Modulation of the internal dynamics of the Homer1 EVH1 domain by putative autism-associated mutations

⚕️

This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: A Broken Key in a Locksmith Shop

Imagine the human brain is a massive, bustling city. In this city, there are tiny "construction sites" called synapses, where brain cells talk to each other. To keep these sites organized, the city uses a master foreman named Homer1.

Homer1's job is to hold different construction tools (proteins) together so they can build strong connections. One of Homer1's most important tools is a specific part of his body called the EVH1 domain. Think of the EVH1 domain as a specialized handshake. It reaches out and grabs a specific "glove" (a proline-rich region) on another protein called Shank3. When Homer1 and Shank3 shake hands, they form a stable team that helps the brain learn, remember, and function correctly.

The Problem: Two Typos in the Blueprint

Scientists have found that in some people with Autism Spectrum Disorder (ASD), the blueprint for making Homer1 has two specific typos (mutations):

M65I: A letter in the code changed from "M" to "I".
S97L: A letter changed from "S" to "L".

The big question was: Do these typos break the handshake?

The Investigation: Checking the Shape and the Grip

The researchers decided to build these "typo" versions of Homer1 in a lab and test them using a high-tech toolkit (NMR spectroscopy, X-ray scattering, and computer simulations). They wanted to see three things:

Does the shape change? (Is the hand still a hand?)
Does the grip change? (Can it still hold the glove?)
Does the movement change? (Does the hand shake or jitter differently?)

1. The Shape: Surprisingly Intact

The Analogy: Imagine you have a clay statue of a hand. If you swap one tiny grain of sand inside the clay for a slightly different color, does the whole statue melt and turn into a blob?
The Result: No. The researchers found that both the M65I and S97L mutants looked almost exactly the same as the normal version. The overall "hand" shape was preserved. The "glove" (Shank3) could still fit into the "hand" (EVH1) just fine. The handshake still happened with nearly the same strength.

2. The Grip: Still Strong

The Analogy: If you try to shake hands with a robot that has a typo in its code, does it squeeze too hard or too weakly?
The Result: The grip strength was very similar to the normal version. The mutants could still hold onto their partner protein effectively. This was surprising because usually, when a mutation causes a disease, you expect the protein to fall apart or lose its ability to grab things.

3. The Movement: The Hidden Glitch

The Analogy: This is where the real story lies. Imagine two dancers.

Dancer A (Normal): Moves with a smooth, rhythmic sway.
Dancer B (M65I Mutant): Looks like Dancer A, but their internal muscles are twitching. They are vibrating at a different frequency. It's like a guitar string that is tuned slightly off; it still looks like a guitar, but the sound (the vibration) is wrong.
Dancer C (S97L Mutant): Also has a different internal vibration, but in a different way than Dancer B.

The Result: The study found that while the shape was fine, the internal dynamics (how the protein wiggles and moves on a microscopic scale) were messed up.

The M65I mutation made the protein a bit "wobbly" and less stable (it fell apart easier when heated). It changed how the protein moved in a way that might expose a "secret pocket" that isn't usually open.
The S97L mutation changed the movement in a different, more subtle way.

Why Does This Matter?

Think of a protein like a lock. Usually, we think a mutation breaks the lock so the key (the partner protein) can't get in. But this study suggests a different problem: The lock is still there, and the key still fits, but the tumblers inside are vibrating at the wrong speed.

In the brain, proteins need to move in a very specific rhythm to send signals correctly. If the "wobble" is wrong, the signal might get garbled.

The M65I mutation might be making the protein "jittery," potentially opening up a secret door that shouldn't be open, or making the protein unstable.
The S97L mutation changes the rhythm in a way that might confuse the protein's ability to coordinate with its partners.

The Conclusion

The researchers concluded that these autism-associated mutations don't break the protein's shape or its ability to grab its partner. Instead, they tune the protein's internal rhythm incorrectly.

It's like a car engine that looks perfect and has all the right parts, but the pistons are firing at the wrong time. The car might still drive, but it runs poorly, sputters, and eventually causes the whole system (the brain's social and communication networks) to malfunction.

In short: The problem isn't that the tool is broken; the problem is that the tool is vibrating at the wrong frequency, and that subtle glitch is enough to contribute to autism.

1. Problem Statement

The Homer1 scaffold protein is a critical component of the postsynaptic density (PSD) in excitatory synapses, where it organizes signaling complexes by binding to proline-rich motifs in partners like Shank3 via its N-terminal EVH1 domain. Mutations in the HOMER1 gene, specifically M65I and S97L within the EVH1 domain, have been identified in patients with Autism Spectrum Disorder (ASD). However, the biophysical mechanisms by which these mutations contribute to pathology were unknown. Specifically, it was unclear whether these mutations caused:

Global structural collapse or misfolding.
Disruption of the canonical ligand-binding interface.
Alterations in the protein's internal dynamics (conformational flexibility) that could affect allosteric regulation or binding kinetics.

2. Methodology

The authors employed an integrated approach combining experimental biophysics, structural biology, and computational modeling to characterize the wild-type (WT) and mutant (M65I, S97L) EVH1 domains.

