The role of circulating lipoprotein lipase and adiponectin on the particle size of remnant lipoproteins in patients with diabetes mellitus and metabolic syndrome

Sex-Related Differences in the Associations between Adiponectin and Serum Lipoproteins in Healthy Subjects and Patients with Metabolic Syndrome

The strong associations between the serum levels of adiponectin and the lipoprotein subclasses observed in healthy subjects are much weaker in patients with metabolic syndrome (MS). However, the impact of sex on these associations remained unexplored. Therefore, in the present study, we examined associations between adiponectin and the lipoprotein subclasses, analyzed by nuclear magnetic resonance spectroscopy, separately in healthy females and males, as well as in females and males with MS. We observed negative correlations between adiponectin and VLDL, IDL, and small-dense LDL in healthy males, but neither in healthy females nor in females or males with MS. Additionally, adiponectin was positively correlated with some HDL subclasses in healthy males and females with MS, but not in healthy females or males with MS. Adjusting for age and either body mass index, waist circumference, C-reactive protein, or interleukin-6 weakened the associations between adiponectin and VLDL and IDL but not small-dense LDL. The adjustment weakened the associations between adiponectin and HDL in healthy males but not in females with MS. Based on our results, we conclude that sex and the presence of MS are strong determinants of the associations between adiponectin and serum lipoproteins and that the complex regulatory network comprising adiponectin and other molecular players involved in the regulation of lipoprotein metabolism is primarily operative in healthy males and females with MS.

Association Between Remnant Cholesterol and Nonalcoholic Fatty Liver Disease: A Systemic Review and Meta-Analysis

Lipid disorders are related to the risk of nonalcoholic fatty liver disease (NAFLD). Remnant cholesterol (RC), a nonclassical and once-neglected risk factor for NAFLD, has recently received new attention. In this study, we assessed the relationship between the RC levels and NAFLD risk. We searched across PubMed, Web of Science, Embase, Cochrane Library, and China National Knowledge Infrastructure, with no restrictions on publication languages. Retrospective cohort studies and cross-sectional studies were enrolled from the inception of the databases until August 6, 2023. A random-effect model was applied to construct the mean difference, and a 95% confidence interval was applied to assess the relationship between the RC levels and NAFLD risk. We used two methods to estimate RC levels: Calculated-1 subtracts low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol from total cholesterol; Calculated-2 uses the Friedewald formula for LDL-C when triglycerides are <4.0 mmol/L, otherwise directly measured. A total of 265 published studies were selected through preliminary retrieval. Of these, six studies met the inclusion requirements and were enrolled in the meta-analysis. The RC level in the NAFLD group was significantly higher than that in the non-NAFLD group (mean difference: 0.18, 95% confidence interval: 0.10-0.26, P < 0.00001). We conducted subgroup analyses of computational methods and geographic regions. Notably, in the subgroup analysis of Calculation Method 2, the NAFLD group had significantly higher RC levels than the non-NAFLD group. On the other hand, in Calculation Method 1, the difference between the two groups was insignificant. In both the Asian and non-Asian populations, the RC levels were significantly higher in the NAFLD group than in the non-NAFLD group. The association of RC with an increased NAFLD risk was not dependent on the triglyceride. This meta-analysis suggests that elevated RC levels are associated with an increased risk of NAFLD. In addition to the conventional risk factors for fatty liver, clinicians should be concerned about the RC levels in the clinic.

Association between RC/HDL-C ratio and risk of non-alcoholic fatty liver disease in the United States

Background

The occurrence of non-alcoholic fatty liver disease (NAFLD) is increasing worldwide. The link between serum remnant cholesterol (RC) to high-density lipoprotein cholesterol (HDL-C) ratio and NAFLD remains unclear. Therefore, we sought to clarify the relationship between the RC/HDL-C ratio and the NAFLD.

Methods

Data for our cross-sectional study came from the 2017-2018 National Health and Nutrition Examination Survey (NHANES) with 2,269 participants. Associations between RC/HDL-C levels and the prevalence of NAFLD and hepatic fibrosis were evaluated using adjusted multivariate logistic regression analyses. A generalized additive model examined the non-linear relationship between RC/HDL-C and the probability of developing NAFLD.

Results

Among 2,269 participants, 893 (39.36%) were diagnosed with NAFLD. In each of the three models, RC/HDL-C and NAFLD had a strong positive statistical relationship: model 1 (OR = 9.294, 95%CI: 6.785, 12.731), model 2 (OR = 7.450, 95%CI: 5.401, 10.278), and model 3 (OR = 2.734, 95%CI: 1.895, 3.944). In addition, the subgroup analysis by gender and BMI suggested that RC/HDL-C showed a positive correlation with NAFLD. The RC/HDL-C ratio was positively correlated with the degree of liver steatosis. There was an inverted U-shaped connection between the prevalence of NAFLD and RC/HDL-C, with an inflection point of 0.619 for all participants and 0.690 for men. Receiver operating characteristic (ROC) analysis showed that the predictive value of RC/HDL-C for NAFLD (area under the curve: 0.7139; 95%CI: 0.6923, 0.7354; P < 0.001), was better than traditional lipid parameters.

Conclusion

Increased RC/HDL-C levels are independently associated with an increased risk of NAFLD and the severity of liver steatosis in the American population. In addition, the RC/HDL-C ratio can be used as a simple and effective non-invasive biomarker to identify individuals with a high risk of NAFLD.

Serum Levels of Adiponectin Are Strongly Associated with Lipoprotein Subclasses in Healthy Volunteers but Not in Patients with Metabolic Syndrome

Metabolic syndrome (MS) is a widespread disease in developed countries, accompanied, among others, by decreased adiponectin serum levels and perturbed lipoprotein metabolism. The associations between the serum levels of adiponectin and lipoproteins have been extensively studied in the past under healthy conditions, yet it remains unexplored whether the observed associations also exist in patients with MS. Therefore, in the present study, we analyzed the serum levels of lipoprotein subclasses using nuclear magnetic resonance spectroscopy and examined their associations with the serum levels of adiponectin in patients with MS in comparison with healthy volunteers (HVs). In the HVs, the serum levels of adiponectin were significantly negatively correlated with the serum levels of large buoyant-, very-low-density lipoprotein, and intermediate-density lipoprotein, as well as small dense low-density lipoprotein (LDL) and significantly positively correlated with large buoyant high-density lipoprotein (HDL). In patients with MS, however, adiponectin was only significantly correlated with the serum levels of phospholipids in total HDL and large buoyant LDL. As revealed through logistic regression and orthogonal partial least-squares discriminant analyses, high adiponectin serum levels were associated with low levels of small dense LDL and high levels of large buoyant HDL in the HVs as well as high levels of large buoyant LDL and total HDL in patients with MS. We conclude that the presence of MS weakens or abolishes the strong associations between adiponectin and the lipoprotein parameters observed in HVs and disturbs the complex interplay between adiponectin and lipoprotein metabolism.

