Topology-aware multiscale modeling of viral genomes reveals stability determinants in circoviruses

This study introduces an integrative computational framework combining AI-based prediction, lattice Monte Carlo simulations, and multiscale molecular dynamics to model the topology of the Porcine Circovirus type 2 genome, revealing that distinct internal genome arrangements with varying stability and stress distributions can produce virions with identical external morphologies.

The Gist

The Big Picture: A Tiny, Tightly Packed Suitcase

Imagine a virus not as a scary monster, but as a tiny, spherical suitcase (the capsid) that is trying to fit a very long, tangled piece of yarn (the viral DNA) inside it.

In the case of the Porcine Circovirus (PCV2), this suitcase is incredibly small—about 20 nanometers wide (roughly the size of a speck of dust). Yet, it has to hold a circular strand of DNA that is about 1,700 "steps" long. If you stretched that DNA out, it would be huge compared to the suitcase. It's like trying to stuff a 100-foot garden hose into a lunchbox.

The Problem:
Scientists have excellent cameras (like Cryo-EM) that can take super-clear photos of the outside of the suitcase. They know exactly what the suitcase looks like. But because the yarn inside is messy, tangled, and doesn't follow a perfect pattern, the cameras can't see it clearly. It's like looking at a closed, opaque suitcase; you know the yarn is in there, but you don't know if it's neatly coiled or knotted into a mess.

The Solution:
Since we can't "see" the yarn clearly, the researchers built a virtual simulation (a computer model) to figure out how the yarn is likely arranged inside. They used a mix of AI, math, and physics to create thousands of different ways the yarn could be packed.

---

The Experiment: Three Ways to Pack the Yarn

The researchers didn't just guess one way the DNA fits. They tested three different "packing strategies" based on how the DNA interacts with the inside of the suitcase:

1. The "Ordered" Pack: Imagine the yarn is neatly coiled, visiting every corner of the suitcase in a specific, logical order. It's like a librarian organizing books on a shelf perfectly.
2. The "Disordered" Pack: Imagine the yarn is thrown in randomly, jumping from one side of the suitcase to the other without a pattern. It's like throwing a ball of yarn into a box and shaking it.
3. The "Intermediate" Pack: A mix of both—some parts are neat, others are messy.

The Surprising Discovery: Same Outside, Different Inside

The most fascinating finding of the paper is this: All three packing methods look exactly the same from the outside.

If you took a photo of the suitcase after packing it using any of the three methods, you wouldn't be able to tell the difference. The suitcase looks round and perfect in all cases.

However, the inside is very different:

• The "Ordered" Suitcase: The yarn is happy. It fits snugly, holds the suitcase walls together, and the whole system is stable. It's like a well-tuned engine.
• The "Disordered" Suitcase: The yarn is stressed. It's pulling on the walls of the suitcase in weird directions. The inside is chaotic, and the suitcase is much more likely to fall apart or break if you heat it up.

The "Melting" Test

To prove this, the researchers simulated heating up the virus, like putting it in an oven.

• The Ordered viruses stayed strong until they reached about 75°C (167°F)—which matches real-world data showing these viruses are very tough.
• The Disordered viruses started falling apart at much lower temperatures.

Why Does This Matter? (The "Quasispecies" Analogy)

Usually, we think of a virus population as a group of clones. But this paper suggests something cooler: A virus population is actually a crowd of twins with different personalities.

Even if every virus has the exact same genetic code (the same DNA sequence), they can pack that DNA differently.

• Think of it like a classroom of students who all have the same textbook.
  • Some students have their book neatly organized (Ordered). They are ready for the test and can handle stress.
  • Others have their book thrown in a messy pile (Disordered). They are stressed and might break under pressure.

The virus doesn't need to mutate (change its DNA) to survive. It just needs to have a mix of packing styles. If the environment gets hot, the "messy" ones might die, but the "neat" ones survive. If the environment changes, maybe the messy ones are better at escaping the host's immune system.

The Takeaway

This study gives us a new way to look at viruses. It's not just about the genetic code; it's about how that code is folded.

By understanding that a virus can exist in many different "folded" states, scientists can better understand:

• Why some viruses are so hard to kill (they have the "neat" packing).
• How they might release their genetic material to infect a cell (uncoiling).
• How to design better drugs or vaccines that target these specific internal structures.

In short: The virus is a master of disguise. It looks the same on the outside, but its internal "personality" (stability) depends entirely on how it packs its luggage.

Technical Summary

1. Problem Statement

Understanding the organization of viral genomes within capsids is a fundamental challenge in structural virology. While high-resolution structures of viral protein capsids (icosahedral symmetry) are readily available via cryo-EM and X-ray crystallography, the internal genome topology remains largely unresolved.

• The Symmetry Problem: Standard reconstruction methods average out asymmetric features. Since viral genomes do not obey icosahedral symmetry, their density is often smeared or eliminated, leaving their 3D coordinates and topological constraints unknown.
• Limitations of Existing Models: Current computational approaches often rely on simplified polymer representations that fail to capture sequence-specific features, topological constraints, and the extreme compaction observed in small DNA viruses like Circoviruses.
• The Case Study: The authors focus on Porcine Circovirus type 2 (PCV2), one of the smallest autonomous mammalian viruses. It packages a circular ~1.7 kb single-stranded DNA (ssDNA) genome into a ~20 nm T=1 icosahedral capsid, achieving one of the highest DNA packing densities in nature. The internal organization of this genome and its impact on virion stability are poorly understood.

