The sleeping beauty: how reproductive diapause affects hormone signaling, metabolism, immune response and somatic maintenance in Drosophila melanogaster

Abstract

Some organisms can adapt to seasonal and other environmental challenges by entering a state of dormancy, diapause. Thus, insects exposed to decreased temperature and short photoperiod enter a state of arrested development, lowered metabolism, and increased stress resistance. Drosophila melanogaster females can enter a shallow reproductive diapause in the adult stage, which drastically reduces organismal senescence, but little is known about the physiology and endocrinology associated with this dormancy, and the genes involved in its regulation. We induced diapause in D. melanogaster and monitored effects over 12 weeks on dynamics of ovary development, carbohydrate and lipid metabolism, as well as expression of genes involved in endocrine signaling, metabolism and innate immunity. During diapause food intake diminishes drastically, but circulating and stored carbohydrates and lipids are elevated. Gene transcripts of glucagon- and insulin-like peptides increase, and expression of several target genes of these peptides also change. Four key genes in innate immunity can be induced by infection in diapausing flies, and two of these, drosomycin and cecropin A1, are upregulated by diapause independently of infection. Diapausing flies display very low mortality, extended lifespan and decreased aging of the intestinal epithelium. Many phenotypes induced by diapause are reversed after one week of recovery from diapause conditions. Furthermore, mutant flies lacking specific insulin-like peptides (dilp5 and dilp2-3) display increased diapause incidence. Our study provides a first comprehensive characterization of reproductive diapause in D. melanogaster, and evidence that glucagon- and insulin-like signaling are among the key regulators of the altered physiology during this dormancy.

Figure 1. Ovarian development and yolk accumulation are arrested during diapause.

A Ovarian developmental stages (%) in virgin female flies Drosophila melanogaster Canton S (A), kept for 1–12 weeks at 11°C and short photoperiod 10L:14D (diapause, D1–D12) or after recovery for 1 week after 3 weeks of diapause (R1′), 1 (R1) or 2 weeks (R2) after 6 weeks of diapause. Recovery was carried out at 25°C and normal photoperiod 12L:12D, light/dark. As comparison we used 3–6 h old (newly eclosed) virgin flies (C0) or virgin flies kept under normal conditions for 1 week (C1). The developmental stage of ovaries was assessed as outlined in Material and methods. Note that one week control flies (C1) have fully developed ovaries. Data are presented as means ± S.E.M, n = 4 independent replicates with 8–12 flies in every replicate. B Yolk accumulation incidence (%) in ovaries of virgin female flies (Canton S) kept at the same conditions as in 1A. 1-week old flies, kept under normal conditions (C1) were used as reference. Yolk accumulation incidence (%) was defined as the presence yolk deposit even in a single oocyte (stage 8) up to formation of one/several chorionated eggs (stages 12–14) in mostly previtellogenic ovaries of diapausing flies. Data are presented as means ± S.E.M, n = 4 independent replicates with 8–12 flies in every replicate. We indicate data that are significantly different from the C1 flies kept for one week at normal conditions (C1) that display 100% yolk incidence (Student t-test with *p<0.05, **p<0.01, ***p<0.001). See Fig. S1 in File S1 for ovary development in dilp mutants and Fig. S6A, B in File S1 for w¹¹¹⁸ flies. C Representative images of ovaries at selected stages of Canton S.

Figure 2. Food intake and body mass are reduced during diapause in D. melanogaster (Canton S).

