Enhanced C/EBPβ function promotes hyperplastic versus hypertrophic fat tissue growth and prevents steatosis in response to high-fat diet feeding

Chronic obesity is correlated with severe metabolic and cardiovascular diseases as well as with an increased risk for developing cancers. Obesity is usually characterized by fat accumulation in enlarged – hypertrophic – adipocytes that are a source of inflammatory mediators, which promote the development and progression of metabolic disorders. Yet, in certain healthy obese individuals, fat is stored in metabolically more favorable hyperplastic fat tissue that contains an increased number of smaller adipocytes that are less inflamed. In a previous study, we demonstrated that loss of the inhibitory protein-isoform C/EBPβ-LIP and the resulting augmented function of the transactivating isoform C/EBPβ-LAP promotes fat metabolism under normal feeding conditions and expands health- and lifespan in mice. Here, we show that in mice on a high-fat diet, LIP-deficiency results in adipocyte hyperplasia associated with reduced inflammation and metabolic improvements. Furthermore, fat storage in subcutaneous depots is significantly enhanced specifically in LIP-deficient male mice. Our data identify C/EBPβ as a regulator of adipocyte fate in response to increased fat intake, which has major implications for metabolic health and aging.


Introduction
Nutrient overload, particularly in combination with a sedentary lifestyle, is the main cause of the increasing incidence of obesity we are facing today. In most cases, chronic obesity provokes the development of metabolic diseases like insulin resistance, type 2 diabetes (T2D), non-alcoholic fatty liver disease (NFALD), and non-alcoholic steatohepatitis (NASH) (Longo et al., 2019). The surplus of fat is stored in the white adipose tissue (WAT) largely without an increase in adipocyte number, resulting in an increase in fat cell size. This hypertrophy is accompanied with reduced vascularization and oxygen

Results
Separate cohorts of male and female Cebpb ΔuORF mice and wild-type (wt) control littermates on a C57BL/6 J background were fed a high-fat diet (HFD; 60% fat) or standard chow (10% fat) for a period of 19 weeks. In males, both genotypes gained weight over the whole experimental period yet with significantly lower body weights for Cebpb ΔuORF mice, suggesting that they are partially protected from HFD induced weight gain ( Figure 1A). Both the food intake and the energy efficiency (energy that was extracted from the food during digestion) were similar in Cebpb ΔuORF and wt males (Figure 1figure supplement 1A, B). Lean mass and fat mass body composition was determined by computer tomography (CT) after 19 weeks of HFD feeding. On standard chow, fat mass of Cebpb ΔuORF males was reduced compared to wt males ( Figure 1B) as we showed before (Zidek et al., 2015). Unexpectedly, we observed a slight but significantly increased overall fat mass volume in the Cebpb ΔuORF males compared to wt controls ( Figure 1B), which correlates with a relative reduction in lean mass ( Figure 1-figure supplement 1C). There was no significant difference in the amount of visceral fat mass between wt and Cebpb ΔuORF males under HFD ( Figure 1C). However, the HFD fed Cebpb ΔuORF males had a significantly increased subcutaneous fat mass compared to the wt controls ( Figure 1D). Taken together, the data show that Cebpb ΔuORF male mice store more fat upon HFD feeding compared to wt littermates, and that this extra fat is mainly stored in the subcutaneous fat depot. Similar to male mice, the body weight gain of Cebpb ΔuORF females after 19 weeks on HFD was smaller compared to the weight gain of wt females ( Figure 1E). However, different from the male mice, in females this was associated with a reduced accumulation of fat particularly in the subcutaneous depot in Cebpb ΔuORF compared to wt females as was determined by fat tissue weight ( Figure 1F, G). Next, we compared histological sections from visceral fat of HFD fed Cebpb ΔuORF mice and wt littermates and observed that the adipocyte size in the Cebpb ΔuORF mice for both sexes is significantly smaller compared to the wt mice (Figure 2A, B). Calculated from visceral fat volume (CT analysis) of males or visceral fat weight of females and the average adipocytes area (histology) per mouse, the number of adipocytes is approximately 3 times higher in Cebpb ΔuORF males and 1.5 times higher in Cebpb ΔuORF females compared to their wt littermates.
