Adipokinetic hormone signaling determines dietary fatty acid preference through maintenance of hemolymph fatty acid composition in the cricket Gryllus bimaculatus

Adipokinetic hormone (AKH), an analog of mammalian glucagon, functions in supplying the required energy by releasing lipids and carbohydrates from the fat body into the hemolymph. Our previous study showed that AKH receptor (AKHR) knockdown in the two-spotted cricket Gryllus bimaculatus decreased hemolymph lipid levels and increased its feeding frequency. To reveal underlying mechanisms by which AKH signaling modulates lipid homeostasis, we analyzed the fatty acid composition as the lipid structure in the crickets. AKH administration significantly increased the proportion of unsaturated fatty acids (USFAs) to total fatty acids with decrease of the saturated fatty acids (SFAs) in hemolymph, while these proportions were inversely changed in RNA interference-mediated AKHR-knockdowned (AKHRRNAi) crickets. Interestingly, knockdown of hormone-sensitive lipase (Hsl) by RNAi (HslRNAi) affected the proportion of USFAs and SFAs in a similar manner to that observed in AKHRRNAi crickets. AKH administration in HslRNAi crickets did not change hemolymph fatty acid composition, indicating that AKH signaling critically altered fatty acid composition in the hemolymph through Hsl. In addition, a choice assay revealed that AKHRRNAi significantly increases the preference of USFAs. These data indicate that hemolymph lipid level and composition were modulated by AKH signaling with a complementary feeding behavior toward USFAs.

carbohydrates and lipids has been observed in various insects, including the yellow fever mosquito Aedes aegypti, the blood-sucking insect Rhodnius prolixus, and the cockroach Blattella germanica [18][19][20] . The phenotypic changes induced by AKHR manipulation at the transcriptional level have highlighted the importance of AKH signaling in altering lipid and carbohydrate levels accompanied by regulating feeding behavior of insects.
Recent studies have demonstrated modification of the fatty acid composition in response to changes in the physiological conditions of several insect species 21,22 . Changes in the fatty acid composition of the codling moth Cydia pomonella reinforce maintenance of cellular membrane structures and stored lipid levels, even under unfavorable conditions such as low temperature 23 . Therefore, lipid composition can be a crucial factor for responding to changes in environmental conditions. The lipid composition in tissues or the hemolymph is influenced by the state of lipid metabolism. In particular, lipid composition is linked to the degree of flux in lipid usage and synthesis and the corresponding status of enzymatic activities in lipogenesis and lipolysis. In addition, the structural characteristics of lipids in the carbon chain length and the saturation degree of fatty acids indicate a temporal influence of lipid metabolism on the nutrient state. Therefore, AKH signaling must modulate lipid biosynthesis and metabolism, eventually causing changes in the fatty acid composition, according to the nutritional state following feeding behavior.
Along with alterations in the hemolymph lipid levels under different physiological and biological conditions, the previous study has demonstrated that the lipid biochemical composition in the fat body changes during adult development in the cricket 24 . Furthermore, AKH signaling has been shown to affect the fatty acid composition of the fat body in the tenebrionid beetle Zophobas atratus 25 . Taken together, these findings suggest that AKH signaling is involved in both qualitative and quantitative aspects of lipid homeostasis in insects. However, the relationship between AKH signaling and feeding behavior and the relationship between lipid metabolism and feeding behavior in insects remains unclear.
In the present study, we explored the contribution of AKH signaling in maintaining the hemolymph fatty acid composition to better understanding of the underlying regulatory mechanisms of insect feeding behavior. We also demonstrate that lipid composition changes are caused by differences in the transcriptional levels of lipogenic and lipolytic enzymes, which in turn is regulated by AKH signaling. In addition, we demonstrate that hemolymph lipid changes resulted in the modification of preferential feeding behaviors to compensate for the imbalanced lipid component.

Effects of AKH injection on lipid mobilization.
We first confirmed the effects of AKH injection on adult crickets, which triggered a significant increase in hemolymph lipid level compared with those of no treatment, while the lipid level of PBS-injected crickets did not alter (Fig. 1A). We also examined the effects of AKH injection on lipid mobilization in cricket nymphs to exclude the influence of reproduction. Injection of AKH into fifth instar cricket nymphs increased hemolymph lipid level by 26.1% compared with lipid level of non-treated crickets, while the lipid level of PBS-injected crickets did not alter (Fig. 1B), consistent with the data using adult crickets (Fig. 1A). Hereafter, we thus used the cricket nymphs for all studies.

