The Notch1/Delta-like-4 axis is crucial for the initiation and progression of oral squamous cell carcinoma

This study demonstrates that the Notch1/Delta-like-4 signaling axis acts as a central oncogenic driver in oral squamous cell carcinoma, characterized by specific expression patterns distinct from Jagged1 that emerge early in tumor development and promote cancer progression.

The Gist

Imagine your mouth is a bustling city, and the cells lining your tongue are the workers building and maintaining the city walls. In a healthy city, these workers follow a strict schedule: they start as apprentices (stem cells) at the bottom, learn their trade, and gradually move up to become skilled, specialized workers (differentiated cells) who form a strong, protective barrier.

This paper is about what happens when the "foreman" of this construction crew gets confused, leading to a chaotic, uncontrolled expansion that turns into a dangerous tumor called Oral Squamous Cell Carcinoma (OSCC).

Here is the story of the discovery, explained simply:

1. The Confused Foreman (The Notch Pathway)

In our city, there is a communication system called the Notch pathway. Think of it as the radio network the foreman uses to tell workers, "Stop building, you're done, now go polish the wall."

Usually, this system works perfectly. But in cancer, the radio gets jammed. The workers never get the "stop" signal. They keep multiplying, staying as unskilled apprentices forever, piling up into a messy, dangerous mound.

2. The Two Walkie-Talkies: Dll4 vs. Jagged1

The foreman has two different walkie-talkies to talk to the workers:

• Walkie-Talkie A (Jagged1): In a healthy mouth, this one is used to tell workers to grow up and become skilled.
• Walkie-Talkie B (Delta-like-4 or Dll4): This one is usually quiet.

The Big Discovery: The researchers found that in the cancer city, the workers completely stop listening to Walkie-Talkie A. Instead, they are constantly screaming into Walkie-Talkie B.

• The Switch: As soon as the cancer starts, the "Jagged1" signal disappears, and the "Dll4" signal takes over 100% of the communication.
• The Result: Because the workers are only listening to the "Dll4" signal, they never learn their trade. They stay as chaotic, unskilled apprentices forever, building a massive, undifferentiated tumor.

3. The City Spreads (Invasion)

As the tumor grows, it doesn't just sit there; it starts to break down the city's foundation (the basement membrane) and invade the neighboring districts (the muscles and lymph nodes). This is called metastasis.

The researchers found that even as the cancer cells break out and run away to invade new areas, they still keep using that same "Dll4" walkie-talkie. They never switch back to the healthy signal. This means the "Dll4" signal is the engine driving the cancer from the very first day until it becomes a deadly invader.

4. The "Off" Switch (The Experiment)

To prove this theory, the scientists tried to jam the "Dll4" walkie-talkie using a special drug called CB103.

• Before the drug: The cancer cells were like wild, fast runners, zooming around and invading everything.
• After the drug: When they jammed the signal, the cells suddenly stopped running. They started acting like normal, healthy workers again. They began to stick together tightly (like a good neighborhood) and stopped moving around.
• The Lesson: If you cut the communication line for "Dll4," the cancer loses its ability to move and invade.

5. Why This Matters

This study is like finding the specific key that unlocks the cancer's power.

• Early Detection: Because this "Dll4" switch flips right at the very beginning of the disease, doctors might be able to use it as a warning light to catch cancer before it spreads.
• New Treatments: Instead of trying to kill all the cells (which hurts healthy ones too), doctors could potentially use drugs to specifically jam the "Dll4" walkie-talkie. This would force the cancer cells to "grow up," stop moving, and become harmless.

In a nutshell: The cancer cells are stuck in a "childish" state because they are listening to the wrong radio station (Dll4). If we can change the station or turn the radio off, the cancer stops being a runaway train and becomes a manageable situation.

Technical Summary

1. Problem Statement

Oral Squamous Cell Carcinoma (OSCC) is the most common malignancy of the oral mucosa and a subset of Head and Neck Squamous Cell Carcinoma (HNSCC). Despite multimodal therapies, prognosis remains poor due to high rates of recurrence and metastasis.

• The Knowledge Gap: The Notch signaling pathway is frequently mutated in OSCC (approx. 15% of cases), yet its precise functional role is controversial. Literature presents conflicting evidence regarding whether Notch acts as an oncogene or a tumor suppressor in this context.
• Specific Question: How do specific Notch receptor-ligand interactions (specifically Notch1 with Delta-like-4 [Dll4] vs. Jagged1 [Jag1]) drive the initiation, maintenance of undifferentiation, and progression (including Epithelial-to-Mesenchymal Transition/EMT) of OSCC?

2. Methodology

The study employed a multi-modal approach combining in vivo murine models, ex vivo organoid cultures, human clinical samples, and in vitro cell line experiments.

