Abstract
ATXN2 gene, encoding for ataxin-2, is located in a trait locus for obesity. Atxn2 knockout (KO) mice are obese and insulin resistant; however, the cause for this phenotype is still unknown. Moreover, several findings suggest ataxin-2 as a metabolic regulator, but the role of this protein in the hypothalamus was never studied before. The aim of this work was to understand if ataxin-2 modulation in the hypothalamus could play a role in metabolic regulation. Ataxin-2 was overexpressed/re-established in the hypothalamus of C57Bl6/Atxn2 KO mice fed either a chow or a high-fat diet (HFD). This delivery was achieved through stereotaxic injection of lentiviral vectors encoding for ataxin-2. We show, for the first time, that HFD decreases ataxin-2 levels in mouse hypothalamus and liver. Specific hypothalamic ataxin-2 overexpression prevents HFD-induced obesity and insulin resistance. Ataxin-2 re-establishment in Atxn2 KO mice improved metabolic dysfunction without changing body weight. Furthermore, we observed altered clock gene expression in Atxn2 KO that might be causative of metabolic dysfunction. Interestingly, ataxin-2 hypothalamic re-establishment rescued these circadian alterations. Thus, ataxin-2 in the hypothalamus is a determinant for weight, insulin sensitivity and clock gene expression. Ataxin-2’s potential role in the circadian clock, through the regulation of clock genes, might be a relevant mechanism to regulate metabolism. Overall, this work shows hypothalamic ataxin-2 as a new player in metabolism regulation, which might contribute to the development of new strategies for metabolic disorders.
Introduction
The hypothalamus is the key brain region for metabolic homeostasis (Carmo-Silva & Cavadas 2017). Obesity-induced hypothalamic dysfunction, caused by inflammation and other cellular distress mechanisms (endoplasmic reticulum stress, oxidative stress, etc.), compromises the autonomic regulation of peripheral organs (liver, pancreas, white and brown adipose tissue) (Carmo-Silva & Cavadas 2017, Lee et al. 2020). This event, especially the loss of liver autonomic regulation, further potentiates features of metabolic dysfunction, such as insulin resistance (Carmo-Silva & Cavadas 2017, Lee et al. 2020). Furthermore, the hypothalamus also maintains metabolic homeostasis through circadian rhythm (Carmo-Silva & Cavadas 2017, Engin 2017, Lee et al. 2020).
Circadian rhythms are responsible for the body’s 24 h adaptation to external factor changes such as light (day and night). This system regulates, within 24 h of the day, biological functions that range from body temperature to hormone release and even food intake (Albrecht 2012). The circadian system is composed of central and peripheral clocks, the central clock being located within the suprachiasmatic nucleus (SCN) of the hypothalamus and synchronizing peripheral clocks in other organs (Albrecht 2012, Blancas-Velazquez et al. 2017, Engin 2017, Oishi & Hashimoto 2018). Although the central clock is mostly affected by light, the SCN can receive direct input from the arcuate nucleus of the hypothalamus (ARH) (regulates feeding and energetic expenditure) and can be affected by nutrient availability (Albrecht 2012, Blancas-Velazquez et al. 2017, Engin 2017, Oishi & Hashimoto 2018). The peripheral clocks can also be (de)synchronized by feeding and nutritional uptake, independently of the SCN (Albrecht 2012, Blancas-Velazquez et al. 2017, Engin 2017, Oishi & Hashimoto 2018). Studies show that obesity and high-fat diet (HFD) can disrupt the circadian clock (central and peripheral) (Kohsaka et al. 2007, Oishi & Hashimoto 2018), and, on other hand, this disruption can also promote changes in metabolism and predispose to obesity (Blancas-Velazquez et al. 2017, Engin 2017, Hatori et al. 2012).
Beyond hypothalamic dysfunction, the genetic contribution is of great importance for obesity development. Several obesity-related genes, such as FTO (fat mass and obesity associated/FTO alpha-ketoglutarate dependent dioxygenase), were found to be modulated in the hypothalamus upon obesity and starvation (Cheung et al. 2013), functioning as nutritional sensors. These observations led to the discovery of new hypothalamic players that can pave the way for new targets for obesity therapeutics. In this context arises the ATNX2 gene and its product, ataxin-2. ATXN2 gene is located at chromosome 12q24.1 within a well-described obesity trait (ATXN2-SH2B3) (Sanpei et al. 1996, Sahba et al. 1998, Auburger et al. 2014). ATXN2 gene and its product have been studied in the context of the neurodegenerative disorder, spinocerebellar ataxia type 2 (SCA2) (Pulst et al. 1996, Sanpei et al. 1996), but have also been implicated in other diseases, such as amyotrophic lateral sclerosis (Elden et al. 2010, Becker et al. 2017) and Machado Joseph Disease (Nobrega et al. 2015, Ding et al. 2016). Ataxin-2 has a ubiquitous expression and is involved in RNA metabolism, translation regulation, stress granules formation, cytoskeleton organization, cytokinesis, calcium signalling and endocytosis (Carmo-Silva et al. 2017). Evidence point to a possible role for ataxin-2 in metabolic regulation (Auburger et al. 2017, Carmo-Silva et al. 2017). ATXN2 polymorphisms have been associated with hypertension, type 1 diabetes and obesity (Figueroa et al. 2009, Auburger et al. 2014, Kraja et al. 2014, Lv et al. 2017). Furthermore, ATXN2 mutation and associated disease (SCA2) lead to an atrophic phenotype with severe weight loss (Pedroso et al. 2017), a feature recapitulated in SCA2 animal models (Damrath et al. 2012, Sen et al. 2019b). Interestingly, individuals from an Egyptian family with SCA2 exhibited polyphagia and increased weight gain in the mid-stage phase of the disease, a feature lost with disease progression (Abdel-Aleem & Zaki 2008). In contrast, loss of ATXN2 in mice (Atxn2 knockout (KO)) leads to obesity, insulin resistance and dyslipidaemia (Kiehl et al. 2006, Lastres-Becker et al. 2008). Others showed involvement of ataxin-2 and its orthologues, in lipid metabolism (Meierhofer et al. 2016, Sen et al. 2019a) and cellular energetics, through regulation of nutrient-sensing pathways, mTOR and AMPK (reviewed (Carmo-Silva et al. 2017)).
Concerning ataxin-2, the hypothalamus, circadian rhythm and metabolism, nothing is known so far. Ataxin-2 ortholog in Drosophila was shown to regulate circadian rhythm, through translation regulation of the core clock gene PER (Lim & Allada 2013, Lee et al. 2017, Xu et al. 2019). Furthermore, SCA2 patients also present disturbances in REM (rapid eye movement) sleep that might be secondary to altered circadian rhythm (Zanatta et al. 2019).
Even though published evidence point to ataxin-2 as a potential metabolic regulator, its hypothalamic role has never been studied before. Considering this, we hypothesize that ataxin-2 might play a determinant role in hypothalamic-regulated functions. Thus, in the present study, we investigated the role of hypothalamic ataxin-2 on body weight, food intake, insulin sensitivity, neuroinflammation, clock genes expression and other peripheral parameters relevant to metabolic dysfunction in animal models. This work provides new insight into ataxin-2 and hypothalamic physiology and may also contribute to the consideration of hypothalamic ataxin-2 as a new target for metabolic dysfunction and/or obesity therapeutics.
Methods
Mouse studies
For the experiments in this study, male and female C57Bl6 J and Atxn2 KO mice were used. Atxn2 KO mice on a C57BI6x129SvJ mixed background were described previously (Scoles et al. 2012). Mice were housed in pairs, under a 12 h light:12 h darkness cycle in a temperature/humidity-controlled room with ad libitum access to water and food. For the experimental procedures, standard chow diet and rodent diet with 45% kcal% fat (D12451 from Research Diets, New Brunswick, NJ, USA) were used. To recapitulate previous studies with the Atxn2 KO mice strain (Kiehl et al. 2006, Scoles et al. 2012), all the experimental procedures were performed in mice fed with rodent diet with 45% kcal% fat (D12451 from Research Diets).
After stereotaxic injection, mice were monitored for 4 (C57Bl6 experiment) and 8 (Atxn2 KO experiment) weeks. Body weight and food intake were measured twice a week. Detailed descriptions can be found in Supplementary Methods (see section on supplementary materials given at the end of this article).
