Olfactory specificity regulates lipid metabolism through neuroendocrine signaling in Caenorhabditis elegans

Abstract

Olfactory and metabolic dysfunctions are intertwined phenomena associated with obesity and neurodegenerative diseases; yet how mechanistically olfaction regulates metabolic homeostasis remains unclear. Specificity of olfactory perception integrates diverse environmental odors and olfactory neurons expressing different receptors. Here, we report that specific but not all olfactory neurons actively regulate fat metabolism without affecting eating behaviors in Caenorhabditis elegans, and identified specific odors that reduce fat mobilization via inhibiting these neurons. Optogenetic activation or inhibition of the responsible olfactory neural circuit promotes the loss or gain of fat storage, respectively. Furthermore, we discovered that FLP-1 neuropeptide released from this olfactory neural circuit signals through peripheral NPR-4/neuropeptide receptor, SGK-1/serum- and glucocorticoid-inducible kinase, and specific isoforms of DAF-16/FOXO transcription factor to regulate fat storage. Our work reveals molecular mechanisms underlying olfactory regulation of fat metabolism, and suggests the association between olfactory perception specificity of each individual and his/her susceptibility to the development of obesity.

Fig. 1. Neuronal guanylate cyclase regulates peripheral fat mobilization.

a SRS microscopy images and quantification of fat content levels in the anterior intestine of 3-day-old adults show increased fat storage in the guanylate cyclase daf-11 mutants, when compared to wild-type worms (WT). Data are mean ± s.e.m., **P < 0.01 by one-way ANOVA with Dunnett’s test, scale bar = 40 μm, dashed lines indicate quantified intestine areas, yellow pixels indicate higher SRS signals. b Triglyceride levels measured using colorimetric enzyme assays are increased in the daf-11 mutants, when compared to WT. Data are mean ± s.d. of five biological replicates with ~5000 worms for each genotype in each replicate, *P < 0.05, ***P < 0.001 by one-way ANOVA with Dunnett’s test. c–e daf-11 mutants do not show any significant changes in their locomotion (c), or in feeding (d) and defecation rates (e) when compared to WT. Statistical analysis with one-way ANOVA with Sidak’s multiple comparison test. The boxes span the interquartile range, median is marked by the line and whiskers indicate the minimum and the maximum measurements. f Labeling/chasing assays with deuterium-labeled oleic acid (OA-D₃₄) determine rates of lipid synthesis and catabolism. Signals derived from deuterium-labeled lipids in intestinal lipid droplets were quantified at indicated time points using SRS microscopy. daf-11 mutants have no changes in the rate of lipid synthesis, but have a significant reduction in the rate of lipid catabolism (P < 0.01 by linear regression analysis). Data are mean ± s.d, dashed lines represent the trendlines and their corresponding equations are indicated. SRS intensity curve represents replicate #1 (Replicate #2 is presented in Supplementary Fig. 2c). g Reduction in oxygen consumption rate is measured after addition of etomoxir, that blocks mitochondrial β-oxidation, using Seahorse extracellular flux analyzer. daf-11 mutants have reduced levels of mitochondrial β-oxidation compared to wild-type worms. Data are mean ± s.e.m. of three independent biological replicates with at least ten microplate wells containing about 25 worms, *P < 0.05 by Student’s two-tailed t-test. For a and c–f numbers of animals used are listed in Supplementary Data 1. For a–g source data are provided as a Source Data file.

Fig. 2. Guanylate cyclase acts in olfactory neurons to regulate fat metabolism.

a The [daf-11::GFP] integrated transgenic array shows the expression of daf-11 in seven pairs of head neurons: ASK, ADL, ASI, AWB, ASH, ASJ, and AWC. Scale bar = 10 μm. b Mosaic analysis using the [daf-11::RFP] extrachromosomal transgenic array discovers that daf-11 expression in AWC olfactory neurons sufficiently rescues the increased fat storage in the daf-11 mutants. Individual worms that are mosaic for the transgenic array expression, are numbered 1–12 in the bar chart and in the table presenting the expression pattern. Error bars represent s.d. Individual worm images are in Supplementary Fig. 1b. c Using the ceh-36 promoter, specific restoration of daf-11 only in AWC neurons is sufficient to suppress the increased fat storage in the daf-11 mutants. Representative SRS images are in Supplementary Fig. 6a. d Using the ssu-1 promoter, restoration of daf-11 expression specifically in ASJ neurons does not suppress the fat storage increase in the daf-11 mutants. Representative SRS images are in Supplementary Fig. 6b. For c and d, data are mean ± s.d., ***P < 0.001, n.s. not significant by two-way ANOVA with Sidak’s multiple comparison test. For c–d, numbers of animals used are listed in Supplementary Data 1. For b–d, source data are provided as a Source Data file.