Adipose tissue composed of small adipocytes is metabolically more active and better supplied with oxygen, and its inflammatory state is usually lower than that of enlarged adipocytes (Ghaben and Scherer, 2019). We therefore analyzed the inflammatory state of visceral white adipose tissue (WAT) by determining the expression of the macrophage marker CD68 and the inflammatory cytokines TNFα, MCP1, IL-1 and IL-6 using quantitative PCR (qPCR) and immunohistochemistry (anti-CD68). Male wt mice on HFD show a strong increase in CD68 mRNA expression compared to wt males on ND indicating increased macrophage infiltration. In contrast, CD68 mRNA expression is much lower in the visceral fat from HFD fed Cebpb ΔuORF males and not significantly different from ND fed mice ( Figure 3A). Accordingly, histological staining of visceral WAT derived from three different mice shows more pronounced macrophage infiltration in wt mice on HFD (19 weeks) compared to Cebpb ΔuORF males ( Figure 3B). In addition, expression of the inflammatory markers TNFα, MCP1, IL-1β, and IL-6 was significantly induced in the visceral fat of wt males on HFD ( Figure 3C). In contrast, in Cebpb ΔuORF (females, week 19; ND wt, n = 7; HFD wt, n = 4; ND and HFD ΔuORF, n = 6). (G) Subcutaneous fat weight (females, week 19; ND wt, n = 7; HFD wt, n = 4; ND and HFD ΔuORF, n = 6). All values are mean ± SEM. p-Values were determined with Student's t-test, *p < 0.05; **p < 0.01; ***p < 0.001.
The online version of this article includes the following source data and figure supplement(s) for figure 1: Source data 1. Raw data related to Figure 1A-G.   . Cebpb ΔuORF mice on high-fat diet (HFD) store fat in hyperplastic adipocytes. Histological hematoxylin and eosin (H&E) staining of epididymal WAT from (A) males (19 weeks HFD) and (B) females (19 weeks HFD). Quantification of the fat cell area is shown at the right (males: wt, n = 7; ΔuORF, n = 4; females: wt, n = 4; ΔuORF, n = 7; 12 adjacent cells are measured per mouse).
The online version of this article includes the following source data for figure 2: Source data 1. Raw data related to Figure 2A   males, HFD feeding did not significantly induce TNFα and IL-1β expression, and the induction of MCP1 was significantly lower compared to HFD fed wt males. Solely IL-6 levels were comparable between the two genotypes on HFD feeding. For female wt mice the mean value of induction of CD68 expression in response to HFD is high but strongly varies between the mice. Therefore, the lower value of CD68 expression in HFD fed Cebpb ΔuORF mice is not statistically significant yet shows much less variation ( Figure 4A). Notwithstanding, the anti-CD68 staining of visceral WAT did show infiltration by macrophages in wt females on HFD although to a lower extend compared to wt males, while in Cebpb ΔuORF females hardly any staining was observed ( Figure 4B). The inflammatory markers TNFα, MCP1 and IL-1 are all induced in wt and Cebpb ΔuORF mice on HFD feeding to similar extends and thus no differences are measured between the genotypes. No induction was observed for IL6 expression in both genotypes upon HFD feeding ( Figure 4C).
Together, these data demonstrate that Cebpb ΔuORF mice on HFD feeding store extra fat through an increase in adipocyte numbers (hyperplasia), which results in smaller sized adipocytes and reduced fat tissue inflammation in males. In Cebpb ΔuORF females, adipocytes are also smaller but consistent differences in the inflammation state of the visceral fat were not measured.
