The complex contributions of genetics and nutrition to immunity in Drosophila melanogaster

Abstract

Both malnutrition and undernutrition can lead to compromised immune defense in a diversity of animals, and "nutritional immunology" has been suggested as a means of understanding immunity and determining strategies for fighting infection. The genetic basis for the effects of diet on immunity, however, has been largely unknown. In the present study, we have conducted genome-wide association mapping in Drosophila melanogaster to identify the genetic basis for individual variation in resistance, and for variation in immunological sensitivity to diet (genotype-by-environment interaction, or GxE). D. melanogaster were reared for several generations on either high-glucose or low-glucose diets and then infected with Providencia rettgeri, a natural bacterial pathogen of D. melanogaster. Systemic pathogen load was measured at the peak of infection intensity, and several indicators of nutritional status were taken from uninfected flies reared on each diet. We find that dietary glucose level significantly alters the quality of immune defense, with elevated dietary glucose resulting in higher pathogen loads. The quality of immune defense is genetically variable within the sampled population, and we find genetic variation for immunological sensitivity to dietary glucose (genotype-by-diet interaction). Immune defense was genetically correlated with indicators of metabolic status in flies reared on the high-glucose diet, and we identified multiple genes that explain variation in immune defense, including several that have not been previously implicated in immune response but which are confirmed to alter pathogen load after RNAi knockdown. Our findings emphasize the importance of dietary composition to immune defense and reveal genes outside the conventional "immune system" that can be important in determining susceptibility to infection. Functional variation in these genes is segregating in a natural population, providing the substrate for evolutionary response to pathogen pressure in the context of nutritional environment.

Fig 1. Correlation of natural log bacterial load (CFU) 24-hours post infection for DGRP lines raised on high glucose and low glucose diets.

Dashed line represents 1 to 1 relationship; solid line from regression analysis. There is strong correlation across diets, but several lines appear to perform disproportionately poorly (i.e. carry high bacterial load) on the high glucose diet. A natural log value of 10 corresponds to about 2.2x10⁴ bacteria, 12 corresponds to about 1.6x10⁵ bacteria, 14 corresponds to 1.2x10⁶ bacteria and 16 corresponds to 8.9 x10⁶ bacteria.

Fig 2. Histograms of estimated line means for nutritional indices for DGRP lines reared on high glucose and low glucose diets.

The differences in distributions on the high glucose and low glucose diets were highly significant (p < 10^-4) in all cases, supporting the assertion that diet significantly alters the metabolic state of the fly.

Fig 3. Correlation between nutritional indices and immune defense (Ln CFU per fly).

Diagonal represents correlation for each index between high and low glucose diets; above diagonal is correlation among indices on the high glucose diet; below diagonal is correlation among indices on the low glucose diet. p<0.0001***, p<0.001**, p<0.05*. Several nutritional indices are correlated with each other but only glucose on the high glucose diet is significantly (negatively) correlated with immune defense.

Fig 4. Correlation between SNP log₁₀ p-values from genome wide associations on high glucose diet and low glucose diet.

Solid line represents 1 to 1 value. While most significant SNPs were significant on both diets, we considered SNPs with p<10^-6 on one diet and p>10^-4 on the other diet to be diet specific.

Fig 5. Crinkled has a diet-specific effect on immune defense.

A.) validation experiment with RNAi knockdown (kd) shows natural log CFU 24 hours post infection with P. rettgeri compared to control (see S1 Table). B.) correlation between principal component 4 of nutritional indices on the high glucose diet and bacterial load on the high glucose diet polarized by allele in crinkled showing that this correlation between nutritional status and immune defense is driven by one allele.