Molecular basis of collagen triple helix recognition by VWF A-like domain 2 of collagen VII: Implications for interlaced anchoring fibril formation

This study elucidates the molecular mechanism by which the von Willebrand factor A-like domain 2 of collagen VII recognizes a Met-Gly-{Phi} motif on fibrillar collagens through specific hydrophobic interactions, thereby facilitating the transient recruitment of collagen I and III into interlaced anchoring fibrils essential for dermal-epidermal stability.

The Gist

The Big Picture: The Skin's "Safety Net"

Imagine your skin is a two-story building. The top floor is the epidermis (the part you see), and the bottom floor is the dermis (the supportive layer underneath).

To keep the roof from falling off the foundation, nature built a special "safety net" called anchoring fibrils. These are made of a protein called Collagen VII. Think of Collagen VII as a giant, flexible lasso or a horseshoe that hooks the roof to the foundation.

However, there's a problem: The dermis is filled with other strong ropes (Collagen I and III). For the safety net to work, the lasso (Collagen VII) needs to catch and wrap around these other ropes. If it doesn't, the skin becomes fragile and blisters easily (a condition called Epidermolysis Bullosa).

The Mystery: Scientists knew the lasso existed, but they didn't know how it grabbed the other ropes. It was like seeing a magnet stick to a fridge, but not knowing which part of the magnet was doing the sticking.

The Investigation: Finding the "Hook"

The researchers wanted to find the specific "hook" on the Collagen VII lasso that grabs the other ropes.

1. The Fishing Trip (Yeast Two-Hybrid Screening):
   Imagine throwing a net full of thousands of random, tiny string patterns into a pond to see what sticks to the hook. The researchers used a computerized "fishing net" (yeast cells) to test millions of random string patterns against the Collagen VII hook.

   • The Catch: They found a specific pattern that stuck very well: Met-Gly-Aromatic.
   • The Surprise: The pattern they found used a "super-sticky" amino acid called Tryptophan (Trp), which doesn't actually exist in human skin ropes. It was like finding a hook that works best with a specific type of super-glue that nature doesn't usually use.

2. The Crystal Snapshot (X-ray Crystallography):
   To see exactly how the hook works, they froze the hook and the super-sticky string together and took a 3D X-ray picture (like a CT scan of a molecule).

   • The Discovery: They saw that the hook has two deep pockets (like a pair of hands).
   • One hand grabs the "Met" part of the string.
   • The other hand grabs the "Aromatic" part (the sticky bit).
   • Crucially, the hook grabs all three strands of the rope at once. It's not just a one-on-one handshake; it's a three-way hug that locks the rope in place.

The Reality Check: Why the Real Skin is Different

The researchers realized that while the "super-sticky" string (with Tryptophan) worked perfectly in the lab, real human skin ropes use Phenylalanine (Phe) instead. Phenylalanine is like a "regular" sticky note, whereas Tryptophan is like "super glue."

• The Simulation: They ran computer simulations to see what happens with the real skin ropes.
• The Result: The hook still grabs the real rope, but it's looser. It's a "transient" or temporary grip. The rope can slip out of one of the hands occasionally, but the other hand holds on tight.

Why is this good?
If the grip were super strong (like super glue), the lasso might get stuck and couldn't move or adjust. Because the grip is "just right" (strong enough to hold, but weak enough to let go), the lasso can slide along the rope, find the right spot, and wrap around it. This is called dynamic binding.

The Solution: How the Skin is Built

Based on these findings, the authors propose a new theory on how the skin's safety net is built:

1. The Dance: Collagen VII (the lasso) is secreted by skin cells. It doesn't just snap onto the foundation immediately.
2. The Catch: It uses its "loose grip" (the A2 domain) to temporarily catch the nearby Collagen I and III ropes.
3. The Weave: Because the grip is temporary, the lasso can slide and weave these ropes through its arches, creating a complex, interlocked structure (like a polyrotaxane, which is a molecular ring on a molecular axle).
4. The Lock: Once the ropes are woven in, the ends of the lasso lock firmly onto the basement membrane (the foundation) using a different, stronger glue.

Why This Matters

• For Disease: In diseases like Epidermolysis Bullosa, mutations break this "hook." Without the ability to grab the ropes, the safety net falls apart, and the skin blisters.
• For the Future: Now that we know exactly what the "hook" looks like and how it grabs, scientists can design drugs or biomaterials that fix broken hooks or create artificial skin that mimics this perfect "loose-but-stable" grip.

In a nutshell: The paper explains that Collagen VII acts like a magnetic lasso that uses a specific "handshake" (the Met-Gly-Phe motif) to temporarily catch and weave other skin fibers together, creating a strong, flexible safety net that keeps your skin from falling apart.

Technical Summary

1. Problem Statement

The dermal–epidermal junction (DEJ) relies on anchoring fibrils, composed primarily of collagen VII, to secure the epidermis to the dermis. These fibrils form a unique arc-shaped structure that encloses mesenchymal collagen fibrils (Collagen I and III). While it is known that the N-terminal non-collagenous (NC1) domain of collagen VII interacts with basement membrane components (like Collagen IV) and that its von Willebrand factor A-like domain 2 (A2 domain) binds to Collagen I and III, the molecular mechanism by which the A2 domain recognizes the triple-helical structure of these collagens remains unclear. Furthermore, the transient nature of these interactions has hindered structural analysis, leaving a gap in understanding how anchoring fibrils recruit and interlace with other collagen fibrils during assembly.

