Alterations of leptin in the course of inflammation and severe sepsis

Cryopreserved Human White Preadipocytes (HWP) from subcutaneous adipose tissue and all culture media were supplied by Promocell (Heidelberg, Germany). They were cultivated as described by Promocell at 37°C in a humidified atmosphere of 95% air / 5% CO₂. Medium was changed every 2–3 days. Preadipocytes were plated at a density of 5000 cells/cm [2] and cultured to total confluency stage in 6-well plates with 3 mL growth medium without antibiotics, containing 5% fetal calf serum. Growth medium was replaced and differentiation medium (including d-Biotin 8 μg/mL, Insulin 0.5 μg/mL, Dexamethasone 400 pg/ml, IBMX 44 μg/mL, L-Thyroxine 9 pg/ml, Ciglitazone 3 μg/mL) was added for 72 h to start the differentiation process. To complete differentiation the cells were then cultured for further 12 days with adipocyte nutrition medium (including fetal calf serum 3%, d-Biotin 8 μg/mL, Insulin 0.5 μg/mL, Dexamethasone 400 pg/ml). During this time lipid droplets developed. Fully differentiated cells at day 15 after starting the differentiation process were preincubated with 1 mL nutrition medium lacking fetal calf serum and Dexamethasone for 24 h before starting the experiment. The dosages of TNF alpha and DAA used for the present analyses accorded to our established protocols within our working group, which have formerly been published, respectively in an inflammatory model with endothelial cells [20], [21]. Several dosages of the highly selective and efficient stimulator TNF alpha resulted in 1 ng/ml as a concentration leading to sub-maximal effects on cytokines like IL-8 or IL-6. Drotrecogin alpha (activated) was added to yield final concentrations of 50 ng/ml and 5 μg/ml. 1 h later TNF alpha was accordingly added for 1 ng/ml. Cells were incubated for 6 h and 24 h and harvested to prepare total RNA. Supernatants of cell medium were frozen for quantification of leptin by ELISA. Treatment with DAA (50 ng/ml or 5 μg/ml) and TNF alpha (1 ng/ml) did not affect adipocyte viability, as assessed by trypan blue exclusion (90% viable cells, no difference to untreated controls).

For quantitative evaluation of DAA-dependent leptin mRNA steady-state expression in adipocyte cultures, total RNA was prepared by using the RNeasy mini column kit (Qiagen, Hilden, Germany), including DNAse treatment. Three micrograms of RNA were reverse transcribed and converted to cDNA with oligo(dT)₁₅ primers using AMV reverse transcriptase according to standard protocols (Roche Applied Science, Germany, AMV Cat. No. 11495062001). The cDNA was amplified by quantitative PCR (qPCR) on the ABI 7000 realtime system (Applied Biosystems, Foster City, USA) using a qPCR-mix with hot start Taq DNA polymerase, SYBR-Green and enzyme system which reduces carryover DNA contamination (SYBR® GreenER™, Invitrogen, Cat.No.11760500) in the presence of sense and antisense primers (400 nM each). The sense- and antisense-primers for leptin were supplied from SABiosciences Corp. (Cat. No. PPH00581A). GAPDH as housekeeping gene was analyzed using primers as follows: [22] sense 5‘-TGCACCACCAACTGCTTAGC-3‘, antisense 5‘-GGCATGGACTGTGGTCATGAG-3‘. The qPCR condition consisted of 50°C for 2 min (reduction of DNA carryover contamination), 95°C for 10 min followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min, followed by melting-curve analysis to verify the correctness of the amplicon. Relative quantification of leptin mRNA expression was calculated as follows: The expression of leptin mRNA relative to the housekeeping gene GAPDH in samples from cells treated with DAA or untreated (Control) was calculated by the ΔΔCT method, based on the threshold cycle (CT), as fold change = 2^−Δ(ΔCT), where ΔCT = CT_leptin − CT_GAPDH and Δ(ΔCT) = ΔCT_DAA − ΔCT _Control[23], [24]. Efficiencies of the amplification reactions were calculated with a typical sample analysing the slope of the regression line of a 10-fold dilution series of cDNA (log10) versus CT: Efficiency =10^-1/slope. Only primer pairs that showed an amplification efficiency between 1.90 – 2.05 (90 –105%) and a coefficient of correlation (r) between 0.90 – 1.0 were used for quantification [25]. For verification of the correct length of amplification products, they were analysed on an ethidium bromide stained 2% agarose gel. All cell-culture experiments have been repeated at least twice to guarantee repeatability.

