Search

Additional file 1: Supplementary Figure S1. Related to Figure 1. Validation and characterizing of our DOX-Inducible system for manipulating FOS expression in Muscle Progenitor Cells Ex Vivo. (A) Representative images of cultured muscle progenitor cells in GM supplemented with increasing amounts of Doxycycline (0, 0.005 μg/ml, 0.01 μg/ml, 0.05 μg/ml, 0.1 μg/ml, 0.5 μg/ml, 1 μg/ml, 2.5 μg/ml, 5 μg/ml) for 48 hours in culture. Scale bars represent 50 microns. (B) Quantification of the total number of Hoechst+ cells after 48 hours in GM supplemented with the indicated concentration of DOX (n=cells from 3mice). (C) Experimental Flowchart for detecting FOS protein (sc-7292) using ProteinSimple WES platform and analysis using the COMPASS software. (D) Virtual bands showing gradual increase of FOS protein with increasing amounts of DOX. Raw area of signal is shown for loading control (GAPDH) and FOS. FOS signal normalized to GAPDH is displayed, highlighting that 1 ug/ml of DOX was the lowest concentration that gave the maximal induction of FOS protein. (E) 20X images of two-week cultured muscle progenitor cells grown in GM for 48 hours or differentiated in DM for 72 hours and stained for PAX7 and MyoG. Nuclei stained with DAPI. Scale bar represents 50 microns. (F) Corrected total cell fluorescence (CTCF) for PAX7 (left) and MYOG (right) quantified in (E). n= 98-4184 cells (PAX7) and n= 2169-46118 cells (MYOG). (G) Corrected total cell fluorescence (CTCF) for PAX7 (left) and MYOG (right) in two-week cultured pSLIK-Fos and pSLIK-Gfp muscle progenitor cells. n=2295-2816 (PAX7) and n=1432-3779 (MYOG). Mean comparisons using One-way ANOVA with post-hoc Tukey test (B) and Mann Whitney U-test (F, G). Supplementary Figure S2. Related to Figure 3. FOS-dependent suppression of myogenic genes in muscle progenitor cells is reversible upon DOX removal. pSLIK-Fos cells were treated with 1 ug/ml DOX for 72-hours and then chased with media devoid of DOX for 3 and 5 days followed by RT-qPCR. Mean relative mRNA expression (+/-SD) and normalized (to Gapdh) for Fos, MyoD, MyoG, Tcap, and Ttn. Data shows that upon removal of DOX, mRNA expression for MyoD, MyoG, Tcap, and Ttn are de-repressed concomitant with FOS returning to basal levels. Mean comparisons were performed with a one-way ANOVA with post hoc Tukey test. N= 5-6 replicate cultures from cells pooled from 4 mice. Supplementary Figure S3. Related to Figure 3. Chromatin Immunoprecipitation (ChIP)-qPCR analysis for H3K27Me3 occupancy at the promoter region of Bmp6, an intergenic region, MyoD, and Ttn in pSLIK-Fos and pSLIK-Gfp cultured muscle progenitor cells. (A) Experimental design: Primary satellite cells were infected with virus expressing pSLIK-Fos or pSLIK-Gfp constructs, selected with 100 μg/mL of hygromycin for 6 days, and expanded for approximately 3 weeks to generate enough cells for ChIP-qPCR. Once at the desired cell density, pSLIK-Fos and pSLIK-Gfp cells were cultured in GM-supplemented with 1μg/mL doxycycline (DOX) for 48 hours. Chromatin was IP’d using an H3K27Me3- or immunoglobulin G (IgG)-specific antibodies followed by qPCR. Plot shows mean percent of input (+/- SD) for each amplicon targeting the first 500bps upstream of the TSS for Bmp6 and an intergenic region (left) and MyoD and Ttn (right). Location of primers relative to gene TSS is shown in schematic above each plot. Statistical comparisons were performed with two-way ANOVA with post hoc Holm-Sidak test. N= 8-9 replicate IPs were performed on cells pooled from 4 mice. Supplementary Figure S4. Related to Figure 4. Quality Metrics of Hi-C Datasets. (A) Pairwise comparison of all Hi-C heatmaps for each replicate of pSLIK-Gfp and pSLIK-Fos datasets showing all the chromosomes and the inter-chromosomal interactions. (B) Scaling plot showing the interaction frequency as a function of genomic distance, demonstrating that pSLIK-Fos and pSLIK-Gfp muscle progenitor cells have similar rates of decay of interaction frequency. (C) Representative higher resolution images of Hi-C datasets of pSLIK-Gfp and pSLIK-Fos muscle progenitor cells. Supplementary Figure S5. Related to Figure 5. Compartment Analysis of Hi-C Data. (A-B) Heatmap for the pSLIK-Fos and pSLIK-Gfp cells showing the compartment strength, quantified by plotting interaction frequencies in 250kb bins arranged by their values along the first eigenvector (PC1/EV1) to obtain compartmentalization saddle plots. The average interaction frequencies of the observed / expected interactions between pairs of loci (250kb bins) were calculated and arranged by their compartment signal (1st eigenvector value) for pSLIK-Fos and pSLIK-Gfp muscle progenitor cells. In these plots the upper left quadrant represents B-B interactions, and the lower right corner represents A-A interactions. (C) Comparison of the intra- and inter-compartmental interaction frequencies indicates that FOS induction (i.e., pSLIK-Fos cells) results in significantly higher interactions within the A-type and B-type compartments (p < 2.2 x 10-16, Wilcoxon rank-sum test). In contrast, the inter-compartmental (between the A-type and B-type compartments) interactions were significantly decreased (p < 2.2 x 10-16, Wilcoxon rank-sum test) in pSLIK-Fos cells when compared to pSLIK-Gfp cells. (D) Gene ontology terms of differentially expressed transcripts whose coding regions have switched from open (A-type) to closed (B-type) compartmentalization. (E). Gene ontology terms of differentially expressed transcripts whose coding regions have switched from closed (B-type) to open (A-type) compartmentalization. Supplementary Figure S6. Related to Figure 6. Analysis of TAD boundaries. (A) A representative Hi-C heatmap at 40kb resolution for a 10 megabase genomic region (chr1:186,000,000-196,000,000) with insulation plots depicted on the bottom showing that most TADs are stable in pSLIK-Fos cells relative to pSLIK-Gfp cells. (B) Violin plot demonstrating similar TAD boundary strengths between pSLIK-Gfp and pSLIK-Fos datasets (p-value: Wilcoxon rank-sum test). Supplementary Figure S7. Related to Figure 6. Analysis of looping events. (A) Hi-C heatmap of pSLIK-Gfp and pSLIK-Fos datasets showing a reduction of a loop formation at the Myl1 gene locus, which is associated with a decrease in Myl1 gene expression (pSLIK-Fos vs. pSLIK-Gfp log2FC = -1.7). The z-scores of the looping interaction are depicted on the figure. (B) Overlap in MyoD-specific looping events in [38] that overlaps with FOS-specific looping events identified in this study.

