SEARCH

SEARCH BY CITATION

Keywords:

  • diabetes;
  • inflammation;
  • long-chain fatty acids;
  • whole blood assay

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

Fatty acids, uric acid and glucose are thought to contribute to subclinical inflammation associated with diabetes mellitus. We tested whether co-incubation of free fatty acids and uric acid or glucose influences the secretion of immune mediators from stimulated human whole blood in vitro. Fresh whole blood samples from 20 healthy subjects, 20 patients with type 1 diabetes and 23 patients with type 2 diabetes were incubated for 24 h with palmitic acid (PAL), linolenic acid (LIN) or eicosapentaenoic acid (EPA) alone or together with elevated concentrations of uric acid or glucose. Concentrations of proinflammatory cytokines interleukin (IL)-1β, IL-2, IL-12(p70), IL-18, IFN-γ, of regulatory cytokines IL-4, IL-10, IL-17 and chemokine CCL2 (MCP-1) were measured by multiplex-bead technology from supernatants. Co-incubation of fatty acids with uric acid resulted in a significant reduction of IL-10, IL-12(p70), IFN-γ and CCL2 (MCP-1) concentrations in supernatants compared to incubation with uric acid alone (P < 0·0001). In contrast, IL-18 was up-regulated upon co-stimulation with fatty acids and uric acid. Similarly, co-incubation of fatty acids with glucose diminished secretion of IL-10, IFN-γ and CCL2 (monocyte chemotactic protein-1), while IL-8 was up-regulated (P < 0·001). Samples from healthy and diabetic subjects did not differ after adjustment for age, sex, body mass index and diabetes type. All three fatty acids similarly influenced whole blood cytokine release in vitro and modulated uric acid or glucose-stimulated cytokine secretion. Although the ω-3-fatty acid EPA showed slightly stronger effects, further studies are required to elaborate the differential effects of PAL, LIN and EPA on disease risk observed previously in epidemiological studies.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

Plasma free fatty acid (FA) composition is influenced by the amount and quality of dietary fat, and plays a role in insulin resistance in muscle and liver as well as in low-grade inflammation in human subjects [1-5]. Whereas unsaturated ω-3 FA such as eicosapentaenoic acid (EPA) are considered protective [6, 7], saturated FA have been shown to associate with inflammation [8, 9]. The consumption of polyunsaturated ω-3 FA has been shown in epidemiological studies to reduce risk factors for cardiovascular disease and type 2 diabetes (T2D), e.g. hypertension, hyperlipidaemia, high LDL levels, insulin resistance and inflammation [10-12].

FA have been shown to mediate their (patho-)physiological effects in multiple ways: they can interact directly with the glucose transporter system and glucose phosphorylation, resulting in reduced glucose-6-phosphate levels and glycogen synthesis [13-16], thereby inducing insulin resistance on a cellular level [13]. Furthermore, FA may also activate inflammatory processes independently [17, 18]. Enhanced CD8 T cell memory was found to be modulated by FA metabolism [19], and FA provide a molecular signal by which they can link nutrition with innate immunity [20]. FA – especially saturated FA – bind to Toll-like receptor (TLR)-4, a membrane-bound pattern recognition receptor expressed on almost all mammalian cells, and thereby can activate proinflammatory pathways in macrophages and adipocytes. This ability is abolished in TLR-4-deficient mice, and partly protects them from lipid-induced insulin resistance [21, 22]. TLR-4 interacts further with various endogenous and exogenous ligands, resulting in activation of the innate immune system [23, 24]. The dominant exogenous TLR-4 ligand is lipopolysaccharide (LPS), a component of the outer cell wall of Gram-negative bacteria such as Escherichia coli [25]. Elevated LPS levels in the peripheral blood are associated with low-grade inflammation in animal models and human subjects [26-28].

In experimental cell culture and animal models, ω-3 FA lead to a down-regulation of proinflammatory cytokines such as interleukin (IL)-1β and tumour necrosis factor (TNF)-α [29-31]. Similarly, preclinical studies with human endothelial cells have shown an anti-inflammatory effect of polyunsaturated ω-3 FA by transcriptional down-regulation of proinflammatory cytokines [31, 32]. FA, glucose and uric acid are thought to contribute to the inflammatory processes observed in T2D and can be influenced by lifestyle factors, especially food composition [33-36]. Interestingly, in an epidemiological survey, Bandaru and Shankar found that higher serum uric acid levels were associated with fewer diabetes cases [37]. Evidence from epidemiological studies points to an association of increased FA levels not only with T2D, but also with immune-mediated type 1 diabetes (T1D), indicating that these results are under debate [38-40]. Findings by Matheus et al. indicated that uric acid is one of the most important independent variables associated with microvascular endothelial dysfunction in T1D [41].

