Influence of dietary saturated fat content on adiposity, macrophage behavior, inflammation, and metabolism: composition matters

METHODS

RESULTS

DISCUSSION

HFDs are strongly linked with the accumulation of excess body fat, chronic inflammation, and metabolic perturbations, ultimately leading to poorer health outcomes. It is well known that ATMs play a central role in this relationship (24). However, the extent to which the FA composition of a HFD influences macrophage behavior and inflammation is still poorly understood; most of the available supporting literature is limited by the lack of control for various nutrients (e.g., ratio of MUFA:PUFA; omega-6:omega-3 FAs; protein:carbohydrate:fat, etc.), the utilization of a single ingredient as the sole source of dietary fat (e.g., corn oil, beef tallow, milk fat, etc.), and the absence of dose response studies. We examined the effect of three HFDs, differing in the percentage of total calories from SF (6%, 12%, and 24%) but identical in total fat (40%), on adiposity (absolute fat mass), macrophage phenotype, inflammation, and metabolism utilizing controlled diets consisting of various lipid-rich ingredients. Our findings indicate that manipulating the SF content without changing the percentage of total calories from fat has a profound effect on these outcomes. The 12%-SF diet, most closely mimicking the standard American diet, led to the greatest adiposity (absolute fat mass), macrophage infiltration, and IR. Although the 24%-SF diet increased adiposity and produced IR, it did not significantly increase macrophage infiltration, it led to a lesser degree of AT inflammation, and it did not raise the TC/HDL-C ratio.

Given that SFAs, in general, are less efficiently oxidized than USFAs, we hypothesized that adiposity would be greatest following consumption of the 24%-SF diet (16, 17). Interestingly, despite the fact that HFD-fed mice consumed similar kilocalories, the 12%-SF diet led to the greatest accumulation of fat and the largest adipocyte size, followed by the 6% and 24%-SF groups, respectively. We also confirmed that plasma leptin was proportional to the degree of fat-mass accumulation as previously reported (9).

The FA composition of AT is primarily dependent on the FA composition of the diet (25). Therefore, lipolysis of white adipose tissue (WAT) composed of a greater proportion of SFAs should lead to higher levels of circulating saturated free-fatty acids and subsequent activation of macrophages through binding to TLRs (10, 13). Given this, we expected that a diet higher in SF content would lead to greater macrophage infiltration and more pronounced inflammation. In agreement with previously reported literature, all HFDs increased expression of F4/80 as well as markers for M1 and M2 macrophages; interestingly, however, these reached statistical significance only in the 12%-SF group. The activation of macrophages in AT is thought to be mediated by TLRs; both TLR-2 and TLR-4 have been implicated in HFD-induced macrophage activation and inflammation given their ability to bind saturated free-fatty acids (13, 26, 27). Consistent with the macrophage data, we show a statistically significant increase in TLR-2 expression only in the 12%-SF-fed mice, but surprisingly there was no significant upregulation of TLR-4 in any of the HFD groups. To our knowledge, there have been no reports of an upregulation of WAT TLR-2 expression without significant changes in WAT TLR-4 expression following HFD feedings. These results warrant further investigation into the role that individual TLRs play in HFD-induced inflammation, and conversely, the effect that varying the composition of HFDs has on TLR activation.

Because macrophages are thought to mediate their inflammatory processes through activation of various transcription factors (10, 12, 13), we next examined phosphorylated NFκB, JNK, and p38 MAPK, and we found that all three HFDs increased activation of JNK, whereas only the 12%-SF diet significantly increased NFκB activation. All HFD-fed-mice exhibited an increase in mRNA expression of MCP-1 in AT, and the 6%-SF and 12%-SF-fed mice, but not the 24%-SF-fed mice, exhibited an increase in TNF-α mRNA expression. Surprisingly, IFN-γ, which has been shown to be upregulated in WAT and plays a key role in macrophage activation in HFD-induced obesity (28), was not statistically different across diets. This may be due to the specific measurement time point, the composition of these novel diets, or most likely, an increase in factors that can regulate expression of IFN-γ. Additionally, we found no changes in AT IL-6 mRNA expression across groups. Although previous studies have shown AT IL-6 mRNA expression to be upregulated as a result of high-fat feeding (29, 30), others have shown that this is not always the case (31). We also measured IL-10, an anti-inflammatory cytokine, and found it to be upregulated in the 12%-SF mice only, which is most likely a compensatory response to the increased inflammation (32). Interestingly, SOCS1, a negative regulator of inflammation (33), appeared to be downregulated in all the HFDs, but this reached significance in the 24%-SF group only. Circulating markers of inflammation (MCP-1, IL-6) were measured to determine whether they mirrored the observed changes in WAT as has previously been reported (30, 34). We found no significant increases in plasma MCP-1 or IL-6 for any of the HFDs. The most likely explanation for the discrepancies between our findings and those of others is the differences in the composition of the diets used.