Effects of AKH injection on fatty acid composition.
To investigate these changes in more detail, we analyzed the hemolymph lipid composition of DAG-and TAG-derived fatty acids. In the present analyses of fatty acid composition using HPLC, we analyzed most abundant seven fatty acids, which were observed in the hemolymph and fat body of cricket nymphs; lauric acid (12:0), myristic acid (14:0), palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), linoleic acid (18:2), and linolenic acid (18:3). AKH injection decreased the proportions of palmitic (16:0) and stearic acids (18:0) to total fatty acids, whereas the proportions of oleic acid (18:1) was significantly increased in the hemolymph (Fig. 1C). In contrast, fatty acids derived from the fat body in the AKH-injected crickets showed negligible changes in their composition (Fig. 1E). Incidentally, AKH injection significantly decreased the proportion of saturated fatty acids (SFAs; 16:0 and 18:0) to the total fatty acids and increased the proportion of unsaturated fatty acids (USFAs; 18:1, 18:2, and 18:3) in the hemolymph, whereas these proportions in the fat body were unchanged ( Fig. 1D and F).
Effects of AKHR knockdown on fatty acid composition. Because the fatty acid composition in the hemolymph was altered by AKH injection, we further analyzed the fatty acid composition by disrupting AKH signaling using cricket after treatment of AKHR RNAi (AKHR RNAi ). The effects of AKHR knockdown were confirmed by qRT-PCR ( Fig. 2A). Similar to the previous observation 17 , hemolymph lipid levels were significantly decreased in AKHR RNAi crickets, whereas the control crickets (DsRed2 RNAi ) did not alter the levels (Fig. 2B). We then analyzed the fatty acid composition in the hemolymph and fat body of AKHR RNAi crickets. In the hemolymph of AKHR RNAi crickets, the proportion of stearic acid (18:0) to total fatty acids was significantly increased, while the proportions of oleic (18:1) and linoleic acid (18:2) were decreased (Fig. 2C). In contrast, in the fat body, the proportion of oleic acid (18:1) was significantly increased (Fig. 2E). Totally, knockdown of AKHR significantly increased the proportion of SFAs to total fatty acids and decreased that of USFAs in the hemolymph (Fig. 2D), whereas the opposing changes in those proportions were observed in the fat body (Fig. 2F).
Effects of AKHR knockdown on lipid metabolism-related genes. As observed above, manipulating AKH signaling drastically changed the fatty acid composition in the hemolymph and fat body. To determine whether the lipid biosynthetic pathway in the fat body mediates AKH signaling-induced changes in hemolymph fatty acid composition, we analyzed the transcriptional levels of genes associated with lipid metabolism in the fat body of AKHR RNAi crickets. We measured the transcriptional levels of three genes encoding enzymes which critically contribute to the fatty acid composition: fatty acid synthase (Fas), stearoyl-CoA 9 desaturase (SCD), and elongation of very long chain fatty acids protein 6 (Elovl6). In addition, we also measured the transcriptional levels of two lipases involved in fatty acid mobilization from DAGs; hormone-sensitive lipase (Hsl) and brummer (bmm). qRT-PCR showed that the transcriptional levels of Fas, SCD, and Elovl6 were significantly increased in AKHR RNAi crickets compared with those in control dsRNA-treated crickets ( Fig. 3A-C). In contrast, the transcriptional levels of Hsl and bmm were decreased in AKHR RNAi crickets compared with those in the control crickets ( Fig. 3D and E). These results showed that AKH signaling modulated the flux of lipids by controlling both lipid biosynthesis in the fat body and lipid mobilization from the fat body into the hemolymph at the transcriptional level.

Involvement of Hsl in the AKH-mediated fatty acid flux. Because AKH injection and AKHR knock-
down demonstrated that AKH signaling influenced the composition of hemolymph fatty acids, we next investigated the relationship between lipase activity and the hemolymph fatty acid composition. Similar to mammalian lipid metabolism in which glucagon activates Hsl activity, AKH regulates Hsl function in D. melanogaster 26,27 . It appears that AKH likewise activates cricket Hsl. We therefore analyzed the fatty acid composition in the hemolymph of crickets following Hsl knockdown (Hsl RNAi ). The proportion of USFAs was significantly decreased Figure 1. Effects of AKH injection on lipid levels and lipid composition in crickets. Hemolymph lipid levels in AKH-injected adult crickets (A) and fifth instar cricket nymphs (B). The lipid levels after AKH injection were compared with those of non-treated crickets. Also, the lipid levels by AKH injection were compared with those of PBS-injected crickets after 90 min. Fatty acid composition in the hemolymph of fifth instar cricket nymphs (C) and the proportions of SFAs and USFAs to total fatty acids (D). Fatty acid composition in the fat body of fifth instar cricket nymphs (E) and the proportions of SFAs and USFAs to total fatty acids in the hemolymph (F). Values are expressed as means ± SD (A; n = 5, B; n = 5-6, C; n = 5, D; n = 5, E; n = 7, F; n = 7, *P < 0.05 for Tukey's post hoc test).
