Exposure to Free Fatty Acid Increases the Transfer of Albumin across Cultured Endothelial Monolayers

Transcription

1 Exposure to Free Fatty Acd Increases the Transfer of Albumn across Cultured Endothelal Monolayers Bernhard Henng, D. Mchael Shasby, Alce B. Fulton, and Arthur A. Spector An ntal exposure to hgh concentratons of free fatty acd ncreased the transfer of albumn across cultured endothelal monolayers. The rate and amount of albumn transfer was dependent on the concentraton to whch the cultures were ntally exposed, wth 300 p.m producng the maxmum transfer. The albumn transfer also ncreased wth the ncreasng tme of exposure to, the maxmum effect occurrng durng the frst 24 hours. An exposure to 300 jum ln produced an even greater ncrease n albumn transfer than dd 300 /um. The ncreased albumn transfer observed when cells were exposed to hgh concentratons of free fatty acd was largely reversble after rencubaton of the cell monolayers n free fatty acd-poor meda. In parallel experments, radoactve ncorporaton nto cell trglycerdes ncreased lnearly as the fatty acd concentraton was rased, wth cell trglycerde content ncreasng up to sevenfold after ncubaton n a medum contanng 300 /nm. A sgnfcant amount of was ncorporated nto phospholpds, and the fatty acd composton of the endothelal trglycerdes and phospholpds was modfed. All these effects of occurred wthout alterng the ncorporaton of leucne nto the cell proten. These results ndcate that exposure to hgh concentratons of free fatty acd can alter endothelal cell lpd composton, and that ths ncreases the albumn transfer across endothelum. Ths process mght permt more macromolecules to enter the arteral wall. (Arteroscleross 4: , September/October 1984) Snce the endothelum acts as a selectve permeablty barrer to plasma components, 1 t s mportant to mantan endothelal ntegrty to lmt the entry of macromolecules nto the arteral wall. Several lnes of evdence suggest that the mechansm of atheroscleross nvolves damage to the endothelum, resultng n ts reduced effectveness as a barrer. 23 Zlversmt has hypotheszed 4 that the lberaton of fatty acd anons durng hydrolyss of lpoproten trglycerdes n proxmty to the arteral surface may cause local endothelal njury. Ths may facltate the penetraton nto the arteral wall of remnants derved from the trglycerde-rch lpoprotens, leadng to lpd accumulaton wthn the ntma. In addton to endothelal njury produced by exposure From the Departments of Bochemstry and Internal Medcne, Unversty of Iowa, Iowa Cty, Iowa. Ths work was supported by Arteroscleross Specalzed Center of Research Grant HL 14230; Postdoctoral Tranng Grant HL from the Natonal Heart, Lung and Blood Insttute, Natonal Insttutes of Health; and Iowa Afflate of the Amercan Heart Assocaton Grant 83-G-36. Address for reprnts: Dr. Arthur A. Spector, Department of Bochemstry, Bowen Scence Buldng, Unversty of Iowa, Iowa Cty, IA Receved December 27, 1983; revson accepted May 3, to hgh concentratons of fatty acd anons, enrchment of endothelal lpds wth selectve fatty acds has been shown 5 to alter cell morphology n cultured monolayers. A method s now avalable to test expermentally whether exposure to elevated concentratons of fatty acd alters the transfer of macromolecules across the endothelum. Endothelal monolayers are grown on mcropore flters that separate an ncubaton flask nto two compartments. A determnaton of albumn transfer through the monolayer provdes an expermental measure of the movement of a hgh molecular weght substance across the endotheum. Usng such a system, Shasby et al. 6 have demonstrated a reversble ncrease n endothelal transfer of albumn when the cells are exposed to cytochalasn B or D. In the present work, cultured porcne pulmonary artery endothelal cells were exposed to relatvely hgh concentratons of fatty acds n an attempt to determne what mght occur at the arteral surface durng the lpolyss of trglycerde-rch lpoprotens. Albumn transfer across endothelal monolayers treated n ths way was examned to ascertan whether such exposure could enhance the movement of a macromolecule nto the arteral wall.

