Boron Trick Stops Enzyme Overreach in Soybean Oil Refining

Every bottle of soybean oil on a supermarket shelf has passed through a refining step that most consumers never think about: degumming. Crude soybean oil carries between one and three percent phospholipids, a family of fatty molecules that includes phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine and phosphatidic acid. Left in the oil, these compounds shorten shelf life, degrade quality and complicate storage. Removing them is the central challenge of oil refining, and a new study published in Food Chemistry: X offers an unusually elegant solution, one that borrows its logic from quantum chemistry and delivers it in the form of a cheap, food-compatible additive.

The traditional approach to degumming is chemical. Refiners hydrate the oil to strip out most phospholipids, then add phosphoric acid to catch the remainder. The method works, but it is wasteful. Chemical degumming sacrifices a significant fraction of the oil, lowers the refining yield and generates substantial environmental pollution. Enzymatic degumming, which uses phospholipase A1 to slice the fatty acid from the Sn-1 position of phospholipids, is cleaner and more precise. The enzyme converts phosphatidylcholine into lysophospholipids, which hydrate readily and can be spun out of the oil by centrifugation. On paper, it is a near-perfect industrial biocatalysis.

In practice, however, the process sabotages itself. The immediate product of the enzyme’s cut is Sn-2-lysophosphatidylcholine, an unstable molecule whose remaining fatty acid sits at the Sn-2 position of the glycerol backbone. Through a well-known rearrangement called acyl migration, that fatty acid drifts to the Sn-1 position, exactly where phospholipase A1 can strike again. The enzyme then hydrolyzes it a second time, producing glycerophosphocholine and releasing an extra free fatty acid into the oil. The result is a cascade of problems: wasted enzyme activity, elevated acid value in the finished oil, more sodium hydroxide needed downstream for neutralization, greater oil loss, and a water-soluble glycerophosphocholine product that vanishes into wastewater, taking valuable phospholipid resources with it.

The research team, led by Ruichi Guo and Dianyu Yu, set out to map this unwanted reaction at the level of atoms and electrons. Their first step was molecular docking. To keep the simulations tractable, they used a four-carbon truncated analogue of the natural Sn-2-lysophosphatidylcholine, retaining the polar headgroup and glycerol backbone that the enzyme recognizes while trimming away the long hydrophobic tail that would otherwise explode the conformational search space. Docking against the crystal structure of phospholipase A1 from the Protein Data Bank revealed the enzyme’s catalytic triad: aspartate 202, histidine 280 and serine 141, with neighboring residues forming an oxyanion hole that stabilizes charged intermediates during the reaction.

With the active site identified, the team turned to density functional theory, using the B3LYP functional and progressively larger basis sets to trace the full acyl migration pathway. The simulation revealed a four-stage journey through four transition states and three intermediates. First, the enzyme deprotonates the Sn-1 hydroxyl group of Sn-2-lysophosphatidylcholine, converting it into a reactive oxygen anion with a remarkably low energy barrier of just 6.6 kilocalories per mole. Second, that anion attacks the carbonyl carbon at Sn-2, forming a five-membered ring intermediate at a cost of 13.2 kilocalories per mole. Third, a proton shuttles back to the carbonyl oxygen with a negligible barrier of 1.8 kilocalories per mole. Finally, the ring collapses, the Sn-2 carbon-oxygen bond cleaves, and the fatty acid emerges at Sn-1, with a closing barrier of 18.2 kilocalories per mole that closely matches an experimentally reported value of 17.8 kilocalories per mole, validating the computational model.

The critical insight came from the first stage. Because the initial deprotonation step has the lowest barrier and generates the nucleophilic lone pair that drives everything downstream, it is the gatekeeper of the entire pathway. Block it, and the whole migration stalls, much as inhibiting a rate-limiting enzyme throttles an entire metabolic route regardless of what happens further along. The team’s hypothesis was that sequestering the Sn-1 oxyanion would cut free fatty acid generation by at least ten percent. Their chosen tool was sodium tetraborate, a compound whose boron atom, with fewer than eight outer-shell electrons and vacant orbitals, is a powerful electron acceptor capable of forming a coordination bond with the oxyanion lone pair and locking it out of the reaction.

The experimental validation was systematic. Working with artificial crude soybean oil, purified triglycerides spiked with two percent phosphatidylcholine, the researchers optimized enzyme loading, inhibitor concentration, temperature, pH and reaction time. Phosphorus removal proved insensitive to the inhibitor, meaning degumming efficiency was preserved, while free fatty acid content fell steadily as borate was added. At 0.8 percent borate by weight, 60 degrees Celsius, pH 6.0 and 0.02 percent enzyme, four hours of reaction drove the phosphorus content down to 12.10 milligrams per kilogram, comfortably meeting commercial degumming standards, while free fatty acids reached only 1.05 grams per 100 grams of oil.

The comparison with conventional complete hydrolysis was striking. Without the inhibitor, the same process yielded 1.24 grams of free fatty acids per 100 grams, a phosphorus content of 21.40 milligrams per kilogram and an acyl migration rate of 39.8 percent. With borate, the migration rate dropped to 22.8 percent, a reduction of 17.0 percentage points, and free fatty acids fell by 0.19 grams per 100 grams, roughly a fifteen percent decrease. Nuclear magnetic resonance confirmed the mechanism directly. Proton NMR showed that without inhibitor, glycerophosphocholine dominated the product mixture, while with inhibitor, Sn-2-lysophosphatidylcholine became the major species. Carbon-13 NMR tracked the characteristic signals at 63.38 and 68.38 parts per million over time, showing that the Sn-2 product remained stable and the Sn-1 isomer was suppressed when borate was present.

Food safety considerations were addressed as well. Sodium tetraborate is water-soluble and insoluble in oil, and no residue was detected in the finished oil after hydrolysis and centrifugation. The moderate hydrolysis oil also showed a lower acid value than the complete hydrolysis product, 1.92 versus 2.32 milligrams of potassium hydroxide per gram, with peroxide values and color essentially unchanged. Lower acidity means less sodium hydroxide in deacidification, less oil lost to soapstock, and better storage stability for the final product.

What makes this work resonate beyond one refinery stream is the strategy itself. Previous attempts to tame acyl migration, including amphoteric hydroxides, borate buffers, process optimization and enzyme immobilization, each carried drawbacks ranging from metal contamination to tedious preparation and activity loss, and none had been systematically evaluated for moderate hydrolysis that preserves phospholipid value. By combining molecular docking with quantum chemical simulation to identify the lowest-barrier, most reactive functional group and then designing a targeted inhibitor against it, the team demonstrated a rational, mechanism-first route to controlling an industrial biocatalytic process. The approach turns a longstanding nuisance in lipid enzymology into a solvable problem at its electronic source, and it suggests that similar gatekeeper-blocking tactics could improve other enzymatic processes where unwanted secondary reactions erode yield, quality and profit.

Subject of Research: Inhibition of acyl migration in phospholipase A1-catalyzed enzymatic degumming of crude soybean oil

Article Title: Acyl transfer activity inhibition and its application in moderate enzymatic degumming of crude soybean oil

Article References: Acyl transfer activity inhibition and its application in moderate enzymatic degumming of crude soybean oil. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: soybean oil, enzymatic degumming, phospholipase A1, acyl migration, phospholipids, lysophosphatidylcholine, molecular docking, density functional theory, sodium tetraborate, free fatty acids, oil refining, food chemistry