Integrated Multi-Omics Enabled by Sequential Extraction for Comprehensive Molecular Profiling of Small Extracellular Vesicles

This paper presents a novel sequential extraction-based multi-omics platform that enables deep, reproducible, and simultaneous characterization of proteins, lipids, and metabolites from a single small extracellular vesicle sample, overcoming material limitations and revealing purification-dependent molecular profiles to advance biomarker discovery and therapeutic applications.

The Gist

Imagine your body is a bustling city, and the cells are the buildings. Sometimes, these buildings send out tiny, floating "messenger balloons" called small extracellular vesicles (sEVs). These balloons carry important messages—proteins, fats, and chemicals—that tell us what's happening inside the cells. If a cell is sick (like in cancer), the message inside its balloon changes.

The problem? These balloons are incredibly tiny. Trying to study them is like trying to read a book by looking at a single grain of sand. Usually, scientists have to destroy the balloon to read the message, and they need so many of them that they often run out of "sand" before they can read the whole story.

This paper introduces a brilliant new way to read the entire story from just one single grain of sand.

The "All-in-One" Sandwich Method

Think of the researchers' new method as a master chef preparing a complex meal from a single, tiny ingredient. Instead of throwing away parts of the ingredient, they use a sequential extraction technique.

1. The First Slice (Lipids): They first peel off the outer skin of the balloon to study the fats (lipids).
2. The Middle Layer (Metabolites): Next, they squeeze out the watery juice inside to study the small chemicals (metabolites).
3. The Core (Proteins): Finally, they take the remaining solid core and break it down to study the proteins.

By doing this, they get a complete "multi-omics" profile (a full report on fats, chemicals, and proteins) from the exact same sample. It's like getting the appetizer, main course, and dessert from a single, tiny crumb of cake, ensuring nothing is wasted.

The "Fishing Net" Experiment

To test if their method works, the scientists tried to catch these balloons from human blood using three different "nets" (purification methods):

1. The Heavy Net (Ultracentrifugation/UC): This spins the blood super fast to spin the heavy balloons to the bottom.
   • Result: It catches the cleanest balloons, but it's slow and you don't catch many of them. It's like using a fine sieve; you get pure sand, but very little of it.
2. The Size Filter (SECUF): This pushes the blood through a special column that only lets things of a certain size through.
   • Result: It catches a medium amount of balloons. It's good at keeping out big junk, but sometimes it accidentally lets in some "look-alike" particles (like low-density lipoproteins, which are like delivery trucks that look like balloons).
3. The Sticky Trap (Polymer Precipitation/PPT): This uses a chemical glue to make all the balloons clump together so they fall out of the liquid.
   • Result: It catches a huge amount of balloons (high yield), but it's messy. It grabs everything, including the "look-alikes" and the glue itself, making the sample very dirty.

What They Found

Using their new "All-in-One" method, the researchers discovered that how you catch the balloon changes what you find inside it:

• The Heavy Net (UC) gave the purest message, but because there were so few balloons, the message was a bit quiet.
• The Sticky Trap (PPT) gave a loud, noisy message. It had tons of data, but it was full of "static" (contaminants) that made it hard to hear the real story.
• The Size Filter (SECUF) was a middle ground, but it had a specific type of noise (the "delivery trucks" or LDLs) that confused the data.

The Big Picture

The most important takeaway is that there is no perfect net. Every method has a trade-off between how many balloons you catch (yield) and how clean they are (purity).

By using their new technique, scientists can now take a tiny drop of blood, extract everything possible from it, and see exactly how the "catching method" changes the results. This is a huge step forward for liquid biopsies (diagnosing diseases from blood). It means doctors can one day get a complete, 360-degree view of a patient's disease from a single, tiny blood sample, helping them find better treatments and catch diseases earlier.

In short: They built a super-efficient machine that reads the entire biography of a tiny cell balloon from a single grain of sand, and they proved that the way you catch the sand changes the story you read.

Technical Summary

1. Problem Statement

Small extracellular vesicles (sEVs), including exosomes and ectosomes, are critical for intercellular communication and serve as promising targets for liquid biopsy and biomarker discovery. However, comprehensive molecular profiling of sEVs faces significant challenges:

• Sample Scarcity: sEVs are extremely small (diameter <200 nm); 10 million sEVs have a biological volume comparable to only a few cells, limiting the material available for analysis.
• Fragmented Analysis: Conventional multi-omics strategies typically require splitting a single sample into separate aliquots for proteomics, lipidomics, and metabolomics. This approach compromises cross-omic biological linkage, introduces technical variability, and wastes precious sample material.
• Purification Variability: sEVs isolated from biofluids (like plasma) using different methods (e.g., ultracentrifugation, size-exclusion chromatography, polymer precipitation) yield populations with distinct purity and yield profiles, yet these differences are rarely characterized simultaneously across all molecular layers.

2. Methodology

The authors developed a unified, low-input multi-omics platform capable of extracting lipids, metabolites, and proteins from a single sEV sample via sequential extraction.

