Stabilization of Leukotriene A4 by Epithelial Fatty Acid-binding Protein in the Rat Basophilic Leukemia Cell*

Leukotriene A4 (LTA4) is a chemically unstable triene epoxide product of 5-lipoxygenase metabolism of arachidonic acid. Despite this chemical reactivity and its synthesis at the perinuclear membrane, LTA4 is enzymatically converted into the cysteinyl leukotrienes and leukotriene B4. Furthermore, LTA4 participates in transcellular biosynthesis and is thus transferred between cells as an intact molecule. A cytosolic fatty acid-binding protein present in the rat basophilic leukemia cells was identified using mass spectrometry. This protein was determined to be the stabilizing factor present in the cell cytosol responsible for increasing the effective chemical half-life of LTA4. Rat epithelial fatty acid-binding protein (E-FABP) was isolated using partial protein purification and immunoprecipitation. In-gel digestion with trypsin followed by peptide fingerprint analysis using matrix-assisted laser desorption ionization mass spectrometry and sequencing the major tryptic peptide obtained from liquid chromatography/mass spectrometry/mass spectrometry analysis identified E-FABP in the active fraction. Semi-quantitative Western blot analysis indicated that E-FABP in the cytosolic fraction of RBL-1 cells was present at ∼1–3 pmol/106 cells. E-FABP (9 μm) was tested for its ability to stabilize LTA4, and at 37 °C E-FABP was able to increase the half-life of LTA4 from the previously reported half-life less than 3 s to a half-life of ∼7 min. These results present a novel function for the well studied fatty acid-binding protein as a participant in leukotriene biosynthesis that permits LTA4 to be available for further enzymatic processing in various cellular regions.

Leukotrienes are a family of biologically active metabolites of arachidonic acid known to play important roles in multiple physiological and pathophysiological processes by acting as lipid mediators through specific G protein-coupled receptors (1,2). The biosynthesis of these 20 carbon fatty acids is regulated within cells that express 5-lipoxygenase, the first committed enzyme of the leukotriene cascade. The 5-lipoxygenase catalyzes two separate reactions; the first is abstraction of a hydrogen atom from carbon 7 of arachidonic acid and the insertion of molecular oxygen at carbon 5 to generate 5-(S)-hydroperoxyei-cosatetraenoic acid (HpETE). 1 The second reaction involves abstraction of a hydrogen atom from carbon 10 of HpETE followed by internal rearrangement of double bonds, loss of the hydroxyl group, and formation of the chemically reactive conjugated triene epoxide, leukotriene A 4 (LTA 4 ) (3,4). LTA 4 is a substrate for two enzymes that form the biologically active leukotrienes. LTA 4 hydrolase converts LTA 4 into leukotriene B 4 (LTB 4 ), a chemotactic factor for human neutrophils (5). Leukotriene C 4 synthase converts LTA 4 into the glutathione adduct leukotriene C 4 (LTC 4 ), which is a myotropic agent (6). Once LTA 4 is formed and released from the active site of 5-lipoxygenase, a competitive nonenzymatic reaction with water can also lead to the hydrolysis of the epoxide in a reaction in buffer measured to have a half-life of less than 30 s at 37°C (7) through formation of a carbocation intermediate (8). The products of this nonenzymatic hydrolysis reaction include biologically inactive but chemical stable isomers of LTB 4 including the ⌬ 6 -trans-5,12-dihydroxyeicosatetraenoic acid and 5,6-dihydroxyeicosatetraenoic acid (8).
In neutrophils, 5-lipoxygenase is found in the cytosol (9). However, in other cell types, including the rat basophilic leukemia (RBL) cell, 5-lipoxygenase is found in the nucleus (10). Upon stimulation of RBL cells, 5-lipoxygenase translocates to the nuclear envelope, where together with 5-lipoxygenase-activating protein (11) and arachidonic acid, which is released by cytosolic phospholipase A 2 (12), the biosynthesis of LTA 4 takes place.
