Liver fatty acid-binding protein expression in transfected fibroblasts stimulates fatty acid uptake and metabolism

BB

etBiochi~ic~a BiophysicaA~ta

ELSEVIER Biochimica et BiophysicaActa 1301 (1996) 191-198

Liver fatty acid-binding protein expression in transfected fibroblasts stimulates fatty acid uptake and metabolism Eric J. Murphy a, Daniel R. Prows b, John R. Jefferson a,b,1, Friedhelm Schroeder

a,*

~ Department of Physiology and Pharmacology, Texas A and M University, TVMC, College Station, TX 77843-4466, USA b Division of Pharmacology and Medicinal Chemistry, Department of Pharmacology and Cell Biophysics, University of Cincinnati Medical Center, Cincinnati, OH 45267-0004, USA

Received 13 July 1995; revised 11 January 1996; accepted 16 January 1996

Abstract

The role of cytosolic liver fatty acid binding protein (L-FABP) in fatty acid uptake and metabolism was examined using cultured L-cell fibroblasts transfected with the cDNA encoding for L-FABP. [3H]Oleic acid was used to determine the effects of intracellular esterification on fatty acid uptake and to determine esterified fatty acid localization to specific lipid classes, cis-Parinaric acid, a poorly esterified fatty acid, was used to determine uptake in the absence of any appreciable esterification. High-expression L-cells had a 80% and 50% greater initial uptake rate for both [3H]oleic acid and cis-parinaric acid, respectively compared to low-expression L-cells. Maximal uptake of [3H]oleic acid did not plateau because of intracellular esterification. In high-expressing cells, maximal cis-parinaric acid uptake rapidly plateaued at a level 34% higher than in low-expression cells. After 1 min of incubation, the majority of cellular [ 3H]oleic acid was unesterified, with the bulk of the esterified portion preferentially localized to phospholipids. After 5 and 30 min, cells expressing L-FABP esterified a significantly greater amount of [3H]oleic acid into both the neutral lipid and phospholipid fractions than did low-expression cells. L-FABP expression also selectively stimulated [3H]oleic acid incorporation into choline glycerophospholipids. Thus, L-FABP expression not only stimulated fatty acid uptake at all time points, but also stimulated intracellular esterification into specific lipid pools. These results show in detail for the first time using an intact cell culture system that L-FABP expression not only stimulated fatty acid uptake, but also increased intracellular esterification of exogenously supplied fatty acids. Keywords: L-cell; Fatty acid binding protein; Oleic acid; cis-Parinaric acid; Fatty acid uptake and esterification; Fluorescence; (Liver)

1. Introduction

The physiological role of fatty acid binding proteins (FABP) in fatty acid uptake and metabolism is unclear (reviewed in [1-5]). FABP are a family of proteins with overlapping ligand specificity, considerable sequence homology, lack of cross-reactivity to respective antisera, and often unique tissue distribution. FABP bind fatty acids and

Abbreviations: L-FABP, liver fatty acid binding protein; cis-parinaric acid; 9Z,11E,13E,15Z-octatetraenoic acid; PBS, phosphate-buffered saline; CerPCho, sphingomyelin; PtdSer, phosphatidylserine; Ptdlns, phosphatidylinositol; ChoGpl, choline glycerophospholipid; EtnGpl, ethanolamine glycerophospholipid;lysoPtdEtn, lysophosphatidylethanolamine; lysoPtdCho, lysophosphatidylcholine. * Corresponding author. Fax: + 1 (409) 845655. i Present address: Department of Chemistry, Luther College, Decorah, IA 52101-1045, USA. 0005-2760/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0005-2760(96)00024-0

in some cases stimulate fatty acid transfer between membranes or enhance fatty acid esterification in vitro. On this basis, it is thought that FABP can act as intracellular fatty acid carriers, supply fatty acyl-CoA to lipogenic enzymes, and buffer the cell against the detergent actions of these ligands. However, ligand binding studies in vitro alone are not capable of elucidating physiological function [6-9]. There is little evidence that fatty acid binding proteins stimulate fatty acid uptake a n d / o r intracellular esterification in intact cells. Fatty acid binding protein expression could stimulate fatty acid uptake simply by providing additional binding sites for internalized fatty acids. For example, intestinal fatty acid binding protein (I-FABP) may act in this manner because it does not stimulate esterification of fatty acids to cholesterol esters or glycerides in vitro [10,11]. Liver fatty acid binding protein (L-FABP) expression stimulates fluorescent parinaric acid uptake, but the potential effect on intracellular esterifica-

192

E.J. Murphy et al. / Biochimica et Biophysica Acta 1301 (1996) 191-198

tion was not reported [5]. Because L-FABP stimulates [3H]oleoyl CoA incorporation into phosphatidic acid in vitro [10,12,13], it seems likely that cellular L-FABP expression may stimulate fatty acid esterification into phospholipids. In another study, adipocyte fatty acid binding protein expression in CHO cells enhanced [3H]oleic acid uptake and esterification into cellular lipids by 1.5- to 2.0-fold over non-transfected cells [14]. However, like other previous studies, the contribution of intracellular esterification to the overall uptake rate was not resolved. The experiments reported here were designed to address this issue. In the present study, L-cell fibroblasts that are transfected with the cDNA encoding for L-FABP [15] were used to determine the effect of L-FABP expression on fatty acid uptake, intracellular esterification, and directing fatty acids into specific lipid fractions in an intact cell system. We found that L-FABP expression not only increased fatty acid uptake, independent of esterification, but also stimulated a preferential increase in [3H]oleic acid esterification into distinct phospholipid fractions.

