In the past decade the image acquisition rates of biological electron microscopy facilities have been scaled to 10⁷–10⁹ pixels per second through parallelization and automation (Hayworth et al., 2014; Schalek et al., 2011; Ren and Kruit, 2016; Denk and Horstmann, 2004; Hayworth et al., 2020; Eberle et al., 2015; Graham et al., 2019; Xu et al., 2017). The increase in imaging throughput for biological samples has been mainly driven by the emerging Neuroscience field of connectomics which aims to densely reconstruct neuronal circuits with synaptic resolution (Briggman et al., 2011; Helmstaedter et al., 2013; Bock et al., 2011; Zheng et al., 2018; Scheffer et al., 2020; Kornfeld et al., 2017; Wanner and Friedrich, 2020; Lee et al., 2016; Kasthuri et al., 2015; Wilson et al., 2019; Schmidt et al., 2017; Svara et al., 2018; Vishwanathan et al., 2017). Also, the analysis of the terabyte-sized electron microscopy datasets produced by these studies is becoming increasingly automated using machine-learning and computer vision (Januszewski et al., 2018; Dorkenwald et al., 2020; Schubert et al., 2019; Dorkenwald et al., 2017; Berning et al., 2015; Staffler et al., 2017; Jain et al., 2010; Buhmann et al., 2020; Vergara et al., 2020; Turner et al., 2020). Both the image acquisition as well as the image analysis for these types of datasets are being scaled by parallelization to large samples on the order of several cubic millimeters or even entire brains. However, a remaining obstacle is the lack of reliable tissue processing and staining protocols for large samples where the smallest dimension is greater than 1 mm.

Existing en bloc electron microscopy staining protocols have been optimized for staining small samples with dimensions of less than 1 mm (Genoud et al., 2018; Hua et al., 2015; Deerinck et al., 2010; Tapia et al., 2012). Using aldehyde-stabilized cryopreservation (McIntyre and Fahy, 2015) the cellular ultrastructure can be preserved even in large tissue blocks with dimensions exceeding 1 mm and entire brains. Despite pioneering work on en bloc staining protocols for whole mouse brains (Mikula and Denk, 2015; Mikula et al., 2012), large en bloc stained samples still suffer from artifacts such as inhomogeneous staining and membrane or tissue cracks.

En bloc sample preparation for electron microscopy generally requires tissue fixation, staining and embedding in resin. First, the macromolecules in the tissue are stabilized via crosslinking by diffusing or perfusing buffered solutions of fixatives such as formaldehyde (Claude and Fullam, 1945; Fox et al., 1985) and glutaraldehyde (Sabatini et al., 1964). Because biological tissue is composed mostly of carbon and other low atomic number elements, the tissue is stained with heavy metals to increase the contrast in electron micrographs (Bahr, 1954; Porter et al., 1945). These heavy metal stains are composed of electron dense atoms such as osmium, lead or uranium. In addition, some of these heavy metals (e.g. osmium) also act as fixatives (Bahr, 1955). Finally, the stained tissue gets dehydrated and embedded in resin. The most commonly used resins are epoxy resins (Maaløe and Andersen, 1956), such as araldite (Glauert et al., 1956) or Epon (Finck, 1960) because of their thermal stability and electron transparency. To inspect the staining quality in an electron microscope, ultrathin sections (<100 nm) are typically collected from the embedded tissue using an ultramicrotome.

Many steps in the classic en bloc electron microscopy staining protocols are based on passive diffusion of chemicals into the tissue. Passive diffusion is one of the main bottlenecks for preparing large samples (smallest dimension >1 mm) for electron microscopy (Burkl and Schiechl, 1968; Medawar, 1941). For small samples (smallest dimension <1 mm) a typical en bloc staining protocol takes about 10–15 days including sectioning (Tapia et al., 2012). However, for large samples or whole brains, a diffusion-based staining protocol takes several weeks or even months (Mikula et al., 2012; Mikula and Denk, 2015; Masís et al., 2018).

To date, electron microscopy staining is a ‘black box’ and changes to staining protocols can only be assessed using an electron microscope, which in turn requires a sample to be run through the entire staining protocol. Conventional approaches for optimizing the parameters of staining protocols rely on sequential screening of hundreds of samples that have been processed end-to-end. However, sequential screening is very inefficient for months-long protocols.

Recently, X-ray based computed micro-tomography (μCT) has been introduced for a relatively fast, macroscopic assessment of staining quality and tissue integrity of resin-embedded whole mouse brains (Mikula and Denk, 2015; Kuan et al., 2020; Dyer et al., 2017). (Mikula and Denk, 2015) showed that the pixel intensities of serial-section EM images and the intensity of the corresponding reslice are similar. Building on this pioneering work, we developed in situ time-lapsed X-ray-assisted staining to observe the staining process while the samples are in the staining solution (Figure 1a). We used X-ray-assisted staining to explore the micro-scale tissue mechanics and the kinetics of the heavy metal diffusion and accumulation in large aldehyde-fixed brain tissue blocks, resulting in new insights on how different staining agents affect the tissue. X-ray-assisted staining opens the ‘black box’ of electron microscopy staining protocols. Each staining step can be monitored and assessed in real time. This enables in silico optimization of electron microscopy staining protocols, which will be particularly useful for the development of staining procedures for large biological samples such as whole brains (Figure 1g).

Figure 1 with 3 supplements see all

Download asset Open asset

Figure 1—source data 1

Experimentally measured accumulation of heavy metals (blue traces) in buffered 2% OsO₄ solution in n=13 different samples.

The corresponding simulated model was fitted individually to the experimental data of each sample (black traces). The first row of every sample shows the temporal profile of osmium accumulation at different cortical depths, while the second row shows the spatial profile of osmium accumulation at different times after staining onset. The black circles indicate the location of axon tracts that tended to accumulate osmium faster and stronger. The third row shows the measured spatial-temporal accumulation of osmium (left), the corresponding fitted model simulations (middle) and the distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig1-data1-v2.pdf

Download elife-72147-fig1-data1-v2.pdf

Figure 1—source data 2

Experimentally measured accumulation of heavy metals in a buffered solution of 2% OsO4 reduced with 2.5% potassium ferrocyanide K4[Fe(CN)6] (Ferro-redOs) in n=6 different samples.

The first row shows the temporal profile of osmium accumulation at different cortical depths, while the second row shows the spatial profile of osmium accumulation at different times after staining onset. The black circles indicate the location of axon tracts that tended to accumulate osmium faster and stronger. The third row shows the measured spatial-temporal accumulation of osmium (left) and the corresponding distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hoursr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig1-data2-v2.pdf

Download elife-72147-fig1-data2-v2.pdf

Figure 1—source data 3

Experimentally measured accumulation of heavy metals in buffered 2% Ferro-redOs solution with 5.56% formamide in n=4 different samples.

The first row of every sample shows the temporal profile of osmium accumulation at different cortical depths, while the second row shows the spatial profile of osmium accumulation at different times after staining onset. The black circles indicate the location of axon tracts that tended to accumulate osmium faster and stronger. The third row shows the measured spatial-temporal accumulation of osmium (left) and the corresponding distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hoursr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig1-data3-v2.pdf

Download elife-72147-fig1-data3-v2.pdf

Figure 1—source data 4

Experimentally measured accumulation of heavy metals in buffered 2% Ferro-redOs solution with 11.25% formamide in n=4 different samples.

The first row of every sample shows the temporal profile of osmium accumulation at different cortical depths, while the second row shows the spatial profile of osmium accumulation at different times after staining onset. The black circles indicate the location of axon tracts that tended to accumulate osmium faster and stronger. The third row shows the measured spatial-temporal accumulation of osmium (left) and the corresponding distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hoursr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig1-data4-v2.pdf

Download elife-72147-fig1-data4-v2.pdf

Figure 1—source data 5

Experimentally measured accumulation of heavy metals in a buffered solution of 2% OsO4 reduced with 2.5% potassium ferricyanide K3[Fe(CN)6] (Ferri-redOs) in n=2 different samples.

The first row shows the temporal profile of osmium accumulation at different cortical depths, while the second row shows the spatial profile of osmium accumulation at different times after staining onset. The black circles indicate the location of axon tracts that tended to accumulate osmium faster and stronger. The third row shows the measured spatial-temporal accumulation of osmium (left) and the corresponding distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hoursr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig1-data5-v2.pdf

Download elife-72147-fig1-data5-v2.pdf

We developed in situ time-lapsed X-ray-assisted staining with the goal of facilitating and accelerating the optimization of staining protocols for large biological samples such as whole mouse brains (Abbott et al., 2020). 4mm-brain punches of transcardially perfused mice were immersed in aldehyde fixatives for 36 hr. After washing the sample blocks with cacodylate buffer, the samples were immersed in staining solution and immediately placed in the acquisition chamber of a Zeiss Xradia 520 Versa 3D for X-ray microscopy (Figure 1a). The Xradia can be operated in two different modes:

• In the μCT mode, projections of the sample are acquired at different rotation angles in order to reconstruct a 3D computed tomograph of the sample. Depending on the required signal to noise and resolution, this mode allowed us to acquire approximately 1–2 computed tomographs per hour.

• In single-projection mode, a projection of the sample is acquired in a fixed position every few seconds without rotating the sample.

The advantage of the μCT mode is that arbitrary virtual reslices can be extracted to assess the staining progression in various parts of the sample. However, with our X-ray microscope, the temporal resolution in μCT mode was limited to 30–60 min per tomograph. Therefore we performed all quantitative experiments in single-projection mode where the temporal resolution for monitoring of the heavy metal diffusion and accumulation was on the order of a few seconds. Note, in this mode, the pixel intensities of the resulting X-ray projection images are accumulated through the entire sample and therefore the correspondence to the pixel intensities in thin serial-section EM images is less accurate than in μCT mode (Figure 1—figure supplement 1). As soon as the samples got placed in the staining solutions, the heavy metals started accumulating in the immersed tissue. The stained tissue absorbs more X-rays than the unstained tissue, resulting in a noticeable intensity difference that can be used to track the diffusion and accumulation of the heavy metals in the tissue (Figure 1b).

