Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Eicosanoids are a family of oxygen-containing metabolites of arachidonic acid (AA) that play a fundamental role in inflammatory responses in mammals (1, 2). AA is stored within inflammatory cells esterified to complex glycerolipids, and is mobilized from those stores during cell stimulation by activated lipases. This event is central to the physiologic control of the inflammatory response (3). Several important factors appear to control the availability of AA for hydrolysis within cells. The first is the glycerolipid in which AA is stored. For example, it is generally assumed that cellular phospholipids are the major mobilizable stores of AA within mammalian cells. The subcellular distribution of AA-containing glycerolipids appears to be a second factor that influences the availability of AA for product formation (4, 5). For example, our studies have shown that there are distinct subcellular pools of AA within mast cells; one pool of AA serves as a source that is used for leukotriene (LT) formation, whereas another pool serves as a source of AA that is destined to be released unmetabolized from the cells (6). Other studies reveal that the cellular nucleus is an important subcellular location for many enzymes such as prostaglandin (PG) G/H synthase II (7), 5-lipoxygenase, 5-lipoxygenase-activating protein (8), and cytosolic phospholipase (PL) A₂ (9) that participate in AA metabolism.

Lipid bodies (LB) are spherical, non-membrane-bound, lipid-rich cytoplasmic inclusions found within various inflammatory cells (10). The number and size of LB increase in leukocytes from patients and animals with a variety of inflammatory disorders (15). For example, neutrophils recovered from the lung lavage fluid of subjects with lung injury contain significantly more LB than blood neutrophils (16). Likewise, increased numbers of LB are found in polymorphonuclear cells from the blood of subjects with rheumatoid arthritis (17) and from peritoneal exudates in rabbits treated with intraperitoneal sodium caseinate (18). In each of these situations, there is a concomitant increase in the number of cells that are characterized as having a "hypodense" phenotype (16).

The fatty acid composition of LB and their role in eicosanoid generation are not completely understood. Radiolabeled AA is incorporated into LB in numerous cell lines including peritoneal macrophages (11), human lung mast cells (19), neutrophils (12), and eosinophils (13). However, the glycerolipid distribution of AA within LB is not clear. For example, Lutas and Zucker-Franklin have shown neutrophil LB to be composed largely of triglycerides (TG) (20). Weller and colleagues (12) have demonstrated that eosinophils incubated with nanomolar concentrations of labeled AA esterify the majority of the label in PL. However, relative mass amounts of AA in the glycerolipid pools after incubation were not addressed (12). This latter group has also localized cystolic PLA₂ and its activating kinases to LB in U937 monocytes (21) and cyclooxygenase to LB in human eosinophils and 3T3 fibroblasts, potentially suggesting a role for LB in eicosanoid production (22). Our studies reveal that in situations where there are large numbers of LB in leukocytes, formed either in vivo or in vitro, AA-containing TG are the major stores of AA, by mass, within the cell (16).

The aforementioned studies have raised the question of whether stores of AA in TG can be mobilized and then further utilized for eicosanoid biosynthesis by leukocytes. This question is particularly important because AA-containing TG appear to be the major store of AA in leukocytes that migrate to sites of inflammation. In the current study, this hypothesis was tested using previously described techniques to generate polymorphonuclear leukocytes (PMN) containing numerous LB (16). Experiments in this study reveal that although large mass quantities of AA are present within PMN containing numerous LB, these pools of AA are not mobilized or utilized for LT generation during PMN activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

