Isolation of Nuclei from Mammalian Cells and Tissues for Single-Nucleus Molecular Profiling

Abstract

Both single-cell RNA sequencing (scRNAseq) and single-nucleus RNA sequencing (snRNAseq) can be used to characterize the transcriptional profile of individual cells, and based on these transcriptional profiles, help define the cell type distribution in mixed cell populations. However, scRNAseq analyses are confounded if some of the cells are large (>50 micron) or if some of cells adhere more tightly to some adjacent cells than to other adjacent cells. Further, single cell isolation for scRNAseq requires fresh tissue, which may not be available for human or animal model tissues. Nuclei for snRNAseq, on the other hand, can be isolated from any cell, regardless of size, and from either fresh or frozen tissues. Additionally, the current enzymatic and mechanical methods for single-cell dissociation can lead to stress-induced transcriptional artifacts.

Here, we describe a time- and cost-effective procedure to isolate nuclei from mammalian cells and tissues that is suitable for separation and subsequent single-nucleus RNA sequencing (snRNAseq) or other high-throughput single-cell molecular profiling techniques. The protocol incorporates steps to mechanically disrupt samples to release nuclei. Compared to conventional nuclei isolation protocols, the approach described here increases overall efficiency, eliminates risk of contaminant exposure, and reduces volumes of expensive reagents. A series of RNA quality control checks are also incorporated to ensure success and reduce costs of subsequent snRNAseq experiments. Nuclei isolated by this procedure can be separated on the 10x Genomics Chromium system for either snRNAseq and/or Single-Nucleus ATAC-Seq (snATAC-Seq), and is also compatible with other single cell platforms.

Basic Protocol 1: Sample preparation and QC via RNA Isolation and Analysis

Basic Protocol 2: Nuclei Isolation Protocol

Introduction:

Heterogenous cell populations define the anatomical and functional complexity of tissues. The proper function of organs across different physiological settings hinges upon the organization of these cell types. Until recently, there has been limited knowledge of the diverse cellular composition of organs and of the molecular signatures and spatial distributions of specific cells that fulfill the different functions in organs. RNA sequencing has revolutionized the ability to visualize differences in gene expression levels (Hwang, Lee, & Bang, 2018).

Bulk-RNA sequencing provides an overview of the average gene expression in each tissue, across all cell types present. Despite its sensitivity, this technique masks cell-specific responses. The introduction of single-cell RNA sequencing (scRNAseq) provides insights into each cell’s role in biological processes, including how changes in gene expression can contribute to disease (Bakken et al., 2018; Hwang et al., 2018).

Despite its potential, scRNAseq has various limitations. Cell re-suspension and library preparation require fresh tissue, and the required droplet-based sequencing platforms have a cell size restriction of 25–50 μm (Del-Aguila et al., 2019). ScRNAseq also requires enzymatic digestion, which can be harsh and cause some more sensitive cell types to be less represented in the final dataset (Bakken et al., 2018). Additionally, current enzymatic and mechanical methods for single-cell dissociation can lead to stress-induced transcriptional artifacts (Wu, Kirita, Donnelly, & Humphreys, 2019).

Single-nucleus RNA sequencing (snRNAseq) can overcome some of the above-mentioned issues by using nuclei rather than whole cells for transcriptional profiling. snRNAseq is able to profile gene expression in difficult-to-isolate cells, including frozen and even long-time archived tissue (Bakken et al., 2018). Compared to whole cells, nuclei are more resistant to mechanical pressures and appear not to exhibit as many cell isolation-based transcriptional artifacts (Bakken et al., 2018). In addition, snRNAseq does not have cell size restrictions and can accommodate large cells, including large cardiomyocytes —the cells that compose the cardiac muscle —, which are 100 μm in length and 25 μm in width(Hwang et al., 2018).

This protocol details the isolation of nuclei, which is essential for single-nucleus molecular profiling, and sets the stage for the success of these downstream experiments. This protocol was designed so there is a tissue sample quality check prior to nuclei isolation. The tissue quality check involves the isolation of RNA from a sample prepared in parallel to the sample prepared for nuclei isolation. As the samples are cut from the same section of tissue (one piece cut in half), the piece for RNA isolation is a good measurement of the quality of the piece for nuclei isolation. The RNA is isolated and evaluated for its purity and integrity (Basic Protocol 1), which is critical for downstream experiments. If this sample passes the quality check, the user then moves on to Basic Protocol 2, which outlines the nuclei isolation procedure.

After tissue disruption and release of the nuclei using a TissueLyser II device, the sample is filtered to remove large debris, and nuclei are purified using flow activated cell sorting (FACS) to separate Hoechst-positive nuclei (previously stained with NucBlue) from small-particle debris. Following FACS, the quantity and quality of the sorted nuclei suspension are further validated using an automated cell counter. The nuclei are also visualized under a microscope to confirm that they are of correct morphology and size and that debris is no longer present in the sample.

This protocol yields high-quality nuclei suitable for separation and subsequent high-throughput single-cell molecular profiling, including snRNAseq and snATAC-Seq. See Figure 1 for an overview of the whole procedure.

Figure 1: Schematic Representation of experimental design and workflow.

Basic Protocol 1 includes the Sample Preparation and QC Check via RNA Isolation and Analysis. Basic Protocol 2 details the Nuclei Isolation Protocol.

