Bhwhimh'a ('t Biophysica .4cta. i 12,'I (19t~2) 267-274 c 1992 Elsevier Science Publishers B.V. All rights rcscr,.'cd (ll]()5-27fil)/t)2/S05.l)t)

267

BBALI P 541)24

Essential fatty acid m tabolism in cultured human airway epithelial cells J i n g X. K a n g , S.F. P a u l M a n , Neff E. B r o w n , P a u l A. L a b r e c q u e , M . L . G a r g and M. Thomas Clandinin Ntttrilhm am/.~h'tabt)li~m Re.warcll (;rmtp. l)('partm('nts ot" .lh'di('im. amf t:~)t,,t~am/Nutrithm, Unirersity al',41berta. [((bm)nlott ¢('ana(ht )

(Rccci,.ctl I9 February 1992)

Key words: Essential fatty acid; [ncorporalion: Dc~,tur~tion: Fatty acid composition: t lumun :fir~vayepithelium

To characterize essential fatty acid metabolism of human airway epithelium, we examined the capacity of epithelial cells to incorporate and dcsaturatc/clongatc 18: 2(n - 6 ) and the turm)vcr t)f phospholipid httty acyl chains in these cells. Epithelial cells wcrc cultured ti)r 5-7 days and incubated with [1-I'tC]lg:2(n - 6) (1 /aCi, tt10 nmol). The essential fatty acid profile of the cells was readily modified by 1 8 : 2 ( n - ¢ 0 supplementation to culture medium. Attot 4 h incubation, 3 2 + 5 . 6 nmol of [I t 4 C ] 1 8 : 2 ( n - 6 ) w a s incorporated into pho.',pholipids (65 +__9.5G. of which 74"~ was incorporated into phosphatidylcholinc (PC)) and neutral lipid (31 + lilt7,;) per mg prt)tcin t)f cuhurcd c c l k 3(I ± SG of [I-t'~C]lS:2(n - fi) incorporated, was converted

to homologous tricncs, Ictraencs and pcntacncs, the major products being 20: 3 ( n - 6) and 211:1(;:- 6,). The c,,).':;cr~.i,,m,,~f 18:2(n - 6 ) was time-dependent and donor age-related, A ifighcr proportion of 211:3(n - 6) and 20:4(n - 6) was incorporated into pht)sphatidylinositol (PI) and phosphatidylcthanolaminc (PE). About 10-15e; of total products formed from 18:21n - 6 ) was released from membrane to culture medium. Both 20:4(n - 6) and 20:5(n 3) inhibited 18:2(n - 6) incorpor~ltion and dcsaturatit)n. Rate of incorporation of 1 8 : 2 ( n - 6 ) w a s more than either 18: l ( n - 9 ) o r 16:1). With pulse-chase studio.',, the

half-life of 1 8 : 2 ( n - 6 ) in PC. PI and PE was estimated to be 5.5, 6.{1 :rod 7.3 h, respectively. These data indicate active metabolism o1" essential fatty acids in human airway epithelial cells. This metabolism may play a key role in the regulation of membrane properties and function i.1 these cells.

Introduction Changes in the composition of phospholipid fatty acid chains can affect a number of cellular functions, including ion transport and transmembrane signal transduction, by altering the conformation, activity and translational diffusion of proteins embedded in the lipid matrix [ 1-3]. Relative availability of saturated and unsaturated fatty acids for phospholipid synthesis is determined by dict, fatty acid incorporation and activity of elongasc and desaturasc enzymes. Thus, the capacity of a particular cellular or subcellular fraction to elongate/desaturate available fatty acid substrate and deacylate/reacylate existing membrane lipid fatty acids, determines in situ control of membrane phospholipid composition.

Corrcspt)ndencc to: M.T. Clandinin. Nutritit)n and Metaboli.sm Research Group, 533 Newton Building, University of Alberta, Edmonton, Alberta, Canada T6G 2('2.

