Investigation of Spatial Interaction Between Astrocytes and Neurons in Cleared Brains

Summary

Combining viral vector transduction and brain clearing using the CLARITY method allows the investigation of a large number of neurons and astrocytes simultaneously.

Abstract

Combining viral vector transduction and tissue clearing using the CLARITY method makes it possible to simultaneously investigate several types of brain cells and their interactions. Viral vector transduction enables the marking of diverse cell types in different fluorescence colors within the same tissue. Cells can be identified genetically by activity or projection. Using a modified CLARITY protocol, the potential sample size of astrocytes and neurons has grown by 2-3 orders of magnitude. The use of CLARITY allows the imaging of complete astrocytes, which are too large to fit in their entirety in slices, and the examination of the somata with all their processes. In addition, it provides the opportunity to investigate the spatial interaction between astrocytes and different neuronal cell types, namely, the number of pyramidal neurons in each astrocytic domain or the proximity between astrocytes and specific inhibitory neuron populations. This paper describes, in detail, how these methods are to be applied.

Introduction

In recent years, the knowledge of astrocyte function and how they interact with neuronal circuits has increased dramatically. Astrocytes can influence plasticity^1,2, assist in neuronal postinjury recovery^3,4, and even induce de novo neuronal potentiation, with recent studies exhibiting the importance of astrocytes in memory acquisition and reward, previously regarded as purely neuronal functions^5,6,7. A feature of particular interest in astrocyte research is the spatial arrangement of the cells, which maintain unique spatial organizations in the hippocampus and other brain structures^8,9,10. Unlike the neuronal dendrites that intertwine between cell somata, hippocampal astrocytes inhabit visually distinguishable territories with slight overlap between their processes, creating distinct domains^8,11,12,13. The evidence supporting the participation of astrocytes in neuronal circuits does not support the lack of detailed anatomical description of such populations and the neurons in their domains¹⁴.

Viral vector transduction procedures, along with transgenic animals (TG), have been popularized as a toolset to investigate brain structures, functions, and cell interactions^15,16. The utilization of different promoters allows the targeting of specific cells according to their genetic properties, activation levels^17,18, or projection targets. Different viruses can express different colored fluorophores in different populations. A virus can be combined with the endogenous expression of fluorophores in TG, or TG animals can be used without the need for viruses. These techniques are widely used for neuronal marking, and some labs have started using them with modifications specialized for targeting other cell types, such as astrocytes^5,9,19.

The CLARITY technique, first described in 2013^20,21, enables the study of thick brain slices by making the entire brain transparent while leaving the microscopic structures intact. By combining the two methods-viral vector transduction and tissue clearing-the option of examining the spatial interactions between different cell type populations is now available. Most astrocyte-neuron interaction studies were performed on thin brain slices, resulting in images of incomplete astrocytes due to their large domains, thus radically restricting the number of analyzed cells. The use of the CLARITY technique allows single-cell resolution characterization of cell populations in large-scale volumes simultaneously. Imaging fluorescently tagged cell populations in clear brains does not deliver synaptic resolution but permits thorough characterization of the spatial interactions between astrocytes and a variety of neuronal cell types.

For that reason, we harnessed these state-of-the-art techniques to investigate the properties of astrocytes throughout the dorsal CA1, imaging all lamina (Stratum Radiatum, pyramidal layer, and Stratum Oriens). We measured tens of thousands of astrocytes (with viral penetrance of >96%⁵), thereby analyzing the information of the entire astrocytic population around CA1. With efficient penetrance of the neuronal markers, we could record the interactions between the entire population of CA1 astrocytes and the four types of neuronal cells-parvalbumin (PV), somatostatin (SST), VIP inhibitory neurons, and excitatory pyramidal cells⁹.

