Pathophysiology of diabetic dyslipidaemia: where are we?

Dietary lipids are absorbed by the enterocytes via passive diffusion or specific transporters (e.g. CD36 for NEFA and Niemann-Pick C1-like 1 protein [NPC1L1] for cholesterol). Within the enterocytes, triacylglycerols (triglycerides), cholesteryl esters and other lipids (phospholipids and small amounts of unesterified cholesterol) are associated with apolipoprotein (Apo)B-48 (as well as ApoA-IV and ApoA-I) to form chylomicrons in a process involving microsomal triacylglycerol transfer protein (MTP) and fatty acid transport proteins. Chylomicrons are then exported into lymph and subsequently into the blood. ApoB-48 synthesis by the gut occurs continuously; however, lipidation to form chylomicrons is dependent on the availability of lipids and occurs mainly after meals.

Lipoprotein lipase (LPL), which is attached to the luminal surface of endothelial cells and present mostly in muscles, the heart and the adipose tissue, plays a major role in chylomicron clearance by hydrolysing triacylglycerols and liberating NEFA into the circulation. The chylomicron remnants produced by the lipolysis of chylomicrons are taken up by the liver via the LDL receptor (Fig. 1) in conjunction with the LDL receptor-related protein (LRP), both of which bind ApoE.

Lipids are exported from the liver into the blood as VLDLs. The first step of VLDL assembly takes place in the rough endoplasmic reticulum (ER) where ApoB-100 is cotranslationally and post-translationally lipidated by MTP, forming pre-VLDL [9, 10]. In the absence of adequate core lipids and/or MTP, partially translocated ApoB is exposed to the cytosol and subjected to degradation via the ubiquitin–proteasome system. During the second step, pre-VLDL is further lipidated late in the ER compartment to form VLDL₂, exiting the ER compartment via Sar1 (a GTPase)/coat protein II (COPII) vesicles that fuse to the cis side of the Golgi apparatus. In the Golgi apparatus, VLDL₂ can be converted into larger VLDL₁ by the addition of lipids (Fig. 2). At this stage, VLDL particles may also be degraded via post-ER presecretory proteolysis (PERPP) [11]. The formation of VLDL₁ depends on factors such as ADP ribosylation factor 1 (ARF-1), phospholipase D1 and extracellular signal-regulated kinase 2 (ERK2), which are involved in membrane trafficking between the ER and the Golgi apparatus or in the formation of cytosolic lipid droplets [10].

As with chylomicrons, triacylglycerols from VLDLs are hydrolysed by LPL in plasma, producing NEFA to be used as fuel in the heart and skeletal muscle or for storage within adipocytes (as triacylglycerols). The progressive triacylglycerol depletion of VLDLs induces the transfer of a portion of the lipoprotein surface layer (including phospholipids, ApoC and ApoE) to HDLs and leads to the formation of IDLs [8]. Approximately 90% of IDLs are converted into LDL through further lipolysis involving hepatic lipase, which has both triacylglycerol lipase and phospholipase activities, whereas the rest is cleared by the liver (via LRP or LDL receptors).

LDL, the major transporter of cholesterol within the blood, comprises a core of esterified cholesterol molecules enclosed in a shell of phospholipids and unesterified cholesterol, together with a single molecule of ApoB-100. LDL is taken up into cells via receptor-mediated endocytosis, which involves, first, the binding of LDL–ApoB-100 to the LDL receptor on the plasma membrane of hepatic and other tissues, then the internalisation of the LDL-receptor complex via endocytosis, followed by fusion with lysozymes, which contain a number of catabolic enzymes. Proprotein convertase subtilisin/kexin type 9 (PCSK9) plays a key role in regulating LDL-receptor activity by binding the LDL-receptor/LDL complex and directing the receptor away from recycling back to the surface and into the lysosomal catabolic pathway.