in the hemolymph of Hsl RNAi crickets accompanied by an increase in the proportion of SFAs to total fatty acids, while little difference was observed in the hemolymph proportions of SFAs and USFAs after injection of AKH into Hsl RNAi crickets compared with those in DsRed2 RNAi crickets ( Fig. 4A and B). These results indicated that Hsl was involved in the maintenance of the fatty acid flux via AKH signaling, possibly with Hsl substrate preference for USFAs as reported previously 28 .
Effects of AKHR knockdown on the feeding preference for fatty acids. Since AKHR RNAi reduced the proportion of hemolymph USFAs, we examined the selective feeding behavior for dietary lipid contents in AKHR RNAi crickets using a two-choice assay with two diet blocks containing different fatty acids: an USFA-included diet (USFA diet) and a SFA-included diet (SFA diet). Time-lapse recording of the cricket movement showed that AKHR RNAi crickets exhibited increased feeding frequency on USFA diet compared with that by DsRed2 RNAi crickets (Fig. 5A). Furthermore, the amount of intake from the SFA diet in the AKHR RNAi crickets was indistinguishable from that of control crickets, whereas the amount of USFA diet intake was significantly increased in AKHR RNAi crickets ( Fig. 5B and C). Also, overall food intake by AKHR RNAi crickets was increased (Fig. 5D) as previous report 17 . These results indicated that AKHR RNAi changed the feeding preference for fatty acids, which apparently complement the imbalanced hemolymph lipid component.

Discussion
The originally characterized function of AKH is mobilization of lipids from storage in the fat body for large energy demands such as long flight during migration 1,2 . Therefore, most studies on AKH regulation have been performed using adult insects such as migratory locusts 6 . The involvement of AKH in lipid metabolism not only for acute energy demands but also in the function of nutritional homeostasis has been highlighted recently, and has shown AKH to be an insect analog of mammalian glucagon 26,29 . These findings indicate that AKH is crucial for maintaining lipid and carbohydrate levels in the insect body throughout its life cycle. In the present study, we confirmed that the effects of AKH on lipid mobilization in pre-adult G. bimaculatus are comparable to those reported for adult crickets 17,30 . Our results using nymphal crickets also indicated that AKH can contribute to lipid homeostasis in the absence of adult specific events such as reproduction required for additional energy demands (Figs 1B, 2A and B).
The current study illustrates that fatty acid composition is correlated with lipid biosynthesis and metabolism, which is influenced by nutritional conditions in the crickets. Indeed, previous studies using several insect species have demonstrated that fatty acid composition is regulated by physiological and biochemical factors, including dietary conditions and surrounding temperature [21][22][23] . Therefore, the composition of fatty acids is a consequence of the metabolic and nutritional states. Unfortunately, except for a recent report using the beetle Zophobas atratus 25 , few studies have reported whether the metabolic regulation involved in AKH signaling can control fatty acid composition. In the present study, AKH stimulation increased the proportion of USFAs composed of oleic acid (18:1), linoleic acid (18:2) and linolenic acid (18:3) in the hemolymph of cricket nymphs (Fig. 1C and D). This finding is consistent with that reported in the beetle study 25 . In contrast, the proportion of USFAs to total fatty acids decreased in the hemolymph of AKHR RNAi crickets ( Fig. 2C and D). These opposing data might be due to differences in the nutritional states. In the AKH-injected crickets, fatty acids in the fat body that respond to AKH stimulation for acute energy demands were mobilized even under normal conditions. In contrast, AKHR RNAi crickets were in a state of chronic AKH signaling deficiency; thus it was difficult for those insects to maintain basal hemolymph lipid levels and may have induced changes in the mechanisms underlying lipid homeostasis maintenance. Data in Fig. 2 implied that hemolymph USFAs were apparently replaced by SFAs. In fact, allocation of more USFAs after AKH stimulation via hemolymph to the organs such as muscle might have physiologically benefits to produce ATP easier via β-oxidation as an energy source to supply energy demand for a long flight. This speculation is consistent with the fact that USFAs is rapidly catalyzed by β-oxidation compared with SFAs 31 .