BASIC PROTOCOL 1:

Sample preparation and QC via RNA Isolation and Analysis

Basic Protocol 1 provides a quality check on the tissue that is used for nuclei isolation in Basic Protocol 2, as the integrity of RNA molecules is essential to the success of downstream transcriptomic profiling experiments (Schroeder et al., 2006). Basic Protocol 1 first details the parallel preparation of the tissue samples for Basic Protocol 1 (Piece 1a) and Basic Protocol 2 (Piece 1b) in which one piece of tissue is cut in half. To check the quality of the tissue of interest, RNA is isolated from Piece 1a and the purity and integrity is analyzed to determine if this tissue is suitable for nuclei isolation and downstream transcriptomic profiling. If the RIN of the RNA from the tissue of Piece 1a is less than five, the tissue sample of interest is not suitable for the nuclei isolation protocol and thus, Piece 1b should not be used for the nuclei extraction. However, if the RIN of the RNA from the tissue of Piece 1a is five or greater, Piece 1b can be used for Basic Protocol 2, nuclei extraction. As Piece 1a and Piece 1b were prepared in parallel from the same original piece of tissue, we assume Piece 1a is a good proxy for the quality of Piece 1b. However, this is under the assumption that there is no mishandling or user error that could affect the quality of RNA in either piece of tissue.

This protocol requires the use of a TissueLyser II device, a refrigerated centrifuge, and a TapeStation (HS or Regular Screen Tape) or other device/materials to quantify RNA integrity, including a Bioanalzyer or gel electrophoresis. The tissues of interest, either fresh or frozen (on dry ice), are retrieved prior to dissection. Two very small pieces of tissue are first cut from the sample of interest and transferred to separate microcentrifuge tubes, each with a stainless-steel bead. A bucket of ice should be prepared and all materials should be located and set up in advance so they are accessible for the protocol. Glycogen should be removed from the −20°C freezer at the beginning of protocol, so the Glycogen can thaw. Ethanol should be kept on ice, so the Ethanol is chilled prior to use. The TissueLyser II adapter plate and rack should be placed in −20°C freezer for fifteen minutes prior to use. The Sample preparation and QC via RNA Isolation and Analysis Protocol, described here, only uses the first tube, with Piece 1a.

Materials for Basic Protocol I

Tissue Sample (Piece 1a), fresh or frozen

Dry Ice, (Fisher Scientific, cat no. NC1296841), stored in airtight container

Dissection tools

Disposable Polystyrene Forceps, Tradewinds Direct (VWR, cat no. 12576–934),

Micro dissecting scissors, 4 1/2”, straight (BRI, cat no. 25–1000),

Scalpel disposable sterile size 11 (VWR, cat no. 89176–382),

Round cell culture petri dish, 6 cm diameter (VWR, cat no. 253840090),

TRIzol Reagent (Invitrogen, cat no. 15596–018), stored at room temperature

Chloroform HPLC Plus (Sigma-Aldrich, cat no. 650498), stored at room temperature

Isopropyl Alcohol Molecular Biology grade (Fisher Scientific, cat no. BP2618500), stored at room temperature

Nuclease-Free Water (not DEPC-Treated) (Ambion, cat no. AM9937), stored at room temperature

Koptec 190 proof Ethanol (VWR, cat no. TX89125164HU), stored at room temperature (chilled on ice prior to use)

Glycogen (for molecular biology) (Roche, cat no. 10901393001), store in −20°C freezer

1.5 ml Non-Stick RNase-free Microfuge Tubes (Ambion, cat no. AM12450),

Serological Pipettes, 10 ml (VWR, cat no. 89130–898),

Serological Pipettes, 5 ml (VWR, cat no. 89130–908),

2200 TapeStation (Agilent Technologies, cat no. G2991AA) or similar fluorimeter,

Qiagen TissueLyser II (Qiagen, cat no. 85300),

Qiagen; Stainless Steel Beads, 5 mm for TissueLyser II (Qiagen, cat no. 69989),

Fisherbrand 2 ml microcentrifuge tube for TissueLyser II (Fisher Scientific, cat no. 02–681-332),

Qiagen TissueLyser II adapter plates and racks (Qiagen, cat no. 11980),

Refrigerated Centrifuge (Beckman, cat no. 8G089),

Vortex Mixer (BenchMixer Vortex Mixer, Genesee Scientific, cat no. 31–100),

Protocol Steps

Preparing Tissue Samples for Basic Protocol 1 & Basic Protocol 2

• 1Retrieve the tissue samples of interest. If frozen, remove the tissue from the −80°C freezer in a bucket with enough dry ice to cover the samples. Cut two equal size pieces, Piece 1a and Piece 1b, (~5×5×5 mm each) from the mammalian tissue sample of interest using sterile dissection tools (forceps, scalpel, scissors) in a cell culture petri dish.

  Tissues are not homogenous - ensure the pieces include all layers or other tissue components relevant for the analysis and appropriately matched for the subsequent nuclei purification.

• 2Using sterile forceps, place each of the samples in two separate 2 ml microcentrifuge tubes, Tube 1a (Piece 1a) and Tube 1b (Piece 1b), each containing one 5 mm bead. Store Tube 1b in −80°C freezer until performing Basic Protocol II (Nuclei Isolation). If not directly proceeding to QC via RNA Isolation and Analysis, store Tube 1a in −80° C freezer as well. (Figure 2)

Figure 2:

Sample tissue pieces to be used. Shown here are two equally-sized frozen human tissue pieces (heart tissue) (~5 × 5 × 5 mm each). By using very small pieces of tissue to isolate nuclei, precious tissues samples can be used for additional experiments.

