Neuroscience experiments, decoded

Concise verdict

The hypothesis is plausible but incompletely supported by current data: (1) local ATP dips can occur and precede aggregation in cells, (2) loss of ATP reduces hydrotropic solubilization capacity and promotes condensate nucleation in vitro and in silico, and (3) seeded condensates can spread cell-to-cell — but direct evidence that physiologic synaptic microdepletion seeds pathological condensates that later propagate transcellularly is lacking and requires targeted experiments. Key supporting and constraining evidence is cited below.

• Single cell yeast work shows transient ATP dips causally precede rapid aggregation events and that stable high ATP prevents aggregationYeast ATP dips cause aggregation
• Multiscale computations and experiments show interprotein beta sheet formation within condensates strongly increases binding/viscosity and that RNA or small molecules can slow maturation from liquid to solid states — connecting loss of hydrotropic effects to nucleation/maturationBeta sheet driven condensate ageing
• Phase diagram and ExVivo PhaseScan studies show RNA, salt, and hydrophobic interactions determine condensate stability and that disease mutations shift condensates toward aggregation; hydrotropic-like interventions (MBP recruitment) can reverse condensates in cellsExVivo PhaseScan maps condensate phase space

Visual summary first

Executive evidence map

1. ATP dips can be fast, local, and precede aggregation — single cell imaging in yeast shows stochastic cytosolic ATP dips precede Hsp104 marked aggregate formation and toxicity; restoring ATP stability prevents aggregation (QUEEN sensor). This provides the clearest demonstration of temporal precedence of ATP loss before aggregation at single-cell resolution, albeit in yeast cytosol rather than mammalian synapsesYeast single cell ATP dips precede aggregation
2. Hydrotropic and solubilizing roles of small molecules and ATP — conceptually, millimolar ATP has hydrotropic properties in vitro and can solubilize proteins; experimental/optogenetic recruitment of solubilizers (MBP) dissolves condensates rapidly in mammalian cells demonstrating that local loss of solubilizing capacity can increase condensate stability and persistenceOptoMBP dissolves condensates in cells
3. Condensate nucleation and maturation governed by kinetics and local environment — in vitro and computational work show nucleation is sensitive to local concentration, multivalent interactions, interfaces, and presence of RNA or other modulators; rapid local increases in effective concentration can nucleate condensates that later age into gel/solid states via interprotein beta sheet formationProtein RNA condensation kineticsKinetic determinants of condensate maturation
4. Liquid to solid maturation can be decelerated by RNA or kept soluble by hydrotropes — simulations and ExVivo PhaseScan show RNA can prevent or slow beta sheet accumulation and maintain fluidity; conversely, depletion of solubilizers or changes in ionic/hydrophobic environment shift condensates toward gels/solidsRNA slows condensate ageingPhaseScan RNA biphasic effects
5. Cell to cell spreading of aggregates is plausible via multiple pathways — prionlike spreading models and experimental evidence demonstrate extracellular release, uptake, and synaptic/trans-synaptic propagation of seeds for tau, alpha-synuclein, and others; macropinocytosis, exosomes, tunneling nanotubes, and synaptic mechanisms can transfer aggregates between cells, enabling transcellular propagation once local seeds existPrion like spread reviewMacropinocytosis enables uptake of aggregates

Synthesis: linking the chain in the hypothesis

The hypothesis proposes the following causal chain:

1. Local synaptic ATP microdepletion occurs before wholesale cellular ATP decline.
2. That local ATP loss reduces local hydrotropic suppression (ATP is a hydrotrope at millimolar levels), enabling local failure of solubilization and increasing effective local protein concentrations and interaction strengths.
3. These microdomains enable nucleation of biomolecular condensates (via LLPS or crystallization-like pathways) which mature into gel/solid seeds through structural transitions (eg interprotein beta sheets).
4. Seeds are released or transferred transcellularly (synaptic release, macropinocytosis, nanotubes, exosomes) leading to network spread.

Which links have direct evidence and which do not?

Critical gaps and alternative explanations

• No direct measurements of ATP microdomains at mammalian synapses showing millimolar local ATP drops that are sufficient to alter hydrotropic suppression locally. Yeast data show cytosolic dips but scaling to synaptic nanodomains with high ATP turnover is uncertainYeast ATP dips precede aggregation.
• ATP as hydrotrope in cells: biochemical studies support hydrotropic behavior at millimolar concentrations but intracellular crowding, protein partners, and organelles may modulate effective hydrotropic action; alternative cellular hydrotropes or chaperone systems (Hsp70/ATPase cycles) could be primary drivers of solubilization rather than ATP per se (and ATP depletion also impairs chaperone activity), confounding causality.
• Aggregate nucleation may be driven by local post-translational modifications (phosphorylation, oxidation) or local increases in client concentration due to trafficking/translation — ATP dips could be a consequence rather than cause in some contexts.
• Some condensates may nucleate at membranes or via lipid interactions independent of ATP hydrotropic effects (membrane-anchored LLPS), so synaptic membranes themselves could be nucleation platformsMembrane anchored condensate nucleation.