PGB₂, octadeuterated arachidonic acid ([²H₈]AA), and trideuterated stearic acid ([²H₃]SA) were obtained from Biomol Research Laboratories, Inc. (Plymouth Meeting, MA). Phospholipid standards used in normal-phase high-performance liquid chromatography (HPLC) were purchased from Avanti Polar Lipids, Inc. (Birmingham, AL). LTB₄, 20-hydroxy-LTB₄ (20-OH-LTB₄), and AA were obtained from Cayman Chemical Co. (Ann Arbor, MI). Monobasic potassium phosphate was obtained from Aldrich Chemical Co. (Milwaukee, WI). Bakerbond octadecyl (C18) bound to silica gel columns was purchased from JT Baker Chemical Co. (Phillipsburg, NJ). Ionophore A23187 was purchased from Calbiochem-Behring Corp. (San Diego, CA). Ficoll-Plaque was purchased from Pharmacia LKB Biotechnology, Inc. (Piscataway, NJ). All solvents (HPLC grade) were purchased from Fisher Scientific (Norcross, GA). Pentafluorobenzyl bromide (20% in acetonitrile) and diisopropylethylamine (20% in acetonitrile) were purchased from Pierce (Rockford, IL). [³H]AA and [¹⁴C]AA were purchased from American Radiolabeled Chemicals (St. Louis, MO). Thin layer chromatography (TLC) plates were purchased from Analtech (Newark, DE). RPMI 1640 medium was purchased from GIBCO BRL (Grand Island, NY). Insulin transferrin selenite solution, fatty acid-depleted human serum albumin (HSA), and fetal calf serum (FCS) were from Sigma Chemical Co. (St. Louis, MO). Antibiotics (penicillin/streptomycin mixture) were purchased from BioWhittaker (Walkersville, MD).

Isolation and Culture of Neutrophils

Venous blood was obtained from healthy human volunteers and PMN were isolated by dextran sedimentation followed by Ficoll-Paque density centrifugation (23). After isolation, PMN were suspended (2 million cells/ml) in RPMI 1640 containing 1% FCS, 10 µg/ml insulin, 5 µg/ml transferrin, 4 ng/ml sodium selenite, and 1% penicillin/ streptomycin. Cells were supplemented with 0, 5, or 20 µM exogenous AA (eAA) bound to HSA (1 mg/ml) and cultured for 12 h at 37°C. After incubation, cells were washed with Hanks' balanced salt solution (HBSS) without Ca²⁺. The percentage of cells recovered ranged from 72 to 86% and cell viability exceeded 90% in all experiments.

Labeling of Cellular AA Pools with [³H]AA and [¹⁴C]AA

Previous studies in our laboratory have revealed that the concentration of eAA provided to PMN dictates the glycerolipids where that AA will be incorporated (24). If high concentrations of AA are provided to PMN, this AA will be incorporated via de novo glycerolipid biosynthesis into predominantly TG pools. In contrast, low concentrations of AA provided to PMN are almost exclusively incorporated into phospholipid pools using the deacylation-reacylation remodeling pathway. These experiments took advantage of this differential incorporation by first labeling predominantly phospholipids in PMN with low concentrations of [³H]AA (5 nM [³H]AA [specific activity 200 Ci/ mmol]), and then labeling predominantly TG in PMN with high concentrations of [¹⁴C]AA (5.5 mM [¹⁴C]AA [specific activity 55 mCi/mmol]). This labeling protocol resulted in a high [³H]AA/[¹⁴C]AA ratio in phospholipids of PMN and a low [³H]AA/[¹⁴C]AA ratio in TG of PMN. PMN were then stimulated and the [³H]AA/[¹⁴C]AA ratios of LT products measured. Specifically, PMN were isolated from venous blood as described above and suspended in HBSS without Ca²⁺. A total of 10 µCi [³H]AA added as an albumin complex was provided to the cells for 5 min. Cells were removed from supernatant fluids and washed with cold HBSS without Ca²⁺. PMN were then suspended in culture media as described previously and incubated with 3 µCi [¹⁴C]AA for 12 h at 37°C. The [³H]AA/[¹⁴C]AA ratios in glycerolipids and LTs before and after stimulation with ionophore were determined after separation as described later.

