Enhancers regulate 3′ end processing activity to control expression of alternative 3′UTR isoforms

Kwon, Buki; Fansler, Mervin M.; Patel, Neil D.; Lee, Jihye; Ma, Weirui; Mayr, Christine

doi:10.1038/s41467-022-30525-y

Download PDF

Article
Open access
Published: 17 May 2022

Enhancers regulate 3′ end processing activity to control expression of alternative 3′UTR isoforms

Nature Communications volume 13, Article number: 2709 (2022) Cite this article

8357 Accesses
15 Citations
21 Altmetric
Metrics details

Subjects

Abstract

Multi-UTR genes are widely transcribed and express their alternative 3′UTR isoforms in a cell type-specific manner. As transcriptional enhancers regulate mRNA expression, we investigated if they also regulate 3′UTR isoform expression. Endogenous enhancer deletion of the multi-UTR gene PTEN did not impair transcript production but prevented 3′UTR isoform switching which was recapitulated by silencing of an enhancer-bound transcription factor. In reporter assays, enhancers increase transcript production when paired with single-UTR gene promoters. However, when combined with multi-UTR gene promoters, they change 3′UTR isoform expression by increasing 3′ end processing activity of polyadenylation sites. Processing activity of polyadenylation sites is affected by transcription factors, including NF-κB and MYC, transcription elongation factors, chromatin remodelers, and histone acetyltransferases. As endogenous cell type-specific enhancers are associated with genes that increase their short 3′UTRs in a cell type-specific manner, our data suggest that transcriptional enhancers integrate cellular signals to regulate cell type-and condition-specific 3′UTR isoform expression.

Improving prime editing with an endogenous small RNA-binding protein

Article Open access 03 April 2024

Nuclear mRNA decay: regulatory networks that control gene expression

Article 18 April 2024

Natural antisense transcripts as versatile regulators of gene expression

Article 17 April 2024

Introduction

Most of our knowledge on gene expression regulation has been gained through analysis of genes that generate mRNAs with constitutive 3′UTRs, meaning that their pre-mRNAs are processed into mRNA isoforms with single 3′UTRs. This class of genes contains widely expressed housekeeping genes but also the majority of developmentally regulated genes whose transcription is switched on or off in a cell type-specific manner^1,2,3. Cell type-specific gene expression is known to be regulated by transcriptional enhancers^1,2,4,5,6. However, approximately half of human genes use alternative cleavage and polyadenylation (APA) to generate mRNA isoforms that encode the same protein but differ in their 3′UTR sequence³. The majority of these genes are widely expressed, but they are characterized by tissue- and cell type-specific expression of specific 3′UTR isoforms. These genes are enriched in regulatory factors, including transcription factors, RNA-binding proteins, kinases, and ubiquitin enzymes³. However, their mode of regulation is largely unknown, and it is currently unclear how cell type-specific expression of individual mRNA isoforms with unique 3′UTRs is achieved.

APA is developmentally regulated and can be dysregulated in disease^7,8. Inclusion of different regulatory elements in 3′UTRs influences mRNA stability, translation, and localization^9,10. A difference in 3′UTR sequence can also determine protein function as alternative 3′UTRs allow newly translated proteins to participate in alternative protein complexes^{11,12,13,14,15}. Alternative 3′UTR isoform usage is regulated by RNA-binding proteins, including polyadenylation and splicing factors, but also by factors that influence transcription elongation^{16,17,18,19,20,21,22,23,24,25}. Knock-down of polyadenylation factors often changes 3′UTR isoform usage of hundreds of genes, but genome-wide analyses of 3′UTR isoform expression across cell types and conditions point to a more fine-grained regulation of APA^3,7,8. It is currently largely unknown how alternative 3′UTRs of individual genes are regulated in a gene- and condition-specific manner.

According to the original definition, transcriptional enhancers are DNA sequences that increase the expression of a reporter gene^4,5,6,26. Currently, increased gene expression is often used interchangeably with increased transcription, thus implying that enhancers mostly affect transcript production^6,27. However, the generation of mature mRNAs requires pre-mRNA production and processing which includes splicing and 3′ end cleavage and polyadenylation (CPA)^8,28. Therefore, when disregarding the contribution of mRNA stability, mRNA production of unspliced transcripts is largely determined by the number of produced transcripts and by the 3′ end processing activity that we call here CPA activity.

Under physiological conditions, it is currently difficult to disentangle the contribution of transcript production and transcript processing to the expression level of single-UTR genes. However, viral infection or osmotic stress impair the transcript processing of cellular genes at a large scale. This leads to massive read-through transcription downstream of polyadenylation signals (PAS), thus illustrating the crucial contribution of transcript processing^29,30,31,32. Moreover, point mutations or genetic variants that occur in PAS or in their surrounding sequence elements reveal the contribution of 3′ end processing activity to mRNA expression^{33,34,35,36,37,38,39,40}. Such mutations result in 1.5 to 2-fold differences in steady-state mRNA levels which is sufficient to cause disease phenotypes, including thalassemia, thrombophilia, or cancer predisposition^{33,34,35,37,38,40}.

Here, we set out to investigate if cell type- or condition-specific expression of 3′UTR isoforms is regulated by transcriptional enhancers. PTEN is a multi-UTR gene and we found that deletion of the endogenous PTEN enhancer did not reduce transcript production but impaired CPA activity at a proximal and intrinsically weak PAS. The enhancer-dependent processing activity regulation was mediated by transcription factors, transcription elongation factors, and chromatin modifiers. Enhancer-mediated regulation of 3′UTR isoform expression is widespread as endogenous, cell type-specific enhancers significantly associate with genes that exclusively upregulate their 3′UTR isoform expression in a cell type-specific manner. Our data indicate that transcriptional enhancers regulate both aspects of mature mRNA generation, namely transcript production and 3′ end processing to regulate mRNA and mRNA isoform expression in a cell type- and condition-specific manner.

Results

The PTEN enhancer induces a 3′UTR isoform switch of endogenous PTEN

PTEN is a tumor-suppressor gene whose expression is altered in a large fraction of cancers. Cells are very sensitive to PTEN dosage as even a small decrease in PTEN expression is cancer-promoting⁴¹. The PTEN gene generates multiple mRNA isoforms with alternative 3′UTRs that encode the same protein. In our previous 3′UTR isoform expression study, PTEN was among the top genes with extensive differences in alternative 3′UTR isoform usage across cell lines and tissues³. To obtain a better understanding of PTEN expression regulation, we applied CRISPR technology to delete the promoter-proximal PTEN enhancer in the breast cancer cell line MCF7 which expresses wild-type PTEN⁴². The boundaries of the PTEN enhancer were determined using ChIP-seq data on transcription factor binding sites and acetylated H3K27 levels (Fig. 1a)^43,44,45,46. We used a pair of guide RNAs to delete the PTEN enhancer. We obtained two control clones with the wild-type (WT) enhancer sequence and two “delta enhancer” (dE) clones with a heterozygous deletion in the region of the PTEN enhancer (Fig. 1b and Supplementary Fig. 1a, b).

Heterozygous deletion of the enhancer increased steady-state PTEN mRNA level by only 1.17-fold and did not affect protein level (Fig. 1c, d). We hypothesized that enhancer activation may be necessary to observe an effect. The PTEN enhancer contains canonical NF-κB binding sites (Fig. 1a)⁴⁷. As cytoplasmic acidification was previously reported to increase NF-κB activity in MCF7 cells⁴⁸, we cultivated the cells in acidified media (pH = 6.5) and measured NF-κB activity by blotting for phosphorylated transcription factor p65 (which is encoded by the RELA gene). Although cultivation of cells in acidified media increased phosphorylated p65 (Fig. 1e and Supplementary Fig. 1c), we did not observe an enhancer-mediated difference in PTEN mRNA and protein levels (Fig. 1c, d).

Promoters were previously implicated in the regulation of mRNA processing^{49,50,51,52,53,54,55,56,57,58,59}. Therefore, we investigated if deletion of the enhancer would change alternative 3′UTR isoform expression of PTEN. Enhancer deletion had little effect on 3′UTR isoform expression under normal cultivation conditions (Fig. 1f and Supplementary Fig. 1d). However, in acidified conditions, we observed a striking switch in 3′UTR isoform expression with increased expression of the short 3′UTR (SU) isoform of PTEN which was fully abrogated in cells with heterozygous deletion of the enhancer (Fig. 1f and Supplementary Fig. 1d).

As the PTEN enhancer overlaps with the KLLN gene, we tested the influence of the KLLN gene on PTEN 3′UTR isoform expression. shRNA-mediated knock-down (KD) of the KLLN gene did not influence PTEN mRNA or 3′UTR isoform expression (Supplementary Fig. 1e–g). This supports our result that the PTEN enhancer is responsible for the 3′UTR isoform change. As cultivation of MCF7 cells in acidified media activates NF-κB, we tested if KD of the transcription factor p65 influences PTEN 3′UTR isoform expression. KD of p65 did not affect steady-state PTEN mRNA levels, but it abrogated the switch in 3′UTR isoform expression (Fig. 1g, h and Supplementary Fig. 1h). This result demonstrates that a transcription factor that binds to the PTEN enhancer is responsible for the 3′UTR isoform change of endogenous PTEN.