2 490 ARTERIOSCLEROSIS VOL 4, No 5, SEPTEMBER/OCTOBER 1984 Cell Culture Methods We obtaned M-199-Earle's Base, BME amno acds, BME vtamns, penclln-streptomycn soluton (10,000 unts penclln/ml, 10,000 /xg streptomycn/ ml) and trypsn from KC Bologcals, Incorporated. (Lenexa, Kansas.) We used fetal bovne serum (Sterle System, Logan, Utah) contanng 0.83 mmol/ lter cholesterol, 0.41 mmol/lter trglycerde, and 0.06 mmol/lter free fatty acds. The composton of the major fatty acds n the serum was 16% palmtc (16:0), 5% palmtolec (16:1), 14% stearc (18:0), 28%olec (18:1), 7% lnolec (18:2), and 13% arachdonc acd (20:4). Endothelal cells were obtaned from porcne pulmonary arteres. 7 After these arteres were resected under sterle condtons and rnsed wth M-199, the lumen was exposed to 0.1% collagenase (Worthngton Bochemcal Company, Freehold, New Jersey) for 5 to 10 mnutes. The lumen was then gently swabbed wth a sterle, cotton-tpped applcator, and the adherent endothelal cells were released nto M- 199 contanng 10% fetal bovne serum. The cells were cultured and subcultured n M-199 contanng 10% fetal bovne serum, usng standard technques and flasks obtaned from Cornng Glass Works (Cornng, New York). The cultures were determned to be pure by morphologcal crtera and by demonstratng angotensn-convertng enzyme actvty. Cells from passages 5-12 were plated on treated flters at a densty of 3X10 5 cells/flter, and the flters were mantaned n the standard culture medum. Polycarbonate flters (Nucleopore Corporaton, Pleasanton, Calforna) 13 mm n dameter and 0.8 ^.m pore sze, were gelatn-mpregnated by the method of Postlethwate et al. 8 as modfed by Shasby et al. 7 These flters were coated wth fbronectn (Collaboratve Research, Incorporated, Waltham, Massachusetts) before cell platng. All studes were started 2 days after the cells were plated on the flter. For lpd analytcal studes, the cells were grown n Costar tssue culture plates (Cambrdge, Massachusetts). Trypan blue (Gbco Laboratores, Grand Island, New York) excluson was used to ndcate cell vablty. Albumn Transfer Studes Treated flters were glued to polystyrene chemotactc chambers (ADAPS, Incorporated, Dedham, Massachusetts) and after ethylene oxde sterlzaton, the endothelal cells were plated on the flters. The flters and chemotactc chambers were then placed nto the wells of a standard 24-well tssue culture plate (Costar Incorporated, Cambrdge, Massachusetts) and ncubated for 48 hours n M-199 enrched wth 10% fetal bovne serum at 37 C n an atmosphere of 95% ar and 5% CO 2. After 48 hours, the chemotactc chambers wth attached flters and endothelal monolayers were removed from the 24- well plate and were washed free of serum by gentle mmerson n M-199; they were then placed nto new 24-well tssue culture plates wth control or fatty acdsupplemented meda. The meda were composed of M-199 enrched wth vtamns, amno acds, 15 mm HEPES, 5% fetal bovne serum, 100 /ULM bovne albumn (Fracton V, fatty acd-free, Mles Laboratores, Elkhart, Indana or crystallne bovne albumn, fatty acd-free, Sgma Chemcal Company, St. Lous, Mssour) and varous concentratons of (NuChek Prep, Elysan, Mnnesota, 98% pure by gas-lqud chromatography). The cells were ncubated n the expermental or control meda for tme perods specfed n the text. After the ncubaton perod, the chemotactc chambers wth attached monolayers were removed from the 24-well plates, washed three tmes by mmerson n M-199 and placed nto new 24-well plates for determnaton of albumn transfer. In the new plates, each well of a 24-well plate was flled wth 1.5 ml of serum-free M-199. The chemotactc chamber wth attached monolayer was placed nto one of these wells, and the nner well of the chamber was flled wth 0.5 ml of M-199 contanng 200 /M albumn along wth vtamns, amno acds, and 15 mm HEPES. After 1 hour, the meda wthn the chemotactc chamber and the surroundng meda n the wells of the 24-well tssue culture plate were sampled, and ther respectve albumn concentratons were determned. Ths was done by followng the change n absorbance at 630 nm after addton of a porton of the sample to a reagent soluton of bromcresol green (Sgma Chemcal Company, St. Lous, Mssour). The separate monolayers were tested for each varable; all cells used n a sngle experment were from the same passage of the same lne and were plated at the same tme. All expermental data were analyzed usng Student's t test. For experments testng the reversblty of the fatty acd effect, the chemotactc chambers and attached monolayers were removed from the plate used to measure albumn transfer and placed nto a new 24- well plate, where the monolayer was rencubated n M-199 enrched wth 10% fetal bovne serum. At the end of the rencubaton, the chemotactc chambers and monolayers were washed and transferred to new plates, and the albumn transfer was agan determned. Leuclne Incorporaton Incorporaton of 4,5-3 H-leucne (Amersham Internatonal Lmted, Amersham, UK, 147 C/mmol) nto total cell proten was determned by the procedure of Cooper. 9 Cells were ncubated for 24 hours n expermental meda, enrched wth ether 100 (xm albumn and 300 f.m, or 100 fm albumn. All of the expermental meda contaned 5% fetal bovne serum. The fatty acds were dssolved n hexane. After the addton of 1 to 2 drops of 6N NaOH and dryng under hgh purty N 2, the resdue was redssolved n a small amount of warm dstlled water. The clear solu-

