1000ExM expands a biological sample a thousandfold; calculations indicate that adjacent amino acids can be resolved
1000ExM expands a biological sample a thousandfold; calculations indicate that adjacent amino acids can be resolved
On July 21, Helena Hu, Ed Boyden, and their coauthors published 1000ExM on bioRxiv. The method expands a fixed biological sample by approximately a thousandfold in each linear dimension. The authors tested structural preservation in several proteins and one peptide, then modeled protein identification from the resulting spatial map.
A light microscope merges closely spaced points into a single spot. Adjacent amino acids in a protein chain are separated by about 0.38 nanometers, a distance that conventional optics cannot resolve. 1000ExM separates molecules until they reach a scale visible with a conventional microscope.
The researchers chemically attach amino acid side chains to a swellable gel, cut the protein chain between the attachment sites, and add water. As the gel expands, it separates the anchored fragments. Four polymer networks repeat this expansion process: the sample grows by approximately a thousandfold along each axis, and its volume increases by approximately a billionfold. A distance of 0.38 nanometers becomes approximately 380 nanometers, which a confocal microscope can resolve.
Earlier forms of expansion microscopy achieved a total linear expansion of about 16–22 times. After each round, they added a neutral gel to hold the expanded sample in place. In the study by Hu and colleagues, each subsequent charged network is formed directly within the gel that has already expanded. This change allowed the expansion process to be repeated four times.
In the Eon experiment, neuronal microcultures were expanded approximately 20 times so that their wiring could be matched to recorded activity. 1000ExM extends the same approach to protein labels: after expansion, the model indicates that conventional optics can resolve their spatial arrangement.
The authors compared the resulting maps with known structures of nanobodies, the mCLING peptide, and GFP, the green fluorescent protein. This comparison tested whether the labels preserved the geometry of the original molecule. They then analyzed 23 391 canonical human proteins, accounting for missing labels and the measured deformation of the gel. The resulting label pattern was unique for most proteins.
In the calculations, the label pattern serves as a protein signature. Fragments bearing anchored side chains leave spatially defined points, and their arrangement can distinguish one protein from another. The gel expanded the original distances to a scale accessible to conventional confocal optics.
The authors tested the preservation of label geometry in individual molecules. Their model suggests that such signatures could eventually be obtained inside cells and tissues.