Fructose is a monosaccharide found mainly in sucrose (50% glucose and 50% fructose), fruits, honey as well as in processed forms like fructose-containing caloric sweeteners (FCCS), high fructose corn syrup (HFCS) and employed in processed foods and beverages called sugar-sweetened beverages (SSBs) (21). Epidemiological studies have related fructose consumption (in sugar, or HFCS form) with obesity (22), MetS , T2DM (23), CVD (24) and NAFLD (6,25-27). The correlation with SSBs is particularly strong (28, 29). The mechanism of how fructose participates in these pathologies is not completely clear yet, however, different studies in both animals and humans (30, 31) has allowed the dissection of some of its metabolic effects.

Free fructose is absorbed directly in the intestinal lumen, whereas from larger molecules like sucrose, both glucose and fructose are acquired by the cleavage of sucrase (invertase), an enzyme found in the brush border of the villi or enterocyte of the small intestine (32, 33). Intestinal fructose is mostly transported via the glucose transporter 5 (GLUT5) via diffusion on the luminal side and glucose transporter 2 (GLUT2) on the basolateral side (33, 34). Fructose enters the liver from the portal circulation (32, 35). The liver contains two glucose and two fructose transporters, GLUT 2 and GLUT 8 respectively (Figure 1). Fructose transport and metabolism within hepatocytes is regulated by GLUT 8. (36, 37). Fructose is metabolized mostly in the liver (more than 80% undergoes first pass extraction), whereas when consumed in isolation, approximately 50% is converted to glucose, 15-20% into hepatic glycogen and 15-25% into lactate or fatty acids (FA) which are secreted as very low-density lipoproteins (VLDL) triglycerides (TG) or stored as intrahepatic fat (38-40).

Figure 1

It’s not all about the calories nor about all carbohydrates but a specific one. This diagram shows the comparison of the major pathways for the fate of either A) 100 g of glucose (from starch) or B) 100 of sugar (50% glucose and 50% fructose). A). After digestion of starch 1), glucose enters the portal vein 2). In the liver it is converted, in part, to glycogen 3) and most of it goes into the bloodstream 4) to feed the tissues as it increases insulin secretion 5) and glucose enters muscle and adipose tissues 6). The rest is used to fuel the liver itself via glycolysis 7) leading to Acetyl coenzyme A (AcCoA) 8) which generates energy in the mitochondria. Very little is converted to fat 9) via the process of de novo lipogenesis (DNL). B). After digestion of sugar 1), glucose and fructose enter the portal vein in equal amounts 2). In the liver, glucose will be turned, in part, into glycogen 3). Most of it enters the bloodstream 4) to feed the tissues as it increases insulin secretion 5) and glucose enters muscle and adipose tissues 6). Fructose does not leave the liver for the most part. Instead, it is quickly phosphorylated by FFK C, bypassing regulatory steps in glycolysis and flooding the system 7), 8). The trapped metabolites have one fate: they are turned into fat by de novo lipogenesis 9) and 10). This process impairs FA oxidation by the mitochondrion because malonylCoA inhibits carnitine palmitoyl transferase I (CPT I) and FA transport into the mitochondrion 9). Some of the trioses are also transformed into the toxic metabolite methylglyoxal (MG), which can be detoxified to D-lactate 11). These processes have dire consequences as explained in the next figures. Finally, as further developed in other diagrams, quick phosphorylation of fructose leads to energy depletion and uric acid production, which in turn stimulates fructose metabolism 12).

In the liver, three key enzymes metabolize fructose. First, fructose is phosphorylated to fructose 1 phosphate (fructose-1-P) by the enzyme fructokinase C (FFK C), also named ketohexokinase (KHK). Fructose-1-P is then converted into di-hidroxyacetone-phosphate (DHAP) by the enzyme aldolase B and glyceraldehyde-3-phosphate (G-3-P) via thiokinase (TKFC). These trioses participate in other metabolic pathways: glycolysis, lipid synthesis, gluconeogenesis, and glycogenesis (41, 42) (Figure 1). It is important to note that fructose enters glycolysis more directly, and consequently is not tightly regulated as glucose (19).

Most of the ingested fructose is extracted from the portal blood via first pass hepatic metabolism while only a small fraction of the ingested fructose will eventually enter the systemic circulation (40). It needs to be said that we rarely consume pure fructose, rather, it is co-ingested with glucose and this makes all the difference. To better highlight these differences, Figure 1 A and B compares what happens with a load of glucose (from pasta, for instance) and the same load of sucrose (fructose and glucose).

