Integration of flow cytometry and chemometric methods to monitor nitrogen-starvation-induced stress in Chlorella sorokiniana

Nitrogen-starvation is a widely exploited stress condition for inducing the accumulation of valuable molecules in microalgae. However, because two-stage batch cultivation is the most commonly adopted configuration, accurately predicting when cells are in nitrogen-starvation remains challenging and strongly affects process performance. This study aims to develop an innovative, rapid and reliable method to determine when cells are in nitrogen-starvation condition, and the associated changes in microalgal biochemical profiles. The method is based on flow cytometric analysis requiring no sample pretreatment other than dilution, followed by chemometric data processing using Partial Least Squares Discriminant Analysis (PLS-DA). The model was developed using 50 samples collected under nitrogen-replete (N+) and 52 under nitrogen-starvation (N-). It efficiently discriminated cells exposed to N- from N+, achieving excellent classification performance (AUC = 0.98) with cross-validated sensitivity, specificity and precision between 0.86 -0.90. PLS-DA was also applied to classify changes in carbohydrate, chlorophyll and carotenoid content, resulting from nitrogen-starvation. The model achieved AUC = 0.95 for carbohydrates and chlorophylls, and 0.93 for carotenoids. Cross-validated sensitivity, specificity and precision were 0.81 -0.95 for carbohydrates and chlorophylls, and 0.69 -0.87 for carotenoids. Classifications based on metabolite content showed greater overlap than those based on N+ versus N-. The proposed method offers a promising tool for monitoring nitrogen-starvation in industrial-scale bioreactors, with potential application in predicting carbohydrate accumulation or chlorophyll reduction.