Optimization of the glucosinolate core pathway for production of simple glucosinolates in Escherichia coli

This study demonstrates the optimized microbial biosynthesis of defined glucosinolates in *Escherichia coli* by engineering P450 expression and sulfate assimilation, achieving a 37-fold titer increase for benzyl glucosinolate and establishing the first production of tyrosine-derived glucosinolates with a record 1250 μM yield of indolyl-3-methyl glucosinolate.

The Gist

Imagine you have a magical factory inside a tiny, single-celled organism called E. coli (a common bacterium found in our gut). Scientists want to turn this bacterium into a mini-chemical plant that can produce glucosinolates.

What are glucosinolates? Think of them as the "superhero vitamins" found in broccoli, kale, and Brussels sprouts. They are the compounds that give these veggies their spicy kick and are known to fight cancer and help with diabetes. Usually, you have to eat a lot of broccoli to get a healthy dose, and even then, the amount is small.

The goal of this paper is to teach E. coli how to make these "superhero vitamins" in a lab, so we can produce them cheaply and in huge quantities, without needing to farm acres of broccoli.

Here is how the scientists did it, explained with some everyday analogies:

1. The Assembly Line Problem

Imagine the production of a glucosinolate is like building a custom car on an assembly line.

• The Raw Material: You start with an amino acid (like a block of clay).
• The Workers (Enzymes): You need a team of specialized workers to shape that clay into the final car.
• The Problem: In nature, plants have a very specific team of workers. But when the scientists tried to put these plant workers into the bacterial factory, the line was slow, the workers were confused, and the final product was rare.

2. Hiring the Best Workers (Combinatorial Screening)

The scientists realized that the plant's "assembly line" wasn't perfectly organized for a bacterial factory. They decided to play "mix-and-match."

• They took workers from the "aromatic" team (good for some smells) and the "aliphatic" team (good for other structures) and tried different combinations.
• The Discovery: They found that the best team wasn't just one type of worker. It was a "Frankenstein" team! For example, to make the best product, they needed a worker from the aromatic team to start the job, but a worker from the aliphatic team to finish it.
• The Result: By finding the perfect team lineup, they made the assembly line run much smoother.

3. Fixing the "Traffic Jam" (The P450 Problem)

Two of the most important workers on the line are called P450 enzymes. In plants, these workers are like heavy-duty construction cranes. But in bacteria, these cranes are too big and heavy; they get stuck in the factory walls (cell membranes) and refuse to work.

• The Fix: The scientists realized these cranes had a "handle" (a membrane anchor) that was too long. They used a pair of metaphorical scissors to trim the handles off the cranes.
• The Result: Once trimmed, the cranes could move freely inside the bacterial factory and started working efficiently.

4. The Missing Fuel (Sulfate and PAPS)

To finish the car, the assembly line needs a specific type of fuel called PAPS (a sulfur donor).

• The Problem: The bacterial factory is stingy with this fuel. It usually keeps it locked away for its own survival, leaving the glucosinolate assembly line starving.
• The First Attempt (The "Blockade"): The scientists tried to break the lock on the fuel tank by disabling the machine that usually uses up the fuel. This worked, but it made the bacteria sick and slow.
• The Better Solution (The "Highway"): Instead of breaking the system, they built a super-highway for the fuel. They added special trucks (transporters) from other bacteria that are really good at grabbing sulfur and bringing it into the factory. They also added more fuel pumps (enzymes) to keep the supply flowing.
• The Result: The assembly line was suddenly flooded with fuel, and production skyrocketed.

5. The Final Result: A Production Boom

By combining all these tricks—hiring the best mixed team, trimming the heavy workers, and building a super-highway for fuel—the scientists achieved something amazing:

• They created the first-ever bacterial factory that can make a specific type of glucosinolate derived from tyrosine (a new discovery).
• They boosted the production of Indolyl-3-methyl glucosinolate (the broccoli super-chemical) by 500 times compared to previous attempts in yeast.
• They reached levels of production that were previously thought impossible in bacteria.

Why Does This Matter?

Think of this as upgrading from a hand-crafted workshop to a massive, automated Tesla factory.

• Before: We had to eat huge bowls of broccoli to get a tiny bit of medicine.
• After: We can grow bacteria in a tank that churns out these health-boosting chemicals in massive amounts. This could lead to cheaper medicines, better supplements, and new ways to treat diseases, all without needing to harvest millions of acres of vegetables.

In short, the scientists took a clumsy, slow bacterial factory, reorganized the staff, fixed the equipment, and poured in extra fuel to turn it into a high-speed production line for life-saving plant chemicals.

Technical Summary

1. Problem Statement

Glucosinolates are sulfur-rich plant secondary metabolites found in brassicaceous plants (e.g., broccoli) with significant health benefits, including cancer prevention and insulin resistance reduction. However, extracting them from plants is inefficient due to complex mixtures and low concentrations. While microbial biosynthesis in Saccharomyces cerevisiae and E. coli has been attempted, titers remain low (e.g., max 8.3 mg/L for benzyl glucosinolate in E. coli).

