Fat test for algae

Although you can’t tell from looking at them, the tiny photosynthetic organisms known as algae are actually rather fatty. Grow algae under the right conditions and they naturally build up high levels of fat in their cells. This makes them a promising biofuel feedstock, as these fats can be extracted and converted into biodiesel. Adding to this promise is the fact that algae grow quickly and abundantly in water and so don’t have any impact on food supplies, unlike crop-based biofuel feedstocks such as corn and soybeans.

To develop an efficient process for producing biodiesel from algae, you obviously want to find those algae strains that are best at producing and storing fat, which means measuring fat concentrations in thousands of different strains. Using conventional analytical techniques, this is a time-consuming process that requires at least 1g of algae. Although this doesn’t sound like too much, it still means growing quite a lot of algae before the analysis can be conducted, preventing scientists from quickly testing lots of different strains to find the best ones.

Now, however, Lieve Laurens and Edward Wolfram at the US National Renewable Energy Laboratory have shown that NIR spectroscopy can accurately measure the concentration of not only fat, but also protein and carbohydrate, in just 10mg of algae. This amount is small enough to fit into the well of a 96-well plate, raising the prospect of measuring fat concentrations in algae on a high-throughput basis.

Laurens and Wolfram first used NIR spectroscopy to probe larger amounts of biomass, around 100–200mg, taken from three different algal species, for which the concentrations of fat, protein and carbohydrate were already known. As they report in the Journal of Agricultural and Food Chemistry, this revealed that the absorption of specific NIR wavelengths did vary in line with the fat, protein and carbohydrate concentrations in the algae.

They then used some of this NIR data to construct a mathematical model based on the detected wavelengths and then tested the model with the rest of the NIR data from the study, finding that their model could accurately predict the three concentrations. Finally, they showed that their model could still produce accurate concentration predictions when presented with NIR spectra derived from analysing just 10mg of algae biomass.