Epithelial-Mesenchymal Wnt Crosstalk Directs Planar Cell Polarity in the Developing Cochlea

This study reveals that Wnt proteins from both the cochlear epithelium and surrounding periotic mesenchyme act redundantly as global instructive cues to direct cochlear outgrowth and planar cell polarity, ensuring a fail-safe developmental program through epithelial-mesenchymal crosstalk.

The Gist

Imagine the inner ear, specifically the cochlea, as a tiny, spiral-shaped piano. To play music (sound), this piano needs two things:

1. Keys: The tiny "keys" are the hair cells. They need to be perfectly lined up in neat rows.
2. Orientation: Every single key must face the exact same direction (like all piano hammers pointing the same way) to catch the sound waves correctly. If they are twisted or facing the wrong way, the piano is out of tune, and you can't hear.

This paper is about how the body builds this perfect piano during embryonic development. The scientists discovered that the "architects" of this process are tiny chemical messengers called Wnt proteins.

Here is the story of their discovery, broken down into simple concepts:

1. The Two Construction Crews

The cochlea is built by two different teams of cells working side-by-side:

• The Inner Team (Epithelium): These cells form the actual piano keys (the hair cells).
• The Outer Team (Mesenchyme): These cells are the scaffolding and support structure surrounding the piano.

For a long time, scientists thought only the Inner Team sent the instructions on how to line up the keys. They thought the Outer Team just held the building together.

2. The "Secret Handshake" (Wnt Proteins)

The Wnt proteins are like construction blueprints or GPS signals. They tell the hair cells: "Face this way!" and "Grow long enough to make a full spiral!"

The researchers used a clever trick to test this. They built mice where they could "turn off" the ability of these cells to send out the Wnt blueprints.

3. The Experiments: What Happened When We Turned Off the Signals?

Experiment A: Turning off the Inner Team's signals only.

• The Result: The piano was a little shorter than usual, and a few keys were slightly crooked.
• The Lesson: The Inner Team does help, but it's not the whole story. The piano still mostly works, just not perfectly. It's like a construction crew missing a few blueprints; they can guess the rest, but the building is a bit off.

Experiment B: Turning off the signals from just ONE specific blueprint (Wnt5a, Wnt7a, or Wnt7b).

• The Result: Nothing happened! The piano looked perfect.
• The Lesson: This revealed a backup system. The cells have multiple blueprints (Wnt5a, Wnt7a, Wnt7b). If you lose one, the others step in and say, "Don't worry, we got this!" This is called redundancy. Nature loves having a "fail-safe" plan.

Experiment C: Turning off the signals from BOTH the Inner and Outer Teams.

• The Result: Disaster. The piano was extremely short (like a tiny, stunted spiral), and the keys were completely twisted and facing the wrong way. The "GPS" was gone entirely.
• The Lesson: This was the big "Aha!" moment. The scientists realized that both teams are sending blueprints. The Outer Team (Mesenchyme) was secretly sending just as many instructions as the Inner Team. When you cut off both sources, the building collapses.

4. The "Global vs. Local" Analogy

The paper proposes a beautiful model for how this works:

• The Core PCP Proteins (The Local Foremen): These are the workers on the hair cells. They talk to their immediate neighbors to make sure the key next to them is aligned. They handle the local details.
• The Wnt Proteins (The Global GPS): These are the signals coming from the surrounding environment. They tell the entire group of keys which way is "North." They provide the global direction.

If you only have the Local Foremen but no Global GPS, the keys might align with each other, but the whole group might be rotated the wrong way. If you have the Global GPS, the whole group knows where to face, even if the local workers are a bit confused.

The Big Takeaway

This study changes how we understand hearing development. It shows that the body doesn't rely on just one source of information. Instead, it uses a redundant, dual-team system:

1. The inner cells talk to each other.
2. The outer cells send signals to the inner cells.
3. They use multiple types of signals (Wnt5a, Wnt7a, Wnt7b) so that if one fails, the others save the day.

Why does this matter?
Understanding this "fail-safe" system helps scientists figure out why some people are born with hearing loss. If a genetic mutation knocks out one of these backup signals, the other signals might compensate, and the person hears fine. But if multiple signals are knocked out, or if the timing is wrong, the "piano" is built crooked, leading to deafness.

In short: Building a perfect ear requires a team effort from both the inside and the outside, with plenty of backup plans to ensure the music plays correctly.

Technical Summary

1. Problem Statement

The mammalian cochlea is a spiral-shaped sensory organ where hair cells must be precisely oriented (Planar Cell Polarity or PCP) to transduce sound. While the core PCP protein machinery (e.g., Vangl2, Fzd6, Dvl2) is well-characterized, the role of secreted Wnt ligands in directing cochlear PCP and outgrowth has been controversial.

• The Controversy: Previous studies showed that deleting Wnt ligands or the Wnt-secretion chaperone Wntless (Wls) in the cochlear epithelium caused only mild hair cell misorientation and partial loss of PCP protein polarization, unlike the severe defects seen in classic PCP mutants (e.g., Vangl2 mutants).
• The Gap: It remained unclear whether Wnts act as global instructive cues for PCP, whether redundancy among Wnt ligands masks their function, and whether Wnts from the surrounding periotic mesenchyme (POM) contribute to epithelial PCP.