Some studies in animals (43, 44) and humans (45, 46) have shown that fructose, compared with glucose or starch in diets with the same number of calories, is able to increase food intake, visceral fat, circulating TGs, blood pressure and reduces fatty acid oxidation, insulin sensitivity and energy metabolism (47, 48). All of these characteristics are related to the presence of MetS and various scientific evidence shows that drinks sweetened with fructose or HFCS have a role in the pathogenesis of MetS and its components (6, 11, 23).

It is noteworthy that while glucose generates energy in the form of ATP during its metabolism, fructose consumption is able to the decrease the levels of intracellular ATP due to the quick process of phosphorylation by FFK C (Figure 1B). As we show further below, replenishing of the ATP increases AMP leading to its catabolism into uric acid. The lack of ATP in turn generates a mitochondrial oxidative stress that favors an increase of lipogenesis, blockade of the oxidation of FA (46) and stimulates gluconeogenesis (49-51) as we further elaborate in section 3.2. On the other hand, while the metabolism of glucose is limited by the amount of ATP and insulin, the metabolism of fructose is not limited by these factors.

As previously mentioned, the result of fructose ingestion (Figure 1) may first be evidenced by an increase in hepatic glucose production and the conversion to lactate in the liver which can be measured in the blood. After this, an increase in plasma lipids is observed due to the production of fat from fructose in the liver. As reported by multiple authors, high fructose concentrations converts pyruvate to acetyl-CoA by the reaction of pyruvate dehydrogenase. The flux of pyruvate dehydrogenase from increased entry into the TCA cycle also results in an increased acetyl-CoA and citrate cycled in the synthesis of fatty acids (52) which are stored as intrahepatic fat and/or secreted into the bloodstream as VLDL triglyceride. Liver fat accumulation is a key link to IR, an entity linked to MetS and NAFLD (53-55).

There is a close relationship between fructose consumption, DNL (FA and TG synthesis) and NAFLD. Fructose increases hepatic FFK C and induces DNL (25) which is increased in NAFLD (56, 57), a process characterized by an imbalance between the lipids synthesized via DNL or lipolysis and lipid oxidation or VLDL export from liver (58). The excess fat in the liver may lead to the development of hepatic IR (59) as well as nonalchoholic steatohepatitis (NASH), a stage that predisposes to cirrhosis (60, 61) and its complications (32, 35, 62).

Isotopic studies have shown that people with NAFLD produce 2 times more liver fat and secrete more VLDL-triglycerides via DNL compared to IR obese subjects and 3 times more compared with healthy subjects (9, 32). Moreover, prolonged exposure of lipids in the liver causes oxidative stress in the endoplasmic reticulum ER and this alters apolipoprotein B100 degradation as well as VLDL secretion (63), a condition described in people with NASH (61).

The main deleterious effects of fructose at the hepatic and systemic level include: insulin resistance, inflammation, stress hepatic, ATP depletion (64), DNL (triglyceride and fatty acid synthesis) (65-69), NAFLD, nonalcoholic steatohepatitis (NASH) (67, 70), acid uric production (47, 71), endoplasmic reticulum stress (ER), fibrosis (9)(71). These will be explored further in the following sections. In addition, we have proposed that the increase in trioses flux that increases lipogenesis should also greatly increase the generation of methylglyoxal (MG) and its detoxification product, D-lactate (72, 73). The importance of fructose metabolism in fatty liver disease is highlighted clearly by the fact that Pfizer is developing (phase 1) PF-06835918 a FK C inhibitor.

Fructose participates in formation of methylglyoxal compound (MG), a powerful precursor of advanced glycation end products (AGEs) formed in vivo (which are described in another section of this review). MG is detoxified as D-lactate. Trioses formed in the unregulated metabolism of fructose may increase the MG production in the liver (74, 75). This increase of MG generates dicarbonyl stress, which is characterized by modification/dysfunction of proteins (MG attacks especially arginine residues) and DNA (76-78). We have advanced the hypothesis that MG inactivates the enzyme adenosyl-monophosphate-dependent kinase (AMPK), which under normal conditions would activate the catabolic pathways in the liver. However, MG may have affinity for the three arginines of the subunit gamma of AMPK. When coupled to them, AMPK is inactivated thereby favoring the anabolic processes including lipogenesis and IR which are widely related to obesity, metabolic syndrome and NAFLD (73).

