A phenotypic Caenorhabditis elegans screen identifies a selective suppressor of antipsychotic-induced hyperphagia

Abstract

Antipsychotic (AP) drugs are used to treat psychiatric disorders but are associated with significant weight gain and metabolic disease. Increased food intake (hyperphagia) appears to be a driving force by which APs induce weight gain but the mechanisms are poorly understood. Here we report that administration of APs to C. elegans induces hyperphagia by a mechanism that is genetically distinct from basal food intake. We exploit this finding to screen for adjuvant drugs that suppress AP-induced hyperphagia in C. elegans and mice. In mice AP-induced hyperphagia is associated with a unique hypothalamic gene expression signature that is abrogated by adjuvant drug treatment. Genetic analysis of this signature using C. elegans identifies two transcription factors, nhr-25/Nr5a2 and nfyb-1/NFYB to be required for AP-induced hyperphagia. Our study reveals that AP-induced hyperphagia can be selectively suppressed without affecting basal food intake allowing for novel drug discovery strategies to combat AP-induced metabolic side effects.

Fig. 1

AP administration results in hyperphagic response in C. elegans. a Schematic diagram of C. elegans based food intake assay. b EC₅₀s of effect of various APs on food intake in C. elegans. c Food intake response of various C. elegans strains with single mutations or d triple/quadruple mutations in dopamine and serotonin signaling pathways compared with N2 (wild type) worms. e Comparison of food intake response in serotonin receptor mutants (ser-1, ser-4, ser-5, and ser-7) after either AP administration or serotonin administration. f Food intake response in a Crispr/Cas9 knock out strain ser-5(vq1) or N2 worms after treatment with various APs (olanzapine, ziprasidone, quetiapine) or serotonin as a hyperphagic control stimulus. g Food intake response in a Crispr/Cas9 knock out strain ser-7(vq2) or N2 worms after treatment with olanzapine or serotonin as a hyperphagic control stimulus. h Food intake response of ser-5(vq1); ser-7(vq2) double knockout or N2 worms after treatment olanzapine or serotonin as a hyperphagic control stimulus. Error bars in dose response plots (b, d, f, g, h) represent +/− S.E.M of 14-21 replicate wells with 5–15 animals each. c and e are Tukey style plots with the box representing the upper and lower quartile. Asterisks (*) represent p < 0.05 (ANOVA, Holm-Sidak corrected), n.s., not significant. Data (c–f) expressed as mean fold change in food intake relative to food intake of untreated N2 (wild-type) C. elegans assayed in parallel. For number of animals and P values see Supplementary Data 2. Source data are provided as a Source Data file

Fig. 2

C. elegans based screen of FDA-approved drugs for potential adjuvant to block AP-induced hyperphagia. a Diagram of screening process. b Food intake based screen of 192 compounds in C. elegans in the presence of 50 µM of the AP loxapine (See Supplementary Data 1 and 2 for details). c Tukey style plot representing effects on food intake of various antibiotics alone (33 µM, black) or in combination with the AP loxapine (50 µM, red). d Structural comparison of primary hits identified from C. elegans based food intake screen. e Effect of primary hits on basal food intake in the absence of AP. f Effect of primary hits on chlorpromazine-induced hyperphagia. g Effect of primary hits on olanzapine-induced hyperphagia. Error bars in e–g represent +/− S.E.M. Food intake was measured using the bacterial clearance assay, feeding dead, γ-irradiated bacteria. Data expressed as mean fold change in food intake relative to untreated N2 (wild-type) after 72 h. Asterisks (*) represent p < 0.05 (ANOVA, Bonferroni corrected), n.s. not significant. Source data are provided as a Source Data file

Fig. 3

Co-administration of minocycline in mice prevents olanzapine-induced hyperphagia and weight gain. a Food intake and (b) Body weight after 1 week of treatment. c Chronic administration of minocycline (MINO) prevents olanzapine-induced (OLZ) weight gain. d, e EchoMRI measurements of fat mass (d) and lean mass (e). f–h Metabolic chamber analysis of food intake (f), and oxygen consumption (VO_2, g) and carbon dioxide production (VCO₂, h). i Plots locomotion as a function of time before and after amphetamine administration (black arrow). Amphetamine induced a significant increase in locomotor activity that was attenuated by OLZ treatment. Co-treatment with MINO did not affect the ability of OLZ to suppress amphetamine-induced locomotor activity. For a–f, data are presented as mean per group +/− S.E.M. a–e. n = 10 per group, f–h. n = 4 per group in metabolic chambers, i n = 12 per group. a–b, d–e *p < 0.05, One-way ANOVA with uncorrected Fishers LSD test. c, f–h, i *p < 0.05 indicates significance using Two-way repeated measures ANOVA with Tukey multiple comparisons test. i Shaded background indicates 95% confidence interval or each cohort. Source data are provided as a Source Data file

Fig. 4

RNA sequencing identifies gene expression profile associated with AP-hyperphagia. a RNA-seq of hypothalamus of mice identifies unique signature of genes associated with AP-induced hyperphagia, n = 4 mice per group, control (CTRL), olanzapine (OLZ), and olanzapine + minocycline (OLZ + MINO). b Comparison of AP-induced profile genes in model of low fat diet (LFD) and high fat diet (HFD) fed mice (n = 4 per group). C. Testing AP-induced hyperphagia in various orthologous C. elegans mutants of genes identified from mouse hypothalamic RNA seq. d, e Comparison of food intake response of nhr-25 mutant (d) and (e) nfyb-1 mutant strains, to olanzapine induced hyperphagia and serotonin-induced hyperphagia compared with N2 (wild type control worms). Error bars in dose response plots (d, e) represent mean and +/− S.E.M of 14–21 replicate wells with 5–15 animals each. Data in c are represented by a tukey style plots with the box representing the upper and lower quartile. ANOVA, Holm-Sidak corrected. For number of animals see Supplementary Data 2