The atheroprotective effect of HDLs related to their role in reverse cholesterol transport is well known, but these lipoprotein particles also have anti-inflammatory, anti-oxidative, anti-thrombotic and anti-apoptotic properties. HDLs are synthesised by both the liver and the intestine. Newly secreted or nascent HDLs contain only apolipoproteins (mainly ApoA-I) and rapidly recruit non-esterified cholesterol and phospholipids from peripheral cells through the binding of ApoA-I to the membrane-associated ATP-binding cassette protein 1 (ABCA1) transporter, allowing for the transport of non-esterified cholesterol and phospholipids from the cytoplasm into HDL (Fig. 1) [12]. Within the HDL particle, non-esterified cholesterol is esterified by lecithin–cholesterol acyltransferase (LCAT). In the circulation, HDLs acquire more cholesterol from peripheral tissues, including macrophages within artery walls, again through the ABCA1 transporter, as well as the ATP-binding cassette G1 (ABCG1) transporter. During this process, small-size HDL (usually called HDL₃) becomes larger (usually called HDL₂). HDL returns cholesterol to the liver via both direct and indirect mechanisms. Through scavenger receptor B1 (SR-B1), HDL-cholesteryl esters are directly taken up by the liver, where they are hydrolysed. HDL also exchanges lipids with VLDL and LDL in a process involving cholesteryl ester transfer protein (CETP), whereby cholesteryl esters are transferred from HDL to VLDL and, reciprocally, triacylglycerols are transferred from VLDL to HDL. In the circulation, HDL also receives ApoC and ApoE from VLDL. In addition, phospholipid transfer protein (PLTP) promotes the transfer of phospholipids from VLDL to HDL. During this process, HDL becomes enriched in triacylglycerols and phospholipids that are both degraded by hepatic lipase, thus forming smaller HDL particles that can be cleared by the liver or again participate in reverse cholesterol transport. During the catabolic process, lipid-poor ApoA-I is formed that can be filtered at the level of the glomerulus and then catabolised by proximal renal tubular epithelial cells after binding to cubilin, a protein localised to the apical surface of the renal tubular cell [13].

CETP and PLTP play key roles in lipoprotein metabolism. CETP facilitates the transport of cholesteryl esters and triacylglycerols between the lipoproteins, resulting in: (1) a net loss of cholesterol esters and gain of triacylglycerols by HDL and LDL; and (2) a reciprocal net gain of cholesterol esters and loss of triacylglycerols by chylomicrons and VLDL. CETP activity is increased by triacylglycerol-rich lipoproteins, NEFA and some phospholipids (phosphatidylcholine) and is inhibited by ApoC-I [14]. Any changes in CETP activity significantly modify the composition and metabolism of lipoproteins. PLTP circulates in the plasma bound to HDL, mediating the transfer of phospholipids (e.g. from VLDL remnants) into these particles, and also the exchange of phospholipids between lipoproteins.

Adiponectin is thought to play a direct role in influencing lipid metabolism. Adiponectin has been shown to be inversely correlated with fasting and postprandial triacylglycerols [15, 16]. It facilitates ApoA-I-mediated cholesterol efflux from macrophages by upregulating ABCA1 expression [17]. In addition, adiponectin is positively correlated with plasma HDL-cholesterol levels, and some data indicate that adiponectin may directly reduce HDL catabolism [18].

Patients with type 2 diabetes have a reduced plasma level of campesterol, a marker of cholesterol absorption, and increased plasma levels of lathosterol, a marker of cholesterol synthesis [37]. Using peroral administration of isotopes, reduced cholesterol absorption and increased cholesterol synthesis have been demonstrated in patients with type 2 diabetes [38]. The mechanisms responsible for these changes in cholesterol homeostasis are not yet clarified. In a study performed in 263 patients with type 2 diabetes, liver fat content was independently associated with plasma lathosterol [39]. It has been suggested that this could be due to increased expression of SREBP2, encoding sterol regulatory element-binding protein, a factor regulating cholesterol uptake and synthesis, observed under conditions of increased liver fat content [40].