Adipocyte hypertrophy under obese conditions is associated with a limit in fat storage capacity of the adipocytes and enhanced lipolysis (Khan et al., 2009;Laurencikiene et al., 2011) The resulting increase in the concentration of fatty acids in the circulation causes lipid accumulation in non-fat tissues like liver, muscle and heart (Longo et al., 2019). In a previous study, we showed that Cebpb ΔuORF males on a C57Bl/6 genetic background are protected against age-related steatosis on a ND, compared to wt mice that do accumulate fat in the liver at an age of 8 months (Zidek et al., 2015). To investigate possible differences in HFD-induced steatosis, we stained histological sections of liver for fat accumulation. Livers of both male and female wt mice showed massive fat accumulation (steatosis) on HFD. Compared to the wt mice, fat accumulation was much lower in Cebpb ΔuORF mice of both sexes ( Figure 5A, B). In agreement with the differences in steatosis, the livers of wt females on HFD are significantly heavier than livers of Cebpb ΔuORF females while in males a trend towards heavier wt livers is visible ( Figure 5C, D). Fat accumulation on HFD was also lower in the heart and skeletal muscle of male mice (Figure 5-figure supplement 1A) and for both males and females the weight of the heart on HFD is higher in wt compared to Cebpb ΔuORF mice ( Figure 1-figure supplement 1B). Altogether, these data show that Cebpb ΔuORF mice are protected against steatosis in the liver and other organs in response to HFD.
Chronic obesity often results in the loss of glucose homeostasis (Abranches et al., 2015). We therefore analyzed glucose tolerance and insulin sensitivity in the HFD fed Cebpb ΔuORF mice and wt littermates. Glucose clearance from the circulation measured by intraperitoneal glucose tolerance test (IPGTT) was impaired in response to 7 weeks HFD feeding for the wt mice of both sexes, as is shown by a significantly increased area under the curve (AUC) ( Figure 6A, B). For the Cebpb ΔuORF male mice, the already significantly better glucose clearance on normal diet does not change on HFD ( Figure 6A). The female Cebpb ΔuORF mice on HFD show reduced glucose clearance in the IPGTT compared to ND but they perform significantly better than the HFD fed wt females ( Figure 6B). At the time of 7 weeks on HFD, both the wt and Cebpb ΔuORF mice of both sexes did not develop insulin insensitivity as measured by intraperitoneal insulin sensitivity test (IPIST) ( Figure 6C, D). Both the Cebpb ΔuORF males and females, however, generally performed better on IPIST than the wt mice.
In conclusion, our data show that Cebpb ΔuORF mice on HFD feeding perform better in a glucose tolerance test, are protected against steatosis and show a lower inflammatory status of WAT, although the latter is less evident in females. These metabolically favorable phenotypes of the Cebpb ΔuORF mice correlate with hyperplastic fat storage and in male mice with more efficient fat accumulation in the subcutaneous depot. levels of the inflammatory cytokines TNFα, MCP1, IL-1β, and IL6 measured in the visceral fat of Cebpb ΔuORF (ΔuORF) and wt male mice on either normal diet (ND) or HFD (19 weeks; wt: ND, n = 5; HFD, n = 6; ΔuORF: ND, n = 6 (for IL-6, n = 4, the results of two mice were excluded due to undetectable signal); HFD, n = 4). All values are mean ± SEM. p-Values were determined with Student's t-test, *p < 0.05; ***p < 0.001.
The online version of this article includes the following source data for figure 3: Source data 1. Raw data related to Figure 3A, C. wt: ND, n = 7; HFD, n = 4; ΔuORF: ND, n = 4 (the result from one mouse was excluded due to undetectable signal); HFD, n = 5). (B) Immunohistological staining of the visceral fat of Cebpb ΔuORF female mice (ΔuORF) and wt mice on HFD (19 weeks) using a CD68-specific antibody (arrow points to specific Figure 4 continued on next page C/EBPβ is a known transcriptional regulator of fat cell differentiation and function (Siersbaek and Mandrup, 2011). We have shown earlier that the truncated C/EBPβ isoform LIP inhibits adipocyte differentiation and that fibroblasts derived from Cebpb ΔuORF mice have an increased adipogenic differentiation potential (Zidek et al., 2015). We therefore analyzed the expression in the visceral WAT of the adipogenic transcription factors C/EBPα and PPARγ, the sterol regulatory element-binding protein 1 c (SREBP1c) as a key transcription factor for lipogenesis, and fatty acid synthase (FAS) as a key lipogenic enzyme, by quantitative PCR. In wt male mice all four transcripts are significantly lower expressed on HFD compared to ND ( Figure 7A). This generally corresponds to their protein levels as determined by immunoblotting, although the expression of PPARγ and SREBP1c varies considerably between the mice (Figure 7-figure supplement 1A). In the WAT of Cebpb ΔuORF males on ND, expression of the four transcripts is similar (C/EBPα and PPARγ) or higher compared to WAT from wt males on ND (SREBP1 and FAS, also shown in Zidek et al., 2015), and its reduction under HFD occurs to a lesser extend compared to wt mice ( Figure 7A). With the exception of C/EBPα, the transcript levels generally correspond to the protein levels, although here variations of in particular PPARγ and SREBP1 levels complicate interpretation (Figure 7-figure supplement 1A). For the wt female mice, only the transcript levels of FAS were downregulated upon HFD and for Cebpb ΔuORF mice only expression of PPARγ and FAS was significantly higher on HFD compared to wt females on HFD ( Figure 7B). The better maintained expression of FAS in HFD fed Cebpb ΔuORF females in the qPCR analysis however could not be recapitulated with immunoblotting ( Figure 7-figure supplement 1B).

