Alterations in fatty acid composition of plasma and erythrocyte lipids in critically Ill patients during sepsis

Overview

The overall fatty acids (FA) composition, and especially proportions of n-6 and n-3 polyunsaturated fatty acids in plasma and membrane lipids, greatly impacts on cell and organ functions as well as on many biological processes. Polyunsaturated FA determine membrane fluidity and thus modulate activities of membrane proteins (enzymes, carriers and receptors). They also are precursors of pro- and anti-inflammatory eicosanoids and other autacoids (resolvins, protectins). Thus, alterations in lipid FA composition of critically ill patients affect reactivity of the organism to numerous pathological stimuli. The objective of this study was to analyse FA composition of plasma triacylglycerols, cholesteryl esters, plasma phospholipids and erythrocyte phospholipids in septic patients. The study group consisted of 30 septic patients, 19 of whom were available for three samplings: Sampling 1 was 24 hours after the onset of sepsis, Sampling 2 was 7 days after Sampling 1, and Sampling 3 was 7 days after recovery from sepsis. Eight septic patients died. Compared to healthy controls, a decrease in n-6 polyunsaturated fatty acids accompanied by increase in monounsaturated fatty acids in cholesteryl esters, plasma phospholipids and erythrocyte phospholipids persisted in all three samplings of septic patients. This effect of sepsis was significantly greater in cholesteryl esters and plasma phospholipids of non-surviving septic patients than in surviving ones. Moreover, non-survivors had lower proportions of n-3 polyunsaturated fatty acids in plasma phospholipids compared to survivors. The significant decrease in proportion of polyunsaturated fatty acids in lipids of septic patients in the course of sepsis reflects the severity of their critical state and supports the importance of appropriate nutritional polyunsaturated fatty acids supplementation.

Key words:
fatty acids of plasma lipids, fatty acids of erythrocyte phospholipids, n-6 PUFA, n-3 PUFA, C-reactive protein, sepsis, critical care

INTRODUCTION

Fatty acids (FA) are fundamental components of the human body. They provide important storage of energy and can influence the structure and function of cell membranes. Moreover, FA are substrates of many substances with very active physiological effects.

The composition of lipid FA is changing all through life due to various specific situations (e.g. starvation, aging, gravidity, changes in diet, and diseases). At the same time, these changes in FA composition and metabolism affect the organism’s reactivity to diverse stimulations. Simple or complex lipids of plasma and different tissues possess their characteristic FA profiles with varying proportions of saturated and unsaturated FA. The position of the first double bond from the methyl-end (ω position) of the hydrocarbon chain of unsaturated FA is used to designate n-7, n-9, n-6 and n-3 series. The n-6 and n-3 series denote polyunsaturated fatty acids (PUFA). Humans and other mammals are only able to synthesize saturated FA (SFA) and monounsaturated FA (MUFA) of the n-7 and n-9 series because they lack delta (Δ)12 and Δ15 desaturases (present in plants) for insertion of the double bond at the n 6 or n-3 positions of PUFA. Thus, mammals must obtain essential n-6 (linoleic acid, LA, 18:2n-6) and essential n-3 (α-linolenic acid, ALA, 18:3n-3) PUFA from the diet. Mammalian cells can interconvert PUFA within each series by elongation, desaturation and retroconversion. After ingestion, LA is metabolized by a series of alternating oxidative desaturation and elongation steps principally to dihomo-gamma-linoleic acid (DHGLA, 20:3n-6) and arachidonic acid (AA; 20:4n-6). The main metabolic pathways for conversion of PUFA are shown in Figure 1. The Δ6 pathway is responsible for conversion of LA to AA and of ALA to eicosapentaenoic acid (EPA, 20:5n-3) and occurs primarily in the endoplasmic reticulum of liver cells. After ingestion of essential PUFA, these can be converted to longer chain and more unsaturated PUFA by the action of elongases and desaturases (1,2). The conversion of ALA to EPA and docosahexaenoic acid (DHA, 22:6n-3) is limited in humans. It is supposed that less than 10% of ingested ALA is converted to EPA. Furthermore, human adipose tissue has only limited storage capacity for long-chain n-3 PUFA. In humans, these facts imply a need for continuous supply of long-chain n-3 PUFA through the diet (3).