Mechanistically, the switch in 3′UTR isoform expression is either caused by a change in alternative PAS usage or by preferential degradation of the long 3′UTR (LU) isoform. However, degradation of the LU isoform would need to be accompanied by increased transcript production to account for the upregulated SU expression. To identify the mechanism by which the enhancer controls the switch in 3′UTR isoform expression, we measured transcript production and stability. We inhibited transcription with actinomycin D and performed northern blot and qRT-PCR analysis at different time points to measure stability of the alternative 3′UTR isoforms. We did not detect an enhancer- or condition-specific difference in stability of the mRNA isoforms (Fig. 1i and Supplementary Fig. 1i). Next, we used metabolic labeling with 4-thiouridine to examine an enhancer-dependent change in transcript production before and after media acidification. We did not observe a significant difference in PTEN pre-mRNA production between WT and dE mutant cells or between normal and acidified conditions (Fig. 1j). Enhancers are further known to regulate alternative splicing⁵⁸, but we did not observe enhancer-dependent alternative splicing of PTEN (Supplementary Fig. 1j). As the PTEN enhancer was required for a switch in 3′UTR isoform expression without substantially regulating transcript production or causing differential stability of the alternative mRNA transcripts, our data suggest that it regulates CPA activity.

The PTEN enhancer increases CPA activity of intrinsically weak PAS in a reporter system

Processing activity of endogenous transcripts cannot be fully disentangled from transcript production and stability. To investigate if transcriptional enhancers indeed control CPA activity, we developed a reporter system. Our luciferase reporter system allows us to separately investigate enhancer-dependent transcript production and transcript processing. The PAS derived from the SV40 late transcript is one of the strongest known PAS^60,61. When used for termination of a luciferase reporter construct it results in processing of all produced transcripts (Fig. 2a)^60,61. Therefore, the SV40 PAS reporter construct measures transcriptional activity of the promoter, a system that has been widely used to measure transcriptional activity⁶². To assess enhancer-dependent CPA activity, we measured luciferase activity of a reporter construct that is nearly identical except that it is terminated by the proximal PAS (PPAS) of PTEN instead of the SV40 PAS. As only processed transcripts contribute to luciferase activity, the ratio of luciferase activities obtained from the PTEN PPAS reporter over the SV40 PAS reporter represents the relative CPA activity of the PTEN PPAS when driven by a specific promoter (Fig. 2a). In this reporter system, CPA activity corresponds to luciferase activity if the different polyadenylation sites do not influence stability of the reporters. To minimize the elements that may affect mRNA stability we used minimal PAS that only contain ~100 base pairs of surrounding sequence to ensure proper 3′ end processing (Supplementary Table 1a)²⁸.

**Fig. 2: The *PTEN* enhancer increases CPA activity of proximal PAS.**

We measured transcriptional activity of the PTEN promoter (Pprom) and observed a slight decrease in transcriptional activity (1.6-fold) in the presence of the enhancer (Penh-Pprom; Fig. 2b). The decrease may be due to the high transcriptional activity of the isolated PTEN promoter and the fact that some transcription factors act as repressors. However, typical enhancers increase transcriptional activity^6,26. When we analyzed the effect of the PTEN enhancer in the context of two weaker core promoters, it indeed increased transcriptional activity, indicating that it acts as a transcriptional enhancer following the original definition (Supplementary Fig. 2a, b and Supplementary Table 1b)^6,26. Intriguingly, the addition of the enhancer increased luciferase activity of the PTEN PPAS reporter 4-fold (Fig. 2c).

To investigate if the increase in luciferase activity is the result of an enhancer-dependent increase in CPA activity, we measured transcript abundance upstream and downstream of the PAS which enables detection of potential differences in the levels of read-through transcripts in the four reporter constructs (Fig. 2d). We compared read-through transcripts of a pair of reporters that are transcribed from the same promoter. We observed a similar amount of read-through transcripts downstream of the SV40 PAS and the PTEN PPAS when the reporters were transcribed from the Penh-Pprom constructs (Fig. 2e). However, we observed a three-fold higher amount of read-through transcripts downstream of the PTEN PPAS compared to the SV40 PAS when the reporters were transcribed from the Pprom promoter (Fig. 2e). This result supports enhancer-dependent regulation of CPA activity of the weak PTEN PPAS.

We performed additional control experiments to gain a better understanding of enhancer-dependent reporter regulation. We integrated the reporter into the genome using Flp-in cells and also observed a significant increase in CPA activity in the presence of the PTEN enhancer (Supplementary Fig. 2c, d). We did not observe enhancer- or PAS-dependent differences in mRNA stability of the transfected reporters, indicating that the difference in luciferase activity correlates with CPA activity (Supplementary Fig. 2e). The addition of the enhancer did not change the transcription start site of the reporter (Supplementary Fig. 2f), indicating that the mature mRNAs produced from the PTEN promoter in the presence or absence of the PTEN enhancer are identical. As circular plasmids were transfected, we measured rolling-circle transcription, as this may affect transcriptional or processing activity. We did not observe a significant difference in rolling-circle transcription of the plasmids that contain or lack the enhancer (Supplementary Fig. 2g). Enhancers regulate transcription independently of their orientation and can be located up- or downstream of genes^6,46. When we placed the reverse complement of the PTEN enhancer downstream of the PAS, it enhanced transcription and PAS cleavage (Supplementary Fig. 2h–j). Taken together, these data suggest that regulation of 3′ end processing activity is a bona fide activity of transcriptional enhancers.

Next, we assessed if the enhancer controls processing activity of additional PAS (Supplementary Table 1a). We tested two PPAS (derived from NUDT21 and DICER1), two distal PAS (DPAS; derived from PTEN and NUDT21), and two PAS derived from housekeeping genes (GAPDH and UBC) that generate constitutive 3′UTRs³. The rules that determine the intrinsic strength of PAS were unknown at the time of PAS selection^63,64. However, the chosen proximal PAS are supposed to be weak as they contribute to less than 30% of SU isoform expression in most cell types analyzed³. PAS strength of distal or single-UTR genes cannot be inferred from isoform expression, but we expected them to be stronger as these PAS need to make sure that mature mRNAs are produced which is consistent with PAS scores estimated by a neural network^40,63,64.

We observed that the PTEN enhancer increased CPA activity of different proximal PAS by up to 3.6-fold (Fig. 2f). The enhancer also influenced CPA activity of two out of four non-proximal PAS (Fig. 2f). However, the activity change of non-proximal PAS was not consistent across them and cannot be predicted with our current knowledge. Therefore, we did not include them in further experiments. Our observations suggest that CPA activity of weak PAS may be low in the absence of an enhancer, but their processing activity can increase in vivo when transcribed from promoters with active enhancers.

Promoter types determine if enhancers regulate transcription of CPA activity

We then asked if other enhancers are also capable of regulating CPA activity and set out to test the enhancer of the NUDT21 gene. The NUDT21 gene encodes an important polyadenylation factor that changes CPA of hundreds of genes, when knocked-down^18,20. Moreover, the NUDT21 gene undergoes APA and similar to PTEN, its alternative 3′UTR isoforms are extensively regulated across samples³. We searched for ChIP-seq peaks with high H3K27 acetylation level in the vicinity of the NUDT21 gene, as high H3K27 acetylation levels are characteristic for enhancers^6,46. H3K27 acetylation levels in the promoter-proximal region of NUDT21 were only intermediate, but we detected a region with very high acetylation levels 80 kb downstream of the NUDT21 gene. We cloned 2 kb of this region and called it distal enhancer (Denh; Fig. 3a, b and Supplementary Table 1c) as we currently have no evidence that this region is an enhancer of the NUDT21 gene. To test if the distal enhancer is functional, we measured enhancer-dependent transcriptional activity of the GAPDH promoter (Gprom) which drives expression of a single-UTR gene³. The distal enhancer upregulated transcriptional activity by more than 4-fold, thus acting as a transcriptional enhancer (Fig. 3c). However, the distal enhancer had no influence on CPA activity in the context of the Gprom (Fig. 3d). Similar results were obtained when using the PTEN enhancer in the context of the Gprom (Supplementary Fig. 2b, k).

**Fig. 3: A distal enhancer regulates CPA activity of proximal PAS.**

In contrast, in the context of two promoters derived from multi-UTR genes, such as the PTEN or NUDT21 promoters (Nprom), the distal enhancer did not significantly affect transcriptional activity (Fig. 3b, e), but instead increased CPA activity of the PTEN PPAS between 3.4 and 5.3-fold (Fig. 3f, g) and CPA activity of the NUDT21 PPAS by 1.8-fold (Supplementary Fig. 2l, m). As the distal enhancer did not affect processing activity of a stronger PAS, our results indicate that transcriptional enhancers regulate CPA activity of weak proximal polyadenylation sites in the context of multi-UTR promoters (Fig. 3f, g).

Transcription factors and transcription elongation factors are widespread regulators of CPA activity

As KD of p65 changed 3′UTR isoform expression of the endogenous PTEN gene, we set out to identify additional transcription factors that may mediate enhancer-dependent regulation of CPA activity. Binding of transcription factors to enhancers results in the recruitment of co-activators to promoters which includes components of the mediator complex, the general transcription machinery, chromatin remodelers, transcription elongation factors, and histone acetyltransferases^65,66,67. We performed a small-scale shRNA screen and knocked-down individual transcription factors or co-activators expressed in MCF7 cells (Table 1 and Supplementary Fig. 3). All sequence-specific transcription factors that were knocked-down were shown by ChIP-seq to bind to the PTEN enhancer (Fig. 4a)⁴⁴. We measured transcriptional and CPA activity in the context of the Penh-Pprom in control (ctrl) KD and transcription factor KD samples (Fig. 4b). As positive control, we knocked-down the CPA factor FIP1L1, which was shown previously to be required for PPAS usage¹⁷. KD of FIP1L1 decreased PTEN PPAS usage from 0.6 to 0.36 without affecting transcriptional activity (Fig. 4c, Table 1, and Supplementary Fig. 4a, b).

Table 1 Transcription factors and co-activators that regulate CPA activity of the proximal PAS of PTEN.

