An esterification protocol for cis-parinaric acid-determined lipid peroxidation in immune cells1,2

An Esterification Protocol for cis-Parinaric Acid-Determined Lipid Peroxidation in Immune Cells1,2 Susan O. McGuirea,*, Marilyn R. James-Krackeb, Grace Y. Sunc, and Kevin L. Fritschea Departments of aAnimal Sciences, bPharmacology and cBiochemistry, The University of Missouri, Columbia, Missouri 65211

ABSTRACT: Loss of fluorescence from cis-parinaric acid (cPnA) is a sensitive indicator of lipid peroxidation. The purpose of this study was to utilize cPnA to determine, at the level of the intact immune cell, whether enrichment of membranes with polyunsaturated fatty acids (PUFA) increased lipid peroxidation. P388D1 macrophages were labeled by addition of cPnA as an ethanolic solution. Within two minutes of addition, in the absence of serum, cPnA rapidly intercalated into the plasma membrane. Lipid peroxidation was initiated by addition of Fe2+EDTA resulting in a dose-dependent decrease in fluorescence with increased oxidant concentration. Cells previously enriched with PUFA and labeled by intercalation showed no differences in spontaneous or Fe2+-induced lipid peroxidation. In separate experiments, 20 µM cPnA in ethanolic solution was injected into cell culture media containing 0.1% essentially fatty acid free bovine serum albumin (BSA). Cells were resuspended and incubated for 90 min at 37°C. After washing with BSA to remove cPnA which had not incorporated, 0.5% (0.1 µM) of the added cPnA was found esterified within cellular lipids. This level of cPnA provided a 100-fold increase over basal autofluorescence levels. Cells labeled in this manner also lost fluorescence in a dose-dependent manner as levels of oxidant stress increased. Cells enriched with PUFA and labeled by esterification had significantly increased rates and total amounts of lipid peroxidation. Co-incubation with α-tocopherol and PUFA resulted in a decrease in lipid peroxidation which was not significantly different from control cells. In conclusion, esterification of cPnA into membrane phospholipids can sensitively detect changes in lipid peroxidation induced by alteration of membrane PUFA and/or vitamin E content. Lipids 32, 219–226 (1997).

Membrane fatty acid composition can be rapidly modified by changing the source of dietary lipid (1,2). In fact, when polyunsaturated oils are increased in the diet as a means of 1

Presented in part at the Experimental Biology Meetings, Anaheim, California, April 1994. 2 Contribution from the Missouri Agriculture Extention Station, Journal #12,495. *To whom correspondence should be addressed at The Department of Biochemistry, M121 Medical Sciences Building, The University of Missouri, Columbia, MO 65212. Abbreviations: AA, arachidonic acid; BSA, bovine serum albumin; cPnA, cis-parinaric acid, 9, 11, 13, 15-cis-trans-trans-cis-octadecatetraenoic acid; DHA, docosahexaenoic aicd; EPA, eicosapentaenoic acid; FBS, fetal bovine serum; PBS, phosphate buffered saline; PUFA, polyunsaturated fatty acid; TLC, thin-layer chromatography; UV, ultraviolet. Copyright © 1997 by AOCS Press