How saturated and unsaturated FA differ in modulating immune cells with regard to cytokine secretion and how co-incubation of FA with uric acid or glucose is altering cytokine secretion in patients with diabetes mellitus is currently unknown. Therefore, the aim of our study was to investigate cytokine secretion of whole blood cultures of patients with T1D, T2D, latent autoimmune diabetes of the adult (LADA) and control subjects under the influence of different FA and co-stimulating conditions with glucose or uric acid, respectively.

Material and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

Study participants

We included 20 Caucasian healthy control subjects, 20 subjects with T1D, 23 subjects with T2D and six subjects with LADA (Table 1). Subjects were recruited for this study within the first 5 years after diabetes diagnosis. Diagnosis of LADA was defined as diabetes not requiring insulin for the first 6 months and positive antibodies for glutamate decarboxylase (GADA). Diagnosis of T1D and T2D was performed according to World Health Organization (WHO) criteria. The study was approved by the local ethical committee of the Heinrich-Heine-University, Düsseldorf, Germany in accordance with the declaration of Helsinki. All participants gave written informed consent for the study.

Table 1. Characteristics of study groups.
 Control groupT2DT1DLADAP-value
  1. Values represent the median (25th percentile, 75th percentile); to detect differences between groups we used analysis of variance (anova). Significant P-values are marked in bold type. BMI: body mass index; HDL: high-density lipoprotein; LDL: low-density lipoprotein; HbA1c: glycosylated haemoglobin 1c; T2D: type 2 diabetes; T1D: type 1 diabetes; LADA: latent autoimmune diabetes in adults.

Male : female10:1012:119:114:2 
Disease duration (month) 12 (7, 27)26 (13, 40)17 (6, 56)0·279
Age (years)50 (40, 53)48 (45, 58)41 (36, 47)44 (41, 52)0·007
Weight (kg)71·0 (67·5, 79·0)98·0 (80·0, 112·7)76·0 (65·0, 85·3)84·0 (80·0, 106·0)<0·001
BMI (kg/m2)24·8 (23·0, 26·6)31·3. (27·5, 36·7)24·5 (21·5, 26·6)26·5 (24·9, 33·6)<0·001
Waist circumference (cm)83 (76, 95)105 (94, 120)82 (77, 89)94 (85, 114)<0·001
Hip circumference (cm)99 (95, 105)109 (103, 123)98 (95, 108)102 (97, 122)0·006
Triglycerides (mg/dl)93 (65, 140)209 (134, 285)60 (48, 119)122 (53, 245)<0·001
LDL-cholesterol (mg/dl)132 (117, 157)139 (107, 161)121 (105, 135)125 (104, 168)0·343
HDL-cholesterol (mg/dl)58 (53, 69)47 (40, 63)64 (49, 76)54 (37, 81)0·028
Uric acid (mg/dl)5·0 (4·4, 5·5)5·5 (5·0, 6·5)3·8 (3·4, 4·6)4·3 (3·8, 5·4)<0·001
HbA1c (%)5·5 (5·4, 5·7)6·7 (6·0, 7·4)6·6 (6·2, 7·9)6·8 (5·5, 8·8)<0·001

Preparation of free FA

Preparation of free FA (Sigma-Aldrich, Taufkirchen, Germany) palmitic acid (PAL) 250 μM, linolenic acid (LIN) 250 μM and eicosapentaenoic acid (EPA) 250 μM were bound to bovine serum albumin (BSA) as described previously, with minor modifications [42]. In brief, free FA dissolved in ethanol were diluted to a final concentration of 6 mM in a 10% BSA solution (low endotoxin ≤ 1 EU/mg; Sigma-Aldrich, Seelze, Germany). The mixture was agitated gently at 37°C under nitrogen overnight. Control medium containing ethanol and BSA was prepared accordingly [42, 43].

