# Critical Evaluation of Hypotheses
---
## Hypothesis 1: TRIM21 Ubiquitin Gradient
### Weak Links
- **Assumed directionality without mechanism**: The model claims steric constraints prevent Ub chain propagation in the core but provides no evidence that chain elongation is physically limited in dense LLPS phases. Chain formation rates and enzymatic processivity in condensed phases remain uncharacterized.
- **Spatial localization of TRIM21 activity**: Source paper demonstrates TRIM21 ubiquitinates G3BP1 but does not establish that this modification occurs preferentially at the periphery. TRIM21 may access and modify core substrates.
- **Chain accumulation vs. chain formation**: Peripheral K63-Ub enrichment could reflect dynamic equilibrium of chain turnover rather than propagation constraints—different hypothesis entirely.
### Counter-Evidence
- Ubiquitination machinery can function in dense chromatin and nuclear pore complexes; steric exclusion is not universal.
- If the core blocks chain elongation, how does TRIM21 (or other E3s) initially access core substrates for ubiquitination?
### Falsifying Experiment
- Perform **nanoSIMS or cryo-EM with K63-Ub specific nanobodies** to directly map Ub chain length distribution (monoubiquitin vs. polymeric) at single-granule resolution. If chains are equally long throughout granules, the steric constraint model is falsified.
- Test whether forced peripheral recruitment of TRIM21 (via peripheral-targeting domain fusion) is necessary vs. sufficient for SQSTM1 peripheral localization.
### Revised Confidence: 0.48
*(Significant mechanistic gap: no direct spatial mapping of Ub chains)*
---
## Hypothesis 2: LLPS Saturation Partitioning
### Weak Links
- **Valency paradox**: SQSTM1 has multivalent PB1 dimerization domain and LC8 dimer binding motifs; CALCOCO2 has coiled-coil self-association. High-valency proteins typically partition *into* dense phases, not out of them. The hypothesis requires these receptors to have "relatively low valency"—but this contradicts established structural biology.
- **Uniform shell vs. selective peripheral localization**: Pure partitioning physics predicts uniform shell distribution around the phase boundary, not selective localization to one region. Does not explain why receptors localize specifically to the *outside* periphery rather than the interface.
- **Initial recruitment mechanism**: If receptors are "squeezed out" by core density, how do they initially recognize and engage granules for clearance?
### Counter-Evidence
- SQSTM1 forms liquid-like droplets for selective autophagy; it actively partitions into dense cargo-phagy receptor condensates.
- CALCOCO2 forms innate immune condensates itself—its inherent LLPS behavior suggests affinity for dense phases.
### Falsifying Experiment
- Measure **partition coefficients (Kₚ) directly** using fluorescence correlation spectroscopy (FCS) inside vs. outside SG cores. If Kₚ > 1 for SQSTM1/CALCOCO2, the hypothesis is falsified.
- Test whether artificially increasing valency (dimerization domain fusions) shifts localization toward SG interior.
### Revised Confidence: 0.42
*(Mechanistic contradictions with receptor structural biology)*
---
## Hypothesis 3: G3BP1 NTF2L Steric Exclusion
### Weak Links
- **Domain localization unverified**: The NTF2L domain architecture places it C-terminal to G3BP1's RGG domain. Whether