Low-input Nucleus Isolation and Multiplexing with Barcoded Antibodies of Mouse Sympathetic Ganglia for Single-nucleus RNA Sequencing

Summary

This protocol describes the detailed, low-input sample preparation for single-nucleus sequencing, including the dissection of mouse superior cervical and stellate ganglia, cell dissociation, cryopreservation, nucleus isolation, and hashtag barcoding.

Abstract

The cardiac autonomic nervous system is crucial in controlling cardiac function, such as heart rate and cardiac contractility, and is divided into sympathetic and parasympathetic branches. Normally, there is a balance between these two branches to maintain homeostasis. However, cardiac disease states such as myocardial infarction, heart failure, and hypertension can induce the remodeling of cells involved in cardiac innervation, which is associated with an adverse clinical outcome.

Although there are vast amounts of data for the histological structure and function of the cardiac autonomic nervous system, its molecular biological architecture in health and disease is still enigmatic in many aspects. Novel technologies such as single-cell RNA sequencing (scRNA-seq) hold promise for the genetic characterization of tissues at single-cell resolution. However, the relatively large size of neurons may impede the standardized use of these techniques. Here, this protocol exploits droplet-based single-nucleus RNA sequencing (snRNA-seq), a method to characterize the biological architecture of cardiac sympathetic neurons in health and disease. A stepwise approach is demonstrated to perform snRNA-seq of the bilateral superior cervical (SCG) and stellate ganglia (StG) dissected from adult mice.

This method enables long-term sample preservation, maintaining an adequate RNA quality when samples from multiple individuals/experiments cannot be collected all at once within a short period of time. Barcoding the nuclei with hashtag oligos (HTOs) enables demultiplexing and the trace-back of distinct ganglionic samples post sequencing. Subsequent analyses revealed successful nuclei capture of neuronal, satellite glial, and endothelial cells of the sympathetic ganglia, as validated by snRNA-seq. In summary, this protocol provides a stepwise approach for snRNA-seq of sympathetic extrinsic cardiac ganglia, a method that has the potential for broader application in studies of the innervation of other organs and tissues.

Introduction

The autonomic nervous system (ANS) is a crucial part of the peripheral nervous system that maintains body homeostasis, including the adaption to environmental conditions and pathology¹. It is involved in the regulation of multiple organ systems throughout the body such as the cardiovascular, respiratory, digestive, and endocrine systems. The ANS is divided into sympathetic and parasympathetic branches. Spinal branches of the sympathetic nervous system synapse in ganglia of the sympathetic chain, situated bilaterally in a paravertebral position. The bilateral cervical and thoracic ganglia, especially the StG, are important components participating in cardiac sympathetic innervation. In disease states, such as cardiac ischemia, neuronal remodeling can occur, resulting in a sympathetic overdrive². The neuronal remodeling has been demonstrated in multiple histological studies in humans and several other animal species^3,4,5,6. A detailed biological characterization of cardiac ischemia-induced neuronal remodeling in cardiac sympathetic ganglia is currently lacking, and the fundamental biological characteristics of specialized neuronal cell types or subtypes within the cardiac sympathetic nervous system (SNS) are not fully determined yet in health and disease⁷.

Novel technologies, such as scRNA-seq, have opened gateways for the genetic characterization of small tissues on a single-cell level^8,9. However, the relatively large size of neurons may impede the optimized use of these single-cell techniques in humans¹⁰. In addition, single-cell sequencing requires a high-throughput of cells to recover a sufficient cell number due to a high loss in the sequencing process. This might prove challenging when studying small tissues that are hard to capture in one session and require multiple samples to introduce enough single cells for sequencing. The recently developed droplet-based snRNA-seq technology (i.e., the 10x Chromium platform) allows the study of biological differences among single nuclei^11,12. snRNA-seq holds an advantage over scRNA-seq for large cells (>30 µm), which may not be captured in Gel Bead in Emulsions (GEMs), as well as improved compatibility with extensive dissociation and/or prolonged preservation^13,14,15.

Heterogeneity, the number of neuronal cells, and other cells enriched in the cardiac SNS are important aspects for the characterization of the ANS in health and disease states. In addition, the organ- or region-specific innervation by each sympathetic ganglion contributes to the complexity of the SNS. Moreover, cervical, stellate, and thoracic ganglia of the sympathetic chain have been shown to innervate different regions of the heart¹⁶. Therefore, it is necessary to perform single-nucleus analysis of ganglionic cells derived from individual ganglia to study their biological architecture.