RNA has been classically known to play central roles in biology, including maintaining telomeres, protein synthesis, and in sex chromosome compensation. While thousands of long noncoding RNAs (lncRNAs) have been identified, attributing RNA-based roles to lncRNA loci requires assessing whether phenotype(s) could be due to DNA regulatory elements, transcription, or the lncRNA. Here, we use the conserved X chromosome lncRNA locus Firre, as a model to discriminate between DNA- and RNA-mediated effects in vivo. We demonstrate that (i) Firre mutant mice have cell-specific hematopoietic phenotypes, and (ii) upon exposure to lipopolysaccharide, mice overexpressing Firre exhibit increased levels of pro-inflammatory cytokines and impaired survival. (iii) Deletion of Firre does not result in changes in local gene expression, but rather in changes on autosomes that can be rescued by expression of transgenic Firre RNA. Together, our results provide genetic evidence that the Firre locus produces a trans-acting lncRNA that has physiological roles in hematopoiesis.

Additional file 15: Fig. S10. Loss of TUG1 expression in infertile human males. Heatmap of microarray data from three different probe sets showing decreased expression of TUG1 in sperm from infertile teratozoospermic men (T) compared to fertile (normospermic) individuals (ND). In all cases, p < 4.39 x10-4.

Additional file 10: Fig. S5. Gene expression of multiple tissues in Tug1 WT and KO mice. Gene expression of multiple samples (columns) were described with log scale of TPM: Log2(TPM+1). Genes (rows) are clustered by hierarchical clustering with Ward’s method based on Euclidean distance. Annotation of samples are provided in the top panel in terms of genotypes and tissues.

Additional file 1. Supplemental Methods, Supplemental Figures S1-S15, and Supplemental Tables S1, S3, and S4.

Additional file 17. Review history.

Additional file 3: Fig. S3. Overview of the Tug1 locus in mouse. UCSC genome browser showing the murineTug1 locus. The three predominate Tug1 isoforms are depicted (black) and the Tug1 transgene (tg(Tug1)) is shown (blue). For Tug1 knockout, the longest annotated Tug1 isoform was replaced by a lacZ reporter cassette, leaving the promotor and first exon intact. The deleted region is indicated by red dashed lines. The open reading frame (ORF) encoding the TUG1-BOAT protein and PhyloCSF scores for the (-2) frame across the locus are depicted (grey). Chromosomal coordinates (mm10) are shown.

Additional file 12: Fig. S7. Changes of gene expression in two comparisons of Tug1-/- (KO) vs. WT and Tug1rescue (Rescue) vs. Tug1-/- (KO). Each dot represents a gene whose x-axis value is the fold change of gene expression between KO and WT and y-axis value shows the fold change in Rescue vs. KO. Color describes statistical significance of fold change (adjusted p-value < 0.05): no statistical significance in either comparison (gray, N=33688); significance in KO vs. WT (blue, N=998); significance in Rescue vs. KO (orange, N=126); significance in both comparisons, KO vs. WT and Rescue vs. KO (red, N= 52).

Additional file 5. Review history.

Additional file 11: Fig. S6. Tug1 transgene expression and fertility assessment. (A) qRT-PCR for Tug1 RNA expression in testes and sorted peripheral blood populations: WT (n = 1), Tug1+/- (n = 1), Tug1-/- (n = 1), and Tug1rescue (n = 1) and sorted peripheral blood populations. Error bars indicate the relative quantification minimum and maximum confidence interval at 98%. Not detected (n.d.). (B) Representative flow cytometry gating strategy for NK, CD4, and CD8 cells in peripheral blood from WT, Tug1+/-, Tug1-/-, and Tug1rescue mice (gating from WT peripheral blood shown). (C) Scatter dot plot (mean with standard error of the mean shown) of the number of pups at birth per copulatory plug for matings using male wild type, Tug1+/-; tg(Tug1); rtTA, Tug1-/-, or Tug1rescue (on dox diet) with wild type C57BL/6J females. Each dot represents a litter from a different mouse. (D) Sperm count from control (WT and Tug1+/-, n = 2), Tug1-/- (n = 2), and Tug1resuce (n = 3) mice. Each dot represents a different mouse and the error bars indicate the standard error of the mean. (E) Hematoxylin and eosin staining in Tug1+/-, Tug1-/-, and Tug1rescue testes and epididymis. (F) Morphological analysis of sperm from Tug1-/- (n = 2), and Tug1rescue (n = 3) mice.

Additional file 2: Fig. S2. In vivo expression pattern of Tug1 during murine embryogenesis. RNA in situ hybridization of Tug1 RNA using a digoxigenin-labeled antisense RNA probe in mouse embryos at different developmental stages. Embryonic day (E)8.5, E9.5, E10.5, E11.5, and E12.5 are shown.