QC via RNA Isolation and Analysis

• 3Place a TissueLyser (I or II) adapter plate and rack in −20°C freezer for a minimum of fifteen minutes prior to use. (Step 5)
• 4Transfer Tube 1a to wet ice and add 1 ml of TRIzol.
• 5Remove the Qiagen TissueLyser II adapter plate and rack from the −20°C freezer, load Tube 1a, and place rack in Qiagen TissueLyser II. Lyse tissue for two minutes at a frequency of 25 Hz. After two minutes, remove sample tube and check if lysis is complete by confirming samples appear to be homogenously disrupted and no large pieces of tissue are present. Run an additional one minute if samples need further disruption to be homogenous.

  Do not lyse tissues more than needed, as doing so can damage the RNA.

• 6Pipette Tube 1a up and down five times to mix the sample and transfer homogenized sample into new 1.5 ml microcentrifuge tube (Tube 2a).
• 7Add 200 μl Chloroform to Tube 2a, cap tube, and shake well by hand for fifteen seconds (the sample should turn milky pink color).
• 8Incubate sample at room temperature (RT, ~24°C) for three minutes.
• 9Centrifuge samples at 12,000 g x fifteen minutes at 4°C.
• 10In the meantime, add 2 μl glycogen to a new 1.5 ml Non-Stick RNase-free Microfuge tube (Tube 3a).
• 11After centrifugation, carefully transfer ~450 μl (the top of aqueous layer) of Tube 2a to Tube 3a. Be careful not to include any of the middle protein layer or bottom TRIzol layer of sample. (Figure 3)
• 12Add an equal volume of isopropyl alcohol to Tube 3a (equal to the aqueous layer volume transferred in step 10 (~450 μl)).
• 13Vortex Tube 3a for 5 seconds, and incubate at RT for 10 minutes.
• 14Centrifuge Tube 3a at 12,000 g x 10 minutes at 4°C.
• 15After centrifugation, visualize the pellet at the bottom of the Tube 3a, and carefully remove supernatant without disturbing the pellet.
• 16Add 500 μl of chilled 75% EtOH (keep on ice) to Tube 3a, and gently vortex for five seconds.
• 17Centrifuge Tube 3a at 10,000 g x five minutes at 4°C.
• 18Remove all of the 75% ethanol supernatant from Tube 3a, and air dry for one minute.

  Do not overdry pellet; as the pellet will be difficult to resuspend.

• 19Add 50 μl nuclease-free water and resuspend the pellet. Gently vortex for 10 seconds and centrifuge briefly for 5 seconds, and keep on ice to avoid RNA degradation.
• 20Using a small aliquot (~1 μl), quantify the amount of RNA from each well (for quality check and concentration measurement).

Figure 3: Transfer of Aqueous Top Layer from Tube 2a to Tube 3a in Step 11 of Basic Protocol 1.

After centrifugation, carefully transfer ~450 μl (the top of aqueous layer) of Tube 2a to Tube 3a. This figure provides an image of the three separate layers. Be careful not to include any of the middle protein layer or bottom TRIzol layer of sample.

Several methods are available for RNA analysis, but we use an Agilent TapeStation and the Agilent RNA Screen Tape. See the Agilent High Sensitivity RNA ScreenTape System Quick Guide for instructions. Other devices for quantifying RNA concentration and quality can also be used, such as a Bioanalyzer (see Agilent RNA 6000 Nano Kit Quick Start Guide), gel electrophoresis (see Agarose Gel Electrophoresis, Current Protocols (Armstrong & Schulz, 2015)), or Nanodrop (see NanoDrop Nucleic Acid Quantification, Thermo Fischer Scientific). If the user performs a gel electrophoresis, intact total RNA on a eukaryotic sample will have sharp, clear 28S and 18S rRNA bands, in which the 28S rRNA band should be approximately twice as intense as the 18S rRNA band. If a Nanodrop device is used, a 260/280 ratio of ~ 2.0 is generally accepted as “pure” for RNA.

For additional details and information, see the appropriate manufacturer’s protocol, depending on which device and method is available at facility. If a methodology is used which does not provide a RIN value, refer to the protocol of that methodology for how to determine the quality of RNA. We do not recommend proceeding to the next step with samples showing RIN values of less than five.

Basic Protocol 2: Nuclei Isolation

The use of isolated nuclei, rather than whole cells, allows for the transcriptomic profiling of difficult-to-isolate cells in both fresh and frozen tissues (Del-Aguila et al., 2019). This protocol details the steps of nuclei isolation and should only be performed if the RIN of the RNA isolated from the tissue in Basic Protocol 1 (Piece 1a) is five or greater. If that is the case, Tube 1b, which contains Piece 1b from the tissue, is removed from the −80°C freezer and used to isolate the nuclei. Through a series of steps including tissue homogenization with a TissueLyser II, filtration, centrifugation, and fluorescence-activated cell sorting for Hoechst-positive nuclei (stained by NucBlue) for nuclei enrichment, a suspension of nuclei is generated. The nuclei suspension concentration is then measured on an automated cell counter device, and the morphology and size of the nuclei are assessed under a microscope. The nuclei can then be used for subsequent high-throughput molecular profiling.

Prior to this protocol, prepare and keep all solutions (see the Reagents and Solution section below) on ice and set the centrifuge to 4°C in advance. Remove DTT (Dithiothreitol) and 50x Protease Inhibitor from −20°C freezer, and let thaw at room temperature, so these solutions can be used to make NIM2. Remove 40 U/μl RNaseIn, 20 U/μl SuperaseIn, and 40 U/μl Protector RNaseIn immediately prior to use from −20°C freezer and return to freezer immediately after use. Locate and set up all materials in advance so they are accessible for the protocol. the TissueLyser II adapter plate and rack should be placed in the −20°C freezer for fifteen minutes prior to use, and the centrifuge should be set to 4°C.

