Abstract
The insulin/IGF-signaling pathway is central in control of nutrient-dependent growth during development, and in adult physiology and longevity. Eight insulin-like peptides (DILP1-8) have been identified in Drosophila and several of these are known to regulate growth, metabolism, reproduction, stress responses and lifespan. However, the functional role of DILP1 is not fully understood. Previous work showed that dilp1/DILP1 is transiently expressed during the non-feeding pupal stage and the first days of adult life. Here we show that mutation of dilp1 diminishes organismal weight during pupal development, whereas overexpression increases it. Overexpression of dilp1 additionally increases body size of flies, but reduces stores of larval-derived energy, leading to increased feeding the first days after eclosion. No effects of dilp1 manipulations were detected during larval development. An earlier study demonstrated interactions between dilp1 and dilp2 in regulation of adult lifespan. Here we monitored the effects of dilp1, dilp2 and dilp1/dilp2 mutations on growth and found that only the single mutants displayed lower body mass. In recently eclosed flies, survival during starvation is strongly diminished in dilp1 mutants, but not in dilp2 and double mutants, whereas in older flies double mutants display reduced starvation resistance. Egg to pupal viability is decreased both after overexpression of dilp1, and in the double mutants. In conclusion, dilp1 promotes growth of adult tissues during the non-feeding pupal stage, likely due to reallocation of stored energy. This results in larger newly-eclosed flies with reduced stores of larval/pupal energy and diminished starvation tolerance and fecundity.
Introduction
The Insulin/IGF signaling (IIS) pathway plays a central role in nutrient-dependent growth control during development, as well as in adult physiology and aging [1-5]. More specifically, in mammals insulin, IGFs and relaxins act on different types of receptors to regulate metabolism, growth and reproduction [6-9]. This class of peptide hormones has been well conserved over evolution and therefore the genetically tractable fly Drosophila is an attractive model system for investigating IIS mechanisms [10,4,11]. Eight insulin-like peptides (DILP1-8), each encoded on a separate gene, have been identified in Drosophila. Based on sequence similarities DILP1-5 are considered related to bona fide insulins, DILP6 is IGF-like, whereas DILP7 and DILP8 are relaxin-like [12,10,13,14]. The genes encoding these DILPs display differential temporal and tissue-specific expression profiles, suggesting that they have different functions [12,14-17]. Specifically, DILP1, 2, 3 and 5 are mainly expressed in median neurosecretory cells located in the dorsal midline of the brain, designated insulin-producing cells (IPCs) [12,18,16,19,20]. From the IPCs the DILPs can be released into the open circulation from axon terminations in the corpora cardiaca, the anterior aorta and crop. Genetic ablation of the IPCs alters growth and metabolism, leads to increased resistance to several forms of stress and prolongs lifespan [21,18]. The functions of the individual DILPs produced by the IPCs may vary depending on the stage of the Drosophila life cycle. Already the temporal expression patterns hint that DILP1-3 and 5 play different roles during development; whereas DILP2 and 5 are relatively highly expressed during larval and adult stages, DILP1 and 6 are almost exclusively expressed during pupal stages under normal conditions [22,15].
DILP1 is unique among the IPC-produced peptides since it can be detected primarily during the non-feeding pupal stage and the first few days of adult life when residual larval/pupal fat body is present [15,16]. Furthermore, in female flies kept in adult reproductive diapause, where feeding is strongly reduced, dilp1/DILP1 expression is also high [16]. Its temporal expression profile resembles that of DILP6 although this peptide is primarily produced by the fat body, not IPCs [22,15]. Since DILP6 was shown to regulate growth of adult tissues during pupal development [22,15], we asked whether also DILP1 plays a role in growth control. It is known that overexpression of several of the DILPs is sufficient to increase body growth through an increase in cell size and cell number, and especially DILP2 produces a substantial increase in body mass [12,23,24]. In contrast, not all single dilp mutants display a decreased body mass. The dilp1, dilp2 and dilp6 single mutants display decreased body weight [10,22,15], whereas the dilp3, dilp4, dilp5 and dilp7 single mutants retain normal body weight, indicating that some of the individual DILPs have redundant functions [10]. However, a triple mutation of dilp2, 3, and 5 causes a drastically reduced body weight, and a dilp1–4,5 mutation results in even smaller flies [10,25].
There is a distinction between how DILPs act in growth regulation. DILPs other than DILP1 and 6 promote growth primarily during the feeding larval stages when their expression is high [12,23]. This nutrient dependent growth is relatively well understood and is critical for production of the steroid hormone ecdysone and thereby developmental timing and induction of developmental transitions such as larval molts and pupariation [26-30]. The growth during non-feeding stages, which affects imaginal discs and therefore adult tissues, is far less studied. In this study, we investigate the role of DILP1 in growth regulation in Drosophila. We found that mutation of dilp1 diminishes body mass and ectopic dilp1 expression promotes organismal growth during the non-feeding pupal stage, similar to DILP6.