A Food intake during 6 h was measured in flies of Canton S strain at different time points by feeding flies dyed food (measured as µg/fly/6 h). We used 1 week old (C1) virgin flies in non-diapause conditions as comparisons. Flies were kept for 1–12 weeks in diapause conditions at 11°C and 10L:14D (D1–D12). We also tested flies (blue bars) placed for 1 and 1–2 weeks at 25°C and 12L:12D after either 3 weeks (R1′) or 6 weeks (R1, R2) of diapause. During diapause food intake decreases drastically already after one week, but increases slightly to a peak after three weeks. Food consumption reverts back to non-diapausing levels after recovery from diapause. Data are presented as means ± S.E.M, n = 5–6 independent replicates with 6–10 flies in each replicate. We indicate significance values for experimental flies compared to the 1 week non-diapausing control (C1, grey bar) or as indicated by connectors (***p<0.001, as assessed by ANOVA followed with Tukey test). B Body mass of whole flies was measured in flies corresponding to the sampling points in 2A. A significant decrease of body mass is seen already after one week of diapause (D1) compared to C1 controls, but no significant gain is seen after recovery from diapause (R1′, R1–R2). Data are presented as means ± S.E.M, n = 8–10 independent replicates with 10–15 flies in each replicate. Statistics as in Fig. 2A (*p<0.05, ***p<0.001). See Fig. S4 and S7C, D in File S1 for comparisons of food intake and body mass in dilp mutants and w¹¹¹⁸. C Representative images of flies at different stages: 3 h normal conditions (C0), 1 week normal conditions (C1), 3 weeks diapause (D3), and 1 week of recovery after one week diapause (R1′). D Abdomen size (mm²) of female Canton S flies, kept under conditions as in A2. The largest abdomens are found in the newly eclosed flies (C0), which are full of pupal fat body. In flies, kept under diapause conditions for 6–12 weeks (D6–D12) the abdomen size is decreased. After recovery from diapause the abdomen size increases and resembles that of 1-week old control flies (C1) kept at non-diapausing conditions. Statistics as in Fig. 2A (*p<0.05, **p<0.001, ***p<0.001). E The crop size increases during diapause with a maximum after 2 weeks, and remains larger than the C1 control during 12 weeks. See Fig. S1 in File S1 for images of crops. We indicate significance values for experimental flies compared to the 1 week non-diapausing control (C1, grey bar) or as indicated by connectors (^# p<0.05, ^## p<0.01 or ^#### p<0.001 as assessed by Kruskal–Wallis test followed by pairwise comparisons using Wilcoxon rank sum test).

Figure 3. Diapause conditions affect circulating and stored carbohydrates and proteins as well as stored lipids in Canton S.

Flies were kept under the same conditions as in Fig. 2A. In the panels data are presented as means ± S.E.M, n = 5–6 independent replicates with 10–15 flies in each replicate. Data significantly different from the flies kept for one week at normal conditions (C1) are indicated * p<0.05, ** p<0.01, ***p<0.001, N.S. not significantly different (ANOVA followed with Tukey test) or with ^# p<0.05, ^## p<0.01 or ^#### p<0.01 (Kruskal–Wallis test followed by pairwise comparisons using Wilcoxon rank sum test). A Glucose concentrations in hemolymph (mM) are significantly increased throughout diapause as compared to the flies kept at normal conditions for one week (C1). After recovery from diapause (R1′–R2) glucose levels remain high. B Trehalose levels in hemolymph increase to a peak at three weeks of diapause (D3) and then return to the control level (C1). Recovery conditions have no effect on trehalose. C Whole body glucose also increases significantly during diapause, but falls after recovery compared to controls (C1). D Whole body trehalose stores increase significantly to a peak at three weeks diapause (D3) and drop thereafter to very low levels, similar to one week controls (C1). E Compared to the 1 week controls (C1) glycogen stores first drop (D1) and then increase with a peak at three weeks of diapause (D3). During recovery (R1′, R1 and R2) flies restore glycogen to the control (C1) value. F Triacylglycerid (TAG) contents increase significantly and also peak after three weeks of diapause (D3) and remain elevated compared to C1. G The total protein in the hemolymph is elevated during 1–3 weeks of diapause (D1–D3) compared to 1 week control flies (C1). After recovery from three weeks of diapause protein levels remain reduced, but after 2 weeks of recovery from 6 weeks of diapause proteins are elevated in the hemolymph. H The total body protein is similar to the one week non-diapausing controls (C1) throughout diapause. See also Fig. S5, S6 and S7E–L in File S1 for comparisons of metabolite levels in dilp mutants and w¹¹¹⁸.

Figure 4. Altered gene expression in diapausing flies assessed by qPCR suggest endocrine diapause phenotypes.

