An Automated HDX-MS Platform for in situ characterisation of Membrane Proteins

This paper introduces the first fully automated HDX-MS platform featuring a two-stage delipidation workflow that enables the characterization of membrane protein dynamics, such as those of the ABC transporter MsbA, within their native lipid environments, revealing physiologically relevant motions obscured by traditional detergent-based methods.

The Gist

The Big Picture: The "Greasy" Problem

Imagine you are trying to study a very important machine part (a membrane protein) that lives inside a cell. This part is essential for keeping the cell alive, but it has a problem: it is permanently stuck in a thick, sticky layer of grease (the lipid bilayer or cell membrane).

For decades, scientists have tried to study these parts by ripping them out of the grease and washing them in a soapy solution (detergent). It's like taking a greasy engine part, scrubbing it with soap, and then trying to figure out how it moves. The problem is, once you wash off the grease, the part changes shape. It doesn't behave the way it does when it's actually working inside the engine.

The Goal: The scientists wanted to study these proteins while they are still sitting in their natural grease, without washing them off.

The Challenge: The "Grease Trap"

They used a powerful tool called HDX-MS (Hydrogen-Deuterium Exchange Mass Spectrometry). Think of this tool as a high-speed camera that takes "snapshots" of how a protein wiggles and moves.

However, there was a major roadblock. The "grease" (lipids) is so thick and messy that it clogs the camera lens. If you try to take a picture of the protein while it's covered in grease, the image comes out blurry, or the machine breaks down. Previous attempts to clean the grease off were like trying to wipe a greasy window with a dirty rag—it was slow, manual, and often left the window still dirty or damaged the glass (the protein).

The Solution: The "Automated Car Wash"

The team built a fully automated platform that acts like a high-tech, two-stage car wash for these proteins.

1. Stage 1: The Magnetic Sponge (ZrO2 Beads): First, the sample goes through a filter made of special beads that act like a magnetic sponge. They specifically grab onto the "sticky" parts of the grease (the charged heads of the lipids) and pull them out, leaving the protein behind.
2. Stage 2: The Size-Selecting Gate (SEC): The sample then flows through a second gate (a Size Exclusion Chromatography column). Imagine a hallway with a bouncer. The bouncer only lets the "small" things (the protein pieces) through, while the "big" clumps of remaining grease and dirt are kicked out the back door.

By combining these two stages, they created a Dual-Mode Delipidation system. It's so efficient that it can handle samples that are 99% grease and still get a clean picture of the protein.

The Experiment: Testing the "MsbA" Machine

To test their new car wash, they used a famous protein called MsbA. MsbA is like a flipper in the bacterial cell wall; it grabs lipids and flips them from the inside to the outside to build the cell's armor.

They tested MsbA in two scenarios:

1. In the Soap (DDM): The old way (protein washed in detergent).
2. In the Grease (IIMVs): The new way (protein still in the native membrane).

They watched how MsbA moved when it was "resting" (apo state) and when it was "working" (using energy from ATP).

The Discovery: The "Hidden" Moves

Here is the exciting part. The new platform revealed things the old method missed:

• The Old Way (Soap): When MsbA was in the soap, it looked very stiff and stable. It seemed to lock into one specific shape when it worked.
• The New Way (Grease): When MsbA was in its natural grease, it was much more dynamic. It showed movements and "wiggles" that the soap had hidden.
  • The "Open" Secret: The data suggested that in the real membrane, the protein opens up wider to release its cargo than scientists previously thought.
  • The "Back" Door: They noticed that the back part of the protein (the C-terminal) became unstable and wiggly when it finished its job. This suggests the protein needs to loosen up to reset itself for the next round.

Why This Matters

Think of it like studying a dancer.

• The Old Method: You take the dancer out of the studio, put them in a stiff suit of armor (detergent), and ask them to dance. They can move, but they look stiff and unnatural.
• The New Method: You watch the dancer in their natural studio, wearing their normal clothes. You see the subtle twirls, the flexible knees, and the way they interact with the floor.

The Takeaway:
This paper introduces the first fully automated "car wash" that lets scientists study membrane proteins in their natural, greasy environment. It proves that context matters. Proteins behave differently in their native home than they do when washed in soap. This new tool will allow scientists to see the "real" movements of these proteins, leading to better drugs and a deeper understanding of how life works at a molecular level.

Technical Summary

1. The Problem

Membrane proteins are critical biological targets, but studying their structural dynamics within their native lipid environment remains a significant challenge.

• Limitations of Current Methods: Traditional structural biology often relies on detergent solubilization or reconstitution into lipid mimetics (e.g., nanodiscs), which can alter protein conformation and dynamics.
• HDX-MS Bottlenecks: While Hydrogen–Deuterium Exchange Mass Spectrometry (HDX-MS) is powerful for probing dynamics, applying it to native membrane vesicles is difficult. The lipid bilayer introduces complexity that interferes with downstream LC-MS analysis, causing:
  • Reduced enzymatic digestion efficiency.
  • Peptide ionization suppression.
  • Increased spectral complexity.
  • Impaired chromatographic performance.
• Existing Solutions: Previous delipidation strategies (e.g., using Zirconium Dioxide beads or manual Size Exclusion Chromatography) are often labor-intensive, lack reproducibility, suffer from non-specific protein binding, or fail to remove all contaminants (non-phospholipids, detergents, chaotropic agents) from complex samples like isolated inner membrane vesicles (IIMVs).
• Automation Gap: There is a lack of fully automated workflows capable of handling high-lipid-content samples (like bacterial inner membranes) while maintaining high peptide coverage and low deuterium back-exchange.

