In the FIRST POST of this two part series, I laid out some facts about Type 2 diabetes which results from insulin resistance, and indicated how non-esterified (‘Free’) Fatty Acids (FFAs) induce chronic inflammation via engagement of TLR4 and the NF-κB pathway, eventually leading to Insulin resistance – and yet, since FFA doesn’t bind TLR4, it’s not known how the twain meets. The elegant set of studies described in the Pal at al. paper in the July 29, 2012 issue of Nature Medicine  provides evidence for a mechanism hypothesized to be active in lipid-induced insulin resistance, i.e., one that can connect the dots.
The hypothesis hinges around Fetuin-A (or FetA, as it’s affectionately called; officially known as α-2-Heremans-Schmid-glycoprotein, AHSG for short), a glycoprotein molecule secreted by the liver. It is a known culprit in genesis of insulin resistance, because it binds to and inhibits the insulin receptors (which belong to the Tyrosine Kinase enzyme family) in hepatocytes and skeletal muscles, hampering insulin signal transduction (response to insulin) in those tissues. FetA also reduces the level of Adiponectin, a protein with protective effects released from adipose (fat) tissues, thereby contributing further to insulin resistance, as well as to certain renal pathologies and a range of liver pathologies designated ‘non-alcoholic fatty liver disease’ . Many of these pathologies involve unmitigated chronic inflammation, in which FetA is considered a biomarker (sort of a disease ‘signature’), since it is known to stimulate the production of inflammatory cytokines from adipocytes and macrophages (a type of cell responsible for innate immune defence).
You know what they say about ‘birds of a feather’ flocking together? Yup, happens a lot in the body. It appears that FFAs significantly enhance FetA expression by increasing NF-κB binding to its promoter , and as mentioned above, elevation of serum and plasma FetA levels have proinflammatory effects, as borne by both in vitro and in vivo experiments. Genetically modified mice that cannot make FetA or lack functional TLR4 show no insulin resistance in response a high fat diet (HFD) containing 65% fat, unlike normal mice.
FetA has also been known – for two decades, no less – as a major carrier for FFAs. This action is beneficial during development; in certain cell types, FetA stimulates the incorporation of fatty acids into mono-, di- and triglycerides, which can act as energy depots. However, like so many other processes in the body, the same action may lead to deleterious effects, as in the scenarios of chronic inflammatory diseases and insulin resistance. Pal and her colleagues considered all the facts about FFA and FetA, and came up with the hypothesis that FetA may be that unknown connector of dots, the molecule that delivers FFA to TLR4. The corroboration of their hypothesis came from both human subjects and mouse experiments; comparing several pertinent variables, namely serum FetA concentration, expression of TLR4 as well as pro-inflammatory cytokines IL-6 and TNF-α and activation of NF-κB in adipocytes, they found significantly higher values in obese diabetic human subjects, mice made insulin resistant by feeding HFD, and mice genetically modified to make dyslipidemic (i.e., with dysregulation of lipid metabolism) than, respectively, in non-obese non-diabetic humans, normal insulin-sensitive mice on regular diet, and normal (‘wild type’; unmodified) mice. This observation indicated a possible association between FetA, TLR4, as well as insulin resistance developed in response to high levels of fat.
The link of FetA with levels of NF-κB, IL6 and TNF-α was additionally corroborated by the observation of significant reduction of the serum levels of these proteins in mice which had a partial hepatectomy (i.e., part of liver excised; less liver = less FetA made by liver) compared to their “fully-livered” counterparts placed on same HFD. Further, the requirement of TLR4 for the stimulatory effects of FetA on NF-κB was established by the finding that macrophages and adipocytes from mice which cannot make TLR4 did not show NF-κB activation in response to FetA.
To establish the mechanisms of complex biological processes, researchers often need to formulate questions in depth and test the hypotheses from multiple angles. In order to demonstrate the correlation between FetA and TLR4 activities, Pal and her colleagues made mice insulin resistant via HFD, and then disabled the genes for FetA or TLR4 (‘knockdown’ or KD model); not surprisingly, the diet could not induce or sustain insulin resistance in FetAKD or TLR4KD mice, as estimated from observations of several variables, such as blood glucose levels; glucose uptake tests; tests of glucose tolerance (GTT) and insulin tolerance (ITT); homeostasis model assessment–insulin resistance (HOMA-IR) scores; and activation (via phosphorylation) of insulin receptor-β (Ir-β) and Akt (both molecules involved in insulin signaling). In addition, FetAKD or TLR4KD mice showed lower NF-κB activity, along with the marked reduction of IL-6 and TNF-α mRNA expression (as well as of TLR4 activation) in their adipocytes (compared to those from their ‘wild type’ counterparts on same diet). Reintroduction of purified FetA in the FetAKD mice resulted in regaining of insulin resistance.
So… Does TLR4-signaling need FetA? More importantly, does TLR4 signaling induced by FFA require FetA? To find out, the researchers used Palmitate, a.k.a hexadecanoic acid, a 16-Carbon free fatty acid that is synthesized first in the Fatty Acid biosynthesis pathway. Their in vitro model system used one of two types of cells, human adipocytes and 3T3-L1 (a mouse cancer cell line with the ability to differentiate into adipocytes), in which they checked for signs of TLR4 signaling and inflammation (i.e., activation of NF-κB, as well as elevation of the expression of pro-inflammatory cytokines IL6 and TNF-α) in presence of only FFA, only FetA or both. They also examined release of IL6 and TNF-α from macrophages (a mouse macrophage cell line, RAW264.7) under these conditions. To establish the TLR4 dependence of these processes, they included conditions in which TLR4 expression was artificially suppressed. The observations were quite illuminating.