Concrete experimental tests that would most powerfully falsify or support the hypothesis

1. Direct measurement of synaptic ATP microdomains + aggregation coupling — Use high spatial resolution ratiometric ATP sensors targeted to presynaptic boutons and postsynaptic densities (eg targeted QUEEN or novel ATP FRET sensors with nanometer targeting) combined with live imaging of endogenous aggregation-prone proteins (tagged at genomic loci) to test whether reproducible local ATP dips precede local condensate nucleation at synapses in cultured neurons or brain slices. If ATP dips temporally and spatially precede nucleation, that supports the hypothesis; absence or mismatch falsifies it. (Feasible with two-photon FLIM-FRET or lattice light sheet for intact tissue.)
2. Local restoration of hydrotropic capacity rescues nucleation — Apply OptoMBP-style recruitment of a solubilizer to single synapses concurrently with induced local ATP depletion (chemogenetic or optogenetic metabolic blockade) to test whether targeted solubilizer rescue prevents local condensate nucleation. If rescue prevents nucleation despite ATP dip, then hydrotropic suppression is mechanistically required; if not, other mechanisms (eg chaperone failure) dominateOptoMBP provides solubilizer rescue approach.
3. PhaseScan-like microdroplet reconstitution with synaptic lysate — Use ExVivo PhaseScan but with synaptic/isolated synaptoneurosome lysate and physiological ATP titration to map whether modest local ATP changes near synaptic concentrations shift phase boundaries for synaptic RBPs (e.g., tau, synapsin, FUS, TDP43). Finding sharp ATP-sensitive binodals at physiological ATP ranges would strongly support in vivo plausibilityPhaseScan for synaptic lysates.
4. Block seed spread pathways after local nucleation — Induce local condensates in one neuron by targeted ATP depletion and test if blocking macropinocytosis, exosome release, or tunneling nanotube formation in adjacent cells prevents transcellular spread; if spread is blocked, pathway established; if spread occurs despite blockade, alternative propagation mechanisms are implicatedPrionlike spread mechanisms and interventions.

Novel experiments I recommend (concise protocols)

• Experiment A Targeted synaptic ATP imaging and aggregation coupling: Express synapse-targeted QUEEN in primary hippocampal neurons together with endogenously tagged synapsin or tau (CRISPR knock-in). Use patterned optogenetic stimulation of mitochondria/Na pump inhibitors to induce transient synaptic ATP dips while imaging. Quantify time-lag distributions between ATP dip onset and local condensate marker appearance; compare with controls. Outcome metric: fraction of synapses where ATP dips precede condensate formation within defined time window.
• Experiment B Rescue with local solubilizer: Combine Experiment A with OptoMBP targeted to the same synapse — trigger MBP recruitment at dip onset; test whether recruitment prevents condensate nucleation and subsequent spread. Metric: incidence of condensate formation with vs without MBP recruitment.

Interpretation and conservative likelihood estimate

Given existing evidence from yeast (ATP dips precede aggregation), computational/biophysical studies (ATP/hydrotrope effects, kinetics of nucleation, role of RNA), and prionlike spread literature, the chain is biologically plausible but currently missing crucial links (direct mammalian synapse ATP microdomain data; demonstration that ATP hydrotropy is the operative mechanism vs co-varying processes like chaperone failure or post-translational modifications). I therefore estimate a conservative likelihood that the full hypothesis (all links true in physiological mammalian neurons and leading to pathological transcellular spread) is 35% (80% CI 20 55). Confidence reflects strong mechanistic plausibility but limited direct in vivo evidence.

Improved, testable restatement of the hypothesis

Improved hypothesis Local, transient synaptic ATP microdepletion reduces local hydrotropic and ATPase-chaperone activities at active synapses, increasing effective local client concentration and promoting nucleation of metastable condensates of synaptic RBPs; these condensates mature by interprotein structural transitions into seeding competent assemblies that can transfer to connected neurons via synaptic release or macropinocytosis and seed further aggregation. This formulation makes the multi-mechanism nature explicit (hydrotrope loss + chaperone impairment) and identifies measurable intermediates (local ATP, chaperone activity, condensate mobility, beta sheet signatures, uptake mechanisms).

Blindspots and biases

• Model systems: many cited studies use yeast, in vitro reconstitution, or simulations — cross-species and in vivo translation is uncertain.
• Confounding causes: ATP depletion impairs energy-dependent chaperones and proteostasis pathways; any causality must separate hydrotropic chemical effects of ATP from energy failure consequences.
• Publication and model bias: much LLPS work uses overexpression or purified systems at concentrations above physiological; must control for stoichiometry.

Key citations used in the analysis

• Yeast ATP and aggregation: Yeast ATP dips cause aggregation
• Condensate structural transitions: Beta sheet driven condensate ageing
• Phase behavior in lysate contexts: ExVivo PhaseScan maps condensates
• OptoMBP solubilizer tool: OptoMBP dissolves condensates
• Nucleation kinetics and filamentous nanoclusters: Protein RNA condensation kinetics
• Prionlike spread mechanisms review: Prion like spread review
• Uptake pathways (macropinocytosis): Macropinocytosis of aggregates
• Membrane assisted nucleation: Membrane anchored condensate nucleation

Practical next steps and resources

Immediate actions to progress the hypothesis experimentally:

1. Develop synapse-targeted high sensitivity ATP sensors (QUEEN variants or FLIM-FRET) and validate dynamic range in neurons. Use Opto tools to induce local metabolic perturbations while imaging aggregation markers.
2. Adapt ExVivo PhaseScan to synaptoneurosome lysates and titrate ATP and chaperone factors to map ATP-sensitive binodals for synaptic RBPs.
3. Combine local OptoMBP rescue with local ATP manipulations to test hydrotropic vs chaperone explanations.
4. Use microfluidic connected-neuron devices to observe local induction and transcellular spread while pharmacologically blocking uptake routes (macropinocytosis, exosome release) to define spread mechanisms.

Summary conclusion (short)

Mechanistically plausible and supported by component evidence (ATP dips can precede aggregation; local changes in solubility and kinetics can nucleate condensates; aggregates can spread cell-to-cell), but the full chain from physiologic synaptic ATP microdepletion → local hydrotropic failure → nucleation → transcellular spread remains unproven in mammalian neurons; targeted experiments outlined above can decisively test the hypothesis.