decreasing saturated fat, tissues become enriched with polyunsaturated fatty acids (PUFA) such as eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), and arachidonic acid (AA). Membranes containing increased levels of these fatty acids are more easily oxidized (3,4). Indeed, increased lipid peroxidation has been observed in rats (5–9) and humans (10,11) upon feeding fish oil which is rich in EPA and DHA. Tissue damage resulting from lipid peroxidation has been implicated in conditions such as atherosclerosis, sickle cell anemia, hemochromatosis, ischemia-reperfusion injury, and arthritis (12,13). Vitamin E is widely accepted to be the primary lipid-soluble antioxidant responsible for protection of unsaturated membrane components. Increasing dietary PUFA has been reported to be antagonistic to vitamin E status (14–17). Both the antagonistic effect of dietary PUFA and the proposed increased requirement at the level of the PUFA-enriched membrane have provided a basis for an increase in the dietary vitamin E requirement (16–19). Increases in membrane unsaturation without concomitant increases in membrane vitamin E could render cells more susceptible to oxidative stress. Macrophages, which produce reactive oxygen species (e.g., O2-., H2O2) for phagocytic killing, could be at potentially greater risk than other cells and tissues (20,21). Consequently, the ability to measure lipid peroxidation at the level of the intact cell becomes important in the determination of an overall dietary vitamin E requirement. Although numerous methods are available for measuring lipid peroxidation, these methods are discontinuous and indirect, requiring time point sampling. Oxidative by-products are monitored rather than the actual, continuous process of lipid peroxidation. These methods require tissue homogenization and sample preparation which may include lipid extraction, denaturing of protein and chemical derivatization, possibly introducing artifacts (13,22,23). Recently, a new method has been described which allows direct measurement of lipid peroxidation at the level of the cell membrane without disruption of cellular integrity and function (24–26). This sensitive method which is both direct and continuous, uses cisparinaric acid (9, 11, 13, 15-cis-trans-trans-cis-octadecatetraenoic acid, cPnA), a naturally occurring fatty acid with conjugated tetraene structure which renders it fluorescent in a

219

Lipids, Vol. 32, no. 2 (1997)

220

METHOD

lipid environment. Exposure of erythrocyte ghosts or intact erythrocytes to ethanolic cPnA results in rapid intercalation of the free fatty acid into the membrane bilayer. Rearrangement of the tetraene structure of cPnA after electron abstraction during lipid peroxidation causes loss of fluorescence and is therefore a direct measure of lipid peroxidation which may be monitored as it occurs. One of the overall goals of our laboratory is to understand the interaction between dietary PUFA and vitamin E, relative to lipid peroxidation, at the level of the immune cell. Our objective was to use cPnA in intact immune cells to assess the effect of modulating membrane components (i.e., PUFA and α-tocopherol), on the susceptibility of living cells to lipid peroxidation. Here, we present evidence that with minor modifications to the published method, useful information about nutritional modulation of lipid peroxidation in immune cells can be obtained. MATERIALS AND METHODS Materials. Solvents were high-performance liquid chromatography grade or Fisher Optima grade (Fisher Scientific Co. St. Louis, MO). Silica gel thin-layer chromatography (TLC) plates were purchased from Fisher Scientific Co. (St. Louis, MO). α-Tocopherol was the kind gift of Eastman Kodak Chemical (Rochester, NY). Arachidonic acid (AA) and eicosapentaenoic acid (EPA) were purchased from Cayman Chemical Co. (Ann Arbor, MI). cPnA was purchased from Molecular Probes Inc. (Plano, TX). Stock solutions of cPnA (2 µM in study #1, 20 µM in study #2), α-tocopherol, and fatty acids were made in absolute ethanol, previously deoxygenated by bubbling with nitrogen gas. Stock solutions were stored at −80°C under nitrogen gas. The murine monocyte cell line, P388D1, was obtained from American Type Culture Collection (ATCC, Rockville, MD). Cell culture reagents were obtained from the Cell and Immunobiology Core Facilities (University of Missouri, Columbia, MO). Cell culture. P388D1 cells were maintained in RPMI 1640 supplemented with 15% fetal bovine serum (FBS), 1 mM glutamine, 100 µg/mL penicillin, and 100 units/mL streptomycin (complete medium). Cells were grown in 75 cm2 flasks with confluent cells subcultured after detaching by mild scraping. Enrichment of P388D1 cells with α-tocopherol and/or AA and EPA. For all experiments involving cellular enrichment with α-tocopherol or the polyunsaturated fatty acids (PUFA), AA and EPA, stock solutions in absolute ethanol were added to complete media which was then used to seed cells grown to near confluence (90%). At no time did total ethanol addition exceed 0.2% of the total volume. Incorporation of these membrane constituents was promoted by allowing cells to incubate at 37°C for 12 h under a humidified atmosphere of 95% air and 5% CO2. After this culture period, cells were harvested by gentle scraping as previously described. Addition of an equal concentration of absolute ethanol into complete media was used in control cell cultures. Viability of cells, as Lipids, Vol. 32, no. 2 (1997)