In individuals with type 2 diabetes and insulin resistance, an increase in chylomicron production is observed, contributing to the postprandial hyperlipidaemia observed in this population [41]. Indeed, patients with type 2 diabetes have an increased rate of intestinal ApoB-48 secretion [42] and augmented expression of MTP (responsible for the addition of triacylglycerols to ApoB-48) within the intestine [43]. Insulin resistance is likely to be involved in the increased chylomicron production, since the normal acute suppression of postprandial chylomicron secretion, by insulin, is absent in patients with type 2 diabetes [44]. In addition, increased plasma NEFA concentrations (as a result of reduced inhibition of hormone-sensitive lipase in type 2 diabetes [45]) may further drive ApoB-48 secretion [46]. The clearance of chylomicrons is also impaired in type 2 diabetes [47]. Several mechanisms are responsible for this delayed chylomicron catabolism. The activity of LPL, the enzyme responsible for chylomicron hydrolysis, is significantly reduced in patients with type 2 diabetes [15, 48]. Insulin resistance is also associated with increased plasma levels of ApoC-III, an inhibitor of LPL [49]. Furthermore, the activation of LRP (one of the hepatic receptors responsible for chylomicron remnant uptake) by insulin is abolished in insulin-resistant mice [26]. The net result of all these changes is a significantly larger pool of chylomicrons (hypertriglyceridaemia) (Fig. 3). Moreover, patients with type 2 diabetes show increased levels of atherogenic remnant particles, including chylomicron remnants and VLDL remnants [50].

Postprandial hyperlipidaemia is likely to promote atherosclerosis and the occurrence of cardiovascular events in patients with type 2 diabetes. The increase in postprandial triacylglycerols has been shown to be correlated with the increase in TNFα, IL-6 and vascular cell adhesion molecule 1 (VCAM-1) values, in patients with type 2 diabetes, indicating a deleterious proinflammatory effect of postprandial hyperlipidaemia [51]. Furthermore, the magnitude of postprandial hypertriglyceridaemia is strongly correlated with the reduction in flow mediated dilatation, in patients with type 2 diabetes, indicating a role for postprandial hyperlipidaemia in endothelial dysfunction [52].

Increased plasma triacylglycerol levels in patients with type 2 diabetes are largely due to an increased number of VLDLs, particularly large VLDL₁ particles [6]. Both increased production and delayed catabolism of VLDL are responsible for the increased VLDL pool. In vivo kinetic studies in patients with type 2 diabetes have shown an augmented production of both VLDL-ApoB and VLDL-triacylglycerols [53‐55]. More precisely, it has been demonstrated that type 2 diabetes is associated with increased production of large VLDL₁ particles [56, 57]. Similar increases in VLDL production have been observed in obese, non-diabetic, insulin-resistant individuals, suggesting a critical role of insulin resistance in the pathophysiology of VLDL overproduction in type 2 diabetes [58, 59]. It has been shown that the VLDL₁ production rate is correlated with insulin resistance and liver fat in patients with type 2 diabetes [57, 60].

Insulin resistance is associated with reduced inhibition of hormone-sensitive lipase in adipose tissue by insulin, leading to increased lipolysis and, thereby, augmented NEFA portal flux to the liver. This has been shown to stimulate synthesis of triacylglycerols in hepatocytes [10]. Furthermore, the normal suppressant effect of insulin on postprandial VLDL (more specifically, VLDL₁) production is blunted by hepatic insulin resistance [45, 56, 61]. Several mechanisms seem to be involved in the overproduction of hepatic VLDL relative to the reduced inhibitory effect of insulin (Fig. 4). First, data from animal studies have provided evidence that insulin resistance is associated with a reduction in ApoB degradation in hepatocytes, leading to an increase in the ApoB pool available for VLDL assembly [62, 63]. Reduced PI3K activity in animal models, secondary to insulin resistance, has been reported to increase the expression of protein-tyrosine phosphatase 1B (PTP-1B), leading to suppression of ER60, a protease associated with the ER, which promotes ApoB degradation via a non-proteasomal pathway [62, 64]. In addition, an increased NEFA level in hepatocytes reduces the post-translational degradation of ApoB. Second, MTP expression is increased in insulin-resistant states and type 2 diabetes [65]. In insulin resistance, the reduced activation of PI3K leads to increased forkhead box protein O1 (FOXO1) activation, which is normally inhibited by activated PI3K. This increased activation of FOXO1 is responsible for augmented transcription of the MTP gene [66]. Third, it has been suggested that insulin resistance could be responsible for the increased activity of two factors involved in the formation of VLDL₁: phospholipase D1 and ARF-1 [21]. It has also been suggested that decreased PIP3, secondary to reduced PI3K activation, may also be involved, since PIP3, a highly negatively charged phospholipid, reduces lipid transfer to VLDL precursor and, thus, the formation of VLDL₁ [22]. It is possible that the increased formation of VLDL₁ automatically reduces the amount of VLDL precursors available for PERPP and, thus, decreases the rate of ApoB degradation [22].