Additional file 13: Fig. S8. The 5’ region of human TUG1 contains a conserved ORF. (A) GWIPS-viz tracks for human TUG1 genomic locus (hg38) is shown. Global aggregate of ribosome occupancy (ribosome profile), RNA-seq (mRNA coverage), and evolutionary protein-coding potential (PhyloCSF) across the TUG1 locus is shown. ORF1 and ORF2 are outlined with red and gray boxes, respectively. Tracks surrounding both ORFs are zoomed in for clarity (bottom). (B) Scheme showing human ORF1 construct design. hORF1 (labeled with a 3xFLAG epitope tag prior the stop codon) with the 5’UTR was inserted into pcDNA3.1(+) and transfected into HeLa cells. 48 hours post-transfection, TUG1-BOAT-3xFLAG localization was analyzed by immunofluorescence (IF) (shown in C). (C) Maximum intensity projection of HeLa cells expressing human 5’UTR-hORF1-3xFLAG. Localization of 3xFLAG tagged TUG1-BOAT was assessed by immunostaining against the 3xFLAG (green). Nucleus was monitored by DAPI (blue) and mitochondria was monitored by immunostaining against mitochondrial membrane translocase TOM20 (gray). Bar plot shows localization analysis of TUG1-BOAT. Scale bar is 5 μm.

Additional file 1: Fig. S1. Mouse and human Tug1 locus and chromatin context in different cell types. (A) Tug1 mouse and (B) human genomic loci. Evolutionary nucleotide conservation (PhyloP) of the locus are presented along with the chromatin context (DNase I hypersensitive regions, histone modifications) and protein binding ChIP-seq peaks (Pol2, CTCF, SIN3A, COREST, SETDB1, HDAC2) from ENCODE (UCSC Genome Browser, mm9) datasets in the indicated cell types.

Additional file 4: Fig. S4. Morphology analysis of Tug1-/- mice and sperm (A) Body mass (g) measurements over 11 weeks of male and female Tug1-/- mice compared to wild type littermates. Males: Tug1-/- (n = 7); WT (n = 8). Females: Tug1-/- (n = 3), WT (n = 7). Significant p values at specific time points are indicated (*). (B) Representative images from adult male mice (12 weeks old) show normal physiological appearance of external genitalia and reproductive tracks in Tug1-/- compared to WT. Seminal vesicles (SV), vas deferens (VD), bladder (B), testicle (T), epididymis (E), anterior prostate (AP). (C) Box plots of body mass (g) (left panel), relative testis mass (testis mass / body mass; middle panel) and total sperm count for wild type (n = 9) and Tug1-/- males. (D) Box plots of the percentage of different sperm morphological abnormalities for wild type (n = 9) and Tug1-/- (n = 8) males. Significant (*) p value (Wilcoxon rank sum test) is indicated.

Transcription initiates at both coding and noncoding genomic elements, including mRNA and long noncoding RNA (lncRNA) core promoters and enhancer RNAs (eRNAs). However, each class has a different expression profile with lncRNAs and eRNAs being the most tissue specific. How these complex differences in expression profiles and tissue specificities are encoded in a single DNA sequence remains unresolved. Here, we address this question using computational approaches and massively parallel reporter assays (MPRA) surveying hundreds of promoters and enhancers. We find that both divergent lncRNA and mRNA core promoters have higher capacities to drive transcription than nondivergent lncRNA and mRNA core promoters, respectively. Conversely, intergenic lncRNAs (lincRNAs) and eRNAs have lower capacities to drive transcription and are more tissue specific than divergent genes. This higher tissue specificity is strongly associated with having less complex transcription factor (TF) motif profiles at the core promoter. We experimentally validated these findings by testing both engineered single-nucleotide deletions and human single-nucleotide polymorphisms (SNPs) in MPRA. In both cases, we observe that single nucleotides associated with many motifs are important drivers of promoter activity. Thus, we suggest that high TF motif density serves as a robust mechanism to increase promoter activity at the expense of tissue specificity. Moreover, we find that 22% of common SNPs in core promoter regions have significant regulatory effects. Collectively, our findings show that high TF motif density provides redundancy and increases promoter activity at the expense of tissue specificity, suggesting that specificity of expression may be regulated by simplicity of motif usage.