Droplet-based snRNA-seq allows transcriptome-wide expression profiling for a pool of thousands of cells from multiple samples at once with lower costs than plate-based sequencing platforms. This approach enables droplet-based snRNA-seq to be more suitable for cellular phenotype classification and new subpopulation identification of cells within the SCG and the StG. Notably, this protocol provides a concise stepwise approach for the identification, isolation, and single-nucleus RNA sequencing of sympathetic extrinsic cardiac ganglia, a method that has the potential for a broad application in studies of the characterization of ganglia innervating other related organs and tissues in health and disease.

Protocol

This protocol describes all steps required for the snRNA-seq of murine cervical and/or cervicothoracic (stellate) ganglia. Female and male C57BL/6J mice (15 weeks old, n = 2 for each sex) were used. One additional Wnt1-Cre;mT/mG mouse was used to visualize the ganglia for dissection purposes^17,18. This additional mouse was generated by the crossbreeding of a B6.Cg-Tg(Wnt1-cre)2Sor/J mouse and a B6.129(Cg)-Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo/J mouse. All animal experiments were carried out according to the Guide for Care and Use of Laboratory Animals published by NIH and approved by the Animal Ethics Committee of the Leiden University (License number AVD1160020185325, Leiden, The Netherlands). See the Table of Materials for details regarding all materials, equipment, software, and animals used in the protocol.

1. Preparations

NOTE: All steps are performed in a cell culture flow cabinet.

1. Clean the forceps and scissors by immersing the instruments in 70% ethanol for 20 min.
2. Prepare the ganglion medium consisting of Neurobasal Medium supplemented with B-27 plus (1x), L-glutamine (2 mM), and 1% Antibiotic-Antimycotic. Prewarm the ganglion medium at room temperature.
3. Prepare the digestion solution: 0.25% Trypsin-EDTA (1:1) and 1,400 U/mL collagenase type 2 dissolved in the ganglion medium.
4. Prepare fresh, cold (4 °C) cell wash buffer (0.4% bovine serum albumin [BSA]) and lysis buffer (10 mM Tris-HCl, 10 mM NaCl, 3 mM MgCl₂, and 0.1% nonionic detergent, 40 U/mL RNAse in nuclease-free water) for the nucleus isolation.
5. Prepare nucleus wash buffer (1x phosphate-buffered saline [PBS] with 2.0% BSA and 0.2U/µL RNase Inhibitor).
6. Prepare ST staining buffer (ST-SB) (10 mM Tris-HCl, 146 mM NaCl, 21 mM MgCl₂, 1 mM CaCl₂, 2% BSA, 0.02% Tween-20 in nuclease-free water).

2. Dissection of adult mouse superior cervical ganglia (SCG)

1. Euthanize the mice and keep them on ice.
   NOTE: In the current study, a total of 4 C57BL6/J mice were euthanized by CO₂ asphyxiation. Alternatively, isoflurane can be used followed by exsanguination when a large amount of blood needs to be collected for other study purposes.
2. Fix the mouse on a dissection board with pins and douse it with 70% ethanol to minimize the dispersion of fur (shaving is not necessary).
3. Under a stereomicroscope, open the skin of the neck region by making a midline cut with scissors, move the submandibular glands aside, and remove the sternomastoid muscle to expose and locate the common carotid artery and its bifurcation (Figure 1A, B, see arrow).
4. Dissect the right and left carotid artery bifurcation and the tissue attached to it. Transfer each dissected piece of tissue to a separate 3.5 cm Petri dish containing cold PBS.
5. Look for the SCG attached to the carotid bifurcation. Clean the SCG further by removing the artery and other attached tissue in the Petri dish (Figure 1E).

3. Dissection of adult mouse stellate ganglia (StG)

1. To dissect the StG, make a midline cut in the abdomen, followed by opening the diaphragm and the ventral thoracic wall.
2. Remove the heart and lungs to expose the dorsal thorax. Look for the left and right StG anterolateral to the musculus colli longus (MCL) at the level of the first rib (Figure 1C, D, indicated by dashed lines).
3. Dissect both left and right StG with forceps and separately transfer them to 3.5 cm Petri dishes containing cold PBS (Figure 1F).
   ​

4. Isolation and cryopreservation of mouse ganglionic cells

Steps 4-6 are summarized in Figure 2.

