Single-Cell Analysis Applications

The diminishing cost of sequencing and increasing throughput is driving the rapid adoption of single-cell technologies. Biocompare recently hosted a Bench Tips webinar where senior postdoctoral fellows discussed how they were using various types of single-cell analysis and combining it with spatial techniques to better understand cell expression, location, and interactions. They shared their experiences and best practices, and this article summarizes their presentations and some of the questions posed.

Single-cell analysis is a great tool for identifying gene-expression pathways, novel cell types and cellular subtypes; tracing cell lineages, studying tissue heterogeneity; understanding developmental and pathological trajectory; and more. “This technology is very powerful, user-friendly, and can be used to study many aspects of biology,” explained Xuan Bao, M.D., Ph.D., Postdoctoral Associate in the Laboratory of Dr. Rui Chen in the Department of Molecular & Human Genetics at Baylor College of Medicine. The Chen lab is using single-cell RNA sequencing (scRNAseq) technology for establishing a cell atlas of the human eye across all ages and ethnicities, and to understand the genetics of eye disease. They are also using single-cell ATAC sequencing (scATACseq) to look at chromatin differences, and spatial transcriptomic profiling to study the differences in cell location.

Have the right sample?

The eye is an extremely complicated organ with many different parts, each part composed of different cell types with different functions. “To generate the cell atlas of the whole eye we have to analyze each part separately,” noted Bao. For single-cell RNA profiling the eyeballs received from donors in the eye bank have to be first dissected into various parts. “We have set certain criteria to achieve good sample quality,” she added. Good sample quality is guaranteed for almost all parts of the eye when the dissection is completed within six hours of death. Less than 12 hours is acceptable for some parts of the eye, but after 12 hours most parts have low cell viability as the RNAs start to degrade. “We snap freeze the samples in liquid nitrogen for single-nuclei RNA sequencing (snRNAseq) and for scRNAseq we usually digest them right away.”

It’s important to know all the factors that could lead to low sample quality and the cell types in the sample that are more vulnerable to dissociation and damage. According to Bao, for sc/snRNAseq you need fresh samples and the right enzymes for dissociation. The cells have to be fully dissociated and the cell viability must be checked. “We always need to check the cells under a microscope during dissociation,” she noted. “This will allow you to make sure that the cells are separated and not clumped together. Sometimes, cells need some cleaning or purification before they can be used.”

To ensure nuclei quality, you need to know the time for lysis, know the right buffer concentration, use RNAase inhibitor, and check the nuclear morphology. For single-nuclei ATAC sequencing (snATACseq), the nuclei can be isolated using a glass homogenizer or a commercial kit. “RNAase inhibitor is highly recommended to avoid RNA degradation,” Bao said. Fluorescence-activated cell sorting (FACS) can be used to clean up the nuclei but is usually not recommended for ATACseq. Before loading the samples, the quality of the nuclei should be checked under the microscope. Ideally the nucleus should be intact, and the nuclear membrane should be continuous. “If it's broken the contents will leak out and the experiment will fail.”

Which single-cell technology to use?

When the COVID-19 pandemic began, Dr. Craig Wilen’s lab at the Yale University School of Medicine started designing experiments to find out which cell types were preferentially infected by SARS-CoV-2, what the important determinants of cell tropism of SARS-CoV-2 in the lung were, and the inflammatory responses it elicited. “At that time, there were limited human samples available for setting up these experiments, so we used human bronchial epithelial cells (HBECs) from healthy and diseased donors, which were commercially available,” explained Mia Madel Alfajaro, DVM, Ph.D., Postdoctoral Associate. “We developed an air liquid interface (ALI) culture system and after 21–25 days when the cells had grown and differentiated we used them for experiments.” The cells were infected with SARS-CoV-2 and harvested at different time points and sent for scRNAseq. Alfajara spent a lot of time optimizing the protocols for cell culture and differentiation, which involved optimizing buffers, reagents, washing times, harvesting cycles, and dissociation steps to generate single cells.

According to Alfajaro, knowing what type of single-cell sequencing technology is suitable for your study is very important. “For the questions that we were trying to answer we chose 3’ scRNAseq as it allowed us to quantify the gene-expression changes and identify the different cell populations. Using the scRNASeq data, we were able to identify the cell type that was most affected by SARS-CoV-2 infection.” Careful handling and processing of cells for scRNAseq is necessary to preserve natural expression in cells and deliver meaningful data. Similarly knowing the sequencing depth you need to get the best results is also critical. “Single-cell experiments are very sensitive but also very expensive. Hence, it’s important that you know your experimental set up.”

Why use multiomics analysis?

Amin Abedini, M.D., Postdoctoral Fellow in the Laboratory of Dr. Katalin Susztak at the Perelman School of Medicine at University of Pennsylvania, uses both single-cell RNA and ATAC sequencing for understanding the pathophysiology of kidney fibrosis. He uses single cells derived from a mouse model of kidney fibrosis, as well as from patients with kidney fibrosis. Abedini explained that chronic kidney disease is common but poorly understood due to the complex 3D architecture of the kidney. Hence, to better understand the disease pathogenesis and to create the first spatially resolved human kidney cell atlas he needed three types of data—scRNAseq, snRNAseq, and snATACseq, along with spatial transcriptomics. “To create a comprehensive cell atlas we needed all three technologies as each one has its own advantages and disadvantages.”

scRNAseq is powerful to capture different types of immune cells, but not stromal or rare cells. snRNAseq is useful for identifying stromal cells and some types of rare cells, but not immune cells. snATACseq is similar to snRNAseq but it can also identify transcription factors. Both snRNAseq and snATACseq can be performed on frozen samples, while scRNAseq requires fresh samples. “However, these three methodologies do not provide any spatial context for the data,” added Abedini. Spatial transcriptomics is critical for providing a comprehensive spatial analysis, but currently it cannot provide data at the single-cell level. Downstream bioinformatic analysis needs to be done to get information at the single-cell level.

Combining and studying the data obtained from various single-cell and spatial techniques is what helped Abedini appreciate the cellular and architectural complexity of the processes involved in kidney fibrosis. Bioinformatic tools helped identify which cell types were injured in chronic kidney disease and the function of each cell type. “For most cells the spatial data coincides with known gene-expression signatures, and we were able to characterize the expression markers for cells that were only known spatially,” noted Abedini. “We were also able to identify which cells were close to each other. The cell-cell communication analysis showed the complex interaction between the various cells.” Although performing single-cell and single-nucleic RNAseq is possible and very useful, Bao cautioned that sometimes only one method works for certain tissues. Also, the transcriptomic profile data can be different for the cell and for the nuclei. Hence, it’s important to try both and see which one works best consistently.