Structure of SARS-CoV-2 spike in complex with its co-receptor the neuronal cell adhesion protein contactin 1

This study elucidates the molecular mechanism of SARS-CoV-2 neurotropism by resolving the cryo-EM structure and quantifying the nanomolar affinity interaction between the viral spike protein and the neuronal co-receptor contactin 1, revealing how the virus binds to the nervous system via a unique interface that overlaps with neutralizing antibody epitopes.

The Gist

The Big Picture: A Virus with a "Backdoor" Key

Imagine the SARS-CoV-2 virus as a burglar trying to break into a house (your body).

• The Main Door: The virus usually tries to get in through the front door, which is a lock called ACE2. Everyone knows about this key.
• The Backdoor: This new study discovered that the virus also has a secret "backdoor" key that fits a specific lock found mostly in the brain and nervous system. This lock is a protein called Contactin 1 (CNTN1).

The researchers wanted to know: How does this backdoor key work? Does it fit perfectly? And why does this matter for the neurological symptoms (like "brain fog" or long-COVID) that many people experience?

The Discovery: Finding the Fit

The scientists used two main tools to solve this mystery:

1. A Molecular Scale (SPR): They measured how tightly the virus's "spike" (the part that grabs onto cells) sticks to the brain protein.

   • The Result: They stick together very tightly, like Velcro. In fact, because the virus is a three-pronged object (a trimer), it can grab onto multiple brain proteins at once, making the grip even stronger. This is called an avidity effect—like a spider using three legs to hold on to a web instead of just one.

2. A Molecular Camera (Cryo-EM): They took a 3D "photo" of the virus spike and the brain protein locked together.

   • The Result: They saw exactly where they touch. It's like seeing a puzzle piece snap into place.

The "How It Works" Analogy: The Umbrella and the Wedge

Think of the virus's Spike protein as a three-pronged umbrella that can open and close.

• The Prongs (RBDs): Each prong can point up (open) or down (closed).
• The Brain Protein (CNTN1): This protein looks like a horseshoe or a curved spoon.

The Interaction:
When the virus tries to enter a nerve cell, one of its prongs opens up. The "horseshoe" brain protein slides in and wedges itself between two of the prongs.

• It doesn't just touch the tip of the prong; it wedges into the base of the prong.
• This is unique! The main door key (ACE2) touches the tip of the prong. This backdoor key (CNTN1) touches the base.

Why This Matters: The "Traffic Jam" Effect

Here is the clever part of the mechanism:
Because the brain protein wedges itself between the prongs, it acts like a doorstop.

1. It locks the prong it is touching in the "open" position.
2. However, because it is wedged in the middle, it physically blocks the neighboring prong from opening up easily.

The Consequence:
This creates a weird traffic jam. The virus can't easily switch all its prongs to the "open" position at once to fuse with the cell. This might explain why the virus is tricky to fight in the brain—it's a slow, complex process that might trigger different immune responses or cause damage over time, contributing to Long COVID.

The "Good News" and "Bad News"

The Good News (The Backdoor is Specific):
The researchers checked if this backdoor key works for the original SARS virus (SARS-CoV-1). It doesn't. The SARS-CoV-1 virus has a slightly different shape on its prongs, so the brain protein (CNTN1) doesn't fit. This explains why the original SARS virus didn't cause as many neurological issues as the new virus does.

The Bad News (The Shield Problem):
The spot where this brain protein grabs the virus is also a spot where some of our body's "security guards" (antibodies) try to attack.

• Imagine the virus is wearing a jacket. The brain protein grabs the jacket near the pocket.
• Some of our antibodies try to grab the jacket at the same pocket.
• If the brain protein is already holding the jacket, the antibody can't get a good grip. This might make it harder for our immune system to neutralize the virus once it's already inside the nervous system.

Summary in One Sentence

This study reveals that the SARS-CoV-2 virus has a special, high-strength "backdoor" key that fits specifically into brain cells, wedging itself in a unique spot that helps explain why the virus can cause long-term neurological damage and why it behaves differently than the original SARS virus.

Technical Summary

1. Problem Statement

While SARS-CoV-2 is primarily known for respiratory symptoms, it exhibits significant neurotropism (affinity for nervous tissue) and is linked to long-term neurological sequelae (Long-COVID). The virus enters cells via the Spike (S) protein binding to the primary receptor, Angiotensin-Converting Enzyme 2 (ACE2). However, the mechanisms facilitating neuroinvasion remain partially understood.

• Knowledge Gap: Previous studies identified Contactin 1 (CNTN1), a neuronal cell adhesion protein, as a potential co-receptor that enhances SARS-CoV-2 infection in neuronal cells. However, the molecular details of the interaction between the SARS-CoV-2 Spike protein and CNTN1, including binding affinity, stoichiometry, and structural interface, were previously elusive.
• Specific Questions: How does Spike bind CNTN1? Which domains are involved? Does this interaction differ from ACE2 binding? Can a single Spike trimer bind multiple CNTN1 molecules or simultaneous co-receptors?

2. Methodology

The authors employed a multi-disciplinary approach combining biophysics and structural biology:

• Protein Expression & Purification:
  • Generated prefusion-stabilized SARS-CoV-2 Spike trimers (6P construct, Wuhan-Hu-1 strain).
  • Produced various CNTN1 constructs (full ectodomain, Ig1-6, and FN1-4) in HEK293 cells.
• Surface Plasmon Resonance (SPR):
  • Used to quantify binding affinity ($K_D$).
  • CNTN1 constructs were immobilized on sensor chips to mimic cell-surface topology; Spike was used as the analyte.
  • Data was fitted to 1:1 Langmuir and two-state binding models to analyze monovalent vs. multivalent interactions.
• Single-Particle Cryo-Electron Microscopy (cryo-EM):
  • Prepared complexes of Spike trimer and full extracellular CNTN1 (CNTN1fe).
  • Data collected on a 300 kV Titan Krios microscope.
  • Processing: Used RELION v5.0.0 for motion correction, CTF estimation, and 3D classification.
  • Refinement: Employed focused 3D classification and refinement to isolate the Spike-CNTN1 complex. Local refinement was applied to the flexible RBD-CNTN1 interface to improve resolution.
• Structural Analysis:
  • Rigid body fitting of existing PDB structures (Spike: 7DWX; CNTN1 Ig1-4: 7OL2) into the cryo-EM density maps.
  • Comparison with known receptor structures (ACE2, TMEM106B) and neutralizing antibody epitopes.