Experiments were performed in accordance with the European Union Directive 86/609/EEC for the care and use of laboratory animals. Researchers received adequate training (Federation of Laboratory Animal Science Associations certified course) and certification to perform the experiments from the Portuguese authorities (Direção Geral de Veterinária). This study was submitted to the ethics committee that approves and oversees all research projects requiring the use of vertebrate animals, Orgão de Bem-Estar e Ética Animal (ORBEA)/Animal Welfare and Ethics Body of the Faculty of Medicine of the University of Coimbra, where our licensed animal facility is located (International Animal Welfare Assurance number 520.000.000.2006) and permission to carry this study was granted (ORBEA number 165).
Production of viral vectors and stereotaxic injection
The cDNA encoding for GFP (Nobrega et al. 2013), and human Atxn2 with 22 CAG (kindly provided by Prof Stefan Pulst) [17], was cloned in a self-inactivating lentiviral vector as described previously (Deglon et al. 2000). All viral vectors encoding for the different constructs were produced in human embryonic kidney 293T cells using a four-plasmid system described previously (de Almeida et al. 2002). Lentiviral vectors encoding for GFP and Ataxin-2 were used for stereotaxic injection as previously described (Nobrega et al. 2013). The ARH was defined by using The Paxino’s Mouse Brain Atlas and the injection was performed bilaterally, as performed in previous studies (Aveleira et al. 2015). Detailed methods are in Supplementary Methods.
Blood and tissue collection and analysis: qRT-PCR, immunohistochemistry and histological staining
Upon euthanization, blood was collected for biochemical analysis, the hypothalamus was harvested and snap frozen for RNA extraction and quantitative real-time PCR (qRT-PCR). Peripheral organs such as the liver, epididymal white adipose tissue (WAT) and brown adipose tissue (BAT) were collected and cut into two separate portions. One portion was snap frozen and stored at −80°C for RNA extraction and the other portion was kept in a 10% neutral buffered formalin solution for 48 h for posterior histological processing and haematoxylin–eosin staining.
For immunohistochemistry purposes, mice were submitted to a pentobarbital overdose and transcardially perfused with 4% paraformaldehyde/phosphate-buffered saline (PFA/PBS) solution. Brains were removed and cryoprotected in a sucrose solution. The brains were then frozen and sectioned using a cryostat and sections were used for immunohistochemistry.
Tissue collection for all studies was performed during the same temporal window of the day, between Zeitgeiber (ZT) = 1 h and ZT = 3 h, time after the beginning of the light period (lights switch on at 07:00 h ZT = 0 h).
Detailed methods of RNA extraction, qRT-PCR, histological processing and staining, and immunohistochemistry are in Supplementary Methods.
Insulin tolerance test
For insulin tolerance test (ITT), mice were fasted for 6 h (from ZT = 1 h to ZT = 7 h) and injected i.p. with 1 U/kg of insulin. Blood samples were collected at 0, 5, 10, 15, 20, 25 and 30 min from the tail for glucose measurement using a glucometer. The area under the curve (AUC) was calculated in the statistic software Graphpad Prism 8.
Glucose tolerance test
For glucose tolerance test (GTT), mice were fasted for 6 h (from ZT = 1 h to ZT = 7 h) and injected i.p. with a 25% glucose solution. Blood samples were collected at 0, 15, 30, 45, 60 and 120 min from the tail for glucose measurement using a glucometer. The AUC was calculated in the statistic software Graphpad Prism 8.
Behavioural test: open field
For the assessment of mice locomotor horizontal activity, the open field test was performed. Open field was performed 42 and 44 days after stereotaxic injection in Atxn2 KO mouse. The open field test was performed the first time within the inactive period of the mice (light – ZT = 3 h after the beginning of the 12 h light cycle) and the second time within the active period (dark – ZT = 15 h, 3 h after the beginning of the 12 h darkness cycle). For all the tests, mice were acclimated into the test room for a 12 h period. Mice were placed in a 50 × 50 cm arena with 50 cm high walls and their movement activity was recorded for 40 min using the Acti-Track System (Panlab, Barcelona, Spain). The activity tracing of the two zones of the box and the mean values for total distance travelled and velocity were analysed. Comparisons between active and inactive periods were performed as well as comparisons between groups.
Statistical analysis
Results are expressed as mean ± s.e.m. in columns plus scatter plot graphs. A detailed description of the statistical analysis performed for each group of animals is provided within the caption of each figure and in the Supplementary Methods section. Outlier identification was performed in Graphpad Prism 8 software; these points were omitted from analysis. When data from the same animal were consistently identified as an outlier by the software in different parameters, data concerning this animal were removed from all the analysis and not considered as part of the study. More information on the statistical analysis can be found in the captions of each figure and in Supplementary Methods.
Results
Ataxin-2 in the hypothalamus, a new nutritional sensor
Evidence in different models suggest that ataxin-2 might be affected by the nutritional status; in worm and yeast mediating caloric restriction effects and in mammalian cells, increasing upon starvation (Carmo-Silva et al. 2017). Here, we show that HFD decreases ataxin-2 mRNA (Fig. 1C) and protein levels (Fig. 1A and B) in the hypothalamus of mice (Fig. 1), suggesting an expression modulated by the nutritional environment, that decreases upon nutrient excess. Furthermore, ataxin-2 mRNA expression in the liver is also decreased (Fig. 1D) and shows a tendency for decrease in WAT BAT (Fig. 1D), further highlighting this potential role as a nutritional sensor.
Ataxin-2 expression is affected by the nutritional status. (A, B and C) HFD decreases ataxin-2 levels (protein) and expression (mRNA) in the hypothalamus. (A, B) Western blotting analysis of ataxin-2 protein levels in the whole hypothalamus of mice fed with an HFD or with chow diet for 4 weeks. (C) Ataxin-2 mRNA expression in the whole hypothalamus is decreased in mice fed an HFD. (D) HFD promotes a trend for decreased ataxin-2 mRNA expression in the liver, WAT and BAT. (A, B, C and D) n = 3–6 per group (males only) *P < 0.05 compared to control (chow) as determined by a t test for each condition. Data are expressed as mean ± s.e.m.
Citation: Journal of Molecular Endocrinology 70, 1; 10.1530/JME-21-0272
Ataxin-2 in the ARH prevents diet-induced obesity
Considering these observations, and the fact that loss of ATXN2 promotes metabolic dysfunction, we aimed to understand if the effects of HFD could be mediated through the decrease of ataxin-2. Hence, we specifically overexpressed ataxin-2 in the ARH of mice fed an HFD. Ataxin-2 overexpression was achieved through lentiviral vectors encoding for ataxin-2 delivered through stereotaxic injection in the ARH (Fig. 2A and B). The transduction targeted all neuronal populations within the ARH.
Ataxin-2 ARH overexpression prevents diet-induced obesity and metabolic dysfunction. (A) Representative images of ataxin-2 immunoreactivity in the mouse ARH 4 weeks after lentiviral injection, in the control group (chow) and ARH Ataxin-2 overexpressing group (Ataxin-2). Brain slices were immunolabelled for Ataxin-2 (red) and nuclei stained with Hoechst 33342 (blue). White tracing defines the ARH area, IIIV marks the third ventricle area. Figures are representative of three/four independent experiments. Scale bar, 100 μM. (B) Quantification of Ataxin-2 immunoreactivity in the ataxin-2 overexpressing group through the anterior-posterior length of the mouse ARH, 4 weeks after lentivirus injection. n = 3–4 per group. *P < 0.05 compared to control (chow) as determined by Student’s t test. Data are expressed as mean ± s.e.m. (C) Cumulative body weight gain (in grams) of mice with Ataxin-2 ARH overexpression fed chow and HFD. n = 6–10 mice per group. **P < 0.01, ***P < 0.001 and ****P < 0.0001 compared to control chow, as determined by a two-way ANOVA test. ##P < 0.01, ###P < 0.001 and ####P < 0.0001 compared to HFD, as determined by a two-way ANOVA test, followed by a Bonferroni’s multiple comparison post hoc test. Data are expressed as mean ± s.e.m. (D) Total food intake (Cal/4 weeks) for all groups. n = 6–10 mice per group (all males). ****P < 0.0001 compared to control Chow, as determined by one-way ANOVA followed by a Dunnet post hoc test. Data are expressed as mean ± s.e.m. (E, F, G and H) Ataxin-2 modulation in the ARH impacts insulin sensitivity and glucose tolerance. (E) ITT, expressed as the glycaemia (mg/mL) measured every 5 min after i.p. injection of insulin. (F) AUC represents glucose values upon insulin administration. (G) GTT, expressed as the glycaemia (mg/mL) measured up to 120 min after i.p. injection of glucose. (H) AUC represents glucose values upon glucose administration. n = 3–10 per group (all males). *P < 0.05, **P < 0.01 and ***P < 0.001, compared to control chow and ###P < 0.001, compared to HFD, as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests. Data are expressed as mean ± s.e.m. (I) ARH Ataxin-2 overexpression improves serum metabolic alterations caused by HFD. n = 6–10 per group (all males). **P < 0.01, ***P < 0.001 and ****P < 0.0001 compared to control chow and P = 0.06 and ##P < 0.01 compared to HFD, as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests. Data are expressed as mean ± s.e.m.
