A novel form of ciliopathy underlies hyperphagia and obesity in Ankrd26 knockout mice

Abstract

Human ciliopathies are genetic disorders caused by mutations in genes responsible for the formation and function of primary cilia. Some are associated with hyperphagia and obesity (e.g., Bardet-Biedl Syndrome, Alström Syndrome), but the mechanisms underlying these problems are not fully understood. The human gene ANKRD26 is located on 10p12, a locus that is associated with some forms of hereditary obesity. Previously, we reported that disruption of this gene causes hyperphagia, obesity and gigantism in mice. In the present study, we looked for the mechanisms that induce hyperphagia in the Ankrd26-/- mice and found defects in primary cilia in regions of the central nervous system that control appetite and energy homeostasis.

Fig. 1

Schematic demonstration of the hypothalamic regulation of appetite. Peripheral satiety hormones reach POMC and NPY/AGRP neuron groups in the arcuate nucleus (ArcN) of the hypothalamus. Activation of POMC neurons by anorexigenic hormones (e.g., leptin) stimulates the hypothalamic MC4 receptor-expressing cells (e.g., CRH neurons) inhibiting appetite, while orexigenic NPY/AGRP neurons in the arcuate nucleus and the dorsomedial hypothalamus (DHth) are inhibited. In contrast, when orexigenic satiety signals (e.g., ghrelin) activate NPY/AGRP neurons, the dorsomedial hypothalamus is stimulated and the activity of PVN neurons is attenuated. As a result food intake increases. PVN paraventricular nucleus, ME median eminence, 3rd ventr third ventricle, CRH corticotrophin-releasing hormone, MC4R melanocortin 4 receptor, NPYR NPY receptor, MCH melanine-concentrating hormone. (Modified from M. Palkovits)

Fig. 2

Expression of Ankrd26 mRNA in the mouse brain. Inverted X-ray image of in situ hybridization demonstrates the distribution of Ankrd26 mRNA in the brain. Left panel representative in situ hybridization images from different brain areas. Right panel schematic drawings of the corresponding brain areas. Areas with marked Ankrd26 mRNA expression are indicated in the schematic drawings by abbreviations. a Ankrd26 expression in the cortex (c), hippocampal CA2 and CA3 regions (Hi), mediodorsal thalamic nucleus (MD), reticular thalamic nucleus (RT), zona incerta (ZI), and in the hypothalamic PVN and ARC. b Prominent signal detected in the hippocampal CA2 and CA3 regions (Hi), paraventricular thalamic nucleus (PV), medial amygdaloid nucleus and posteromedial cortical amygdaloid nucleus (a), and in the hypothalamic DM, VM and ARC (indicated by asterisk) nuclei. c Midbrain. Raphe dorsalis nucleus (RD) shows high signal. d Medulla oblongata. The nucleus of the solitary tract (NTS) and the cerebellum (CB) has visible radiolabeling. Scale bar 1 mm

Fig. 3

Expression of Ankrd26 in the melanocortin pathway and in the pituitary gland. Dual labeling IHC demonstrating Ankrd26 expression in the key cell populations of the melanocortin pathway. Ankrd26 immunostaining (green) is shown in the upper row (a, e, i, m). a–d In the ARC, LepR-positive cells (b, red) are co-labeled with Ankrd26 (a, green) as shown in the merged image (d). The magnified area in the insets is indicated by dashed boxes. Arrows show double-labeled cells. Scale bar 100 µm. e–h Immunostaining for POMC (f, red) and Ankrd26 (e, green) in the ARC shows many double-labeled cells (h, merged image) (marked by arrows). 409 magnification, scale bar 15 µm. i–l Immunostaining for MC4R (j, red) and Ankrd26 (i, green) in the PVN reveals numerous double-stained cells (l, merged image). The magnified area in the insets is indicated by dashed boxes. Arrows show double-labeled cells. Scale bar 100 µm. m–p In the anterior lobe of the pituitary gland all POMC-positive cells (n, red) are stained for Ankrd26 (p, merged image). Scale bar 100 µm. Inset in o is the schematic drawing of the pituitary gland. Al anterior lobe, il intermediate lobe, pl posterior lobe. In c, g, k and o nuclei are visualized with DAPI (blue). Asterisk in c, g, and k indicates the third ventricle