In addition, de novo lipogenesis is increased in individuals with insulin resistance [67]. This increased de novo lipogenesis is secondary to augmented expression of both carbohydrate responsiveness element-binding protein (ChREBP) and sterol regulatory element-binding protein (SREBP)-1c in insulin resistance and type 2 diabetes [10]. It is suspected that, in patients with type 2 diabetes, hyperglycaemia directly activates ChREBP [68]. The increase in SREBP-1c expression could be related to the augmented ER stress observed in insulin resistance and type 2 diabetes [69]. Based on animal studies, it has been proposed that hyperinsulinaemia observed in insulin resistance and type 2 diabetes might be responsible for increased SREBP-1c expression because insulin stimulates SREBP-1c transcription. However, this is not supported by data in humans, since hyperinsulinaemia in patients with insulinoma is not associated with increased VLDL production [70]. Furthermore, insulin treatment in patients with type 2 diabetes induces a reduction in liver fat rather than an increase [71]. Moreover, reduced plasma adiponectin levels observed in type 2 diabetes may also promote VLDL production by increasing plasma NEFA levels, as a consequence of reduced muscle NEFA oxidation, and by inducing a decrease in AMP-kinase activation in the liver, which promotes de novo lipogenesis [10].

As assessed by kinetic studies using radioisotopes [54] and stable isotopes [53], catabolism of VLDLs is reduced in patients with type 2 diabetes, which also promotes hypertriglyceridaemia. This defect in VLDL catabolism mainly reflects the reduced activity of LPL in type 2 diabetes, particularly in adipose tissue [48]. Since insulin is an activator of LPL, it has been suggested that the diminution of LPL activity may be due to a relative insulin deficiency and/or insulin resistance. In addition, increased plasma levels of ApoC-III (an inhibitor of LPL) could also contribute to the decreased catabolism of VLDL in patients with type 2 diabetes, since increased plasma levels of ApoC-III were associated with impaired VLDL clearance in obese insulin-resistant men [72]. Moreover, postprandially, chylomicrons and VLDL compete for LPL, which exacerbates postprandial hypertriglyceridaemia. Most of the time, hypertriglyceridaemia observed in patients with type 2 diabetes is mild to moderate but the severity of hypertriglyceridaemia is also influenced, in each patient, by the genetic susceptibility [73]. In some sporadic cases of high genetic susceptibility, severe hypertriglyceridaemia may develop, with attendant risk of pancreatitis, particularly in hyperglycaemia [73].

The type of VLDL particles produced in type 2 diabetes is also altered, with a shift towards a greater proportion of those of a larger particle size (VLDL₁) [32, 74]. Relative to smaller VLDL particles, these are enriched with cholesterol esters and phospholipids. Larger triacylglycerol-enriched VLDL particles are potentially more atherogenic, as indicated by their significant association with endothelial dysfunction [52], and their preferential uptake by macrophages, leading to the formation of foam cells in vessel walls [75]. It has recently been demonstrated that VLDL from patients with type 2 diabetes has increased diacylglycerol content and reduced sphingomyelin, as well as increased palmitic acid-containing species [76]. Since palmitate is the major fatty acid synthesised during de novo lipogenesis, the increased palmitic acid content in VLDL-triacylglycerols might be a consequence of increased de novo lipogenesis, as suggested by a recent study that reported increased levels of circulating triacylglycerol-containing palmitic acid in insulin-resistant obese individuals with non-alcoholic fatty liver disease [77]. In addition, increased palmitic acid in VLDL-triacylglycerols has been shown to promote secretion of proinflammatory mediators by human smooth muscle cells [76]. Glycation of apolipoproteins in VLDL (ApoB, ApoCs, ApoE) may occur in diabetes. This may reduce VLDL binding to the ApoB/E receptor and hence impair its catabolism [78]. Furthermore, it has been suggested that glycation of ApoC-II, a cofactor of LPL, could contribute to lesser LPL activation [79].

