Generation of single-cell and single-nuclei suspensions from embryonic and adult mouse brains

Summary Efficient protocols to generate single-cell and single-nuclei suspensions are critical for the burgeoning field of single-cell/single-nuclei sequencing. Here we describe procedures to generate single-cell and single-nuclei suspensions from embryonic and adult mouse brains. This protocol can be modified for any brain region and/or neural cell type. For complete details on the use and execution of this protocol, please refer to Lee et al. (2022),1 Rhodes et al. (2022),2 Mahadevan et al. (2021),3 Ekins et al. (2020),4 and Wester et al. (2019).5

tissue or cell-types before experimentation because distinct brain regions contain different proportions of cells and myelin, which may affect the amount of starting material required and the efficiency of the cleanup procedure.
For anyone using these protocols for single-cell sequencing using the 10x Genomics platform, we recommend exploring the 10x Genomics Support Documentation resources as they regularly provide updated protocols for tissue processing with their single-cell sequencing platforms. Many of our nuclei preparation solutions are in part based on these 10x Genomics protocols. One notable difference is that we utilize mechanical dissociation via Dounce homogenizers to lyse brain tissue whereas most current 10x Genomics protocols digest tissue with detergents on ice followed by trituration. In our experience, we obtain better nuclei recovery and higher quality cDNA libraries and sequencing results using the mechanical dissociation in lysis solution to generate nuclei suspensions from both embryonic and adult brain samples.

Institutional permissions
All experimental procedures were conducted in accordance with the National Institutes of Health guidelines and were approved by the NICHD Animal Care and Use Committee. Both male and female embryonic and adult mice were used without bias for all experiments. Housing conditions: 12/12 h light/dark cycle, humidity between 30%-50%, temperature 20 C-22 C. All researchers will need to acquire permission from their institution to perform these experiments on mice.

Preparation of equipment and solutions
Timing: 30-60 min 1. All dissection tools and Dounce homogenizers should be autoclaved or sterilized using acceptable practices, and proper PPE should be worn throughout the protocol. 2. Prepare isoflurane chamber setup and/or load 0.05 mL Euthasol per mouse into syringe(s) for anesthesia. 3. If generating single-cell suspensions, use a Bunsen burner to fire polish the tips of Pasteur pipettes, creating some with large bore openings ($600 mm, opening should be clearly visible) and some with small bore openings ($300 mm, opening should be barely visible). Test the flow speed by pipetting water up and down through fire polished Pasteur pipette. 4. Prepare the following solutions fresh on the day of experiment (depending on the desired procedure) and keep all solutions on ice throughout the procedure: a. Embryonic brain dissections: Prepare the artificial cerebral spinal fluid (ACSF) solution and carbogenate (bubble) with 5%:95% CO 2 :O 2 . If generating single-cell suspensions, be prepared to make the Pronase Solution and Cell Reconstitution Solution after brain dissection. b. Adult brain dissections: Prepare NMDG High-Mg 2+ Solution and NMDG Activity-blocking Solution, place on ice and keep carbogenated with 5%:95% CO 2  Alternatives: Many reagents in the 'Other' and 'Chemicals, peptides and recombinant proteins' categories can be found from other vendors than then ones listed above.

MATERIALS AND EQUIPMENT
Solutions for embryonic brain dissections CRITICAL: Add CaCl 2 and MgCl 2 after bubbling with carbogen to avoid precipitation of these reagents.
Note: pH must be adjusted to 7.4 and solution should always be kept on ice. Final volume of 500 mL can be adjusted depending on volume needed.
Solutions for embryonic brain dissections -Single-cell suspensions Note: 10x Genomics has 2 relevant protocols for their Multiome kit, one general for all tissue (CG000365, Rev C) and one specific for embryonic mouse brains (CG000366, Rev D). The only difference between these two protocols is the Lysis buffer, with the general protocol using 13 lysis buffer and the embryonic mouse brain protocol using 0. 13

STEP-BY-STEP METHOD DETAILS
We describe multiple strategies to generate single-cell or single-nuclei suspensions from either embryonic or juvenile/adult mouse brain tissue. If harvesting tissue from embryonic mouse brains: steps #1-21 describe the procedure to harvest specific regions of the embryonic brain, steps #32-37 describe generation of single-cell suspensions, and steps #46-57 describe generation of singlenuclei suspensions. If harvesting tissue from juvenile/adult mouse brain: steps #22-31 describe the procedure to harvest specific of the adult brain via either brain slices (#22-26) or whole-mount (#27-31), steps #38-45 describe generation of single-cell suspensions, and steps #46-61 describe generation of single-nuclei suspensions. If proceeding with single-cell sequencing reactions using the 10x Genomics platform: steps #62-63 describe how to prepare cell or nuclei suspensions for RNA-seq, ATAC-seq or Multiome analysis.

