p27Kip1 Deficiency Impairs Brown Adipose Tissue Function Favouring Fat Accumulation in Mice

The aim of this work was to investigate the effect of the whole-body deletion of p27 on the activity of brown adipose tissue and the susceptibility to develop obesity and glucose homeostasis disturbances in mice, especially when subjected to a high fat diet. p27 knockout (p27−/−) and wild type (WT) mice were fed a normal chow diet or a high fat diet (HFD) for 10-weeks. Body weight and composition were assessed. Insulin and glucose tolerance tests and indirect calorimetry assays were performed. Histological analysis of interscapular BAT (iBAT) was carried out, and expression of key genes/proteins involved in BAT function were characterized by qPCR and Western blot. iBAT activity was estimated by 18F-fluorodeoxyglucose (18FDG) uptake with microPET. p27−/− mice were more prone to develop obesity and insulin resistance, exhibiting increased size of all fat depots. p27−/− mice displayed a higher respiratory exchange ratio. iBAT presented larger adipocytes in p27−/− HFD mice, accompanied by downregulation of both Glut1 and uncoupling protein 1 (UCP1) in parallel with defective insulin signalling. Moreover, p27−/− HFD mice exhibited impaired response to cold exposure, characterized by a reduced iBAT 18FDG uptake and difficulty to maintain body temperature when exposed to cold compared to WT HFD mice, suggesting reduced thermogenic capacity. These data suggest that p27 could play a role in BAT activation and in the susceptibility to develop obesity and insulin resistance.


Introduction
Over recent decades overweight/obesity has become a worldwide health concern [1], having rapidly increased in prevalence in the last 30 years [2]. Many factors are involved in the development of obesity, including genetics, epigenetics and lifestyle determinants [1,3,4]. Excessive fat accumulation in adipose tissue is related to a wide range of metabolic diseases, cardiovascular disorders, fatty liver and some types of cancer [5]. Therefore, more efficient therapeutic strategies targeting adipose tissue for prevention of obesity are needed [6].
Adipocytes are the main lipid storage cells in adipose tissue. White adipocytes store energy in the form of triglycerides and provide fatty acids and glycerol to other tissues, whereas brown adipocytes dissipate energy as heat in a process called thermogenesis [7], which is highly dependent on the activity of the uncoupling protein 1 (UCP1) [8]. Brownlike adipocytes have been described in the white adipose tissue (WAT), were named brite (browning-in-white) or beige, and respond to thermogenic stimuli [9]. Stimulating brown adipose tissue (BAT) activity and/or browning of WAT has been proposed as a strategy against overweight/obesity [10]. WAT expansion during obesity is due to an increase in the number (hyperplasia) and size (hypertrophy) of adipocytes [11]. In contrast, hyperplasia of BAT occurs mainly upon cold stimulation [12,13].
Hyperplasia and hypertrophy are mainly regulated by a family of proteins known as cyclin dependent kinases (CDKs), well-known for regulation of the cell cycle progression. CDKs are kinases whose activity depends on the formation of a complex with a regulatory subunit (cyclin) [14]. The CDK-cyclin complexes can be inhibited by CDK inhibitors (CDKIs) [15]. These CDKIs can be classified in two families: CIP/KIP inhibitors, which are able to inhibit a wide range of CDK-cyclin complexes, and the INK4 family, which specifically inhibit complexes formed by CDK4 or CDK6 [16]. Within the CIP/KIP family, CDKI 1B (CDKN1B), also known as p27 Kip1 , prevents the activation of the complexes formed by cyclin E-CDK2 and cyclin D-CDK4 by binding to them and stopping the cell cycle progression during the G1 phase [17,18]. A study showed that the overexpression of p27 in Chinese hamster ovary cells (CHO) cells increased cell size, and was followed by an increase in oxygen and glucose consumption [19]. Moreover, a recent study of our group has shown that p27 was overexpressed in the subcutaneous (sc)WAT of obese female mice and in the visceral WAT of obese men and women. In humans, higher expression of p27 in scWAT was associated with a larger area of this depot [20]. This suggests that p27 could have an important role in the regulation of the expansion of adipose tissue and its metabolism during obesity [20]. Interestingly, the p27 knockout mouse model presents a markedly increased white adipose tissue mass [21], and in preadipocytes, progression of hyperplasia occurs upon a decrease of p27 levels [22]. The absence of p27 also caused a moderate increase of brown fat mass (25% vs. WT) [21]. In contrast, the overexpression of p27 in adipose tissue led to an impairment of brown adipose activity tissue in mice, with a smaller BAT depot and less UCP1 [23]. Even though BAT accounts for a small proportion of whole-body fat mass, it has a crucial role in thermogenesis, body weight control and insulin sensitivity [24]. Therefore, it is necessary to expand in the mechanisms controlling BAT function. Additionally, several CDK complexes inhibited by p27 have been related to glucose and adipose tissue metabolism [25][26][27]. Although the literature points to p27 as a possible player in glucose and adipose tissue metabolism, its role is yet to be fully understood. In the present study we analysed the susceptibility of p27 −/− mice to develop HFD-induced obesity and insulin resistance. We also assessed the effects of whole-body p27 deficiency on BAT morphology and function, and its potential relationship with changes in energy expenditure, respiratory exchange ratio and the development of obesity.

