Postnatal Dynamic Ciliary ARL13B and ADCY3 Localization in the Mouse Brain

Abstract

Primary cilia are hair-like structures found on nearly all mammalian cell types, including cells in the developing and adult brain. A diverse set of receptors and signaling proteins localize within cilia to regulate many physiological and developmental pathways, including the Hedgehog (Hh) pathway. Defects in cilia structure, protein localization, and function lead to genetic disorders called ciliopathies, which present with various clinical features that include several neurodevelopmental phenotypes and hyperphagia-associated obesity. Despite their dysfunction being implicated in several disease states, understanding their roles in central nervous system (CNS) development and signaling has proven challenging. We hypothesize that dynamic changes to ciliary protein composition contribute to this challenge and may reflect unrecognized diversity of CNS cilia. The proteins ARL13B and ADCY3 are established markers of cilia in the brain. ARL13B is a regulatory GTPase important for regulating cilia structure, protein trafficking, and Hh signaling, and ADCY3 is a ciliary adenylyl cyclase. Here, we examine the ciliary localization of ARL13B and ADCY3 in the perinatal and adult mouse brain. We define changes in the proportion of cilia enriched for ARL13B and ADCY3 depending on brain region and age. Furthermore, we identify distinct lengths of cilia within specific brain regions of male and female mice. ARL13B+ cilia become relatively rare with age in many brain regions, including the hypothalamic feeding centers, while ADCY3 becomes a prominent cilia marker in the mature adult brain. It is important to understand the endogenous localization patterns of these proteins throughout development and under different physiological conditions as these common cilia markers may be more dynamic than initially expected. Understanding regional- and developmental-associated cilia protein composition signatures and physiological condition cilia dynamic changes in the CNS may reveal the molecular mechanisms associated with the features commonly observed in ciliopathy models and ciliopathies, like obesity and diabetes.

Keywords: ADCY3; ARL13B; brain; cilia; energy homeostasis; hypothalamus; nucleus accumbens.

Figure 1

AI identification and analysis of ciliary ARL13B and ADCY3. Cilia identification and training: To identify and assess cilia localization of ARL13B and ADCY3 in a robust, non-biased manner, Nikon Elements Analysis (NIS Elements) was used to recognize proteins on sample images. Hand-drawn binaries were created on the sample set of data and then ran overnight using Segment.ai. Here, the computer uses half of the training set to teach itself the characteristics of the identified cilia and the other half to test itself on accuracy. The training loss calculated at the end of our run was found to be 0.01. White arrow indicates an example of a selected cilia binary. Experimental analysis: Once the training set is established (goal of training loss to be <0.05), experimental image sets can be run through NIS Elements using this trained Segment.ai to identify both ARL13B and ADCY3 cilia. Experimental z-stacks are made into sum projection images and converted to microns based on image capture settings. Separate cilia binaries are identified for each cilia marker on their respective channels. To enhance accuracy, only cilia binaries in contact with an FOP basal body marker are selected for final analysis. General analysis scheme (GA3) and final results: Using the finalized cilia binaries, a GA3 is constructed to measure specific cilia characteristics, such as frequency and length. Separate GA3s were constructed to assess the cilia characteristics of single binaries and overlapping cilia binaries. All data are exported and assessed into Excel and GraphPad Prism for further analysis.

Figure 2

Assessing CNS cilia with an ARL13B fluorescent allele. (A) Schematic: Cilia and basal body. Visualization: Fluorescent transgenic alleles. Their localization to cilia membrane-associated ARL13B-mCherry and basal body-associated Centrin2-GFP. Immunofluorescence. Proteins ADCY3 and ARL13B enriched along the cilia membrane and FOP to the basal body analyzed throughout. Distribution: Cilia with ARL13B alone (ARL13B+), with ADCY3 alone (ADCY3+), or with both ARL13B and ADCY3 (COLO). Cilia lengths assessed in two ways for COLO cilia. ARL13B measured length (ARL13B in COLO) or ADCY3 measured length (ADCY3 in COLO). (B) Immunofluorescence for ARL13B in the suprachiasmatic nucleus (SCN) of 8-week-old C57BL/6J control and Arl13b-mCherry;Centrin2-EGFP mice. Cilia markers are indicated with ADCY3 (green), ARL13B antibody labeled cilia (red), and basal body marker FOP (white) in C57BL/6J control animals and Arl13b-mcherry and Centrin2-GFP animals. Boxes indicate examples of cilia with different membrane compositions: ARL13B+ (red), ADCY3+ (green), and COLO (orange) boxes, respectively. Scale bar 10 µm. Hoechst-stained nuclei blue. (C) Brain regions assessed. (D) Analysis of cilia lengths between C57BL/6J control (open circles) and Arl13b-mCherry;Centrin2-EGFP (closed circles) brain regions. Significant length differences are indicated * p < 0.05, ** p < 0.01, *** p < 0.001, using nested t-test. N = 10 animals (6 C57BL/6J control and 4 Arl13b-mCherry;Centrin2-EGFP). (E) Colorimetric RNAscope in situ hybridization in the SCN of 8-week-old animals for Arl13b (red) in C57Bl/6J (control) and Arl13b-mCherry;Centrin2-EGFP animals. Black boxes indicate region for inset images. Counter-stain hematoxylin (purple). Scale bar 60 µm. N = 4 animals (2 C57BL/6J control and 2 Arl13b-mCherry;Centrin2-EGFP). All animals are 8-week-old adults.

