High-resolution cryoEM structure determination of soluble proteins after soft-landing electrospray ion beam deposition

This paper presents a novel ESIBD-based cryoEM workflow that enables high-resolution structure determination of chemically selected soluble proteins by precisely controlling deposition energy to embed them in vitreous ice, yielding near-atomic resolution maps while revealing how solvent exposure influences dehydration-induced structural rearrangements.

The Gist

Imagine you have a very delicate, intricate origami sculpture made of protein. You want to take a perfect, high-resolution photograph of it to see every tiny fold and crease. But there's a problem: to take the photo, you have to put it in a vacuum (like outer space) and freeze it.

Usually, scientists use a method called "plunge freezing," where they dip the protein into liquid nitrogen. It's like dunking a wet sponge into ice water. It works well, but it's a bit chaotic. You can't pick just the specific type of protein you want if your sample is a mix, and the freezing process can sometimes distort the shape.

This paper introduces a new, ultra-precise way to do this called ESIBD (Electrospray Ion Beam Deposition). Think of it as a high-tech, molecular 3D printer that works in a vacuum.

Here is how they did it, broken down into simple steps:

1. The "Molecular Mailman" (The Ion Beam)

Instead of dunking the proteins in water, the scientists turn them into a mist of charged particles (ions) using a gentle spray.

• The Analogy: Imagine a conveyor belt carrying thousands of different toys. Some are the exact toy you want; others are broken or the wrong color.
• The Magic: Before the toys reach the destination, they pass through a "smart filter" (a mass spectrometer). This filter only lets the exact toy you want pass through and blocks everything else. This ensures the sample is 100% pure.

2. The "Gentle Landing" (Soft Landing)

Once the pure proteins are selected, they need to be placed onto a special grid.

• The Problem: If you drop a fragile glass vase from a height, it shatters. If you drop it gently onto a pillow, it stays intact.
• The Solution: The scientists control the speed of the proteins so they "land" on the grid with the gentleness of a feather falling on a pillow. They land at a temperature of -158°C (115 Kelvin), which instantly freezes them in place without damaging their shape.

3. The "Ice Blanket" (Vitreous Ice)

Here is the biggest breakthrough. When proteins land in a vacuum, they are dry. If you just take a picture of a dry protein, the edges look fuzzy and blurry because the air around it is gone.

• The Analogy: Imagine trying to take a photo of a snowflake. If it's dry and floating in the air, it's hard to see the details. But if you gently cover it with a thin, clear sheet of glass, it becomes sharp and stable.
• The Innovation: The scientists figured out exactly how to grow a layer of perfectly smooth, glass-like ice over the proteins. They controlled the temperature and the amount of water vapor like a master chef controlling a soufflé.
  • If it's too cold, the ice grows in weird, bumpy columns (like stalagmites).
  • If it's too warm, the ice turns into crunchy, crystalline snow (which ruins the photo).
  • At just the right temperature (115 K), the ice grows as a flat, smooth, glassy sheet that perfectly wraps the protein.

4. The "X-Ray Vision" (Cryo-EM)

Now that the proteins are pure, gently landed, and wrapped in a perfect ice blanket, they are put under a super-powerful electron microscope.

• The Result: They took pictures of four different complex proteins (including one that helps digest sugar and another that helps plants make food).
• The Discovery: They got incredibly sharp images, almost like seeing individual atoms.
  • The Twist: They noticed that the "insides" of the proteins looked perfect, but the "outsides" (the parts that usually touch water) looked a bit wobbly or rearranged.
  • Why? In water, the outside of a protein is happy and relaxed. When you take away the water (dehydration), the protein tries to hug itself to stay warm, causing the outer parts to shift. It's like a person taking off a heavy winter coat; they might shrug their shoulders or pull their arms in. The scientists could actually see this shrinking and shifting happening in their 3D models.

Why Does This Matter?

This is a game-changer for two reasons:

1. Purity: You can now pick only the specific protein you want from a messy soup of molecules, ignoring the junk.
2. Structure: It links the chemical identity (what the molecule is made of) directly to its 3D shape with extreme precision.

In a nutshell: The scientists built a robotic, vacuum-sealed factory that gently selects, places, and wraps proteins in a perfect ice blanket, allowing us to see their atomic structure with a clarity we've never had before. It's like upgrading from a blurry Polaroid to a 4K HDR photo of the molecular world.

Technical Summary

1. Problem Statement

The integration of Native Mass Spectrometry (MS) and Cryo-Electron Microscopy (cryoEM) has long been a goal in structural biology to link precise chemical identity (stoichiometry, ligand binding, heterogeneity) with near-atomic structural resolution.

• Limitations of Current Methods: Native MS excels at chemical analysis but provides only low-resolution structural data (via ion mobility). CryoEM provides high-resolution structures but requires extensive sample purification and cannot easily distinguish between chemically distinct species within a heterogeneous mixture.
• Limitations of Previous ESIBD Attempts: Soft-landing Electrospray Ion Beam Deposition (ESIBD) allows for the gentle deposition of intact gas-phase ions onto surfaces. However, previous attempts to combine ESIBD with cryoEM suffered from:
  • Inconsistent Ice Embedding: Proteins were often embedded in polycrystalline ice of varying thickness or nanostructured ice, leading to poor contrast and resolution.
  • Contamination: Transfer of samples from the vacuum deposition chamber to the cryoEM microscope introduced contamination and devitrification.
  • Structural Artifacts: Dehydration during the ESI process caused surface rearrangements and loss of resolution, particularly in solvent-exposed regions.