Blood collection and assay procedure

Venous blood was drawn into lithium–heparin tubes (BD Vacutainer System, Heidelberg, Germany) in the morning after an overnight fast. Within 3 h, whole unseparated blood was diluted 1:3 with medium [Dulbecco's modified Eagle's medium (DMEM) and HEPES 2·4%; Invitrogen, Karlsruhe, Germany], and agitated gently in 50-ml tubes (Greiner Bio-one, Solingen, Germany); 200 μl aliquots were seeded per well of 96-well round-bottomed plates (Nunc, VWR International GmbH, Langenfeld, Germany) and cultured for 24 h at 37°C and 5% CO2. This whole blood assay procedure was established from pilot studies with blood from four healthy subjects. We pretested different dilutions (whole blood undiluted and 1:3 diluted), different incubation times (6 and 24 h for undiluted blood; 24, 48 and 72 h for diluted blood). As 24-h incubation with a 1:3 dilution gave cytokine concentrations in ranges detectable by our multiplex assay used, these conditions were taken as standard protocol for the experiments. Initially we measured a wide range of cytokines and chemokines [IL-1β, IL-1ra, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12 (p70), IL-13, IL-15, IL-17, IL-18, IFN-γ, TNF-α, IP-10, CCL2, CCL3]. From these immune mediators we chose IL-1β, IL-2, IL-12 (p70), IL-18, IFN-γ, CCL2 and IL-4, IL-10 and IL-17 for further read-out, as these mediators were detectable in our assay and had been reported to be potentially relevant in diabetes.

All parameters of each subject included for blood withdrawal were measured in one assay on the same plates in order to ensure intra-individual cross-comparison of results. From each blood drawing we performed in parallel triplicate incubations with positive and negative control, cultures with single FA and FA in combination with uric acid and glucose. Samples were incubated in triplicate with negative control (BSA 1·25%; Sigma-Aldrich), positive control [phytohaemagglutinin (PHA) 5 pg/ml; Remel Europe, Kent, UK], in the presence of BSA-coupled FA, PAL 250 μM (Sigma-Aldrich), LIN 250 μM (Sigma-Aldrich), EPA 250 μM (Sigma-Aldrich, Taufkirchen, Germany), uric acid (100 μg/ml, Invitrogen) and glucose (200 mg/dl; Merck, Darmstadt, Germany) as single stimuli or in various combinations of the stimuli, as indicated in the figure legends (either uric acid and PAL, uric acid and LIN, uric acid and EPA or glucose and PAL, glucose and LIN, glucose and EPA, as well as PHA and PAL, PHA and LIN, PHA and EPA). The same lots of FA and BSA were used for all experiments. Supernatants were removed from each well after centrifugation at 700 g for 5 min at 20°C, aliquoted and stored at −20°C until further analysis.

Samples from freshly drawn blood were used for the determination of blood cell counts and HbA1c at the central laboratory of the German Diabetes Center and for quantification of the concentrations of triglycerides, uric acid, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol in the central laboratory of the Heinrich-Heine-University Hospital.

Endotoxin concentration

Endotoxin (LPS) content of the supernatants of the blood cultures was determined using a chromogenic limulus amoebocyte lysate (LAL) assay (QCL-1000 assay; Lonza Ltd, Verviers, Belgium), according to the manufacturer's instructions [44].

Concentrations of cytokines and chemokines

Concentrations of proinflammatory cytokines IL-1β, IL-2, IL-12(p70), IL-18, IFN-γ, chemokine CCL2 [monocyte chemotactic protein-1 (MCP-1)] and anti-inflammatory regulatory cytokines IL-4, IL-10 and IL-17, respectively, were measured by multiplex-bead technology, using matched antibody pairs (R&D Systems, Minneapolis, MN, USA). Concentrations of immune mediators lower than the detection limit, a value half the detection limit, was assigned as described previously [45-47]. The lower detection limits of the assays were 21·50 pg/ml for IL-1β, 0·69 pg/ml for IL-2, 0·67 pg/ml for IL-4, 0·67 pg/ml for IL-10, 0·67 pg/ml for IL-12(p70), 10·75 pg/ml for IL-17, 2·68 pg/ml for IL-18, 10·75 pg/ml for IFN-γ and 10·75 pg/ml for CCL2 (MCP-1). The upper detection limits were 5441 pg/ml for IL-1β, 6603 pg/ml for IL-2, 5220 pg/ml for IL-4, 6092 pg/ml for IL-10, 5440 pg/ml for IL-12(p70), 5950 pg/ml for IL-17, 5948 pg/ml for IL-18, 5615 pg/ml for IFN-γ and 5441 pg/ml for CCL2 (MCP-1). As the reader program provided extrapolated values if they were above the upper detection level, we used these values for comparisons. The immunoassays showed interassay variations of < 20%. The mean coefficient of variation was ≤ 1, further confirming the quality of the assays.