The neocortex contains an unparalleled diversity of neuronal subtypes, each defined by distinct traits that are developmentally acquired under the control of subtype-specific and pan-neuronal genes. The regulatory logic that orchestrates the expression of these unique combinations of genes is unknown for any class of cortical neuron. Here, we report that Fezf2 is a selector gene able to regulate the expression of gene sets that collectively define mouse corticospinal motor neurons (CSMN). We find that Fezf2 directly induces the glutamatergic identity of CSMN via activation of Vglut1 (Slc17a7) and inhibits a GABAergic fate by repressing transcription of Gad1. In addition, we identify the axon guidance receptor EphB1 as a target of Fezf2 necessary to execute the ipsilateral extension of the corticospinal tract. Our data indicate that co-regulated expression of neuron subtype–specific and pan-neuronal gene batteries by a single transcription factor is one component of the regulatory logic responsible for the establishment of CSMN identity.

Noncoding RNAs have emerged as important key players in the cell. Understanding their surprisingly diverse range of functions is challenging for experimental and computational biology. Here, we review computational methods to analyze noncoding RNAs. The topics covered include basic and advanced techniques to predict RNA structures, annotation of noncoding RNAs in genomic data, mining RNA-seq data for novel transcripts and prediction of transcript structures, computational aspects of microRNAs, and database resources., Austrian Science Fund (Schrodinger Fellowship J2966-B12), German Research Foundation (grant WI 3628/1-1 to SW), National Institutes of Health (U.S.) (NIH award 1RC1CA147187)

Massively parallel cDNA sequencing (RNA-Seq) provides an unbiased way to study a transcriptome, including both coding and noncoding genes. Until now, most RNA-Seq studies have depended crucially on existing annotations and thus focused on expression levels and variation in known transcripts. Here, we present Scripture, a method to reconstruct the transcriptome of a mammalian cell using only RNA-Seq reads and the genome sequence. We applied it to mouse embryonic stem cells, neuronal precursor cells and lung fibroblasts to accurately reconstruct the full-length gene structures for most known expressed genes. We identified substantial variation in protein coding genes, including thousands of novel 5′ start sites, 3′ ends and internal coding exons. We then determined the gene structures of more than a thousand large intergenic noncoding RNA (lincRNA) and antisense loci. Our results open the way to direct experimental manipulation of thousands of noncoding RNAs and demonstrate the power of ab initio reconstruction to render a comprehensive picture of mammalian transcriptomes.

The regulated recruitment and differentiation of multipotent bone marrow-derived cells (BMDCs) to sites of injury are critical for efficient wound healing. Previously we demonstrated that sustained expression of HOXA3 both accelerated wound healing and promoted angiogenesis in diabetic mice. In this study, we have used green fluorescent protein-positive bone marrow chimeras to investigate the effect of HOXA3 expression on recruitment of BMDCs to wounds. We hypothesized that the enhanced neovascularization induced by HOXA3 is due to enhanced mobilization, recruitment, and/or differentiation of BMDCs. Here we show that diabetic mice treated with HOXA3 displayed a significant increase in both mobilization and recruitment of endothelial progenitor cells compared with control mice. Importantly, we also found that HOXA3-treated mice had significantly fewer inflammatory cells recruited to the wound compared with control mice. Microarray analyses of HOXA3-treated wounds revealed that indeed HOXA3 locally increased expression of genes that selectively promote stem/progenitor cell mobilization and recruitment while also suppressing expression of numerous members of the proinflammatory nuclear factor kappaB pathway, including myeloid differentiation primary response gene 88 and toll-interacting protein. Thus HOXA3 accelerates wound repair by mobilizing endothelial progenitor cells and attenuating the excessive inflammatory response of chronic wounds.

RNA has been classically known to play central roles in biology, including maintaining telomeres, protein synthesis, and in sex chromosome compensation. While thousands of long noncoding RNAs (lncRNAs) have been identified, attributing RNA-based roles to lncRNA loci requires assessing whether phenotype(s) could be due to DNA regulatory elements, transcription, or the lncRNA. Here, we use the conserved X chromosome lncRNA locus Firre, as a model to discriminate between DNA- and RNA-mediated effects in vivo. We demonstrate that (i) Firre mutant mice have cell-specific hematopoietic phenotypes, and (ii) upon exposure to lipopolysaccharide, mice overexpressing Firre exhibit increased levels of pro-inflammatory cytokines and impaired survival. (iii) Deletion of Firre does not result in changes in local gene expression, but rather in changes on autosomes that can be rescued by expression of transgenic Firre RNA. Together, our results provide genetic evidence that the Firre locus produces a trans-acting lncRNA that has physiological roles in hematopoiesis.

