Search

Background The traditional approach to studying the epigenetic mechanism CpG methylation in tissue samples is to identify regions of concordant differential methylation spanning multiple CpG sites (differentially methylated regions). Variation limited to single or small numbers of CpGs has been assumed to reflect stochastic processes. To test this, we developed software, Cluster-Based analysis of CpG methylation (CluBCpG), and explored variation in read-level CpG methylation patterns in whole genome bisulfite sequencing data. Results Analysis of both human and mouse whole genome bisulfite sequencing datasets reveals read-level signatures associated with cell type and cell type-specific biological processes. These signatures, which are mostly orthogonal to classical differentially methylated regions, are enriched at cell type-specific enhancers and allow estimation of proportional cell composition in synthetic mixtures and improved prediction of gene expression. In tandem, we developed a machine learning algorithm, Precise Read-Level Imputation of Methylation (PReLIM), to increase coverage of existing whole genome bisulfite sequencing datasets by imputing CpG methylation states on individual sequencing reads. PReLIM both improves CluBCpG coverage and performance and enables identification of novel differentially methylated regions, which we independently validate. Conclusions Our data indicate that, rather than stochastic variation, read-level CpG methylation patterns in tissue whole genome bisulfite sequencing libraries reflect cell type. Accordingly, these new computational tools should lead to an improved understanding of epigenetic regulation by DNA methylation.

S-palmitoylation is a reversible posttranslational modification that plays an important role in regulating protein localization, trafficking, and stability. Recent studies have shown that some proteins undergo extremely rapid palmitoylation/depalmitoylation cycles after cellular stimulation supporting a direct signaling role for this posttranslational modification. Here, we investigated whether β-adrenergic stimulation of cardiomyocytes led to stimulus-dependent palmitoylation of downstream signaling proteins. We found that β-adrenergic stimulation led to rapidly increased Gαs and Gαi palmitoylation. The kinetics of palmitoylation was temporally consistent with the downstream production of cAMP and contractile responses. We identified the plasma membrane-localized palmitoyl acyltransferase DHHC5 as an important mediator of the stimulus-dependent palmitoylation in cardiomyocytes. Knockdown of DHHC5 showed that this enzyme is necessary for palmitoylation of Gαs, Gαi, and functional responses downstream of β-adrenergic stimulation. A palmitoylation assay with purified components revealed that Gαs and Gαi are direct substrates of DHHC5. Finally, we provided evidence that the C-terminal tail of DHHC5 can be palmitoylated in response to stimulation and such modification is important for its dynamic localization and function in the plasma membrane. Our results reveal that DHHC5 is a central regulator of signaling downstream of β-adrenergic receptors in cardiomyocytes.

By combining next generation whole exome sequencing and induced pluripotent stem cell (iPSC) technology we found that an Alu repeat inserted in exon 9 of the MAK gene results in a loss of normal MAK transcript and development of human autosomal recessive retinitis pigmentosa (RP). Although a relatively rare cause of disease in the general population, the MAK variant is enriched in individuals of Jewish ancestry. In this population, 1 in 55 individuals are carriers and one third of all cases of recessive RP is caused by this gene. The purpose of this study was to determine if a viral gene augmentation strategy could be used to safely restore functional MAK protein as a step toward a treatment for early stage MAK-associated RP. Patient iPSC-derived photoreceptor precursor cells were generated and transduced with viral vectors containing the MAK transcript. One week after transduction, transcript and protein could be detected via rt-PCR and western blotting respectively. Using patient-derived fibroblast cells and mak knockdown zebra fish we demonstrate that over-expression of the retinal MAK transgene restored the cells ability to regulate primary cilia length. In addition, the visual defect in mak knockdown zebrafish was mitigated via treatment with the retinal MAK transgene. There was no evidence of local or systemic toxicity at 1-month or 3-months following subretinal delivery of clinical grade vector into wild type rats. The findings reported here will help pave the way for initiation of a phase 1 clinical trial for the treatment of patients with MAK-associated RP.

Congenital eye defects represent a large class of disorders affecting roughly 21 million children worldwide. Microphthalmia and anophthalmia are relatively common congenital defects, with approximately 20% of human cases caused by mutations in SOX2. Recently, we identified the RNA-binding motif protein 24a (Rbm24a) which binds to and regulates sox2 in zebrafish and mice. Here we show that morpholino knockdown of rbm24a leads to microphthalmia and visual impairment. By utilizing sequential injections, we demonstrate that addition of exogenous sox2 RNA to rbm24a-deplete embryos is sufficient to suppress morphological and visual defects. This research demonstrates a critical role for understanding the post-transcriptional regulation of genes needed for development.