Citation: Journal of Molecular Endocrinology 70, 1; 10.1530/JME-21-0272
Mice with ataxin-2 ARH overexpression were exposed to HFD in order to understand if replenishing ataxin-2 in the hypothalamus could prevent the effects of the diet. Indeed, ataxin-2 overexpression prevented HFD-induced weight gain (Fig. 2C). However, this was unrelated to food intake, since both groups (HFD-fed) showed similar caloric intake (Fig. 2D and B). Data from HFD-fed group showed higher expression of orexigenic neuropeptides in the hypothalamus (Supplementary Fig. 1A), which was prevented by ARH ataxin-2 (Supplementary Fig. 1A). However, the effects of ataxin-2 on orexigenic neuropeptide expression were not sufficient to decrease food intake (Fig. 2D). Furthermore, increased locomotor activity could also explain the prevention of weight gain with similar food intake; however, we observed no differences on locomotor activity between groups (Supplementary Fig. 1B and C).
Since insulin resistance and glucose intolerance are major hallmarks of metabolic dysfunction, and insulin resistance a described characteristic of mice lacking ATXN2 (Lastres-Becker et al. 2008), we performed ITT and GTT. Under HFD, as expected, we observe a decrease in insulin sensitivity (Fig. 2E), as shown by the increase of the AUC of the ITT test in the HFD group (Fig. 2F). Interestingly, the HFD with ARH ataxin-2 overexpression group shows similar response as the chow group (Fig. 2E), with a decrease in the AUC, compared to HFD group (Fig. 2F). Regarding GTT, both HFD groups show lower tolerance (Fig. 2G); however, with ARH ataxin-2, the response is significantly better (Fig. 2G and H). These observations suggest that replenishing ataxin-2 in the ARH of HFD-fed mice prevents insulin resistance and glucose intolerance.
HFD feeding and consequent diet-induced obesity (DIO) are also characterized by alterations in several parameters such as increased cholesterol, triglycerides, alteration in hepatic enzymes and others. In HFD-fed mice, as expected, we observed an increase in these parameters (Fig. 2I), prevented with ARH ataxin-2 overexpression (Fig. 2I). Overall data suggest that ataxin-2 in ARH can prevent diet-induced weight gain and also other metabolic features that occur upon DIO. This might indicate that the decrease of ataxin-2 in the hypothalamus, occurring upon DIO, might be potentiating the effects of the diet, and its re-establishment might prevent some metabolic consequences of HFD feeding.
Ataxin-2 ARH upregulation prevents central and peripheral diet-induced alterations
DIO and its associated insulin resistance have been associated with peripheral organ alterations, namely with liver steatosis and inflammation, BAT whitening and WAT enlargement and inflammation (Glass & Olefsky 2012, Mendes et al. 2018, Lee et al. 2020). Indeed, we observed that livers from HFD-fed mice presented qualitative signs of hepatocyte hypertrophy and microvesicular steatosis (Fig. 3A), prevented by ataxin-2 ARH overexpression. Furthermore, we assessed the mRNA expression of a set of genes encoding for enzymes determinant for hepatic lipogenesis (Fig. 3E). The HFD-fed group showed significant upregulation of hepatic mRNA of medium-chain acyl-coenzyme A dehydrogenase (MCAD – mitochondrial fatty acid β oxidation enzyme) and phosphoenolpyruvate carboxykinase (PEPCK – gluconeogenesis-related enzyme) (Fig. 3E). Interestingly, ARH ataxin-2 overexpression prevented HFD-induced MCAD and PEPCK mRNA increase (Fig. 3E), showing potential effects also on fatty acid synthase (FASN) and sterol regulatory element-binding protein (SREBP) (Fig. 3E). Considering inflammation as a major hallmark of DIO (Monteiro & Azevedo 2010), we also evaluated mRNA expression of pro-inflammatory markers. Indeed, in the liver of HFD mice, the levels of tumour necrosis factor alpha (TNFα) were higher, when compared to chow diet group, significantly prevented by the ARH ataxin-2 overexpression (Fig. 3F).
Ataxin-2 ARH overexpression impacts the physiology of main peripheral organs, possibly through prevention of hypothalamic inflammation. (A) Representative images of H&E-stained livers, white and brown adipose tissues. For each group n = 3–4, scale bar 100 µM. (B, C and D) HFD increases WAT amount (in % of overall body weight) (B) and adipocyte area (µM2) (C), prevented by ataxin-2 ARH overexpression. Overexpression increases overall amount of BAT (D). n = 6–10 per group (all males). *P < 0.05, **P < 0.01 and ***P < 0.001 compared to control chow and #P < 0.05 compared to HFD as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m. (E) Ataxin-2 overexpression partially prevents changes in liver gene expression. mRNA quantification of genes involved in liver physiology (MCAD; PEPCK; FASN; ATP citrate lyase (ACLY); glucose-6-phosphate (G6P) and SREBP). n = 3–7 per group. **P < 0.01 and ****P < 0.0001 compared to control chow and P = 0.07, #P < 0.05 and ####P < 0.0001 compared to HFD as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m. (F) Ataxin-2 overexpression in ARH partly prevents diet-induced inflammation in the liver, decreasing TNFα, but showing no effect on IL1β and TGF1β. n = 3–7 per group (all males). *P < 0.05 compared to control chow; #P < 0.05 compared to HFD as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m. (G) Ataxin-2 overexpression in ARH prevents diet-induced inflammation in the hypothalamus (TNFα, IL1β, TGF1β and GFAP). n = 3–7 per group (all males). P = 0.053, ***P < 0.001 and ****P < 0.0001 compared to Control Chow; ####P < 0.0001 compared to HFD as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m.
Citation: Journal of Molecular Endocrinology 70, 1; 10.1530/JME-21-0272
Adipose tissue (WAT and BAT) changes were also evaluated. Under HFD, ARH ataxin-2 was able to decrease WAT% in concern to whole body weight (Fig. 3B), concordant with the decreased weight gain of the mice in this group (Fig. 2C). In accordance, adipocyte area, increased upon HFD, was also decreased with ARH ataxin-2 overexpression (Fig. 3A and C). As expected, BAT from chow-fed mice showed regular multilocular adipocytes (Fig. 3A) and, with HFD, an increase in lipid droplet size (Fig. 3A). Furthermore, ARH ataxin-2 rescued BAT structural changes induced by HFD (Fig. 3A). Interestingly, mice from this group show an increase in BAT amount (%) compared to chow and HFD-fed mice (Fig. 3D). To better understand the changes occurring in WAT and BAT, we screened the expression of genes involved in mitochondrial activity and catabolic activity of the adipose tissues, uncoupling protein-1, peroxisome proliferator-activated receptor gamma and peroxisome proliferator-activated receptor-gamma coactivator alfa. However, we observed no significant changes between groups (Supplementary Fig. 2A and B). Furthermore, we also evaluated the expression of pro-inflammatory markers in WAT and BAT and found a significant increase in TNFα mRNA levels in the HFD group, which was prevented by ARH ataxin-2 overexpression (Supplementary Fig. 2C and D). These results present high variability between subjects of the same group which can dampen possible significant differences that could be found between groups. However, data suggest a beneficial effect of ARH ataxin-2 overexpression in these organs.
Hypothalamic inflammation, a determinant hallmark of obesity, can lead to hypothalamic neuronal networks dysfunction, disrupting autonomic regulation of peripheral organs (Monteiro & Azevedo 2010, Carmo-Silva & Cavadas 2017, Lee et al. 2020). Overall, ARH ataxin-2 overexpression prevented HFD-induced inflammation, observed by the decrease in pro-inflammatory cytokines mRNA, such as TNFα, interleukin-1β (IL1β), transforming growth factor 1β (TGF1β) and also in glial fibrillary acidic protein (GFAP) mRNA levels (Fig. 3G).
