Neuroendocrine regulation of fat metabolism by autophagy gene atg‐18 in C. elegans dauer larvae

In environments with limited food and high population density, Caenorhabditis elegans larvae may enter the dauer stage, in which metabolism is shifted to fat accumulation to allow larvae to survive for months without food. Mutations in the insulin‐like receptor gene daf‐2 force C. elegans to constitutively form dauer larva at higher temperature. It has been reported that autophagy is required for fat accumulation in daf‐2 dauer larva. However, the mechanism underlying this process remains unknown. Here, we report that autophagy gene atg‐18 acts in a cell nonautonomous manner in neurons and intestinal cells to mediate the influence of daf‐2 signaling on fat metabolism. Moreover, ATG‐18 in chemosensory neurons plays a vital role in this metabolic process. Finally, we report that neuronal ATG‐18 functions through neurotransmitters to control fat storage in daf‐2 dauers, which suggests an essential role of autophagy in the neuroendocrine regulation of fat metabolism by insulin‐like signaling.

In environments with limited food and high population density, Caenorhabditis elegans larvae may enter the dauer stage, in which metabolism is shifted to fat accumulation to allow larvae to survive for months without food. Mutations in the insulin-like receptor gene daf-2 force C. elegans to constitutively form dauer larva at higher temperature. It has been reported that autophagy is required for fat accumulation in daf-2 dauer larva. However, the mechanism underlying this process remains unknown. Here, we report that autophagy gene atg-18 acts in a cell nonautonomous manner in neurons and intestinal cells to mediate the influence of daf-2 signaling on fat metabolism. Moreover, ATG-18 in chemosensory neurons plays a vital role in this metabolic process. Finally, we report that neuronal ATG-18 functions through neurotransmitters to control fat storage in daf-2 dauers, which suggests an essential role of autophagy in the neuroendocrine regulation of fat metabolism by insulin-like signaling.
Living organisms accumulate fat as an energy resource to prevent food deprivation. In the presence of food, Caenorhabditis elegans goes through four larval stages and develops to a fertile adult. However, in an environment with limited food and high population density, C. elegans larvae may arrest development during the second molt and enter the dauer stage [1]. The dauer larva is a dispersal stage that is stable for months under adverse environmental conditions and is an example of facultative diapause [1]. A dauer larva is radially shrunken with a constricted intestine, closed buccal cavity, and specialized cuticle morphology and is resistant to detergents such as sodium dodecyl sulfate. Dauer larvae store lipids in intestinal and hypodermal cells and can survive for months without feeding [2].
Molecular studies of dauer mutants have revealed that three functionally overlapping neural pathways, including the insulin-like growth factor (IGF) [3,4], transforming growth factor-b [5,6], and the cyclic guanosine monophosphate [7] pathways, control dauer formation in response to dauer-inducing environmental cues. DAF-2, an insulin/IGF receptor, regulates fat metabolism and dauer morphogenesis by inhibiting the activity of DAF-16, a member of the forkhead family of transcription factors [3,8,9]. The daf-2 pathway also regulates adult lifespan, and mutations in daf-2 increase lifespan [10]. DAF-2 was reported to work in both neurons and intestinal cells, while DAF-16 acts mainly in the intestine to control C. elegans lifespan [11][12][13]. It has been shown that mutations of bec-1, the worm ortholog of autophagy gene atg6, suppress fat accumulation, dauer morphogenesis, and the extended lifespan of daf-2 mutants [14].
Autophagy is an evolutionarily conserved lysosomal degradation pathway that promotes degradation of cytosolic components. Macroautophagy (hereafter referred to as autophagy) shuttles cytosolic components to lysosomes using a double membrane-bound vesicle called an autophagosome [15]. The fusion of the lysosomes and the autophagosome results in an autolysosome, the inside of which contains lysosomal hydrolases that proceed to hydrolyze the shuttled cytosolic components. The new carbohydrates, amino acids, nucleosides, and fatty acids produced by the degradation of cytosolic components can be used by the cell to maintain cellular metabolism [15]. Studies have shown that the inhibition of autophagy leads to decreased lipid accumulation in C. elegans daf-2 mutant dauer larva [14].
Autophagy gene atg-18 encodes a protein that belongs to the WD repeat protein interacting with phosphoinositides protein family [16]. We recently reported that mutations in the atg-18 gene can suppress autophagy induction in daf-2 mutants, and tissue-specific expression of atg-18 can restore autophagy activity in corresponding tissues of daf-2;atg-18 mutants [17]. Here, we examined the tissue-specific requirement of atg-18 for fat accumulation in daf-2 mutant dauer larvae. Our results suggest that autophagy in chemosensory neurons and intestinal cells plays an important role in DAF-2-regulated fat metabolism in C. elegans dauer larva.