Materials for Basic Protocol 2

Tissue Sample (Piece 1b), stored in −80°C freezer immediately prior to use,

Reagents for Nuclei Isolation Buffer 1 (NIM1)

1.5 M Sucrose (Sigma-Aldrich, cat no. S0389), stored at room temperature,

2 M KCl (Thermo Fisher Scientific, cat no. AM9640G), stored at room temperature,

1 M MgCl₂ (Thermo Fisher Scientific, cat no. AM9530G), stored at room temperature,

1 M Tris, pH 8 (Thermo Fisher Scientific, cat no. AM9855G), stored at room temperature,

Reagents for Nuclei Isolation Buffer 2 (NIM2) (see Reagents and Solutions)

NIM1 (see above)

1 mM DTT (Thermo Fisher Scientific, cat no. P2325), stored in −20°C freezer,

50x Protease Inhibitor (Roche, cat no. 11873580001), stored in −20°C freezer,

Reagents for Homogenization Buffer (HB)

NIM2 (see Reagents and Solutions)

40 U/μl RNaseIn (Thermo Fisher Scientific, cat no. AM2684), stored in −20°C freezer,

20 U/μl SuperaseIn (Invitrogen, cat no. AM2696), stored in −20°C freezer,

10% (v/v) Triton X-100 (Sigma-Aldrich, cat no. T8787–100ML), stored at room temperature,

Reagents for Storage Buffer (SB)

PBS (1X) 10×500ml (Invitrogen, cat no. 10010–049), stored in 4°C refrigerator,

Bovine Serum Albumin (BSA) (Sigma-Aldrich, cat no. A4503–10G), stored in 4°C refrigerator,

40 U/μl Protector RNaseIn (Sigma-Aldrich, cat no. 03335402001), stored in −20°C freezer,

Nuclease free Water (not DEPC-Treated) (Ambion, cat no. AM9937), stored in RT

NucBlue Live ReadyProbes Reagent (Thermo Fisher Scientific, cat no. R37605), stored at room temperature,

Trypan Blue Stain 0.4% (Fisher Scientific, cat no. T8154–100M), stored at room temperature,

1.5 ml Non-Stick RNase-free Microfuge Tubes (Ambion, cat no. AM12450),

15 ml Falcon Centrifuge Tubes (Corning, cat no. 352097),

50 ml Falcon Centrifuge Tubes (VWR, cat no. 21008–951),

Serological Pipettes, 10 ml (VWR, cat no. 89130–898),

Serological Pipettes, 5 ml (VWR, cat no. 89130–908),

Cell strainers, Falcon, 40 μm (Corning, cat no. 352340),

FACS tubes (Corning, cat no. 352235),

Qiagen TissueLyser II (Qiagen, cat no. 85300),

Qiagen; Stainless Steel Beads, 5 mm for TissueLyser II (Qiagen, cat no. 69989),

Fisherbrand 2 ml microcentrifuge tube for TissueLyser II (Fisher Scientific, cat no. 02–681-332),

Qiagen TissueLyser II adapter plates and racks (Qiagen, cat no. 11980),

Refrigerated Centrifuge (Beckman, cat no. 8G089),

Refrigerated Ultracentrifuge (Centrifuge 5424 R, Eppendorf, cat no. 2231000655),

Vortex (BenchMixer Vortex Mixer, Genesee Scientific, cat no. 31–100),

Microscope and imaging system (Keyence, cat no. BZ -X710),

Countess II Automatic Cell Counter (Thermo Fisher Scientific, AMQAX1000),

Counting slides for Countess II (Invitrogen, cat no. C10228),

Florescence Activated Cell Sorting Machine (FACS) (FACSAria II Flow Cytometer, BD Biosciences, cat no. 650033).

Protocol Steps

Nuclei Isolation

1. Remove Tube 1b from the −80°C freezer and place it on wet ice immediately.

2. Add 1 ml of Homogenization Buffer (HB).

3. Remove Qiagen TissueLyser II adapter plate and rack from the −20°C freezer and load with samples (up to eight).

4. Run TissueLyser II for one minute at a frequency of 25 Hz. Assess the samples to ensure all samples are adequately lysed. If samples appear to be homogenously disrupted, proceed to the next step. Run an additional one minute if samples need further disruption to be homogenous.

   Do not lyse tissues more than needed, as nuclei damage may occur.

5. Remove samples from the TissueLyser II adapter rack and place on ice. Allow the samples to lyse on ice for five minutes.

6. Filter the homogenized tissue through 40 μm cell strainer into a 50 ml tube.

7. Centrifuge at 500 g x 5 minutes at 4°C to pellet the nuclei.

8. Transfer supernatant to a separate 1.5 ml Eppendorf tube and keep on ice.

   The supernatant can be used for RNA Isolation for additional Bulk-RNA experiments.

9. Resuspend the nuclei pellet from step 7 in 500 μl of Storage Buffer (SB) and transfer all sample through the filter of a FACS tube (35 μm filter).

10. Prepare unstained, negative control by pipetting 100 μl of the filtered sample to a separate FACS sorting tube.

11. Add two drops of NucBlue to the original filtered suspension (Nuclei resuspended in 500 μl of Storage Buffer from Step 9). Vortex for five seconds and place on ice. Wait for 15 minutes for staining to appropriately take place.