Congenital eye defects represent a large class of disorders affecting roughly 21 million children worldwide. Microphthalmia and anophthalmia are relatively common congenital defects, with approximately 20% of human cases caused by mutations in SOX2. Recently, we identified the RNA-binding motif protein 24a (Rbm24a) which binds to and regulates sox2 in zebrafish and mice. Here we show that morpholino knockdown of rbm24a leads to microphthalmia and visual impairment. By utilizing sequential injections, we demonstrate that addition of exogenous sox2 RNA to rbm24a-deplete embryos is sufficient to suppress morphological and visual defects. This research demonstrates a critical role for understanding the post-transcriptional regulation of genes needed for development.

Additional file 6. Review History.

Additional file 5: Supplementary Figures. Supplementary Figure 1. Comparison of existing read-based metrics. Schematic depicting a genomic region containing two CpG sites with different patterns of methylation, but same average methylation, in two different samples. Calculated methylation haplotype load, methylation entropy, and epi-polymorphism metrics are shown to compare how they differentiate (or fail to differentiate) the two samples. Supplementary Figure 2. Genome-wide calculation of sample-specific CluBCpG clusters. (a) Bar chart showing total numbers of clusters identified genome wide between ENCODE B cells and monocytes. Bars show total number of clusters found in both samples, B cell only, and monocyte only. (b) Histogram depicting the distribution of the number of clusters per bin from ENCODE B cells and monocytes; x-axis truncated at 10 for clarity. Supplementary Figure 3. Predominant CpG patterns differ between shared and cell type-specific clusters. Bar plots showing the total counts of different CpG patterns identified across the full genome. Shared clusters (left) were clusters with patterns found in both B cells and monocytes. Unique clusters (right) were found only in one cell type. Plots are separated by CpG density (i.e. 2, 3, or 4 CpGs/bin). On the y-axes a number 1 indicates a methylated CpG site, 0 is unmethylated. Supplementary Figure 4. CluBCpG identified regions are predominantly found outside of DMRs. (a) Venn diagram showing the overlap of CluBCpG-identified bins with read clusters and DSS-identified DMRs using different p-value thresholds. Size of circles scale with number of regions. (b-c) Bar plots showing the ratio of overlapping regions in (a) to (b) total bins with B cell or monocyte specific read clusters and (c) the ratio of overlapping regions to total DMRs. (d) Venn diagrams depicting the overlap between DMRs and bins when adjusting the minimum length threshold within DSS. (e) Bar plot showing the percentage of bins with a cell type-specific cluster (green) and DMRs (purple) overlapping annotated genic features. (f) Bar plot of the odds ratio calculated from the overlaps in (d); annotated genomic features are defined as: promoter = transcription start site (TSS) +/− 3 kb; intragenic = TSS- transcription end site (TES); 3′ = TES +/− 3 kb; intergenic = all other genomic regions. (g) The odds ratio of the overlap between cell type-specific bins and enhancer regions. Non-DMR bins have had all bins overlapping a DMR removed from the analysis. (h) Odds ratio of the overlap between cell type-specific clusters and cell type-specific active enhancers. Supplementary Figure 5. CluBCpG informative bins as a function of read coverage. Proportion of bins with ≥10 fully covered reads vs. average read depth of the sequencing data. Calculations were performed on chromosome 19 from ENCODE B cells. Supplementary Figure 6. Comparison of multiple machine learning algorithms. Box and whisker plots showing the area under the receiver operating characteristic curve (AUC) for imputation by multiple machine learning algorithms; KNN=K nearest neighbors, LR = logistic regression, NN = neural network, RF = random forest, NB=Naive Bayes. AUCs were calculated from 5-fold cross validation of each model on data from mouse neurons (a) and glia (b). (c) PReLIM was compared against a Naive Bayes model which uses only the Column mean (average methylation at each CpG site) as a feature. No confidence filtering was performed for these comparisons. Supplementary Figure 7. PReLIM cross-tissue and cross-chromosome performance. (a) Heatmap showing how a PReLIM model trained on one library performs when predicting on a different library. (b) Heatmap showing accuracy of PReLIM when trained on one mouse neuron chromosome and predicted on all other chromosomes. Supplementary Figure 8. Imputation gains in informative bins vs. average sequencing depth. (a-b) Line graphs showing the proportion of bins with ≥10 reads covering all CpGs before and after imputation with PReLIM on human (a) neurons and (b) glia. Calculations were performed using chromosome 19. Supplementary Figure 9. Genome-wide imputation gains in informative bins. (a-b) Bar plots showing the number of bins with ≥10 reads genome-wide when analyzing human neuron and glia (a) and human B cells and monocytes (b) before and after imputation. (c-d) Heatmap showing the top 30 GO biological process terms with bins unique to either neuron (c) or glia (d) before and after imputation. Analysis was performed using GREAT, colors represent -log10 of the q-value. (e) Pyrosequencing plots of novel DMRs found only post-imputation using PReLIM on the mouse neuron and glia WGBS data. Each point represents one CpG site and are connected by a line. Horizontal dotted lines indicate average cell type-specific methylation across the DMR, from the WGBS data following PReLIM imputation. Supplementary Figure 10. No effect of imputation on DSS P-value distribution under the null. P value distributions for differentially methylated locus (DML) test statistic from DSS, in self vs. self comparisons before and after imputation by PReLIM. Each histogram shows the mean and standard deviation (error bars) across 10 random 50:50 splits of WGBS reads from mouse neuron or glia [27]. In each panel, the full P value distribution (0–1) is shown on the left and the low range (0–0.1) on the right. (a) Neuron vs. neuron. (b) Neuron vs. neuron following imputation by PReLIM. (c) Glia vs. glia. (d) Glia vs. glia following imputation by PReLIM. Supplementary Figure 11. Deconvolution PCA and permutation analyses. (a) Scree plot with bars showing proportion of explained variance provided by each of the 20 principal components (PCs). (b-c) Scatter plot of PC1 vs PC2 of the B cell:monocyte synthetic mixtures from ENCODE (b) and Blueprint (c). Colors represents the proportion of the mixture. (d) Scatter plot of PC1 vs PC2 on randomly generated data. (e-f) The predicted proportion vs true proportion on the randomly generated data (e) and Blueprint data with permuted proportion labels (f). Supplementary Figure 12. Evaluating alternate strategies to predict gene expression from promoter methylation. Random forest models using methylation data from promoter windows of +/− 3 kb from the transcription start site (TSS) (green) and 500 bp upstream of the TSS (purple) were compared against each other. Box and whisker plots overlaid with individual points showing the area under the ROC curve (AUC) for 100 random train-test splits (a) and 10-fold nested cross validation (b). To test the effect of including methylation levels at each of the individual CpG sites, the +/− 3 k promoter window was broken down into methylation frequencies in decile blocks at the individual CpG-level and tested using 100 random train-test splits (c). Whiskers extend to 1.5x the intra-quartile range.

