While it is known that certain diets are associated with common changes in animals — diets rich in sugar can promote hyperglycemia, for example — the molecular and genetic basis for the variation in dietary effects across individuals is largely unknown. Most research points to a handful of so-called nutrient sensing pathways – in particular, the insulin/insulin-like (IIS) and the target of rapamycin (TOR) signaling pathways – as solely responsible for the dietary response. These pathways have been identified through single-gene knockout studies, an approach that involves “turning off,” or perturbing, a gene and observing its effects. Scientists, however, have been unable to easily connect genes that have been implicated through these mutation-type studies to natural variation in natural populations.

The disconnect, according to Ng’oma, may stem partly from the use of inbred lines. Inbred lines are organisms that have been bred to be physiologically and genetically identical.  These single genotypes are often used precisely because their responses show little variability among individuals.

“This approach is useful, and it has given us clues about what processes are important to coordinate the effects of nutrition in general. But, a very small fraction of the genetic variability that’s available in nature is tested in these single genotype experiments. Moreover, previous studies have shown the different inbred lines can vary widely in how they respond to diet changes, meaning that the findings from a single genotype could represent a highly specific response and thus not be broadly applicable,” he explained.

In this new study, Ng’oma and his colleagues took a different approach. They used a genetically diverse population of fruit flies reconstituted from a large multiparent population of inbred lines, and looked at the response of every expressed gene to diet. They looked at gene expression because biological traits are determined as much by when, where, and how genes are expressed as by the identity of the gene per se. This approach, they hypothesized, would get them closer to what would be seen in nature than if they just focused on a single genotype and one or two genes.

“We expected that characterizing the overall transcriptional response to diet in a multiparent population would better capture the average response of the population and have the potential to be more broadly applicable than those characterized by only a few genotypes,” Ng’oma said.

To test their idea, the researchers reared the population of fruit flies on three diets — restricted, control, or high sugar — that varied in carbohydrates and protein. After ten days, the flies were dissected, RNA was isolated from the heads, ovaries, and bodies, and changes in gene expression were measured. The gene expression data were analyzed to see which genes responded to the different diets and also the effects of interactions between diet and the different tissues. An example of an interaction would be if a given gene is upregulated on restricted diets in ovaries but downregulated in heads.

“A major finding was that we found a huge number of genes were regulated in response to diet,” said Ng’oma. “If we just look at the effect of diet, we found 2,471 genes were regulated, or about 20 percent of the genes in the study. Likewise, for the interaction, close to a thousand genes, or about 8%, were regulated.”