FOXO regulates organ-specific phenotypic plasticity in Drosophila

Abstract

Phenotypic plasticity, the ability for a single genotype to generate different phenotypes in response to environmental conditions, is biologically ubiquitous, and yet almost nothing is known of the developmental mechanisms that regulate the extent of a plastic response. In particular, it is unclear why some traits or individuals are highly sensitive to an environmental variable while other traits or individuals are less so. Here we elucidate the developmental mechanisms that regulate the expression of a particularly important form of phenotypic plasticity: the effect of developmental nutrition on organ size. In all animals, developmental nutrition is signaled to growing organs via the insulin-signaling pathway. Drosophila organs differ in their size response to developmental nutrition and this reflects differences in organ-specific insulin-sensitivity. We show that this variation in insulin-sensitivity is regulated at the level of the forkhead transcription factor FOXO, a negative growth regulator that is activated when nutrition and insulin signaling are low. Individual organs appear to attenuate growth suppression in response to low nutrition through an organ-specific reduction in FOXO expression, thereby reducing their nutritional plasticity. We show that FOXO expression is necessary to maintain organ-specific differences in nutritional-plasticity and insulin-sensitivity, while organ-autonomous changes in FOXO expression are sufficient to autonomously alter an organ's nutritional-plasticity and insulin-sensitivity. These data identify a gene (FOXO) that modulates a plastic response through variation in its expression. FOXO is recognized as a key player in the response of size, immunity, and longevity to changes in developmental nutrition, stress, and oxygen levels. FOXO may therefore act as a more general regulator of plasticity. These data indicate that the extent of phenotypic plasticity may be modified by changes in the expression of genes involved in signaling environmental information to developmental processes.

Figure 1. The genitalia of male Drosophila are nutrition- and insulin-insensitive.

(A) The scaling relationship for male genital (closed circles) and wing size (open circles) against body size, where size variation is due to variation in developmental nutrition. Each line is the standardized major axis and the slope of this line – the allometric coefficient – captures the nutritional plasticity of wing and genital size relative to the nutritional plasticity of body size. (B) The allometric coefficient is significantly lower for the male genitals than for the wings or the maxillary palps, indicating a reduced nutritional plasticity. *** common slope test, p<0.001. (C) Flies that are homozygous for mutations of Inr or its substrate chico show a significantly smaller reduction in genital size than wing or maxillary palp size, relative to wild-type controls, genocopying starvation (2% diet) (*** Tukey HSD, P<0.001 for all). (D) 48 h wild-type, Inr^E19 and Akt¹ clones in wing and genital discs. Within genotypes, discs are from the same fly. Mutation of Inr or Akt has a greater effect on clone size in the wing disc than in the genital disc. Clones were induced by the MARCM system and express GFP. (E) Inr^E19 and Akt¹ mutant clones proliferate at a slower rate in the eye-antennal and wing imaginal disc than in the genital imaginal disc (*** Tukey HSD, P<0.001 for all). (F) Cell size within Inr^E19 and Akt¹ mutant clones is reduced by more-or-less the same degree in all discs (* Tukey HSD, P<0.05, non-significant comparisons not shown) Error bars are 1 standard error.

Figure 2. The mechanisms that reduce the insulin sensitivity of the genitalia act at FOXO in the IIS pathway.

(A) The insulin-signaling pathway. (B) The effect of different IIS pathway mutations/perturbations on wing and genital size. White circle is well-fed wild-type size, and dotted line indicates where perturbation reduces genital and wing size equally. Perturbations of IIS upstream of activated FOXO, including expression of wild-type FOXO (FOXO.wt) (white squares) causes less of a size reduction of the genitalia than the wing of well fed flies and genocopy dietary restriction. In contrast, expression of constitutively active FOXO (FOXO.TM) (black squares) causes an equal reduction in both organs. Multiple markers of the same color refer to different perturbations of the same gene. (C) The effect of FOXO.wt and FOXO.TM expression on wing and genital arch size. Note that only expression of FOXO.TM causes a substantial reduction in genital size. Magenta shading shows area measured on control organ. Within an organ, all images are at the same scale. (D) In contrast, all these IIS pathway mutations/perturbations, including expression of FOXO.TM (black square) cause a more-or-less equal reduction in the size of the maxillary palps and the wings. Error bars are 1 standard error.

Figure 3. FOXO is necessary to maintain organ-specific nutritional plasticity and insulin sensitivity.

(A) The genital imaginal discs of fed larvae have low levels of activated FOXO compared to the wing discs and show less of an increase in activated FOXO after 24 hours of starvation. (B) The scaling relationships between wing and genital size in wild-type and FOXO-mutant flies, where variation in size is due to variation in developmental nutrition. In control flies, starvation has less of an effect on genital size than wing size (slope = 0.55, 95% C.I. = 0.45–0.68). In contrast, starvation has does not have a significantly different effect on the size of the wing and genital in FOXO-mutant flies (slope = 0.93, 95% C.I. = 0.69–1.27). (B) Mutation of FOXO-attenuates the effect of Inr-mutation on the rate of cell proliferation in clones generated in the eye-antennal and wing imaginal discs, but not in the genital discs. The rate of cell proliferation in Inr-FOXO double mutant clones is not significantly different among discs (mixed model ANOVA, P = 0.771). Columns with the same letter are not significantly different (Tukey HSD, P>0.05). Error bars are 1 standard error.

Figure 4. Organ-specific nutritional plasticity is regulated by differential expression of FOXO.

(A) The genital imaginal discs express an unusually low level of FOXO relative to the wing and eye-antennal imaginal discs (* Tukey HSD, P<0.05). (B) The scaling relationship between genital and body size for flies reared under different nutritional conditions. Driving expression of FOXO.wt in the genitalia (NP6333>FOXO.wt) increases the slope of the scaling relationship, and hence the genitalia's nutritional plasticity, relative to wild-type controls (NP6333>GFP). (C) Up-regulating FOXO expression in the genitalia (NP6333>FOXO.wt) significantly increases their nutritional plasticity while down-regulating FOXO expression in the wing (NP6333>FOXO.RNAi) significantly decreases their nutritional plasticity, compared to wild-type controls (NP6333>GFP) (** Common Slope Test, p<0.01). Because we used the temperature-dependence of GAL4 activity to modulate expression, experimental temperatures are indicated in parentheses (see Materials and Methods). Controls were reared at the experimental temperature.

Figure 5. There is a non-linear relationship between FOXO expression and nutritional plasticity.

(A) A moderate increase in FOXO expression in the wing results in an increase in its nutritional plasticity, while a more substantial increase causes a decrease in its nutritional plasticity. Expression levels are normalized to wild-type expression at 25°C. Plasticity is the allometric coefficient of the wing-pupal scaling relationship, uncorrected for temperature. Line is quadratic regression for raw data, gray shading is 95% CI, N = 49. Circles indicate flies in which we have manipulated FOXO expression using UAS-GAL4, diamonds indicate wild-type flies reared at different temperatures (see Figure S5). (B) The nutritional static allometry of wings with different levels of FOXO expression shows that at very high levels of FOXO expression (>FOXO.wt at 25°C), wing size is small even in well-fed flies with large bodies, while at very low levels of FOXO expression (>FOXO.RNAi at 20°C) wing size is large even in poorly-fed flies with small bodies. Note that in both situations the nutritional plasticity of the wing is reduced. Data is normalized to control for the effects of temperature on scaling (see Materials and Methods). Marker color in (B) refer to data points in (A).