determined by trypan blue exclusion, exceeded 90% under all culture conditions described. Intercalation of fluorescent probe in cellular membranes. P388D1 cells were found to have a low level of autofluorescence (less than four arbitrary fluorescence units) at 324 and 413 nm for excitation and emission, respectively. Injection of 2 µM cPnA into the cell suspension resulted in a dramatic and immediate rise in fluorescence (i.e., 400-fold increase). In the presence of a 30% filter, fluorescence stabilized within two minutes and was determined to be a linear function of probe concentration between 1 and 4 µM cPnA. A cell concentration of 4 × 105 cells/mL was sufficient to completely incorporate 4 µM cPnA such that cPnA remaining in the water phase was minimal. Therefore for our studies with the monocyte cell line, we chose injection of 2 µM ethanolic cPnA solution into a cell suspension containing 4 to 6 × 105 cells/mL of phosphate buffered saline (PBS) without Ca+2 and Mg+2 followed by a two-minute equilibration period for the intercalation experiments. Each compound used in the assay (e.g., ethanol, FeCl2 · 4 H2O, ethylenediaminetetraacetate, vitamin E) was tested individually to check for autofluorescence and/or quenching of the cPnA signal prior to establishing individual protocols. Incorporation of fluorescent probe into cellular lipids. The labeling protocol for esterification experiments was established as 20 µM cPnA in 0.1% essentially fatty acid-free bovine serum albumin (BSA) in RPMI 1640 for 90 min. The increase in cPnA concentration was necessary to provide a 100-fold increase in fluorescent signal over the basal autofluorescence. cPnA in ethanolic solution was injected into complete media which was then used to resuspend cells for incubation. After a 90-min incubation at 37°C, cells were washed three times with RPMI 1640 containing albumin (0.1% BSA), which was sufficient to remove nonesterified cPnA from the cells. The amount of cPnA not incorporated into the cells was determined by separate extraction of the lipids in the original supernatant and each of the three wash supernatants with 4 vol of chloroform/methanol (2:1, vol/vol). The cPnA in each fraction was quantified by ultraviolet (UV) absorbance (304 nm) after the sample was resuspended in absolute ethanol. The incorporation of cPnA into the cellular lipids of P388D1 was assessed by extracting total lipids from cell pellets. Briefly, 4 vol of chloroform/methanol (2:1, vol/vol) were mixed with the thawed cell pellet. The lower phase was placed into a glass tube and evaporated in a 40°C water bath under a stream of N2 gas. The total lipid extract was resuspended in a small volume of chloroform/methanol (2:1, vol/vol), and applied to a 2-cm lane of a TLC plate (27). Total cellular lipids were separated into major lipid classes (i.e., phospholipid, triglyceride, cholesterol ester, free cholesterol, free fatty acids) using hexane/diethyl ether/glacial acetic acid (80:20:2, by vol) as the developing solvent (28). Identity of lipid classes was verified with commercial standards. Individual lipid classes were removed from the plate by scraping and then extracted with 750 µL double-distilled H2O and 3 mL chloroform/methanol (2:1, vol/vol). Individual lipid classes

