A cell-level model to predict the spatiotemporal dynamics of neurodegenerative disease
Shih-Huan Huang, Matthew W. Cotton, Tuomas P.J. Knowles, David Klenerman, Georg Meisl · 2025
Read A cell-level model to predict the spatiotemporal dynamics of neurodegenerative disease by Shih-Huan Huang, Matthew W. Cotton, Tuomas P.J. Knowles, David Klenerman, Georg Meisl (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-15046.txt.
Introduction
Shih-Huan Huang and co-authors model neurodegenerative disease cell by cell to explain its slow start and sudden rush, in a 2025 paper. Their physical model gives each cell two ways to turn diseased: on its own, or by catching the runaway protein aggregation from a sick neighbor. Which driver dominates has long been disputed.
The combined model grows pathology slowly for a long while, then accelerates sharply, matching the diseases' familiar course. It also reveals a switch point where spread takes over from spontaneous starts. The authors argue the framework can sort diseases by their driver and so point to the treatment class, blocking formation or blocking spread, most likely to help each one.
Catalog record
| CREATOR | Shih-Huan Huang, Matthew W. Cotton, Tuomas P.J. Knowles, David Klenerman, Georg Meisl |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-15046
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A cell-level model to predict the spatiotemporal dynamics of neurodegenerative disease Preprint: APS/123-QED Shih-Huan Huang, Matthew W. Cotton, Tuomas P.J. Knowles, David Klenerman, Georg Meisl Abstract A central challenge in modeling neurodegenerative diseases is connecting cellular-level mechanisms to tissue-level pathology, in particular to determine whether pathology is driven primarily by cell-autonomous triggers or by propagation from cells that are already in a pathological, runaway aggregation state. To bridge this gap, we here develop a bottom-up physical model that explicitly incorporates these two fundamental cell-level drivers of protein aggregation dynamics. We show that our model naturally explains the characteristic long, slow development of pathology followed by a rapid acceleration, a hallmark of many neurodegenerative diseases. Furthermore, the model reveals the existence of a critical switch point at which the system’s dynamics transition from being dominated by slow, spontaneous formation of diseased cells to being driven by fast propagation. This framework provides a robust physical foundation for interpreting pathological data and offers a method to predict which class of therapeutic strategies is best matched to the underlying drivers of a specific disease. I Introduction Current understanding of neurodegenerative diseases is limited by the challenge of connecting the microscopic mechanisms of protein aggregation to the macroscopic patterns of pathology observed in patients. While the general link between protein aggregation and disease is well-established Chiti and Dobson 2017 , the physical principles governing the emergence of disease-specific patterns, such as the distinct spatial patterns or the characteristic long, slow initial phase followed by rapid progression Jack et al. 2010 , remain poorly understood, hindering the development of rationally designed therapies Meisl et al. 2020 . This difficulty stems from the fundamental multiscale nature of the problem (Fig. 1 A). Current kinetic theories describe protein filament assembly at the molecular level Knowles et al. 2009 ; Cohen et al. 2011 ; Meisl et al. 2016 and network models capture phenomenological spreading at the organ level Vogel et al. 2023 ; Fornari et al. 2020 , but the crucial mesoscopic, or tissue, scale is left without a first-principles framework. There is thus currently no theory that connects the behaviour of individual cells to the emergent aggregation patterns observed in patient tissue. This theoretical gap is particularly problematic as it hinders our ability to resolve a central mechanistic question: does pathology arise primarily from cell-autonomous triggers within vulnerable cells, or is the major driver a spreading mechanism from cells that are already in a pathological, runaway aggregation state Mudher et al. 2017 ; Clavaguera et al. 2009 ? The challenge in distinguishing the contributions of these two fundamental mechanistic pathways from pathological data has been a major barrier to progress in the field Meisl et al. 2021 as effective therapies should target fundamentally different processes in the two scenarios. To bridge this gap, we develop a bottom-up physical multiscale model connecting molecular, cellular, and tissue levels. Our framework explicitly incorporates