Emerging evidence suggests that pro-inflammatory M1 macrophages may play a role in inducing IR, whereas alternatively activated M2 macrophages may help to maintain insulin sensitivity (29). Our data somewhat supports this hypothesis as the degree of M1 macrophage mRNA expression corresponded well with the level of IR: the 12%-SF diet resulted in the most severe IR, followed by the 6%-SF and 24%-SF diets, respectively. However, given that M1 macrophage quantification in this study was limited to mRNA expression of CD11c, this association should be interpreted with caution. Adiponectin, well characterized as an anti-inflammatory adipocytokine known to promote insulin sensitivity, has been shown to be inversely correlated with body fat accumulation (9). Interestingly, there were no changes in the concentration of plasma adiponectin across groups. It is likely that changes in adiponectin would have been observed if it had been measured in the WAT (35); however, our analysis was limited to plasma levels. Concerning lipid metabolism, all three HFDs increased TC and LDL-C levels. Of interest was the finding that although the 24%-SF diet increased LDL-C and elevated plasma triglycerides, it also resulted in a higher HDL-C level, producing a more favorable TC/HDL-C ratio similar to that of control diets.

A possible explanation for the discrepancies in adiposity (absolute fat mass), IR, inflammation, and the TC/HDL-C ratio between the 12%-SF-fed and the 24%-SF-fed mice may be the difference in the content of medium-chain fatty acids (MCFA, C8:0–C12:0) (3.6%, and 11.8% of total caloric intake for the 12%-SF and 24%-SF diets, respectively). These variations exist because it was not possible for the composition of the SF in each of the HFDs to be consistent while utilizing various lipid-rich ingredients and simultaneously controlling for the omega-6:omega-3 and MUFA:PUFA ratios. A previous study reported that a diet composed of 12% caprylic (C8:0) and capric acids (C10:0), both MCFAs, augments the rate of fat-mass loss compared with a diet composed of 12% olive oil that contains mostly MUFAs (36). Furthermore, there is substantial evidence demonstrating that small doses of MCFAs and HFDs rich in MCFAs can effectively reduce body weight, reduce fat-mass gains, and minimize IR compared with HFDs rich in LCFAs (37–41). Additionally, MCFA-rich diets have been associated with an improved cholesterol profile (42–45) and reduced inflammation (46, 47) compared with other isocaloric diets. The reported reduction in body weight produced by MCFAs is associated with higher energy expenditure and upregulated FA oxidation (48–50). As a result of their shorter chain length, MCFAs can be transported from the intestines directly to the liver where they can be quickly oxidized (37, 51). LCFAs, on the other hand, are first incorporated into chylomicrons before they leave the intestine via the lymphatic system and travel through the blood to extrahepatic tissues to be stored or metabolized. Not only do MCFAs and LCFAs differ in their digestive routes but their propensity to be oxidized is also dissimilar (16). Once inside a cell, MCFAs are less likely to be stored as AT as they can enter the mitochondria to be oxidized independent of carnitine palmitoyltransferase 1, unlike LCFAs.

It is important to point out that the MCFA argument may not be valid when comparing the 6%-SF and 12%-SF diets (0.1% versus 3.6% of total calories from MCFAs, respectively). It is likely that the MCFA-caloric content of the diet would need to be higher to produce similar effects generated by the 24%-SF diet. The differences in adiposity we observed between the 6%-SF and 12%-SF-fed mice may be due to the fact that the 12%-SF diet was composed of more obesogenic LCSFAs (8.4% versus 5.9% of total calories). However, we find this unlikely as the 24%-SF diet had the largest percentage of LCSFAs (12.2% of total calories) and resulted in a similar level of adiposity as in the mice consuming the 6%-SF diet.

