Understanding in 3-D how molecules are configured throughout neurons, and how neurons are configured in circuits, may not only enable the discovery of new targets and technologies for treating neural diseases, but could help reveal fundamental principles of neural computation. Since biomolecules are nanoscale, however, and configured with nanoscale precision, this has remained difficult to study. For example, with electron microscopy, fantastic spatial resolution is possible, but it is difficult to identify the biomolecules in a protein complex. On the other hand, with optical microscopes the spatial resolution is limited to 300 nm due to the diffraction of light waves. Optical super-resolution techniques that overcome this limit face challenges in 3D scalability and require expensive hardware and/or are slow to image large scale specimens, which limits their application.

We recently discovered that it was possible to beat the diffraction limit through physical magnification of biological specimens, by embedding them in dense, swellable polyelectrolyte gels (Science (2015) 347(6221):543-548). The original process, which we called expansion microscopy (ExM), achieved a 4.5x linear expansion (i.e., a 300 nm diffraction limited lens would now have a resolution of 300 / 4.5 ≈ 60 nm). We also showed that, by iterating the polymerization and expansion process (iExM), we could achieve higher expansion factors (4.5 x 4.5 ~20x, enabling a resolution of 300/20 ~15 nm; Nature Methods (2017) 14, 593–599). However, iExM required us to discard the original biomolecules, replacing them with a polymer-anchored DNA oligo. This results in limited resolution due to the fact that the antibody must be administered first, and thus the size of the antibody becomes the key factor limiting resolution.

Here, we report a new form of iterative expansion microscopy which addresses these problems – enabling the preservation of biomolecules throughout the entire process, and also allowing for antibodies and other probes to be delivered at the end of the process, greatly improving resolution. Our new method, which we call iterated direct ExM (idExM), enables high expansion factors (20x to 100x) to be achieved, and may lead to resolution on the scale of individual biomolecules.

IdExM overcomes the limit of all previous super-resolution techniques, where the effective resolution is limited by the size of the labels (eg. primary and secondary antibodies which are about 20-30 nm in total size). This may, in principle, result in resolutions of 1 nm or less. IdExM also de-crowds biomolecules through iterative expansion, allowing access of antibodies to epitopes that may not otherwise be accessible for viewing by existing super-resolution methods. Using this technology, we revealed for the first time, detailed synaptic architectures in intact brain circuits, which influences the transport of neurotransmitters accross the synapse as well as nanoscale organization of transcriptomes.