METHOD

were resuspended in absolute ethanol and the cPnA concentration determined by UV absorbance at 304 nm. In order to provide gas chromatographic verification that cPnA was esterified into membrane phospholipids, it was necessary to modify the labeling conditions. P388D1 cells were harvested at 90% confluence and pelleted by centrifugation at 250 × g for 10 min. Cells were resuspended at a concentration of 106 cells/mL in complete media to which 40 µM cPnA had been added then incubated. After a 24-h incubation at 37°C, cells were harvested and cellular lipids were extracted using the methods described previously. Lipid classes (i.e., phospholipid, free fatty acid, triglyceride, cholesterol ester) were separated as previously described. Isolated lipids were transmethylated with 4% H2SO4 in methanol (88°C, 1 h). Some cell lipid extracts were subjected to two-dimensional TLC in order to determine the incorporation of cPnA into individual phospholipid classes. In this case, methyl esters were prepared by transmethylation using 0.5M NaOH in methanol (28). Fatty acid methyl esters were extracted and then injected into a Hewlett-Packard gas chromatograph (Sunnyvale, CA) Model 5890A with a 30-m capillary column (Supelcowax 10; Supelco, Bellefonte, PA). Results, expressed as the percentage of total fatty acids, were determined by electronic integration (Hewlett-Packard 3380A integrator). Lipid peroxidation assay. Two mL of cells in PBS (0.4 to 0.6 × 106 cells/mL), which had been labeled with cPnA either by intercalation or esterification, were placed in each of four cuvettes in a Perkin-Elmer spectrofluorometer MPF-66 (Norwalk, CT), equipped with a stirring apparatus and a jacketed, automated four-chamber cuvette holder to maintain the cells at 37°C for the time course of the experiment. Fluorescence was monitored by excitation at 324 nm and determining emission at 413 nm (29). Intercalation experiments were conducted in the presence of a 30% light filter. After measurement of initial fluorescence from each cuvette, lipid peroxidation was initiated by addition of an equimolar solution of FeCl2 · 4H2O and EDTA such that final concentrations were 10 and 100 µM Fe+2 in 10 and 100 µM EDTA, respectively (30). Lipid peroxidation has been shown to be initiated by hydroxyl radicals derived from a Fenton-type reaction when ferrous iron and EDTA are present in equimolar concentrations (30–32). Loss of fluorescence, which served as a measure of lipid peroxidation, was monitored simultaneously every 2 min for each of four cuvettes, in real time and is expressed as the percentage of initial fluorescence remaining at a given time point. Since cPnA fluorescence has been found to undergo spontaneous decay after exposure to light, a cuvette containing cPnA-labeled cells with no additives or added oxidant stress was monitored simultaneously as a control for each experiment (24). The rate of spontaneous decay of cPnA signal in control cells was measured with and without enrichment with α-tocopherol. The presence of this antioxidant did not change either the rate or amount of fluorescence lost, indicating that loss of signal in control cells was probably due to photobleaching rather than lipid peroxidation of the probe. Statistical analysis. All data were analyzed as a repeated,

221

split-plot analysis of variance as outlined by Gil and Hafs (33). The linear statistical model contained the effect of treatment, replicate within treatment, time and treatment × time (34,35). The main effect of treatment was tested using replicate within treatment as the denominator of F. All trend analyses were tested using the residual error term. Each experiment was run in duplicate with values averaged for each treatment. Each experiment was replicated from three to six times as indicated in the figure legends. A one-degree-of-freedom polynomial linear and quadratic contrast was used for testing trend analyses (36). RESULTS Location of esterified label. To determine the quantity and location of cPnA esterified into cellular lipids, a series of experiments were conducted utilizing the UV properties of this fatty acid. This approach is more sensitive than gas chromatography. After 90 min incubation with 20 µM cPnA in the presence of 0.1% essentially fatty acid free BSA, the incorporation of cPnA into cellular lipids was less than 1% (0.6%) of the added cPnA. In order to ensure that the fluorescent signal was coming from esterified rather than intercalated cPnA, cells were washed with RPMI which contained 0.1% essentially fatty acid-free BSA to facilitate removal of any cPnA intercalated into the plasma membrane. Total lipids from wash supernatants were extracted and the cPnA concentration determined. Over 98% of the cPnA originally added was recovered in the incubation supernatant with less than 1% recovered in the first wash and less than 0.2% in each of the subsequent two washes. Results from five separate experiments showed that under these conditions, approximately 50% of the incorporated cPnA was found in the phospholipids, 20% in the triglycerides, 18% in the cholesterol esters, and 12% in the free fatty acid fraction. Although less than 1% of the added cPnA was esterified into cellular lipids (0.1 µM), this amount was sufficient to provide an adequate fluorescence with which to conduct these studies. Furthermore, the majority of the cPnA was located in the membrane phospholipid fraction of these cells. Chromatographic evidence for cPnA incorporation into phospholipid is provided in Figure 1 and was accomplished by increasing both the concentration of cPnA and the length of incubation time. The peak identified as cPnA was present in the TLC isolated phospholipid fraction of cells incubated with cPnA, but was absent in cells incubated without exogenous cPnA. The cPnA peak in enriched cells eluted at the same time as did the methyl ester of the free fatty acid. Separation of P388D1 phospholipid classes by two-dimensional TLC gave evidence that cPnA was in the phosphatidylcholine and phosphatidylethanolamine fractions. In contrast, when P388D1 cells were incubated with cPnA in medium without a source of albumin (e.g., no FBS), incorporation of cPnA into phospholipids was not observed (data not shown). The effect of labeling method on the sensitivity of cPnA to increasing levels of oxidant stress. P388D1 cells were labeled Lipids, Vol. 32, no. 2 (1997)