A total of 104 patients suffering from severe sepsis were prospectively enrolled at the First Department of Medicine, University Medical Centre Mannheim (UMM) (Germany) and at the Department of Anaesthesiology and Intensive Care Medicine, University Hospital Bonn (Germany) from March 2001 until October 2006. The study was carried out according to the principles of the declaration of Helsinki and was approved by the medical ethics commission II of the Medical Faculty Mannheim, University of Heidelberg, Germany. Written informed consent was obtained from all participating patients or their legal representatives.

The diagnosis of severe sepsis was based on criteria established by the American College of Chest Physicians and the Society of Critical Care Medicine Consensus Conference 1992: [5] Patients presented with a proven infection, three or more criteria of the systemic inflammatory response syndrome (SIRS criteria) and at least 1 of the following newly developed, sepsis-induced organ failures: cardiovascular organ failure with need for vasopressors, pulmonary organ failure defined as PaO₂/FiO₂ <250, renal organ failure with urine output <0,5 ml/kg/h, hematological organ failure with platelet count <80000/mm [3] or an unexplained metabolic acidosis with pH <7.3 and lactate levels >1.5 times of the upper limit of normal. Sepsis-induced organ failures in these patients were strongly connected to infection and were present for less than 24 h. Severity of sepsis was defined by the Acute Physiology and Chronic Health Evaluation II (APACHE II) score [26].

Out of a total of 104 patients suffering from severe sepsis 26 patients were treated with drotrecogin alpha (activated) (DAA) and 78 patients were not treated. The decision to treat patients with DAA or without was based on site-specific treatment modalities concerning the inclusion and exclusion criteria for the administration of DAA [27]. Twenty-six septic patients received a 96-hour infusion of DAA (XIGRIS®: 24 μg per 1 kg body weight per hour). 78 patients with severe sepsis did not receive DAA because of contraindications for drotrecogin alpha (activated) such as platelet count 30000/mm, need for therapeutic anticoagulation with heparin, increased risk of bleeding, recent gastrointestinal bleeding, or stroke within the last 3 months [28]. Additionally 45 healthy humans were included to serve as a control group.

Baseline characteristics, such as APACHE II scores, body mass index (BMI), creatinine levels, white blood cell count, platelet count, C-reactive protein (CRP), bilirubin, international normalized ratio (INR), activated partial thromboplastin time (a-PTT), antithrombin III (AT III), D-dimer as well as body temperature were determined on day 1 of severe sepsis. Total clinical follow-up lasted over 28 days and defined clinical outcomes such as death or survival were documented.

Leptin measurement was performed with a solid phase two-site immunosorbent assay (ELISA) (Quantikine® human leptin immunoassay, R&D Systems Inc., Minneapolis, USA) [29].

For data with normal distribution, the Student t test was applied. Otherwise, the Mann–Whitney U test was used as a nonparametric test. Deviations from a Gaussian distribution were tested by the Kolmogorov-Smirnov test. To generate hypotheses we used multiple t-tests with Holm-Bonferroni adjustments. For cell culture experiments the most important groups were specified in advance to be compared by statistical analysis: i) Control versus TNF alpha; ii) TNF alpha versus TNF alpha + DAA 50 ng/mL; iii) TNF alpha + DAA 50 ng/mL versus TNF alpha + DAA 5 μg/mL. Pearson’s correlation was used for normally distributed data. Spearman’s rank correlation was used for nonparametric data. Non-continuous variables were analyzed by the use of a 2x2 table and Fisher’s exact test. Data are presented as mean ± standard deviation (SD) or standard error of mean (SEM) as indicated. Values of P < 0.10 (two-tailed) were considered tending to be significant and P < 0.05 (two-tailed) were considered statistically significant. The calculations were performed with InStat (GraphPad Software) and SPSS software (SPSS Software GmbH).

In-vitro analyses showed final cell differentiation of human adipocytes after 15 days of cell growth (Figure 1a-c). Stimulating human adipocytes with tumor necrosis factor alpha alone (concentration TNF alpha, 1 ng/ml) over 6 hours resulted in leptin mRNA transcripts decreased by 46% compared to un-stimulated controls (p < 0.05) (Figure 2). Adding DAA at 50 ng/ml or 5 μg/ml to TNF alpha at 1 ng/ml, a significant change of leptin mRNA expression could not be detected (Figure 2). Stimulating human adipocytes with TNF alpha alone at the same concentration of 1 ng/ml for a longer time of 24 hours resulted in leptin mRNA transcripts decreased by 59% compared to un-stimulated controls (p < 0.05) (Figure 2). Adding DAA at 50 ng/ml or 5 μg/ml to TNF alpha at 1 ng/ml, a significant upregulation of leptin mRNA expression was detected after 24 hours from DAA at 50 ng/ml to 5 μg/ml (p < 0.05) (Figure 2).

