Background

The human liver performs critical functions ranging from lipid metabolism, amino acid synthesis, and drug processing to protection from portal venous bacteria and viruses (Trefts et al., 2017). Consistent with this diverse range of functions, its spatial organization is specifically tailored (Ben-Moshe and Itzkovitz, 2019). For example, lobule zonation between a central vein and portal triad divides hepatocyte and endothelial functions (Halpern et al., 2017). Diverse and specialized cell types populate the liver (Aizarani et al., 2019), and thus studies involving the liver would benefit greatly from single-cell resolution.

The field of single-cell genomics has accelerated cellular composition studies of tissues and samples. Single-cell RNA sequencing (RNA-seq) has been used to discover subtypes within blood monocytes and dendritic cells (Villani et al., 2017), as well as to discover myeloid differentiation pathways (Drissen et al., 2016). In solid tissues, single-cell RNA-seq has provided insight into fibrosis mechanisms in lung (Xu et al., 2016) and kidney (Kuppe et al., 2021). Single-cell RNA-seq has also been used in liver to transcriptionally characterize the landscape of cell types (MacParland et al., 2018), as well as the immune landscape in hepatocellular carcinoma (HCC) (Zhang et al., 2019).

One adaptation of single-cell RNA-seq is the use of nuclei instead of whole cells. Single-nucleus RNA-seq (snRNA-seq) has been shown to produce similar results as scRNA-seq (Habib et al., 2017). Furthermore, it can identify a greater expression diversity within cell types than scRNA-seq (Andrews et al., 2022). A main advantage of snRNA-seq is its applicability to frozen tissues. Archived tissues remove the need to coordinate tissue processing immediately after biopsy collection, which is challenging for human tissues. Archived tissues may also have additional phenotype and molecular data readily available, allowing for more thorough analyses and controlling of confounders.

However, the isolation of nuclei from frozen tissues for snRNA-seq presents distinct challenges. Snap freezing can result in the formation of ice crystals, disrupting the tissue and possibly decreasing the yield of intact nuclei (Larson and Chin, 2021). This can also result in higher amounts of tissue debris and contaminating ambient RNA in the lysate and nucleus suspension. We present a specifically tailored protocol for snRNA-seq of human frozen liver tissue (Rao et al., 2021), balancing requirements for yield and debris removal. For an overview of the protocol, see Figure 1. The advantage of our protocol is the rapid isolation of nuclei with minimum equipment, which prevents RNA degradation and increases throughput. Our protocol avoids labor-intensive flow cytometry sorting, which may require high yields and decrease throughput. We also avoid density-based centrifugation, which increases run times and susceptibility to RNA degradation. In addition, we provide details on how to isolate nuclei suitable for droplet-based microfluidics applications, as well as data analysis guidelines. While optimized for snap-frozen liver tissue, this protocol can be adapted to other frozen tissue samples, such as adipose tissue (Alvarez et al., 2020).

Figure 1. Overview of nucleus isolation from frozen liver. The flowchart outlines the steps from tissue lysis to nuclear suspension.