Relative steady state expression of genes encoding DILPs, InR and AKH in virgin female flies (Canton S), kept for 1–9 weeks under diapause conditions (11°C and 10L:14D), and for one week recovery (R1′) under non-diapause conditions (25°C and 12L:12D) after 3 weeks of diapause. Virgin flies kept for one week at normal conditions (C1) were used as comparison. The expression values were calculated with the 2^−ΔΔCt method relative to that of the 3–6-h old virgin flies (C0) in each assay. Data are presented as means ± S.E.M, n = 4 independent replicates with 10–15 flies in each replicate. Significance of differences from 1 week non-diapause control (1) is indicated as well between groups indicated by connectors, p<0.05, ** p<0.01, *** p<0.001, N.S. not significantly different (ANOVA followed with Tukey test) or alternatively ^# p<0.05, ^## p<0.01 or ^#### p<0.01 (Kruskal–Wallis test followed by pairwise comparisons using Wilcoxon rank sum test). A Compared to 1-week old non-diapausing flies (C1) the dilp2 expression increased significantly after one week of diapause (D1) and remained significantly higher over diapause (D3–D9). After one week of recovery from diapause (R1′) the dilp2 expression decreased significantly back to the C1 level. Interestingly the recently hatched control flies (C0) display a significantly higher dilp2 expression compared to 1-week non-diapausing flies (C1). B Compared to C1 controls dilp3 increased after one week of diapause and remained higher during 9 weeks (D1–D9). After one week of recovery (R1′) from diapause dilp3 decreased back to the control level (C1). The dilp3 expression is significantly higher in recently eclosed flies (C0) than in one week normal controls (C1). C The dilp5 expression showed a profile similar to that of dilp3. D The dilp6 expression was significantly higher throughout diapause than in C1 controls, but with a slight decrease after 6 weeks (D6). After one week of recovery from diapause dilp6 decreased significantly. Recently hatched flies (C0) display higher dilp6 expression than the one week old ones (C1). E The insulin receptor (InR) transcript displayed no significant differences between the treatments. F The Akh mRNA increased drastically after one week in diapause (D1) and peaked after three weeks (D3). The transcript level returned to the one week control (C1) level after recovery from diapause. The Akh expression is not significantly different in recently eclosed flies (C0) and one week normal controls (C1).

Figure 5. Changes in gene expression determined by qPCR suggest metabolic diapause phenotypes.

Relative expression of genes encoding the cytokine Unpaired-2 (Upd2), the translational inhibitor 4EBP (4ebp), the α-1,4-glucosidase Target of brain insulin (Tobi), Brummer TAG lipase (bmm) and the lipid metabolism regulator PEPCK (phosphoenolpyruvate carboxykinase) in virgin female flies (Canton S), kept for 1–9 under diapause conditions (D1–D9) and for one week recovery (R1′) at normal conditions after 3 weeks of diapause, as well as in virgin flies 3–6 h after eclosion (C0) and one week old flies (C1) kept under non-diapausing conditions. Sampling and statistics are the same as in Fig. 4. A The Upd2 expression in 9-week diapausing flies (D9) and after recovery (R1′) are significantly above the level in non-diapausing controls (C1). The Upd2 expression was lower in one week controls (C1) than newly hatched ones (C0). B 4ebp expression increases during diapause with a maximum at three weeks (D3) compared to C1 flies, but decreases back to the control (C1) level after recovery (R1′). C The tobi mRNA level in 1-week diapausing flies (D1) drastically exceeds that seen in 1-week non-diapausing flies (C1) and remains higher until 6 weeks of diapause (D6). After recovery (R1′) tobi expression is back to the control level (C1). D During diapause there is no significant change in bmm expression compared to the non-diapausing control (C1), with exception of 1-week diapause (D1), where the transcript level is increased. E The level of PEPCK mRNA is increased during diapause (D1, D3 and D9) compared to (C1) flies, and after recovery (R1′) it returns to the level of C1. The bmm expression is much higher in 3–6 h old flies (C0) than in one week old control flies (C1).

Figure 6. Effects of diapause conditions on gut-related structures in Canton S flies.

A The intestinal epithelium appears to age more slowly during diapause. The epithelial cells (EC) were marked with NP1-Gal4 driven expression of GFP (green) in combination with nuclear staining with Hoechst 33342 (blue). The insets display enlarged view of nuclear staining. The flies tested were kept under control and diapause conditions as described earlier for 3 h (3 h N) or 3 days normal conditions (3 dN), 3–9 weeks normal conditions (3–9 w N), 3 days (3 d D) or 3–9 weeks diapause conditions (3–9 w D) and finally for 1–9 weeks recovery conditions after 6 weeks of diapause (1–9 w R). The age-associated changes in growth of EC size and disruption of the EC monolayer in the midgut are delayed by at least 3 weeks in diapausing flies. The yellow asterisks indicate small polyploid cells (sign of intestinal dysplasia). B The length of the midgut was measured during diapause (conditions as in Fig. 2A). The midgut is significantly longer in flies kept 1-week at normal conditions (C1), know to feed properly, than in 3–6 h old flies (C0). In diapausing flies (D1–D12) the midgut is shorter than in C1 controls and then becomes significantly longer after recovery from diapause (R1′, R1–2). Data are presented as means ± S.E.M, n = 6–9 randomly selected flies for each sample point. Significance of differences from the 1-week control (C1) is indicated, as well as between groups indicated by connectors, * p<0.05, ** p<0.01, *** p<0.001, N.S. not significantly different (ANOVA followed with Tukey test). C The width of the midgut did not change much during diapause, except a significant increase after 1-week recovery from diapause (R1′, R1). Data are presented as means ± S.E.M, n = 5–8 randomly selected flies for each sample point. Significance of differences from the 1-week control (C1) is indicated, as well as between groups indicated by connectors, ^# p<0.05 (Kruskal–Wallis test followed by pairwise comparisons using Wilcoxon rank sum test).