Transcription initiates at both coding and noncoding genomic elements, including mRNA and long noncoding RNA (lncRNA) core promoters and enhancer RNAs (eRNAs). However, each class has a different expression profile with lncRNAs and eRNAs being the most tissue specific. How these complex differences in expression profiles and tissue specificities are encoded in a single DNA sequence remains unresolved. Here, we address this question using computational approaches and massively parallel reporter assays (MPRA) surveying hundreds of promoters and enhancers. We find that both divergent lncRNA and mRNA core promoters have higher capacities to drive transcription than nondivergent lncRNA and mRNA core promoters, respectively. Conversely, intergenic lncRNAs (lincRNAs) and eRNAs have lower capacities to drive transcription and are more tissue specific than divergent genes. This higher tissue specificity is strongly associated with having less complex transcription factor (TF) motif profiles at the core promoter. We experimentally validated these findings by testing both engineered single-nucleotide deletions and human single-nucleotide polymorphisms (SNPs) in MPRA. In both cases, we observe that single nucleotides associated with many motifs are important drivers of promoter activity. Thus, we suggest that high TF motif density serves as a robust mechanism to increase promoter activity at the expense of tissue specificity. Moreover, we find that 22% of common SNPs in core promoter regions have significant regulatory effects. Collectively, our findings show that high TF motif density provides redundancy and increases promoter activity at the expense of tissue specificity, suggesting that specificity of expression may be regulated by simplicity of motif usage.

Deriving mechanisms of immune-mediated disease from GWAS data remains a formidable challenge, with attempts to identify causal variants being frequently hampered by strong linkage disequilibrium. To determine whether causal variants could be identified from their functional effects, we adapted a massively parallel reporter assay for use in primary CD4 T cells, the cell type whose regulatory DNA is most enriched for immune-mediated disease SNPs. This enabled the effects of candidate SNPs to be examined in a relevant cellular context and generated testable hypotheses into disease mechanisms. To illustrate the power of this approach, we investigated a locus that has been linked to six immune-mediated diseases but cannot be fine-mapped. By studying the lead expression-modulating SNP, we uncovered an NF-κB-driven regulatory circuit which constrains T-cell activation through the dynamic formation of a super-enhancer that upregulates TNFAIP3 (A20), a key NF-κB inhibitor. In activated T cells, this feedback circuit is disrupted-and super-enhancer formation prevented-by the risk variant at the lead SNP, leading to unrestrained T-cell activation via a molecular mechanism that appears to broadly predispose to human autoimmunity.

Deriving mechanisms of immune-mediated disease from GWAS data remains a formidable challenge, with attempts to identify causal variants being frequently hampered by linkage disequilibrium. To determine whether causal variants could be identified via their functional effects, we adapted a massively-parallel reporter assay for use in primary CD4 T-cells, key effectors of many immune-mediated diseases. Using the results to guide further study, we provide a generalisable framework for resolving disease mechanisms from non-coding associations – illustrated by a locus linked to 6 immune-mediated diseases, where the lead functional variant causally disrupts a super-enhancer within an NF-κB-driven regulatory circuit, triggering unrestrained T-cell activation.