Analysis of Neutrophil Lipids

To determine the fatty acid composition of glycerolipids, total lipids were extracted by the method of Bligh and Dyer (25). [²H₈]AA and [²H₃]SA (100 ng each) were added to aliquots of the total extract as internal standards. Initially, fatty acyl chains were hydrolyzed from glycerolipids with 0.6 N KOH in methanol/water (75/25, vol/vol) at 60°C for 30 min. After 30 min, water was added to samples and the reaction mixtures were acidified with 6 N HCL. A fatty acid-enriched fraction was obtained from this mixture by a previously described column extraction procedure using a C-18 octadecyl column (23). Fatty acids were then converted to pentafluorobenzyl esters using 20% pentafluorobenzyl bromide and 20% diisopropylethylamine in acetonitrile at 40°C for 40 min. Solvents were then evaporated under a stream of nitrogen and the samples were suspended in hexane. Fatty acid quantities were then determined by negative ion chemical ionization-gas chromatography mass spectrometry (NICI-GC/MS) as described later.

In some experiments, total lipids were separated into glycerolipid classes. Extracts were loaded onto an Ultraphere Silica column (4.6 × 250 mm; Rainin Instrument Co., Woburn, MA) and eluted with hexane/2-propanol/ ethanol/phosphate buffer (pH 7.4)/acetic acid (490:367:100: 30:0.6, vol/vol) for 5 min at a flow rate of 1 ml/min. The amount of phosphate buffer was increased to 5% over 10 min and this composition was maintained until all major gylcerolipids were eluted from the column.

Neutral lipids (NL), separated from other glycerolipid classes using HPLC as described previously, were further separated on silica gel G developed in hexane/ethyl ether/ formic acid (90:60:6, vol/vol/vol) as a mobile phase. Once separated, quantities of fatty acids in glycerolipids such as monoacyl-, diacyl-, and triacylglycerides were determined following base hydrolysis by NICI-GC/MS as described later.

Stimulation of PMN and Analysis of Products

PMN, suspended at 10 million/ml in HBSS containing calcium, were preincubated for 5 min at 37°C and then incubated with ionophore A23187 (2 µM) for an additional 5 min. Reactions were terminated by the addition of methanol/chloroform (2:1, vol/vol) or methanol alone for fatty acid or LT analysis, respectively. To determine the quantity of fatty acids released from glycerolipids during cell activation, 100 ng each of [²H₈]AA and [²H₃]SA were added to the terminated reaction mixture and lipids were extracted by the method of Bligh and Dyer (25). Fatty acids were then converted to pentafluorobenzyl esters as described previously and analyzed by NICI-GC/MS as described later.

To determine the quantity of LTs produced after cell stimulation, PGB₂ (250 ng) was added to each terminated reaction as an internal standard. Protein debris was removed by centrifugation (400 × g, 10 min) and an aliquot was then loaded onto a reverse-phase octadecylsilica (ODS) HPLC column (Ultrasphere, 4.6 × 250 mm; Rainin Instrument Co.). Lipids were loaded on the column with methanol/water/phosphoric acid (550:450:0.2, vol/vol/vol, pH 5.5) for 5 min at a flow rate of 1 ml/min. At 5 min, the composition of methanol was increased to 100% over 30 min. Appropriate peaks for LTB₄ were identified based on the elution times of standards run in the same system. Quantities of LTs were determined by comparing their areas with that of PGB₂ added as an internal standard and using appropriate standard curves to normalize for recoveries. [³H]/[¹⁴C] ratios of products were determined in fractions collected from the HPLC by liquid scintillation spectrometry.

