Fluorescence-Activated Nuclei Negative Sorting of Neurons Combined with Single Nuclei RNA Sequencing to Study the Hippocampal Neurogenic Niche

The strategy to exclude neurons using negative FACS sorting generates high-quality RNA sequencing rates of low abundance cell population, like neural stem cells, astrocytes, and oligodendrocytes, to study their role in the modulation of adult hippocampal neurogenesis. With this method, it is possible to adapt the choice of antibody in the FACS gating, which makes this protocol very versatile in addressing various biological questions related to the dentate gyrus. This method is primarily designed to investigate the adult hippocampal niche.

However, it could easily be adapted for studying other stem cell niches. Demonstrating the procedure will be Sara Ahmed de Prado, post-doctoral training fellow at the Francis Crick Institute. To begin, transfer the brain dissected from the euthanized mouse into a 10 centimeter Petri dish filled with ice-cold PBS.

Then place the Petri dish on ice and cut the brain in two halves along the sagittal axis. Using a scalpel, remove the cerebellum. Transfer one half of the brain to the new 10 centimeter Petri dish placed on ice, containing ice-cold PBS.

Dissect the dentate gyrus, or DG, using binoculars, and repeat this step to obtain the second DG from the second half of the brain. Transfer the two DGs into the pre-cooled Dounce homogenizer and add 1 milliliter of cold homogenization buffer, or HB.Homogenize the tissue with 10 strokes of the loose A pestle, followed by 15 strokes of the tight B pestle. Transfer the homogenate into a prechilled 15 milliliter tube.

Rinse the Dounce homogenizer with 1 milliliter of cold HB and combine it with the same tube. Add 3 milliliters of HB to the 15 milliliter tube, and incubate for 5 minutes on ice. Mix this nuclei suspension twice by inverting the tube gently.

Using 0.5 milliliters of HB, pre-wet the 70 micrometer strainer cap placed over a 50 milliliter test tube, then strain the nuclei suspension before washing the cell strainer with 0.5 milliliters of HB.Next, remove the cell strainer and centrifuge the test tube in a swing bucket centrifuge at 500 x g for 5 minutes at 4 degrees Celsius. Discard the supernatant. Gently resuspend the pellet in 4 milliliters of HB using a P1000 pipette.

After 5 minutes of incubation on ice, centrifuge the suspension at 500 x g for 10 minutes, at 4 degrees Celsius. Discard the supernatant and resuspend the pellet in 3 milliliters of wash media. Use 0.5 milliliters of wash media to pre-wet a 35 micrometer strainer cap over a 50 milliliter test tube.

Then strain the nuclei suspension by pipetting 0.5 milliliters of suspension at a time, using a P1000 pipette. After washing the strainer cap with 0.5 milliliters of wash media, place the tube on ice. Transfer the filtrate to a new 15 milliliter tube and centrifuge it for 5 minutes and 4 degrees Celsius, at 500 x g.

Discard the supernatant, and resuspend the pellet in 3 milliliters of wash media. Repeat the centrifugation and resuspend the pellet in 1 milliliter of wash media with the mouse anti-NeuN, Alexa Fluor 488-conjugated antibody and 1 microgram per milliliter DAPI. Incubate the reaction for 45 minutes on ice in the dark.

For fluorescence activated nuclei sorting, or FANS, transfer the immuno-stained nuclei suspension to a 5 milliliter test tube, and place it on ice until the start of the flow cytometry procedure. Vortex the samples for 3 seconds, with mild intensity, before placing the tubes into the FACS instrument. To acquire the data from stained nuclei suspension, set the gates in DAPI-height and the DAPI-area to exclude the cell debris and the aggregated nuclei.

Then set the gates in the log side scatter, or SSC, area, and the log forward scatter, or FSC, area, to separate the single nuclei from the remaining DAPI stained aggregate or cell debris. Then isolate the NeuN-AF488-negative population by setting the gates for the anti-NeuN-AF488 and FSC area. After the analysis, sort the anti-NeuN-AF488 negative population in a 1.5 milliliter collection tube filled with 50 microliters of wash media using the gating strategy.

Once the sorting is done, add 1 milliliter of PBS containing 1%bovine serum albumin, or BSA, to the collection tube to collect droplets from the wall of the tube and centrifuge the tube at 500 x g, for 5 minutes, at 4 degrees Celsius. Discard the supernatant, leaving 50 microliters solute to resuspend centrifugated nuclei. In a 0.5 milliliter microtube, add 5 microliters of the nuclei suspension to 5 microliters of Trypan blue.

Measure the concentration and assess the viability of the single cell suspension using an automated cell counter before performing the library preparation and sequencing of the nuclei. Bioinformatic clustering revealed well-separated groups of nuclei corresponding to known cell types within the DG, with or without FANS. Within the non-FACS-sorted sample, most of the high-quality sequenced nuclei were 84.9%of neurons.

It comprised of three groups of neurons, indicating the possibility of the most representative cell populations in the dentate gyrus to be granule neurons, other excitatory neurons, and inhibitory neurons. Within the non-FACS-sorted sample, the identified non-neuronal clusters were mostly made of 11.1%of glial cell types, including astrocytes, oligodendrocytes, and oligodendrocyte precursor cells, 3.3%immune cells, and 0, 6%Cajal-Retzius cells. While performing FANS to exclude NeuN positive populations, the clusters of glial cells became prominent, with 81.3%of the total nuclei.

For the samples sequenced at 50, 000 reads per nuclei, 25, 010 genes were detected per nuclei for the non-FACS-sorted sample, and 1, 665.5 genes for the NeuN-negative FANS sample, confirming high-quality transcriptomic profiling of single nuclei and FACS sorting does not damage the nuclei for subsequent snRNA-seq. The high proportion of neurons in the non-FACS-sorted sample had a higher transcriptional activity of 2, 660 genes per nuclei and 6, 170 transcripts per nucleus in the non-FACS-sorted sample than the non-neuronal cell types with average transcriptional activity of 1, 090 genes per nucleus and 1, 785 transcripts per nucleus. Once the data has been generated with this protocol, it is worth considering these new orthogonal methods such as special symptomics or in vivo studies to validate any findings.