3 FATTY ACIDS AND ENDOTHELIAL ALBUMIN TRANSFER Henng et al. 491 ton was added wth strrng to M-199 supplemented wth fetal bovne serum plus albumn, and the ph was adjusted mmedately to 7.4. Before ncubaton wth 3 H-leucne n Dulbecco's phosphate-buffered salne (DPBS), the cells were washed wth warm DPBS contanng 50 /xm fatty acd-free albumn, followed by warm DPBS. The cells were ncubated wth the 3 H-leucne for 2 hours, washed once wth ce-cold DPBS contanng 50 /AM albumn, and then twce wth ce-cold DPBS. After the addton of 5% (wt/vol) trchloroacetc acd, the cells were harvested and placed n a bolng water bath for 20 mnutes. The tubes were cooled and the precptates were collected on glass fber flters. These flters were rnsed wth ce-cold 100% ethanol before beng placed nto countng vals. Ten mllters of Budget-Solve (Research Products Internatonal, Incorporated, Mt. Prospect, Illnos) was added to each val, and radoactvty was determned n a Beckman LS7000 lqud scntllaton spectrometer equpped wth a 137 Cs external standard to montor quenchng. Incubaton and Llpld Analyses Confluent monolayers n Cornng tssue culture flasks were washed wth warm DPBS contanng 50 nm albumn and then DPBS. The cells were ncubated for the ndcated tme at 37 C wth expermental meda contanng 9,10-3 H-, 6-8 x 10 s dpm/flask (New England Nuclear, Boston, Massachusetts, 9.5 C/mmol). The expermental meda also contaned 5% fetal bovne serum. After ncubaton, the monolayers were washed wth ce-cold DPBS contanng 50 ^M albumn and then ce-cold DPBS. The cells were scraped nto tubes, sedmented at 600 g for 10 mnutes, and then resuspended n DPBS. Alquots were taken for proten determnaton by a modfcaton of the Lowry method. 10 The remanng cells were extracted wth CHCI3/CH3OH, 2:1 (vol/ vol), 11 and the CHCI 3 phase was solated by addng 9 mm NaCI contanng 0.04 N HCI. To measure the Utlzaton of the ncorporated radoactve fatty acd, we exposed confluent monolayers for 16 hours at 37 C to meda contanng 150 fm 9,10-3 H-. The cells were then washed wth DPBS contanng 50 /xm albumn, followed by DPBS. Four of the cultures were extracted at ths tme to measure the amount and dstrbuton of the ncorporated radoactvty. The remanng cultures were ncubated for varous tmes at 37 C wth meda that was not supplemented wth fatty acds. After ncubaton, the monolayers were washed and harvested to determne the amount and dstrbuton of the radoactvty remanng n the cells. To determne the ncorporaton of radoactvty nto cell lpds, we dred alquots of the CHCI 3 extract under N 2 and dssolved them n 5 ml of Budget-Solve. The dstrbuton of radoactvty n the cellular lpd fractons was determned by thn-layer chromatography on slca gel G plates (Analabs, North Haven, Connectcut). The chromatograms were developed n heptane/dethyl ether/acetc acd/methanol, 85: 20:2:2 (vol/vol). 12 The plates were staned wth l 2 and after sublmaton, bands of slca gel correspondng to the dfferent lpd fractons were scraped nto scntllaton vals contanng 5 ml of Budget-Solve and were assayed n the lqud scntllaton spectrometer. For fatty acd compostonal analyss, the endothelal cultures were ncubated and the lpds were extracted and separated by thn-layer chromatography. No radoactve sotopes were present n these experments and these chromatograms were not exposed to any stan. The trglycerde and phospholpd bands were localzed from correspondng chromatograms contanng standards, and the segments of slca gel were scraped nto separatory funnels. Lpds were extracted wth CHCIg/CHaOH, 1:1 (vol/vol) and the phases were separated wth 9 mm NaCI contanng 0.04 N HCI. After chloroform extracts were dred under N 2,1 ml of 14% BF 3 n CH 3 OH, 1 ml of acetontrle, and 2 ml of CH 3 OH were added to each tube. The samples were mxed and heated at 95 C for 10 mnutes, and the methyl esters were extracted nto n-heptane. The fatty acd methyl esters were separated by gas-lqud chromatography usng a 2 mm x 1.9 m glass column packed wth 10% SP2330 on 100/120 mesh Chromasorb W- AW; 12 peak areas were measured wth a Hewlett- Packard 3380A ntegrator. Addtonal chromatograms were run wth use of fatty acd methyl ester standards obtaned from Nu Check Prep. In the lpd extract obtaned from other cultures, the trglycerde