Monitoring staining kinetics and tissue deformation

After 20 hr of incubation the staining density of osmium is homogeneous across the first 1000 μm of cortical depth (Figure 1e). In addition, the ultrastructure is well preserved (Figure 1—figure supplement 2). Reduced osmium is another commonly used staining agent that is known to result in higher contrast for electron microscopy images than non-reduced OsO₄. Typically, a buffered 2% OsO₄ solution is reduced with 2.5% potassium ferrocyanide (K₄[Fe(CN)₆]) (Hua et al., 2015; Willingham and Rutherford, 1984; Mikula and Denk, 2015), which we will call Ferro-redOs throughout this manuscript. Consistent with previous reports on Ferro-redOs staining in large samples (Hua et al., 2015; Mikula and Denk, 2015) we found that for Ferro-redOs the accumulation of heavy metals peaks at a depth between 300 and 800 μm (Figure 1e), whereas the tissue above or below that depth is stained less. Traditionally, this band of more heavily stained tissue has been associated with precipitated osmium that hinders the diffusion and prevents homogenous staining (Genoud et al., 2018; Hua et al., 2015; Mikula and Denk, 2015), in particular deeper in the tissue. No such band is present in tissue immersed in non-reduced OsO₄. (Mikula and Denk, 2015) reported that homogeneous staining of large samples with Ferro-redOs was achieved by adding formamide to the Ferro-redOs solution. However, the underlying mechanisms by which formamide acts are not known (Mikula and Denk, 2015; Genoud et al., 2018). (Mikula and Denk, 2015) hypothesized that formamide might prevent precipitation by generally solubilizing compounds or that it might allow highly charged molecules to cross membranes more easily. Adding formamide to the Ferro-redOs solution indeed resulted in more homogeneous staining (Figure 1e, ), but the heavy metal density was lower than in osmium only. (Mikula and Denk, 2015) reported that the tissue tends to expand at high concentrations of formamide (>50%). But we found that the tissue expands even at lower formamide concentrations, while the amount of expansion depends on both the formamide concentration and the incubation duration (Figure 1f). For a concentration of 11.25% formamide, the sample height increased by about 15% within 20 hr of incubation, whereas for a formamide concentration of 5.56%, the tissue height expanded only by approximately 10% (Figure 1f). Note, in the first 3 hr of incubation in Ferro-redOs with 5.56% formamide the tissue actually shrank, suggesting that there are opposite forces acting on the tissue. Indeed, in Ferro-redOs solution without any formamide, the tissue shrank by about 5% in height (Figure 1f). In contrast, we found that in 2% buffered OsO₄ solution the sample height expands by about 5% (Figure 1f). However, we found no sample expansion if the reduced osmium solution was prepared with 2.5% potassium ferricyanide K₃[Fe(CN)₆]+ 8 a,b), another commonly used reducing agent for OsO₄.

Hua et al., 2015 suggested an alternative approach to achieve homogeneous staining in 1 mm brain punches with Ferro-redOs without formamide, in which osmium and the reducing agent are applied separately. First, the samples are immersed in buffered 2% OsO₄ solution for 90 min. Subsequently, the samples are placed for 90 min in buffered 2.5% K₄[Fe(CN)₆] without any washing step in-between. We repeated this procedure for 4 mm brain punches: First, the samples were incubated in 2% buffered OsO₄ for 22 hours resulting in homogeneous staining throughout the sample. Next, we placed the sample directly without any washing step in 2.5% buffered K₄[Fe(CN)₆] solution. (Hua et al., 2015) hypothesized that the main effect of reducing agents such as K₄[Fe(CN)₆] is to convert VIII-oxidized water-soluble osmium into an VI-oxidized water-soluble form which is thought to generate additional, non-polar OsO₂ to be deposited in the membrane increasing the heavy metal content and the contrast in electron microscopy. However, we found that the heavy metal density in the K₄[Fe(CN)₆] immersed samples did not increase, but rather decrease from the sample surface towards the center as time progressed (Figure 2a). This suggests that K₄[Fe(CN)₆] removes or “washes out” osmium from the sample (Litman and Barrnett, 1972) resulting in an inverse staining gradient. This “washing effect” is stronger for longer incubation times as well as close to the sample surface and around blood vessels (anecdotal observation in electron microscopy images, data not shown). For incubation times longer than 12 hr, K₄[Fe(CN)₆] readily dissolves and disintegrates the tissue (Figure 2—figure supplement 1). Interestingly, no ‘washing effect’ was observed if the same procedure was repeated with 2.5% potassium ferricyanide K₃[Fe(CN)₆], but the sample height shrank by about 2% (Figure 1—figure supplement 3,f, ). Similarly, only a small reduction in heavy metal density could be observed when the sample was washed with double-distilled H₂O for 20 hr (Figure 2—figure supplements 2 and 3).

Figure 2 with 5 supplements see all

Download asset Open asset

Figure 2—source data 1

Experimentally measured washout of heavy metals in buffered 2.5% potassium ferrocyanide K₄[Fe(CN)₆] solution in n=6 different samples that have previously been stained with buffered OsO₄ for 22 hr.

The first row of every sample shows the temporal profile of osmium washout at different cortical depths, while the second row shows the spatial profile of osmium washout at different times after staining onset. The black circles indicate the location of axon tracts. The third row shows the measured spatial-temporal washout of osmium (left) and the corresponding distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig2-data1-v2.pdf

Download elife-72147-fig2-data1-v2.pdf

Figure 2—source data 2

Experimentally measured washout of heavy metals in buffered 2.5% potassium ferricyanide K3[Fe(CN)6] solution in n=4 different samples that have previously been stained with buffered OsO4 for 22 hrs.

The first row of every sample shows the temporal profile of osmium washout at different cortical depths, while the second row shows the spatial profile of osmium washout at different times after staining onset. The black circles indicate the location of axon tracts. The third row shows the measured spatial-temporal washout of osmium (left) and the corresponding distribution of osmium in projection images of the sample at the beginning of the experiment and after about 22 hoursr of incubation (right).

https://cdn.elifesciences.org/articles/72147/elife-72147-fig2-data2-v2.pdf

Download elife-72147-fig2-data2-v2.pdf

In situ X-ray-assisted staining is a new tool for monitoring the different steps of en bloc staining protocols for electron microscopy. By opening the ‘black box’ of staining protocols, X-ray-assisted staining enables, for the first time, to observe directly the diffusion of heavy metals and their microscopic effects on the tissue. While the technique will be useful to develop novel staining protocols for small to intermediate sized samples with dimensions less than 1 mm, it will particularly accelerate and facilitate the development and optimization of new staining protocols for large biological samples such as whole brains (Figure 1g).

We monitored the accumulation of osmium in different commonly used staining solutions and found significant differences in the accumulation densities. In addition, we found that the tissue expands and shrinks significantly depending on the composition of the staining solution. In particular, we found that the tissue height increased by 5% in buffered osmium solution, while the tissue height decreased by 5% in buffered Ferro-redOs solution. We hypothesize that this shrinkage could provide a mechanistic explanation for the long standing issue of diffusion barrier formation in large samples during Ferro-redOs incubation.

The method also revealed that K₄[Fe(CN)₆], a reducing agent that was believed to enhance the accumulation of osmium in membranes, in contrast can wash out osmium from stained tissue if applied separately. If K₄[Fe(CN)₆] removes osmium from the sample, why does the protocol modification of Hua et al., 2015 result in enhanced electron microscopy contrast for samples with dimensions of up to 1 mm? A potential explanation could be that K₄[Fe(CN)₆] clears the cytosol more efficiently than the cell membranes. As a consequence, the osmium content in the cytosol gets washed out more/faster, which might result in an overall electron microscopic contrast increase between the membranes and the cytosol. However, the chemical mechanisms by which K₄[Fe(CN)₆] interact with the tissue remain poorly understood and shining a light on those processes will require additional work.

X-ray assisted staining opens the door for in silico staining protocol optimization, by enabling to measure important heavy metal staining parameters such as the effective diffusion coefficients and reaction rate constants directly in situ. We derived a diffusion-reaction-advection model that describes the accumulation of osmium in homogeneous tissue by four distinct, but coupled, processes. Osmium diffuses passively into the tissue and binds there quickly to available binding sites, for example unsaturated lipids. Thereby the tissue expands by about 5% in height. Additional binding sites are being unmasked over the course of hours, which enables the tissue to accumulate additional osmium. While the binding of osmium appears to be nearly instantaneous, the passive diffusion and unmasking reactions are slow and appear to be rate-limiting for the staining. The mechanisms and reactions underlying the unmasking of additional binding sites for osmium are not known. Unsaturated lipids, one of the prominent binding sites for osmium, can be masked by lipid-protein complexes. In the presence of osmium, these complexes could dissociate and make the previously masked unsaturated lipid available for binding osmium (Clayton, 1959; Lauder and Beynon, 1993; Ashhurst, 1961; Wigglesworth, 1975). A recent study used time-of-flight secondary ion mass spectrometry to analyze the colocalization of different fatty acid species with osmium at the tissue surface (Belazi et al., 2009). As expected, the unsaturated fatty acids colocalize well with the tissue-bound osmium. However, they also report a remarkable complementary clustering of saturated and unsaturated fatty acids in osmium stained tissue that appears not to be present in unstained tissue. Hence, it could be that osmium can convert saturated lipids to unsaturated lipids through hydrogenation/oxidation. Alternatively, the masked binding sites could also be interpreted as a different binding partner that reacts much slower with osmium.

Biological samples such as brains are non-homogeneous media. They are composed of different tissue and cell types and therefore the diffusion of staining agents is expected to be heterogeneous. Nevertheless, the fitted model parameters reproduce the observed kinetics of osmium accumulation relatively well up to a depth of about 700–800 μm from the surface. Deeper in the tissue osmium that diffused from secondary sources such as ventricles or axon tracts appear to contribute to the accumulation of osmium. For example, we found that white matter and myelinated axon tracks tend to accumulate osmium faster and stronger compared to gray matter (, ). Such secondary sources and tissue inhomogeneities are not considered in our model. Still, the fitted model parameters appear consistent with comparable values from the literature, although we also find some discrepancies. For example, in mouse brains we expect to find about 117.975 nmol/mm³ double-bonds on unsaturated lipids (see methods for the derivation). Because the dominant binding reaction for osmium is assumed to be the formation of mono- and di-ester bonds with unsaturated lipids which has been described to result in the deposition of OsO₂, the density of double bonds should roughly correspond to the density of available binding sites. But the proposed model predicts 2.999 times more binding sites, even without unmasking. One interpretation would be that unsaturated lipids are not the dominant reaction partner and that other reaction partners such particular amino acids or nucleotides account for about two third of the available binding sites. Alternatively, this could indicate that each unsaturated lipid double bond gets associated with on average 3 osmium atoms, for example through Os₃O₅^- and Os₃O₅H^- isotopes that have previously reported to exist in osmium-stained adipose mouse tissue (Belazi et al., 2009). Similarly, it has been reported that if the mono- and di-ester products are left standing in solution, a highly insoluble trimer osmium product is formed (Schroeder, 1980).