Hypothalamic inflammation can directly and negatively impact the physiology of peripheral organs. Ataxin-2 ARH overexpression prevents hypothalamic inflammation and some changes to the peripheral organs. These observations need further studying on whether hypothalamic ataxin-2 is also interfering with the peripheral organs, or if the beneficial changes derive from ataxin-2-promoted prevention of hypothalamic inflammation.
Hypothalamic ataxin-2 re-establishment in Atxn2 KO prevents insulin resistance and partly rescues hepatic steatosis
Atxn2 KO mice were generated to study ataxin-2 loss of function in the context of SCA2 (Kiehl et al. 2006). Interestingly, Atxn2 KO mice are viable, although not fertile, and show no neurodegenerative signs or locomotor impairments (Kiehl et al. 2006, Lastres-Becker et al. 2008). However, Atxn2 KO mice are insulin resistant and prone to obesity (Kiehl et al. 2006, Lastres-Becker et al. 2008). Considering results obtained with C57Bl6 mice (Figs 1, 2 and 3), we hypothesized that this phenotype might be partly due to ataxin-2’s absence in the hypothalamus.
To evaluate this hypothesis, we re-established ataxin-2 levels in the ARH of Atxn2 KO through the same approach as before. Atxn2 KO strains are more prone to obesity compared to other strains, either naturally at older age or under HFD feeding (Kiehl et al. 2006, Scoles et al. 2012). For this reason, mice were all fed an HFD, following previously reported protocols (Kiehl et al. 2006, Lastres-Becker et al. 2008, Scoles et al. 2012). As described by others (Kiehl et al. 2006, Lastres-Becker et al. 2008), our Atxn2 KO mice showed higher body weight gain, when compared to wild-type (WT) littermates (Fig. 4A). The ARH ataxin-2 reestablishment in Atxn2 KO mice did not reduce body weight, compared to control Atxn2 KO mice (Fig. 4A). Surprisingly, we did not observe hyperphagic behaviour in Atxn2 KO mice (Fig. 4B) that could explain body weight gain, but on the contrary, Atxn2 KO mice eat less than the WT littermates, and ARH ataxin-2 had no impact on food intake (Fig. 4B). Compared to WT, Atxn2 KO mice also showed no significant changes in the expression of hypothalamic genes involved in food intake regulation that could account for decreased food intake (Supplementary Fig. 3A).
Re-establishment of ARH ataxin-2 in Atxn2 KO mouse improves insulin sensitivity and hepatic lipid accumulation, with mild impact on hypothalamic inflammation, but no weight changes. (A) Cumulative weight gain (in grams) for different timepoints throughout the 8 weeks of the study. n = 5–6 per group (WT – 4 males + 1 female; KO – 2 males + 3 females; KO+Atxn2 – 3 males + 3 females). *P < 0.05, compared to WT, as determined by two-way ANOVA. Data are expressed as mean ± s.e.m. (B) Total food intake expressed in total calories ingested (Cal/8 weeks). n = 5–6 per group. *P < 0.05, compared to WT, as determined by one-way ANOVA followed by Dunnet’s post hoc test. Data are expressed as mean ± s.e.m. (C, D, E and F) Ataxin-2 hypothalamic re-establishment in the KO mice improves fasting glycaemia, insulin sensitivity and promotes modest improvements in glucose tolerance. (C) Fasting glycaemia (mg/mL) in all mice. (D) GTT, expressed as the glycaemia (mg/mL) measured up to 120 min after i.p. glucose injection. (E) ITT, expressed as the glycaemia (mg/mL) measured every 5 min after i.p. injection of insulin. (F) AUC represents glucose values upon insulin administration. n = 5–8 per group (WT – 7 males + 1 female; KO – 2 males + 5 females; KO+Atxn2 – 3 males + 3 females). *P < 0.05 compared to WT and ##P < 0.01 compared to KO, as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc test. Data are expressed as mean ± s.e.m. (G) Representative images from H&E-stained sections from livers of WT, Atxn2 KO mice and Atxn2 KO + ARH Ataxin-2, fed with HFD for 8 weeks n = 5–6 per group (WT – 4 males + 1 female; KO – 2 males + 3 females; KO+Atxn2 – 3 males + 3 females). Scale bar 100 µM. (H) Atxn2 KO shows significant changes in the expression of genes participating in the physiological role of the liver. mRNA quantification of genes involved in liver physiology MCAD, ACLY, FASN, PEPCK, G6P and SREBP. n = 3–6 per group (WT- 2 males + 1 female; KO – 2 males + 3 females; KO+Atxn2 – 3 males + 3 females). P = 0.052, *P < 0.05, **P < 0.01 and ***P < 0.001 compared to WT as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m. (I and J) Atxn2 KO mice show no signs of inflammation in the liver (I) n = 3–6 per group (WT – 2 males + 1 female; KO – 2 males + 3 females; KO+Atxn2 – 3 males + 3 females), but display signs of possible hypothalamic inflammation (J). mRNA quantification of TNFα, IL1β, TGF1β and GFAP. n = 3–6 per group (WT – 4 males + 1 female; KO – 2 males + 1 female; KO+Atxn2 – 2 males + 1 female). *P < 0.05 compared to WT and P = 0.052 and #P < 0.05 compared to KO, as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset.
Citation: Journal of Molecular Endocrinology 70, 1; 10.1530/JME-21-0272
Since insulin resistance is one of the most striking aspects of this mouse model (Lastres-Becker et al. 2008), we started by evaluating fasting glucose, glucose tolerance and insulin sensitivity. Indeed, we observe a trend for higher fasting glycaemia (Fig. 4C) and higher glucose intolerance (Fig. 4D and Supplementary Fig. 3B) in the KO. The strongest observation was related to ITT, where we observed an impaired response to insulin in Atxn2 KO (Fig. 4E), observed by the increase in the AUC of the ITT (Fig. 4F) and a decrease in glucose clearance rate (Supplementary Fig. 3C). All these changes were prevented with ataxin-2 ARH re-establishment (Fig. 4E, F and Supplementary Fig. 3C). Furthermore, ARH ataxin-2 also promoted a trend for improvement on parameters such as cholesterol (Supplementary Fig. 3D).
Previous studies reported severe lipid accumulation in the liver of Atxn2 KO mice (Kiehl et al. 2006, Lastres-Becker et al. 2008). We also observed macro and microvascular steatosis (Fig. 4G). Surprisingly, ARH ataxin-2 re-establishment in Atxn2 KO mice rescued lipid accumulation (Fig. 4G). We further assessed mRNA expression of genes encoding for enzymes determinant for hepatic lipogenesis (Fig. 4H). We found significant changes inn mRNA levels of MCAD, ACLY and SREBP of Atxn2 KO mice, not affected by ataxin-2 ARH re-establishment (Fig. 4H). We did not observe changes in mRNA levels of proinflammatory genes in the liver (Fig. 4I). Following the steps of the C57Bl6 study, and considering hypothalamic inflammation as a driver for metabolic dysfunction, we studied pro-inflammatory markers in the hypothalamus. Indeed, we observed a significant increase in GFAP and a trend for increased TNFα in the Atxn2 KO (Fig. 4J), similar to what was observed in HFD-fed C57Bl6 mice (Fig. 2G). Hypothalamic inflammation might promote hepatic lipid accumulation; however, this does not seem to be the main mechanism in Atxn2 KO mice.
Regarding WAT, we observed no significant changes in structure (Supplementary Fig. 4A), amount (Supplementary Fig. 4B) or gene expression (Supplementary Fig. 4B) between genotypes. In BAT, we observed a ‘whiter’ appearance in Atxn2 KO mice, that was somewhat prevented with ARH ataxin-2 (Supplementary Fig. 4D). Both Atxn2 KO groups showed increased BAT amount when compared to WT (Supplementary Fig. 4E). In both WAT and BAT of Atxn2 KO mice, it was possible to observe a trend for mRNA increase of pro-inflammatory markers and some prevention in ARH-ataxin-2 re-establishment group; however, these changes were not statistically significant (Supplementary Fig. 4G and H).
These results suggest that absence of ataxin-2 in the hypothalamus is determinant for major characteristics of the Atxn2 KO phenotype, insulin resistance and hepatic steatosis. ARH ataxin-2 can prevent these, independently of changes on body weight.
