# Targeted metabolomics to quantitatively profile changes in amino acids and phenolics at every step of amahewu production from two Zea Mays L. maize types (white and yellow)

## TL;DR

This study uses targeted metabolomics to track amino acid and phenolic changes during the fermentation of white and yellow maize into amahewu, a traditional African beverage.

## Contribution

The study provides a detailed quantitative analysis of metabolite shifts during amahewu production using targeted LC-MS/MS and chemometric methods.

## Key findings

- Proline was the most abundant amino acid in raw maize but decreased significantly during fermentation.
- Apigenin dominated phenolic compounds, while caffeic acid decreased sharply after cooking.
- Correlation analysis revealed feed-forward amino acid associations and feedback-driven phenolic transformations.

## Abstract

Most existing research on amahewu has primarily examined the microbial and physicochemical properties of beverages, leaving the overall metabolite profile insufficiently characterized. Metabolomic analysis enables the monitoring of metabolite shifts throughout fermentation, revealing how microbial processes drive variations in bioactive compounds. Therefore, this study employed targeted metabolomics to investigate dynamic changes in amino acids and phenolic compounds during the production of amahewu, a traditional Southern African fermented maize beverage. White and yellow Zea mays L. varieties were fermented with malted sorghum inoculum, and samples were analyzed using liquid chromatography–tandem mass spectrometry (LC-MS/MS) with multiple reaction monitoring across four production stages. Chemometric analyses (PCA, OPLS-DA) revealed distinct stage- and maize type-dependent clustering with model validation (Q2 > 0.5; CV-ANOVA, p < 0.05). Across both maize types, seven metabolites discriminated raw from cooked maize, with valine (white maize) and alanine (yellow maize) serving as unique markers. Proline was the most abundant amino acid (>1.0 × 108 ng/g in raw maize), but declined by >80% during fermentation, consistent with microbial utilization. Conversely, threonine and valine increased >3-fold in fermented samples, reflecting proteolysis and microbial synthesis. Apigenin represented >70% of quantified phenolics at all stages, whereas caffeic acid decreased sharply post-cooking (FC = 0.07, p = 7.03 × 10−9), and avenanthramides emerged de novo in fermented samples. Correlation analysis showed strong positive feed-forward associations among amino acids (r > 0.8) and feedback-driven phenolic transformations, while pathway enrichment identified isoquinoline alkaloid biosynthesis (impact = 0.41) in white maize and alanine/aspartate/glutamate metabolism (impact = 0.13) in yellow maize as dominant routes. Collectively, this study provides insight into the quantitative changes in metabolic composition at different stages of production, which reflect changes in nutritional, biochemical, and health-promoting properties, as well as microbial communities.