We expect that X-ray-assisted staining will be very useful in establishing a more refined picture of the heterogeneity of the heavy metal diffusion processes in different tissue types and compartments as well as in different conditions. For example, we also measured and modeled the accumulation of osmium in liver tissue (Figure 2—figure supplement 5). It will also be interesting to examine how changes in the extracellular space and different fixation protocols or microwave-assisted incubation affect the diffusion dynamics (Korogod et al., 2015; Fulton and Briggman, 2020; Jensen and Harris, 1989; Login and Dvorak, 1988).

Furthermore, X-ray-assisted staining can be used at higher resolution and in combination with computed tomography for producing a more detailed picture of the tissue mechanics involved in different staining steps. We and several other labs have noticed that brain tissue expands significantly during incubation in ‘osmium amplifiers’ such as the osmiophilic thiocarbohydrazide (TCH) (Seligman et al., 1966). (Mikula and Denk, 2015) reported that nitrogen bubbles are formed during TCH incubation that could be responsible for the formation of tissue cracks in large samples. X-ray-assisted staining could be used to detect and monitor the formation of tissue cracks and to explore ways to reduce these artifacts in large samples.

Aside from accelerating the development of novel staining protocols this new method will also allow for improved closed-loop quality control of the staining process. Large samples tend to be biologically more variable and more difficult to precisely fine-tune all the parameters in the staining protocol ahead of time to account for these individual differences. Hence, X-ray-assisted staining can be used to adjust the parameters in real time to compensate for these individual differences. This is particularly important for experiments in which the sample yield is low or the individual sample is precious, as for example in the case of human brains or functionally imaged brains of behaving animals that require several months of behavioral training. However, this will require calibrating what is measured with X-ray microscopy to the corresponding appearance at an electron microscopic level, because the image formation process and the resulting contrast, signal-to-noise, etc. for both imaging modalities depend not only on the staining protocol, but also on the imaging modality, energy, flux, etc. (Du and Jacobsen, 2018).

The code to analyze the X-ray projection images and to model the accumulation of heavy metals can be found on https://github.com/adwanner/XrayAssistedStaining, (copy archived at swh:1:rev:62c9930d08bbdd1602614d515cd20b1cb1f89d15). All X-ray data is available for download on https://www.ebi.ac.uk/empiar/ with dataset ID EMPIAR-10782.

1. 1. Ströh S
   2. Hammerschmith EW
   3. Tank DW
   4. Seung HS
   5. Wanner AA

   (2022) EMPIAR

   ID EMPIAR-10782. In situ X-ray assisted electron microscopy staining for large biological samples.

   https://www.ebi.ac.uk/empiar/EMPIAR-10782/

1. 1. Abbott LF
   2. Bock DD
   3. Callaway EM
   4. Denk W
   5. Dulac C
   6. Fairhall AL
   7. Fiete I
   8. Harris KM
   9. Helmstaedter M
   10. Jain V
   11. Kasthuri N
   12. LeCun Y
   13. Lichtman JW
   14. Littlewood PB
   15. Luo L
   16. Maunsell JHR
   17. Reid RC
   18. Rosen BR
   19. Rubin GM
   20. Sejnowski TJ
   21. Seung HS
   22. Svoboda K
   23. Tank DW
   24. Tsao D
   25. Van Essen DC

   (2020) The mind of a mouse

   Cell 182:1372–1376.

   https://doi.org/10.1016/j.cell.2020.08.010

   • PubMed
   • Google Scholar
2. 1. Ashhurst DE

   (1961) The cytology and histochemistry of the neurones of Periplaneta americana

   Journal of Cell Science s3-102:399–405.

   https://doi.org/10.1242/jcs.s3-102.59.399

   • Google Scholar
3. 1. Badea A
   2. Ali-Sharief AA
   3. Johnson GA

   (2007) Morphometric analysis of the C57BL/6J mouse brain

   NeuroImage 37:683–693.

   https://doi.org/10.1016/j.neuroimage.2007.05.046

   • PubMed
   • Google Scholar
4. 1. Bahr GF

   (1954) Osmium tetroxide and ruthenium tetroxide and their reactions with biologically important substances. electron stains. III

   Experimental Cell Research 7:457–479.

   https://doi.org/10.1016/s0014-4827(54)80091-7

   • PubMed
   • Google Scholar
5. 1. Bahr GF

   (1955) Continued studies about the fixation with osmium tetroxide

   Experimental Cell Research 9:277–285.

   https://doi.org/10.1016/0014-4827(55)90100-2

   • PubMed
   • Google Scholar
6. 1. Barceló-Coblijn G
   2. Golovko MY
   3. Weinhofer I
   4. Berger J
   5. Murphy EJ

   (2007) Brain neutral lipids mass is increased in alpha-synuclein gene-ablated mice

   Journal of Neurochemistry 101:132–141.

   https://doi.org/10.1111/j.1471-4159.2006.04348.x

   • PubMed
   • Google Scholar
7. 1. Belazi D
   2. Solé-Domènech S
   3. Johansson B
   4. Schalling M
   5. Sjövall P

   (2009) Chemical analysis of osmium tetroxide staining in adipose tissue using imaging TOF-SIMS

   Histochemistry and Cell Biology 132:105–115.

   https://doi.org/10.1007/s00418-009-0587-z

   • PubMed
   • Google Scholar
8. 1. Berning M
   2. Boergens KM
   3. Helmstaedter M

   (2015) SegEM: efficient image analysis for high-resolution connectomics

   Neuron 87:1193–1206.

   https://doi.org/10.1016/j.neuron.2015.09.003

   • PubMed
   • Google Scholar
9. 1. Bock DD
   2. Lee W-CA
   3. Kerlin AM
   4. Andermann ML
   5. Hood G
   6. Wetzel AW
   7. Yurgenson S
   8. Soucy ER
   9. Kim HS
   10. Reid RC

   (2011) Network anatomy and in vivo physiology of visual cortical neurons

   Nature 471:177–182.

   https://doi.org/10.1038/nature09802

   • PubMed
   • Google Scholar
10. 1. Briggman KL
    2. Helmstaedter M
    3. Denk W

    (2011) Wiring specificity in the direction-selectivity circuit of the retina

    Nature 471:183–188.

    https://doi.org/10.1038/nature09818

    • PubMed
    • Google Scholar
11. Preprint

    1. Buhmann J
    2. Sheridan A
    3. Gerhard S
    4. Krause R
    5. Nguyen T
    6. Heinrich L
    7. Schlegel P
    8. Lee WCA
    9. Wilson R
    10. Saalfeld S
    11. Jefferis G
    12. Bock D
    13. Turaga S
    14. Cook M
    15. Funke J

    (2020) Automatic Detection of Synaptic Partners in a Whole-Brain Drosophila EM Dataset

    bioRxiv.

    https://doi.org/10.1101/2019.12.12.874172

    • Google Scholar
12. 1. Burkl W
    2. Schiechl H

    (1968) A study of osmium tetroxide fixation

    The Journal of Histochemistry and Cytochemistry 16:157–161.

    https://doi.org/10.1177/16.3.157

    • PubMed
    • Google Scholar
13. 1. Carnevale JH
    2. Cole ER
    3. Crank G

    (1979) Fluorescent light catalyzed autoxidation of beta-carotene

    Journal of Agricultural and Food Chemistry 27:462–463.

    https://doi.org/10.1021/jf60222a042

    • PubMed
    • Google Scholar
14. 1. Claude A
    2. Fullam EF

    (1945) AN electron microscope study of isolated mitochondria

    Journal of Experimental Medicine 81:51–62.

    https://doi.org/10.1084/jem.81.1.51

    • Google Scholar
15. 1. Clayton B-P

    (1959) The action of fixatives on the unmasking of lipid

    Journal of Cell Science s3-100:269–274.

    https://doi.org/10.1242/jcs.s3-100.50.269

    • Google Scholar
16. 1. Deerinck TJ
    2. Bushong EA
    3. Lev-Ram V
    4. Shu X
    5. Tsien RY
    6. Ellisman MH

    (2010) Enhancing serial block-face scanning electron microscopy to enable high resolution 3-D nanohistology of cells and tissues

    Microscopy and Microanalysis 16:1138–1139.

    https://doi.org/10.1017/S1431927610055170

    • Google Scholar
17. 1. Denk W
    2. Horstmann H

    (2004) Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure

    PLOS Biology 2:e329.

    https://doi.org/10.1371/journal.pbio.0020329

    • PubMed
    • Google Scholar
18. 1. Dorkenwald S
    2. Schubert PJ
    3. Killinger MF
    4. Urban G
    5. Mikula S
    6. Svara F
    7. Kornfeld J

    (2017) Automated synaptic connectivity inference for volume electron microscopy

    Nature Methods 14:435–442.

    https://doi.org/10.1038/nmeth.4206

    • PubMed
    • Google Scholar
19. Preprint

    1. Dorkenwald S
    2. McKellar C
    3. Macrina T
    4. Kemnitz N
    5. Lee K
    6. Lu R
    7. Wu J
    8. Popovych S
    9. Mitchell E
    10. Nehoran B
    11. Jia Z
    12. Bae JA
    13. Mu S
    14. Ih D
    15. Castro M
    16. Ogedengbe O
    17. Halageri A
    18. Ashwood Z
    19. Zung J
    20. Brittain D
    21. Collman F
    22. Schneider-Mizell C
    23. Jordan C
    24. Silversmith W
    25. Baker C
    26. Deutsch D
    27. Encarnacion-Rivera L
    28. Kumar S
    29. Burke A
    30. Gager J
    31. Hebditch J
    32. Koolman S
    33. Moore M
    34. Morejohn S
    35. Silverman B
    36. Willie K
    37. Willie R
    38. Yu S
    39. Murthy M
    40. Seung HS