serum (Gibco, Grand Island, NY) as described earlier [17,15]. Clonal variation in the parent L-cell population was determined using six different subclones and mocktransfected cell lines. No differences were found between clones for endogenous FABP levels or on the rate or the amount of fatty acid taken up. 2.3. L-FABP Expression in transfected L-cells

2. Materials and methods

Recombinant rat liver L-FABP, was isolated [ 18,19] and was used to generate polyclonal antisera in rabbits [17,15]. Quantitative Western blot assays were performed on L-cell cytosol as described earlier [17,15]. Clones expressing 0.4% of cytosolic protein as L-FABP and clones expressing less than 0.008% of cytosolic protein as L-FABP were designated as high- and low-expression cells, respectively. It should be noted that L-cells contain an endogenous FABP that differs from L-FABP and does not cross react with antisera to L-FABP. The quantity of total FABP, the summation of endogenous FABP and L-FABP levels, was 0.50 + 0.04% and 0.11 + 0.03% of cytosolic protein in high- and low-expression clones, respectively. Thus, highexpression clones had 4.5-fold higher total fatty acid binding protein content.

2.1. Materials

2.4. Fluorescent fatty acid uptake assay

cis-Parinaric acid (MW 276 g/mol) was from Molecular Probes, Eugene, OR. Oleic acid was obtained from Sigma (St. Louis, MO). [3H]Oleic acid (10 Ci/mmol) was purchased from New England Nuclear (Boston, MA). Neutral lipid standards (triglyceride, diglyceride, monoglyceride, cholesteryl ester, and oleic acid) for high-performance liquid chromatography were obtained from NuChek Prep (Elysian, MN) and cholesterol was obtained from Steraloids (Wilton, NH). Phospholipid standards for highperformance liquid chromatography were obtained as follows: choline glycerophospholipids and ethanolamine glycerophospholipids from Serdary (London, Ont., Canada); sphingomyelin, phosphatidylserine, and cardiolipin from Avanti Polar Lipids (Birmingham, AL); and phosphatidylglycerol, phosphatidylinositol, and phosphatidylinositol-4phosphate from Sigma (St. Louis, MO). Lysophosphatidylethanolamine and lysophosphatidylcholine were prepared from brain ethanolamine plasmalogen and heart choline plasmalogen (Serdary, London, Ont., Canada) by acid hydrolysis yielding the respective lysophospholipid. All high-performance liquid chromatography solvents were purchased from E.M. Science (Cherry Hill, N J). All other chemicals used were of reagent grade or better.

Cells were harvested with a rubber policeman, sedimented, and washed two times with phosphate-buffered saline, (PBS, 8 mM sodium phosphate, 130 mM NaC1, 2.7 mM KC1, pH 7.3), and resuspended in PBS at a concentration of 3 × 105 cells/ml. Fatty acid uptake was continuously monitored by rapidly adding cis-parinaric acid (0.5 ug/ml, 1.8 /zM) to a continuously stirred (Bel-Art Cell Spinbar, Fisher Scientific, Pittsburgh, PA) suspension of cells (3 × 105cells/ml). This concentration of cis-parinaric acid was in the linear portion of the uptake curve when either initial rate or maximal uptake were plotted versus cis-parinaric acid concentration. The final ethanol concentration was maintained below 0.5%. Because cis-parinaric acid is only fluorescent in a hydrophobic environment, this assay permits sensitive, continuous measurement of fatty acid uptake without separating the free fatty acid in solution from that taken up by the cells. Thus, an increase in fluorescence represents increased cis-parinaric acid uptake. This change was monitored using a SLM 4800 Fluorimeter (SLM/Aminco, Urbana, IL) in T-format. Excitation was at 324 nm with a 450 watt Xe-Arc lamp as a light source. Fluorescence emission was measured through a GG-375 sharp cut-off filter (Janos Technology, Townshend, VT) in order to eliminate highly polarized scattered light. Absorbance was maintained below 0.15 in order to eliminate the inner filter effect. Fluorescence data were collected continuously at a rate of 50 points/min with the aid of a modified version of the SLM 4800 data collection program.

2.2. Cells

Both high- and low-expression cells were grown on 75 cm 2 plastic tissue culture dishes (Coming Corp., Coming, NY) with Higuchi medium [16] containing 10% fetal calf

E.J. Murphyet aL/ Biochimicaet BiophysicaActa 1301 (1996) 191-198 2.5. Fat~ acid esterification High- and low-expression cell lines were grown to confluency and used to measure fatty acid uptake and esterification by incubating with either cis-parinaric acid or [3H]oleic acid. Prior to incubation with fatty acids, the medium was removed, the dishes rinsed with an equal volume of phosphate-buffered saline (PBS), and then incubated with 30 ml of buffer. To this buffer, 150/zl of either fatty acid (0.1 /zg//zl ethanol) was added to the cells to give a final fatty acid concentration of 1.8 /zM and incubated at 37°C. The [3H]oleic acid added was diluted from a stock solution with a specific activity of 10 Ci/mmol using cold oleic acid to give a final specific activity of 1 /zCi/nmol.