Figure 3 continued
To examine whether HFD feeding is associated with changes in LAP/LIP expression, we analyzed extracts from WAT isolated from wt and Cebpb ΔuORF mice of both sexes under ND and HFD conditions. In wt males, both LAP and LIP isoforms were upregulated upon HFD feeding as shown in the immunoblot ( Figure 8A) and determined by quantification of blot signals from a cohort ( Figure 8B, C). The quantification reveals a small but significant decrease in the LAP/LIP ratio ( Figure 8B), indicating that the inhibitory function of LIP increases upon HFD in wt males. Due to the LIP-deficiency caused by the Cebpb ΔuORF mutation, the LAP/LIP ratio is very high for the Cebpb ΔuORF males and does not change upon HFD feeding (some residual LIP expression is usually visible in Cebpb ΔuORF mice due to leaky scanning over the not-optimal AUG-start codons for the LAP proteins). For the females, a significant increase in both LAP and LIP expression in response to HFD feeding was only observed in the Cebpb ΔuORF mice ( Figure 8D, F). However, no significant changes were measured in LAP/LIP expression ratios, despite an overall trend towards a lower LAP/LIP ratio on HFD in wt females ( Figure 8D, E). Taken together, LAP/LIP isoform ratios decline in response to HFD feeding due to higher increase of LIP expression which is significant for males but not for females.

Discussion
The Cebpb ΔuORF mutation prevents expression of the inhibitory C/EBPβ protein isoform LIP which results in unconstrained function of the C/EBPβ transactivator isoform LAP. In two previous reports, we have shown that Cebpb ΔuORF mice display metabolic improvements and a delay in the onset of age-related conditions, collectively resulting in an extended lifespan in females (Müller et al., 2018;Zidek et al., 2015). Here, we demonstrate that Cebpb ΔuORF mice are protected against the development of metabolic disturbances in response to HFD feeding. In males, this improved metabolic phenotype occurs although the total fat mass in Cebpb ΔuORF mice is increased in response to HFD to a greater extent than in wt mice. Our data indicate that two special features of the white adipose tissue (WAT) in Cebpb ΔuORF males contribute to these metabolic improvements. Firstly, Cebpb ΔuORF males on a HFD store the surplus of nutrients in fat depots that expand through hyperplasia; they increase the number of adipocytes and thus the individual cells have to store less fat. These smaller adipocytes staining). Histological sections from three individual mice per genotype are shown. (C) Relative mRNA expression levels of the inflammatory cytokines TNFα, MCP1, IL-1β, and IL6 measured in the visceral fat of Cebpb ΔuORF female mice (ΔuORF) and wt mice on either normal diet (ND) or HFD (19 weeks; wt: ND, n = 6; HFD, n = 4; ΔuORF: ND, n = 5; HFD, n = 5). All values are mean ± SEM. p-Values were determined with Student's t-test, *p < 0.05; **p < 0.01.