p27 −/− Mice are More Prone to Develop Obesity Than WT Mice
Changes in body weight and body composition were assessed in WT and p27 knockout mice fed with a normal chow diet (NCD) or a high fat diet (HFD) for 10 weeks. A two-way ANOVA test was used to establish if the obtained results were a consequence of the effect of the genotype (p27 deficiency), of the diet, or of an interaction between both. Figure 1a shows that body weight gain was markedly enhanced by the lack of p27 independently of the type of diet. Body composition analysis (Figure 1b) revealed that the HFD significantly increased the body fat mass and, interestingly, the lack of p27 markedly stimulated body fat accumulation (weight and percentage), suggesting that p27 deficiency confers susceptibility to develop obesity. Although p27 null mice also exhibited a higher total lean mass content, the lean mass percentage was reduced, suggesting that the increase in fat mass is more Figure 1. Mice lacking p27 gain more weight and have larger WAT and BAT depots than WT mice. (a) Body weight gain after 10 weeks of study in WT and p27 -/-mice fed with normal chow diet (NCD) or a high fat diet (HFD). (b) Whole body fat and lean mass (in g or as percentage of body weight) in WT and p27 -/-mice fed with NCD or HFD for 10 weeks. (c) Weights of different adipose tissue depots: gonadal (gWAT) and inguinal (iWAT) and interscapular brown (iBAT) collected after the sacrifice in WT and p27 -/-mice fed with NCD or HFD for 10 weeks. (d) Weights of different organs collected after the sacrifice in WT and p27 -/-mice fed with NCD or HFD for 10 weeks. Data are expressed as mean ± SEM (n = [11][12][13][14][15][16][17][18][19][20][21][22][23][24][25]. The legend above subfigures shows the result of the two-way ANOVA analysis (* p < 0.05, ** p < 0.01, *** p < 0.001, ns: non-significant). Gen, genotype; GenxDiet, interaction between genotype and diet. If an interaction (GenxDiet) was found, differences between groups were analysed by Student's t-test. ### p < 0.001 vs. WT HFD. $ p < 0.05 vs. p27 -/-NCD.

p27 −/− Mice Exhibit Increased Respiratory Exchange Ratio
To explore the possible causes of the increased susceptibility to diet-induced obesity of the p27 −/− mice, food intake ( Figure S1), whole-body oxygen consumption (Figure 2a), energy expenditure ( Figure 2b) and the respiratory exchange ratio (RER) (Figure 2c) were measured. In this regard, there were no significant differences on daily food intake between WT and p27 −/− mice, independently of feeding with NCD or HFD ( Figure S1). Interestingly, p27 −/− mice showed a tendency for reduced oxygen consumption and energy expenditure during the dark phase (Figure 2a,b). As expected, the animals fed with the HFD exhibited a reduced RER and, conversely, p27 deficiency accounted for an increased total RER (Figure 2c), suggesting that p27 −/− mice may display a lower lipid oxidation capacity.

p27 -/-Mice Exhibit Increased Respiratory Exchange Ratio
To explore the possible causes of the increased susceptibility to diet-induced obesity of the p27 -/-mice, food intake ( Figure S1), whole-body oxygen consumption (Figure 2a), energy expenditure (Figure 2b) and the respiratory exchange ratio (RER) (Figure 2c) were measured. In this regard, there were no significant differences on daily food intake between WT and p27 -/-mice, independently of feeding with NCD or HFD ( Figure S1). Interestingly, p27 -/-mice showed a tendency for reduced oxygen consumption and energy expenditure during the dark phase (Figure 2a,b). As expected, the animals fed with the HFD exhibited a reduced RER and, conversely, p27 deficiency accounted for an increased total RER (Figure 2c), suggesting that p27 -/-mice may display a lower lipid oxidation capacity. . Legend above subfigures shows the result of the two-way ANOVA analysis; * p < 0.05, *** p < 0.001, ns: non-significant. Gen, genotype; GenxDiet, interaction between genotype and diet.

