(158b) CAM/Camkii Mediated Remodeling of PDZ Domains and the Inhibitory Influence of Syngap in the Post-Synaptic Density

font-family:" times new roman>CAM/CAMKII mediated remodeling of PDZ
domains and the inhibitory influence of SynGAP in the post-synaptic density

Patrick
A. Giolando, Barrett F. Davis, Matthew C. Pharris, Tamara L. Kinzer-Ursem

Weldon
School of Biomedical Engineering, Purdue University

Keywords: CAMKII, PSD-95, synGAP

" times new roman>Introduction:
Approximately, 1 in every 60 children in the United States have been diagnosed
with Autism Spectrum Disorder (ASD) (1). Of the hundreds of genes associated
with ASD, those most highly correlated with the disorder produce proteins
involved in neuronal synapse function. Among these, SYNGAP1 is a major
constituent of the post-synapse and is a downstream target of the excitatory
neurotransmitter glutamate (via NMDA receptor activation). Equally highly
expressed is the scaffolding protein PSD-95, which binds synGAP at all three of
its PDZ domains. synGAP has been found to occupy up to 15% of these domains at
one time (2), potentially restricting the binding of other proteins to these
sites including TARP and LRRTM proteins involved in regulating synaptic
functions. To gain a full understanding of how synGAP competes with other
synaptic proteins for binding to PDZ domains, we designed a computational model
that captures the dynamic binding of synGAP and other key synaptic proteins
(TARP, and LRRTM) with PDZ domains. Our model focuses on two events influencing
the binding of synGAP to PDZ domains: the binding of Ca²⁺/Calmodulin
(Ca²⁺/CaM) to synGAP, and phosphorylation of synGAP by Ca²⁺/calmodulin-dependent
protein kinase II (CaMKII). Both of these calcium
dependent events act as a switch to free up PDZ domains and allow for receptors
to bind and be held at the synapse.

Materials
and Methods:

The
dynamic interaction of proteins within the post-synaptic density, including synGAP,
CaM, CaMKII, TARP, and LRRTM, is modeled according to mass action kinetics. The
resulting system of ordinary differential equations (ODEs) is numerically
solved in Python 3.7. The model is parameterized using literature values. A
global sensitivity analysis is performed using Latin hypercube sampling and
partial rank correlation coefficients (LHS/PRCC).

Results
and Discussion:

To
verify our model, we compared the concentration of synGAP bound to PSD-95 at
basal conditions with known values and found synGAP bound up to 15% of the
PSD-95 domains, in agreement with previous studies. The model accurately
predicted the reduction in synGAP bound to PSD-95 by the activation of CaM and
CaMKII, which previous in vitro studies determined to be 20% and 54%
respectively. To test whether the phosphorylation of synGAP by CaMKII causes a
significant rearrangement of proteins bound to PDZ domains, we stimulated the
activation of CaMKII by Ca²⁺/CaM at a 100 Hz Ca²⁺ spike
frequency for one second, to mimic long-term potentiation (LTP), and at 10 Hz
for long-term depression (LTD), similar to typical experimental protocols. We
find significant rearrangement in the binding patterns of synGAP, TARP and
LRRTM occur on the order of seconds and thus on a timescale relevant for AMPAR
receptor trafficking to the synapse.

Conclusion:

The
use of a system of ODEs to model the reaction kinetics among post-synaptic
proteins and PSD-95 allows for a more complete understanding of the role of
synGAP, Ca²⁺/CaM, and CaMKII on inhibiting or stimulating change of
protein composition within the synapse. In the future this model will be used
to determine the effect of disease states, particularly SYNGAP1 mutations, on
the composition of proteins in the synapse, and the anchoring of AMPA
receptors.

References:

1.     Centers for
Disease Control and Prevention (CDC). 2019. font-family:" times new roman>https://www.cdc.gov/ncbddd/autism/data

2.     Walkup WG, et al
Elife. 2016;5. https://doi.org/10.7554/eLife.16813.