Differentiated human stem cells resemble fetal, not adult, β cells Siniša Hrvatina, Charles W. O’Donnella,b, Francis Denga, Jeffrey R. Millmana, Felicia Walton Pagliucaa, Philip DiIorioc, Alireza Rezaniad, David K. Giffordb, and Douglas A. Meltona,e,1 aDepartment of Stem Cell and Regenerative Biology, Harvard Stem Cell Institute and eHoward Hughes Medical Institute, Harvard University, Cambridge, MA 02138; bComputer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA 02139; cDiabetes Center of Excellence, University of Massachusetts Medical School, Worcester, MA 01605; and dBetaLogics Venture, Janssen Research and Development, LLC, Raritan, NJ 08869 Contributed by Douglas A. Melton, January 14, 2014 (sent for review September 14, 2013) Human pluripotent stem cells (hPSCs) have the potential to Pancreatic β cells are a cell type responsible for the regulation generate any human cell type, and one widely recognized goal is of serum glucose levels through secretion of the hormone insulin. to make pancreatic β cells. To this end, comparisons between dif- Despite productive studies on new cell-surface markers for β ferentiated cell types produced in vitro and their in vivo counter- cells, insulin expression remains the best specific β-cell marker. parts are essential to validate hPSC-derived cells. Genome-wide β + Although adult cells express only the hormone insulin, it is transcriptional analysis of sorted insulin-expressing (INS ) cells de- important to note that some fetal β cells are polyhormonal and rived from three independent hPSC lines, human fetal pancreata, express other hormones, in addition to insulin, including gluca- and adult human islets points to two major conclusions: (i) Differ- + gon and somatostatin (17–19). Unlike glucose-responsive adult β ent hPSC lines produce highly similar INS cells and (ii) hPSC- + + cells, β cells present during early fetal development do not re- derived INS (hPSC-INS ) cells more closely resemble human fetal β cells than adult β cells. This study provides a direct comparison spond to glucose with increased insulin secretion (20, 21). + of transcriptional programs between pure hPSC-INS cells and true Cell line variability and the lack of a clear understanding of β cells and provides a catalog of genes whose manipulation may human early (fetal) development have hampered efforts to gen- + β BIOLOGY convert hPSC-INS cells into functional β cells. erate pancreatic cells from hPSCs. Current directed differen- + DEVELOPMENTAL tiation protocols generate insulin-expressing (INS ) cells (hPSC- + transcriptional profiling | differentiation | beta cells | MARIS INS ) that lack expression of several key β-cell genes and fail to properly secrete insulin in response to glucose (14, 22–32). uman pluripotent stem cells (hPSCs), including embryonic Additionally, careful analysis of select chromatin modifications Hstem cells (hESCs) and induced pluripotent stem cells that accompany β-cell development has shown that Polycomb- (hiPSCs), are characterized by their capacity for unlimited self- mediated differences in chromatin remodeling are deficient in in renewal and the ability to differentiate into any human cell type vitro differentiated cells (33). These observations led to the + (1–4). Stepwise differentiation protocols, designed to mimic se- speculation that hPSC-INS cells are more similar to fetal β cells quential developmental signals, attempt to generate specific cell than to adult β cells (27). It is, however, equally possible that + types from hPSC lines for use in transplantation therapy and hPSC-INS cells represent a different in vitro-derived cell type disease modeling (5–7). unlike fetal or adult β cells. A thorough analysis as to whether Significant variation in differentiation efficiencies has been hPSC-directed differentiation generates cell types found during observed between different hPSC lines, with some lines more normal human development has been long awaited. readily differentiating into a particular cell type than others – (8 11). The reasons for this variation have not been completely Significance explained, but studies point to variation in genetic, epigenetic (12), and cell cycle patterns (13). Owing to these differences in Human pluripotent stem cells (hPSCs) can be produced from differentiation propensity, directed differentiation protocols of- any person and have the potential to differentiate into any cell ten require laborious optimization for specific hPSC lines (14). type in the body. This study focuses on the generation of The use of different protocols and different cell lines calls into insulin-expressing cells from hPSCs and compares their gene question the degree to which