Clock gene expression alterations as potential promoters of metabolic dysfunction
Obesity without increased food intake can occur through circadian alterations (Oishi & Hashimoto 2018), in fact, food ingestion during the rest phase promotes obesity (Kohsaka et al. 2007, Hatori et al. 2012). Others described a hyperactive phenotype of Atxn2 KO (Lastres-Becker et al. 2008, Scoles et al. 2012, Pfeffer et al. 2017) and suggested disrupted circadian rhythms in this mouse model; however, the results were not conclusive (Pfeffer et al. 2017). Hence, we hypothesized that Atxn2 KO obesity, without hyperphagia (Fig. 3A, B and C), could be related to circadian disturbances. We started by evaluating locomotor activity/behaviour during the rest phase (day) and active phase (night) in the three groups.
In the open field test, the normal pattern of behaviour is represented by higher locomotor activity (distance travelled) in the dark period (active phase), compared to the light period (rest phase). In fact, as expected, we observed that WT presents this pattern (Fig. 5A and B). Interestingly, Atxn2 KO mice showed higher activity in the light period (rest) compared to WT (Fig. 5B), while Atxn2 KO with ARH ataxin-2 showed a trend for reduced locomotor activity when compared to other KO (Fig. 5A and B). Atxn2 KO mice present a non-significant fold-change between light and dark period activity, which can be observed in WT (****P < 0.0001) (Fig. 5B). Interestingly, in Atxn2 KO with ARH ataxin-2, changes between rest and active phase were restored (**P < 0.01) (Fig. 5B).
Atxn2 KO mice display altered light–dark behaviour that might be related to disturbances in the central circadian clock. (A and B) Open field test was performed to evaluate locomotor activity in all groups during the rest (light) and active (dark) periods. (A) Representative activity tracing of the two zones analysed in the open field test for all groups in both active and inactive periods. Tracing obtained by the analysis of the ActiTrack software. (B) Total distance travelled (cm) in the open field test in all groups for light (ZT = 3) and dark period (ZT = 15). Bars with no pattern represent the light/rest period, whereas patterned bars represent the dark/active period. n = 6 per group (WT – 5 males + 1 female; KO – 2 males + 4 females; KO+Atxn2 – 3 males + 3 females). **P < 0.01 and ****P < 0.0001 relative to the fold change between ‘Active’ and ‘Rest’ periods of the respective group, as determined by t test. #P < 0.05 relative to WT ‘Active period’, as determined by one-way ANOVA followed by Dunnett’s post hoc test. Data are expressed as mean ± s.e.m. (C and D) Atxn2 KO mice present disruption in mRNA expression of clock genes in the hypothalamus and in the liver. mRNA content of clock genes in the whole hypothalamus (C) and liver (D) of WT, KO and KO+Ataxin-2 at ZT = 1–3 h: Bmal1, CLOCK, Cry and Per1 and Per2. n = 3–6 per group (WT – 4 males + 1 female; KO – 2 males + 2 females; KO+Atxn2 – 2 males + 1 female). *P < 0.05, **P < 0.01 and ****P < 0.0001 compared to WT and #P < 0.05 and ##P < 0.01 compared to KO as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m. (E and F) HFD feeding in C57Bl6 promotes alterations on clock gene expression in the hypothalamus (E) and liver (F) ZT = 1–3 h. n = 3–6 per group (all males). *P < 0.05 and**P < 0.01 compared to Chow and #P < 0.05 and ##P < 0.01 compared to HFD as determined by one-way ANOVA followed by Dunnet’s and Bonferroni’s post hoc tests, for each gene subset. Data are expressed as mean ± s.e.m.
Citation: Journal of Molecular Endocrinology 70, 1; 10.1530/JME-21-0272
Others showed that clock gene alterations can promote the loss of light–dark behavioural adaptations, which can predispose to insulin resistance and obesity (Ribas-Latre & Eckel-Mahan 2016). Hence, we assessed the expression of main circadian clock genes, aryl hydrocarbon receptor nuclear translocator-like protein 1 (ARNLT or Bmal1), circadian locomotor output cycles kaput (Clock), cryptochrome (Cry) and period 1/2 (Per1 and Per2) in mice hypothalami. We observed higher mRNA levels of the clock genes Bmal1, Cry and Per2 in Atxn2 KO (Fig. 5C), compared to WT. Interestingly, ataxin-2 ARH re-establishment significantly recuperated mRNA levels of these clock genes to levels similar to WT. These results might shed some light on Atxn2 KO obese phenotype and the beneficial metabolic outcome of ARH ataxin-2 re-establishment.
To confirm the hypothesis that ataxin-2 could regulate clock gene expression, we evaluated clock genes in the hypothalamus of the previous cohort of C57Bl6 mice (Figs 2 and 3). We observed an upregulation of Bmal1 and Per1 in HFD, prevented by ARH ataxin-2 overexpression (Fig. 5E). Furthermore, ARH ataxin-2 significantly counteracted the effect of HFD on Per2 (Fig. 5E), similar to what happens in Atxn2 KO.
ARH ataxin-2, in both C57Bl6 and Atxn2 KO mice, had a significant effect on the liver (Figs 3A, E, F and 4G, H, I). The liver circadian rhythm is determinant for overall metabolic homeostasis (Tahara & Shibata 2016); thus, we evaluated clock gene expression in the livers of both mice cohorts. The first striking observation was the distinct expression patterns of clock genes in the hypothalamus and liver of the mice composing each experiment (Fig. 5C vs D, E vs F). In Atxn2 KO mice liver, Clock was upregulated while Cry and Per2 were downregulated, and ARH ataxin-2 rescued Clock expression (Fig. 5D). In HFD-fed C57Bl6, we observed a significant decrease in Bmal1 expression and increase of Per1 (Fig. 5F). Interestingly, ataxin-2 ARH upregulation had no effect on these gene's expression, but impacted Clock expression (Fig. 5F), similar Atxn2 KO (Fig. 5D). Since changes on WAT and BAT physiology were not as marked as those in the liver, clock gene alterations in these organs were not as expected. However, in C57Bl6 mice, we observed an increase in the expression of Per genes (Per1 and Per2) in WAT (Supplementary Fig. 2E) and downregulation of BAT’s Bmal1 and Cry (Supplementary Fig. 2F), not reverted by ARH ataxin-2 overexpression. In Atxn2 KO mice, we observed lower levels of Per1 and Per2 mRNA in BAT (Supplementary Fig. 4J), unchanged by ARH ataxin-2. In WAT of Atxn2 KO mice, ARH ataxin-2 impacts Bmal1, Clock and Cry expression (Supplementary Fig. 4I).
Overall, we show that ataxin-2 absence (KO) and/or decrease (HFD) promotes changes in clock genes expression, both central and peripheral, which can per se potentiate metabolic dysfunction. Re-establishment of ataxin-2 in the ARH rescues these alterations, mostly at central level, which might be a possible cause for the beneficial metabolic effects observed in both experimental approaches.
Discussion
Here, we provide new evidence for ataxin-2 as a player in central energy balance regulation and overall metabolism. We describe for the first time that DIO depletes ataxin-2 levels in mouse hypothalamus and liver. Furthermore, we show that increasing hypothalamic (ARH) ataxin-2 levels prevents HFD-induced weight gain, insulin resistance and glucose intolerance. Moreover, we deepen studies relating Atxn2 KO mice, where the reestablishment of ataxin-2 in the ARH has no effect on body weight, but reverts insulin resistance and hepatic steatosis, phenotypic characteristics of this mouse model. We further describe a role for ataxin-2 in the modulation of genes involved in circadian rhythm, suggesting that the beneficial metabolic effects of ARH ataxin-2 might, at least partly, be linked to central circadian clock resetting.
The hypothalamus exerts a special unifying role concerning metabolism, regulating directly metabolic processes in peripheral organs, but also indirectly, through the circadian clock (Albrecht 2012, Carmo-Silva & Cavadas 2017, Engin 2017, Lee et al. 2020). Hence, modulating hypothalamic physiology might exert an impact on metabolism in a more direct or indirect manner. In this study, we observed that modulating ataxin-2 in the hypothalamus indeed exerts a direct effect on peripheral organs such as the liver and also has an impact on clock gene expression both centrally (in the hypothalamus) and peripherally (in the liver).