DNA methylation regulates cell type-specific gene expression. Here, in a transgenic mouse model, we show that deletion of the gene encoding DNA methyltransferase Dnmt3a in hypothalamic AgRP neurons causes a sedentary phenotype characterized by reduced voluntary exercise and increased adiposity. Whole-genome bisulfite sequencing (WGBS) and transcriptional profiling in neuronal nuclei from the arcuate nucleus of the hypothalamus (ARH) reveal differentially methylated genomic regions and reduced expression of AgRP neuron-associated genes in knockout mice. We use read-level analysis of WGBS data to infer putative ARH neural cell types affected by the knockout, and to localize promoter hypomethylation and increased expression of the growth factor Bmp7 to AgRP neurons, suggesting a role for aberrant TGF-β signaling in the development of this phenotype. Together, these data demonstrate that DNA methylation in AgRP neurons is required for their normal epigenetic development and neuron-specific gene expression profiles, and regulates voluntary exercise behavior., AgRP neurons in the hypothalamic arcuate nucleus (ARH) are involved in regulating hunger and energy balance. Here the authors show that knockout of the DNA methyltransferase Dnmt3a in AgRP neurons of the ARH leads to a reduction in voluntary exercise along with numerous epigenetic and gene expression changes in ARH neurons.

Mutations in the key transcription factor, SOX2, alone account for 20% of anophthalmia (no eye) and microphthalmia (small eye) birth defects in humans—yet its regulation is not well understood, especially on the post-transcription level. We report the unprecedented finding that the conserved RNA-binding motif protein, RBM24, positively controls Sox2 mRNA stability and is necessary for optimal SOX2 mRNA and protein levels in development, perturbation of which causes ocular defects, including microphthalmia and anophthalmia. RNA immunoprecipitation assay indicates that RBM24 protein interacts with Sox2 mRNA in mouse embryonic eye tissue. and electrophoretic mobility shift assay shows that RBM24 directly binds to the Sox2 mRNA 3’UTR, which is dependent on AU-rich elements (ARE) present in the Sox2 mRNA 3’UTR. Further, we demonstrate that Sox2 3’UTR AREs are necessary for RBM24-based elevation of Sox2 mRNA half-life. We find that this novel RBM24–Sox2 regulatory module is essential for early eye development in vertebrates. We show that Rbm24-targeted deletion using a constitutive CMV-driven Cre in mouse, and rbm24a-CRISPR/Cas9-targeted mutation or morpholino knockdown in zebrafish, results in Sox2 downregulation and causes the developmental defects anophthalmia or microphthalmia, similar to human SOX2-deficiency defects. We further show that Rbm24 deficiency leads to apoptotic defects in mouse ocular tissue and downregulation of eye development markers Lhx2, Pax6, Jag1, E-cadherin and gamma-crystallins. These data highlight the exquisite specificity that conserved RNA-binding proteins like RBM24 mediate in the post-transcriptional control of key transcription factors, namely, SOX2, associated with organogenesis and human developmental defects.

Background DNA methylation is thought to be an important determinant of human phenotypic variation, but its inherent cell type specificity has impeded progress on this question. At exceptional genomic regions, interindividual variation in DNA methylation occurs systemically. Like genetic variants, systemic interindividual epigenetic variants are stable, can influence phenotype, and can be assessed in any easily biopsiable DNA sample. We describe an unbiased screen for human genomic regions at which interindividual variation in DNA methylation is not tissue-specific. Results For each of 10 donors from the NIH Genotype-Tissue Expression (GTEx) program, CpG methylation is measured by deep whole-genome bisulfite sequencing of genomic DNA from tissues representing the three germ layer lineages: thyroid (endoderm), heart (mesoderm), and brain (ectoderm). We develop a computational algorithm to identify genomic regions at which interindividual variation in DNA methylation is consistent across all three lineages. This approach identifies 9926 correlated regions of systemic interindividual variation (CoRSIVs). These regions, comprising just 0.1% of the human genome, are inter-correlated over long genomic distances, associated with transposable elements and subtelomeric regions, conserved across diverse human ethnic groups, sensitive to periconceptional environment, and associated with genes implicated in a broad range of human disorders and phenotypes. CoRSIV methylation in one tissue can predict expression of associated genes in other tissues. Conclusions In addition to charting a previously unexplored molecular level of human individuality, this atlas of human CoRSIVs provides a resource for future population-based investigations into how interindividual epigenetic variation modulates risk of disease. Electronic supplementary material The online version of this article (10.1186/s13059-019-1708-1) contains supplementary material, which is available to authorized users.

BACKGROUND Modern genomics has enabled the identification of an unprecedented number of genetic variants, which in many cases are extremely rare, associated with blinding disorders. A significant challenge will be determining the pathophysiology of each new variant. The Zebrafish is an excellent model for the study of inherited diseases of the eye. By 5 days post-fertilization (dpf), they have quantifiable behavioral responses to visual stimuli. However, visual behavior assays can take several hours to perform or can only be assessed one fish at a time. RESULTS To increase the throughput for vision assays, we used the Viewpoint Zebrabox to automate the visual startle response and created software, Visual Interrogation of Zebrafish Manipulations (VIZN), to automate data analysis. This process allows 96 Zebrafish larvae to be tested and resultant data to be analyzed in less than 35 minutes. We validated this system by disrupting function of a gene necessary for photoreceptor differentiation and observing decreased response to visual stimuli. CONCLUSIONS This automated method along with VIZN allows rapid, high-throughput, in vivo testing of Zebrafish's ability to respond to light/dark stimuli. This allows the rapid analysis of novel genes involved in visual function by morpholino, CRISPRS, or small-molecule drug screens. Developmental Dynamics 245:605-613, 2016. © 2016 Wiley Periodicals, Inc.

Wnt proteins regulate diverse biological responses by initiating two general outcomes: β-catenin-dependent transcription and β-catenin-independent activation of signaling cascades, the latter including modulation of calcium and regulation of cytoskeletal dynamics (Planar Cell Polarity, PCP). It has been difficult to elucidate the mechanisms by which Wnt signals are directed to effect one or the other outcome due to shared signaling proteins between the β-catenin-dependent and -independent pathways, such as the Dishevelled binding protein Naked. While all Naked paralogs contain a putative calcium-binding domain, the EF-Hand, Drosophila Naked does not bind calcium. Here we find a lineage-specific evolutionary change within the Drosophila Naked EF-hand that is not shared with other insects or vertebrates. We demonstrate the necessary role of the EF-hand for Nkd localization changes in calcium fluxing cells and using in vivo assays, we identify a role for the zebrafish Naked EF-hand in PCP but not in β-catenin antagonism. In contrast, Drosophila-like Nkd does not function in PCP, but is a robust antagonist of Wnt/β-catenin signaling. This work reveals that the zebrafish Nkd1 EF-hand is essential to balance Wnt signaling inputs and modulate the appropriate outputs, while the Drosophila-like EF-Hand primarily functions in β-catenin signaling.

