Nutritional immunotherapy: dietary fatty acids

Most people familiarized with a paleo-type lifestyle already know that a healthy diet should include good fats, rich in saturated (SFAs) and monounsaturated fatty acids (MUFAs). This profile is characteristic of animal fats. Conversely, intake of polyunsaturated fatty acids (PUFAs) should be limited. These recommendations are in accordance with the biological, chemical and evolutionary aspects of fatty acids in the human diet. We might see this dietary fatty acid profile as a "natural" one. 

As I stated in my last post, I support the notion that some inflammatory/autoimmune disorders deserve a different approach, owing to the body's unnatural state. This deviation from normality implies that we cannot expect the same response to some nutritional components in diseased people. 

Background

For understanding the basis of my recommendations, we must review a little bit about Toll-like receptors (TLRs) and the classic cellular inflammatory cascade, the NFkB pathway. TLRs are pattern recognition receptors which bind pathogen associated molecular patterns (PAMPs). PAMPs are conserved molecular motifs found in a broad range of pathogens that are recognized by receptors mediating an innate-type immune response (like TLRs). PAMPs include LPS, lipoproteins, peptidoglycan, lipoteichoic acid and other molecules capable of binding to these receptors and trigger a response. There are different type of TLRs, with different cellular localizations and associated intracellular pathways, but most converge in the activation of NFkB. The final effect of ligand binding and protein signaling is the expression of inflammatory genes. 

TLRs have evolved not only a function in immunity, but recent evidence suggests a pivotal role in metabolism. 

TLR4/MD-2 binds to LPS

LPS, the classic TLR ligand, binds and activates signaling through TLR4 (1). This interaction is mediated by the lipid moiety, Lipid A. The toxicity of LPS is thought to be due to the interaction of TLR4 with Lipid A, and the shape and conformation of this lipid may determine the toxicity of a given pathogen (LPS are not created equal) (2). The differences arise from the number and length of fatty acid chains. 

LPS general structure. Supplemental Material: Annu. Rev. Biochem. 2011. 80:917-941. Link.

Typically, 4 to 7 lipid chains with 12 to 14 carbons of length are anchored to the glucosamine backbone. 

For being able to recognize LPS, TLR4 binds MD-2 and forms the complex responsible for the interaction. Only one third of MD-2 is involved in the interaction with TLR4 and the remaining part is available for interaction with other ligands. The presence of LPS is necessary for TLR4/MD-2 dimerization (3). 

TLR4/MD-2 complex. Annu. Rev. Biochem. 2011. 80:917-941.

Structural studies have proposed that certain factors determine the endotoxic activity of lipid A. The number of lipid chains seems to be the most important: Six lipid chains of 12 to 14 carbons each promote an optimal inflammatory activity. Because of this, the deletion of two lipid chains can destroy agonistic activity, making some LPS TLR4 antagonists (4). The SFAs linked to the Lipid A backbone are 3-O-acylated by lauric acid, myristic acid or palmitic acid, which seem to be important for the immunological effects of LPS, given that their loss abolishes LPS endotoxic properties and produces an antagonist of Lipid A (5, 6). The logical inference of these facts is that these SFAs are important for the binding and/or LPS-induced TLR4 activation.

The role for TLR4 in metabolic abnormalities has been corroborated using animal models. Mice with a loss-of-function mutation in TLR4 or TLR4-null mice are protected against the development of diet-induced obesity and insulin resistance (7, 8, 9). Deletion of CD14, a TLR2/TLR4 co-receptor, attenuates the cardiovascular and metabolic complications of obesity (10). TLR4 has been involved in other metabolic complications which are less understood, like the progression of simple steatosis to non-alcoholic steatohepatitis (NASH) associated with obesity (11). 

From the above evidence, researchers thought they had found the link between dietary fat and inflammation. Nevertheless, Erridge and Samani (12) found that the in vitro evidence showing a direct activation of TLRs (they tested TLR2, TLR4 and TLR5) was caused by contamination of fatty-acid-free BSA (used to present SFAs to cells) with LPS and lipopeptide (although some authors show that BSA alone is not sufficient to activate TLR4, see below). Others have suggested that SFAs activate TLR4 signaling indirectly, promoting TLR4 dimerization and association of TLR4 with MD-2 and downstream adaptor proteins (TRIF and MyD88) into lipid rafts (13). This latter explanation seems to be more plausible.