    (2020) FlyWire: Online Community for Whole-Brain Connectomics

    bioRxiv.

    https://doi.org/10.1101/2020.08.30.274225

    • Google Scholar
20. 1. Du M
    2. Jacobsen C

    (2018) Relative merits and limiting factors for X-ray and electron microscopy of thick, hydrated organic materials

    Ultramicroscopy 184:293–309.

    https://doi.org/10.1016/j.ultramic.2017.10.003

    • PubMed
    • Google Scholar
21. 1. Dyer EL
    2. Gray Roncal W
    3. Prasad JA
    4. Fernandes HL
    5. Gürsoy D
    6. De Andrade V
    7. Fezzaa K
    8. Xiao X
    9. Vogelstein JT
    10. Jacobsen C
    11. Körding KP
    12. Kasthuri N

    (2017) Quantifying mesoscale neuroanatomy using X-ray microtomography

    ENeuro 4:ENEURO.0195-17.2017.

    https://doi.org/10.1523/ENEURO.0195-17.2017

    • PubMed
    • Google Scholar
22. 1. Eberle AL
    2. Mikula S
    3. Schalek R
    4. Lichtman J
    5. Tate MLK
    6. Zeidler D

    (2015) High-resolution, high-throughput imaging with a multibeam scanning electron microscope

    Journal of Microscopy 259:114–120.

    https://doi.org/10.1111/jmi.12224

    • PubMed
    • Google Scholar
23. 1. Finck H

    (1960) Epoxy resins in electron microscopy

    The Journal of Biophysical and Biochemical Cytology 7:27–30.

    https://doi.org/10.1083/jcb.7.1.27

    • PubMed
    • Google Scholar
24. 1. Fitzner D
    2. Bader JM
    3. Penkert H
    4. Bergner CG
    5. Su M
    6. Weil MT
    7. Surma MA
    8. Mann M
    9. Klose C
    10. Simons M

    (2020) Cell-type- and brain-region-resolved mouse brain lipidome

    Cell Reports 32:108132.

    https://doi.org/10.1016/j.celrep.2020.108132

    • PubMed
    • Google Scholar
25. 1. Fox CH
    2. Johnson FB
    3. Whiting J
    4. Roller PP

    (1985) Formaldehyde fixation

    The Journal of Histochemistry and Cytochemistry 33:845–853.

    https://doi.org/10.1177/33.8.3894502

    • PubMed
    • Google Scholar
26. Preprint

    1. Fulton KA
    2. Briggman KL

    (2020) Permeabilization-Free En Bloc Immunohistochemistry for Correlative Microscopy

    bioRxiv.

    https://doi.org/10.1101/2020.10.07.327734

    • Google Scholar
27. 1. Genoud C
    2. Titze B
    3. Graff-Meyer A
    4. Friedrich RW

    (2018) Fast homogeneous en bloc staining of large tissue samples for volume electron microscopy

    Frontiers in Neuroanatomy 12:76.

    https://doi.org/10.3389/fnana.2018.00076

    • PubMed
    • Google Scholar
28. 1. Glauert AM
    2. Glauert RH
    3. Rogers GE

    (1956) A new embedding medium for electron microscopy

    Nature 178:803.

    https://doi.org/10.1038/178803a0

    • PubMed
    • Google Scholar
29. Preprint

    1. Graham BJ
    2. Hildebrand DGC
    3. Kuan AT
    4. Maniates-Selvin JT
    5. Thomas LA
    6. Shanny BL
    7. Lee WCA

    (2019) High-Throughput Transmission Electron Microscopy with Automated Serial Sectioning

    bioRxiv.

    https://doi.org/10.1101/657346

    • Google Scholar
30. 1. Hayworth KJ
    2. Morgan JL
    3. Schalek R
    4. Berger DR
    5. Hildebrand DGC
    6. Lichtman JW

    (2014) Imaging ATUM ultrathin section libraries with wafermapper: a multi-scale approach to em reconstruction of neural circuits

    Frontiers in Neural Circuits 8:68.

    https://doi.org/10.3389/fncir.2014.00068

    • PubMed
    • Google Scholar
31. 1. Hayworth KJ
    2. Peale D
    3. Januszewski M
    4. Knott GW
    5. Lu Z
    6. Xu CS
    7. Hess HF

    (2020) Gas cluster ion beam SEM for imaging of large tissue samples with 10 nm isotropic resolution

    Nature Methods 17:68–71.

    https://doi.org/10.1038/s41592-019-0641-2

    • PubMed
    • Google Scholar
32. 1. Helmstaedter M
    2. Briggman KL
    3. Turaga SC
    4. Jain V
    5. Seung HS
    6. Denk W

    (2013) Connectomic reconstruction of the inner plexiform layer in the mouse retina

    Nature 500:168–174.

    https://doi.org/10.1038/nature12346

    • PubMed
    • Google Scholar
33. 1. Hua Y
    2. Laserstein P
    3. Helmstaedter M

    (2015) Large-Volume en-bloc staining for electron microscopy-based connectomics

    Nature Communications 6:7923.

    https://doi.org/10.1038/ncomms8923

    • PubMed
    • Google Scholar
34. 1. Jain V
    2. Seung HS
    3. Turaga SC

    (2010) Machines that learn to segment images: a crucial technology for connectomics

    Current Opinion in Neurobiology 20:653–666.

    https://doi.org/10.1016/j.conb.2010.07.004

    • PubMed
    • Google Scholar
35. 1. Januszewski M
    2. Kornfeld J
    3. Li PH
    4. Pope A
    5. Blakely T
    6. Lindsey L
    7. Maitin-Shepard J
    8. Tyka M
    9. Denk W
    10. Jain V

    (2018) High-Precision automated reconstruction of neurons with flood-filling networks

    Nature Methods 15:605–610.

    https://doi.org/10.1038/s41592-018-0049-4

    • PubMed
    • Google Scholar
36. 1. Jensen FE
    2. Harris KM

    (1989) Preservation of neuronal ultrastructure in hippocampal slices using rapid microwave-enhanced fixation

    Journal of Neuroscience Methods 29:217–230.

    https://doi.org/10.1016/0165-0270(89)90146-5

    • PubMed
    • Google Scholar
37. 1. Kasthuri N
    2. Hayworth KJ
    3. Berger DR
    4. Schalek RL
    5. Conchello JA
    6. Knowles-Barley S
    7. Lee D
    8. Vázquez-Reina A
    9. Kaynig V
    10. Jones TR
    11. Roberts M
    12. Morgan JL
    13. Tapia JC
    14. Seung HS
    15. Roncal WG
    16. Vogelstein JT
    17. Burns R
    18. Sussman DL
    19. Priebe CE
    20. Pfister H
    21. Lichtman JW

    (2015) Saturated reconstruction of a volume of neocortex

    Cell 162:648–661.

    https://doi.org/10.1016/j.cell.2015.06.054

    • PubMed
    • Google Scholar
38. 1. Khan AA
    2. Riemersma JC
    3. Booij HL

    (1961) The reactions of osmium tetroxide with lipids and other compounds

    The Journal of Histochemistry and Cytochemistry 9:560–563.

    https://doi.org/10.1177/9.5.560

    • PubMed
    • Google Scholar
39. 1. Korn ED

    (1967) A chromatographic and spectrophotometric study of the products of the reaction of osmium tetroxide with unsaturated lipids

    The Journal of Cell Biology 34:627–638.

    https://doi.org/10.1083/jcb.34.2.627

    • PubMed
    • Google Scholar
40. 1. Kornfeld J
    2. Benezra SE
    3. Narayanan RT
    4. Svara F
    5. Egger R
    6. Oberlaender M
    7. Denk W
    8. Long MA

    (2017) EM connectomics reveals axonal target variation in a sequence-generating network

    eLife 6:e24364.

    https://doi.org/10.7554/eLife.24364

    • PubMed
    • Google Scholar
41. 1. Korogod N
    2. Petersen CCH
    3. Knott GW

    (2015) Ultrastructural analysis of adult mouse neocortex comparing aldehyde perfusion with cryo fixation

    eLife 4:e05793.

    https://doi.org/10.7554/eLife.05793

    • PubMed
    • Google Scholar
42. 1. Kuan AT
    2. Phelps JS
    3. Thomas LA
    4. Nguyen TM
    5. Han J
    6. Chen C-L
    7. Azevedo AW
    8. Tuthill JC
    9. Funke J
    10. Cloetens P
    11. Pacureanu A
    12. Lee W-CA

    (2020) Dense neuronal reconstruction through X-ray holographic nano-tomography

    Nature Neuroscience 23:1637–1643.

    https://doi.org/10.1038/s41593-020-0704-9

    • PubMed
    • Google Scholar
43. 1. Lauder RM
    2. Beynon AD

    (1993) Unmasking and abolition of sudanophilia of the mineralizing front

    Biotechnic & Histochemistry 68:180–185.

    https://doi.org/10.3109/10520299309104693

    • PubMed
    • Google Scholar
44. 1. Lee WCA
    2. Bonin V
    3. Reed M
    4. Graham BJ
    5. Hood G
    6. Glattfelder K
    7. Reid RC

    (2016) Anatomy and function of an excitatory network in the visual cortex

    Nature 532:370–374.

    https://doi.org/10.1038/nature17192

    • PubMed
    • Google Scholar
45. 1. Litman RB
    2. Barrnett RJ

    (1972) The mechanism of the fixation of tissue components by osmium tetroxide via hydrogen bonding

    Journal of Ultrastructure Research 38:63–86.

    https://doi.org/10.1016/s0022-5320(72)90084-6

    • PubMed
    • Google Scholar
46. 1. Login GR
    2. Dvorak AM

    (1988) Microwave fixation provides excellent preservation of tissue, cells and antigens for light and electron microscopy

    The Histochemical Journal 20:373–387.

    https://doi.org/10.1007/BF01002732

    • PubMed
    • Google Scholar
47. 1. Maaløe O
    2. Andersen B

    (1956)

    Symposium of the Society for General Microbiology

    Bacterial anatomy, Symposium of the Society for General Microbiology, Cambridge, University Press.