Figure 4 continued
The online version of this article includes the following source data and figure supplement(s) for figure 5: Source data 1. Raw data related to Figure 5C, D.   Intraperitoneal glucose tolerance test (IPGTT) with the calculated area under the curve (AUC) of Cebpb ΔuORF (A) male and (B) female (ΔuORF) and wt mice injected i.p. with glucose (2 g/kg) after a 16 hr fast (7 weeks; males: wt ND, n = 6; wt HFD, n = 9; ΔuORF ND, n = 6; ΔuORF HFD, n = 7; females: wt ND, n = 7; wt HFD, n = 6; ΔuORF are metabolically more active and less inflamed compared to the inflated wt adipocytes residing in a hypertrophic fat depot. Hypertrophic adipocytes are known to secrete inflammatory cytokines that promote insulin resistance and other metabolic disturbances (Reilly and Saltiel, 2017;Weisberg et al., 2003). Furthermore, since the number of adipocytes in hypertrophic fat tissue does not increase and the amount of fat that can be stored in an adipocyte is limited, fat starts to accumulate in ectopic tissues like liver or muscle, compromising metabolic health (Frasca et al., 2017). Accordingly, in wt males on HFD we observed pronounced inflammation and macrophage infiltration in the visceral WAT and severe steatosis. In contrast, Cebpb ΔuORF males on HFD are protected against these metabolic disturbances, which also correlated with better maintenance of glucose tolerance. We have shown earlier that the expression of genes related to fatty acid oxidation is enhanced in the liver of Cebpb ΔuORF mice accompanied by a significant increase in fatty acid oxidation (Zidek et al., 2015). This enhanced fat utilization likely contributes to the reduced lipid accumulation in the liver of Cebpb ΔuORF mice on HFD, and the healthier metabolic phenotype of Cebpb ΔuORF mice is presumably the result of the combination of an increase in WAT function and fat utilization. In addition, the Cebpb ΔuORF males store relatively more fat in the subcutaneous compartment than wt mice, which relieves the fat storage pressure for the visceral depots. Fat storage in the subcutaneous fat depot is associated with a better metabolic health status in humans and mice, while fat storage in the visceral fat depot is associated with insulin resistance and inflammation (Carey et al., 1997;McLaughlin et al., 2011;Tran et al., 2008). In contrast to the males, female Cebpb ΔuORF mice showed reduced fat accumulation in the subcutaneous fat depot upon HFD compared to wt mice, and the visceral fat depot showed a trend towards a reduced fat storage although this difference was not statistically significant ( Figure 1F and G). However, similar to the Cebpb ΔuORF males, the adipocyte cell size in the visceral fat from Cebpb ΔuORF females was significantly reduced and the calculated number of adipocytes was higher compared to wt females revealing increased adipocyte hyperplasia also in HFD fed Cebpb ΔuORF females. Accordingly, also the female mice showed an improved metabolic phenotype on HFD including reduced hepatic steatosis and better maintained glucose tolerance. The difference in inflammation between wt and Cebpb ΔuORF females was less pronounced compared to males and only visible in antibody staining of the macrophage marker CD68 indicating reduced macrophage infiltration in the visceral fat of Cebpb ΔuORF females. However, macrophage infiltration seemed to be less pronounced in wt females on HFD than in wt males based on the CD68 immunohistological staining (compare Figures 3B and  4B), which might be explained by the known anti-inflammatory function of β-estradiol (Camporez et al., 2019). The generally lower vulnerability for inflammation in females may mitigate the differences in inflammatory cytokines between the two genotypes.