In patients with type 2 diabetes, the mean LDL-cholesterol level is comparable or slightly elevated relative to that in individuals without diabetes [6, 8, 32]. However, the catabolism of LDL is substantially reduced [53, 80], inducing a longer duration of LDL in plasma that may promote lipid deposition within artery walls. In patients with type 2 diabetes, the number of LDL B/E cell-surface receptors is significantly reduced, which may be due to reduced insulin-mediated expression and could be responsible for observed impairments in LDL catabolism [81]. It has also been suggested that reduced LDL catabolism could be partly attributed to a decreased affinity of LDL for its receptor following ApoB glycation [82].

As a consequence of hyperglycaemia, increased glycation of LDL is observed in individuals with type 2 diabetes [83]. Glycated LDL has reduced affinity for LDL B/E receptors [84] and is preferentially taken up by macrophages, leading to the formation of foam cells [85]. Another lipoprotein modification observed in type 2 diabetes that has marked atherogenic potential is increased LDL oxidation. Patients with type 2 diabetes show increased oxidisability of LDLs and have an augmented number of oxidised LDL particles in their plasma [8]. Oxidised LDLs demonstrate a decreased affinity for the LDL receptor, and are preferentially taken into macrophages via specific oxidised LDL receptors prior to foam cell development [86]. In addition, they have chemoattractant effects on monocytes by increasing the formation of adhesion molecules, such as intercellular adhesion molecule 1 (ICAM-1), by endothelial cells, and by stimulating the formation of cytokines, such as TNFα or IL-1, by macrophages, which amplifies the inflammatory atherosclerotic process [86].

Small, dense, triacylglycerol-rich LDL particles (known as subclass B particles) are more prevalent in type 2 diabetes [87]. This is mainly related to hypertriglyceridaemia, and VLDL₁ triacylglycerol is the major predictor of LDL size in patients with type 2 diabetes and in non-diabetic individuals [6]. The characteristic hypertriglyceridaemia observed in patients with type 2 diabetes stimulates CETP, leading to the preferential formation of triacylglycerol-rich small, dense LDL particles over larger ones [74]. The presence of small, dense LDL particles has been reported to be associated with increased cardiovascular risk and progression of atherosclerosis [88]. Small, dense LDL particles are more atherogenic. They are more likely to undergo glycation and oxidation than larger LDL particles, which promotes the generation of foam cells [8, 89]. In addition, they show increased affinity for intimal proteoglycans, which may favour the penetration of LDLs into the arterial wall [90]. Individuals with small, dense LDL particles have an impaired response to the endothelium-dependent vasodilator acetylcholine [91]. Furthermore, as with VLDL, LDL from patients with type 2 diabetes shows significant changes in both lipid class (increased diacylglycerol and reduced sphingomyelin) and lipid species (increased palmitic acid-containing species), and the level of palmitic acid in LDL is correlated with insulin resistance [76].