If SFAs promote inflammation and seem to act in part through TLR4 in in vivo studies, then maybe they are acting as carriers of another molecules which bind and activate TLR4. This seems to be the case. As Peter has blogged briefly before, Ghoshal et al. (14) demonstrated that chylomicron formation in the gut promotes LPS absorption. However, this effect is not exclusive to SFA, but to long chain fatty acids (LCFA) as they need chylomicrons for absorption/transport, in contrast to short- and medium-chain fatty acids. 

Does this makes LCFAs inherently unhealthy? No. Appropriate TLR stimulation is important of adequate innate responses to pathogens and for maturation and development during childhood. Excessive uptake of LPS, promoted either by calorie excess and/or overgrowth of gram-negative gut bacteria, seem to contribute to chronic endotoxemia and disease. Endotoxin overload is particularly problematic in adults with inflammation and/or immune related diseases. 

Association between postprandial lipemia and inflammation

Fisher-Wellman & Bloomer found that isocalorically, high-fat meals promote a stronger postprandial oxidative stress than carbohydrate and/or protein meals in healthy subjects (13). However, they used heavy whipping cream as the fat source, dextrose powder as carbohydrate and casein-whey protein powder as the protein source. This helps isolating variables (macronutrients) but doesn't help much for assessing the effect of a mixed meal composed of real food. Moreover, they didn't control for calories, as the "lipid meal" contained more calories than any other meal:

Fisher-Wellman & Bloomer, 2010

In fact, the "protein meal" included more fat than the "lipid meal" (98g/38% vs. 93g/34%). The carbohydrate percentage was the same between the carbohydrate and the lipid meal, but the absolute amount was higher in the latter because of the calorie content. If dietary fat were indeed to blame, it can be expected that the higher the fat content, the higher the measures of oxidative stress. But the meal with higher fat content was the mixed meal, which didn't produce significantly different results than the carbohydrate or the protein meal.  

Another measure of inflammatory changes induced by specific macronutrients or meals is the activation of transcription factors and proteins involved in cellular inflammatory pathways, such as NFkB. For example, glucose ingestion (75g in 300ml water) stimulated nuclear transport of NFkB, reduced IKB-alpha protein levels and increased the activity and expression of IKK-alpha and IKK-beta in mononuclear cells (15). This was paralleled by an increased expression of TNF-alpha and activation of NADPH oxidase. Similar results have been found after a 900kcal mixed meal (81g carbohydrate, 51g fat and 32g protein) (16). This is to be expected as energy intake will increase mitochondrial respiration, stimulating mitochondrial ROS production. Some ROS are capable of activating NFkB, so any increase in intracellular ROS can increase NFkB-mediated signaling (17). 

Ingestion of 300kcal of cream or glucose stimulated NFkB binding, expression of SOCS3, TNF-alpha and IL-1beta in mononuclear cells of healthy subjects, but only cream increased plasma LPS and TLR4 expression (18). This was not seen with ingestion of orange juice, probably due to the increase in uric acid associated with fructose ingestion. These results are in accordance with chylomicron-mediated transport of LPS through the gut. As expected, high FFA plus high glucose amplify the inflammatory response (19).

Hypertriglyceridemia increases endotoxemia

The human body must be able to cope with acute increases in LPS in plasma, attenuating the inflammatory response induced by fat ingestion. Clemente-Postigo, et al. (20) showed that in morbidly obese subjects, endotoxin increases were strongly correlated to the difference between baseline and postprandial triglyceride levels. They also found that baseline triglyceride level was the best variable that predicted basal LPS level in serum. In this regard, very low carbohydrate diets have shown to reduce baseline triglyceride levels and postprandial lipemia (21, 22). In the metabolically healthy, the immune system is capable of attenuating postprandial endotoxemia (as with inflammation induced by any meal). The inflammatory nature of absorption, digestion and metabolism of macronutrients must be coupled with an anti-inflammatory period, such as fasting (depending on the inflammatory load of the diet, an overnight fast might work). By this way, the body's inflammatory balance is maintained in a healthy range. 

