Benefits of Single-Nucleus RNA Sequencing
Progress in biomedical research may be measured at some level by the shrinking nature and massive parallelism of experimentation. Instruments that once occupied an entire room are now handheld or sit on silicon or glass chips; milliliters have given way to microliters or picoliters, and two or three replicates have become 96, 384, or 1536 microwells. Consequently, the focus has shifted from organisms to tissues to cells and organelles, including nuclei. Underlying all these emerging possibilities have been greater adaptability of methods, generally, to the requirements of ‘omics disciplines, refinements in polymerase chain reaction (PCR) and related gene amplification methods, and lower-cost,, more-accessible next-generation sequencing (NGS).
Studying nuclei harvested from cultured cells isn’t new, as the isolation of organelles has been a staple in bioscience labs for decades. Similarly, performing PCR on single cells has been possible since the late 1990s.
But beginning around 2011, well into the age of NGS, papers began appearing on the sequencing of single nuclei. The first one located through a Google Scholar search was a 2011 study of tumor evolution in single nuclei, an analysis of gene targeting in moss was published later that year, followed by a report on copy number variants in tumors in 2012. The field rapidly developed momentum, to where in the past calendar year no fewer than 162 papers appeared, not on just any single-nucleus method, but specifically single-nucleus RNA-seq (snRNA-seq).
snRNAseq is a relatively new family of methods that analyze nuclei instead of intact cells. snRNAseq profiles gene expression in cells that are difficult to isolate, for example from archived tissue. The method uses one of several forms of droplet microfluidics, which enables high-throughput screening of individual cells in microdroplets, ranging in volume from 1 pL to 10 nL. Droplets are suspended in an immiscible oil. Droplet microfluidics essentially creates nanoscopic “test tubes” containing a single entity, a format that enables nucleus isolation from complex tissues.
Single-cell RNA sequencing methods are most appropriate in situations when cells cannot be harvested intact and viable, which is often true for preserved tissues, and always true for some cell types (e.g. neurons, adipocytes). These cells do not readily undergo successful dissociation since the enzymes and disruptive forces used to separate cells also affect the contents of other cellular compartments.
To prepare single nuclei, cells are lysed with detergent and homogenized using the Dounce mechanical homogenization method. Nuclei are purified using flow cytometry or gradient centrifugation.
Advantages of snRNAseq
Intuitively, working with cells should be easier than hunting down nuclei. Obtaining nuclei from cells involves at least two steps (lysis, centrifugation). However, the preparation process for single-cell studies often are a significant source of variability. One difficulty with single-cell genomics on solid tissues is obtaining single-cell suspensions of sufficient quality, particularly for rare cells, or those that are difficult to dissociate.
Tissues vary in the composition of extracellular matrix, which can carry over into single-cell preps. Tissues also differ in mechanical properties, requiring individualized optimization of dissociation protocols. Sample prep involves the usual steps of tissue harvest, grinding/homogenization, enzymatic breakdown, and for rare cells, enrichment. Each step affects the cells’ expression signatures, while the dissociation protocol often biases sample prep toward the isolation of one or more cell types, to the detriment of others that may be of greater interest.
Researchers and vendor companies have introduced tools and methods for minimizing dissociation bias in single-cell research. A group at the University of California, Irvine, has developed a microfluidic device that uses flow restriction and regions of high-shear force to create “hydrodynamic micro-scalpels” that separate cells from their matrix, rendering them suitable to single-cell analysis. Miltenyi Biotec sells a line of tissue dissociators and enzyme kits, thus combining mechanical disruption and enzymatic digestions for this purpose. A very recent advance uses cold protease dissociation, a method that requires no special skills or equipment. These techniques may work with some preserved tissues but certainly not with all.
Which is better: single cells, or nuclei?
It would be fair to ask how well the nuclear transcriptome represents that of the whole cell, and if not whether genes uncovered during snRNAseq are more or less relevant to the particular study than those in the cytoplasm. Studies comparing scRNAseq and snRNAseq show that transcripts are expressed equally in whole cells and nuclei.
A group at Washington University, St. Louis, recently compared single-cell sequencing using DropSeq to a single-nucleus method based on sNuc-DropSeq, a method that employs neither enzymatic dissociation nor nucleus sorting. The target tissues were fibrotic adult mouse kidney. Out of the more than 11,000 transcriptomes isolated with DropSeq investigators found that glomerular cells were absent, and one group of cells appeared to have experienced dissociation-induced stress. By contrast, the sRNAseq method uncovered the full diversity of kidney cell types that single-cell methods could not detect.
These researchers concluded that while comparable in cell type coverage, sRNAseq had the advantages of “reduced dissociation bias, compatibility with frozen samples, elimination of dissociation-induced transcriptional stress responses, and successful performance on inflamed fibrotic kidney.”
A definitive method
Quite often, after a research field reaches the proof of concept stage, a development comes along that defines it. For snRNAseq that technique was described in a seminal paper by Naomi Habib at MIT. DroNcSeq eliminates the significant issues with dissociation bias and has led to several methods for preparing and analyzing single nuclei. DroNc-Seq works with cells that are either fixed or that do not readily dissociate from their matrices. Based on DropSeq, DroNc-Seq is massively parallel modification that accounts for the smaller size and lower RNA content of nuclei compared with whole cells. DroNcSeq is the basis for the microfluidics-based Nadia instrument platform by Dolomite Bio.
DroNc-sequencing applies advances in single-cell methods to nuclei by encapsulating single cells and DNA-barcoded beads within microdroplets, then processes them microfluidically. According to Haib et al: “DroNc-seq opens the way to systematic single nucleus analysis of complex tissues that are inherently challenging to dissociate or already archived, helping create vital atlases of human tissues and clinical samples.”
scRNAseq has broadened the capabilities of single-cell research, particularly for preserved or frozen samples by uncovering cell types, states, genetic diversity, and interactions that were previously inaccessible in complex tumors. It overcomes issues with dissociation bias or, practically speaking, with coordinating sample analysis—usually performed in a laboratory— with collection in medical settings. snRNA-Seq also handles difficult-to-process fresh samples. The much lower mRNA content of nuclei compared with intact cells, matrix differences, and the overlap (or lack thereof) between nuclear and cellular RNA expression are issues that continue to be addressed, both at research laboratories and commercially.