1. Carefully transfer all individual SCG and StG to separate 1.5 mL microcentrifuge tubes with forceps.
   NOTE: Do not use pipette tips to transfer the ganglia because the ganglia are prone to adhere to the wall of plastic pipette tips.
2. Add 500 µL of 0.25% trypsin-EDTA solution to each microcentrifuge tube and incubate in a shaking water bath at 37 °C for 40 min.
   NOTE: This step is aimed to facilitate the digestion and cell release hereafter in the collagenase type 2 solution.
3. Prepare a 15 mL tube containing 5 mL of ganglion medium for each sample. Allow the ganglia to settle down at the bottom of the microcentrifuge tubes. Collect the supernatant, which contains very few ganglionic cells, transfer the supernatant to the prepared 15 mL tubes, and label each tube. Alternatively, to save some time, aspirate the trypsin-EDTA supernatant without collection as very few dissociated cells can be detected in it.
   NOTE: A small amount of trypsin-EDTA solution (~10-30 µL) can be left in the microcentrifuge tube to avoid removal of the ganglia. Avoid pipetting at this step because it may damage the ganglia and lead to low output of ganglionic cells afterward.
4. Add 500 µL of collagenase type 2 solution into each microcentrifuge tube and incubate in a shaking water bath at 37 °C for 35-40 min. Try to resuspend the ganglion after 35 min; if the ganglion is still intact and does not dissociate, prolong the incubation time or increase the concentration of collagenase type 2 solution as necessary.
   NOTE: The incubation time could vary depending on the ganglion size.
5. Resuspend the ganglia in collagenase solution by pipetting up and down ~10 times or until tissue clumps are no longer detected.
6. Transfer the cell suspension to the previously used 15 mL tube that contains the ganglia culture medium and the trypsin-EDTA suspension from the same ganglion. Spin down the cell suspension with a swinging bucket rotor centrifuge for 10 min, 300 × g at room temperature. Carefully discard the cell supernatant.
   NOTE: Because the ganglionic cells are dissociated from a single ganglion, the cell pellet may be too small to detect by eye; a small amount of supernatant can be left in the tube to avoid accidental removal of the cell pellet.
7. Resuspend the ganglionic cells in 270 µL of fetal bovine serum (FBS, low endotoxin) and transfer each cell-FBS suspension into a 1 mL cryovial.
8. To count the cells, mix 5 µL of the ganglionic cell suspension with 5 µL of 0.4% trypan blue dye and load the mixture into a hemocytometer. Count the total and live-cell numbers under a microscope.
   NOTE: The cell viability (live cell count/total cell count = viability %) is usually above 90% with this dissociation protocol. The live cell count of a single ganglion (either SCG or StG) usually falls within the range of 9,000-60,000 cells when the ganglion is isolated from a mouse aged 12 to 16 weeks.
9. Add 30 µL of dimethyl sulfoxide (DMSO) to each cell-FBS suspension in the cryovials, mix well, and transfer the cryovials into a cell freezing container, which is kept at room temperature. Store the cryovial loaded container at -80 °C overnight, and transfer the cryovials to liquid nitrogen the next day for long-time preservation before sequencing.

5. Nucleus isolation

NOTE: Left and right SCG isolated from four mice (in total 8 samples) were used as an example in the following nucleus isolation and sequencing preparation. Keep everything on ice during the whole procedure. Because of the invisibility of small nucleus pellets, a centrifuge with swinging buckets is highly recommended to facilitate supernatant removal throughout the whole procedure.

1. Prepare 15 mL tubes with a strainer (30 µm) on top and prerinse the strainer with 1 mL of the ganglion medium.
2. Take out the cryovials from liquid nitrogen and immediately thaw them in a water bath at 37 °C. When a small pellet of ice is left in the cryovial, take the cryovials out of the water bath.
3. Recover the ganglionic cells by dropping 1 mL of the ganglion medium into each cryovial while shaking carefully by hand. Optional: To evaluate cell recovery, mix the cell suspension after recovery and take 5 µL of cell suspension out for live-cell counting, as described in step 4.8.
4. Load each ganglionic cell suspension on a separate strainer (prepared in step 5.1) and rinse each strainer with 4-5 mL of ganglion medium.
5. Centrifuge the strained cell suspension for 5 min at 300 × g, remove the supernatant carefully, and resuspend the cells in 50 µL of cell wash buffer.
6. Transfer the cell suspension to a low-binding DNA/RNA 0.5 mL microcentrifuge tube.
7. Centrifuge the cell suspension at 500 × g for 5 min at 4 °C.
8. Remove 45 µL of the supernatant without touching the bottom of the tube to avoid dislodging the cell pellet.
9. Add 45 µL of chilled Lysis Buffer and gently pipette up and down using a 200 µL pipette tip.
10. Incubate the cells for 8 min on ice.
11. Add 50 µL of cold nucleus wash buffer to each tube. Do not mix.
12. Centrifuge the nucleus suspension at 600 × g for 5 min at 4 °C.
13. Remove 95 µL of the supernatant without disrupting the nucleus pellet.
14. Add 45 µL of chilled nucleus wash buffer to the pellet. Optional: Take 5 µL of nucleus suspension, mix with 5 µL of 0.4% trypan blue to count, and check the quality of nuclei under a microscope with a hemocytometer.
15. Centrifuge the nucleus suspension at 600 × g for 5 min at 4 °C.
16. Remove the supernatant without touching the bottom of the tube to avoid dislodging the nucleus pellet.