The HFD induced adipocytic hyperplasia in Cebpb ΔuORF mice indicates that unconstrained LAP functionality -through loss of inhibitory function of LIP -stimulates adipocyte differentiation and function. It may explain why Cebpb ΔuORF male mice store more fat in WAT on a HFD than wt littermates assuming that efficient fat storage by adaptive increase of the number of adipocytes prevents relocation of fat to peripheral tissues. Although fat storage in female Cebpb ΔuORF mice upon HFD was rather reduced compared to wt littermates, also their adipocyte numbers in the visceral fat depot were increased. These observations are in line with our previous experiments showing that mouse embryonic fibroblasts (MEFs) derived from Cebpb ΔuORF mice are much more efficiently induced to undergo adipogenesis than wt MEFs, and differentiation of 3T3-L1 preadipocytes is strongly suppressed upon ectopic induction of LIP (see data in Expanded View Figure 3 B, C of Zidek et al., 2015). Furthermore, the Kirkland lab has shown that endogenous LIP levels increase upon ageing in pre-adipocytes and in isolated fat cells from rats which correlated with reduced C/EBPα expression and reduced differentiation potential of the pre-adipocytes and that overexpression of LIP in preadipocytes from young ND and HFD, n = 6). Intraperitoneal insulin sensitivity test (IPIST) with the calculated area under the curve (AUC) of Cebpb ΔuORF (C) male and (D) female (ΔuORF) mice and wt mice injected i.p. with insulin (0.5 IU/kg) (7 weeks; males: wt ND, n = 6; wt HFD, n = 8; ΔuORF ND, n = 5; ΔuORF HFD, n = 7; females: wt ND, n = 7; wt HFD, n = 7; ΔuORF ND, n = 6; ΔuORF HFD, n = 7). All values are mean ± SEM. p-Values were determined with Student's t-test, *p < 0.05; ***p < 0.001.
The online version of this article includes the following source data for figure 6: Source data 1. Raw data related to Figure 6A-D.  (Karagiannides et al., 2001). Similarly, we observed a shift of the C/ EBPβ isoform ratio towards more LIP expression in visceral fat from wt males on HFD (Figure 8) which probably inhibits adipogenesis and thereby might contribute to adipocyte hypertrophy observed in wt mice.
In the current model of the regulatory cascade of adipocyte differentiation C/EBPβ and C/EBPδ induce the expression of C/EBPα and PPARγ, which by positive feedback stimulate each other's expression (Siersbaek and Mandrup, 2011). Pharmacological activation of PPARγ by thiazolidines similarly to the Cebpb ΔuORF mutation stimulates adipocyte differentiation, results in fat storage in hyperplastic adipocytes and in a shift to fat storage in the subcutaneous compartment, resulting in improved metabolic health (Adams et al., 1997;Fujiwara et al., 1988;McLaughlin et al., 2010;Okuno et al., 1998). Our finding that the mRNA expression of adipogenic transcription factors PPARγ, SREBP1 and C/EBPα in the visceral fat of Cebpb ΔuORF males is better maintained on HFD compared to wt mice also fits to these observations. However, in contrast to our qPCR data, C/EBPα protein levels were rather downregulated in Cebpb ΔuORF males both at ND and HFD conditions upon HFD (Figure 7-figure supplement 1) suggesting counter-regulation at the level of translation. This might be an adaptive response to increased LAP function yet does not seem to affect the adipocyte hyperplasia phenotype.
In the visceral WAT of wt females, LIP levels and the LAP/LIP isoform ratios do not significantly change in response to HFD feeding, in contrast to males. Accordingly, the mRNA expression levels of the adipogenic transcription factors are maintained upon HFD feeding in wt females. Whether this sex-specific difference might be due to the less pronounced inflammation (macrophage infiltration) observed in females or to other sex-specific responses to HFD feeding has to be examined in future studies. Probably, the slight increase in PPARγ expression together with the increased LAP function might be sufficient for the observed adipocyte hyperplasia in female Cebpb ΔuORF mice on HFD, which however, seems to be less pronounced compared to HFD fed Cebpb ΔuORF males. The downregulation of fatty acid synthetase (FAS) mRNA levels that we observe in wt mice upon HFD seems to be independent from the expression of the adipogenic transcription factors tested because at least in females these regulatory events were uncoupled and might be due to other effects of HFD feeding. Furthermore, in Cebpb ΔuORF females the protein expression of FAS on HFD does not correspond to the FAS mRNA levels, it is efficiently reduced despite almost completely maintained mRNA levels suggesting interfering, post-transcriptional effects. What these effects are and why they are only observed in female mice is so far not known.