Mice Lacking p27 Are More Susceptible to HFD-Induced Insulin Resistance
It is well established that obesity plays a critical role in impairing glucose homeostasis and insulin resistance [28]. Therefore, we decided to analyse blood glucose and insulin levels. HFD in general elevated fasting serum glucose levels, which were not significantly affected by the lack of p27 (Figure 3a). Concerning serum insulin levels, there was a significant interaction between the effects of the HFD and the lack of p27 (Figure 3b). Indeed, p27 -/-NCD mice had elevated fasting serum insulin levels compared to the WT NCD mice, and WT HFD mice also tended to have increased insulin levels, suggesting the . Legend above subfigures shows the result of the two-way ANOVA analysis; * p < 0.05, *** p < 0.001, ns: non-significant. Gen, genotype; GenxDiet, interaction between genotype and diet.

Mice Lacking p27 Are More Susceptible to HFD-Induced Insulin Resistance
It is well established that obesity plays a critical role in impairing glucose homeostasis and insulin resistance [28]. Therefore, we decided to analyse blood glucose and insulin levels. HFD in general elevated fasting serum glucose levels, which were not significantly affected by the lack of p27 (Figure 3a). Concerning serum insulin levels, there was a significant interaction between the effects of the HFD and the lack of p27 (Figure 3b). Indeed, p27 −/− NCD mice had elevated fasting serum insulin levels compared to the WT NCD mice, and WT HFD mice also tended to have increased insulin levels, suggesting the development of insulin resistance. Strikingly, p27 −/− HFD mice presented similar insulin levels to WT NCD mice ( Figure 3b); however, the glucose levels were not normalized, suggesting that the reduced insulin levels in the p27 null mice fed with a HFD are probably the consequence of deficient pancreatic insulin production as a result of a more aggravated insulin resistance. To further investigate this possibility, an insulin tolerance test (ITT) and a glucose tolerance test (GTT) were performed. In the ITT (Figure 3c), no differences were noticed between WT NCD and p27 −/− NCD groups. As expected, the WT HFD group exhibited impaired insulin response compared to WT NCD. Furthermore, p27 −/− HFD mice displayed severe insulin resistance when compared to the WT HFD mice (Figure 3c), suggesting that the lack of p27 interferes with insulin signalling in the context of obesity. The differences were not as marked in the GTT (Figure 3d), which showed that the p27 −/− HFD mice tended to be more glucose intolerant than the WT HFD mice. However, there was no difference between the WT NCD and p27 −/− NCD regarding glucose tolerance. No influence of the p27 deficiency was noticed on other serum parameters such as triglycerides, total cholesterol, LDL-cholesterol, HDL-cholesterol, β hydroxybutyrate, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) ( Table 1). development of insulin resistance. Strikingly, p27 -/-HFD mice presented similar insulin levels to WT NCD mice ( Figure 3b); however, the glucose levels were not normalized, suggesting that the reduced insulin levels in the p27 null mice fed with a HFD are probably the consequence of deficient pancreatic insulin production as a result of a more aggravated insulin resistance. To further investigate this possibility, an insulin tolerance test (ITT) and a glucose tolerance test (GTT) were performed. In the ITT (Figure 3c), no differences were noticed between WT NCD and p27 -/-NCD groups. As expected, the WT HFD group exhibited impaired insulin response compared to WT NCD. Furthermore, p27 -/-HFD mice displayed severe insulin resistance when compared to the WT HFD mice (Figure 3c), suggesting that the lack of p27 interferes with insulin signalling in the context of obesity. The differences were not as marked in the GTT (Figure 3d), which showed that the p27 -/-HFD mice tended to be more glucose intolerant than the WT HFD mice. However, there was no difference between the WT NCD and p27 -/-NCD regarding glucose tolerance. No influence of the p27 deficiency was noticed on other serum parameters such as triglycerides, total cholesterol, LDL-cholesterol, HDL-cholesterol, β hydroxybutyrate, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) ( Table 1).  (c) An insulin tolerance test (ITT) was performed after 6 h of fasting, and then mice were injected i.p. with 0.375 mU/g of body weight of rapid insulin and blood glucose levels were measured before and 30, 60 and 120 min after injection. (d) A glucose tolerance test performed after overnight fasting, and then mice were injected i.p. with 1 g/kg of body weight of glucose and blood glucose levels were measured before and 30, 60 and 120 min after injection. Data are expressed as mean ± SEM (n = 5-27). The legend above subfigures shows the result of the two-way ANOVA analysis; * p < 0.05, ** p < 0.01, *** p < 0.001, ns: non-significant. If an interaction (GenxDiet) was found, differences between groups were analysed by Student's t-test. * p < 0.05, ** p < 0.01 vs. WT NCD; ## p < 0.01, ### p < 0.001, a p = 0.05 vs. WT HFD. $ p < 0.05, $$ p < 0.01, $$$ p < 0.001 vs. p27 −/− NCD. Gen, genotype; AUC, Area Under the Curve.