12. Pipette 150 μl of SB to a new 1.5 ml Non-Stick RNase-free Microfuge Tube, and place on ice to bring to FACS device.

    This tube will be used to collect the sample following FACS sorting. Prepare one tube for each sample being processed, up to eight.

13. Proceed to FACS (which should occur at least fifteen minutes after staining with NucBlue). Follow instructions for flow activated cell sorting to separate Hoechst-positive nuclei from small-particle debris in the BD FACSAria II User’s Guide with Nozzle size 70 μm, Amplitude 4.6 Volts, and Frequency 87 kHz. See the appropriate manufacturer’s protocol, depending on which FACS device is available at facility.

14. Sort through the entire sample to maximize the yield of nuclei. Each sample should run through the FACS machine for approximately thirty minutes. (Figure 5).

15. Keep samples (which are now in the 1.5 ml Non-Stick RNase-free Microfuge Tube with SB from Step 12) on ice until placed in refrigerated centrifuge.

16. Centrifuge sample at 500xg for five minutes at 4°C to produce nuclei pellet.

17. Slowly remove the supernatant to reduce the risk of disturbing nuclei.

18. Collect and save supernatant in another 1.5 ml Non-Stick RNase-free Microfuge tube.

19. Re-suspend nuclei pellet in SB according to the Resuspension volumes shown below.

    SB volumes are based on the sort count provided from FACS. The volumes provided were developed to achieve a concentration of approximately 1.0×10⁶ nuclei.

    FACS Cell Numbers	SB VOLUME (μl)
    50,000 or less	30
    50,000 to 100,000	50
    100,000 to 150,000	70
    150,000 or more	90

20. For each sample, take 6 μl of the nuclei suspension and mix with 6 μl of Trypan blue (1:1 dilution) in a new 1.5 ml Non-Stick RNase-free Microfuge Tube. Load 10 μl of each sample with Trypan blue into one well on a Countess Cell Counting Chamber Slide. Load 10 μl of each sample without Trypan blue into the other well on the slide.

21. Count the nuclei 3x using Countess (Figure 6) using the Trypan blue side, moving the slide around while in the Countess to capture the concentration in different areas. The ideal concentration of nuclei should be between 8×10⁵ to 1.2×10⁶.

    Users should move the region of the slide that is being captured, while the slide is inserted into the Countess machine, so the concentration can be measured at different regions within the sample. Users should first push the slide into the Countess, allowing the slide to click into the normal position. The concentration should be measured in this position. Then the user should slightly move the slide from this position, allow the machine to focus on the new position, and count the concentration in the new position. The slide should be counted at three positions and these values should be averaged.

    If countess numbers are low, stain the supernatant (from Step 17) with Trypan blue (1:1 dilution). Load 10 μl of the supernatant with Trypan blue onto a Countess Cell Counting Chamber Slide, and repeat Step 21 to check if nuclei were accidentally transferred with the supernatant.

22. Use the side of the Countess slide without Trypan blue to visualize the nuclei under a microscope.

    A standard light microscope can be used with a 10x or 20x lens to view the nuclei. Microscopes with automated platforms, such as a Keyence, can also be used, to observe the presence of the fluorescent DAPI label in the nuclei, as the sample had been previously stained with NucBlue. Check nuclei morphology to confirm presence of nuclei and removal of debris. An intact nucleus should be round and 8–12 μm in diameter (e.g. nuclei from heart tissue) (Figure 7)

    Nuclei isolated by this procedure are ready to be separated on the 10x Genomics Chromium system for either snRNAseq and/or Single-Nucleus ATAC-Seq (snATAC-Seq), and are also compatible with other single cell platforms. We recommend proceeding immediately to the protocol for the single cell platform available at the user’s facility.

Figure 5:

Flow Activated Cell Sorting (FACS) for capturing nuclei populations and removing all unstained debris. This figure provides an example of the results of FACS for nuclei enrichment, in which a nuclei suspension from frozen mammalian heart tissue was sorted. A BD FACSAria II was used with Nozzle size 70 μm, Amplitude 4.6 Volts, and Frequency 87 kHz A) The measurements of forward (FSC) and sideward scatter (SSC) allow for the discrimination of cells by size, separating the smaller nuclei from the larger debris. B) The nuclei also separate from the debris. After the nuclei are stained with NucBlue, they are Hoechst-A-positive, and are able to be discriminated from the debris, which are Hoechst-A-negative. (left) A “Sort Count” is provided, which includes the number of nuclei, which separated from debris. (right) C) This panel provides the overall results of the gating, and the percentage of events (nuclei), which meet the given cell size and Hoechst-A parameters.

Figure 6: Measurement of Nuclei Count.

Following FACS, which separates nuclei from debris, the user must obtain an accurate estimate of nuclei concentration in the sample. After resuspending the nuclei pellet in storage buffer, the resuspended pellet is loaded onto a slide with Trypan blue in an automated cell counter device (Countess). A) The automated cell counter is used to visualize the nuclei suspension and determines the concentration of nuclei in the suspension. The nuclei counted are circled in red. B) The automated cell counter also provides a “Results” section, which represents the total concentration of nuclei in the nuclei suspension. We measure this three times at different regions of the slide to obtain an accurate concentration. The expected concentration of nuclei should be between 8×10⁵ to 1.2×10⁶, which falls into the ideal range for downstream experiments, which is designed to maximize the likelihood of achieving the desired recovery target. This number is determined from the 10X Genomics 3’ Library Manual (User should see manual for given downstream experiment). The Countess device is also able to provide the percent of nuclei, which are live and dead, as a result of the Trypan blue stain. Trypan blue is a charged cell impermeant stain, which is only able to enter cells or nuclei if the membrane is compromised. As the starting material is frozen mammalian tissue, the samples are expected to have ~5% live nuclei and ~95% dead nuclei.

