Neural differentiation, selection and transcriptomic profiling of human neuromesodermal progenitor-like cells in vitro

ABSTRACT Robust protocols for directed differentiation of human pluripotent cells are required to determine whether mechanisms operating in model organisms are relevant to our own development. Recent work in vertebrate embryos has identified neuromesodermal progenitors as a bipotent cell population that contributes to paraxial mesoderm and spinal cord. However, precise protocols for in vitro differentiation of human spinal cord progenitors are lacking. Informed by signalling in amniote embryos, we show here that transient dual-SMAD inhibition, together with retinoic acid (dSMADi-RA), provides rapid and reproducible induction of human spinal cord progenitors from neuromesodermal progenitor-like cells. Using CRISPR-Cas9 to engineer human embryonic stem cells with a GFP-reporter for neuromesodermal progenitor-associated gene Nkx1.2 we facilitate selection of this cell population. RNA-sequencing was then used to identify human and conserved neuromesodermal progenitor transcriptional signatures, to validate this differentiation protocol and to reveal new pathways/processes in human neural differentiation. This optimised protocol, novel reporter line and transcriptomic data are useful resources with which to dissect molecular mechanisms regulating human spinal cord generation and allow the scaling-up of distinct cell populations for global analyses, including proteomic, biochemical and chromatin interrogation.


Figure S3 Robust differentiation of spinal cord progenitors from NMP-like cells in an iPS cell line
A. Schematic representation of the differentiation protocol used on the ChiPS4 cell line, including a dSMADi step from end of day 2 to end of day 4. B. Expression profile of selected genes over time in chiPS4 cells submitted to the differentiation protocol presented in A. Graphs represent the expression of each individual gene normalized to Gapdh and relative to hiPSC levels. RTqPCR data represents average of 3 independent experiments, error bars SEM. GAGAGTCCGGGATTTTCTCCCTCCGTTTCTGAGACAGCAGGATGTACTAAAAAAGCA  CTGACTGGTCCAGTAGAAGACCGAGGTCCAAACCCAGACTCTGTCACCAACTCACAG  TGACCCTGGGGAAATCTTTTCTTACCTTTGAACCTCGATTTCCTCATCTTTAAAACG  GGGACAGTGGTCTGTGCCACGTGACGCCCATCTCACAGGGATGCTCAAATAATCAAA  AGAGATCGTGCAAGCCTCAGGGCTTTGTGAACGCTAAACTGTGAGAACGTGAGGGAT  TTTACCTCCGAGGTAACCGGGTCTGAAGCTATTACAGTAATTCACTGGCGGGGAAGG  AGATGCGCTGAGCATTGCCTGGGAGTAAGCAGTCCTGGCCTCAGTTGCATCCCCAAG  TCTAAGGCGGGTGCACCGGAGAAAGGGAACAAACTCAAGTCACAGAGGTGTGTGTGT  CGGGGTGAGGGATCCCCGGGATGGAAGCCTCCCTCGCGCCCTCGGAGAGTCCCAGAG  GGTTGGGCGGAGGCGCGCGGAGACGACAACACTGTCCCCGCGGTCGCGCGCACCGGG  CGCGCGGAGGCTTCCCCGAGCCCAGGCAAGCGGCCGCGGCACAGCGCCTGATAGTCC  CGAGGCTGGCCCGGGCTGCGCCGGTGCCAATCGGCGCGCAGCCCCCCGCGGCGCTCT  CCCCGCCCCGCCTCCCCGCCCCCCTCCCCAGCTTCACTTGGCAGCGCGGACCCGGCT  CCTGGCTGGAAAGCTACCGCCAAGCCACAGCCGAAGGCAAGCCCGAGCGGCGCCATC  CCCAACCCCGCGCCGCCGACCGCCGGCCCGTGGGCGACGGGCATGGTGAGCAAGGGC  GAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAGCTGGACGGCGACGTAAAC  GGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTACGGCAAGCTG  ACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACCCTCGTG  ACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAGCAG  CACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGCACCATCTTC  TTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGACACC  CTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAACATCCTG  GGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCGACAAG  CAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGCAGC  GTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTG  CTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAG  AAGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGC  ATGGACGAGCTGTACAAGTCCGGACTCGGATCCGAGGGCAGAGGAAGTCTTCTAACA  TGCGGTGACGTGGAGGAGAATCCCGGCCCACTGGCATGGCAaGAtGGtGGtGCCAAG  GCGGCTCCCTCCCACCACAAGATTTCTTTCTCTGTCCTGGACATCCTGGACCCACAG  AAATTCACCCGCGCAGCGCTCCCTGCCGTGCGCCCGGCTCCCCGGGAAGCCAGGAAA  AGTTTGGCGGAGGTCGAAGCGGGGAAAGATGCCAGCTCCAGGGACCCTGTCCGACAG  CTGGAGACCCCTGGTAAGATGCAAGGCGGCCCCGGCCCCAGGAGGCCTCAGCCCCAA  CAATGCGGAGTGTATGGGGGAACAGCCGGGCCCGGTGAGTGGCCCTTAACAGCGTCT  TCCTCAGAGAGAAGGCGCACGGGACCGGGTGCGAAGTGTAGCCCCCGCCTCGGACTT  GGATAGAGGCAGAGAGGAGGCTCCCCGCATTACAGGGCAGGGATTTGCCGCATCCCT  GCTCACCCGCCAAGCTCACCCGCACCACAGTTCTGATGCTCGCGGTGGAAACTTACC  TGGCGCCTGTCTTGCCAGGCTTACTCATTTATCGGGCATTTAATGCGCTTGCCACGT  GCTAGGACCTGGGCTAAGGGCTGGGATAAAGGTGATGAAAACTTCGGAACCTGAGAG  ATGGACAGCATCATTAACATCACCTCCATTTTATGGATGGGGGAGCTGACGCTAAGG  CTTGCGCCGGGGGTCTTCGTGAGTAGCGAGGTCAGGTGCCACCCGGAGACGCCTGCG  GGGCTGGGCTGCCCCAGGCGCTCAAGCTCCCCAGACTCAACTGCCCGCTACCTCGAG GGCGCCACGCCGGCGAGATCTTGATCACCTAGGGGGCCCGACGTCGCTGGTTACACC TTAAGCGGGAA