We also demonstrated the role of a lipase in the fatty acid composition changes observed in the hemolymph. It has been reported so far that insects have at least two lipolytic systems that are mediated by AKH signaling and non-AKH signaling, which includes bmm activity 10,32 . In D. melanogaster, they function differently based on the nutritional state, including the trans-state as they shift from hungry to starved 10 . Based on the different composition of fatty acids in the hemolymph after AKH injection, it can be concluded that AKH injection activated Hsl in the fat body. Hsl activation might cause a biased-preference in the enzymatic substrate, which would eventually change the fatty acid composition in the hemolymph. This is also consistent with previous reports showing that Hsl in the rat Rattus norvegicus and the Antarctic fish Trematomus newnesi has characteristic catabolism in substrate specificity 33,34 . This speculation agrees with our data showing an increase in the proportion of SFAs with the simultaneous reduction of USFAs in the hemolymph of Hsl-knockdown crickets (Fig. 4). Moreover, Figure 5. Effects of RNAi-mediated AKHR knockdown on the feeding preference. Representative feeding behaviors for 9 h (A). Colors represent the duration of time a cricket spent within a particular area with thicker red indicating longer periods of time. The amount of food intake from the USFA or SFA diets (B). PI for the USFA diet (C). Total food intake of SFA and USFA diets (D). Values are expressed as means ± SD (B,C, and D; n = 9, **P < 0.01 and *P < 0.05 for Student's t-test).
SCieNtifiC REPORTS | (2018) 8:4737 | DOI:10.1038/s41598-018-22987-2 AKH injection in Hsl RNAi crickets did not affect the proportions of SFAs and USFAs to total fatty acids in the hemolymph, indicating that AKH signaling influences the hemolymph fatty acid composition via Hsl activity in the fat body.
In AKHR RNAi crickets, significant changes were observed in the transcriptional levels of lipogenic and lipolytic genes (Fig. 3), indicating that the decreased level of hemolymph lipids and accumulation of lipid level in the fat body are involved in AKH signaling 17,30 . In the present study, AKH-injected crickets showed an acute lipid mobilization response within 90 minutes, indicating that such acute transcriptional changes in the lipogenic and lipolytic genes likely have little role in affecting changes in the nutritional state of these crickets. In contrast, AKHR RNAi crickets showed a chronic deficiency in AKH signaling, indicating that the transcriptional levels of the lipogenic and lipolytic genes in the fat body of these nymphs were altered. These data revealed that AKH plays roles in both acute and chronic lipid mobilization, although the transcriptional contribution to lipogenesis and lipolysis are different.
Similar to other animals, insects need to discriminate and select the appropriate nutrition from their diet to complete various life events including development and reproduction. In D. melanogaster, the feeding preference for dietary fatty acids was shifted according to nutritional demands during development 35 . Because hemolymph components, such as lipids, are the critical factors that reflect the individual nutritional states, feeding behaviors to seek the appropriate nutrition must be affected by the nutritional state of the hemolymph as observed in Locust migratoria 36 . In this study, AKHR RNAi decreased the proportion of USFAs in the hemolymph and increased the amount of food intake of a USFA diet (Figs 2D and 5). This result indicated that AKHR RNAi crickets discriminated the different chemical structures of fatty acids and selectively ingested the USFA diet to compensate for the nutritional loss in the hemolymph. However, it remains unknown whether increased feeding motivation for USFAs is regulated by AKH signaling directly at a molecular level.
In summary, we demonstrated that AKH signaling mobilizes lipid from the fat body with changes of fatty acid composition in the hemolymph of cricket nymphs. The mobilized lipids by AKH stimulation might be composed of more USFAs such as oleic acid (18:1) in the hemolymph. In the case of activated AKH signaling, those changed composition of fatty acids is possibly mediated by regulation of lipid metabolic gene expression and Hsl activation. In addition, AKH signaling can modulate the feeding preference for SFAs or USFAs directly or indirectly, possibly according to the required nutrients in various biological processes. Understanding these behavioral changes will help further elucidate the mechanisms linking lipid homeostasis and insect feeding behavior.