Harvesting embryonic mouse brains
Timing: 15 min Embryos are removed from the dam and brains are collected in ACSF.
1. Anesthetize timed pregnant dams with either isoflurane (4% for induction) or Euthasol (270 mg/ kg, 50 mL/30 g mouse, intraperitoneal (IP) injection). 2. After confirming no response to painful toe pinch stimuli, perform cervical dislocation and place dam on petri dish. 3. Use sterilized scissors to cut an incision along the midline of lower abdominal to expose the uterus and embryos. 4. Cut through connective tissue under the embryonic chain to extract the uterus with embryos and transfer to carbogenated ACSF in Petri dish on ice. 5. Using Dumont #2 forceps, remove each embryo from the uterus and amniotic sac and transfer embryos to a new Petri dish with carbogenated ACSF on ice. 6. Decapitate each embryo and transfer heads to a new Petri dish with carbogenated ACSF on ice. 7. Select 1 head and under a dissecting microscope, use Dumont #5 forceps and/or curved micro forceps to remove the skin and skull and expose the brain using 1 of 2 strategies: a. With the head dorsal side up, stabilize the head of embryo by holding the nasal and Maxillary region with forceps. i. Using fine forceps in other hand, peel away skin to expose the skull.
ii. Pierce the skull with forceps and peel away skull to expose the brain ( Figure 1A). b. With head lying on its side, use forceps to grasp skin between brain and base of skull to minimize damaging the brain. Peel away skin and skull to expose brain, rotating head as needed to remove skin and skull ( Figure 1B). 8. When skin and skull are removed from the top and sides of the brain ( Figure 1C), slide forceps underneath the brain (along the base of the skull) and 'pinch' off brain to remove it from head, then transfer brain to a new Petri dish with carbogenated ACSF on ice.
CRITICAL: Repeat steps 6-8 for all embryos before moving on to the next steps to ensure that brains are incubated in ACSF as fast as possible.

Microdissection of embryonic forebrain regions
Timing: 1-2 h depending on litter size, number of regions being dissected, etc.
Specific forebrain regions are dissected out from the embryonic mouse brain. Here we describe how to harvest the MGE, LGE, CGE and cortex from an individual brain. Researchers can harvest the ll OPEN ACCESS STAR Protocols 4, 101944, March 17, 2023 specific brain regions based on their experimental plans. All dissections performed in ice-cold carbogenated ACSF.
9. Transfer 1 brain to a new Petri dish with carbogenated ACSF on ice. 10. To harvest specific forebrain regions, remove any midbrain and hindbrain by pinching off any tissue caudal to the cerebral hemispheres ( Figure 1D).
Note: If uncertain about embryonic neuroanatomy, please consult a reference atlas 15 or online resources such as the Allen Brain Developing Mouse Brain Atlas.
11. Using forceps, pinch between the hemispheres to hemisect the forebrain.
12. Position one hemisphere medial side up and remove any diencephalon tissue medial to the lateral ventricle. The ganglionic eminences (GEs) should be visible through the lateral ventricle ( Figure 1E). 13. Gently pinch the dorsal cortex with forceps at 2-3 locations and peel it backwards to further expose the GEs ( Figure 1F). 14. To remove the MGE, place one prong of the forceps into the posterior portion of the MGE-LGE sulcus and the other prong ventro-caudal to the MGE. Pinch off the MGE to separate it from the CGE ( Figure 1G). 15. Continue separating the MGE from the LGE and septum by gently pinching around the MGE.
When fully separated ( Figure 1H), collect the MGE with either forceps or a pipette and transfer it to a 'MGE'-labeled 5 mL round bottom tube with carbogenated ACSF on ice.