Figure 7: Nuclei from Human Heart Tissue Sample.

Following the determination of the concentration of nuclei on the automated cell counter device, the user should confirm the morphology of nuclei and removal of debris, on a microscope (Basic Protocol 2, Step 22). Any microscope can be used. In this image, we used a Keyence microscope and visualized the sample on 20x using Brightfield, while overlaying the image with DAPI (as the sample was previously stained with NucBlue) to highlight the nuclei in blue.

REAGENTS AND SOLUTIONS

Nuclei isolation buffer 1 (NIM1):

This solution is used prior to Basic Protocol 2: Nuclei Isolation. It is used to make NIM2, which is used to make HB (Step 2, Basic Protocol 2). Make in advance, mixing the components listed below, and store up to six months at 4°C. This protocol is based on 8 samples for analysis by the 10x Chromium chip, which has eight lanes. Of course, volumes should be adjusted based on the number of samples being processed.

Nuclei Isolation Buffer 1 (NIM1)
Component	Volume [μl] 1 sample	Final Concentration [mM]
1.5 M Sucrose	625	250
2 M KCl	46.875	25
1 M MgCl2	18.75	5
1 M Tris, pH 8	37.5	10
NFW	3021.875	-
Total	3750	-

Nuclei isolation buffer 2 (NIM2):

This solution is used prior to Basic Protocol 2: Nuclei Isolation. It is used to make HB (Step 2, Basic Protocol 2). Prepare fresh, mixing the components listed below, and keep on ice. These numbers are calculated so that there is spare volume available, if needed. Adjust according to number of samples being processed.

Nuclei isolation buffer 2 (NIM2)
Component	Volume [μl] 1 sample	Final Concentration [mM]
NIM1	1246.25	-
1 mM DTT	1.25	1 μM
50xProtease Inhibitor	25	-
Total	1250	-

Homogenization Buffer (HB):

This solution is used in Step 2 of Basic Protocol 2: Nuclei Isolation. Prepare fresh, mixing the components listed below, and keep on ice. These numbers are calculated so that there is spare volume available, if needed. Adjust according to number of samples being processed. See table below.

Homogenization Buffer
Component	Volume [μl] 1 sample	Final Concentration [mM]
NIM2	1212.5	-
40 U/μl RNaseIn	12.5	0.4 U/μl
20 U/μl SuperaseIn	12.5	0.2 U/μl
10% (v/v) Triton X-100	12.5	0.10%
Total	1250	-

Storage Buffer (SB):

This solution is used in Steps 9, 12, and 19 of Basic Protocol 2: Nuclei Isolation. Prepare fresh, and keep on ice. These numbers are calculated so that there is spare volume available, if needed. Adjust according to number of samples being processed. See table below.

Storage Buffer
Component	Volume [μl] (1 sample)	Final Concentration [mM]
PBS (-)	995	-
BSA	40 mg	4%
40 U/μl Protector RNaseIn	5	0.2 U/μl
Total	1000	-

Commentary

Background

The anatomical and functional complexity of tissues is determined by heterogenous cell populations. In order for there to be normal function of any organ, the highly heterogenous cell populations within the organ must properly perform their specialized functions, which are governed by differential gene expression (Hwang et al., 2018). There have been recent advances of methods for high-throughput transcriptome profiling, including bulk-RNA-sequencing, single-cell RNA sequencing (scRNAseq), and single-nucleus RNA sequencing (snRNAseq) (Hwang et al., 2018).

Bulk RNA sequencing often masks the uniqueness of each cell, as the technique only provides an overview of the average gene expression within the tissue, concealing cell-specific responses. This presents challenges when the biological effect of interest is limited to a subpopulation of cells, which is critical to understand the pathogenesis of a certain disease. Single-cell RNA sequencing provides insights into RNA expression at the level of an individual cell and has been instrumental to the Human Cell Atlas Project, which aims to identify each cell population in the human body (Regev et al., 2017). However, as mentioned above, single-cell RNA sequencing also has limitations including the requirement of fresh tissue, cell size restriction of 25–50 μm, harsh enzymatic digestion, and the potential for stress-induced transcriptional artifacts (Wu et al., 2019).

By using nuclei rather than whole cells for transcriptional profiling, snRNAseq is able to profile gene expression in difficult-to-isolate cells, including frozen and even long-time archived tissue, without cell-size restrictions or the presence of transcriptional artifacts. (Bakken et al., 2018). snRNAseq, however, profiles less mRNA, as nuclei contain 20% of all cellular transcripts (Bakken et al., 2018). Despite this, in most organ systems, there is evidence that individual nuclei provide appropriate gene expression information to define cell types and related sub-populations (Lake et al., 2016).

While there are many protocols to isolate nuclei for RNA sequencing, this protocol provides a novel approach by incorporating quality control steps and using an efficient and uniform method of tissue dissociation, the TissueLyser II device. This protocol was developed to optimize a nuclei isolation protocol by Litvinukova et al. (Litvinukova, 2018) We increase the number of samples that can be processed at a time, reduce the time to process each sample, limit the risk of contamination, and lower cost compared to prior protocols (Litvinukova, 2018; Ohkawa, Mallappa, Vallaster, & Imbalzano, 2012; Tadevosyan, Allen, & Nattel, 2015; Tan, Li, & Lim, 2010).