Full size table

**Fig. 4: Transcription and transcription elongation factors are widespread regulators of CPA activity.**

Seventeen out of 22 tested sequence-specific transcription factors significantly changed CPA activity (Fig. 4c and Table 1). Four of them appear to be CPA repressors as their KD increased CPA activity, but suppression of the majority of transcription factors decreased CPA activity of the PTEN PPAS (Fig. 4c). Several transcription factors predominantly changed CPA activity rather than transcriptional activity in this assay. They include RELA (NF-κB p65), MYC, RXRA, and FOX1 (Fig. 4c and Table 1). KD of general transcription factors led to a binary pattern as their suppression only affected transcriptional activity or strongly repressed CPA activity, as was observed for TFIIF, TBP, TAF1, and MED1 (Fig. 4c and Table 1). Downregulation of histone acetyltransferases, chromatin remodelers, or transcription elongation factors mostly repressed CPA activity and the repressive effect was similar in strength to KD of the CPA factor FIP1L1 (Fig. 4c and Table 1). Taken together, we found that a large fraction of tested transcription factors, transcription elongation factors, and chromatin modifiers significantly influences CPA activity in the context of a reporter.

Mutation of MYC-binding sites in the enhancer decreases CPA activity

RELA and MYC KD also decreased CPA activity of additional PPAS but did not significantly affect cleavage activity of the strong PTEN DPAS (Fig. 4d and Supplementary Fig. 4c). Similar results were obtained after mutation of the two highly conserved MYC-binding sites in the PTEN enhancer (Fig. 4d, e). Mutation of the MYC-binding sites had no significant influence on transcriptional activity (Supplementary Fig. 4d), did not affect CPA activity of the stronger distal PAS of PTEN, but decreased CPA activity of weaker proximal PAS, thus phenocopying the effect of MYC KD (Fig. 4d). These results suggest that binding of transcription factors to conserved motifs in the PTEN enhancer regulates CPA activity of weak proximal PAS.

CPA activity is regulated by active enhancers

Enhancer activation often results in histone acetylation and KD of histone acetyltransferases such as TIP60 and PCAF (encoded by KAT5 and KAT2B, respectively) interferes with enhancer activation⁶⁵. Their shRNA-mediated suppression reduced CPA activity of several proximal PAS in the context of the Penh-Pprom reporter (Supplementary Fig. 4e). KD of histone acetyltransferases also decreased CPA activity in the context of the distal enhancer but had no effect on CPA activity in reporters that lack the enhancers (Supplementary Fig. 4f–i). These results suggest that regulation of CPA activity of proximal PAS by active enhancers has the potential to be widespread as two out of two tested enhancers regulated 3′ end processing activity in the context of reporters.

Cell type-specific enhancers preferentially associate with genes that upregulate SU isoforms in a cell type-specific manner

Next, we set out to investigate if endogenous enhancers are widespread regulators of CPA activity. The analysis of multi-UTR genes allows us to distinguish transcriptional and CPA activity. Transcriptional upregulation will increase SU and LU isoform expression to a similar extent, whereas increased CPA activity of proximal PAS will only increase SU, but not LU isoform expression.

To examine if cell type-specific enhancers are associated with changes in 3′UTR isoform expression, we used a dataset that mapped erythroblast-specific enhancers and associated them with individual genes⁶⁸. To identify genes that become upregulated in erythroblasts, we compared gene expression between erythroblasts and hematopoietic stem cells^69,70. We analyzed single- and multi-UTR genes separately, but as expected, we observed that genes that increase their expression in erythroblasts preferentially associate with enhancers that are active in erythroblasts (Fig. 5a, b)⁷¹. The association with erythroblast-specific enhancers was not observed for genes whose expression did not increase (Fig. 5a, b).

**Fig. 5: Cell type-specific enhancers are associated with genes that upregulate SU isoforms.**

Next, we focused on all multi-UTR genes and separated them into two groups based on their gene expression change between erythroblasts and hematopoietic stem cells. The two groups (“gene up”, “gene not up”) were further subdivided to identify among them the genes with significantly increased SU isoform expression (“SU up” with upregulated SU isoform counts and upregulated SU isoform ratio). For the four groups, we visualized their gene expression changes, their SU and LU isoform expression changes, and their 3′UTR isoform ratio changes (Fig. 5c–e). Next, we associated the four groups with cell type-specific enhancers that are active in erythroblasts (Fig. 5f). As expected, the groups with increased gene expression were significantly associated with cell type-specific enhancers (Fig. 5f). Importantly, also genes with exclusive upregulation of their SU isoforms without leading to a significant upregulation in overall gene expression were significantly associated with cell type-specific enhancers (Fig. 5f).

To fully control for the slightly increased gene expression in the “SU up” group compared to the “gene not up” group, we performed stratified random sampling. To make sure that the genes in the control group do not upregulate their SU isoforms, we excluded all genes with a slight upregulation of SU from the control group, thus strongly reducing the number of genes available for comparison (Supplementary Fig. 5a–d). For three stratified random samples, our analysis shows that the fraction of genes associated with erythroblast-specific enhancers is lower in the control group (mean, 12.2%) than in the “SU up” group (mean, 15.4%), but the Chi-square test does not reach statistical significance (Supplementary Fig. 5d). This analysis revealed that it is difficult to disentangle increased gene expression from exclusive SU upregulation as this usually also slightly upregulates gene expression.

Model of enhancer-mediated control of 3′UTR isoform expression

It is well-known that active enhancers upregulate gene expression of single-UTR genes in a cell type-specific manner (Fig. 5g, top)⁷¹. In contrast, multi-UTR genes are usually transcribed in the majority of cell types³. We found that enhancer activation of multi-UTR genes can have three potential outcomes. It can result in upregulation of gene expression with SU and LU isoforms increasing similarly. It can result in a combination of change in gene and isoform expression or it can result in upregulation of SU isoforms without a change in gene expression (Fig. 5g, bottom).

Discussion

Mature mRNA production is determined by the extent of transcript production and transcript processing. Here, we show that transcriptional enhancers regulate both stages of mature mRNA production, but they differentially control them for different classes of genes. At single-UTR genes, enhancers increase transcript production, whereas at multi-UTR genes, transcriptional enhancers increase transcript production or transcript processing, thus resulting in a gene expression change, in a cell type- or condition-specific 3′UTR isoform expression change or in a change of both parameters.

With our newly developed reporter assay, we were able to measure separately the two parameters of mRNA expression. Our results are consistent with the expression pattern of endogenous single- and multi-UTR genes. Whereas single-UTR genes are often transcribed in a cell type-specific manner, meaning that their expression is “off” in some cell types and “on” in others, most multi-UTR genes are always “on” as they are transcribed in the majority of cell types^3,72. Despite their near ubiquitous expression, they encode regulatory factors and show cell type-specific 3′UTR isoform expression³. This expression pattern was mirrored in the reporter assays, where two transcriptional enhancers increase transcript production when the reporter gene is driven by a single-UTR gene promoter. In contrast, the same enhancers did not affect overall transcription but instead increased the expression of a transcript terminated by a weak polyadenylation site when the reporter gene was driven by promoters derived from multi-UTR genes (Fig. 3). As the enhancer-dependent increase in isoform expression was associated with decreased read-through transcription downstream of weak polyadenylation sites, we conclude that enhancers regulate 3′ end processing activity (Fig. 2). Our reporter data are further supported by results obtained at endogenous gene loci, where enhancer deletion of the multi-UTR gene PTEN did not change gene expression but altered 3′UTR isoform expression. Furthermore, KD of the transcription factor p65 which binds to the enhancer did not change steady-state mRNA level but induced a 3′UTR ratio change (Fig. 1). Moreover, in a transcriptome-wide analysis, we found that cell type-specific enhancers were significantly associated with genes that, during differentiation, exclusively upregulated their short 3′UTR isoforms without leading to a significant overall increase in gene expression (Fig. 5). However, as upregulation of SU isoforms usually results in a slight increase in gene expression, SU upregulation and a change in gene expression cannot be fully disentangled at endogenous genes (Supplementary Fig. 5a–d).

Our study further revealed that transcription and transcription elongation factors are responsible for increased transcript processing at weak polyadenylation sites and subsequent upregulation of mRNA isoform expression (Fig. 4). Moreover, as suppression of histone acetyltransferase complexes decreases CPA activity, our data suggest that active enhancers regulate 3′ end processing (Fig. 4). However, the exact mechanism by which enhancers regulate CPA activity is currently unknown as we did not detect a linear relationship between overall gene expression and SU isoform usage across cell types (Supplementary Fig. 5e). Based on the literature, several potential mechanisms exist that are not mutually exclusive. It is established that RNA-binding proteins that bind to the polyadenylation signal and the surrounding sequence determine 3′ end CPA activity^{7,8,16,17,18,19,20,21,22,59,61,73}. One model by which these RNA-binding proteins bind to polyadenylation sites is through the promoter loading model: Active enhancers recruit these factors to promoters which allows them to travel with RNA polymerase II and to bind to a newly transcribed polyadenylation site, thus increasing 3′ end processing activity locally^{49,50,51,52,53,55,56,57,58}. Such a mechanism has been proposed for promoter-dependent regulation of post-transcriptional processes in yeast, including the regulation of mRNA stability, cytoplasmic localization, and translation^{74,75,76,77,78}. This model is supported by the presence of a variety of RNA-binding proteins at 80% of human promoters and by the observation that many RNA-binding proteins bind to transcription factors^55,56.