Studies of the fate of LTA 4 produced within the human neutrophil revealed that greater than 50% of this lipid generated after cell activation is released to participate in the process of transcellular metabolism (13). Although many of the details of this process are unclear, it is now established that LTA 4 produced in the neutrophil can appear within other cell types that express synthetic enzymes for the biologically active leukotrienes. For example, erythrocytes, which express LTA 4 hydrolase but do not express 5-lipoxygenase, have been shown to convert LTA 4 into LTB 4 (14). Endothelial cells and platelets, which express LTC 4 synthase but have no 5-lipoxygenase, have been shown to convert LTA 4 derived from the neutrophil into LTC 4 (15,16). Because of the chemical instability of LTA 4 , it is clear that some mechanism must protect LTA 4 from exposure to water, preventing the nonenzymatic hydrolysis during transit between cells.
Both the chemical instability and the extent of transcellular * This work was supported by National Institutes of Health Grants HL25785 and DK53189. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
ʈ To whom correspondence should be addressed: National Jewish Medical and Research Center, 1400 Jackson St., Denver, CO 80206. Tel.: 303-398-1849; Fax: 303-398-1694; E-mail: murphyr@njc.org. 1 The abbreviations used are: HpETE, 5-(S)-hydroperoxyeicosatetraenoic acid; LTA 4 , leukotriene A 4 ; LTB 4 , leukotriene B 4 ; LTC 4 , leukotriene C 4 ; RBL, rat basophilic leukemia; FABP, fatty acid-binding protein; E-FABP, epithelial FABP; A-FABP, adipocyte FABP; HPLC, high pressure liquid chromatography; MALDI, matrix-assisted laser desorption ionization; MS, mass spectrometry; TOF, time-of-flight; LC, liquid chromatography. biosynthesis taking place in multi-cellular compartments suggest that LTA 4 is stabilized by binding to an intracellular protein that protects LTA 4 from water. Previous work has shown that serum albumin from various species can increase the half-life of LTA 4 to more than 20 min at concentrations that are found in plasma (7). However, intracellular proteins that function to protect LTA 4 from hydrolysis have not been identified (17). The purpose of our study was to critically test for the presence of LTA 4 stabilizing proteins within RBL cells and identify the stabilizing proteins. RBL-1 cells were chosen for these studies because this cell line expresses 5-lipoxygenase and therefore is likely to have a protein that functions in lipid stabilization (18). In addition, leukotriene biosynthesis has been extensively studied and characterized in these cells (19). Our findings now identify epithelial fatty acid-binding protein as an important molecule for stabilizing LTA 4 in these cells.

EXPERIMENTAL PROCEDURES
Materials-LTA 4 ethyl ester was a generous gift from Dr. Joseph Mancini at Merck-Frosst Canada (Pointe-Claire, Canada). All other eicosanoids were obtained from the Cayman Chemical Company (Ann Arbor, MI). LTA 4 -free acid was prepared as previously described (20). Anti-epithelial fatty acid-binding protein (anti-E-FABP) and anti-adipocyte fatty acid-binding protein (anti-A-FABP) antibodies were prepared as previously described (21). Nonspecific rabbit IgG and triethylamine were purchased from Aldrich. All other solvents and reagents were HPLC grade and were purchased from Fisher.
Cytosol Preparation-Rat basophilic leukemia cells (RBL-1) were cultured at the National Cell Culture Center (Minneapolis, MN). The cells were grown in suspension to a density between 0.85 ϫ 10 6 and 1.5 ϫ 10 6 cells/ml in Joklik's medium supplemented with 10% fetal bovine serum. The cells were harvested by centrifugation at 2500 ϫ g. The pellet was resuspended and washed twice with phosphate-buffered saline. The final cell pellet was snap frozen and stored at Ϫ70°C until use. Cytosol preparation was performed as previously described (22).
Protein Purification-Ammonium sulfate precipitation was performed as previously described (23) to remove any proteins insoluble at 30% saturation (w/v) or soluble above 75% saturation (w/v). The pellet from this centrifugation was resuspended in a total volume of 2 ml of 50 mM NaOAc, pH 5.0, and injected onto a Superose 12 prep grade column (16 ϫ 30 mm) (Amersham Biosciences) at a flow rate of 1 ml/min. A mixture of known proteins was chromatographed under the same conditions to create a molecular mass calibration curve. The active fractions from the size exclusion chromatography were pooled and exchanged into 25 mM Tris, pH 7.2, using a desalting column (Econo-Pac® 10DG disposable desalting column; Bio-Rad). These fractions were injected onto an anion exchange column (Econo-Pac® High Q, 5-ml bed volume; Bio-Rad) with a flow rate of 2 ml/min and a gradient from 0 to 1 M NaCl in 25 min. The active fraction in this purification step was concentrated by ultrafiltration with a YM1 membrane (Amicon stirred ultrafiltration cell; Millipore Corp., Bedford, MA). In some purifications, the pooled fractions from size exclusion were exchanged into 50 mM NaOAc, pH 5.0. Some of these samples were delipidated as previously described (24). The samples (delipidated or not) were applied to a cation exchange column (Econo-Pac® CM, 5-ml bed volume; Bio-Rad) with a flow rate of 2 ml/min 50 mM NaOAc, pH 5.0, and a gradient from 0 to 1 M NaCl in 25 min.
