Distinct gene expression program dynamics during erythropoiesis from human induced pluripotent stem cells compared with adult and cord blood progenitors

Abstract

Background: Human-induced pluripotent stem cells (hiPSCs) are a potentially invaluable resource for regenerative medicine, including the in vitro manufacture of blood products. HiPSC-derived red blood cells are an attractive therapeutic option in hematology, yet exhibit unexplained proliferation and enucleation defects that presently preclude such applications. We hypothesised that substantial differential regulation of gene expression during erythroid development accounts for these important differences between hiPSC-derived cells and those from adult or cord-blood progenitors. We thus cultured erythroblasts from each source for transcriptomic analysis to investigate differential gene expression underlying these functional defects.

Results: Our high resolution transcriptional view of definitive erythropoiesis captures the regulation of genes relevant to cell-cycle control and confers statistical power to deploy novel bioinformatics methods. Whilst the dynamics of erythroid program elaboration from adult and cord blood progenitors were very similar, the emerging erythroid transcriptome in hiPSCs revealed radically different program elaboration compared to adult and cord blood cells. We explored the function of differentially expressed genes in hiPSC-specific clusters defined by our novel tunable clustering algorithms (SMART and Bi-CoPaM). HiPSCs show reduced expression of c-KIT and key erythroid transcription factors SOX6, MYB and BCL11A, strong HBZ-induction, and aberrant expression of genes involved in protein degradation, lysosomal clearance and cell-cycle regulation.

Conclusions: Together, these data suggest that hiPSC-derived cells may be specified to a primitive erythroid fate, and implies that definitive specification may more accurately reflect adult development. We have therefore identified, for the first time, distinct gene expression dynamics during erythroblast differentiation from hiPSCs which may cause reduced proliferation and enucleation of hiPSC-derived erythroid cells. The data suggest several mechanistic defects which may partially explain the observed aberrant erythroid differentiation from hiPSCs.

Keywords: Erythropoiesis; SMART and Bi-CoPaM; Transcriptome; hiPSC.

Fig. 1

Gene expression during erythroid differentiation from adult stem cells in SEM-F. a PCA of differential gene expression in the triplicate AB FBS samples transforms the data into a series of uncorrelated variables made up from linear combinations and shows, in an unsupervised analysis, the progression of the differentiating erythroid cells through gene expression state-space. Genes reaching a minimum linear expression value of 100 in all replicates of at least one sample group were selected as differentially-expressed (DE) between any two stages during erythroid differentiation if they met the following criteria: p ≤ 0.01, fold change (FC) ≥ 2, B > 2.945 (Additional file 2: Table S1A). The union of all DE genes was used in the PCA. The Euclidean distances relating to this PCA are available in Additional file 19: Table S6. b Hierarchical clustering analysis of the differentially-expressed genes used in (A), clustering by gene only according to Euclidean distance. The colour bar on the left hand side denotes clusters of co-regulated genes. c Plot showing the number of differentially-expressed genes between consecutive populations in the adult FBS time-course data. d Genes described to be preferentially expressed in primary human erythroid cells [32] were examined in our current adult SEM-F dataset. Where these genes were expressed in our data, their expression pattern is shown as a hierarchical clustering, clustered by gene

Fig. 2

Gene expression during erythroid differentiation from adult and cord blood stem cells. a PCA of DE gene expression in cord blood- (CB-) derived differentiations in SEM-F. Genes were selected if they were DE between any two AB populations, or between any two CB populations (Additional file 5: Table S3A). The union of these two DE gene sets were then used to arrange the samples in the PCA, clustering the CB-erythroblasts together with the AB-erythroblasts shown in Fig. 1. CB-erythroblast populations were isolated with the same gating strategy used for AB-erythroblasts. The Euclidean distances relating to this PCA are available in Additional file 19: Table S7. b Hierarchical clustering analysis of the same AB or CB DE gene set, clustering by Euclidean distance and by gene. The colour bar on the left hand side denotes clusters of co-regulated genes. c The number of DE transcripts between AB-erythroblasts and the same CB-erythroblast population is shown to examine the fold change at each point during erythropoiesis. Blue bars depict genes that are expressed more abundantly in CB-erythroblasts, and red, in AB-erythroblasts. d Mean expression values of selected key erythroid genes during erythropoiesis in SEM-F, +/- standard error of the mean. ACTB and PAFAH1B2 were used to normalise the data since they were consistently expressed throughout erythropoiesis in this data (Additional file 4: Tables S2)