NICI-GC/MS of Fatty Acids

NICI-GC/MS analysis was carried out on a Hewlett Packard 5989A single-stage quadrapole mass spectrometer (23). The gas chromatography was performed using a 30-m DB-17 fused silica column (SPB-5; 0.25-mm inner diameter, 0.25-mm film thickness; Supelco, Inc., Bellafonte, PA) on a Hewlett Packard 5890 GC. The initial column temperature was 60°C. The column was heated to 220°C at a rate of 40°C/min with subsequent increase in temperature to 280°C at a rate of 5°C/min. The injector temperature was maintained at 250°C. Each injection was performed in the splitless mode. A volume of 1 µl of 200 µl of recovered material dissolved in hexane was injected. Helium was used as the carrier gas. The pentafluorobenzyl esters were analyzed using selected ion-recording techniques to monitor for AA (m/z 303) and [²H₈]AA (m/z 311). A standard mixture of the aforementioned fatty acids was injected and analyzed by NICI-GC/MS before each biologic sample to obtain precise retention times.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

A Model in PMN for Increasing Levels of AA within PMN Glycerolipids

Initial experiments were designed to obtain a cellular model to test whether AA in cellular TG can be mobilized and further metabolized during cell activation. Isolated PMN were cultured for 12 h with 0, 5, and 20 µM eAA. Our previous studies suggested that a large proportion of AA presented to PMN at high concentrations (> 1 µM) is incorporated into a neutral lipid fraction. Figure 1 illustrates that the total mass of AA in all cellular lipids increased approximately 2- and 4-fold after incubation of PMN with 5 and 20 µM AA, respectively. This represents an increase from 14.9 ± 0.8 nmol/10 million in control PMN to 54 nmol/10 million in PMN after incubation with 20 µM eAA. These in vitro increases are similar to those observed when neutrophils move into airways of patients with adult respiratory distress syndrome (ARDS).

View larger version (11K): [in this window] [in a new window]   Figure 1.   Effect of culture with exogenous AA on total AA content of PMN glycerolipids. PMN were isolated from healthy volunteers and incubated for 12 h at 0, 5, and 20 µM eAA. Lipids were extracted from isolated PMN preparations by the method of Bligh and Dyer (25). After base hydrolysis and column extraction, the quantities of AA were determined by NICI-GC/MS. Data represent means ± standard error of the mean (SEM) for three separate experiments. *P  0.05 compared with 0 µM; *P < 0.05 compared with 5 µM.

Distribution of Arachidonate within Phospholipid and NL Classes

HPLC and NICI-GC/MS analysis of the AA content in glycerolipid classes revealed that PMN cultured with no eAA contain the bulk of their AA mass within phospholipids (phosphatidylethanolamine [PE] > phosphatidylcholine [PC] = phosphatidylinositol [PI] > NL), with AA in the neutral fraction accounting for only 10% of total cellular AA. There was a dramatic difference in lipid distribution of AA after culturing cells with 20 µM eAA. For example, NL contained the bulk (70%) of cellular AA after incubation of PMN with 20 µM eAA (Figure 2A). By contrast, there were only modest increases in the AA content of any phospholipid class after culturing cells with 20 µM eAA. These data indicate that almost all of the increase in the cellular content of AA observed after exposure of PMN to high concentrations of AA can be accounted for by an increase in cellular NL.

View larger version (14K): [in this window] [in a new window]   Figure 2.   Effect of culture in exogenous AA on the distribution of AA in glycerolipid classes (A) and neutral lipid subclasses (B). (A) PMN were isolated from healthy volunteers and incubated for 12 h at 0 and 20 µM AA. Lipids were extracted from isolated neutrophil preparations by the method of Bligh and Dyer (25) and further separated into glycerolipid classes by normal-phase HPLC. The AA content in glycerolipid classes was determined by NICI-GC/MS. NL, PE, PI/PS, and PC represent fatty acids in NL, PE, PI/phosphatidylserine (PS), and PC, respectively. *P  0.05 compared with 0 µM. (B) Neutral lipid subclassess were further separated by TLC. Mono, Di, FFA, and Tri represent the quantity of AA in monoglycerides, diglycerides, free fatty acids, and triglycerides, respectively. These data are means ± SEM of three separate experiments. *P  0.05 compared with 0 µM.

