citations  Citation Browser
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network;
Brain Region(s)/Organism: Hippocampus;
Cell Type(s): Hippocampus CA1 pyramidal GLU cell; Hippocampal CA1 CR/VIP cell; Hippocampus CA1 axo-axonic cell; Hippocampus CA1 basket cell; Hippocampus CA1 basket cell - CCK/VIP; Hippocampus CA1 stratum oriens lacunosum-moleculare interneuron ; Hippocampus CA1 bistratified cell;
Channel(s): I A; I h; I K,Ca; I K; I CAN; I M; I Sodium; I_AHP; I Calcium;
Gap Junctions:
Receptor(s): AMPA; GabaA; GabaB; NMDA;
Gene(s):
Transmitter(s):
Simulation Environment: NEURON; Brian;
Model Concept(s): Spatial Navigation;
Implementer(s): Chavlis, Spyridon [schavlis at imbb.forth.gr]; Pandi, Ioanna ; Poirazi, Panayiota [poirazi at imbb.forth.gr];
Search NeuronDB for information about:  Hippocampus CA1 pyramidal GLU cell; GabaA; GabaB; AMPA; NMDA; I A; I K; I M; I h; I K,Ca; I CAN; I Sodium; I Calcium; I_AHP; 
# README file for Shuman et al, 2019 Breakdown of spatial coding and interneuron synchronization in epileptic mice. Nature Neuroscience

# UPDATED: MArch 20, 2021: all scripts are running under python 3.

# for more information, refer to the comments inside the scripts or contact me in: chavlis [DOT] spiros [AT] gmail [DOT] com

# Scripts' author: S. Chavlis, PhD, I. Pandi, M.Sc.


#################### INPUT CREATION #################################################################################
# First you have to create the Inputs, go in make_inputs_linear_track directory
cd make_inputs_linear_track

# In a command line execute

python make_grid_like_inputs.py <run_number>
python sp_make_place_inputs.py <run_number>
python glim_shuf_new_noisy.py <run_number> <desynch_level> <jitter_source>

# <run_number> is a specific run from one edge of the track to the other. To replicate the figures one needs 10 runs
# <desynch_level>: ms of desynchronization
# <jitter_source>: which input to randomize. Valid options: EC or CA3. Use EC to replicate the paper figures.

# e.g., python glim_v2_prelearning.py 1 20 EC

# Then enter background_noise directory
cd ../background_noise

# create the background noise by executing 

python poisson_input.py <total_number_of_runs> <poisson_rate>  # e.g., python poisson_input.py 1 5 --> creates run1 poisson random noise with lambda 5 Hz


#################### MAIN SIMULATIONS #######################################################################
# return to main directory
cd ../

# Compile all mechanisms (mod files)
nrnivmodl mechanisms/

# Run the simulation
./x86_64/special -nogui -nopython -c n_runs=<run_number> -c n_trials=<virtual_mouse_id> -c desynch=<desynch_level_in_ms> -c n_neuron=<deletion_type> -c factor=<reduction_factor> Network_CA1.hoc

# e.g., ./x86_64/special -nogui -nopython -c nruns=1 -c ntrials=1 -c n_neuron=0 -c desynch=0 -c factor=1 Network_CA1.hoc

# to replicate the results of the paper you need 10 runs/trial and 10 trials and all possible deletions (see below)

# Valid deletions: 
# Control: All connections and cells, default  -- option: 0
# SOMred:  SOMs are removed by a specific percentage  -- option: 1
# PVred:   PVs are removed by a specific percentage  -- option: 2
# Desynch: Desynchronization of EC/CA3 inputs by a specific amount  -- option: 3
# ALL:     SOMred, PVred and Desynch simultaneously  -- option: 4
# SOMdel:  All PVs are removed  -- option: 5
# PVdel:   All PVs are removed  -- option: 6

** Notice: it is highly recommended to run all inputs first, and then run the simulations.
# Output of the simulation is saved into Simulation_Results/

#################### ANALYSIS OF LOCOMOTION DATA BEFORE PROCEEDING #####################################################

# First, one needs to extract the spiketimes for neurons in order to analyze them
# Go to AnalysisRawData directory
cd AnalysisRawData

# Exctract spike times

python spiketimes_analysis.py <neuron_type> <deletion_type> <number_of_trial> <number_of_run>

# Valid deletions_types: 
# Control
# SOMred
# PVred
# Desynch
# ALL
# SOMdel
# PVdel

# Valid <neuron_type> values:

# _pvsoma_   : Pyramidal cells
# _aacell_   : Axoaxonic cells
# _bcell_    : Basket cells
# _bscell_   : Bistratified cells
# _olmcell_  : OLM cells
# _vipcck_   : VIP/CCK cells
# _vipcr_    : VIP/CR PVM cells
# _vipcrnvm_ : VIP/CR NVM cells

# After the analysis for all trials, runs and deletions execute:

python all_path_all_spiketimes.py <deletion_type> # e.g., python all_path_all_spiketimes.py Control

# This will create the subfolder final_results/metrics_permutations
# where the spiketimes and the path for all cases is stored (for better handling)

# Permutations for all cells to find spatial information and stability null distributions
python permutations_analysis_peyman.py <virtual_mouse> <pyramidalID> <deletion_type>

# Data save for using in GraphPad Prism and basic plotting
python analysis_path.py <virtual_mouse> <deletion_type>

python all_trials_paper_all.py

python all_trials_per_animal.py

# for more information, refer to the comments inside the scripts or contact me in: chavlis [DOT] spiros [AT] gmail [DOT] com