Figure 7. Selective effects of diapause conditions on expression of innate immune genes.

The relative expression of four immune genes was determined in six groups of female Canton S flies: virgin 3–6 h old flies (C0), one week old uninfected and non-diapausing flies (C1), uninfected flies kept for 3 weeks either at 11°C and 10L:14D (diapause, D3) or at 25°C and 12L:12D, (normal conditions, N3) and infected flies (cross hatched bars) kept under diapause (D3) or non-diapause conditions (N3) for 3 weeks. Infected flies were injected with a suspension of Micrococcus luteus and Escherichia coli and kept for an additional 3 hours before freezing and RNA extraction. Data are presented as means ± S.E.M, n = 3–4 replicates with 10–15 flies in each. Significance compared to the newly hatched control (C1) which was set at one, or as indicated by connectors: * p<0.05, ** p<0.01, *** p<0.001, N.S. – not significant (ANOVA followed with Tukey test) or ^# p<0.05, ^## p<0.01 (Kruskal–Wallis test followed by pairwise comparisons using Wilcoxon rank sum test). A The Drosomycin expression was significantly upregulated in flies diapausing for 3 weeks (D3) in both infected and uninfected flies compared to non-diapausing 1-week old (C1) and 3-week old flies (N3). Infection further increased transcripts in both N3 and D3 flies. B Cecropin A1 was significantly upregulated during diapause (D3) versus normal conditions (C1), but not N3, in uninfected flies only if comparing C1 to D3. Infection drastically increased transcripts in both diapausing and nondiapausing flies. C The peptidoglycan recognition protein SB1 (PGRP-SB1) transcript level are not affected by diapause, only infection increased it. D Diptericin also increased only due to infection. For the investigated immune genes we did not find a differences in expression levels between 1-week old (C1) and newly eclosed (C0) flies. Similar results were obtained with flies that were reared on food supplemented with antibiotics (see Fig. S2 in File S1).

Figure 8. Reproductive diapause is an adaptive life-history trait in D. melanogaster induced by unfavourable environmental conditions.

Under favourable conditions (25°C, 12L:14D) newly eclosed virgin flies (C0) with previtellogenic ovaries develop into a reproductive mature flies with vitellogenic ovaries (C1). However, if the initial C0 flies are exposed to unfavourable conditions (11°C, 10L:14D), they enter reproductive diapause (D) characterized by arrested ovarian development. Diapause in D. melanogaster is characterized by slowed maturation and aging, as indicated by previtellogenic ovaries and slowed aging of the gut epithelium. During diapause (D1–12) flies ingest very little food compared to control flies (C1). In diapausing flies there is an increase in hemolymph glucose (hGlu) and trehalose (hTre), body glucose (bGlu) and trehalose (bTre), as well as stored glycogen (Gly) and triacylglycerides (TAG), but neither in hemolymph (hPro) nor total body protein (bPro). Diapausing flies also display increased levels of some immune gene transcripts. The metabolic homeostasis is regulated by insulin producing cells (IPCs), endocrine cells of the corpora cardiaca (CC) and fat body cells. During diapause the expression profiles of several genes are altered: including dilps (Drosophila insulin-like peptides), Akh (adipokinetic hormone), tobi (target of brain insulin), bmm (Brummer lipase), PEPCK (phosphoenolpyruvate carboxykinase), Upd2 (unpaired-2) and 4ebp (eukaryotic initiation factor 4 binding protein). Flies recover from diapause within a week of return to normal temperature and photoperiod (recovery), emphasizing that the dormancy serves during periods of unfavourable conditions. The outlines of the fly CNS, intestine and ovaries are redrawn from Toivonen and Partridge (2009) with permission from L. Partridge.