There are many biological effects related to PUFA. PUFA of both families are bound preferentially in the sn-2 position of membrane phospholipids, contribute to the biophysical properties of membrane (fluidity, thickness and deformability), and regulate the activity of trans-membrane proteins (enzymes, carriers and receptors). Membrane phospholipids are also the source of lipid signalling molecules such as diacylglycerols, phosphatidic acid, inositol-1,4,5-triphosphate, ceramides and eicosanoids. The relative ratios of n-6 PUFA (LA, DHGLA, AA) and n-3 PUFA (ALA, EPA, DHA) determine their availability as substrates for cyclooxygenases (COX) and lipoxygenase (LOX) after phospholipase A₂ cleavage and hence the balance of eicosanoids (pro- and anti-inflammatory) and other autacoids (resolvins and neuroprotectins). Finally, PUFA are ligands for nuclear receptors such as peroxisome proliferator-activated receptor (PPAR) and retinoid X receptor that influence gene regulation. Both n-6 and n-3 PUFA, after desaturation and elongation to 20-carbon PUFA, can be metabolized to prostaglandins (PG), thromboxanes (TX), hydroxyeicosatetraenoic acids (HETE) and leukotrienes (LT) by the enzymatic activity of COX and LOX (4). Eicosanoids modulate cell growth and differentiation, inflammation, immunity, platelet aggregation and many other functions. Generally, eicosanoids derived from n-6 PUFA have pro-inflammatory effects whereas those derived from n-3 PUFA precursors have the opposite effects. DHGLA can be metabolized to form 1-series prostaglandins (PGE₁, a potent vasodilator and antiaggregator) and to 3-series LT (e.g. 15-hydroxyeicosatrienoic acid, HEtrE) that inhibit the formation of leukotriene B₄. Eicosanoids produced from AA by COX and LOX, respectively, are 2-series PG and 4-series LT that act as mediators of inflammatory processes per se (both series of mediators have pro-inflammatory effects, such as to induce pain, fever, vasodilatation). Moreover, eicosanoids derived from AA are able to modify the responses of other mediators (e.g. PGE₂ potentiates the pain caused by bradykinin) and can regulate other processes, such as leucocyte chemotaxis, blood clotting, platelet aggregation, cytokine production and immune functions (5-7). Inasmuch as PGE₂ suppresses synthesis of TNF-α, IL-1, IL-2 and INF-γ, it is in this respect an immunosuppressive agent. PGE₂ also inhibits 5-LOX, thus preventing generation of the inflammatory 4-series of LT. Furthermore, PGE₂ induces 15-LOX to produce the lipoxin A₄, an inflammation “stop signal”. Although PGE₂ has a distinct pro-inflammatory effect, it is also involved in mediating the resolution of inflammation through effects on the generation of other eicosanoids (8). EPA is a precursor of 3-series PG and TX that have antiaggregatory effects without immunosuppressive activity. LT of the 5-series derived from EPA display anti-inflammatory and antiaggregatory activities (9).

In addition to the biological effects of eicosanoids derived from n-3 PUFA, these FA exert indirect actions by inhibiting synthesis of eicosanoids derived from n-6 PUFA. The mode of n-3 PUFA functioning includes replacement of AA and reduction of the n-6 PUFA pool in membrane phospholipids (PL) by n-3 PUFA, which competes with n-6 PUFA for desaturases, elongases, COX and LOX (10). These processes result in decreased synthesis of PGE₂, TXA₂ and LTB₄. Thus, supplementation with n-3 PUFA can be associated with reduced platelet aggregation, blood clotting and modulation of inflammatory cytokine production and immune function (11). EPA can affect immune cell response through regulation of gene expression via PPAR and sterol regulatory element binding protein (SREBP). It can suppress expression of NF κB caused by lipopolysaccharide (9).