Development • Supplementary information
Legend:

Plamid sequence
Primers used for cloning gRNA (antisense and sense) GFP sequence T2A sequence Nkx1.2 Exon1 repeats Figure S4 Sequencing of the GFP-T2A insertion site in the correctly targeted clone used in this study. Grey: plasmid sequence, highlighted yellow: Primers used for cloning, highlighted blue: gRNA (antisense and sense), highlighted green: GFP sequence, orange: T2A sequence, bold black: Nkx1.2 Exon1, peach: repetitive sequences.   A. Flow cytometry analysis of GFP expression at D3 (NMP-like) and D7 of the differentiation protocol. % of maximum intensity for GFP channel is plotted, geometric mean for each peak is indicated. B. Expression of selected marker genes analyzed by RTqPCR during differentiation of the second GFP-Nkx1.2 clone following the protocol presented Figure 1B. Graphs represent the expression of each individual gene normalized to Gapdh and relative to hESC levels. Average of 3 independent RTqPCR experiments, error bars are SEM.

Figure S7 Calcium signalling increases as NMP-like cells differentiate into neural progenitors
Calcium signalling was assessed using Fluo-3, AM in D3 NMP-like cells and in D8 neural progenitors (NP). A. D3 NMP-like cells exposed to Ca 2+ indicator, Fluo-3, AM (in DMSO/medium), medium alone, medium with vehicle DMSO alone, or the calcium ionophore A23187 in the presence of Fluo-3, AM; B. D8 NPs treated with Fluo-3, AM (in DMSO), medium alone, medium with vehicle DMSO alone, or the calcium ionophore A23187 in the presence of Fluo-3, AM. Green emission of Fluo-3, AM excited at 488 nm has been pseudo-coloured and presented as a heat-map using the HeatMap Histogram plugin for Image J (red=high and blue=low fluorescence); C): Quantification and comparison of calcium fluorescence in D3 NMP-like cells and D8 NPs exposed to Fluo-3, AM, medium alone, medium with DMSO or the calcium ionophore A23187 in presence of Fluo-3, AM. Quantification was made using the total fluorescence intensity from 3 images for each condition from 4 independent experiments. Data were analysed using the non-parametric Mann-Whitney test with Graphpad Prism V6. Error bars are ± standard deviations. p-value **p<0.01. Scale bar = 50 μm. These data show that calcium signalling is higher in D8 neural progenitors in comparison with D3 NMP-like cells from which they are derived.

Supplementary Materials and methods
Quality control and passage numbers for pluripotent cells used in this study H9 (WA09) hES cells were purchased from Wicell and were supplied at passage 24. The cells were thawed transferred to DEF-CS and cell banks prepared at passage 29. For routine production the cells were used between passage 29 and 39. SA121 hES cells were purchased from Cellartis AB and were supplied at passage 9. The cells were thawed and cell banks prepared at passage 13. For routine production the cells used between passage 13 and 23.
ChiPS4 hiPS cells were purchased from Cellartis AB and were supplied at passage 9. The cells were thawed and cell banks prepared at passage 13. For routine production the cells used between passage 13 and 23.
For making the Nkx1.2 GFP knock in line H9 (WA09) cells were transfected at passage 33 and monoclonal cell lines banked at passage 40. For routine production the cells used between passage 40 and 50 For quality control purposes, representative lots of each cell bank were thawed and tested for postthaw viability, and to ensure sterility and absence mycoplasma contamination. After 2 passages the cell lines were tested for the expression of pluripotency markers (Oct4, Sox2, Nanog, SSEA-3, SSEA-4, TRA-1-60 and TRA-1-81) and differentiation markers (SSEA-1, HNF-3 beta, beta-III-tubulin and smooth muscle alpha-actinin) by immunofluorescence, and the ability to form all three germ layers when embryoid bodies are allowed to spontaneously differentiate in culture (immunofluorescence for HNF-3 beta, beta-III-tubulin and smooth muscle alpha-actinin). Table S1.
List of genes specifically enriched in human NMP-like cells.
Human NMP-like genes (Fig. 3A) were determined by selection of RNAseq data using criteria indicated in the methods section (at least 10 read counts in D3(NMP-like), significantly enriched (p value < 0.01) in D3(NMP-like) compared to both hESC and hD8 samples, with a foldchange >2). Using these criteria, 1348 genes were identified as highly expressed in human NMP-like cells (D3). Full list including information on the 1348 hNMP-like genes are included in sheet 1. Sheet 2 contains the selected dataset used to make figure 3A. Both tables present for each gene: gene names and description, mean read counts from independent experiments in hESC, D3(NMP-like) and D8, fold change between (D3)NMP-like and hESC conditions (FC NMPlike/hESC), fold change between (D3)NMP-like and D8 conditions (FC NMP-like/D8), and pvalues associated (p_hECS.NMPlike, p_NMPlike.D8, p_hESC.D8).

Click here to Download Table S1
Development • Supplementary information Table S2. Primers for qPCR Click here to Download Table S2