Another model suggests that active enhancers regulate transcription elongation rate which could result in differential usage of polyadenylation sites^{23,24,79,80,81}. This is supported by our data showing that silencing of several factors involved in transcription elongation, including SPT4/5, NELF, and the PAF complex affect CPA activity of proximal polyadenylation sites (Table 1)⁸². Finally, the integrator complex has been shown to associate with active enhancers and increases enhancer-promoter communication⁸³. Decreased expression of INTS11, the catalytic subunit of the complex, promotes read-through transcription at polyadenylation sites and shifts alternative isoform expression towards the distal isoform^32,84. This suggests that increased association of integrator at active enhancers could prevent read-through and could increase CPA activity. Integrator is also known to regulate transcription elongation^85,86, but it is currently unclear if its role in transcription elongation is required for integrator-dependent regulation of 3′ end processing of protein-coding genes.

Currently, the in vivo usage of a given PAS cannot be predicted. Intrinsically weak sites, meaning PAS that are surrounded by a poor sequence context may get upregulated upon binding of suitable factors, whereas intrinsically strong PAS may have a lower potential for regulation. This is supported by our reporter results which revealed that enhancers or transcription factors upregulate CPA activity of 6/6 tested proximal PAS, whereas their influence on distal or single-UTR PAS was unpredictable (Fig. 2f). This work, together with work from others suggests that in vivo usage of PAS is regulated by the sequence context as well as by transcription factors, transcription elongation factors, chromatin regulators and expression levels of RNA-binding proteins, including CPA factors^{7,8,23,24,25,40,82}.

Although APA is widespread and is regulated in a cell type- and condition-specific manner, for most genes the consequences of a change in alternative 3′UTR isoform expression are unknown^3,7,8. Nevertheless, striking examples exist in the literature that revealed that altering alternative 3′UTR isoform expression can result in substantial changes in protein expression caused by 3′UTR-dependent control of mRNA stability or translation^36,87,88. However, alterations in alternative 3′UTR isoform expression often do not change overall protein levels (Fig. 1)^{12,14,20,89,90}. For those cases, it has been demonstrated that isoform-specific differences in protein localization or function occur through 3′UTR-mediated formation of alternative protein complexes^{11,12,13,14,15}. For example, the ubiquitin ligase BIRC3 switches from predominant SU isoform expression in normal B cells to LU isoform expression in malignant B cells. Normal B cells mostly form protein complexes that are independent of LU isoforms and they mediate BIRC3’s tumor-suppressive functions. In contrast, malignant B cells preferentially form long 3′UTR-dependent protein complexes that have tumor-promoting roles¹². So far, most studies that investigated the functional consequences of alternative 3′UTR isoform expression have relied on using expression constructs, but more recently CRISPR-mediated deletions of 3′UTRs were added to the tool kit^{11,12,13,14,15,28,89,90,91}. Alternatively, as we showed here, 3′UTR isoform expression can be altered through enhancer deletion (Fig. 1). This strategy will allow researchers to study the resulting functional consequences at endogenous gene loci and has the advantage of keeping 3′UTR cis-elements intact while only changing the relative expression of 3′UTR isoforms. Although we only generated a heterozygous deletion of the PTEN enhancer, it was sufficient for detecting a molecular phenotype as shown by a switch in 3′UTR isoform expression. The switch induced by the heterozygous deletion was only partial. In an ideal case, we expect that enhancer activation will result in SU-only expression, whereas full deletion of the enhancer will result in LU-only expression.

3′UTR length has expanded substantially during evolution of more complex animals and it correlates with the number of cell types observed in an organism^92,93. At the same time, the number of enhancers has increased with organismal complexity¹. Therefore, we speculate that increased regulation by enhancers has co-evolved with increased regulation by 3′UTRs to integrate intrinsic and extrinsic signals to change gene and mRNA isoform expression².

Methods

Cells used

For all experiments, the human breast cancer cell line MCF7 was used which was a gift from the laboratory of Robert Weinberg (Whitehead Institute, Cambridge, USA). For the experiments with luciferase reporters integrated into the genome, the MCF7/FRT cell line was used which was a gift from the laboratory of Reuven Agami (Netherlands Cancer Institute, Amsterdam, Netherlands)⁹⁵. All cells were cultured in DMEM supplemented with 10% FCS and 1% Pen-strep (normal condition). For experiments using acidified media, MCF7 cells were cultured for 24 h in DMEM supplemented with HCl (pH = 6.5). MCF7 cells were transfected with Lipofectamine 2000 (Invitrogen).

PTEN enhancer deletion using CRISPR-Cas9

All primers are listed in Supplementary Table 2. The guide RNAs targeting the 5′ and 3′ ends of the PTEN enhancer were designed using GuideScan and cloned into pX330 as previously described^96,97. MCF7 cells were transfected with pmaxGFP and the two pX330 plasmids. After 4 days, GFP-positive cells were sorted into 96-well plates and grown into single cell colonies. Enhancer deletion was tested by PCR and sequencing.

Screening of cells with PTEN enhancer deletion

Genomic DNA was isolated from each clone and the deletion of the PTEN enhancer was assessed by PCR with primers (PcE-FS2 and PcE-RS2) flanking the enhancer region and Sanger sequencing. As MCF7 cells are not diploid at the PTEN locus, the presence of the alleles with PTEN enhancer sequences was further assessed by qPCR using genomic DNA isolated from WT and dE cells to amplify two regions outside of the PTEN enhancer (chr10:89,618,718-89,618,843 and chr10:89,620,263-89,620,387) and two regions inside PTEN enhancer (chr10:89,621,562-89,621,699 and chr10:89,621,320-89,621,440) using FastStart universal SYBR green master mix (Roche). The qPCR results from the amplicons inside the PTEN enhancer were normalized to those from the amplicons outside of the PTEN enhancer and were compared between WT and enhancer deletion cells to determine the extent of enhancer loss.

ShRNA-mediated knock-down

For the knock-down of RELA and KLLN, lentiviral vectors (pLKO-puro) containing shRNAs were purchased from Sigma (RELA, TRCN0000014684, TRCN0000014686; KLLN, TRCN0000339049, TRCN 0000339051). As control, an shRNA with scrambled sequence was used (Addgene, #1864)⁹⁸.

Northern blotting

Northern blotting was performed as described previously with modifications⁸⁷, dx.doi.org/10.17504/protocols.io.bqqymvxw. Briefly, total RNA was isolated using Tri reagent (Invitrogen). PolyA+ mRNA was purified with Oligotex (Qiagen) and 2 µg of polyA+ mRNA was loaded in each lane. The single-stranded probe was generated by unidirectional PCR reaction as described⁹⁹ with a slight modification. As input for the unidirectional PCR 30 ng of DNA template was used. This consisted of a 724-bp fragment of the PTEN coding region amplified from MCF7 cDNA (PTEN-NB-F and PTEN-NB-R). The unidirectional PCR reaction (20 µl) was conducted in buffer (100 mM Tris-HCl/pH 8.3, 500 mM KCl, 15 mM MgCl₂, 1% Triton-X-100) containing 0.2 mM each of dCTP, dGTP, and dTTP, 0.5 unit of Taq polymerase, the reverse primer from above, and 6 µl of 3000 Ci/mmol [α-³²P]dATP (Perkin Elmer). The reaction mixture was initially boiled for 10 min at 95 °C and subjected to 35 thermal cycles (95 °C for 30 s; 45 °C for 30 s; 72 °C for 1 min), which was followed by 5 min incubation at 72 °C. After the PCR reaction, 5 µl of 0.2 mM EDTA was added to the mixture and boiled for 5 min at 95 °C, followed by 2 min chilling on ice. The PCR product was then used for hybridization to probe the PTEN transcript and the blot was scanned by Fuji phosphorimager.

Quantification of endogenous PTEN transcripts

For qRT-PCR, total RNA was isolated using Tri reagent (Invitrogen). cDNA synthesis was performed using qScript cDNA supermix (Quantabio), followed by qRT-PCR was performed using FastStart universal SYBR green master mix (Roche). RPL19 mRNA was used as loading control.

To quantify steady-state total PTEN mRNA level, primers PTEN-qP-F and PTEN-qP-R were used. To quantify the fraction of PTEN-LU over total PTEN mRNA, PTENLU-qP-F and -R primers were used for RT-qPCR. The relative abundance of PTEN-LU using this primer pair corresponds to the relative abundance of PTEN-LU obtained by northern blot analysis (e.g., Fig. 1f). To obtain the fraction of PTEN-LU, the abundance of PTEN-LU was normalized to the abundance of total PTEN mRNA detected by primers PTEN-qP-F and -R.

Half-life of PTEN-LU

MCF7 WT and dE2 cells were treated with or without 25 mM HCl overnight. The cells were then treated with either DMSO or 4 μg/ml actinomycin D for the indicated time points. The half-lives of PTEN-LU mRNA in each condition were determined as described previously^100,101. Briefly, the amounts of PTEN-LU mRNA remaining were determined at each time point relative to that at 0 h. The relative amounts of PTEN-LU mRNA over time were subjected to nonlinear regression, given that the amount of PTEN-LU remaining at a certain time point can be calculated by the equation: remaining amount = e^kt (where k and t are decay rate and time, respectively). The half-lives were then calculated by the equation: half-life = ln(0.5)/k.

PTEN splicing pattern

A primer pair (PTEN-sp-F and PTEN-sp-R) located in the first and last exon of PTEN (NM_000314) was used on the cDNA described above and visualized on 1% agarose gels.

