Revealing pH-dependence and independence of the characteristics of a β sheet-forming antimicrobial peptide

This study utilizes constant pH molecular dynamics simulations to reveal how pH-dependent deprotonation of lysine residues, particularly LYS11, modulates the conformational dynamics of the antimicrobial peptide GL13K, identifying a therapeutically relevant $\beta$-hairpin configuration stabilized under conditions of partial protonation.

The Gist

Imagine a tiny, 13-step-long molecular "soldier" called GL13K. Its job is to hunt down and destroy harmful bacteria. But this soldier has a secret superpower: it can change its shape depending on the acidity (pH) of the water it's swimming in.

This paper is like a high-tech detective story where scientists used a computer simulation to watch this soldier change its outfit and behavior as the water got more and more alkaline (less acidic).

Here is the story of what they found, explained simply:

1. The Soldier's "Charge" and the "Acidity Switch"

Think of the GL13K peptide as a magnet. It has four special spots (called Lysine residues) that act like little switches.

• In normal water (neutral pH): These switches are "ON." They are positively charged, like four tiny magnets all pushing away from each other. Because they repel each other, the soldier stays stretched out and floppy, like a loose piece of string.
• In alkaline water (high pH): The water acts like a chemical eraser. It turns these switches "OFF." The positive charges disappear. Suddenly, the magnets stop pushing away, and the soldier can curl up, fold, or snap into a tight shape.

2. The Mystery of the "Special Switch" (Lysine 11)

The scientists had a hunch. They knew that one specific switch, located near the soldier's tail (called Lysine 11), was the "boss" that controlled whether the soldiers would stick together to form a deadly net (fibrils) or stay apart.

• The Old Theory: They thought Lysine 11 was a "weak" switch that turned off first (at a lower pH) compared to the others. They thought, "If we turn off the boss first, the whole group changes shape."
• The Investigation: They ran a super-accurate computer simulation (called CpHMD) that let the switches flip on and off in real-time as they changed the water's pH.
• The Surprise: The old theory was wrong! Lysine 11 didn't turn off first. In fact, it was the stubbornest of all. It held onto its charge longer than the others. It was actually the last to turn off, not the first.

The Analogy: Imagine a group of four friends holding hands. You thought the friend at the end (Lysine 11) would let go first. But instead, the friends in the middle let go first, and the friend at the end held on tight until the very last moment.

3. The Shape-Shifting Dance

As the water became more alkaline and the switches turned off, the soldier didn't just shrink; it did a specific dance:

• Low pH (Switches ON): The soldier is a floppy, random noodle.
• Medium pH (The "Sweet Spot"): Around pH 10.5 (where the switches are halfway between on and off), the soldier suddenly snaps into a β-hairpin.
  • What is a β-hairpin? Imagine a piece of string folding back on itself to look like a hairpin clip. This is a very specific, stable shape.
• High pH (Switches OFF): The soldier becomes a tight, collapsed ball.

Why does this matter? The scientists found that this "hairpin" shape is the Goldilocks zone. It's the shape that seems best at grabbing onto the enemy (bacteria) and neutralizing their toxins. It happens right when the switches are partially turning off, not when they are fully on or fully off.

4. The "Hook" in the Hairpin

Here is the most exciting part. When the soldier folds into that perfect hairpin shape, a specific part of its body (a residue called Serine 9) ends up right at the tip of the hook.

• Scientists already knew this "hook" is what grabs onto the bacteria's poison (LPS).
• The study suggests that this perfect hook shape only forms when the pH is just right (between 10 and 11).

The Big Takeaway

This paper tells us that to design better medicine, we can't just assume the soldier is always in one shape. We have to know exactly what the environment (pH) is doing.

• The Lesson: If you want to stop bacteria, you might need to create a drug that works best when the environment is slightly alkaline, because that's when the soldier folds into its most effective "hairpin" weapon.
• The Future: The scientists are now suggesting we test this in real life: "Does the soldier grab the bacteria's poison best at pH 10.5?" If yes, we can design new drugs that target this specific moment to disarm bacteria without killing them (which stops them from becoming resistant).

In a nutshell: The scientists used a computer to watch a tiny protein soldier change its outfit. They found a "sweet spot" in the water's acidity where the soldier folds into a perfect weapon shape, and they discovered that one specific part of the soldier is stubborn and holds its ground longer than the rest, playing a key role in this transformation.

Technical Summary

1. Problem Statement

Antimicrobial peptides (AMPs), such as the synthetic 13-residue peptide GL13K, are promising candidates to combat antimicrobial resistance (AMR). Their efficacy relies on their ability to disrupt bacterial membranes, a process governed by their charge state and conformational dynamics.

• The Challenge: The charge state of AMPs is highly sensitive to local pH. Traditional Molecular Dynamics (MD) simulations typically fix amino acid charges at their most probable state (e.g., protonated lysine at neutral pH), failing to capture the dynamic protonation/deprotonation events that occur as pH changes.
• Specific Context: GL13K (sequence: GKIIKLKASLKLL) exhibits pH-dependent behavior: it remains a random coil at neutral pH but forms $\beta$-sheets and fibrils at higher pH. Previous hypotheses suggested that a specific lysine residue on the hydrophobic face (LYS11) controls this transition by having a significantly lower pKa than other lysines, allowing it to deprotonate first and trigger aggregation.
• Goal: To rigorously characterize the pH-dependent protonation states of GL13K's lysine residues and their impact on the peptide's structural dynamics using advanced computational methods.

