Electrophysiological evidence for successful neuronal development. Cells were differentiated on glass coverslips towards the neuronal lineage for 20–24 days and then placed into a temperature-controlled recording chamber for whole-cell patch-clamp studies. (a) Representative example for the currents observed during the 20 ms voltage steps of the whole-cell voltage clamp recording protocol displayed in (b). Note that Na⁺ currents (downwards deflection) are observed at voltages ⩾−40 mV (solid line). Strong depolarizing and repolarizing (K⁺ currents; upwards deflection) are observed at depolarization to 0 mV (dashed line). (c) For voltage clamp recording (voltage step from −80 to 0 mV) of Ca²⁺ channels, Na⁺ and K⁺ channels were blocked by addition of tetrodotoxin, tetraethylammoniumchloride (5 mM), 4-aminopyridine (10 mM) and substitution of intracellular K⁺ ions by 120 mM Cs⁺. Moreover, the measurement of Ca-currents was favoured by a bath solution containing barium ions (10 mM) instead of calcium ions. Current traces were obtained without Ca²⁺-channel blocker, or with the blockers nimodipine (1 μM) or Cd²⁺ (1 mM) added. Current data at 15 ms after the voltage step were corrected for cell capacitance (indirect measure for cell size) and displayed. Data represent means±S.D. ^**P<0.01. (d) Spontaneous action potentials were recorded in current clamp mode (0 pA). At the time indicated by an arrow, tetrodotoxin was added. The dashed line indicates 0 mV membrane potential. The scale bars indicate the dimensions of the membrane potential and the time domain. (e) Recordings at individual neurons excited with specific glutamate receptor agonists in the presence or absence of blockers. Current traces were recorded after application of N-methyl--aspartate (NMDA) or kainic acid. All agonists were also tested in the presence of their respective specific antagonist (traces with 5-aminophosphovalerate (AP-5), 6,7-dinitroquinoxalin-2,3-dione (DNQX)). The scale bars represent the current and time dimensions of the experiment. Data are representative for n⩾10 neurons (for agonists) and n=3 for antagonists (on neurons with positive agonist response)

Cluster analysis of mRNA time course profiles, and their association with distinct phases of differentiation. (a) Gene expression kinetics was determined for all genes represented on the chip. An unbiased clustering analysis of the kinetic profiles of all regulated genes was performed. For each cluster (named Ia, Ib, IIa, IIb, IIIa, IIIb, IV, V), the means of the absolute expression level of all genes in the respective cluster, for each analysis time point, are displayed and plotted on a logarithmic scale; n: number of genes in the cluster (b). Number of genes expressed in mESCs (ESCs=40 evaluated, 33 found), NPCs (73 evaluated, 63 found) and developing neurons (N) were analyzed by extensive literature search (mESCs, NPCs) or GO-analysis (N). The relative distribution of these genes across the different clusters was calculated (in %) and displayed (e.g. 65% of all ESC markers was found in cluster Ia, 35% of all N markers in cluster III)

Correlation of neural precursor formation with a transiently upregulated group of genes. (a) On DoD7, cultures were immunostained for the neural stem cell marker nestin (green) and DNA (red). Scale bar: 100 μm. (b) For quantification of nestin-positive NPCs, cells were immunostained for nestin on DoD7 and analyzed by flow cytometry. Data are means±S.D. of seven independent differentiations. ^***P<0.001. (c) Relative expression profiles of genes from cluster IIa were calculated by normalization of expression of each gene to DoD0 expression, which was arbitrarily set to 1. The expression kinetics for each gene within that cluster is displayed. (d) Genes upregulated during neuronal differentiation of mESCs were analyzed for their role in regional specification of the brain and classified accordingly (colour-coding). The number of genes associated with each of the three chosen subregions of the brain are displayed separately for each regulation cluster. A detailed list of genes with their regional assignment is given in Supplementary Figure S5

Functional assignment of neuronal genes upregulated in different waves. A combination of bioinformatics tools and literature information was used to search all upregulated clusters for conspicuous biological themes and for genes associated with them. Themes are displayed and corresponding genes (with original NCBI gene names) are colour-coded according to the clusters they were found in (displayed graphically besides the legend, with dots on the lines representing DoD0, DoD7, DoD15 and DoD20). (a) Core neurochemical themes. Note a relatively early induction of receptors and channels, compared with late emergence of genes coding for transporters and synaptic vesicles, and those related to neurodegenerative disease. (b) Themes related to neurite growth indicate an early focus on growth cone formation and guidance. (c) Genes related to extracellular matrix are displayed. (d) The cells were treated with a non-cytotoxic concentration (assessed by resazurin reduction and LDH release, data not shown) of lead (1 μM) only during the last phase of differentiation (DoD14–DoD20). RNA was isolated on DoD20 and used for quantitative RT-PCR analysis of genes associated with neurodevelopment and known to be involved in neuronal disease. Pink1 and Snca were not affected. Also, their relative increase with respect to the pan-neuronal marker synaptophysin was not significant. Data indicate relative expression levels in % compared with the untreated controls on DoD20 and are means±S.D. (n=2). Significance levels (ANOVA) are indicated (^*P<0.05, ^**P<0.01, ^***P<0.001)