6. Nucleus barcoding with hashtag oligos (HTOs) and multiplexing

NOTE: HTO staining steps were modified and optimized for nuclear labeling of very low amounts of (ganglionic) nuclei according to the previous application in cortical tissue by Gaublomme et al.¹⁵.

1. Add 50 µL of ST-SB buffer to the nucleus pellet, gently pipette 8-10 times until the nuclei are completely resuspended.
2. Add 5 µL of Fc Blocking reagent per 50 µL of the ST-SB/nuclei mix and incubate for 10 min on ice.
3. Add 1 µL (0.5 µg) of single-nucleus hashtag antibody per tube of the ST-SB/nuclei mix and incubate for 30 min on ice.
   NOTE: Shorter incubation time leads to lower efficiency of hashtag labeling, as demonstrated in the representative results.
4. Add 100 µL of ST-SB to each tube. Do not mix.
5. Centrifuge the nucleus suspension for 5 min, 600 × g at 4 °C.
6. Remove 145 µL of the supernatant without disrupting the nucleus pellet.
7. Repeat steps 6.4 and 6.5. Remove the supernatant as much as possible without touching the bottom of the tube to avoid dislodging the nucleus pellet.
8. Resuspend the nucleus pellet in 50 µL of ST-SB, and gently mix the nuclei.
9. Take 5 µL of the nucleus suspension and mix it with 5 µL of 0.4% trypan blue to count the nuclei under a microscope. See Figure 3A for a representative image of nuclei mixed with trypan blue and loaded in a hemocytometer.
10. Centrifuge the nucleus suspension for 5 min at 600 × g at 4 °C.
11. Resuspend the nuclei in ST-SB to achieve a target nucleus concentration of 1,000-3,000 nuclei/µL for each sample according to the corresponding nucleus count.
12. Pool the samples to achieve the desired number of cells.
    NOTE: For example, in this experiment, 8 samples were equally pooled to achieve a total of 25,000 nuclei to immediately proceed to 10x Genomics Chromium and snRNA-seq afterward. The nucleus count usually falls within the range of 6,000-40,000 cells when the ganglion is isolated from a mouse aged 12 to 16 weeks. Only around half of the total loaded nuclei can be captured by droplet-based snRNA-seq. For example, a 25,000 nucleus mixture was prepared to ensure capture of 10,000 nuclei, which is needed for further library preparation and sequencing.

Representative Results

Quality control analysis of the single-nucleus cDNA library preparation and snRNA-seq
Representative results describe sequencing results of 10,000 captured nuclei in a single pool with a 25,000 reads/nucleus gene expression library and a 5,000 reads/nucleus hashtag library. Figure 3B illustrates the quality control results of the 1^st strand cDNA, gene expression (GEX) library, and HTO library, which were checked with Bioanalyzer. The HTO-derived cDNAs are expected to be smaller than 180 bp, whereas mRNA-derived cDNAs are larger than 300 bp. A high-quality GEX library can be detected as a broad peak from 300 to 1,000 bp, and the HTO library is detected as a specific peak of 194 bp. Cell Ranger was used for demultiplexing, fastq file generation, and read alignment by default setting. Seurat R package¹⁹ was subsequently used for quality control and downstream analyses.

Demultiplexing of the snRNA-seq data was performed by identifying HTOs using the Seurat built-in demultiplexing strategy. The demultiplexing ability for each HTO was first visualized with HTO expression files (Supplemental Figure S1). In the heatmap of Figure 4A, singlets are detected as nuclei with specific HTO expression, while doublets and negatives show nonspecific expression of multiple or no HTOs. Of note, approximately 33% of the nuclei were detected as negatives using a 10 min HTO antibody incubation approach. Prolongation of the incubation time (step 6.3) from 10 min to 30 min in a subsequent experiment revealed a remarkable decrease in negatively labeled nuclei (Supplemental Figure S2). These findings indicate that prolonging antibody incubation time may improve hashtag efficiency.

