Engineering S. cerevisiae extracellular vesicles using synthetic biology

This study establishes a synthetic biology framework for engineering *Saccharomyces cerevisiae* extracellular vesicles as therapeutic delivery vehicles by developing a modular cloning strategy to identify and validate protein scaffolds, such as the yeast ortholog Bro1, that efficiently sort cargo into yeast EVs.

The Gist

Imagine you have a tiny, natural delivery service that exists inside every living thing, from bacteria to humans. These are called Extracellular Vesicles (EVs). Think of them as microscopic "bubble mailers" or "envelopes" that cells use to send messages, tools, and packages to their neighbors.

Scientists love these bubbles because they are perfect for delivering medicine. They are biodegradable, safe, and can sneak into cells that other drugs can't reach. But there's a catch: Nature didn't design these bubbles to carry our specific medicines. They are like generic shipping containers; we need to modify them to carry specific cargo (like a cancer-fighting drug) and ensure they go to the right destination.

This is where the paper by Bouffard and his team comes in. They wanted to turn a common baker's yeast (Saccharomyces cerevisiae) into a super-factory for these custom medicine bubbles.

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

1. The Problem: We Need a Better Factory

Usually, scientists try to make these bubbles using human cells. But human cells are picky, expensive to grow, and hard to engineer.
The Solution: They chose Baker's Yeast.

• Why? You already know yeast! It's used to make bread and beer. It's cheap, grows fast, and is "Generally Recognized As Safe" (GRAS) by the FDA. Plus, yeast is a genetic engineer's dream; it's very easy to rewrite its DNA.

2. The Toolkit: "EVclo" (The Lego Set)

To build these custom bubbles, the team didn't just hack the yeast randomly. They used a method called Synthetic Biology.

• The Analogy: Imagine you are building a house. Instead of carving bricks out of stone one by one, you have a Lego set with standardized pieces (walls, windows, doors) that snap together perfectly.
• The Tool: They created a system called "EVclo." This is their custom Lego set for yeast. It allows them to snap together different genetic "parts" (promoters, genes, tags) quickly and reliably. This follows a cycle called DBTL (Design, Build, Test, Learn), which is the standard way engineers solve problems.

3. The Mission: Finding the "Loading Dock"

The biggest challenge is getting the medicine inside the bubble. You can't just dump it in; the cell has to actively load it.

• The Metaphor: Think of the bubble as a train car. You need a specific conductor (a "scaffold" protein) to tell the cargo, "Get in here!"
• The Experiment: The team tested several "conductors." Some were from humans (like CD63 and ExoSignal), and one was from the yeast itself (Bro1). They attached a glowing green tag (GFP) to these conductors to see if they could successfully load the green tag into the bubbles.

4. The Results: Who Got the Job?

They built the yeast factories, turned on the lights, and waited for the bubbles to be released. Here is what they found:

• The Human Candidates: Some human proteins (like CD63) didn't work well in yeast. It's like trying to use a human train conductor on a medieval horse cart; the systems just didn't speak the same language.
• The Winner (Bro1): The yeast's own conductor, Bro1, was the star of the show. It was incredibly efficient at grabbing the cargo and shoving it into the bubbles.
• The Surprise: A human peptide called ExoSignal also worked surprisingly well! This suggests that even though yeast and humans are very different, they share some ancient, deep-down rules for how to pack these bubbles.

5. The Quality Control

They checked the bubbles to make sure they were real:

• Size: They were the right size (tiny, like viruses).
• Shape: They looked like perfect little spheres under a microscope.
• Cargo: They confirmed the green glowing cargo was actually inside the bubbles, not just stuck to the outside.

Why Does This Matter?

This paper is a proof of concept. It's like building a prototype car to prove that an electric engine can run on a chassis made of recycled materials.

• The Big Picture: They have proven that we can use cheap, safe baker's yeast to mass-produce custom "medicine bubbles."
• The Future: Now that they have the "Lego set" (EVclo) and know which "conductors" work (Bro1), they can swap the green glow for real drugs. Imagine a future where we brew yeast in giant tanks, and the yeast spits out millions of tiny, custom-made bubbles ready to deliver cancer drugs or repair damaged tissue in the human body.

In a nutshell: The team turned a humble baker's yeast into a high-tech delivery drone factory, proving that nature's smallest couriers can be engineered to save lives.

Technical Summary

1. Problem Statement

Extracellular vesicles (EVs), specifically exosomes, are promising natural drug delivery vehicles due to their ability to mediate intercellular communication and their biocompatibility. However, realizing their full therapeutic potential requires overcoming significant barriers:

• Scalability: Producing sufficient quantities of EVs for clinical use is difficult with mammalian cell lines.
• Engineering Limitations: Current methods for loading therapeutic cargoes (proteins, RNAs) into EVs are often inefficient or lack precision.
• Lack of Standardization: There is no unified synthetic biology framework to iteratively design, build, test, and learn (DBTL) to optimize EV engineering.
• Chassis Selection: While human mesenchymal stem cells are commonly used, they are difficult to scale. Saccharomyces cerevisiae (baker's yeast) is a "Generally Regarded As Safe" (GRAS) organism used in biomanufacturing, but its potential as a chassis for engineering designer EVs remains underexplored.

