A joint model of network spread and selective vulnerability explains tau pathology spread in Alzheimer's disease mouse models.

Background: The propagation and spread of pathology in Alzheimer's disease (AD) results from the interplay of many complex biological processes. Two key hypotheses have been proposed to help explain the characteristic Braak stageing patterns that define typical AD tau pathology1: network spread, which posits that trans‐synaptic spread of tau determines how pathology progresses2, and selective vulnerability, which states that the regions most affected by tau pathology contain high densities of cell types particularly sensitive to the presence of tau3. Here we explore the explanatory power of each hypothesis in the context of mouse tauopathy models. Methods: We implemented a mathematical network spread model that extends the Network Diffusion Model2 by adding a linear accumulation term (sNDM). We also utilized the cell type distributions recently made available by the MISS algorithm, which deconvolves spatial transcriptomic data into cell type densities using scRNAseq information4. We fit the sNDM to the Iba, et al. PS19 tauopathy model seeded at the hippocampus5 (IbaHippInj) and the Hurtado, et al. unseeded PS19/PDAPP model6 (Hurtado). Results: We find that although network spread is the single‐best correlate of tau pathology, there are many cell types that are highly correlated at the regional level, particularly Layer‐2/3 glutamatergic neurons and Lamp5‐expressing GABAergic neurons (Figure 1). A combined linear model using sNDM predictions and these two cell types (Figure 2) yields higher R2 values than the sNDM predictions alone (Figure 3). Conclusion: We demonstrate that both the network spread and selective vulnerability hypotheses contribute meaningful information for explaining tau pathology in mouse models, indicating that more sophisticated modeling approaches taking into account the full biological complexity of AD have more explanatory power. References: 1. Braak, H. & Braak, E. Acta Neuropathol. 82, 239‐259 (1991). 2. Raj, A., Kuceyeski, A. & Weiner, M. Neuron (2012). 3. Fu, H., Hardy, J. & Duff, K. E. Nat. Neurosci. 21, 1350‐1358 (2018). 4. Mezias, C., Torok, J., Maia, P. D., Markley, E. & Raj, A. PNAS. (In press). 5. Iba, M. et al. J. Neurosci. 33, 1024‐1037 (2013). 6. Hurtado, D. E. et al. Am. J. Pathol. 177, E4376–E4385 (2010). [ABSTRACT FROM AUTHOR]