222

METHOD

FIG. 1. Gas chromatogram of phospholipid extracted from P388D1 cells and separated by one-dimensional thin-layer chromatography, incubated without (A) or with (B) 40 µM cis-parinaric acid, 9, 11, 13, 15cis-trans-trans-cis-octadecatetraenoic acid (cPnA) in the presence of 0.1% essentially fatty acid-free bovine serum albumin for 24 h; DHA, docosahexaenoic acid; AA, arachidonic acid.

by the intercalation method, and then subjected to increasing levels of oxidative stress (i.e., Fe2+-EDTA complex). The rate at which fluorescence was lost increased in a dose-dependent fashion (Fig. 2A). Both the rate at which lipid peroxidation occurred and the total amount of probe oxidized were significantly affected by treatment as evidenced by significant differences in linear and quadratic function (P < 0.05). Cells labeled by esterification of cPnA (Fig. 2B) responded in a similar manner to increasing levels of oxidant stress as provided by ferrous-EDTA. Each increasing dose of ferrous-EDTA resulted in a significant increase in both rate and total fluorescence lost (P < 0.05). Cells labeled by esterification lost from 13 to 17% more fluorescence (10 and 100 µM Fe2+–EDTA complex, respectively) than did cells labeled by intercalation. When differences in the kinetics of photobleaching were taken into consideration, this difference was reduced to only 5%. Oxidant stress provided by enrichment with long-chain PUFA. To further evaluate the use of cPnA as a means to assess lipid peroxidation in immune cells, we incubated P388D1 cells in media supplemented with two long-chain PUFA (i.e., 5 µM EPA plus 5 µM AA). After 12 h, the degree of membrane unsaturation was greatly increased. This was primarily a result of the replacement of linoleic acid with EPA Lipids, Vol. 32, no. 2 (1997)

FIG. 2. (A) Effect of increasing oxidant stress on loss of fluorescence in P388D1 cells, labeled with 2 µM cPnA by intercalation for two minutes. Cell concentration was 0.5 × 106 cells/mL in phosphate-buffered saline without calcium or magnesium. Lipid peroxidation was induced at time zero by addition of 10 or 100 µM FeCl2 · 4 H2O in equimolar EDTA. Zero µM represents spontaneous decay of signal from control cells in the presence of the light source. Values used are means ± SEM for three separate experiments. Curves assigned different superscript letters are significantly different (P < 0.05) based on linear and quadratic analysis. (B) Effect of increasing oxidant stress on loss of fluorescence in P388D1 cells, labeled by esterification with 20 µM cPnA in 0.1% bovine serum albumin for 90 min. Cell concentration was 0.5 × 106 cells/mL in phosphate-buffered saline without calcium or magnesium. Lipid peroxidation was induced at time zero by addition of 10 or 100 µM FeCl2 · 4 H2O in equimolar EDTA. Zero µM represents spontaneous decay of signal from control cells in the presence of the light source. Values used are means ± SEM for three separate experiments. Curves assigned different superscript letters are significantly different (P < 0.05) based on linear analysis. Quadratic analysis shows that the shapes of the curves for the 0 and 100 µM treatment groups were not different from each other, but were significantly different from those exposed to 10 µM FeCl2 (P < 0.05). See Figure 1 for abbreviations.