Further TLC analysis (Figure 2B) of the NL fractions obtained from HPLC revealed that AA was distributed in several NL including diacylglycerides (10% of total), free fatty acids (32% of total), and TG (48% of total) in cultured cells without exposure to AA. After culturing cells with 20 µM eAA, the mass quantities of AA increased almost exclusively in cellular TG, going from 0.8 ± 0.6 nmol/ 10 million control PMN to 33.1 ± 4.2 nmol/10 million PMN. These data reveal that the observed increase in the neutral lipid fraction was almost exclusively due to an increase of AA in cellular TG. Again, this increase of AA in cellular TG along with a large number of cytoplasmic LB and a hypodense phenotype is observed in vivo in PMN that reside in the lungs of patients with ARDS (16).

Correlation between Mass of AA in TG and Capacity of PMN to Mobilize and Metabolize AA

The aforementioned experiments established a cellular model in PMN whereby cells could be generated that contained various quantities of AA esterified to TG. This provided a means to determine whether AA could be mobilized from TG pools and, if so, whether this AA was further metabolized to eicosanoids. Two sets of experiments were performed to test this hypothesis. In the first set of experiments, TG pools within PMN were preloaded to varying degrees with eAA. These cells were then stimulated and the influence of different mass quantities of AA within TG on the capacity of PMN to mobilize AA and synthesize LTs was determined. Figure 3A shows that, after culture with 20mM eAA, PMN mobilized increased amounts of free AA (0.636 nmol/10⁷ PMN; 1.2% of total cellular AA) as compared with control cells (0.333 nmol/ 10⁷ PMN; 2.2% of total cellular AA). However, this increase in AA release was modest compared with the 43-fold increase in the mass of AA found within TG pools after incubation of cells with 20 µM AA. Similarly, the capacity of PMN, with preloaded AA-containing TG pools, to synthesize LTB₄ was monitored. Quantities of LTB₄ increased approximately 1.5- and 2-fold after preloading TG pools of PMN with 5 and 20 µM AA, respectively (Figure 3B). Again, the enhanced capacity of PMN to synthesize LTB₄ was low relative to the large increase in the pool size of AA within cellular TG.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Release of AA (A) and LTB4 biosynthesis (B) after stimulation of PMN. After incubation in 0, 5, or 20 µM eAA, neutrophils were isolated, suspended at 10 million/ml, and stimulated with 2 µM ionophore A23187 for 5 min at 37°C. (A) AA levels: Reactions were terminated and fatty acids were extracted by the method of Bligh and Dyer (25), and AA levels were quantitated by NICI-GC/MS. These data represent means ± SEM of three separate experiments. *P  0.05 compared with 0 µM; #P < 0.05 compared with 5 µM. (B) LTB4 levels: Reactions were terminated and cells removed from supernatant by centrifugation. After extraction, an aliquot of the supernatant fluid was analyzed by reverse-phase HPLC. Quantities of LTs were determined by comparing their areas under the curve with that of PGB2 added as an internal standard. These data are means ± SEM of three separate experiments. *P  0.05 compared with 0 µM.