Sepsis is an important cause of morbidity and mortality in industrialized countries. Severe sepsis and septic shock remain relevant public health problems due to their frequency and high mortality rate. Sepsis, like other systemic inflammatory response syndromes (SIRS), is characterized by increased secretion of stress hormones (e.g. catecholamines and cortisol), cytokine overproduction, complement activation and mitochondrial dysfunction with decreased availability of ATP. Sepsis-related inflammation causes microcirculatory dysfunction, inadequate tissue oxygen supply, plus subcellular and cellular dysfunction (12,13). There are increasing data supporting the key role of reactive oxygen and nitrogen substances (RONS) and oxidative stress in endothelial damage in the pathogenesis of sepsis (14).

Numerous studies have shown substantial alterations in FA composition in the critically ill, including septic patients. Fatty acid oxidation rates, free FA turnover, and lipolysis are elevated, suggesting that enhanced mobilization and oxidation of fat is one of the fundamental responses to metabolic stress (15-17). It has been published that plasma levels of PUFA are reduced while SFA and MUFA are increased in cases of burn injury, as well as in patients with sepsis and acute respiratory distress syndrome (ARDS). This suggests an essential FA deficiency followed by increased oxidative stress (18-20).

Erythrocyte phospholipid FA composition partly reflects dietary fat intake. It also provides insight into PUFA metabolism and information about the incorporation of these FA into cell membranes. Thus, erythrocyte FA analysis can detect PUFA deficiency and imbalance caused by the diet but also by metabolic abnormalities and lipid peroxidation (21). PUFA erythrocyte alterations have been observed in cases of coronary heart disease, stroke, hypertension, inflammatory diseases and cancer (22-27).

The objective of the study was to analyse FA composition of plasma and erythrocyte lipids in the course of sepsis with the aim to identify potentially useful metabolic markers and predictors of sepsis outcome in critically ill patients.

PATIENTS AND METHODS

Patients: This was a prospective case control study in an adult medical intensive care unit (ICU) of the 4th Department of Internal Medicine of the First Faculty of Medicine and General University Hospital in Prague. A group of 30 septic patients (SP) and 30 age- and sex-matched healthy controls (HC) were included into the study. SP had to fulfil the criteria of sepsis according to the Society of Critical Care Medicine/American College of Chest Physicians (SCCM/ACCP) definitions (28) together with the following inclusion criteria: Acute Physiologic and Chronic Health Evaluation II (APACHE II) score > 10 (29) and C reactive protein (CRP) in serum > 20 mg/l. The primary source of sepsis in SP was lungs (20 cases). Other sources of sepsis were: central venous catheter (4 cases), abdominal infection (4 cases), and urinary tract infection (2 cases). All patients had standardized nutritional care according to ESPEN guidelines (30). Exclusion criteria for all patients in the study were: antioxidant therapy, chronic dialysis, diabetes, generalized tumours, immunosuppressive therapy and chemotherapy. In the study, 22 SP fully recovered from sepsis, 3 of whom were transferred to another health care facility while the remaining 19 were available for all three samplings (S1–S3). SP enrolled within 24 hours after the onset of sepsis (S1), then were sampled 7 days after S1 (S2), and finally one week after the clinical and laboratory cessation of sepsis symptoms (S3). Eight SP died due to sepsis either after S1 or S2 (non-survivors S1, n=8; non-survivors S2, n=4). Healthy subjects were defined as individuals without clinical and laboratory signs of sepsis, inflammation or known major disease. Written informed consent was obtained from all participants. The study protocol was approved by the Ethical Committee of the 1st Faculty of Medicine, Charles University in Prague.