Figure 3

Assessing CNS cilia with an antibody approach. (A) Immunofluorescence for mCherry (green) and ARL13B (red) in the suprachiasmatic nucleus (SCN) of adult Arl13b-mCherry;Centrin2-EGFP mice. Centrin2-EGFP basal body marker in white. (B) Cilia distribution in the ARC, PVN, SCN, NAc, and NAs of Arl13b-mCherry;Centrin2-EGFP animals with mCherry cilia (mCherry ARL13B+, green) and ARL13B antibody labeled (ARL13B Ab+, red) and cilia localizing both (COLO). N = 4 to 8 Arl13b-mCherry;Centrin2-EGFP animals per region. (C) Immunofluorescence for ARL13B (red) and cilia marker ADCY3 (green) in the SCN of mice homozygous for the cilia restricted Arl13b allele (Arl13b^V358A). Basal body staining for FOP in white. (A,C) Scale bars 10 µm. Hoechst-stained nuclei blue. (D) Cilia distribution analysis shows complete absence of ARL13B positive cilia and no changes in ADCY3 positive cilia in Arl13b^V358A homozygous animals in the hypothalamus (ARC, PVN, SCN) and accumbens (NAc, NAs). N = 4 animals (2 Arl13b^+/+ and 2 Arl13b^V358A). All animals are 8-week-old adults.

Figure 4

Region-dependent cilia signatures in the adult mouse brain. (A) Immunofluorescence for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the arcuate nucleus (ARC), paraventricular nucleus (PVN), and suprachiasmatic nucleus (SCN) of the hypothalamus and nucleus accumbens core (NAc) and shell (NAs). Scale bar 10 µm. Hoechst-stained nuclei blue. (B) Cilia distribution analysis of membrane composition: ARL13B+, ADCY3+, and COLO. Two-way ANOVA and Tukey’s multiple comparisons analyses revealed differences among all regions analyzed, ARC, PVN, SCN, NAc, and NAs, in 8-week-old mice. Significant differences are indicated * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. N = 6 animals (3 males and 3 females). All animals are 8-week-old adults.

Figure 5

Region-dependent cilia signatures in perinatal mouse brain. (A) Immunofluorescence for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the arcuate nucleus (ARC), paraventricular nucleus (PVN), and suprachiasmatic nucleus (SCN) of the hypothalamus and nucleus accumbens core (NAc) and shell (NAs) on the day of birth (P0) in males and females. Scale bar 10 µm. Hoechst-stained nuclei blue. (B) Cilia distribution analysis of cilia membrane composition: ARL13B+, ADCY3+, and COLO. Two-way ANOVA and Tukey’s multiple comparisons analyses revealed differences among specific regions in P0 mice. Significant differences are indicated ** p < 0.01, *** p < 0.001, **** p < 0.0001. N = 6 animals (3 males and 3 females). All animals are postnatal day 0 (P0).

Figure 6

Sex differences in cilia length in early postnatal development. Analysis of cilia length across regions at different ages. (A) Immunofluorescence and cilia length analysis for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the arcuate nucleus (ARC) at the day of birth (P0). Scale bar 10 µm. Hoechst-stained nuclei blue. (B) Statistically significant differences in cilia lengths between males and females in the ARL13+ cilia lengths in the P0 PVN and SCN. (C) Immunofluorescence and cilia length analysis for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the arcuate nucleus (ARC) at 3 weeks of age (3 WKS). Scale bar 10 µm. Hoechst-stained nuclei blue. (D) ARL13B in COLO cilia length in 3 WKS NAs and ADCY3 in COLO cilia length in 3 WKS PVN. Nested t-test analysis revealed significant differences, which are indicated ** p< 0.01, *** p< 0.001, **** p< 0.0001. N = 6 animals (3 males and 3 females).