3. Key Contributions

• First Structural Resolution: This study provides the first high-resolution structural view of the SARS-CoV-2 Spike protein in complex with the neuronal co-receptor CNTN1.
• Binding Interface Definition: It precisely maps the interaction site, revealing that CNTN1 binds to a unique, previously uncharacterized interface at the base of the Receptor Binding Domain (RBD) in the "up" conformation.
• Mechanism of Avidity: The study elucidates how the trimeric nature of Spike and the flexibility of CNTN1 allow for multivalent binding, significantly enhancing affinity through an avidity effect.
• Variant Analysis: The authors analyzed how mutations in Variants of Concern (Alpha, Beta, Gamma, Omicron) and differences between SARS-CoV-1 and SARS-CoV-2 affect CNTN1 binding.

4. Key Results

A. Binding Affinity and Stoichiometry

• Affinity: SPR analysis revealed a high-affinity interaction with a $K_D$ of 67 nM.
• Avidity Effect: At higher concentrations, a two-state binding model indicated a secondary, lower-affinity interaction ($K_D \approx 12 \mu M$). This suggests that while a single Spike-CNTN1 interaction is strong, the trimeric Spike can bind multiple CNTN1 molecules (up to three) simultaneously, driving the high overall avidity observed.
• Domain Specificity: Binding is strictly dependent on the N-terminal Ig1-4 domains of CNTN1 (forming a "horseshoe" shape). Constructs lacking these domains (FN1-4) showed no interaction.

B. Structural Insights (Cryo-EM)

• Complex Architecture: The structure shows one CNTN1 molecule wedged between two RBDs of the Spike trimer: one RBD in the up-conformation ($RBD_{up}$) and one in the down-conformation ($RBD_{down}$).
• Primary Interface: The main contact occurs between the CNTN1 Ig3 domain and the $RBD_{up}$.
  • Buried Surface Area: ~1,500 Ų.
  • Key Residues: Involves RBD $\beta$-strand $\beta2$, helices $\alpha2$ and $\alpha5$, interacting with CNTN1 Ig3 $\beta$-strands A and G.
• Secondary Interface: A secondary, weaker interface exists between the $RBD_{down}$ and CNTN1 Ig2-3 (~1,400 Ų), which appears flexible and less critical than the primary $RBD_{up}$ interaction.
• Conformational Dynamics: Binding of CNTN1 does not induce major conformational changes in CNTN1 itself but requires the RBD to be in the "up" state. The binding site is located at the base of the RBD, distinct from the tip where ACE2 binds.

C. Comparison with Other Receptors and Antibodies

• Unique Site: The CNTN1 binding site is distinct from the ACE2 binding site (which targets the Receptor Binding Motif at the RBD tip) and the TMEM106B site.
• Co-receptor Compatibility: Structural modeling suggests a single Spike trimer can simultaneously bind CNTN1 and ACE2 on different protomers (e.g., one RBD bound to CNTN1, another to ACE2), provided multiple RBDs are in the "up" state.
• Antibody Interference: The CNTN1 interface significantly overlaps with Class 4 neutralizing antibodies (e.g., CR3022, S309, VHH-72). This suggests that these antibodies could potentially block CNTN1-mediated neuroinvasion. Class 1 and 3 antibodies may also cause steric clashes due to their extended Fab structures.

D. SARS-CoV-1 and Variant Analysis

• SARS-CoV-1: The study confirms why SARS-CoV-1 does not bind CNTN1. Five specific residue differences in the RBD interface (T372A, F373S, A384P, R439N, I503V) and the absence of a specific glycan motif likely prevent binding.
• Omicron Variant: The Omicron variant carries mutations (S373P, S375F) within the CNTN1 interface. These mutations, along with S371L, stabilize the "one-RBD-up" conformation, potentially reducing the flexibility required for CNTN1 binding and possibly explaining reduced neurotropism in some neuronal models compared to earlier variants.

5. Significance

• Mechanism of Neurotropism: This work provides a structural basis for understanding how SARS-CoV-2 invades the central nervous system, identifying CNTN1 as a critical co-receptor that facilitates infection in neuronal cells.
• Therapeutic Implications:
  • The identification of the specific interface offers a target for small molecule inhibitors or peptide blockers designed to disrupt Spike-CNTN1 binding without necessarily blocking ACE2 (potentially preserving ACE2's physiological functions).
  • The overlap with Class 4 antibodies suggests that existing or engineered monoclonal antibodies could be repurposed to mitigate neurological complications of COVID-19.
• Viral Evolution: The findings highlight how viral evolution (mutations in variants like Omicron) may inadvertently alter neurotropism by changing the accessibility of co-receptor binding sites.
• Multivalency: The study underscores the importance of avidity in viral entry, demonstrating how the trimeric nature of Spike allows for robust binding to cell-surface co-receptors even if individual affinities are moderate.

In summary, this paper resolves the structural mystery of SARS-CoV-2's interaction with a key neuronal co-receptor, offering critical insights into the molecular drivers of neurological disease and potential avenues for therapeutic intervention.