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 far from understood. Previous work has shown that dilp1/DILP1 is transiently expressed mainly 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, similar to dilp6 manipulations. No growth effects of dilp1 or dilp6 manipulations were detected during larval development. We next show that dilp1 and dilp6 increase metabolic rate in the late pupa and promote lipids as the primary source of catabolic energy. This lipid mobilization in the pupa is not correlated with transcriptional changes of adipokinetic hormone. The effects of dilp1 manipulations carry over to the adult fly. In newly eclosed flies, survival during starvation is strongly diminished in dilp1 mutants, but not in dilp2 and dilp1-dilp2 double mutants, whereas in older flies only double mutants display reduced starvation resistance. In conclusion, dilp1 and dilp6 promote growth of adult tissues during the non-feeding pupal stage, likely by utilization of stored lipids. This results in larger newly-eclosed flies with reduced stores of pupal-derived nutrients 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 [4,10,11]. Eight insulin-like peptides (DILP1-8), each encoded on a separate gene, have been identified in Drosophila [10,12-14]. 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) [12,16,18-20]. The IPC derived DILPs can be released into the open circulation from axon terminations in the corpora cardiaca, the anterior aorta and the crop. Genetic ablation of the IPCs reduces growth and alters metabolism, and results in increased resistance to several forms of stress and prolongs lifespan [18,21].
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. Thus, 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 [15,22].
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 [15,22]. Since DILP6 was shown to regulate growth of adult tissues during pupal development [15,22], 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 weight [12,23,24]. In contrast, not all single dilp mutants display a decreased body mass. The dilp1, dilp2 and dilp6 single mutants display slightly decreased body weight [10,15,22], whereas the dilp3, dilp4, dilp5 and dilp7 single mutants display normal body weight [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/DILP1 in growth regulation in Drosophila in comparison to dilp6/DILP6. We found that mutation of dilp1 diminishes body weight and ectopic dilp1 expression promotes organismal growth during the non-feeding pupal stage, similar to dilp6. Determination of metabolic rate and respiratory quotient as well as TAG levels during late pupal development provides evidence that dilp1 and dilp6 increase the metabolic rate and that the associated increased metabolic cost is fueled by increased lipid catabolism. We, however, find no evidence for a role of the lipid mobilizing adipokinetic hormone (AKH) [31-33] in the altered lipid catabolism in pupae.
We also investigated the role of dilp1 mutation and overexpression on early adult physiology. Interestingly, the newly eclosed dilp1 mutant flies are less resistant to starvation than controls and dilp2 mutants. Thus, dilp1 acts differently from other dilps for which it has been shown that reduced signaling increases survival during starvation [21]. The decreased starvation resistance in newly hatched flies after dilp1 overexpression may be a consequence of diminishment of stored nutrients in the pupa during the increased growth of adult tissues, and thus less residual pupal fat body in newborn flies. Also early egg laying and fecundity are affected by dilp1.
Taken together, our data suggest that dilp1/DILP1 promotes growth of adult tissues during the non-feeding pupal stage, and that this process mainly utilizes stored lipids to fuel the increased metabolic rate. The effect of this increased metabolic rate in the pupa carries over to affect the metabolism in the young adult fly. We suggest that dilp1, similar to dilp6 [15], ensures that if a larva is exposed to poor nutritional conditions it will after pupariation utilize stored nutrients for growth of adult tissues, rather than keeping these stores for the first days of adult life.
Results
Mutation of dilp1 decreases body weight
Growth in Drosophila is in part regulated by several of the DILPs through activation of the canonical IIS/TOR (target of rapamycin) pathway [11,12,28]. It was previously reported that decreased dilp1 activity reduces adult body weight 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 [34]. The efficacy of these mutants was confirmed by qPCR in stage 8-9 pupae and immunolabeling in one-week-old mated female flies (S1 Fig). 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 (S1A-C Fig). At the protein level DILP2 but not DILP3 immunofluorescence increased in dilp1 mutants (S1D-G Fig). These findings suggest only minor compensatory changes in other dilps/DILPs in dilp1 mutants during the pupal stage.
We monitored the body weight (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 weight 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 weight (Fig 1C). To determine whether decreased organismal growth was responsible for the lower body weight we measured wing size in the female mutant flies and found no significant difference to controls (Fig 1D). Thus, the decreased weight of the flies does not seem to reflect a significant 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 a 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 (S1C Fig).
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 [15]. This suggests that dilp6 is important for assuring growth of adult tissues under low protein conditions. 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 weight and female dilp1 mutants weighed less than controls under protein starvation (Fig 1F). In male flies this latter effect was not seen in the mutants (Fig 1G).
We then asked whether mutation of both dilp1 and dilp6 would result in a further decrease of body weight and generated a recombinant dilp1-dilp6 mutant. Using qPCR we found that these flies displayed virtually no detectable dilp1 and dilp6 RNA (S2A Fig.). The weights of dilp1/dilp6 mutants were significantly reduced compared to controls (S2B Fig.). However, their weights were not diminished more that those of the single dilp1 and dilp6 mutants, indicating that there was no additive effect of loss of both dilps.
