May 05, 2019
Single-nucleus isolation from frozen human lung tissue for single-nucleus RNA-seq
- Nikita Joshi1,
- Alexander Misharin1
- 1Northwestern University
- Human Cell Atlas Method Development Community
Protocol Citation: Nikita Joshi, Alexander Misharin 2019. Single-nucleus isolation from frozen human lung tissue for single-nucleus RNA-seq. protocols.io https://dx.doi.org/10.17504/protocols.io.zu8f6zw
Manuscript citation:
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: In development
This protocol is working for our group, however, we are still optimizing this protocol.
Created: April 07, 2019
Last Modified: May 05, 2019
Protocol Integer ID: 22144
Keywords: single nucleus RNA-seq, lung, HCA
Abstract
We aimed to develop a protocol for isolation of the single nuclei from the archival frozen human lung tissue suitable for single cell RNA-seq using standard 10x Genomics chemistry. The protocol utilizes standard nuclei isolation buffer (Nuclei EZ buffer) supplemented with RNase inhibitor and a protease inhibitor.
For the homogenization step, we elected to use C Tube and GentleMACS tissue dissociation as a way to standardize the homogenization procedure. To maintain nuclei integrity we skipped washing steps and instead use large volumes of washing buffer and proceed to FACSorting immediately after lysis for nuclei purification.
This protocol produces a good yield of nuclei and diverse libraries, with multiple cell types being detected. While the number of detected genes and UMI per nuclei is lower than for single cell RNAseq, it is sufficient for identification of the specific cell types, including rare populations. Moreover, snRNAseq allowed resolution of the cell types that are intimately integrated into the lung matrix and otherwise hard to resolve: fibroblasts and type 1 alveolar epithelial cells.
We thank Oni Basu (UChicago), Luciano Martelotto (Monash University), Nicole Abreu (10x Genomics) and Sharmila Chatterjee (10x Genomics) for their advice.
Materials
MATERIALS
RNasin(R) Plus RNase Inhibitor, 10,000uPromegaCatalog #N2615
Nuclei EZ lysis buffer Merck MilliporeSigma (Sigma-Aldrich)Catalog #EZ PREP NUC-101
DAPI Thermo Fisher ScientificCatalog #D1306
DPBS (no Ca, no Mg)ThermofisherCatalog #14190144
Albumin, Bovine Serum, 10% Aqueous Solution, Nuclease-FreeMerck MilliporeSigma (Sigma-Aldrich)Catalog #126615-25ML
cOmplete™ EDTA-free Protease Inhibitor CocktailMerck MilliporeSigma (Sigma-Aldrich)Catalog #11873580001
C TubeMiltenyi BiotecCatalog #130-096-334
Pre-Separation Filters (30 µm)Miltenyi BiotecCatalog #130-041-407
Before start
| Buffers | ||||
| 1. cOmplete stock (10x) | 1 ml | |||
| Nuclei EZ Prep buffer | 1000 | ul | ||
| cOmplete | 1 | tablet | ||
| 2. Lysis buffer (1x) | 1 ml | Final concentration | ||
| cOmplete stock 10x | 100 | ul | 1x | |
| Nuclei EZ Prep buffer | 875 | ul | 1x | |
| RNasin Plus (40 U/ul) | 25 | ul | 1 U/ul | |
| 3. Wash buffer (1x) | 1 ml | Final concentration | ||
| PBS | 875 | ul | 1x | |
| BSA 10% | 100 | ul | 1% | |
| RNasin Plus (40 U/ul) | 25 | ul | 1 U/ul | |
| 4. Resuspension buffer (1x) | 1 ml | Final concentration | ||
| PBS | 974 | ul | 1x | |
| BSA 10% | 1 | ul | 0.1% | |
| RNasin Plus (40 U/ul) | 12.5 | ul | 0.5 U/ul | |
| 5. Capture buffer (1x) | 1 ml | Final concentration | ||
| PBS | 775 | ul | 1x | |
| BSA 10% | 200 | ul | 2% | |
| RNasin Plus (40 U/ul) | 25 | ul | 1 U/ul |
General Protocol
General Protocol
Prepare cOmplete stock, 10x, keep on ice.
Prepare Lysis buffer, 2 ml per sample, keep on ice.
Prepare Wash buffer, 4 ml per sample, keep on ice.
Prepare Resuspension buffer, 0.3 ml per sample, keep on ice.
Take human lung sample from -80C or LN2 storage, cut ~5-73 mm piece, keep on dry ice until ready.
Place human lung sample on a small plastic weighing boat, keep on ice, let thaw almost completely (~30-60 sec).