2. Methodology

The authors developed a synthetic biology-based workflow named "EVclo" to engineer EVs in S. cerevisiae. The approach followed a single Design-Build-Test-Learn (DBTL) cycle:

• Design & Modularity:

  • Utilized the Yeast Toolkit (YTK) MoClo assembly standard for modular cloning.
  • Created transcriptional units (TUs) consisting of a constitutive promoter (pTDH3), coding sequences (CDS) for EV scaffolds fused to cargo proteins (EGFP or mRuby2), and a terminator (tADH1).
  • Integrated Gateway cloning cassettes into the system to allow for the rapid swapping of cargo genes in the future.
  • Designed constructs for both plasmid-based expression (high copy, URA3 selection) and genomic integration (using the MyLO CRISPR-Cas9 toolkit for markerless integration at the HIS3 or FF18 loci).

• Scaffold Candidates:

  • Tested a yeast ortholog: Bro1 (an ALIX homolog).
  • Tested human EV scaffolds: CD63 (tetraspanin), PDGFR (transmembrane domain), and ExoSignal (a short peptide motif).
  • Constructed various fusion strategies (N-terminal, C-terminal, lumenal, and membrane-localized) to determine optimal cargo loading.

• Build & Expression:

  • Generated multiple yeast strains expressing these scaffold-cargo fusions.
  • Induced EV release using mild heat stress (42°C for 30 minutes), a known trigger for EV secretion in yeast.

• Isolation & Characterization:

  • Isolation: EVs were isolated via differential centrifugation, ultrafiltration (100 kDa cutoff), and size exclusion chromatography (SEC).
  • Validation:
    • Flow Cytometry: Measured cellular expression levels of GFP-tagged scaffolds.
    • Confocal Microscopy: Assessed colocalization of scaffolds with Nhx1-mRuby2, a specific marker for Multivesicular Bodies (MVBs), the site of exosome biogenesis.
    • Nanoparticle Tracking Analysis (NTA): Measured EV size and concentration.
    • Transmission Electron Microscopy (TEM): Visualized EV morphology.
    • Nano-flow Cytometry: Stained EVs with membrane dye FM4-64 to confirm lipid-bound particles and detect GFP signal within EVs.
    • Western Blot: Confirmed the presence of intact scaffold-GFP fusion proteins in purified EV fractions.

3. Key Contributions

• Development of the EVclo System: Established a modular, synthetic biology framework specifically for engineering yeast EVs, enabling iterative optimization of cargo loading.
• Identification of Effective Scaffolds: Successfully identified Bro1 (yeast) and ExoSignal (human) as effective scaffolds for sorting proteins into yeast EVs.
• Cross-Kingdom Sorting Evidence: Demonstrated that human-derived EV sorting motifs (ExoSignal) can function in yeast, suggesting a conservation of the underlying sorting machinery (likely HSP70-mediated).
• Proof-of-Concept for Designer EVs: Validated that S. cerevisiae can serve as a scalable platform for producing "designer" EVs loaded with specific cargoes.

4. Key Results

• Expression Levels: All candidate scaffolds were expressed in yeast. ExoSignal-GFP showed the highest cellular fluorescence intensity, while Bro1-GFP showed moderate expression. Genomic integration significantly increased the percentage of cells expressing detectable GFP compared to plasmid expression for Bro1 and CD63.
• Subcellular Localization:
  • Soluble scaffolds (Bro1-GFP, ExoSignal-GFP) colocalized with the MVB marker Nhx1-mRuby2, confirming they are present at the site of EV biogenesis.
  • Membrane scaffolds (PDGFR, CD63) showed punctate localization consistent with plasma membrane or lipid raft association but did not show strong MVB colocalization in this context.
• EV Biogenesis: Overexpression of scaffolds did not alter the size (median ~142.5 nm), morphology, or total titer of EVs released, indicating the engineering process is non-disruptive to native EV production.
• Cargo Loading Efficiency:
  • Bro1-GFP (both plasmid and integrated) and ExoSignal-GFP were successfully detected in purified EV fractions via fluorometry, Western blot, and nano-flow cytometry.
  • Bro1 was the most abundant scaffold in EVs.
  • ExoSignal showed the highest enrichment among the human candidates.
  • CD63 and PDGFR were not detected in EVs, suggesting that human tetraspanin sorting mechanisms (involving syntenin/syndecan) are not conserved or functional in S. cerevisiae.
• Quantitative Yield: Nano-flow cytometry estimated that Bro1-GFP was present in approximately 6% of the total lipid-bound EV particles in the sample.

5. Significance

• Platform Advancement: This study establishes S. cerevisiae as a viable, scalable, and genetically tractable chassis for the biomanufacturing of therapeutic EVs, addressing the scalability limitations of mammalian systems.
• Mechanistic Insight: The success of Bro1 and ExoSignal, but not CD63, highlights that while some EV sorting mechanisms (ESCRT/HSP70 pathways) are evolutionarily conserved, others (tetraspanin-mediated) are not. This guides future engineering strategies toward conserved pathways.
• Therapeutic Potential: By proving that human proteins can be sorted into yeast EVs, the authors demonstrate that yeast EVs can be "humanized." This opens the door for using yeast to produce large-scale, low-cost EVs loaded with therapeutic proteins, antibodies, or nucleic acids for treating cancers, neurodegenerative diseases, and other conditions.
• Future Workflow: The EVclo system provides a standardized, modular tool for the synthetic biology community to screen new scaffolds and optimize cargo loading, accelerating the development of next-generation drug delivery vehicles.