Obesity is a multifactorial disorder characterized by chronic, low-grade inflammation throughout the body, especially in insulin-responsive organs, such as adipose tissues, liver, muscle or the pancreas. This chronic inflammatory state is caused by activated immune cells, mainly macrophages accumulating in these tissues, leading to an increased expression of pro-inflammatory cytokines [22]. In turn, many of the diseases associated with obesity, such as diabetes, NAFLD, or even cancer can be traced back to this low-grade inflammation. However, the exact mechanism and the causality of these processes are not yet fully understood.
In this study, we used two diet-induced mouse models, HFD and HFD + FR, to investigate the pathological changes and inflammatory processes induced by obesity at systemic level and in individual organs. HFD treatment is widely used to model obesity, hyperlipidemia and hyperglycemia in mice [9]. On the other hand, dietary sugars have been shown to influence the effects of HFD [23]. Fructose metabolism is largely different from that of glucose. For example, fructose does not trigger insulin release, and the main site of fructose metabolism is the liver [24]. Moreover, it has potent lipogenic effects, functioning as a substrate for fatty acid synthesis and activating the associated enzymes [23]. Consequently, excessive fructose intake strongly promotes the development of NAFLD and hepatic IR [25]. On the other hand, fructose-feeding alone did not lead to increased serum cholesterol level and seemed to induce lipid accumulation only in the liver [26]. Therefore, the combination of high-fat feeding with fructose supplementation is probably more effective to initiate metabolic syndrome in animal models.
Accordingly, we found remarkable differences between the two diet models regarding certain obesity-related pathological changes. Most interestingly, although the 5-month-long HFD resulted in significant weight gain and a mild NAFLD, it did not induce IR. In contrast, mice receiving fructose-supplemented water in addition to HFD showed higher peak glucose concentrations in the OGTT. This indicates that HFD + FR has a more significant negative effect on glucose metabolism than HFD alone, leading to the development of IR. In addition, compared to HFD, HFD + FR resulted in even higher weight gain, more severe symptoms of hepatic steatosis and lipid accumulation in the BAT. As opposed to WAT, the main function of BAT is not energy storage but non-shivering thermogenesis, therefore brown adipocytes contain high number of small lipid droplets. However, in response to excessive calorie intake, lipid accumulation can also be observed in the BAT, resulting in a structure that resembles WAT with adipocytes containing fewer but larger lipid droplets [11]. Such morphological changes of the BAT were observed only in the HFD + FR group. On the other hand, we did not find differences in serum lipid parameters between the two models. While the serum triglyceride level did not increase significantly by either diet, the LDL- and HDL-cholesterol levels elevated in similar extents in response to both diets. This suggests that serum hyperlipidemia is mainly induced by dietary fats, and fructose consumption has no significant additional effect, at least in the studied mouse model.
Because fructose supplementation exacerbated metabolic disturbances, a more severe inflammation was expected in these animals. Therefore, we examined several characteristics of inflammation, such as cytokine expression and immune cell activation in different organs and in the blood. NAFLD is a spectrum of changes caused by triglyceride accumulation from early-stage steatosis to chronic inflammation (steatohepatitis), which can finally lead to fibrosis and cirrhosis [27, 28]. In our experiments, we found that HFD alone caused mild steatosis, but more severe symptoms were observed when it was combined with fructose supplementation. The considerably higher level of NAS was a result of the more advanced lipid accumulation and hepatocyte ballooning. Consistent with this, several genes that are involved in the regulation of lipid accumulation, such as lipoprotein lipase, the fatty acid translocase Cd36 or the gene encoding the hormone-like molecule irisin, were induced in response to HFD alone, and showed an even higher increase in the HFD + FR animals. On the other hand, we could not detect histopathological signs of inflammation or increased hepatic cytokine expression. Therefore, these results show that fructose supplementation worsened lipid accumulation but, at this stage of the disease, did not cause inflammation in the liver.
According to our current knowledge, the main contributors to obesity-induced systemic inflammation are the cytokines and other hormone-like molecules released by the WAT [29]. Indeed, the gene expression level of Lep was increased in parallel with higher body weight in our models, although we could not detect significant differences between the two diets. Three cytokines (Tnf, Il10 and Tgfb) showed elevated expression in response to HFD, which increased further upon fructose supplementation. On the other hand, the mRNA levels of Il6 and Il1b did not elevate in the WAT of animals with obesity. In recent years, increasing evidence supports that the molecular chaperone heat shock proteins are involved in the regulation of inflammation and certain metabolic disturbances [30,31,32]. In the current study, we also observed significant increase in the small heat shock protein Cryab (and to a smaller extent in Hsp25) mRNA level in the WAT of the HFD and HFD + FR animals, while the level of Hsp70 rather decreased. In addition, we registered strong positive linear correlation between the expression levels of Lep and Cryab or Hsp25. These results suggest a specific role of small heat shock proteins in the regulation of obesity-related metabolic alterations, which is in agreement with previous research revealing that αB-crystallin functions as an adipokine [33] and might be involved in the pathogenesis of diet-induced diabetes [34].