and AA in membrane phospholipids. For example, the AA content of the phosphatidylcholine fraction from these cells was increased from 0.915 to 2.08 nmoles/106 cells and EPA from 0.0 to 0.41 nmoles/106 cells for EPA with a concomitant

METHOD

223

30% decrease in linoleic acid. We hypothesized that such an enrichment in unsaturation would lead to significantly greater loss of cPnA upon oxidative challenge. However, the increased levels of long-chain PUFA did not increase the loss of basal fluorescence or loss of signal in the presence of an oxidant stress when cells were labeled by the intercalation method (data not shown). In contrast, upon exposure to the light source, without addition of our exogenous oxidant stress, cells enriched with PUFA and then labeled by esterification were found to have a significantly greater lipid peroxidation (i.e., loss of fluorescence) as compared to control cells

FIG. 4. Effect of incubation with increasing concentrations of d-αtocopherol: 0.0, 0.4, 0.8, or 1.6 mM for 12 h, on the loss of fluorescence in P388D1 cells, labeled by esterification with 20 µM cis-parinaric acid, 9, 11, 13, 15-cis-trans-trans-cis-octadecatetraenoic acid (cPnA) in 0.1% bovine serum albumin for 90 min. Cell concentration was 0.5 × 106 cells/mL in PBS. Lipid peroxidation was induced at time zero by addition of 100 µM FeCl2 · 4 H2O in equimolar EDTA. Values used are least square means ± SEM for three separate experiments. Curves assigned different superscript letters are significantly different based on linear analysis (P < 0.05). Quadratic analysis shows that the shapes of the curves from the 0.0 and 0.4 mM additions of d-α-tocopherol were significantly different from that of the 1.4 mM d-α-tocopherol, but not from the 0.8 mM d-α-tocopherol.

FIG. 3. (A) Effect of incubation with or without 5 µM arachidonic acid (AA) and 5 µM eicosapentaenoic acid on the spontaneous loss of fluorescent signal in P388D1 cells, labeled by esterification with 20 µM cPnA in 0.1% bovine serum albumin for 90 min. Cell concentration was 0.5 × 106 cells/ mL in phosphate-buffered saline. Values used are means ± SEM for three separate experiments. Curves assigned different superscript letters are significantly different based on linear and quadratic analysis (P < 0.05). (B) Effect of incubation with or without 5 µM arachidonic and 5 µM eicosapentaenoic acid and 250 µM d-α-tocopherol for 12 h on the loss of fluorescent signal in P388D1 cells, labeled by esterification with 20 µM cPnA in 0.1% bovine serum albumin for 90 min. Cell concentration was 0.5 × 106 cells/ mL in phosphatebuffered saline. Lipid peroxidation was induced at time zero by addition of 10 µM FeCl2 · 4 H2O in equimolar EDTA. Values used are means ± SEM for six separate experiments. Curves assigned different superscript letters are significantly different based on linear and quadratic analysis (P < 0.05); PUFA, polyunsaturated fatty acids. See Figure 1 for other abbreviations.