3"-Dcsaturasc converts 18:2(n - 6) to 18:3(n - 6) and is thought to be a rate-limiting step in the synthesis of arachidonic acid (20:4(n - 6)) from 18: 2(n - 6 ) [ 4 5]. 20:4(n - 6) is generally present in the sn - 2 position of membrane phospholipid, plays at1 important role for membrane physico-chemical properties and is a common precursor for synthesis of prostaglandlns, prostacyclin and thromboxanes, in the ainvay epithelium, prostaglandin production has been implicated as a mediator of ion transport, particularly CI- secretion [6-8]. Thus, whether or not the airway epithelium can utilize 18 : 2(n - 6) and synthesize 21): 4( n - 6) is critical for a variety of cell functions in this cell type. ..l~'-Desaturase activity is present in some tissues and cultured cells of mammalian origin [9,10] and within a given species, expression of A"-dcsaturase activity may vary. For example, .,l'"-desaturase activity is present in human lung fibroblasts [9,I0], but is evidently absent in cultured endothelial cells from human umbilical vein, which are incapable of converting [l-v~C]linoleic acid to arachidonic acid [11]. Furthermore, Aipert and

268 Walenga [12] recently postulated, in light of the low level of 2 0 : 4 ( n - 6 ) found in cultured human airway epithelial cells, that these cultured cells may lack A6desaturase. Should a deficiency in A6-desaturase activity in fact exist in human airway epithelial cells, it would imply that not only do these cells require a source of es.,;ential fatty acids, but such cells might also be dependent on an exogenous source: of 20 : 4(n -- 6). Whether human airway epithelial cells lack A6-de saturase a n d / o r AS-desaturase activity, remains to be determined by direct experiments with appropriate precursor fatty acids. Moreover, incorporation of essential fatty acid into these cell membrane lipids must be characterized. Understanding airway epithelial lipid metabolism may be of clinical importance in terms of dietary fal supplementation to patieltts with pulmonary diseases such as cajstic fibrosis. The objective of the present stud.v is to characterize incorporation and desaturation of 1 8 : 2 ( n - 6) in cultured human airway epithelial cells, to assess the turnover rate of this fatty, acid in epithelial phospholipids and to determine the effect of fatty acid supplementation on fatty acid composition of membrane phospholipid. Materials and M e t h o d s

This research was approved by the Research Ethics Board of the Faculty of Medicine, University of Alberta, Edmonton, Canada. Materials Radioactive materials, [1-14C]18 : 2(n - 6) (50 mCi/mmol), [1-n~C]I8: l(n - 9) (57 m C i / m m o l ) and [1-t4C]16:0 (56 m C i / m m o l ) of 98.5% radiochcmical purity and 99% chemical purity were purchased from NEN, Canada and used without further purification. Unlabelled fatty acids (18: 2(n - 6), I 8 : 3 ( n - 6), 20:3(n - 6), 20:4(n - 6)), lipid standards and other biochemicals were obtained from Sigma Chemical (St. Louis, MO). All solvents were redistiltcd before use. Cell isolation and culture Primary cultures of epithelial cells from human nasal turbinates were obtained by methods similar to those described by Yankaskas et al. [13]. Turbinates, received 1-4 h following surgery, were placed in Joklik's modified minimum essential medium (MEM) supplemented with antibiotics (100 U / m l penicillin G, 100 ~tg/ml streptomycin sulphate and 50 ~tg/ml gentamycin) at 4°C for transport from surgery. Tissues were transferred to a solution of 1 m g / m l Type 14 proteinase in MEM with the same antibiotics and digested overnight at 4°C with mild agitation. After raising the tempera-