iBAT of p27 −/− HFD Mice Exhibits Whitening and an Impaired Insulin Signalling
BAT activation plays an important role in weight gain control because obesity is associated with decreased activity in BAT [29,30]. The BAT depot is also relevant for glucose homeostasis and insulin sensitivity [31]. Therefore, as the insulin resistance was exacerbated by the HFD in the mice lacking p27 and the NCD had no effect, we focused on studying the effect of the absence of p27 in combination with HFD. The interscapular BAT (iBAT) from the p27 −/− HFD mice was heavier ( Figure 1c) and visually larger than their WT HFD counterparts (Figure 4a). Interestingly, the histological examination of the iBAT revealed a different morphology of the brown adipocytes between p27 −/− HFD and WT HFD mice, with larger and had bigger lipid droplets in the p27 deficient mice, confirmed by the histological analysis of the brown adipocyte size (Figure 4b,c). This suggests that p27 deficiency promotes brown adipocyte hypertrophy. However, considering the role of p27 in cell cycle regulation, we cannot rule out that p27 deficiency can also affect hyperplasia. Indeed, the mRNA expression of the Ccna, Ccne and Cdk2, which are involved in cell cycle progression and inhibited by p27, was significantly increased in p27 −/− HFD mice compared to WT HFD mice ( Figure S2). Moreover, we tested the mRNA expression levels of two proliferation markers Ki67 [13] and Tpx2 [32], which were increased in the iBAT of p27 −/− HFD mice, suggesting that the increased size of iBAT observed in p27 −/− mice is secondary to both hyperplasia and hypertrophy ( Figure S2). To obtain a broader image of the condition of the iBAT, the expression of genes relevant for different BAT functions was measured ( Figure 4d). Interestingly, Pparg was augmented in p27 −/− HFD mice, suggesting increased adipogenesis. The lipogenesis-related gene, Fas, was downregulated and Scd1 showed a similar tendency in p27 −/− HFD, probably due to the higher content of stored fat. Regarding fatty acid oxidation, Acox levels were downregulated, while no changes were observed in Cpt1a or in Ppara. There were no significant changes for genes typically involved in BAT development and activity (Prdm16, Dio2, Pgc1a). On the other hand, the glucose transporter Glut1 was dramatically downregulated in p27 −/− HFD mice, but Glut4 presented no change in these mice. However, a diminished expression of UCP1 at protein level was observed in iBAT of p27 −/− HFD mice (Figure 4e), which could lead to impaired thermogenic capacity. Although, the mRNA levels of the insulin-dependent glucose transporter Glut4 were not affected by p27 deficiency, it is important to note that the p27 −/− HFD mice exhibited impaired insulin response in iBAT, as demonstrated by the reduced phosphorylation of AKT, and therefore a reduced pAKT S473 /AKT ratio (Figure 4f). All these data suggest that p27 −/− HFD mice may display impaired glucose uptake in BAT, which could also account for a defective thermogenic response. Similar to the results observed in iBAT, histological analysis of the iWAT revealed larger adipocytes in the p27 −/− HFD mice ( Figure S3a,b). Gene expression analysis revealed a marked downregulation of Ucp1 and Glut1 in this depot ( Figure S3c), which may also suggest a reduced capacity in the adaptive thermogenesis of this depot. Interestingly, the histological analysis of the liver and the determination of the hepatic triglyceride content, also suggest that p27 −/− HFD mice are more prone to accumulate fat and develop liver steatosis ( Figure S4a,b). thermogenic capacity. Although, the mRNA levels of the insulin-dependent glucose transporter Glut4 were not affected by p27 deficiency, it is important to note that the p27 -/-HFD mice exhibited impaired insulin response in iBAT, as demonstrated by the reduced phosphorylation of AKT, and therefore a reduced pAKT S473 /AKT ratio (Figure 4f). All these data suggest that p27 -/-HFD mice may display impaired glucose uptake in BAT, which could also account for a defective thermogenic response. Similar to the results observed in iBAT, histological analysis of the iWAT revealed larger adipocytes in the p27 -/-HFD mice ( Figure S3a,b). Gene expression analysis revealed a marked downregulation of Ucp1 and Glut1 in this depot ( Figure S3c), which may also suggest a reduced capacity in the adaptive thermogenesis of this depot. Interestingly, the histological analysis of the liver and the determination of the hepatic triglyceride content, also suggest that p27 -/-HFD mice are more prone to accumulate fat and develop liver steatosis ( Figure S4a,b).