AMPK is a master regulatory enzyme that controls the cellular energy state (79-81). A decrease in energy activates AMPK by initiating catabolic pathways and inhibiting anabolic pathways (79, 82). AMPK is comprised of three sub units: alpha, beta and gamma (its allosteric site). The epsilon subunit is linked to AMP by 3 Arginine residues (79, 82). The allosteric regulation is influenced by the AMP/ATP ratio and blocking the allosteric site of AMP can inhibit activation of AMPK. Related to this, as previously described, the particular metabolism of fructose leads to the formation of triose (catalyzed by FFK C), a process that favors a rapid depletion of ATP (51, 83), while at the same time, AMP production forms uric acid. This change in proportion of ATP/AMP should activate AMPK with its consequent effects, however, under the consumption of fructose this does not happen. To explain this flagrant metabolic paradox, as shown in Figure 2.

Figure 2

Some of the deleterious actions of fructose on the liver may be due to the actions of methylglyoxal on master regulatory enzyme AMPK. As shown in Figure 1 B, fructose metabolism is largely hepatic. we have proposed that a surge of fructose (40 g in liquid form which is not uncommon in our diet), through unregulated metabolism, generates MG 1), which can be detoxified to D-lactate 2), and we use as a marker of this flux (76), a process that may be overwhelmed. MG is very reactive and may bind to the 3 key arginine residues in the allosteric site of AMPK, rendering it non responsive 3). AMPK favors energy generation, its cumulative actions may be summarized as anti-diabetic. If rendered inactive, the processes favored are gluconeogenesis 4), increasing hepatic production of glucose even in the fed state, lipogenesis 5) and cholesterol synthesis 6). These are precisely the processes which research has shown are stimulated by fructose, with the consequences of ectopic fat accumulation 7), hyperlipidemia 8), insulin hypersecretion 9) and therefore insulin resistance (53, 84). Further research is needed to fully establish the above as a clinically relevant mechanism.

It has been demonstrated that MG is metabolized by the glyoxalases system which is diminished in the presence of clinical obesity and glyceroneogenesis (74-76,85). Large loads of fructose can alter the metabolism of MG (86) increasing the excess of triose, MG and D-Lactate. D-lactate is of particular interest since its plasma levels have been used as a surrogate marker of MG flux (87-89). In support of our contention, Thornalley has found increased MG and D-lactate in obese adults (74, 76) and we have shown the same in adolescents in a cross sectional study (Reyna Rodriguez, Claudia Luevano, Sergio Solorio, Russell Caccavello, Yasmin Bains, Ma. Eugenia Garay and Alejandro Gugliucci, CCLM, in press 2018). Further, in an intervention study, fructose restriction resulted in a 38% decrease in D-lactate levels in just 10 days) (Yasmin Bains, Caccavello Russell, Michael Wen, Susan Noworolski, Kathleen Mulligan, Viva Tai, Jean-Marc Schwarz, Ayca Erkin-Cakmak, Robert Lustig and Alejandro Gugliucci unpublished results 2018).

Therefore, fructose, by increasing MG (and its product D-lactate) may play a key role in obesity and metabolic syndrome through the MG postulated mechanism on AMPK (73). More research is needed to ascertain this contention. It has also been proposed that the dicarbonyl stress promoted by MG (acting on many other proteins) can play an important role in the development of IR in obesity and increase the risk of developing DMT2 and NAFLD (76).

Uric acid, produced from the AMP generated by the metabolism of fructose, activates the renin-angiotensin system and inhibits endothelial nitric oxide (NO) a vasodilator, causing an increase in blood pressure (45,90-92) which, together with the mitochondrial effects of fructose contribute to MetS development (45,93-95). The fluxes of uric acid generated by fructose are a result of transient energy deficits generated by quick unregulated phosphorylation of fructose as depicted in Figure 1B. In Figure 3, we expand on the details of this process.

Figure 3

Some of the deleterious actions of fructose on the liver may be due to ATP depletion and uric acid formation. As depicted in Figure 1B, fructose bypasses the 2 key regulatory steps in glycolysis because The liver has the very active FFK C 1), that floods the cytosol with trioses and AcCoA 2), which lead to lipogenesis 3), fatty liver 4) and hyperlipidemia 5), especially because fructose is co-ingested with glucose which is used for glycogen production (instead of fructose, Figure 1 A) vs 1 B)). This drives insulin secretion which enhances lipogenesis 3). The rapid, unregulated phosphorylation of fructose leads to quick cytosolic ATP depletion 6). In order to replenish the cytosolic ATP the cells use adenylate kinase 7) to generate 1 ATP and 1 AMP from 2 ADP 8). AMP is an endproduct that is degraded into uric acid. Uric acid quenches NO, leading to impaired vascular tone and hypertension 10). One important feature of uric acid is that it has been shown to be a FFK C activator, and therefore a perpetuator of this cycle (93). Actually, the mutation by which we lost uricase during evolution has been proposed as an evolutionary advantageous feature, facilitating fruit fructose assimilation in times of plenty (96). Uricase expression in experimental animals reduces fructose deleterious effects (97). The importance of FFK C is further evidenced by recent studies showing that its activity is enhanced in both obese humans with NASH (98) and fructose-fed mice (99). Its knockdown prevents fructose-induced steatosis and IR 11) (45, 93).