Overexpression of dilp1 promotes growth during the non-feeding pupal stage
Having shown effects of the dilp1 null mutation on adult flyweight we next explored the outcome of over-expressing dilp1, either in IPCs, or more broadly. For this we generated several UAS-dilp1 lines [see [34]]. These UAS-dilp1 lines were verified by DILP1 immunolabeling after expression with several Gal4 drivers (S3A-D Fig) and by qPCR in stage 8-9 pupae (S4A-F Fig). Overexpression of dilp1 in fat body (ppl-Gal4 and to-Gal4) and IPCs (dilp2-Gal4) results in a drastic upregulation of dilp1 RNA (S4A, D Fig), but has no effect on dilp2 and dilp6 expression (S4B, C, E, F Fig), except a minor decrease in dilp2 for ppl-Gal4 (S4B Fig). 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 (S5A-E Fig). 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 weight of female flies (Fig 2A). We then expressed dilp1 in the fat body, the insect functional analog of the liver and white adipocytes in mammals [35-37]. The fat body displays nutrient sensing capacity, and is an important tissue for regulation of growth and metabolism in Drosophila [15,37-41]. It is also the tissue where DILP6 is produced and released [15,38]. 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 (S3D Fig). In ppl>dilp1 flies we also found DILP1 labeling in the nephrocytes (not shown), which are highly endocytotic cells located close to the heart [42]. The immunoreactive DILP1 is likely to have accumulated from the circulation after release from the fat body since the ppl-Gal4 is not expressed in the nephrocytes.
Before monitoring the effect of dilp1 overexpression in the fat body on adult body weight and organismal size, we wanted to determine whether dilp1 has an effect on larval development. We therefore measured the time to pupariation and size of pupae to determine whether dilp1 overexpression affected timing of larval development and growth during this stage. 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 [15,38], 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 [43], 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 weight of mated 6-7 d old flies. Body weight increased significantly in ppl>dilp1 flies compared to the controls 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 weight (Fig 2F). However, dilp2>dilp1-RNAi induced a decrease in body weight (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. Since we see no effect of dilp1 expression on developmental time or pupal volume, but register increased body weight and size of adults, 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 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 than controls and that the older flies were heavier than the younger ones (Fig 2K). In male flies ppl>dilp1 also increased the body weight, but there was a loss of weight for all genotypes over the first 6-7 days of adult life (S6A Fig). As a comparison dilp2>dilp1 had only minor effects on body weight of female flies, only in 6-7 d old flies there was an increase (S6B Fig), whereas in males a significant increase was noted at both ages for dilp2>dilp1, and a loss of weight over the next six days for all genotypes (S6C Fig).
Using the to-Gal4 fat body driver to express dilp1 we also noted an increase in weight of recently emerged female and male flies (S6D, E Fig), but no change in body size except a minor increase in thorax length in females (S6F, G Fig). The female to>dilp1 flies increased further in weight the first 6-7 days of adult life, but not later (S6D Fig), whereas the males did not (S6E Fig). Furthermore, with the to-Gal4 driver there was no increase in pupal volume, supporting that dilp1 does not affect larval growth (S6H Fig).
Ectopic expression of dilp1 in neuroendocrine cells by means of the c929-Gal4 increased adult body weight (S7A Fig), but had no effect on wing size in males and females or food intake in young flies (S7B, C Fig), suggesting that dilp1 expression (and/or systemic release) was not strong enough to yield major effects. Also dilp2>dilp1 flies were tested in food intake and no effect was seen (S7C Fig).
Overexpression of dilp1 increases the size of the adult brain and neuroendocrine cells
It was previously shown that signaling through the Drosophila insulin receptor (dInR) can lead to an enlargement 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 [44,45]. Since dilp1 has a temporal 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 neuroendocrine cell driver c929-Gal4 [43,46], and monitored the cell body size of several groups of neuroendocrine cells in the adult CNS with specific peptide antisera. We found that the cell body size of IPCs increased in adult c929>dilp1 flies, as shown by anti-DILP2 staining (S8A1-3 Fig, Table 1). 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 (S8B1-3 Fig, Table 1). Next, 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 (S8C1-3 Fig, Table 1).
However, the observed increase in cell body size appears to 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 (S8D Fig, Table 1). The c929-Gal4 is expressed in IPCs and several other groups of peptidergic neurosecretory cells [43,46], which could be the source of systemic release of ectopic DILP1 that affects brain and cell growth. To support that systemic DILP1 is required to promote this growth we employed the ppl-Gal4 to drive dilp1 in the fat body and found an increase in the size of the PDF expressing clock neurons (S8F1-3 Fig, Table 1) and the brain (S8G Fig, Table 1). In contrast, we found that expressing dilp1 in interneurons, such as PDF-expressing clock neurons does not induce growth of brain neurons (S9A, B Fig, Table 1) or size of the brain (S9C Fig, Table 1), but affected the intensity of PDF immunolabeling (S9D Fig). 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 (S9E-G Fig, Table 1), further supporting that dilp1 overexpression has no effect on cell growth during the larval stage. Finally, since overexpression of dilp6 in glial cells by Repo-Gal4 promotes increase in size of neuronal cell bodies [45], we tested overexpression of dilp1 in these cells, but found no significant effect on the cell-body size of PDF neurons (S10A, B Fig, Table 1). This again indicates that to affect cell/tissue growth DILP1 must act systemically rather than in a paracrine fashion.