We incorporated the QC check via RIN Isolation and Analysis (Basic Protocol 1) into our protocol following multiple unsuccessful snRNAseq experiments with frozen human heart samples with poor quality RNA. We established a RIN cutoff of less than 5 based on the results of many transcriptomic profiling experiments which were successful with a RIN ≥ 5, but not <5.

Droplet-based scRNAseq techniques are still very expensive. In order to maximize the use of the 10x Chromium chip, which has eight lanes, this protocol is able to easily process eight samples at once. This decreases both the time and cost associated with each experiment.

Additionally, the mechanical disruption of tissues to isolate nuclei in this protocol is more time- and cost-effective than in prior methods (Litvinukova, 2018; Ohkawa et al., 2012; Tadevosyan et al., 2015; Tan et al., 2010). This protocol was optimized to use small amounts of tissues (5×5×5 mm each), preserving precious tissue for future experiments. In the prior protocol by Litvinukova et al, the user would break apart each piece of tissue with a mortar and pestle under liquid nitrogen followed by further mechanical disruption with a dounce homogenizer and homogenization buffer (Litvinukova, 2018). Not only did each of these materials need to be cleaned after each run, which increased time and potential contamination, but those protocols also used larger volumes of buffer reagents (Litvinukova, 2018; Ohkawa et al., 2012; Tadevosyan et al., 2015; Tan et al., 2010). This protocol uses the TissueLyser II to mechanically release nuclei, which is a uniform and efficient method of tissue disruption (see TissueLyser II Handbook, Qiagen).

This uniform and rapid method of tissue homogenization coupled with the additional quality control checks have exponentially increased the success of our molecular profiling experiments as well as the potential tissues available for transcriptomic profiling. This protocol has been successfully applied to very small pieces (~5×5×5 mm) of fresh and frozen cell lines and tissue derived from humans, murine and porcine organs. These include heart tissue, liver tissue, and vasculature tissue, such as the portal and pulmonary veins.

Critical Parameters & Troubleshooting:

Assessment of RNA Quality Prior to Nuclei Isolation

The assessment of total RNA quality by RNA integrity is an important step for any RNA sequencing experiment. Although intact nuclei can be isolated from many tissues, in some instances, the RNA associated with the nuclei will be degraded and not suitable for subsequent analyses. We feel it is essential to assess RNA quality (a function of both integrity and purity) in a tissue prior to nuclei isolation, which is the goal of Basic Protocol 1. If RNA obtained from a previously frozen tissue is low quality, the experiment should be aborted as successful single-nucleus molecular profiling will be impossible.

Transfer of Aqueous Layer to Tube with Glycogen to Separate RNA

In Step 7 of Basic Protocol 1, chloroform is added to the sample, and centrifuged, which creates three layers: a top aqueous layer containing the RNA, a middle layer with protein, and a bottom layer with TRIzol. The user should be careful to only remove the aqueous top layer, and not the middle or bottom layer, as contaminants from the other layers can lead to sample degradation and adversely affect the final quality of RNA. Tilt Tube 1a to one side, and pipette the aqueous layer from that same side, so there is a larger volume to remove. (Figure 3)

Removing Ethanol from Pellet

In Step 18 of Basic Protocol 1, the 75% ethanol is removed from Tube 3a, so that the pellet can be resuspended in nuclease-free water. The user should keep the nuclease-free water nearby and also on ice, so the water can be immediately added to the pellet. If the pellet is left dry for too long, the pellet will be difficult to resuspend.

Removal of Supernatant from Nuclei Pellet

In Steps 6–8 of Basic Protocol 2, homogenized tissues are filtered through 40 μm cell strainers into a 50 ml tube. The samples are then centrifuged. Nuclei pellets are formed and loosely anchored to the bottom of the tube. To remove the supernatant, the user should visualize the pellet and pour off the supernatant immediately following the above-mentioned centrifugation. If the pellet moves while one is attempting to transfer the supernatant, the user should repeat the original centrifugation to allow the pellet to reform at the bottom of the tube and attempt this step again. If the pellet is still not securely anchored to the bottom of the 50 ml tube, remove the supernatant with a pipette.

Use of Supernatant from Nuclei Isolation for Future Experiments

An additional step included in this protocol is Step 8, in which the isolation of RNA from the cytoplasmic material when the nuclei are pelleted. In prior protocols, the supernatant would be discarded after the nuclei were pelleted (Litvinukova, 2018). In this protocol, the cytoplasmic RNA is extracted for additional experiments, including bulk RNA sequencing.

Staining the Isolated Nuclei with NucBlue

In Step 11 of Basic Protocol 2, the user stains the original filtered suspension with two drops of NucBlue. The user should wait at least fifteen minutes to run the sample through the FACS machine after staining the sample with two drops of NucBlue. This will allow the fluorescent probe in NucBlue to attach to the nuclei prior to FACS sorting.

FACS (Flow Activated Cell Sorting)

While we recommend thirty minutes to sort through each sample in Step 14 of Basic Protocol 2, the FACS machine may need more time to run through the entire sample. The entire samples must undergo FACS to maximize the yield of nuclei. This amount of time for FACS is directly related to the amount of tissue used.

