Article Figures & Data
Figures
- Figure EV1. Diagram of meta‐analysis flowchart showing our search for potential CNS iPSC papers spanned over three years
Pipeline includes our inclusion and exclusion criteria for selection of candidate papers; 93 of which met these requirements and were examined in our study.
- Figure 1. Diversity of methods utilized to model neurological diseases using patient‐derived iPSCs
Pie charts illustrating the percentage distribution of disease groups and specific diseases.
Venn diagram showing unique and shared experimental techniques among the examined articles.
Minimal information about iPSC experiments (MiPSCE). Percentage of studies with or without the information for each category is depicted by color. Complete descriptions of the categories can be found in Materials and Methods. SMA, spinal muscular atrophy; ALS, amyotrophic lateral sclerosis; FTD, frontotemporal dementia.
- Figure EV2. Taxonomy of phenotype classes established to classify our 663 phenotypes
Phenotype classes include definitions and examples from our analysis. Colors of each phenotype class are preserved through all of the Figures.
- Figure 2. Distribution of phenotype clusters and ideogram of iPSC phenotype–genotype associations
Taxonomy of classes from 663 iPSC‐derived cellular phenotypes.
Chromosomal maps illustrating the locations of the 42 mappable loci involved in this study. Locus label colors are indicative of the phenotype class observed for each respective locus. These colors were used in the succeeding figures.
- Figure 3. Phenotypic classes by patient‐derived cell type from 663 annotated phenotypes and phenotype: paper metric
Distribution of the phenotype classes within each CNS cell type with total number of phenotypes listed above each respective column.
Circos plot of phenotype classes by cell type and vice versa is depicted by connecting ribbons, with the width of each band proportional to the percent composition and the top‐most ribbons highlighted. The neuronal ribbons (blue) were found to connect to and be largest for almost every phenotypic class. The outer track indicates the numeric percentage of phenotypic classes comprising each cell type.
Metric of total phenotypes per cell type with respect to the total number of studies that investigated that particular cell type.
- Figure 4. Quantification of phenotypes by genes and developmental stage
A. Schematic diagram depicting developmental timeline of iPSC‐derived cells included in analysis.
B–F Percent distribution plots of (B) iPSC, (C) NSC, (D) astrocyte, (E) oligodendrocyte, and (F) neuronal phenotypes reported for genes linked to neurodegenerative, neurodevelopmental, or other (psychiatric and viral‐induced) disorders. Each data point represents a specific disease. One‐way analysis of variance (ANOVA) with Bonferroni multiple comparisons tests was performed (NDeg, n = 13; NDev, n = 15; Other, n = 3). Data are expressed as mean percentage ± s.e.m., *P < 0.05, **P < 0.01.
G. Distribution of phenotypes by pluripotent, progenitor, and postmitotic cell type for Alzheimer's disease (AD), Parkinson's (PD), Huntington's disease (HD), and Rett syndrome. Two‐way ANOVA with Tukey's multiple comparisons test was performed. *P < 0.05.
H. Quantification of number of genes observed by phenotype.
I. Quantification of observed phenotypes per gene.
- Figure 5. A network view of overlapping phenogenetic associations
The network view was built by combining genetic and phenotypic associations. Diamond nodes represent gene loci and are labeled in red while elliptical nodes indicate phenotypes, which are colored by their phenotypic class as described by Fig EV2. Each phenotypic node is represented by a distinct number and, for identification purposes, can be found in Appendix Table S10. The full version of this network, generated through an identical method of analysis, is included in Appendix Fig S3.
- Figure EV3. Phenogenetic networks of genes linked to Alzheimer's and Parkinson's disease reveal concordant phenotypes
A, B A nuanced phenogenetic network view of genes associated with (A) Alzheimer's disease and (B) Parkinson's disease. The number of concordant phenotypes shared by gene pairs of AD and PD is outlined in tables, with APP‐Sporadic and APP‐PSEN1 having the most in AD and LRRK2‐PINK1 in PD.
- Figure EV4. Comparison between phenotype and gene ontology
Pairwise statistical comparison of functional annotations derived from well‐established gene ontologies and phenotype ontology, which notably includes several novel phenotype ontology terms, n = 15, that have yet to be reported in gene ontologies.
- Figure 6. Phenogenetic correlation of neurons in neurodegenerative and neurodevelopmental disorders
Representative treemaps of genes linked to neurodegenerative and neurodevelopmental diseases show significant molecular alterations at the neuronal level, reflecting the presence of reported cellular phenotypes and therefore high phenogenetic correlation.
A. Downregulated functional annotation of neurons with mutant SMN1.
B, C (B) Downregulated and (C) upregulated functional annotations of neurons with mutated SNCA show decreased expression of SUV39H1, a regulator of neuronal survival (Liu et al, 2005), and upregulation of AGPS, a gene involved in lipid biosynthesis (Brites et al, 2004), respectively.
