A comparative study of endoderm differentiation in humans and chimpanzees

Abstract

Background: There is substantial interest in the evolutionary forces that shaped the regulatory framework in early human development. Progress in this area has been slow because it is difficult to obtain relevant biological samples. Induced pluripotent stem cells (iPSCs) may provide the ability to establish in vitro models of early human and non-human primate developmental stages.

Results: Using matched iPSC panels from humans and chimpanzees, we comparatively characterize gene regulatory changes through a four-day time course differentiation of iPSCs into primary streak, endoderm progenitors, and definitive endoderm. As might be expected, we find that differentiation stage is the major driver of variation in gene expression levels, followed by species. We identify thousands of differentially expressed genes between humans and chimpanzees in each differentiation stage. Yet, when we consider gene-specific dynamic regulatory trajectories throughout the time course, we find that at least 75% of genes, including nearly all known endoderm developmental markers, have similar trajectories in the two species. Interestingly, we observe a marked reduction of both intra- and inter-species variation in gene expression levels in primitive streak samples compared to the iPSCs, with a recovery of regulatory variation in endoderm progenitors.

Conclusions: The reduction of variation in gene expression levels at a specific developmental stage, paired with overall high degree of conservation of temporal gene regulation, is consistent with the dynamics of a conserved developmental process.

Keywords: Comparative genomics; Functional genomics; Gene expression.

Fig. 1

Study design and quality control analyses. a Study design. Samples from four chimpanzees and six humans were studied at four time points during endoderm development. We included two technical replicates from each of the chimpanzees and two technical replicates for two of the six humans. iPSC induced pluripotent stem cell, PS primitive streak, EP endoderm progenitor, DE definitive endoderm. b Purity analysis. Cell type composition at each day based on FACS analysis (see “Methods”), estimated by k-means clustering. c Heat map of normalized log₂(CPM) as a measure of expression levels of TFs that are known to be highly expressed in one or more stages in the differentiation to endoderm [7]. Generally, samples from the same day, regardless of species, cluster together

Fig. 2

General patterns in the data. a Normalized log₂(CPM) expression measurements for all genes projected onto the axes of the first two PCs. Color indicates day. Shape represents species. PC1 is highly correlated with differentiation day (r = 0.92). PC2 is highly correlated with species (r = 0.93). b Box plots of normalized expression values for genes with known roles in endoderm development

Fig. 3

Number of DE genes in pairwise analyses. Venn diagrams of a DE genes at each day, b DE genes between consecutive time points in humans, c DE genes between consecutive time points in chimpanzees, d genes with a significant species–time point interaction effect at each day (DE was classified at FDR of 5% in all cases)

Fig. 4

High sharing of DE genes across species. A circos diagram with the number of shared, human-specific, and chimpanzee-specific DE genes across time points. There is a high degree of sharing of DE genes (yellow ribbon), particularly from day 0 to 1 and day 1 to 2

Fig. 5

Gene expression motifs. Correlation motifs based on the probability of differential expression across days for each species with the number of genes assigned to each correlation motif. The shading of each box represents the posterior probability that a gene is DE between two time points in a given species. Each row (“correlation motif”) represents the most prevalent expression patterns. Out of 10,304 genes, 8004 were assigned to one correlation motif in this model

Fig. 6

Global reduction of variation in gene expression from the iPSCs to primitive streak state. Box plot of the log₂ variance of expression levels for each gene. Variation in gene expression levels are significantly reduced from iPSCs to primitive streak (P < 10⁻¹⁵ in both species) but not in subsequent time points (P > 0.5 in both species)

Fig. 7

Conserved patterns of reduced variation in gene expression at primitive streak. We plotted the P value distributions of F tests of the null hypothesis that there is no reduction in variation in gene expression levels as samples progress along the time course in human (a) and chimpanzee (b) samples. $\widehat{\mathrm{\pi }}$₀ is the estimated proportion of null tests in each distribution. In the next four panels, we plotted the P value distribution for the same test, but included only genes whose variation was classified as reduced between states in the other species; thus in (c) we plotted P value distributions of F tests in chimpanzees only for genes who variation was classified as reduced (P < 0.05) in humans and in (d) we did the reverse. In (e) (chimpanzee conditional on human) and (f) (human conditional on chimpanzee), we plotted the P value distributions of F tests of the null hypothesis that there is increase in variation in gene expression levels as the samples progress from day 1 to 2