content was determned by a modfcaton of the procedure of Kessler and Lederer, adapted to the Techncon Autoanalyzer II Trglycerdes were measured fluorometrcally after hydrolyss to release glycerol. The glycerol was subsequently oxdzed to formaldehyde, followed by couplng wth acetylacetone to gve a fluorescent product, 3,5-dacetyl-4-dhydrolutdne. Results Albumn Transfer Studes Intal experments were conducted to test the valdty of the endothelal monolayer system used to measure albumn transfer. It was also found that the presence of varous amounts of fatty acd dd not nfluence the bromcresol green colormetrc assay used to measure albumn concentraton. No loss n cell vablty was observed, and the cell number was not reduced due to fatty acd exposure durng culture. Ths was establshed wth trypan blue excluson studes and morphologcal assessment by phase contrast mcroscopy and scannng electron mcroscopy. Furthermore, there was no reducton n the ncorporaton of g.o-^h-leucne nto cellular proten when the cultures were exposed to. For example, the ncorporaton of radoactvty nto the total cell proten after a 2-hour ncubaton perod was 8360 ± 360 dpm/mg proten (mean ± SE) n control

4 492 ARTERIOSCLEROSIS VOL 4, No 5, SEPTEMBER/OCTOBER r - In Transfe > / / / Albumn (MM) Fgure 1. The effect of albumn concentraton on net transfer across cultured endothelal monolayers. Before measurng albumn transfer over a 1-hour perod, cells were washed wth M-199 and ncubated for 24 hours wth M-199 enrched wth 5% fetal bovne serum and 100 /xm albumn, wth or wthout 300 /xm. The albumn concentraton of the M-199 added to the upper compartment of the test chamber vared from 50 to 300 fm. Values are means ± SE, n = 6. cultures and 8930 ± 690 n cultures exposed for 24 hours to 300 ^M (n = 6). Fgure 1 shows the effect of ncreasng albumn concentraton on net transfer across the endothelal cell monolayer durng a 1-hour test perod. The ncrease n albumn transfer was concentraton-dependent. Compared wth control cultures, the albumn transfer was always sgnfcantly greater across monolayers that were exposed durng the prevous 24 hours to 300 /JLM. Fgure 2 shows the effect of concentraton durng the ntal 24-hour perod on albumn transfer. The amount of albumn transferred across the endothelal cell monolayer durng a subsequent 1-hour ncubaton was dependent on the concentraton to whch the cultures were exposed, up to 300 fm where maxmum transfer appears to have been reached. Table 1 shows that the fatty acd effect on albumn transfer was largely reversble. Cells were frst ncubated for 24 hours wth M-199 that was enrched wth 5% fetal bovne serum and 100 /M albumn wthout Olec Acd (MM) Fgure 2. The effect of concentratons on subsequent net transfer of albumn across cultured endothelal monolayers. Before measurng albumn transfer over a 1- hour perod, cells were washed wth M-199 and then ncubated for 24 hours wth M-199 enrched wth 5% fetal bovne serum and 100 fxm albumn contanng varous concentratons of. The concentraton of the albumn added to the upper chamber durng the subsequent 1 - hour ncubaton was 200 /xm. Values are means ± SE, n = 6. Table 1. Reversblty of Fatty Acd Effect on Albumn Transfer across Endothelal Cell Monolayers Supplemental Intal ±0.9 Albumn transfer nmol After rencubaton 4.4 ±0.6* 4.4±0.3t Cells were ncubated ntally for 24 hours wth M-199 enrched wth 5% fetal bovne serum and 100 /M albumn wthout (control) or wth ether 100 or 300 ^M. Albumn net transfer was measured durng a 1-hour ncubaton perod wth M-199 contanng 200 /xm albumn (ntal). Subsequently, all cells were rencubated for 8 hours wth M-199 enrched wth 10% fetal bovne serum, and albumn net transfer was measured agan. Each value s the mean ± SE wth n = 6. 'Not sgnfcantly dfferent from ntal value, p > tsgnfcantly less than ntal value after exposure to 100 MM, p < ^Sgnfcantly less than ntal value after exposure to 300 fm, p <

5 FATTY ACIDS AND ENDOTHELIAL ALBUMIN TRANSFER Henng et al. 493 (control) or wth ether 100 or 300 /AM. Albumn transfer, measured over a 1-hour perod, ncreased wth ncreasng concentraton. To study reversblty, we washed all cell monolayers wth M-199 and rencubated them for 8 hours wth M- 199 enrched wth 10% fetal bovne serum. Albumn transfer was measured agan as above after the second ncubaton. The fatty acd effect on albumn transfer was completely reversble when the monolayers were ntally exposed to the lower fatty acd concentraton (100 fm, 100 /*M albumn) and partally reversble when the ntal exposure was to the hgher fatty acd concentraton (300 /xm olec acd, 100 /u.m albumn). Fgure 3 llustrates the effect of the tme of exposure to an -supplemented medum on albumn transfer across the endothelal cell monolayers. The cultures were ncubated wth or wthout 300 /M for perods of tme rangng from 2 to 