p27 −/− HFD Mice Have a Less Functional BAT and a Diminished Thermogenic Capacity
Since there was a decrease in UCP1 and Glut1 as well as impaired insulin signaling observed in the BAT of p27 −/− HFD group, pointing towards a compromised thermogenic capacity in these mice, their response to cold exposure was explored. Upon the cold challenge, the p27 −/− HFD mice could not maintain their body temperature as did the WT HFD (Figure 5a), and after 2 h of cold exposure showed a higher decrease of body temperature (Figure 5b). Additionally, to better evaluate BAT functionality, a microPET study was carried out. 18 F-fluorodeoxyglucose ( 18 F-FDG) uptake was evaluated after one hour of cold exposure, showing a significant decrease in the 18 F-FDG uptake of iBAT area in the p27 −/− HFD mice (Figure 5c), suggesting that the iBAT of the p27 −/− HFD mice had a lower capacity of taking up glucose than the WT HFD mice, which could be the reason for the defective response to adapt to a cold exposure challenge.
Since there was a decrease in UCP1 and Glut1 as well as impaired insulin signaling observed in the BAT of p27 -/-HFD group, pointing towards a compromised thermogenic capacity in these mice, their response to cold exposure was explored. Upon the cold challenge, the p27 -/-HFD mice could not maintain their body temperature as did the WT HFD (Figure 5a), and after 2 h of cold exposure showed a higher decrease of body temperature (Figure 5b). Additionally, to better evaluate BAT functionality, a microPET study was carried out. 18 F-fluorodeoxyglucose ( 18 F-FDG) uptake was evaluated after one hour of cold exposure, showing a significant decrease in the 18 F-FDG uptake of iBAT area in the p27 -/-HFD mice (Figure 5c), suggesting that the iBAT of the p27 -/-HFD mice had a lower capacity of taking up glucose than the WT HFD mice, which could be the reason for the defective response to adapt to a cold exposure challenge. Rectal temperature measured at different timepoints during cold exposure to 4 °C including AUC (area under the curve) (a) and changes in rectal temperature before and after 2 h of cold exposure (b) in WT and p27 -/-mice fed with HFD for 10 weeks (n = 5-10). (c) 18 F-FDG uptake in iBAT using microPET 1 h after 18 F-FDG injection in WT and p27 -/-male mice fed with HFD for 10 weeks and preexposed for 1 h at 4 °C. Left panel: representative microPET coronal slices of the iBAT area (red arrow) co-registered with a CT 3D image of another animal used as anatomical reference. Right panel: maximum standardized uptake value (SUVmax) (n = 3). Data are expressed as mean ± SEM. # p < 0.05, ## p < 0.01, vs. WT HFD.

Discussion
This study supports that p27 deficiency causes increased susceptibility to obesity, especially when mice are subjected to an obesogenic diet. This was demonstrated by the higher percentage of whole-body fat mass, and the larger size of all fat depots analysed (gWAT, iWAT and BAT).
The increased size of the fat depots could be expected due, in part, to the absence of an appropriate regulation of the cell cycle progression, which would lead to higher cell proliferation rates, and greater growth and organomegaly, as previously described [33]. Indeed, it has been previously reported that loss of CDK inhibitors also produces adipocyte hyperplasia and obesity [21]. Furthermore, our current data suggest that the increase of adipose tissue mass could be due to the presence of hypertrophic adipocytes. Indeed, the analysis of iWAT and iBAT morphology clearly showed a higher adipocyte hypertrophy in the p27 null mice fed with the HFD compared to the WT mice.

