FLIM-FRET as a Molecular Filter for Membrane-Induced Aggregation

Membrane binding is thought to trigger early aggregation of alpha-synuclein (aSyn) in neurons. However, live-cell measurements of membrane-proximal aggregation with high specificity remain challenging. We combine three- channel fluorescence lifetime imaging microscopy (FLIM) with Förster resonance energy transfer (FRET) in a model that uses FRET as a proximity filter. A membrane-tethered donor reports membrane-aSyn separation through lifetime changes, while an aSyn-labeled acceptor reports the aggregation state through its lifetime. We estimate per-cell lifetimes and mixture fractions for membrane-bound and membrane-unbound populations using a hierarchical expectation-maximization (EM) algorithm that pools information across pixels. We validate the estimator using Monte Carlo studies. Using experimental neuronal data, this method resolves changes in membrane-proximal aggregation and aggregate-associated lifetimes. This framework provides quantitative per-cell metrics linking membrane proximity and aggregation for comparative live-cell studies.

Many studies map FRET signals pixel by pixel or average signals over whole cells. Neither approach reliably answers key biological questions: how much membrane-proximal aggregation exists in a given cell, and how does it change across conditions? Here, we combine three-channel FLIM-FRET with a hierarchical analysis that estimates per-cell lifetimes and fractions for bound and unbound populations by pooling information across pixels. This method shifts the focus from noisy images to cell-level metrics that support comparisons across treatments, time points, or genotypes. Simulations and neuronal data show improved accuracy under realistic photon budgets and reveal membrane-proximal aggregation effects that were unattainable otherwise. This approach is broadly applicable and extends beyond alpha-synuclein.