(A-40) Systems Biology Modeling of Inflammasomes Impacting Hypertension

Hypertension is a leading health concern and has been shown to be linked to chronic stress. Stress leads to the creation of reactive oxygen species in the mitochondria of endothelial cells. This causes the oxidation, segmentation, and release into the cytosol of mitochondrial DNA [1]. This mtDNA is a nucleation point for inflammasome sensor proteins which polymerize on the mtDNA fragments and recruit ASC which polymerizes and binds to procaspase-1 [2,3,4]. Procaspase-1 is cleaved to form mature caspase-1 which induces the formation of the inflammatory cytokines IL-18 and IL-1β. This leads to inflammation, pyroptosis, and hypertension [2,3,4]. Even though inflammasomes are widely implicated as a sensing platform in immune responses they have not been the subject of heavy quantitative study and modeling [3,4]. The multi-step process of their assembly informed by the inflammasome proteins' underlying structure makes them an excellent target for kinetic and systems biology modeling [3,4]. Systems biology looks to model complex molecular pathways by constructing them out of their most basic components [5]. The mechanism of AIM2 involving its interactions with dsDNA and ASC has been modeled through an Occam's razor minimalistic model previously in Digital signaling network drives the assembly of the AIM2-ASC inflammasome [4]. In this work, we look to expand on this model by applying a mass balance, and DNA length-dependent rate terms to create a model for inflammasome assembly that predicts its binding curves under different concentrations of AIM2 monomer and DNA as well as different lengths of DNA.

The kinetic models developed in this work were created in Python and utilized SciPy libraries for numerical integration of the differential equations. The models were fit to binding curves of AIM2 polymerization on dsDNA reported in Digital signaling network drives the assembly of the AIM2-ASC inflammasome [4]. To analyze length and concentration-dependent binding, the model created expands on the simultaneous set of ordinary differential equations from reference four in two significant ways. First, mass balance equations were added to monitor the concentration of AIM2 monomers and free dsDNA over time. Second, DNA length dependence was explicitly incorporated into the rate equations. This was accomplished by only allowing DNA-bounded AIM2 reactions when the two proteins were attached to the same strand of DNA. To model this, the concentration of dsDNA strands bound to zero, one, or multiple AIM2 proteins was tracked. In this model the binding of AIM2 to DNA was reversible and the aggregation of DNA-bound AIM2 is irreversible. The rate of oligomerization between two dsDNA bound AIM2 monomers on the same DNA strand was assumed to be large enough that it can be modeled as instantaneous. The rate constants within this model were fitted using data from an AIM2 polymerization curve on 600bp dsDNA. To test this model these rate constants were used to create binding curves for dsDNA of different lengths as well as different initial concentrations of both AIM2 free monomers and dsDNA. These predictions were then compared to experimental data of corresponding initial conditions.

The assembly of AIM2 polymers on DNA is a complicated set of reactions. We show here that it can be effectively represented through a systems biology model created from kinetic rate equations that represent the fundamental molecular interactions that comprise the AIM2 inflammasome activation mechanism.  From the model, as the initial concentration of either AIM2 monomers or DNA is increased the rate of polymerization is also increased. This is likely due to a greater number of binding interactions between the two species. The developed model also shows an increased rate of reaction for increasing lengths of DNA. This suggests that AIM2 polymerization's dependence on DNA length is due to an increased likelihood of two monomers being attached to the same strand of DNA.  This matches experimental data for dsDNA lengths greater than or equal to 72bp.  The predictive capability of this model suggests that the mechanism the model is based on is an adequate approximation of the actual molecular mechanism taking place. This model, however, does not accurately predict binding curves for 39bp and 24bp dsDNA strands. Possible causes of this could be due to the disassociation between oligomers smaller than four and dsDNA or the occurrence of different molecular structures in small oligomers as suggested by reference four [4]. Further MD simulations would need to be performed to analyze the validity of this model on the atomic scale. The next steps in this project will investigate testing this model through MD simulation and generalizing it to include other inflammasomes of interest in hypertension like NLRP3 and NLRP10.

This project would not have been possible without the time and support of Dr. Melissa Moss and the University of South Carolina College of Engineering and Computing.

1. Xian H, Watari K, Sanchez-Lopez E, Offenberger J, Onyuru J, Sampath H, Ying W, Hoffman HM, Shadel GS, Karin M: Oxidized DNA fragments exit mitochondria via mPTP- and VDAC-dependent channels to activate NLRP3 inflammasome and interferon signaling. Immunity 2022, 55(8):1370-1385.e13781. Geula S, Naveed H, Liang J, Shoshan-Barmatz V: Structure-based Analysis of VDAC1 Protein. Journal of Biological Chemistry 2012, 287(3):2179-2190.

2. Wang B, Bhattacharya M, Roy S, Tian Y, Yin Q: Immunobiology and structural biology of AIM2 inflammasome. Mol Aspects Med 2020, 76:100869.

3. Morrone SR, Matyszewski M, Yu X, Delannoy M, Egelman EH, Sohn J: Assembly-driven activation of the AIM2 foreign-dsDNA sensor provides a polymerization template for downstream ASC. Nature Communications 2015, 6(1):7827.

4. Matyszewski M, Morrone SR, Sohn J: Digital signaling network drives the assembly of the AIM2-ASC inflammasome. Proceedings of the National Academy of Sciences 2018, 115(9):E1963-E1972.

5. Morris AM, Watzky MA, Finke RG: Protein aggregation kinetics, mechanism, and curve-fitting: A review of the literature. Biochimica et Biophysica Acta (BBA) - Proteins and Proteomics 2009, 1794(3):375-397.