Robust induction of primordial germ cells of white rhinoceros on the brink of extinction

Abstract

In vitro gametogenesis, the process of generating gametes from pluripotent cells in culture, is a powerful tool for improving our understanding of germ cell development and an alternative source of gametes. Here, we induced primordial germ cell-like cells (PGCLCs) from pluripotent stem cells of the northern white rhinoceros (NWR), a species for which only two females remain, and southern white rhinoceros (SWR), the closest species to the NWR. PGCLC differentiation from SWR embryonic stem cells is highly reliant on bone morphogenetic protein and WNT signals. Genetic analysis revealed that SRY-box transcription factor 17 (SOX17) is essential for SWR-PGCLC induction. Under the defined condition, NWR induced pluripotent stem cells differentiated into PGCLCs. We also identified cell surface markers, CD9 and Integrin subunit alpha 6 (ITGA6), that enabled us to isolate PGCLCs without genetic alteration in pluripotent stem cells. This study provides a first step toward the production of NWR gametes in culture and understanding of the basic mechanism of primordial germ cell specification in a large animal.

Fig. 1.. Induction of OGBT-positive cells from SWR-ESCs.

(A) OGBT reporter SWR-ESCs. Scale bar, 200 μm. BF, bright field; WT, wild type. (B) Pluripotent state of SWR-ESCs. The Venn diagrams identify genes up-regulated in the naïve or primed state in human and mouse pluripotent stem cells (52, 53). The heatmap shows correlation coefficients of gene expression representing the pluripotent state. n = 2, biologically independent experiments. mEpiSCs, mouse epiblast stem cells. (C) Time course of SWR-PGCLC induction. FN, fibronectin. (D) Effect of CHIR dosage during preinduction. OGBT SWR-ESC derivatives at day 4 of PGCLC induction are shown. The graphs show a summary of the percentage and number of OGBT-positive cells. n = 3, biologically independent experiments shown with SE. (E) Effect of BMP4 dosage during preinduction. OGBT SWR-ESC derivatives at day 4 of PGCLC induction are shown. n = 2, biologically independent experiments. (F) Morphology after preinduction with 6 μM CHIR. Scale bar, 200 μm. (G) Formation of BT-positive cells at day 4 of PGCLC induction. BT-positive cells (arrowheads) and an autofluorescence signal presumably derived from dead cells (asterisk) are shown. Scale bars, 200 μm. (H) Effect of IWR1 during PGCLC induction. The images and FACS analyses of OGBT SWR-ESC derivatives at day 4 of PGCLC induction. The box-and-whisker plots are a summary of the percentage and number of OGBT-positive cells. n = 5, biologically independent experiments. Scale bar, 200 μm. *P < 0.05 and **P < 0.01, Student’s t test. (I) Effect of the duration of preinduction. FACS plots show OGBT SWR-ESC derivatives at day 4 of PGCLC induction. n = 2, biologically independent experiments.

Fig. 2.. Gene expression profiling of OGBT-positive cells from SWR-ESCs.

(A) Gene expression dynamics during differentiation. Shown are averaged ΔC_t values with SE determined by qPCR analysis. n = 3, biologically triplicated samples. (B) Coexpression of transcription factors in OGBT-positive cells. Immunofluorescent analyses of the transcription factors involved in PGC specification are shown. Scale bar, 20 μm. (C) Unsupervised hierarchical clustering of derivatives from SWR-ESCs. (D) PCA of transcriptomes of SWR-ESC derivatives. (E) Scatterplot of the z scores of the genes for the PC1 and PC2. Genes with a radius of >3 SDs (566 genes) are shown. The colors of the plots correspond to the clusters shown in fig. S3B. (F) PCA of transcriptomes of ESC derivatives in humans, mice, and SWR. Note that the PCA using PC1 and PC3 showed a similar direction of differentiation. (G) Unsupervised hierarchical clustering. The colors correspond to those in (E). (H) Scatterplot of the z scores of the ortholog gene set involved in PGC specification. In the Venn diagram, the blue and orange circles include 456 human genes involved in PGC specification (29) and 395 SWR-PC1/2 genes whose orthologs exist in the gene list of (29), respectively. The bar graphs show the distribution of 395 SWR-PC1/2 genes or 116 genes commonly assigned to the gene clusters in the scatterplot of the z scores in (E). The right scatterplot shows the z scores of the 395 SWR-PC1/2 genes with the colored plot for the 116 genes. All results are based on biologically duplicated samples.