2.6. Cell lipid extraction Immediately following incubation, extracellular cisparinaric or [3H]oleic acid was removed by washing the cells two times with PBS. Further washes recovered insignificant additional free fatty acid. After the second PBS wash was removed, the culture dishes were immediately floated on liquid nitrogen and frozen to minimize acylhydrolase activity [20]. The cellular lipids were extracted directly from the frozen monolayer of cells using a modified n-hexane/2propanol (3:2, v / v ) extraction [21]. This method quantitatively extracts both neutral lipids and phospholipids with decreased protein levels in the lipid extract. Cellular lipids were extracted by adding 4 ml of 2-propanol to the frozen monolayer and the cells were removed from the substratum by scraping with a Teflon cell scraper. The 2-propanol containing the cells was transferred directly to 9 ml of n-hexane. The cell plate was washed with another 2 ml aliquot of 2-propanol to remove any remaining cells and this wash was added to the n-hexane for a final ratio of n-hexane/2-propanol (3:2, v / v ) . The cell extract was centrifuged at 2500 rpm in an HN-S table top centrifuge (International Equipment Company, Needham, MA) to pellet the cellular protein. The solvent containing the lipids was decanted and the protein saved for quantitation [22] following digestion in 0.2 M KOH at 65°C overnight. The lipid extract was filtered through a Rainin (Woburn, MA) 0.2 /zm Nylon filter prior to high-performance liquid chromatography (HPLC). The sample volume was reduced by N 2 evaporation and redissolved in 200 /zl of HPLC grade n-hexane/2-propanol (3:2, v / v ) . The neutral lipid and phospholipid fractions were separated by HPLC and [3H]oleic acid incorporation determined by liquid scintillation counting using Scintiverse E scintillation cocktail (Fisher, Pittsburgh, PA) and an LS-7000 liquid scintillation counter (Beckman, Fullerton, CA). cis-Parinaric acid incorporation was measured using a Shimadzu RS-535 fluorescence detector (Kyoto, Japan).

193

2.7. High-performance liquid chromatography of L-cell phospholipids The cell lipid extracts were separated by HPLC into all major phospholipid classes including: phosphatidylserine (PtdSer), phosphatidylinositol (Ptdlns), and lysophosphatidylethanolamine (lysoPtdEtn) [23]. The lower limit of detection for optimal resolution is 100 nmol of total lipid phosphorus per injection. Phospholipid class elution was monitored by absorption at 205 nm and phospholipid classes were quantified by measuring lipid phosphorus [24]. Elution order and retention times were confirmed by use of commercially purchased standards except for lysophospholipids which were produced by acidic hydrolysis of the alkenyl ether linkage resulting in 1-1ysophospholipids [25]. The HPLC system consisted of two Altex (Berkeley, CA) 100A pumps, an Altex 420/421 controller, and an Altex model 210 injection port. A Dupont (Wilmington, DE) Zorbax silica column (4.5 mm X 250 mm, 5 - 6 /zm) was used in the separation and maintained at 34°C with a Jones Chromatography (Littleton, CA) heating block. An ISCO (Lincoln, NE) V4 ultraviolet variable wavelength detector was used to monitor absorption at 205 nm. All solvents were HPLC grade and purchased from E.M. Science (Cherry Hill, NJ). Solvent A was n-hexane/2-propanol (3:2, v / v ) and solvent B was n-hexane/2-propanol/ water (56.7:37.8:5.5, v/v). Initial solvent proportions were 65% A / 3 5 % B with a flow rate of 1.5 ml/min. A step gradient was used to separate the phospholipids, with proportions of solvent B increasing until 100% B was reached. For [3H]oleic acid containing samples individual phospholipid classes were collected into 20 ml vials, evaporated in an oven, and 20 ml of Scintiverse E was added to each sample. Samples were counted by liquid scintillation counting with an efficiency of 56%. This high efficiency can be attributed to the low amount of chemical and physical quench in the samples. Disintegrations per minute (dpm) were calculated for each phospholipid class and specific activity calculated based on the number of nmoles for each phospholipid class. To calculate the amount of cis-parinaric acid incorporated into the phospholipid fraction, a Shimadzu RS-535 fluorescent detector equipped with a 150 watt Xenon arc and 10 nm band widths was used in line with the ultraviolet absorption detector. The wavelengths for excitation and emission were 310 nm and 420 nm, respectively. Fluorescence was independent of changing solvent proportions during the phospholipid separation.