ture to 37°C for I - 2 h in a 5% CO 2 incubator, the cell suspension was collected and 10% fetal bovine serum (FBS) was added to neutralize the proteinase. Cells were filtered through a 60 tzm Nitex mesh, centrifuged at 275 × g for 10 min and washed once in 10% FBSD M E / F I 2 . The resulting cells were suspended in culture medium ( D M E / F I 2 + hormones (insulin, 2 /xg/ml; transferrin, 7.5 ptg/ml; endothelial cell growth factor, 7.5 /.tg/ml; epidermal cell growth factor, I8 n g / m l ; hydrocortisone, 36 m g / m l ; triiodo-L-thyronine (T3), 2 n g / m l ) + antibiotics (cholera toxin, 10 n g / m l ; penicillin G, 60 /xg/ml; streptomycin sulfate, 100 ~ g / m l ; gentamycin, 5 0 / x g / m l ) + I% FBS [12,13]. Cells were plated on collagen (type VII, from rat tail)-coated plates, then cultured at 37°C, 5% CO 2, 98% relative humidity in a tissue culture incubator (Model 3173, Forma Scientific). The culture medium was changed every other day. Alter 5 - 7 days culture, cells had grown to confluence and were used for biochemical assay. Isotope incubation Confluent cells were detached from the plate surface with 0.25% trypsin solution. Cells were collected in a centrifuge tube, neutralized with 10% FBS, washed once with culture medium and suspended in fresh culture medium (without FBS) and counted in a hemocytometer ((3-5)-106 cells/culture with a purity of more than 90%), An aliquot was removed for analysis of protein [14] with bovine serum albumin as standard. The remaining cells were replated on 60 mm diameter plastic culture dishes at a density of ( 5 - 8 ) . 105 cells in 2 ml medium. Mixtures of l-n4c-labelled fatty acids and unlabelled fatty acids, suspended by sonication at 37°C in sterile 5% bovine serum albumin [15], were added in a 100 ttl vol. to give a final fatty acid concentration of 10 p.M and 2.2.106 DPM of ~'~C-labelled fatty acid and 40 ttM for unlabelled fatty acids (2.2- 107 D P M / n m o l fatty acids). Cells were incubated for various lengths of time under conditions described above. After incubation the cells were harvested with either plastic policemen or 1 ml of I).25% trypsin. Trypsin was used only when incubation was longer than 4 h in which a large population of the cells attached to the plates (recovery 99-100%). Culture medium was removed by centrifugation, cell pellets were washed twice with phosphatc-bttffered saline and used for lipid analysis. For turnover studies, cells were pulse-labelled with 1-14C-labelled fatty acid (18:2(n - 6)) (1/.tCi/dish, 30 ttM) for 2 h. After incubation, the radioactive medium was removed. Cells were washed once with 0.5% (w/v) bovine serum albumin in phosphate-buffered saline and were further incubated with fresh medium containing 100 p.M unlabeUed 1 8 : 2 ( n - 6) for various times. After the chase, cells were harvested and the medium

269 was removed by centrifugation. Cells were washed once and subjected to lipid analysis.

tion.~hip between desaturase activity and age of cell donors was analyzed by linear regression [22].