Fig. 3.. Characterization of OGBT-positive cells from SWR-ESCs.

(A) SOX17-dependent differentiation of OGBT-positive cells. Shown are images and FACS plots of wild-type and SOX17^−/− OGBT SWR-ESC derivatives at day 4 of PGCLC induction. n = 2, biologically independent experiments. Scale bar, 200 μm. (B) Restoration of OGBT-positive cell differentiation by exogenous SOX17 expression. Shown are images and FACS plots of SOX17-transgenic OGBT SWR-ESC derivatives at day 4 of PGCLC induction with or without Dox. n = 2, biologically independent experiments. Scale bar, 200 μm. (C) Comparison of transcriptomes of OGBT-positive cells induced by the cytokines or exogenous SOX17 expression. Shown are scatter plots of transcriptomes of OGBT-positive cells with DEGs (>4 times, FDR < 0.001, logCPM > 4). (D) Epigenetic features in OGBT-positive cells. Shown are the results of immunofluorescence analysis of the epigenetic modification indicated and SOX17 expression in OGBT SWR-ESC derivatives at day 4 of PGCLC induction by cytokines. SOX17-positive cells are enclosed by dashed lines. Scale bar, 20 μm. (E) Quantification of the epigenetic modifications in SOX17-positive cells. The box-and-whisker plots show the values relative to the averaged value in SOX17-negative cells. ***P < 0.001, Mann-Whitney U test.

Fig. 4.. Robust induction of SWR-PGCLCs.

(A) Refinement of basal medium for SWR-PGCLC induction. Shown are images and FACS plots of PGCLCs at day 4 of induction in the medium indicated. GK15, GMEM containing KSR at 15%; aRK5, aRK10, and aRK15, advanced RPMI 1640 containing KSR at 5, 10, and 15%, respectively. The right box-and-whisker plots summarize the percentage and number of PGCLCs per aggregate. n = 5, biologically independent experiments. Scale bar, 200 μm. *P < 0.05 and **P < 0.01, Turkey-Kramer test. (B) Effect of cyclic adenosine 5′-monophosphate activation during PGCLC induction. Shown are images and FACS plots of PGCLCs at day 4 of induction with forskolin and/or rolipram. The right box-and-whisker plots summarize the percentage and number of PGCLCs per aggregate. n = 5, biologically independent experiments. Scale bar, 200 μm. ***P < 0.001, Turkey-Kramer test. (C) Effect of bpV during PGCLC induction. Shown are images and FACS plots of PGCLCs at day 4 of induction with or without bpV. The right box-and-whisker plots summarize the percentage and number of PGCLCs per aggregate. n = 5, biologically independent experiments. Scale bar, 200 μm. *P < 0.05, Student’s t test. (D) Effect of the length of the PGCLC induction period. Shown are images and FACS plots of PGCLCs at the indicated days of induction. The PGCLC induction condition in this experiment was aRK10 + BLSEYF2i. The right box-and-whisker plots summarize the percentage and number of PGCLCs per aggregate. Scale bar, 200 μm. n = 7, biologically independent experiments.

Fig. 5.. Propagation of SWR-PGCLCs in culture.

(A) Effect of IWR1 and CHIR on the propagation of SWR-PGCLCs. The fold change (FC) of the SWR-PGCLC number, compared with the initialSWR-PGCLC number spread on the plate, is shown in each FACS plot. The plots at right are a summary of the percentage and FC of SWR-PGCLCs in the aggregates. n = 3, biologically independent experiments with SE. *P < 0.05, Tukey-Kramer test. (B) Effect of the inhibition of MAPK signaling pathways. (C) Effect of SB590885 (SB59). The plots at right are a summary of the percentage and FC of SWR-PGCLCs. n = 3, biologically independent experiments. *P < 0.05, Student’s t test. (D) FACS plot of the cell population during long-term culture. Days of analysis are shown on the top of each FACS plot. (E) Biexponential proliferation of SWR-PGCLCs. The plots show the FC increase of SWR-PGCLCs from culture day 0 and the days of culture indicated. (F) Comparison of transcriptomes of SWR-PGCLCs in the long-term culture. Shown are the scatter plots of transcriptomes of propagated SWR-PGCLCs. Colored plots show DEGs (>4 times, FDR < 0.001, logCPM > 4). (G) PCA and unsupervised hierarchical clustering of derivatives from SWR-ESCs, preinduced cells, and SWR-PGCLCs at the days of culture indicated. (H) Images of SWR-PGCLCs at 8 days of long-term culture. The bottom is a high-magnification image of the dotted box in the middle panel. Arrowheads indicate filopodia. Scale bars, 200 μm. (I) Tracking of SWR-PGCLC locomotion. Shown are images from the tracking of SWR-PGCLC (left) and SWR-ESC (right) locomotion for 12 hours. Scale bar, 100 μm.