48 hours, and albumn transfer was tested durng a subsequent 1-hour ncubaton. Albumn net transfer ncreased wth ncreasng tme of exposure to olec acd, the maxmum effect occurrng durng the frst 24 hours. Olec acd treatment for 2 or 6 hours, however, dd not produce any sgnfcant ncrease n albumn net transfer across endothelal cell monolayers. Some ncrease n albumn transfer also occurred as a result of the 48-hour ncubaton even when the medum contaned no supplemental fatty acd, but the ncrement was consderably smaller than when the ncubaton medum contaned supplemental olec acd. All of the ncrease that occurred between 24 and 48 hours of exposure to could be accounted for by the ncrease observed n the cultures ncubated wth the control medum. Table 2 llustrates that an ntal ncubaton of the cultures wth a medum contanng supplemental lnolec acd also produced an ncrease n albumn transfer across endothelal cell monolayers. A greater ncrease n albumn transfer occurred wth 300 ^M ln than wth 300 /AM. The acute addton of varous amounts of fatty acd to the medum used to measure albumn transfer, whch contaned 200 tm albumn, dd not affect albumn transfer across the endothelal monolayer. For example, the amount of albumn transferred over a 1-hour perod was 4.0 ± 0.5 nmol (mean ± SE) (no addtonal fatty acd), 4.0 ± 0.2 nmol (50 /um olec acd), 3.9 ± 0.2 nmol (100 M oec acd) and 4.3 ± 0.3 nmol (200 LIM ). Fatty Add Uptake and Utlzaton Intal experments ndcated that 9,10-3H-olec acd was ncorporated nto the trglycerdes and phospholpds of the endothelal cells n a tme-dependent manner. The uptake was essentally complete wthn 24 hours. Based on these fndngs, a 24- hour ncubaton perod was chosen for most experments. The effect of concentraton on ts ncorporaton nto the endothelal cell trglycer- o S s 20 c 10 1 I I I I Control O Tme (h) Fgure 3. The effect of tme of exposure to on subsequent transfer of albumn across cultured endothelal monolayers. Albumn transfer was measured over a 1- hour perod wth M-199 contanng 200 /xm albumn. The cells had been washed earler wth M-199 and then ncubated for the tmes ndcated n M-199 enrched wth 5% fetal bovne serum and 100 /AM albumn, wth or wthout 300 fj.m. The earlest tme at whch albumn transfer was measured was 2 hours. Values are means ± SE, n = 6. Table 2. Comparson of Albumn Transfer across Endothelal Cell Monolayers Exposed to Ether Olelc or Lnolec Acd Fatty acd Albumn transfer (nmol) Control 4.7 ±0.2* Olec acd 7.8±0.5f Ln 13.5 ±1.2$ Cells were ncubated ntally for 24 hours wth M-199, whch was enrched wth 5% fetal bovne serum and 100 um albumn wthout (control) or wth ether 300 pm olec or ln. Subsequently, albumn net transfer was measured durng a 1-hour ncubaton perod wth M-199 contanng 200 fj.m albumn. 'Each value s the mean ± SE wth n = 6. \p < vs control. Xp < vs.

6 494 ARTERIOSCLEROSIS VOL 4, No 5, SEPTEMBER/OCTOBER 1984 des and phospholpds after a 24-hour ncubaton s shown n Fgure 4. Each expermental medum contaned 5% fetal bovne serum and 100 fm fatty acdfree albumn, n addton to the supplemental olec acd. Incorporaton of radoactvty nto trglycerdes ncreased lnearly as the concentraton was rased. No ndcaton of saturaton was evdent even when the concentraton was 400 /xm. By contrast, more than one-half of the total radoactvty recovered n phospholpds occurred at the lowest concentraton tested, 50 /xm. Up to an concentraton of 100 /xm, ncorporaton of radoactvty nto phospholpds was greater than nto trglycerdes. Above ths concentraton, however, consderably more radoactvty appeared n trglycerdes. To assess the utlzaton of the ncorporated olec acd, we labeled cultures wth 9,10-3 H- for 16 hours. At ths tme, about 65% of the ncorporated radoactvty was present n trglycerdes and about 20%, n phospholpds. After washng the cells wth buffer contanng albumn to remove any adherent free fatty acd radoactvty, we ncubated the cultures n M-199 contanng fetal bovne serum and albumn wthout supplemental fatty acd. As seen n Fgure 5, there was a rapd declne n trglycende radoactvty durng ths ncubaton. Decreases of 18%, 55%, and 78% occurred at 2, 8 and 24 hours, respectvely. By o 500 I 1= Trlglyc«rld«s Ptotphollplds Olec Acd (um) Fgure 4. Incorporaton of 9,10-3 H- nto ertdothelal cell lpds. Incubaton tme was 24 hours. Values are the means ± SE of results from three separate cultures T1me(h) 24 Fgure 5. Utlzaton of radoactve fatty acds n endothelal cell lpds. The cells were ncubated for 16 hours n M- 199 enrched wth fetal bovne serum, fatty acd-free albumn, and 150 t*.m 9,10-3 H-. After washng, the cultures were ncubated for the tmes ndcated n M-199 contanng fetal bovne serum and 50 pm fatty acd-free albumn, but no supplemental fatty acd. Values are means ± SE of results from three separate cultures. contrast, radoactvty n phospholpds