2.8. High-performance liquid chromatography of L-cell neutral lipids The neutral lipid fraction from the phospholipid separation was saved and separated using an isocratic HPLC

194

E.J. Murphy et al. / Biochimica et Biophysica Acta 1301 (1996) 191-198

method that resolved all the major neutral lipids including: triglycerides, 1,3-diacylglycerol, 1,2-diacylglycerol, free fatty acids, cholesteryl esters, and cholesterol. Peaks were identified using commercially prepared standards. Peak area for each fraction was calculated using a Nelson Analytical (Cupertino, CA) 760 series intelligent analog to digital interface box followed by peak area integration using Nelson Analytical 2600 software. The HPLC system consisted of an Altex 420/421 controller, two Beckman l14M pumps, a Beckman 210A injector, and ISCO V4 variable wavelength detector, a Shimadzu RS-535 fluorescence detector, a Jones Chromatography heating block and a Dupont Zorbax silica column (4.6 mm × 250 mm, 5 - 6

~m). This separation used a two solvent system. Solvent A was 1.2% 2-propanol in hexane containing 0.01% acetic acid and Solvent B was n-hexane. The separation conditions were as follows: solvent proportions 90% A / 1 0 % B; column temperature 55°C; and a flow rate of 0.6 ml/min. The ultraviolet and fluorescence detectors were set up in-line to continuously monitor the eluent. Each neutral lipid class eluting from the column was collected separately (except the 1,2- and 1,3-diacylglycerols which were combined) in a 20 ml scintillation vial, solvent was removed by evaporation, and 20 ml of Scintiverse E scintillation cocktail was added. A 56% counting efficiency was used to calculate dpm. Neutral lipid class radioactivity was normalized to either cholesterol (mg) or total lipid phosphorus (nmol). Both cholesterol and cis-parinaric acid concentrations were calculated based on peak areas. Calculations were done using a standard curve generated by running authentic cholesterol and cis-parinaric acid standards. There was a wide detectable range for both lipids.

2.9. Statistical analysis

Table 1 Esterified, unesterified, and total [3H]oleic acid levels in cells a [3H]Oleic acid taken up ( n m o l / m g ) protein

Time (min)

low-expression

high-expression

Total [3H]oleic acid 1

8.6_+ 1.8

15.3_+ 1.9 *

5 30

15.2__ 1.5 101.5 _ 10.7

26.0_+3.3 * 180.4 _+ 12:8 *

Unesterified [ ~H]oleic acid 1 5 30

4.6+_0.9 3.5_+0.3 10.1 +_ 1.1

10.9+_ 1.3 * 6.8_+0.9 * 18.0_+ 1.3 *

Esterified: neutral lipids 1 5 30

0.9+_0.1 2.3+_0.2 13.2+_ 1.4

0.5_+0.1 * 3.1 +_0.4 25.3+0.5 *

Esterified: phospholipids 1 5 30

3.2+_0.7 + 9.4+_0.9 ÷ 78.2+_8.2 ÷

4.0+_0.5 + 16.1+_2.0 *,+ 137.1 +_9.7 * ,+

a Values represent the mean + S.E.M., n = 3-9. * P < 0.05 vs. low expression. + P < 0.01 vs. esterified neutral lipids.

cence intensity reaching a plateau after 1 minute as shown in representative spectra (Fig. 1). This fluorescence intensity remained constant for 30 min (not shown). L-FABP expression significantly increased the initial rate of cisparinaric acid uptake by 50% from 230 AF/min to 346 AF/rffln ( P < 0.025). Maximal cis-parinaric acid uptake was also increased by 30% ( P < 0.025) in the high-expression cells compared to low-expression cells. Thus, the increase in the initial rate and maximal uptake of cisparinaric acid in high-expression cells was consistent with the increase observed after 1 rain of [3H]oleic acid uptake.

Values represent the m e a n _ S.E.M. Statistical significance was evaluated with Students t-test. 120 d ,,

3. Results

~ .',"" .'.'*.....,.........

High

Expcession

,....,.*-..'..

~',,.~..'.......~o

0

90.

Low

0 .,...,..,

3.1. Fatty acid uptake into L-cells The effect of L-FABP expression on fatty acid uptake was determined using [3H]oleic acid. In high-expression cells, total [3H]oleic acid levels were significantly increased compared to low-expression cells (Table 1). Highexpression cells had a 1.8-, 1.7-, and 1.8-fold increase in [3H]oleic acid uptake compared to low-expression cells after 1, 5, and 30 min of incubation, respectively. cis-Parinaric acid was also used to measure fatty acid uptake. In high- and low-expression cells, cis-parinaric acid uptake was rapid and monophasic with the fluores-

,., , , . , . . . . . . . - . ,

Expression

. . . . . . . . . . . . . . . . . . , . ,. . . . . . . . . . . . .

0

,'7

60 -

"e er

305511r **.*

| 75

| 150

! 225

T i m e Isecl

Fig. 1. A representative cis-parinaric acid uptake curve for high and low expression mouse L-cell fibroblasts, cis-Parinaric acid uptake was determined as described in Section 2,

E.J. Murphy et al./ Biochimica et BiophysicaActa 1301 (1996) 191-198 3.2. Effect of esterification on fatty acid uptake in L-cells

had 71 + 7% of the [3H]oleic acid unesterified compared to 53 + 3% in low-expression cells. Thus, although there was an L-FABP induced increase in fatty acid uptake after 1 min, there was a much smaller increase in esterification of [3H]oleic acid at this same time point, such a difference may reflect an increase in intracellular fatty acid binding sites in high-expression cells. At later incubation times, the early distributional differences in [3H]oleic acid were not evident, although there was an approximate 1.8-fold increase in [3H]oleic acid levels in all lipid fraction (Table 1). Apparently, after 5 and 30 min of incubation, sufficient [3H]oleic acid was taken up to saturate the fatty acid binding sites, thereby increasing the amount of free fatty acid to be diverted into esterification pathways. Collectively, these data suggest that L-FABP expression increases cellular fatty acid uptake and esterification, and may result in an increase in the number of intracellular fatty acid binding sites. Once these sites are sufficently occupied, fatty acid is then available for esterification.