Removing Supernatant from Pellet after FACS Sorting

The cell count after FACS (Step 13–14, Basic Protocol 2) can be low (<50,000) if small tissue samples are used. As a result, pellet visualization following centrifugation at 4°C could be difficult (Basic Protocol 2, Step 16). If the pellet is not easily visualized, users should merely reduce the volume of supernatant rather than remove the entire supernatant volume. Be sure to remove liquid from the opposite side of the anticipated pellet. By leaving behind supernatant, the risk of discarding the nuclei pellet is reduced. Additionally, collect and save supernatant in another 1.5 ml Non-Stick RNase-free Microfuge tube (Step 18). If Countess numbers are low, stain the supernatant (from Step 18) with Trypan blue (1:1 dilution). Load 10 μl of the supernatant with Trypan blue onto a Countess Cell Counting Chamber Slide, and repeat Step 21 to check if nuclei were accidentally transferred with the supernatant.

Obtaining Ideal Nuclei Concentration for Downstream Transcriptomic Profiling Experiments

Count the nuclei 3x using Countess (Figure 6) using the side Trypan blue side (Step 21, Basic Protocol 2). The ideal average concentration of nuclei should be between 8×10⁵ to 1.2×10⁶. If the concentration is below this range and the RIN of the sample is greater than 5, Basic Protocol 2 should be repeated with a larger piece of tissue. If the concentration is above this range and the RIN of the sample is greater than 5, the nuclei suspension should be diluted with Storage Buffer, based on the table provided. (Step 19)

Freezing Isolated Nuclei Following Nuclei Isolation

We recommend proceeding immediately to subsequent experiments following nuclei isolation. We have discovered a loss of >50% of nuclei with freezing and thawing, which significantly impacts future experiments and is not recommended.

Understanding Results

Assessment of RNA Integrity is required regardless of successful nuclei isolation. The purpose of Basic Protocol 1 is to prepare the tissue and QC the tissue sample by isolating its RNA and determining the RNA integrity. This methodology and RIN cutoff of less than five were developed following unsuccessful transcriptomic profiling experiments with frozen heart samples before the RIN QC check was added to the protocol. When the RIN of these tissues was subsequently checked, the experiments using tissue with RIN <5 were the ones that failed and those with RIN ≥ 5 were successful.

Following FACS, the user will obtain a suspension of sorted nuclei and a measurement of the cell count of nuclei sorted in that sample. Based on the starting tissue amount described here (~5×5×5 mm of tissue), this number should be between 50,000 and 150,000. The sample is re-pelleted, resuspended in storage buffer, and loaded onto a Countess slide, which will determine the concentration of nuclei in the suspension. A successful run from a ~5×5×5 mm piece of tissue will produce a suspension with a concentration of nuclei between 8×10⁵ to 1.2×10⁶, as this is the required nuclei concentration for downstream sequencing platforms. These nuclei suspensions can be used for high-throughput molecular profiling modalities, including snRNAseq and snATACseq. If this concentration is not achieved, the experiment may need to be repeated with a larger piece of tissue, as the initial piece of tissue used did not contain a high enough concentration of suitable nuclei.

	Expected Result
RNA Integrity Measurement (RIN)	≥ 5
Sort Count (post-FACS)	50,000–150,000
Concentration of Nuclei (Automatic Cell Counter)	8x105 – 1.2x106

Time Considerations

Sample preparation and QC via RNA Isolation and Analysis, Nuclei isolation, and FACS should take no more than four to five hours to complete. Preparing the tissue samples should take approximately thirty minutes, depending on the number of samples to be processed. Basic Protocol 1 should take between 45–90 minutes and is also dependent on the number of samples assessed. One sample should take approximately 45 minutes. Each additional sample should add 10 minutes to the protocol, to allow for the pipetting steps required.

Basic Protocol 2 should take 3–4 hours depending on the number of samples and the concentration of nuclei in each sample. The user should plan to spend 30 minutes to set up materials and prepare the solutions. Mechanical tissue lysing is performed simultaneously on all samples for one minute and followed by further lysing of cell membranes on ice for additional five minutes. Filter purification and subsequent centrifugation should take five minutes and not exceed 15 minutes depending on the number of samples. There should be enough time for the FACS device to process the entire sample. This typically requires 30 minutes per sample.

Figure 4: Assessment of RNA Integrity is required regardless of successful nuclei isolation.

This figure highlights the importance of the assessment of RNA Integrity prior to nuclei isolation. Nuclei isolation can be successful even in the presence of degraded RNA. The RNA quality of Sample 1 and Sample 2 was assessed using a TapeStation (see Basic Protocol 1). A) Electropherogram of the RNA of Sample 1, which lacks significant 18S and 28S peaks, indicating that Sample 1 does not contain high-quality RNA. B) The electropherogram of the RNA of Sample 2, however, contains these peaks, which indicates that this sample has high-quality RNA and will be suitable for the preparation of nuclei suspensions compatible with subsequent RNA library preparation. C) Sample 1 had a RIN of 3.6, which is less than 5, and therefore not high-quality RNA, despite displaying perfect nuclei under microscopy (as seen in D). By contrast, Sample 2 has a RIN of 6.5. D) This is a picture of nuclei from Sample 1, which are visualized on a microscope. Each nucleus is the correct shape and size (round and 6 μm in diameter), and there are no debris present in this sample. The nuclei are highlighted by DAPI, as the sample had been stained with NucBlue. While these nuclei were able to be isolated, they do not contain good quality RNA, as evidenced in panel A, which lacks strong 18S and 28S peaks, and panel C, which demonstrates a RIN <5 (3.6)