Lipid extraction and analysis

Results

Cell pellets and culture medium plus washed buffer were extracted with chloroform/methanol (2: 1, v/v) [16] containing 0.005% (w/v) butylated hydroxytoluene as antioxidant. Total phospholipid and triacylglyeerol were separated by thin-layer chromatography (TLC) on silica-gel G plates using a solvent system comprised of petroleum ether/diethyl ether/acetic acid (80:20:1 by vol.) [17], Indiviciual phospholipids were separated by TLC on silica-gel H plates using the following solvent system: chloroform/methanol/2-propanol/ 0.25% (w/v) KCl/triethylamine ( 3 0 : 9 : 2 5 : 6 : 1 8 by vol,) [18]. Fatty acids were methylated with 14% (w/w) BF3-methanol reagent and heated for 1 h at 100°C [19]. Fatty acid methyl esters were separated according to degree of unsaturation on silica-gel TLC plates impregnated with AgNO 3 to a final concentration of 30% (w/w) AgNO 3 in silica using h e x a n e / d i e t h y l ether/acetic acid (94:4:2 by vol.) for 35 rain and the same solvents (90:10:2 by vol.) as sequential solvent systems for 25 rain. Bands containing saturated, monoenoie, dienoic, trienoic and tetraenoic fatty acids could be identified by comparison with reference standards. Analysis of phosphoiipid fatty acid composition was carried out by gas-liquid chromatography using a fully automated Varian Vista 6000 GLC equipped with a flame-ionization detector [20]. The chromatography utilized a fused silica BP,o capillary column (25 M × 0.25 mm I.D.). Helium was used as the carrier gas at a flow rate of 1.8 m l / m i n using a splitless injection. The initial oven temperature was 150°C, increased to 190°C at 20°C/rain and held for 23 min, then increased to 220°C at 2°C/rain for a total analysis time of 40 rain. These analytical conditions separated aii saturated, mono-, di- and polyunsaturated fatty acids from C,4 to C25 carbons in chain length. A Varian Vista 402 data system was used to analyze area percent for all resolved peaks and to quantify sample size based on external standards when added. Quantity of phospholipids was determined by detection of phosphorus content using a modified Bartlett assay for microphosphorus analysis 1"21].

Liquid scintillation counting All labelled samples separated by TLC were scraped directly from plates into scintillation vials containing 5 ml of scintillation cocktail (Aquasol, NEN). Samples were counted in a Beckman LS-5801 liquid scintillation counter. Counting efficiency was more than 95%.

Statistical analysis Results are presented as means 4-S.D. Data were analyzed by analysis of variance procedures. The rela-

hworpomtion of labelled fatty acid m cells Cultured human airway epithelial cells efficiently i,',corporatcd 18:2(n - 6) and its desatur~tcd preducts into both the neutral lipids and phospho[ipids of the ceils, Following 4 h incubation the total amount of [ I- I~C118 : 2(n - 6) incoq:mrated into celluiar lipids was 32 _+5.6 nmol per mg protein of cu!tured cells with 65 +__9.5% in phospholipid and 31 ± 10% in neutral lipid. Fatty acid incorporation into cell membrane was time-dependent (Fig. IA). Within 4 h [l-t4C]18:2(n 6) incorporation into phospholipid increased linearly, followed by a relatively slower increase from 4 h to 8 h. During the first 2 h the amount of fatty acid incorporated into triacyiglycerol was the same as that incorporated into phospholipid. After 2 h the rate of [I~C]18: 2 ~ n - 6)incorporation into triacylglycerol was low (Fig. IA) Separation of total phospholipid into specific phospholipid fractions showed that PC contained most of the radioactivity and that the pattern of label incorporation into individual phospholipids was different over 8 h of incubation (Fig. IB). The radioactivity in PC increased rapidly, while that in PI and PE increased slowly. As a result, an alteration in distribution of labelled fatty acid among phospholipid fractions was seen during the incubation period (Fig. I B). Overall, the percentage of radioactivity in PC increased in the 8th h compared with the Ist h, from 65.5% to 76.5%, while the percentage of radioactivity in P! and PE decreased from 19.5% to 13% and 15% to 10.5%, respectively. To compare 18:2(n - 6) incorporation with non-essential fatty acid incorporation in the cells, [lCla]18: l(n - 9) or [1-'4C]16:0 was incubated with the cells under identical conditions used for incubation of [I-14C]18:2(n- 6). The amount of fatty acid incorporated into phospholipid was in the order: 1 8 : 2 ( n - 6) ( 16.4 + 0.5 nmol/mg protein) > 18: l(n - 9) ( 13.6 +_0.3 nmol/mg protein) > 16: 0 (1 !.2 ± 0.3 nmol/mg protein), suggesting that these cultured cells preferentially incorporate polyunsaturated fatty acid. To test the effect of other supplementation with polyunsaturated fatty acids on 18: 2(n - 6) incorporation in the cells, [I-14C]18:2(n-6) plus 50 # M of 2 0 " 5 ( n - 3 ) or 2 0 : 4 ( n - 6 ) was incubated with the cells. After 4 h incubation with 2 0 : 5 ( n - 3 ) , [1~ C ] 1 8 : 2 ( n - 6 ) incorporation into phospholipid and triacylglycerol was reduced by 72% and 16%, respectively, while 2 0 : 4 ( n - 6) supplementation reduced [lI'~C]18:2(n-6) incorporation into phospholipid by 15% but increased incorporation of I 8 : 2 ( n - 6 ) into