Key bottlenecks identified include:

• P450 Expression: Cytochrome P450 enzymes (CYP79 and CYP83 families) are notoriously difficult to express and fold correctly in bacterial hosts.
• Cofactor Limitation: The pathway requires 3'-phosphoadenosine-5'-phosphosulfate (PAPS) as a sulfate donor. E. coli does not naturally accumulate PAPS for sulfotransferase reactions, creating a metabolic bottleneck.
• Enzyme Compatibility: Previous engineering strategies rigidly followed the plant's "aromatic" vs. "aliphatic" pathway classification, potentially missing optimal enzyme combinations from mixed origins.

2. Methodology

The authors employed a systematic metabolic engineering approach in E. coli BL21 (DE3) to optimize the production of three simple glucosinolates: Benzyl glucosinolate (BGLS), p-hydroxybenzyl glucosinolate (pOHB), and indolyl-3-methyl glucosinolate (I3M).

Key Engineering Strategies:

• Combinatorial Screening: Instead of adhering strictly to aromatic/aliphatic classifications, the team assembled pathway modules using homologs from both pathways. They tested various combinations of post-aldoxime enzymes (CYP83, GSTF, UGT, SOT) to identify the most efficient route for each specific product.
• P450 Optimization: To improve expression and membrane integration, the N-terminal transmembrane helices of CYP79 and CYP83 enzymes were truncated based on TMHMM predictions.
• Sulfate Assimilation Engineering:
  • Knockout Strategy: They initially attempted to overaccumulate PAPS by knocking out cysH (PAPS reductase) in a $\Delta$cysH strain.
  • Overexpression Strategy: Realizing the knockout impaired cell health, they shifted to overexpressing the cysDNCQ operon (sulfate to PAPS conversion and recycling) and heterologous sulfate permeases (cysZ from Corynebacterium glutamicum and cysP from Bacillus subtilis) in wild-type cells to ensure PAPS supply without disrupting sulfur metabolism.
  • Media Supplementation: Fermentation media was supplemented with sulfate (5 mM) and thiosulfate (0.8 mM) to support both PAPS production and glutathione synthesis (required for the pathway).
• Fermentation Conditions: The study utilized autoinduction media (TB-based) at 18°C for 72 hours to ensure proper P450 folding, rejecting higher temperature pre-induction protocols that proved detrimental.

3. Key Contributions

• Pathway Reclassification: Demonstrated that the strict aromatic/aliphatic enzyme classification is not required for biosynthesis; optimal routes often involve "cross-pathway" enzyme combinations (e.g., using aliphatic SOTs with aromatic CYPs).
• Enzyme Identification: Identified that UGT74C1 (aliphatic) is non-functional in E. coli for this pathway and that GSTF9 outperforms GSTF11 in full pathway contexts, likely due to protein-protein interactions (metabolon formation) rather than just catalytic activity.
• PAPS Supply Solution: Proved that overexpressing sulfate transporters and assimilation genes in wild-type cells is superior to knocking out cysH, which caused metabolic stress and reduced flux.
• First Microbial Synthesis: Established the first microbial production of tyrosine-derived p-hydroxybenzyl glucosinolate (pOHB).

4. Results

The study achieved significant titer improvements through iterative optimization:

• Autoinduction: Maintaining cultures at 18°C for the entire 72-hour fermentation yielded the highest titers; any time spent at 28°C reduced productivity by up to 90%.
• Enzyme Screening:
  • For BGLS, the optimal module (BG6) combined CYP79A2, CYP83A1, GSTF9, UGT74 (aliphatic), and SOT18 (aliphatic).
  • For I3M and pOHB, similar mixed-pathway combinations were found to be optimal.
• Sulfate Engineering:
  • The $\Delta$cysH strain showed poor flux despite high conversion rates of intermediates, confirming that disrupting sulfate assimilation harms the cell.
  • Co-expression of cysDNCQ and the permease cysP in wild-type cells restored PAPS availability, eliminating the accumulation of desulfo-glucosinolates (dsBGLS) and driving complete conversion to final products.
• Final Titers:
  • I3M: 1250 ± 91 µM (approx. 561 mg/L). This represents a 500-fold increase over previous yeast production and a massive improvement over prior E. coli attempts.
  • BGLS: 744 ± 36 µM (approx. 335 mg/L), a ~63% improvement over the PAPS-engineered strain alone.
  • pOHB: 628 ± 13 µM.
  • All products were produced with negligible accumulation of desulfo-intermediates, indicating a balanced pathway flux.

5. Significance

This work establishes a robust, high-yield platform for the microbial production of glucosinolates in E. coli. By decoupling enzyme selection from traditional plant pathway classifications and solving the PAPS supply bottleneck through transport/assimilation engineering rather than gene knockout, the authors achieved titers sufficient for industrial relevance.

The successful production of I3M at >1 g/L levels suggests that microbial cell factories can now compete with plant extraction for high-value glucosinolates. Furthermore, the identification of specific enzyme homologs (like GSTF9 and SOT18) and the strategy of N-terminal P450 truncation provide a blueprint for optimizing other complex plant secondary metabolite pathways in bacteria. This paves the way for sustainable production of health-promoting compounds currently limited by agricultural constraints.