Fig. 6.. Induction of PGCLCs from OGBT NWR-iPSCs.

(A) Subtle induction of PGCLCs from OGBT NWR-iPSCs under the aRK10 + BLSEYF2i condition. Shown are images and a FACS plot of PGCLCs at day 4 of induction under condition aRK10 + BLSEYF2i. n = 2, biologically independent experiments. Scale bar, 200 μm. (B) Optimization of PGCLC induction form NWR-iPSCs. Shown are images and FACS plots of PGCLCs at day 4 of induction without the factor indicated. Note that PGCLCs were efficiently induced under the PGCLC induction condition without BMP4. n = 2, biologically independent experiments. Scale bar, 200 μm. (C) Coexpression of transcription factors in NWR-PGCLCs. Shown are the results of immunofluorescent analysis of the transcription factors indicated. Note that SOX17 and TFAP2C are coexpressed in BT-positive cells. Scale bar, 20 μm. (D) Comparison of transcriptomes of SWR-PGCLCs and NWR-PGCLCs. Shown are scatterplots of the transcriptomes of SWR-PGCLCs and NWR-PGCLCs with DEGs (>4 times, FDR < 0.001, logCPM > 4). (E) PCA of the differentiation trajectory from NWR-iPSCs to NWR-PGCLCs. Note that the transcriptomes of derivatives of NWR-iPSCs were similar to the corresponding derivatives of SWR-ESCs. (F) Comparison of the expression dynamics of the genes involved in PGC specification between derivatives of NWR-iPSCs and SWR-ESCs. Shown are each value and the averaged values of gene expression based on transcriptome analyses using biologically duplicated samples.

Fig. 7.. Identification of surface marker proteins for isolation of PGCLCs.

(A) Specific expression of CD9 and ITGA6 in OGBT SWR-PGCLCs. Shown are FACS plots for CD9 and ITGA6 expression in OGBT SWR-ESCs and their derivatives at day 6 of induction (left). Note that 94% of the gated CD9 and ITGA6 double-positive cells were also OGBT-positive (right). APC;,Allophycocyanin. (B) Expression of CD9 and ITGA6 in derivatives from nonreporter SWR-ESCs. Shown are FACS plots for CD9 and ITGA6 expression in nonreporter SWR-ESCs and their derivatives at day 6 of induction. (C) Specific expression of CD9 and ITGA6 in OGBT NWR-PGCLCs. Shown are FACS plots of OGBT expression (left) and CD9 and ITGA6 expression (middle) in OGBT NWR-iPSC derivatives at day 4 of induction. Note that 88% of the gated CD9 and ITGA6 double-positive cells were also OGBT-positive (right). (D) Induction of CD9 and ITGA6 double-positive cells from nonreporter NWR-iPSCs. Shown are FACS plots and immunofluorescence analysis results for the expression of key transcription factors. 6A, 6B, 60R, and 60X are independent nonreporter NWR-iPSC lines. Note that SOX17-expressing cells also expressed TFAP2C and POU5F1. Scale bar, 20 μm. In addition, note that FACS plots and the percentages of CD9 and ITGA6 double-positive cells were varied among the cell lines. (E) Enrichment of NWR-PGCLCs in CD9 and ITGA6 double-positive cells. Shown are the percentages of SOX17, TFAP2C, and POU5F1 triple-positive cells among CD9 and ITGA6 double-positive cells induced from the indicated NWR-iPSCs. All results in this figure are based on experiments using biologically duplicated samples.