Expansion Microscopy
Beat the diffraction limit by inflating the specimen — a swellable gel that physically blows tissue up four-fold, then forty-fold.
Linearly enlarges fixed biological samples 4×–~20× by infusing them with a polyelectrolyte hydrogel that swells uniformly in water. Because every fluorescently labelled molecule's relative position is preserved during the expansion, a normal diffraction-limited microscope (≈250 nm resolution) becomes effectively a super-resolution instrument (≈70 nm in 4×ExM, ≈25 nm in 10×ExM, ≈10 nm in iterated 20–40× ExM). Applies to cultured cells, brain slices, whole zebrafish embryos, and CLARITY-cleared tissue.
(1) Anchoring: a sample is fixed (typically PFA), labelled with fluorescent antibodies or DNA probes whose tags carry an acrylamide-bearing moiety, then exposed to an anchor reagent (Acroloyl-X, GMA) that covalently links biomolecules to the future hydrogel. (2) Gelation: the sample is soaked in a monomer mix containing sodium acrylate (the swellable monomer), acrylamide, and a crosslinker (BIS); polymerisation produces a polyelectrolyte hydrogel densely interpenetrating the tissue. (3) Mechanical homogenisation: proteinase K (or autoclaving in some variants) digests proteins to break tissue mechanical integrity while leaving the gel-anchored fluorophores intact. (4) Expansion: replacing the buffer with low-salt water causes the sodium-acrylate-rich gel to swell. The original Chen, Tillberg, Boyden 2015 protocol gave ~4.5× linear expansion, equivalent to ~70 nm effective resolution from a confocal microscope. Iterative-ExM (Chang et al. 2017) re-embeds the expanded gel in a second swellable matrix and expands again, reaching ~16–20× linear (effective ~25 nm) and even ~40× in proof-of-concept. Magnify-ExM (2022) reaches ~10× in a single round using a tougher gel chemistry that also retains lipid signals. The expanded sample can be imaged on any confocal, light-sheet, or even spinning-disk microscope — the magnification is physical, not optical.
In plain terms
Imagine taking a piece of cake and replacing all the crumbs with the kind of polymer that's inside a wet diaper. When you add water, the gel swells — the cake balloons up to four times its original size in every direction, but every cherry stays in exactly the same relative position. Now you can pick out individual cherries that used to look like a single blob. Expansion microscopy does the same trick with cells.
Also called: ExM, physical magnification microscopy, sample-expansion super-resolution.
Why it matters
- Democratises super-resolution microscopy: any lab with a confocal can image at sub-100 nm resolution.
- Compatible with multi-colour antibody labelling, RNA-FISH, and lipid markers (in newer variants).
- Stacks with light-sheet microscopy and CLARITY for synaptic-resolution whole-circuit imaging.
- Decouples resolution from optical NA limits — extending what visible-light microscopy can image at all.
Where it lives
- Boyden lab original ExM, iExM, Magnify-ExM (MIT)
- Standard ExM kits from many vendors (Sigma, Bio-Techne)
- Routine in neuroscience and structural-biology labs worldwide
Key numbers
- 4× ExM linear expansion: ~4–4.5× (Effective resolution ~70 nm)
- iExM linear expansion: ~16–20× (Effective resolution ~25 nm)
- Magnify-ExM linear expansion: ~10× (Single round, lipid-compatible)
- Distortion error: ≤4 % over 10 µm
- Hydrogel swelling agent: Sodium acrylate
Surprising facts
- Expansion microscopy turns a 250 nm-resolution confocal microscope into an effective 70 nm imaging instrument — and iterative ExM pushes that to ~25 nm, comparable to STED.
- The hydrogel used in ExM is chemically related to the absorbent material in baby diapers — both are polyelectrolyte sodium-polyacrylate networks that swell about 100× in volume in pure water.
- Expanded brain slices have been imaged at ~10 nm effective resolution, resolving individual proteins on the post-synaptic density — a structure ~300 nm across.
- ExM linearly preserves geometry to within a few percent over distances of tens of micrometres — meaning the swelling is uniform enough that distance measurements remain quantitatively meaningful after expansion.
Open questions
- Anisotropic swelling near membranes: ExM swelling is nearly isotropic over micrometre scales but exhibits ~5–10 % local anisotropy near hydrophobic membranes and lipid-rich structures — a complication for the most demanding sub-100 nm measurements.
- Fluorophore quenching loss: A non-trivial fraction of fluorophores are quenched or detached during proteinase digestion; the surviving fraction depends on antibody, fluorophore, and protocol in ways not fully predictable from chemistry.
A short history
Edward Boyden's group at MIT introduced expansion microscopy in Science in 2015 (Chen, Tillberg, Boyden). Iterative ExM followed in 2017 (Chang, Boyden et al.). Magnify-ExM (2022) and ExLR variants combining ExM with light-sheet microscopy enabled large-volume nanoscale imaging. As of 2026 ExM is a standard wet-lab method taught in graduate-level cell-biology courses.
Sources
- Chen, Tillberg & Boyden, Science (2015) — original ExM paper (https://www.science.org/doi/10.1126/science.1260088)
- Chang et al., Nat. Methods (2017) — iterative ExM (https://www.nature.com/articles/nmeth.4261)
- Klimas et al., Nat. Biotechnol. (2022) — Magnify-ExM (https://www.nature.com/articles/s41587-022-01546-1)
- Wassie, Zhao & Boyden, Nat. Methods (2019) — ExM applications review (https://www.nature.com/articles/s41592-018-0219-4)