Discussion
This study supports that p27 deficiency causes increased susceptibility to obesity, especially when mice are subjected to an obesogenic diet. This was demonstrated by the higher percentage of whole-body fat mass, and the larger size of all fat depots analysed (gWAT, iWAT and BAT).
The increased size of the fat depots could be expected due, in part, to the absence of an appropriate regulation of the cell cycle progression, which would lead to higher cell proliferation rates, and greater growth and organomegaly, as previously described [33]. Indeed, it has been previously reported that loss of CDK inhibitors also produces adipocyte hyperplasia and obesity [21]. Furthermore, our current data suggest that the increase of adipose tissue mass could be due to the presence of hypertrophic adipocytes. Indeed, the analysis of iWAT and iBAT morphology clearly showed a higher adipocyte hypertrophy in the p27 null mice fed with the HFD compared to the WT mice.
Interestingly, our study showed that p27 deficient mice presented a trend for a moderate decrease of oxygen consumption and energy expenditure, which are typically observed in obesity [34]. Additionally, the increased RER levels observed in the p27 −/− mice suggest a reduced capacity for lipid oxidation, which could partly explain why fat accumulation in these mice is greater [35,36].
It is already known that BAT activity is critical for the improvement of insulin sensitivity and glucose metabolism [37]. In addition a lower BAT activity can contribute to bringing on obesity [38,39]. Therefore, a plausible explanation for the phenotype observed in the p27 −/− HFD mice could be defective BAT activity. First, we observed that p27 −/− HFD mice had a light brown coloured and larger and heavier iBAT depot with greater adipocyte size. The "whitening" of iBAT has already been described and causes a decline in its activity [40]. Expression analysis also revealed alterations in genes involved in peroxisomal fatty acid oxidation, such as Acox. Indeed, ACOX-dependent peroxisomal fatty acid oxidation can modulate BAT activity [41]. The downregulation of Glut1 could also compromise the glucose uptake capacity in the iBAT of p27 −/− HFD mice, as it is highly dependent on GLUT1 transcription, synthesis and translocation to the membrane [42]. In fact, glucose is an important fuel source for BAT [43] and, as a result of the lowered uptake, BAT function is diminished. In addition, we explored insulin pathway activation, and the iBAT of the p27 −/− HFD mice exhibited a diminished pAKT S473 /AKT ratio in response to acute insulin challenge, suggesting an impairment of the insulin signalling in this tissue, probably as a consequence of the higher fat accumulation [44]. Although the mRNA levels of the insulin-dependent glucose transporter Glut4 were not altered in iBAT of the p27 null mice, the alteration in insulin signalling suggests that the translocation of GLUT4 to the membrane to promote glucose uptake into the cell [45] could be also affected. Thus, both insulin-dependent and insulin-independent glucose transporters could be impaired in p27 null mice, which may explain the reduced thermogenic capacity in these animals.
Curiously, the relation between p27 and glycaemia is rather contradictory. In this regard, a study [46] evaluated the effect on glucose metabolism of the deficiency of p27, finding that young p27 −/− mice fed with NCD had enhanced growth of pancreatic beta cells and elevated insulin levels, as occurred in the present study, together with improved glucose tolerance. In contrast, other studies have shown that 10-to 12-week-old p27 −/− mice were glucose intolerant and insulin insensitive [21], defining p27 as a crucial mediator in beta cell generation [47]. Our study shows no changes in glucose or insulin tolerance in the p27 −/− mice fed on NCD, but the dramatic impairment of insulin response to the ITT observed in the p27 −/− HFD mice could be a consequence of the exhaustion of the pancreatic cells, as a significant reduction of the insulin circulating levels was observed in these mice [48] in combination with an increased resistance to insulin.
Moreover, downregulation of UCP1 levels in the iBAT of the p27 −/− HFD mice, which is critical for BAT activity and body temperature maintenance [49], could compromise the cold adaptation capacity of p27 −/− HFD mice. In fact, these animals had a reduced capacity to keep their body temperature after cold exposure. This observation was combined with a lower iBAT glucose uptake capacity of the p27 −/− HFD in response to acute cold exposure. These facts suggest that the impairment of iBAT observed in p27 −/− mice could affect whole-body oxygen consumption [50], favouring increased body fatness and impaired insulin sensitivity and glucose metabolism. Indeed, a less active BAT is linked to a higher susceptibility to develop obesity and insulin resistance, and studies propose BAT activation as a strategy to fight these diseases [24]. However, studies assessing BAT activity and function in humans are complex, and cannot be carried out easily, therefore studies in mice have more relevance.
Our results suggest that the lack of p27 impairs iBAT activity, which apparently contrasts with a previous study reporting that transgenic mice with adipose tissue-specific overexpression of p27 showed a defect of body temperature maintenance when exposed to cold temperature. This circumstance was accompanied by a lower response of BAT to β-adrenergic stimulation [23]. The fact that the adipose-specific overexpression of p27 results in a similar loss of BAT activity than the full-body ablation of p27, suggests that the impairment of BAT activity in these mice could be secondary to the alterations induced by the lack of p27 in other tissues/systems with key roles in controlling the thermogenic activity of BAT, such as the sympathetic nervous system [51], which, if affected by the lack of p27, could alter the normal-functioning of this fat depot. In fact, it has already been reported that another cell cycle regulator, CDK4, modulates BAT thermogenesis via regulation of the sympathetic innervation of BAT through hypothalamic nuclei [52]. Furthermore, p27 seems to be relevant in the hypothalamus [53] and for the development of the central nervous system [54] and its absence may disrupt its normal functioning. Thus, these observations suggest that cell cycle regulators might regulate BAT activity also acting at a nervous system level. However, further future studies are necessary to better clarify this. Indeed, the main limitation of this study is that a whole-body knockout mouse model was used. Therefore, it cannot be discerned whether the changes observed in iBAT activity and metabolic phenotype are caused directly by the lack of p27 in adipose tissue, as effects in other non-adipose tissues are also probably involved. To clarify the role of p27 in iBAT function and its relationship with whole-body metabolism, a study including brown adipose tissue-specific p27 knockout mice should be carried out. It would be of interest to characterize this in parallel if the observations found at mRNA levels in p27 null mice are also observed at protein or enzyme activity level.
Taken together, our data show that the whole-body deficiency of p27 favors the development of obesity, increasing the size of all fat depots independently of the diet. The deficiency of p27 combined with a HFD causes severe insulin resistance. In iBAT, p27 deficiency, together with HFD, promotes adipocyte hypertrophy, while it decreases the mRNA levels of the glucose transporter glut1 and disrupts the insulin signaling pathway (pAKT/AKT ratio). Moreover, p27 −/− HFD mice display lower levels of the thermogenic protein UCP1 in iBAT and an impaired response to cold exposure in parallel with a difficulty to maintain body temperature, suggesting reduced thermogenic capacity. These data suggest that the cell cycle inhibitor p27 could play a role in adipose tissue metabolism, especially in BAT. This highlights the importance of investigating the metabolic functions of p27, and other cell cycle regulators, with respect to their function in BAT and the development of obesity.
At the endpoint, body composition was measured by magnetic resonance technology (EchoMRI-100-700; Echo Medical Systems, Houston, TX, USA), as previously described [57]. Mice were sacrificed, blood was extracted, and serum samples were obtained. Different tissues and organs (gWAT, iWAT, iBAT, liver, spleen, gastrocnemius, heart, pancreas, and kidney) were immediately collected, weighed and snap-frozen in liquid nitrogen and kept at −80 • C for further determinations, or fixed with 4% formalin to perform histological analyses. Experiments and analyses were performed in both male and female mice, except for the in vivo study of iBAT activation by positron emission tomography (PET), which was carried out only in males All experimental procedures were performed under protocols approved by the University Ethics Committee for the use of laboratory animals according to the national and institutional guidelines for animal care and use (protocols: 048-17, E41-19 (048-17E2), 073-22, 049-17, 072c-22).