Statistical methods

Statistical analysis was performed using sas version 9·1·3 (SAS Institute Inc., Cary, NC, USA) and GraphPad Prism version 4·0 for Windows (GraphPad Software, San Diego, CA, USA). Statistical significance was defined as a P-value < 0·05 for two-sided testing. Gaussian distribution of the data was assessed using the Kolmogorov–Smirnov test. Continuous variables, e.g. cytokine concentrations, are presented as medians and interquartile ranges (IQR) (Q1, 25th percentile; Q3, 75th percentile). The Wilcoxon test for paired data was used for exploratory analysis. We used multivariate regression models to investigate differences of log-transformed cytokine concentrations (dependent variables) in three different models: model 1, adjustment for age, sex, BMI, diabetes status; model 2, adjustment for age, sex, waist-to-hip ratio and diabetes status; and model 3, adjustment for age, sex, triglycerides (TG), HDL-cholesterol, LDL-cholesterol, urea and HbA1c (independent variables).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

Study participants

Healthy control subjects and patients with T1D, T2D and LADA showed significant differences for age, weight, BMI, waist and hip circumferences, as expected. Patients with T1D were younger compared to healthy control subjects and patients with T2D or LADA (P < 0·008). There were significant differences between study groups in levels of TG (P < 0·001), HDL cholesterol (P < 0·03), uric acid (P < 0·001) and HbA1c (P < 0·001). Patients with T1D showed significantly higher levels of HDL cholesterol than healthy control subjects and patients with T2D and LADA (P < 0·05), and lower levels of TG when compared to patients with T2D (P < 0·001). The (median) diabetes duration was [median (IQR): 25th percentile, 75th percentile] 12 (7, 27) months for T2D, 26 (13, 40) months for T1D and 17 (6, 56) months for LADA (Table 1).

Influence of 24-h incubation of whole blood with medium and BSA

In a first approach investigating whether the basal incubation conditions have an impact on the release of cytokines from whole blood samples, no differences in cytokine and chemokine secretion were found when comparing the healthy control group and the three patient groups (see exemplary for IL-1β Supplementary Figs S1 and S2). Therefore, further analyses were performed with all four groups combined.

Elevated concentrations of IL-1β, IL-10, IL-18, IFN-γ and chemokine CCL2 (MCP-1) were measured after incubation under basal conditions in medium (DMEM without BSA) for 24 h [baseline (0 h) versus 24 h (P < 0·05, Table 2)], whereas IL-2, IL-4, IL-12(p70) and IL-17 were not stimulated by incubation for 24 h. Incubation for 24 h with BSA-containing medium led to increased concentrations of the cytokines IL-1β, IL-2, IL-10, IL-12(p70), IL-18, IFN-γ and chemokine CCL2 (MCP-1) (Table 2). The concentrations of IL-4 and IL-17 were not altered under basal conditions. The spontaneous secretion of IL-1β, IL-10, IL-12, IL-18, and IFN-γ after 24 h incubation was significantly higher after incubation with BSA (P < 0·002), compared to 24-h incubation with medium DMEM alone. Only chemokine CCL2 (MCP-1) showed significantly lower concentrations upon BSA stimulation compared to medium DMEM after 24 h (P < 0·001).

Table 2. Influence of 24 h of incubation with medium [Dulbecco's modified Eagle's medium (DMEM)] and bovine serum albumin (BSA) on cytokine release.
 DMEM 0 hDMEM 24 hP-value1BSA 24 hP-value2
  1. Values represent the median (25th percentile, 75th percentile); to detect differences between groups we used Wilcoxon's matched-pairs test. P-value1 describes differences between DMEM 0 h and DMEM 24 h; P-value2 gives differences between DMEM 0 h and BSA 24 h; significant P-values are marked in bold type. Cytokine concentrations are given in pg/ml.