To determine the contribution of esterification to the increase in fatty acid uptake, cis-parinaric acid, a poorly esterified fatty acid, was used. Both [3H]oleic acid (Table 1) and cis-parinaric acid (Fig. 1) were rapidly taken up. In contrast to the 1.3-fold increase in cis-parinaric acid maximal uptake, [3H]oleic acid maximal uptake was increased 1.8-fold in high expression compared to low-expression cells. In order to determine if this difference in maximal fatty acid uptake may be the result of intracellular esterification, the incorporation of both fatty acids into glycerolipids was determined. [3H]Oleic acid was extensively incorporated into phospholipids as well as neutral glycerides (Table 1), but cis-parinaric acid primarily appeared as the unesterified fatty acid with less than 3% esterified into cellular lipids even after 30 min uptake. These results indicate that the increase in cis-parinaric acid uptake in high-expression cells is representative of fatty acid uptake in the absence of appreciable esterification. Hence, part of the increase in [3H]oleic acid uptake could be accounted for by [3H]oleic acid esterification into the phospholipid and neutral lipid pools.

3.4. Neutral lipid [ 3H]oleic acid distribution To further resolve whether L-FABP may target [ 3H]oleic acid esterification into specific neutral lipids, the neutral lipid fraction was separated into triglyceride, diglyceride, and cholesteryl ester fractions (Table 2). [3H]Oleic acid percent distribution was calculated by dividing the radioactivity of each individual neutral lipid class by the total radioactivity of the neutral lipid fraction. After 1 min of incubation, the proportions of [3H]oleic acid esterified in triglycerides and cholesteryl esters was decreased 4 - 5 - f o l d in high-expression cells compared to low-expression cells. There were no differences in diglyceride [3H]oleic acid proportions. As fatty acid binding sites became saturated with increasing time, [3H]oleic acid was in excess and available for esterification into the neutral lipid fractions in high-expression cells. This was observed after 5 min, when no significant differences in radiolabel distribution were

3.3. Increased intracellular fatty acid binding sites In both high- and low-expression cells, [3H]oleic acid was initially found in the free fatty acid pool. At early time points, the [3H]oleic acid was primarily esterified into the phospholipids, but by 30 min the [3H]oleic acid was evenly distributed in esterified form among both the neutral and phospholipid classes (Table 1). L-FABP expression increased free [3H]oleic acid levels by 2.4-, 1.9-, and 1.8-fold after 1, 5, and 30 min of incubation, respectively (Table 1). The 2.4-fold increase in free fatty acid levels seen after 1 min of incubation, corresponds with a shift in the distribution of label into the free fatty acid pool in high-expression cells. After 1 rain, high-expression cells

Table 2 [3H]Oleic acid distribution in the free fatty acid and esterified neutral lipids Neutral lipid species (% distribution) Time (rain)

195

a

free fatty acids

triglycerides

diglycerides

Low expression 1 5 30

85.0 _+2.3 61.0 _+4.0 43.5 + 7.8

9.1 _+ 1.7 26.0 _+0.6 53.2 _+7.0

1.4 _+0.3 2.6 + 0.6 2.0 _+0.3

4.5 _+ 1.5 8.7 + 3.4 1.2 _+ 1.0

High expression 1 5 30

95.7-t-3.1 * 69.4 _+3.8 39.4_+3.1

2.2-1-0.5 * 24.1 -+ 5.8 47.1 _+5.8

1.4_+0.1 1.8 _+0.2 1.8_+0.6

0.6_+0.1 * 4.7 _+2.0 11.7_+3.6 *

a Values represent the mean + S.E.M., n = 3-10. * P < 0.05 vs. low expression.

cholesteryl esters

E.J. Murphy et al./Biochimica et Biophysica Acta 1301 (1996) 191-198

196

Table 3 [3H]Oleic acid relative specific activity in phospholipid fractions a Time (min)

Phospholipid class ( d p m / / z m o l phospholipid class) ChoGpl

EtnGpl

Ptdlns

PtdSer

CerPCho

lysoPtdEtn

lysoPtdCho

273 + 61 960 + 68 8891 + 335

82 + 13 373 + 36 3287 + 114

490 + 69 1018 + 33 6930 + 222

345 + 36 471 + 45 1162 + 167

235 + 3l 545 + 69 1326 + 180

114 + 14 382 + 10 3377 + 61

1033 + 76 4374 + 318 4407 + 310

351 + 52 13,068 + 137 * 12,210 + 642 *

86 _+ 11 424 + 55 3736 +__353

346 + 42 877 _ 124 7969 + 669

146 + 21 * 209 + 41 * 7761 + 123 *

211 + 44 519 _+ 106 1878 + 80 *

189 __+19 477 + 45 4548 + 751

1293 + 97 3425 + 523 4092 + 487

Low expression 1 5 30

High expression 1 5 30

a Values iepresent the mean :1: S.E.M., n = 3-10. * P < 0.05 vs. low expression.

found between high- and low-expression cells (Table 2). However, by 30 min the proportion of [3H]oleic acid esterified into the cholesteryl ester fraction was 11-fold higher in high-expression than low-expression ceils. Hence, at longer times L-FABP expression stimulated an increase in fatty acid esterification into cholesteryl esters, while esterification into other neutral lipid fractions was not enhanced.