Taken together, our data propose pharmacological reduction of LIP expression as an approach to switch the unhealthy metabolic phenotype of obese individuals into a healthy obese phenotype to prevent the development of metabolic disease (possibly together with pharmacologic PPARγ activation). We have shown that a search for such an intervention is feasible through the identification of drugs that inhibit LIP expression similar to mTORC1-inhibition (Zaini et al., 2017). One drug that we identified as an inhibitor of LIP expression, the antiviral drug adevovir dipivoxil (Zaini et al., 2017), was recently tested on female wt mice upon ND and HFD feeding (Bitto et al., 2021). This study showed that adevovir is effective in increasing the LAP/LIP isoform ratio also in vivo (although in the mice LIP expression was not affected but rather LAP expression was increased), which resulted in increased C/EBPβ target gene activation and increased expression of β-oxidation genes in the liver similar to what we observed in the Cebpb ΔuORF mice (Zidek et al., 2015). Remarkably, adevovir treatment resulted in a significant reduction of body weight and fat content particularly in the HFD fed mice (Bitto et al., 2021) similar to what we found with the Cebpb ΔuORF females on HFD. Whether this effect on body weight and fat accumulation is caused only by increasing C/EBPβ function or whether additional not yet identified effects of adevovir contribute in addition is not known so far. ND, n = 6; wt HFD, n = 4; ΔuORF ND and HFD, n = 5). All values are mean ± SEM. p-Values were determined with Student's t-test, *p < 0.05; **p < 0.01; ***p < 0.001.
The online version of this article includes the following source data and figure supplement(s) for figure 7: Source data 1. Raw data related to Figure 7A, B.  Furthermore, it will be interesting to examine whether adevovir treatment of males results in reduced fat accumulation like in females or in increased (subcutaneous) fat accumulation as we observed in the Cebpb ΔuORF males. Since the increased C/EBPβ function in Cebpb ΔuORF mice also has the potential to extend the healthspan (Müller et al., 2018), therapeutic interference with the C/EBPβ isoform ratio may represent a promising strategy to attenuate the metabolic disturbances not only associated with overnutrition but also with ageing.
The online version of this article includes the following source data for figure 8: Source data 1. Raw data related to Figure 8B, C, E and F.

Continued on next page
Key resources

Determination of body composition
Mice were anesthetized and the abdominal region from lumbar vertebrae 5-6 was analyzed using an Aloka LaTheta Laboratory Computed Tomograph LCT-100A (Zinsser Analytic) as described in Zidek et al., 2015. Determination of caloric utilization Both the feces and samples of the HFD food were collected, dried in a speed vacuum dryer at 60 °C for 5 hr, grinded and pressed into tablets. The energy content of both the feces and food samples was determined through bomb calorimetry using an IKA-Calorimeter C5000. The energy efficiency was calculated through subtraction of the energy loss in the feces from the energy consumed.

IP glucose tolerance and insulin sensitivity tests
For the determination of glucose tolerance, mice were starved overnight (16 hr) and a 20% (w/v) glucose solution was injected i.p., using 10 μl per gram body weight. After different time points, the blood glucose concentration was measured using a glucometer (AccuCheck Aviva, Roche). For the determination of insulin sensitivity, an insulin solution (0,05 IU/ml insulin in 1xPBS supplemented with 0.08% fatty acid-free BSA) was i.p. injected into non-starved mice using 10 μl per gram body weight and the blood glucose concentration was measured as described above.

Histological staining
Tissue pieces were fixed for 24 hr with paraformaldehyde (4%) and embedded in paraffin. Sections (5 μm) were stained with hematoxylin and eosin (H&E) in the Autostainer XL (Leica). Adipocyte area was determined using the ImageJ software from 12 adjacent cells per mouse. For CD68 staining, sections (5 μm) from paraffin embedded tissue were dried for 2 hr at 55 °C, deparaffinized and rehydrated. For antigen retrieval, sections were incubated for 25 min in 10 mM citrate buffer, pH 6.0. Endogenous peroxidase was blocked in 1% H 2 O 2 in methanol for 30 min. After blocking with normal goat serum (1:10 in PBS), sections were incubated with a CD68 specific antibody (E307V, #97,778 from Cell Signaling, 1:200) over night at 4 °C followed by incubation with a biotin-conjugated secondary goat anti rabbit antibody (Vector Labs, BA-1000, 1:250) for 30 min and incubation with reagents of the Vectastain ABC HRP kit (Vector Labs, PK-4000) according to the manufacturer's instruction. Slides were stained with DAB and counterstained with hematoxylin, dehydrated and covered using Eukitt. A Hamamatsu scanner was used to take images. For lipid staining with Sudan III, cryosections (10 μm) were fixed with paraformaldehyde (4%) and stained with Sudan-III solution (3% (w/v) Sudan-III in 10% ethanol and 90% acetic acid) for 30 min. For lipid staining with Oil-Red-O, cryosections (10 μm) were fixed with paraformaldehyde (10%), washed shortly in 60% isopropanol and stained with Oil-Red-O solution (3 mg/ml isopropanol stock solution diluted to 1.8 mg/ml with H 2 0) for 15 min. After shortly washing first with isopropanol and then with water, cells were counterstained with hematoxylin and covered with 10 mM Tris HCl pH 9.0 in glycerol.