Violin plots in Figure 4B-D demonstrate the number of genes (nFeature_RNA), number of unique molecular identifiers (UMI) (nCount_RNA), and the percentage of mitochondrial counts (percent.mt) within the snRNA-seq dataset to identify outliers and low-quality nuclei. Nuclei were only included in downstream analysis when the following criteria were met: i) nFeature_RNA > 500 and nCount_RNA < 20,000; ii) percent.mt < 5%; iii) the individual nucleus showed clear expression of a single HTO. Gene expression counts were normalized using the default method in Seurat: 875 (2.71%) genes were detected as highly variable genes (Figure 4E). snRNA-seq GEX was scaled, was performed and the elbow plot was used to assess the inclusion of principal components that would be used for downstream analyses (Figure 4F). In total, 18 PCs were included. Clustering was performed with a resolution of 0.4.

The nuclei were clustered, and dimension reduction (UMAP) was performed for visualization of the 12 individual clusters (Figure 5A). The median raw gene count per cluster varies between 991.5 and 4,586 (Supplemental Figure S3A, B). Visualizing the division of the HTO antibodies within the UMAP reveals a clear distribution of ganglia, indicating that all clusters are presented in each ganglion (Figure 5B). To validate the accuracy of HTO sample segregation, the expression of X Inactive Specific Transcript (Xist, expressed in the inactive female X chromosome) was assessed to identify the male samples and the female samples (Figure 5C). Xist expression was in accordance with the hashtag labeling, showing that HTO 1-4 labeled samples were female samples, and HTO 5-8 labeled samples were male samples. This suggests that the curated HTO labeling is highly specific.

To further verify the quality and resolution of the sequencing data with the present method, some key transcripts of sympathetic neurons were first examined. The results show the presence of sympathetic neurons that highly express Th, Dbh, and Snap25 in clusters 5 and 7 (Figure 5D-F). Satellite glial cells were detected with the expression of S100b in clusters 0-3 (Figure 5G)²⁰. Endothelial cells were detected in cluster 4 with a high expression of Pecam1 (Figure 5H) and stromal cells in cluster 8 with a high expression of Acta2 (Figure 5I). These results support the successful nuclei capture of neuronal, satellite glial, endothelial, and stromal cells of the sympathetic ganglion using snRNA-seq.

Figure 1: Dissection of adult mouse superior cervical ganglia and stellate ganglia. (A) Brightfield image of the location of the SCG. (B) To facilitate visualization, a Wnt1Cre;mT/mG mouse was used. Asterisks indicate the SCG (eGFP⁺), arrowheads indicate the bifurcation of the carotid artery. Left panel, phase contrast image; right panel, fluorescent image. (C) Brightfield image of the location of the StG. (D) Asterisks indicate the StG (eGFP⁺), dashed lines indicate the MCL. Left panel, phase contrast image; right panel, fluorescent image. (E) Dissected ganglia are transferred into a Petri dish separately for further cleaning under a stereomicroscope. (F) Left panel, the dissected SCG with the carotid artery still attached. Dashed outline indicates the SCG. Right panel,the dissected and cleaned StG has the shape of an inverted triangle, as indicated by the dashed outline. Scale bar = 1,000 µm. Abbreviations: SCG = superior cervical ganglia; StG = stellate ganglia; MCL = musculus colli longus; eGFP = enhanced green fluorescent protein. Please click here to view a larger version of this figure.

Figure 2: Workflow of sample preparation and hashtag staining-based multiplexing for snRNA-seq. The flowchart depicts the steps from the dissociation of ganglionic cells (orange), nucleus isolation (blue), and hashtag antibody staining and multiplexing (green) that are carried out for snRNA-seq. Abbreviations: FBS = fetal bovine serum; ST-SB = ST staining buffer; snRNA-seq = single-nucleus RNA sequencing. Please click here to view a larger version of this figure.

Figure 3: Quality control of nucleus isolation and gene expression library preparation. (A) Phase-contrast image of the HTO-stained nuclei. Nuclei are indicated with arrows. Scale bar = 100 µm. (B) Bioanalyzer results of 1^st strand cDNA (top), GEX library (middle), and HTO library (bottom). Abbreviations: GEX = gene expression; HTO = hashtag oligo. Please click here to view a larger version of this figure.