Figure 7

Cilia length is similar between sexes across neuroanatomical regions past 5 WKS of age. (A,B) Analysis of cilia length across regions at 5 and 8 WKS. Immunofluorescence and cilia length analysis for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the arcuate nucleus (ARC) of 5- and 8-week-old animals (5 WKS, 8 WKS). Scale bar 10 µm. Hoechst-stained nuclei blue. Scale bar 10 µm. Nested t-test analysis did not reveal differences between males and females. N = 6 animals (3 males and 3 females).

Figure 8

Age-dependent cilia signatures in the mouse brain. (A) Immunofluorescence and cilia distribution analysis for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the nucleus accumbens core (NAc) at the day of birth, 3 weeks, 5 weeks, and 8 weeks of age (P0, 3 WKS, 5 WKS, and 8 WKS). Scale bar 10 µm. Hoechst-stained nuclei blue. (B) Cilia distribution analysis of ciliary protein enrichment: ARL13B+, ADCY3+, and COLO. Two-way ANOVA and Tukey’s multiple comparisons analyses revealed age differences within a region. Significant differences are indicated * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. N = 6 animals per age (3 males and 3 females). Male and female data were combined for this analysis.

Figure 9

ARL13B ciliary lengths are similar throughout postnatal development. Analysis of ARL13B cilia lengths in each brain region (ARC, PVN, SCN, NAc and NAs) at the day of birth, 3 weeks, 5 weeks, and 8 weeks of age (P0, 3 WKS, 5 WKS, and 8 WKS). No significant differences were observed between male (♂) and female (♀) animals in the ARC and NAc. Datasets were combined when no sex-dependent differences were observed. (A) Cilia with only ARL13B (ARL13B+) showed no difference. (B) Cilia length of the ARL13B mask in cilia with both ARL13B and ADCY3 (ARL13B in COLO). Two-way ANOVA and Tukey’s multiple comparisons analysis revealed significant differences, which are indicated * p < 0.05, ** p < 0.01, **** p < 0.0001. N = 6 (3 males and 3 females).

Figure 10

ADCY3 ciliary lengths are significantly longer after 3 WKS of age in different neuroanatomical regions. Analysis of ADCY3 cilia lengths in each brain region (ARC, PVN, SCN, NAc and NAs) at different ages (P0, 3 WKS, 5 WKS, and 8 WKS). Significant differences between male and female cilia lengths are indicated in separate graphs for the PVN, SCN, and NAs (♂ male, ♀ female). Datasets were combined when no sex-dependent differences were observed. (A) Cilia with only ADCY3 (ADCY3+). (B) Cilia length of the ADCY3 mask in cilia with both ARL13B and ADCY3 (ADCY3 in COLO). Two-way ANOVA and Tukey’s multiple comparisons analyses revealed significant differences, which are indicated * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. N = 6 (3 males and 3 females).

Figure 11

Physiological condition-dependent cilia analysis in the mouse brain. Analysis of cilia under different feeding and body composition conditions, ad libitum fed, high-fat diet-induced obesity (HFD) and pair-fed caloric restricted (Cal Restrict). (A) Weekly body weight measurements in grams (g) beginning at 8 weeks of age. ANOVA analyses revealed significant differences, which are indicated with blue * and red # when * p < 0.05, (B) Immunofluorescence for ADCY3 (green), ARL13B (red), and basal body marker FOP (white) in the paraventricular nucleus (PVN) for each condition. Scale bar 10 µm. Hoechst-stained nuclei blue. (C) Analysis of cilia lengths between conditions. Significant length differences are indicated * p < 0.05 using two-way ANOVA and Tukey’s multiple comparisons. (D) Analysis of cilia distribution: ARL13B+, ADCY3+, and COLO. Two-way ANOVA and Tukey’s multiple comparisons analyses revealed no significant differences among specific regions in ad libitum, Cal Restrict, or HFD. Trends in PVN cilia distribution between ad libitum and HFD are indicated (p = 0.0523 for COLO and p = 0.1 for ARL13B+). N = 4 male animals per physiological condition (4 ad libitum, 4 Cal Restrict, and 4 HFD). All mice went on diet or caloric restriction at 8 weeks of age and were analyzed at 19 weeks of age.