Another possible rationale for the differences in IR and inflammation across the diets may be the disparities in the linoleic content (C18:2), (11.4%, 9.4%, and 5.4% of total caloric intake for the 6%-SF, 12%-SF, and 24%-SF diets, respectively). Even when controlling for various ratios within an isocaloric diet, the manipulation of one macronutrient, or subset of macronutrients, results in an uncontrollable alteration of another. As such, in the current study, alterations in the percentage of SF across diets also resulted in changes in the percentage of unsaturated fat. Thus, although the ratio of omega-6:omega-3 FAs was the same for each of the HFDs, the absolute quantity of linoleic acid in the 6%-SF and 12%-SF diets was greater than in the 24%-SF diet. Linoleic acid serves as a short-chain, parent omega-6 FA necessary for the synthesis of essential omega-6 LCFAs (52) that can play an important role in the promotion of inflammatory processes through the production of various eicosanoids (52–54). In fact, others have shown that a HFD rich in omega-6 FAs can increase inflammation more than a HFD rich in SFAs (55). Further, omega-6 FAs are prone to peroxidation, leading to accumulation of 4-hydroxy-2-nonenal (4-HNE), which has been shown to induce IR (56).

It is also likely that the differences in MCFAs, LCFAs, and linoleic content across the diets can explain some of the other unexpected reported findings. For instance, the discrepancies in MCFA and linoleic content between the 6%-SF and 24%-SF diets may explain the similar adiposity but different levels of IR seen in these groups; previous work has shown that MCFA-rich diets can hinder the development of IR without influencing body weight (57). And it may be that the greater content of linoleic acid in the 6%-SF diet produced significantly more 4-HNE, leading to a greater IR compared with the 24%-SF diet. Also, it is certainly possible that the relatively high content of both linoleic acid and LCSFAs in 12%-SF diet may explain the high degree of adiposity, macrophage infiltration, and IR compared with the 6%-SF diet. Clearly, the disparate findings across the three HFDs do not result from variations of a single group of FAs but instead stem from alterations in the content of several classes of FAs.

It is well established that SFAs play a role in inflammatory signaling (10, 12, 58, 59). However, the degree to which individual FAs activate pro-inflammatory pathways remains somewhat controversial, as there has been at least one study to report that SFAs do not activate TLRs in vitro (60). Further, it is evident from our findings as well as from previous research that varying the composition of dietary SFAs can differentially regulate inflammatory processes in vivo compared with in vitro. For example, Lee and others have shown that lauric acid (C12:0), a MCFA that varies considerably with increasing percentage of SF in our diets, can serve as a potent agonist of TLR signaling in vitro (10, 12). On the contrary, Rivera et. al. (46) reported that a MCFA-rich diet more effectively attenuated nonalcoholic steatohepatitis and reduced hepatic TLR-4 expression versus a PUFA-rich diet. These convergent findings may be explained by the inclination of lauric acid to be oxidized in vivo, thus limiting its ability to serve as a ligand for TLRs (16). These findings highlight the importance of additional research to better understand the role of various SFAs on inflammatory-related signaling, and further, to determine whether the effects observed in vitro are actually reflected in vivo.

Additionally, our data and that of others suggest that isocaloric diets with a greater content of omega-6 PUFAs can produce greater IR, result in a poorer TC/HDL-C ratio, and may even increase inflammation (46, 55), more so than an isocaloric diet composed of significantly more SF. Meanwhile, others have shown that high consumption of omega-6 FAs has been associated with reduced inflammation, a more favorable TC/HDL-C ratio (55), and no negative effects on inflammatory markers (61, 62). It should be noted, however, that the majority of these studies were performed in humans, in which it is extremely difficult to control for the nutrient composition of the diet and the activity level of the subjects--two factors known to greatly influence metabolism and inflammation (63, 64). This affirms the need for future research utilizing various controlled diets to better understand the role that omega-6 FAs have on physiological processes.

In summary, we examined the influence of three 40% HFDs, which differed in the percentage of total calories from SF (6%, 12%, and 24%), on body composition, macrophage behavior, inflammation, and metabolic dysfunction in mice. In general, the 12%-SF diet, most closely mimicking the standard American diet, led to the greatest adiposity, macrophage infiltration, and IR, whereas the 24%-SF diet had the lowest levels of these outcomes. In conclusion, our findings support previously published data that ad libitum, high-fat feeding can lead to an increased risk of obesity and obesity-related side effects. However, the extent of excess fat accumulation and adverse health perturbations is not necessarily proportional to the percentage of SF in the diet. Although SFAs and omega-6 FAs have been implicated as pro-inflammatory molecules, future research should examine the degree to which diets differing in SFA and omega-6 FA content impact individual TLR activation, macrophage behavior, inflammatory signaling, and subsequent metabolic dysfunction, as well as the effect that manipulating the absolute quantities of omega-6 and omega-3 FAs, without changing the ratio between these FAs, has on these outcomes.