D, E Neurons containing DISC1 mutations show (D) downregulation of genes associated with neurogenesis, like OTX2 (Puelles et al, 2004), and (E) upregulation of gene expression, including NEDD4L, related to neurotransmission (Laedermann et al, 2013).
- Figure 7. Spatial and temporal expression patterning of dysregulated genes from NSCs in vivo
A–C Heatmaps show localization of dysregulated gene expression from NSCs with mutant LRRK2 in the (A) developing cortex (CP) and progenitor zones (IZ, SVZ) of the prenatal human brain, (B) temporal expression predominates in the brain during the late months of development through the years of adulthood, and (C) spatial gene expression in the adult human brain localizes to the mesencephalon (Mes), myencephalon (My), and cerebellum (Cb). CP, Cortical plate; Cx, Cortex; CxN, Subcortical Nuclei; Die, Diencephalon; HPC, Hippocampal Formation; HTS, Hindbrain transient structures; IZ, Intermediate zone; Met, Metencephalon; Mos., Months; MZ, Marginal zone; SG, Subpial granular zone; SP, Subplate zone; SS, Sulci and spaces; SVZ, Subventricular zone; WM, White Matter.
- Figure 8. Spatial and temporal expression patterning of dysregulated genes from neurons in vivo
A–C Heatmaps show localization of dysregulated gene expression from neurons with mutations in DISC1 in the (A) developing cortex (CP, SP) and progenitor zones (SVZ) of the prenatal human brain, (B) temporal expression predominates in the early and late weeks of postconception brain during development, and (C) spatial gene expression in the adult human brain localizes to the subcortical nuclei (CxN). CP, Cortical plate; HTS, Hindbrain transient structures; IZ, Intermediate zone; Mos, Months; MZ, Marginal zone; SG, Subpial granular zone; SP, Subplate zone; SS, Sulci and spaces; SVZ, Subventricular zone.
- Figure 9. Novel principles of phenogenetic correlations of iPSC‐derived cellular phenotypes derived from patients included in our meta‐analysis of iPSC models of neurological disorders
Tables
- Table 1. Association of phenotype ontology and gene ontology
Phenotype ontology term Gene P‐value Gene ontology functional annotation P‐value SMN protein SMN1 5.50 × 10−15 Absent Absent Aβ protein PSEN1 6.66 × 10−14 Absent Absent Interferons UNC‐93‐B 9.33 × 10−14 Absent Absent TDP inclusions TDP‐43 9.63 × 10−14 Formation of cytoplasmic inclusions 6.02 × 10−4 Neurofilaments SOD1 4.79 × 10−10 Formation of neurofilament inclusions 1.09 × 10−4 Caspase‐4 APP 4.79 × 10−10 Absent Absent RNA foci C9ORF72 1.12 × 10−9 Absent Absent Motor neurons SOD1 3.31 × 10−9 Neurodegeneration of motor neurons 3.28 × 10−4 Glutamatergic neurons MeCP2 3.47 × 10−9 Absent Absent Binding immunoglobin protein APP 1.35 × 10−7 Absent Absent Cellular autophagy NPC1 3.23 × 10−7 Autophagy 1.66 × 10−2 Glutathione PINK1 4.50 × 10−7 Absent Absent ATP levels HTT 5.17 × 10−7 Depletion of ATP 2.74 × 10−4 Mitochondrial membrane PINK1 5.69 × 10−7 Function of mitochondria 7.66 × 10−4 Tau filaments APP 2.75 × 10−6 Generation of tau filament 1.09 × 10−4 Nuclear morphology LRRK2 1.16 × 10−5 Organization of nuclear envelope 4.38 × 10−4 Neural rosettes ATP7A 7.51 × 10−5 Absent Absent GABAergic neurons SCN1A 7.51 × 10−5 Absent Absent Lamin LRRK2 1.35 × 10−4 Absent Absent Aβ protein APP 1.85 × 10−4 Aggregation of amyloid fibrils 5.47 × 10−5 Alpha‐Synuclein SNCA 2.06 × 10−4 Absent Absent Increased susceptibility to chemical exposure LRRK2 4.13 × 10−4 Absent Absent Caspase‐3 activation LRRK2 4.57 × 10−4 Absent Absent Cellular autophagy GBA1 1.90 × 10−3 Absent Absent Motor neurons SMN1 2.85 × 10−3 Loss of motor neurons 2.74 × 10−4 Table comparing phenotype ontologies and gene ontology functional annotations with respective P‐values. If the phenotype ontology was not reported in the current gene ontology, it was termed “Absent”.
Expanded View Figures PDF [emmm201708191-sup-0002-EVFigs.pdf]
Web Resources EV1 [emmm201708191-sup-0003-Web_Resources_EV1.pdf]
Movie EV1 [emmm201708191-sup-0004-MovieEV1.zip]
