P-126 Spiking network models of developing mouse prefrontal cortex

Background: The prefrontal cortex (PFC) is a hub of cognitive processes
and exhibits prominent synchronization in the gamma frequency
band (30–80Hz). Disruptions in the performance and
development of these oscillations are related to models of
schizophrenia and autism.
It has been shown in the mouse PFC that fast oscillations become
prominent in the second week of postnatal development
(Bitzenhofer et al., 2020). The peak network frequency starts from
the beta frequency range (15–30Hz) and increases until stabilizing
in the gamma range in the fourth postnatal week.
Objective: Our objective is to identify potential drivers of peak frequency
acceleration in the PFC local field potential through the simulation
of recurrent cortical networks whose biophysical parameters
change over development as reported in experimental data.
Methods: We first implemented a ‘‘mature” recurrent cortical model
using a network of leaky integrate-and-fire neurons based on parameter
values found in the literature for gamma oscillations models
(Brunel and Wang, 2003; Cavallari et al., 2014; Martínez-Cañada
et al., 2021; Mazzoni et al., 2008). The conductance-based model
network consists of 4000 excitatory neurons and 1000 inhibitory
connected with a 20% probability. We estimated the LFP generated
from this simulated network as the sum of the absolute values of
the synaptic currents (Mazzoni et al., 2008). This model generated
gamma oscillations with a spectral peak at 70Hz.
To mimic the effects of development, we then altered the following
parameters across simulations: the synaptic time constants, the E/I
balance, and the membrane electrical properties. The parameter
deviations from the mature model were analyzed by changing
parameters one at a time and the range of values analyzed was chosen
from published results regarding mouse PFC development.
Results: Changing GABA_A synaptic decay time constants decreased
the peak frequency from 70Hz at the mature state to 20Hz using the
‘‘youngest” values of GABA_A rise and decay times (Bosman et al.,
2005; Doischer et al., 2008; Kroon et al., 2019a), in agreement with
the experimental data (Bitzenhofer et al., 2020). Modifying AMPA
synaptic times (Kroon et al., 2019b) slightly decreased the peak frequency.
Altering the excitation-inhibition balance towards excitation
(Chini et al., 2021) modulated the intensity of the spectral
power at the peak frequency but did not change the peak frequency
itself. Modifying the membrane capacitance or the membrane time
constant (Bosman et al., 2005; Doischer et al., 2008) increased the
peak frequency, opposite from the shift seen in experimental data.
Conclusions: The computational modeling study suggests that
changes in the network peak frequency as seen in developing mouse
PFC could be explained principally by changes in the GABA_A synaptic
dynamics. These results could help identify the molecular mechanisms
underlying the physiological and pathological gamma
development.