Effects of dietary PUFAs on immune cells

It is important to note that, although LCFA (which means they stimulate chylomicron formation and thus, LPS transport), MUFAs and PUFAs exhibit different effects than long chain SFAs. This could be related to their effects on membrane fatty acid composition. Cellular membranes are highly structured, and subtle variations in the unsaturation of phospholipids can have diverse but important molecular consequences. It is well known by now that dietary fatty acids alter the composition of membrane lipids, as they are incorporated. In immune cells, this is extremely important for the overall response to a certain stimulus, from phagocytosis against a pathogen to secretion of cytokines for proliferation and clonal expansion. Fatty acids incorporated to immune cell membranes act through different mechanisms:

• Altering the composition of lipid rafts. This, in turn, influences protein-protein interactions, as well as coupling ligand-receptor interaction with scaffold and intracellular signaling proteins.
• Producing intermediate molecules, such as prostaglandins. The final effect is difficult to assess, but there seem to be clear differences between the action of metabolites produced from omega 6 (O6) vs. omega 3 (O3).
• Altering membrane permeability. A higher unsaturation index (that is, the degree of unsaturation of phospholipid chains) renders a more fluid membrane, being the opposite true for a low unsaturation index. Increasing the proportion of SFA in cell membranes decreases permeability because unsaturated fatty acids chains form a "kink", increasing the degrees of freedom of the molecules and its physicochemical characteristics (both individually and as a group). 
• Providing energy.

Animal and some human studies have shown that altering the amount of either O6 or O3 can affect the function of immune cells (23; references therein):

• Increased dietary intake of EPA (2.7g/day) has shown to reduce PGE2 production (a metabolite of arachidonic acid) in human mononuclear cells (MNCs). 
• Fish oil ingestion has shown to increase the production of 5-series leukotrienes, products derived from EPA.
• EPA/DHA or fish oil also induces the production of resolvins, which have anti-inflammatory properties.
• Increasing membrane permeability by increasing the unsaturation index might increase phagocytosis by MNCs. The phagocytic index of neutrophils and monocytes has shown to be negatively correlated with palmitic acid content, but positively correlated with the content of PUFAs, specifically O3. In healthy humans, 1.5g/day of EPA+DHA for 6 months increased the phagocytic activity in monocytes and neutrophils by 200% and 40%, respectively. 
• Arachidonic acid, EPA and DHA have shown to inhbit T-cell proliferation and IL-2 production in vitro. This has been replicated in animal models with fish oil and/or EPA/DHA in high doses. O3 might also affect the composition (and hence function) of lipid rafts, as treatment of T-cells with O3 displaces acylated proteins anchored to the inner lipid leaflet from lipid rafts, but not GPI-anchored proteins. This displacement (probably as a direct consequence of incorporation of EPA and DHA into membranes) affects the intracellular signaling pathway associated with the protein being displaced, such as LAT. 
• Increasing the amount of dietary fish oil in rats causes a reduction in MHC II expression on dendritic cells, as well as levels of CD2, CD11a and CD18. Arachidonic acid and DHA, by slowing the transit of new MHC I molecules from the endoplasmic reticulum to Golgi, have shown to decrease surface MHC I expression, decreasing cytotoxic T-cell mediated lysis of target cells enriched in these fatty acids. 

The acute effect of increasing doses of animal O3 is a reduction in arachidonic acid-derived inflammatory metabolites, increases in membrane permeability and anti-inflammatory molecules derived from EPA/DHA, as well as reduction in T-cell activation and antigenic stimulation. O3 also have direct effects: inhibition of LPS or lipopeptide-stimulated COX2 expression and LPS-induced NFkB activation (24, 25). Interestingly, there is evidence that the anti-inflammatory effects seen for O3 are dependent on their oxidation. Oxidized EPA, but not unoxidized EPA, inhibits NFkB activation and expression of inflammatory molecules in a PPARa dependent manner, as well as chemotaxis (26, 27, 28). Oxidized, but not unoxidized DHA, inhibits polychlorinated biphenyl-induced NFkB activation and MCP-1 expression, effects probably mediated by its oxidation products (A4/J4 neuroprostanes) (29). Thus, it seems that contrary to what is believed, oxidation of O3 PUFA is necessary to mediate their beneficial biological effects. 