Quantification of lipoprotein lipase in mouse plasma with a sandwich enzyme-linked immunosorbent assay

To support in vivo and in vitro studies of intravascular triglyceride metabolism in mice, we created rat monoclonal antibodies (mAbs) against mouse LPL. Two mAbs, mAbs 23A1 and 31A5, were used to develop a sandwich ELISA for mouse LPL. The detection of mouse LPL by the ELISA was linear in concentrations ranging from 0.31 ng/ml to 20 ng/ml. The sensitivity of the ELISA made it possible to quantify LPL in serum and in both pre-heparin and post-heparin plasma samples (including in grossly lipemic samples). LPL mass and activity levels in the post-heparin plasma were lower in Gpihbp1-/- mice than in wild-type mice. In both groups of mice, LPL mass and activity levels were positively correlated. Our mAb-based sandwich ELISA for mouse LPL will be useful for any investigator who uses mouse models to study LPL-mediated intravascular lipolysis.

Not only baseline but cumulative exposure of remnant cholesterol predicts the development of nonalcoholic fatty liver disease: a cohort study

BACKGROUND AND AIM

Remnant cholesterol (remnant-C) mediates the progression of major adverse cardiovascular events. It is unclear whether remnant-C, and particularly cumulative exposure to remnant-C, is associated with nonalcoholic fatty liver disease (NAFLD). This study aimed to explore whether remnant-C, not only baseline but cumulative exposure, can be used to independently evaluate the risk of NAFLD.

METHODS

This study included 1 cohort totaling 21,958 subjects without NAFLD at baseline who underwent at least 2 repeated health checkups and 1 sub-cohort totaling 2,649 subjects restricted to those individuals with at least 4 examinations and no history of NAFLD until Exam 3. Cumulative remnant-C was calculated as a timeweighted model for each examination multiplied by the time between the 2 examinations divided the whole duration. Cox regression models were performed to estimate the association between baseline and cumulative exposure to remnant-C and incident NAFLD.

RESULTS

After multivariable adjustment, compared with the quintile 1 of baseline remnant-C, individuals with higher quintiles demonstrated significantly higher risks for NAFLD (hazard ratio [HR] 1.48, 95%CI 1.31-1.67 for quintile 2; HR 2.07, 95%CI 1.85-2.33 for quintile 3; HR 2.55, 95%CI 2.27-2.88 for quintile 4). Similarly, high cumulative remnant-C quintiles were significantly associated with higher risks for NAFLD (HR 3.43, 95%CI 1.95-6.05 for quintile 2; HR 4.25, 95%CI 2.44-7.40 for quintile 3; HR 6.29, 95%CI 3.59-10.99 for quintile 4), compared with the quintile 1.

CONCLUSION

Elevated levels of baseline and cumulative remnant-C were independently associated with incident NAFLD. Monitoring immediate levels and longitudinal trends of remnant-C may need to be emphasized in adults as part of NAFLD prevention strategy.

Remnant Cholesterol Is Associated With Gestational Diabetes Mellitus: A Cohort Study

CONTEXT

The association between remnant cholesterol (RC) and gestational diabetes mellitus (GDM) risk is unclear.

OBJECTIVE

This study investigated the association between RC and GDM.

METHODS

We used data from the Tongji Maternal and Child Health Cohort, a prospective cohort study in China. Fasting lipid concentrations were measured around 16 weeks' gestation. RC was calculated as total cholesterol minus low-density lipoprotein cholesterol and high-density lipoprotein cholesterol. GDM was diagnosed by a 75-g oral glucose tolerance test at 24 to 28 weeks' gestation. Log-Poisson regression models were performed to estimate relative risks (RRs) of GDM across quartiles of RC levels and triglyceride (TG) levels after adjustment for potential confounders. TG and RC were mutually adjusted.

RESULTS

Among 2528 women, 256 (10.1%) developed GDM. The adjusted RRs (95% CIs) for GDM across increasing quartiles of RC were 1.00 (reference), 1.35 (0.91, 1.99), 1.68 (1.16, 2.45), and 1.73 (1.19, 2.50), respectively. Compared to pregnant women without 3 risk indicators (TG <2.08 mmol/L, RC <0.40 mmol/L, and pre-BMI <24.0 kg/m2), the risk of GDM was elevated in those with normal pre-BMI but high RC (aRR: 1.54; 95% CI: 1.08, 2.19) or high TG (aRR: 2.15; 95% CI: 1.33, 3.49). For those with all 3 risk indicators, the risk of GDM was the highest (aRR: 4.80; 95% CI: 3.20, 7.18).

CONCLUSION

Elevated RC levels were associated with the increased risk of GDM and independent of traditional risk factors. Pregnant women with high pre-BMI, high TG, and high RC were at greatly increased risk of GDM.

Multivessel Coronary Artery Disease Complicated by Diabetes Mellitus Has a Relatively Small Effect on Endothelial and Lipoprotein Lipases Expression in the Human Atrial Myocardium and Coronary Perivascular Adipose Tissue

Endothelial (EL) and lipoprotein (LPL) lipases are enzymes involved in lipoproteins metabolism and formation of atherosclerosis, a pathological feature of coronary artery disease (CAD). This paper examines the role of the lipases in the right atrial appendage (RAA) and coronary perivascular adipose tissue (PVAT) of patients with CAD alone or with accompanying diabetes. Additionally, correlation analysis for plasma concentration of the lipases, apolipoproteins (ApoA-ApoJ) and blood lipids (Chol, HDL-C, LDL-C, TAG) was performed. We observed that CAD had little effect on the lipases gene/protein levels in the RAA, while their transcript content was elevated in the PVAT of diabetic CAD patients. Interestingly, the RAA was characterized by higher expression of EL/LPL (EL: +1-fold for mRNA, +5-fold for protein; LPL: +2.8-fold for mRNA, +12-fold for protein) compared to PVAT. Furthermore, ApoA1 plasma concentration was decreased, whereas ApoC1 and ApoH were increased in the patients with CAD and/or diabetes. The concentrations of ApoC3 and ApoD were strongly positively correlated with TAG content in the blood, and the same was true for ApoB with respect to LDL-C and total cholesterol. Although plasma concentrations of EL/LPL were elevated in the patients with diabetes, CAD alone had little effect on blood, myocardial and perivascular fat expression of the lipases.