Western Blots-SDS-PAGE was performed as previously described with a 16% acrylamide gel and a 4% acrylamide stacking gel (25). The gel was then transferred to nitrocellulose, and a Western blot was performed according to the manufacturer's directions (Bio-Rad; horseradish peroxidase-conjugated antibody), using a 1:10,000 dilution of the anti-E-FABP antibody. The Western blot was quantified by densitometry using Alpha Ease on an Alpha Imager (Alpha Innotech Co, San Leandro, CA).
Immunoprecipitation-Immunoprecipitation of solutions containing LTA 4 stabilizing activity were performed by adding anti-E-FABP or anti-A-FABP antibodies or nonspecific rabbit IgG at a dilution of 1:1000. The antibody was allowed to incubate with the protein solution overnight at 4°C; then 50 l of a 50% slurry of Protein A Sepharose beads (Zymed Laboratories, Inc., South San Francisco, CA) was added; and the mixture was allowed to incubate further for 2 h at 4°C. The mixture was then briefly centrifuged at 12,000 ϫ g to remove antigenantibody complexes, and the supernatant was examined for LTA 4 stabilizing activity. The protein A beads containing immune complexes were boiled in SDS loading buffer for analysis by SDS-PAGE to correlate antigen depletion and loss of LTA 4 stabilizing activity. The proteins were visualized using Coomassie Blue G-250 (Bio-Rad).
LTA 4 Stabilization Assays-To determine chemical half-life, LTA 4 (500 ng) was added to 250 l of buffer or protein solution and incubated at 4°C. Aliquots (50 l) were removed at various time points between 2 and 30 min and added to ethanol (100 l) containing 100 ng of internal standard, which was either 15-oxo-ETE or 20-trifluoro LTA 4 . These samples were analyzed using either an in-line photo diode array or a triple quadrupole mass spectrometer (API-3000; PE-Sciex, Thornhill, Canada) as previously described (26). In either case, the log ratio (peak area of LTA 4 )/(peak area of its internal standard) versus time was plotted, and the half-life was calculated using the slope of the resulting line.
Assessment of the stabilization capacity (units of protein activity) was used to follow protein purification. For these assays, LTA 4 (100 ng) was added to either buffer or protein fraction (50 l). Each sample was allowed to incubate at 4°C for 20 min, and then ethanol (100 l) containing 15-oxo-ETE (100 ng) was added. The sample was brought to initial HPLC conditions by the addition of 150 l of 10 mM triethylamine, and the UV absorbance (280 nm) from LTA 4 and 15-oxo-ETE was determined at their corresponding retention times. A reversed phase XTerra MS column (2.1 ϫ 50 mm, 3.5 m C18; Waters Corporation, Milford, MA) was used at a flow rate of 200 l/min with a linear gradient using a mobile phase A consisting of 10 mM triethylamine at pH 11 and a mobile phase B consisting of acetonitrile:methanol (65:35 v/v) containing 10 mM triethylamine. The gradient started at 30% B for initial conditions and increased to 80% B in 5 min. Units of protein activity were defined as 10 times the ratio of milli absorbance unit LTA 4 /milli absorbance unit 15-oxo-ETE after the subtraction of the same ratio measured in buffer in the absence of protein. One complicating factor in these analyses was the lipophilicity of LTA 4 . Approximately 5% of the LTA 4 added to the protein or buffer solutions adhered to the wall of the polypropylene tube (data not shown); this added to the variability of the measurements of stabilizing units.