Blood samples collection: Blood samples were taken from SP at S1, S2 and S3 and from HC after fasting overnight. All samples were marked with unique identification numbers to ensure anonymity, and data was merged only after assays had been completed. Blood samples were processed immediately after collection. For plasma, K2EDTA was used as anticoagulant. Erythrocytes, separated from plasma, were washed three times with saline and separated by centrifugation at 2500 x g for 5 min at laboratory temperature. Serum was prepared, following coagulation in vacutainer^® tubes, by centrifugation at 2500 x g at 4 °C for 10 min. The aliquots of serum, plasma and erythrocyte samples were stored in a freezer below −80 °C until analysis. The chemicals were purchased from Sigma (USA), unless otherwise indicated.

CRP concentration: The concentration in the serum was measured by immunoturbidimetric method using the K-ASSAY CRP kit (Kamiya Biomedical Company, USA) on a Hitachi Modular analyser (Japan).

Fatty acid composition of plasma and erythrocyte lipids: Plasma lipids were extracted according to a modified method from Folch (31). Plasma (1 ml) was dissolved in 21 ml of a chloroform-methanol mixture (2:1) and shaken in a pear-shaped flask. The serum protein precipitate was removed by filtration: 10 ml of chloroform-methanol-water mixture (3:48:47 v/v/v) was added and after a vigorous shaking, the lower lipid layer was separated and dried at 40 °C under a stream of nitrogen. Individual plasma lipids, i.e. total PL, triacylglycerols (TAG) and cholesteryl esters (CE), were separated by one-dimensional thin-layer chromatography (0.5 mm Silica Gel H, Merck, Germany) using the solvent mixture hexane-ether-acetic acid (70:30:1 v/v/v), detected by 2,7-dichlorofluorescein (0.005% in methanol), then scraped out and stored in a nitrogen atmosphere at −20 °C. On the next day, FA methyl esters were prepared and separated by gas chromatography (32). Erythrocyte lipids were extracted according to the method of Rose and Oklander (33). The same procedure as used for plasma PL was performed for separating total erythrocyte phospholipids (EPL) and their fatty acid analysis.

Statistical analysis: Data are expressed as mean ± standard deviation (SD) for parametric and as median and interquartile range (25th–75th percentiles) for non-parametric variables. Normality of data distribution was tested by Shapiro-Wilk test. Differences among compared groups were tested using one-way ANOVA with Scheffé and Newman-Keuls post-hoc comparisons. For nonparametric analysis, Kruskal-Wallis ANOVA was used. Friedman ANOVA was used for dependent analysis. All statistical analyses were performed using version 8.0 of StatSoft software Statistica (2007, CZ) and P < 0.05 was considered to be statistically significant.

RESULTS

Critically ill patients in the course of sepsis

In the study, 22 septic patients (SP) fully recovered from sepsis and 19 of them were available for three samplings. Table 1 presents the demographic and clinical characteristics and CRP in samplings S1, S2 and S3. The increased value of CRP persisted in all three samplings in comparison with HC. The CRP value was highest for S1 and gradually decreased in S2 and S3, but S3 did not reach the HC value.

Figure 2 demonstrates the proportion of FA classes in TAG, CE, plasma phospholipids (PPL) and EPL during the course of sepsis. Figure 2A shows slight increase in the MUFA proportion in the TAG of SP that is significantly higher in S2 and S3 compared to HC. A decrease in the n-6 PUFA proportion in CE of SP in all three samplings is accompanied by an increase in SFA and MUFA proportions (Figure 2B). These changes are mainly due to a decreased proportion of LA (18:2n-6) and enhanced proportions of palmitic (PA, 16:0), palmitoleic (POA, 16:1n-7), oleic (OA, 18:1n-9) and vaccenic ( VA, 18:1n-7) acids in SP (Table 2). In contrast to the fall in 18:2n-6 there is an enhanced proportion of DHA (22:6n-3) in CE of SP compared to HC in all three samplings, thereby causing a decline of the n-6/n-3 PUFA ratio. Similarly in PPL, there is a decline in n-6 PUFA and rise in MUFA that copy the movements observed in CE of SP. Moreover, increase in the n-3 PUFA proportion in PPL of SP is seen in S3 as compared to S2 and HC that is caused by increase in the proportion of 22:6n-3 in SP (Figure 2C, Table 3). The 16:1n-7/16:0 and 18:1n-9/18:0 ratios (Δ9 desaturase) and 18:3n-6/18:2n-6 ratio (Δ6 desaturase) are increased in CE and PPL of SP (Tables 2, 3). Changes in the proportions of individual FA classes in EPL of SP are similar to those of PPL but less significant (Figure 2D).

