High-contrast en bloc staining of mouse whole-brain and human brain samples for EM-based connectomics

Abstract

Connectomes of human cortical gray matter require high-contrast homogeneously stained samples sized at least 2 mm on a side, and a mouse whole-brain connectome requires samples sized at least 5–10 mm on a side. Here we report en bloc staining and embedding protocols for these and other applications, removing a key obstacle for connectomic analyses at the mammalian whole-brain level.

A Scalable Staining Strategy for Whole-Brain Connectomics

“…Using sodium cacodylate, the second osmication took over 14 days to complete (Figure S6), while sodium chloride reduced this duration to just 6 days. It has been hypothesized that cacodylate anions may coordinate with osmium ( 12 ), potentially increasing the hydrodynamic radius of osmium tetroxide and thereby reducing its diffusivity. Coordination with large anions may also introduce steric hindrance, which could decrease the reactivity of aqueous osmium with the densely packed osmium in the tissue.…”

“…Although this expansion can be accommodated by the elasticity of unstained tissue, samples stained with heavy metals become brittle and lose their elastic properties, resulting in disintegration. Previous whole-brain staining protocols have used water transitions as rinses before ( 12 ) and after ( 11 ) the pyrogallol step. In these protocols membrane breaks, fissures, fragility, and disintegration have been reported previously and confirmed by us (Figure S9).…”

“…In particular, the introduction of two rounds of osmication for enhanced membrane contrast ( 29, 40 ), the separation of osmium and ferrocyanide treatments to increase the penetration depth of staining ( 9 ), and the preservation of ECS to improve staining efficiency ( 36 ) have all advanced the field. Moreover, the recent attempts at staining a whole mouse brain have replaced TCH with pyrogallol for tissue conditioning ( 11, 12 ) and employed microCT X-ray monitoring of the staining process ( 42 ), greatly improving large volume staining and its analysis. Nevertheless, multiple challenges persist because, for a sample the size of a mouse brain, improving whole-brain staining quality often has deleterious effects on tissue integrity.…”

“…Previous whole-brain staining attempts employed Spurr's resin for tissue embedding and used block-face scanning electron microscopy for imaging (10)(11)(12). The lower viscosity of Spurr's resin, compared to the more commonly used Epon epoxy resin, allowed for easier and more complete infiltration into large brain samples.…”

“…A few years later, the BROPA protocol, while being able to reliably stain neuropil and myelin, generated suboptimal membrane contrast and large cracks in the stained brain ( 11 ). The recently developed whole-brain staining protocol by Song et al ( 12 ) is a big improvement over previous whole-brain staining, but still creates microbreakages that may extend over a length of 100 μm, and more problematic macroscopic tissue fragmentation in the anterior and posterior poles of the mouse brain. Without a completely intact and well stained brain, a large number of neuronal pathways will be disrupted.…”