Several experiments were performed using a combination of fluorescence from TG animals and differently colored viral vectors (all inhibitory cells), while others (excitatory) utilized two viral vectors expressing different fluorophores under different promoters⁹. This paper presents a detailed protocol, including the tagging of the desired cells in the brain, making the brain transparent using a modified CLARITY procedure, as well as imaging and analyzing complete brain structures, using various procedures and software.

Protocol

Experimental protocols were approved by the Hebrew University Animal Care and Use Committee and met the guidelines of the National Institute of Health Guide for the Care and Use of Laboratory Animals.

1. Viral vector transduction

NOTE: Viral vector transduction is used to express fluorophores in the brain.

1. Use an atlas (e.g., Allen Brain Atlas) to locate the relevant coordination of the target area.
   NOTE: 3D atlases can be found online (e.g., http://connectivity.brain-map.org/3d-viewer).
2. Using stereotactic surgery, inject the viral vectors into the relevant brain structure. See⁹ for a detailed protocol.
3. Wait for 3-6 weeks for fluorophore expression.
   NOTE: A short period of time is enough if only cell bodies need to be marked. A long period of time will be needed if the axonal projections are relevant to the question, as it takes longer for the fluorophores to express in axons, which can be several millimeters long.
4. Validate the specificity and penetrance of fluorophore expression (both of viral expression and TG animals) beforehand (Figure 1A). Before proceeding with the CLARITY procedure, designate at least one brain for thin slices and make sure the fluorophore expression is both strong and specific to the target cell feature.

2. CLARITY

NOTE: This method renders the brain transparent within 2-6 weeks.

1. Perform transcranial perfusion on animals using cold Phosphate-buffered Saline (PBS) followed by 4% Paraformaldehyde (PFA) in PBS. Remove the brain and keep it in PFA overnight at 4 °C in a 50 mL tube or a similar container.
   NOTE: Before transcranial perfusion is performed, the animal must be deeply anesthetized. In the examples presented in this protocol, all animals were anesthetized using Ketamine and Xylazine (90% and 10%, respectively).
2. Replace PFA with Hydrogel Solution (HS; see Table 1) for 48 h at 4 °C.
   NOTE: Do not allow the materials to become warmer than the refrigeration temperature (4-8 °C) at this stage, or the Hydrogel will polymerize. The HS used in this protocol contains 2% acrylamide, unlike previous protocols suggesting 4%²² or 1%¹⁷. The benefit of 2% is detailed in the discussion section. When preparing the HS, work on an ice-cold surface. Store it at -20 °C.
3. Degassing
   NOTE: The purpose of this stage is to remove all free oxygen from the tissue as O₂ interferes with the polymerization process. Any nonO₂ gas can be used (e.g., N₂, CO₂); N₂ is recommended.
   1. Transfer the N₂ from the tank via a 5 mm (internal) flexible tubing, connected to a 19 G needle at its end (Figure 1B).
   2. Make two small holes (needle-wide) in the cap of the tube: one to introduce the gas and the other to allow the air to leave the tube (Figure 1C).
   3. Attach the pipe from the N₂ gas to the tube and replace the gases for approximately 30 min at room temperature (RT).