Figure 4: Quality control of the hashtag oligo labeling efficiency and quality control of snRNA-seq. (A) Heatmap of HTO staining achieved with an incubation time of 10 min with hashtag antibodies. (B-D) Violin plots of quality control metrics depicting the number of genes (nFeature_RNA, B); the number of UMIs (nCount_RNA, C); the percentage of mitochondrial counts (percent.mt, D). (E) Of the total 32,285 genes sequenced, 875 (2.71%) were identified as highly variable features as visualized in the scatter plot. (F) Elbow plot of PCs to determine inclusion of true signal used for clustering. Abbreviations: snRNA-seq = single-nucleus RNA sequencing; HTO = hashtag oligo; PCs = principal components. Please click here to view a larger version of this figure.

Figure 5: Representative results of the analysis of snRNA-seq. (A) UMAP plot of the clustered snRNA-seq dataset. (B) UMAP plot visualizing the distribution of the pooled HTO samples. (C) Violin plot validating female samples (high Xist expression) after demultiplexing. (D-I) UMAP plots displaying selected marker genes; the clusters highly express the corresponding genes which are indicated within the red circles: D-F, Sympathetic neurons (Th, Dbh, Snap25); (G) Satellite glial cells (S100b); (H) Endothelial cells (Pecam1); (I) Stromal cells (Acta2). Abbreviations: snRNA-seq = single-nucleus RNA sequencing; HTO = hashtag oligo; UMAP = uniform manifold approximation and projection. Please click here to view a larger version of this figure.

Supplemental Figure S1: Quality control metrics of a representative single-nucleus sequencing result. HTO expression profiles of individual samples visualizing the demultiplexing ability of the used hashtag antibodies. Abbreviation: HTO = hashtag oligo. Please click here to download this File.

Supplemental Figure S2: HTO expression heatmap of subsequent samples incubated with HTO antibodies for 30 min. Comparison of the number of negative-labeled nuclei after HTO demultiplexing with Figure 4A shows a marked improvement of nucleus labeling after prolonging the antibody incubation time. Abbreviation: HTO = hashtag oligo. Please click here to download this File.

Supplemental Figure S3: Median and quantiles of gene expression per cluster. (A) Median nonzero raw gene expression of each cluster. (B) Descriptive statistics of nonzero raw gene expression of each cluster. Q1: 25^th percentile, Q3: 75^th percentile. Please click here to download this File.

Discussion

Here, a detailed protocol is described that focuses on i) the dissection of adult mouse superior cervical and stellate sympathetic ganglia, ii) the isolation and cryopreservation of the ganglionic cells, iii) nucleus isolation, and iv) nucleus-barcoding with HTO labeling for multiplexing purposes and snRNA-seq.

With this protocol, sympathetic ganglionic cells can easily be obtained by dissociating individual ganglia using commonly used trypsin and collagenase. Long-term preservation of isolated ganglionic cells is also readily achieved by freezing cells in FBS supplemented with 10% DMSO, which showed a high quality of recovery after thawing. Moreover, compared to conventional single-cell RNA sequencing, the use of droplet-based snRNA-seq of single murine sympathetic ganglia combined with the application of HTO staining-based nucleus-barcoding has the following advantages: i) samples can be preserved for a long time until all samples are ready for further nucleus isolation; ii) nuclei of good quality isolated from multiple small-size ganglia can be pooled together for sequencing without a batch effect caused by separated sample preparation; iii) the ability to trace back the distinct ganglionic origin after sequencing by using nucleus barcodes; and iv) cost-effectiveness, since only single library preparation is needed. Importantly, the described isolation and cell culture protocol provides a single uniform method for both murine cervical and stellate ganglia and is potentially applicable to other ganglia, such as dorsal root ganglia and other species, e.g., human ganglia.

One of the major advantages of scRNA-seq is the ability to identify (novel) cell types and to reveal rare cell populations that could not be detected by bulk RNA-seq. The droplet-based scRNA-seq platform facilitates the capture of more cells. It thus can provide an aggregate view of the cell (sub)types and transcriptional heterogeneity of a large cell population compared to a plate-based sequencing platform. However, the droplet-based (e.g., 10x Chromium) platform is not suitable for cells larger than 50 µm, limiting its application in large cells such as human neurons (~100 µm). The availability of the snRNA-seq technique overcomes this drawback because of the small size of a nucleus. Moreover, snRNA-seq is a useful method for gene expression studies of highly interconnected and low-yield cells such as neurons and frozen tissues.

