You are here

| Pluripotent Stem Cells

BMP-ing Up Endoderm Efficiency – Activin and BMP4 Synergistically Promote Formation of Definitive Endoderm in hESCs

From Stem Cells 
By Stuart P. Atkinson

Protocols for the production of hepatocytes and insulin-secreting pancreatic beta cells from human embryonic stem cells (hESCs) for treatment of disease and injury, begins with the production of definitive endoderm (DE), but while factors such as Activin A, which activates SMAD2/3, FGF, WNT and BMP (Arnold and Robertson, and Tam and Loebel) are thought to be important for DE differentiation, further differentiation of these cells into pancreatic progenitors, for example, is not efficient. This suggested to the group of N. Ray Dunn atA*STAR, Singapore, that if initial DE differentiation protocols could be made more efficient, this could allow for subsequent high efficiency differentiation of DE-derivatives (Teo and Aliet al).

Previous studies by the same group (Phillips et al) had described a moderately efficient DE differentiation protocol from hESC, but this did not allow for the efficient production of PDX1+ pancreatic progenitors. As the conditions used (embryoid body growth in 20% KSR and growth factor-reduced Matrigel) subjected hESCs to multiple and possibly opposing growth factor signals, a chemically defined culture medium was sought. Experimentation with several different growth substrates suggested that the most efficient means of differentiating DE from hESCs was through feeder depletion and replating of cell clumps on fibronectin-coated dishes in RPMI medium containing 2% B-27 (serum-free, chemically defined growth supplement (Brewer et al) and recombinant Activin A. This was confirmed through Q-PCR analysis which demonstrated that by day 6 of this protocol nearly 74% of cells were CXCR4+/SOX17+ DE cells. As the previous study indicated an important role for BMP4 signalling during DE differentiation, this was further studied using the new protocol, finding that irrespective of the Activin A concentration, addition of BMP4 resulted in a greater decrease in OCT4+ cells and elevation of SOX17 levels. Addition of 50ng/ml of Activin A and BMP4 led to the maximal suppression of OCT4 and near-peak expression of SOX17. OCT4 levels are minimal at day 1 when both these factors are added at day 0, while SOX17 and FOXA2 levels reach maximal levels under similar conditions. These conditions also led to the rapid upregulation of genes characteristically expressed in the mouse primitive streak (PS) and early DE progenitors such as NODAL, EOMESODERMIN, MIXL1, GOOSECOID and CERBERUS. 86% of cells were CXCR4+ SOX17+ DE by day 6, as found by FACS analysis and only 15.9% of cells were OCT4+ by three days.

To understand the role of BMP4, the BMP-binding protein Noggin was added on day 0 of differentiation in the presence of 50 ng/ml Activin A and absence of BMP4 to understand the role of endogenous BMP4 which was shown to be readily expressed in feeder-depleted undifferentiated hESC, along side BMP2 and their activated effector proteins phospho-Smad1/5/8. At day 3 of the differentiation process, the addition of Noggin led to higher levels of pluripotency-associated gene expression confirming previous observations that BMP proteins are pro-differentiation factors, although BMP signalling has not been previously indicated to have a role in DE formation. Differentiation in the presence of Activin A and BMP4 led to increase in BMP2, BMP4 and phospho-Smad1/5/ levels compared to Activin A alone; suggesting that robust DE formation in the presence of Activin A and BMP4 s promoted by continuous BMP signaling. Comparison of gene expression in DE cells derived under conditions of Activin A and BMP4 vs. Activin A only, found that genes involved in formation and patterning of the mouse PS (EOMES, MIXL1, BRA, FGF8, MESP1, WNT3, CER1, BMP2, BMPR2, LHX1, and FGFR4) and DE-specific genes (CXCR4, SOX17, FOXA3, and GATA4) were significantly upregulated in AB, while pluripotency associated genes were downregulated. The top 40 genes upregulated in Activin A and BMP4 treated cells compared to Activin A treated cells included the endodermal marker SOX17 but also ID1, PRDM1, and SMAD6, which are known BMP target genes and thus serve as an internal control for BMP stimulation (Moustakas and Heldin, and  Ohinata et al), alongside other less well-characterised genes (GRP, APLNR, PPAPDC1A, MFAP4, and PRSS35). Q-PCR analysis confirmed their relative over-expression and also demonstrated that these genes are expressed in the mouse embryo to varying degrees on embryonic day 7.5, a stage at which embryonic endoderm genes like Cxcr4, Sox17, and Foxa2 are robustly activated (Arnold and Robertson, and Tam and Loebel). Of the top 40 downregulated genes were OCT4, NANOG, HOXA5 and TXNIP which all showed greater downregulation in the Activin A and BMP4 treated cells compared to Activin A treated cells only.

