A Surfactant Cocktail Overcomes Air-Water Interface Artifacts in Single-Particle CryoEM

The authors introduce SurfACT, a versatile cocktail of diverse surfactants that effectively mitigates air-water interface-induced denaturation and preferred orientation bias in single-particle cryoEM, thereby enabling high-quality 3-D reconstructions of challenging biomolecules without the need for laborious, sample-specific screening.

The Gist

The Big Problem: The "Sticky Ceiling" of Science

Imagine you are trying to take a 3D photo of a tiny, delicate sculpture (a protein) using a super-powerful camera. To get a perfect 3D picture, you need to take photos of the sculpture from every possible angle: top, bottom, left, right, and tilted.

In the world of Cryo-EM (a high-tech way to take these photos), scientists freeze these proteins in a thin layer of ice so fast that they don't form crystals. They then shoot electrons at them to take thousands of pictures.

The Catch: When the scientists put the water drop on the grid to freeze it, the water has a top surface and a bottom surface. Think of the top surface as a sticky, invisible ceiling made of air.

When the proteins float in the water, they get sucked up to this "sticky ceiling." Once they hit it, they stick to it like flies on flypaper. Because they are stuck, they all lie flat in the exact same position.

• Result: The camera only sees the "top" of the proteins. It never sees the sides or the bottom.
• The Consequence: When scientists try to build the 3D model, it looks like a flattened pancake. They are missing huge chunks of information, and the model is blurry or incomplete. This is called "Preferred Orientation."

The Old Solution: The "One-Size-Fits-All" Mistake

For years, scientists tried to fix this by adding surfactants (detergents) to the water. Think of surfactants like soap. The idea was that the soap would coat the sticky ceiling, making it slippery so the proteins wouldn't get stuck.

The Problem: Just like how you wouldn't use the same dish soap to wash a delicate silk scarf and a greasy frying pan, different proteins react differently to different soaps.

• Soap A might work great for Protein X but destroy Protein Y.
• Soap B might make Protein Z clump together.
• Scientists had to spend months "shopping around" for the perfect soap for every single new protein they studied. It was slow, expensive, and frustrating.

The New Solution: The "SurfACT" Cocktail

The authors of this paper asked a simple question: "What if we don't pick just one soap? What if we mix a little bit of many different soaps together?"

They created a SurfACT cocktail. Imagine a "smoothie" made of four different types of mild surfactants.

• The Logic: Instead of relying on one ingredient to do everything (which is risky), they use a team. Each ingredient in the mix does a tiny bit of work to stop the proteins from sticking to the ceiling, but none of them are strong enough to hurt the protein on their own.
• The Magic: Because it's a mix, it works like a universal shield. It protects the proteins from the sticky ceiling without needing to be tweaked for every single new protein.

How They Tested It (The "Proof")

They tested this "magic smoothie" on four very different proteins that were known to be difficult:

1. The Influenza Virus (Hemagglutinin):

   • Without the cocktail: The virus looked like a flat coin. You could only see the top.
   • With the cocktail: The virus started tumbling in the ice! The scientists could finally see the "stalk" (the stem) of the virus, which is crucial for making vaccines. It went from a flat pancake to a 3D object.

2. The Nitrogen-Eater (MoFeP):

   • Without the cocktail: A specific part of the protein (the "wing") kept disappearing because it was getting damaged by the sticky ceiling.
   • With the cocktail: The "wing" stayed intact. The scientists could finally see the whole machine, which helps us understand how plants get nutrients from the air.

3. The Sugar-Cutter (Aldolase):

   • Without the cocktail: One of the four legs of this protein was broken or missing in the photos. Scientists usually just "glued" it back together using computer tricks (symmetry), which isn't real science.
   • With the cocktail: All four legs were visible and healthy. The cocktail stopped the protein from breaking in the first place.

The "Secret" Discovery: Where do the proteins go?

The team used a special 3D camera (Cryo-ET) to look inside the ice cubes.

• Before: They saw that almost 100% of the proteins were huddled right against the top and bottom surfaces (the sticky ceilings).
• After: With the SurfACT cocktail, the proteins were happily floating in the middle of the ice (the "bulk ice"). They were tumbling around randomly, just like people at a dance party, instead of being stuck to the walls.

Why This Matters

This paper is a game-changer because:

1. It saves time: Scientists no longer need to spend months hunting for the perfect detergent. They can just add the SurfACT cocktail.
2. It saves money: You get better data with fewer failed experiments.
3. It opens doors: Proteins that were previously considered "too hard to study" or "impossible to image" can now be solved.

In a nutshell: The scientists realized that trying to find one perfect detergent was like trying to find one key that opens every door. Instead, they made a "Master Key" (the cocktail) that works on almost every door, letting us see the hidden 3D structures of life's tiny machines clearly for the first time.

Technical Summary

1. Problem Statement

Single-particle cryogenic electron microscopy (cryoEM) relies on the random distribution of biomolecules in vitreous ice to generate sufficient 2D views for 3D reconstruction. However, a major bottleneck is the Air-Water Interface (AWI). During the vitrification process, particles often interact with the AWI, leading to:

• Preferred Orientation: Particles align in specific orientations (e.g., "top-down" views), resulting in incomplete sampling of Fourier space and missing structural domains.
• Sample Denaturation: Interaction with the AWI can destabilize protein subdomains or cause partial unfolding, leading to "broken" particles or missing density in the final map.
• Current Limitations: Existing solutions involve single surfactant additives (detergents or polymers). However, these require laborious, sample-specific screening. Single surfactants often have stochastic effects, causing aggregation, destabilization, or micelle formation at effective concentrations. Furthermore, there is no consensus on the optimal surfactant or concentration for a given target.