The most striking effect of ARH ataxin-2 (both in HFD and Atxn2 KO mice) was the prevention of insulin resistance, which might be partly related to its anti-inflammatory effect in the hypothalamus. Ataxin-2 promotes a robust decrease of pro-inflammatory cytokines (TNF-α, IL-1β), TGF1β and glial markers (GFAP) even in HFD. Others showed that hypothalamic inflammation contributes to insulin resistance and that suppressing it can improve insulin sensitivity (Monteiro & Azevedo 2010, Glass & Olefsky 2012, Mendes et al. 2018, Lee et al. 2020). This is the first report of ataxin-2’s anti-inflammatory potential. Increased mTOR activity can also promote insulin resistance (Yoon & Choi 2016), and others have shown that silencing ataxin-2 in mammalian cells increases mTOR activation, whereas ataxin-2 orthologue in yeast (Pbp1) can sequester mTOR components, preventing its activation (Carmo-Silva et al. 2017). This mechanistic aspect might also be involved in the effects of ataxin-2 that we show here.
Others showed that blocking hypothalamic inflammation prevents hepatic steatosis and improves liver physiology (Monteiro & Azevedo 2010, Glass & Olefsky 2012, Mendes et al. 2018). Here we showed that replenishing ARH ataxin-2 prevented hepatic lipid accumulation and mRNA alterations of hepatic genes encoding for lipogenic enzymes, thus improving liver metabolism and insulin sensitivity. Moreover, insulin resistance can also be induced by the deregulation of liver lipogenic pathways (Glass & Olefsky 2012). Evidence shows that ataxin-2 depletion modifies lipid droplet regulator levels, perilipin-3 (Plin3), and apolipoprotein secretor factor microsomal triglyceride transfer protein (Mttp), suggesting ataxin-2 as a regulator of lipid energy reserves (Carmo-Silva et al. 2017).
Atxn2 KO under HFD gained more weight when compared to their WT littermates, as described by others (Kiehl et al. 2006, Lastres-Becker et al. 2008). However, we did not observe hyperphagia, as shown by others (Kiehl et al. 2006, Scoles et al. 2012); on the contrary, Atxn2 KO mice eat less. These results suggest that body weight gain could be secondary to factors beyond food intake. On one hand, Atxn2 KO mice might present blunted reward responses that could affect food intake. Scoles et al., observed a marked expression of ataxin-2 in the olfactory bulb (OB) in mice (Scoles et al. 2012), evidence from flies implicates ataxin-2 ortholog in olfactory habituation (McCann et al. 2011) and, SCA2 patients show olfactory impairments (Pedroso et al. 2017). Olfaction and food consumption are tightly interconnected. In Atxn2 KO, ataxin-2 absence in OB might blunt smell, decreasing smell-associated reward signalling, resulting in decreased food intake. Another explanation for increased body weight with decreased food intake would be decreased physical activity. However, on the contrary, we observed that Atxn2 KO mice show signs of hyperactivity when compared to WT, as described by others (Lastres-Becker et al. 2008, Pfeffer et al. 2017). Interestingly, higher activity of Atxn2 KO mice was mostly observed during the rest phase (day), which could suggest deregulation of circadian behaviour. In fact, Atxn2 KO mice show altered expression of clock genes (Bmal1, Cry and Per2 mRNA) in the hypothalamus, which were rescued upon ataxin-2 ARH re-establishment. Furthermore, Per2 liver expression was altered in Atxn2 KO, and rescued upon ARH ataxin-2 re-establishment. Alterations in clock genes, central and peripheral, are directly linked to insulin resistance and obesity (Kohsaka et al. 2007, Albrecht 2012, Hatori et al. 2012, Blancas-Velazquez et al. 2017, Engin 2017, Oishi & Hashimoto 2018) and are a typical observation of DIO (Kohsaka et al. 2007, Carmo-Silva & Cavadas 2017, Engin 2017).
The role of ataxin-2 on circadian rhythm was described in Drosophila, as a regulator of PER translation, homologue for PER gene in mice (encodes PER1, PER2 and PER3) (Lim & Allada 2013, Lee et al. 2017, Xu et al. 2019). Others showed a decrease of RORA, involved in circadian rhythm regulation, in the cerebellum of Atxn2 KO mice (Halbach et al. 2017). Here, we show further evidence that could point to circadian rhythm upon loss of ataxin-2 (genetic KO and caused by DIO). Considering the connection between circadian disruption and metabolic disturbances (Kohsaka et al. 2007, Albrecht 2012, Hatori et al. 2012, Blancas-Velazquez et al. 2017, Engin 2017, Oishi & Hashimoto 2018), the regulatory effect of ataxin-2 over clock genes expression might be linked to its beneficial metabolic effects.
Some limitations need to be addressed in a following study, namely regarding circadian rhythm studies. The results concerning clock gene expression were interesting but unexpected; hence, our experiments were not designed to properly evaluate 24 h circadian rhythms. Initially, this evaluation was not even predicted, but our group is exploring this avenue in further studies, focusing on the mechanisms involved in this triad: hypothalamic ataxin-2, circadian rhythm and metabolism. Furthermore, our approach to the ataxin-2 overexpression in the hypothalamus was an exploratory one. We transduced all neuronal populations within the hypothalamus with ataxin-2. Although this gives us a more physiological understanding of the role of ataxin-2 in this brain region, some information regarding the intrinsic neuronal networks is missing. Transducing NPY/AgRP or POMC/CART neurons differentially could provide us with not only a better understanding of the results obtained but also the specific role of ataxin-2 on weight regulation at a hypothalamic level. We are now performing more detailed experiments using neuron activity assessment within the ARH. All this suggests that ataxin-2 could be used as a potential target for metabolic dysfunction. Nonetheless, the upregulation of ataxin-2 needs further investigation, considering its multiple roles in physiology and also in disease. Interestingly, our approach of ataxin-2 overexpression in C57Bl6J mice fed with chow exerted no effects on metabolism (Supplementary Fig. 5), which shows us that increasing ataxin-2 in the hypothalamus per se does not seem to be deleterious in a non-pathological condition. Nonetheless, upregulation of ataxin-2 specifically in the hypothalamus is most probably a non-translation approach, and the administration of compounds and caloric restriction that increase ataxin-2 levels could be an interesting strategy to investigate. The increase of ataxin-2 specifically in the hypothalamus using these strategies or gene delivery approaches would be a challenge that deserves further investigation. Our group is currently exploring if fucoxanthin administration can increase ataxin-2 in the hypothalamus. Other approaches known to increase ataxin-2 in mammalian cells in vitro are starvation that may also promote a specific increase of this protein in the hypothalamus; thus, starvation or caloric restriction mimetics could also be a potential pharmaceutical tool to achieve ataxin-2 increase.
In conclusion, our findings provide evidence that ataxin-2 in the hypothalamus regulates energy balance, acting through different mechanisms: anti-inflammatory pathways, insulin signalling and circadian rhythm. The major effects of ataxin-2 might all derive from the latter, since the circadian clock is directly connected to hypothalamic and overall inflammation, and also to insulin signalling (Fig. 6). Since DIO decreases ataxin-2 levels in the hypothalamus which, per se, promotes metabolic dysfunction (as seen in Atxn2 KO), the re-establishment of ataxin-2 levels specifically in this brain region might prove to be an innovative approach to prevent diet-induced metabolic dysfunction.
Ataxin-2 in the hypothalamus plays a pivotal role in metabolism. We describe for the first time that diet-induced-obesity promotes depletion of ataxin-2 within the hypothalamus. Ataxin-2 decrease within the hypothalamus, the main regulator of numerous functions such as food intake, body weight, insulin sensitivity and circadian rhythm, amongst others, prompted us to further study the role of this protein in metabolism. As expected, DIO was characterized by increased body weight, insulin resistance, neuroinflammation, inflammation in peripheral tissues and alterations in the expression of clock genes. Interestingly, ataxin-2 ARH overexpression in HFD-fed mice prevented the settlement of obesity and other features. Atxn2 KO mice were described before as obese and insulin resistant (Kiehl et al. 2006, Lastres-Becker et al. 2008). We hypothesized that this phenotype might be related to the absence of ataxin-2 in the hypothalamus; thus, we re-established hypothalamic ataxin-2 in the Atxn2 KO. Although, with no changes body weight, we observed improved insulin sensitivity, decreased hepatic steatosis and improved neuroinflammation. We also gathered evidence of altered day/night behaviour that could suggest alterations in the circadian system. Accompanying this, we observed changes in expression of clock genes in the Atxn2 KO, rescued upon ataxin-2 ARH re-establishment. With this study, we show that ataxin-2 plays a pivotal role in the hypothalamus, regulating body weight and insulin sensitivity. Interestingly, the impact of ataxin-2 in the circadian clock suggests that we might have uncovered a possible mechanism through which ataxin-2 plays a role in metabolism. Nonetheless, this was a pilot study that needs further experiments to prove this connection.
