Journal: Science (New York, N.Y.)

Article Title: mRNA initiation and termination are spatially coordinated

doi: 10.1126/science.ado8279

Figure Lengend Snippet: (A) Schematic of a gene with alternative start and end sites. Alternative first exons (AFEs , dark grey ) and their respective transcription start sites (TSSs) are shown on left, while alternative last exons (ALEs, light grey ) and their respective polyadenylation sites (PASs) are shown on right. Exons are numbered by the order in which they appear in the direction of transcription (ordinal position). (B) The total number of protein coding genes ( y-axis ) that use a given number of annotated AFEs ( black ) or ALEs ( grey ), aggregated across all GTEx tissues. (C) Distribution of Pearson’s r between the number of expressed AFEs and ALEs per gene in all GTEX samples (n= 17,350, mean r = 0.53). Inset shows a representative ovary sample, marked by the dashed line of the distribution plot, in which genes expressing n alternative first exons (AFEs, x-axis ) and n alternative last exons (ALEs, y-axis ) are depicted. Color intensity represents the number of genes exhibiting each unique AFE–ALE count combination. The trend line ( black ) reflects the correlation between the number of expressed AFEs and ALEs in genes using multiple AFE and ALEs (Pearson’s r = 0.55, p-value = 1.85 × 10−⁸). (D) Heatmap of Pearson’s r for pairwise correlations between the relative usage (Ψ) of AFEs and ALEs based on their ordinal position for genes with multiple first and last exons ( left panel , n= 1,560,899 exons) and genes with exactly 3 AFEs and 3 ALEs ( right panel , n = 63,811 exons).

Article Snippet: For the second batch ( MAST1 , ZNF638 , and control RNA), extracted RNA was sent to Novogene Corporation for polyA enriched mRNA library preparation and high-throughput sequencing on a NovaSeq 6000 with 150nt paired-end reads.

Techniques: Expressing

Journal: Science (New York, N.Y.)

Article Title: mRNA initiation and termination are spatially coordinated

doi: 10.1126/science.ado8279

Figure Lengend Snippet: ( A ) Annotated isoforms ( top , orange ), a subset of LRS reads ( middle , purple with introns in thin black lines ), and annotated protein domains ( bottom , light blue ) in H9 cells for MYO10 . ( B ) Correlation between LRS read start ( x-axis ) and end coordinates ( y-axis ) for MYO10 (Spearman’s ρ = 0.66). ( C ) Schematic of expected genome-wide distributions for Spearman’s ρ showing possible shifts towards negative correlations ( orange , anti-PITA), positive correlations ( blue , PITA), or an unbiased distribution ( grey , no PITA) using fictional data. ( D ) Schematic of expected ΔAUCs for the categories outlined in C. ΔAUC is defined as AUC ρ>0 - AUC ρ<0 . Distribution of Spearman’s ρ (E) and the mean ΔAUC across samples (F) for solo termini genes (soloT, grey ) and dual alternative termini genes (dualT, blue ) in five human tissue types. Kolmogorov-Smirnov test * p-value < 10 −8 ; ** p-value < 10 −16 for E . ( G ) Distribution of ΔAUC values across 109 long-read sequencing samples for soloT ( grey ) and dualT genes ( blue ). T-test * p-value < 10 −16 .

Article Snippet: For the second batch ( MAST1 , ZNF638 , and control RNA), extracted RNA was sent to Novogene Corporation for polyA enriched mRNA library preparation and high-throughput sequencing on a NovaSeq 6000 with 150nt paired-end reads.

Techniques: Genome Wide, Sequencing

Journal: Science (New York, N.Y.)