We also investigated the role of dilp1 mutation and overexpression on early adult physiology, and found that dilp1 manipulation affects starvation resistance in newly eclosed flies, but less so with increasing age. Testing flies that are three days or a week old shows that dilp1/dilp2 double mutants are more sensitive to starvation than the other mutants and controls. The diminished starvation resistance in newly hatched flies after dilp1 overexpression is probably a consequence of nutrient reallocation during the increased growth of adult tissues leading to newborn flies with low energy stores. Conditional dilp1 overexpression in young adults results in a slightly decreased survival during starvation suggesting a need for intact DILP signaling also in adult homeostasis. Interestingly, the newly eclosed dilp1 mutant flies are less resistant to starvation than controls and dilp2 mutants; the common notion is that reduced IIS increases survival during starvation [21].
Taken together, our study suggests that DILP1 promotes growth of adult tissues during the non-feeding pupal stage, and this carries over to affect the metabolism in the young adult fly. Most of the phenotypes from dilp1 knockout and overexpression are stronger in female flies. We suggest that dilp1, similar to dilp6 [15], may be a hormonal factor that ensures that a larva exposed to poor nutritional conditions will as a pupa utilize stored nutrients for growth of adult tissues, rather than keeping these stores for the first days of adult life. Our findings indicate a separate functional role of dilp1 in the adult fly in metabolism, and lifespan.
Experimental procedures
Fly lines and husbandry
Parental flies were reared and maintained at 18°C with 12:12 Light: Dark cycle on a food recipe from Bloomington Drosophila Stock Center (BDSC) (http://fly-stocks.bio.indiana.edu/Fly_Work/media-recipes/bloomfood.htm). The experimental flies were reared and maintained at 25°C, with 12:12 Light:Dark cycle on an agar-based diet with 10% sugar and 5% dry yeast.
The following Gal4 lines were used in this study: dilp2-Gal4 [[18] from E. Rulifson, Stanford, CA], Pdf-Gal4 (obtained from BDSC, Bloomington, IN), Ppl-Gal4 [[31] from M.J. Pankratz, Bonn, Germany], To-Gal4 [[32] from B. Dauwalder, Houston, TX], DaGS-Gal4 [Daughterless gene switch; [33] obtained from V. Monnier, Paris, France], c929-Gal4 [[34] from Paul H. Taghert], yw; UAS-dilp6, yw; UAS-dilp2;+ [[23] from H. Stocker, Zürich, Switzerland]. Several UAS-dilp1 lines were produced is a previous study [35] and two of them, UAS-dilp1 (II) and UAS-dilp1 (III), were used here. As controls we used w1118 or yw obtained from BDSC, crossed to Gal4 and UAS lines. All flies (except yw; UAS-dilp6, and yw; UAS-dilp2;+) were backcrossed to w1118 for at least 6 generations.
We used a double null mutation of dilp1/dilp2 that was previously generated by homologous recombination and verified as described by Post et al. [35]. Also single dilp1 and dilp2 null mutants were employed. As described earlier [35]; these were obtained from BDSC and a residual w+ marker was Cre excised followed by chromosomal exchange to remove yw markers on chromosomes 2 and X.
To activate the GeneSwitch-Gal4 driven dilp1 expression in the adult stage, DaGS-Gal4>UAS-dilp1 flies were raised on normal food until two days of age to allow them to mate. Thereafter the flies were transferred either to food that contained RU486 (mifepristone; Sigma, St. Louis, MO, USA) at a final concentration of 20 µM dissolved in EtOH, or food containing the same concentration of solvent. The flies were kept on this food for five days after which the experiments were performed.
Antisera and immunocytochemistry
For immunolabeling, tissues from larvae or female adults were dissected in chilled 0.1 M phosphate buffered saline (PBS). They were then fixed for 4 hours in ice-cold 4% paraformaldehyde (PFA) in PBS, and subsequently rinsed in PBS three times for 1 h. Incubation with primary antiserum was performed for 48 h at 4°C with gentle agitation. After rinse in PBS with 0.25% Triton-X 100 (PBS-Tx) four times, the tissues were incubated with secondary antibody for 48 h at 4°C. After a thorough wash in PBS-Tx, tissues were mounted in 80% glycerol with 0.1 M PBS.
The following primary antisera were used: Rabbit or guinea pig antiserum to part of the C-peptide of DILP1 diluted 1:10000 [16]. Rabbit antisera to A-chains of DILP2 and DILP3 [36] and part of the C-peptide of DILP5 [37] all at a dilution of 1:2000, rabbit anti-AKH (1:1000) from M.R. Brown, Athens, GA, rabbit anti-pigment-dispersing hormone (1:3000) from H. Dircksen, Stockholm, Sweden [38], rabbit antiserum to cockroach leucokinin I (LK I) at 1:2000 [39], mouse anti-green fluorescent protein (GFP) at 1:000 (RRID: AB_221568, Invitrogen, Carlsbad, CA).
The following secondary antisera were used: goat anti-rabbit Alexa 546 antiserum, goat anti-rabbit Alexa 488 antiserum, and goat anti-mouse Alexa 488 antiserum (all from Invitrogen). Cy3-tagged goat anti-guinea pig antiserum (Jackson ImmunoResearch, West Grove, PA). All were used at a dilution of 1:1000.
Image analysis
Images were captured with a Zeiss LSM 780 confocal microscope (Jena, Germany) using 10×, 20× and 40× oil immersion objectives. The projection of z-stacks was processed using Fiji (https://imagej.nih.gov/ij/). The cell body outlines were extracted manually and the size and staining intensity were determined using ImageJ (https://imagej.nih.gov/ij/). The background intensity for all samples was recorded by randomly selecting three small regions near the cell body of interest. The final intensity value of the cell bodies was determined by subtracting the background intensity.
