Diabetes, Chronic Inflammation and Neutrophils – another part of the story

I hope by now you, my gentle reader(s), are familiar with the story of diabetes, insulin resistance and their link with chronic inflammation. Let me emphasize once more. Inflammation per se is not a bad thing; it is an essential process in the immune defence of the body. Whenever there is an invasion of a tissue, internal or external, by a foreign substance – such as a pathogen, the inflammatory process, initiated by signals from the primary defence cells in the tissue (such as Macrophages), calls in reinforcements from bone marrow derived immune cells, the secondary defence system of the body – in form of leukocytes (a.k.a. White Blood Cells/WBCs; e.g. neutrophils, T- and B-lymphocytes, monocytes, and sometimes basophils and eosinophils) and/or mast cells, from blood into the tissue. These defence cells communicate with each other and perform their action by releasing chemical substances known as cytokines and chemokines, that act as messengers.

Neutrophils, in particular, are excellent defenders. They can kill pathogens such as bacteria by engulfing (a process called phagocytosis) and destroying them using highly corrosive chemical substances (such as superoxide anions); sometimes, they also break open (a process called degranulation) and release various enzymes and other proteins which can poke holes into bacteria. Within the last decade, another role of neutrophils has come to light, the Spiderman of the host immune defence! It appears that neutrophils can weave an extracellular web (a.k.a. neutrophil extracellular trap) composed of chunks of chromosomal proteins, DNA and antimicrobial proteins and enzymes – which can trap and kill microbial pathogens [1]. Neat, huh?!

Neutrophils in blood smear stained with Giemsa Stain; Image adapted from Original Image in Wikimedia Commons licensed under CC-by-SA)

And yet, like in most processes in the body, a delicate balance is required in the inflammatory process, too. Dysregulation of the inflammatory process, such as non-specific activation, or activation in response to the host body’s own components, can lead to extensive tissue damage and various types of inflammatory diseases. Insulin resistance, along with a decrease in Glucose tolerance, is just one of them.

Chronic, low-grade inflammation, occurring in the liver and adipose (fat-containing) tissue, is causally linked to systemic insulin resistance and low tissue insulin sensitivity, both hallmarks of Type 2 Diabetes, as well as of obesity. Adipose tissue in mice which became obese as a result of a diet high in fat (High Fat Diet, or HFD) contain a type of inflammation-causing (‘pro-inflammatory’) macrophages, which secrete several pro-inflammatory cytokine/chemokines (such as TNFα, IL-1β and IL-6, never mind the acronyms) with the ability to directly reduce insulin sensitivity (check out this brilliant review by Olefsky et al. [2]). Several other immune cell types may contribute to this process, too.

However, when there is inflammation, can neutrophils – the pre-eminent inflammatory mediators – be far behind? It appears not. A recent study in Nature Medicine by Talukdar et al. (Professor Olefsky’s group), published online on 5 August 2012, implicates neutrophils in the genesis of insulin resistance, via the action of a proteolytic (protein-destroying) enzyme, neutrophil elastase (a serine protease secreted by neutrophils during inflammation) [3].

Let’s look at the evidence the researchers accumulated. It was observed earlier than mice on HFD show a recruitment of neutrophils to the intra-abdominal adipose tissue (sort of murine love-handles!). Talukdar and his associates followed the time course of movement of neutrophils (a process called infiltration) into adipose tissue of HFD (60% fat)-fed obese mice. After the mice ate the HFD for 3 days, the number of neutrophils in the adipose tissue increased rapidly (corroborating earlier reports), remaining chronically high (for up to 90 days) on the same diet; these numbers were a lot lower in mice on regular diet.

Caveolin (blue) stained adipocytes, showing neutrophil localization (yellow-orange, merge of red and green) in HFD-fed mice; chow = regular diet; Figure ©Nature; adapted for educational purpose from Figure 1 of Talukdar et al. (2012) Nature Medicine 18, 1407–1412

So, once in the tissues, what do the neutrophils do to cause insulin resistance and reduction of glucose tolerance? Talukdar thinks that the neutrophil elastase enzyme, mentioned above, is to blame. After all, the expression and activity of neutrophil elastase in the adipocytes of HFD-fed obese mice went way up, and when treated with an inhibitor of the neutrophil elastase enzyme for 14 days (whether administered at the beginning of the HFD regimen or later), these mice showed substantial improvement in glucose tolerance with no change in body weight; in the reverse experiment, mice on regular diet given treatment with neutrophil elastase showed marked reduction in glucose tolerance compared to untreated mice.