    • Google Scholar
48. 1. Masís J
    2. Mankus D
    3. Wolff SBE
    4. Guitchounts G
    5. Joesch M
    6. Cox DD

    (2018) A micro-CT-based method for quantitative brain lesion characterization and electrode localization

    Scientific Reports 8:5184.

    https://doi.org/10.1038/s41598-018-23247-z

    • Google Scholar
49. 1. McIntyre RL
    2. Fahy GM

    (2015) Aldehyde-stabilized cryopreservation

    Cryobiology 71:448–458.

    https://doi.org/10.1016/j.cryobiol.2015.09.003

    • PubMed
    • Google Scholar
50. 1. Medawar P

    (1941) III.-the rate of penetration of fixatives

    Journal of the Royal Microscopical Society 61:46–57.

    https://doi.org/10.1111/j.1365-2818.1941.tb00884.x

    • Google Scholar
51. 1. Mikula S
    2. Binding J
    3. Denk W

    (2012) Staining and embedding the whole mouse brain for electron microscopy

    Nature Methods 9:1198–1201.

    https://doi.org/10.1038/nmeth.2213

    • PubMed
    • Google Scholar
52. 1. Mikula S
    2. Denk W

    (2015) High-Resolution whole-brain staining for electron microscopic circuit reconstruction

    Nature Methods 12:541–546.

    https://doi.org/10.1038/nmeth.3361

    • PubMed
    • Google Scholar
53. 1. Palade GE

    (1952) A study of fixation for electron microscopy

    The Journal of Experimental Medicine 95:285–298.

    https://doi.org/10.1084/jem.95.3.285

    • PubMed
    • Google Scholar
54. 1. Porter KR
    2. Claude A
    3. Fullam EF

    (1945) A study of tissue culture cells by electron microscopy

    Journal of Experimental Medicine 81:233–246.

    https://doi.org/10.1084/jem.81.3.233

    • Google Scholar
55. 1. Ren Y
    2. Kruit P

    (2016) Transmission electron imaging in the delft multibeam scanning electron microscope 1

    Journal of Vacuum Science & Technology B, Nanotechnology and Microelectronics 34:06KF02.

    https://doi.org/10.1116/1.4966216

    • Google Scholar
56. 1. Sabatini DD
    2. Miller F
    3. Barrnett RJ

    (1964) ALDEHYDE fixation for morphological and enzyme histochemical studies with the electron microscope

    The Journal of Histochemistry and Cytochemistry 12:57–71.

    https://doi.org/10.1177/12.2.57

    • PubMed
    • Google Scholar
57. 1. Schalek R
    2. Kasthuri N
    3. Hayworth K
    4. Berger D
    5. Tapia J
    6. Morgan J
    7. Turaga S
    8. Fagerholm E
    9. Seung H
    10. Lichtman J

    (2011) Development of high-throughput, high-resolution 3D reconstruction of large-volume biological tissue using automated tape collection ultramicrotomy and scanning electron microscopy

    Microscopy and Microanalysis 17:966–967.

    https://doi.org/10.1017/S1431927611005708

    • Google Scholar
58. 1. Scheffer LK
    2. Xu CS
    3. Januszewski M
    4. Lu Z
    5. Takemura S-Y
    6. Hayworth KJ
    7. Huang GB
    8. Shinomiya K
    9. Maitlin-Shepard J
    10. Berg S
    11. Clements J
    12. Hubbard PM
    13. Katz WT
    14. Umayam L
    15. Zhao T
    16. Ackerman D
    17. Blakely T
    18. Bogovic J
    19. Dolafi T
    20. Kainmueller D
    21. Kawase T
    22. Khairy KA
    23. Leavitt L
    24. Li PH
    25. Lindsey L
    26. Neubarth N
    27. Olbris DJ
    28. Otsuna H
    29. Trautman ET
    30. Ito M
    31. Bates AS
    32. Goldammer J
    33. Wolff T
    34. Svirskas R
    35. Schlegel P
    36. Neace E
    37. Knecht CJ
    38. Alvarado CX
    39. Bailey DA
    40. Ballinger S
    41. Borycz JA
    42. Canino BS
    43. Cheatham N
    44. Cook M
    45. Dreher M
    46. Duclos O
    47. Eubanks B
    48. Fairbanks K
    49. Finley S
    50. Forknall N
    51. Francis A
    52. Hopkins GP
    53. Joyce EM
    54. Kim S
    55. Kirk NA
    56. Kovalyak J
    57. Lauchie SA
    58. Lohff A
    59. Maldonado C
    60. Manley EA
    61. McLin S
    62. Mooney C
    63. Ndama M
    64. Ogundeyi O
    65. Okeoma N
    66. Ordish C
    67. Padilla N
    68. Patrick CM
    69. Paterson T
    70. Phillips EE
    71. Phillips EM
    72. Rampally N
    73. Ribeiro C
    74. Robertson MK
    75. Rymer JT
    76. Ryan SM
    77. Sammons M
    78. Scott AK
    79. Scott AL
    80. Shinomiya A
    81. Smith C
    82. Smith K
    83. Smith NL
    84. Sobeski MA
    85. Suleiman A
    86. Swift J
    87. Takemura S
    88. Talebi I
    89. Tarnogorska D
    90. Tenshaw E
    91. Tokhi T
    92. Walsh JJ
    93. Yang T
    94. Horne JA
    95. Li F
    96. Parekh R
    97. Rivlin PK
    98. Jayaraman V
    99. Costa M
    100. Jefferis GS
    101. Ito K
    102. Saalfeld S
    103. George R
    104. Meinertzhagen IA
    105. Rubin GM
    106. Hess HF
    107. Jain V
    108. Plaza SM

    (2020) A connectome and analysis of the adult Drosophila central brain

    eLife 9:e57443.

    https://doi.org/10.7554/eLife.57443

    • PubMed
    • Google Scholar
59. 1. Schmidt H
    2. Gour A
    3. Straehle J
    4. Boergens KM
    5. Brecht M
    6. Helmstaedter M

    (2017) Axonal synapse sorting in medial entorhinal cortex

    Nature 549:469–475.

    https://doi.org/10.1038/nature24005

    • PubMed
    • Google Scholar
60. 1. Schroeder M

    (1980) Osmium tetraoxide cis hydroxylation of unsaturated substrates

    Chemical Reviews 80:187–213.

    https://doi.org/10.1021/cr60324a003

    • Google Scholar
61. 1. Schubert PJ
    2. Dorkenwald S
    3. Januszewski M
    4. Jain V
    5. Kornfeld J

    (2019) Learning cellular morphology with neural networks

    Nature Communications 10:2736.

    https://doi.org/10.1038/s41467-019-10836-3

    • Google Scholar
62. 1. Seligman AM
    2. Wasserkrug HL
    3. Hanker JS

    (1966) A new staining method (OTO) for enhancing contrast of lipid--containing membranes and droplets in osmium tetroxide--fixed tissue with osmiophilic thiocarbohydrazide(tch)

    The Journal of Cell Biology 30:424–432.

    https://doi.org/10.1083/jcb.30.2.424

    • PubMed
    • Google Scholar
63. 1. Staffler B
    2. Berning M
    3. Boergens KM
    4. Gour A
    5. Smagt P
    6. Helmstaedter M

    (2017) SynEM, automated synapse detection for connectomics

    eLife 6:e26414.

    https://doi.org/10.7554/eLife.26414

    • PubMed
    • Google Scholar
64. 1. Svara FN
    2. Kornfeld J
    3. Denk W
    4. Bollmann JH

    (2018) Volume em reconstruction of spinal cord reveals wiring specificity in speed-related motor circuits

    Cell Reports 23:2942–2954.

    https://doi.org/10.1016/j.celrep.2018.05.023

    • PubMed
    • Google Scholar
65. 1. Tapia JC
    2. Kasthuri N
    3. Hayworth KJ
    4. Schalek R
    5. Lichtman JW
    6. Smith SJ
    7. Buchanan J

    (2012) High-Contrast en bloc staining of neuronal tissue for field emission scanning electron microscopy

    Nature Protocols 7:193–206.

    https://doi.org/10.1038/nprot.2011.439

    • PubMed
    • Google Scholar
66. Website

    1. The Jackson Laboratory

    (2021) jax-mice-and-services

    Accessed November 8, 2022.

    https://www.jax.org/de/-/media/jaxweb/files/jax-mice-and-services/b6j-data-summary.xlsx
67. Preprint

    1. Turner NL
    2. Macrina T
    3. Bae JA
    4. Yang R
    5. Wilson AM
    6. Schneider-Mizell C
    7. Lee K
    8. Lu R
    9. Wu J
    10. Bodor AL
    11. Bleckert AA
    12. Brittain D
    13. Froudarakis E
    14. Dorkenwald S
    15. Collman F
    16. Kemnitz N
    17. Ih D
    18. Silversmith WM
    19. Zung J
    20. Zlateski A
    21. Tartavull I
    22. Yu S
    23. Popovych S
    24. Mu S
    25. Wong W
    26. Jordan CS
    27. Castro M
    28. Buchanan J
    29. Bumbarger DJ
    30. Takeno M
    31. Torres R
    32. Mahalingam G
    33. Elabbady L
    34. Li Y
    35. Cobos E
    36. Zhou P
    37. Suckow S
    38. Becker L
    39. Paninski L
    40. Polleux F
    41. Reimer J
    42. Tolias AS
    43. Reid RC
    44. da Costa NM
    45. Seung HS

    (2020) Multiscale and Multimodal Reconstruction of Cortical Structure and Function

    bioRxiv.

    https://doi.org/10.1101/2020.10.14.338681

    • Google Scholar
68. Preprint

    1. Vergara HM
    2. Pape C
    3. Meechan KI
    4. Zinchenko V
    5. Genoud C
    6. Wanner AA
    7. Titze B
    8. Templin RM
    9. Bertucci PY
    10. Simakov O
    11. Machado P
    12. Savage EL
    13. Schwab Y
    14. Friedrich RW
    15. Kreshuk A
    16. Tischer C
    17. Arendt D

    (2020) Whole-Body Integration of Gene Expression and Single-Cell Morphology

    bioRxiv.

    https://doi.org/10.1101/2020.02.26.961037

    • Google Scholar
69. 1. Vishwanathan A
    2. Daie K
    3. Ramirez AD
    4. Lichtman JW
    5. Aksay ERF
    6. Seung HS