3.5. Phospholipid [ 3H]oleic acid distribution and specific activity Potential L-FABP mediated targeting of [3H]oleic acid to specific phospholipid classes was determined. The specific activity of [3H]oleic acid in each phospholipid class was calculated. As expected, very little differences in [3H]phospholipid specific activity were noted between high- and low-expression cells after 1 min of [3H]oleic acid uptake (Table 3). However, 3H specific activity in choline glycerophospholipids, phosphatidylserine, and sphingomyelin was increased in high-expression cells, especially after 5 or 30 min of incubation (Table 3). These changes in specific activity were up to 13-fold, depending on the time point and specific phospholipid fraction examined. In contrast, total cellular fatty acid uptake was stimulated much less, 1.5- to 1.8-fold (Table 1). The larger differences in phospholipid class specific activity may be the result of increased acyl chain turnover in specific phospholipid classes and/or may represent L-FABP dependent stimulation of specific phospholipid class synthesis. To determine whether L-FABP expression stimulated phospholipid synthesis, phospholipid levels (nmol phospholipid/mg protein) were determined for both highand low-expression cells (Table 4). These data show that choline glycerophospholipids, ethanolamine glycerophospholipids, and sphingomyelin levels were significantly increased by 31%, 81%, and 94%, respectively, in high-expression cells compared to low-expression cells (Table 4).

These increases in certain phospholipid class levels (nmol phospholipid/mg protein) may account for the increase in fatty acid esterification into phospholipids that was observed in high-expression cells. However, because phosphatidylserine levels (nmol phospholipid/mg protein) were not increased, the higher [3H]oleic acid specific activity observed in the phosphatidylserine fraction (Table 3) must result from enhanced acyl group turnover rather than increased phosphatidylserine levels (nmol phospholipid/mg protein). Conversely, because ethanolamine glycerophospholipid specific activity was unaffected by L-FABP expression (Table 3), then [3H]oleic acid esterification into this class must be decreased in spite of an overall increase in levels (nmol phospholipid/mg protein). After 30 min, choline glycerophospholipids specific activity was increased to the same extent as the increased choline glycerophospholipid levels in the high-expression cells (Table 4). This is consistent with an L-FABP induced increase in choline glycerophospholipid synthesis, not an increase in acyl chain turnover in high-expression cells. In short, L-FABP expression modulated acyl chain turnover and/or increased synthesis of specific phospholipid classes in high-expression cells. The net result was an approximate 40% increase in total cellular phospholipid levels (nmol phospholipid/mg protein). Table 4 Phospholipid levels in high- and low-expression L-cells a Phospholipid fraction

Low expression ( n m o l / m g protein)

High expression ( n m o l / m g protein)

ChoGpl EtnGpl Ptdlns PtdSer CerPCho lysoPtdEtn lysoPtdCho

61.4 -t- 0.9 23.4 _+0.7 5.4 + 0.2 1.9+0.4 12.7 _ 0.7 2.6 + 0.5 2.1 + 0.7

80.8 + 42.4 + 9.0 + 5.5 + 24.4 + 2.7 + 3.9 +

Values represent the m e a n + S.E.M., n = 4. * P < 0.01 by Student's t-test.