Images of pupae, adult flies and fly wings were captured with a Leica EZ4HD light microscope (Wetzlar, Germany). The size of the adult fly body and wings were determined using Fiji. The pupal volume (v) was calculated using the equation v = 4/3 π (L/2) × (l/2)2, in which L = length and l = width [40]. Thorax length was measured from the posterior tip of the scutellum to the base of the most anterior point of the humeral bristle.
Pupariation time, egg to pupae viability and adult body weight
To determine time to pupariation, 6-7 day old adult females were crossed in the evening. The following morning, adult flies were transferred to vials with fresh food on which they were allowed to lay eggs for four hours. Two hours after the initiation of egg laying was considered time “0”, and thereafter the number of pupae was monitored at 6 or 12-hour intervals. To investigate the viability of egg to pupae formation, one pair of 6-7 day old adult flies was allowed to lay eggs for 24 hours after which the total number of eggs was counted. Subsequently, the total number of pupae was counted and the viability of egg to pupae was determined as pupa number/egg number × 100%. The body weight (wet weight) of single adult flies was determined using a Mettler Toledo MT5 microbalance (Columbus, USA).
Starvation survival assay
Newly hatched and mated 6-7 day old adults were used for starvation resistance experiments. For newly hatched flies, we collected virgin flies every 4 hours, to be used for starvation experiments. The flies were kept in vials containing 5 ml of 0.5% aqueous agarose (A2929, Sigma-Aldrich). The number of dead flies was counted at least every 12 hours until all the flies were died. At least 110 flies from 3 replicates were used for the analysis.
Capillary feeding (CAFE) assay
Food intake was measured using a slightly modified capillary feeding (CAFE) assay following Ja et al. [41]. In brief, female flies were placed into 1.5-ml Eppendorf micro centrifuge tubes with an inserted capillary tube (5 µl, Sigma) containing 5% sucrose, 2% yeast extract and 0.1% propionic acid. To estimate evaporation, three food-filled capillaries were inserted in identical tubes without flies. The final food intake was determined by calculating the decrease in food level minus the average decrease in the three control capillaries. Food consumption was measured daily and calculated cumulatively over four consecutive days. For this assay we used 8-10 flies in each of three biological replicates.
Metabolite quantification
Glycogen and triacyl glyceride (TAG) levels were assayed as previously described [42,43,35]. For glycogen assays, 5-6 adult female flies per sample were homogenized in PBS and quantified using the Infinity Glucose Hexokinase reagent by spectrophotometry. For TAG assays, 5-6 adult female flies per sample were homogenized in PBS + 0.05% TBS-T and quantified using the Infinity Triglycerides reagent by spectrophotometry. The fly lysate protein levels were determined by BCA assay (Thermo Fisher) and metabolite levels were normalized to protein level.
Statistical analysis
All results are presented as means ± SEM. We first investigated normality of data using Shapiro-Wilk’s normality test, then used one-way analysis of variance (ANOVA) or Student’s t-test, followed by Tukey’s multiple comparisons test. Lifespan data were subjected to survival analysis (Log rank tests with Mantel-Cox post-test) and presented as survival curves. Prism GraphPad version 6.00 (La Jolla, CA, USA) was used for generating the graphs.
Results
Mutation of dilp1 decreases body mass
Growth in Drosophila is in part regulated by several of the DILPs through activation of the canonical IIS/TOR (target of rapamycin) pathway [12,11,28]. It was previously reported that decreased dilp1 activity reduces adult body mass in Drosophila, but it was not investigated at what developmental stage this occurred [10,19]. This is relevant to ask since dilp1 displays a restricted temporal expression during the Drosophila life cycle (see Fig. 1A). To analyze growth effects of dilp1 and possible interactions with its tandem-encoded paralog dilp2, we employed recently generated dilp1, dilp2 and double dilp1-dilp2 null mutants [35]. The efficacy of these mutants was confirmed by qPCR in stage 8-9 pupae and immunolabeling in one-week-old mated female flies (Suppl Fig. 1). It can be noted that in dilp1 mutant pupae the mRNA levels of dilp2, dilp3 (not shown) and dilp6 were not altered, but in dilp6 mutants the dilp1 level was upregulated (Suppl Fig. 1A-C). At the protein level DILP2 but not DILP3 immunofluorescence increased in dilp1 mutants (Suppl Fig. 1D-G). These findings suggest only minor compensatory changes in other dilps/DILPs in dilp1 mutants during the pupal stage.
We monitored the body mass (wet weight) of dilp1, dilp2 and dilp1/dilp2 double mutants. First we measured the body weight both in newborn and 6-7 day old adult mated dilp1 mutant flies. In female flies the newly hatched dilp1 mutants displayed a decrease in body weight compared to controls (Fig. 1B). However, this difference in body weight was no longer detectable in 6-7-day-old mated flies kept under normal feeding conditions; a significant weight increase was observed (Fig. 1B). Also dilp2 mutant female flies have significantly lower body weight than controls one day after emergence, but in contrast to dilp1 mutants they did not increase the mass over 6-7 days of feeding (Fig. 1B). Interestingly the weight of dilp1/dilp2 double mutants was not significantly affected compared to the single mutants (and control) and no weight increase was seen the first week, except in control flies (Fig. 1B). Thus, there was no additive effect of the two mutations. In male flies none of the mutant flies displayed altered body mass (Fig. 1C). To determine whether decreased organismal growth was responsible for the lower body mass we measured wing size in the female mutant flies and found no significant difference to controls (Fig. 1D). Thus, the decreased mass of the flies does not seem to reflect a decrease in organismal size.