2. Methodology

The authors employed a multi-faceted approach combining computational biology, genetic mouse models, and rigorous phenotypic analysis:

• Computational Cell-Cell Communication Analysis:

  • Integrated single-cell RNA sequencing (scRNA-seq) datasets from embryonic mouse cochleae (E13.5–E16.5).
  • Used the CellChat algorithm to infer Wnt ligand-receptor interactions between specific cell clusters: Hair Cells (HCs), Prosensory cells, Cochlear Epithelium (roof, LER, GER), and Periotic Mesenchyme (POM).
  • Identified top candidate Wnt ligands acting on hair cells from both epithelial and mesenchymal sources.

• Genetic Mouse Models (Conditional Knockouts):

  • Epithelial-specific deletion: Used Emx2-Cre to delete Wntless (Wls) or specific Wnt ligands (Wnt5a, Wnt7a, Wnt7b) in the cochlear epithelium.
  • Mesenchymal-specific deletion: Used Dermo1-Cre to delete Wnt5a specifically in the POM.
  • Dual Epithelial-Mesenchymal deletion: Used Sox9-CreERT2 (inducible) to delete Wls in both the cochlear epithelium and POM simultaneously. Tamoxifen was administered at E10.5–E11.5 (early) or E11.5–E12.5 (late) to test temporal sensitivity.
  • Combinatorial Deletion: Generated single, double, and triple conditional knockouts (cKO) of Wnt5a, Wnt7a, and Wnt7b in the epithelium to test for redundancy.

• Phenotypic Analysis:

  • Morphometry: Measured cochlear duct length and hair cell counts at E18.5.
  • PCP Assessment: Used phalloidin staining to measure hair bundle orientation (circular mean and variance) and immunofluorescence to assess the polarized localization of core PCP proteins (Fzd6, Dvl2, Vangl1, Vangl2, Celsr1).
  • Proliferation: Assessed cell proliferation via Ki67 and EdU labeling to distinguish between growth defects and cell death/proliferation failure.
  • Statistical Analysis: Employed Bayesian mixed models for circular data (hair cell orientation) and standard t-tests/ANOVA for linear measurements.

3. Key Contributions & Results

A. Identification of Candidate Wnts via Computational Analysis

• Analysis of scRNA-seq data predicted that Wnt5a, Wnt7a, and Wnt7b are the dominant Wnt ligands secreted by the cochlear epithelium, while Wnt5a is the primary ligand secreted by the POM acting on hair cells.

B. Redundancy in the Cochlear Epithelium

• Single/Double Deletions: Deleting Wnt5a, Wnt7a, or Wnt7b individually, or in double combinations, resulted in no significant defects in hair cell orientation or PCP protein polarization, though some double deletions caused mild cochlear shortening.
• Triple Deletion: Simultaneous deletion of Wnt5a, Wnt7a, and Wnt7b in the epithelium caused significant cochlear shortening (~30%) and reduced hair cell counts. However, hair cell orientation remained largely normal, and core PCP proteins (Fzd6, Dvl2) retained their polarized localization.
• Conclusion: Epithelial Wnts are additive for cochlear outgrowth but are insufficient to cause severe PCP defects when mesenchymal Wnts are present.

C. Mesenchymal Wnts are Insufficient Alone

• Deleting Wnt5a specifically in the POM (Dermo1-Wnt5a cKO) resulted in no defects in cochlear length, hair cell orientation, or PCP protein polarization.
• Conclusion: Mesenchymal Wnts alone cannot drive PCP; they require epithelial input.

D. Severe Defects upon Dual Ablation (The "Fail-Safe" Mechanism)

• Sox9-Wls cKO (Epithelium + POM): When Wnt secretion was blocked in both compartments (using Sox9-CreERT2 at E10.5), the phenotype was dramatically severe:
  • Cochlear Shortening: ~70% reduction in length (more severe than Vangl2 mutants).
  • Hair Cell Counts: ~60% reduction.
  • Convergent Extension: Hair cells formed 5–6 rows instead of the normal 4, indicating a failure in tissue narrowing.
  • PCP Defects: Hair cells were severely malrotated, predominantly toward the apex (unlike the random rotation seen in Vangl2 mutants).
  • Protein Mislocalization: All core PCP proteins (Fzd6, Dvl2, Vangl1, Vangl2, Celsr1) lost their polarized expression.
• Temporal Sensitivity: Delaying the deletion to E11.5–E12.5 resulted in milder defects, indicating a critical window for Wnt signaling around E10.5–E11.5.
• Proliferation: Proliferation rates (Ki67/EdU) were unchanged, confirming that the defects were due to disrupted morphogenesis (outgrowth/convergent extension) rather than cell loss.

4. Significance and Model

• Resolution of Controversy: The study resolves the debate regarding Wnts in cochlear PCP. Previous mild phenotypes were due to redundancy between epithelial and mesenchymal Wnt sources. When both sources are removed, Wnts are revealed as essential global instructive cues.
• Dual Regulation Model: The authors propose a model where:
  1. Core PCP proteins (e.g., Vangl2) provide local instructive cues to limit variance in hair cell orientation.
  2. Secreted Wnts from both epithelial and mesenchymal compartments provide a global instructive cue that directs the overall axis of hair cell polarization (preventing apex-directed malrotation) and drives cochlear outgrowth.
• Developmental Insight: The findings highlight a "fail-safe" developmental program where distinct Wnt ligands across different tissue compartments ensure robust cochlear patterning. Disruption of this crosstalk leads to severe structural and functional hearing deficits.

In summary, this paper demonstrates that epithelial-mesenchymal Wnt crosstalk is non-redundant in the context of global tissue organization, acting as a master regulator for both cochlear elongation and the precise orientation of sensory hair cells.