Similar expression profiles as with leptin mRNA were seen when measuring leptin levels out of supernatants of the adipocytes by ELISA. Leptin levels of supernatants of adipocyte culture decreased by 25% (p < 0.05) after incubation with TNF alpha alone after 6 hours (Figure 3). After 6 hours incubation with TNF alpha at 1 ng/ml and DAA at 50 ng/ml or 5 μg/ml leptin levels increased from DAA at 50 ng/ml to 5 μg/ml (p < 0.05) (Figure 3). After 24 hours leptin levels significantly decreased by 23% after incubation with TNF alpha alone (p < 0.05) (Figure 3). But a significant change of leptin levels from DAA at 50 ng/ml to 5 μg/ml could not be detected after 24 hours (Figure 3).

Treatment with DAA (50 ng/ml or 5 μg/ml) and TNF alpha (1 ng/ml) did not affect adipocyte viability, as assessed by trypan blue exclusion (90% viable cells, no difference to untreated controls; data not shown).

Baseline characteristics of 104 patients suffering from severe sepsis are given in Table 1. Mean age of all study patients was 64 years. Patients treated with DAA (n = 26) were younger than patients not treated (n = 78) (57 years vs. 67 years; p = 0.002). The most common primary site of infection was the lung (n = 51), followed by cardiac (n = 12) and intra-abdominal infections (n = 7) (p = 0.3). The mean APACHE II score was 28.3 (SD = 5.4). Mean BMI levels were similar in all patient groups (DAA+, mean = 25.2, ±SEM = 2.5, DAA-, mean = 26.6, ±SEM = 0.9, p = 0.6).

Leptin serum levels significantly correlated with age, body temperature, BMI and APACHE II score. Additionally leptin levels inversely correlated with adiponectin, and correlated positively with white blood cells, antithrombin III, interleukin 6 and creatinine (Table 2). However, there were no significant differences of leptin levels in patients with impaired renal function (defined as a serum creatinine value above 1.1 mg/dl) compared to patients with a regular renal function (defined as a serum creatinine value below or equal than 1.1 mg/dl) (patients with renal failure, leptin mean 29282 pg/ml, ± SEM 4398 pg/ml, n = 75; patients without renal failure, leptin mean 21896 pg/ml, ± 5631 pg/ml, n = 29; p = 0.3). No association was found between leptin and C reactive protein (Table 2).

As demonstrated in Figure 4, patients suffering from severe sepsis without treatment of DAA revealed significantly higher serum concentrations of leptin compared to healthy controls (n = 45, mean = 12116 pg/ml, ± SEM = 1945 pg/ml) (leptin concentrations, Day 1: DAA-, mean = 30175 pg/ml, ±SEM = 4203 pg/ml. Day 3: DAA-, mean = 32670 pg/ml, ± SEM = 6173 pg/ml. Day 5: DAA-, mean = 32731 pg/ml, ±SEM = 6220 pg/ml) (p < 0.05).

At the beginning of treatment with DAA (day 1) leptin levels showed a tendency (p < 0.10) to be lower (day 1: DAA+, mean = 18991 pg/ml, ±SEM = 6009 pg/ml, nc = 26) compared to patients without DAA (Figure 4). From day 1 to day 3 of sepsis, leptin levels increased in patients treated with DAA compared to patients without treatment of DAA (p < 0.10) (leptin concentrations, Day 1: DAA+, mean = 18991 pg/ml, ±SEM = 6009 pg/ml. Day 3: DAA+, mean = 57315 pg/ml, ± SEM = 17075 pg/ml. Day 5: DAA+, mean = 54980 pg/ml, ±SEM = 17660 pg/ml).

There was no difference of leptin concentrations between patients surviving 28 days follow-up (mean = 26073 pg/ml, ±SEM = 4704 pg/ml, n = 51) compared to those who were deceased (mean = 28824 pg/ml, ±SEM = 5442 pg/ml, n = 53) (p > 0.10).

The exact mechanisms contributing to increased leptin mRNA and protein expressions after administration of DAA in human adipocytes and patients suffering from severe sepsis were not determined in this study. However, we performed cell culture experiments with adipocytes in order to confirm our clinical findings in septic patients with a more straightforward system and proving the involvement of adipocytes. Instead of LPS we used TNF alpha as a more selective and efficient pro-inflammatory stimulus for sepsis according to our established protocols, respectively in an inflammatory model with endothelial cells having been published already [20], [21]. It was our intention to apply the same conditions as in the above mentioned inflammatory models in order to guarantee comparability of our own experimental approach. However, TNF alpha serum levels were not measured in our septic patients. Regarding clinical benefits of septic patients increased leptin levels could not show a survival benefit over 28-days follow-up in our study. Nevertheless, our results need to be confirmed by ongoing basic research analyses and larger prospective clinical studies to explain the more detailed mechanisms beyond our findings.