Comparison of surviving and non-surviving septic patients

This part of the study compares septic patients who survived until recovery (survivors 1st sampling [S1, n=22] and survivors 2nd sampling [S2, n=21]) with those who died after S1 or S2 and never recovered from sepsis (non-survivors S1, n=8 and non-survivors S2, n=4). Table 4 presents some clinical characteristics of septic patients, survivors/non-survivors. APACHE II in non-survivors is significantly higher compared to survivors. Figure 3 illustrates the proportions of fatty acid classes in CE and PPL in survivors and non-survivors. The proportion of MUFA is higher and both n-6 and n-3 PUFA lower in non-survivors compared to survivors in PPL from S1 and S2 samplings. In CE, higher MUFA and lower n-6 PUFA proportions in non-survivors compared to survivors are found in S1 only.

DISCUSSION

The most consistent finding of the present study of SP was decrease in n-6 PUFA levels in plasma and erythrocyte lipids (mainly due to the fall of LA) that was accompanied by a proportional increase in MUFA (mainly due to the rise of POA (16:1n-7), OA (18:1n-9), and VA (18:1n-7). The aforementioned changes were observed especially in CE and PPL. Similar shifts in FA proportions were observed also in EPL of SP in comparison with HC. The observed changes were present at the onset of sepsis (S1), culminated seven days after the onset (S2), and lasted until recovery (S3). Comparing SP survivors and SP non-survivors, the decrease in n-6 PUFA and increase in MUFA were significantly greater in non-survivors than in survivors. Moreover, a significantly lower proportion of n-3 PUFA was found in PPL of non-survivors than in survivors. The composition of FA in CE and PPL reflects both dietary intake of FA over the previous few weeks, as well as endogenous FA metabolism (de novo synthesis of FA, β-oxidation, enzymatic desaturation and elongation as well as lipoperoxidation) (34,35). In humans, there is only limited storage capacity for n-3 PUFA. The most abundant PUFA in human adipose tissue TAG is LA, which contributes 14% to the total amount of FA. On the other hand, the concentration of ALA (18:3n-3), the predominant n-3 PUFA in TAG of adipose tissue, comprises only 1% of total FA. In contrast to AA (20:4n-6), which contributes 7% of total FA in adipose tissue, long-chain n-3 PUFA (EPA, 20:5n-3 and DHA, 22:6n-3) are present there only in minute amounts (3). The aforementioned facts concerning the FA composition of adipose tissue imply a need for continuous supply of n-3 PUFA through the diet. There is low probability, however, that n-6 PUFA deficiency in SP is due only to decreased LA dietary intake. The decrease in n-6 PUFA can be generally caused by several mechanisms or their combinations: (i) reduced dietary intake of LA, (ii) impaired conversion of LA to its metabolites, (iii) increased demands of long-chain PUFA (especially AA) for PL and eicosanoid synthesis, and (iv) increased degradation of PUFA by β-oxidation and/or peroxidation. It is known that during critical illness (e.g. sepsis and trauma), activities of Δ-6- as well as Δ-5-desaturases are compromised (9). Furthermore, although n-3 PUFA contain more double bonds, they are thought to be more resistant to oxidation than LA (36). Our results are consistent with those of Pratt et al., who described reduced proportions of n-6 and n-3 PUFA and higher levels of MUFA and SFA in plasma and in erythrocyte lipids in burn patients early after injury. The decreased levels of AA and n-3 PUFA in plasma lipids of burn patients suggest increased demand for these FA for wound healing and immune processes following burn injury (20). Similarly, patients with ARDS revealed decreased plasma