ncreased durng ths ncubaton, the ncreases beng 19%, 42%, and 53% at 2, 8 and 24 hours, respectvely. Thus, a large percentage of the radoactvty orgnally contaned n trglycerdes was conserved n cell phospholpds durng trglycerde breakdown. Radoactvty also was recovered as free fatty acd released nto the medum. The amount of free fatty acd radoactvty present after 24 hours was 2.9 tmes greater than that present after 2 hours. Trglycerde Content Table 3 shows the changes n the chemcally measured trglycerde content of the endothelal cells. After ncubaton for 24 hours wth 100 ym olec acd, there was a twofold ncrease n trglycerde content as compared wth cultures ncubated wthout supplemental fatty acd, and a seven-fold ncrease occurred as a result of ncubaton wth a medum contanng 300 p,m. As wth, the ncubatng of cells wth ln also resulted n an elevated trglycerde content. The ncrease was

7 FATTY ACIDS AND ENDOTHELIAL ALBUMIN TRANSFER Henng et al. 495 Table thelal Type None Olec Olec Lnolec 3. Changes Cells Supplemental In Trlglycerlde Content of Endo- fatty acd Amount (nm) Cellular trglycerde content (ng/mg proten) 19±1 38±5 125±3 95±2 Before measurement of trglycerde content, cells were ncubated for 24 hours wth M-199 enrched wth 5% fetal bovne serum, 100 M fatty acd-free albumn, and varous concentratons of. Values are means ± SE; n = 3. fvefold after a 24-hour ncubaton when 300 pm of ln was compared to correspondng control cultures. Fatty Acd Composton The fatty acd composton of the trglycerde fracton after a 24 hour ncubaton wth meda contanng 0, 100 or 300 JAM s shown n Table 4. These values do not add up to 100% because only the major fatty acds are lsted. Compared wth control cultures, the percentage of ncreased by 180% followng ncubaton wth 100 xm and by 240% when the meda contaned 300 /xm. The percentage of most other fatty acds, except olec and arachdonc acd, decreased followng ncubaton wth supplemental. Table 4 also shows that a consderable enrchment of the endothelal cell phospholpds wth occurred durng a 24-hour ncubaton wth the meda contanng supplemental. In contrast to ts effects upon the trglycerde fracton, however, the 100 AAM concentraton produced nearly maxmal changes n the fatty acd composton of the phospholpd fracton. Table 4. Fatty acd 16:0 18:0 18:1 18:2 204 Dscusson A number of studes 1-3 suggest that atheroscleross may be ntated by damage to the endothelum and the resultng dsturbance n endothelal ntegrty. The endothelum normally s the only cell n the arteral wall exposed to hgh concentratons of lpoprotens that are rch n trglycerdes and cholesterol. 15 When the plasma trglycerde-rch lpoprotens, ether chylomcrons derved from detary fat or very low densty lpoprotens of hepatc orgn, are elevated, trglycerde hydrolyss by lpoproten lpase occurs n proxmty to the endothelal surface Ths may expose the endothelum to an excessve local concentraton of fatty acd anons and lead to an accumulaton of fatty acd wthn the endothelal cells. 4 ' 19 It has been suggested 24 ' 19 that excessve amounts of fatty acd anons lberated durng lpoproten trglycerde hydrolyss may cause endothelal njury, allowng ncreased penetraton of cholesteryl ester-rch remnant lpoprotens derved from chylomcrons or very low densty lpoprotens nto the arteral wall. Accordng to ths hypothess, 19 the subsequent events leadng to atherosclerotc leson formaton are ntated by the accumulaton of these cholesterol-rch remnants n the arteral ntma. In the present study usng cultured endothelal cell monolayers derved from porcne pulmonary arteres, we obtaned some expermental evdence consstent wth the general concept of ths hypothess. Our fndngs ndcate that exposure to hgh concentratons of fatty acd, even when bound to plasma albumn, can affect endothelal lpd content and composton. These changes were accompaned by an ncrease n the transfer of albumn, a proten havng a molecular weght of S/.OOO, 20 across the endothelal monolayer. When the monolayers were ntally exposed to moderate concentratons of fatty acd, the effect was completely reversble wthn 8 hours after removal of the medum supplemented wth fatty acd, but when the ntal exposure was to hgh concentratons of fatty acd, the effect was only partally reversble after 8 hours. Fatty Acd Composton of Endothelal Cell Trglycerldes and Phospholpds Fatty acd composton (%) Trglycerdes Phospholpds Medum wthout 33.4 ± ± ± ± /AM 20.3 ± ± ± ± ± /xM 11.6± ± ± ± ±0.2 Medum wthout 18.6± ± ± ± ± /xM 13.5± ± ± ± ± nm 10.9± ± ± ± ±0.2 The expermental desgn was the same as that descrbed n Table 1. The values for the fatty acd composton do not add up to 100% because only major fatty acds of nterest are lsted. Each value represents the mean ± SE, n = 3. Olec acd was added to the culture medum n amounts of ether 100 M or 300 nm, as ndcated.