27(I

triacylglyccrol 0y 20%. These results are similar to those found in cultured human fibroblasts [23].

Desattt,'aEon and elongation of labelled fatty acid The convcrsion of [t-=4C]fatty acid to its desaturation and elongation products was determined by detec-

O'~

TABLE 1

~Tmvcr.~itm of hJbt'lh'd flirty acids to dcsatttrathm trod ehmgation pro&tots hy ctdtured epithelial cells ('ells were cultured for 5 - 7 days with MEM/FI_, + H o r m o n e s + lC; FBS and incubated with radiolabelled fatty acids 11 /.tCi. 1(~i1 nmoll for 4 h ill it similar medium without serum. Lipids were extracted and methylated. The labelled fatty acids were separated on AgNO3-TLC plates. Radioactivity was determined by scintill,:tion coumcr, n = cell cultures from different .-,ub.,ects. Substrate provided

Distribution of radioactivity ( r ; )

Fatty acid fraction

[ t- 14C] i8:2(n-61 (n=ll]

30

Saturated Monoenes Dienes Trienes Tetra,:nes Pentacnes L.._

'T--

0

2

1

F

I

4

6

8

B

20 ._o -~

m9

16

o .t~

++1+ 4 0

0

2

4

6

8

2O ~

15

"0

~

lO

.,....

t--

5 c 0

0

I

I

I

t

2

4

6

8

Incubation Time (hours) Fig. t. Time-courses of [ 1 - 1 4 C } 1 8 : 2 ( n - 6 ) incorporation into total ph',~spholipid (e-e) and triacylglycerol ( A - A ) (A); distribution of radiolahel aTntmg phospholipid fractions to-o), phosphatidylcholine: ( A- A ), pht~sphatidylinositol: ( • - • ) phosphatidylcth:mol:tmine (B): and convcrsi,n ~f [ I - 1 4 C ] 1 8 : 2 ( n - 6 ) to its products (C) in human ai~vay ephhelial cells. ('ells were incubated with I v.Ci [|14(']1 lq : 2( n - 6) ( 1011 nmol) for time period.,, :is indicated. Lipids were ext ratted, mcthylatcd (for fatty acid analysis only), separated on T L C plates and labcl-inc~rpt~rated was counted with u scintitla;,:,:'n counter. Values are mean _+S.D. of cells cultured from 4 (A), and 3 1(') individuMs, respectively. Values in Fig. IB represent average cells cultured from two different subjects; each vaht¢ was within 10'7/of the other.

71t + 1.2 19.8±2.6 9.0 + 2.fi 1.1 _-+i1.6

[ i- t4C_'] Ifl:(I (n=2)

[ I- t~c] 18.1(n-9) (n=2)