Using 3 ml syringe and 30G needle, inject ~1 ml of the Lysis buffer into the tissue, move the needle to distribute solution evenly ("inflate" the tissue).
Chop with scissors into ~1.5-2.0 mm pieces (~1 min).
Transfer chopped lung tissue and solution into C tube, add the rest of the lysis buffer, final vol 2 ml. Close C tube, invert, make sure that all small pieces are at the base, keep on ice.
Place C tube on MACS Tissue Dissociator and run m_lung_01 program, then run m_lung_02 program for 20 sec, stop, immediately place the tube on ice. The solution will be foamy and will contain small pieces of tissue.
To bring foam down - briefly spin the tube in the swinging bucket rotor centrifuge (~30 sec, at 4C).
Set 30 um filter on top of the 15 ml polypropylene tube (on ice). Using wide bore tip transfer lysis buffer and remaining pieces of the lung tissue on top of the filter.
Rinse the filter with 4 ml of Wash buffer, remove the filter, close the tube, mix by inverting. Keep on ice.
Take 20 ul aliquot for counting: use AO/PI solution, count on Nexcellom K2 Cellometer.
No AO-positive cells should be present, PI-positive nuclei should be uniformly round or oval. Some debris will be present.
Add 2 ul of DAPI (5 mg/ml stock) per 1 ml of the solution (~12 ul). Mix, incubate on ice for 5 min, proceed to cell sorting.
On the sorter: make sure the UV laser is on, use 450/50 filter for DAPI. Gate on DAPI+ cells on log scale, then switch to linear and gate on G0/G1 and G2/M events, exclude subG0 events and doublets.
Optional: SYTO RNA Select Dye can be used instead or in addition to DAPI to identify nuclei containing RNA. Acquire with the Blue (488 nm) laser and 530/50 filter.
Option A: Sort 10K events into RT mix, adjust the volume to 90 ul with H2O, add RT enzyme, proceed to emulsion generation using 10x Chromium.
This step is based on Luciano Martelotto's protocol: https://community.10xgenomics.com/t5/Customer-Developed-Protocols/ct-p/customer-protocols (pdf).
Option B: Sort nuclei into 200 ul of Capture buffer, use protein lo-bind 1.5 ml tubes, after sorting keep on ice.
Pellet nuclei using swinging bucket rotor centrifuge, 300 rcf, 5 min, 4C.
Remove supernatant add appropriate volume of Resuspension buffer (based on the number of sorted nuclei, add enough buffer to obtain 1000 nuclei/ul, adjust for 20% loss during sorting and centrifugation), let sit on ice for 1 min before gently resuspending the pellet. Filter through 30 um filter if necessary.
Count nuclei, adjust concentration if necessary, proceed with 10x Chromium.
Expected results
Expected results
Typical flow cytometry plots:
Typical flow cytometry plots for nuclei prep. Gate P1 separates nuclei from debris. Gate P3 allows exclusion of doublets. Gate P2 allows focusing on the nuclei with high DNA/RNA content. In high qulity tissues SYTO RNA Select and DAPI signals correlate well.
UMAP plot demonstrating clustering of 4,342 nuclei into 16 clusters.
Explanted lung from patient with systemic sclerosis-associated interstitial lung disease was frozen in liquid nitrogen and stored at -80C for several months prior to processing according to the protocol above. Librariy was prepared using 10x Genomics V2 3' chemistry, sequenced on Illumina HiSeq 4000, initial processing was performed using CellRanger package, both exonic and intronic reads were mapped to GRCh38 reference. Secondary analysis was performed using Seurat V3.0 R toolkit.
UMAP plot demonstrating clustering of 4,342 nuclei into 16 clusters, corresponding to alveolar epithelial type II cells (cluster O), alveolar epithelial type I cells (clusters 6 and 9), club (cluster 11) and ciliated (cluster 15) cells, endothelial cells (cluster 13), fibroblasts (cluster 5), macrophage subsets (clusters 1 and 3), monocytes (cluster 4), dendritic cells (cluster 12), mast cells (cluster 14) , plasma cells (cluster 10), T and NK cells (clusters 2 and 7).
Steps that require further optimization:
- Tissue cutting: starting from thick (60-100 um) cryostat sections from the frozen tissue may improve sampling of the cell types deeply integrated into lung matrix.
- Decreasing concentration of the EZ lysis buffer 10 times does not affect nuclei yield. Impact on nuclei quality and cell type bias needs to be tested and validated.
- Decreasing centrifugation speed from 400 rcf to 300 rcf results in nuclei with better morphology and less "blebbing".
- Use of 70 um nozzle instead of 100 um nozzle will decrease amount of ambient RNA.
- The impact of wash vs no wash protocol on ambient RNA needs to be validated.