Future environmental and economic sustainability of aquaculture will depend to a large extent on improvements to large-scale hatchery production. Currently, the high variability in both quantity and quality of larval yields limits industry growth. However, new biotechnological advances in metabolomics promise to revolutionize the way we assess and solve bottlenecks in hatchery production and other aquaculture sectors.
At the European Aquaculture Society meetings held last fall in San Sebastián, Spain, Dr. Andrea Alfaro of the Institute for Applied Ecology in Auckland, New Zealand, described an innovative approach to understanding the physiological condition of mussel larvae by using metabolomics, which provides an instantaneous snapshot of larval physiology through metabolite profile analysis.
By using a variety of univariate- and multivariate methods for feature-selection, she and her colleagues, T. Young and S. Villas-Bôas, identified biomarkers that reflect the variation in larval quality during hatchery production. Analysis of the underlying biochemical pathways may explain how even slight modifications in culture conditions could result in significantly different hatchery production outcomes.
Green lipped mussel (Perna canaliculus) larvae were grown to umbo-stage larvae (4-days post-fertilization) and separated by size (larger or smaller than 120 μm) into poor quality (slow growing) and high quality (fast growing) groups. About 80,000 larvae from each of the two size groups were pooled and their metabolites extracted by GC/MS. A variety of feature-selection methods were used to identify potential biomarkers for determining larval quality.
Unsupervised hierarchical clustering and a partial least squares-discriminate analysis (PLS-DA) model revealed clear separation of larval quality classes based on metabolite ratios.
A range of biomarker identification methods and statistical analyses were used to identify four metabolite-metabolite ratios involving levels of succinate, glycine, alanine, pyroglutamate and myristic acid as candidate biomarkers for assessing mussel larval quality. These metabolites have roles in energy metabolism, osmotic regulation, immune function and cell-to-cell communication.
The study demonstrated the potential application of metabolomics to:
• Classify mollusc larvae based on their physiological condition;
• Construct prediction models for larval quality assessment, and;
• Identify biochemical pathways that may be under differential regulation and would reveal important avenues for future investigation.
Analysis of individual metabolites and their ratios may also be integrated with gene- and protein expression data, providing new avenues for selective breeding programs that would consistently yield high quality larvae.
— David Scarratt