OPEN ACCESS
Note: After removing MGE, the remaining LGE-CGE structure is symmetric and one can lose track of the orientation. Keep in mind which side of the brain is anterior (LGE) and posterior (CGE). 16. To cleanly separate LGE from CGE, make two incisions through the LGE-CGE structure to roughly split the structure into thirds ( Figure 1I), with the anterior third being the LGE, the posterior third being the CGE, and the middle third being the LGE-CGE boundary region. 17. Using curved micro forceps, place the prongs dorsal and ventral to the LGE, then pinch off the LGE and transfer to a 'LGE'-labeled 5 mL round bottom tube with carbogenated ACSF on ice ( Figure 1J). 18. Following the same process, remove the CGE and transfer to a 'CGE-labeled 5-mL round bottom tube with carbogenated ACSF on ice.
CRITICAL: Refrain from placing the forceps too deep when harvesting the LGE and CGE, as this will increase the likelihood of removing lateral cortex along with the LGE and/or CGE and contaminating the sample with glutamatergic cortical cells.
19. Remove a chunk of the cortex (e.g., somatosensory) and transfer to a 'Cortex'-labeled 5 mL round bottom tube with carbogenated ACSF on ice ( Figure 1J). 20. Repeat steps 12-19 for the other hemisphere, noting that the anterior and posterior positions are flipped. 21. Repeat steps 9-20 for all embryonic mouse brains.
Optional: If harvesting nuclei, collected tissue can be transferred directly to a 1.5 mL conical tube, flash frozen on dry ice, and stored at À80 C. This allows for collection and storage of tissue from multiple animals for future processing. We do not recommend freezing tissue when collecting cells as cell membranes will fracture without using cryopreservation media.
Note: If generating single-cell suspensions, prepare the Pronase Solution and Cell Reconstitution Solution at this point, after dissecting all embryonic brains.

Adult mouse brain preparation -Brain sections
Timing: $20-30 min/brain for brain matrix, $30-45 min/brain for vibratome Harvesting specific regions from the adult mouse brain by generating brain slices.
22. P21 and older mice are deeply anesthetized with Euthasol (270 mg/kg, 50 mL/30 g mouse, intraperitoneal (IP) injection). After confirming no response to painful toe pinch stimuli, decapitate mouse. 23. Peel skin forward and use forceps to pinch and remove skull starting from posterior end. When brain is fully exposed, remove and place into Petri dish with carbogenated NMDG High Mg 2+ Solution on ice. 24. To generate brain sections using brain matrix: a. Place brain in a pre-chilled stainless steel 0.5 mm mouse brain matrix on ice (or similar matrix based on desired orientation and tissue thickness). b. Firmly depress razor blades through all slots encompassing desired brain regions (Figure 2A). c. Carefully remove razorblades, noting that tissue often sticks to one of the blades. Transfer desired slices to a Petri dish with carbogenated NMDG Activity-blocking Solution on ice (Figure 2B). 25. To generate brain sections using vibratome: a. If desired, remove excess tissue from anterior or posterior brain to minimize cutting through unneeded tissue. b. Glue brain(s) to vibratome plate and allow glue to dry for 30-60 s. Then transfer plate into the vibratome chamber surrounded by ice and filled with NMDG High Mg 2+ Solution that can be continuously carbogenated ( Figure 2C). c. Cut 300-400 mm sections and transfer desired sections to a Petri dish with carbogenated NMDG Activity-blocking Solution ( Figure 2D). Adult mouse brain dissection (A and B) Adult mouse brain in matrix with razor blades for coronal brain slices. (C and D) Vibratome setup to generate coronal brain slices. (E and F) Whole mount preparation to remove intact brain regions (e.g., hippocampus (Hippo), striatum (Str) and cortex (Ctx)).

OPEN ACCESS
26. Use forceps to remove desired regions (striatum, cortex, hippocampus, etc.) from individual brain slices and transfer to properly labeled 5 mL round bottom tube or Petri dish.
Optional: If using a transgenic fluorescent reporter mouse line, dissecting desired regions can be done under a fluorescent dissecting scope to harvest fluorescent brain region/cells of interest.
Optional: If harvesting nuclei, tissue can be directly added to a 1.5 mL conical tube, flash frozen on dry ice, and stored at À80 C. This allows for collection and storage of tissue from multiple animals for future processing. We do not recommend freezing tissue when collecting cells as cell membranes will fracture without using cryopreservation media.

Adult mouse brain preparation -Whole mount
Timing: $10 min/brain, depending on number of brain regions being collected Harvesting specific regions from the adult mouse brain by removing whole intact structures.
27. Follow steps 22 and 23 above to remove brains from juvenile/adult mice. 28. Place brain ventral side down in a Sylgard-coated petri dish containing NMDG High Mg 2+ Solution and insert pins through cerebellum and anterior forebrain. 29. Use curved forceps to peel cortex antero-lateral and lay flat onto petri dish to expose underlying hippocampus and ventral structures ( Figure 2E). 30. If collecting cortex and/or hippocampus, remove these structure and place into petri dish containing NMDG Activity-blocking Solution on ice ( Figure 2F). 31. For striatum and other ventral brain structures, scrape away unwanted tissue, remove desired structure and transfer to petri dish containing NMDG Activity-blocking Solution on ice (Figure 2F).
Optional: If using a transgenic fluorescent reporter mouse line, this dissection can be done under a fluorescent dissecting scope to harvest fluorescent brain region/cells of interest.
Optional: If harvesting nuclei, tissue can be directly added to a 1.5 mL conical tube, flash frozen on dry ice, and stored at À80 C. This allows for collection and storage of tissue from multiple animals for future processing. We do not recommend freezing tissue when collecting cells as cell membranes will fracture without using cryopreservation media.
Generation of single-cell suspensions from embryonic brain tissue  Generation of single-cell suspensions from adult brain tissue