We next asked whether the weight gain over the first 6-7 days seen in Fig. 1B was caused by increased feeding. Using the capillary feeding (CAFE) assay over four days we found that during the first day of assay the dilp1 mutant flies actually fed less than the other mutants and control flies (Fig. 1E). The subsequent days food intake was not significantly different between the genotypes. Thus, the food intake profile does not explain the weight gain over the 6-7 days (Fig. 1E); possibly the female dilp1-/- flies excrete less waste or spend less energy. It was shown earlier that 1 week old dilp1 mutant flies display a two-fold increased expression of dilp6 transcript [16], that might compensate for the loss of dilp1. However, in the midpupal stage there is no significant upregulation of dilp6 in dilp1 mutants (Suppl Fig. 1C).
In a study of dilp6 it was shown that if third instar larvae (after reaching critical size) were put on a low protein diet they emerged as smaller adults and that this was accentuated in dilp6 mutants, suggesting that dilp6 is important for assuring growth of adult tissues under low protein conditions [15]. We, thus, performed a similar experiment with dilp1 mutant larvae kept on normal food or low protein diet. Flies emerging from larvae on restricted protein indeed displayed significantly lower body mass and female dilp1 mutants weighed less than controls under protein starvation (Fig. 1F). It was shown that dilp6 transcript increased in larvae exposed to protein starvation [15]. Third instar larvae exposed to similar starvation did not display any change in DILP1 immunolevels in IPCs (not shown).
Overexpression of dilp1 promotes growth during the non-feeding pupal stage
Having shown effects of the dilp1 null mutation on adult fly mass we next explored the outcome of over-expressing dilp1, either in IPCs, or more broadly. For this we generated several UAS-dilp1 lines [see [35]]. These UAS-dilp1 lines were verified by DILP1 immunolabeling after expression with several Gal4 drivers (Suppl Fig. 2A-D) and by qPCR in stage 8-9 pupae (Suppl Fig. 3A-F). Overexpression of dilp1 in fat body (ppl-Gal4 and to-Gal4) and IPCs (dilp2-Gal4) results in a drastic upregulation of dilp1 RNA (Suppl Fig. 3A, D), but has no effect on dilp2 and dilp6 expression (Suppl Fig. 3B, C, E. F), except a minor decrease in dilp2 for ppl-Gal4 (Suppl Fig. 3B). At the protein level dilp1 overexpression resulted in minor changes in DILP2, 3 and 5 immunolevels in IPCs of one week old adult female flies (Suppl Fig. 4A-E). One line, UAS-dilp1 (III), was selected for subsequent experiments since it generated the strongest DILP1 immunolabeling.
First we used a dilp2-Gal4 driver to express dilp1 in the IPCs and detected a significant increase in body mass of female flies (Fig. 2A). Next we expressed dilp1 in the fat body, the insect functional analog of the liver and white adipocytes in mammals [44,45]. The fat body displays nutrient sensing capacity, and is an important tissue for regulation of growth and metabolism in Drosophila [46,15,47-49]. It is also the tissue where DILP6 is produced and released [46,15]. To investigate the effect of ectopic dilp1 expression in the fat body, we used the fat body-specific pumpless (ppl) and takeout (to) Gal4 drivers. The efficiency of the drivers was confirmed by DILP1 immunostaining of larval fat body of ppl>dilp1 and to>dilp1 flies, but not in the control flies (Suppl. Fig. 2D). In ppl>dilp1 flies we also found DILP1 labeling in the nephrocytes, which are highly endocytotic cells located close to the heart [50] (not shown). Possibly the immunoreactive DILP1 has accumulated from the circulation after release from the fat body.
The effect of dilp1 overexpression in the fat body was monitored both on adult body mass and organismal size. We also measured the time to pupariation and size of pupae to determine whether dilp1 overexpression affected larval development and growth. Using the ppl-Gal4 driver we did not observe any effect on the time from egg to pupa compared to controls (Fig. 2B). Pupal volume, as a measurement of larval growth, was not altered by ppl-Gal4>dilp1 (Fig. 2C). As expected [15,46], over-expression of dilp6 also had no effect on pupal size (Fig. 2C). However, as shown earlier for ubiquitously expressed dilp2 [23], dilp2 expression in the fat body generated a strong increase in pupal volume, suggesting growth during the larval feeding stage (Fig. 2C). Driving dilp1 with the c929 Gal4 line, that directs expression to several hundred dimm-expressing peptidergic neurons including IPCs [51], we did not observe any effect on time to pupariation or pupal volume (Fig. 2B, C). Taken together our data suggest the ectopic dilp1 does not affect larval growth or developmental time.