Metabolic rate and respiratory quotient in pupae of different genotypes
To investigate the role of dilp1 in utilization of nutrients during pupal development we determined metabolic rate (MR) and respiratory quotient (RQ) in pupae of different genotypes. First we characterized the metabolic trajectory in control pupae (w1118) by measuring cumulative MR daily throughout pupal development (Fig 3A). These data show the exponential MR curve typical for developing insects, including D. melanogaster [47]. To minimize handling stress, we chose to investigate only the end of pupal development in more detail and measured MR and RQ in 4-day-old pupae (that is the cumulative MR between hours 96 and 120 after pupation). For this experiment we used only ppl-Ga4 overexpression animals, since the mutant animals displayed high mortality in the respirometry setup used here. As can be seen in Fig 3B and 3C the ppl>dilp1 and ppl>dilp6 differed significantly from the controls. The MR was higher and RQ lower in the overexpression flies than in the control flies. RQ values, around 0.6 in both overexpression lines, suggest pure lipid metabolism [48], and lipids are known to be a major or sole fuel during metamorphosis of insects [49,50]. Our findings strongly suggest that dilp1 (and dilp6) affects metabolism in the pupa, maybe to ensure that enough fuel is allocated for growth of adult tissues.
TAG, carbohydrates and AKH signaling in pupae of different genotypes
To determine whether it indeed are lipids that fuel growth of adult tissues in 4 day old pupae we determined TAG levels after over expression of dilp1 and dilp6 in fat body (ppl-Gal4). Pupae of both genotypes displayed increased weight (Fig. 3D) and also significantly reduced TAG levels (Fig. 3E), compared to controls of the same age. The decreased glycogen levels in pupae after ectopic expression of dilp1 and dilp6 were not significant (Fig. 3F) and glucose levels were not significantly changed (S11 Fig).
Since AKH is known to mobilize lipids in insects, including Drosophila [31-33,51], we determined levels of akh transcript in pupae with dilp1 and dilp6 overexpression (using ppl-Gal4) and in dilp mutants, at two different time points (2 and 4 d old pupae). There was no significant alteration in Akh transcript after dilp1 or dilp6 overexpression; the only phenotype was a slight upregulation in dilp1 mutants in 4 d pupae (S11A-D Fig). Next we analyzed levels of transcript of brummer (bmm), a lipase known to promote TAG mobilization [52], in pupae of the same stages and found no significant change in expression for any genotype (S11E-H Fig). We also measured transcript of the α–glucosidase tobi, which regulates glycogen levels and is a target of both DILPs and AKH [53], and found no effect of overexpression (not shown) or loss of function of dilp1 at either stage (S11 I, J Fig).
Effects of dilp1 manipulations on metabolism in newly eclosed and young flies
To investigate whether energy reallocation during pupal development affects adult physiology and metabolism, we monitored the levels of triacylglycerids (TAG), glycogen and glucose in recently emerged and three day old dilp mutant and dilp1-overexpressing female flies (Fig 4). 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 4A-C). In the three-day-old flies dilp1 and double mutants displayed reduced glycogen, whereas in dilp1/dilp2 double mutants TAG was increased (Fig 4D-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 4G-I). Thus, it appears that intact dilp1 signaling is required for mobilization of glycogen stores in newly emerged and young flies. This supports that dilp1 signaling in the late pupa affects metabolism and that this is carried over into the young adult.
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 pupal 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 newly emerged, three days old and one-week-old female dilp1, dilp2 and dilp1/dilp2 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 starvation 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 5B, Table 1). In a separate study [34] 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 (see 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 (S12A Fig).
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 emerged and three day old virgin flies. As seen in Fig 5C (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 starvation 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 change in the role of dilp1 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 (S13A Fig) 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 (S13B Fig), 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 [10,21,55,56]. 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 action is 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 5D). In males the double mutants also displayed higher mortality during desiccation, whereas the two single mutants were more resistant than controls (S12B Fig). 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, using ppl-and c929-Gal4 drivers (S13C-D Fig). We furthermore investigated starvation resistance in flies overexpressing dilp1 in IPCs (dilp2>dilp1) and in most neuroendocrine cells (c929>dilp1) and found that in newborn flies overexpression reduced survival (Fig 6C, E), whereas in a week old flies all genotypes displayed the same survival (Fig 6D, F). 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 7A), suggesting that metabolism is still altered in older flies. 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 [34]. Here, we expressed dilp1 in fat body (ppl-Gal4) and detected an increase in number of eggs laid over 24 h in 6-7 d old flies (Fig. 7B). Both ppl-Gal4-and c929-Gal4-driven dlip1 decreased the viability of eggs laid as monitored by numbers of eggs that developed into pupae (Fig 7C). As a comparison we noted no difference in number of eggs in 3-day-old dilp1 mutant flies (Fig. 7D).