Causal associations of remnant cholesterol with cardiometabolic diseases and risk factors: a mendelian randomization analysis

BACKGROUND

Emerging evidence suggests that remnant cholesterol (RC) is strongly associated with an increased incidence of cardiometabolic diseases (CMD). However, the causality have not been confirmed. We aimed to evaluate the causal associations of RC with CMD and the relative risk factors using two-sample Mendelian randomization (MR) methods.

METHODS

Summary-level statistics of RC, CMD, and cardiometabolic risk factors were obtained from the published data from individuals with a predominantly European ancestry mainly from the UK Biobank and the FinnGen biobank. Univariable and multivariable MR analyses were used to evaluate the causal relationships between RC and CMD. A bidirectional MR analysis was performed to estimate the causality between RC and cardiometabolic risk factors. The main MR method was conducted using the inverse-variance weighted method.

RESULTS

Univariable MR analyses showed that genetically predicted RC was causally associated with higher risk of ischemic heart disease, myocardial infarction, atrial fibrillation and flutter, peripheral artery disease, and non-rheumatic valve diseases (all P < 0.05). Multivariable MR analyses provided compelling evidence of the harmful effects of RC on the risk of ischemic heart disease (P < 0.05). Bidirectional MR analysis demonstrated that RC was bidirectionally causally linked to total cholesterol, triglycerides, low-density lipoprotein cholesterol, hypercholesterolemia (all P < 0.05). However, no genetic association was found between RC and metabolic disorders or the other cardiometabolic risk factors.

CONCLUSIONS

This MR study demonstrates that genetically driven RC increases the risk of several CMD and cardiometabolic risk factors, suggesting that targeted RC-lowering therapies may be effective for the primary prevention of CMD.

Genetic association of lipids and lipid-lowering drug target genes with non-alcoholic fatty liver disease

BACKGROUND

Some observational studies found that dyslipidaemia is a risk factor for non-alcoholic fatty liver disease (NAFLD), and lipid-lowering drugs may lower NAFLD risk. However, it remains unclear whether dyslipidaemia is causative for NAFLD. This Mendelian randomisation (MR) study aimed to explore the causal role of lipid traits in NAFLD and evaluate the potential effect of lipid-lowering drug targets on NAFLD.

METHODS

Genetic variants associated with lipid traits and variants of genes encoding lipid-lowering drug targets were extracted from the Global Lipids Genetics Consortium genome-wide association study (GWAS). Summary statistics for NAFLD were obtained from two independent GWAS datasets. Lipid-lowering drug targets that reached significance were further tested using expression quantitative trait loci data in relevant tissues. Colocalisation and mediation analyses were performed to validate the robustness of the results and explore potential mediators.

FINDINGS

No significant effect of lipid traits and eight lipid-lowering drug targets on NAFLD risk was found. Genetic mimicry of lipoprotein lipase (LPL) enhancement was associated with lower NAFLD risks in two independent datasets (OR₁ = 0.60 [95% CI 0.50-0.72], p₁ = 2.07 × 10^-8; OR₂ = 0.57 [95% CI 0.39-0.82], p₂ = 3.00 × 10^-3). A significant MR association (OR = 0.71 [95% CI, 0.58-0.87], p = 1.20 × 10^-3) and strong colocalisation association (PP.H₄ = 0.85) with NAFLD were observed for LPL expression in subcutaneous adipose tissue. Fasting insulin and type 2 diabetes mediated 7.40% and 9.15%, respectively, of the total effect of LPL on NAFLD risk.

INTERPRETATION

Our findings do not support dyslipidaemia as a causal factor for NAFLD. Among nine lipid-lowering drug targets, LPL is a promising candidate drug target in NAFLD. The mechanism of action of LPL in NAFLD may be independent of its lipid-lowering effects.

FUNDING

Capital's Funds for Health Improvement and Research (2022-4-4037). CAMS Innovation Fund for Medical Sciences (CIFMS, grant number: 2021-I2M-C&T-A-010).

Remnant Cholesterol Predicts Long-term Mortality of Patients With Metabolic Dysfunction-associated Fatty Liver Disease

CONTEXT

Elevated serum remnant cholesterol independently predicts risks of cardiovascular diseases. However, the association between remnant cholesterol and metabolic dysfunction-associated fatty liver disease (MAFLD) remains unclear.

OBJECTIVE

This study aimed to explore the association of remnant cholesterol with MAFLD and its long-term mortality.

METHODS

We extracted data from the NHANES III, 1988 to1994 and the linked mortality data until December 31, 2015. The association between remnant cholesterol and MAFLD was analyzed by multivariable logistic regression. Cox proportional hazards regression was performed to assess whether elevated remnant cholesterol increased all-cause and cause-specific mortalities in MAFLD patients.

RESULTS

At baseline, 28.6% (1474/5156) of participants had MAFLD. In multivariable logistic regression, the fourth quartile of remnant cholesterol was associated with an increased risk of MAFLD compared with the first quartile (odds ratio [OR]: 1.714; 95% CI, 1.586-1.971; P < .001). In participants with normal levels of triglycerides, low-density lipoprotein and high-density lipoprotein cholesterol, the relationship between remnant cholesterol and MAFLD risk remained significant (OR: 1.346; 95% CI, 1.248-1.761; P < .001). During a median follow-up of 307 months, MAFLD patients with serum remnant cholesterol in the fourth quartile were associated with a higher risk of all-cause mortality (hazard ratio [HR]: 2.183; 95% CI, 1.825-2.407; P < .001), as well as a higher risk of cardiovascular mortality (HR: 2.346; 95% CI, 2.046-2.885; P < .001) and cancer-related mortality (HR: 2.366; 95% CI, 1.864-2.932; P < .001) compared with MAFLD patients in the first quartile.

CONCLUSION

Remnant cholesterol was independently associated with the risk of MAFLD and predicted all-cause, cardiovascular, and cancer-related mortalities in MAFLD patients.