A little background on liver here. Hepatocytes (the cells of the liver) essay an important role in maintaining the balance of glucose in blood and tissues, by virtue of their abilities to take up (‘uptake’) glucose to store it (in form of glycogen), and release it either by breaking down glycogen (a process called glycogenolysis) or by making it de novo from non carbohydrate sources, such as amino acids and lipids (a process called gluconeogenesis). Although it appears that the liver cells can respond adequately to blood glucose levels by modifying the activities of its own enzymes, two hormones, Insulin and Glucagon, released respectively from the β- and α-cells of the pancreas, play very important roles by their direct and indirect actions on hepatocytes; their detailed roles are for another discussion. Suffice it to say, glucagon stimulates hepatic glucose production (or HGP, for short) by enhancing glycogenolysis and limiting glucose uptake by hepatocytes, thereby increasing blood glucose – which comes in handy during states of low blood sugar (hypoglycemia) or high blood sugar demand, as in exercise. In vitro, glucagon treatment may almost double the glucose output from primary (i.e., taken directly from the tissue and not from an immortalized cell line) hepatocytes. As mentioned earlier, insulin’s action is exactly opposite. By direct (interacting directly with hepatocytes via liver insulin receptors) and/or indirect methods (suppressing glucagon release from α-cells, and non-esterified/’free’ fatty acid (FFA, remember?) release from adipocytes; reducing mobilization of lipids to liver; asking the brain to shake a peremptory finger at liver and say, “Oh no, you don’t!”), insulin enhances glucose uptake by hepatocytes and reduces HGP to basal levels.

Keeping these facts in mind, Talukdar and associates devised experiments to determine how neutrophil elastase fitted into this picture. They had access to mice that have been genetically modified to lack a functional neutrophil elastase (a.k.a. NEKO mouse) – a valuable tool in the study of the function of that enzyme. They found that NEKO mice, when placed on HFD, gained somewhat less weight compared to wild type (naïve and unmodified) mice; liver weight of NEKO mice was substantially lower, whereas weight of White Adipose Tissue (WAT; one of body’s energy reservoirs) was higher, than in their wild type counterparts. NEKO mice were also more resistant to the effects of HFD; after 10 weeks of HFD, NEKO mice still had higher glucose tolerance, better insulin sensitivity, and lower fasting insulin concentration (indicative of better maintenance of glucose homeostasis) than wild type mice.

Not surprisingly, NEKO mice, when fed on HFD, managed to hold off neutrophil infiltration into adipose tissue much better than did wild type mice. Correspondingly, obese wild type mice in which neutrophil elastase is inactivated by treating with an inhibitor also show a marked reduction in adipose tissue neutrophils. Not just in adipose tissue, obese humans with an inflammatory condition called non-alcholic fatty liver disease (formally, ‘steatohepatitis’) have neutrophil infiltration into liver. Talukdar and associates noted that wild type mice, if given HFD instead of regular diet, had higher neutrophil numbers in liver, but NEKO mice, with no neutrophil elastase, did not respond to HFD with elevation of hepatic neutrophils.

Stained sections of liver from regular diet (chow)-fed wild type mice, HFD-fed wild type mice and HFD-fed NEKO. Red indicates lipid content in hepatocytes, blue shows the nuclei of hepatocytes, and green indicates neutrophils recruited to the tissue, seen only in liver of wild type mice on HFD;
Figure ©Nature; adapted for educational purpose from Figure 3 of Talukdar et al. (2012) Nature Medicine 18, 1407–1412

It is not immediately clear why inhibition of neutrophil elastase or its absence (as in NEKO mice) results in a decrease of neutrophil infiltration into tissue. Talukdar and associates surmised that it may be a feedback/feed-forward issue; generally in the immune system, cells are recruited at the site of injury by release and amplification of signals from chemical messengers (cytokines and chemokines). So less neutrophil activation in the target tissue may mean less release of neutrophil elastase, resulting in less stimulation (via the TLR4 pathway, on which depends the pro-inflammatory effects of the elastase – as observed by Talukdar’s group and others) of chemokine release from the cells, which has an overall dampening effect on further neutrophil recruitment. However, clearly, the enzyme neutrophil elastase was involved in the glucose metabolism process; its absence improved glucose tolerance, and vice versa.