    (2017) Electron microscopic reconstruction of functionally identified cells in a neural integrator

    Current Biology 27:2137–2147.

    https://doi.org/10.1016/j.cub.2017.06.028

    • PubMed
    • Google Scholar
70. 1. Wanner AA
    2. Genoud C
    3. Masudi T
    4. Siksou L
    5. Friedrich RW

    (2016) Dense EM-based reconstruction of the interglomerular projectome in the zebrafish olfactory bulb

    Nature Neuroscience 19:816–825.

    https://doi.org/10.1038/nn.4290

    • PubMed
    • Google Scholar
71. 1. Wanner AA
    2. Friedrich RW

    (2020) Whitening of odor representations by the wiring diagram of the olfactory bulb

    Nature Neuroscience 23:433–442.

    https://doi.org/10.1038/s41593-019-0576-z

    • PubMed
    • Google Scholar
72. 1. Watson ML

    (1958) Staining of tissue sections for electron microscopy with heavy metals

    The Journal of Biophysical and Biochemical Cytology 4:727–730.

    https://doi.org/10.1083/jcb.4.6.727

    • Google Scholar
73. 1. White DL
    2. Andrews SB
    3. Faller JW
    4. Barrnett RJ

    (1976) The chemical nature of osmium tetroxide fixation and staining of membranes by X-ray photoelectron spectroscopy

    Biochimica et Biophysica Acta 436:577–592.

    https://doi.org/10.1016/0005-2736(76)90442-9

    • PubMed
    • Google Scholar
74. 1. Wigglesworth VB

    (1975) Lipid staining for the electron microscope: a new method

    Journal of Cell Science 19:425–437.

    https://doi.org/10.1242/jcs.19.3.425

    • PubMed
    • Google Scholar
75. 1. Willingham MC
    2. Rutherford AV

    (1984) The use of osmium-thiocarbohydrazide-osmium (OTO) and ferrocyanide-reduced osmium methods to enhance membrane contrast and preservation in cultured cells

    The Journal of Histochemistry and Cytochemistry 32:455–460.

    https://doi.org/10.1177/32.4.6323574

    • PubMed
    • Google Scholar
76. 1. Wilson AM
    2. Schalek R
    3. Suissa-Peleg A
    4. Jones TR
    5. Knowles-Barley S
    6. Pfister H
    7. Lichtman JW

    (2019) Developmental rewiring between cerebellar climbing fibers and Purkinje cells begins with positive feedback synapse addition

    Cell Reports 29:2849–2861.

    https://doi.org/10.1016/j.celrep.2019.10.081

    • PubMed
    • Google Scholar
77. 1. Xu CS
    2. Hayworth KJ
    3. Lu Z
    4. Grob P
    5. Hassan AM
    6. García-Cerdán JG
    7. Niyogi KK
    8. Nogales E
    9. Weinberg RJ
    10. Hess HF

    (2017) Enhanced FIB-SEM systems for large-volume 3D imaging

    eLife 6:e25916.

    https://doi.org/10.7554/eLife.25916

    • PubMed
    • Google Scholar
78. 1. Zheng Z
    2. Lauritzen JS
    3. Perlman E
    4. Robinson CG
    5. Nichols M
    6. Milkie D
    7. Torrens O
    8. Price J
    9. Fisher CB
    10. Sharifi N
    11. Calle-Schuler SA
    12. Kmecova L
    13. Ali IJ
    14. Karsh B
    15. Trautman ET
    16. Bogovic JA
    17. Hanslovsky P
    18. Jefferis GSXE
    19. Kazhdan M
    20. Khairy K
    21. Saalfeld S
    22. Fetter RD
    23. Bock DD

    (2018) A complete electron microscopy volume of the brain of adult Drosophila melanogaster

    Cell 174:730–743.

    https://doi.org/10.1016/j.cell.2018.06.019

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "In situ X-ray assisted electron microscopy staining for large biological samples" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Catherine Dulac as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

As you can see from the detailed reviews, all reviewers (and myself as well) felt that this work is potentially an important contribution.

However, as the central claim of the work is to use CT as a benchmark for the quality of staining for EM, it will be critical to directly relate those two measures, i.e. to perform EM imaging on previously CT imaged samples and relate the two measures of contrast to each other. This will be the critical experimental dataset for revision. There are other valuable comments below. Please edit the manuscript accordingly.

Reviewer #1 (Recommendations for the authors):

The authors are in a very good position to disambiguate current divergences in the initial steps of osmium staining of soft tissues for electron microscopy imaging purposes.

First, it would be very informative to assess the effect of a different iron salt commonly used to reduce osmium: potassium ferricyanide (e.g. by adding the experimental group 2% OsO4, 2.5% potassium ferricyanide to figure panels 1e-f).

Second, ref. Hua et al., 2015 implemented the osmium reduction posterior to the osmium incubation, by adding potassium ferrocyanide without intermediate washes. The authors have already explored this staining protocol in Supp Figure 5: samples being stained with OsO4 for 22h were transferred to a potassium ferrocyanide solution. It would be very informative to view those results in summary (ie averaging the 20h timepoint datasets across samples) along with the osmium and reduced osmium baselines shown in Figure 1e, and to complement the insight on sample expansion for that protocol, showing it against the same baselines, as in Figure 1f. An additional control group could be beneficial here: transferring the osmium-stained samples to a water vial, to account for decrease of staining by dilution and diffusion of unbound osmium. In this case it would not be expected to see sample degradation regardless of the incubation time.

Third, after staining with osmium or reduced osmium, samples are meant to be washed with water before proceeding to the next staining step, aiming to remove any unbound osmium from the media. The goal of optimizing the first staining step would therefore be to maximize the bound osmium remaining in the sample after the water wash(es). To that end, it would be interesting to see the evolution of intensity vs depth for samples placed in staining solution for 20h and then placed in water, for at least the most typical stains explored (osmium and reduced osmium).

Finally, it would be very useful if the authors released the code used for pre-processing single X-ray projections. Given the popularity of benchtop X-ray uCT devices, doing so could simplify and accelerate the adoption of this monitoring method of sample preparation protocols by the user community.

These additions could slightly improve the scope and completeness of an already robust and solid study.

Reviewer #2 (Recommendations for the authors):

Reference, page 4: Claude and Fullam (1945) used formaldehydes + osmium fixation. Only Sabatini introduced glutaraldehyde. It might be good to slightly modify the text.

Page 9: the authors refer to figure 1b to comment the staining after 20hours. Yet, fig 1b shows example at 10 hours max.

Page 9: "we found that for reduced osmium the accumulation of heavy metals peaks at a depth between 300- 800 μm (Figure 1e, Supplementary Figure 2), whereas the tissue above or below that depth is stained less." Could this be due to the anatomy/heterogeneity of the various cortical layers?

Page 9: about the shrinking/expansion of the tissue upon fixation with osmium or reduced osmium. This is very interesting as it could explain a few artifacts observed with several sample types. Yet in this experiment, the cortex biopsy is fitted inside a hole, of the same shape, made inside the sylgard support layer. It is thus possible that the tissue is constrained and can't expand in the lateral directions. Would this enhance the vertical expansion? And therefore such expansion would be less than 5%? Please comment on this.

In line with this, it might be good to adapt the drawings of figure 1a. The way the biopsy is displayed, it seems as if it was just placed onto the vial bottom, when in fact is is embedded in sylgard.

Page 12: about the osmium density: if I am not wrong these numbers are estimates, extrapolated from many assumptions re the density of unsaturated lipids, double bonds etc ... this should be made clear in the main text and not only in the methods. Even though the authors made their best possible efforts to describe the reactions within the tissues and to estimate the amount of binding sites, I doubt we have a solid proof of the actual concentration of bound or free osmium in this experiment.

Page 12: the authors write "While we assume that the tissue density is homogeneous across the cortical thickness ..." I don't think this is a fair assumption. The cortex is organized in layers of different cellular densities.

Page 14: discussion. what is not addressed in this paper is the secondary osmium blackening during dehydration. This has been described previously and attributed to unbound osmium and later reduction during the process. IT might be good to elaborate a bit on this. Of course, this leads to the comments made above, which highlight the need to check the samples in EM, i.e. to go, for some of them, through the full sample preparation.

Page 16: methods.

- The concentration ranges are surprising. Can you explain: "fixative solution 1.3 - 1.5% glutaraldehyde (GA) and 2.5 - 2.6% paraformaldehyde (PFA) in 0.14 - 0.15M cacodylate buffer with 2.0 - 2.1mM CaCl2"

- What is the rationale behind these incubation times. Intuitively, one would expect to keep whole brains longer in the fixative? "The samples were kept in fixative solution at 4 °C for 36 hours (4-mm biopsy punches) or for 12 hours (whole mouse brain)."

- About the sylgard molds: can you comment on the osmium infiltration. The sides of the biopsy might not be as well exposed as the top.

Page 17: X-ray microscopy: the working temperature is not indicate, though it is a very important parameter for the osmium fixation.

Reviewer #3 (Recommendations for the authors):

The manuscript would be strengthened if the authors compared the actual membrane contrast observed in cross-sections from their samples to demonstrate if these intensities can be quantitatively interpreted (beyond the description of staining gradients). I suggest to cut thin sections from a reduced osmium and osmium tissue cylinder and quantify the staining intensity/contrast across the 4mm diameter to compare with the x-ray profiles.

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "In situ X-ray assisted electron microscopy staining for large biological samples" for further consideration by eLife. Your revised article has been evaluated by Catherine Dulac (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed by small textual adjustments, as outlined below:

Please take note of the comments made by reviewer 3 regarding the strict correlation of X-ray and EM measure. Please adjust your textual description accordingly. While e.g. Mikula and Denk do provide some indication that this is the case (as does your work here), I would not want to give the impression that this is a question that has experimentally conclusively and quantitatively been settled. There is still room for direct experimental comparison for X-rays with EM under different conditions (although not necessarily in this manuscript).

In short, please tune down the corresponding comments and add a sentence or two in the discussion to mention the caveats and limitations discussed below and maybe point towards further such experiments. In this context it could be useful to expand a bit on the physics background and mention how X-ray contrast will depend e.g. on photon energy and that e.g. at the absorption edges the relationship of X-ray absorption and signal in the EM can become more complicated.