Next we determined the body mass of mated 6-7 d old flies. Body weight increased significantly in ppl>dilp1 flies compared to the control flies both in female (Fig. 2D) and male flies (Fig. 2E). Here we additionally noted increased weight for ppl>dilp2 and ppl>dilp6 flies. We also monitored the weight of one day old flies and found that ppl>dilp1, but not dilp2>dilp1 flies displayed increased mass (Fig. 2F). Moreover, organismal size, estimated by wing size (Fig. 2G,H) and thorax length (Fig. 2G, I), increased after ectopic expression of dilp1 in the fat body (see also Fig. 2H). Since we see no effect of dilp1 expression on developmental time or pupal volume, but register increased body mass and size, we propose that dilp1, like dilp6, promotes growth of adult tissues during the pupal stage.
It was suggested that dilp6 promotes growth of adult tissues during pupal development by utilizing nutrients stored in the larval fat body, which is carried into the pupa [15]. This may be the case also for dilp1, and if so, newly hatched dilp1 overexpressing flies should have less energy stores in the form of residual larval fat body. To test this idea we monitored feeding in recently hatched dilp1 mutant flies and controls. Indeed, flies overexpressing dilp1 displayed increased food ingestion over the first four days after adult emergence compared to controls (Fig. 2J). Next we compared the weights of one day old and 6-7 day old flies after dilp1 overexpression with ppl-Gal4 and found that at both ages the female ppl>dilp1 flies weighed more and that the older flies were heavier than the younger ones (Fig. 2K). In male flies ppl>dilp1 also increased the body mass, but there was a loss of weight for all genotypes over the first 6-7 days of adult life (Suppl. Fig. 5A). As a comparison dilp2>dilp1 had only minor effects on body mass of female flies, only in 6-7 d old flies there was an increase (Suppl. Fig. 5B), whereas in males a significant decrease was noted at both ages (Suppl. Fig. 5C).
Using the to-Gal4 fat body driver to express dilp1 we also noted an increase in weight of recently emerged female and male flies (Suppl. Fig. 5D, E), but no change in body size except a minor increase in thorax length in females (Suppl. Fig. 5F, G). The female to>dilp1 flies increased further in weight the first 6-7 days of adult life, but not later (Suppl. Fig. 5D), whereas the males did not (Suppl. Fig. 5E). Furthermore, with the to-Gal4 driver there was no increase in pupal volume, indicating that dilp1 does not affect larval growth (Suppl. Fig. 5H).
Ectopic expression of dilp1 in neuroendocrine cells by means of the c929-Gal4 did increase body weight (Suppl. Fig. 6A), but had no effect on wing size in males or females or food intake in young flies (Suppl. Fig. 6B, C), suggesting that dilp1 expression (or release) was not strong enough to yield major effects. Also dilp2>dilp1 flies were tested in food intake and no effect was seen (Suppl. Fig. 6C).
Effects of dilp1 manipulations on metabolism in newborn and older flies
To investigate whether energy is reallocated during pupal development we monitored the levels of triacylglycerids (TAG), glycogen and glucose in recently emerged and three day old dilp mutant and dilp1-overexpression female flies (Fig. 3). In newborn dilp1 mutant flies glycogen was significantly lowered, whereas glucose and glycogen was diminished in dilp2 mutants while in the dilp1/dilp2 double mutants all three compounds were decreased (Fig. 3A-C). In the three-day-old flies dilp1 and double mutants displayed reduced glycogen, whereas in dilp1/dilp2 double mutants TAG was increased (Fig. 3D-F). Using ppl-Ga4 to express dilp1 we found that the only effect was a reduction of glycogen in newborn flies; at 3 or 7 days of age no effect was noted (Fig. 3G-I). Thus, it appears that intact dilp1 signaling is required for mobilization of glycogen stores in young flies.
dilp1 overexpression increases the size of the adult brain and neuroendocrine cells
It was previously shown that signaling through the Drosophila insulin receptor (dInR) can regulate growth of cell bodies of neuroendocrine cells in a cell autonomous manner, and that dilp6 in glial cells is a candidate ligand to mediate this dInR dependent growth [52,53]. Since dilp1 has an expression profile similar to dilp6 and promotes growth of adult tissues in the pupal stage we asked whether dilp1 also affects size of neuroendocrine cells that differentiate in the pupa. Thus, we overexpressed dilp1 with the broad driver c929-Gal4, and monitored the cell body size of several groups of neuroendocrine cells in the adult with specific peptide antisera. We found that the cell body size of IPCs increased in c929>dilp1 adult flies, as shown by anti-DILP2 staining (Fig. 4A1-A3). Furthermore, the cell bodies of the adult-specific pigment dispersion factor (PDF) expressing clock neurons (l-LNvs), as shown here by anti-PDF staining, were enlarged in c929>dilp1 flies compared to the controls (Fig. 4B1-B3). Finally, we monitored the cell-body size of leucokinin (LK) producing neurons in the abdominal ganglia (ABLKs), and found that the adult-specific anterior, but not the larval-derived posterior ABLKs, displayed increased size in c929>dilp1 flies (Fig. 4C1-C3). However, the observed increase in cell body size may be partly due to a broader growth of the adult fly tissues, since we found that also the size of the brain increased in c929>dilp1 flies (Figure 4D). The c929-Gal4 is expressed in IPCs and several other groups of neurosecretory cells that could underlie systemic release of DILP1, which affects growth. In contrast, we found that expressing dilp1 in interneurons, such as PDF-expressing clock neurons does not induce growth of brain neurons (Suppl. Fig. 7A,B) or size of the brain (Suppl. Fig. 7C), but affected the intensity of PDF immunolabeling (Suppl. Fig. 7D). Thus, paracrine release of DILP1 in the brain does not seem to affect growth of neurons. Interestingly, we found that in third instar larvae, the cell body size of ABLK neurons or the size of the CNS were not different in c929>dilp1 larvae compared to controls (Fig. 4F), suggesting that dilp1 overexpression has no effect on neuron growth during the larval stage. Using the ppl-Gal4 to drive dilp1 in the fat body we also found an increase in the size of the PDF expressing clock neurons (Fig. 4G1-G3) and the brain (Fig. 4H) supporting the proposal that systemic DILP1 is required to promote this growth. Finally, since overexpression of dilp6 in glial cells by Repo-Gal4 promotes neuronal cell body growth [53], we tested overexpression of dilp1 in these cells, but found no significant effect on the cell-body size of PDF neurons (Suppl. Fig. 8), again indicating that to affect growth DILP1 must act systemically rather than in a paracrine fashion. Finally, there was no effect on body mass after expressing dilp1 with the Repo-Gal4 (Suppl. Fig. 8C, D).