2. Methodology

The authors developed a rational, combinatorial approach to mitigate AWI artifacts, termed SurfACT (SURFactants for Air-water interface ConTrol).

• Rational Design: Instead of relying on a single additive, the authors formulated a low-concentration cocktail of four distinct surfactants with diverse physicochemical properties:
  1. FOM: Fluorinated octyl maltoside (small, fluorinated non-ionic).
  2. Brij-35: Large non-ionic detergent.
  3. CHAPSO: Zwitterionic detergent.
  4. Amphipol A8-35: Amphiphilic polymer.
• Concentration Strategy: All components are used at 0.1-fold of their Critical Micelle Concentration (CMC) (or 0.1-fold protein-solubilizing concentration for A8-35). This ensures the mixture prevents micelle formation (which causes false particle picking) while providing synergistic interfacial coverage.
• Application Protocol: The SurfACT cocktail is pre-diluted into the sample buffer and mixed 1:1 (v/v) with the protein immediately prior to grid freezing. This prevents local high-concentration microenvironments that could destabilize weak protein complexes.
• Validation Systems: The method was tested on four distinct proteins known for AWI issues:
  • Two Influenza Hemagglutinin (HA) trimers (H3N2 and H1N1).
  • Molybdenum-iron protein (MoFeP) from nitrogenase.
  • Rabbit muscle aldolase.
• Techniques:
  • Single-Particle CryoEM (SPA): Used for high-resolution reconstruction and assessment of viewing distribution (using the Conical FSC Area Ratio, cFAR).
  • Cryo-Electron Tomography (cryoET): Used to visualize particle localization within the ice thickness (AWI vs. bulk ice) and to determine if missing density was due to orientation bias or actual denaturation.
  • Freezing Modalities: Tested across SPT Labtech Chameleon (blot-free), Vitrobot (traditional blot-and-plunge), and manual plunge freezing.

3. Key Contributions

• SurfACT Cocktail: The first generalizable, rationally designed surfactant mixture that eliminates the need for extensive screening of individual surfactants for preferred orientation.
• Mechanistic Insight via cryoET: Provided direct visual evidence that SurfACT redistributes particles from the AWI into the bulk ice, rather than just altering surface tension.
• Differentiation of Artifacts: Demonstrated that "missing density" in cryoEM maps can stem from two distinct causes: (1) lack of views (preferred orientation) and (2) physical denaturation at the AWI. SurfACT addresses both.
• Universal Applicability: Validated the method across different freezing hardware (Chameleon, Vitrobot, manual) and diverse protein sizes/complexities.

4. Results

A. Hemagglutinin (HA) Trimers

• Without SurfACT: HA exhibited extreme preferred orientation (dominated by top-down views), resulting in missing stalk domain density and poor cFAR scores (~0.01).
• With SurfACT (0.25×): Drastically improved viewing distribution (side and tilt views became abundant). The cFAR score jumped to ~0.8. The stalk domain density was fully recovered.
• CryoET Findings: Without SurfACT, >90% of particles were localized at the AWI. With SurfACT, particles were redistributed throughout the bulk ice, correlating with the recovery of diverse views.

B. Molybdenum-Iron Protein (MoFeP)

• Without SurfACT: The dynamic $\alpha$III domain (essential for catalysis) was lost due to AWI-induced instability, even at high nominal resolutions.
• With SurfACT (1×): The $\alpha$III domain was fully recovered, allowing visualization of the active site (FeMoco). The map completeness improved significantly, and cFAR increased from 0.29 to 0.80.

C. Rabbit Muscle Aldolase

• Without SurfACT: Exhibited a "broken protomer" phenotype where one subunit lacked continuous density. This was previously attributed to data processing issues but was shown to be AWI-induced denaturation.
• With SurfACT (1×): The broken protomer was fully restored.
• CryoET Confirmation: Particles at the AWI showed the "broken" phenotype and preferred orientation. Particles in the bulk ice (induced by SurfACT) showed intact tetramers and random orientations.
• Resolution: Achieved near-atomic resolution (~1.98–2.16 Å) without enforcing symmetry, whereas previous studies required symmetry enforcement to recover the missing density.

D. General Observations

• Particle Pushing: SurfACT causes a "particle pushing" effect where particles are less concentrated in the center of the grid hole. The authors recommend increasing protein concentration (2x) and imaging slightly off-center to compensate.
• Hardware Independence: SurfACT performed consistently across Chameleon, Vitrobot, and manual plunge freezing, proving it is a sample-additive solution rather than a hardware-specific fix.

5. Significance

• Paradigm Shift: Moves the field from empirical, single-additive screening to a rational, cocktail-based strategy for sample preparation.
• Data Quality: Enables the determination of high-resolution structures for "difficult" targets that were previously considered unsuitable for cryoEM due to preferred orientation or instability.
• Efficiency: Reduces the time and resources required for method development, as a single standard protocol (SurfACT) can be applied to a wide range of targets.
• Structural Biology Impact: By recovering missing domains (like the HA stalk or MoFeP $\alpha$III domain) and preventing denaturation, SurfACT allows for more accurate biological insights into dynamic and fragile macromolecular complexes.

In conclusion, SurfACT represents a versatile, generalizable solution to the pervasive problem of air-water interface artifacts in cryoEM, significantly expanding the scope of biomolecules amenable to near-atomic resolution structure determination.