Transcript production rate for endogenous PTEN

The nascent transcripts were extracted following the protocol described by ref. ¹⁰² with minor modifications. Briefly, MCF7 cells were treated with 400 µM 4-thiouridine for 2 h (4sU, Fisher Scientific). After labeling, Tri reagent (Invitrogen) was used to extract total RNA from cells, followed by DNase I treatment (NEB). Next, 50 µg of total RNA was mixed with 0.5 µg of 4sU-labeled yeast RNA as a spike-in. The mixture was biotinylated with 10 µg of MTSEA biotin-XX (Biotium) in 100 mM HEPES/pH 7.5 and 10 mM EDTA for 30 min at 25 °C in dark with gentle agitation and the biotinylated RNA was recovered by ethanol precipitation. Then, half of the biotinylated RNA was mixed with 100 µM DTT and retained as the total RNA fraction. The other half was incubated with Dynabeads M-280 Streptavidin (Invitrogen) in bead wash buffer (100 mM Tris-HCl/pH 7.5, 1 M NaCl, 0.1% Tween-20, 10 mM EDTA), washed twice and eluted twice in 100 µl of 0.1 M DTT at 25 °C for 5 min, followed by ethanol precipitation. cDNA generated using Superscript III reverse transcriptase (Invitrogen) and random hexamers (Invitrogen). PTEN transcript abundance was measured by qPCR using an amplicon in the first intron of the PTEN gene (PTEN-int1-qP-F and PTEN-int1-qP-R) normalized to the abundance of yeast ACT1 transcripts from the spike-in RNA (using primers yACT1-qP-F and yACT1-qP-R). The rate of transcript production was determined by the ratio of the abundance of the nascent PTEN transcripts to the total PTEN transcripts.

Western blotting

Cell pellets were lysed in 2x Laemmli buffer (Alfa Aesar) and performed as described previously¹² with the following antibodies: anti-PTEN (A2B1, Santa Cruz Biotechnology, sc-7974, 1:1000), anti-GAPDH (V-18, Santa Cruz Biotechnology, sc-20357, 1:500), anti-P65 (L8F6, Cell Signaling Technology, 6956, 1:1000), and anti-phospho-P65 Ser536 (93H1, Cell Signaling Technology, 3033, 1:1000). As secondary antibodies anti-mouse IRDye 800 (1:5000; Li-Cor Biosciences, Cat# 926-68072), anti-rabbit IRDye 680 (1:5000; Li-Cor Biosciences, Cat# 926–6807), and anti-goat IRDye 680 (1:5000; Li-Cor Biosciences, Cat# 926-32224) were used. The blot was analyzed by Odyssey CLx imaging system (Li-Cor).

Luciferase reporter assays

Luciferase reporter constructs

All luciferase constructs were derived from PIS1 vector⁸⁷. It contains the thymidine kinase promoter of Herpes simplex virus, followed by a renilla luciferase open reading frame, followed by the late SV40 PAS. To obtain reporter constructs with different promoters or PAS, the promoter or the SV40 PAS were exchanged using restriction enzyme digest or Gibson cloning. To obtain reporter constructs with different enhancers, the enhancer sequences (Supplementary Table 1) were cloned upstream in the sense orientation of the respective promoters. In the case of Penh1, the reverse complement of the sequence was cloned downstream of the PAS. MYC-binding sites (E-boxes with the sequence CACGTG at positions −1149 bp and −1353 bp upstream of the transcription start site of the PTEN gene) were mutated to CAAGAA using Quikchange Lightening kit (Agilent). The TATA synthetic promoter sequence was derived from pGL firefly reporter with a minimal promoter.

Luciferase assay

Luciferase assays were performed in 24-well plates as described previously⁸⁷. The number of experiments listed in the figures corresponds to biological replicates. In each well, 100 ng of firefly luciferase control reporter plasmid PISO together with 400 ng of renilla luciferase plasmid were transfected. Same molar amounts of plasmid were transfected to account for different construct sizes (400 ng were used for a plasmid of 5000 bp). Firefly and renilla luciferase activities were measured with the Dual-luciferase assay (Promega) 24 h after transfection using Glomax 96 microplate luminometer and Glomax 96 software (Promega). Renilla activity was normalized to firefly activity to control for transfection efficiency. Transcriptional activity of a promoter corresponds to renilla luciferase activity (normalized by firefly luciferase activity) after transfection of the reporter containing the promoter and the SV40 PAS. CPA activity of a test PAS was obtained by dividing the luciferase activities of the constructs with the test PAS by the SV40 PAS in the context of the same promoter.

To assess CPA activity after knock-down of transcription factors, luciferase constructs were transfected into MCF7 cells stably expressing control (ctrl) shRNAs or shRNAs targeting specific transcription factors or co-activators. PAS usage was calculated as described above. When several shRNAs against a specific factor were available, the results were pooled. The fold change in transcriptional activity upon factor knock-down was obtained by normalization to the transcriptional activity of the SV40 PAS reporter in cells expressing ctrl shRNAs that were performed in parallel.

mRNA stability of the reporter constructs

Luciferase plasmids were transfected as described above. After 24 h, cells were either treated with DMSO or with actinomycin D (4 µg/ml; Sigma) for the indicated time points. Total RNA was extracted using Tri reagent and was used to generate cDNA using qScript qScript cDNA supermix (Quantabio). qRT-PCR was performed using FastStart universal SYBR green master mix (Roche) on a 7500 HT Fast Real-Time PCR System (Applied Biosystems). The primer pairs used to quantify total PTEN mRNA (PTEN-stability-F and -R), luciferase mRNA (Rluc-stability-F and -R) and GAPDH mRNA (GAPDH-F and GAPDH-stability-R) (for normalization) are listed in Supplementary Table 2.

5’ RACE

Renilla luciferase plasmids were transfected as described above and total RNA was extracted after 24 h using Tri reagent (Invitrogen). 5’ RACE was performed with the 5’ RACE kit (Roche) using gene-specific reverse primers: PTEN-5’RACE-R and R2 (for nested PCR).

Measurement of read-through transcripts of the reporter constructs

Luciferase reporter constructs were linearized by BglII, which cuts upstream of the enhancer/promoter in the renilla luciferase constructs. 650 fmol of the linearized plasmids were transfected as described above. Total RNA was extracted after 24 h using Tri reagent (Invitrogen), followed by DNase I (NEB) treatment and ethanol precipitation. 1 µg of total RNA was used to generate cDNA using Superscript III reverse transcriptase (Invitrogen) and random hexamers (Invitrogen). Read-through transcripts were detected with a primer pair that localizes to the vector backbone (1040 bp downstream of the PAS; RenRT-qP-F and RenRT-qP-R). Read-through values were normalized to total transcript levels obtained by a primer pair that localizes to the renilla open reading frame (941 bp upstream of the PAS; RenORF-qP-F and RenORF-qP-R).

Measurement of rolling-circle transcription in luciferase reporter constructs

Luciferase plasmids were transfected as described above and total RNA was extracted after 24 h using Tri reagent, followed by DNase I treatment and ethanol precipitation. 1 µg of total RNA was used to generate cDNA using Superscript III reverse transcriptase and random hexamers. To detect the transcripts produced in vivo in a manner dependent on rolling-circle transcription on the renilla luciferase plasmids, PCR was conducted using a primer pair (Ren-Rolling-F and -R) that binds 147-bp upstream of BglII site which marks the start of the Penh. The total transcript levels were determined by PCR using a primer pair that localizes to the renilla open reading frame (Rluc-stability-F and RenORF-Rolling-R). The PCR products were resolved in 2% agarose gel and quantified using Multi-Gauge (Fuji). The amounts of the rolling-circle transcription were normalized to the total transcript levels.

Small-scale shRNA screen

shRNAs were designed using the siRNA selection program from the Whitehead Institute and cloned into pSUPERretropuro. Retroviral particles were obtained as described before⁸⁷. Knock-down efficiency was tested by RT-PCR with gene-specific primers and primers for GAPDH.

Luciferase assay using reporters integrated into the genome

The expression cassettes of the renilla luciferase reporters (from Pprom or Penh-Pprom to PAS (SV40 PAS or PTEN PPAS) in PIS1 plasmids) were inserted by Gibson cloning into pcDNA5-FRT plasmid digested by BglII and SphI. The integration of each pcDNA5-FRT plasmid with the renilla luciferase reporters into MCF7-FRT cells was carried out by Flp-in system as described elsewhere⁹⁵. Luciferase assays were conducted as described above after transfection of the cells with PIS0 plasmid.

Association of SU usage and gene expression

To assess if high levels of gene expression are associated with increased SU isoform usage, a dataset generated by Lianoglou et al. (2013) was reanalyzed³. The 10% highest and lowest expressed multi-UTR genes in each cell type were selected and the short 3′UTR index (SUI) which is the fraction of SU isoform expression divided by the total 3′UTR isoform expression, was plotted.

Association of cell type-specific enhancers with alternative 3′UTR isoforms

Dataset selection

To obtain genes regulated by cell type-specific enhancers, we used a publicly available dataset generated by Cai et al. (2020) who determined cell type-specific enhancers in murine definitive erythroblasts (here: Ery) and assigned them to target genes using a combination of methods including transcriptome analysis, chromatin accessibility, histone modifications, transcription factor occupancy, and 3D chromatin interactions⁶⁸. We used the assigned target genes from Supporting Information Dataset_S01.

To identify genes that increase their expression in erythroblasts, we compared gene expression between erythroblasts and murine hematopoietic stem cells (HSC). We set out to use replicates (two HSC datasets^69,94 and two Ery datasets^68,70). The accession numbers are listed above. To perform quality control for the samples, we aligned them to mm10 using HISAT2 v2.1.0. Gene body coverage was estimated against the mm10 housekeeping gene annotation provided by RSeQC using the geneBody_coverage.py script from RSeQC v4.0.0¹⁰³. MultiQC v1.10.1 was used to compare coverages¹⁰⁴. This analysis revealed uneven coverage among the samples (Supplementary Fig. 5f). Uneven coverage will confound the analysis on 3′UTR isoform expression as QAPA determines 3′UTR isoform ratios by using the coverage of reads that fall into the region of the SU and LU isoforms. Moreover, the different datasets use different library preparation methods (pair-end reads, single-end reads, stranded, not stranded) which result in technical differences across the samples independent of their cell type-specific differences. To identify technically similar datasets, we performed principal component analysis (using the top 1000 genes of highest variance in the gene expression matrix after DESeq2’s rlog transformation) on the four datasets. We observed that the datasets generated by Ling (2019) and Lee (2018) had a minimal distance in PC2 which reflects technical variation (Supplementary Fig. 5g)^69,70. Therefore, these samples were used for downstream pairwise analysis as representative Ery and HSC cell types, respectively.