Fat staining
Sudan black B staining daf-2(e1370);atg-18(gk378) mutants are lethal at 25 and 20°C. The animals arrest development at egg and L1 larval stages. Therefore, all strains were grown at 15°C. L4 hermaphrodites were picked up and allowed to develop at 15°C for 24 h. The 1-day-old adults were transferred to fresh food plates and allowed to lay eggs at 15°C for 16 h. The adults were removed, and eggs/L1s were shifted to 25°C and incubated for 3 days (72 h). The dauer animals were picked up for staining. At least one hundred dauer larvae for each strain were picked up for Sudan Black B staining. For N2 and atg-18(gk378) animals, L3-stage larvae that were comparable to dauer larvae were used for staining. Collected animals were washed two to three times with M9 buffer. Paraformaldehyde stock solution (10%) was added to a final concentration of 1%. The samples were frozen in dry ice/ethanol and then thawed under a stream of warm water. After a total of three freeze-thaw cycles, the worms were dehydrated through ethanol solutions and then stained with Sudan Black B as described by Kimura et al. [3]. After staining, all animals were examined for fat accumulation. The stained worms were mounted on a 2% agarose pad and observed under a Zeiss upright fluorescence microscope (Axio Imager A2, Zeiss, Oberkochen, Germany). To compare the fat content in different strains, the pictures were taken with the same camera setting using the Zeiss AxioCam ICm1 digital camera at 10009 magnification. The DIC filter and the Zeiss AxioVision 4.8 were used for imaging.

Nile red staining
Nile red staining of fixed worms was performed as described by Pino et al. [26]. Worm samples were collected as described in the Sudan Black staining. Animals were washed twice with M9 buffer. After the final wash, worms were fixed in 40% isopropanol at room temperature for 3 min. The fixed worms were stained in Nile red/isopropanol solution for 30 min at room temperature with gentle rocking. The stained worms were washed once with 1 mL M9 buffer and mounted on a 2% agarose pad for microscopy under the fluorescence channel. To compare the fat content in different strains, the pictures were taken with the same camera settings under 10009 magnification as described in the Sudan Black B staining.

Image quantitation
All quantification was done using FIJI/IMAGEJ [27]. Images were imported as TIFF images, converted to 8-bit, and then run through Fiji's native threshold algorithm to isolate the lipid droplets. The size of the isolated lipid droplets was then quantified by taking an area measurement immediately posterior to the second bulb of the pharynx. GRAPH-PAD PRISM 5 (GraphPad Software, La Jolla, CA, USA) was used to generate column graphs and to perform Student's t-test.
These observations were confirmed by Nile red staining ( Fig. 4F-J). Thus, ATG-18 influences fat metabolism through a neuroendocrine mechanism in daf-2 mutant dauer larva.