Debilitating visual impairment caused by cataracts or microphthalmia is estimated to affect roughly 20 million people in the United States alone. According to the National Eye Institute, by 2050 that number is expected to more than double to roughly 50 million. The identification of candidate disease-causing alleles for cataracts and microphthalmia has been accelerated with advanced sequencing technologies creating a need for verification of the pathophysiology of these genes. Zebrafish pose many advantages as a high-throughput model for human eye disease. By 5 days post-fertilization, zebrafish have quantifiable behavioral responses to visual stimuli. Their small size, many progeny, and external fertilization allows for rapid screening for vision defects. We have adapted the OptoMotor Response to assay visual impairment in zebrafish models of cataracts and microphthalmia. This research demonstrates an inexpensive, high-throughput method for analyzing candidate genes involved in visual impairment.

Bardet Biedl syndrome (BBS) is a multisystem genetically heterogeneous ciliopathy that most commonly leads to obesity, photoreceptor degeneration, digit anomalies, genito-urinary abnormalities, as well as cognitive impairment with autism, among other features. Sequencing of a DNA sample from a 17-year-old female affected with BBS did not identify any mutation in the known BBS genes. Whole-genome sequencing identified a novel loss-of-function disease-causing homozygous mutation (K102*) in C8ORF37, a gene coding for a cilia protein. The proband was overweight (body mass index 29.1) with a slowly progressive rod-cone dystrophy, a mild learning difficulty, high myopia, three limb post-axial polydactyly, horseshoe kidney, abnormally positioned uterus and elevated liver enzymes. Mutations in C8ORF37 were previously associated with severe autosomal recessive retinal dystrophies (retinitis pigmentosa RP64 and cone-rod dystrophy CORD16) but not BBS. To elucidate the functional role of C8ORF37 in a vertebrate system, we performed gene knockdown in Danio rerio and assessed the cardinal features of BBS and visual function. Knockdown of c8orf37 resulted in impaired visual behavior and BBS-related phenotypes, specifically, defects in the formation of Kupffer's vesicle and delays in retrograde transport. Specificity of these phenotypes to BBS knockdown was shown with rescue experiments. Over-expression of human missense mutations in zebrafish also resulted in impaired visual behavior and BBS-related phenotypes. This is the first functional validation and association of C8ORF37 mutations with the BBS phenotype, which identifies BBS21. The zebrafish studies hereby show that C8ORF37 variants underlie clinically diagnosed BBS-related phenotypes as well as isolated retinal degeneration.

Retinitis pigmentosa (RP) is a highly heterogeneous group of disorders characterized by degeneration of the retinal photoreceptor cells and progressive loss of vision. While hundreds of mutations in more than 100 genes have been reported to cause RP, discovering the causative mutations in many patients remains a significant challenge. Exome sequencing in an individual affected with non-syndromic RP revealed two plausibly disease-causing variants in TRNT1, a gene encoding a nucleotidyltransferase critical for tRNA processing. A total of 727 additional unrelated individuals with molecularly uncharacterized RP were completely screened for TRNT1 coding sequence variants, and a second family was identified with two members who exhibited a phenotype that was remarkably similar to the index patient. Inactivating mutations in TRNT1 have been previously shown to cause a severe congenital syndrome of sideroblastic anemia, B-cell immunodeficiency, recurrent fevers and developmental delay (SIFD). Complete blood counts of all three of our patients revealed red blood cell microcytosis and anisocytosis with only mild anemia. Characterization of TRNT1 in patient-derived cell lines revealed reduced but detectable TRNT1 protein, consistent with partial function. Suppression of trnt1 expression in zebrafish recapitulated several features of the human SIFD syndrome, including anemia and sensory organ defects. When levels of trnt1 were titrated, visual dysfunction was found in the absence of other phenotypes. The visual defects in the trnt1-knockdown zebrafish were ameliorated by the addition of exogenous human TRNT1 RNA. Our findings indicate that hypomorphic TRNT1 mutations can cause a recessive disease that is almost entirely limited to the retina.