both cell-autonomous triggers and cell-to-cell propagation to simulate the stochastic evolution of aggregation at the single-cell level. We showcase the varied behaviour this minimal model gives rise to in different regimes and discuss the mathematical relationship to reaction diffusion equations and epidemiological models of the spread of disease in a susceptible population. By deriving analytical expressions for key spatial statistics, we provide a quantitative toolkit to infer the dominant aggregation mechanism from static patterns of aggregates, culminating in the derivation of a critical switch fraction that predicts when the system transitions from being cell-autonomous to propagation-dominated. Figure 1 : Model description. (A) A central problem in neurodegenerative research is the lack of theory that can describe different systems across a wide range of temporal and spatial scales. Scale bar: cellular level: 30 \mu m ; tissue level: 2 mm. The brain image is adapted from Human Brain Project Amunts et al. 2021 . (B) In our model, the fundamental unit is a cell. Aggregation can begin autonomously within a cell at a rate k_{a} (measured in inverse time). Additionally, cells that have already aggregated can trigger aggregation in neighboring cells. This interaction depends on the coupling strength or rate k_{s} (also measured in inverse time) and a length-dependent factor characterized by \sigma (with units of length). II A cell-level model for neurodegenerative diseases To bridge the gap between molecular events and tissue-level pathology, we developed a biophysical model to describe the spread of disease within a tissue. We model the tissue as an ensemble of discrete cells, the fundamental unit of biological organization Brenner 2010 . Each cell i is treated as a point-like compartment at position \mathbf{x}_{i} and is described by a single binary state variable, \phi_{i} . This variable denotes whether the cell is ‘healthy’ (\phi_{i}=0) or has transitioned to a state of pathological protein aggregation (\phi_{i}=1) . We simulate the evolution of tissue by prescribing \phi for every cell at some initial time and proceed in discrete time steps, \Delta t , during which a healthy cell can transition to the aggregated state through two distinct mechanistic pathways. The first pathway, the cell-autonomous triggers , represents the spontaneous formation of protein aggregates within a single cell, independent of its neighbours. Depending on the disease, this cell-autonomous transition, could be caused by a stochastic nucleation event, or via overwhelming of the cellular aggregate removal mechanisms Meisl 2024 . The effect of rare nucleation events has been well established in vitro Knowles et al. 2011 and in vivo Sinnige et al. 2021 , and is likely the key driver of sporadic prion disease. By contrast, an overwhelming of aggregate removal mechanisms is a more likely scenario in common neurodegenerative disorders such as Alzheimer’s disease. This process may be driven by factors such as age-related accumulation of somatic mutations in neurons Bae et al. 2022 , or other stressors leading to a temporary failure of protein homeostasis Labbadia and Morimoto 2015 and ultimately runaway aggregation Thompson et al. 2021 . We coarse grain these distinct processes into a single time-independent cell-autonomous rate, k_{a} . The probability p_{a,i} that any given healthy cell i spontaneously transitions to the aggregated state in a small time interval \Delta t is therefore given by: p_{a,i}=1-\exp(-k_{a}\Delta t)\approx k_{a}\Delta t for the short times \Delta t . The second pathway for a cell to transition to the aggregated state is via cell-to-cell propagation , which models the spread of pathology from an already aggregated cell to its healthy neighbours. This can represent various biological processes, including the direct transport of pathogenic protein seeds between cells or the triggering of inflammatory responses Weickenmeier et al. 2018 ; Frost et al. 2009 , which in turn can compromise the protein quality control systems of otherwise healthy cells, increasing their vulnerability to aggregation. We model this coupling as a radially isotropic interaction function that decays with distance. Specifically, we use a Gaussian kernel for the majority of this work, as it represents a simple, canonical model for a diffusive process, but we also show that the conclusions are robust regardless of the choice of the specific interaction function. The probability, p_{s,i} , that a healthy cell i is triggered by its aggregated neighbours during \Delta t is then p_{s,i}=1-e^{-\lambda_{i}\Delta t}, where \lambda_{i} is the local ‘cell-to-cell