SRSF1 regulates polyadenylation site selection independently of and through coordination with U1 snRNP

This study reveals that the splicing factor SRSF1 regulates polyadenylation site selection and RNA isoform determination through a dual mechanism involving direct RNA binding independent of U1 snRNP and coordinated interactions with U1 snRNP and RNA Polymerase II to modulate transcription elongation.

The Gist

Imagine your DNA is a massive, ancient library containing the instructions for building a human. When a cell needs to make a specific protein, it doesn't just copy the whole book; it takes a specific chapter, edits it, and adds a "The End" marker (called a polyadenylation site or PAS) to know where to stop.

The problem is, sometimes a chapter has multiple "The End" markers. If the cell picks the wrong one, it might cut the story short (missing important plot points) or keep reading into the next chapter (adding gibberish). This is called Alternative Polyadenylation, and getting it right is crucial for health.

This paper investigates a specific editor in the library called SRSF1. Scientists wanted to know: How does SRSF1 decide where to put the "The End" marker?

Here is the story of their discovery, explained with some analogies:

1. The Two Ways SRSF1 Works

The researchers found that SRSF1 acts like a master editor using two different strategies to decide where a story should end.

Strategy A: The "Local Landmark" (Independent Mode)

• The Analogy: Imagine you are walking down a street with two "Stop" signs. One is right in front of you, and the other is three blocks away. If you are a walker who likes to stop early, you need a reason to stop at the first sign.
• The Science: SRSF1 acts like a friendly guide who stands right near the first (proximal) "Stop" sign in the 3' UTR (the end of the gene). By physically grabbing the RNA right there, it tells the cellular machinery, "Stop here! This is the right place."
• The Result: When SRSF1 is present, the cell stops early. When SRSF1 is removed, the cell ignores the first sign, keeps walking, and stops at the distant sign instead. This changes the final product of the protein.

Strategy B: The "Traffic Cop" (Coordinated Mode)

• The Analogy: Now imagine a busy intersection where a traffic cop (U1 snRNP) is directing cars. SRSF1 doesn't just stand alone; it works with the traffic cop.
• The Science: SRSF1 and U1 snRNP are old friends who work together. U1 snRNP is known to speed up the transcription machine (Pol II). SRSF1 helps U1 snRNP do its job by physically linking up with the machine.
• The Result: Together, they regulate where the story ends for genes located deeper inside the "book" (the gene body). They act as a team to ensure the machine doesn't run off the rails.

2. The Connection to Cancer

The paper also looked at breast cancer.

• The Analogy: In a healthy library, the editors (SRSF1) are present in the right amounts, ensuring stories end correctly. In some cancer libraries, there are too many editors (SRSF1 is overproduced).
• The Science: The researchers found that breast cancer tumors with high levels of SRSF1 had very different "ending choices" compared to tumors with low levels.
• The Takeaway: Because SRSF1 changes where the story ends, it can create "mutant" versions of proteins that help cancer grow. This suggests that SRSF1 isn't just a splicing editor; it's a key player in how cancer cells rewrite their own instructions.

3. The "Speed Limit" Effect

One of the coolest discoveries was about speed.

• The Analogy: Think of the transcription machine (Pol II) as a train reading the DNA track.
  • U1 snRNP is like a green light that tells the train to speed up.
  • SRSF1, however, acts like a gentle brake. It slows the train down slightly.
• Why slow down? If the train goes too fast, it might zoom past the "Stop" sign at the beginning of the 3' UTR. By slowing the train down, SRSF1 gives the machinery enough time to see the first sign and stop there.
• The Consequence: When SRSF1 is missing, the train speeds up, zooms past the first stop, and keeps going until it hits a stop sign way down the track. This causes "transcription readthrough," where the train keeps going into areas it shouldn't, potentially causing chaos in the cell.

Summary

This paper tells us that SRSF1 is a multitasking editor that shapes our genetic stories in two ways:

1. Directly: It stands next to the "Stop" sign and says, "Stop here!"
2. Indirectly: It teams up with the traffic cop (U1 snRNP) to slow down the transcription train, ensuring the machinery has time to pick the right stop.

When this system goes wrong (like in breast cancer), the stories get cut in the wrong places, leading to the production of faulty proteins that drive disease. This research helps us understand the complex "traffic control" system inside our cells that keeps our genetic code running smoothly.

Technical Summary

1. Problem Statement

Polyadenylation site (PAS) selection is a critical step in pre-mRNA processing that determines RNA isoform diversity, stability, and translation. While the core spliceosome component U1 snRNP is known to regulate PAS choice (specifically by inhibiting intronic premature cleavage and polyadenylation, or PCPA), it remains unclear how U1 snRNP coordinates with other splicing factors to shape these decisions. SRSF1 (Serine/Arginine-Rich Splicing Factor 1) is a well-characterized splicing factor that interacts with both U1 snRNP and RNA Polymerase II (Pol II). The central question of this study is whether SRSF1 regulates PAS selection independently of U1 snRNP, or if it acts in concert with U1 snRNP to couple transcription dynamics with 3'-end processing.