Protein Preparation: Murine Homer1 EVH1 (residues 1–118) mutants were generated via PCR mutagenesis, expressed in E. coli (unlabeled and isotopically labeled for NMR), and purified using IMAC, ion exchange, and size exclusion chromatography.
Structural Characterization:
- NMR Spectroscopy: $^{1}$ H- $^{15}$ N HSQC spectra were recorded for resonance assignment and Chemical Shift Perturbation (CSP) analysis. Backbone dynamics were assessed via $^{15}$ N relaxation ( $T_1$ , $T_2$ , heteronuclear NOE) and model-free analysis. Chemical Exchange Saturation Transfer (CEST) experiments were attempted to probe slow-timescale dynamics.
- Circular Dichroism (CD): Far-UV CD spectroscopy was used to monitor secondary structure and thermal stability.
- Small-Angle X-ray Scattering (SAXS): Used to determine the overall shape and radius of gyration in solution.
- Thermal Shift Assay (TSA): Differential scanning fluorimetry to measure melting temperatures ( $T_m$ ).
Binding Affinity: Biolayer Interferometry (BLI) was used to measure the dissociation constant ( $K_d$ ) of the variants binding to a Shank3 peptide (LVPPPEEFANG).
Computational Modeling:
- Molecular Dynamics (MD): 1 $\mu$ s simulations using GROMACS (AMBER99SB-ILDN force field) to analyze Root Mean Square Fluctuation (RMSF) and conformational stability.
- Gaussian Network Model (GNM): Used to identify hinge residues and collective motions within the domain architecture.

3. Key Contributions

First Biophysical Characterization: This study provides the first comprehensive structural and dynamic analysis of the ASD-associated M65I and S97L mutations in the Homer1 EVH1 domain.
Decoupling Structure from Dynamics: The authors demonstrate that these pathogenic mutations do not alter the global fold or the canonical binding affinity significantly, but rather perturb the internal dynamics of the protein on the microsecond-to-millisecond ( $\mu$ s-ms) timescale.
Identification of Dynamic Hinges: By correlating NMR relaxation data with GNM predictions, the study identifies specific hinge residues (including the mutation site M65) that are critical for the domain's collective motions, suggesting a mechanism for allosteric dysregulation.

4. Key Results

A. Structural Integrity

Global Fold: Both CD and SAXS data confirmed that the overall secondary structure and global shape of the M65I and S97L mutants are virtually identical to the wild type. SAXS fits to the WT structural model yielded low $\chi^2$ values (1.12 for M65I, 1.10 for S97L).
Local Perturbations: NMR $^{1}$ H- $^{15}$ N HSQC spectra revealed that while S97L showed minimal chemical shift changes, M65I caused widespread CSPs, particularly in the 60–78 residue segment, indicating a significant change in the local chemical environment without global unfolding.

B. Thermal Stability

M65I Destabilization: The M65I mutant exhibited thermal destabilization, showing a lower second transition temperature ( $T_m$ ) compared to WT. Its melting curve displayed a distinct two-step unfolding profile with a prominent low-temperature transition.
S97L Stability: The S97L mutant showed a slightly elevated second $T_m$ compared to WT, suggesting it is not thermally destabilized in the same manner as M65I.
Ligand Effect: Binding to the Shank3 peptide did not significantly stabilize or destabilize any of the variants, indicating the mutations do not drastically alter the thermodynamics of the canonical binding event.

C. Binding Affinity

BLI Results: All variants bound the Shank3 peptide with comparable affinity.
- WT $K_d$ : ~5.15 $\mu$ M
- S97L $K_d$ : ~3.88 $\mu$ M (slightly stronger)
- M65I $K_d$ : ~8.88 $\mu$ M (slightly weaker)
The differences were minor (approx. 2-fold), suggesting the mutations do not abolish the primary interaction with Shank3.

D. Internal Dynamics (The Core Finding)

Relaxation Analysis: Model-free analysis of $^{15}$ $^{15}$ N relaxation revealed distinct changes in the $R_{ex}$ term (exchange contribution), which reports on conformational exchange on the $\mu$ $μ$ s-ms timescale.
- M65I: Showed altered $R_{ex}$ patterns near the mutation site (residues 60–78) and in the N-terminal region (residues 14–16).
- S97L: Showed altered $R_{ex}$ near residue 97 and in the N-terminal region, but with a different pattern than M65I.
GNM Correlation: The GNM analysis identified Met65 as a hinge residue for the second major collective motion mode. The mutation M65I likely disrupts this hinge, altering the "breathing" or relative motion between the upper and lower parts of the domain. The M65I mutation affects residues that correspond to hinge sites in the GNM model, whereas S97L affects residues less central to these global modes.

5. Significance and Conclusion

Mechanism of Pathogenicity: The study concludes that the ASD-associated mutations in Homer1 EVH1 likely cause disease not by destroying the protein's structure or its ability to bind Shank3, but by modulating its internal dynamics.
Allosteric Implications: The perturbation of $\mu$ s-ms motions suggests that these mutations may disrupt allosteric communication between the canonical binding site and other regions of the protein.
Cryptic Binding Sites: The authors speculate that the M65I mutation, by altering the dynamics of the $\beta$ 4- $\beta$ 5 loop and adjacent regions, might expose a cryptic binding site or alter the conformation of a secondary binding site (similar to mechanisms seen in ENAH and N-WASP EVH1 domains). This could lead to aberrant interactions with unknown partners in the PSD, disrupting synaptic organization.
Differential Effects: The M65I and S97L mutations have distinct biophysical signatures (M65I is thermally destabilizing and affects global hinges; S97L is more stable but alters local dynamics), implying they may contribute to ASD through different molecular mechanisms.

In summary, this work shifts the paradigm for understanding these ASD mutations from a "structural defect" model to a "dynamical defect" model, highlighting the importance of protein flexibility in synaptic signaling fidelity.