2. Methodology: The Dual-Mode Delipidation (DMD) Platform

The authors developed the first fully automated HDX-MS platform featuring a two-stage delipidation workflow integrated into a LEAP HDX robot system.

• Hardware Configuration:
  • A dual-head LEAP HDX robot equipped with a filtration system.
  • A novel fluidic system utilizing three valves and three LC pumps to control online Size Exclusion Chromatography (SEC).
• The Dual-Mode Delipidation (DMD) Workflow:
  1. Stage 1: ZrO₂ Filtration: Samples are passed through Zirconium Dioxide (ZrO₂) coated silica beads. These beads bind negatively charged phospholipid head groups, removing the bulk of the lipid content.
  2. Stage 2: Online SEC: The filtrate is immediately subjected to online Size Exclusion Chromatography. The SEC column separates remaining lipids and contaminants (which elute later) from intact proteins and peptides (which elute earlier).
  3. Valve Switching: A precise valve switching mechanism diverts the lipid-rich fraction to waste while directing the protein/peptide fraction to the protease column for digestion and subsequent LC-MS analysis.
• Automation Features:
  • The system automates sample handling, mixing, delipidation, digestion, and column regeneration/cleaning.
  • Includes rigorous cleaning protocols (pepsin wash, SEC wash, and gradient cleaning) to prevent carry-over between runs.
• Model Systems:
  • Validation: Used Bovine Carbonic Anhydrase (BCA) and POPC (lipid) to test efficiency.
  • Biological Application: Applied to Escherichia coli Isolated Inner Membrane Vesicles (IIMVs) containing the ABC transporter MsbA and the sugar transporter XylE.
  • States Analyzed: MsbA was studied in its Apo state (no nucleotide) and Post-hydrolytic state (ATP + Vanadate/ADP*Vi) to capture conformational changes during the transport cycle.

3. Key Contributions

• First Fully Automated In Situ HDX-MS Platform: The paper reports the first end-to-end automated system capable of processing complex, lipid-dense membrane vesicles without manual intervention.
• Dual-Mode Delipidation (DMD): Demonstrated that combining ZrO₂ filtration with online SEC significantly outperforms either method alone, achieving near-complete lipid removal (99% of 3500 pmol POPC) while maintaining sufficient protein recovery.
• Robustness and Reproducibility: The platform minimizes deuterium back-exchange (average ~47%, within community recommendations) and ensures high peptide sequence coverage even in complex lipid environments.
• Standardization: The authors provide a framework for reporting protein:lipid ratios and expression levels, addressing a gap in reproducibility for membrane protein HDX-MS.

4. Key Results

• Efficiency: The DMD workflow removed 99% of model lipids with 44–48% protein recovery, which is sufficient for high-quality LC-MS analysis. Back-exchange levels remained low (<5% increase compared to non-delipidated controls).
• Peptide Coverage:
  • Achieved 98.3% sequence coverage for MsbA and 93.9% for XylE in IIMVs.
  • Successfully identified 64 peptides for MsbA in IIMVs (60.3% coverage after curation), covering both Transmembrane Domains (TMDs) and Nucleotide Binding Domains (NBDs).
• MsbA Dynamics (In Situ vs. In Vitro):
  • Apo State: IIMVs showed a mixture of inward-facing (IF) conformations, with some regions showing lower deuterium uptake compared to detergent-solubilized MsbA, suggesting a more stable or heterogeneous population in the native membrane.
  • ATP/Vi State (Stabilized):
    • Conserved Stabilization: Both DDM and IIMV samples showed protection (stabilization) at the NBD dimerization interface and core TMDs upon nucleotide binding.
    • Novel In Situ Features: IIMVs revealed deprotection (increased dynamics) in specific regions not seen in detergent:
      • The surface of the NBDs (non-interface regions).
      • The C-terminal $\alpha$-helix of the NBD.
      • The extracellular loop between TMH1 and TMH2.
  • Conformational Insight: The data supports an Open Outward-Facing (OFopen) conformation in the native membrane, distinct from the fully occluded states often seen in nanodiscs. The increased dynamics at the NBD surface and C-terminus in IIMVs suggest a mechanism for NBD separation and substrate release that is masked in detergent environments.

5. Significance

• Physiological Relevance: The study proves that studying membrane proteins in their native lipid environment yields distinct dynamic features that are lost or altered in detergent or mimetic systems.
• Technological Advancement: By automating the removal of lipids and contaminants, this platform makes HDX-MS accessible for high-throughput analysis of complex biological systems (e.g., bacterial inner membranes, potentially mammalian membranes).
• Biological Discovery: The platform uncovered previously uncharacterized dynamics in the ABC transporter MsbA, specifically regarding the role of the native membrane in regulating NBD flexibility and the transition between transport states.
• Future Impact: This workflow sets a new standard for "in situ" structural biology, enabling researchers to resolve conformational ensembles and protein-lipid interactions with peptide-level resolution, paving the way for drug discovery and mechanistic studies of membrane proteins in their true physiological context.