We next asked whether there is any physiological trigger of increased dilp1 expression in adult flies, except for diapause [16] and experimental ones such as ectopic expression of sNPF or knockdown of dilp6, dilp2 and dilp2,3,5 [16,34,57]. Although diminished protein diet in larvae had no effect on dilp1 expression measured by immunolabeling (not shown), we found that 40 h starvation of 10 d old flies (w1118) leads to a significant increase in dilp1, but not in dilp2 or dilp6 (Fig 7E). 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, AKH, plays important roles both in lipid mobilization, metabolism and regulation of lifespan [31,51,58,59]. A previous paper showed that in dilp1 mutant flies levels of AKH were not affected [34]. Here we found that dilp1 overexpression with the c929-Gal4 driver induced an increase in AKH immunolabeling in one-week-old flies (Fig 7F). Thus in adult flies (in contrast to larvae) there appears to be an interaction between dilp1 and AKH that may underlie some of the effects of this DILP on metabolism and stress tolerance.
Discussion
Our study indicates a role for dilp1 in regulation of adult tissue growth during the pupal stage, as well as a function 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 utilization in the pupa to support 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 [34] 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, the diminishing dilp1/DILP1 expression during the first few days of adult life may be associated with a metabolic transition (fat body remodeling; [60]) and the onset of sexual maturation.
In Drosophila, the final body size is determined mainly during the larval feeding stage [11,12,23,29]. However, regulation of adult 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 [15,38]. 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. This is supported by RQ-data that clearly shows a shift from mixed-energy substrate energy metabolism in control flies towards almost pure lipid catabolism at the end of pupal development in the dilp1 overexpression flies (also seen for dilp6). Furthermore, TAG (but not carbohydrate) levels in dilp1 overexpression flies were clearly decreased, which likely reflects the shift in catabolic energy substrate also seen in the R/Q using respirometry. It should be noted that insects predominantly use lipids as fuel during metamorphosis [49,50] and dilp1 overexpression increases lipid catabolism. 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 weight. Both alterations in dilp1 expression influence the metabolic balance in early adults as manifested in reduced starvation resistance at this stage. Our study hence suggests that dilp1 parallels dilp6 [15,38] in balancing adult tissue growth and storage of nutrient resources during pupal development, and thereby probably affecting adult physiology. This is interesting since dilp6 is an IGF-like peptide that is produced in the nutrient sensing fat body [15,38], 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. Furthermore, overexpression of dilp1 in fat body or IPCs has no effect 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, under normal conditions dilp6 levels are far higher than those of dilp1 [38] (see also modENCODE_mRNA-Seq_tissues [61]), which could buffer the effects of changes in dilp1 signaling.
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 promotes growth mainly during the non-feeding pupal stage. Interestingly, restricted protein diet during the later larval stage diminished the body weight 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 responses to starvation, desiccation and oxidative stress in Drosophila [see [10,21,62]]. Flies with ablated IPCs or genetically diminished IIS display increased resistance to several forms of stress, including starvation [10,21]. Conversely, overexpression of dilp2 increases mortality in Drosophila [24]. We found that dilp1 mutant flies displayed diminished starvation resistance. Both in newborn and 3 day old flies, mutation of dilp1 decreased survival during starvation (but not in 6-7 day old ones). 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 increase in utilization of nutrient stores. 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 [34]. Thus, dilp1 activity is beneficial also during adult life, even though its expression under normal conditions is very low [15,16,38]. 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 of 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 over expression specifically decreased starvation resistance only in female flies. Thus, taken together, we found earlier that dilp1 displays a sex-specific expression [16] and here we show sex-specific function in young adult Drosophila, and the dilp1 mutation affects body weight 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 metabolism associated with reproductive physiology and early ovary maturation, which is also reflected in the dilp1 upregulation during reproductive diapause [16]. In fact, we show here that egg-laying increased after dilp1 overepression, and an earlier study demonstrated a decreased egg laying in dilp1 mutant flies [16].
This study demonstrates that dilp1 promotes growth of adult tissues during the pupal stage, and in females it influences starvation resistance during the young adult stage, and affects fecundity. Like dilp6, perhaps dilp1 acts as a signal promoted by nutrient shortage during the late larval stage to ensure growth of adult tissues by recruiting 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-produced dilp6, including temporal expression pattern, growth promotion, effects on adult stress resistance and lifespan. Additionally dilp1 may play a role in regulation of nutrient utilization/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 fuel/nutrient store [54] and also contribute to egg development [64]. Curiously, 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 signaling pathways downstream the dInR 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 [37,54]. Also there seems to be a difference in the interactions with AKH signaling. During pupal development we did not see any effect of dilp1 on transcripts of Akh or tobi, whereas in adult flies Akh expression is induced by dilp1 [34]. This is in agreement with earlier work, which showed that AKH plays no role in development or lipid and carbohydrate metabolism in the pupa [51]. In the future it would be interesting to investigate whether DILP1 acts differently on larval/pupal and adult fat body, or act on different downstream signaling in the two stages of the life cycle, and whether dilp1 and dilp6 interact to regulate growth and metabolism in Drosophila.