Differential gene and 3′UTR isoform analysis

To identify genes with a significant difference in gene expression between Ery and HSC, we used DESeq2 v1.28.0^105,106. To identify genes with a significant difference in 3′UTR isoform expression, we first used QAPA to identify multi-UTR genes and to obtain TPM values for SU and LU isoforms¹⁰⁷. Samples were pseudoaligned for transcript quantification in Salmon v1.3.0 (“salmon quant–gcBias–validateMappings -l A”) to the pre-compiled QAPA v1.3.0 mm10 3′UTR annotation with the full mm10 genome as decoy¹⁰⁸. Multi-UTR genes used in the analysis were filtered based on Num_Events > 1 (more than one annotated isoform) and at least 3 TPM in one or more samples. Single-UTR genes have Num_Events = 1. SU isoforms were identified by QAPA APA_IDs ending in “_P”. To identify statistically significant changes in 3′UTR isoform usage we required a dPAU > 0.1 with a significant isoform ratio change according to DEXSeq v1.34.0 with 10% FDR. Normalized expression values were computed by rescaling QAPA TPM values per sample by estimated size factors from DESeq2. We further required a minimal fold change >2 for SU isoform expression to be considered significantly upregulated. All analysis was performed in R v4.0.2 and Bioconductor v1.13.

Multi-UTR gene categories

We first separated the multi-UTR genes into two categories based on their gene expression changes using DESeq2 (“gene up”, “gene not up”) using 10% FDR and at least 2-fold change in gene expression in erythroblasts. Next, we subdivided these groups with respect to “SU up” and identified the genes with at least 2-fold upregulation of SU isoform expression using QAPA TPM values and a significant 3′UTR isoform usage change with dPAU values obtained by QAPA of >0.1 and being significant according to DEXseq.

Stratified random sampling

The range of gene expression log2 fold-changes was subdivided into 13 strata of equal width, with 1 as an upper bound and the minimum of the “SU up” genes as the lower bound (approximately −4). For each stratum, we randomly subsampled the two groups (“gene not up” and “SU up”) to have matching numbers of genes. The number of genes per stratum was determined by the group with the least genes in that stratum.

Statistics

For all pairwise comparisons of PAS usage or transcriptional activity a 2-sided, 2-sample unequal variance t-test for independent samples was applied. When comparing several samples, a One-way ANOVA was performed. Mann–Whitney tests were used to determine if gene expression influences SU usage. Chi-square (X²) tests were used to test for significant enrichment of genes associated with Ery-specific enhancers. The Pearson value was reported. Statistical tests were performed using Excel, R, and SPSS.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The data supporting the findings of this study are available from the corresponding authors upon reasonable request. To identify the enhancers and promoters used in this study, levels of acetylated H3K27 and transcription factor binding sites in MCF7 cells were visualized using published ChIP-seq data (GSM946850), generated by the Encode project^43,44. Binding of MYC to the PTEN promoter was assessed by using published ChIP-seq data (GSE33213). FASTQ files for bulk RNA-seq samples were obtained from the Sequence Read Archive. Mouse definitive erythroblasts: SRR6946157-9⁶⁸, SRR8945139-41,44-45⁷⁰, mouse hematopoietic stem cells: SRR7946616-7⁹⁴, SRR6458998-9000⁶⁹. 3′-seq data were obtained from SRP029953³. Source data for the figures and Supplementary Figures are provided as a Source Data file.

Code availability

Our code is available at https://github.com/Mayrlab/utr-enhancers.