73 13.3 111.5 2.8

82 7J, 111.4

tion of 1-14C in fatty acid methyl esters. Conversion was compared to control samples in which cells werc heated in boiling water for 5 min before addition of labelled substrate fatty acid. it is apparent that human airway epithelial cells are capable of chain elongating/ desaturating [ 1-14C] 18 : 2(n - 6) to synthesize homologous trienes, tetraenes and pentaenes (Table I), indicating that the cells in culture contain active A "- and £-desaturasc activity. Produt.ts compatible with A4-desaturase activity or an alternate pathway [24,25] were also observed. After 4 h incubation approx. 8.2 nmol of total products/rag cellular protein, accounting for 2(130% of fatty acid incorporated, was formed from 18:2(n - 6). The major products were 20:3(n - 6) and 20:4(n - 6). with less 18:3(n - 6), suggesting that elongation from 18:3 to 20: 3(n - 6) occurs very quickly (Tab!e l). This finding is similar to that found in cultured gtioma cells [15]. The ratio of trienes Io tetracnes was about 2:1. A time-dependent increase in total desaturation products was observed during 8 h incubation, suggesting active desaturase activities (Fig. 1C). A linear relationship was apparent between cell donor's age and desaturase activity observed ( Y = 0.12X + 12.55, r = 0.89). The total desaturation products of 1 8 : 2 ( n - 6) decreased as the donor's age advanced. This result is consistent with previous findings in other cell types [23,26-28]. When compared to [ I - t ' : c ] 1 8 : 2 ( n - 6), the percent conversion rate ~ff [l-t4C]lS: 1(9) was relatively lower (Table !). When [l-t~C]16:0 was provided to the cells, 26% of [l-L4C]16:0 incorporated was converted to higher homologous fatty acids, the major product being 16: l, indicating that a high level of J%desaturase activity also occurs in the epithclial cells.

271 T A B L E Ii

Cells were incubated with [ I - H C ] I S : 2 ( n - 6 ) fi=r 4 h. Lipids were extracted and separated on silica-gel-tl plates. Fatty acid chains of lipid fractions ~¢re methylalcd, separated on AgNO3-TLC plates and counted. Values art: means 4- S.E. , = 5 cell cultures from different subjects. Radioactivity U'} )

Phosphatidylcholine Phosphatidylethanolamine Phosphatidylinngltt,I Triacylglycerol

- 6) was followed by chase periods of I-4 h. Following a 4 h chase the radioactivity in the major phospholipids (PC, PE and Pi) :;ignificantly decreased, particularly in PC, while the fatty acid label in phosphatidyIserine and triacylglycerol increased slightly (Fig. 2~. The label decrease in PC, P! and PE during the first 2 h of chase was faster than during the last 2 h (2-4 h). Accordingly, fatty acid label released to the medium increased with the decrease in phospholipids (Fig. 2). The half-life of fatty acyl chains in PC, Pi and PE was estimated to be approx. 5.5, 6 and 7.4 h, respectively. Analysis of lipid mass (phosphorus content) and label distribution in individual phospholipids showed that PC, PI, PE, and PS comprise 53.2%, 12%, 28.3%

[1-:4]!8:2(n

Distribt~tion of labelled faro' acid in irtdiridtt,t/ lipid .fractions of ctdtared epithelial cells

18:2(n --6)

Trienes

85.(J 4- 3.

I i .7 + 2.3

59. i + 7." 52.0 + 7.,'; 82.4 _ 7.:;

23.0 4- 3.8 31.5 ± 6.4 14.1) _+5.2

Tclraenes 2.6 + 0.5 15.7 4- 4.0 16.5 +_ 1.5 3.6 _+2.0

When supplemented to cul'.ure medium, 20 : 4(n - 6) and 20:5(n - 3) reduced conversion of [ 1-t4C]18 : 2(n - 6 ) to its products by 5(1% and 70%, respectively. This effect of 20:5(n - 3) on desaturation of 18: 2(n 6) has been suggested in previous studies using microsomes isolated from rats fed a fish-oil enriched diet [29,30] and using cultured human fibroblasts [23]. 2 0 : 4 ( n - 6 ) inhibition of 1 8 : 2 ( n - 6) conversion has been reported in other celt types [15,23,31].

Distribution of labelled fatty acyl chabt, hz bulit'idual lipid fractions The distribution of [1-t4C118:2 and its products among phospholipid fractions and triacylglycerol after incubation of 100 nmol [I-t'~C]18:2(n-6) with 5" 105 cells for 4 h is shown (Table !!). in PC and triacylglycerol, the major labelled fatty acid was 18 : 2(n - 6), whereas PI and PE exhibited relatively higher proportions of 20:4 and 20:3, suggesting that the more polyunsaturated fatty acids are preferentially incorporated into PI and PE.