Since uric acid is a consequent product of ATP depletion and increased AMP in fructose metabolism, it has been used as a marker of hepatic decrease of ATP (100, 101). Studies have reported that acid uric induces oxidative stress and inflammation increasing lipogenesis, decreasing FA oxidation as well as AMPK activity (49, 50, 100) similar to what happens when systems such as nicotinamide adenine dinucleotide phosphate-oxidase (NADPH oxidase) and nuclear factor-kappa B (NF-kB) are activated (102, 103).

Studies where the uric acid synthesis inhibitor allopurinol has been used, have shown that a decrease in uric acid improves the MetS induced by fructose (93) and that this decrease in uric acid has a beneficial impact on both blood pressure and IR in humans (104-106), besides, acid uric promotes to NALFD due it effect of increase lipogenesis (49, 50, 107).

The group of Lustig et al has shown that SSBs have an impact on uric acid levels and blood pressure even in adolescents (90), highlighting the need for timely interventions in this age group to prevent future complications. Furthermore, these authors have highlighted that in addition to a regulation of salt intake, a regulation in sugar consumption, and therefore fructose, should be a treatment goal to prevent both hypertension and the metabolic syndrome (91).

Fructose exerts effects on both adipose tissue and liver, including adipogenesis, oxidative stress, inflammation, and glucocorticoid activation (71, 108, 109) which induces an increase in proliferation and differentiation of adipocytes (110).

The activation of inactive glucocorticoids such as cortisone in humans and 11-dehydrocorticosterone in rodents to their active forms, cortisol and corticosterone respectively (110) refers to an increase in bioavailability of these active forms within cells (7, 111). This glucocorticoid transformation is exerted by the enzyme 11 beta-hydroxysteroid dehydrogenase (11-beta-OHSDH), which is expressed both in the liver and in adipose tissue (and in other tissues such as the kidney and skeletal muscle) and is found in the luminal membrane of the endoplasmic reticulum (ER) (110, 112).

This enzyme is crucial for glucocorticoid activation via its reductase activity, which is dependent on NADPH (7, 71, 113). This reductase activity is increased in the presence of hexose 6 phosphate dehydrogenase (H6PDH), which forms NADPH in the ER lumen and therefore maintains the reducing power (109,114-116). In addition to these cofactors, 11-beta-OHSDH is induced in the presence of pro-inflammatory cytokines (114,117-119). Thus, the way in which fructose activates glucocorticoids is via stimulating an inflammatory state and activating NADPH, which in turn induces 11-beta-OHSDH (110). In addition, studies have shown that fructose is capable of affecting the gene expression of 11-beta-OHSDH (120, 121). We summarize these data in Figure 4.

Figure 4

Fructose and cross-talk between visceral adipose tissue and hepatocytes. The role of cortisol. Fructose in hepatocytes (over 90% percent of intake) or visceral adipose tissue (minor but non-negligible concentrations) also induces local inflammation and activation of G6PDH 1). This leads to activation of 11βOHSDH which turns inactive cortisone into cortisol 2). Cortisol stimulates fat synthesis 3) in both tissues as well as deposit of ectopic fat 4). Visceral adipose tissue cross-talks via portal circulation with inflammatory molecules as well as FA which enhance liver IR. Cortisol stimulates gluconeogenesis and hepatic glucose output 7), as wells as hyperlipidemia 8).

Regarding the inflammation caused by fructose (122), studies have reported that its consumption can lead to infiltration of macrophages in adipocytes, which promotes release of pro-inflammatory cytokines and increased inflammation (116,119,123-125). In addition, fructose also participates in the development of inflammation and insulin resistance which induces ER stress in adipocytes (122) which depletes the expression of endoplasmic reticulum oxidoreductase 1 alpha (ERO-1alpha), an ER chaperon responsible for regulating the secretion of adiponectin, adipokine considered anti-inflammatory and insulin sensitizer (126). This state of inflammation caused by fructose stimulates 11-beta-OHSDH and elevates cortisol within cells with its consequent effects involved in different metabolic alterations including components of MetS (113, 124, 127, 128).

