Investigating the Bioactive Conformation of Angiotensin II Using Markov State Modeling Revisited with Web-Scale Clustering

Emmanouil Christoforou, Hari Leontiadou, Frank Noé, Jannis Samios, Ioannis Z. Emiris, Zoe CourniaJournal of Chemical Theory and Computation 2022 · DOI

Abstract

Molecular dynamics simulation is a powerful technique for studying the structure and dynamics of biomolecules in atomic-level detail by sampling their various conformations in real time. Because of the long timescales that need to be sampled to study biomolecular processes and the big and complex nature of the corresponding data, relevant analyses of important biophysical phenomena are challenging. Clustering and Markov state models (MSMs) are efficient computational techniques that can be used to extract dominant conformational states and to connect those with kinetic information. In this work, we perform Molecular Dynamics simulations to investigate the free energy landscape of Angiotensin II (AngII) in order to unravel its bioactive conformations using different clustering techniques and Markov state modeling. AngII is an octapeptide hormone, which binds to the AT1 transmembrane receptor, and plays a vital role in the regulation of blood pressure, conservation of total blood volume, and salt homeostasis. To mimic the water-membrane interface as AngII approaches the AT1 receptor and to compare our findings with available experimental results, the simulations were performed in water as well as in water-ethanol mixtures. Our results show that in the water-ethanol environment, AngII adopts more compact U-shaped (folded) conformations than in water, which resembles its structure when bound to the AT1 receptor. For clustering of the conformations, we validate the efficiency of an inverted-quantized k-means algorithm, as a fast approximate clustering technique for web-scale data (millions of points into thousands or millions of clusters) compared to k-means, on data from trajectories of molecular dynamics simulations with reasonable trade-offs between time and accuracy. Finally, we extract MSMs using various clustering techniques for the generation of microstates and macrostates and for the selection of the macrostate representatives.

Introduction

Understanding how flexible peptides like Angiotensin II (AngII) transition between conformational states is a cornerstone of rational drug design. AngII is a key hormone in the renin-angiotensin system, responsible for regulating blood pressure. However, identifying its "bioactive" conformation—the specific shape it adopts to trigger the AT1 receptor—has been a long-standing challenge due to the peptide's high flexibility.

To address this, researchers utilize Molecular Dynamics (MD) simulations to generate massive datasets of conformational snapshots. Analyzing these datasets to find stable states and transition kinetics requires Markov State Models (MSMs). A significant bottleneck in this process is the clustering of millions of data points into "microstates," where traditional algorithms like $k$-means often fail to scale efficiently.

Methodology

This study combines high-resolution MD simulations with advanced data science techniques to map the free energy landscape of AngII.

Simulation Environments

AngII was simulated in two distinct environments to observe conformational shifts:

  1. Pure Water: A standard aqueous environment.
  2. Water-Ethanol Mixture: A 35% ethanol solution used to mimic the dielectric environment of a cell membrane interface.

Simulations were conducted across a range of temperatures (278K to 323K) to capture temperature-dependent folding dynamics, totaling 8 $\mu$s of simulation time.

Web-Scale Clustering with IQ-means

The technical novelty of this work lies in the application of IQ-means (Inverted-Quantized $k$-means). Originally designed for massive image databases, IQ-means uses an inverted index and a 2D grid to drastically accelerate clustering. Unlike standard $k$-means, which requires exhaustive distance calculations for every point, IQ-means provides a "web-scale" solution that scales more effectively as the number of simulation snapshots grows into the millions.

Results and Discussion

The Bioactive U-Shape and Membrane Mimicry

The simulations revealed that the solvent environment plays a critical role in "pre-organizing" the peptide's structure. In pure water, AngII exists primarily in extended (E) or intermediate (I) $eta$-structures. However, the membrane-mimicking water-ethanol environment acts as a catalyst for folding.

Comparison of Angiotensin II conformational populations (Extended, U-shaped, and Intermediate) in water versus water-ethanol environments.

As illustrated above, the presence of ethanol encourages the peptide to adopt a compact U-shaped (U) conformation. This folded state is believed to be the bioactive form required for binding to the AT1 receptor. At human body temperature (310K) and above, this U-shaped state becomes significantly more stable in the membrane mimic, suggesting that AngII prepares its shape as it approaches the cell surface, even before direct receptor contact.

Computational Performance

IQ-means demonstrated a clear advantage in processing speed. For a dataset of 100,000 snapshots, IQ-means achieved a ~2.5x speedup over standard $k$-means (5.6 seconds vs. 14.2 seconds). Crucially, the MSMs constructed using these approximate clusters were statistically consistent with those built using traditional methods, confirming that the speedup does not come at the cost of biological or kinetic accuracy.

Conclusion

By adapting "web-scale" clustering algorithms to the field of biophysics, this research provides a scalable framework for analyzing massive MD datasets. The findings support the hypothesis that the membrane interface stabilizes the bioactive U-shaped conformation of Angiotensin II. Future work will look to apply these accelerated modeling techniques to even more complex systems, such as explicit phospholipid bilayers, to further refine our understanding of peptide-receptor signaling.

Summary updated

Abstract

Molecular dynamics simulation is a powerful technique for studying the structure and dynamics of biomolecules in atomic-level detail by sampling their various conformations in real time. Because of the long timescales that need to be sampled to study biomolecular processes and the big and complex nature of the corresponding data, relevant analyses of important biophysical phenomena are challenging. Clustering and Markov state models (MSMs) are efficient computational techniques that can be used to extract dominant conformational states and to connect those with kinetic information. In this work, we perform Molecular Dynamics simulations to investigate the free energy landscape of Angiotensin II (AngII) in order to unravel its bioactive conformations using different clustering techniques and Markov state modeling. AngII is an octapeptide hormone, which binds to the AT1 transmembrane receptor, and plays a vital role in the regulation of blood pressure, conservation of total blood volume, and salt homeostasis. To mimic the water-membrane interface as AngII approaches the AT1 receptor and to compare our findings with available experimental results, the simulations were performed in water as well as in water-ethanol mixtures. Our results show that in the water-ethanol environment, AngII adopts more compact U-shaped (folded) conformations than in water, which resembles its structure when bound to the AT1 receptor. For clustering of the conformations, we validate the efficiency of an inverted-quantized <i>k</i>-means algorithm, as a fast approximate clustering technique for web-scale data (millions of points into thousands or millions of clusters) compared to <i>k</i>-means, on data from trajectories of molecular dynamics simulations with reasonable trade-offs between time and accuracy. Finally, we extract MSMs using various clustering techniques for the generation of microstates and macrostates and for the selection of the macrostate representatives.
BibTeX
@article{christoforou2022investigating,
  author = {Emmanouil Christoforou and Hari Leontiadou and Frank Noé and Jannis Samios and Ioannis Z. Emiris and Zoe Cournia},
  title = {{Investigating the Bioactive Conformation of Angiotensin II Using Markov State Modeling Revisited with Web-Scale Clustering}},
  journal = {Journal of Chemical Theory and Computation},
  year = {2022},
  doi = {10.1021/acs.jctc.1c00881},
}
Figure