Calculation of adipocyte number
The mean adipocyte area from the visceral fat per mouse was used to calculate the adipocyte volume per mouse with r (radius) = and V (volume) = π r 3 . For males, the mean volume of the visceral fat as determined by CT analysis was then divided by the mean adipocyte volume to get the cell number. For females, adipocyte weight was calculated by multiplying the calculated cell volume with 0.915 (the density of triolein). Then, the mean weight of the visceral WAT tissue was divided by the calculated adipocyte weight.

Determination of organ weight
After termination of the mice organs were collected and cleaned from surrounding fat or connective tissue and their weight was determined using an analytical balance.

qRT-PCR analysis
Tissue pieces were homogenized using the Precellys 24 system (Peqlab) in the presence of 1 ml QIAzol reagent (QUIAGEN). The RNA was isolated using the RNeasy Lipid Tissue Mini kit (QUIAGEN) according to the protocol of the manufacturer, incubated with RQ1 RNase-free DNase (Promega) for 30 min at 37 °C and purified further using the RNeasy Plus Mini kit (QUIAGEN) starting from step 4. One μg RNA was reverse transcribed into cDNA with Oligo(d)T primers using the Transcriptor First Strand cDNA Synthesis kit (Roche). The qRT-PCR was performed with the LightCycler 480 SYBR Green I Master mix (Roche) using the following primer pairs: CD68: 5'-GCC CAC CAC CAC CAG TCA CG-3' and 5'-G TG G TC C AG G GT G AG G GC C A-3 ', PPARγ 5'-GCC CTT TGG TGA CTT TAT GG-3' and 5'-C AG C AG G TT G TC T TG G AT G T-3 ', C/EBPα 5'-C AA G AA C AG C AA C GA G TA C CG-3' and 5'-G TC A CT G GT C AA C TC C AG C AC-3', SREBP1c: 5'-AAC GTC ACT TCC AGC TAG AC-3' and 5'-C CA C TA A GG T GC C TA C AG A GC-3', FAS: 5'-A CA C AG C AA G GT G CT G GA G -3' and 5'-G TC C AG G CT G TG G TG A CT C T-3 ', TNFα: 5'-CCA GAC CCT CAC ACT CA-3' and 5'-C AC T TG G TG G TT T GC T AC G AC-3', MCP1: 5'-G CT G GA G AG C TA C AA G AG G AT C A-3 ' and 5'-A CA G AC C TC T CT C TT G AG C TT G GT, IL-1β: 5'-G AA A TG C CA C CT T TT G AC A GT G -3' and 5'-T GG A TG C TC T CA T CA G GA C AG-3', IL-6: 5'-C CG G AG A GG A GA C TT C AC A G-3 ' and 5'-T TC T GC A AG T GC A TC A TC G T-3 ', GAPDH: .5'-ATTG TCAG CAAT GCAT CCTG -3' and 5'-ATGG ACTG TGGT CATG AGCC -3' and β-actin: 5'-AGA GGG AAA TCG TGC GTG AC-3' and 5'-C AA T AG T GA T GA C CT G GC C GT-3'.

Statistical methods
The number of biological replicates is indicated as n = x. All graphs show average ± standard error of the mean (SEM). To calculate statistical significance of the obtained results the Student's t-Test was used with *p < 0.05; **p < 0.01 and ***p < 0.001. Single mice were excluded when results indicated technical failure of the experimental performance. Furthermore, extreme outliers were excluded from the analysis. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.