The effects of MUFAs (mainly oleate) have not been studied in detail as with PUFAs. In contrast to palmitate and stearate, oleate do not seems to induce TLR2/4 activation in monocytes (19) (in this case, the authors used a BSA-only control, showing no activation of TLR). This makes sense, as oleate is the main FFA in human circulation (30). However, oleate has a strong inflammatory effect on human islet cells, increasing the levels of IL-1beta mRNA, IL-6 mRNA and IL-8 mRNA compared to palmitate and stearate, effect which was amplified by high glucose levels (31). In contrast with the latter two, the expression of IL-1Ra (antagonist of IL-1) was lower with oleate. The authors suggested that oleate-mediated islet inflammation could be hormetic (which makes complete sense).

Oleate levels in circulation are determined by oral intake as well as de novo synthesis from SFA. This process is mediated by stearoyl-CoA desaturases (SCD), specially SCD-1 in humans (32). This enzyme catalizes the introduction of a single double bond at the delta9, 10 position of long chain acyl-CoAs, preferentially to stearoyl-CoA and palmitoyl-CoA. Over-stimulation of SCD-1 increases the synthesis of MUFAs (like palmitoleoyl-CoA and oleoyl-CoA), affecting intermediary metabolism and promoting obesity, pathological insulin resistance, hypertriglyceridemia and hepatic steatosis (33). SCD-1 expression is induced by SREBP-1, LXR, and inhibited by PPARbeta/delta and PPARgamma. Accordingly, SCD-1 expression and activity is increased with high carbohydrate diets (34, 35), because insulin activates SREBP-1 and glucose (actually glucose-6-phosphate and/or xylulose-5-phosphate) activates ChREBP, which increases SCD-1 expression (36). However, it is important to interpret this data with caution, as lipogenic/lipolytic enzymes in rodents are more active than humans. Nevertheless, a high carbohydrate diet can contribute to the pool of MUFA, thereby influencing the secretion and expression of pro-inflammatory cytokines.

In contrast, EPA has shown to decrease the level of SCD-1 mRNA and SREBP-1c mRNA in Hep G2 cells (37) and omega 3 status is important for controlling the activity and expression of SCD-1 in rats (38). 

Albumin binds fatty acids and LPS

In vitro, one of the most inflammatory fatty acids is lauric acid, which activates NFkB, partially mediated by the TLR4-MyD88/PI3K/Akt pathway, while DHA inhibits this effect (39). SFAs released by adipocytes (mainly palmitate) are also able to activate TLR4 in macrophages, activating NFkB by a mechanism shared partially with LPS (40). It seems that activation of inflammatory genes in different immune cells is related to chain length (41). This suggests that in addition to promoting dimerization and organization of TLR4 with adaptor and co-stimulatory molecules into lipid rafts, some SFA could indeed activate TLR4 independently. What is really interesting is that free fatty acids travel in the bloodstream bound to albumin (and the levels of individual fatty acids correlate with those found free in plasma) (42), and recently, analysis by surface plasmon resonance found that albumin not only binds to LPS, but also modulates its interaction with TLR4 and MD-2, and thus, controls the inflammatory response to a given endotoxin load (results not published)*. So we have a situation in which increasing the dose of O3 PUFA might increase the relative proportion of O3 bound to albumin, thus inhibiting the interaction of palmitic or stearic acid with LPS, or at least, ameliorating it. On the other hand, we can decrase the proportion of saturated fatty acids in circulation and bound to albumin by dietary means (43, 44). Ultimately, the balance between lipolysis and oxidation determines the level of free fatty acids in circulation. A high lipolytic environment uncoupled to mitochondrial oxidation contributes to lipotoxicity and inflammation. This also holds true for LPS-induced inflammation. The perfect balance between hydrolysis of stored fatty acids and oxidation is achieved under fasting conditions. 

