Baker’s yeast turns benzaldehyde into a chiral building block in student lab

Baker’s yeast has long been prized for turning sugar into alcohol, but a laboratory study highlights a second, less celebrated talent: forging carbon–carbon bonds. Researchers have detailed a whole-cell biocatalytic procedure in which Saccharomyces cerevisiae converts benzaldehyde into (R)-1-hydroxy-1-phenylpropan-2-one, commonly known as (R)-PAC, through an enzymatic acyloin condensation. The work, implemented as an upper-division laboratory experiment, combines fermentation, crude-product recovery, a classic colorimetric assay, and quantitative proton nuclear magnetic resonance analysis to measure how much of the substrate actually follows the desired carboligase pathway rather than competing reductive routes.

The chemistry at the heart of the transformation is carried out by pyruvate decarboxylase, an essential enzyme in the central carbon metabolism of microorganisms. Using thiamine pyrophosphate and magnesium ions as cofactors, the enzyme decarboxylates pyruvate to generate an unusual acetaldehyde–thiamine pyrophosphate species known as activated acetaldehyde. When a measured portion of benzaldehyde is added to fermenting yeast, the enzyme attaches the two molecules, creating a new carbon–carbon bond and yielding the ketol (R)-PAC. In the biocatalysis literature this green carbon–carbon bond-forming reaction has classically been referred to as carboligation, and it proceeds in water under mild conditions without the reactive reagents demanded by conventional synthetic methods.

The industrial pedigree of this reaction is remarkable. Because the ketol product can be converted to ephedrine and pseudoephedrine with the aid of methylamine and appropriate reduction steps, the semi-synthetic manufacturing procedure became the first industrially significant biocatalytic process of its kind. In that approach, one step of the synthesis is executed by microorganisms while the second relies on established chemical methodology. High titers of (R)-PAC are achieved industrially using engineered strains of Saccharomyces cerevisiae or Candida flareri, after which a second chiral center is introduced at the prochiral carbonyl carbon through reductive amination, and ephedrine undergoes N-acetylation followed by acid treatment to give the desired diastereomer, pseudoephedrine.

In practice, adding benzaldehyde to a yeast fermentation broth produces several by-products alongside (R)-PAC. These arise primarily from oxidoreductase enzymes, particularly alcohol dehydrogenase, and can be organized into three levels according to their biosynthetic distance from the original substrate. Level one products are only one reaction away: (R)-PAC from the carboligase activity of pyruvate decarboxylase, benzyl alcohol from the reduction of benzaldehyde by alcohol dehydrogenase, and benzoic acid if the substrate is oxidized. Level two products are two enzymatic steps away, including the reduction of (R)-PAC to the vicinal diol (1R,2S)-PAC-diol in accordance with the Prelog rule, and oxidation of the ketol to the 1,2-dicarbonyl compound PAC-dione. A third-level isomer, (S)-2-hydroxy-1-phenylpropane-1-one, can arise from further transformations in certain strains, although the authors note it was not encountered under their experimental conditions.

The experimental workflow was designed for senior undergraduate students in chemical engineering or biotechnology who have a working knowledge of organic chemistry and NMR spectroscopy. It unfolds across two workshop sessions totaling five hours. In the first, whole-cell biotransformation is initiated by adding measured amounts of benzaldehyde to yeast in YPD broth under controlled pH, temperature, agitation, and optical density, with vigorous shaking on an orbital shaker. After one hour the reaction is halted by centrifugation, and a sample of the medium is subjected to the creatine–α-naphthol test. The second session is devoted to solvent extraction with dichloromethane and vacuum evaporation, after which the crude yellow oily extract is analyzed by proton NMR at 500 MHz in CDCl3 doped with D2O to determine yield and substrate conversion. Pairs of students proved sufficient to run the biotransformation sessions.

The colorimetric assay carries a long scientific legacy. Before modern chromatography and NMR spectroscopy, an enhanced version of the Voges–Proskauer test, originally developed to detect acetoin, the aliphatic homologue of (R)-PAC, was adapted by Gröger and Erge into a colorimetric assay for the ketol. Mechanistic work has shown that color development proceeds in two steps: under alkaline conditions, 2-hydroxy ketones are oxidized by atmospheric oxygen to the corresponding dicarbonyl compound under catalytic influence of α-naphthol, with oxygen diffusion through the medium acting as the rate-limiting step; guanidino compounds such as creatine or arginine then react with this intermediate to form the characteristic chromophore. In the present study, development of a red–purple ring at the interface between the reaction mixture and air indicated a positive result, providing qualitative evidence for (R)-PAC in the fermentation broth, with intense coloration discernible after four hours.