Experimental procedures
Fly lines and husbandry
Parental flies were reared and maintained at 18°C with 12:12 Light:Dark cycle on food based on a 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 [[65] from M.J. Pankratz, Bonn, Germany], To-Gal4 [[66] from B. Dauwalder, Houston, TX], c929-Gal4 [[46] from Paul H. Taghert], yw; UAS-dilp6, and yw; UAS-dilp2;+ [[23] from H. Stocker, Zürich, Switzerland]. Several UAS-dilp1 lines were produced for a previous study [34] and two of them, UAS-dilp1 (II) and UAS-dilp1 (III), were used here. UAS-dilp1-RNAi flies were from Vienna Drosophila Resource Center (VDRC), Vienna, Austria. 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. [34]. Also single dilp1 and dilp2 null mutants were employed. We refer to these three null mutants as dilp1, dilp2 and dilp1/dilp2 mutants for simplicity. As described earlier [34], 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 generate a recombinant dilp6;;dilp1 double mutant, the dilp1 and dilp668 mutants [10] were used for crossing with a double balancer fly, 4E10D/FM7,dfd;;Vno/TM3,dfd, obtained from Dr. Vasilios Tsarouhas (Stockholm University). The efficiency of the dilp6;;dilp1 double mutant was validated by qPCR.
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 [67] and part of the C-peptide of DILP5 [68] 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 [69], rabbit antiserum to cockroach leucokinin I (LK I) at 1:2000 [70], 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, goat anti-rabbit Alexa 488, and goat anti-mouse Alexa 488 (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 projections of z-stacks were 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 [71]. 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). The number of eggs of stage 10-14 in ovaries was counted in 3-day-old flies.
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 dead. 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. [72]. 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.
Quantitative real-time PCR (qPCR)
Total RNA was extracted from whole bodies of middle or late pupal stages of Drosophila by using Trizol-chloroform (Sigma-Aldrich). Quality and concentration of the RNA were determined with a NanoDrop 2000 spectrophotometer (Thermo Scientific). The concentration of the RNA was adjusted to 400 ng/µl. Totally 2ug RNA was used for cDNA synthesis. The cDNA syntheses were performed by using random hexamer primer (Thermo Scientific) and RevertAid reverse transcriptase (Thermo Scientific). The cDNA products were then diluted 10 times and applied for qPCR using a StepOnePlus™ instrument (Applied Biosystem, USA) and SensiFAST SYBR Hi-ROX Kit (Bioline) followed the protocol from the manufacturer. The mRNA abundance was normalized to ribosomal protein (rp49) levels in the same samples. Relative expression values were determined by the 2-ΔΔCT method [73]. The sequences of primers used for qPCR were those used previously [16,34,74]:
dilp1 F: CGGAAACCACAAACTCTGCG
dilp1 R :CCCAGCAAGCTTTCACGTTT
dilp2 F: AGCAAGCCTTTGTCCTTCATCTC
dilp2 R: ACACCATACTCAGCACCTCGTTG;
dilp6 F: CCCTTGGCGATGTATTTCCCAACA
dilp6 R: CCGACTTGCAGCACAAATCGGTTA
akh F: GCGAAGTCCTCATTGCAGCCGT
akh R: CCAATCCGGCGAGAAGGTCAATTGA
tobi F: CCACCAAGCGAGACATTTACC
tobi R: GAGCGGCGTAGTCCATCAC
bmm F: GGT CCC TTC AGT CCC TCC TT
bmm R: GCT TGT GAG CAT CGT CTG GT
rp49 F: ATCGGTTACGGATCGAACAA
rp49 R: GACAATCTCCTTGCGCTTCT
Metabolite quantification
Glycogen and triacyl glyceride (TAG) levels were assayed as previously described [34,75,76]. 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.
To measure the amount of TAG during late pupal stages, 6 replicates with 4 pupae in each were collected and then homogenized in PBS + 0.05% Triton-X 100 with a tissuelyser II from Qiagen. The TAG levels was determined with a Liquick Cor-TG diagnostic kit (Cormay, Poland) using a linear regression coefficient from a standard curve made with 2.2 µg/µl TAG standard (Cormay, Poland). Absorbance of samples was measured at 550 nm with a micro-plate reader (Thermo scientific). Data are expressed as micrograms of TAG related to protein levels. Protein levels were determined using a Bradford assay according to Diop et al. [77].
Dynamic injection respirometry
Carbon dioxide (CO2) production and oxygen (O2) consumption of individual pupae of both sexes were measured during pupal development at 25°C to assess metabolic rate (MR) as described previously [49]. Pupae were placed in 1 ml syringes (i.e. respirometry chambers) that were filled with air scrubbed of CO2 with ascarite (Acros Organics, USA) that then passed through filtered acidified water (pH < 4.5, checked weekly), closed with three-way luer valves, and kept for roughly 24 hours at 25°C with 12:12 Light:Dark cycle. An empty syringe served as control. CO2 production was measured using a Sable Systems (Las Vegas, NV, USA) differential respirometry setup. Two independent lines of outdoor air scrubbed of H2O and CO2, using drierite (WA Hammond Drierite, USA) and ascarite scrubbers respectively, were pushed at a steady rate of 150 ml min-1 using a SS-4 pump (Sable Systems) and two separate mass flow controllers (840 Series; Sierra Instruments Inc, California, USA). The syringes containing pupae were placed after the mass valve controllers in the first line (sample) and 0.45 ml pushed into the airflow. The push rate was recorded through a second flowmeter downstream of the syringe and approximated a flow rate of 162 ml min-1 downstream of the syringe. The line was then scrubbed of H2O with drierite and entered the sample line of a Li-7000 CO2 analyser (LiCor, Lincoln, NE, USA). The second line (reference) proceeded the same way, mimicking the exact length of the sample line (including an empty measurement chamber), entering the reference line of the CO2 analyser. The lines then proceeded through a second set of ascarite CO2 scrubbers and entered an Oxzilla FC-2 O2 analyser (Sable Systems) after which air was ejected. Preliminary measurements were performed to ensure stability of flow rate through either channel by measuring the flow rate of air ejected from the O2 analyser. After the measurement pupae were weighed using a Mettler Toledo MT5 microbalance (Columbus, USA) and left at 25°C with 12:12 Light:Dark cycle until adult eclosion, at which point they were sexed.