Protein Identification-In-gel tryptic digests were performed using a method previously described (27,28). Resulting peptide samples were analyzed by time-of-flight mass spectrometry after being spotted on a MALDI plate and cocrystallized with an equal volume of ␣-cyano-4hydroxycinnamic acid. This plate was then loaded into an Applied Biosystems Q-Star XL mass spectrometer (Applied Biosystems, Foster City, CA). The samples were analyzed using a N 2 laser intensity of 2.4 J and a pulser frequency of 4.993 kHz. The mass spectrometer was operated in enhance mode over the entire mass range (m/z 400 -3800). Autolytic trypsin peaks were used as internal standards for the samples; this resulted in mass accuracies of less than 10 ppm.
Some peptide samples were also injected onto a 75-m internal diameter column with a gravity pulled tip containing 7.5 cm of C18 packing (Jupiter 10 m, 300 Å; Phenomenex, Torrance, CA) and eluted using reversed phase solvents consisting of mobile phase solvent A, water containing 0.5% HOAc, and mobile phase solvent B, acetonitrile containing 0.5% HOAc. The gradient started at 5% B and ramped to 85% B in 40 min at a flow rate of 500 ml/min achieved with a passive split from pumps flowing at 0.5 ml/min. The effluent was introduced into a Finnigan LCQ Deca ion trap mass spectrometer (San Jose, CA) using a source voltage of 1.6 kV. The stainless steel capillary held a potential of 38 V, the heated capillary was set at 200°C, and the tube lens offset was Ϫ10 V. The MS/MS experiments were carried out with a scan speed of 3 s and a mass isolation width of 2 Da; the collision energy for these experiments was set at 35% relative collision energy, and the collision gas was helium. Data from both MALDI-TOF and LC/MS/MS experiments were entered into the Mascot search algorithm (www.matrixscience.com) (29).

RESULTS
Initial studies were designed to critically examine the hypothesis that cells involved in LTA 4 synthesis contain proteins that stabilize this leukotriene and protect it from spontaneous hydrolysis. Cytosol obtained from RBL-1 cells was incubated with LTA 4 , and the chemical half-life of LTA 4 in this preparation was compared with the chemical half-life in buffer. The experiments were performed at 4°C because the half-life of LTA 4 had been previously shown to be substantially longer at lower temperatures (17). As shown in Fig. 1A, RBL-1 cytosol was able to increase the half-life of LTA 4 to 7.5 Ϯ 0.67 min, a 5.4-fold increase over lysis buffer alone (t1 ⁄2 ϭ 1.4 Ϯ 0.27 min).

Standardized units of stabilizing activity (see "Experimental
Procedures") were used to monitor the protein preparations through subsequent purification steps. Ammonium sulfate precipitation of the cytosolic proteins demonstrated that the activity precipitated between 30 and 75% salt saturation. The number of total activity units found in cytosol decreased when the protein was subjected size exclusion chromatography as shown in Fig. 1 (B and C). However, the specific activity at this stage increased 3.6-fold following size exclusion chromatography. The active fraction eluted from the size exclusion column at a volume corresponding to a molecular mass of 10 -25 kDa (Fig.  1B). Without delipidation, the active cytosolic component eluted in the flow through from both cation and anion exchange columns (Fig. 1C and Table I) with up to a 23-fold increase in specific activity. When the protein preparations were delipidated prior to cation exchange, the active component bound to the column and eluted with 160 -200 mM sodium chloride (Table I) with essentially the same specific activity.
The molecular mass of the stabilizing protein together with its physiochemical behavior on ion exchange and behavior following delipidation were consistent with a known family of low molecular mass proteins known as the fatty acid-binding proteins (FABPs). To test whether FABPs accounted for the activity found in the anion exchange flow through, this fraction was subjected to an immunoprecipitation, and the supernatant from the immunoprecipitation was tested for any LTA 4 stabilizing activity (Fig. 2). Treatment of the anion exchange flow through with the antibody to E-FABP, followed by precipitation of the antibody complex with protein A-Sepharose, completely abolished the stabilizing activity that was previously found in the anion exchange flow through; the units per milliliter decreased from 56 Ϯ 5 to 1.9 Ϯ 2.5 units/ml. When the partially purified proteins present in the anion exchange flow through fraction were incubated with either nonspecific rabbit IgG or with an antibody to the A-FABP (21), the stabilizing activity remained in the supernatant following the precipitation.