2. Methodology

The authors developed an integrative multiscale modeling protocol combining AI-based prediction, lattice Monte Carlo (MC) simulations, and molecular dynamics (MD) to generate and simulate 3D topological models of the complete PCV2 virion.

• Capsid Modeling:

  • Protein Reconstruction: Used comparative modeling (Modeller) and AI-based prediction (ColabFold/AlphaFold 2) to reconstruct the missing N-terminal residues (1–35) of the Cap protein, which are disordered in experimental structures but crucial for genome interaction.
  • Relaxation: Performed energy minimization and structural relaxation (using AMBER22 with the FF14SB force field) to resolve steric clashes in the N-termini while maintaining the overall capsid geometry.

• Genome Topology Generation:

  • Packaging Signals: Identified 60 tetranucleotide packaging signals (Py-Py-Pu-Pu motifs) from experimental structures. These were clustered and mapped to the capsid interior.
  • Lattice Monte Carlo: Represented the genome as a polymer on a lattice defined by the centers of mass of the 60 packaging signals (forming an irregular truncated icosahedron).
  • Hamiltonian Paths: Generated distinct genome topologies by defining Hamiltonian paths that visit all 60 lattice sites exactly once. Three topological classes were defined based on how the ssDNA visits the pentameric symmetry axes:
    1. Ordered: Visits all 12 pentagons consecutively.
    2. Intermediately Ordered: Visits some pentagons consecutively (e.g., 6 or 1).
    3. Disordered: Visits packaging signals without completing any pentamer consecutively (jumping between distant sites).
  • Backmapping: Converted low-resolution lattice models to nucleotide-level coarse-grained (CG) models (using the SPQR model) and finally to all-atom representations.

• Molecular Dynamics Simulations:

  • Force Field: Used the SIRAH 2.0 force field (optimized for DNA-protein interactions) within the AMBER22 engine.
  • Simulation Protocol: Conducted 2 µs production runs at 300K, followed by heating simulations (up to 350K) to assess thermal stability.
  • Analysis: Calculated RMSD, secondary structure content, native contact conservation, and potential energy to compare the stability of different topological arrangements.

3. Key Contributions

• Novel Methodology: Introduced a workflow that bridges AI structural prediction, lattice-based topology sampling, and multiscale MD to model entire viral genomes inside capsids, explicitly accounting for circular topology and sequence-specific packaging signals.
• Topological Diversity: Demonstrated that a single viral genome sequence can adopt multiple distinct 3D topological arrangements (Ordered, Intermediate, Disordered) while maintaining the same external capsid morphology.
• Energetic Heterogeneity: Established that these topologically distinct virions, despite looking identical externally, possess significantly different internal stress distributions and thermodynamic stabilities.

4. Key Results

• Decoupled Dynamics: While the external capsid size (diameter) remained consistent across all models (~19.4 nm, matching experimental data), the internal dynamics differed.
  • Ordered Topologies: Showed high structural stability, low RMSD, and conservation of native protein-protein and protein-DNA contacts. Their behavior closely resembled a capsid filled only with neutralizing ions, suggesting an optimized balance of electrostatics and topology.
  • Disordered Topologies: Exhibited higher structural deviations (RMSD), faster loss of native contacts (especially around 2-fold symmetry axes), and higher internal energy.
• Stability and Melting Points:
  • Energy Analysis: Ordered and Intermediate virions had negative (stable) DNA-protein interaction energies. Some Disordered virions had marginally stable or even positive internal energies, relying heavily on solvation to counteract internal repulsions.
  • Thermal Stability: Heating simulations revealed that Ordered and Intermediate virions have melting temperatures consistent with the known inactivation temperature of PCV2 (~75°C). Disordered virions melted at significantly lower temperatures, indicating they are thermodynamically less stable.
• Contact Conservation: Ordered genomes maintained better conservation of experimental DNA-protein contacts, particularly in the disordered N-terminal regions of the capsid proteins, suggesting that orderly genome packaging stabilizes these flexible regions.

5. Significance and Implications

• Conformational Quasispecies: The study proposes an extension of the "viral quasispecies" concept. Beyond genetic mutations, a viral population may exist as an ensemble of conformational quasispecies—particles with identical genomes but different topological organizations and functional properties (stability, uncoating efficiency).
• Biological Fitness: This topological heterogeneity could provide a fitness advantage. While ordered particles may be more stable and persistent in the environment, disordered particles might uncoat more easily or have different infection kinetics, allowing the virus to adapt to varying selective pressures without genetic mutation.
• General Framework: The methodology provides a generalizable framework for modeling genome topology in other confined viral systems, particularly where experimental data is limited to surface contacts. It highlights the necessity of accounting for physical constraints and topology when interpreting viral architecture and designing viral vectors or nanocarriers.