Although it is possible to directly isolate nuclei from tissues without prior cell dissociation, it was beneficial for the yield to take a two-step nucleus isolation method that first dissociates the ganglion into single cells (which can be preserved in liquid nitrogen) followed by the nucleus isolation. Because of the small size of a mouse sympathetic ganglion, it was found that more nuclei were obtained using a two-step nucleus isolation method than with a one-step nucleus isolation approach. The quality check of the cDNA, library, and sequencing analyses indicates good nuclei/RNA quality. In addition, logistics are improved, since samples can be collected and stored at different time points before collective sequencing. Single-nucleus analysis also revealed the successful recovery and capture of neurons and glial cells, suggesting that the described two-step nucleus isolation approach might be better suited for the application in small tissues.

Another advantage of this protocol is the multiplexing with barcoded antibodies²¹. The mouse sympathetic ganglion is a tiny tissue (average size 0.1 mm³), and the low number of cells derived from an individual ganglion is insufficient for droplet-based sequencing by itself. However, pooling several ganglia of different mice or different ganglia of the same mouse will cause the loss of either individual mouse information or individual ganglion information. As a solution, the HTO staining step is easy to perform and enables the barcoded labeling of nuclei derived from different mice or different ganglia before nucleus pooling. The accuracy of HTO demultiplexing is verified in this protocol by matched Xist expression in known female nuclei populations. Nucleus multiplexing with barcoded antibodies, therefore, reduces batch effects and lowers the sequencing cost.

A potential limitation of snRNAseq might be that there may be differences between the RNA composition in the nucleus and cytoplasm due to the natural presence of nascent transcripts in the nucleus, associated with early response to neuronal activities^22,23. The nucleus and cytoplasm may also differ in transcripts depending on the cell cycle state²⁴. Fewer transcripts were detected in an individual nucleus (~7,000 genes) than in a cell (~11,000 genes)¹⁴. Therefore, scRNA-seq and snRNA-seq may yield different results at a transcript level. Nevertheless, the comparison between scRNA-seq and snRNA-seq demonstrated a similar capability to discriminate neuronal cell types of brain tissue¹⁴. To improve the discrimination between highly similar cell types or subtypes by snRNA-seq, more nuclei might be needed to compensate for the lower gene detection ability compared to scRNA-seq. Furthermore, although the accuracy of HTO demultiplexing is sufficient, the loss of some data is inevitable as not all nuclei show HTO specificity. Further optimization of the antibody incubation could minimize the number of nuclei with double or negative expression of HTOs.

Taken together, this protocol provides the experimental procedure for sequencing neuronal nuclei from sympathetic ganglia by means of an easy-to-follow workflow starting from ganglion isolation to nuclei preparation of low input of cells, followed by HTO staining-based nucleus labeling for snRNA-seq. The protocol provides a detailed overview of all key steps that can be easily performed and applied to various ganglia in murine and other species.