Deriving mechanisms of immune-mediated disease from GWAS data remains a formidable challenge, with attempts to identify causal variants being frequently hampered by strong linkage disequilibrium. To determine whether causal variants could be identified from their functional effects, we adapted a massively parallel reporter assay for use in primary CD4 T cells, the cell type whose regulatory DNA is most enriched for immune-mediated disease SNPs. This enabled the effects of candidate SNPs to be examined in a relevant cellular context and generated testable hypotheses into disease mechanisms. To illustrate the power of this approach, we investigated a locus that has been linked to six immune-mediated diseases but cannot be fine-mapped. By studying the lead expression-modulating SNP, we uncovered an NF-κB-driven regulatory circuit which constrains T-cell activation through the dynamic formation of a super-enhancer that upregulates TNFAIP3 (A20), a key NF-κB inhibitor. In activated T cells, this feedback circuit is disrupted-and super-enhancer formation prevented-by the risk variant at the lead SNP, leading to unrestrained T-cell activation via a molecular mechanism that appears to broadly predispose to human autoimmunity.

RNA has been classically known to play central roles in biology, including maintaining telomeres, protein synthesis, and in sex chromosome compensation. While thousands of long noncoding RNAs (lncRNAs) have been identified, attributing RNA-based roles to lncRNA loci requires assessing whether phenotype(s) could be due to DNA regulatory elements, transcription, or the lncRNA. Here, we use the conserved X chromosome lncRNA locus Firre, as a model to discriminate between DNA- and RNA-mediated effects in vivo. We demonstrate that (i) Firre mutant mice have cell-specific hematopoietic phenotypes, and (ii) upon exposure to lipopolysaccharide, mice overexpressing Firre exhibit increased levels of pro-inflammatory cytokines and impaired survival. (iii) Deletion of Firre does not result in changes in local gene expression, but rather in changes on autosomes that can be rescued by expression of transgenic Firre RNA. Together, our results provide genetic evidence that the Firre locus produces a trans-acting lncRNA that has physiological roles in hematopoiesis., LncRNA loci potentially contain multiple modes that can exert function, including DNA regulatory elements. Here, the authors generated genetic models in mice to dissect the role of the syntenically conserved lncRNA Firre in the context of hematopoiesis.

Transcription factors (TFs) direct developmental transitions by binding to target DNA sequences, influencing gene expression and establishing complex gene-regultory networks. To systematically determine the molecular components that enable or constrain TF activity, we investigated the genomic occupancy of FOXA2, GATA4 and OCT4 in several cell types. Despite their classification as pioneer factors, all three TFs exhibit cell-type-specific binding, even when supraphysiologically and ectopically expressed. However, FOXA2 and GATA4 can be distinguished by low enrichment at loci that are highly occupied by these factors in alternative cell types. We find that expression of additional cofactors increases enrichment at a subset of these sites. Finally, FOXA2 occupancy and changes to DNA accessibility can occur in G1-arrested cells, but subsequent loss of DNA methylation requires DNA replication.

While gene expression dynamics have been extensively cataloged during hematopoietic differentiation in the adult, less is known about transcriptome diversity of human hematopoietic stem cells (HSCs) during development. To characterize transcriptional and post-transcriptional changes in HSCs during development, we leveraged high-throughput genomic approaches to profile miRNAs, lincRNAs, and mRNAs. Our findings indicate that HSCs manifest distinct alternative splicing patterns in key hematopoietic regulators. Detailed analysis of the splicing dynamics and function of one such regulator, HMGA2, identified an alternative isoform that escapes miRNA-mediated targeting. We further identified the splicing kinase CLK3 that, by regulating HMGA2 splicing, preserves HMGA2 function in the setting of an increase in let-7 miRNA levels, delineating how CLK3 and HMGA2 form a functional axis that influences HSC properties during development. Collectively, our study highlights molecular mechanisms by which alternative splicing and miRNA-mediated post-transcriptional regulation impact the molecular identity and stage-specific developmental features of human HSCs. Human hematopoietic stem cells (HSCs) display substantial transcriptional diversity during development. Here, we investigated the contribution of alternative splicing to such diversity by analyzing the dynamics of a key hematopoietic regulator, HMGA2. Next, we showed that CLK3, by regulating the splicing pattern of HMGA2, reinforces an HSC-specific program.

