Single-cell sequencing reveals the existence of fetal vascular endothelial stem cell-like cells in mouse liver

Abstract

Background: A resident vascular endothelial stem cell (VESC) population expressing CD157 and CD200 has been identified recently in the adult mouse. However, the origin of this population and how it develops has not been characterized, nor has it been determined whether VESC-like cells are present during the perinatal period. Here, we investigated the presence of perinatal VESC-like cells and their relationship with the adult VESC-like cell population.

Methods: We applied single-cell RNA sequencing of endothelial cells (ECs) from embryonic day (E) 14, E18, postnatal day (P) 7, P14, and week (W) 8 liver and investigated transcriptomic changes during liver EC development. We performed flow cytometry, immunofluorescence, colony formation assays, and transplantation assays to validate the presence of and to assess the function of CD157⁺ and CD200⁺ ECs in the perinatal period.

Results: We identified CD200^- expressing VESC-like cells in the perinatal period. These cells formed colonies in vitro and had high proliferative ability. The RNA velocity tool and transplantation assay results indicated that the projected fate of this population was toward adult VESC-like cells expressing CD157 and CD200 1 week after birth.

Conclusion: Our study provides a comprehensive atlas of liver EC development and documents VESC-like cell lineage commitment at single-cell resolution.

Keywords: Development; Endothelial stem cells; Heterogeneity; Mouse liver; RNA velocity; Single-cell transcriptomic sequencing; Trajectory analysis.

Fig. 1

Single-cell analysis of liver ECs across development. a Schematic of experimental design for the single-cell analysis pipeline. Livers from E14, E18, P7, P14, and W8 were dissected and dissociated into single cells. Sorted CD31⁺CD45⁻ ECs were collected and analyzed using the 10X Genomics Chromium platform. For each time point, one scRNA-seq library was generated using pooled tissues dissected from ± 6 individual animals. b UMAP visualization of ECs from all time points (n = 20.799 cells) colored and labeled based on cell clusters (left) and time point (right). c Dot plot of differentially expressed genes in each cluster. Gene expression in Log10 (TPM)

Fig. 2

CD157⁺CD200⁺ VESCs emerge from CD157⁻CD200⁺ ECs during the perinatal period. a UMAP projection of each EC clusters from E14 to W8. b Feature plot showing relative distribution and expression of CD157/Bst1 (top) and CD200 (bottom) for each time point. Each purple dot represents a single cell. Black arrows: CD200 + ECs; black arrowheads: CD157 + ECs. c Quantification of total ECs among all cells in liver. d FACS analysis of the liver from different time points showing the emergence of CD157⁺CD200⁺ among CD31⁺CD45⁻ ECs during early postnatal period. e, f Quantification of CD157⁺CD200⁺ ECs (e) and CD157⁻CD200⁺ ECs (f) among total ECs. g Immunofluorescence analysis of CD157 and CD200 in portal vein of E18, P7, and P21 liver. White arrowheads show the expression of CD157 within the PV. PV, portal vein. Data are shown as means ± SEM. Statistical analysis using unpaired two-tailed t-test, *** p < 0.0001 and **p < 0.005, ns p > 0.05

Fig. 3

Trajectory analysis reveals stem cell-like populations in perinatal micro- and macrovascular ECs. a UMAP visualization of ECs from this study and published data [10]. b Trajectory analysis of ECs from E12 until W8 projected onto the UMAP plot. c UMAP embedding of PV EC clusters colored and labeled by cell type. d Dot plot of known endothelial stem cell or progenitor marker expression in PV EC clusters. e Comparison of GO biological process enrichment in each cluster. f RNA velocity from E14 through P14 illustrating development of fetal VESCs at E14 and transition from fetal VESCs to adult VESCs at P7. g Model of EC development in the portal vein and hepatic artery. Arrows indicate the possible differentiating and reprogramming directions

Fig. 4

Perinatal CD157⁻CD200⁺ ECs and adult CD157⁺CD200⁺ ECs possess similar endothelial colony-forming ability. a Schematic overview. A total of 1000 liver ECs from fetal, neonatal, and adult liver were cultured on OP9 feeder cells for 10 days. b Cultured ECs from different time points were stained for CD31 (low-power view). The colony is defined as a cluster of > 50 ECs. There are no CD157 + CD200 + ECs at day E18. Experiments were repeated at least three times. c Quantification of the number of EC colonies formed by each fraction of ECs across development. d Schematic overview. Limiting dilution analysis of liver ECs from GFP mice cultured on OP9 feeder cells for 10 days. Experiments were repeated at least three times. e ECFCs frequency in each EC fraction calculated using the online algorithm, n = 3. Data are shown as means ± SEM. Statistical analysis using unpaired two-tailed t-test, ***p < 0.0001, **p < 0.005, and *p < 0.05, ns p > 0.05

Fig. 5

Perinatal CD157⁻CD200⁻ ECs and CD157⁻CD200⁺ ECs generate CD157⁺CD200⁺ ECs in vivo. a Schematic depicting the EC transplantation workflow. Experiments were repeated at least three times. b Representative images of recipient liver transplanted with perinatal CD157⁻CD200⁻ ECs and CD157⁻CD200⁺ ECs after 2 months. c FACS analysis of the recipient mouse liver showing the development of CD157⁺CD200⁺ ECs from perinatal CD157⁻CD200⁻ ECs and CD157⁻CD200⁺ ECs