Materials

Name	Company	Catalog Number	Comments
Chemicals and reagents
0.25% Trypsin-EDTA	Thermo Fisher Scientific	25200056
0.4% trypan blue dye	Bio-Rad	1450021
Antibiotic-Antimycotic	Gibco	15240096
B-27	Gibco	A3582801
Collagenase type 2	Worthington	LS004176	use 1,400 U/mL
Dimethyl sulfoxide	Sigma Aldrich	67685
Ethanol absolute ≥99.5%	VWR	VWRC83813.360
Fetal bovine serum (low endotoxin)	Biowest	S1810-500
L-glutamine	Thermo Fisher Scientific	25030024
Neurobasal Medium	Gibco	21103049
Bovine Serum Albumin 10%	Sigma-Aldrich	A1595-50ML	Cell wash buffer
DPBS (Ca2+, Mg2+free)	Gibco	14190-169	Cell wash buffer
Magnesium Chloride Solution, 1 M	Sigma-Aldrich	M1028	Nucleus Lysis buffer
Nonidet P40 Substitute (nonionic detergent)	Sigma-Aldrich	74385	Nucleus Lysis buffer
Nuclease free water (not DEPC-treated)	Invitrogen	AM9937	Nucleus Lysis buffer
Protector RNase Inhibitor, 40 U/µL	Sigma-Aldrich	3335399001	Nucleus Lysis buffer
Sodium Chloride Solution, 5 M	Sigma-Aldrich	59222C	Nucleus Lysis buffer
Trizma Hydrochloride Solution, 1 M, pH 7.4	Sigma-Aldrich	T2194	Nucleus Lysis buffer
Bovine Serum Albumin 10%	Sigma-Aldrich	A1595-50ML	Nucleus wash
DPBS (Ca2+, Mg2+free)	Gibco	14190-169	Nucleus wash
Protector RNase Inhibitor,40 U/µL	Sigma-Aldrich	3335399001	Nucleus wash
Bovine Serum Albumin 10%	Sigma-Aldrich	A1595-50ML	ST staining buffer (ST-SB)
Calcium chloride solution, 1 M	Sigma-Aldrich	21115-100ML	ST staining buffer (ST-SB)
Magnesium Chloride Solution, 1 M	Sigma-Aldrich	M1028	ST staining buffer (ST-SB)
Nuclease free water (not DEPC treated)	Invitrogen	AM9937	ST staining buffer (ST-SB)
Sodium Chloride Solution, 5M	Sigma-Aldrich	59222C	ST staining buffer (ST-SB)
Trizma Hydrochloride Solution, 1M, pH 7.4	Sigma-Aldrich	T2194	ST staining buffer (ST-SB)
Tween-20	Merck Millipore	822184	ST staining buffer (ST-SB)
TotalSeq-A0451 anti-Nuclear Pore Complex Proteins Hashtag 1 Antibody	Biolegend	682205	Hashtag antibody
TotalSeq-A0452 anti-Nuclear Pore Complex Proteins Hashtag 2 Antibody	Biolegend	682207	Hashtag antibody
TotalSeq-A0453 anti-Nuclear Pore Complex Proteins Hashtag 3 Antibody	Biolegend	682209	Hashtag antibody
TotalSeq-A0461 anti-Nuclear Pore Complex Proteins Hashtag 11 Antibody	Biolegend	682225	Hashtag antibody
TotalSeq-A0462 anti-Nuclear Pore Complex Proteins Hashtag 12 Antibody	Biolegend	682227	Hashtag antibody
TotalSeq-A0463 anti-Nuclear Pore Complex Proteins Hashtag 13 Antibody	Biolegend	682229	Hashtag antibody
TotalSeq-A0464 anti-Nuclear Pore Complex Proteins Hashtag 14 Antibody	Biolegend	682231	Hashtag antibody
TotalSeq-A0465 anti-Nuclear Pore Complex Proteins Hashtag 15 Antibody	Biolegend	682233	Hashtag antibody
TruStain FcX (human)	Biolegend	422302	FC receptor blocking solution
Equipment and consumables
Bright-Line Hemacytometer	Merck	Z359629-1EA
Centrifuge 5702/R A-4-38	Eppendorf	EP022629905
CoolCell LX Cell Freezing Container	Corning	CLS432003-1EA
Cryovial	Thermo Scientific	479-6840
DNA LoBind 0.5 mL Eppendorf tube	Eppendorf	EP0030108035-250EA
Eppendorf Safe-Lock Tubes 1.5 mL	Eppendorf	30121872
Falcon 35 mm Not TC-treated Petri dish	Corning	351008
Falcon 15 mL Conical Centrifuge Tubes	Fisher scientific	10773501
Forceps Dumont #5	Fine science tools	11252-40
Hardened Fine Scissors	Fine science tools	14091-09
Ice Pan, rectangular 4 L Orange	Corning	CLS432106-1EA
Leica MS5	Leica		Microscope
Moria MC50 Scissors	Fine science tools	15370-50
Noyes Spring Scissors	Fine science tools	15012-12
Olympus CK2 ULWCD	Olympus		Microscope
P10	Gilson	F144802
P1000	Gilson	F123602
P200	Gilson	F123601
Preseparation Filters (30 µm)	Miltenyi biotec	Miltenyi biotec130-041-407
Shaking water bath	GFL	1083
Silicon plate	RubberBV	3530	Dissection board
Software and packages
Cell ranger	V4.0.0
R programming	V4.1.1
R sudio	V1.3.1073
Seurat	V4.0
tydiverse	V1.3.1
Animals
B6.Cg-Tg(Wnt1-cre)2Sor/J mouse	The Jackson Laboratory	JAX stock #022501
B6.129(Cg)-Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo/J mouse	The Jackson Laboratory	JAX stock #007576
C57BL/6J mice	Charles River
Code for the data analysis
https://github.com/rubenmethorst/Single-cell-SCG_JoVE