Effects of dilp1 on adult physiology
Genetic ablation of the IPCs, which produce DILP1, 2, 3 and 5, results in enhanced starvation resistance in adult flies [21]. Thus, we asked whether the alterations of dilp1 expression during development have effects on adult physiology such as survival during starvation or desiccation (as a proxy for effects on metabolism). We investigated the starvation resistance in newborn, three days old and one-week-old female dilp1, dilp2 and double mutant flies. The newly eclosed dilp1 mutant flies display strongly reduced survival during starvation and double mutants increased survival compared to control flies, whereas the stress resistance of dilp2 mutants is similar to the controls (Fig. 5A, Table 1). In three days old virgin flies the dilp1 and dilp1/dilp2 mutants display reduced survival during starvation, whereas the dilp2 mutants perform similar to the controls (Fig 4B, Table 1). In a separate study [35] it was shown that 6-7 day old female flies display a similar response to starvation: the dilp1/dipl2 mutants exhibit the strongest reduction in survival, followed by dilp1 mutants that also are much less stress tolerant, whereas dilp2 mutants and control flies perform very similar (Table 1). Here we tested also 6-7 day old male flies and found that they survived starvation in a manner different from females with dilp2 and double mutants displaying diminished stress resistance whereas dilp1 mutants survive similar to controls (Fig. 4C).
As seen above, our data suggest a change in the response to loss of dilp function over the first week of adult life. It is known that newly hatched wild type flies are more resistant to starvation than slightly older flies [54]. Thus, we compared the survival during starvation in recently hatched and three day old virgin flies. As seen in Fig. 5D (based on data in Fig. 5A and B), recently hatched control flies (w1118) indeed exhibit increased starvation resistance compared to controls that were tested when three days old. Also the dilp1 mutant flies are more stress resistant when tested as newly hatched than as older flies, and the mutants perform less well than controls at both ages (Fig. 5D). However, the most drastic change within the first week is that dilp1 mutants yield the strongest phenotype as newborn flies and then in 3d and 6-7 d old flies the dilp1/dilp2 mutants are the ones with the lowest stress resistance. Thus, a shift in dilp function seems to occur as the fly matures during the first few days of adult life. To provide additional evidence that dilp1 impairs starvation resistance we performed dilp1-RNAi using a dilp2-Gal4 driver. The efficiency of the dilp2>dilp1-RNAi was tested by qPCR (Suppl. Fig. 9A) where a strong decrease in dilp1, but not dilp2 or dilp6 was seen. The dilp1-RNAi resulted in newly eclosed flies that displayed reduced survival during starvation (Suppl. Fig. 9B), similar to dilp1 mutant flies.
It is also interesting to note that the diminished starvation resistance in dilp1 and dilp1/dilp2 mutants is opposite to the phenotype seen after IPC ablation, mutation of dilp1-4, or diminishing IIS by other genetic interventions [21,10,55,56]. Thus, in recently hatched flies dilp1 appears to promote starvation resistance rather than diminishing it. Furthermore, the decreased survival during starvation in female dilp1 mutants is the opposite of that shown in dilp6 mutants [15] indicating that dilp1 acts by mechanisms different from the other insulin-like peptides.
Next we investigated the effect of the mutations on the flies’ response to desiccation (dry starvation). One-week-old flies were put in empty vials and survival recorded. Female dilp1/dilp2 mutants were more sensitive to desiccation than controls and the single mutants (Fig. 5E). In males the double mutants also displayed higher mortality during desiccation, whereas the two single mutants were more resistant than controls (Fig. 5F). Thus, there is a sex dimorphism in how the different mutants respond to both desiccation and starvation.
When overexpressing dilp1 with the fat body driver ppl-Gal4 newly eclosed and 6-7 d old female flies become less resistant to starvation compared to parental controls (Fig. 6A, B). However, in 6-7-day-old male flies there is no difference between controls and flies with ectopic dilp1 (Fig. 6C). Thus, in females it appears as if both knockout and over expression of dilp1 reduces starvation resistance, maybe due to offsetting a narrow window of homeostasis. It was shown earlier that conditional knockdown of dilp6 by RNAi during the pupal stage resulted in newborn flies with increased survival during starvation [15], suggesting that the effect the dilp1 null mutation is different. After ectopic expression of dilp1 in the fat body there was an increase in food intake (cumulative data) in one-week-old flies over four days (Fig. 6D), suggesting that metabolism is still altered in older flies.