Homozygous familial lipoprotein lipase deficiency without obvious coronary artery stenosis

The prevalence of familial lipoprotein lipase deficiency (LPLD) is approximately one in 1,000,000 in the general population. There are conflicting reports on whether or not LPLD is atherogenic. We conducted coronary computed tomographic (CT) angiography on two patients in their 70 s who had genetically confirmed LPLD. Patient 1 was a 73 year old woman with a body mass index (BMI) of 27.5 kg/m², no history of diabetes mellitus and no history of drinking alcohol or smoking. At the time of her first visit, her serum total cholesterol, triglycerides and high-density lipoprotein cholesterol levels were 4.8 mmol/L, 17.3 mmol/L, and 0.5 mmol/L, respectively. She was treated with a lipid-restricted diet and fibrate but her serum TG levels remained extremely high. Next-generation sequencing analysis revealed a missense mutation (homo) in the LPL gene, c.662T>C (p. Ile221Thr), leading to the diagnosis of homozygous familial LPL deficiency (LPLD). Patient 2 was another 73- year- old woman. She also had marked hypertriglyceridemia with no history of diabetes mellitus, drinking alcohol, or smoking. Previous genetic studies showed she had a nonsense mutation (homozygous) in the LPL gene, c.1277G>A (p.Trp409Ter). To clarify the degree of coronary artery stenosis in these two cases, we conducted coronary CT angiography and found that no coronary artery stenosis in either the right or left coronary arteries. Based on the findings in these two elderly women along with previous reports on patients in their 60 s with LPLD and hypertriglyceridemia, we suggest that LPLD may not be associated with the development or progression of coronary artery disease.

The role of remnant cholesterol beyond low-density lipoprotein cholesterol in diabetes mellitus

BACKGROUND

Previous research has linked elevated low-density lipoprotein cholesterol (LDL-C) and remnant cholesterol (RC) with diabetes mellitus (DM). The present study aims to estimate the RC-related DM risk beyond LDL-C, and to investigate the extent to which the association of RC and DM is mediated via insulin resistance and inflammation.

METHODS

We enrolled 7308 individuals without previous history of DM into the present study from the China Health and Nutrition Survey. Fasting RC was calculated as total cholesterol minus LDL-C and high-density lipoprotein cholesterol. Subjects were divided into four groups according to their LDL-C (100 mg/dL) and RC (24 mg/dL) levels to evaluate the role of LDL-C vs. RC on DM. A logistic regression analysis was then employed to evaluate the relationships between the discordant/concordant LDL-C and RC and DM. A mediation analysis was undertaken to identify potential mediators.

RESULTS

Of all the participants, a total of 625 (8.55%) patients were newly diagnosed with DM. Compared to the high LDL-C/low RC group, the low LDL-C/high RC group was more common in DM patients. After a multivariate adjustment, elevated LDL-C and RC were associated with DM. Moreover, the low LDL-C/high RC group and the high LDL-C/low RC group manifested a 4.04-fold (95% CI 2.93-5.56) and 1.61-fold (95% CI 1.21-2.15) higher risk of DM, relative to those with low LDL-C/low RC. The subgroup analysis indicated that low LDL-C/high RC was more likely to be related to DM in females. Similar results were also shown when the sensitivity analyses were performed with different clinical cut-points of LDL-C. Insulin resistance and inflammation partially mediated the association between RC and DM.

CONCLUSIONS

Our findings provided evidence for RC beyond the LDL-C associations with DM that may be mediated via insulin resistance and the pro-inflammatory state. In addition, women are more susceptible to RC exposure-related DM.

Heparin binding triggers human VLDL remodeling by circulating lipoprotein lipase: Relevance to VLDL functionality in health and disease

Hydrolysis of VLDL triacylglycerol (TG) by lipoprotein lipase (LpL) is a major step in energy metabolism and VLDL-to-LDL maturation. Most functional LpL is anchored to the vascular endothelium, yet a small amount circulates on TG-rich lipoproteins. As circulating LpL has low catalytic activity, its role in VLDL remodeling is unclear. We use pre-heparin plasma and heparin-sepharose affinity chromatography to isolate VLDL fractions from normolipidemic, hypertriglyceridemic, or type-2 diabetic subjects. LpL is detected only in the heparin-bound fraction. Transient binding to heparin activates this VLDL-associated LpL, which hydrolyses TG, leading to gradual VLDL remodeling into IDL/LDL and HDL-size particles. The products and the timeframe of this remodeling closely resemble VLDL-to-LDL maturation in vivo. Importantly, the VLDL fraction that does not bind heparin is not remodeled. This relatively inert LpL-free VLDL is rich in TG and apoC-III, poor in apoE and apoC-II, shows impaired functionality as a substrate for the exogenous LpL or CETP, and likely has prolonged residence time in blood, which is expected to promote atherogenesis. This non-bound VLDL fraction increases in hypertriglyceridemia and in type-2 diabetes but decreases upon diabetes treatment that restores the glycemic control. In stark contrast, heparin binding by LDL increases in type-2 diabetes triggering pro-atherogenic LDL modifications. Therefore, the effects of heparin binding are associated negatively with atherogenesis for VLDL but positively for LDL. Collectively, the results reveal that binding to glycosaminoglycans initiates VLDL remodeling by circulating LpL, and suggest heparin binding as a marker of VLDL functionality and a readout for treatment of metabolic disorders.

Adiponectin, May Be a Potential Protective Factor for Obesity-Related Osteoarthritis

Osteoarthritis (OA) is the most common joint disease in elderly individuals and seriously affects quality of life. OA has often been thought to be caused by body weight load, but studies have increasingly shown that OA is an inflammation-mediated metabolic disease. The current existing evidence suggests that OA is associated with obesity-related chronic inflammation as well as abnormal lipid metabolism in obesity, such as fatty acids (FA) and triglycerides. Adiponectin, a cytokine secreted by adipose tissue, can affect the progression of OA by regulating obesity-related inflammatory factors. However, the specific molecular mechanism has not been fully elucidated. According to previous research, adiponectin can promote the metabolism of FA and triglycerides, which indicates that it is a potential protective factor for OA through many mechanisms. This article aims to review the mechanisms of chronic inflammation, FA and triglycerides in OA, as well as the potential mechanisms of adiponectin in regulating chronic inflammation and promoting FA and triglyceride metabolism. Therefore, adiponectin may have a protective effect on obesity-related OA, which could provide new insight into adiponectin and the related mechanisms in OA.