TLR4 is modulated by specific fatty acids. Saturated fatty acids (SFAs) activate TLR4 either by direct interaction or by promoting its dimerization and association with MD-2 into lipid rafts. Conversely, omega 3 polyunsaturated fatty acids (PUFAs) inhibit this effect, and reduces SFA-induced TLR4 activation. Both O3 PUFAs as well as O6 PUFAs are able to modulate the immune response by affecting membrane phospholipid composition. Increased levels of arachidonic acid (ARA) derived from O6 metabolism, as well as its metabolites are mainly pro-inflammatory, while those derived from O3 PUFAs are anti-inflammatory. Increasing the proportion of unsaturated fatty acids also changes membrane fluidity, affecting immune functions such as phagocytosis. Activation of TLR4 promotes the association of adaptor molecules such as TIRAP, MyD88 and IRAK. IRAK activates TRAF6, which interacts with TAK1, leading to the phosphorylation (denoted by the yellow P) of IKK/NEMO and its activation. This complex is able to phosphorylate IkB, thus letting transcriptionally active NFkB migrate into the nucleus. Once activated, NFkB increases the expression of pro-inflammatory genes, such as TNFa, IL-1 and IL-6. See text for details.

Summary

The composition of different fatty acids in the diet modulate endotoxemia. From the available evidence, there is consistent research which shows that:

• Saturated fatty acids (SFAs) activate TLR4 and the downstream signaling pathway, ultimately leading to the activation of NFkB, which increases the expression of pro-inflammatory molecules (TNFa, IL-6, etc.).
• SFAs might contribute directly (by interacting with LPS and/or TLR4-MD-2) or indirectly (by reorganizing lipid rafts). In either case, an increase in the level of SFA promotes the activation of this pathway. 
• The activation of TLR4 has been shown to be important for the onset and development of metabolic diseases such as obesity, diabetes and non-alcoholic hepatic steatosis.
• LPS uptake is mediated through chylomicrons and is promoted by a loss of barrier function of the small intestine.
• The level of endotoxemia correlates with baseline and post-prandial triglyceride levels.
• O3 PUFAs (EPA and DHA) have shown an inhibitory effect on LPS and LPS plus SFA-induced TLR4 activation. 
• The oxidation of O3 PUFAs seems to be necessary for their anti-inflammatory effects.
• The level of SFA in the bloodstream is controlled by diet as well as the cellular energy status. 
• MUFAs, in most studies, seem to be neutral. However, there is some evidence linking excess oleate and SCD-1 activity to cellular dysfunction, particularly beta-cell abnormalities. EPA has opposite effects and reduces SCD-1 expression.
• Albumin binds both fatty acids and LPS, and modulates the inflammatory response to a given LPS load. The relative proportion of individual fatty acids bound to albumin might influence the binding of LPS to TLR4, thus affecting the activation of the downstream signaling pathway. 

Practical recommendations

For people with autoimmune and/or inflammatory problems, I recommend the following measures to be taken with respect to fatty acids in the diet:

1. Reduce and control the amount of O6 PUFA, specially from vegetable sources (linoleic acid).
2. Control the amount of SFAs. Consumption of dairy fat seems to be protective against endotoxemia (45). Ghee might be a better option than butter. Better to avoid protein-rich dairy.
3. Increase the amount of marine EPA and DHA (O3 PUFAs). This should work best increasing the consumption of marine foods, but might be a problem for those with leaky gut given the presence of some metals in seafood. Individual tolerance must be assessed. If very sensitive, start with dietary supplements. A high dose (3-5g/day of EPA + DHA) might work first, and the those should be lowered afterwards (46). The higher the baseline triglyceride levels, the higher the dose. Additionally, the worse the inflammatory/immune status, the higher and longer the supplementation. This can be assessed using traditional blood markers (C-reactive protein, etc.) and symptoms. It has been shown that the effects of O3 supplementation are influenced by the O3 status of the subject (47). High inflammatory markers and/or symptoms might reflect O3 status.
4. Avoid industrial trans-fatty acids.

* Work was presented in the Inmunoperu 2012 conference. More information when available.