References

Levine, M. & Tjian, R. Transcription regulation and animal diversity. Nature 424, 147–151 (2003).
Article ADS CAS PubMed Google Scholar
Levine, M., Cattoglio, C. & Tjian, R. Looping back to leap forward: transcription enters a new era. Cell 157, 13–25 (2014).
Article CAS PubMed PubMed Central Google Scholar
Lianoglou, S., Garg, V., Yang, J. L., Leslie, C. S. & Mayr, C. Ubiquitously transcribed genes use alternative polyadenylation to achieve tissue-specific expression. Genes Dev. 27, 2380–2396 (2013).
Article CAS PubMed PubMed Central Google Scholar
Pennacchio, L. A. et al. In vivo enhancer analysis of human conserved non-coding sequences. Nature 444, 499–502 (2006).
Article ADS CAS PubMed Google Scholar
Visel, A. et al. ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature 457, 854–858 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Gasperini, M., Tome, J. M. & Shendure, J. Towards a comprehensive catalogue of validated and target-linked human enhancers. Nat. Rev. Genet. 21, 292–310 (2020).
Article CAS PubMed PubMed Central Google Scholar
Tian, B. & Manley, J. L. Alternative polyadenylation of mRNA precursors. Nat. Rev. Mol. cell Biol. 18, 18–30 (2017).
Article CAS PubMed Google Scholar
Gruber, A. J. & Zavolan, M. Alternative cleavage and polyadenylation in health and disease. Nat. Rev. Genet. 20, 599–614 (2019).
Article CAS PubMed Google Scholar
Mayr, C. What Are 3’ UTRs Doing? Cold Spring Harb Perspect Biol 11, a034728 (2019).
Tushev, G. et al. Alternative 3’ UTRs modify the localization, regulatory potential, stability, and plasticity of mRNAs in neuronal compartments. Neuron 98, 495–511 (2018).
Article CAS PubMed Google Scholar
Berkovits, B. D. & Mayr, C. Alternative 3’ UTRs act as scaffolds to regulate membrane protein localization. Nature 522, 363–367 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Lee, S. H. & Mayr, C. Gain of additional BIRC3 protein functions through 3’-UTR-mediated protein complex formation. Mol. Cell 74, 701–712 e709 (2019).
Article CAS PubMed PubMed Central Google Scholar
Ma, W. & Mayr, C. A membraneless organelle associated with the endoplasmic reticulum enables 3’UTR-mediated protein-protein interactions. Cell 175, 1492–1506.e1419 (2018).
Article CAS PubMed PubMed Central Google Scholar
Fernandes, N. & Buchan, J. R. RPS28B mRNA acts as a scaffold promoting cis-translational interaction of proteins driving P-body assembly. Nucleic Acids Res 48, 6265–6279 (2020).
Article CAS PubMed PubMed Central Google Scholar
Moretti, F. et al. Growth cone localization of the mRNA encoding the chromatin regulator HMGN5 modulates neurite outgrowth. Mol. Cell Biol. 35, 2035–2050 (2015).
Article CAS PubMed PubMed Central Google Scholar
Jenal, M. et al. The poly(A)-binding protein nuclear 1 suppresses alternative cleavage and polyadenylation sites. Cell 149, 538–553 (2012).
Article CAS PubMed Google Scholar
Lackford, B. et al. Fip1 regulates mRNA alternative polyadenylation to promote stem cell self-renewal. EMBO J. 33, 878–889 (2014).
Article CAS PubMed PubMed Central Google Scholar
Masamha, C. P. et al. CFIm25 links alternative polyadenylation to glioblastoma tumour suppression. Nature 510, 412–416 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Gruber, A. J. et al. A comprehensive analysis of 3’ end sequencing data sets reveals novel polyadenylation signals and the repressive role of heterogeneous ribonucleoprotein C on cleavage and polyadenylation. Genome Res 26, 1145–1159 (2016).
Article CAS PubMed PubMed Central Google Scholar
Brumbaugh, J. et al. Nudt21 controls cell fate by connecting alternative polyadenylation to chromatin signaling. Cell 172, 629–631 (2018).
Article CAS PubMed Google Scholar
Zhu, Y. et al. Molecular mechanisms for CFIm-mediated regulation of mRNA alternative polyadenylation. Mol. Cell 69, 62–74.e64 (2018).
Article CAS PubMed Google Scholar
Chatrikhi, R. et al. RNA binding protein CELF2 regulates signal-induced alternative polyadenylation by competing with enhancers of the polyadenylation machinery. Cell Rep. 28, 2795–2806.e2793 (2019).
Article CAS PubMed PubMed Central Google Scholar
Dubbury, S. J., Boutz, P. L. & Sharp, P. A. CDK12 regulates DNA repair genes by suppressing intronic polyadenylation. Nature 564, 141–145 (2018).
Article ADS CAS PubMed PubMed Central Google Scholar
Krajewska, M. et al. CDK12 loss in cancer cells affects DNA damage response genes through premature cleavage and polyadenylation. Nat. Commun. 10, 1757 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Nanavaty, V. et al. DNA methylation regulates alternative polyadenylation via CTCF and the cohesin complex. Mol. Cell 78, 752–764.e756 (2020).
Article CAS PubMed PubMed Central Google Scholar
Banerji, J., Rusconi, S. & Schaffner, W. Expression of a beta-globin gene is enhanced by remote SV40 DNA sequences. Cell 27, 299–308 (1981).
Article CAS PubMed Google Scholar
Pennacchio, L. A., Bickmore, W., Dean, A., Nobrega, M. A. & Bejerano, G. Enhancers: five essential questions. Nat. Rev. Genet. 14, 288–295 (2013).
Article CAS PubMed PubMed Central Google Scholar
Mitschka, S., Fansler, M. M. & Mayr, C. Generation of 3’UTR knockout cell lines by CRISPR/Cas9-mediated genome editing. Methods Enzymol. 655, 427–457 (2021).
Article CAS PubMed Google Scholar
Rutkowski, A. J. et al. Widespread disruption of host transcription termination in HSV-1 infection. Nat. Commun. 6, 7126 (2015).
Article ADS PubMed Google Scholar
Eaton, J. D., Francis, L., Davidson, L. & West, S. A unified allosteric/torpedo mechanism for transcriptional termination on human protein-coding genes. Genes Dev. 34, 132–145 (2020).
Article CAS PubMed PubMed Central Google Scholar
Wang, X. et al. Herpes simplex virus blocks host transcription termination via the bimodal activities of ICP27. Nat. Commun. 11, 293 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Rosa-Mercado, N. A. et al. Hyperosmotic stress alters the RNA polymerase II interactome and induces readthrough transcription despite widespread transcriptional repression. Mol. Cell 81, 502–513.e504 (2021).
Article CAS PubMed PubMed Central Google Scholar
Higgs, D. R. et al. Alpha-thalassaemia caused by a polyadenylation signal mutation. Nature 306, 398–400 (1983).
Article ADS CAS PubMed Google Scholar
Orkin, S. H., Cheng, T. C., Antonarakis, S. E. & Kazazian, H. H. Jr. Thalassemia due to a mutation in the cleavage-polyadenylation signal of the human beta-globin gene. EMBO J. 4, 453–456 (1985).
Article CAS PubMed PubMed Central Google Scholar
Gehring, N. H. et al. Increased efficiency of mRNA 3’ end formation: a new genetic mechanism contributing to hereditary thrombophilia. Nat. Genet 28, 389–392 (2001).
Article CAS PubMed Google Scholar
Graham, R. R. et al. Three functional variants of IFN regulatory factor 5 (IRF5) define risk and protective haplotypes for human lupus. Proc. Natl Acad. Sci. USA 104, 6758–6763 (2007).
Article ADS CAS PubMed PubMed Central Google Scholar
Stacey, S. N. et al. A germline variant in the TP53 polyadenylation signal confers cancer susceptibility. Nat. Genet 43, 1098–1103 (2011).
Article CAS PubMed PubMed Central Google Scholar
Johnston, J. J. et al. NAA10 polyadenylation signal variants cause syndromic microphthalmia. J. Med Genet 56, 444–452 (2019).
Article CAS PubMed Google Scholar
Li, L. et al. An atlas of alternative polyadenylation quantitative trait loci contributing to complex trait and disease heritability. Nat Genet 53, 994–1005 (2021).
Bogard, N., Linder, J., Rosenberg, A. B. & Seelig, G. A deep neural network for predicting and engineering alternative polyadenylation. Cell 178, 91–106 e123 (2019).
Article CAS PubMed PubMed Central Google Scholar
Alimonti, A. et al. Subtle variations in Pten dose determine cancer susceptibility. Nat. Genet 42, 454–458 (2010).
Article CAS PubMed PubMed Central Google Scholar
Mendes-Pereira, A. M. et al. Synthetic lethal targeting of PTEN mutant cells with PARP inhibitors. EMBO Mol. Med. 1, 315–322 (2009).
Article CAS PubMed PubMed Central Google Scholar
Frietze, S. et al. Cell type-specific binding patterns reveal that TCF7L2 can be tethered to the genome by association with GATA3. Genome Biol. 13, R52 (2012).
Article CAS PubMed PubMed Central Google Scholar
Consortium, E. P. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012).
Article ADS CAS Google Scholar
Zhang, J. et al. An integrative ENCODE resource for cancer genomics. Nat. Commun. 11, 3696 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Shlyueva, D., Stampfel, G. & Stark, A. Transcriptional enhancers: from properties to genome-wide predictions. Nat. Rev. Genet. 15, 272–286 (2014).
Article CAS PubMed Google Scholar
Vasudevan, K. M., Gurumurthy, S. & Rangnekar, V. M. Suppression of PTEN expression by NF-kappa B prevents apoptosis. Mol. Cell Biol. 24, 1007–1021 (2004).
Article CAS PubMed PubMed Central Google Scholar
Gupta, S. C., Singh, R., Pochampally, R., Watabe, K. & Mo, Y. Y. Acidosis promotes invasiveness of breast cancer cells through ROS-AKT-NF-κB pathway. Oncotarget 5, 12070–12082 (2014).
Article PubMed PubMed Central Google Scholar
Dantonel, J. C., Murthy, K. G., Manley, J. L. & Tora, L. Transcription factor TFIID recruits factor CPSF for formation of 3’ end of mRNA. Nature 389, 399–402 (1997).
Article ADS CAS PubMed Google Scholar
McCracken, S. et al. The C-terminal domain of RNA polymerase II couples mRNA processing to transcription. Nature 385, 357–361 (1997).
Article ADS CAS PubMed Google Scholar
Rosonina, E., Bakowski, M. A., McCracken, S. & Blencowe, B. J. Transcriptional activators control splicing and 3’-end cleavage levels. J. Biol. Chem. 278, 43034–43040 (2003).
Article CAS PubMed Google Scholar
Calvo, O. & Manley, J. L. The transcriptional coactivator PC4/Sub1 has multiple functions in RNA polymerase II transcription. EMBO J. 24, 1009–1020 (2005).
Article CAS PubMed PubMed Central Google Scholar
Nagaike, T. et al. Transcriptional activators enhance polyadenylation of mRNA precursors. Mol. Cell 41, 409–418 (2011).
Article CAS PubMed PubMed Central Google Scholar
Ji, Z. et al. Transcriptional activity regulates alternative cleavage and polyadenylation. Mol. Syst. Biol. 7, 534 (2011).
Article PubMed PubMed Central CAS Google Scholar
Oktaba, K. et al. ELAV links paused pol II to alternative polyadenylation in the drosophila nervous system. Mol. Cell 57, 341–348 (2015).
Article CAS PubMed Google Scholar
Xiao, R. et al. Pervasive chromatin-RNA binding protein interactions enable RNA-based regulation of transcription. Cell 178, 107–121.e118 (2019).
Article CAS PubMed PubMed Central Google Scholar
Bentley, D. L. Coupling mRNA processing with transcription in time and space. Nat. Rev. Genet. 15, 163–175 (2014).
Article CAS PubMed PubMed Central Google Scholar
Rambout, X., Dequiedt, F. & Maquat, L. E. Beyond transcription: roles of transcription factors in Pre-mRNA splicing. Chem. Rev. 118, 4339–4364 (2018).
Article CAS PubMed Google Scholar
Soles, L. V. & Shi, Y. Crosstalk between mRNA 3’-end processing and epigenetics. Front. Genet. 12, 637705 (2021).
Article CAS PubMed PubMed Central Google Scholar
Carswell, S. & Alwine, J. C. Efficiency of utilization of the simian virus 40 late polyadenylation site: effects of upstream sequences. Mol. Cell Biol. 9, 4248–4258 (1989).
CAS PubMed PubMed Central Google Scholar
Chao, L. C., Jamil, A., Kim, S. J., Huang, L. & Martinson, H. G. Assembly of the cleavage and polyadenylation apparatus requires about 10 seconds in vivo and is faster for strong than for weak poly(A) sites. Mol. Cell Biol. 19, 5588–5600 (1999).
Article CAS PubMed PubMed Central Google Scholar
Smale, S. T. Luciferase assay. Cold Spring Harb. Protoc. 2010, pdb prot5421 (2010).
Article PubMed Google Scholar
Beaudoing, E., Freier, S., Wyatt, J. R., Claverie, J. M. & Gautheret, D. Patterns of variant polyadenylation signal usage in human genes. Genome Res 10, 1001–1010 (2000).
Article CAS PubMed PubMed Central Google Scholar
Martin, G., Gruber, A. R., Keller, W. & Zavolan, M. Genome-wide analysis of Pre-mRNA 3’ end processing reveals a decisive role of human cleavage factor I in the regulation of 3’ UTR length. Cell Rep. 1, 753–763 (2012).
Article CAS PubMed Google Scholar
Spiegelman, B. M. & Heinrich, R. Biological control through regulated transcriptional coactivators. Cell 119, 157–167 (2004).
Article CAS PubMed Google Scholar
Perales, R. & Bentley, D. “Cotranscriptionality”: the transcription elongation complex as a nexus for nuclear transactions. Mol. Cell 36, 178–191 (2009).
Article CAS PubMed PubMed Central Google Scholar
Jonkers, I. & Lis, J. T. Getting up to speed with transcription elongation by RNA polymerase II. Nat. Rev. Mol. cell Biol. 16, 167–177 (2015).
Article CAS PubMed PubMed Central Google Scholar
Cai, W. et al. Enhancer dependence of cell-type-specific gene expression increases with developmental age. Proc. Natl Acad. Sci. USA 117, 21450–21458 (2020).
Article CAS PubMed PubMed Central Google Scholar
Lee, S. C. et al. Synthetic lethal and convergent biological effects of cancer-associated spliceosomal gene mutations. Cancer Cell 34, 225–241 e228 (2018).
Article CAS PubMed PubMed Central Google Scholar
Ling, T. et al. Chromatin occupancy and epigenetic analysis reveal new insights into the function of the GATA1 N terminus in erythropoiesis. Blood 134, 1619–1631 (2019).
Article PubMed PubMed Central Google Scholar
Creyghton, M. P. et al. Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proc. Natl Acad. Sci. USA 107, 21931–21936 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Fansler, M. M., Zhen, G. & Mayr, C. Quantification of alternative 3′UTR isoforms from single cell RNA-seq data with scUTRquant. bioRxiv, 2021.2011.2022.469635 (2021).
Takagaki, Y., Seipelt, R. L., Peterson, M. L. & Manley, J. L. The polyadenylation factor CstF-64 regulates alternative processing of IgM heavy chain pre-mRNA during B cell differentiation. Cell 87, 941–952 (1996).
Article CAS PubMed Google Scholar
Harel-Sharvit, L. et al. RNA polymerase II subunits link transcription and mRNA decay to translation. Cell 143, 552–563 (2010).
Article CAS PubMed Google Scholar
Trcek, T., Larson, D. R., Moldon, A., Query, C. C. & Singer, R. H. Single-molecule mRNA decay measurements reveal promoter- regulated mRNA stability in yeast. Cell 147, 1484–1497 (2011).
Article CAS PubMed PubMed Central Google Scholar
Bregman, A. et al. Promoter elements regulate cytoplasmic mRNA decay. Cell 147, 1473–1483 (2011).
Article CAS PubMed Google Scholar
Zid, B. M. & O’Shea, E. K. Promoter sequences direct cytoplasmic localization and translation of mRNAs during starvation in yeast. Nature 514, 117–121 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Haimovich, G., Choder, M., Singer, R. H. & Trcek, T. The fate of the messenger is pre-determined: a new model for regulation of gene expression. Biochim Biophys. Acta 1829, 643–653 (2013).
Article CAS PubMed PubMed Central Google Scholar
Cramer, P. et al. Coupling of transcription with alternative splicing: RNA pol II promoters modulate SF2/ASF and 9G8 effects on an exonic splicing enhancer. Mol. Cell 4, 251–258 (1999).
Article CAS PubMed Google Scholar
de la Mata, M. et al. A slow RNA polymerase II affects alternative splicing in vivo. Mol. Cell 12, 525–532 (2003).
Article PubMed Google Scholar
Fong, N. & Bentley, D. L. Capping, splicing, and 3’ processing are independently stimulated by RNA polymerase II: different functions for different segments of the CTD. Genes Dev. 15, 1783–1795 (2001).
Article CAS PubMed PubMed Central Google Scholar
Rozenblatt-Rosen, O. et al. The tumor suppressor Cdc73 functionally associates with CPSF and CstF 3’ mRNA processing factors. Proc. Natl Acad. Sci. USA 106, 755–760 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Lai, F., Gardini, A., Zhang, A. & Shiekhattar, R. Integrator mediates the biogenesis of enhancer RNAs. Nature 525, 399–403 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Barra, J. et al. Integrator restrains paraspeckles assembly by promoting isoform switching of the lncRNA NEAT1. Sci. Adv. 6, eaaz9072 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Gardini, A. et al. Integrator regulates transcriptional initiation and pause release following activation. Mol. Cell 56, 128–139 (2014).
Article CAS PubMed PubMed Central Google Scholar
Elrod, N. D. et al. The integrator complex attenuates promoter-proximal transcription at protein-coding genes. Mol. Cell 76, 738–752 e737 (2019).
Article CAS PubMed PubMed Central Google Scholar
Mayr, C. & Bartel, D. P. Widespread shortening of 3’UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells. Cell 138, 673–684 (2009).
Article CAS PubMed PubMed Central Google Scholar
Yang, Y. et al. Maternal mRNAs with distinct 3’ UTRs define the temporal pattern of Ccnb1 synthesis during mouse oocyte meiotic maturation. Genes Dev. 31, 1302–1307 (2017).
Article CAS PubMed PubMed Central Google Scholar
Bae, B. et al. Elimination of Calm1 long 3’-UTR mRNA isoform by CRISPR-Cas9 gene editing impairs dorsal root ganglion development and hippocampal neuron activation in mice. RNA 26, 1414–1430 (2020).
Article CAS PubMed PubMed Central Google Scholar
Mitschka, S. & Mayr, C. Endogenous p53 expression in human and mouse is not regulated by its 3’UTR. eLife 10, e65700 (2021).
Zhao, W. et al. CRISPR-Cas9-mediated functional dissection of 3’-UTRs. Nucleic Acids Res 45, 10800–10810 (2017).
Article CAS PubMed PubMed Central Google Scholar
Chen, C. Y., Chen, S. T., Juan, H. F. & Huang, H. C. Lengthening of 3’UTR increases with morphological complexity in animal evolution. Bioinformatics 28, 3178–3181 (2012).
Article CAS PubMed Google Scholar
Mayr, C. Regulation by 3’-untranslated regions. Annu. Rev. Genet. 51, 171–194 (2017).
Article CAS PubMed Google Scholar
Numata, A. et al. Lysine acetyltransferase Tip60 is required for hematopoietic stem cell maintenance. Blood 136, 1735–1747 (2020).
Slobodin, B. et al. Transcription impacts the efficiency of mRNA translation via co-transcriptional N6-adenosine methylation. Cell 169, 326–337.e312 (2017).
Article CAS PubMed PubMed Central Google Scholar
Ran, F. A. et al. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8, 2281–2308 (2013).
Article CAS PubMed PubMed Central Google Scholar
Perez, A. R. et al. GuideScan software for improved single and paired CRISPR guide RNA design. Nat. Biotechnol. 35, 347–349 (2017).
Article CAS PubMed PubMed Central Google Scholar
Sarbassov, D. D., Guertin, D. A., Ali, S. M. & Sabatini, D. M. Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098–1101 (2005).
Article ADS CAS PubMed Google Scholar
Marquardt, S., Hazelbaker, D. Z. & Buratowski, S. Distinct RNA degradation pathways and 3’ extensions of yeast non-coding RNA species. Transcription 2, 145–154 (2011).
Article PubMed PubMed Central Google Scholar
Moon, S. L. et al. A noncoding RNA produced by arthropod-borne flaviviruses inhibits the cellular exoribonuclease XRN1 and alters host mRNA stability. Rna 18, 2029–2040 (2012).
Article CAS PubMed PubMed Central Google Scholar
Lugowski, A., Nicholson, B. & Rissland, O. S. Determining mRNA half-lives on a transcriptome-wide scale. Methods 137, 90–98 (2018).
Article CAS PubMed Google Scholar
Russo, J., Heck, A. M., Wilusz, J. & Wilusz, C. J. Metabolic labeling and recovery of nascent RNA to accurately quantify mRNA stability. Methods 120, 39–48 (2017).
Article CAS PubMed PubMed Central Google Scholar
Wang, L., Wang, S. & Li, W. RSeQC: quality control of RNA-seq experiments. Bioinformatics 28, 2184–2185 (2012).
Article CAS PubMed Google Scholar
Ewels, P., Magnusson, M., Lundin, S. & Kaller, M. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047–3048 (2016).
Article CAS PubMed PubMed Central Google Scholar
Anders, S., Reyes, A. & Huber, W. Detecting differential usage of exons from RNA-seq data. Genome Res 22, 2008–2017 (2012).
Article CAS PubMed PubMed Central Google Scholar
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Article PubMed PubMed Central CAS Google Scholar
Ha, K. C. H., Blencowe, B. J. & Morris, Q. QAPA: a new method for the systematic analysis of alternative polyadenylation from RNA-seq data. Genome Biol. 19, 45 (2018).
Article PubMed PubMed Central CAS Google Scholar
Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419 (2017).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank all members of the Mayr lab for helpful discussions and Kristian Helin for critical reading of the manuscript. This work was funded by the NIH Director′s Pioneer Award (DP1-GM123454), Damon Runyon-Rachleff Innovation Award, the Pershing Square Sohn Cancer Research Alliance, and the NCI Cancer Center Support Grant (P30 CA008748).