Konjac Glucomannan Attenuated Triglyceride Metabolism during Rice Gruel Tolerance Test

In a recent study, we showed that konjac glucomannan (KGM) inhibits rice gruel-induced postprandial increases in plasma glucose and insulin levels. To extend this research, we investigated the effects of KGM addition to rice gruel on pre- and postprandial concentrations of circulating lipoprotein lipase (LPL), glycosylphosphatidylinositol-anchored high-density lipoprotein-binding protein 1 (GPIHBP1), hepatic triglyceride lipase (HTGL), free fatty acids (FFA), and triglycerides (TG). A total of 13 Japanese men, without diabetes, dyslipidemia, or gastrointestinal diseases, interchangeably ingested rice gruel containing no KGM (0%G), rice gruel supplemented with 0.4% KGM (0.4%G), and rice gruel supplemented with 0.8% KGM (0.8%G), every Sunday for 3 weeks. Blood samples were obtained at baseline and at 30, 60, and 120 min after ingestion to measure the abovementioned lipid parameters. Lipid parameters showed small, but significant, changes. Significant reductions were found in circulating FFA levels among all participants. Circulating TG levels significantly declined at 30 min and then remained nearly constant in the 0.8%G group but exhibited no significant difference in the 0%G and 0.4%G groups. Although circulating levels of LPL and GPIHBP1 significantly decreased in the 0%G and 0.4%G groups, they increased at 120 min in the 0.8%G group. Participants in the 0%G and 0.4%G groups showed significant decreases in circulating HTGL levels, which was not observed in the 0.8%G group. Our results demonstrate the novel pleiotropic effects of KGM. Supplementation of rice gruel with KGM powder led to TG reduction accompanied by LPL and GPIHBP1 elevation and HTGL stabilization, thereby attenuating TG metabolism.

Remnant Cholesterol is an Independent Predictor of New-Onset Diabetes: A Single-Center Cohort Study

PURPOSE

Remnant cholesterol (RC) is the cholesterol of triglyceride-rich lipoproteins, which has a high degree of atherogenic effect. To date, epidemiological evidence supports that higher RC levels lead to a greater risk of adverse cardiovascular events in patients with diabetes, but the nature of the association between RC levels and diabetes risk remains unclear. This study was designed to assess the association of RC with the risk of new-onset diabetes and to investigate whether there is a causal relationship between the two.

PATIENTS AND METHODS

The subjects included 15,464 individuals of the general population who participated in a health examination. Subjects were quartered according to the RC quartile, and the Cox proportional hazard model was used to assess the independent association between RC and new-onset diabetes.

RESULTS

During an average observation period of 6.13 years, 2.41% of the subjects were diagnosed with new-onset diabetes. Kaplan-Meier analysis showed that the 13-year cumulative diabetes rates corresponding to the RC quartile were 8.62%, 2.49%, 12.78%, and 17.91%. Multivariate Cox regression analysis indicated that higher RC levels were independently associated with an increased risk of new-onset diabetes (HR: 2.44, 95% CI: 1.50-3.89). Additionally, according to the results of receiver operating characteristic curve analysis, RC had the largest area under the curve (0.7314) compared to traditional lipid parameters in predicting new-onset diabetes.

CONCLUSION

These results indicated that RC is an important independent predictor of new-onset diabetes in the general population.

Association between postprandial lipoprotein subclasses and Framingham cardiovascular disease risk stratification

OBJECTIVE

To determine the ability of postprandial lipoprotein subclass concentrations to stratify patients with respect to their risk for cardiovascular disease (CVD).

METHODS

Using the Framingham cardiovascular disease risk score (FRS) algorithm, a total of 112 consecutive patients referred for community health screening were stratified into two groups: (a) low-risk (FRS < 10%) and (b) intermediate/high-risk (FRS ≥ 10%). Serum lipoprotein subclass concentrations were determined by Vertical Auto Profile (VAP-II).

RESULTS

Fasting and postprandial levels of LDL4, HDL2, VLDL1 + 2, VLDL3, and RLP, as well as fasting levels of ApoB and postprandial levels of LDL3 and IDL1, were significantly different in the intermediate/high risk FRS group vs. the low-risk group (P < 0.05). Correlations between Framingham CVD risk and LDL3, LDL4, IDL1, VLDL1 + 2, VLDL3, RLP, and ApoB were positive while negative for HDL2 in both the fasting and postprandial states. Intermediate/high risk for CVD was shown to be significantly associated with both fasting and postprandial levels of VLDL1 + 2 and RLP, as well as with postprandial LDL4 and VLDL3, as determined using forward conditional logistic regression analysis. Postprandial levels of VLDL1 + 2 were better at identifying patients in the intermediate/high-risk FRS group than fasting levels, although the differences were not significant due to overlapping reference intervals. In addition, the association between RLP and VLDL subclasses relative to Framingham CVD risk increased significantly in the postprandial state (ΔR² = 0.023; ΔF = 7.178; ΔP = 0.025) but not in the fasting state.

CONCLUSIONS

The use of postprandial lipoprotein subclass concentrations is not inferior to the use of fasting levels in identifying intermediate/high-risk FRS individuals. In addition, changes in RLP and VLDL subclass concentrations in fasting vs. postprandial states may reveal lipid metabolic mechanisms associated with CVD.

Association between skeletal muscle mass and serum concentrations of lipoprotein lipase, GPIHBP1, and hepatic triglyceride lipase in young Japanese men

Background

Two important regulators for circulating lipid metabolisms are lipoprotein lipase (LPL) and hepatic triglyceride lipase (HTGL). In relation to this, glycosylphosphatidylinositol anchored high-density lipoprotein binding protein 1 (GPIHBP1) has been shown to have a vital role in LPL lipolytic processing. However, the relationships between skeletal muscle mass and lipid metabolism, including LPL, GPIHBP1, and HTGL, remain to be elucidated. Demonstration of these relationships may lead to clarification of the metabolic dysfunctions caused by sarcopenia. In this study, these relationships were investigated in young Japanese men who had no age-related factors; participants included wrestling athletes with abundant skeletal muscle.

Methods

A total of 111 young Japanese men who were not taking medications were enrolled; 70 wrestling athletes and 41 control students were included. The participants’ body compositions, serum concentrations of lipoprotein, LPL, GPIHBP1 and HTGL and thyroid function test results were determined under conditions of no extreme dietary restrictions and exercises.