Citation: Journal of Molecular Endocrinology 70, 1; 10.1530/JME-21-0272
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/JME-21-0272.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was co-funded by the European Regional Development Fund (ERDF), through the Centro 2020 Regional Operational Programme under project CENTRO-01-0145-FEDER-000012 (HealthyAging 2020) and through the COMPETE 2020 – Operational Programme for Competitiveness and Internationalization and Portuguese national funds via FCT – Foundation for Science and Technology under the Strategic Project (UID/NEU/04539/2013), the interchange program Portugal-Brazil Convénio FCT/CAPES (340/13) and the fellowships SFRH/BD/89035/2012, SFRH/BPD/62945/2009, SFRH/BPD/73942/2010, SFRH/BD/73004/2010. Work from Pulst lab was supported by NINDS (USA) grants R37NS033123 and R21NS103009.
Author contribution statement
S C S: conceptualization, methodology, investigation, formal analysis, data curation, writing – original draft preparation; M F M: investigation; C N: conceptualization, methodology, writing – reviewing and editing; M B: investigation; D C: investigation; C A: conceptualization, writing – reviewing and editing; S P: resources, writing – reviewing and editing; L P A: conceptualization, supervision, funding acquisition, writing – reviewing and editing; C C: conceptualization, supervision, funding acquisition, writing – reviewing and editing. L P A and C C: contributed equally.
Acknowledgements
The authors wish to thank Janssen Pharmaceutical for their acknowledgement of the present work, through the attribution of an award in ‘Prémio Inovação Janssen’. The authors also wish to acknowledge Research Diets, Inc. for providing a diet grant and Abbott for providing glucometers and glycemia strips for the preliminary studies.
References
Abdel-Aleem A & Zaki MS 2008 Spinocerebellar ataxia type 2 (SCA2) in an Egyptian family presenting with polyphagia and marked CAG expansion in infancy. Journal of Neurology 255 413–419. (https://doi.org/10.1007/s00415-008-0690-4)
Albrecht U 2012 Timing to perfection: the biology of central and peripheral circadian clocks. Neuron 74 246–260. (https://doi.org/10.1016/j.neuron.2012.04.006)
Auburger G, Gispert S, Lahut S, Omur O, Damrath E, Heck M & Basak N 2014 12q24 locus association with type 1 diabetes: SH2B3 or ATXN2? World Journal of Diabetes 5 316–327. (https://doi.org/10.4239/wjd.v5.i3.316)
Auburger G, Sen NE, Meierhofer D, Basak AN & Gitler AD 2017 Efficient prevention of neurodegenerative diseases by depletion of starvation response factor ataxin-2. Trends in Neurosciences 40 507–516. (https://doi.org/10.1016/j.tins.2017.06.004)
Aveleira CA, Botelho M, Carmo-Silva S, Pascoal JF, Ferreira-Marques M, Nobrega C, Cortes L, Valero J, Sousa-Ferreira L & Alvaro AR et al.2015 Neuropeptide Y stimulates autophagy in hypothalamic neurons. PNAS 112 E1642–E1651. (https://doi.org/10.1073/pnas.1416609112)
Becker LA, Huang B, Bieri G, Ma R, Knowles DA, Jafar-Nejad P, Messing J, Kim HJ, Soriano A & Auburger G et al.2017 Therapeutic reduction of ataxin-2 extends lifespan and reduces pathology in TDP-43 mice. Nature 544 367–371. (https://doi.org/10.1038/nature22038)
Blancas-Velazquez A, Mendoza J, Garcia AN & la Fleur SE 2017 Diet-induced obesity and circadian disruption of feeding behavior. Frontiers in Neuroscience 11 23. (https://doi.org/10.3389/fnins.2017.00023)
Carmo-Silva S & Cavadas C 2017 Hypothalamic dysfunction in obesity and metabolic disorders. Advances in Neurobiology 19 73–116. (https://doi.org/10.1007/978-3-319-63260-5_4)
Carmo-Silva S, Nobrega C, Pereira de Almeida L & Cavadas C 2017 Unraveling the role of ataxin-2 in metabolism. Trends in Endocrinology and Metabolism 28 309–318. (https://doi.org/10.1016/j.tem.2016.12.006)
Cheung MK, Gulati P, O’Rahilly S & Yeo GS 2013 FTO expression is regulated by availability of essential amino acids. International Journal of Obesity 37 744–747. (https://doi.org/10.1038/ijo.2012.77)
Damrath E, Heck MV, Gispert S, Azizov M, Nowock J, Seifried C, Rub U, Walter M & Auburger G 2012 ATXN2-CAG42 sequesters PABPC1 into insolubility and induces FBXW8 in cerebellum of old ataxic knock-in mice. PLoS Genetics 8 e1002920. (https://doi.org/10.1371/journal.pgen.1002920)
de Almeida LP, Ross CA, Zala D, Aebischer P & Deglon N 2002 Lentiviral-mediated delivery of mutant huntingtin in the striatum of rats induces a selective neuropathology modulated by polyglutamine repeat size, huntingtin expression levels, and protein length. Journal of Neuroscience 22 3473–3483. (https://doi.org/10.1523/JNEUROSCI.22-09-03473.2002)
Deglon N, Tseng JL, Bensadoun JC, Zurn AD, Arsenijevic Y, Pereira de Almeida L, Zufferey R, Trono D & Aebischer P 2000 Self-inactivating lentiviral vectors with enhanced transgene expression as potential gene transfer system in Parkinson’s disease. Human Gene Therapy 11 179–190. (https://doi.org/10.1089/10430340050016256)
Ding D, Li K, Wang C, Chen Z, Long Z, Peng Y, Zhou X, Peng H, Qiu R & Xia K et al.2016 ATXN2 polymorphism modulates age at onset in Machado-Joseph disease. Brain 139 e59. (https://doi.org/10.1093/brain/aww176)
Elden AC, Kim HJ, Hart MP, Chen-Plotkin AS, Johnson BS, Fang X, Armakola M, Geser F, Greene R & Lu MM et al.2010 Ataxin-2 intermediate-length polyglutamine expansions are associated with increased risk for ALS. Nature 466 1069–1075. (https://doi.org/10.1038/nature09320)
Engin A 2017 Circadian rhythms in diet-induced obesity. Advances in Experimental Medicine and Biology 960 19–52. (https://doi.org/10.1007/978-3-319-48382-5_2)
Figueroa KP, Farooqi S, Harrup K, Frank J, O’Rahilly S & Pulst SM 2009 Genetic variance in the spinocerebellar ataxia type 2 (ATXN2) gene in children with severe early onset obesity. PLoS ONE 4 e8280. (https://doi.org/10.1371/journal.pone.0008280)
Glass CK & Olefsky JM 2012 Inflammation and lipid signaling in the etiology of insulin resistance. Cell Metabolism 15 635–645. (https://doi.org/10.1016/j.cmet.2012.04.001)
Halbach MV, Gispert S, Stehning T, Damrath E, Walter M & Auburger G 2017 Atxn2 knockout and CAG42-knock-in cerebellum shows similarly dysregulated expression in calcium homeostasis pathway. Cerebellum 16 68–81. (https://doi.org/10.1007/s12311-016-0762-4)
Hatori M, Vollmers C, Zarrinpar A, DiTacchio L, Bushong EA, Gill S, Leblanc M, Chaix A, Joens M & Fitzpatrick JA et al.2012 Time-restricted feeding without reducing caloric intake prevents metabolic diseases in mice fed a high-fat diet. Cell Metabolism 15 848–860. (https://doi.org/10.1016/j.cmet.2012.04.019)
Kiehl TR, Nechiporuk A, Figueroa KP, Keating MT, Huynh DP & Pulst SM 2006 Generation and characterization of Sca2 (ataxin-2) knockout mice. Biochemical and Biophysical Research Communications 339 17–24. (https://doi.org/10.1016/j.bbrc.2005.10.186)