2. Methodology

The authors employed Constant pH Molecular Dynamics (CpHMD) simulations, an advanced technique that allows titratable residues to change protonation states dynamically during the simulation based on the environmental pH.

• Simulation Setup:
  • System: A single GL13K peptide in a water box (TIP3P model) with physiological salt (0.15 M NaCl).
  • Force Field: Charmm36m (modified for CpHMD).
  • Software: GROMACS 2021 with the $\lambda$-dynamics package.
  • Conditions: 10 distinct pH levels (8.0 to 12.5) centered around the standard lysine pKa (10.5).
  • Replicates: 5 independent replicates per pH level, totaling 50 simulations. Each production run was 100 ns (analyzing the last 90 ns).
• Analysis Techniques:
  • pKa Calculation: Used the Henderson-Hasselbalch equation fitted to the average deprotonation fraction ($S_{dep}$) derived from $\lambda$-dynamics trajectories.
  • Structural Metrics: Calculated Radius of Gyration ($R_g$), End-to-End distance ($r_{e2e}$), Asphericity, and Acylindricity to assess compactness and shape.
  • Secondary Structure: Used DSSP to quantify $\alpha$-helix, $\beta$-strand, and random coil content.
  • Dynamics & Landscapes: Constructed Markov State Models (MSMs) using Time-lagged Independent Component Analysis (TICA) and Principal Component Analysis (PCA) to identify metastable states and free energy landscapes (FES).

3. Key Contributions

1. Methodological Application: Successfully applied CpHMD to a short AMP to resolve the dynamic interplay between protonation states and conformational ensembles, overcoming the limitations of fixed-charge simulations.
2. Refutation of Prior Hypotheses: Challenged the prevailing hypothesis that LYS11 has a uniquely lower pKa driving aggregation.
3. Discovery of Differential pKa: Demonstrated that while most lysines share similar pKa values, LYS11 actually possesses a slightly higher pKa than LYS5, leading to a distinct deprotonation profile.
4. Conformational Mapping: Mapped the transition from random coils to $\beta$-hairpins as a function of pH, identifying a specific "therapeutically relevant" $\beta$-hairpin configuration stabilized near the pKa of the lysines.

4. Key Results

A. pKa Values and Protonation States

• Calculated pKa Values:
  • LYS2: ~10.21
  • LYS5: ~10.12
  • LYS7: ~10.19
  • LYS11: ~10.31
• Statistical Finding: Contrary to the initial hypothesis, LYS11 does not have a lower pKa. Instead, it has a statistically significant but small higher pKa compared to LYS5.
• Implication: LYS11 remains protonated (charged) slightly longer than LYS5 as pH increases. This suggests that LYS11 is not the primary trigger for deprotonation-driven aggregation; rather, the simultaneous deprotonation of LYS2, LYS5, and LYS7 is likely the controlling factor.

B. Conformational Dynamics vs. pH

• Compactness: As pH increases (crossing the pKa ~10.5), the peptide becomes more compact.
  • Radius of Gyration ($R_g$) and End-to-End distance ($r_{e2e}$) decrease significantly at pH > 10 due to the neutralization of positive charges, reducing electrostatic repulsion.
• Secondary Structure:
  • pH < 10: The peptide is predominantly a random coil with some transient $\alpha$-helical content.
  • pH $\ge$ 10.5: A distinct increase in $\beta$-strand content is observed.
  • Critical Transition: The emergence of $\beta$-sheet structures coincides with the partial deprotonation of the lysine residues.
• Metastable States (MSM Analysis):
  • Free Energy Landscapes revealed subtle shifts in accessible states.
  • At pH 10.5, a stable $\beta$-hairpin configuration emerges (specifically populated when PC1 $\approx$ 1, PC2 $\approx$ -1).
  • This $\beta$-hairpin is distinct from the random coils seen at lower pH and the collapsed states at very high pH.

C. The Role of LYS11 and SER9

• The $\beta$-hairpin configuration places the residue SER9 at the "top of the hook."
• SER9 is known to be critical for binding to Lipopolysaccharide (LPS), a component of Gram-negative bacterial cell walls.
• The study suggests that the $\beta$-hairpin, stabilized by the specific protonation state of lysines near pH 10–11, may be the active conformation for LPS binding.

5. Significance and Future Directions

• Theoretical Foundation: The study provides a rigorous basis for modeling GL13K, establishing that fixed-charge simulations are insufficient near pH 10.5 where protonation states are dynamic and critical for structure.
• Therapeutic Insight: The identification of a pH-dependent $\beta$-hairpin offers a new target for drug design. If this conformation is indeed the LPS-binding state, optimizing peptides to stabilize this hairpin at physiological or infection-site pH could enhance efficacy against endotoxins without necessarily killing bacteria (reducing resistance pressure).
• Experimental Proposal: The authors propose a testable hypothesis: LPS binding affinity should peak between pH 10 and 11, corresponding to the stabilization of the $\beta$-hairpin.
• Limitations & Next Steps: The current work is limited to a single peptide in solution. Future work must address peptide aggregation and membrane interactions at these specific pH levels to fully understand the mechanism of action in biological environments.

In summary, this paper utilizes CpHMD to correct misconceptions about the pKa hierarchy of GL13K lysines and reveals a nuanced pH-dependent mechanism where a specific $\beta$-hairpin, stabilized near the lysine pKa, may be crucial for the peptide's antimicrobial and endotoxin-neutralizing functions.