IL-1β10·74 (10·74, 10·74)76·27 (10·74, 828·30)<0·0013292·00 (839·00, 5480·00)<0·001
IL-20·34 (0·34, 0·34)0·34 (0·34, 0·34)0·6500·34 (0·34, 3·33)<0·001
IL-40·34 (0·34, 0·34)0·34 (0·34, 0·34)0·9980·34 (0·34, 0·34)0·364
IL-100·34 (0·34, 0·34)2·61 (0·41, 20·54)<0·001387·20 (138·30, 896·80)<0·001
IL-120·34 (0·34, 0·34)0·34 (0·34, 0·34)0·64913·19 (5·89, 25·68)<0·001
IL-175·38 (5·38, 5·38)5·38 (5·38, 5·38)0·9985·38 (5·38, 5·38)0·877
IL-187·58 (4·70, 12·74)10·00 (6·98, 15·05)0·02230·28 (16·82, 48·87)<0·001
IFN-γ5·38 (5·38, 5·38)5·38 (5·38, 11·95)0·020514·20 (357·50, 855·30)<0·001
CCL237·46 (23·94, 50·02)1543·00 (849·20, 2322·00)<0·001429·4 (204·50, 967·01)<0·001

In order to understand why BSA incubation leads to cytokine responses, we measured LPS sample concentrations. Measurements of 38 supernatants from stimulated whole blood cultures of two randomly chosen subjects showed LPS concentrations of 13·4 ± 3·4 EU/ml [mean ± standard deviation (s.d.)] in all measured samples except the whole blood control sample not containing BSA. As all these positive samples contained BSA, we concluded that the BSA preparation used for FA coupling contained LPS. LPS concentrations of samples measured were in the same range for all stimuli that contained BSA (FA, glucose, uric acid).

FA-stimulated cytokine and chemokine secretion

Compared to incubation with BSA for 24 h, significantly higher concentrations of the two proinflammatory cytokines IL-2 and IL-18 were measured in whole blood samples incubated with the three FA for 24 h (P < 0·01, data not shown). Exposure to FA was associated with significantly reduced concentrations of the anti-inflammatory cytokines IL-10 and chemokine CCL2 (MCP-1) (P < 0·01, data not shown).

When comparing cytokine concentrations induced by equimolar levels of LIN, PAL and EPA, the lowest concentrations of anti-inflammatory cytokines IL-10 and chemokine CCL2 (MCP-1) were measured upon EPA exposure compared to PAL and LIN [median (IQR: 25th percentile, 75th percentile) IL-10: EPA 98·23 pg/ml (45·9, 176·4) versus PAL 167·1 pg/ml (93·1, 244·0) versus LIN 160·1 pg/ml (107·3, 301·9), P < 0·001; and CCL2: EPA 185·4 pg/ml (113·1, 288·7) versus PAL 315·6 pg/ml (992·7, 478·6) versus LIN 312·9 pg/ml (195·2, 579·8), P < 0·001].

Effect of glucose on FA-stimulated cytokine and chemokine secretion from whole blood

Compared to BSA exposure alone, stimulation with high concentrations of glucose (11·1 mmol/l, equal to 200 mg/dl) for 24 h was associated with diminished IL-10 production (P < 0·02, Fig. 1b), while the levels of all other cytokines were unaltered (Fig. 1a,c–f).

figure

Figure 1. Influence of glucose on spontaneous and fatty acid (FA)-stimulated whole blood cytokine and chemokine secretion. Whole blood cells were incubated for 24 h in the presence of glucose (Glu) and palmitic acid (PAL), linolenic acid (LIN) or eicosapentaenoic acid (EPA). The concentrations of pro- and anti-inflammatory mediators accumulated in the culture supernatants were quantified by multiplex-bead technology (a–f). Shown are data of individual study participants (dots) and medians (horizontal lines).