   4. Remove the pipe and immediately seal the holes with modeling clay on every tube (Figure 1D).
4. Transfer the degassed sealed tubes to a 37 °C bath for 3.5 h to polymerize the HS, which will become a gel. Be careful not to shake the tubes (Figure 1E).
5. Extract the brain from the tube and gently remove the polymerized gel from around it using laboratory wipes. Make sure no residual gel remains attached to the surface of the tissue, as it might react with the solution in the following steps, inhibiting the tissue clearing process (Figure 1F).
   NOTE: Because the gel contained PFA, the extraction should be done under a fume hood.
6. Slice the brain if the question at hand requires only a part of it. Divide it in half or into very thick slices that contain all areas of interest (Figure 1G).
7. Place the slices in the First Clearing Solution (CS1, see Table 2) in a new container and shake at 70 rpm for 24 h at 37 °C.
8. Follow the steps described below to transfer the brain from CS1 to the second Clearing Solution (CS2, see Table 2).
   1. Prepare a perforated tube for the brain in advance (Figure 1H).
   2. Preheat CS2 to 40-45 °C in a beaker large enough to contain the tube. Do this on a hotplate with a stirrer. Ensure the temperature does not reach 55 °C to prevent the bleaching of the expressed proteins.
   3. Place the beaker filled with CS2 and the perforated tube on a stirring device (a 2 L beaker in Figure 1I). Set a moderate stirring rate that will cause the liquid to flow without distorting the tissue.
      NOTE: Within 2-6 weeks, the tissue will become transparent (the process starts at the periphery and moves inward toward the central brain structures). The decision as to when the tissue is "clear enough" is up to personal judgment. The brain should be sufficiently transparent for imaging under the microscope without interference (Figure 1J not clear enough; Figure 1K clear enough). Surpassing the point at which the tissue is clear enough may cause loss of tissue rigidity.
9. Transfer from CS2 into PBST (0.5% Triton X-100 in PBS, see Table 2) in a new container at 37 °C with mild shaking (70 rpm) for 24 h.
   NOTE: In the PBST, the tissue will become whitish (Figure 1L). The clarity will return (and improve) when embedded in the Refractive Index Matching Solution (RIMS, see step 2.14).
10. Replace the PBST with new PBST. Keep under the same conditions (37 °C with mild shaking) for another 24 h.
11. Replace PBST again with new PBST and keep at RT for 24 h.
12. Transfer to PBS at RT for 24 h.
13. Replace the PBS and leave at RT for an additional 24 h.
14. Remove the brain from PBS and transfer it to RIMS at 37 °C overnight.
    NOTE: Initially, ripples in the RIMS may surround the brain. This protocol was designed and validated using two specific commercial RIMS (see Table 3). However, the protocol should not differ when using other RIMS (commercial or self-made).
15. Keep the tissue at RT for another 24 h or until the solution reaches full equilibrium, i.e., when the tissue becomes transparent, and the solution no longer contains any visible ripples (Figure 1M).
16. If the tissue is transparent, proceed to the chamber preparation. If at this point, the tissue becomes white instead of transparent (due to aggregates of residual SDS molecules), clean the sample again by repeating steps 2.9 and 2.10, and transfer the tissue to CS2 (step 2.8) for a few days, followed by all the steps until step 2.15.