Fig. 3. Specific odors and responding olfactory neural circuit regulate fat metabolism.

a Genetic ablation of AWC neurons increases fat storage, while the loss of the asymmetric AWC^OFF (nsy-1 = AWC^ON/ON) or AWC^ON (nsy-5 = AWC^OFF/OFF) neuron leads to reciprocal changes, decreasing or increasing fat levels, respectively. Representative SRS images are in Supplementary Fig. 6c. b Scheme of the odorants and their responding neural circuit with asymmetric AWC olfactory neurons and AIY interneurons. c Exposing WT to 2-butanone that inhibits AWC^ON neurons is sufficient to increase fat levels. Other odors that target AWC^OFF, both AWC, or AWB neurons, do not change fat storage. Worms are exposed to odorants in the absence of food for 4 h. BUT 2-butanone, PENT 2,3-pentanedione, BENZ benzaldehyde, IAA isoamyl alcohol, OCT 1-octanol, NON 2-nonanone. Representative SRS images are in Supplementary Fig. 6d. d Exposure of WT to 2-butanone increases fat storage within 4 h. Upon odor removal, the fat levels are restored back to normal within 4 h. e Lack of AWC (AWC abl.) or AIY (ttx-3 mutation) suppresses the increased fat storage conferred by 2-butanone. EtOH, negative control. Representative SRS images are in Supplementary Fig. 6e. f ChR2-mediated optogenetic activation of AWC^ON is sufficient to decrease fat storage levels, but optogenetic stimulation of AWC^OFF does not affect fat levels. 1 h blue light stimulation and 1 h in dark in the absence of food followed by SRS imaging. Representative SRS images are in Supplementary Fig. 6f. g Lack of AIY suppresses the increased fat storage in the nsy-5 mutants. Hyperactivity of AIY (glc-3 mutation) recapitulates the fat storage increase in the nsy-5 mutants. Representative SRS images are in Supplementary Fig. 6g. h bPAC-mediated optogenetic activation of AIY is sufficient to increase fat storage. Representative SRS images are in Supplementary Fig. 6h. For a and c data are mean ± s.d., *P < 0.05, **P < 0.01, ***P < 0.001, n.s. not significant by one-way ANOVA with Dunnett’s test and for d–h by two-way ANOVA with Sidak’s test. Numbers of animals used are listed in Supplementary Data 1. For a and c–h, source data are provided as a Source Data file.

Fig. 4. Specific FOXO isoforms mediate the olfactory regulation of fat metabolism.

a Inactivation of the nuclear hormone receptor DAF-12 does not suppress the increased fat storage in the daf-11 mutants with only AWC^OFF neurons. Representative SRS images are in Supplementary Fig. 6i. b The daf-2 mutation leads to fat storage increase, and further enhances the fat storage increase in the daf-11 mutants. Representative SRS images are in Supplementary Fig. 6j. c DAF-16 is required for the increased fat storage in the daf-11 and nsy-5 mutants, that lack AWC^ON neurons, and in the glc-3 mutants with activated AIY neurons. Representative SRS images are in Supplementary Fig. 7a. d The daf-16 deletion fully suppresses the increased fat storage in the daf-11 mutants, which can be rescued by restoration of the daf-16d/f, but not the daf-16a or daf-16b isoform. Intestine-only but not neuron-only restoration of the daf-16d/f isoform sufficiently rescues the fat storage increase in the daf-11; daf-16 double mutants. Representative SRS images are in Supplementary Fig. 7b. For a, b, c data are mean ± s.d., **P < 0.01, ***P < 0.001, n.s. not significant by two-way ANOVA with Sidak’s multiple comparison test. For d, data are mean ± s.d., ***P < 0.001, n.s. not significant by one-way ANOVA with Tukey’s multiple comparison test. Numbers of animals used are listed in Supplementary Data 1. For a–d source data are provided as a Source Data file.