Fig. 3

The gene expression profile of hiPSC derived erythroblasts is independent of the media used as evidenced by comparison of erythroid differentiation in SEM-i with SEM-F. a Hierarchical clustering analysis by Euclidean distance of adult erythroid differentiations performed using media for adult derived HSPCs (SEM-F), or hiPSC conditions (SEM-i). The gene set used contained genes which were DE during maturation in any of the two media settings (Additional file 20: Table S4). The colour bar on the left hand side denotes clusters of co-regulated genes. Samples cluster together by time and by the immunophenotype of the developing erythroid cells. There are no coherent subclusters formed according to the type of media used. b PCA of the samples shown in (A) with the same gene set. Populations are represented as follows: SEM-F, red symbols; SEM-i, blue symbols; d0 CD34⁺, black circles; day 4 CD36⁺CD71⁺CD235a⁻, triangles; day 7 CD36⁺CD71⁺CD235a⁻, thin diamond; day 7 CD36⁺CD71⁺CD235a⁺, square; day 7 CD71⁺ beads, fat diamond; day 14 CD71⁺CD235a⁺, pentagon; day 14 CD235a⁺ beads, hexagon. c Proliferation in SEM-i of erythroid AB-erythroblasts (black triangles), CB-erythroblasts (green squares) and hiPSC-erythroblasts (blue) where hiPSCs were specified from CD34⁺ peripheral blood (squares), erythroid cells (triangles) or fibroblasts (circles). Error bars indicate standard error of the mean of 3 or more cultures. d Morphological changes observed in erythroblasts of hiPSC and AB origin cultured in SEM-i. Representative images of Giemsa-benzidine stained cytospins of cultures on day 7, day 14 and day 20. Scale bar is ~10 μm. “m” is the stromal cell line MS-5. AB-derived differentiations, cultured further until day 20/21, were typically 70-80 % enucleated (see also Additional file 21: Figure S12), whereas the hiPSC-erythroblast cultures failed to enucleate

Fig. 4

Elaboration of functional gene clusters in erythroid cells derived from adult and hiPSC progenitors in SEM-i. a PCA of genes differentially expressed in AB-erythroblasts (red symbols) or hiPSC-erythroblasts (blue symbols) during erythroid maturation in SEM-i. The gene set was any gene DE during maturation of erythroblasts from either source. The Euclidean distances relating to this PCA are available in Additional file 19: Table S8. b Expression profiles of globin genes in AB-erythroblasts and hiPSC-erythroblasts. Mean expression +/- standard error of the mean is plotted. c Expression profiles of genes encoding proteins with key roles in the regulation of erythroid development in the adult and hiPSC-derived settings. Mean expression +/- standard error of the mean is plotted. d Hierarchical clustering analysis by Euclidean distance of AB-erythroblasts and hiPSC-erythroblasts, clustered by gene only. The colour bar on the left hand side denotes clusters of co-regulated genes. e Gene expression profiles for c-KIT and NFIA in AB-erythroblasts and hiPSC-erythroblasts. Mean expression +/- standard error of the mean is plotted. Given these findings, we validated the microarray-based gene expression measurements of selected transcripts using quantitative RT-PCR (in Additional file 13: Figure S10B)

Fig. 5

Transduction of hiPSC hematopoietic progenitors with lentivirus expressing c-KIT enhances erythroblast expansion. a Proliferation of hiPSC-erythroblasts transduced with either the GFP tagged lentiviral construct (GFP c-KIT) or with the control vector (GFP) was monitored after selection of GFP positive cells by FACS on day 2 of hiPSC erythroblast culture. Results from 2 independent experiments are shown in (i) and (ii). b Increased expression of c-KIT in erythroblasts grown from hiPSC progenitors transduced with the GFP c-KIT lentivral construct. Error bars are SDs of quadruplicate measurements by semi-quantitative PCR from cells taken on day 10 of culture showing a 1.5 fold increase that is comparable to the increase in cell expansion. c Protein expression of c-KIT at the surface of erythroblasts grown from hIPSC progenitors transduced with the GFP-c-KIT or control vector expressing GFP alone. GFP-c-KIT erythroblasts collected at the end of experiment (i) show a small increase in c-KIT expression where mean fluorescence intensity (MFI) is 47 compared with control MFI of 37 and the proportion of c-KIT positive cells is also modestly increased (highlighted in blue)

Fig. 6

Differential expression of co-ordinately regulated genes relevant to erythroid expansion and terminal differentiation identified using SMART. a Standard hierarchical clustering analysis by Euclidean distance of the union of DE genes from all cells of origin, in all media, clustered by gene only, of all samples utilised in the study, regardless of medium type or cell-of-origin, to visualise clusters of robustly co-regulated genes. The colour bar on the left hand side denotes clusters of co-regulated genes. b Transcription factor binding site analysis of 1kb of genomic DNA sequence upstream of the TSS for genes in the DNA repair/cell-cycle cluster. Matches from MEME/TOMTOM analysis are depicted: upper panels are motifs from the database; lower panels are the enriched motif detected within cluster 20 from our SMART analysis. Vertical axes are scaled to 2 bits in all images, and horizontal axes show sequential bases in the relevant motifs. c Examples of genes in SMART cluster 1, important in autophagic processes, which are differentially expressed between AB-erythroblasts and hiPSC-erythroblasts during the last phases of differentiation