Differential CO2 and O2 were calculated by subtracting the output of the reference line from the output of the sample line. For all measurements sampling rate was 1 Hz. In the program Expedata (version 1.9.10) the raw output was baseline corrected against the reference line value, fractioned and multiplied with flow rate to yield CO2 and O2 in ml min-1 [78]. The values were then corrected by subtracting the readings from the empty control syringe from the sample values. MR was calculated by first integrating the fractioned CO2 and O2 (ml min-1) values against time to yield CO2 and O2 in ml produced while pupae were in the syringes. Next VCO2 and VO2 were corrected by accounting for the fraction of air that was still left in the syringe and the time spent in the syringe using the formula (only calculation for VCO2 is shown) VCO2 = (CO2 × (0.6 / 0.45)) / hours in syringe (Lighton, 2008). Then the respiratory quotient (RQ) was calculated as RQ = VCO2 / VO2. RQ values provide an estimate on what energy source is being catabolized to fuel metabolism [48]. MR (in Watts = Joules s-1) was converted from VO2 using the formula MR = (VO2 × (16 + (5.164 × RQ))) / (60 × 60) (Lighton 2008) and finally divided by body weight in mg to yield MR mg-1.
In the present study we monitored single identified individuals throughout pupal development, and sexed them after eclosion. For the vast majority eclosion was successful and therefore we could use the true weight of the individual for the calculation above. However, for individuals that failed to eclose properly we instead used the average weight for that sex and treatment to calculate MR.
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.
For the respirometry data we used the natural logarithm of MR mg-1 due to deviations from normality. A factorial two-way ANOVA was used with MR mg-1 or RQ as dependent variable, and sex and treatment as factorial explanatory variables. Non-significant interactions and main effects were removed from final models [79]. The respirometry data were analyzed with the IBM SPSS statistics 23.0 (IBM SPSS Inc., Chicago, IL, USA) statistical software package. Prism GraphPad version 6.00 (La Jolla, CA, USA) was used for generating all the graphs.
Supplemental material figures
S1 Fig. Evaluation of mutant efficiency. A. qPCR reveals that in stage P8-9 pupae the dilp1 and dilp1/dilp2 mutants display dilp1 levels that are close to zero, whereas in the dilp6 mutant dilp1 is upregulated and in dilp2 mutant slightly reduced. B. In the dilp2 and dilp1/dilp2 mutants dilp2 levels are not detectable. C. The dilp6 levels are only affected in the dilp6 mutants. Data are presented as means ± S.E.M, n = 6 replicates for each genotype with 6 pupae in each replicate. (*p < 0.05, compared with w1118 flies, unpaired Students’ t-test). D. Using immunocytochemistry with antisera to DILP1-3 it can be shown that labeling of IPCs in 1-week-old flies is not detectable for anti-DILP1 in dilp1 and double mutants and for DILP2 in dilp2 and double mutants. DILP3 is upregulated in dilp2 mutants. E-G. Quantification of immunofluorescence shows that DILP1 labeling is not affected in dilp2 mutants (E), DILP2 is increased in dilp1 mutants (F) and DILP3 strongly increased only in dilp2 mutants (G). Data are presented as means ± S.E.M, n = 9-12 flies from 3 replicates. (**p < 0.01, ***p < 0.001, compared with w1118 flies, unpaired Students’ t-test).
S2 Fig. Recombinant dilp1/dilp6 mutant flies display reduced body mass. A. Transcripts of dilp1 and dilp6 in one-day-old dilp1/dilp6 mutant flies. Data are presented as means ± S.E.M, n = 3 replicates for each genotype with 6 pupae in each replicate. (**p < 0.001, ***p < 0.0001, compared with w1118 flies, unpaired Students’ t-test). B. Body weights are significantly reduced in single mutants and recombinant double mutants, but no additive effect of the double mutation was detected. (n = 11–16 flies per genotype from three replicates, One-way ANOVA followed by Tukey’s test).