Release of products formed from / 1- r~C/18: 2(n - O) Released fatty acids formed from [!-'4C118 : 2(n - 6) were determined by quantitation of labelled trienoic and tetraenoie fatty acids (18: 3, 20: 3, 20: 4) extracted from culture medium following 4 h incubation. Approx. 1,2 nmol of fatty acids (except 1 8 : 2 ( n - 6))/ms protein was released to culture medium, accounting for 10-15% of total products formed. The major fatty acid was 1 8 : 3 ( n - 6 ) (60%), while 2 0 : 3 ( n - 6 ) a n d 20:4(n - 6 ) accounted for 25% and 15%, respectively. Release of fatty acid was not due to cell death or injury since cell viability was found to be 98-I(X)% by trypan blue stain after incubation.

Tumot'er rate of phospholipid fatty acyl chain Turnover t,i phospholipid fatty acyl chains in cultured airway epithelial cells was dc~ermined by a pulse-chase experiment in which a 2 h pulse of

1,6

-

if=-

X ¢z

1.4

0

.t-a,,,../,.

1.2 I

E

13. Q

1.0 ! O.

._1 ¢..m

(D e~

0.3

._1

'~

o.2

..Q 0.1 .

B

_.___=.__..-i

¢t)

r',t

0

0

.=

I

I

I

i

1

2

3

4

_

Chase Time (h) Fig. 2. Distribution of radiolabel [I-14Clh"[:2(nl~6) in the lipid fractions of cultured airway epithelial ceils during chase with 18:2 (n-hi. Cells were pulse-labelled with [ ! - " = C ] 1 8 : 2 ( n - 6 ) ( 1 p.Ci/dish) for 2 h. Radioactive medium win; removed and cells were washed once with 0.5% b~wine serum albumin in phosph:,te-buffercd saline. Frt:sh medium containing Ill(I p.M unlabclled 18: 2(n - 6) was added and incubated for time periods as indic:ltcd. After the chase, harvested cells were washed once. Lipids were extracted and separateJ into neutral and phospholipid classes by TLC. Results tire expressed as total dpm in lipid f r a c t i o n / m s protein, e-e. phosphatidylehohne: & - A. phosphalidylinosilol; D- t'l. phogphalidylcthanolamine: ==-Ill. phosphatidylserine" o - o . triacylglyccroh ~ ' - v . culture medium. Values are mean of observations of cells cullured from fimr individuals.

272 TABLE tit

Fatty acid profile of phospholipids of cultured human airway ('pidtelial cells with attd without 18: 2(n - 6) supplementation Fatty acid

"

C o m p o s i t i o n (C4 w / w )

None " (n = 4) 16:11 16: I_ 18:0 18:!(n-91

15.8 4.5 13.4 33.()

_+2.6, _+1.2 +3.3 +5.1

18:1(n -7) 18:21n-6) 18;3(n -3) 20: I(n - 9 )

I1.0 _+_2.1 2,6 _i-0.2 11.57±0.1 0.74±0.t

Linolcic acid (18:2(n -6)) " (n = 3) 16.t _+1.2 5.! ±03 11.3_+0.2 23.5-4-2.0 * 10.I _+11.9 111.9_+11.7 * 0.3-t-0.1 -

20:31n-9) 20:3(n-6)

1.2 _+0.43 2.6 -I-0.4

0.2_+0.1 * 4.2_+ 1.5 *

20'.4(n - 6 1 20: 501 - 3) 22:4(n-61 22: 51n - 6)