Download figure to PowerPoint

The simultaneous presence of elevated glucose and FA levels for 24 h altered the secretion of proinflammatory cytokines IL-2, IL-12(p70), IL-18 and IFN-γ of anti-inflammatory cytokine IL-10 and chemokine CCL2 (MCP-1). All other cytokines measured (IL-1β, IL-4, IL-17) remained unaltered (IL-1β, IL-4, IL-17 data not shown). Concentrations of proinflammatory cytokines IL-2 and IL-18 were higher upon co-stimulation with FA and glucose compared to glucose only (Fig. 1a,d), whereas IFN-γ and anti-inflammatory cytokine IL-10 and chemokine CCL2 (MCP-1) showed lower concentrations under these conditions (Fig. 1b,e,f). Co-incubation with EPA and glucose was associated with a maximum decrease in concentrations of proinflammatory cytokines IL-2 (P < 0·05) and IFN-γ (P < 0·05), as well as of IL-10 (P < 0·01) and chemokine CCL2 (MCP-1, P < 0·001), when compared to PAL and LIN (Fig. 1a,b,e,f).

Effect of uric acid on FA-stimulated whole blood cytokine and chemokine secretion from whole blood

Incubation of whole blood with uric acid at a concentration of 100 μg/ml (equal 10 mg/dl), which is close to pathophysiological concentrations in human peripheral blood, did not induce changes in the cytokine or chemokine production from whole blood incubated under standard conditions (Fig. 2a–f). Upon co-incubation of uric acid with FA, secretion of proinflammatory cytokines IL-12(p70), IL-18 and IFN-γ, and anti-inflammatory cytokine IL-10 and chemokine CCL2 (MCP-1), was altered, whereas secretion of proinflammatory mediators IL-2 (Fig. 2a) and IL-1β, and anti-inflammatory regulatory mediators IL-4 and IL-17, remained unaltered (IL-1β, IL-4, IL-17, data not shown). Concentrations of proinflammatory cytokine IL-18 were higher upon co-stimulation with FA and uric acid compared to uric acid only, whereas anti-inflammatory cytokine IL-10, chemokine CCL2 (MCP-1) and proinflammatory cytokines IL-12(p70) and IFN-γ showed lower concentrations under these conditions (Fig. 2b–f). EPA incubation was associated with significantly lower concentrations of IL-10, IFN-γ and chemokine CCL2 (MCP-1) when compared to PAL and LIN (Fig. 2b,e,f).

figure

Figure 2. Influence of uric acid on spontaneous and on fatty acid (FA)-stimulated whole blood cytokine and chemokine secretion. Whole blood cells were incubated for 24 h in the absence or presence of uric acid (UA) in combination with palmitic acid (PAL), linolenic acid (LIN) or eicosapentaenoic acid (EPA). The concentrations of pro- and anti-inflammatory mediators accumulated in the culture supernatants were quantified by multiplex-bead technology (a–f). Shown are data of individual study participants (dots) and medians (horizontal lines).

Download figure to PowerPoint

Effect of PHA-stimulated whole blood cytokine and chemokine secretion by FA from whole blood

Incubation with mitogen PHA and FA was performed to investigate the immunomodulatory effect of FA on stimulated whole blood. As expected, PHA led to significantly higher concentrations of proinflammatory cytokines IL-2, IL-12(p70) and IFN-γ and anti-inflammatory/regulatory cytokines IL-4 and IL-17 when compared to the background responses after 24 h of incubation with BSA (Fig. 3a,c,e,g,h).

figure

Figure 3. Cytokine and chemokine secretion of phytohaemagglutinin ( PHA)-activated and fatty acid (FA)-stimulated whole blood. Whole blood cells were incubated for 24 h in the absence or presence of PHA in combination with palmitic acid (PAL), linolenic acid (LIN) or eicosapentaenoic acid (EPA). The concentrations of pro- and anti-inflammatory mediators accumulated in the culture supernatants were quantified by multiplex-bead technology (a–h). Shown are data of individual study participants (dots) and medians (horizontal lines). Not significant: n.s.

Download figure to PowerPoint

Similar to co-incubation experiments with glucose or uric acid, concentrations of proinflammatory cytokines IL-2, IL-12(p70) and IFN-γ and anti-inflammatory/regulatory cytokines IL-4, IL-10, IL-17 and chemokine CCL2 (MCP-1) were lower upon co-incubation of PHA with PAL, LIN or EPA compared to incubation with PHA alone. All cytokines except IL-12(p70) and IL-18 showed the lowest concentrations upon co-stimulation with EPA and PHA compared to PAL and LIN (Fig. 3c,d). The proinflammatory cytokine IL-18 was the only cytokine tested that showed higher concentrations upon stimulation with PHA and co-incubation with FA compared to stimulation with PHA alone (Fig. 3d).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

The consumption of FA and increased postprandial concentrations of FA in the blood are associated with elevation of the systemic concentrations of inflammatory mediators [48, 49]. Permanently increased levels of these mediators cause a state of chronic low-grade inflammation that may contribute to the development of metabolic disorders, such as impaired glucose tolerance, insulin resistance and diabetes [5, 50, 51].

