PREreview of Massively parallel genetic perturbation reveals the energetic architecture of an amyloid beta nucleation reaction

Based on: https://www.biorxiv.org/content/10.1101/2024.07.24.604935v1.full

The manuscript from Lehner and colleagues presents a wealth of mutagenesis information on amyloid aggregation. The central premise of the paper is to use a yeast selection based around the oligmerizaton/aggregation of Sup35 fused to a peptide (in this case abeta) as a proxy for amyloid forming potential. This is cool information on its own and the experimental analysis and computational framework for linking to energies is top notch. The point of using double mutants to enhance the dynamic range is very well explained and will solidify their approach to impactfully link DMS experiments to thermodynamic concepts. 

The major framing of the paper revolves around an analytical protein folding/engineering concept of phi-values that highlights energetic differences in the importance of interactions for forming the transition state vs. the ground state. For the textual interpretation of the results, one must buy into the energetic effects of the Sup35 system as a readout of the transition state (and secondarily for FoldX calculations on various PDBs of abeta polymorphs to be a readout of the ground state). The major issue is that an alternative (and perhaps simpler) explanation is that mutations in APR2 are more disruptive to the Sup35 oligomerization process in the screen and that this reflects amyloid/oligomerization propensities and not strictly TS of initial nucleation. The data from previous studies that is used to draw correlations to justify their interpretation around the TS is buried in extended data figures and is a bit all over the place, especially the deconvolution of primary vs. secondary nucleation. The existence of multiple polymorphs in human cells (and populations), which may or may not have related transition states - and the exact conformational requirements of the Sup35 activation mechanism - further complicate this interpretation.

In summary - this contains amazing data, but I do not see the language of the interpretations lining up with the strength of the specificity of the claims about the transition state.  A fuller discussion of the limitations of the prior low throughput assays that are referenced in extended data 1 and 2 and detailed kinetic characterization of some of the more surprising mutants in a biochemically defined system would greatly improve the match between the data and the claims. These issues should not stop others from building on this beautiful work - but in doing so, other investigators should note that there remains ambiguity as to whether the effects are truly on the TS or on the ground state.

• Avi Samelson and James Fraser

Competing interests

The authors declare that they have no competing interests.