The pelleted protein A beads from the immunoprecipitation were boiled with SDS loading buffer and separated by SDS-PAGE gel electrophoresis. This gel was stained (Fig. 3A), and regions covering the entire gel were analyzed by tryptic digestion followed by MALDI-TOF to determine those proteins bound to the E-FABP antibody. The tryptic peptides obtained from bands at 25 and 50 kDa (bands I and II) in lane 2 of the gel (Fig. 3A) were found to contain peptides derived from rabbit IgG. The only protein identified from the gel, other than rabbit antibody and low levels of keratin, was E-FABP (Band III). The MALDI-TOF mass spectrum of band III in Fig. 3A is shown in Fig. 3B. All peaks marked with asterisks in Fig. 3B were found to match peptides from E-FABP with an average error of 4 ppm. These peptides provided 69% sequence coverage of the protein. Other ions in the mass spectrum result from the autolytic trypsin fragments (m/z 842.49, 1045.57, and 2211.10, marked with T) and tryptic fragments of keratin (m/z 1307.67, 1475.76, and 1791.76).
A separate in-gel tryptic digest of the same gel region was analyzed by electrospray ionization LC/MS/MS, and the resultant spectrum of the HPLC that corresponded to the most abundant MALDI ion at m/z 927.56 was observed as a doubly charged ion (m/z 464.3). Collisional activation of this ion gen-erated a family of product ions consistent with specific peptide cleavages corresponding to y 1 -y 7 for a nonapeptide (Fig. 3C). The most abundant product ions were observed at m/z 529.4 (y 5 ) and m/z 359.2 (y 3 ) (Fig. 3C and Table II). The remaining fragment ion in the y ion series was not seen when the doubly charged peptide was collisionally activated, but this y 8 ion (m/z 798.5) was seen in the MS/MS spectrum of the singly charged peptide (data not shown) consistent for the tryptic fragment ELGVGLALR. The other major ions seen in the MALDI spectrum (Fig. 3B) were also observed by LC/MS/MS, and the abundant fragment ions from two of these peptides are listed in Table II. A semi-quantitative Western blot was used to estimate the amount of E-FABP present in RBL-1 cytosol (Fig. 4). Based upon densitometry, the cytosol contained ϳ0.4 ng of E-FABP/g of total protein corresponding to ϳ1-3 pmol/10 6 RBL-1 cells.
For stabilization of LTA 4 by E-FABP, the stabilizing properties of purified delipidated E-FABP were examined using LTA 4  Separate studies compared the ability of E-FABP to stabilize LTA 4 relative to bovine serum albumin at both 4 and 37°C (Fig. 6). The half-life of LTA 4 in buffer alone at 4°C was 0.6 Ϯ 0.01 min; at 37°C, the half-life of LTA 4 in buffer was too short to measure. As a negative protein control, ovalbumin was found to have no significant stabilizing activity (data not shown). When bovine serum albumin (10 M) was tested for its ability to stabilize LTA 4

DISCUSSION
Fatty acid-binding proteins are a family of low molecular mass proteins (ϳ15 kDa) found in the cytosol of most cells (30). The members of this family have between 20 and 70% sequence identity; however, they all share similar tertiary structures consisting of 10 antiparallel ␤-strands linked by hydrogen bonds to form a ␤-barrel (31). The fatty acid-binding protein family has been studied for its involvement in transport of long chain fatty acids as well as various hydrophobic ligands in a number of different tissues (32). This protein family has been implicated as a soluble carrier of insoluble lipids found in the cytoplasm. FABPs also have been proposed to facilitate transport of lipids into cells by sequestration of these hydrophobic a Specific activity units correspond to an assessment of LTA 4 remaining after 20 min in incubated buffer, normalized to the protein content as measured by the bicinchoninic acid assay. b ND, not detected. molecules in the cytosol. Even though the binding properties of these proteins with various ligands have been extensively examined in vitro, there is still some controversy as to their exact physiological role in vivo (33).
Initial studies with crude RBL-1 cell cytosol suggested the presence of a factor that was able to stabilize LTA 4 as measured by an increased half-life of the leukotriene in the cytosol, compared with buffer alone (Fig. 1A). The specific activity of this stabilization was increased by greater than 10-fold using a combination of ammonium sulfate precipitation, size exclusion, and anion chromatography and almost 20-fold using size exclusion chromatography in combination with cation exchange chromatography ( Fig. 1B and Table I). The physical and molecular properties of this factor including its molecular mass (10 -25 kDa) and retention upon cation exchange chromatography after delipidation suggested that it might be a member of the FABP family. Immunoprecipitation of partially purified RBL-1 cytosol (Fig. 2)  properties, which would suggest that either E-FABP was uniquely active in LTA 4 stabilization or more likely that little if any A-FABP is expressed in RBL-1 cells. Although the antibody to A-FABP cross-reacts with E-FABP protein only slightly (2%), the antibody to E-FABP does not show any measurable cross-reactivity with any other member of the FABP family. Analysis of the antibody-precipitated proteins removed from the purified fraction by the E-FABP antibody led to the identification of rat E-FABP based upon MS analysis of several tryptic peptides (Fig. 3).