In our current study we investigated the effects of FA on the secretion of cytokines and chemokines from peripheral blood cells in order to assess their contribution to FA-induced accumulation of inflammatory mediators in human blood. We selected a whole blood culture system optimized for studies on the release of immune mediators from unseparated cell populations of peripheral blood samples, similar to other protocols [52, 53]. Using this experimental in-vitro system we exposed human blood samples to FA at concentrations which correspond largely to the systemic levels detectable after a (lipid-rich) meal [33].

To investigate the modulatory capacity of FA on a stimulated immune system, a basic characteristic of subjects with metabolic dysregulation or diabetes, we analysed the effects of FA on blood cells in the presence of stimuli of the adaptive and/or the innate immune system. As a stimulus for the adaptive immune system we chose PHA [54]. Stimulation of the innate immune system was achieved by endotoxin (LPS), a potent activator of cells of the macrophage/monocyte lineage [55] that was present unintentionally in the albumin fraction used for fatty acid coupling. In our assays, the LPS contents of the fatty acid–albumin preparations contributed to an LPS concentration that corresponds to endotoxin levels typically detectable in the periphery of diabetic subjects in the postprandial state [56, 57]. LPS represents a strong immunostimulatory signal that induces a systemic inflammatory response with increased proinflammatory cytokines such as TNF-α and IL-6 [28]. The simultaneous presence of FA and LPS reflects the situation in diabetic patients who experience a strong increase of peripheral LPS levels after a lipid-reach meal [56]. However, despite the stimulatory effect of LPS, we were still able to detect differential stimulatory effects of the three FA.

Interestingly, recent findings suggest that FA and LPS exert their biological activities by stimulating the TLR-4–MyD88-dependent pathway, which results in the production of proinflammatory mediators [58]. Further, the liver secretory protein fetuin-A has been shown recently to act as an endogenous adaptor protein between FA and TLR-4, providing the missing link between FA and chronic low-grade inflammation that impairs insulin sensitivity [59]. Overall, not all aspects of FA-induced inflammation are understood fully, and it remains to be elucidated how different FA exert their distinctive roles in inflammation-induced insulin resistance [60].

Our results demonstrate that all three FA tested (PAL, LIN, EPA) have an overall effect on the release of cytokines from human whole blood cultures, leading to increased concentrations of IL-2 and IL-18 and decreased IL-10, IL-12(p70) and chemokine CCL2 (MCP-1) contents of blood cell supernatants. This cytokine pattern is suggestive of up-regulation of proinflammatory mediators, on one hand, and down-regulation of anti-inflammatory (IL-10), and chemokine [CCL2 (MCP-1)] or proinflammatory mediators [IL-12(p70)], on the other hand, by FA. Our experiments revealed a significant and pronounced down-regulation of proinflammatory cytokines by EPA. However, further experiments, including studies on time- and dose-dependency, will be required to confirm a potential beneficial effect of this polyunsaturated FA on the cytokine release pattern of peripheral blood cells. Nevertheless, our studies demonstrate that human peripheral blood cells are able to respond to FA by modulating their cytokine release pattern and may contribute thereby to the alterations of the systemic levels of (pro-)inflammatory mediators observed after a lipid-rich meal.

As comparable reactivity to FA was observed in cells from healthy subjects and in patients with T1D, T2D and LADA, FA-responsiveness has to be regarded as a general property of peripheral blood cells which seems to be independent of the metabolic state of the donor. It should be taken into account that the absence of different reactivity to FA in cells from healthy subjects and in patients with T1D, T2D and LADA could be explained by the low group size and relatively short duration of diabetes in this study or that the blood cells still act in a similar manner at this time of disease duration and under these in-vitro conditions.

