Abnormal expression of splicing regulators RBFOX and NOVA is associated with aberrant splicing patterns at the Neurexin-3 gene in a monogenic autism spectrum disorder

This study demonstrates that in Pitt-Hopkins Syndrome, a monogenic form of autism caused by TF4 haploinsufficiency, the dysregulation of splicing regulators NOVA and RBFOX leads to aberrant splicing of the Neurexin-3 gene, specifically reducing secreted isoform expression and contributing to impaired synapse organization and decreased neural activity.

The Gist

The Big Picture: A Glitch in the Brain's "Copy Machine"

Imagine your body's DNA as a massive Master Instruction Manual for building a human. But here's the catch: the manual is written in a code that needs to be edited before it can be used. This editing process is called RNA splicing.

Think of splicing like a film editor cutting a movie. The raw footage (DNA) has long scenes (introns) that need to be cut out, and the good scenes (exons) need to be stitched together. Sometimes, the editor has to choose between two different endings for a scene. This is Alternative Splicing. It allows one gene to create different "versions" of a protein, just like a movie can have a "Director's Cut" or a "Theatrical Cut."

Autism Spectrum Disorder (ASD) is often a complex puzzle with many pieces. But this study focuses on a specific, rare type of autism called Pitt-Hopkins Syndrome (PTHS). In this case, the puzzle has a single, clear missing piece: a broken gene called TCF4.

The researchers asked: If we break the TCF4 gene, how does that mess up the brain's "film editing" process, and why does that lead to autism symptoms?

---

The Plot Twist: The Editors Got Fired

In a healthy brain, there are specialized "editors" (proteins) whose job is to tell the splicing machine exactly which scenes to keep and which to cut. Two famous families of these editors are called RBFOX and NOVA. They are like the showrunners of the brain's construction site, ensuring that the connections between brain cells (synapses) are built correctly.

What the researchers found:
In the brains of people with Pitt-Hopkins Syndrome, the TCF4 gene is broken. Because of this, the brain cells don't produce enough of the RBFOX and NOVA editors.

• The Analogy: Imagine a construction site where the foreman (TCF4) is missing. Because the foreman is gone, the specialized foremen who manage the electrical wiring (RBFOX and NOVA) never show up to work.
• The Result: Without these editors, the "film editing" goes haywire. The brain starts cutting out the wrong scenes or keeping the wrong ones.

The Specific Crime Scene: The Neurexin-3 Gene

The researchers zoomed in on one specific gene called Neurexin-3 (NRXN3). This gene is crucial for building the "glue" that holds brain cells together so they can talk to each other.

Normally, the splicing machine creates two versions of the Neurexin-3 protein:

1. The Anchor (Transmembrane): This version is like a bolt that screws the brain cell into the wall. It stays put and helps form stable connections.
2. The Messenger (Secreted): This version is like a loose note passed between cells. It floats around and helps signal that a connection needs to be made or adjusted.

What went wrong in PTHS:
Because the editors (RBFOX/NOVA) were missing, the splicing machine got confused.

• In healthy brains, there is a good balance of Anchors and Messengers.
• In PTHS brains, the machine started cutting out the "Messenger" version almost entirely. It was like the film editor decided to delete all the "Director's Cuts" and only release the "Theatrical Cuts."

The Consequence:
The brain cells ended up with too many "Anchors" and not enough "Messengers." The connections became rigid and unbalanced. The brain cells couldn't communicate effectively, leading to the reduced electrical activity observed in these patients. This lack of communication is likely why the brain struggles with the complex wiring needed for social interaction and behavior.

The Detective Work: How They Proved It

The scientists didn't just guess; they acted like forensic detectives:

1. The Crime Scene (Organoids): They grew tiny, 3D "mini-brains" (organoids) from the skin cells of PTHS patients. These mini-brains acted like a test kitchen to see how the disease behaves in real-time.
2. The Evidence (RNA Sequencing): They read the genetic "scripts" of these mini-brains and found that the "editors" (RBFOX/NOVA) were missing and the "Neurexin" script was being edited wrong.
3. The Lab Test (PCR & Western Blot): They used chemical tools to physically measure the proteins. They found that the "Messenger" protein (the one that should float around) was significantly weaker in the patient samples compared to healthy controls.