We furthermore investigated starvation resistance in flies overexpressing dilp1 in IPCs (dilp2>dilp1) and found that in newborn flies overexpression reduced survival (Fig. 6E), whereas in a week old flies all genotypes displayed the same survival (Fig. 6F).
Since the effect of dilp1 manipulations seems stronger in female flies we asked whether fecundity is affected by overexpression of dilp1. An earlier study showed that dilp1 mutant flies are not deficient in number of eggs laid, or the viability of offspring (egg to pupal viability), although the dilp1/dilp2 double mutants displayed a reduction in viability of these eggs [35]. Here, we expressed dilp1 in fat body (ppl-Gal4) and neuroendocrine cells (c929-Gal4) and both lines resulted in flies that laid eggs that exhibited decreased viability as monitored by numbers of eggs that developed into pupae (Fig. 6G).
We next asked whether there is any physiological trigger of increased dilp1 expression, except for diapause [16] and experimental ones such as ectopic expression of sNPF or knockdown of dilp6, dilp2 and dilp2,3,5 [57,16,35]. Although diminished protein diet in larvae had no effect on dilp1 expression measured by qPCR (not shown), we found that 40 h starvation of 10 d old flies (w1118) leads to a significant increase in dilp1, but not dilp2 or 6 (Fig. 6H). Thus, at a time (12 d) when dilp1 is very low under normal conditions, it is upregulated four times during starvation, further suggesting that the peptide indeed plays a role also in older adult flies.
The functional homolog of glucagon in flies, adipokinetic hormone (AKH), plays important roles both in metabolism and regulation of lifespan [58-60]. A previous paper showed that in dilp1 mutant flies levels of AKH were not affected [35]. Here we found that dilp1 overexpression with the c929-Gal4 driver induced an increase in AKH immunolabeling in one-week-old flies (Fig. 6I). Thus there appears to be an interaction between dilp1 and AKH that may underlie some of the effects of this insulin on metabolism and stress tolerance.
To further test whether dilp1 has direct effects on metabolism/ starvation resistance in the adult fly we over-expressed dilp1 with the drug inducible gene-switch (GS) system [61] in one-week-old adults. We used a daughterless (Da)-Gal4GS driver to conditionally express dilp1 and fed flies RU486 to activate the Gal4 in the adult stage. Starvation resistance was significantly reduced in flies fed RU486 compared to controls that were not exposed to the drug (Fig. 7A). The Da-Gal4GS driven expression of DILP1 was monitored by immunolabeling, and found to include median neurosecretory cells and several other neuron types in the brain, including mushroom bodies and the central body (Fig. 7B,C).
Discussion
Our study indicates a role for dilp1 in regulation of adult tissue growth during the pupal stage, as well as roles in adult physiology, especially during the first days of adult life. The experiments herein suggest that the developmental role of dilp1 may be to ensure nutrient reallocation in the pupa toward growth of adult tissues if the larva was exposed to restricted food sources. In the adult dilp1 is upregulated during starvation and genetic gain and loss of function of dilp1 signaling alters the flies’ survival under starvation conditions. These novel findings combined with previous data showing high levels of dilp1 during adult reproductive diapause [16] and its role as a pro-longevity factor during aging [35] demonstrate a wide-ranging importance of this signaling system. Not only does dilp1 expression correlate with stages of non-feeding (or reduced feeding), these stages are also associated with lack of reproductive activity, and encompass the pupae, newly eclosed flies, and diapausing flies. Under normal conditions, diminished dilp1/DILP1 expression during the first few days of the adult could relate to a metabolic transition and the onset of sexual maturation.
In Drosophila, the final body size is determined mainly during the larval feeding stage [23,12,11,29]. However, regulation of body size can also occur after the cessation of the feeding stage, and this process is mediated by dilp6 acting on adult tissue growth in the pupa in an ecdysone-dependent manner [15,46]. This is likely a mechanism to ensure growth of the adult tissues if the larva is exposed to shortage of nutrition during its feeding stage. Our findings suggest that dilp1 is another regulator of growth during the pupal stage. We show here that dilp1 promotes organismal growth in the non-feeding pupa at the cost of stored nutrients derived from the larval stage. As a consequence large dilp1-overexpressing flies display increased food ingestion over the first four days as adults and an altered response to starvation. Conversely dilp1 mutants hatched as flies with significantly smaller mass. Thus, both alterations in dilp1 expression influence the metabolic balance in early adults as seen in reduced starvation resistance. Our study suggests that dilp1 parallels dilp6 [15,46] in balancing adult tissue growth and storage of resources during pupal development, and affecting adult physiology. This is interesting since dilp6 is an IGF-like peptide that is produced in the nutrient sensing fat body [15,46], whereas the source of the insulin-like dilp1 is the brain IPCs. We showed earlier that young adult dilp1 mutant flies display increased dilp6 and vice versa [16], suggesting feedback between these two peptide hormones. This feedback appears less prominent in dilp1 mutants during the pupal stage with no effects on dilp2, dilp3 or dilp6 levels. However, dilp1 is slightly upregulated in dilp6 mutant pupae. As well, overexpression of dilp1 in fat body of IPCs has no effects on pupal levels of dilp2 and dilp6. Thus, at present we cannot postulate any compensatory changes in other DILPs in pupae with dilp1 manipulations. However, normally dilp6 levels are far higher than those of dilp1 [15,46], which could balance the effects of changes in dilp1 signaling. In adults, DILP6 is released from the fat body and is known to affect the brain IPCs to diminish DILP2 production/release and thereby extending lifespan [42]. It is not known whether DILP6 affects DILP1 release in the pupal stage.