Therefore, possibly because of the absence of the elastase, NEKO mice generally have a low inflammatory phenotype, with lower serum levels of various pro-inflammatory chemical messengers compared to wild type mice. Both liver and adipose tissue from NEKO mice (in comparison to wild type mice) show substantially lower expression of pro-inflammatory genes (i.e., genes responsible for TNFα, F4/80, KC, IL1R1, IL1β, MCP1, CD68, IRF4, IRF5, and so forth) and higher expression of anti-inflammatory genes (such as Arginase, MGL1 and IL4). Even isolated adipocytes and stromal vascular cells (neutrophils from adipose tissue) from NEKO mice expressed lower levels of inflammatory markers than those from wild type mice. Not only that, the pro-inflammatory macrophages reported to be recruited to adipose tissue during inflammation [2] were lower in abundance in NEKO mice, possibly because of marked reduction in recruitment and polarization of those cells in absence of neutrophil elastase. Adipose tissue from NEKO mice (in comparison to wild type mice) generally show greater insulin sensitivity, with higher expression of insulin-sensitive genes (such as GLUT4, responsible for a glucose transporter) and genes for various lipolytic enzymes, and better glucose uptake both basally and in response to insulin.

Comparison of the NEKO mice with their wild type counterparts yielded valuable information about the function of neutrophil elastase. In its absence, insulin acted faster to push glucose away from blood, and also had a greater degree of inhibition of HGP, as well as better suppression of FFAs. It appeared that in NEKO mice, the genetic deletion of this enzyme confers improved insulin sensitivity to both hepatocytes and adipose tissue. In response to insulin delivered intravenously, NEKO mice showed evidence of higher levels of insulin signaling in hepatic and adipose tissue – as measured by greater activation (via phosphorylation) of an enzyme called AKT (a.k.a. Protein Kinase B, a key component in the signaling pathway that ultimately promotes glycogen formation in response to insulin).

A few words about a unique signal transduction protein Insulin Receptor Substrate 1 (IRS1) wouldn’t be remiss here. This protein plays an important role in transmitting signals from certain extracellular molecules (such as insulin; or Insulin-like Growth Factor 1) into the cell where the signals are further processed. For example, binding of insulin to parts of the insulin receptors exposed on the cell surface causes IRS1 to bind to the same receptors inside the cell, which, in turn, activates several signaling pathways.

You do realize where this is going, right? Yes, the extracellular neutrophil elastase can reach in from without to mediate the degradation of IRS1 within the cell. Cue scary music. That sounds a bit dramatic, no? IRS1 is one of the substrates for neutrophil elastase, and in vitro, one unit of the enzyme can destroy (by hydrolysis) 100 units of IRS1. However, IRS1 isn’t expressed on the cell surface; then how does the elastase interact with IRS1? It appears that at physiologic concentrations, neutrophil elastase may be transported actively (i.e., spending cellular energy) to inside the cell [4], and once in… You get the picture.

Predictably, then, NEKO mice was found to have a higher concentration of IRS1 protein (as well as expression of the Irs1 gene) in liver and adipose tissue. But does the elastase inhibit hepatic IRS1-mediated insulin signaling? The researchers took fasting mice (so that basal insulin secretion is at a minimum), and gave them neutrophil elastase, following it up with injection of insulin in some. Regardless of whether insulin was at basal levels or acutely administered, its effects were countered by neutrophil elastase treatment; in wild type mice given the elastase, IRS1 was reduced in liver and adipose tissue, and phosphorylated (=activated) AKT levels were down. Insulin administration in fasting NEKO mice, however, resulted in higher IRS1 expression in liver and adipose tissue, as expected. These observations were corroborated in vitro. Neutrophil elastase, added directly on to primary mouse or human hepatocytes, as well as to 3T3-L1 adipocyte cell line, markedly reduced IRS1 protein content, which in turn suppressed the ability of these cells to respond to insulin by reducing the activation of AKT.