(Fig. 3A). When these cells were exposed to an additional oxidant stress, the PUFA-enriched cells lost significantly more fluorescence than did control cells. The addition of 250 µM α-tocopherol in the incubation media along with the longchain PUFA, prevented the increase in lipid peroxidation noted when cells were incubated with PUFA alone (Fig. 3B). Protection against lipid peroxidation by α-tocopherol. Cells were enriched with vitamin E by incubating them in complete media to which 250 µM α-tocopherol had been added. Control cells were incubated in complete media without the addition of α-tocopherol. After 12 h, control and αtocopherol-enriched cells were split and labeled by intercalation and by esterification, then subjected to different levels of oxidant stress. At either level of oxidative stress, loss of fluorescence in the cells labeled by intercalation was not significantly affected by overnight incubation of cells in tocopherolenriched media. In contrast, cells enriched with α-tocopherol and labeled by esterification of cPnA lost significantly less fluorescence than did control cells (P < 0.05, data not shown). The effect of increasing α-tocopherol concentration in the incubation media from 0, 0.4, 0.8, and 1.2 mM for 12 h on lipid peroxidation induced by 100 µM ferrous-EDTA is shown in Figure 4. Although cells labeled by intercalation showed no differences in lipid peroxidation due to increasing concentrations of vitamin E (data not shown), when cells were labeled by esterification, vitamin E significantly reduced the rate at which fluorescence was lost, in a dose-dependent fashion (P < 0.05). Lipids, Vol. 32, no. 2 (1997)

224

METHOD

DISCUSSION One of the overall goals of our laboratory is to understand the interaction between dietary PUFA, particularly fish oil, and vitamin E. Increases in dietary PUFA are rapidly reflected in the type of fatty acids found in the membrane. Since part of the suggested increase in dietary requirement of vitamin E is based on increased lipid peroxidation in membranes enriched with PUFA, it was particularly relevant for us to be able to examine lipid peroxidation at the level of the immune cell. Our initial objective was to use the published method of intercalating cPnA into intact immune cells to assess the effects of modulation of membrane components (i.e., PUFA and αtocopherol) on the susceptibility of living cells to lipid peroxidation. The intercalation of cPnA as a method for monitoring lipid peroxidation, originally published for use in liposomes (24), has been successfully adapted for use in erythrocyte ghosts (25,37,38), sarcoplasmic reticulum vesicles (39,40), intact erythrocytes (25,26,41,42), and Chinese hamster ovary cells (43). Tsuchiya and coworkers (44) were able to utilize cPnA and a model membrane system (i.e., liposomes) to characterize the antioxidant properties of α-tocopherol, βcarotene, and ubiquinol in response to a lipid-soluble, free radical initiation system. In an attempt to adapt the method to immune cells, we chose to work with the murine monocyte cell line P388D1, with which we have experience modulating the cellular content of both α-tocopherol and PUFA. We observed increased lipid peroxidation in response to increased levels of oxidant stress provided by chelated ferrous iron. When ferrous iron and EDTA are present in equimolar concentrations, lipid peroxidation is initiated by hydroxyl radicals. Cells labeled with cPnA by intercalation and esterification showed increased rate and total amount of peroxidation as the oxidant stress was increased from 10 to 100 µM Fe-EDTA. These data parallel the effect of in vivo administration of ferrous iron. For example, Hu and coworkers (7) found a positive correlation between the concentration of ferrous iron injected intraperitoneally into rats and indices of lipid peroxidation in various tissues such as serum, liver, and kidney. In comparing both methods of labeling our cells, we observed that the esterified cPnA label was slightly more sensitive to the same level of oxidant stress than cells labeled by intercalation. When corrected for basal loss of fluorescence, the 13 to 17% increased sensitivity observed in the esterified cells is reduced to a 5% difference. Although we have characterized the loss of cPnA fluorescence in untreated cells as photobleaching because it was not altered by exogenous vitamin E, it is also possible that this is a background oxidation. Greater oxidation of cPnA in intercalated cells where the label is closer to the surface of the cell and therefore the interface of cell to media may alter the kinetics of background loss of fluorescence due to lipid peroxidation not initiated by the experimenter. Since the goal of these experiments was to investigate whether we could show greater susceptibility of PUFA-enriched cells toward oxidation relative to cells with Lipids, Vol. 32, no. 2 (1997)

