Cryo-electron tomography reveals paracellular claudin-15 pores at the tight junction

Using cryo-electron tomography on a claudin-deficient epithelial model expressing only CLDN15, this study provides the first direct visualization of paracellular pores within tight junctions, confirming their structural organization and dimensions as predicted by previous models.

The Gist

Imagine your body is a bustling city made of billions of tiny rooms (cells). To keep the city safe, these rooms are separated by walls. But these aren't just solid brick walls; they have special "gates" and "fences" that control what can pass between the rooms.

This paper is about a specific type of gate called the Tight Junction. Think of these as the security checkpoints on the fence between two neighboring houses. Their job is to decide who gets in and who stays out.

The Mystery of the "Claudin" Gates

For a long time, scientists knew that a family of proteins called Claudins acts as the building blocks for these gates. Some Claudins act like solid concrete, sealing the fence completely so nothing gets through (the "Barrier" Claudins). Others act like selective turnstiles, letting specific things like salt or water pass through while blocking others (the "Pore" Claudins).

One specific protein, Claudin-15, is famous for acting like a turnstile that lets Sodium (a type of salt) pass through. Scientists had built computer models and taken 2D pictures suggesting that Claudin-15 proteins line up to form tiny, invisible holes (pores) in the fence.

The Problem: Even though we had blueprints and computer simulations, no one had ever actually seen these holes in a real, living cell. It was like having a map of a secret tunnel but never having walked through it to confirm it exists.

The Experiment: Building a "Super-Simple" City

To solve this, the researchers needed to take a picture of these gates without the "noise" of a busy city. In a normal cell, there are dozens of different types of Claudins mixed together, making it impossible to tell which protein is doing what.

So, they created a special model cell line (a "city") where they deleted almost all the natural gate proteins. Then, they added back only one type: the glowing green Claudin-15.

• The Analogy: Imagine a city where every house has a mix of different colored doors. To study the red doors, the scientists painted every house red and removed all other colors. Now, if they see a red door, they know for sure it's the one they are studying.

The High-Tech Camera: Cryo-Electron Tomography

To see these tiny pores, they couldn't use a regular microscope. The pores are smaller than a virus! Instead, they used a technique called Cryo-Electron Tomography (Cryo-ET).

• The Analogy: Imagine trying to take a 3D photo of a snowflake. If you touch it, it melts. If you use a flash, it might distort.
  • Cryo: They flash-froze the cells instantly (in liquid nitrogen) so they stay in their natural, "alive" state without any chemicals or heat damaging them.
  • Tomography: Instead of taking one flat photo, they took hundreds of pictures from different angles (like spinning a camera around a statue) and used a computer to stitch them together into a 3D movie.

They also used a "FIB" (Focused Ion Beam), which is like a microscopic laser scalpel, to slice the frozen cells into thin, transparent sheets (lamellae) so the camera could see inside.

The Discovery: Seeing the Invisible Holes

When they looked at the 3D models of these frozen cells, they found exactly what they were looking for!

1. The "Kissing" Walls: In the cells with the special gates (Claudin-15), the two cell membranes (the walls of the houses) were pulled very close together, almost kissing.
2. The Pores: Between these two walls, they saw a series of tiny, dark spots arranged in a neat line.
   • The Analogy: Imagine two people standing face-to-face, very close. If you look closely, you can see a string of tiny, glowing beads connecting their noses. Those beads are the pores.
3. The Proof: In the control cells (where they added a different protein, mCherry-ZO-1, instead of Claudin-15), these holes were completely missing. The walls were just smooth. This proved that the holes were made specifically by Claudin-15.

Why This Matters

This is a huge "Aha!" moment for science.

• Before: We had a theory and a computer model saying, "The pores look like this."
• Now: We have a photograph saying, "Yes, the pores are actually there, and they look exactly like the model predicted."

The measurements of the holes (about 1.5 nanometers wide) matched the computer simulations perfectly. This confirms that our understanding of how these biological gates work is correct.

The Future

Now that we have the first clear "photo" of these pores, scientists can use this knowledge to:

• Fix Broken Gates: Many diseases (like inflammatory bowel disease) happen when these gates get damaged or leaky. Knowing exactly what the healthy gate looks like helps us figure out how to fix the broken ones.
• Design New Meds: We can design tiny drugs that act like keys to open or close these specific gates, helping to control how much salt or water moves through our intestines.

In short: The scientists built a simplified model, froze it in time, and took a 3D selfie that finally showed us the secret "turnstiles" that control what passes between our cells. It's the difference between guessing how a lock works and finally seeing the keyhole.

Technical Summary

1. The Problem

Tight junctions (TJs) are critical structures in epithelial tissues that regulate paracellular permeability. While functional studies and molecular dynamics (MD) simulations have long hypothesized that specific claudin proteins (such as CLDN15) form charge- and size-selective ion channels (pores) within TJ strands, these pores have never been directly visualized in their native, intact cellular environment.