Article Title: mRNA initiation and termination are spatially coordinated

doi: 10.1126/science.ado8279

Figure Lengend Snippet: ( A ) Schematic of mRNA isoforms based on PITA classification. PITA isoforms preferentially use TSSs and PASs that are ordinally similar ( light blue ), while anti-PITA isoforms preferentially use ordinally different TSSs and PASs ( light orange ). ( B ) Annotated mRNA isoforms ( orange ) and a randomly subsampled proportion of reads for DNAJC11 in lung ( green ), iPSCs ( pink ), and astrocytes ( purple ). ( C ) ΔAUC values for genes with dual alternative termini across tissues. Error bars represent standard error across 100 samples of reads across tissues. t-test * p-value < 10 −16 . ( D ) Conservation scores (mean phastCons score, y-axis ) in a 400nt region around each terminal site of the two most highly expressed isoforms for genes with dual alternative termini. K-S test * p-value < 10 −3 ; ** p-value < 10 −7 . ( E ) Percentage of dual alternative termini genes ( y-axis ) whose isoforms overlap different annotated protein domains. t-test * p-value < 10 −7 ; ** p-value < 10 −16 . (F) CRISPR modulation of a given first exon drives concordant changes in the corresponding last exon of the same ordinal position in three protein-coding genes expressed in HEK293T-A2 cells. Changes in exon expression were quantified relative to control samples ( Methods ). Error bars depict the standard error of means. A t-test measured whether the corresponding last exon of the same ordinal position exhibited the larger directional change than the non-corresponding last exon. ( left ) CRISPR activation of ZNF638 AFE1 resulted in an increase in ALE1, ( middle ) CRISPR activation of MAST1 AFE2 resulted in an increase in ALE2, and ( right ) CRISPR interference of SWI5 AFE2 resulted in a decrease in ALE2. * p-value < .05, ** p-value <.01, *** p-value < 0.001.

Article Snippet: For the second batch ( MAST1 , ZNF638 , and control RNA), extracted RNA was sent to Novogene Corporation for polyA enriched mRNA library preparation and high-throughput sequencing on a NovaSeq 6000 with 150nt paired-end reads.

Techniques: Expressing, CRISPR, Control, Activation Assay

Journal: Science (New York, N.Y.)

Article Title: mRNA initiation and termination are spatially coordinated

doi: 10.1126/science.ado8279

Figure Lengend Snippet: ( A ) Distribution of the lengths ( y-axis ) of dual alternative termini genes ( left ) and PITA pre-mRNAs ( right ) for genes within bins of Spearman’s ρ values ( colors ). ( B ) Distribution of the maximum distances ( y-axis ) between TSSs ( left ), the downstream-most TSS and upstream-most PAS (internal pre-mRNA, middle ), and PASs ( right ) for dual alternative termini genes within varying bins of Spearman’s ρ values ( colors ). ( C ) Distribution of the change in feature lengths between human and mouse orthologs ( y-axis ) for genes that are not PITA in either species, PITA only in mice, or PITA only in humans. To account for global differences in gene lengths between species, distances were first normalized by the mean distance within each species for each feature. Kolmogorov-Smirnov Test * p-value < 10 −3 ; ** p-value < 10 −7 . ( D ) Correlation between the TSS intervals ( x-axis ) and PAS intervals ( y-axis ) for dual alternative termini genes non-PITA genes ( grey , Spearman ρ < 0.3, Pearson’s r = 0.59; p-value < 2.2 × 10 −16 ) and PITA genes ( blue , Spearman ρ >= 0.3, Pearson’s r = 0.82; p-value < 2.2 × 10 −16 ). ( E ) ΔAUC values binned by distances between FEs or PASs ( black and grey , respectively) across pairwise comparisons between mRNA isoforms. Error bars represent bootstrapped 95% confidence intervals.

Article Snippet: For the second batch ( MAST1 , ZNF638 , and control RNA), extracted RNA was sent to Novogene Corporation for polyA enriched mRNA library preparation and high-throughput sequencing on a NovaSeq 6000 with 150nt paired-end reads.

Techniques:

Journal: Science (New York, N.Y.)

Article Title: mRNA initiation and termination are spatially coordinated

doi: 10.1126/science.ado8279

Figure Lengend Snippet: Our results support a model in which longer PITA genes (right) exhibit faster elongation rates when transcription initiates at downstream TSSs. Faster RNAPII trafficking persists through upstream weaker PASs, leading to skipping of these sites in favor of downstream stronger PASs.

Article Snippet: For the second batch ( MAST1 , ZNF638 , and control RNA), extracted RNA was sent to Novogene Corporation for polyA enriched mRNA library preparation and high-throughput sequencing on a NovaSeq 6000 with 150nt paired-end reads.

Techniques: