Differential BMP signaling controls formation and differentiation of multipotent preplacodal ectoderm progenitors from human embryonic stem cells

Abstract

Sensory and endoneurocrine tissues as diverse as the lens, the olfactory epithelium, the inner ear, the cranial sensory ganglia, and the anterior pituitary arise from a common pool of progenitors in the preplacodal ectoderm (PPE). Around late gastrulation, the PPE forms at the border surrounding the anterior neural plate, and expresses a unique set of evolutionarily conserved transcription regulators including Six1, Eya 1 and Eya2. Here, we describe the first report to generate and characterize the SIX1(+) PPE cells from human embryonic stem (ES) cells by adherent differentiation. Before forming PPE cells, differentiating cultures first expressed the non-neural ectoderm specific transcriptional factors TFAP2A, GATA2, GATA3, DLX3, and DLX5, which are crucial in establishing the PPE competence. We demonstrated that bone morphogenetic protein (BMP) activity plays a transient but essential role in inducing expression of these PPE competence factors and eventually the PPE cells. Interestingly, we found that attenuating BMP signaling after establishing the competence state induces anterior placode precursors. By manipulating BMP and hedgehog signaling pathways, we further differentiate these precursors into restricted lineages including the lens placode and the oral ectoderm (pituitary precursor) cells. Finally, we also show that sensory neurons can be generated from human PPE cells, demonstrating the multipotency of the human ES-derived PPE cells.

Figure 1. Derivation of preplacodal and non-neural ectoderm from human ES cells

(A) Semi-quantitative RT-PCR of undifferentiated human ES cells (ES) and differentiation day-7 cultures under SF, neural (NSB) and epidermal (Epi) differentiation conditions. (B) Histogram presentation of the percentage of TFAP2A⁺ and GATA3⁺ cells in human ES cells cultured in the SF medium at day 8. Error bars are standard errors. (C–G) Immunofluorescence analyses of adherent human ES cells cultured in the SF medium (C–F) and neural differentiation medium (NSB)(G) on day 8 with the antibodies indicated. Arrows indicate cells negative for both SIX1 and TFAP2A; arrowheads show SIX1⁻/TFAP2A⁺ and SIX1⁺/GATA3⁻ cells.

Figure 2. Cell density affects the differentiation of distinct ectodermal linages in adherent human ES cell culture

(A) Schematic diagram for generating placodal ectoderm cells from human ES cells, and the change of ectodermal differentiation in relation to the increase of cell density. (B) Histogram presentation of the percentage of SIX1⁺ cells on differentiation day 8 cultures started with different seeding density or different pre-differentiation periods as indicated in (A). Asterisks indicate significant differences (p < 0.05, ANOVA test). (C and D) Immunofluorescence of day-8 cultures with 3.4 × 10⁴ cells/cm² seeding density and 3-day pre-differentiation. Antibodies are indicated on the top. Asterisks indicate cells lacking ECAD, SIX1, and TFAP2A; dashed lines demarcate area containing PAX7⁺ cells. (E) Semi-quantitative RT-PCR of day-8 cultures with the indicated seeding densities and a 3-day pre-differentiation period.

Figure 3. Development of the PPE competence and definitive PPE identity during human ES cell differentiation

(A) Semi-quantitative PCR of human ES cells cultured under SF condition at day (D) 0, 3, 5, and 8. (B and C) Histogram of relative expression of PPE competence and definitive PPE transcripts using qPCR. Expression data were presented in log scales with three independent biological replicates. Asterisks indicate significant differences (p < 0.05, ANOVA test) from that of day 0 cultures. Note that DACH1 expression was significantly down-regulated from day 0 to day 3 before increasing until day 8. (D–G) Immunofluorescence analyses on day (D) 0, 3, and 5 with the indicated antibodies. Note the progressive decrease of OCT4 and increase of TFAP2A, GATA3 and DLX5 signals (green) from day 0 to day 5.

Figure 4. Endogenous BMP signaling is essential for the production of PPE cells in adherent culture of human ES cells

(A) Schematic representation of the manipulations of BMP signaling during human ES cell differentiation. (B) qPCR analysis ID1 expression in day-5 cultures. (C) Semi-quantitative RT-PCR of selected genes in day-5 cultures. (D) Immunofluorescence of day-5 cultures without (Ctrl) or with 300 ng/ml NOGGIN (NOG) treatment. (E) Immunofluorescence for SIX1 in day-8 cultures treated with 20 ng/ml BMP4 or 300 ng/ml NOGGIN. (F–J) Relative expression levels of PPE (F), neural/neural crest (G), extracellular BMP ligand/regulator (H), BMP receptor (I), and the BMP target ID1 (J) transcripts examined by qPCR. Expression data were presented in log scales, and generated from three independent biological replicates. Error bars are S.E. Increasing amount of BMP4 ligands (20, 50, 150 ng/ml) as indicated by the wedged shaped bar in F and G. Asterisks indicate significant differences (p < 0.05, ANOVA test).

Figure 5. Attenuation of BMP is essential for differentiation of the anterior placodal ectoderm

(A and B) qPCR analyses of ID1 (A) and placodal ectoderm genes (B) in day-9 cultures treated with DMSO or LDN193189 from day 6 to day 9. (C) Histogram of the percentage of cells positive for SIX1, PAX6, or PAX3 in day-9 cultures with DMSO or LDN193189 treatment from day 6 to day 9. (D–I) Immunofluorescence of day-9 cultures with the indicated antibodies. Arrowheads show induction of PAX6 or PAX3 in SIX1+ cells; arrows denote the colocalization of PAX6 with DLX5 or ECAD; asterisks indicate significant differences (p < 0.05, t-test).

Figure 6. Differentiation of APE cells by manipulating BMP and hedgehog signaling

(A) Semi-quantitative RT-PCR of day 0, 8, and 18 cultures under SF condition. (B) Conditions to examine the role of BMP signaling in the induction of lens placode. (C) RT-PCR analysis of markers for lens placode. (D) Schematic presentation of conditions to examine the role of BMP and hedgehog signaling in the differentiation of APE cells. (E) qPCR analysis of selected genes in day-13 cultures. Asterisks indicate statistical significant difference (p < 0.05, ANOVA test) from the control (Ctrl). Abbreviations: LDN, LDN193189; CP, cyclopamine; Pur, purmorphamine. (F) Immunofluorescence for FOXE3 and PAX6 on day-13 cultures. Note that cells co-expressing FOXE3 and PAX6 were clustered together to form rosette like structures demarcated by dashed lines. (G and H) Bright-field images of lens placode cell structures (asterisks) observed at day 14 (G) and day 22 (H). Dashed lines in panel H highlight the separation of the lens-like structures from the surrounding cell monolayer.

Figure 7. Generation of placode derived sensory neurons from human ES cells

(A–C) Immunofluorescence analysis of placode-derived sensory neurons. Antibodies used are indicated in the images. Note that the ES-derived TuJ1⁺ neurons (arrows) express markers for peripheral sensory neuron markers BRN3A and peripherin as well as markers characteristic of placode-derived neurons, such as SIX1 and CK8. (D–K) Immunofluorescence on sections of E10.5 mouse embryo. Note that neurons (TuJ1⁺) (arrows) in the olfactory (D, H), the trigeminal (E, I), and the otic (F, J) placode expressed both Six1 and Ck8, whereas neural crest derived sensory neurons (asterisks) in the dorsal root ganglia (G, K) only expressed Six1 but not CK8. Abbreviations: of, olfactory epithelium; tg, trigeminal ganglion; ot, otic vesicle, cvg, cochleovestibular ganglion; nt, neural tube; drg, dorsal root ganglion.