• Murine Models:
  • Lineage Tracing: Used Notch1CreERT:R26mTmGfl/fl transgenic mice. Tamoxifen (TAM) induction activated Cre recombinase, labeling Notch1-expressing cells with GFP (switching from mTomato to mGFP).
  • Carcinogenesis: OSCC was induced via chronic administration of 4-Nitroquinoline-1-oxide (4NQO) in drinking water (50 µg/mL) for 16 weeks, followed by a washout period. This model recapitulates the progression from hyperplasia to dysplasia, carcinoma in situ, and invasive carcinoma (analyzed at 22, 32, and 38 weeks).
  • Imaging: Utilized Mesoscale Selective Plane Illumination Microscopy (mesoSPIM) for 3D whole-organ imaging of cleared tongue tissues to visualize basal lamina integrity and tumor invasion.
• Human Samples & Cell Lines:
  • Analyzed HPV-negative HNSCC patient tissues (surgical samples) for Jag1 and Dll4 expression via immunohistochemistry.
  • Used the human OSCC cell line SCC-25 for in vitro functional assays.
• Pharmacological Intervention:
  • Treated cells and organoids with CB103, a pan-Notch inhibitor targeting the Notch transcriptional activation complex.
• Omics & Molecular Analysis:
  • Transcriptomics: Bulk RNA sequencing (RNA-seq) of 4NQO-treated vs. control mouse epithelia and CB103-treated vs. DMSO-treated SCC-25 cells.
  • Bioinformatics: Gene Set Enrichment Analysis (GSEA), Gene Ontology (GO) analysis, and Reactome pathway enrichment to identify dysregulated pathways (e.g., differentiation, EMT, adhesion).
  • Validation: qRT-PCR, Western blot/Immunofluorescence for markers (K5, K10, Sox2, p63, Runx1, Vimentin, Snail, E-Cadherin).
• Functional Assays:
  • Organoid Formation: Assessed stem cell capacity in tongue epithelial organoids.
  • Migration: Scratch assays and time-lapse imaging to measure cell motility.
  • Adhesion: Immunofluorescence for focal adhesions (Vinculin) and desmosomes.

3. Key Contributions

• Resolution of Notch Paradox: The study clarifies the conflicting role of Notch in OSCC by demonstrating that the pathway acts as an oncogenic driver specifically through the Notch1/Dll4 axis, rather than a general Notch activation.
• Ligand Specificity: It identifies a mutually exclusive expression pattern between Dll4 and Jag1. While Jag1 is downregulated in malignant tissue, Dll4 is upregulated and confined to the tumor mass from the earliest hyperplastic stages through to invasive carcinoma.
• Mechanistic Link to Undifferentiation: Establishes a direct correlation between the Notch1/Dll4 axis and the maintenance of a progenitor-like, undifferentiated state (high Sox2, K5, p63; low K10).
• Therapeutic Validation: Demonstrates that pharmacological inhibition of Notch (via CB103) reverses the malignant phenotype, reducing cell motility and promoting differentiation.

4. Detailed Results

A. Physiological vs. Pathological Distribution

• Normal Tissue: Notch1 is broadly expressed in the stratified epithelium. Ligands show compartmentalization: Jag1 is found in inter-papillary spaces, while Dll4 is restricted to the outer surface of filiform papillae.
• OSCC Initiation (22 weeks): In 4NQO-treated mice, hyperplastic regions showed specific Notch1 activation. Crucially, Dll4 was upregulated in the hyperplastic mass, while Jag1 was downregulated (mutually exclusive). The mass consisted of undifferentiated cells (Sox2+, K5+, K10-).
• Progression (32-38 weeks): The Notch1/Dll4 signature persisted through carcinoma in situ and invasive stages.
  • EMT Onset: By 32 weeks, basal lamina degradation was observed via SPIM. By 38 weeks, invasive clusters were found in the mesenchyme.
  • Invasive Phenotype: Invasive cells retained Dll4 expression. They exhibited EMT markers (Snail, Vimentin, ZEB1/2) and co-expression of E-Cadherin and Vimentin, confirming an ongoing EMT process.
  • Transcriptional Regulation: RNA-seq revealed upregulation of Runx1 (linked to Notch) and downregulation of Hey1/Hey2 (Notch targets), while Hes family genes remained stable. This suggests a specific rewiring of the Notch downstream circuitry in OSCC.

B. Human Relevance

• Immunohistochemistry on human HPV-negative OSCC samples confirmed the murine findings: DLL4 showed diffuse, high expression in hyperplastic and carcinoma nests, whereas JAG1 expression was variable and often reduced in tumor cores.

C. Functional Impact of Notch Inhibition (CB103)

• Organoids: CB103 treatment completely disrupted organoid formation (average diameter dropped from ~122 µm to ~20 µm), indicating a loss of stem cell self-renewal capacity.
• Gene Expression: In SCC-25 cells, CB103 treatment:
  • Downregulated DLL4, SOX2, and p63.
  • Upregulated differentiation markers and keratinization genes.
  • Induced a gene signature associated with desmosomes and cell-cell adhesion (e.g., desmoplakin, desmoglein).
• Cell Motility:
  • Scratch Assay: CB103-treated cells failed to close the wound gap over 10 days, unlike DMSO controls.
  • Time-lapse: Quantitative tracking showed significantly reduced displacement length in treated cells.
  • Mechanism: Inhibition of Notch increased the number of adhesive complexes (focal adhesions and desmosomes), effectively "locking" cells in place and preventing the migration required for invasion.

5. Significance and Conclusion

• Diagnostic Potential: The mutually exclusive expression of Dll4 (high) and Jag1 (low) serves as a robust molecular signature for early detection of OSCC, detectable even at the hyperplastic stage.
• Therapeutic Target: The study validates the Notch1/Dll4 axis as a central oncogenic driver in OSCC. Unlike other cancers where Notch inhibition might be detrimental, blocking this axis in OSCC promotes differentiation and suppresses metastasis.
• Clinical Implication: The pan-Notch inhibitor CB103 shows promise as a therapeutic agent. By targeting the Notch1/Dll4 axis, it can potentially reverse the undifferentiated state of cancer cells and inhibit their invasive motility, offering a new strategy to improve prognosis in OSCC patients.

In summary, this paper redefines the role of Notch in oral cancer, moving from a general "pathway dysregulation" view to a specific Notch1-Dll4 driven mechanism that sustains stemness and drives invasion, providing a clear target for future clinical interventions.