Reviewer #2 (Recommendations for the authors):

The revised version of the paper is efficiently addressing the points that I have raised and I am satisfied with the rebuttal answers as well.

In my opinion, the paper is ready for publication and expect the scientific community, especially the EM community to greatly benefit from this contribution.

Reviewer #3 (Recommendations for the authors):

My recommendation to the authors was to correlate their observed X-ray intensities to what is actually observed by EM in thin sections. In response, they cited Figure 4b from Mikula and Denk, 2015 as evidence that the X-ray reconstructed pixel intensities are correlated with what is observed in EM sections and state "Mikula and Denk (2015) showed that the pixel intensities of serial-section EM images and the intensity of the corresponding reslice are highly correlated."

I think this is an overstatement for several reasons:

1) As far as I can tell, Mikula and Denk Figure 4b is a n=1 measurement

2) The figure in question doesn't report a correlation coefficient, indeed their figure shows a peak SEM normalized intensity at the core of their sample and the X-ray normalized intensity has a peak at the edge of the sample

3) Mikula and Denk themselves are far more cautious in discussing this: "…X-ray microcomputed tomography, which exhibits an image contrast similar to that of SEM (Figure 4b). This allowed us to quickly test an embedded brain for defects and distortion before deciding whether to finish preparing it for electron microscopy." Mikula and Denk were clearing using X-ray reconstructions to check for gross defects/cracks, not to make quantitative comparisons.

Stroh et al., do provide a comparison of X-ray vs. EM intensities (new Figure 1 —figure supplement 1) with the caveat that the X-ray intensities are measured from a projection instead of a computed slice reconstruction. Perhaps due to this caveat the intensities do not appear to be very well correlated (although it would be preferable to actually show the correlation scatter plot and the measured correlation coeff). In light of this (and the above comments regarding the Mikula and Denk data), I remain unsure to what degree the measured X-ray intensities that form the foundation of the manuscript are informative about actual contrast in EM sections.

My original point was related to the author’' Figure 1e: the figure shows that reduced osmium stains tissue with a maximum intensity of ~1.1 (a.u.) compared to osmium alone at ~0.9 (a.u.). Because these intensities are not calibrated against the appearance of the staining in EM sections, their interpretation is limited. The authors state they were not able to test whether this particular finding is consistent with the intensity/contrast of EM sections due to embedding issues of reduced osmium samples.

Essential revisions:

As you can see from the detailed reviews, all reviewers (and myself as well) felt that this work is potentially an important contribution.

However, as the central claim of the work is to use CT as a benchmark for the quality of staining for EM, it will be critical to directly relate those two measures, i.e. to perform EM imaging on previously CT imaged samples and relate the two measures of contrast to each other. This will be the critical experimental dataset for revision. There are other valuable comments below. Please edit the manuscript accordingly.

We added Figure 1 —figure supplement 3 showing the ultrastructural preservation after 20 hours of incubation in OsO₄ and Figure 1 —figure supplement 1 that shows the covariation of the EM and X-ray pixel intensities. Note, in the presented study we acquired X-ray projection images that represent the cumulative tissue and heavy metal density along the direction of projection through a 4 mm thick tissue punch. Therefore our X-ray projection pixel intensities cannot directly be compared to the EM pixel intensity of a thin section. As has been shown previously (see for example Figure 4b in Mikula and Denk 2015), in computed X-ray tomograms the intensity scales linearly with EM intensity, if the pixel intensity in an EM section is compared to the corresponding reslice of a computed X-ray tomogram.

Reviewer #1 (Recommendations for the authors):

The authors are in a very good position to disambiguate current divergences in the initial steps of osmium staining of soft tissues for electron microscopy imaging purposes.

First, it would be very informative to assess the effect of a different iron salt commonly used to reduce osmium: potassium ferricyanide (e.g. by adding the experimental group 2% OsO4, 2.5% potassium ferricyanide to figure panels 1e-f).

We added Figure 1 —figure supplement 7 and 8 as well as Figure 2 —figure supplement 3 that show the spatio-temporal accumulation of osmium reduced with 2.5% potassium ferricyanide K₃[Fe(CN)₆]. The accumulation of heavy metals for the buffered solution with 2% OsO₄ and 2.5% potassium ferricyanide is not as homogeneous as for 2% OsO₄ (Figure 1 —figure supplement 8a). But in contrast to the experiments where osmium was reduced with 2.5% potassium ferrocyanide, the sample does not shrink if incubated in buffered solution of 2% osmium reduced with 2.5% potassium ferricyanide (Figure 1 —figure supplement 8b). In contrast to potassium ferrocyanide, no ‘washout’ effect was observed if the sample was placed in buffered solution of 2.5% potassium ferricyanide after 20hrs of incubation in buffered 2% OsO₄ (Figure 1 —figure supplement 8e).

Second, ref. Hua et al., 2015 implemented the osmium reduction posterior to the osmium incubation, by adding potassium ferrocyanide without intermediate washes. The authors have already explored this staining protocol in Supp Figure 5: samples being stained with OsO4 for 22h were transferred to a potassium ferrocyanide solution. It would be very informative to view those results in summary (ie averaging the 20h timepoint datasets across samples) along with the osmium and reduced osmium baselines shown in Figure 1e, and to complement the insight on sample expansion for that protocol, showing it against the same baselines, as in Figure 1f.

We added a corresponding Figure 1 —figure supplement 8.

An additional control group could be beneficial here: transferring the osmium-stained samples to a water vial, to account for decrease of staining by dilution and diffusion of unbound osmium. In this case it would not be expected to see sample degradation regardless of the incubation time.

Third, after staining with osmium or reduced osmium, samples are meant to be washed with water before proceeding to the next staining step, aiming to remove any unbound osmium from the media. The goal of optimizing the first staining step would therefore be to maximize the bound osmium remaining in the sample after the water wash(es). To that end, it would be interesting to see the evolution of intensity vs depth for samples placed in staining solution for 20h and then placed in water, for at least the most typical stains explored (osmium and reduced osmium).

We added Figure 2 —figure supplement 4 and 5 that show the dynamics of heavy metal washout in double distilled H₂O for 22 hours after 22 hrs of incubation in 2% buffered OsO₄. First, the sample was quickly flushed with double distilled H₂O four times to completely remove any buffered OsO₄ solution. Subsequently, the sample was immersed in double distilled H₂O for 22 hrs. The accumulation of heavy metals decreases by about 6% in a depth of 100-1200μm.

Our hypothesis is that this can be explained by unbound osmium that diffuses out of the sample and by the slight expansion of the sample (Figure 2 —figure supplement 5b). But in contrast to the washout effect in K₃[Fe(CN)₆], the reduction in heavy metal density appears to be small.

Finally, it would be very useful if the authors released the code used for pre-processing single X-ray projections. Given the popularity of benchtop X-ray uCT devices, doing so could simplify and accelerate the adoption of this monitoring method of sample preparation protocols by the user community.

The python and MATLAB code for processing the X-ray data is now available on https://github.com/adwanner/XrayAssistedStaining.

Reviewer #2 (Recommendations for the authors):

Reference, page 4: Claude and Fullam (1945) used formaldehydes + osmium fixation. Only Sabatini introduced glutaraldehyde. It might be good to slightly modify the text.

We corrected the text and citations accordingly.

Page 9: the authors refer to figure 1b to comment the staining after 20hours. Yet, fig 1b shows example at 10 hours max.

We corrected it accordingly.

Page 9: "we found that for reduced osmium the accumulation of heavy metals peaks at a depth between 300- 800 μm (Figure 1e, Supplementary Figure 2), whereas the tissue above or below that depth is stained less." Could this be due to the anatomy/heterogeneity of the various cortical layers?

Indeed, the effects of reduced osmium accumulation could depend at least in parts of the tissue composition, i.e. the anatomy/heterogeneity of the cortical layers in the tissue block. It is, however, important to note that this effect is not present in non-reduced osmium immersed brain tissue samples that have the same composition. We added the sentences that ‘No such band is present in tissue immersed in non-reduced OsO₄.’

Page 9: about the shrinking/expansion of the tissue upon fixation with osmium or reduced osmium. This is very interesting as it could explain a few artifacts observed with several sample types. Yet in this experiment, the cortex biopsy is fitted inside a hole, of the same shape, made inside the sylgard support layer. It is thus possible that the tissue is constrained and can't expand in the lateral directions. Would this enhance the vertical expansion? And therefore such expansion would be less than 5%? Please comment on this.

In line with this, it might be good to adapt the drawings of figure 1a. The way the biopsy is displayed, it seems as if it was just placed onto the vial bottom, when in fact is is embedded in sylgard.

The contribution to the vertical expansion from the embedding in sylgard is expected to be minimal, because the sylgard layer was typically only 2 mm thick, whereas the height of the tissue samples was typically about 6.75 mm tall. The upper end of the sylgard layer can be seen in Figure 1b at the very bottom of the image (small arrowhead). Hence, only a small fraction of the sample sits in sylgard. We adapted Figure 1 accordingly.

That the contribution from the restriction in sylgard is small can also be seen from the following considerations:

${\displaystyle 2mm\times \pi {\left(a\times \frac{1}{2}4mm\right)}^2-2mm\times \pi {\left(\frac{1}{2}4mm\right)}^2=2mm\times {(a-1)}^2\times \pi {\left(\frac{1}{2}4mm\right)}^2}$${\displaystyle =2mm\times {(1-1/a)}^2\times \pi {\left(a\times \frac{1}{2}4mm\right)}^2}$

For buffered 2% OsO₄ we measured an expansion of 5%, i.e. $b$=1.05 and according to 3. above the ‘true’ homogeneous expansion factor would have been ${\displaystyle a=1.049}$.

Page 12: about the osmium density: if I am not wrong these numbers are estimates, extrapolated from many assumptions re the density of unsaturated lipids, double bonds etc ... this should be made clear in the main text and not only in the methods. Even though the authors made their best possible efforts to describe the reactions within the tissues and to estimate the amount of binding sites, I doubt we have a solid proof of the actual concentration of bound or free osmium in this experiment.

The osmium density (as shown in Figure 2b) was measured using the pixel intensities which were calibrated using the concentration series shown in Figure 1c. No further assumptions are being made. In the discussion section highlighted by the reviewer we compare this measured osmium density to the density of double bonds of unsaturated lipids in the brain as estimated from numbers reported in the literature. The double bonds of unsaturated lipids are thought to be the main binding site for osmium. Comparing the measured density of osmium to the estimated density of binding sites we found that the measured osmium density is 2.999 times higher than the estimated density of double bonds in the brain. We therefore hypothesize that with long osmium incubation, 3 osmium atoms get associated with 1 double bond.