Timing: $90 min
Generation of single-cell suspensions from adult mouse brain regions.
38. Cell dissociations are performed with Worthington's Papain Dissociation System following manufacturer's instructions with minor modifications described below.
CRITICAL: Maximum size of tissue for 1 reaction is 100 mg. We observed significant decrease in efficiency with tissue weighing more than 100 mg. If tissue weighs more CRITICAL: If analysis includes chromatin accessibility (e.g., ATAC-seq or Multiome), total time from Douncing samples (step 46) to final nuclei dilution (step 63) should take no more than 60 min. Sample preparation intervals greater that 60 min produce a noticeable decrease in library quality.
46. Add 1 mL of Nuclei Lysis Buffer (RNA-only, DNA-only or Multiome depending on application) to a pre-chilled Dounce homogenizer on ice, 1 Dounce per tissue region/sample. a. If using fresh tissue harvested as described above, transfer tissue pieces to Dounce. b. If using frozen tissue stored at À80 C, thaw for 30-60 s at 20 C-22 C and transfer tissue to Dounce.
CRITICAL: Maximum size of tissue for 1 reaction is 100 mg. We observed significant decrease in efficiency with tissue weighing more than 100 mg. If tissue weighs more than 100 mg, split tissue into multiple Dounces and run as separate reactions, combining sample replicates at the end.
47. Slowly dounce $10 times with A pestle and another $10 times with B pestle (1-2 s per up-or down-stroke), trying to minimize introduction of bubbles into the suspension ( Figure 3B). Note: Depending on tissue quantify and pellet size, it might be preferrable to use less than 250 mL Nuclei Resuspension or Nuclei Wash buffer per tube to resuspend the pellet after the last spin.
57. To remove debris for downstream applications, we strongly recommend sorting the nuclei suspensions. a. Add 1 mL DAPI (or DRAQ5) to nuclei suspension and filter suspension through a pre-wetted 35 mm filter. Proceed to cell sorter, collecting DAPI+ nuclei (or fluorescent nuclei if using a reporter mouse line).
Note: If performing 10x Genomics ATAC-seq or Multiome reactions, we recommend collecting > 50,000 nuclei if possible, as this will provide sufficient nuclei for these reactions after loss of nuclei from spinning and reconstitution.
Optional: Some samples from adult tissue with high myelin content (e.g., adult spinal cord) contain a significant amount of debris and harvesting nuclei via FACS was extremely inefficient. In these instances, we recommend performing the following optional additional cleanup procedure using the Nuclei Pure Prep Isolation Kit prior to FACS to remove excess myelin.
CRITICAL: Sucrose Gradient clean-up has only been tested for analyzing mRNA alone. It has NOT been validated for analyzing DNA alone or Multiome experiments and is not expected to work well in such cases. If debris or myelin removal is needed for DNA only or Multiome (mRNA + DNA) samples, we strongly recommend purifying the nuclei suspensions via flow cytometry.

EXPECTED OUTCOMES
Researchers should observe very clean single-cell or single-nuclei suspensions, with minimal debris and dead cells. Cleaning up both cell and nuclei suspensions via FACS will remove significant amounts of debris and improve the purity of the samples ( Figures 4A and 4B), which can be visualized on a hemocytometer or automated cell counter ( Figure 4C). The total number of cells/nuclei will depend on the amount of starting material and efficiency of the cleanup procedure. Generation of cDNA libraries for downstream applications should yield clean chromatograms of ample cDNA for sequencing ( Figures 4D and 4E).