Fig. 5. SGK-1 kinase regulates fat metabolism in response to olfactory signaling.

a Neither DAF-16A nor DAF-16D/F nuclear localization is strongly enhanced in the daf-11 mutants, as it does in the daf-2 mutants. Scale bar = 40 μm, yellow dashed lines circle nuclei of the intestinal cells. b Inactivation of SGK-1 fully suppresses the increased fat storage in the daf-11 mutants. Intestine-specific restoration of sgk-1 rescues the fat storage increase in the daf-11; sgk-1 double mutants. Representative SRS images are in Supplementary Fig. 7c. c, d Neither the JNK homolog jnk-1 (c), nor the AMPK homolog aak-2 (d) affects the fat storage level in WT or in the daf-11 mutants. Representative SRS images are in Supplementary Fig. 7d, e. e Knockdown of sgk-1 in the daf-11 and nsy-5 mutants, that lack AWC^ON neurons, or in the glc-3 mutants with active AIY neurons suppresses the increased fat storage. Representative SRS images are in Supplementary Fig. 7a. For b–e data are mean ± s.d., ***P < 0.001, n.s. not significant by two-way ANOVA with Sidak’s multiple comparison test. Numbers of animals used are listed in Supplementary Data 1. For b–e source data are provided as a Source Data file.

Fig. 6. Neuropeptide signals link olfactory neural circuit and fat metabolism.

a Inactivation of either neuropeptide processing enzyme egl-21–carboxypeptidase E homolog, or neuropeptide secretion factor unc-31–CAPS homolog suppresses the increased fat storage in the daf-11 mutants. Representative SRS images are in Supplementary Fig. 6k. b Among the neuropeptides that are secreted from AIY interneurons, deletion of flp-1, not flp-18, suppresses the increased fat storage in the daf-11 mutants. Representative SRS images are in Supplementary Fig. 6l. c AIY-specific overexpression of flp-1 increases fat storage under WT conditions. Restoration of flp-1 expression only in AIY neurons is sufficient to rescue fat storage increase in the daf-11; flp-1 double mutants. Representative SRS images are in Supplementary Fig. 6m. d Mutation of either flp-1 neuropeptide or its putative receptor npr-4 fully suppresses the fat storage increase conferred by the 2-butanone odor exposure. Representative SRS images are in Supplementary Fig. 6n. e Scheme of FLP-1 neuropeptides released from AIY interneurons acting through their putative receptor, NPR-4 that cell-autonomously regulates fat levels in the lipid storage tissue. f Inactivation of the npr-4 neuropeptide receptor suppresses the fat storage increase in the daf-11 mutants. Restoration of npr-4 expression specifically in the intestine is sufficient to increase the fat storage in the daf-11; npr-4 double mutants. Representative SRS images are in Supplementary Fig. 6o. g NPR-4 is required for the increased fat storage in the daf-11 and the nsy-5 mutants, that lack AWC^ON neurons, and in the glc-3 mutants with active AIY neurons. Representative SRS images are in Supplementary Fig. 7a. For a and b, data are mean ± s.d., **P < 0.01, ***P < 0.001, n.s. not significant by one-way ANOVA with Tukey’s multiple comparison test. For c, d, f, g data are mean ± s.d., ***P < 0.001, n.s. not significant by two-way ANOVA with Sidak’s multiple comparison test. Numbers of animals used are listed in Supplementary Data 1. For a–d, f, g source data are provided as a Source Data file.

Fig. 7. Olfaction regulates lipid metabolism through neuroendocrine signaling mechanisms.

Between two asymmetric AWC olfactory neurons, only AWC^ON regulates peripheral lipid metabolism. In response to specific environmental odors, this olfactory neuron signals through a specific downstream neural circuit and neuroendocrine signaling pathway (conserved components in mammals are indicated) to directly control lipid homeostasis in peripheral metabolic tissues. DCV, dense core vesicles.