2. Methodology

The authors developed a dedicated instrumentation workflow and a general method to overcome these limitations, enabling high-resolution cryoEM of soluble proteins prepared via ESIBD.

Instrumentation and Workflow

• Modified Mass Spectrometer: A Thermo Scientific Q Exactive UHMR instrument was modified with two differential pumping stages to reach Ultra-High Vacuum (UHV, $p < 10^{-9}$ mbar).
• Ion Beam Control: Intact protein ions generated by native ESI are mass-filtered, thermalized via low-energy collisions with nitrogen gas (narrowing energy distribution to ~1 eV/charge), and steered by electrostatic optics onto a cryogenic sample stage.
• Cryogenic Deposition Stage:
  • Temperature Control: A reverse Stirling cryocooler maintains the sample stage between 60 K and 350 K with $\pm 10$ mK precision.
  • Ice Growth: After protein deposition, water vapor is introduced at a controlled partial pressure ($5 \times 10^{-5}$ mbar) while the stage is held at 115 K. This specific temperature and pressure regime were optimized to grow a thin (10–50 nm), homogeneous, vitreous (amorphous) ice film rather than polycrystalline or nanostructured ice.
  • Contamination Shielding: The stage is designed with no line-of-sight to room-temperature surfaces. A custom cryoshuttle (compatible with Thermo Aquilos) protects grids during transfer, keeping them below the devitrification temperature ($T_g \approx 136$ K) and minimizing atmospheric ice contamination.

Sample Preparation

• Proteins: Four soluble protein complexes were tested: $\beta$-Galactosidase (homotetramer), Glutamate Dehydrogenase (GDH, homohexamer), RuBisCo (hetero-16-mer), and GroEL (homo-14-mer).
• Conditions: Proteins were sprayed at gentle conditions (50°C capillary, no desolvation voltage) to preserve native-like conformations. Mass selection was used to isolate specific oligomeric states and remove impurities.
• Deposition: Ions were deposited at low landing energies (<2 eV/charge) onto amorphous carbon films. Approximately 5–10 billion molecules were deposited to achieve optimal particle density (10–25% monolayer coverage).

3. Key Contributions

1. Generalized ESIBD-cryoEM Workflow: Established a robust, reproducible protocol for preparing soluble proteins for high-resolution cryoEM directly from the gas phase.
2. Optimized Ice Growth: Demonstrated that precise control of stage temperature (115 K) and water vapor pressure yields flat, vitreous ice films, resolving the issue of polycrystalline ice that plagued previous ESIBD-cryoEM attempts.
3. Chemical Purity via Mass Selection: Showed that mass filtering prior to deposition removes stoichiometric impurities and contaminants, resulting in highly homogeneous samples suitable for single-particle analysis (SPA).
4. Correlation of Solvent Exposure and Resolution: Introduced a quantitative "Solvent Exposure Score" to explain structural changes, linking gas-phase dehydration effects directly to local resolution loss.

4. Results

• High-Resolution Structures: The workflow yielded near-atomic resolution cryoEM maps for all four proteins:
  • $\beta$-Galactosidase: 3.09 Å
  • GDH: 2.54 Å (D3 symmetry)
  • RuBisCo: 2.58 Å
  • GroEL: 4.80 Å
• Structural Integrity:
  • Core Retention: Secondary and tertiary structures in the protein cores were largely retained, with high-resolution density matching solution-phase models.
  • Surface Rearrangements: Regions with high solvent exposure in the native state showed lower local resolution or disorder. This is attributed to the loss of water-mediated interactions and the strengthening of intramolecular electrostatic forces upon dehydration.
• Dehydration-Induced Changes:
  • Coherent Changes: Some proteins (e.g., GroEL, $\beta$-Galactosidase) underwent coherent compaction (e.g., closing of cavities, twisting of rings) where the entire domain moved uniformly. These changes were resolvable.
  • Incoherent Changes: Highly flexible, solvent-exposed regions (e.g., GDH "antennae") underwent random, heterogeneous rearrangements, leading to a loss of resolution in those specific areas.
• Ice Morphology Control: Cryo-electron tomography confirmed that ice grown at 115 K was flat and vitreous, whereas ice grown at 93 K formed nanostructured columns and ice grown >120 K was polycrystalline.

5. Significance

• Bridging MS and CryoEM: This work establishes ESIBD+cryoEM as a viable method for structural biology, directly linking the chemical selectivity of native MS (e.g., specific stoichiometry, ligand binding) with high-resolution structural data.
• Understanding Gas-Phase Dynamics: The study provides a microscopic explanation for structural changes observed in the gas phase, demonstrating that dehydration drives compaction and rearrangement primarily in solvent-exposed regions.
• Future Applications: The ability to mass-select and deposit specific species opens doors for studying:
  • Low-abundance protein complexes.
  • Transient protein-ligand interactions.
  • Membrane proteins (an ongoing extension of this work).
  • Structural biology of samples that are difficult to purify via traditional liquid-phase methods.

In conclusion, the authors have successfully transitioned ESIBD from a proof-of-concept for scanning probe microscopy to a robust tool for high-resolution cryoEM, overcoming previous hurdles of ice quality and sample contamination while providing new insights into the structural consequences of protein dehydration.