Fig. 4

Schematic drawing summarizing the physiological regulation of the melanocortin pathway and pathological findings in the in Ankrd26−/− mice. a Under physiological conditions leptin activates STAT3 phosphorylation and POMC is cleaved to α-MSH in ARC POMC neurons. α-MSH activates MC4R-expressing CRH-producing neurons in the PVN. As a result, appetite is inhibited and the stress axis is activated. The stress response is terminated by the negative feedback of CORT. b In Ankrd26−/− mice higher leptin levels are accompanied by STAT3 phosphorylation and increase in hypothalamic α-MSH. In spite of reduced MC4R mRNA in the PVN and elevated serum CORT levels, CRH expression is increased. This is associated with hyperphagia and overactivation of the stress pathway in Ankrd26−/− mice. Note, that neurons with MC4 receptors that make CRH in the PVN are ciliated. In Ankrd26 knockout mice, these PVN cells lack AC3 positive primary cilia (see Fig. 10)

Fig. 5

Characterization of the melanocortin pathway in Ankrd26−/− and WT mice. a– c Graphs show that at the age of 6 weeks, the significantly higher body weight of the Ankrd26−/− (KO) mice (a) is accompanied by significantly elevated serum leptin levels (b) and hypothalamic α-MSH concentration (c) as compared to the WT mice. Data are representative of 2 individual observations (n = 5 in each groups, from 2 different litters). Values are expressed as mean ± SEM. *p < 0.05. d IHC demonstrates nuclear labeling of phospho-STAT3 (red) in the ARC in leptintreated WT and Ankrd26−/− (KO) mice. Scale bar 100 µm. Asterisk indicates the third ventricle. e Effect of α-MSH treatment on the body weight of Ankrd26−/−, Pomc−/− and WT mice. Comparison of weight gain in experimental groups in the pretreatment period and during the α-MSH treatment is shown. The weight gain during the pretreatment and the α-MSH treatment period was significantly less in the WT than in the Ankrd26−/− mice. Pomc−/− mice lost body weight during α-MSH administration. Data are representative of 2 individual observations (n = 5 in each groups, from 2 different litters). For each experiment, values are expressed as mean ± SEM. *p < 0.05

Fig. 6

Expression of primary ciliary markers in the feeding centers in Ankrd26−/− and WT mice. a, b Show immunostaining of AC3 in the PVN of the WT (a) and Ankrd26−/− (b) mice. Note the almost absent immunolabeling in the PVN of the Ankrd26−/− mice. c, d The number of AC3-positive primary cilia is comparable in the ARC in WT (c) and Ankrd26−/− (d) mice. Scale bar 60 µm. Asterisk indicates the third ventricle. e–f IHC of Mchr1 (green) in the PVN in WT and Ankrd26−/− mice. e Neurons of the PVN express Mchr1 in the primary cilia (arrowhead) and in the cytoplasm (arrow). f Ankrd26−/− mice lack Mchr1-positive primary cilia in this region, but the immunostaining is present in the cytoplasm (arrow). Scale bar 10 µm. g–l Expression of Sstr3 (green) and AC3 (red) in the VM nucleus in WT (g–i) and in Ankrd26−/− (j–l) mice. g, j In WT mice (g) Sstr3 is localized to the primary cilia (arrows) and to the cytoplasm, while in the Ankrd26−/− mice (j) the ciliary staining is almost absent. h, k Numerous AC3-positive primary cilia are present in the WT mice (h), while reduced number and shorter AC3-positive primary cilia can be seen in the Ankrd26−/− mice (k). i, l Merged images indicate numerous Sstr3 and AC3 co-labeled primary cilia in the WT mice (i). In Ankrd26−/− mice (l) the number of Sstr3 and AC3-positive primary cilia is markedly reduced. Scale bar 15 µm. Nuclei are visualized with DAPI (blue). Images are representative of four independent observations (n = 3 in each group, four different litters)