- FFA alone was rather marginally stimulatory towards pro-inflammatory processes, but this marginal effect seemed to be independent of TLR4; TLR4 suppression had no effect.
- FetA alone increased gene expression, as well as protein level, of NF-κB, and enhanced its binding to IL6 promoter. It also upregulated the secretion of IL6 and TNF-α from macrophages. And all its effects were TLR4 dependent, abrogated by TLR4 suppression/inactivation. And not only in these few cell types, similar observations about TLR4-dependent FetA activity was made in primary macrophages and hepatocytes isolated from both standard-diet- and HFD-fed mice, where absence of a functional TLR4 (in TLR4KD mice) prevented FetA from activating NF-κB.
- However, significant signs of inflammatory activity was observed only when both FFA and FetA were present; the combination markedly upregulated NF-κB, IL6 and TNF-α at the level of both gene expression and protein. Suppression of TLR4 expression, or the absence of functional molecules downstream of TLR4 in the signal transduction pathway (such as MyD88 or MAPK Kinase; refer to schematic diagram in the FIRST POST), predictably reversed these changes and prevented NF-κB activation.
These observations indicated that both FFA and FetA were necessary for inflammatory response via TLR4 signaling. They also found that in human adipocytes, the combination of FFA and FetA greatly upregulated (= increased) the gene expression of various other pro-inflammatory cytokines that depend upon TLR4, whereas FFA alone was marginally effective. In addition, in insulin sensitivity assays (where 3T3-L1 adipocytes take up glucose under influence of insulin), FFA-FetA combination exerted the greatest inhibitory effect on this process, compared to either used alone.
In scientific research, observations made at any given time are often rechecked subsequently under different conditions by other investigators, and refined if necessary. That’s how knowledge progresses. So, if both FFA and FetA are necessary, what about the previous studies which concluded FFAs alone could activate TLR4 signaling in adipocytes? Pal and her colleagues have a possible answer. FBS (fetal bovine serum) is a commonly used nutrient enrichment for culture media used the world over. However, FBS may contain significantly high concentrations of FetA of bovine origin – which, understandably, may have influenced the earlier observations. Indeed, when Pal used a special type of medium that required no exogenous serum, FFAs alone appeared to have only a marginal stimulatory effect, and one which is not mediated via TLR4.
FFAs, then, sorely need FetA to induce insulin resistance via the TLR4 pathway. Do FFAs shake hands with FetA and go their merry ways? Evidence suggests that they hug, tightly. Pal and her colleagues found evidence that FFAs physically bind FetA, which interacts with TLR4 to form a ternary complex.
Of course, not content to leave matters d’ménage à trois as they are, Pal and her colleagues further delved painstakingly into the actual interaction between FetA and TLR4, revealing a high affinity binding between a terminal carbohydrate (α-galactoside moiety) present on FetA and two specific amino acid regions (Leucine-rich repeats, or LRRs) on the extracellular domain of TLR4.
Long as it is, my description of this work is rather trivial compared to the amazing amount of work that has gone into this paper. Pal and her colleagues hypothesized that an endogenous TLR4 ligand presented the FFAs to the TLR4 for their known action on insulin resistance, thereby connecting the dots, so to speak. In this study, they laid out FetA as a plausible candidate for that ligand. Of course, this opens up further possibilities of FetA being a potential target for therapeutic drugs against insulin resistance and type 2 diabetes.
However, nothing – sadly – is ever absolute in biological systems. Like many, many other biological molecules, FetA has distinct physiological functions, particularly in development of brain and bone tissue; in addition, FetA prevents pathological soft-tissue calcification, and exerts a profound anti-inflammatory effect associated with maternal tolerance of the fetus during pregnancy . Some researchers even doubt the exact role of FetA in insulin resistance as described in the discussed paper; the in vitro and ex vivo measurements made are stark in their outcome, but FetA regulation in the scenario of HFD may need to be reconciled, in view of the conflicting data, in terms of timing of events and causality . As Pal and her colleagues found, the amount of fat present in the diet (more or less than 60%) appears to influence the genesis of insulin resistance, which may – in turn – have influenced the earlier studies.
Nevertheless, this study by Pal et al. represents an important step in understanding the biology associated with inflammatory mediators of insulin resistance and type 2 diabetes, especially in the context of obesity and dietary lipid content.
- Pal D, Dasgupta S, Kundu R, Maitra S, Das G, Mukhopadhyay S, Ray S, Majumdar SS, & Bhattacharya S (2012). Fetuin-A acts as an endogenous ligand of TLR4 to promote lipid-induced insulin resistance. Nature medicine, 18, 1279-1285 PMID: 22842477
- Ix JH, & Sharma K (2010). Mechanisms linking obesity, chronic kidney disease, and fatty liver disease: the roles of fetuin-A, adiponectin, and AMPK. Journal of the American Society of Nephrology : JASN, 21 (3), 406-12 PMID: 20150538
- Dasgupta S, Bhattacharya S, Biswas A, Majumdar SS, Mukhopadhyay S, Ray S, & Bhattacharya S (2010). NF-kappaB mediates lipid-induced fetuin-A expression in hepatocytes that impairs adipocyte function effecting insulin resistance. The Biochemical journal, 429 (3), 451-62 PMID: 20482516
- Jahnen-Dechent W, Heiss A, Schäfer C, & Ketteler M (2011). Fetuin-A regulation of calcified matrix metabolism. Circulation research, 108 (12), 1494-509 PMID: 21659653
Neither the mouse model here nor the conclusions appear particularly helpful. Disappointing. I’m expecting more new findings to come out regarding t2d, obesity, tlr4, inflammation, bacterial endotoxinemia etc.
Thanks for the summary.