What were the biological significances of the neutrophil elastase-mediated suppression of insulin signaling? To investigate, Talukdar and associates measured in vitro glucose output (a surrogate for HGP) in primary mouse hepatocytes with and without neutrophil elastase treatment. As I mentioned above, this is a function ordinarily stimulated by glucagon and suppressed by insulin. However, neutrophil elastase administration alone – without the presence of glucagon in the system – could stimulate basal glucose output by 50% – an effect that insulin did not inhibit; neutrophil elastase treatment also diminished the usual insulin-mediated inhibition of glucagon-induced HGP.

In vivo, in livers of NEKO mice (but not in their wild-type counterparts with functional neutrophil elastase), the genes responsible for making lipids and cholesterol were suppressed, whereas those responsible for energy metabolism (incidentally, same genes that are activated by exercise) were activated. In fact, liver from NEKO mice was making far less lipids. Even when maintained on HFD, which induces obesity in wild type mice, NEKO mice were protected from adipose tissue and liver inflammation, having retained insulin sensitivity in adipose tissue and hepatocytes. In addition, obese wild type mice showed marked improvement of glucose tolerance following treatment with a small-molecule neutrophil elastase inhibitor.

In summary, therefore, neutrophils appear to be suspect number one in the chronic inflammation that characterizes obesity and insulin resistance (and therefore, type 2 diabetes). From the observations in this study, Talukdar and his fellow authors conclude that neutrophils should be considered a key component in this phenomenon, in which the elastase secreted from neutrophils may be prime coordinator in the inflammatory process. The neutrophil elastase appears to (a) cause IRS1 degradation, (b) reduce insulin signaling, (c) enhance glucose production, and (d) derange lipid metabolism, all of which contribute towards an increase in cellular insulin resistance, and these effects are observed in primary mouse and human hepatocytes, adipocyte cell lines, as well as in the liver and adipose tissues of obese mice (and lean mice treated with the elastase enzyme). That’s a pretty serious indictment, I should say.

So why not get rid of the neutrophil elastase altogether? Sadly, in the body, things are hardly ever that simple. The elastase enzyme, packed into azurophilic granules, forms an important tool in neutrophils’ anti-microbial arsenal; specifically, this elastase is known to target and degrade virulence proteins from various nasty Enterobacteria, such as E. coli, Shigella, Yersinia and Salmonella [5]. A much better option to counter the effects of obesity and stave off type 2 diabetes seems to be regular exercise; apart from its obvious, overt effects, exercise appears to be active in regulating expression of the genes of pathways associated with glucose and lipid metabolism, and may effectively circumvent the negative effects of neutrophil elastase. Yup, no escape from those morning walks and jumping jacks and squats for me.

Sources:

1. Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss DS, Weinrauch Y, & Zychlinsky A (2004). Neutrophil extracellular traps kill bacteria. Science (New York, N.Y.), 303 (5663), 1532-5 PMID: 15001782
2. Olefsky JM, & Glass CK (2010). Macrophages, inflammation, and insulin resistance. Annual review of physiology, 72, 219-46 PMID: 20148674
3. Talukdar S, Oh DY, Bandyopadhyay G, Li D, Xu J, McNelis J, Lu M, Li P, Yan Q, Zhu Y, Ofrecio J, Lin M, Brenner MB, & Olefsky JM (2012). Neutrophils mediate insulin resistance in mice fed a high-fat diet through secreted elastase. Nature medicine PMID: 22863787
4. Houghton AM, Rzymkiewicz DM, Ji H, Gregory AD, Egea EE, Metz HE, Stolz DB, Land SR, Marconcini LA, Kliment CR, Jenkins KM, Beaulieu KA, Mouded M, Frank SJ, Wong KK, & Shapiro SD (2010). Neutrophil elastase-mediated degradation of IRS-1 accelerates lung tumor growth. Nature medicine, 16 (2), 219-23 PMID: 20081861
5. Weinrauch Y, Drujan D, Shapiro SD, Weiss J, & Zychlinsky A (2002). Neutrophil elastase targets virulence factors of enterobacteria. Nature, 417 (6884), 91-4 PMID: 12018205