Protein and mRNA-based markers of robust neuronal differentiation of mESCs. (a) Cultures of mESCs were fixed and stained on day 20 of differentiation. DNA (blue) was stained with H-33342. Proteins are indicated as text on the micrograph in the same color as used for the display of their staining pattern. Tuj1: neuronal form of beta-III tubulin; NeuN: neuron-specific nuclear antigen, encoded by fox3;⁴⁰ GAD65: glutamate decarboxylase 2; SV2: synaptic vesicle glycoprotein 2a; PSD95: post-synaptic density protein 95. Scale bars: 20 μm. (b) mESC cultures (n=5 biological experiments) were differentiated towards neurons, and RNA was prepared at the indicated days of differentiation. Gene expression of the stemness factor Oct4, the neural stem cell marker Nestin (nes), the mature neuronal marker synaptophysin and the glial marker gfap was quantified by quantitative RT-PCR. The mean±S.D. of the relative expression compared with day 0 (set to 1 on each diagram) was calculated and displayed (dotted lines). Relative gene expression data were also obtained by chip analysis and the means (n=2) are displayed (solid line). Note the different scaling of the axes for chip or RT-PCR analysis, which was chosen for reasons of better comparability of the overall curve shapes. The figures in the diagram indicate the relative expression level on DoD20 (DoD7 for nestin) versus DoD0, and thus define the axis scaling

Detection of non-cytotoxic developmental disturbances by transcriptional analysis. Cultures of mESCs were neuronally differentiated for 7, 15 or 20 days as indicated in a–d. They were exposed to RA or cyclopamine (Cyclo) for the time periods indicated by the hatched boxes. (A) RNA was isolated at the indicated days (diamond) and used for quantitative RT-PCR analysis of selected differentiation and patterning markers. Headings indicate the overall biological effect, such as accelerated neuronal differentiation (e.g. neuronal diff. (+)) or altered patterning (e.g. caudalization). Names are the official gene names, apart from the following: Vglut1=Slc17a7, HB9=Mnx1. The data indicate relative expression levels in % compared with untreated controls at the same time point and are means±S.D. from two to three independent experiments for each treatment and exposure schedule. Significance levels (by ANOVA within a given experimental condition) are indicated (^*P<0.05, ^**P<0.01, ^***P>0.001). The complete data set with S.D. is given in Supplementary Figure S2. (B) Representative images of cultures on DoD15 in condition a. RA- and cyclopamine-treated cultures were viable, indistinguishable from controls (ctrl.)

Indication of a progressive change in chromatin organization and epigenetic factors in waves of fast and slow downregulation. Gene lists of relevant processes were assembled both with the help of the GO database and extensive literature search. The clusters were then queried for the presence of these genes. (a) Processes linked to chromatin or DNA repair and DNA replication are displayed, and for each of them, the number of genes found to be regulated during neuronal differentiation of mESCs is displayed in brackets. The individual genes are listed in Supplementary Figure S4. Among the identified genes, four (Smarca1, Myst4, Jmjd3 and Hdac11) are known to be neurospecific and five (Suz12, Ezh2, Bmi1, Cbx2 and Cbx8) are components of the polycomb repressor complexes, which play an important role in differentiation-related control of gene promoters. These genes could serve as sensitive markers to detect negative effects of compounds on early developmental processes. For each process, the percentage of genes present in the different clusters is indicated by colour-coded pie charts. All green shades represent clusters of genes downregulated from DoD0 to DoD20. (b) Changes in chromatin structure during differentiation were visualized by DNA staining with DAPI (green) and confocal microscopic analysis. Left panel: undifferentiated mESCs; right panels: neuronally differentiated cells on DoD20 that were stained with neuron-specific beta-III tubulin antibody (red). Scale bar: 10 μm

Analysis of glia-associated genes. (a) DoD20 cultures were fixed and stained for GFAP (green; to identify astrocytes) and Tuj1 (red; to identify neurons). The left image shows a representative overview with large neuronal areas and one typical astrocytic island. The right image shows an astrocytic island in greater detail. Scale bars=100 μm. (b) The table indicates the glia-related genes identified in this study, sorted by the cluster of expression kinetics they fell into. Astrocyte-related genes searched for, but not identified here were glutamine synthetase (Glul), S100b, Slc1a2 (Glt-1, Eaat2), Connexin 30/43 (Gjb6/Gja1), NfiA (also found in oligodendrocytes). Oligodendrocyte-related genes not found here were ATP-binding cassette, sub-family A (Abca2), CNPase (Cnp1), a microtubule-associated protein (Mtap4), myelin-glycoproteins (Omg and Mog), Olig2/3 (Olig2, Olig3), myelin protein zero (Mpz), Ng2 (Cspg4), NfiA. (c) Expression of selected astrocyte-related genes was monitored by qPCR on day 0, 7, 15 and 20 of two differentiations. Data for each differentiation are given individually. The lines indicate the resulting mean values