a

3.3 * 1.5 * 2.7 1.9 1.5 * 0.3 0.2

E.J. Murphy et al. / Biochimica et Biophysica Acta 1301 (1996) 191-198

4. Discussion

The importance of L-FABP expression to cellular fatty acid uptake and metabolism is unknown in intact cells. To address this issue, high-expression and low-expression LFABP clones were used. The effects of L-FABP expression on fatty acid uptake and esterification were studied using [3H]oleic acid, while cis-parinaric acid was used to determine uptake rates in the absence of any appreciable esterification. The results show that L-FABP expression dramatically modified three aspects of lipid metabolism: fatty acid uptake, fatty acid esterification, and synthesis of certain phospholipid classes. L-FABP expression increased the initial rate of cellular fatty acid uptake. The initial cis-parinaric acid uptake rate, 7.0 nmol/mg protein X min, did not significantly differ from that for [3H]oleic acid, 5.8 nmol/mg protein X min. The initial cis-parinaric acid uptake rate was 1.5-fold higher in high expression compared to low-expression cells. Because of intracellular esterification, the initial oleic acid uptake rate was slightly more enhanced than that for cis-parinaric acid and was elevated 1.8-fold in high-expression compared to low-expression L-cells (Table 1). These results from an intact cell system are supported by numerous studies in vitro, that indicate FABP stimulate fatty acid desorption from membranes, membrane fatty acid loading, and membrane fatty acid transfer (reviewed in [3,4,26]). The magnitude of the effect of L-FABP expression on fatty acid uptake was similar to that reported for adipocyte lipid binding protein expression in transfected CHO cells, where uptake was increased 1.5- to 2-fold [14]. Finally, as shown by cis-parinaric acid maximal binding and increased unesterified [3H]oleic acid levels in high-expression cells, L-FABP expression not only increased the initial rate of fatty acid uptake, but also appears to increase the cellular fatty acid binding capacity. A physiological role for L-FABP in fatty acid esterification has not been established in intact cells. The data with transfected L-cells indicate that after short incubation times (1 min), L-FABP expression inhibited intracellular fatty acid esterification into the neutral lipids. After 5 rain of incubation, L-FABP expression resulted in an 1.3- to 1.9-fold increase in [3H]oleic acid esterification into the neutral lipids and a 1.7-fold increase into the phospholipids. Immunolabeling data indicate that a significant proportion of FABP may be near or associated with microsomes and mitochondria, organelles important to fatty acid esterification [12,27,28]. Whether this localization results in inhibition or stimulation of fatty acid esterification in intact cells is unclear. However, results obtained from experiments in vitro are contradictory. Several investigators reported inhibitory effects of FABP in vitro on steps involved in fatty acid activation to fatty acyl-CoA; namely, the inhibition of microsomal acyl-CoA synthase [29,30] and mitochondrial long-chain acyl-CoA synthase [31 ]. In contrast, others have

197

reported that FABP stimulate microsomal enzymes in vitro that are involved in the synthesis of triglycerides and phospholipids, including acyl-CoA:glycerol-3-phosphate acyltransferase [10,32,33], diacylglycerol acyltransferase [34,35], phosphatidate phosphohydrolase [36] and acyl CoA:cholesterol acyltransferase [11]. L-FABP, but not intestinal FABP, stimulate microsomal acyl CoA:cholesterol acyltransferase [11]. This enhanced stimulation may account for the increased [3H]oleic acid esterification into cholesteryl esters in high-expression cells. Furthermore, when L-FABP expressing cells are stimulated to produce cholesteryl esters by treating the cells with sphingomyelinase, significantly more cholesteryl esters are formed in high-expression compared to low-expression cells [17]. Consequently, L-FABP expression results in elevated cholesteryl ester levels [15]. Collectively, these results implicate a role for L-FABP in modulating both cholesteryl ester levels as well as fatty acid esterification to cholesterol. L-FABP expression stimulated phospholipid synthesis as indicated by increased phospholipid levels. L-FABP expression selectively increased the choline glycerophospholipid, sphingomyelin, and ethanolamine glycerophospholipid levels. In addition, the specific activity data were consistent with L-FABP expression enhancing the acyl chain turnover in phosphatidylserine. These results suggest that L-FABP expression establishes a new level of phospholipid synthetic homeostasis as is reflected in the increased phospholipid levels. Previously, L-FABP expression did not alter plasma membrane phospholipid composition [37], however, this does not imply that there is not an increase in membrane phospholipid levels [38]. L-FABP expression increased total cellular phospholipid content by 46% in these same L-cell clones [15]. The results presented herein confirm that L-FABP expression resulted in a dramatic 40% increase in phospholipid levels. L-FABP expression also provided substrate for increased esterification into specific phospholipid classes. This apparent dualistic role indicates the possible regulation of multiple events in phospholipid synthesis by L-FABP. In summary, transfected L-cells provide an excellent model for examining the role of L-FABP in fatty acid uptake and intracellular esterification in intact cells. The use of both [3H]oleic acid and cis-parinaric acid permits differentiation between these two related events. These data presented herein show for the first time in detail the relationship of L-FABP expression to specific fatty acid uptake, trafficking, and esterification into specific phospholipid classes. These results support our preliminary study that showed L-FABP, but not I-FABP expression increased fatty acid uptake into transfected L-cells [39]. Thus, L-FABP expression not only stimulated esterification of [3H]oleic acid, but also directed [3H]oleic acid primarily into phospholipid rather than neutral lipid synthetic pathways in the time frame of our experiments. The mechanisms whereby L-FABP stimulates [3H]oleic acid

198

E.J. Murphy et al. / Biochimica et Biophysica Acta 1301 (1996) 191-198

esterification into specific phospholipid classes is unknown, but is the subject of future studies.

Acknowledgements This work was supported in part by a grant from the USPHS [DK41402]. JRJ was supported in part by a fellowship from the American Heart Association-Ohio Affiliate. The generous assistance of Dr. Lloyd Horrocks, The Ohio State University, in providing access to the HPLC systems described herein was much appreciated.