Correlation between Isotopic Ratios of Glycerolipid Precursors and LT Products

The aformentioned data reveal that while there were increases in both AA mobilized and LTB₄ synthesized, these increases did not correlate with the mass increase of AA within cellular TG pools observed after preloading cells with high concentrations of eAA, suggesting that TG-AA did not participate in eicosanoid generation. However, despite this inconsistency, this experiment did not rule out the possibility that AA from TG was contributing to eicosanoid generation. To address this question more definitively, the large, AA-containing pools within phospholipid classes and TG were double-labeled with [³H]AA and [¹⁴C]AA in such a manner that phospholipid classes and TG had significantly different [³H]AA/[¹⁴C]AA ratios. This was accomplished by using low concentrations of [³H]AA predominantly to label phospholipid class and high concentrations of [¹⁴C]AA predominantly to label TG. Once cells were prelabeled in this manner, the [³H]AA/ [¹⁴C]AA ratios in TG and PL pools were measured. PMN were subsequently stimulated and the [³H]AA/[¹⁴C]AA ratios in PL, TG, and LTB₄ and its major metabolite 20-OH-LTB₄ were measured. The [³H]AA/[¹⁴C]AA ratios in TG and PL did not change with stimulation. As shown in Figure 4, after stimulation, the [³H]AA/[¹⁴C]AA ratios of LTB₄ and 20-OH-LTB₄ match closely and are similar to the ratio in PL classes. This is particularly the case with PI, where the ratios of LTB₄, 20-OH-LTB₄, and PI are 30, 31, and 32, respectively. In contrast, the [³H]AA/[¹⁴C]AA ratio in TG is significantly different from those seen in LTB₄ or 20-OH-LTB₄. Because there is a large mass of AA within cellular TG, even a small mass contribution of the low [³H]AA/[¹⁴C]AA pool within TG would have lowered the [³H]AA/[¹⁴C]AA ratio of LTB₄ and 20-OH-LTB₄ below that seen within phospholipids. Thus, these experiments provide strong evidence that cellular TG do not serve as stores of AA for LT generation.

View larger version (11K): [in this window] [in a new window]   Figure 4.   Correlation between the isotopic ratios of glycerolipid precursors and LT products. Isolated PMN were labeled with [3H]AA and [14C]AA as described in MATERIALS AND METHODS. Cells were then stimulated with A23187 and the [3H]AA/[14C]AA ratio in phospholipid pools (PC, PI, and PE), triglycerides (TG), LTB4, and 20-OH-LTB4 was determined by liquid scintillation spectrometry after isolation. These data are means ± SEM of three separate experiments. *P  0.05 compared with LTB4 and 20-OH-LTB4.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Previous studies reveal that migration of inflammatory cells into the lung or other sites of inflammation where there are large concentrations of AA induces an accumulation of AA into TG pools, leading to morphologic and densitometric changes in cells. These studies have raised the question of whether AA-containing TG in LB are major sources of AA for eicosanoids. Alternatively, arachidonyl-TG in LB may serve as an expandable pool of AA to protect cells and tissues from extremely high concentrations of AA present in inflammatory fluids. Regarding the former, this question is extremely important because PLA₂ has historically been viewed as the major mechanism by which AA is mobilized from cellular glycerolipids, and thus inhibitors of PLA₂ have been seen as a major therapeutic target.

The current study has developed in vitro models to examine whether arachidonyl-TG in LB serves as major sources of AA for eicosanoid generation in PMN. This and previous studies reveal a direct correlation between the exposure of neutrophils to different concentrations of eAA and the appearance of AA in TG as well as the number of LB. However, there was not a good correlation between the size of AA pools within TG of PMN and the capacity of PMN to mobilize AA and synthesize LTB₄ or 20-OH- LTB₄. Despite increasing the pool sizes of AA-containing TG 43-fold (by preloading PMN with eAA), released AA, LTB₄, and 20-OH-LTB₄ increased only modestly after cell activation. However, these experiments still did not rule out the possibility that some AA could have been mobilized from AA-containing TG. Further experiments measuring [³H]AA/[¹⁴C]AA ratio of potential glycerolipid precursors and LT products revealed that TG clearly did not serve as a major source of AA for LT generation.

Therefore, this study indicates that cellular phospholipids and not AA present in TG pools within LB are the major, direct sources of AA available for eicosanoid formation within cells, even when the bulk of the cell's endogenous pools of AA reside in TG. Moreover, inhibitors of certain PLA₂ isotypes, and not TG lipase(s), are likely to inhibit the mobilization of AA at sites of inflammation. Thus, these studies have led us to favor the alternative hypothesis that TG pools within LB serve as expandable storage pools for AA to prevent extensive tissue damage associated with inflammation caused by AA and its metabolites. Other potential roles of arachnidonyl-TG in LB need to be more fully explored.