Molecular evolution and functional characterization of Drosophila insulin-like peptides

Abstract

Multicellular animals match costly activities, such as growth and reproduction, to the environment through nutrient-sensing pathways. The insulin/IGF signaling (IIS) pathway plays key roles in growth, metabolism, stress resistance, reproduction, and longevity in diverse organisms including mammals. Invertebrate genomes often contain multiple genes encoding insulin-like ligands, including seven Drosophila insulin-like peptides (DILPs). We investigated the evolution, diversification, redundancy, and functions of the DILPs, combining evolutionary analysis, based on the completed genome sequences of 12 Drosophila species, and functional analysis, based on newly-generated knock-out mutations for all 7 dilp genes in D. melanogaster. Diversification of the 7 DILPs preceded diversification of Drosophila species, with stable gene diversification and family membership, suggesting stabilising selection for gene function. Gene knock-outs demonstrated both synergy and compensation of expression between different DILPs, notably with DILP3 required for normal expression of DILPs 2 and 5 in brain neurosecretory cells and expression of DILP6 in the fat body compensating for loss of brain DILPs. Loss of DILP2 increased lifespan and loss of DILP6 reduced growth, while loss of DILP7 did not affect fertility, contrary to its proposed role as a Drosophila relaxin. Importantly, loss of DILPs produced in the brain greatly extended lifespan but only in the presence of the endosymbiontic bacterium Wolbachia, demonstrating a specific interaction between IIS and Wolbachia in lifespan regulation. Furthermore, loss of brain DILPs blocked the responses of lifespan and fecundity to dietary restriction (DR) and the DR response of these mutants suggests that IIS extends lifespan through mechanisms that both overlap with those of DR and through additional mechanisms that are independent of those at work in DR. Evolutionary conservation has thus been accompanied by synergy, redundancy, and functional differentiation between DILPs, and these features may themselves be of evolutionary advantage.

Figure 1. Phylogeny of Drosophila insulin-like peptides.

The evolutionary history of the seven DILP families from 12 fully sequenced Drosophila species inferred using the Neighbour-Joining method. Bootstrap replicate support percentages are shown next to the branches. The tree is drawn to scale, with branch lengths proportional to evolutionary distance computed using the JTT matrix-based method. 7 DILPs have remained present and clearly differentiated from each other during the more than 40 million years of evolution of Drosophilidae flies. D. grimshawi contains two DILP2 peptides (shaded in red).

Figure 2. Gene locus organization and generation of dilp mutants.

(A–C) Null mutations for dilp1, 2, 3, 4, 5, 7, 2–3, and 1–4 were generated by ends-out homologous recombination and by imprecise P-element excision for dilp6 (D). (A) The dilp1–4 genes cluster at cytological position 67C8 is separated from dilp5 by two intervening genes of unknown function (Flybase: CG32052 and CG33205). (B) dilp5 is located within an intron of the CG33205 gene. (C) dilp7 is located on the X-chromosome at 3E2 immediately downstream of the essential Poly(ADP-ribose) glycohydrolase (Parg) gene. Donor constructs (dilp ko) for ends-out homologous recombination are indicated by grey bars. Gap between grey bars indicates the genomic region replaced by a white^hs marker gene. Coding parts of exons are marked in black, non-coding parts by white boxes. (D) Transposon integration line KG004972 was used to generate dilp6 deletion mutants dilp6⁴¹ and dilp6⁶⁸. Note: both dilp6 deletion alleles contain remaining P-element sequence, hatched line: region of breakpoint in dilp6⁴¹. (E) PCR on genomic DNA of dilp mutants with dilp specific primer combinations shows homologous recombination mediated replacement of dilp genes by the white^hs marker gene and deletion of dilp6 in dilp6⁶⁸ mutants (M: mutant, C: control). (F) RT–PCR analysis confirms that dilp mutants are transcript null alleles. dilp6⁴¹ mutants express ectopic dilp6 transcripts that lack the first exon but contain the full ORF. dilp6⁶⁸ mutants are dilp6 transcript null alleles.

Figure 3. Compensatory regulation of gene expression in dilp mutants.

