Alex De Nardi 1,2, Alessio Paris 1, Mario Lauria 1,2, Luca Marchetti 1,3
1 COSBI (fondazione The Microsoft Research - University Of Trento Centre For Computational And Systems Biology) (Rovereto, Italy), 2 Department of Mathematics, University of Trento (Trento, Italy), 3 Department of Cellular, Computational and Integrated Biology (CIBIO), University of Trento (Trento, Italy)
Introduction: Amyotrophic lateral sclerosis (ALS) is a multifactorial neurodegenerative disease marked by motoneuron loss and progressive muscle atrophy. Familial ALS is linked to mutations in multiple genes, and core molecular hallmarks include protein aggregate accumulation and neuroinflammation [1]. The mechanistic target of rapamycin (mTOR) pathway integrates nutrient and stress signals, regulating the balance between anabolic growth programs and catabolic processes such as autophagy [2]. mTOR inhibition is therefore a promising therapeutic strategy in ALS, as it can enhance autophagic clearance and improve neuronal survival in cellular models [3]. QSP models of the mTOR pathway are available due to its central role in diseases such as cancer; however, most of them focus on starvation-induced bulk autophagy and omit selective autophagy pathways [4]. This is a critical gap for ALS, where disease-associated mutations frequently impact selective autophagy, which is responsible for aggregate clearance (aggrephagy) [1].
Objectives: Develop a parsimonious model of aggrephagy in ALS that can be incorporated as a selective autophagy module within published mTOR QSP models. In this study, we aim to:
• identify and model a mechanistic link between aggrephagy and the mTOR pathway;
• provide a plausible literature-based parameterization and calibration strategy;
• characterize the model’s qualitative dynamical behavior.
Methods: We performed an extended literature search to identify cellular processes connecting selective autophagy to mTOR signaling, with emphasis on ALS-relevant mechanisms and mutations. Using a subset of the extracted interactions, we built a minimal ODE model. Where available, literature data were used to assign plausible parameter values. Parameter calibration was performed by fitting the disease-free steady state to in vitro measurements. To enable qualitative analysis, model variables were scaled to reduce dependence on unidentifiable parameters. The model was implemented in MATLAB and simulated numerically. Qualitative dynamics were assessed through bifurcation analysis using numerical continuation of steady states in MatCont [5].
Results: The literature review identified multiple ALS-mutated proteins implicated in selective autophagy, including C9orf72, OPTN, p62/SQSTM1, and TBK1. To enable qualitative analysis and improve parameter calibration, we adopted a parsimonious modeling strategy: we made several reactions implicit, retaining only the minimum set required to reproduce biologically plausible dynamical regimes. We therefore focused on p62 as the key mechanistic bridge, since it functions both as an aggrephagy receptor and as an activator of lysosomal mTOR signaling [6,7]. The final model consists of 4 ODEs and 13 parameters describing cytoplasmic p62 mRNA and protein levels, aggregate mass, and aggregate-bound p62. The network captures a negative feedback loop regulating p62 expression via mTOR and transcription factor EB (TFEB) [6]. mTOR activity is represented as a state-dependent variable governed by p62 through a Michaelis-Menten relationship. Proteotoxic stress is modeled as an input S(t) driving aggregate formation; aggregates are cleared by proteasomal degradation and by p62-mediated selective autophagy, which depends on p62 polymerization on the aggregate surface [7]. Literature data allowed estimation of synthesis and degradation rates for p62 mRNA and protein, as well as the kinetics of p62-aggregate interactions [6,8,9]. Biologically plausible values were assigned to unidentified parameters. Assuming constant stress S(t)=S, the model displays distinct behaviors as S varies: a single stable equilibrium, sustained oscillations, or loss of steady state with unbounded aggregate growth.
Conclusions: With a minimal reaction set, we developed a mechanistic model designed for integration within existing mTOR QSP frameworks that links aggrephagy activation to mTOR inhibition, providing a basis for autophagy activation in nutrient-rich conditions. The model predicts qualitatively different behaviors across proteotoxic stress levels. Under low stress, the system converges to a steady state with negligible autophagy. Under high stress, the model predicts continuous autophagy activation and persistent p62-bound aggregates, consistent with the observed p62-positive inclusions in post-mortem ALS neurons [10]. At intermediate stress, the model exhibits oscillatory autophagy dynamics, a relevant in vivo behavior that can be protective by limiting apoptosis associated with continuous autophagy activation [3]. Oscillatory dynamics have been reported for bulk and oxidative stress-induced autophagy [11,12]; experimental validation will be required to determine whether similar periodicity occurs in aggrephagy. Future work will test pharmacological mTOR inhibition and ALS-associated mutations by perturbing specific model parameters and evaluating their effects on aggregate burden and qualitative dynamics. Integrated into a broader mTOR QSP framework, this model will enable system-level evaluation of mTOR-targeting drugs in ALS.
References:
[1] Ling SC, Polymenidou M, Cleveland DW. Converging mechanisms in ALS and FTD: disrupted RNA and protein homeostasis. Neuron. 2013
[2] Goul, C., Peruzzo, R. & Zoncu, R. The molecular basis of nutrient sensing and signalling by mTORC1 in metabolism regulation and disease. Nat Rev Mol Cell Biol (2023)
[3] Lin, LW., Shenouda, M., McGoldrick, P. et al. C9orf72 deficiency impairs the autophagic response to aggregated TDP-25 and exacerbates TDP-25-mediated neurodegeneration in vivo. acta neuropathol commun (2025).
[4] Liu, B., Oltvai, Z.N., Bayır, H. et al. Quantitative assessment of cell fate decision between autophagy and apoptosis. Sci Rep (2017).
[5] A. Dhooge, W. Govaerts, and Yu. A. Kuznetsov. 2003. MATCONT: A MATLAB package for numerical bifurcation analysis of ODEs. ACM Trans. Math. Softw. (2003)
[6] Duran A, Amanchy R et al. p62 is a key regulator of nutrient sensing in the mTORC1 pathway. Mol Cell. 2011
[7] Vargas, J.N.S., Hamasaki, M., Kawabata, T. et al. The mechanisms and roles of selective autophagy in mammals. Nat Rev Mol Cell Biol (2023)
[8] Zaffagnini G, Savova A et al. p62 filaments capture and present ubiquitinated cargos for autophagy. EMBO J. 2018
[9] Pan B, Li J, Parajuli N et al. The Calcineurin-TFEB-p62 Pathway Mediates the Activation of Cardiac Macroautophagy by Proteasomal Malfunction. Circ Res. 2020
[10] Pinkerton M, Lourenco G et al. Survival in sporadic ALS is associated with lower p62 burden in the spinal cord. J Neuropathol Exp Neurol. 2023
[11] Holczer, M., Hajdú, B., Lőrincz, T. et al. Fine-tuning of AMPK–ULK1–mTORC1 regulatory triangle is crucial for autophagy oscillation. Sci Rep (2020)
[12] Kapuy O, Papp D et al. Systems-Level Feedbacks of NRF2 Controlling Autophagy upon Oxidative Stress Response. Antioxidants (Basel). 2018
Reference: PAGE 34 (2026) Abstr 12241 [www.page-meeting.org/?abstract=12241]
Poster: Drug/Disease Modelling - CNS