Plasma levels of HDL-cholesterol and ApoA-I are reduced in patients with type 2 diabetes [6‐8]. In particular, the proportion of circulating smaller HDL particles (HDL₃) is increased, while there are far fewer large HDL particles (HDL₂); therefore, the overall number of HDL particles is reduced [74]. Reduced levels of HDL₂ in patients with type 2 diabetes have been reported to be associated with both hypertriglyceridaemia and obesity [92]. Kinetic studies using radioisotopes [93] and stable isotopes [94] have demonstrated that the decrease in HDL-cholesterol in patients with type 2 diabetes is due to increased catabolism of HDLs. The activity of hepatic lipase, the enzyme controlling HDL catabolism, is augmented in insulin-resistant states, which is likely to be responsible for the observed increase in HDL catabolism [95]. Hypertriglyceridaemia is a major contributing factor to the accelerated HDL catabolism observed in type 2 diabetes. It has recently been demonstrated that both increased VLDL₁ production and reduced VLDL₁ catabolism are independent factors associated with increased HDL catabolism in insulin-resistant states [96]. It is suggested that the increased pool of triacylglycerol-rich lipoproteins (mainly VLDL₁), observed in type 2 diabetes, promotes CETP-mediated triacylglycerol enrichment of HDL particles and, as a consequence, enhances HDL catabolism, since HDL-rich particles are very good substrates for hepatic lipase. The reduction in plasma adiponectin levels observed in individuals with insulin resistance and type 2 diabetes may be another mechanism involved in the diminution of HDL-cholesterol levels. Indeed, a significant negative correlation has been reported between the rate of HDL-ApoA-I catabolism and plasma levels of adiponectin, independently of abdominal obesity, insulin sensitivity, age, sex and plasma lipids, suggesting a direct effect of adiponectin on HDL metabolism [18]; however, its precise role has yet to be determined.

Several qualitative abnormalities in HDLs have been described in patients with type 2 diabetes. They are enriched in triacylglycerols, and this enrichment is responsible for increased HDL catabolism, as previously discussed. Furthermore, HDLs are glycated in type 2 diabetes, although the exact consequences of this glycation remain unknown. The reduction in phospholipids in large HDL particles in patients with type 2 diabetes is associated with increased arterial stiffness [97]. Patients with type 2 diabetes also have reduced ApoE content in large HDL particles, which may have an atherogenic effect, since large, ApoE-rich HDL usually prevents LDL binding to proteoglycans in the vessel wall [97]. ApoM, which is mainly associated with HDL, is reduced in patients with type 2 diabetes due to diabetes-associated obesity [97, 98]. ApoM promotes the formation of pre–β-HDL from α-HDL, and lower levels of ApoM may explain why pre-β-HDL formation is not increased in type 2 diabetes, despite increased PLTP activity. ApoM mediates the enrichment in sphingosine-1-phosphate in HDL, which promotes arterial vasodilation by stimulating endothelial nitric oxide formation [99].

In patients with type 2 diabetes, HDL has a reduced capacity to promote ex vivo cholesterol efflux from cells. This may be due to reduced expression of ABCA1, which is the membrane transporter responsible for the first step of cholesterol transfer from cell membranes to HDL [100]. Furthermore, glycation of ABCA1 has been shown to reduce its activity [101]. In addition, expression of ABCG1, another transporter involved in reverse cholesterol transport, is reduced in monocytes from patients with type 2 diabetes, and this is associated with impaired cholesterol efflux [102].

Reductions in the antioxidative effects of HDLs, promoted by hyperglycaemia and triacylglycerol enrichment, have been reported in patients with type 2 diabetes [103]. Furthermore, the ability of HDL to counteract the inhibition of endothelium-dependent vasorelaxation induced by oxidised LDL is impaired in type 2 diabetes. This reduction in HDL vasorelaxant effects is inversely correlated with HDL triacylglycerol content [104]. In line with these data, HDL from patients with type 2 diabetes has a weaker stimulatory effect on endothelial nitric oxide synthesis [105].

The qualitative lipoprotein abnormalities observed in patients with type 2 diabetes, such as increased triacylglycerol content of LDL and HDL particles, indicate increased CETP activity [106]. The main factor responsible for the increased CETP activity in type 2 diabetes is the augmented pool of triacylglycerol-rich lipoproteins (mainly VLDL), which directly stimulate CETP. However, hyperglycaemia per se could also activate CETP, since glycation of lipoproteins increases CETP activity [107]. In addition, a recent study in patients with diabetes reported that glycation of ApoC-I reduces its inhibitory effect on CETP [108]. Increased PLTP mass and PLTP activity have also been reported in patients with type 2 diabetes [109], and this is associated with increased intima–media thickness [110].