8 496 ARTERIOSCLEROSIS VOL 4, No 5, SEPTEMBER/OCTOBER 1984 The albumn transfer measurement was made after the medum contanng the supplemental fatty acd had been removed, not whle the fatty acd stll was present. Fatty acd-free albumn was used for the transfer measurement. Although albumn may have taken up some of the free fatty acd released from the lpd-enrched cells durng the tme of the transfer measurement, addtonal studes revealed that nether the rate of albumn transfer nor the method used to assay albumn were affected by bound fatty acd. For these reasons, t seems unlkely that the ncreased transfer of albumn s due to an nteracton between fatty acd and albumn durng the tme of the transfer measurement. Whether the transfer of remnant lpoprotens or low densty lpoprotens (whch have a much hgher molecular weght) would also ncrease when the endothelal monolayer s exposed to elevated concentratons of free fatty acd has not yet been tested. Such evdence would be needed to fully support the hypothess 19 that fatty acds nduce ncreased transfer of macromolecules nto the arteral wall. However, the postve results obtaned wth albumn, a moderately large proten, suggest that t would be worthwhle to extend these studes to lpoprotens. In the present experments the endothelal monolayer was drectly exposed to albumn-bound fatty acds. Ths s a smpler approach than attemptng to generate fatty acds from trglycerde-rch lpoprotens n a cell culture system. Furthermore, ths desgn permts a more certan concluson that free fatty acd s the substance medatng the endothelal changes. In attemptng to relate these fndngs to the hydrolyss of trglycerde-rch lpoprotens, we do not know the free fatty acd concentraton at the endothelal surface. The concentratons of plasma-free fatty acd producng the ncreased transfer are hgh, but stll reasonable, relatve to values that can occur n humans. Plasma-free fatty acd concentratons can range from 180 to 1,650 ^M, 21 the usual value n the basal state beng about 300 to 500 /AM. 22 Assumng normal albumn concentraton of 600 /xm, the molar rato of free fatty acd to albumn n human plasma wll vary between 0.3 and 2.8. Hgh plasmafree fatty acd concentratons, and hence molar ratos, occur durng severe stress, 2324 uncontrolled dabetes, 25 M starvaton, 2127 and after prolonged exercse. 28 Metabolc studes 29 ' x ndcate that the molar rato s the man factor controllng free fatty acd avalablty to the tssues. Although most of our experments were done wth fatty acd solutons havng a molar rato of 3.0 (a value slghtly above the hghest values n humans), ncreased transfer of albumn across the endothelum occurred at molar ratos as low as 1. Moreover, the concentraton-dependence data (Fgure 2) suggest that greater albumn transfer may occur as a result of exposure to fatty acd solutons that are n the upper part of the human range (molar ratos between 2.0 and 2.8) as compared wth those of molar rato 0.5 to 1.0. A large ntracellular trglycerde accumulaton occurred when the porcne endothelal cells were ncubated wth hgh concentratons of free fatty acd. Ths was assocated wth the appearance of cytoplasmc nclusons. Smlar fndngs were made prevously 5 wth cultured bovne aortc endothelal cells. The trglycerde probably serves as a storage form of fatty acd for the endothelal cells. As was noted by Tsa and Geyer 31 n murne L fbroblasts, some of the fatty acd present n trglycerdes s transferred nto cellular phospholpds when the trglycerdes are catabolzed (Fgure 5). Lkewse, as reported 32 for macrophages that contan trglycerde accumulatons, the endothelal cells released some fatty acd nto the medum as free fatty acd durng trglycerde catabolsm. It appears unlkely, however, that the trglycerde accumulaton tself s the drect cause of the ncreased endothelal albumn transfer. For example, ln caused a greater ncrease n albumn transfer even though t produced less total trglycerde accumulaton than (Table 3). It also seems unlkely that any free fatty acd released from the cells as a result of trglycerde catabolsm mght be drectly responsble for the ncrease n albumn transfer. The data n Fgure 5 suggest that trglycerde catabolsm occurs durng the flux measurement. Whle some of the released fatty acd probably bnds to the fatty acd-free albumn present n the medum, ths should not affect the transfer because drect measurements ndcated that the acute addton of up to 200 /J.M does not alter the rate of albumn transfer. Whle our current data do not permt a defnte concluson about the exact mechansm causng ncreased albumn transfer across endothelal monolayers, t s probably related to a change n cell membrane characterstcs secondary to the altered lpd composton. Studes wth fbroblasts, retnoblastoma cells, Ehrlch asctes cells, and L1210 cells ndcate that the phospholpd modfcatons reflect the fatty acd compostonal changes that occur n cell membranes Such compostonal changes are suffcent to alter membrane fludty A mechansm nvolvng membrane fatty acd compostonal changes mght result n ncreased transcellular movement of albumn n vescles or transcellular channels 3839 or t mght alter cell-cell adheson and ncrease transfer of macromolecules between cells. Our data demonstrate an assocaton between alteratons n cell lpd composton and the ntegrty of the endothelum as a barrer to macromolecules. Altered cell membrane characterstcs are the most lkely explanaton for ths assocaton. References 1. Renkln EM, Curry TE. Endothelal permeablty: pathways and modulatons. Ann NY Acad Sc 1982;401: Ross R, Harfcer L Hyperpdema and atheroscleross. Chronc hyperpdema Intates and mantans lesons by endothelal cell desquamaton and lpd accumulaton. Scence 1976;193: Alavl M, Dunnett CW, Moore S. Lpd composton of rabbt