Author information

Authors and Affiliations

Cancer Biology and Genetics Program, Memorial Sloan Kettering Cancer Center, New York, NY, 10065, USA
Buki Kwon, Mervin M. Fansler, Neil D. Patel, Jihye Lee, Weirui Ma & Christine Mayr
Tri-Institutional Training Program in Computational Biology and Medicine, Weill Cornell Graduate College, New York, NY, 10021, USA
Mervin M. Fansler & Christine Mayr

Authors

Buki Kwon
View author publications
You can also search for this author in PubMed Google Scholar
Mervin M. Fansler
View author publications
You can also search for this author in PubMed Google Scholar
Neil D. Patel
View author publications
You can also search for this author in PubMed Google Scholar
Jihye Lee
View author publications
You can also search for this author in PubMed Google Scholar
Weirui Ma
View author publications
You can also search for this author in PubMed Google Scholar
Christine Mayr
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

B.K. designed and performed all experiments regarding endogenous PTEN. B.K. measured read-through of the reporter and performed experiments on the integrated reporter, M.M.F. performed all computational analyses, N.P. and J.L. cloned the reporter constructs and performed the reporter assays, J.L. evaluated the knock-down efficiency of the shRNAs, W.M. performed the mRNA stability experiment of the reporter. C.M. conceived and supervised the project, designed and analyzed the reporter experiments, and wrote the manuscript with input from all authors.

Corresponding author

Correspondence to Christine Mayr.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Bin Tian and the other, anonymous, reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Kwon, B., Fansler, M.M., Patel, N.D. et al. Enhancers regulate 3′ end processing activity to control expression of alternative 3′UTR isoforms. Nat Commun 13, 2709 (2022). https://doi.org/10.1038/s41467-022-30525-y

Download citation

Received: 30 July 2021
Accepted: 02 May 2022
Published: 17 May 2022
DOI: https://doi.org/10.1038/s41467-022-30525-y

This article is cited by

Context-specific regulation and function of mRNA alternative polyadenylation
- Sibylle Mitschka
- Christine Mayr
Nature Reviews Molecular Cell Biology (2022)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.