Simultaneous isolation of parenchymal and non-parenchymal cells from healthy and diseased mice liver

Background and Aim:

Liver structure is highly complex consisting of two principal cellular compartments: parenchymal cells (PCs) and non-parenchymal cells (NPCs). PCs include hepatocytes (HCs) and bile duct epithelial cells, while NPCs comprise Kupffer cells (KCs), liver sinusoidal endothelial cells (LSECs), and hepatic stellate cells (HSCs). The separation of these liver cells in high yield, optimal purity and viability from healthy and diseased mouse liver is a prerequisite to investigate their role in normal and pathological conditions in subcellular and molecular detail. Traditional isolation methods usually purify mostly single cell types to obtain high yield for subsequent analyses. Recently, we identified CD271 as a promising cell surface marker for isolation of HSCs from mouse liver using magnetic activated cell sorting (MACS). Based on this finding, we combined CD271 with other well-known NPCs markers to simultaneously purify different liver cells from healthy and diseased livers upon magnetic beads labelling.

Methods:

Livers of adult wildtype mice (Balb/C or C57Bl/6N) were digested by two step EGTA and collagenase II perfusion. HCs were separated by low-speed centrifugation and further purified with a percoll gradient. NPCs were purified with AutoMACS using CD271, CD11b, and CD146 for HSCs, KCs, and LSECs, respectively. All antibody-labelled microbeads were purchased from Miltenyi Biotec GmbH, DE. Carbon tetrachloride (CCl4)-treated C57Bl/6N mice were included as an example of a diseased liver. Purity of cells was confirmed by co-immunofluorescence (Co-IF), immunoblot, qPCR, and FACS.

Results:

Using the current protocol, we successfully isolate parenchymal (PCs) and non-parenchymal cells (NPCs) from healthy mouse liver. The yield of HC, HSCs, KCs, and LSECs are ≈30 × 106, ≈8 × 105, ≈1.5 × 106, and ≈2 × 106 cells per mouse liver. Viability was ≥85% as shown by trypan-blue staining. Purity of isolated cells is ≥90%, as confirmed by Co-IF analysis for HC, HSCs, KCs, and LSECs, or additionally with FACS, qPCR, and UV fluorescence for HSCs. The same protocol was successfully applied to isolate all four cell types from livers three days after CCl4 administration. Interestingly, KC number is 2.5 fold higher in CCl4-intoxicated mice as compared to the corresponding C57Bl/6N control. In addition, the phenotype of HSCs from CCl4-intoxicated presents as activated as compared to non-treated mice.

Conclusions:

The current protocol represents a well-reproducible, automated, and flexible method to isolate PCs and NPCs with high yield and purity from the same healthy or diseased liver. Currently, the protocol is tested to isolate PCs and NPCs from other animal models of liver disease (e.g. Abcb4-KO mice) and from human patient livers.