References [1] Dempsey, M.E. (1984) Curt. Top. Cell. Regul. 24, 63-86. [2] Bass, N.M. (1988) Int. Rev. Cytol. 3, 143-184. [3] Spener, F., Borchers, T. and Mukherjea, M. (1989) FEBS Lett. 244, 1-5. [4] Paulussen, R.J.A. and Veerkamp, J.H. (1990) in Subcellular Chemistry (Hilderson, H.J., ed.), pp. 175-226, Plenum Press, New York. [5] Schroeder, F., Jefferson, J.R., Powell, D., Incerpi, S., Woodford, J.K., Colles, S.M., Myers-Payne, S., Emge, T., Hubbell, T., Moncecchi, D., Prows, D.R. and Heyliger, C.E. (1993) Mol. Cell. Biochern. 101,73-83. [6] Sziegoleit, A. (1982) Biochem. J. 207, 573-582. [7] Habig, W.H., Pabst, M.J., Fleischner, G., Gatmaitan, Z., Arias, I.M. and Jakoby, W.B. (1974) Proc. Natl. Acad. Sci. USA 71, 3879-3882. [8] Ho, M.-T.P., Massey, J.B., Pownall, H.J., Anderson, R.E. and Hollyfield, J.G. (1989) J. Biol. Chem. 264, 928-935. [9] Van Bennekum, A.M., Blaner, W.S., Seifert-Bock, I., Moukides, M., Brouwer, A. and Hendriks, H.F.J. (1993) Biochemistry 32, 1727-1733. [10] Hubbell, T., Behnke, W.D., Woodford, J.K. and Schroeder, F. (1994) Biochemistry 33, 3327-3334. [11] Nemecz, G. and Schroeder, F. (1991) J. Biol. Chem. 266, 1718017186. [12] Bordewick, U., Heese, M., Borchers, T., Robenek, H. and Spener, F. (1989) Biol. Chem. Hoppe-Seyler 370, 229-238. [13] Vancura, A. and Haldar, D. (1992) J. Biol. Chem. 267, 14353-14359. [14] Sha, R.S., Kane, C.D., Xu, Z., Banaszak, L.J. and Bernlohr, D.A. (1993) J. Biol. Chem. 268, 7885-7892.

[15] Jefferson, J.R., Powell, D.M., Rymaszewski, Z., Kukowska-Latallo, J., Lowe, J.B and Schroeder, F. (1990) J. Biol. Chem. 265, 1106211068. [16] Schroeder, F., Perlmutter, L, Glaser, M. and Vagelos, P.R. (1976) J. Biol. Chem. 251, 5015-5026. [17] Jefferson, J.R., Slotte, J.P., Nemecz, G., Pastuszyn, A., Scallen, T.J. and Schroeder, F. (1991) J. Biol. Chem. 266, 5486-5496. [18] Nemecz, G., Hubbell, T., Jefferson, J,R., Lowe, J.B. and Schroeder, F. (1991) Arch. Biochem. Biophys. 286, 300-309. [19] Nemecz, G., Jefferson, J.R. and Schroeder, F. (1991) J. Biol. Chem. 266, 17112-17123. [20] Demediuk, P., Anderson, D.K., Horrocks, L.A. and Means, E.D. (1985) In Vitro Cell. Dev. Biol. 21,569-574, [21] Hara, A. and Radin, N.S. (1978) Anal. Biochem. 90, 420-426. [22] Bradford, M. (1976) Anal. Biochem. 72, 248-254. [23] Dugan, L.L., Demediuk, P., Pendley II, C.E. and Horrocks, L.A. (1986) J. Chromatogr. 378, 317-327. [24] Rouser, G., Siakotos, A. and Fleiscber, S. (1969) Lipids 1, 85-86. [25] Murphy, E.J., Stephens, R., Jurkowitz, M.S. and Horrocks, L.A. (1993) Lipids 28, 565-568. [26] Peeters, R.A., Veerkamp, J.H. and Demel, R.A. (1989) Biochim. Biophys. Acta 1002, 8-13. [27] Borchers, T., Unterberg, C., Rudel, H., Robenek, H. and Spener, F. (1989) Biochim. Biophys. Acta 1002, 54-61. [28] Spener, F., Unterberg, C., Borchers, T. and Grosse, R. (1990) Mol. Cell. Biochem. 98, 57-68. [29] Noy, N. and Zakim, D. (1985) Biochemistry 24, 3521-3525. [30] Burrier, R.E., Mansson, C.R. and Brecher, P. (1987) Biochim. Biophys. Acta 919, 221-230. [31] Wu-Rideout, M.Y.C., Elson, C. and Shrago, E. (1976) Biochem. Biophys. Res. Commun. 71, 809-816. [32] Burnett, D.A., Lysenko, N., Manning, J.A. and Ockner, R.K. (1979) Gastroenterology 77, 247-249. [33] Haq, R.U., Tsao, F. and Shrago, E. (1987) J. Lipid Res. 28, 216-220. [34] O'Doherty, P.J.A. and Kuksis, A. (1975) FEBS Lett. 60, 256-258. [35] Iritana, N., Fukuda, E. and Inoguchi, K. (1980) J. Nutr. Sci. Vitaminol. 26, 271-277. [36] Roncari, D.A.K. and Mack, E.Y.W. (1981) Can. J. Biochem. 59, 944-950. [37] Woodford, J.K., Jefferson, J.R., Wood, G.W., Hubbell, T. and Schroeder, F. (1993) Biochim. Biophys. Acta 1145, 257-265. [38] Incerpi, S., Jefferson, J.R, Wood, G.W., Ball, W.J. and Schroeder, F. (1992) Arch. Biochem. Biophys. 298, 35-42. [39] Prows, D.R., Murphy, E.J. and Schroeder, F. (1995) Lipids 30, 907-910.