Fig 2: EMP transcriptional memory is enriched in metastatic castrate-resistant prostate cancer. a Principal component analysis as cells undergo transition from an epithelial (No Dox; D0) to a mesenchymal (EMT5; E5) and back to an epithelial (MErT; M3–20) state. b Heatmap showing differential transcript expression between MErT20 (M20) and No Dox (D0) (1564 gene identities can be found in the supplementary dataset 1). Heatmap color refers to normalised z-score. c Activation scores of upstream regulators in transcripts differential in EMT5 (E5) vs No Dox (D0) and MErT20 (M20) vs No Dox (D0). d Transcriptional clusters representing persistent transcript alterations. Cluster VII shows transcripts that remained upregulated and Cluster VIII shows transcripts that remained downregulated during the 20 day MErT. Cluster IX shows transcripts that were unaltered during EMT (E5) and then became newly activated by 3 days of MErT (M3) and remained for 5 and 20 days (M3,5–20). Cluster X shows transcripts that were unaltered during EMT (E5) and then became newly repressed by 3 days of MErT (M3) and remained for 5 and 20 days (M5–20). Cluster avg is the average transcriptional pattern of the transcript cluster. Heatmaps represent gene expression within Clusters VII and VIII, and within clusters IX and X. Heatmap color refers to normalised z-score. e Enrichment score of the (ei) EMT persistent (Clusters VII/VIII combined) and (eii) MErT unique (Clusters IX/X combined) signatures in samples of localised PCa and mCRPC from the Grasso dataset [23]. f Correlation plot of the EMT persistent and MErT unique signature scores in (fi) localised PCa and mCRPC samples (fii), and benign prostate tissue from the Grasso dataset. Error bars indicate SEM. **** p < 0.0001; unpaired t-test. g Immunohistochemical staining of POU4F1 and TUBB3 in samples from treatment naïve patients, and during neoadjuvant hormone therapy (NHT). Breakdown of patient numbers can be found in Table S3. Scale bars are 100 µm

Fig 3: Optimized induction condition dramatically improved iPSC-MN survival and yield(A) Phase contrast micrographs of iPSC-MNs induced via indicated combinations of transcription factors at 7 dpi. Scale bars, 100 µm.(B) Survived neurons under indicated conditions. The number of neurons at 1 wpi (weeks post-viral infection) was set as 100%. N, NGN2; S, mSox11; I, ISL1; L, LHX3. N (neurons) > 3,000 from triplicates. ** p<0.01; *** p<0.001.(C) Representative micrographs of iPSC-MNs induced via indicated transcription factors at 3 wpi with immunostaining of TUBB3. Scale bars, 100 µm.(D) The yield of iPSC-MNs at 3 wpi from 1 million NPCs at transduction. N (biological replicates) = 3. **** p<0.0001. Adapted with permission from (Sepehrimanesh and Ding, 2020).

Fig 4: TDP-43 depletion in neurons leads to altered splicing in synaptic genes UNC13A and UNC13B.a, Differential splicing analysis by MAJIQ33 in control (n = 4) and CRISPRi TDP-43 depleted (KD) (n = 3) iPS cell-derived cortical-like i3Neurons. Each point denotes a splice junction. b, Representative sashimi plots showing cryptic exon (CE) inclusion between exons 20 and 21 of UNC13A upon TDP-43 knockdown. c, f, Schematics showing intron retention (IR) (orange; bottom), TDP-43 binding region22(green), and two ALS- and FTLD-associated SNPs (red) in UNC13A (c) and UNC13B (f). d, LocusZoom plot of the UNC13A locus in the most recent ALS GWAS15; the dashed line indicates the risk threshold used in that study. Lead SNP rs12973192 is plotted as a purple diamond, other SNPs are coloured by linkage disequilibrium (LD) with rs12973192 in European individuals from 1000 Genomes. Ref. var., reference variant. e, Representative sashimi plot of UNC13B showing inclusion of the FSE upon TDP-43 knockdown. g, BaseScope detection of UNC13A CE (white puncta) in control (top) and TDP-43-knockdown (bottom) i3Neurons co-stained for TDP-43 (green), neuronal processes (stained for TUBB3, pink) and nuclei (blue). Scale bar, 5 µm. h, Quantification of RT–PCR products using iPS cell-derived neurons made from an independent iPS cell line, NCRM5, with a non-targeting control short guide RNA (sgRNA) (sgTARDBP-), an intermediate TDP-43 knockdown (sgTARDBP+) or stronger TDP-43 knockdown (sgTARDBP++). Data are mean ± s.e.m. sgControl, n = 6; sgTARDBP+, n = 5; sgTARDBP++, n = 6; one-way ANOVA with multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. i, Schematic of nanopore long reads quantified in j, Extended Data Figs. 2d, e, 5e, f. j, Percentage of targeted UNC13A long reads with TDP-43-regulated splice events that contain CE, intron retention or both in TDP-43-knockdown SH-SY5Y cells.Source data

Fig 5: Epidermal melanoma cells invade the dermis and undergo phenotypic switch to a neuronal-like state. a Schematic showing TAM treatment and skin implantation onto nude mice for b–h. b, c Immunofluorescence for GFP at 10 days b and 17 days c after skin implantation in c-Kit-CreER:Braf:Pten:GFP mice. Arrowheads point to GFP+ melanoma cells that invade into the dermis. d Immunofluorescence for GFP (green) and E-cadherin, MCAM (red) at 17 days after skin implantation in c-Kit-CreER:Braf:Pten:GFP mice. e Immunofluorescence for Sox10 (red) and E-cadherin, MCAM (green) in human melanoma specimen during the invasive radial growth phase. f, g Immunofluorescence images for GFP (green) and Dct, MITF, Nestin, GFAP, Tubb3 (red) and brightfield images in areas of dermal melanoma at 22 days f and 30 days g after skin implantation in c-Kit-CreER:Braf:Pten:GFP mice. h Dot plot showing percentage of pigment, Dct, Nestin, GFAP and Tubb3 in GFP+ dermal melanoma cells (mean ± s.d.; 3 independent tumors from 3 mice were analyzed in each group, n = the number of randomly selected tumor areas analyzed. At least 1500 GFP+ tumor cells were analyzed in each group). i Schematic model illustration showing that upon oncogenic induction, McSCs first give rise to epidermal melanoma. Then melanoma cells in the epithelial compartments invade into the dermis. Finally, dermal melanoma cells undergo a phenotypic switch to acquire neuronal signatures. j Schematic showing 4-hydroxytamoxifen (4HT-TAM) treatment and analysis regimen. k Immunofluorescence images for Tomato (red) and Sox10, Dct, Tubb3, Nestin (green) and brightfield images in c-Kit-CreER:Braf:Pten:Tomato mouse skin at 47 days. l Immunofluorescence images for Tomato (red), Dct (green) and brightfield images in lymph node of c-Kit-CreER:Braf:Pten:Tomato mouse at 47 days. m Survival graph showing the days of survival of indicated mice (n = the number of mice). Dashed line outlines the boundary of epithelium and dermis. Scale bars, 50 µm. Source data are provided as a Source Data file