Epithelial FABP (also called keratinocyte FABP) was first identified in murine skin carcinomas and is normally expressed to high levels in epithelial cells of the skin, tongue, lens, lung, and retina and to lesser levels in adipocytes, macrophages, and mammary tissue. Ligand binding studies have revealed high affinity for oleate and arachidonate but little or no affinity for prostaglandin E 2 (34). Palmitic acid binding to E-FABP has been structurally defined by a combination of x-ray crystallography and high field NMR (35,36). The region of the protein in proximity to the C-5 C-6 methylene groups of the bound fatty acid is structurally restricted and excludes crystallographically ordered water. Of particular interest is the side chain hydroxyl group of Tyr 22 that is 4.6 Å from the C-5 carbon. Assuming that the orientation of bound LTA 4 is similar to that of palmitic acid, the Tyr 22 -OH would be in close proximity of the bridging epoxide oxygen. Interestingly, Tyr 22 is structurally conserved in a number of FABPs, including A-FABP.
Recently, E-FABP was also found to covalently bind to 4-hydroxynonenal, suggesting this unique fatty acid-binding protein could play a role in lowering the concentration of some chemically reactive lipid intermediates (37). Here we present an additional novel function for E-FABP as a binding protein that can stabilize the epoxide intermediate, LTA 4 , by markedly reducing the nonenzymatic hydrolysis within the RBL cell. Involvement of E-FABP in the leukotriene biosynthetic pathway has not been previously described, but this protein could play a critically important role in the ultimate production of the biologically active leukotrienes (LTB 4

and LTC 4 ).
Biochemical studies of the effect of E-FABP on LTA 4 stability revealed a saturable process in prolonging the existence of the conjugated triene when present in aqueous solutions. The concentration of E-FABP estimated in the RBL-1 cells (1-3 pg/10 6 cells) should extend the half-life of this triene epoxide up to ϳ5 min at 37°C. Within such a time frame it would then be possible for LTA 4 to distribute throughout the RBL cell cytosol and reach membrane-bound LTC 4 synthase, as well as cytosolic LTA 4 hydrolase. Furthermore, this prolonged half-life would be consistent with the ability of LTA 4 to participate in transcellular biosynthetic reactions, which involve the transfer of intact LTA 4 from the cell of synthesis to a partner cell expressing LTA 4 hydrolase or LTC 4 synthase.
Previous studies have examined the binding of different eicosanoids with fatty acid-binding proteins. E-FABP has been shown to bind 5-lipoxygenase products (HpETE and 5-hydroxyeicosatetraenoic acid) with reasonably high affinity (34). Epoxyeicosatrienoic acids, products of cytochrome P-450 metabolism of arachidonic acid, have also been found to be ligands for other FABPs. Recent investigations have shown that epoxyeicosatrienoic acids can be bound to heart FABP and protect the epoxyeicosatrienoic acid from hydrolysis when soluble epoxide hydrolase is added to the buffer (38). Thus, there is a growing body of evidence to suggest that the fatty acid-binding proteins play an important role in eicosanoid biosynthesis and metabolism.
In summary, after purification and immunoprecipitation, a cytosolic protein present in the RBL-1 cell that increased the half-life of LTA 4 from less than 3 s in buffer at 37°C to greater than 7 min was unambiguously identified as the rat keratinocyte lipid-binding protein. Although the role of E-FABP in leukotriene biosynthesis has not been previously recognized, it is clear that it can play a central role in facilitating distribution of the reactive chemical intermediate, LTA 4 , into cellular compartments where subsequent chemical transformations take place to yield the biologically active leukotrienes, LTC 4 and LTB 4 . It is also possible that such fatty acidbinding proteins are also critically important in the process termed transcellular biosynthesis where LTA 4 is transferred from one cell to another.