Serum concentrations of leptin and TNFα were consistent with the mRNA levels measured in the adipose tissues. In contrast, Il1b gene expression was not induced by the applied diets either in the WAT or in the BAT but the protein was present in higher concentrations in the blood of animals with obesity compared with the control group, suggesting other potential sources of circulating Il-1β. According to some assumptions, certain cytokines secreted into the blood are responsible for the development of IR. Both TNFα and Il-1β were found to be able to influence insulin signaling, although it was also shown that circulating level of TNFα is lower than the effective concentration even in patients with obesity [22]. In our experiments, serum TNFα concentration was higher in the HFD + FR animals in parallel with the higher level of IR compared with the HFD group. In contrast, serum Il-1β concentration was not different between the two diet groups, which implies that Il-1β alone can not be responsible for the development of IR.
Because increased serum concentration of pro-inflammatory cytokines indicates a chronic systemic inflammation, we analyzed the systemic immune changes by single-cell phenotyping of all major immune cell populations in the blood, bone marrow and spleen. The most prominent changes were found in the surface expression levels of CD69 and CD44. CD69 is an early marker of immune cell activation and an important regulator of immune responses [35]. Indeed, in the animals with obesity we found its increased surface expression on T-cells and NK cells in the bone marrow samples, and on B-cells of the blood and spleen. On the other hand, circulating NK cells showed decreased surface expression of CD69 in response to obesity. CD69 on NK cells appears to play a crucial role in initiating tumor cell lysis [36], therefore its reduced expression may impair the anti-tumor response. Indeed, increasing evidence suggests that obesity-related alterations in NK cell physiology and function may influence tumor development [37]. The surface expression of CD44 was also significantly increased in response to the applied diets in various cell types, such as B-cells, macrophages or CD4+ and CD8 + T-cells. CD44 is a multifunctional cell surface glycoprotein, a receptor for hyaluronan and osteopontin. It participates in the activation and proliferation of T-cells and NK cells, therefore these results also confirm the development of systemic inflammation in the animals with obesity. However, CD44 is expressed by other cell types as well, such as hepatocytes or adipocytes, which has an important role in the diet-induced adipose inflammation [38]. Indeed, our qPCR results showed that Cd44 expression in the vWAT displayed a robust elevation in accordance with the higher level of cytokine expression. Previously, an increased frequency of CD44 + T-cells was observed in subcutaneous adipose tissue of HFD-fed mice [39] which can contribute to the increased Cd44 mRNA level in the vWAT of our mice as well. However, in line with the findings of previous research [40, 41], the results of the CD44 immunostaining indicated that endothelial cells, mesenchymal stem cells, and to a lesser extent, adipocytes may also be important sites of Cd44 expression. Moreover, it should be mentioned that CD44 is also involved in the regulation of cell adhesion and migration, and it is an important regulator of cancer cell progression and metastasis [42]. We can therefore suppose that the increased CD44 expression in the WAT might be another important link between obesity-induced inflammation and the higher risk of cancer. Interestingly, we did not find significant differences in the surface expression of CD69 and CD44 between the two obesity models, suggesting that even mild obesity can induce these immunophenotypic changes.
It is important to acknowledge the limitations of the present study, in particular the fact that only males were included. In our previous studies we found that females are more resistant to the development of diet-induced obesity and related disorders, such as NAFLD [11, 12]. Because here our major aim was to investigate obesity-related inflammation, we decided to use only male animals, which have a higher degree of obesity. Moreover, this allowed us to increase the number of animals per groups.
In conclusion, although HFD alone is suitable for examining certain obesity-related parameters, its supplementation with fructose led to a better model of Western-diet. HFD + FR resulted in a more substantial increase in weight and in the level of hepatic steatosis, and only these animals showed IR. Accordingly, fructose supplementation resulted in increased serum TNFα concentration and expression of specific cytokines in the WAT (and BAT), indicating enhanced systemic inflammation. However, despite these differences, both models showed immunophenotypic alterations that may be linked to an increased risk of obesity-related cancer. On the other hand, the extent of inflammation did not correlate with all symptoms of metabolic syndrome as neither serum triglyceride increased nor steatohepatitis was detected in either obesity models.