So far, no such experiments with fatty acid-stimulated blood from patients with T1D, T2D and LADA have been reported in the literature. Instead, data on systemic cytokine concentrations measured from serum and stimulated T cell responses have been published. The main outcome of these recent studies is that systemic cytokines and chemokines do not differ comparing T1D and LADA. An increase of CCL3 has been seen for T1D/LADA compared to T2D; cytokines IL-1ra, IL-6 and TNF-α were lower in LADA/T1D compared to T2D; and in a Chinese study, IL-6 was higher in T2D compared to LADA and T1D [61-64]. Therefore, it is not unexpected that LADA does not differ from T1D, but we had expected that T2D might be different from LADA/T1D because they have different metabolic conditions.

In a further approach we studied whether the effect of FA on cytokine release from peripheral blood cells is influenced by increased levels of glucose and uric acid, as they occur frequently in (pre-) diabetic subjects. In our study a combination of FA and increased glucose levels induced a differential pattern of cytokine release. In a previous study, Kempf et al. also described a stimulatory effect of glucose, and showed that the presence of high concentrations of glucose in vitro led to increased proinflammatory mediators expressed by peripheral blood cells from patients with metabolic syndrome [35]. Further, Kempf et al. described a glucose-induced immune mediator expression in leucocytes, resulting in altered levels of cytokines, e.g. intercellular adhesion molecule (ICAM-1), TNF-α and IL-6 obtained from subjects with three or more parameters of the metabolic syndrome [according to International Diabetes Federation (IDF) criteria, including fasting plasma glucose, triglyceride, HDL cholesterol [65]] compared to those with fewer than three metabolic syndrome parameters [35]. Similarly, a study by Ruggiero et al., with 957 human subjects, showed that the uric acid concentration in serum is associated positively with neutrophil counts as well as with concentrations of proinflammatory (C-reactive protein, IL-6, IL-18 and TNF-α) and anti-inflammatory (IL-1ra) immune mediators [36]. These data indicate that the systemic presence of uric acid is associated with an up-regulated immune response in vivo. Our experimental setting, testing the combination of glucose, uric acid and FA in vitro, extends these observations and suggests that the inflammatory status results from a complex interplay of different parameters contributing to metabolic disease.

In conclusion, our data show an immunomodulatory capacity of the FA PAL, LIN and EPA on pro- and anti-inflammatory immune mediators that are modulated by metabolic stimuli such as uric acid and glucose, which are frequently increased in patients with diabetes mellitus. Further studies are necessary to address the kinetics and effects of different concentrations of stimuli.

Acknowledgement

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

We thank W. Fingberg, C. Bünting and Y. Sagik for excellent technical assistance. This study was supported in part by the German Center for Diabetes Research (DZD e.V.). N. C. Schloot is currently employed by Lilly Deutschland GmbH and guest scientist at the German Diabetes Center, Institute for Clinical Diabetology.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosures
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
cei12071-sup-0001-si.tif2528K

Fig. S1. No effect of the type of diabetes on the release of the cytokine interleukin (IL)-1β from stimulated whole blood. Whole blood cells from healthy controls and patients with type 1 diabetes (T1D), type 2 diabetes (T2D) and latent autoimmune diabetes of the adult (LADA) were incubated for 24 h in the presence of bovine serum albumin (BSA) (a), phytohaemagglutinin (PHA) (b), glucose (c), uric acid (d) or fatty acids [palmitic acid (e), linolenic acid (f), eicosapentaenoic acid (g)]. The concentrations of the cytokine IL-1β in the culture supernatants were quantified by multiplex-bead technology. Depicted are individual data (dots) and medians of each group (horizontal lines).

cei12071-sup-0002-si.tif2418K

Fig. S2. No effect of the type of diabetes on the release of the cytokine interleukin (IL)-1β from stimulated whole blood. Whole blood cells from healthy controls and patients with type 1 diabetes (T1D), type 2 diabetes (T2D) and latent autoimmune diabetes of the adult (LADA) were incubated for 24 h in the presence of glucose (Glu) or uric acid (UA) in combination with palmitic acid [PAL+Glu (a), PAL+UA (b)], linolenic acid [LIN+Glu (c), LIN+UA (d)], eicosapentaenoic acid [EPA+Glu (e), EPA+UA (f)]. Concentrations of the cytokine IL-1β in the culture supernatants were quantified by multiplex-bead technology. Depicted are individual data (dots) and medians of each group (horizontal lines).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.