(A) Q-RT-PCR analysis of 4E-BP expression on fly bodies and fly heads of dilp single mutants. Expression of the IIS downstream target gene 4E-BP is not changed in dilp single mutants. (B) Q-RT-PCR analysis demonstrates up-regulation of 4E-BP and dilp6 and down-regulation of ImpL2 in dilp2–3,5 mutants independent of Wolbachia infection status (w^DahT: Wolbachia-, w^Dah: Wolbachia+). (C) Compensatory regulation among MNC-expressed DILPs. Expression levels of DILP2, 3, and 5 were measured by QRT-PCR on RNA extracted from heads of dilp mutants. (D) Western blot analysis of DILP2 expression in total head protein confirms lack of DILP2 expression in dilp2 and dilp2–3 mutants and downregulation of DILP2 levels in dilp3 mutants. M: homozygous mutant, C: w¹¹¹8 control. An anti-Tubulin antibody was used as loading control. (E) Diagram summarizing the feedback system involved in the control of DILP expression levels. DILP3 is part of a feedback system that acts in an autocrine/paracrine manner to regulate DILP expression in the MNC. DILP expression between MNC in the central nervous system and the peripheral fat body tissue is regulated by a negative feedback system that involves DILP6. See text for further details. Arrows denote activation, blunted lines denote repression. All experiments in (A–D) were done using 8–10 day old females reared on 1x SY-A food. Expression level of mutants were normalised to the corresponding wild type control, which by default was set to 1. * p<0.05, ** p<0.01.

Figure 4. Development time and body weight of dilp mutants.

(A) Egg-to-adult development time of dilp mutants. Only the hatching period of the adult flies is shown. Rectangle: males (m), circle: females (f). (Note: number of heterozygous flies was halved to adjust to the number of homozygous and w control flies.) (B) Body weight of dilp mutant (red) males (m) and females (f) compared to controls (black). (n = 20 flies). (C) Body weight of female flies that lack multiple dilp genes. (n = 40 f; dilp1–4,5 n = 20 f; dilp7,1–4,5 n = 18 f). Wolbachia+ flies (w^Dah) weigh more than Wolbachia- flies (w^DahT). Body weight in (B,C): w^DahT background, except for dilp1–4 mutants: w¹¹¹⁸. ** p<0.01, t-test.

Figure 5. dilp2 mutant flies are long-lived.

Survival curves of dilp mutant flies on standard food. Lack-of DILP2 extends median lifespan in two independent genetic backgrounds (w¹¹¹⁸ and w^DahT) and in both sexes. Combined knock out of DILP2 and 3 results in increased lifespan. In contrast to MNC-ablated flies, homozygous w¹¹¹⁸; dilp2–3,5 mutants do not show increased median lifespan. However, median lifespan is slightly increased in dilp2–3,5 heterozygous flies. ** p<0.001, log-rank test.

Figure 6. Wolbachia-dependent lifespan extension of dilp2–3,5 mutants is correlated with xenobiotic resistance.

Wolbachia affects longevity (A) and xenobiotic resistance (E) of dilp2–3,5 mutants, but not fecundity (B), starvation (C) or oxidative stress resistance (D). (A) Median and maximum lifespan of w^Dah;dilp2–3,5 females (83/92,5 days) compared to w^Dah (64,5/76 days) control females is increased by 29% and 22%, respectively. ** p<0.001, log rank test. In contrast, w^DahT;dilp2–3,5 (62,5/81 days) only show a small increase in maximum, but not in median lifespan compared to w^DahT (64,5/74 days) females. (B) Index of lifetime fecundity of dilp2–3,5 females is reduced to 30% but not affected by Wolbachia infection. Shown is the cumulative number of eggs laid by an average female. ** p<0.01, Wilcoxon rank sum test. (C–E) Survival of dilp2–3,5 females with and without Wolbachia on (C) 1,5% agarose, (D) 5% hydrogen peroxide and (E) DDT (275 mg/l). ** p<0.01, log rank test. (F) PCR analysis confirms the presence of Wolbachia in w^Dah flies and the absence in Tetracycline-treated w^DahT flies. (G) Q-RT-PCR analysis shows that 4E-BP expression is not affected by Wolbachia-infection status. 4E-BP transcript level in Wolbachia- flies was normalised to the corresponding Wolbachia+ sample, which by default was set to 1.

Figure 7. Dietary restriction in Drosophila is mediated by DILPs.

(A–C) dilp 2, 3 or 5 single mutants exhibit a normal response to DR compared to wild type controls. (D) Compensatory regulation among MNC-expressed DILPs on a high yeast diet. Expression levels of DILP2, 3 and 5 were measured by Q-RT-PCR on RNA extracted from heads of 10 day old dilp mutant females kept on 2.0x food. * p<0.05. (E–H) In two independent trials dilp2–3,5 mutants failed to show a normal response to DR. (E,F) w^DahT; dilp2–3,5 mutants (Wolbachia-). (G,H) w^Dah; dilp2–3,5 mutants (Wolbachia+) Bars: index of lifetime fecundity±standard error of mean; connected points: median lifespan in days. dilp2–3,5 mutants in red, w^DahT controls in black/blue.