Page 12: the authors write "While we assume that the tissue density is homogeneous across the cortical thickness ..." I don't think this is a fair assumption. The cortex is organized in layers of different cellular densities.

We agree that the presented model and the underlying assumptions significantly simplify the actual bio-chemical environment. But despite these oversimplifications, the resulting model presented in Figure 2 captures the dynamics of osmium accumulation well. While modeling the tissue densities of different cortical layers in detail and calculating individual diffusion constants is beyond the presented study, we do agree that it would be very interesting to investigate how the diffusion constants differ in different tissue types. To illustrate this, we measured and modeled the staining kinetics of 2% buffered OsO₄ in 4mm punches of liver tissue (Figure 2 - Figure Supplement 7). Interestingly, the effective diffusion coefficient appears to be >4X larger in liver tissue compared to brain tissue.

Page 14: discussion. what is not addressed in this paper is the secondary osmium blackening during dehydration. This has been described previously and attributed to unbound osmium and later reduction during the process. IT might be good to elaborate a bit on this. Of course, this leads to the comments made above, which highlight the need to check the samples in EM, i.e. to go, for some of them, through the full sample preparation.

We added Figure 1 - Figure Supplement 3that shows the ultrastructural preservation after 20 hours of osmium incubation. In addition, we performed additional experiments in which we measured how the osmium density changes during the H₂O washing step after the OsO₄ incubation. These results of the wash-out experiments are shown in Figure 2 - Figure Supplement 4 and 5. We do find a slight decrease in heavy metal density that can likely be explained on one hand by the wash out of unbound osmium from the sample as well as by expansion of the tissue which reduces the local heavy metal density.

Page 16: methods.

- The concentration ranges are surprising. Can you explain: "fixative solution 1.3 - 1.5% glutaraldehyde (GA) and 2.5 - 2.6% paraformaldehyde (PFA) in 0.14 - 0.15M cacodylate buffer with 2.0 - 2.1mM CaCl2"

Some of the mice have been perfused by different experimenters that used slightly different effective fixative formulations.

- What is the rationale behind these incubation times. Intuitively, one would expect to keep whole brains longer in the fixative? "The samples were kept in fixative solution at 4 °C for 36 hours (4-mm biopsy punches) or for 12 hours (whole mouse brain)."

The 36 hours of post-perfusion incubation in fixative corresponds to the duration that we and other labs typically use for fixing brain tissue. The 12 hours post-perfusion incubation in fixative of the whole mouse brain might indeed be on the lower end for what might be necessary for thoroughly fixing a whole mouse brain after perfusion.

- About the sylgard molds: can you comment on the osmium infiltration. The sides of the biopsy might not be as well exposed as the top.

The sylgard layer was typically only 2 mm thick, whereas the height of the tissue samples was typically about 6.75 mm tall. The upper end of the sylgard layer can be seen in Figure 1b at the very bottom of the image (small arrowhead). Hence, only a small fraction of the sample sits in sylgard and the measurements have been taken several millimeters away from the sylgard, therefore we do not expect any major impact of the sylgard onto the osmium infiltration measurements. We adapted Figure 1 accordingly.

Page 17: X-ray microscopy: the working temperature is not indicate, though it is a very important parameter for the osmium fixation.

All staining and washing steps have indeed been done at room temperature. We added that information accordingly in the sample preparation section of the methods.

Reviewer #3 (Recommendations for the authors):

The manuscript would be strengthened if the authors compared the actual membrane contrast observed in cross-sections from their samples to demonstrate if these intensities can be quantitatively interpreted (beyond the description of staining gradients). I suggest to cut thin sections from a reduced osmium and osmium tissue cylinder and quantify the staining intensity/contrast across the 4mm diameter to compare with the x-ray profiles.

We added Figure 1 —figure supplement 3showing the ultrastructural preservation after 20 hours of incubation in OsO₄ and Figure 1 —figure supplement 1 that shows the covariation of the EM and X-ray pixel intensities. Note, in the presented study we acquired X-ray projection images that represent the cumulative tissue and heavy metal density along the direction of projection through a 4 mm thick tissue punch. Therefore our X-ray projection pixel intensities cannot directly be compared to the EM pixel intensity of a thin section. As has been shown previously (see for example Figure 4b in Mikula and Denk 2015), in computed X-ray tomograms the intensity scales linearly with EM intensity, if the pixel intensity in an EM section is compared to the corresponding reslice of a computed X-ray tomogram.

We also tried to acquire EM images of tissue punches stained with reduced osmium, but unfortunately the resin could not infiltrate the samples sufficiently well to allow us to collect serial sections.

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are some remaining issues that need to be addressed by small textual adjustments, as outlined below:

Please take note of the comments made by reviewer 3 regarding the strict correlation of X-ray and EM measure. Please adjust your textual description accordingly. While e.g. Mikula and Denk do provide some indication that this is the case (as does your work here), I would not want to give the impression that this is a question that has experimentally conclusively and quantitatively been settled. There is still room for direct experimental comparison for X-rays with EM under different conditions (although not necessarily in this manuscript).

In short, please tune down the corresponding comments and add a sentence or two in the discussion to mention the caveats and limitations discussed below and maybe point towards further such experiments. In this context it could be useful to expand a bit on the physics background and mention how X-ray contrast will depend e.g. on photon energy and that e.g. at the absorption edges the relationship of X-ray absorption and signal in the EM can become more complicated.

Reviewer #3 (Recommendations for the authors):

My recommendation to the authors was to correlate their observed X-ray intensities to what is actually observed by EM in thin sections. In response, they cited Figure 4b from Mikula and Denk, 2015 as evidence that the X-ray reconstructed pixel intensities are correlated with what is observed in EM sections and state "Mikula and Denk (2015) showed that the pixel intensities of serial-section EM images and the intensity of the corresponding reslice are highly correlated."

I think this is an overstatement for several reasons:

1) As far as I can tell, Mikula and Denk Figure 4b is a n=1 measurement

2) The figure in question doesn't report a correlation coefficient, indeed their figure shows a peak SEM normalized intensity at the core of their sample and the X-ray normalized intensity has a peak at the edge of the sample

3) Mikula and Denk themselves are far more cautious in discussing this: "…X-ray microcomputed tomography, which exhibits an image contrast similar to that of SEM (Figure 4b). This allowed us to quickly test an embedded brain for defects and distortion before deciding whether to finish preparing it for electron microscopy." Mikula and Denk were clearing using X-ray reconstructions to check for gross defects/cracks, not to make quantitative comparisons.

Stroh et al., do provide a comparison of X-ray vs. EM intensities (new Figure 1 —figure supplement 1) with the caveat that the X-ray intensities are measured from a projection instead of a computed slice reconstruction. Perhaps due to this caveat the intensities do not appear to be very well correlated (although it would be preferable to actually show the correlation scatter plot and the measured correlation coeff). In light of this (and the above comments regarding the Mikula and Denk data), I remain unsure to what degree the measured X-ray intensities that form the foundation of the manuscript are informative about actual contrast in EM sections.

My original point was related to the authors' Figure 1e: the figure shows that reduced osmium stains tissue with a maximum intensity of ~1.1 (a.u.) compared to osmium alone at ~0.9 (a.u.). Because these intensities are not calibrated against the appearance of the staining in EM sections, their interpretation is limited. The authors state they were not able to test whether this particular finding is consistent with the intensity/contrast of EM sections due to embedding issues of reduced osmium samples.

Thank you very much for the constructive feedback on the revised manuscript. We agree with Reviewer 3 that adding some quantitative evidence for the correlation between electron microscopy images and X-ray images would strengthen the revised manuscript. On one hand, we adjusted the wording regarding the similarity/correlation between EM and micro-CT in the introduction. On the other hand, we added a note of

caution to the discussion that for a qualitative comparison between EM and micro-CT appearance the two imaging modalities need to be calibrated and the precise image acquisition parameters need to be considered.

In addition, we added additional data comparing EM and micro-CT appearance similar to Mikula et al., to Figure 1 —figure supplement 1. In two different en bloc stained brain samples we show the correspondence between the EM and micro-CT appearance. As suggested by Reviewer 3 we included the (indeed high) correlation statistics and corresponding scatter plots.

To conclude, we have tuned down the claims regarding the correspondence between EM and X-ray imaging and at the same time added more examples and quantifications supporting the claims. We believe that these changes and additional information have improved the revised manuscript.

Author details

1. Sebastian Ströh

   Princeton Neuroscience Institute, Princeton University, Princeton, United States

   Contribution

   Conceptualization, Data curation, Validation, Investigation, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Eric W Hammerschmith

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8128-2617

2. Eric W Hammerschmith

   Princeton Neuroscience Institute, Princeton University, Princeton, United States

   Contribution

   Conceptualization, Data curation, Validation, Investigation, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Sebastian Ströh

   Competing interests

   No competing interests declared

3. David W Tank

   Princeton Neuroscience Institute, Princeton University, Princeton, United States

   Contribution

   Formal analysis, Supervision, Funding acquisition, Validation, Methodology, Writing – review and editing

   Competing interests

   is an inventor of US Patent Application 16/681,028

   "This ORCID iD identifies the author of this article:" 0000-0002-9423-4267

4. H Sebastian Seung

   Princeton Neuroscience Institute, Princeton University, Princeton, United States

   Contribution

   Formal analysis, Supervision, Funding acquisition, Validation, Methodology, Writing – review and editing

   Competing interests

   is an inventor of US Patent Application 16/681,028

5. Adrian Andreas Wanner

   1. Princeton Neuroscience Institute, Princeton University, Princeton, United States
   2. Paul Scherrer Institute, Villigen, Switzerland

   Contribution

   Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing

   For correspondence

   adrian.wanner@psi.ch

   Competing interests

   is an inventor of US Patent Application 16/681,028

   "This ORCID iD identifies the author of this article:" 0000-0002-5864-8577

• 725

  Page views

• 114

  Downloads

• 5

  Citations

Article citation count generated by polling the highest count across the following sources: PubMed Central, Crossref, Scopus.