concentrations of LA compensated by an increase in OA and POA. In these patients, decreased plasma LA concentration was connected with an increase in the concentration of plasma 4-hydroxy-2-nonenal, one of its specific peroxidation products, thus suggesting severe oxidative stress during ARDS leading to the peroxidative changes of lipid molecules (37). Das et al. demonstrated negative correlation between free radical generation and the levels of n-6 and n-3 PUFA in PPL of patients with pneumonia, sepsis, and collagen vascular diseases as rheumatoid arthritis and systemic lupus erythematosus (2). Clinically, sepsis, SIRS and ARDS are characterized by disordered vascular control, which is initiated through the excessive production of RONS leading to redox imbalance and oxidative stress (38,39). The increased proportion of MUFA observed in SP resulted from a rise in concentrations of POA, OA and VA. Heightened concentrations of MUFA have been described in various pathological conditions, such as protein kwashiorkor malnutrition, anorexia nervosa, obesity, diabetes, pancreatic cancer, and cardiovascular disease, that are associated with decreased n-3 and n-6 PUFA concentrations. Increased OA and POA are surrogate markers of Δ9-desaturase, also known as stearoyl-CoA desaturase 1 (SCD1). This is a rate-limiting enzyme in MUFA synthesis that introduces a double bond in the Δ-9 position of acyl-CoA (40). SCD1 is the predominant liver isoform and appears to be a major regulator of energy metabolism by means of several mechanisms: (i) SCD1 activity could decrease FA oxidation via malonyl-CoA accumulation and inhibition of carnitine palmitoyl transferase-1; (ii) by altering the ratio of MUFA to SFA, it can be associated with obesity, diabetes, cardiovascular disease and cancer; and (iii) SCD1 alters expression of the genes for lipid metabolism. SCD1 activity is increased by such nutrients as glucose, fructose and dietary cholesterol, whereas PUFA decreases its activity. Insulin increases SCD1 activity and leptin has the opposite effect (41). In our setting of SP, there was an increased proportion of MUFA in lipids, caused by increased Δ9-desaturase activity that could compensate a loss of membrane fluidity due to the decrease in the content of PUFA in cell membranes. Characterizing the differences between SP survivors and non-survivors, the metabolic predictor in SP that did not survive sepsis is, in addition to an increased proportion of MUFA, a decreased proportion of n-3 PUFA. N-3 PUFA display pleiotropic beneficial effects, such as anti-inflammatory, vasodilatory, and antiaggregatory properties, that attenuate the severity and outcome of sepsis.

CONCLUSIONS

We have demonstrated a decrease in n-6 PUFA in septic patients that was accompanied by a proportional increase of MUFA in plasma lipids (mainly in CE and PPL) in comparison with HC. These changes were observed at the onset of sepsis (S1), culminated 7 days after the onset (S2), and lasted until recovery (S3). Whereas septic survivors lost n-6 PUFA only, deprivation of both n-6 and n-3 PUFA was observed in non-survivors, most probably due to intensified oxidative stress. Our finding of PUFA deficit, especially severe in septic non-survivors, points to a necessity for careful management of lipid supplementation in septic patients with respect to individual fatty acid classes.

ACKNOWLEDGEMENTS
This study was supported by a research grant of the Czech Ministry of Health (Project No. IGANR/8943-4).

Correspondence to:
František Novák, M.D., Ph.D.
4^th Department of Internal Medicine, General University Hospital in Prague
U Nemocnice 2, 12808, Prague, Czech Republic
Tel.: 420-224-962-506; Fax: 420-224-923-524
E-mail: fnova@lf1.cuni.cz