2. Methodology

The authors employed a multi-faceted approach combining genomic, biochemical, and computational techniques in human HEK293T cells and breast cancer tumor data:

• Genetic Perturbation: SRSF1 and U1-70K (a core component of U1 snRNP) were depleted using inducible shRNA and siRNA in HEK293T cells.
• 3'-End Sequencing: Chromatin-associated RNA was isolated to capture nascent transcripts, followed by 3'-end sequencing (QuantSeq) to map PAS usage genome-wide.
• Validation: Changes in specific genes (ANAPC7, IFNAR1) were validated using 3'-RACE.
• Binding Analysis: Integration of existing PIE-seq data (SRSF1 RNA binding sites) with PAS usage data to determine spatial relationships between SRSF1 binding and PAS selection.
• Clinical Correlation: Analysis of The Cancer Genome Atlas (TCGA) breast cancer data to correlate SRSF1 expression levels with alternative last exon (ALE) usage using the HITindex pipeline.
• Biochemical Interaction Studies:
  • Co-Immunoprecipitation (Co-IP): To test physical interactions between Pol II, SRSF1, and U1 snRNP.
  • AMO Treatment: Use of an Antisense Morpholino Oligonucleotide (AMO) targeting U1 snRNA to disrupt U1 snRNP interactions with nascent RNA and Pol II.
  • IP-Mass Spectrometry: To identify the broader network of factors dependent on U1 snRNP for Pol II association.
• Transcription Dynamics Assays:
  • TT-seq (Transient Transcriptome Sequencing): To measure nascent RNA synthesis rates.
  • PRO-seq (Precision Run-On Sequencing): To map the precise positions of Pol II.
  • Elongation Index (EI) & Readthrough Index (RTI): Calculated to assess Pol II elongation speed and transcription termination efficiency.
• Structural Modeling: AlphaFold3 was used to predict interactions between SRSF1 and U1-70K, mapped onto cryo-EM structures of the Pol II-U1 snRNP complex.

3. Key Contributions

The study establishes that SRSF1 regulates PAS selection through two distinct, complementary mechanisms:

1. U1 snRNP-Independent Mechanism: Direct RNA binding near proximal PASs in 3' UTRs.
2. U1 snRNP-Dependent Mechanism: Coordination with U1 snRNP to modulate Pol II elongation dynamics and termination across the gene body.

4. Key Results

A. SRSF1 Regulates PAS Selection

• Depletion of SRSF1 caused genome-wide shifts in PAS usage, affecting ~1,284 sites.
• General Trend: SRSF1 depletion led to a shift toward distal PAS usage (downstream sites), indicating SRSF1 normally promotes the usage of proximal PASs.
• Validation: 3'-RACE confirmed increased proximal PAS usage in ANAPC7 and decreased usage in IFNAR1 upon SRSF1 knockdown (KD).

B. Distinct vs. Shared PAS Regulation

• Comparison of SRSF1 KD and U1-70K KD revealed:
  • SRSF1-specific sites: Enriched in 3' UTRs and distal to the Transcription Start Site (TSS).
  • U1-specific sites: Enriched in introns (PCPA events).
  • Shared sites: A significant overlap (~222 sites) where both factors influence PAS choice. These shared sites are located more proximally within gene bodies.
• Directionality: At shared sites, SRSF1 and U1 KDs could shift PAS usage in the same direction (e.g., both increasing usage) or opposite directions, suggesting distinct regulatory roles rather than simple redundancy.

C. Mechanism 1: Direct RNA Binding in 3' UTRs

• SRSF1 binding sites are enriched near the start of 3' UTRs.
• SRSF1-sensitive, downregulated PASs are located immediately downstream of these binding sites.
• Model: SRSF1 binds near the start of the 3' UTR and directly promotes the usage of nearby proximal PASs. In breast cancer tumors with low SRSF1 levels, a distal shift in last exon usage was observed, linking this mechanism to oncogenic isoform production.

D. Mechanism 2: Coordination with U1 snRNP and Pol II

• Physical Interaction: Co-IP experiments showed that SRSF1 associates with Pol II, but this interaction is dependent on U1 snRNP. Disrupting U1 snRNP (via U1-70K KD or U1 AMO) significantly reduced SRSF1-Pol II binding.
• Specificity: U1 snRNP acts as a "gatekeeper," facilitating the recruitment of a select set of factors to Pol II, including SRSF1, SCAF4, SCAF8, and SRPK1. Other SR proteins not dependent on U1 snRNP were unaffected.
• Pol II Dynamics:
  • Elongation Index (EI): SRSF1 depletion increased the EI, meaning SRSF1 normally slows down Pol II elongation. (Contrast: U1 snRNP accelerates elongation).
  • Readthrough: SRSF1 depletion led to increased transcription readthrough (failure to terminate), particularly in GC-rich, intron-containing genes.
• Synthesis: SRSF1 slows Pol II elongation, providing a kinetic "window of opportunity" for the cleavage and polyadenylation machinery to recognize proximal PASs.

5. Significance

• Dual Role of Splicing Factors: The paper demonstrates that splicing factors like SRSF1 function beyond splicing, acting as critical regulators of 3'-end processing and transcription termination.
• Kinetic Coupling: It provides a mechanistic link between transcription elongation rates and PAS selection, showing how SRSF1 and U1 snRNP tune Pol II speed to ensure accurate isoform determination.
• U1 snRNP as a Hub: The study redefines U1 snRNP not just as a spliceosome component, but as a central hub that recruits specific regulatory factors (like SRSF1) to the transcription machinery to coordinate processing events.
• Cancer Relevance: The findings suggest that altered SRSF1 levels in breast cancer drive oncogenic isoform production via dysregulated PAS selection, offering potential new avenues for understanding cancer transcriptomics.
• Model Integration: The results support a "U1 relay" model where U1 snRNP, SRSF1, and CPA machinery cycle on and off Pol II to dynamically shape 3'-end processing.

In conclusion, SRSF1 shapes RNA isoform diversity through a dual strategy: direct RNA binding to promote proximal PAS usage in 3' UTRs, and U1 snRNP-dependent modulation of Pol II elongation to facilitate co-transcriptional 3'-end processing throughout the gene body.