Insects. Crickets were reared in plastic containers (55 × 39 × 31 cm) at 28 ± 1 °C under long-day lighting conditions (16-h light/8-h dark cycle). Fifth instar cricket nymphs had free access to water and were fed ad libitum a standard diet of rabbit food ORC4 (Oriental Yeast, Tokyo, Japan) and cat food at a 3:1 ratio.

Preparation and injection of AKH.
Gryllus bimaculatus AKH (p-QVNFSTGW-NH 2 ) was chemically synthesized as described previously 17,30 . AKH was injected into the hemolymph through the abdomen with 20 pmol AKH in 5 μl of Phosphate-buffered saline (PBS) or 5 μl of PBS injected into each cricket. To measure hemolymph lipids, 5 μl of hemolymph was collected from crickets prior to injection (0 min) and 90 min after injection.
Lipid extraction from the hemolymph and fat body. Lipids were extracted as described by Lorenz with minor modification 17,30,37 . Briefly, 5 μl of hemolymph or the fat body surrounding the testis were collected in centrifuge tubes containing 20 mg of sodium sulfate and 200 μl of 75% methanol in water. Next, the hemolymph and fat body were homogenized in 600 μl chloroform: methanol (1:1) by sonication and centrifuged at 15,000 × g at 4 °C for 10 min. The supernatant was then transferred to a new tube, vortexed, and centrifuged at 15,000 × g at 4 °C for 10 min after adding 500 μl of chloroform and 300 μl of 1 M NaCl. The organic layer containing lipids was dried under vacuum and centrifugation. The resulting lipid fraction was then used for lipid quantification.

Measurement of lipid.
The lipids extracted from the hemolymph and fat body were quantified using the previously reported sulfo-phospho-vanillin method 17,30,38 . The extracted lipid fractions in chloroform/methanol (1 μl) were mixed with 50 μl of sulfuric acid and heated at 100 °C for 10 min. After cooling the solution to room temperature, 500 μl of vanillin reagent (0.2% vanillin in 67% ortho-phosphoric acid) was added. The resulting samples were measured at 540 nm in a spectrophotometer. Cholesterol (Sigma-Aldrich Japan, Tokyo, Japan) was used as a standard. The lipids from the hemolymph and the fat body were quantified as DAG and TAG, respectively.
Fatty acid composition analysis by high-performance liquid chromatography. An HPLC (Jasco SC-802, PU-880, UV-875; Jasco Inc., Tokyo, Japan) equipped with a fluorescence detector (excitation at 365 nm, emission at 412 nm; RF-20A; Shimadzu Co., Kyoto, Japan) was used for analyzing the fatty acid composition. HPLC was performed using a reversed-phase column (PEGASIL C 8 SP100, 4.6φ × 250 mm; Senshu Scientific Co., Tokyo, Japan) at 40 °C in an oven (865-CO; JASCO, Tokyo, Japan). After saponification of the lipid fraction, fatty acids were identified and quantified by comparing retention times with those of commercially available standard fatty acids. The mobile phase was acetonitrile: water (80:20 [v/v]) at a flow rate of 1.5 ml/min. Fluorescent 9-anthryldiazomethane (ADAM) ester derivatives of fatty acids were prepared as described previously 39  were incubated in the dark with 50 μl of 0.1% ADAM in acetone: methanol (1:9) for 1 h at room temperature. The mixture was then subjected to HPLC.
Two choice assay. To analyze the dietary preference, we prepared two nutritionally different diets as described by Tsukamoto et al. with minor modification 41 . To normalize the dietary quantity, indigestible cellulose was mixed with two different nutrients (USFAs or SFAs), which were isocaloric (4 kj/g). All examined diets contained 33% casein and 33% dextrin for total energy. The USFA diet containing 15% oleic acid (18:1) and 15% linoleic acid (18:2), whereas the SFA diet consisted of 15% palmitic (16:0) acid and 15% stearic acid (18:0). The feeding frequency was represented by a heat map generated from 9 h continuous recorded video data as described by Fukumura et al. 42 . The Preference Index (PI) was defined by calculating the amount of USFA diet intake/the total amount of USFA and SFA diet intake × 100 (%) from 2 days post-dsRNA treatment for 3 days. The amount of food intake was determined as the dry weight by weighing the diet tablets before and after consumption. The diet tablets were dried by baking at 80 °C for 2 hours as described previously 41 . Statistical analyses. Statistical comparison of two groups was performed using Student's t-test. Multiple comparisons were analyzed using Tukey's test. P-values less than 0.05 were considered to be statistically significant.