the final differentiation products expression, as assayed by transcriptional gene products, to that resemble each other. This has been difficult to address because, of insulin-expressing β cells from human fetal and adult sam- even with an optimized protocol and cell line, only a fraction of ples. We employ a new method to isolate and profile insulin- the hPSCs achieve the desired cell fate. Thus, direct comparison expressing cells and conclude that several different hPSC lines of cells, which are present in a mixed population, is not generally generate very similar insulin-expressing cells, cells whose possible, except in those rare instances where appropriate re- transcripts resemble fetal rather than adult β cells. This study porter lines have been constructed to facilitate cell sorting. advances the possibility of directing the differentiation of stem In addition to questions of variability, the extent to which any cells into functional β cells by comparing and cataloging dif- differentiated cell produced in vitro resembles its counterpart ferences between hPSC-derived insulin-expressing cells and produced during normal human development remains unknown. human β cells. Directed differentiation protocols for human cells are often generated using mouse embryonic development as a guide. Al- Author contributions: S.H., C.W.O., F.D., D.K.G., and D.A.M. designed research; S.H., C.W.O., though there are many similarities, significant differences in F.D., J.R.M., and F.W.P. performed research; J.R.M., F.W.P., P.D., and A.R. contributed new reagents/analytic tools; S.H., C.W.O., F.D., D.K.G., and D.A.M. analyzed data; and S.H., C.W.O., transcriptional regulation exist between these two species (15, and D.A.M. wrote the paper. 16). Although it is obvious that documenting transcriptional Conflict of interest statement: A.R. is an employee of Janssen Research and Develop- changes that accompany human development would greatly ment, LLC. benefit directed differentiation of human stem cells, the lack of 1To whom correspondence should be addressed. E-mail: [email protected]. suitable cell surface markers makes it very difficult to isolate and This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. purify most human fetal and adult cell types for analysis. 1073/pnas.1400709111/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1400709111 PNAS Early Edition | 1of6 Downloaded by guest on September 25, 2021 Previous attempts to analyze genome-wide transcription in 100 iPS_ins.1 A C iPS_ins.2 100 enriched populations of adult β cells relied on FACS enrichment 100 H1_ins.1 β 95 H1_ins.2 of cells using either Newport Green dye (34) or a series of cell- 100 HUES8_ins.1 65 HUES8_ins.2 surface markers (35). The extent to which the small proportion 100 H1_S6.2 β 100 HUES8_S6.1 of non- cells present in the sorted population affects tran- 63 HUES8_S6.2 INS-APC H1_S6.1 72 scriptional analysis, and the applicability of these sorting meth- iPS_S6.1 ods to the isolation of human fetal β cells, is unknown. Sorting of 100 HUES8_S0.2 + 90 H1_S0.2 hPSC-INS cells also has been a significant challenge. Although 100 100 HUES8_S0.1 H1_S0.1 100 one hPSC insulin-GFP knock-in reporter line has been recently iPS_S0.2 + GCG/SST-FITC iPS_S0.1 generated (28, 29), isolating INS cells from multiple genetically 0.2 0.15 0.1 0.05 0.0 unmodified hPSC lines is necessary to evaluate the gene ex- B D Correlation value + + pression signature of hPSC-INS cells. Finally, to our knowledge Sorted ins cells * * * * no one has yet purified and transcriptionally profiled human ** fetal β cells. H1_S0.1 H1_S0.2 Hues8_S0.1 Hues8_S0.2 iPS_S0.1 iPS_S0.2 Hues8_S6.1 Hues8_S6.2 H1_S6.1 H1_S6.2 iPS_S6.1 Hues8_ins.1 Hues8_ins.2 H1_ins.1 H1_ins.2 iPS_ins.1 iPS_ins.2 H 1_S0.1 1 0.98 0.98 0.94 0.93 0.94 0.76 0.78 0.62 0.85 0.76 0.46 0.54 0.43 0.5 0.58 0.52 Here we make use of our newly developed Method for Ana- H1_S0.2 0.95 1 0.96 0.97 0.91 0.95 0.76 0.79 0.64 0.87 0.76 0.48 0.55 0.45 0.52 0.58 0.52 fold change Hues8_S0.1 0.98 0.96 1 0.95 0.94 0.94 0.75 0.77 0.62 0.84 0.75 0.46 0.53 0.43 0.5 0.57 0.51 lyzing RNA following Intracellular Sorting (MARIS) (36) to over unsorted S6 cells 2 Hues8_S0.2 0.94 0.97 0.95 1 0.91 0.94 0.77 0.8 0.66 0.87 0.78 0.51 0.58 0.48 0.55 0.6 0.55 analyze the global gene expression profile of three types of Log iPS_S0.1 0.93 0.91 0.94 0.91 1 0.92 0.78 0.82 0.68 0.86 0.83 0.54 0.61 0.53 0.59 0.67 0.61 + sorted INS cells: those differentiated from hPSC lines and hu- iPS_S0.2 0.94 0.95 0.94 0.94 0.92 1 0.8 0.82 0.7 0.88 0.81 0.56 0.63 0.53 0.6 0.65 0.6 Hues8_S6.1 0.76 0.76 0.75 0.77 0.78 0.8 1 0.96 0.88 0.9 0.87 0.77 0.82 0.72 0.77 0.79 0.76 man fetal and human adult pancreata.