• Sample Preparation & Sequential Extraction:

  • Input: 10 million sEVs (isolated from MDA-MB-231 cells or human plasma).
  • Extraction Strategy: A modified Matyash biphasic extraction was employed. This process simultaneously extracts lipids and metabolites while precipitating proteins.
    1. Organic Phase: Contains lipids.
    2. Aqueous Phase: Contains metabolites.
    3. Protein Pellet: The precipitated protein fraction is recovered for proteomic analysis.
  • Proteomics: Proteins were processed using an in-house MS-compatible surfactant (Azo), reduced, alkylated, and digested using S-Trap technology. Peptides were analyzed via nano-flow LC-MS using DIA-PASEF (Data-Independent Acquisition Parallel Accumulation Serial Fragmentation) on a timsTOF Pro instrument to maximize depth and sensitivity.
  • Lipidomics & Metabolomics:
    • Lipids: Analyzed using 2D-LC coupled to a QTOF mass spectrometer with iterative tandem MS to enhance small-molecule fragmentation and identification.
    • Metabolites: Analyzed using HILIC chromatography coupled to QTOF in both positive and negative ion modes.
  • Data Integration: Proteins and small molecules were integrated using Joint Pathway Analysis (MetaboAnalyst) to identify enriched biological pathways.

• Comparative Study:
  The platform was applied to sEVs isolated from human plasma using three distinct purification methods to evaluate their impact on molecular profiles:

  1. Ultracentrifugation (UC): The reference standard for purity.
  2. Size-Exclusion Chromatography with Ultrafiltration (SECUF): Separates based on hydrodynamic size.
  3. Polymer Precipitation (PPT): Uses volume-exclusion effects for high yield.

3. Key Contributions

• Sequential Extraction Platform: Established a robust workflow that maximizes sample utilization by deriving proteomic, lipidomic, and metabolomic data from a single sEV aliquot, eliminating the need for sample splitting.
• Low-Input Sensitivity: Demonstrated deep molecular coverage using only 10 million sEVs, a quantity previously insufficient for comprehensive multi-omics.
• Systematic Purification Comparison: Provided the first integrated multi-omics comparison of UC, SECUF, and PPT methods, revealing how isolation mechanisms dictate the resulting molecular signatures, purity, and yield.
• Iterative MS2 Strategy: Enhanced lipid and metabolite identification through iterative fragmentation, improving coverage in low-input scenarios.

4. Key Results

A. Performance on Cell Culture sEVs (MDA-MB-231)

• Proteomics: Identified an average of 5,623 protein groups and 43,152 peptides. The majority of detected proteins overlapped with established sEV databases (ExoCarta and Vesiclepedia).
• Lipidomics & Metabolomics: Detected ~4,349 lipid features (162 annotated) and ~6,333 metabolite features (35 annotated). The lipid profile was dominated by membrane phospholipids (PC, PE, SM), with minimal triacylglycerol (TG) contamination, indicating high purity.
• Pathway Analysis: Integrated data highlighted "Endocytosis" as the most significantly enriched pathway, consistent with the endosomal origin of exosomes, alongside breast cancer-specific metabolic pathways.

B. Comparative Analysis of Plasma-Derived sEVs

The study revealed distinct trade-offs between yield and purity across the three purification methods:

Feature	Ultracentrifugation (UC)	SECUF	Polymer Precipitation (PPT)
Yield	Lowest (Reference)	Intermediate	Highest (>10x SECUF, ~100x UC)
Proteomic Coverage	Moderate (1,260 proteins)	Highest (1,582 proteins)	Lowest (959 proteins)
Purity (Biomarkers)	High (High CD9/CD63/TSG101)	High (Comparable to UC)	Low (Significantly reduced biomarkers)
Contaminants	Minimal	LDL Contamination (High APOB, TG)	High Plasma/Lipoprotein Contamination
Metabolomics	Compatible	Challenging (Salt interference)	Challenging (Polymer interference)

• Proteomics: SECUF showed the highest protein identification count due to better separation from abundant plasma proteins. However, PPT samples showed the lowest abundance of canonical sEV biomarkers (CD9, CD81, TSG101) and the highest abundance of lipoprotein markers (APOB), indicating significant non-sEV contamination.
• Lipidomics: PPT yielded the highest number of lipid features but included a high proportion of TGs (indicative of LDL contamination). UC showed the highest relative abundance of membrane-type phospholipids.
• Metabolomics: UC provided the cleanest data. SECUF suffered from salt carryover affecting retention times, while PPT suffered from polymer-induced ion suppression.
• Pathway Integration: "Complement and coagulation cascades" were highly enriched in PPT samples (reflecting plasma protein contamination), whereas "Endocytosis" was most enriched in UC and SECUF samples.

5. Significance

• Methodological Advancement: This work establishes a "gold standard" for low-input sEV analysis, proving that deep, reproducible multi-omics characterization is possible from a single, limited sample without splitting.
• Clinical Implications: The findings highlight that the choice of sEV isolation method drastically alters the molecular profile.
  • UC remains the optimal choice for high-purity studies requiring accurate biomarker discovery, despite lower yield.
  • PPT offers high yield but introduces substantial noise (contaminants), making it less suitable for precise multi-omics unless contamination is rigorously accounted for.
  • SECUF offers a balance but requires careful handling of salt buffers to avoid metabolomic artifacts.
• Future Applications: This integrated platform provides a robust foundation for systems-level biology, enabling the discovery of synergistic multi-component biomarkers and facilitating the study of sEV biology in clinical settings where sample volume is often the limiting factor.