Insulin and Glucose Tolerance Test
For the ITT, a week prior to sacrifice mice were fasted for 6 h, and baseline blood glucose levels were measured with a standard glucometer (Accu-Check Advantage blood glucose meter, Roche, Manheim, Germany). Animals were then injected (i.p. 0.375 mU/g of body weight) with human rapid insulin (Actrapid ® Innolet ® , Novo Nordisk A/S, Bagsvaerd, Denmark) and blood glucose levels were measured 30, 60 and 120 min af-ter injection. For the GTT, mice were fasted overnight and after measuring fasted blood glucose levels, mice were injected i.p. with 1 g glucose/kg of body weight (Merck KGaA, Darmstadt, Germany) and blood glucose levels were then measured 30, 60 and 120 min after the injection.

Biochemical Analysis
Fasted glucose serum levels were measured with a standard glucometer (Accu-Check Advantage blood glucose meter, Roche) and serum levels of total cholesterol, HDLcholesterol, triglyceride, β hydroxybutyrate, AST and ALT were determined using a Pentra C200 autoanalyser (Roche Diagnostic, Basel, Switzerland) following the manufacturer's instructions, as described elsewhere [58]. LDL-cholesterol was calculated using Friedewald's equation [59]. Serum insulin was measured using a commercially available ELISA kit (Mercodia, Uppsala, Sweden), according to the manufacturer's guidelines.

Insulin Signalling Assessment
To evaluate the activation of the insulin signalling pathway, overnight fasted mice were injected i.p. with human rapid insulin (Actrapid ® Innolet ® , Novo Nordisk A/S) (1 mU/g body weight) and sacrificed 15 min later. Then, fat depots were excised and frozen in liquid nitrogen. The insulin signalling pathway was evaluated by measuring the phosphorylation of AKT S473 vs. total AKT levels by Western blot (see below).