Fig. 7

Expression and distribution of BBSome protein Bbs4 in the brain of Ankrd26−/− and WT mice. IHC demonstrates the expression and staining pattern of Bbs4 (red) and AC3 (green) in WT (a–j) and Ankrd26−/− (k–o) mice. The middle panel (f–j) displays representative cells from WT mice (white boxes in a–e). a, f, k represent the cortex, b, g, l the CA3 region of the hippocampus, c, h, m the PVN, d, i, n the VM and e, j, o the DM. Note the punctate staining of Bbs4 in the WT mice, the labeling is often localized to the base of AC3-positive primary cilia, as demonstrated in the middle panels (f–j). Bbs4 immunostaining in Ankrd26−/− mice shows a strong nuclear/perinuclear pattern in the indicated areas, accompanied by a significantly reduced number of AC3-positive primary cilia. Nuclei are visualized with DAPI (blue). Scale bar 15 µm in a and k, 2 µm in f. Images are representative of four independent observations (n = 3 in each group, four different litters)

Fig. 8

Expression of appetite- and stress-related neuropeptides in Ankrd26−/− and WT mice. a–h Representative autoradiographs demonstrate mRNA expression of neuropeptides in WT and Ankrd26−/− (KO) mice at 4 weeks. NPY (a–b) and AgRP (c–d) are expressed in the ARC. CRH (e–f) and AVP (g–h) are expressed in the PVN. Note that in the Ankrd26−/− mice, the number of NPY (b) and AgRP (d) cells and the signal intensity is reduced. CRH− (f) and VP− (h) positive cells in the PVN in the Ankrd26−/− mice. Red areas in the schematic insets indicate the corresponding brain areas. Dark field microscopy images, scale bar 200 µm. i–n Representative autoradiographs demonstrate neuropeptides mRNA expression in WT (upper panel, i and l) and Ankrd26−/− (j, m) mice at 3 months. i–k Images show absence of NPY in the hypothalamic DM in the WT mice (i), while many NPY-positive cells are present here in the Ankrd26−/− mice (j). Arrows indicate NPY-positive cells. mt mammillothalamic tract, f fornix. Scale bar 200 µm. k The box in the schematic drawing demonstrates the corresponding brain area. l–n TRH-expressing neurons are present in the brainstem (raphe obscurus, raphe pallidus and parapyramidal nuclei) in the WT (l) and Ankrd26−/− (m) mice. Note that the number of TRH-positive cells and the intensity of the signal are increased in the Ankrd26−/− (m) mice. Arrows indicate the parapyramidal nucleus; arrowheads indicate the midline raphe nuclei. (n) The box in the schematic drawing demonstrates the corresponding brain area. Py, pyramidal tract. Scale bar 500 µm. Bright field microscopy images

Fig. 9

Characterization of the HPA axis in Ankrd26−/− and WT mice. a, b Serum ACTH and CORT levels of Ankrd26−/− and WT mice. a Upper panel basal levels of serum ACTH in Ankrd26−/− (KO) and WT mice. Lower panel serum ACTH levels following 15 min of restraint stress. b Upper panel basal levels of serum CORT in Ankrd26−/− (KO) and WT mice. Lower panel serum CORT levels after 15 min of restraint stress. Bars represent mean values (n = 10 for each group) ± SEM. *p < 0.05, **p < 0.01. c Effect of adrenalectomy on the body weight of Ankrd26−/− and WT mice. Weight gain of experimental mice from day 14 until the end of the experiment is demonstrated. Note, that the body weight gain of the Ankrd26−/− mice is comparable in each group, while CORT-supplemented (ADX + CORT) and sham-operated WT mice gained significantly more weight compared to the non-replaced WT mice (ADX). *p < 0.05 WT ADX + CORT vs. WT ADX, ‡p < 0.05 WT SHAM vs. WT ADX. Values are expressed as mean ± SEM (n = 7 in each group)

Fig. 10

The expression of AC3 primary ciliary marker in the CRH neurons in the PVN in Ankrd26−/− and WT mice. Triple IHC demonstrates that in WT mice (a–d) CRH-producing (b, green) MC4R-expressing (c, red) cells in the PVN express AC3 (a, yellow). This cell population lacks AC3-positive primary cilia in Ankrd26−/− mice (e–h). Nuclei are visualized with DAPI (blue). 409 magnification; scale bar 15 µm; n = 3 in each group