Finally, the formation of PDX1+ cells was analysed, through the assessment of the ability of the new protocol to allow for further differentiation of specialized endodermal derivatives. Activin A and BMP4 treated DE cells were  treated with All-trans retinoic acid, FGF and Nicotinamide, well know pancreatic inducers (Mfopou et al, Sulzbacher et al, Vaca et al and Van Hoof et al), and at day 12 of the treatment, PDX1 expression was 7 times greater in the Activin A and BMP4 treated cells compared to Activin A treatment alone, and 4 times greater at 17 days. At 17 days, only rare OCT4+ cells existed and PDX1+ cells were also FOXA2 and SOX9 positive suggesting that these cells, which also form a honeycomb network of tubular cells, were mitotically active (Pan and Wright).

Overall this data suggests a complex synergism between Activin A and BMP4 in the differentiation of hESCs into DE cells which allows for the efficient production of a therapeutically relevant cell type. The authors note that future research will hopefully more fully elucidate how BMP and Activin/Nodal signalling synergise, through genome-wide occupancy studies of the SMAD1/5/8 effectors in hESCs, which will also aid our understanding of normal human development.



Arnold SJ, Robertson EJ.
Making a commitment: cell lineage allocation and axis patterning in the early mouse embryo.
Nat Rev Mol Cell Biol. 2009;10:91-103.

Brewer GJ, Torricelli JR, Evege EK, et al.
Optimized survival of hippocampal neurons in B27- supplemented Neurobasal, a new serum-free medium combination.
J Neurosci Res. 1993;35:567-576.

Mfopou JK, Chen B, Sui L, et al.
Recent advances and prospects in the differentiation of pancreatic cells from human embryonic stem cells.
Diabetes. 2010;59:2094-2101.

Moustakas A, Heldin CH.
The regulation of TGFbeta signal transduction.
Development. 2009;136:3699- 3714.

Phillips BW, Hentze H, Rust WL, et al.
Directed differentiation of human embryonic stem cells into the pancreatic endocrine lineage.
Stem Cells Dev. 2007;16:561-578.

Pan FC, Wright C.
Pancreas organogenesis: from bud to plexus to gland.
Dev Dyn.240:530-565.

Tam PP, Loebel DA.
Gene function in mouse embryogenesis: get set for gastrulation.
Nat Rev Genet. 2007;8:368-381.

Ohinata Y, Ohta H, Shigeta M, et al.
A signaling principle for the specification of the germ cell lineage in mice.
Cell. 2009;137:571-584.

Sulzbacher S, Schroeder IS, Truong TT, et al.
Activin A-Induced Differentiation of Embryonic Stem Cells into Endoderm and Pancreatic Progenitors-The Influence of Differentiation Factors and Culture Conditions.
Stem Cell Rev. 2009;5:159-173.

Vaca P, Berna G, Araujo R, et al.
Nicotinamide induces differentiation of embryonic stem cells into insulin-secreting cells.
Exp Cell Res. 2008;314:969- 974.

Van Hoof D, D'Amour KA, German MS.
Derivation of insulin-producing cells from human embryonic stem cells.
Stem Cell Res. 2009;3:73-87.