Results

Compared with the control participants, wrestling athletes had significantly higher skeletal muscle index (SMI) (p < 0.001), higher serum concentrations of LPL (p < 0.001) and GPIHBP1 (p < 0.001), and lower fat mass index (p = 0.024). Kruskal–Wallis tests with Bonferroni multiple comparison tests showed that serum LPL and GPIHBP1 concentrations were significantly higher in the participants with higher SMI. Spearman’s correlation analyses showed that SMI was positively correlated with LPL (ρ = 0.341, p < 0.001) and GPIHBP1 (ρ = 0.309, p = 0.001) concentration. The serum concentrations of LPL and GPIHBP1 were also inversely correlated with serum concentrations of triglyceride (LPL, ρ = − 0.198, p = 0.037; GPIHBP1, ρ = − 0.249, p = 0.008). Serum HTGL concentration was positively correlated with serum concentrations of total cholesterol (ρ = 0.308, p = 0.001), low-density lipoprotein-cholesterol (ρ = 0.336, p < 0.001), and free 3,5,3′-triiodothyronine (ρ = 0.260, p = 0.006), but not with SMI.

Conclusions

The results suggest that increased skeletal muscle mass leads to improvements in energy metabolism by promoting triglyceride-rich lipoprotein hydrolysis through the increase in circulating LPL and GPIHBP1.

Serum Triglyceride Lipase Concentrations are Independent Risk Factors for Coronary Artery Disease and In-Stent Restenosis

Aim: Endothelial lipase (EL), hepatic lipase (HL), and lipoprotein lipase (LPL) are all triglyceride lipases and are associated with coronary artery disease (CAD). However, whether they can be simultaneous independent risk factors for CAD is unknown. In the present study, we investigated whether the three lipases can be independent risk factors simultaneously for CAD and whether combining these lipases could provide greater predictive power than high-density lipoprotein cholesterol (HDL-c) for the development of CAD.

Methods: Eighty-six patients with CAD and 65 healthy controls were enrolled in the study. Additionally, 38 patients who underwent one-year follow-up angiography after percutaneous coronary intervention with stent implantation were collected to investigate in-stent restenosis. Serum EL, HL, and LPL concentrations were measured and compared with other coronary risk factors.

Results: Serum EL and HL concentrations were both significantly increased in patients with CAD or in-stent restenosis, whereas serum LPL concentration was reduced significantly in patients with CAD. Multivariate logistic regression analysis indicated that the three lipases were simultaneous independent risk factors for CAD. However, only serum EL concentration was considered an independent risk factor for in-stent restenosis. Importantly, the receiver operating characteristic curve showed that the combined measurement of the three lipases displayed better predictive power than HDL-c or any one of the three lipases for CAD.

Conclusions: Serum EL concentration was an independent risk factor for both CAD and in-stent restenosis. Moreover, the combined assessment of serum EL, HL, and LPL concentrations as multiple risk factors provided potent predictive power for CAD.

Atherogenic postprandial remnant lipoproteins; VLDL remnants as a causal factor in atherosclerosis

Oxidized LDL (Ox-LDL) and chylomicron (CM) remnants have been suggested to be the most atherogenic lipoproteins that initiate and exacerbate coronary atherosclerosis. In this review, we propose a hypothesis of the causal lipoproteins in atherosclerosis based on our recent findings on postprandial remnant lipoproteins (RLP). Plasma RLP-C and RLP-TG increased significantly after food intake, especially a fat load. More than 80% of the TG increase after the fat load consisted of the TG in RLP, which contained significantly greater apoB100 than apoB48 particles as VLDL remnants. The majority of the LPL in non-heparin plasma was found in RLP as an RLP-LPL complex and released into the circulation after hydrolysis. Plasma LPL did not increase after food intake, which may have caused the partial hydrolysis of CM and VLDL as well as the significant increase of RLP-TG in the postprandial plasma. LPL was inversely correlated with the RLP particle size after food intake. We showed that VLDL remnants are the major atherogenic lipoproteins in the postprandial plasma associated with insufficient LPL activity and a causal factor in the initiation and progression of atherosclerosis. We also propose "LPL bound TG-rich lipoproteins" as a new definition of remnant lipoproteins based on the findings of the RLP-LPL complex in the non-heparin plasma.

Pre-heparin lipoprotein lipase mass as a potential mediator in the association between adiponectin and HDL-cholesterol in type 2 diabetes

AIM

Lipoprotein lipase (LPL) is a major enzyme in lipid metabolism. Dyslipidemia, characterized by decreased high-density lipoprotein cholesterol (HDL-C), is prevalent in persons with type 2 diabetes mellitus (T2DM). The aim of this study was to determine whether pre-heparin LPL mass mediates the association between adiponectin and HDL-C in individuals with T2DM.

METHODS

Pre-heparin LPL mass was measured via an enzyme-linked immunosorbent assay, adiponectin by radioimmunoassay, and HDL-C was determined enzymatically. Participants' (n = 50) demographics, HbA1c, adiposity, homeostasis model assessment for insulin resistance (HOMA-IR), serum creatinine, and lipids were measured. Path analysis was utilized to test whether pre-heparin LPL mass is a mediator in the relationship between adiponectin and HDL-C.

RESULTS

All four criteria for mediation were satisfied in the path analysis. The indirect effect of adiponectin on HDL-C through pre-heparin LPL mass was significant, p = 0.001, whereas the direct effect of adiponectin on HDL-C was not significant, p = 0.074. These results remained consistent even after adjustments for age, gender, body mass index, HOMA-IR, and serum creatinine in the model.

CONCLUSION

The findings in this study suggest that pre-heparin LPL mass may mediate the association between adiponectin and HDL-C in T2DM. This relationship for measures of HDL-C functionality requires future investigation.

The effect of combined diet and exercise intervention on body weight and the serum GPIHBP1 concentration in overweight/obese middle-aged women

BACKGROUND

The relationship between the effects of diet and exercise intervention and the body weight associated with the serum lipoprotein lipase (LPL), hepatic triglyceride lipase (HTGL) and glycosylphosphatidylinositol anchored high density lipoprotein binding protein 1 (GPIHBP1) concentrations has not been elucidated.