Discussion
Caenorhabditis elegans has a conserved insulin-like signaling pathway, and the daf-2 gene encodes the single insulin-like receptor tyrosine kinase. Here, we show that mutations of autophagy gene atg-18 completely block fat accumulation in daf-2 mutant dauer larva, which is consistent with the previous report that inactivation of autophagy gene bec-1 suppresses fat storage in daf-2 dauers [14]. Moreover, our data indicate that ATG-18 acts primarily in neurons and intestinal cells to mediate the influence of DAF-2 signaling on fat metabolism in dauer larvae. By contrast, ATG-18 in hypodermis and body wall muscles plays a minor role. We recently reported that ATG-18 in neurons, intestinal cells, and the hypodermis can fully restore the lifespan of daf-2;atg-18 mutants to daf-2 mutant level, while ATG-18 in body wall muscles only modestly increases the lifespan of daf-2;atg-18 mutants [17]. Thus, although neuronal and intestinal ATG-18 functions similarly in DAF-2-regulated lifespan and fat metabolism, hypodermal ATG-18 has a different role in these two processes. Moreover, ATG-18 in body wall muscle is not essential for both of these two daf-2 mutant phenotypes.
Caenorhabditis elegans utilizes chemosensory neurons to detect environmental cues [29]. We found that expression of the atg-18 gene under the control of gpa-3 and daf-11, but not unc-42, promoters significantly increases fat storage in daf-2;atg-18 mutants. These data indicate that ATG-18 in chemosensory neurons except ASH mediates the influence of DAF-2 signaling on fat metabolism in dauer larvae. Our recent report shows that ATG-18 in ASE, ASI, ASJ, ASK, AWB, and AWC neurons has no statistically significant influence on DAF-2-regulated lifespan extension [17]. Thus, ATG-18 in chemosensory neurons functions differently in regulating fat metabolism in dauer larvae and adult lifespan. Indeed, the longevity phenotype can be uncoupled from fat accumulation in C. elegans [11,12]. We reported previously that ATG-18 in ASG neurons is required for DAF-2-regulated longevity. Interestingly, expression of ATG-18 in only ASG neurons (Podr-2::atg-18) significantly increases fat accumulation in daf-2;atg-18 mutants, which suggests ATG-18 in some chemosensory neurons, such as ASG, can regulate both adult lifespan and fat metabolism in dauer larvae.
Neurons communicate through neurotransmitters. The release of neurotransmitters is blocked by unc-64 mutations [28]. We found that unc-64 mutations have no obvious effect on fat accumulation in daf-2 dauer larvae (Fig. 4). However, unc-64 is epistatic to atg-18, as daf-2unc-64;atg-18 mutants store a significantly higher amount of fat compared to daf-2;atg-18 mutants. The genetic interactions of these genes suggest a model illustrated in Fig. 4K. Essentially, DAF-2 negatively regulates the autophagy process that, in turn, negatively influences the availability of neurotransmitters that suppress fat accumulation. Autophagy could influence biosynthesis of neurotransmitters, package of neurotransmitters into synapse vesicles, and/or release of these chemicals into the synaptic cleft through UNC-64. It has been reported that worms deficient in biosynthesis of serotonin accumulate fat, and exogenous administration of 5-hydroxytryptamine (serotonin) increases fat storage in C. elegans [30,31]. These findings suggest that serotonin could be a candidate neurotransmitter that is regulated by autophagy to influence fat metabolism. Interestingly, ASG neurons, where ATG-18 acts to control fat metabolism, can communicate with other neurons through serotonin [32]. Of note, in the present work, we only examine the tissue-specific role of atg-18 in fat metabolism in daf-2 mutant dauer larvae. Thus, the role of atg-18 in wild-type worms, in adult animals, and in other developmental stages of C. elegans remains to be determined. Nevertheless, our data demonstrate that atg-18 in chemosensory neurons can mediate the influence of insulin-like signaling on fat metabolism in dauer larvae. In mammals, insulin signaling in the central nervous system also controls fat homeostasis. Similar to daf-2 mutants, knockout mice without neuronal insulin receptors are obese [33]. Thus, it is possible that autophagy is downstream of neuronal insulin signaling in controlling fat metabolism in mammals.