S3 Fig. Verification of ectopic dilp1 expression by DILP1 immunolabeling. A. After dilp2-Gal4-driven dilp1 expression strong DILP1 immunolabeling can be detected in IPCs of 3rd instar larvae as well as 1 and 3 week old adults, but not in controls (dilp2>w1118). B. Quantification of DILP1 immunofluorescence in IPCs of one-week-old adults, using two different UAS-dilp1 (2 and 3). Data are presented as means ± S.E.M, n = 5-7 flies from 3 replicates. (***p < 0.001, compared with control flies, unpaired Students’ t-test). C. Using the c929 driver DILP1 immunolabeling can be detected in numerous neuroendocrine cells in the CNS of larvae and brain of adults, but not in controls (c929>w1118). D. Using two different fat body Gal4 drivers (ppl and to) DILP1 immunolabeling can be detected in adipocytes.
S4 Fig. Verification of ectopic dilp1 expression by qPCR in stage P8-9 pupae. A. Using the fat body Gal4 drivers ppl and to a drastic increase of dilp1 transcript was seen. B. The dilp2 level was diminished after ppl-driven dilp1. C. No significant effect was seen on dilp6 levels after dilp1 expression. D-F. Driving dilp1 in IPCs with dilp2-Gal4 drastically increases dilp1, but has no effect on dilp2 or dilp6. Data are presented as means ± S.E.M, n = 5-6 replicates per genotype with 10 pupae in each replicate. (*p < 0.05, **p < 0.01, ***p < 0.01, compared with w1118 flies, unpaired Students’ t-test).
S5 Fig. Effects of ectopic dilp1 expression on peptide levels of DILPs in one-week-old adults. A. Expressing dilp1 in IPCs (dilp2>dilp1) increases DILP2 immunolabeling and decreases DILP3. B and C. Quantification of immunolabeling. Data are presented as means ± S.E.M, n = 7-10 per genotype from 3 replicates. (**p < 0.01, compared with w1118 flies, unpaired Students’ t-test). D. Using the broader c929-Gal4 to drive dilp1 the DILP5 immunolabeling of IPCs increase. E. Quantification of DILP5 immunolabeling. Data are presented as means ± S.E.M, n = 9-12 from 3 replicates. (**p < 0.01, compared with w1118 flies, unpaired Students’ t-test).
S6 Fig. Effects of ectopic dilp1 expression on body weight and organismal size. A. Expressing dilp1 in the fat body (ppl-Gal4) of male flies leads to increased weight compared to controls in both young and older flies. However, in contrast to female flies, shown in Fig. 2K, there is no gain in weight over the first 5-6 days as adults, rather a decrease. Data are presented as medians ± range, n = 14–25 flies from three independent replicates (*p < 0.05, **p < 0.01, ***p < 0.001, two-way ANOVA followed with Tukey’s test). B. Driving dilp1 in IPCs with dilp2-Gal4 in females increases the weight compared to controls in older flies. Data are presented as medians ± range, n = 14–23 flies from three independent replicates (*p < 0.05, **p < 0.01, two-way ANOVA followed with Tukey’s test). C. Driving dilp1 in IPCs increases the weight of one-day-old and 6-7 day old male flies, compared to both controls. Furthermore the younger flies weigh more than the older ones for all genotypes. Data are presented as medians ± range, n = 14–24 flies per genotype from three independent replicates (**p < 0.01, ***p < 0.001, two-way ANOVA followed with Tukey’s test). D and E. Using to-Gal4 the body masses show the same patterns as with ppl-Gal4 (Fig. 2K and S5A Fig), where body masses increase after dilp1 over expression, and in females there is an additional weight gain over the first 5-6 days. The following days (13-14 d) no additional increase is seen. Data are presented as medians ± range, n = 9–27 flies per genotype from three independent replicates (*p < 0.05, **p < 0.01, ***p < 0.001, two-way ANOVA followed with Tukey’s test). F-H. The dilp1 expression obtained with the to-Gal4 does not result in a significant increase in wing area, (n = 16–22 flies per genotype from three replicates, One-way ANOVA followed with Tukey’s test), whereas thorax length increased slightly (n = 19-32 flies per genotype (*p < 0.05 unpaired Students’ t-test), but no effect on pupal volume (n = 29 flies per genotype from three replicates, One-way ANOVA followed with Tukey’s test).
S7 Fig. Effects of dilp1 expression on weight, wing area and food intake. A The body weight increased in male flies after ectopic dilp1 expression with c929-Gal4, *** p<0.001, data are presented as means ± S.E.M, n = 16–29 flies per genotype from three independent replicates (One-way ANOVA followed with Tukey’s test). B. The wing area is not affected by c929-driven dilp1 expression. Data are presented as means ± S.E.M, n = 15 flies from three independent replicates (One-way ANOVA followed with Tukey’s test). C. Driving dilp1 with dilp2-and c929-Gal4 does not affect food intake. Data are presented as means ± S.E.M, n = 24 flies from three independent replicates (two-way ANOVA followed with Tukey’s test). dent replicates, as assessed by log-rank (Mantel–Cox) test.