LIMITATIONS
Some neuronal subtypes appear to be more sensitive to these dissociation processes than others, which can introduce bias into the number and/or ratio of cell types recovered. Also, smaller cell bodies such as glia may be more likely to survive the dissociation process and be overrepresented in the cell population, whereas neuronal cell types with large cell bodies (e.g., Purkinje cells and motor neurons) may be less likely to survive. Researchers should keep in mind that the proportion of cell types recovered with these dissociation protocols may not be fully representative of the actual proportions in the brain region of interest.
Additionally, some cell types represent only a small fraction of total neurons in a particular region and thus harvesting ample numbers of these cells/nuclei without fluorescent reporters can be challenging. For example, single-nuclei RNA-seq experiments on the spinal cord recovered very few cholinergic motor neurons due to their relative sparsity, 16 but using fluorescent reporters allowed for the enrichment of motor neurons. 9,17 Whenever possible, using fluorescent reporters is highly recommended to enrich target cells, with the Cre-dependent Sun1-sfGFP 'INTACT' mouse being one excellent tool to fluorescently label nuclei. 18 TROUBLESHOOTING Problem 1 Dirty sample contaminated with debris from adult brain dissociations (steps #38-45 and #58-61).

Potential solution
Samples harvested from adult brains contain significant amounts of myelin and cellular debris that can be challenging to remove. This is especially for regions with a high ratio of axons to cell bodies such as the striatum and spinal cord. Our main recommendation is to pass these single-cell or -nuclei suspensions through a cell sorter to remove nearly all debris, using DRAQ5+/DAPI-to collect live cells or just DAPI+ to collect all nuclei. Another critical step is to ensure that the layer of debris from the Papain dissociation for adult brains (step #43) is completely removed; if not, this will increase the amount of debris in downstream applications. Additionally, one could reduce the amount of starting tissue, or split the tissue samples into multiple tubes and increase the number of cleanup reactions. Another alternative is to utilize specific reagents to remove myelin, such as Myelin Removal Beads (Miltenyi Biotec), but this adds additional time to the procedure that could result in further cell/nuclei loss.

Potential solution
Inefficient cell or nuclei dissociation can result in clumps of cells or nuclei sticking together, leading to cell/nuclei loss during filtration and/or cell sorting steps. Sometimes clumps may be visible during cell counting. Other times cell/nuclei clumps will not pass through the filters and thus would not be visibly detected, but instead would result in lower than expected yields. To minimize clumping, investigators can extend the incubation time of pronase for embryonic tissue (step #32), papain for adult tissue (step #40), increase the number of triturations (steps #34 and 41) and increase the number of dounces with the homogenizer (step #47). Depending on the source of excess cell/nuclei clumping, any of these alterations should reduce the likelihood of clumping and increase singlecell/nuclei dissociation efficiency.

Problem 3
Evidence of unhealthy, dying or dead cells on hemocytometer or automated cell counter after procedure (steps #37 and 45).

Potential solution
Cells that have been damaged during the preparation can lead to poor results for downstream applications. If single-cell solutions contain a high amount (>10%) of dead or dying cells based on live/ dead cell marker analysis during counting, there are several steps that one can take. First, ensure that cell triturations (steps #34 and 41) are done with fire-polished Pasteur pipettes and that no bubbles are introduced during trituration, as this can damage cells. Second, make sure cells are gently added to filters and solution allowed to pass through via gravity or gentle tapping of the tube (steps #35 and 45); do not use excessive force to pass cell solutions through filter as this can lead to shearing and damage cells. Third, for adult tissue (steps #38-45), brain regions can be cut up into smaller pieces to ensure efficient dissociation with Papain kit and minimize cell damage during this procedure.

Potential solution
This can be an issue when using fluorescent reporters to select specific neuronal subtypes in a particular brain region. For nuclei preparations, we recommend flash freezing desired brain regions on dry ice and then storing tissue at À80 C. This allows researchers to collect sufficient tissue samples and pool them together for processing, thus increasing the starting number of desired cells and ultimate yield. As noted in step #46, we recommend using < 100 mg of tissue for each cleanup reaction. If combination of frozen tissue samples exceeds this weight, split sample into 2 (or more) tubes for cleanup process.

Potential solution
In our experience using the Sony SH800 Cell Sorter, the number of cells or nuclei obtained after sorting is $85%-100% of the expected number based the sorter information. If the output is lower than expected, researchers can increase the number of sorted cells collected (if possible) to increase total yield. We have found that additional spins and reconstitutions after sorting can lead to loss of 20%-40% of cells or nuclei, so we recommend collecting sorted cells in small volumes of desired solutions for downstream applications whenever possible to eliminate the need for additional spins and reconstitutions.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to lead contact Timothy J. Petros (tim.petros@nih.gov).

Materials availability
This study did not generate any novel reagents.

Data and code availability
This study did not generate any datasets or original code.