NMR metabolomics of plant and yeast-based hydrolysates for cell culture media applications - A comprehensive assessment

Abstract

Cultivated meat products, generated by growing isolated skeletal muscle and fat tissue, offer the promise of a more sustainable and ethical alternative to traditional meat production. However, with cell culture media used to grow the cells accounting for 55-95% of the overall production cost, achieving true sustainability requires significant media optimization. One means of dealing with these high costs is the use of low-cost complex additives such as hydrolysates to provide a wide range of nutrients, from small molecules (metabolites) to growth factors and peptides. Despite their potential, most hydrolysate products remain poorly characterized and many are thought to suffer from persistent issues of high batch-to-batch variability. Although there have been a number of isolated efforts to determine metabolic profiles for a handful of hydrolysate products, we present the first attempt at a more comprehensive metabolomic characterization of nine different products (four plant and five yeast-based) from two to four different lots each. NMR analysis identified 90 unique metabolites, with only 15 metabolites common to all hydrolysate products (including eight of the nine essential amino acids), and 16 metabolites found in only a single hydrolysate product. The different hydrolysate products were found to have substantial differences in metabolite concentrations (as a fraction of overall mass), ranging from a high of 43% in yeast extract to a low of 14% in soy hydrolysates. The proportion of various metabolites also varied between products, with carbohydrate concentrations particularly high in soy hydrolysates and nucleosides more prominent in two of the yeast products. Overall, yeast extract generally had higher metabolite concentrations than all the other products, whereas both yeast extract and cotton had the largest variety of metabolites. A direct calculation of batch-to-batch variability revealed although there are significant differences between lots, these are largely driven by a relatively small fraction of compounds. This report will hopefully serve as a useful starting point for a more nuanced consideration of hydrolysate products in cell culture media optimization, both in the context of cultivated meat and beyond.

Keywords: Cultivated meat; Hydrolysates; Metabolomics; Nuclear magnetic resonance.

Graphical abstract

Fig. 1

Contribution of each metabolite category to the mass of metabolites as a percentage of the quantifiable hydrolysate sample mass. For visual clarity only the top error bars, representing standard deviation, are shown; note that error bars are symmetrical around each bar.

Fig. 2

Concentrations of the top 80% (by mass) metabolites identified in plant hydrolysates organized by the overall average metabolite concentration. Error bars correspond to the batch standard deviation.

Fig. 3

Concentrations of the top 80% (by mass) metabolites identified in yeast hydrolysates, organized by the overall average metabolite concentration. Error bars correspond to the batch standard deviation.

Fig. 4

The principal component analysis score (A) and loading (B) plots for plant and yeast hydrolysates using the first two principal components.