The Takeaway: Why This Matters

This paper is a big deal because it connects the dots between three things that were previously separate:

1. The Genetic Cause: A broken TCF4 gene.
2. The Molecular Mechanism: Missing "editor" proteins (RBFOX/NOVA) causing bad splicing.
3. The Physical Symptom: A brain that can't fire its electrical signals properly because the "glue" (Neurexin) is the wrong shape.

The Metaphor for the Future:
Think of autism research like trying to fix a car that won't start. For a long time, we knew the engine was broken, but we didn't know which part. This study says, "The spark plugs (RBFOX/NOVA) aren't firing because the ignition switch (TCF4) is broken, which is why the fuel mixture (Neurexin splicing) is wrong."

By understanding this specific chain of events, scientists can now look for drugs or therapies that might bypass the broken ignition switch and manually fix the spark plugs, potentially restoring the brain's ability to communicate. It turns a vague diagnosis into a specific, targetable problem.

Technical Summary

1. Problem Statement

Autism Spectrum Disorders (ASD) are complex conditions involving genetic and environmental factors. While aberrant RNA splicing has been implicated in ASD, the specific molecular mechanisms linking monogenic causes of autism to downstream splicing dysregulation remain poorly understood. Specifically, it is unclear how mutations in specific transcription factors lead to altered expression of RNA-binding proteins (RBPs) like RBFOX and NOVA, and how these alterations subsequently affect the splicing of synaptic genes, ultimately contributing to neural dysfunction. The study focuses on Pitt-Hopkins Syndrome (PTHS), a monogenic form of autism caused by haploinsufficiency of the transcription factor TCF4, to dissect these causal links.

2. Methodology

The researchers employed a multi-omics approach combining patient-derived cellular models with computational and experimental validation:

• Cellular Models:
  • 2D Cultures: Neurons and neural progenitor cells (NPCs) derived from induced pluripotent stem cells (iPSCs) of 4 PTHS patients and their same-sex parental controls.
  • 3D Organoids: Cortical organoids (CtOs) and subpallial organoids (sPOs) derived from the same subjects to model complex neural tissue architecture and cell-type diversity.
• Transcriptomics:
  • Bulk RNA-Seq: Used to identify differentially expressed genes (DEGs) and alternative splicing events in 2D neuronal cultures.
  • Single-Cell RNA-Seq (scRNA-Seq): Analyzed previously generated data from organoids to assess cell-type-specific expression of splicing regulators (NOVA1, RBFOX1/2/3) across glutamatergic and GABAergic lineages.
• Splicing Analysis:
  • rMATS-turbo: Used to quantify alternative splicing events (Skipped Exon, Retained Intron, etc.) and calculate Percent Spliced In (PSI) values and $\Delta$PSI (difference between patient and control).
  • Enrichment Analysis: Over-Representation Analysis (ORA) and Gene Set Enrichment Analysis (GSEA) were performed using WebGestalt to link splicing events to biological pathways (GO, KEGG).
• Binding Site Analysis:
  • Motif Enrichment: Used MEME suite (SEA and FIMO) to detect NOVA1 and RBFOX1/2/3 binding motifs in the genomic regions of genes with aberrant splicing.
  • CLIP-Seq Integration: Cross-referenced findings with published HITS-CLIP datasets to confirm direct binding of these RBPs to target transcripts.
• Experimental Validation:
  • RT-qPCR: Validated expression levels of splicing regulators and specific NRXN3 transcript variants.
  • PCR & Sanger Sequencing: Designed primers to distinguish between NRXN3 isoforms (with or without exon 20.5) and confirmed splicing patterns via cloning and sequencing.
  • Western Blotting: Analyzed protein isoforms in 2D cultures and organoids to detect differences in transmembrane vs. soluble Neurexin-3 forms.

