MagIC-Cryo-EM: Structural determination on magnetic beads for scarce macromolecules in heterogeneous samples

Reviewer #1 (Public review):

Summary:

In this manuscript, Arimura et al describe MagIC-Cryo-EM, an innovative method for immune-selective concentrating of native molecules and macromolecular complexes for Cryo-EM imaging and single-particle analysis. Typically, Cryo-EM imaging requires much larger concentrations of biomolecules than that are feasible to achieve by conventional biochemical fractionation. Overall, this manuscript is meticulously and clearly written and may become a great asset to other electron microscopists and chromatin researchers.

Strengths:

Previously, Arimura et al. (Mol. Cell 2021) isolated from Xenopus extract and resolved by Cryo-EM a sub-class of native nucleosomes conjugated containing histone H1.8 at the on-dyad position, similar to that previously observed by other researchers with reconstituted nucleosomes. Here they sought to analyze immuno-selected nucleosomes aiming to observe specific modes of H1.8 positioning (e.g. on-dyad and off-dyad) and potentially reveal structural motifs responsible for the decreased affinity of H1.8 for the interphase chromatin compared to metaphase chromosomes. The main strength of this work is a clever and novel methodological design, in particular the engineered protein spacers to separate captured nucleosomes from streptavidin beads for a clear imaging. The authors provide a detailed step-by-step description of MagIC-Cryo-EM procedure including nucleosome isolation, preparation of GFP nanobody attached magnetic beads, optimization of the spacer length, concentration of the nucleosomes on graphene grids, data collection and analysis, including their new DUSTER method to filter-out low signal particles. This tour de force methodology should facilitate considering of MagIC-Cryo-EM by other electron microscopists especially for analysis of native nucleosome complexes.
In pursue of biologically important new structures, the immune-selected H1.8-containing nucleosomes were solved at about 4A resolution; their structure appears to be very similar to the previously determined structure of H1.8-reconstituted nucleosomes. There were no apparent differences between the metaphase and interphase complexes suggesting that the on-dyad and off-dyad positioning does not explain the differences in H1.8 - nucleosome binding. However, they were able to identify and solve complexes of H1.8-GFP with histone chaperone NPM2 in a closed and open conformation providing mechanistic insights for H1-NPM2 binding and the reduced affinity of H1.8 to interphase chromatin as compared to metaphase chromosomes.

Weaknesses:

Still, I feel that there are certain limitations and potential artifacts resulting from formaldehyde fixation, use of bacterial-expressed recombinant H1.8-GFP, and potential effects of magnetic beads and/or spacer on protein structure, that should be more explicitly discussed. Also, the GFP-pulled down H1.8 nucleosomes should be better characterized biochemically to determine the actual linker DNA lengths (which are known to have a strong effect of linker histone affinity) and presence or absence of other factors such as HMG proteins that may compete with linker histones and cause the multiplicity of nucleosome structural classes (such as shown on Fig. 3F) for which the association with H1.8 is uncertain.

Reviewer #2 (Public review):

Summary:

The authors present a straightforward and convincing demonstration of a reagent and workflow that they collectively term "MagIC-cryo-EM", in which magnetic nanobeads combined with affinity linkers are used to specifically immobilize and locally concentrate complexes that contain a protein-of-interest. As a proof of concept, they localize, image, and reconstruct H1.8-bound nucleosomes reconstructed from frog egg extracts. The authors additionally devised an image-processing workflow termed "DuSTER", which increases the true positive detections of the partially ordered NPM2 complex. The analysis of the NPM2 complex {plus minus} H1.8 was challenging because only ~60 kDa of protein mass was ordered. Overall, single-particle cryo-EM practitioners should find this study useful.

Strengths:

The rationale is very logical and the data are convincing.

Weaknesses: I have seen an earlier version of this study at a conference. The conference presentation was much easier to follow than the current manuscript. It is as if this manuscript had undergone review at another journal and includes additional experiments to satisfy previous reviewers. Specifically, the NPM2 results don't seem to add much to the main story (MagIC-cryo-EM), and read more like an addendum. The authors could probably publish the NPM2 results separately, which would make the core MagIC results (sans DusTER) easier to read.

Reviewer #3 (Public review):

Summary:

In this paper, Arimura et al report a new method, termed MagIC-Cryo-EM, which refers to the method of using magnetic beads to capture specific proteins out of a lysate via, followed immunoprecipitation and deposition on EM grids. The so-enriched proteins can be analzyed structurally. Importantly, the nanoparticles are further functionalized with protein-based spacers, to avoid a distorted halo around the particles. This is a very elegant approach and allows the resolution of the stucture of small amounts of native proteins at atomistic resolution.
Here, the authors apply this method to study the chromatosome formation from nucleosomes and the oocyte-specific linker histone H1.8. This allows them to resolve H1.8-containing chromatomosomes from oocyte extract in both interphase and metaphase conditions at 4.3 A resolution, which reveal a common structure with H1 placed right at the dyad and contacting both entry-and exit linker DNA.
They then investigate the origin of H1.8 loss during interphase. They identify a non-nucleosomal H1.8-containing complex from interphase preparations. To resolve its structure, the authors develop a protocol (DuSTER) to exclude particles with ambiguous center, revealing particles with five-fold symmetry, that matches the chaperone NPM2. MS and WB confirms that the protein is present in interphase samples but not metaphase. The authors further separate two isoforms, an open and closed form that coexist. Additional densities in the open form suggest that this might be bound H1.8.

Strengths:

Together this is an important addition to the suite of cryoEM methods, with broad applications. The authors demonstrate the method using interesting applications, showing that the methods work and they can get high resolution structures from nucleosomes in complex with H1 from native environments.

Weaknesses:

The structures of the NPM2 chaperone is less well resolved, and some of the interpretation in this part seems only weakly justified.