3. Chamber preparation

NOTE: Each sample requires a slide with an imaging chamber in which the sample will be placed.

1. Place the sample in the middle of the slide.
2. Using a hot glue gun, create walls at the edges of the slide, almost as high as the tissue. Make sure to leave a small gap (approximately 5 mm) at one of the corners (Figure 2A,B).
3. Apply 1-2 drops of the RIMS on the sample to keep the upper surface moist and prevent bubbles forming between the coverslip and the tissue.
4. Immediately after applying the RIMS drops, add the last layer of hot glue to the walls (so that they reach the height of the brain/slice) and progress immediately (while the hot glue is still liquid) to step 3.5.
5. Seal the top with a coverslip, placing it as evenly as possible on the top of the still-warm hot glue layer (Figure 2C).
6. Fill the chamber with the RIMS through the gap left in the hot glue walls (Figure 2D,E).
7. Close the gap with hot glue. Leave no air inside (Figure 2F).
8. If the hot glue walls extend beyond the borders of the slide, cut the extending edges (Figure 2G).
9. If the objective used for imaging is immersed, add another 2-3 mm of glue to the walls above the coverslip so that the immersion solution will last for a longer period (Figure 2H).

4. Imaging (confocal or two-photon)

1. Check whether the stage can hold the chamber, as the thickness of the chamber can reach several millimeters.
   NOTE: For example, the confocal microscope (see the Table of Materials) used in this protocol is equipped with several stages (e.g., circular, rectangular). Whichever stage is chosen to hold the chamber is irrelevant as long as its parameters fit the chamber sizes.
2. Work with an objective with a sufficient working distance, i.e., a minimal working distance ≥3 mm, as the brain region of interest may be a few millimeters in depth, and the coverslip may not be completely straight as it was placed manually.
   NOTE: As a demonstration, the results presented in Figure 1, Video 1, and Video 2 were obtained under a two-photon microscope using a water immersion 16x objective with a 3 mm working distance and magnification of 2.4 to obtain a 652 x 483 µm field of view and a z-stack with 0.937 µm intervals between the planes.
3. Perform multiple-area imaging, which may take a few hours or even days. Make sure there is ample free storage in the computer as the size of each image can reach up to tens of gigabits.
   NOTE: CLARITY imaging may result in significantly more image sections in depth. Therefore, the intensities used to capture the expression should be relatively low to prevent overexpression during the image summation.
4. For immersion objectives, check the liquid routinely while imaging and supplement with additional liquid if needed.
   NOTE: During the imaging of the data presented in Video 3, water was added to the surface of the chamber every 6-8 h to combat evaporation.
5. After the image has been acquired, view and analyze it using several different software packages.
   NOTE: Recommended software includes Imaris and SyGlass for both visualization and analysis. The precise pipeline for optimized reconstruction of astrocytes using Imaris software has been described previously⁹. In short, the Imaris features used in this study were Filaments to fully reconstruct cell structures and Spots to extract the position in the space of all the somata.

Results

Successful clearing of thick brain tissue slices results in a new range of questions that can be asked regarding the properties of large cell populations as opposed to the properties of single cells or neighboring groups of cells. To achieve successful results, one should strictly adhere to the CLARITY protocol, as there is a wide range of parameters that need to be considered to reduce the variance between samples (e.g., percentage of clarity, fluorescence information, swollenness parameters).

Discussion

Tissue clearing methods present a revolutionary tool in brain research, inviting questions that could not previously have been asked. From targeting the properties of a small group of cells, a single cell, or even a single synapse, CLARITY now enables the targeting of total cell populations or long-range connectivity features by using relevant fluorophores.

The outcome of the fluorophore expression and CLARITY procedure combination is not binary; many factors may interfere with the procedure l...

Materials

Name	Company	Catalog Number	Comments
AAV1-GFAP::TdTomato	ELSC Vector Core Facility (EVCF)		viral vector used to detect astrocytes
AAV5-CaMKII::eGFP	ELSC Vector Core Facility (EVCF)		viral vector used to detect neurons
AAV5-CaMKII::H2B-eGFP	ELSC Vector Core Facility (EVCF)		viral vector used to detect neuronal nuclei
AAV5-CaMKII::TdTomato	ELSC Vector Core Facility (EVCF)		viral vector used to detect neurons
Acrylamide (40%)	Bio-rad	#161-0140
Bisacrylamide (2%)	Bio-rad	#161-0142
Boric acid	Sigma	#B7901	Molecular weight - 61.83 g/mol
Confocal microscope, scanning, FV1000	Olympus		4x objective (UPlanSApo, 0.16 NA)
Imaris software	Bitplane, UK		A software that allows 3D analysis of images
NaOH	Sigma	#S5881
PBS
PFA 4%	EMS	#15710
RapiClear	SunJin lab	#RC147002
RapiClear CS	SunJin lab	#RCCS002
SDS	Sigma	#L3771
SyGlass software			A software that allows 3D analysis of images using virtual reality
Tris base 1 M	Bio-rad	#002009239100	Molecular weight - 121.14 g/mol
Triton X-100	ChemCruz	#sc-29112A
Two photon microscope	Neurolabware		Ti:sapphire laser (Chameleon Discovery TPC, Coherent), GaAsP photo-multiplier tubes (Hamamatsu, H10770-40) , bandpass filter (Semrock), water immersion 16x objective (Nikon, 0.8 NA)
VA-044 Initiator	Wako	#011-19365