METHODS

Sixty-six overweight/obese middle aged women were assigned to the diet and exercise intervention for 4months. They were divided into 2 groups followed by the body mass index (BMI) decreased >3% (n=41) and <3% (n=25). Serum lipids, lipoproteins and the LPL, HTGL, GPIHBP1 concentrations were determined.

RESULTS

The cases in which the BMI decreased >3% exhibited significant improvement of diagnostic markers compared with the cases with <3% decrease after the intervention. The LPL concentration did not significantly change, but GPIHBP1 increased significantly after the intervention. The increased GPIHBP1 was significantly associated with decreased body weight. Multiple regression analysis indicated a strong association between GPIHBP1 and percentage of body fat.

CONCLUSIONS

The diet and exercise intervention significantly increased the serum GPIHBP1 concentration in association with a decrease in body weight and percentage of body fat. These results suggest that GPIHBP1 is a better marker for body weight decrease than LPL.

Adipose Tissue LPL Methylation is Associated with Triglyceride Concentrations in the Metabolic Syndrome

BACKGROUND

DNA methylation is one of the epigenetic mechanisms that regulate gene expression. DNA methylation may be modified by environmental and nutritional factors. Thus, epigenetics could potentially provide a mechanism to explain the etiology of metabolic disorders, such as metabolic syndrome (MetS). The aim of this study was to analyze the level of DNA methylation of several lipoprotein lipase (LPL)-promoter-CpG dinucleotides in a CpG island region and relate this to the gene and protein expression levels in human visceral adipose tissue (VAT) from individuals with and without MetS.

METHODS

VAT samples were collected from laparoscopic surgical patients without and with MetS, and levels of LPL mRNA, LPL protein, and LPL DNA methylation were measured by qPCR, western blot, and pyrosequencing. Biochemical and anthropometric variables were analyzed. Individuals included in a subset underwent a dietary fat challenge test, and levels of postprandial triglycerides were determined.

RESULTS

We found higher levels of DNA methylation in MetS patients but lower gene expression and protein levels. There was a negative association between LPL methylation and LPL gene expression. We found a positive association between LPL methylation status and abnormalities of the metabolic profile and basal and postprandial triglycerides, whereas LPL gene expression was negatively associated with these abnormalities.

CONCLUSIONS

We demonstrate that LPL methylation may be influenced by the degree of metabolic disturbances and could be involved in triglyceride metabolism, promoting hypertriglyceridemia and subsequent associated disorders, such as MetS.

Nonalcoholic fatty liver disease associated with metabolic syndrome: Influence of liver fibrosis stages on characteristics of very low-density lipoproteins

BACKGROUND

We evaluated possible changes in VLDLcharacteristics, and metabolic related factors, in MetS-associated NAFLD and accompanying liver fibrosis.

METHODS

We studied 36 MetS patients with biopsy-proven NAFLD (MetS+NAFLD) and 24 MetS without ultrasound NAFLD evidence. Further, MetS+NAFLD was sub-divided according to fibrosis stage into, non-to-moderate (F0-F2, n=27) and severe (F3-F4, n=9) fibrosis. We measured: lipid profile, VLDL composition and size (size exclusion-HPLC), CETP and lipoprotein lipase (LPL) activities and adiponectin. Additionally, in MetS+NAFLD type IV collagen 7S domain was measured.

RESULTS

MetS+NAFLD showed increased VLDL-mass, VLDL particle number, VLDL-triglyceride% and large VLDL-% (p<0.04). CETP activity tended to increase in MetS+NAFLD (p=0.058), while LPL activity was unchanged. Moreover, in MetS+NAFLD, adiponectin was decreased (p<0.001), and negatively correlated with VLDL-mass and VLDL particle number (p<0.05), independently of insulin-resistance. Within MetS+NAFLD group, despite greater insulin-resistance, patients with severe fibrosis showed lower plasma triglycerides, VLDL-mass, VLDL-triglyceride%, large VLDL-% and CETP activity (p<0.05), while type IV collagen was increased (p=0.009) and inversely correlated with large VLDL-% (p=0.045).

CONCLUSIONS

In MetS, NAFLD is associated with larger and triglyceride over-enriched circulating VLDLs, of greater atherogenicity. However, when NAFLD progresses to severe fibrosis, circulating VLDL features apparently improved, probably due to early alterations in hepatic synthetic function.

Adiponectin, lipids and atherosclerosis

PURPOSE OF REVIEW

Adiponectin is an adipokine with anti-inflammatory, antioxidant, antiatherogenic, pro-angiogenic, vasoprotective and insulin-sensitizing properties. Several factors may influence adiponectin levels, such as genetic polymorphisms, obesity / body fat distribution, diet and exercise as well as cardiovascular risk factors such as sleep deprivation and smoking as well as medications. Adiponectin has been proposed as a potential prognostic biomarker and a therapeutic target in patients with cardiometabolic diseases.

RECENT FINDINGS

This narrative review discusses the associations of adiponectin with obesity-related metabolic disorders (metabolic syndrome, nonalcoholic fatty liver disease, hyperuricaemia and type 2 diabetes mellitus). We also focus on the links between adiponectin and lipid disorders and with coronary heart disease and noncardiac vascular diseases (i.e. stroke, peripheral artery disease, carotid artery disease, atherosclerotic renal artery stenosis, abdominal aortic aneurysms and chronic kidney disease). Further, the effects of lifestyle interventions and drug therapy on adiponectin levels are briefly reviewed.

SUMMARY

Based on available data, adiponectin represents a multifaceted biomarker that may beneficially affect atherosclerosis, inflammation and insulin resistance pathways. However, there are conflicting results with regard to the associations between adiponectin levels and the prevalence and outcomes of cardiometabolic diseases. Further research on the potential clinical implications of adiponectin in the diagnosis and treatment of such diseases is needed.

A new enzyme-linked immunosorbent assay system for human serum hepatic triglyceride lipase

There is no established method for measuring human hepatic triglyceride (TG) lipase (HTGL) concentration in serum. In this study, we developed new monoclonal Abs (MoAbs) (9A1 mouse MoAb and 141A1 rat MoAb) that react with HTGL both in serum and in postheparin plasma (PHP) and established a novel ELISA system for measuring serum HTGL and PHP-HTGL concentrations. To confirm the specificity of MoAbs, we performed immunoprecipitation-immunoblotting analysis. Both 9A1 mouse MoAb and 141A1 rat MoAb were able to immunoprecipitate not only recombinant HTGL and PHP-HTGL but also serum HTGL, demonstrating that HTGL exists in serum obtained without heparin injection. This method yielded intra- and interassay coefficients of variation of <6% and showed no cross-reactivity with LPL or endothelial lipase. In clinical analysis on 42 male subjects with coronary artery disease, there were strong positive correlations of serum HTGL concentration to PHP-HTGL concentration (r = 0.727, P < 0.01). Serum HTGL concentrations showed positive correlations to serum TGs (r = 0.314, P < 0.05) and alanine aminotransferase (r = 0.406, P < 0.01), and tendencies toward positive correlations to LDL cholesterol, small dense LDL, and γGTP. These results suggest that this new ELISA method for measuring serum HTGL is applicable in daily clinical practice.