• Limitations of prior knowledge: Existing structural data relies heavily on X-ray crystallography of isolated linear CLDN15 protomers, which lack complete extracellular loops and do not capture interactions between adjacent cells.
• The Gap: There is a lack of direct, high-resolution evidence confirming the existence, geometry, and spatial organization of these postulated paracellular pores within the complex 3D architecture of a living epithelium.

2. Methodology

To overcome the limitations of 2D imaging and the complexity of native tissues containing multiple claudin isoforms, the authors employed a rigorous, multi-modal approach:

• Model System: They utilized a MDCK II quintuple-knockout (quinKO) cell line, where endogenous claudins (CLDN1, 2, 3, 4, 7) are deleted. This eliminates heterogeneity.
  • Experimental Group: Inducible expression of EGFP-CLDN15 (the pore-forming claudin).
  • Control Group: Inducible expression of mCherry-ZO-1 (a scaffolding protein, no pore-forming claudin).
• Sample Preparation:
  • Cells were seeded directly onto EM grids and vitrified via plunge freezing to preserve native state.
  • Correlative Light and Electron Microscopy (Cryo-CLEM): Used to locate specific TJ domains based on fluorescent signals (EGFP or mCherry) before electron microscopy.
  • Cryo-Focused Ion Beam (FIB) Milling: Generated 200–220 nm thick lamellae from the vitrified cells to make them electron-transparent.
• Imaging:
  • Cryo-Electron Tomography (Cryo-ET): Performed on a Titan Krios microscope with a K3 detector. Tilt series were acquired (±54°) to reconstruct 3D volumes of the TJ strands.
• Analysis:
  • Segmentation: Manual 3D segmentation of plasma membranes and identification of low-density features.
  • Quantitative Metrics: Measurement of intermembrane spacing, pore diameter, pore-to-pore distance, and membrane thickness using FIJI and Python.
  • Validation: Comparison of experimental measurements against previously published all-atom molecular dynamics models (8-pore and 3-pore CLDN15 strands).

3. Key Contributions

• First Direct Visualization: This study provides the first direct structural evidence of paracellular pores formed by claudin-15 within intact epithelial tight junctions.
• Methodological Advancement: Demonstrates the feasibility of combining Cryo-CLEM, Cryo-FIB milling, and Cryo-ET to resolve macromolecular assemblies (specifically ~2 nm pores) in their native cellular context.
• Structural Validation: Confirms that the geometric parameters of the observed pores (diameter, spacing, and arrangement) align closely with theoretical MD simulations and crystallographic models, validating the "double-row" assembly model of claudins.

4. Key Results

• Membrane Apposition:
  • Expression of CLDN15 significantly reduced the median intermembrane distance at TJ sites to 4.57 nm (compared to 5.88 nm in controls), confirming CLDN15's role in bringing opposing membranes closer to facilitate TJ formation.
• Identification of Pores:
  • In CLDN15-expressing samples, researchers identified linearly distributed, low electron-density foci located between the closely apposed plasma membranes.
  • These features were absent in the mCherry-ZO-1 control samples, confirming they are CLDN15-dependent.
• Quantitative Measurements:
  • Pore Diameter: Median of 1.55 nm (IQR = 0.92), consistent with the ~1.6 nm predicted by MD simulations.
  • Pore Spacing: Median center-to-center distance of 2.25 nm (IQR = 1.83). While slightly variable, this is comparable to the 2.83 nm (3-pore model) and 3.18 nm (8-pore model) predicted computationally.
  • Paracellular Gap Width: Median width of 4.33 nm.
• Statistical Significance: The differences in intermembrane spacing between CLDN15-expressing and control groups were statistically significant ($p < 0.05$).
• Functional Correlation: Electrophysiological data confirmed that CLDN15 expression increased transepithelial resistance and Na+ selectivity, linking the observed structural pores to functional ion conductance.

5. Significance

• Paradigm Shift: This work moves the understanding of tight junctions from theoretical models and indirect functional assays to direct structural observation. It confirms that claudins physically assemble into pore-like structures in situ.
• Foundation for Future Research: The study establishes a validated platform for:
  • Comparing pore-forming (e.g., CLDN2, 15) vs. barrier-forming (e.g., CLDN1, 4) claudins at the ultrastructural level.
  • Investigating how disease states (e.g., inflammatory bowel disease) or inflammation remodel TJ architecture and pore organization.
  • Developing rational therapeutics (small molecules, peptides, antibodies) to modulate epithelial barrier function by targeting specific claudin pore structures.
• Technical Benchmark: It sets a new standard for resolving small (~2 nm) protein complexes within the crowded environment of the cell membrane using cryo-ET, overcoming previous resolution and sampling limitations.

In summary, this paper bridges the gap between computational modeling and biological reality, providing the "smoking gun" evidence that claudin-15 forms the physical pores responsible for paracellular ion selectivity in epithelial tissues.