S8 Fig. The brain and neuronal cell bodies grow after dilp1 overexpression in neuroendocrine cells. A-C. Using the c929-Gal4 line to drive dilp1 in neuroendocrine cells leads to increased size of the cell bodies of DILP2 immunolabeled insulin producing cells (A1-A3), PDF labeled l-LNv clock neurons (B1-B3) and abdominal leucokinin (LK) immunoreactive neurons, ABLK (C1-C3). D. The entire brain also increases in size in c929>dilp1 flies. E. Expression of dilp1 in IPCs with the dilp2-Gal4 line is not sufficient to obtain an increase in size of IPCs. F1-F3. Expression of dilp1 in the fat body (ppl-Gal4) increases the size of the l-LNv clock neurons and the entire brain (G). Data are presented as means ± S.E.M, n = 8–10 samples for each genotype from three independent replicates (*p < 0.05, **p < 0.01, ***p < 0.001, as assessed by unpaired Students’ t-test).
S9 Fig. Ectopic expression of dilp1 in clock neurons or larval neuroendocrine cells does not affect cell size. A-D. Expression of dilp1 with the clock neuron driver pdf-Gal4 does not affect the size of the PDF-immunolabeled large LNvs quantified in B. The brain size is also not affected (C). However the PDF immunolabeling is strongly increased (D). Data are presented as means ± S.E.M, n = 8 for each genotype from 3 replicates. (**p < 0.01, compared with w1118 flies, unpaired Students’ t-test). E. Ectopic expression of dilp1 with the c929-Gal4 line does not affect the size of leucokinin (LK)-immunolabeled neuronal cell bodies in the third instar larvae (quantified in F) or the size of the larval CNS (G). Data are presented as means ± S.E.M, n = 6-9 for each genotype from 3 replicates.
S10 Fig. Ectopic expression of dilp1 in glial cells with repo-Gal4 does not affect growth of neuronal cell bodies. A. DILP1 immunolabeling appears in cells after Repo>dilp1, but has no effect on the size of l-LNv clock neurons labeled with anti-PDF (quantified in B). Data are presented as means ± S.E.M, n = 9 for each genotype from 3 replicates.
S11 Fig. Overexpression and mutation of dilps have little effect on AKH signaling as determined by qPCR. A -J. Transcripts of AKH (Akh), brummer lipase (bmm) and the glucosidase target of brain insulin (tobi) were measured by qPCR in different genotypes at two stages of pupal development: two day old and 4 day old. For overexpression in fat body we used ppl-and to-Gal4 drivers. We analyzed 3 replicates with 6 pupae in each replicates (*p < 0.05, one-way ANOVA followed by Tukey’s test). K. Glucose levels were determined in 4 d pupae; 6 replicates per genotype with 4 pupae in each replicate (No significant differences; analysis by one-way ANOVA followed by Tukey’s test). This panel is associated with Fig. 3D – F.
S12 Fig. Effect on starvation and desiccation in male dilp mutant flies. A. In 6-7 days old male flies dilp1-dilp2 mutants are least resistant to starvation (p<0.001), followed by dilp2 mutants (p<0.001), whereas dilp1 mutants perform as controls; n = 125-141 flies from three independent replicates. However 6-7 d female flies perform as 3 d virgin females (see [34] and Table 1). B. In males double mutants are less (p<0.001), and the other two mutants more resistant (p<0.001) to desiccation than controls, n = 134-135 flies from three independent replicates. Data are presented in survival curves and the error bars means S.E.M, as assessed by log-rank (Mantel–Cox) test].
S13 Fig. Targeted dilp1-RNAi in IPCs reduces survival in flies exposed to starvation. A. The efficiency of dilp2>dilp1-RNAi on dilp1 levels was monitored by qPCR. A strong reduction in dilp1 was noted, but no effect was seen on levels of dilp2 or dilp6. Data are presented as means ± S.E.M, n = 3 replicates per genotype with 10 pupae in each replicate. (*p < 0.05, compared with control flies, unpaired Students’ t-test). B. In newly eclosed female flies dilp2>dilp1-RNAi resulted in reduced survival during starvation. n = 148-170 flies from three independent replicates. Data are presented in survival curves and the error bars means S.E.M [***p < 0.001, as assessed by log-rank (Mantel–Cox) test]. C. In 6-7 d old males dilp1 overexpression in fat body (ppl-Gal4) has no effect on starvation response. n = 117-128 flies from three independent replicates. E. c929-driven dilp1 does not affect the response to starvation, n = 132-135 flies per genotype from three independent replicates.
Acknowledgements
We thank the Bloomington Drosophila Stock Center (Bloomington, IN), the Vienna Drosophila Resource Center (Vienna, Austria) and Drs. M.J. Pankratz, B. Dauwalder, V. Monnier, P.H. Taghert, V. Tsarouhas and H. Stocker for flies, and Dr. M. R. Brown for AKH antiserum.
Footnotes
dnassel{at}zoologi.su.se, sifang.liao{at}zoologi.su.se, stephanieepost{at}gmail.com, philipp.lehmann{at}zoologi.su.se, jan-adrianus.veenstra{at}ubordeaux.fr, marc_tatar{at}brown.edu
We performed respirometry mesaurements in pupae to determine metabolic rate and respiratory quotient. Thus, we found that dilp1 and dilp6 increase metabolic rate in the late pupa and promote lipids as the primary source of catabolic energy. This was confirmed by determination of triacyl glyceride levels levels in late pupae. A new figure (Fig 3) was added.