Lipoprotein lipase does not increase significantly in the postprandial plasma

BACKGROUND

Previous reports have shown that lipoprotein lipase (LPL) activity significantly increases in the postprandial plasma associated with the increase of TG-rich lipoproteins. Therefore, we have reexamined those relationships using newly developed LPL assay with the different kinds of food intake.

METHODS

Standard meal (n=81), 50g of fat (n=54), 75g of glucose (n=25) and cookie (25g fat and 75g carbohydrate fat) (n=28) were administered in generally healthy volunteers. Plasma LPL, HTGL and TC, TG, LDL-C, HDL-C, RLP-C and RLP-TG were determined at subsequent withdrawal after the food intake.

RESULTS

Plasma TG, RLP-C and RLP-TG were significantly increased at 8PM (2h after dinner of standard meal) compared with 8AM before breakfast within the same day. Also those parameters were significantly increased in 2-6h after fat load. However, the concentrations and activities of LPL and HTGL did not significantly increase in association with an increase in the TG and remnant lipoproteins. Also LPL concentration did not significantly increase after glucose and "cookie test" within 4h.

CONCLUSION

No significant increase of LPL activity was found at CM and VLDL overload after different kinds of food intake when reexamined by newly developed assay for LPL activity and concentration.

Choline supplementation improves the lipid metabolism of intrauterine-growth-restricted pigs

Objective

The objective of this study was to investigate the effects of dietary choline supplementation on hepatic lipid metabolism and gene expression in finishing pigs with intrauterine growth retardation (IUGR).

Methods

Using a 2×2 factorial design, eight normal birth weight (NBW) and eight IUGR weaned pigs were fed either a basal diet (NBW pigs fed a basal diet, NC; IUGR pigs fed a basal diet, IC) or a diet supplemented with two times more choline than the basal diet (NBW pigs fed a high-choline diet, NH; IUGR pigs fed a high-choline diet, IH) until 200 d of age.

Results

The results showed that the IUGR pigs had reduced body weight compared with the NBW pigs (p<0.05 from birth to d 120; p = 0.07 from d 120 to 200). Increased (p<0.05) free fatty acid (FFA) and triglyceride levels were observed in the IUGR pigs compared with the NBW pigs. Choline supplementation decreased (p<0.05) the levels of FFAs and triglycerides in the serum of the pigs. The activities of malate dehydrogenase and glucose 6-phosphate dehydrogenase were both increased (p<0.05) in the livers of the IUGR pigs. Choline supplementation decreased (p<0.05) malate dehydrogenase activity in the liver of the pigs. Gene expression of fatty acid synthase (FAS) was higher (p<0.05) in the IC group than in the other groups, and choline supplementation decreased (p<0.05) FAS and acetyl-CoA carboxylase α expression in the livers of the IUGR pigs. The expression of carnitine palmitoyl transferase 1A (CPT1A) was lower (p<0.05) in the IC group than in the other groups, and choline supplementation increased (p<0.05) the expression of CPT1A in the liver of the IUGR pigs and decreased (p<0.01) the expression of hormone-sensitive lipase in both types of pigs. The gene expression of phosphatidylethanolamine N-methyltransferase (PEMT) was higher (p<0.05) in the IC group than in the other groups, and choline supplementation significantly reduced (p<0.05) PEMT expression in the liver of the IUGR pigs.

Conclusion

In conclusion, the lipid metabolism was abnormal in IUGR pigs, but the IUGR pigs consuming twice the normal level of choline had improved circulating lipid parameters, which could be related to the decreased activity of nicotinamide adenine dinucleotide phosphate-generating enzymes or the altered expressions of lipid metabolism-related genes.

Changes in lipoprotein lipase and endothelial lipase mass in familial hypercholesterolemia during three-drug lipid-lowering combination therapy

Background

This study was performed to compare the effects of three different lipid-lowering therapies (statins, ezetimibe, and colestimide) on lipoprotein lipase and endothelial lipase masses in pre-heparin plasma (pre-heparin LPL and EL mass, respectively) from patients with familial hypercholesterolemia (FH). FH is usually treated by coadministration of these three drugs.

Methods

The pre-heparin LPL and EL masses were measured in fresh frozen plasma drawn and stored at various time points during coadministration of the three drugs from patients with heterozygous FH harboring a single mutation in the LDL receptor (n = 16, mean age 63 years). The patients were randomly divided into two groups based on the timing when ezetimibe was added.

Results

Plasma LPL mass concentration was significantly reduced by rosuvastatin at 20 mg/day (median = 87.4 [IQR: 71.4–124.7] to 67.5 [IQR: 62.1–114.3] ng/ml, P < 0.05). In contrast, ezetimibe at 10 mg/day as well as colestimide at 3.62 g/day did not alter its level substantially (median = 67.5 [IQR: 62.1–114.3] to 70.2 [IQR: 58.3–106.2], and to 74.9 [IQR: 55.6–101.3] ng/ml, respectively) in the group starting with rosuvastatin followed by the addition of ezetimibe and colestimide. On the other hand, the magnitude in LPL mass reduction was lower in the group starting with ezetimibe at 10 mg/day before reaching the maximum dose of 20 mg/day of rosuvastatin. Plasma EL mass concentration was significantly increased by rosuvastatin at 20 mg/day (median = 278.8 [IQR: 186.7–288.7] to 297.0 [IQR: 266.2–300.2] ng/ml, P < 0.05), whereas other drugs did not significantly alter its level.

Conclusion

The effects on changes of LPL and EL mass differed depending on the lipid-lowering therapy, which may impact the prevention of atherosclerosis differently.

Electronic supplementary material

The online version of this article (doi:10.1186/s12944-016-0238-z) contains supplementary material, which is available to authorized users.