A close relationship has been reported between 11-beta-OHSDH, cortisol, obesity and MetS (127-130) since the cellular bioavailability of cortisol induces processes involved in the components of MetS. In fact studies have reported that people with metabolic syndrome show an increased expression of 11-beta-OHSDH and intracellular cortisol (131), a state similar to key metabolic processes present in Cushing's syndrome, which is characterized by an excess of glucocorticoids (132).

In regards to the effects of glucocorticoids active in adipose tissue, it has been documented that there is an increase of intracellular cortisol in subcutaneous adipocytes (110) (where 11-beta-OHSDH activity is doubled) (133, 134). This can be induced by the effects of fructose, resulting in insulin resistance in subcutaneous adipocytes, thereby inhibiting the entry of FA and promoting greater flow and storage of unesterified FA in visceral deposits, mainly liver and visceral adipose tissue (VAT) (110,130,135-137).

11-beta-OHSDH also has an effect on hypertension since it is expressed in vascular tissue and can influence the homeostasis of blood pressure. It has been described that glucocorticoids produce a vasoconstrictor effect (138) which can induce endothelial dysfunction (139).

It is important to note that unlike glucose, fructose induces this glucocorticoid activation; in vitro studies have reported that glucose-6-phosphate (G6P) and fructose-6-phosphate (F6P) stimulate the reductase activity of 11-beta-OHSDH (71, 110) both in liver microsomes and in adipose tissue microsomes. In the latter, the presence of ER-luminal F6P isomerase forms G6P through the formation of NADPH dependent of hexose-6-phosphate dehydrogenase (H6PDH) (109). In addition, studies have reported that fructose compared to glucose generates more ER-luminal NADPH since fructose is easily transported through the plasma membrane and F6P through the ER membrane compared to glucose and G6P (71).

Finally, although more studies are required, it has been suggested that cortisol effects induced by fructose can be mediated by the activity of FFK-C from the liver through a metabolic crosstalk and inflammation (126).

Therefore, when 11-beta-OHSDH increases cortisol within the cells, it plays a role in the increase of visceral fat, inflammation, IR, hyperlipidemia and hypertension; characteristics of MetS.

Fructose-mediated advanced glycation endproducts (AGEs) formation via the Maillard reaction in foods may also be implicated in inflammation and MetS (140). Though well known by food chemists for decades, the Maillard reaction by fructose at physiological temperatures and pressures was studied starting only in the 80’s (141). These early studies helped establish the potential harmful effects of fructose on proteins as far more potent than those from glucose. The Maillard reaction (adduct formation between reactive carbonyls in glucose, fructose and their metabolites-such as methylglyoxal or deoxyglucosone-with amino groups in protein, DNA and lipids) has been implicated in diabetes complications. Fructose is 8 to10 times more reactive than glucose for Maillard reaction product formation as a result of the higher stability of its open chain form and its keto group. It does not form the Amadori but the Heyns product (141). The common methods employed for glucose glycation do not detect the Heyns products and/or other fructose-mediated adducts which has slowed down research on the potential role of fructose glycation in the pathogenesis of chronic disease in humans. Fructose-AGE concentration was measured in more than 100 commercial products (142). The highest levels of Fructose-AGE were shown in yoghurt beverages. Glycation adducts in food can be absorbed (up to 10% of dietary AGEs are absorbed) and exert their deleterious effects via engagement of the pro-inflammatory receptor for advanced glycation end products (RAGE) (141). In Figure 5 we summarize two pathways by which fructose may exert pro-inflammatory effects by yet another mechanism.

Figure 5

Fructose-derived advanced glycation products: the role of dietary AGEs in inflammation and insulin resistance. The Maillard reaction between carbonyls and proteins has been implicated in the pathogenesis of diabetic complications. Fructose (10 times more reactive than glucose) forms AGEs in processed foods. 1) Preformed AGE proteins in food (processed food is very high in fructose) results in intestinal digestion and absorption of AGE peptides 2) which bind to RAGE and are pro-inflammatory and generators of IR 3). Another putative pathway, for which epidemiological evidence (143, 144) and our own in vitro (145, 146) data vouch for is intestinal formation of AGEs 5) and 6) when excess fructose and amino acids or peptides are found in the intestinal lumen as a result of the co-ingestion of sugar and proteins 4). These AGEs, when absorbed, will generate the same effects as shown in 3). More evidence should be forthcoming on the relative role of these processes in fructose-induced pathogenesis of IR.