lower concentrations of PUFA, we chose not to pursue the question of differences between methods. Using the intercalation method, we were not able to detect increases in lipid peroxidation due to membrane PUFA enrichment. No significant differences in loss of basal fluorescence or loss of fluorescence in the presence of an exogenous oxidant stress were found in the cells PUFA-enriched vs. nonenriched cells labeled with cPnA by intercalation. Our data are in agreement with those of Van den Berg and coworkers (26) who reported that cPnA intercalated into intact erythrocytes was not able to detect increased lipid peroxidation associated with membrane PUFA enrichment. However, in that study erythrocytes from fish oil-fed rabbits lost significantly more total membrane PUFA as determined by lipid extraction and gas chromatographic analysis than those from control rabbits. To our knowledge, this is the first report of lipid peroxidation determined by cPnA that did not have a strong positive correlation with other methods of lipid peroxidation that were run simultaneously. In the direct measurement of lipid peroxidation of submitochondrial particles, de Hingh and coworkers (45) found excellent positive correlation between values determined by cPnA and oxygen consumption. We concluded that although intercalated cPnA works well as an indicator of lipid peroxidation in most cases, it does not reflect increased lipid peroxidation due to PUFA enrichment. Rather than abandoning the protocol, we examined the potential for cPnA incorporation into membrane phospholipids. We hypothesized that altering the membrane location such that cPnA was incorporated as a normal membrane component would allow greater sensitivity in detection of lipid peroxidation. Previously, Chinese hamster ovary cells have been reported to be capable of incorporating exogenously supplied cPnA into their membrane phospholipids. That cPnA was provided with a source of albumin such as fetal calf serum was essential (28). Sklar and Dratz (46) reported similar incorporation into bovine retinal rod outer segments. Harris and Stahl (47) suggested that the uptake of cPnA into phospholipids is an enzymatic process mediated by acyl-CoA acyltransferase and have shown its uptake is competitively inhibited by oleic acid. They further report that 71% of the esterified cPnA was found in phosphatidylcholine and 20% in phosphatidylethanolamine. Stuhne-Sekalec et al. (48), using guinea pig liver microsomes, have shown that cPnA is preferentially esterified at the sn-2 position although small quantities were found at the sn-1 position. Further, they showed that cPnA could be readily cleaved by phospholipase A2. Several authors have also shown that as a naturally occurring fatty acid, cPnA can be bound to cellular lipid transfer proteins such as fatty acid binding protein and sterol carrier protein (49) and can be esterified into cellular phospholipids (50). Prows and coworkers (52) have demonstrated that uptake of cPnA into the plasma membrane in the absence of serum is very rapid. Further, the total quantity of cPnA incorporated was significantly increased when fibroblasts were transfected with liver fatty acid binding protein. Although this technique provides a ready source of cPnA for incorporation into mem-

METHOD

branes, less than 3% of cPnA was esterified after a 30-min incubation. Our experience with P388D1 cells showed that they could rapidly esterify PUFA into membrane phospholipids when these fatty acids were introduced as fatty acid–albumin complexes. It must be noted, however, that only 0.5% of the total cPnA was esterified in 90 min in contrast to the rapid uptake of 2 µM cPnA when no albumin is present. Therefore, although cPnA can be esterified into mammalian cells, it must be emphasized that the total amount is very small. Also, the cPnA that is incorporated by esterification is distributed throughout all cellular membranes (43) rather than concentrated into the plasma membrane. The idea of esterifying the fluorescent probe to the membrane phospholipid was a particularly appealing one. Researchers have proposed that membrane α-tocopherol is physically associated with certain membrane phospholipids (52–56). Therefore, esterification of cPnA into the sn-2 position of the membrane phospholipid renders the cPnA in proper orientation for interaction with membrane α-tocopherol. Thus, this approach provides a physiologically relevant model for monitoring lipid peroxidation within cell membranes and the effect of alterations in membrane phospholipid PUFA and α-tocopherol on this process. We believe that cPnA orientation during intercalation may result in a lack of proximity among intercalated cPnA, membrane phospholipids, and vitamin E and that these factors are important determinants for success in examining differences in lipid peroxidation due to alterations of cell membrane components. In conclusion, we have shown that esterification of low concentration of cPnA (