Kohsaka A, Laposky AD, Ramsey KM, Estrada C, Joshu C, Kobayashi Y, Turek FW & Bass J 2007 High-fat diet disrupts behavioral and molecular circadian rhythms in mice. Cell Metabolism 6 414–421. (https://doi.org/10.1016/j.cmet.2007.09.006)
Kraja AT, Chasman DI, North KE, Reiner AP, Yanek LR, Kilpelainen TO, Smith JA, Dehghan A, Dupuis J & Johnson AD et al.2014 Pleiotropic genes for metabolic syndrome and inflammation. Molecular Genetics and Metabolism 112 317–338. (https://doi.org/10.1016/j.ymgme.2014.04.007)
Lastres-Becker I, Brodesser S, Lutjohann D, Azizov M, Buchmann J, Hintermann E, Sandhoff K, Schurmann A, Nowock J & Auburger G 2008 Insulin receptor and lipid metabolism pathology in ataxin-2 knock-out mice. Human Molecular Genetics 17 1465–1481. (https://doi.org/10.1093/hmg/ddn035)
Lee J, Yoo E, Lee H, Park K, Hur JH & Lim C 2017 LSM12 and ME31B/DDX6 define distinct modes of posttranscriptional regulation by ataxin-2 protein complex in drosophila circadian pacemaker neurons. Molecular Cell 66 129.e7–140.e7. (https://doi.org/10.1016/j.molcel.2017.03.004)
Lee CH, Suk K, Yu R & Kim MS 2020 Cellular contributors to hypothalamic inflammation in obesity. Molecules and Cells 43 431–437. (https://doi.org/10.14348/molcells.2020.0055)
Lim C & Allada R 2013 Ataxin-2 activates PERIOD translation to sustain circadian rhythms in Drosophila. Science 340 875–879. (https://doi.org/10.1126/science.1234785)
Lv WQ, Zhang X, Zhang Q, He JY, Liu HM, Xia X, Fan K, Zhao Q, Shi XZ & Zhang WD et al.2017 Novel common variants associated with body mass index and coronary artery disease detected using a pleiotropic cFDR method. Journal of Molecular and Cellular Cardiology 112 1–7. (https://doi.org/10.1016/j.yjmcc.2017.08.011)
McCann C, Holohan EE, Das S, Dervan A, Larkin A, Lee JA, Rodrigues V, Parker R & Ramaswami M 2011 The ataxin-2 protein is required for microRNA function and synapse-specific long-term olfactory habituation. PNAS 108 E655–E662. (https://doi.org/10.1073/pnas.1107198108)
Meierhofer D, Halbach M, Sen NE, Gispert S & Auburger G 2016 Ataxin-2 (Atxn2)-knock-out mice show branched chain amino acids and fatty acids pathway alterations. Molecular and Cellular Proteomics 15 1728–1739. (https://doi.org/10.1074/mcp.M115.056770)
Mendes NF, Gaspar JM, Lima-Junior JC, Donato Jr J, Velloso LA & Araujo EP 2018 TGF-beta1 down-regulation in the mediobasal hypothalamus attenuates hypothalamic inflammation and protects against diet-induced obesity. Metabolism: Clinical and Experimental 85 171–182. (https://doi.org/10.1016/j.metabol.2018.04.005)
Monteiro R & Azevedo I 2010 Chronic inflammation in obesity and the metabolic syndrome. Mediators of Inflammation 2010 289645. (https://doi.org/10.1155/2010/289645)
Nobrega C, Nascimento-Ferreira I, Onofre I, Albuquerque D, Conceicao M, Deglon N & de Almeida LP 2013 Overexpression of mutant ataxin-3 in mouse cerebellum induces ataxia and cerebellar neuropathology. Cerebellum 12 441–455. (https://doi.org/10.1007/s12311-012-0432-0)
Nobrega C, Carmo-Silva S, Albuquerque D, Vasconcelos-Ferreira A, Vijayakumar UG, Mendonca L, Hirai H & de Almeida LP 2015 Re-establishing ataxin-2 downregulates translation of mutant ataxin-3 and alleviates Machado-Joseph disease. Brain 138 3537–3554. (https://doi.org/10.1093/brain/awv298)
Oishi K & Hashimoto C 2018 Short-term time-restricted feeding during the resting phase is sufficient to induce leptin resistance that contributes to development of obesity and metabolic disorders in mice. Chronobiology International 35 1576–1594. (https://doi.org/10.1080/07420528.2018.1496927)
Pedroso JL, Braga-Neto P, Escorcio-Bezerra ML, Abrahao A, de Albuquerque MV, Filho FM, de Souza PV, de Rezende Pinto WB, Borges Jr FR & Saraiva-Pereira ML et al.2017 Non-motor and extracerebellar features in spinocerebellar ataxia type 2. Cerebellum 16 34–39. (https://doi.org/10.1007/s12311-016-0761-5)
Pfeffer M, Gispert S, Auburger G, Wicht H & Korf HW 2017 Impact of ataxin-2 knock out on circadian locomotor behavior and PER immunoreaction in the SCN of mice. Chronobiology International 34 129–137. (https://doi.org/10.1080/07420528.2016.1245666)
Pulst SM, Nechiporuk A, Nechiporuk T, Gispert S, Chen XN, Lopes-Cendes I, Pearlman S, Starkman S, Orozco-Diaz G & Lunkes A et al.1996 Moderate expansion of a normally biallelic trinucleotide repeat in spinocerebellar ataxia type 2. Nature Genetics 14 269–276. (https://doi.org/10.1038/ng1196-269)
Ribas-Latre A & Eckel-Mahan K 2016 Interdependence of nutrient metabolism and the circadian clock system: importance for metabolic health. Molecular Metabolism 5 133–152. (https://doi.org/10.1016/j.molmet.2015.12.006)
Sahba S, Nechiporuk A, Figueroa KP, Nechiporuk T & Pulst SM 1998 Genomic structure of the human gene for spinocerebellar ataxia type 2 (SCA2) on chromosome 12q24.1. Genomics 47 359–364. (https://doi.org/10.1006/geno.1997.5131)
Sanpei K, Takano H, Igarashi S, Sato T, Oyake M, Sasaki H, Wakisaka A, Tashiro K, Ishida Y & Ikeuchi T et al.1996 Identification of the spinocerebellar ataxia type 2 gene using a direct identification of repeat expansion and cloning technique, DIRECT. Nature Genetics 14 277–284. (https://doi.org/10.1038/ng1196-277)
Scoles DR, Pflieger LT, Thai KK, Hansen ST, Dansithong W & Pulst SM 2012 ETS1 regulates the expression of ATXN2. Human Molecular Genetics 21 5048–5065. (https://doi.org/10.1093/hmg/dds349)
Sen NE, Arsovic A, Meierhofer D, Brodesser S, Oberschmidt C, Canet-Pons J, Kaya ZE, Halbach MV, Gispert S & Sandhoff K et al.2019a In human and mouse spino-cerebellar tissue, ataxin-2 expansion affects ceramide-sphingomyelin metabolism. International Journal of Molecular Sciences 20 5854. (https://doi.org/10.3390/ijms20235854)
Sen NE, Canet-Pons J, Halbach MV, Arsovic A, Pilatus U, Chae WH, Kaya ZE, Seidel K, Rollmann E & Mittelbronn M et al.2019b Generation of an Atxn2-CAG100 knock-in mouse reveals N-acetylaspartate production deficit due to early Nat8l dysregulation. Neurobiology of Disease 132 104559. (https://doi.org/10.1016/j.nbd.2019.104559)
Tahara Y & Shibata S 2016 Circadian rhythms of liver physiology and disease: experimental and clinical evidence. Nature Reviews: Gastroenterology and Hepatology 13 217–226. (https://doi.org/10.1038/nrgastro.2016.8)
Xu F, Kula-Eversole E, Iwanaszko M, Lim C & Allada R 2019 Ataxin2 functions via CrebA to mediate huntingtin toxicity in circadian clock neurons. PLoS Genetics 15 e1008356. (https://doi.org/10.1371/journal.pgen.1008356)
Yoon MS & Choi CS 2016 The role of amino acid-induced mammalian target of rapamycin complex 1 (mTORC1) signaling in insulin resistance. Experimental and Molecular Medicine 48 e201. (https://doi.org/10.1038/emm.2015.93)
Zanatta A, Camargo CHF, Germiniani FMB, Raskin S, de Souza Crippa AC & Teive HAG 2019 Abnormal findings in polysomnographic recordings of patients with spinocerebellar ataxia type 2 (SCA2). Cerebellum 18 196–202. (https://doi.org/10.1007/s12311-018-0982-x)