2. Methodology

The authors employed a multi-disciplinary approach combining bioinformatics, structural biology, biophysics, and computational modeling:

• Yeast Two-Hybrid (Y2H) Screening: A triple-helical random peptide library was screened against the human collagen VII A2 domain (bait) to identify binding motifs.
• Peptide Synthesis & Biochemical Assays:
  • Synthetic triple-helical peptides based on Y2H hits and native collagen sequences were created.
  • Competitive ELISA: Used to determine relative binding affinities ($IC_{50}$) of various peptides against the recombinant GST-tagged A2 domain.
  • Structure-Activity Relationship (SAR): Systematic substitution of residues (Pro/Hyp scanning) in high-affinity peptides to identify critical contact points.
• X-ray Crystallography: The A2 domain was co-crystallized with a high-affinity triple-helical peptide containing a Met-Gly-Trp motif (using Norleucine, Nle, as an oxidation-resistant surrogate for Met). Structure was solved at 1.47 Å resolution.
• Molecular Dynamics (MD) Simulations: 100-ns simulations were performed to compare the stability of the A2 domain binding to the high-affinity unnatural peptide (Met-Gly-Trp) versus the native collagen sequence (Met-Gly-Phe).
• Binding Specificity Assays: Pull-down assays using peptide-conjugated beads compared the binding of Collagen VII A2 against homologous domains (VWF A3 and Integrin $\alpha2$I) to assess specificity.

3. Key Contributions & Results

A. Identification of the Binding Motif

• Y2H screening identified a consensus motif: Met-Gly-$\Phi$ (where $\Phi$ is an aromatic amino acid: Trp, Phe, or Tyr).
• While native collagens contain Met-Gly-Phe, the screening yielded high-affinity binders with Met-Gly-Trp.
• Affinity Hierarchy: Biochemical assays revealed that Trp > Phe in terms of binding affinity to the A2 domain. The Met-Gly-Trp peptide showed ~100-fold higher affinity than the Met-Gly-Phe native sequence.

B. Structural Basis of Recognition (Crystal Structure)

• The crystal structure of the A2 domain complexed with the Met-Gly-Trp peptide revealed a unique binding mode:
  • The peptide binds along the $\beta3$ strand of the A2 domain.
  • Two distinct hydrophobic pockets on the A2 domain accommodate the side chains of the motif.
  • Pocket 1: Accommodates the Trp residue from the trailing chain and Met residues from the middle chain.
  • Pocket 2: Accommodates the Trp residue from the middle chain and Met residues from the leading chain.
  • Key Interactions: The interaction involves all three chains of the triple helix. Critical interactions include CH–$\pi$ interactions and a specific hydrogen bond between the indole N$\epsilon1$ of the Trp residue and Ser57 of the A2 domain. This H-bond is absent in Phe, explaining the lower affinity of native sequences.

C. Native Sequence Recognition (MD Simulations)

• MD simulations of the A2 domain with the native Met-Gly-Phe sequence (from Collagen III) showed that while the middle chain Phe remains stably bound in Pocket 2, the trailing chain Phe gradually dissociates from Pocket 1.
• This dissociation occurs because Phe lacks the indole nitrogen required for the hydrogen bond with Ser57.
• Conclusion: The A2 domain recognizes native mesenchymal collagens via a transient, lower-affinity interaction, whereas the high-affinity Trp variant forms a stable complex.

D. Specificity and Disease Relevance

• Specificity: The A2 domain binds the Met-Gly-Phe motif in Collagen III, whereas the homologous VWF A3 domain binds an overlapping region but with different orientation and specificity.
• Disease Link: The A2 domain is the major autoantigen in Epidermolysis Bullosa Acquisita (EBA). Understanding this interface clarifies how autoantibodies disrupt the structural integrity of the DEJ.

4. Significance and Proposed Models

The study proposes a mechanistic model for anchoring fibril assembly:

1. Transient Recruitment: During skin development, the A2 domain of collagen VII (secreted by keratinocytes/fibroblasts) engages in transient, low-affinity interactions with the Met-Gly-Phe motifs on Collagen I and III fibrils.
2. Interlacing: These transient interactions allow the arc-shaped anchoring fibrils to "thread" or recruit mesenchymal collagen fibrils into their structure.
3. Stabilization: Once the fibrils are positioned, the anchoring fibrils are secured to the basement membrane via high-affinity, stable interactions between the NC1 domain and Collagen IV/Laminin 332.
4. Avidity Effect: Although individual A2-Collagen I/III interactions are weak, the multivalency (clustering of multiple A2 domains at the fibril termini) likely provides sufficient avidity to stabilize the assembly.

Clinical Impact:
This work provides the first structural explanation for how collagen VII integrates with the extracellular matrix. It offers a molecular basis for understanding Dystrophic Epidermolysis Bullosa (DEB) (caused by COL7A1 mutations) and EBA (autoimmune). The findings suggest that therapeutic strategies could focus on stabilizing these transient interactions or designing mimetics that enhance the recruitment of collagen fibrils to restore skin integrity.

5. Data Availability

• PDB ID: 9X02 (Collagen VII A2 domain in complex with Short-XGWXGA peptide).
• License: CC-BY 4.0 International.