Ectopic overexpression of dilp1 in neuroendocrine cells or fat body not only increases growth of wings and thorax, but also increases the size of the brain and the cell bodies of several kinds of neuroendocrine cells in adult flies. However, there was no change in the size of neuronal cell bodies or CNS during larval development after overexpression of dilp1. Thus, taken together, our findings suggest that dilp1/DILP1 is able to promote growth mainly during the non-feeding pupal stage. However, restricted protein diet during the later larval stage diminished the body mass of adult flies more in dilp1 mutants than in controls, similar to findings for dilp6 [15]. This suggests that dilp1 function is accessory to dilp6 in maintaining growth of adult tissues in situations where larvae obtain insufficient protein in their diet.
DILPs and IIS are involved in modulating stress responses in Drosophila [see [10,21,62]]. Flies with ablated IPCs or genetically diminished IIS display increased resistance to several forms of stress, including starvation [21,10]. Conversely, overexpression of dilp2 causes lethality in Drosophila [24]. We found that dilp1 mutant flies displayed decreased starvation resistance. Both in newborn, 3 day and 6-7 day old flies, mutation of dilp1 decreased survival during starvation. Curiously, overexpression of dilp1 in the fat body also resulted in decreased survival during starvation in young and older flies. The effects on adult physiology of dilp1 manipulations may be a consequence of the altered adult tissue growth during pupal development and associated reallocation of energy stores. However, we could show that conditional overexpression of dilp1 in the one week old flies also reduces starvation resistance indicating action of the peptide also at this stage. Action of dilp1 in the adult fly is also linked to reproductive diapause in females, where feeding is strongly reduced [63], and both peptide and transcript are upregulated [16]. Related to this we found here that dilp1 mRNA is upregulated during starvation in 12 d old flies. Furthermore, it was shown that expression of dilp1 increases lifespan in dilp1-dilp2 double mutants, suggesting that loss of dilp2 induces dilp1 as a factor that promotes longevity [35]. Thus, dilp1 activity is beneficial also during adult life, even though its expression under normal conditions is very low [16,46,15]. This pro-longevity effect of dilp1 is in contrast to dilp2, 3 and 5 and the mechanisms behind this effect are of great interest to unveil.
A previous study showed that in wild-type (Canton S) Drosophila DILP1 expression in young adults is sex-dimorphic with higher levels in females [16]. In line with this, we show here that increase in body weight the first week or adult life occurs only in female dilp1 mutant flies, and also that starvation survival in one-week-old flies is diminished only in females. Finally, we found that dilp1 overexpression specifically decreased starvation resistance only in female flies both in non-conditional and conditional experiments. Thus, taken together, we find that dilp1 displays a sex-specific expression and function in young adult Drosophila, and the dilp1 mutation affects body mass of newly eclosed flies mainly in females. It is tempting to speculate that the more prominent role of dilp1 in female flies is linked to reproductive physiology and early ovary maturation, which is also reflected in the upregulation during reproductive diapause [16].
This study demonstrates that dilp1 promotes growth during the pupal stage, and in females it regulates starvation resistance during the young adult stage, and affects fecundity. Like dilp6, perhaps dilp1 acts as a signal promoted by nutrient shortage during the larval stage to ensure growth of adult tissues by reallocating nutrient stores from larval fat body. This in turn results in depleted pupal-derived nutrient stores in young adults. Thus, IPC-derived dilp1 displays several similarities to the fat body-derived dilp6, including temporal expression, growth promotion, effects on adult stress resistance and lifespan. Additionally dilp1 may play a role in regulation of nutrient allocation/metabolism during the first few days of adult life, especially in females. At this time larval fat body is still present and utilized as energy/nutrient store [54]. There is a change in the action of DILP1 between the pupal and adult stages from being a stimulator of growth (agonist of dInR) in pupae, to acting opposite to DILP2 and other DILPs in adults in regulation of lifespan and stress responses. It is not known what mechanism is behind this switch in function of DILP1 signaling, but one possibility is that DILP1 acts via different signal pathways in pupae and adults. One obvious difference between these two stages is the presence of larval fat body in the pupa and first few days of adults and its replacement by functional adult fat body in later stages. In the future it would be interesting to investigate if DILP1 act differently on larval and adult fat body and whether dilp1 and dilp6 interact to regulate growth and metabolism in Drosophila.
Acknowledgements
We thank Bloomington Drosophila Stock Center (Bloomington, IN) and Drs. M.J. Pankratz, B. Dauwalder, V. Monnier, Paul H. Taghert and H. Stocker for flies, and Dr. M. R. Brown for antiserum.