Real-Time Quantitative PCR
Total RNA from mouse samples was extracted with TRIzol™ reagent (Invitrogen, ThermoFisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. RNA quality and concentrations were measured by Nanodrop Spectrophotometer ND1000 (Nanodrop Technologies, Inc., Wilmington, NC, USA). RNA (1-5 µg) was then incubated with DNase I (RapidOut DNA Removal kit, Thermo Fisher Scientific) for 30 min at 37 • C and reverse transcribed to cDNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems; Waltham, MA, USA, Thermo Fisher Scientific) according to the manufacturer's instructions in a Touch PCR system (C1000, BIO-RAD, Hercules, CA, USA). Real-time PCR was performed using the Touch Real-Time PCR System (CFX384, BIO-RAD). The expression of genes was determined using Power SYBR Green PCR Master Mix (BIO-RAD) or TaqMan master mix (Applied Biosystems). SYBR Green primers were tested with Primer-Blast software (National Center for Biotechnology Information, Bethesda, MD, USA; https://www.ncbi.nlm.nih.gov/tools/primer-blast, accessed on 13 January 2023). Primer sequences are shown in Table 2. For relative quantitation of gene expression, the comparative Ct method was used [2 −∆∆Ct , where ∆Ct represents the difference in threshold cycle between the target and a housekeeping gene (36b4)] [60].

Acute Cold Exposure Challenge
Mice were housed at 4 • C for 2 h, and rectal temperature was measured at basal state and every 30 min upon cold exposure rectal thermometer probe (Panlab, Barcelona, Spain) [61].

In Vivo Study of iBAT Activation by Positron Emission Tomography (PET)
iBAT activation was studied in vivo by microPET with the radiotracer 18 F-FDG. BAT was stimulated by cold exposure for 1 h at 4 • C [62]. After cold exposure, 18 F-FDG (10.1 ± 0.9 MBq) was injected through the tail vein. MicroPET static images were acquired at room temperature 1 h after 18 F-FDG injection in a small animal microPET scanner (Mosaic, Philips, Amsterdam, The Netherlands). For each study, animals were anesthetized with 2% isoflurane in 100% O 2 gas. All studies were analysed using PMOD software (PMOD Technologies Ltd., Adliswil, Switzerland). For semiquantitative analysis, 18 F-FDG uptake by iBAT was evaluated drawing volume-of-interest (VOIs) on coronal PET images. From each VOI, the maximum standardized uptake value (SUVmax) was calculated using the formula SUV = (tissue activity concentration (Bq cm 3 )/injected dose (Bq)] × body weight (g)).

Indirect Calorimetry and Activity Measurements
Animals were placed in a comprehensive laboratory animal monitoring system for measurements at 20 • C attached to custom-built oxygen and carbon dioxide monitoring system (PhenoMaster, TSE Systems, Berlin, Germany). Airflow rates were 400 mL/min for measurements of oxygen concentration, and carbon dioxide concentrations in room air and air leaving each the cage were measured every 20 min as described before [63]. Energy expenditure (EE) was calculated from the amount of oxygen consumed (VO 2 ) and the amount of carbon dioxide produced (VCO 2 ) using the equation: EE (J) = 15.818 VO 2 + 5.176 VCO 2 . The RER was calculated as the ratio of VCO 2 /VO 2 . Food intake of each animal was also evaluated using these metabolic cages. Indirect calorimetry studies were carried out in the LAFEMEX Laboratory (Universidad Rey Juan Carlos, Alcorcón, Spain).

Histological Analysis
Sections of paraffin-embedded tissues were stained with H&E. Representative images were taken using a Nikon eclipse e800 microscope and a Nikon DS-Qi1Mc camera at 20× and 40× (Nikon, Tokyo, Japan). Adipocyte cell size was determined using ImageJ software [64].

Liver Triglyceride Content Assay
Liver pieces (50 mg) were homogenized in buffer (Tris 10 mM, EDTA 2 mM, Sacarose 0.25 M; pH = 7) to determine TG content of the samples. Following the manufacturer's instructions, the InfinityTM Triglycerides Liquid Stable Reagent (Thermo Fisher Scientific) was used to hydrolyze the TGs of the samples. This method is based on that of Wako with the modifications of McGowan et al. [65] and Fossati et al. [66]. The triglyceride concentration was normalized to protein content.

Statistical Analysis
Statistical analyses were carried out with GraphPad Prism 9 software (Graph-Pad Software, La Jolla, CA, USA). Data are presented as mean ± SEM. The effects of "Genotype" (p27 deficiency) and "Diet" were analysed by two-way ANOVA. If an interaction was found between the two factors (Genotype × Diet), Student's t-test was used for further analysis of differences between groups. Comparisons between two groups were analysed by Student's t-tests or Mann Whitney's tests according to parametric or nonparametric distribution, respectively. Differences between groups were set up as statistically significant at p < 0.05.