The quantitative NMR analysis told a more nuanced story. Resonances at 4.68 and 4.60 parts per million confirmed the formation of benzyl alcohol from benzaldehyde and of (1R,2S)-PAC-diol from (R)-PAC through reductive biotransformation. A characteristic doublet near 8.09 parts per million, corresponding to the ortho aromatic protons of benzoic acid, suggested that prolonged storage of benzaldehyde had allowed gradual chemical oxidation by atmospheric oxygen. A sharp resonance near 2.5 parts per million was assigned to the methyl ketone group of PAC-dione, shifted slightly downfield relative to (R)-PAC because of the deshielding effect of an additional carbonyl group. Students were asked to build a quantification table from integrated signal values, excluding the regions from 0.0 to 1.5 and 7.0 to 7.75 parts per million, where extensive overlap made unambiguous assignment impractical.

The resulting numbers quantify the competition between pathways. Benzaldehyde conversion reached 0.26, meaning roughly a quarter of the substrate was consumed under the conditions used. Selectivity toward (R)-PAC was 0.58, while benzyl alcohol claimed a selectivity of 0.36, PAC-dione 0.06, and the diol a negligible amount below the limit of reliable quantification. The reductive-to-carboligase ratio, defined as the moles of reductive products divided by the moles of carboligase products, came to 0.56. Benzoic acid was excluded from the quantitative calculations because it forms predominantly through chemical oxidation of the substrate rather than biochemical transformation. These results are consistent with the literature, which establishes that process selectivity for (R)-PAC remains well below 100 percent with whole-cell yeast biocatalysts, and that concurrent reduction of benzaldehyde to benzyl alcohol is the primary cause of the diminished selectivity.

The authors point to plausible process improvements, while keeping them as expectations rather than demonstrated results. Substrate pulse-feeding would likely increase benzaldehyde utilization by mitigating the toxic effects of high benzaldehyde concentrations, thereby promoting greater participation of the substrate in the desired biotransformation. The choice of microbial strain may also shift the product distribution toward (R)-PAC or the diol, although strain selection fell outside the scope of the study. Because no chiral chromatographic or polarimetric analysis was performed, the product is referred to simply as PAC in the analytical discussion, with the expected R configuration assigned from established literature rather than measured directly. The space–time yield of the whole-cell process remains relatively low, and research efforts have increasingly shifted toward purified enzymes with the required cofactors in vitro to enhance catalytic efficiency, productivity, and process control.

Beyond the chemistry, the exercise changed how students think about catalysis. Those trained in chemical engineering typically associate the word catalyst with heterogeneous materials such as platinum on activated carbon or alumina; after the sessions, many students who had never encountered biotransformation recognized whole-cell biocatalysis as an alternative paradigm in catalytic synthesis, one that achieves carbon–carbon bond formation without the stringent conditions or highly reactive reagents of methods like the Grignard reaction, Claisen condensation, or Corey–House synthesis. The procedure also adheres to green chemistry principles: it avoids hazardous reagents, runs in aqueous medium under mild, energy-efficient conditions, and uses a renewable, generally recognized as safe biocatalyst handled at biosafety level 1, though the dichloromethane extraction could be replaced by greener solvents such as 2-methyltetrahydrofuran or ethyl acetate. Safety precautions remain essential, since dichloromethane is carcinogenic and an inhalation hazard whose overexposure can cause headaches and symptoms resembling carbon monoxide poisoning, and the alkaline α-naphthol reagent, prepared with 10 percent sodium hydroxide, is toxic and corrosive, so instructors are advised to prepare the reagents in advance and all solvent work belongs in a fume hood, with biomass autoclaved at 121 degrees Celsius for 15 minutes before disposal.

Subject of Research: Whole-cell biocatalytic acyloin condensation of benzaldehyde by Saccharomyces cerevisiae with 1H NMR analysis of product selectivity

Article Title: Whole-cell biocatalytic acyloin condensation of benzaldehyde by baker's yeast: C C bond formation and 1 H NMR analysis of product selectivity

Article References: Majd, A., Hakim, M., & Hamedi, J. (2026). Whole-cell biocatalytic acyloin condensation of benzaldehyde by baker's yeast: CC bond formation and 1H NMR analysis of product selectivity. Results in Chemistry, 31, Article 103957. https://doi.org/10.1016/j.rechem.2026.103957

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103957

Keywords: biocatalysis, baker's yeast, Saccharomyces cerevisiae, pyruvate decarboxylase, acyloin condensation, benzaldehyde, R-PAC, 1H NMR, green chemistry, carboligation, undergraduate laboratory, ephedrine semi-synthesis