3. Key Contributions

• Mechanistic Link in Monogenic Autism: The study establishes a direct causal chain in PTHS: TCF4 haploinsufficiency $\rightarrow$ dysregulation of splicing regulators (RBFOX/NOVA) $\rightarrow$ aberrant splicing of synaptic genes $\rightarrow$ impaired neural function.
• Cell-Type Specificity: It highlights that splicing regulator dysregulation is specific to differentiated neurons rather than neural progenitors, and that 3D organoids reveal complex expression patterns distinct from 2D cultures.
• Novel Splicing Event in Neurexin-3: The authors identified a specific exon skipping event in the NRXN3 gene that alters the ratio of transmembrane to soluble Neurexin-3 isoforms, a mechanism not previously linked to PTHS pathology.

4. Key Results

A. Dysregulation of Splicing Regulators

• 2D Neurons: Bulk RNA-Seq and qPCR revealed significant downregulation of RBFOX1, RBFOX2, RBFOX3, and NOVA1 in PTHS neurons compared to controls. This was not observed in neural progenitors.
• Organoids: scRNA-Seq showed a more complex picture. While RBFOX1/2/3 expression remained generally lower or unchanged in specific neuronal subpopulations, NOVA1 expression was paradoxically higher in PTHS organoids. The authors suggest this discrepancy arises from the complex cellular interactions in 3D organoids versus isolated 2D cultures.

B. Aberrant Splicing Landscape

• Global Impact: rMATS-turbo analysis identified thousands of differential splicing events (SE, RI, A5SS, A3SS, MXE) in PTHS neurons.
• Pathway Enrichment: Genes with aberrant splicing were significantly enriched for pathways related to synapse formation, axon projection, and cell adhesion.
• RBP Involvement: Motif enrichment analysis confirmed that the genomic regions of these aberrantly spliced genes are significantly enriched for NOVA1 and RBFOX1/2/3 binding sites. This was further supported by overlaps with published CLIP-Seq data.

C. The Neurexin-3 (NRXN3) Case Study

• Specific Event: A specific Skipped Exon (SE) event in NRXN3 (involving a non-canonical exon, 20.5) was identified.
  • Control: High inclusion of exon 20.5 (SE-NRXN3-i).
  • PTHS: Significant exclusion of exon 20.5 (SE-NRXN3-e).
• Protein Consequence:
  • SE-NRXN3-i (Control): Includes exon 20.5, which contains a premature stop codon. This results in a truncated protein lacking the transmembrane (TM) domain, producing a soluble/secreted Neurexin-3 isoform.
  • SE-NRXN3-e (PTHS): Excludes exon 20.5, resulting in a full-length transmembrane Neurexin-3 isoform.
• Validation:
  • qPCR: Confirmed a lower ratio of the soluble isoform (SE-NRXN3-i) to the transmembrane isoform (SE-NRXN3-e) in PTHS samples.
  • Western Blot: Showed a significant reduction in the ~40 kDa band (soluble isoform) in PTHS organoids compared to controls, while the ~50 kDa band (transmembrane) was less affected.

5. Significance and Implications

• Pathophysiological Insight: The study provides a molecular explanation for the impaired electrical activity and reduced synapse formation previously observed in PTHS neural tissue. The shift in the Neurexin-3 isoform ratio (less soluble, more transmembrane) likely disrupts the delicate balance required for proper synapse organization and plasticity.
• Broader Relevance: Since Neurexins interact with Neuroligins to control synapse specificity, this mechanism may be relevant to other forms of ASD and neurodevelopmental disorders where splicing regulators are dysregulated.
• Therapeutic Potential: By identifying specific splicing regulators (RBFOX/NOVA) and target genes (NRXN3) as drivers of the phenotype, the study suggests that splicing correction strategies (e.g., antisense oligonucleotides) could be a viable therapeutic avenue for monogenic and potentially multifactorial autism spectrum disorders.

In conclusion, this paper successfully bridges the gap between a monogenic mutation (TCF4), the dysregulation of master splicing regulators, and the functional consequences on synaptic protein isoforms, offering a concrete molecular mechanism for the neural deficits in Pitt-Hopkins Syndrome.