9 FATTY ACIDS AND ENDOTHELIAL ALBUMIN TRANSFER Henng et al. 497 aortc wall followng removal of endothelum by balloon catheter. Arteroscleross 1983;3: Zlversmlt DB. Role of trglycerde-rch lpoprotens n atherogeness. Ann NY Acad Sc 1976;275: Dennng GM, Flgard PH, Kaduce TL, Spector AA. Role of tnglycerdes n endothelal cell arachdonc acd metabolsm. J Lpd Res 1983;24: Shasby DM, Shasby SS, Sullvan JM, Peach MJ. Role of endothelal cell cytoskeleton n control of endothelal permeablty. Crc Res 1982;51: Shasby DM, Shasby SS, Peach MJ. Granulocytes and phorbol myrstate acetate ncrease permeablty to albumn of cultured endothelal monolayers and solated perfused lungs. Am Rev Resp Ds 1983;127: Postlethwate AE, Snyderman R, Kang AH. The chemotactc attracton of human fbroblasts to a lymphocyte-derved factor. J Exp Med 1976:144: Cooper TG. The tools of bochemstry New York: John Wley, 1977: Lees MB, Paxman S. Modfcaton of the Lowry procedure for the analyss of proteolpd protens. Anal Bochem 1972; 47: Folch J, Lees M, Sloane Stanley GH. A smple method for the solaton and purfcaton of total lpds from anmal tssues. J Bol Chem 1957;226: Spector AA, Klser RE, Dennng GM, Koh S-W, DeBaurt LE. Modfcaton of the fatty acd composton of cultured human fbroblasts. J Lpd Res 1979;20: Kessler G, Lederer H. Fluorometrc measurement of tnglycerdes. In: Skeggs LT, ed Automaton n analytcal chemstry. Techncan Symposum. New York: Medad, Lpd Research Clncs Program. Manual of laboratory operatons. Vol 1: Lpd and lpoproten analyss DHEW publcaton (NIH) Bethesda, Maryland Natonal Insttutes of Health, Feldng CJ. The endothelum, trglycerde-rch lpoprotens, and atheroscleross Insghts from cell bology and lpd metabolsm. Dabetes 1981;30: Blanchette-Mackle EJ, Scow RO. Stes of lpoproten lpase actvty n adpose tssue perfused wth chylomcrons. J Cell Bol 1971 ; Pedersen ME, Cohen M, Schotz MC. Immunocytochemcal localzaton of the functonal fracton of lpoproten lpase n the perfused heart. J Lpd Res 1983, Nllsson-Ehle P, Garflnkel AS, Schotz MC. Lpolytc enzymes and plasma lpoproten metabolsm. Annu Rev Bochem 1980:49: Zllversmt DB. A proposal lnkng atherogeness to the nteracton of endothelal lpoproten lpase wth trglycerde-rch lpoprotens. Crc Res 1973;33: Brown JR. Structural orgns of mammalan albumn. Fed Proc 1976;35: Fredrlckson DS, Gordon RS Jr. The metabolsm of albumn-bound, C-' 4 -labeled unesterfed fatty acds n normal human subjects. J Cln Invest 1958:37: Spector AA. Fatty acd bndng to plasma albumn. J Lpd Res 1975; 16: Knltza R, Clasen R, Kunz C. Serum concentratons of nonesterfed fatty acds durng operatve stress and blockade of beta adrenergc receptors. Kln Wochenschr 1978;56: Batt RAL, Toppng DL. Acute effects of ncotne on plasma free fatty acd concentratons and on the response to cold stress, n lean and obese (genotype ob/ob) mce. Int J Obes 1979;3: Galton DJ, Reckless JPD, Cllfton-BIIgh P, Glbert CH. The control of fatty acd metabolsm n adult dabetes. Proc Nutr Soc 1975;34: Hall SEH, Saunders J, Sonksen PH. Glucose and free fatty acd turnover n normal subjects and n dabetc patents before and after nsuln treatment. Dabetologa 1979;16: Sawln CT, Wlllard DA. Normal rse n plasma free fatty acds durng fastng n patents wth hypoptutarsm. J Cln Endocr 1970:31: Havel RJ, Namark A, Borchgrevlnk CF. Turnover rate and oxdaton of free fatty acds of blood plasma n man durng exercse: Studes durng contnuous nfuson of oalmtate-1- C 14. J Cln Invest 1963,42: Spector AA. The transport and utlzaton of free fatty acd. Ann NY Acad Sc 1968:149: Spector AA. Metabolsm of free fatty acds. Progr Bochem Pharmacol 1971:6: Tsal P-Y, Greyer RP. Effect of exogenous fatty acds on the retenton of phospholpd acyl groups by mouse L fbroblasts. Bochm Bophys Acta 1978:528: Von Hodenberg E, Khoo JC, Vance J, Wtztum JL, Stenberg D. Free fatty acd moblzaton from tnglycerde-loaded macrophages catalyzed by a neutral trglycerde lpase. Crculaton 1983:68: Hyman BT, Spector AA. Cholne uptake n cultured human Y79 retnoblastoma cells: Effect of polyunsaturated fatty acd compostonal modfcatons. J Neurochem 1982:38: Yorek MA, Hyman BT, Spector AA. Glycne uptake by cultured human Y79 retnoblastoma cells: Effect of changes n phospholpd fatty acd unsaturaton. J Neurochem 1983; 40: Awad AB, Spector AA. Modfcaton of the fatty acd composton of Ehrlch asctes tumor cell plasma membranes. Bochm Bophys Acta 1976;426: Burns CP, Luttenegger DG, Dudley DT, Buettner GR, Spector AA. Effect of modfcaton of plasma membrane fatty acd composton on fludty and methotrexate transport n L1210 munne leukema cells Cancer Res 1979;39: Kng ME, Spector AA. Effect of specfc fatty acyl enrchments on membrane physcal propertes detected wth a spn label probe. J Bol Chem 1978;253: Slmlonescu N, Slmlonescu M, Palade GE. Recent studes on vascular endothelum. Ann NY Acad Sc 1976;275: Bundgaard M. Vescular transport n capllary endothelum: Does t occur? Fed Proc 1983;42: Index Terms: endothelum fatty acd transfer albumn lpds