Summary Cellular reprogramming through manipulation of defined factors holds great promise for large-scale production of cell types needed for use in therapy and for revealing principles of gene regulation. However, most reprogramming systems are inefficient, converting only a fraction of cells to the desired state. Here, we analyze MYOD-mediated reprogramming of human fibroblasts to myotubes, a well-characterized model system for direct conversion by defined factors, at pseudotemporal resolution using single-cell RNA-seq. To expose barriers to efficient conversion, we introduce a novel analytic technique, trajectory alignment, which enables quantitative comparison of gene expression kinetics across two biological processes. Reprogrammed cells navigate a trajectory with branch points that correspond to two alternative decision points, with cells that select incorrect branches terminating at aberrant or incomplete reprogramming outcomes. Analysis of these branch points revealed insulin and BMP signaling as crucial molecular determinants of reprogramming. Single-cell trajectory alignment enables rigorous quantitative comparisons between biological trajectories found in diverse processes in development, reprogramming, and other contexts.

Defining the transcriptional dynamics of a temporal process such as cell differentiation is challenging owing to the high variability in gene expression between individual cells. Time-series gene expression analyses of bulk cells have difficulty distinguishing early and late phases of a transcriptional cascade or identifying rare subpopulations of cells, and single-cell proteomic methods rely on a priori knowledge of key distinguishing markers. Here we describe Monocle, an unsupervised algorithm that increases the temporal resolution of transcriptome dynamics using single-cell RNA-Seq data collected at multiple time points. Applied to the differentiation of primary human myoblasts, Monocle revealed switch-like changes in expression of key regulatory factors, sequential waves of gene regulation, and expression of regulators that were not known to act in differentiation. We validated some of these predicted regulators in a loss-of function screen. Monocle can in principle be used to recover single-cell gene expression kinetics from a wide array of cellular processes, including differentiation, proliferation and oncogenic transformation.

Human pluripotent stem cell derived models that accurately recapitulate neural development in vitro and allow for the generation of specific neuronal subtypes are of major interest to the stem cell and biomedical community. Notch signaling, particularly through the Notch effector HES5, is a major pathway critical for the onset and maintenance of neural progenitor cells (NPCs) in the embryonic and adult nervous system1-3. This can be exploited to isolate distinct populations of human embryonic stem (ES) cell derived NPCs4. Here, we report the transcriptional and epigenomic analysis of six consecutive stages derived from a HES5-GFP reporter ES cell line5 differentiated along the neural trajectory aimed at modeling key cell fate decisions including specification, expansion and patterning during the ontogeny of cortical neural stem and progenitor cells. In order to dissect the regulatory mechanisms that orchestrate the stage-specific differentiation process, we developed a computational framework to infer key regulators of each cell state transition based on the progressive remodeling of the epigenetic landscape and then validated these through a pooled shRNA screen. We were also able to refine our previous observations on epigenetic priming at transcription factor binding sites and show here that they are mediated by combinations of core and stage- specific factors. Taken together, we demonstrate the utility of our system and outline a general framework, not limited to the context of the neural lineage, to dissect regulatory circuits of differentiation.

The human genome contains hundreds of regions in which the patterns of genetic variation indicate recent positive natural selection, yet for most of these the underlying gene and the advantageous mutation remain unknown. We recently reported the development of a method, Composite of Multiple Signals (CMS), that combines tests for multiple signals of natural selection and increases resolution by up to 100-fold. Applying CMS to candidate selected regions from the International Haplotype Map, we localized several hundred signals to ~50-100 kb, identifying individual gene and polymorphism targets of selection. These regions included genes involved in processes known to be targets of selection, such as infectious disease, skin pigment, metabolism, and hair and sweat. We further identified many candidates that are similar to regulatory elements. In several regions, we identified variants that are significantly associated with the expression of nearby genes in the selected population. Moreover nearly half of the ~200 regions we examined localized to regions with no genes. Thirty of the regions contain long non-coding RNAs that have been shown to often regulate nearby genes, suggesting that variation within the RNAs might have functional consequences. With preliminary data now available from the 1000 Genomes Project, we are beginning to explore full sequence data, which should contains most if not all of the causal selected polymorphisms. We extended the CMS method to the preliminary data set, validating our previously identified candidates and identifying many new intriguing coding and regulatory variants.