4.3 1.4 0,5 0,25

6.1 __+I).3 * (1.2 + (I. l 1.3_+I).2 * 0.6 _+O, 15

22:51n - 31 22 .... a~

1.1 +_I).5 1.7 4-0.65

+ l.t) _=_0.3 __I),2 4- O. 1

I.I)+0.2 1.9+0.6

VSFA ~'MUFA

31,3 _+3.5 54.2 ± 5 . 9

29.4_+2,1 43.9-+3.5 *

Yn-6 2"n - 3

10.2 ±2.7 3.6 + !.4

23.5± !.2 * 3.3 -I-0.5

" Fatty acids analyzed included all saturated and unsaturated fatty acids of 16 to 22 carbons in chain length. h Growth medium containing no supplemental fatty acid. c Growth medium supplemented with 18:21n-6) (lift) /.tM). The cells were exposed to the 18:21n -6) for 4 h. XSFA = iolal saturated hffly acids; 2MUFA = total monounsalurated fatty acids: -vn - 6 = total n - 6 fatty acids; ~-'n- 3 = total n - 3 fatty acids. values are means_+S.D.. * P < I).115. n = cell cultures from different subjects. and 6.5% of total lipid mass, respectively, while accounting for 73%, 13%, 12% and 2%, respectively, of total labelled fatty acids incorporated during 4 h incubation. T h e s e data also suggest that the o r d e r of turnover rate of phospholipid fatty acid chains is PC > PI > PE > PS.

Effect of fatty acid supplementation on fatty acM composition To examine the effect of fatty acid s u p p l e m e n t a t i o n to culture m e d i u m on phospholipid fatty acid composition, the fatty acid profile of total phospholipids from cells incubated with or without 100 ~ M 1 8 : 2 ( n - 6 ) was analyzed (Table III). It is apparent that after fatty acid s u p p l e m e n t a t i o n the levels of both 18:2(n - 6) and 20: 4(n - 6) !ncreased significantly and are close to those found in freshly isolated cells [12]. Discussion

A l t h o u g h fatty acid metabolism of cultured mammalian cells has b e e n extensively investigated [5,11,32]

and the capability to incorporate and desaturate exogenous fatty acids ha '~ 6eel. de.~ons'.rated in many animal tissues [3,29,33] as well as culture cell lines [11t,23,3437], there is little information characterizing essential fatty acid metabolism in h u m a n epithelial cells. This study has clearly d e m o n s t r a t e d that h u m a n airway epithelial ceils in primary cultures are able to rapidly incorporate exogenous fatty acyl chains into cellular iipids and actively desaturate and elongate t h e m to higher polyunsaturated fatty acids; and fatty acyl chains of the epithelial phospholipid are in a dynamic state of rapid turnover. From the patterns of change in cellular incorporation of the [ 1 - n 4 C ] 1 8 : 2 ( n - 6 ) over time at least two distinctive phases for metabolism of label were identified. T h e first phase, during the initial 2 h after addition, is characterized by rapid maximum labelling of neutral lipid followed by slower labelling of p h o s p h o lipid. T h e second phase, from 3 - 8 h, is characterized by a slow, constant increase in labelling of triacylglycerol and a continuing increase in labelling of phospholipids, particularly PC. T h e s e results are similar to those in previous studies using different cell types and labelled fatty acids [38,39]. T h e pattern of incorporation observed in the second phase may result from a larger fatty acyl chain pool in phospholipid a n d / o r a shift of fatty acyl chains from triacylg!ycerol to phospholipid following longer incubation. Analysis of label distribution a m o n g phospholipids shows that the a m o u n t of 1 8 : 2 1 n - 6 ) incorporated into individual phospholipid classes in the cultured cells was in the order: PC > PI > PE. This result is similar to that found previously in the same cell type incubated with [ l - t 4 C ] 2 0 : 4 ( n - 6 ) [40], but different from that found in cells of neural origin (PC > P E > PI) [33]. T h e present study also d e m o n s t r a t e s that the composition of labelled fatty acid o f individual phospholipid classes is distinctive: PI and to a lesser extent PE has higher proportions of 2 0 : 4 ( n - 6 ) and 2 0 : 3 ( n - 6) c o m p a r e d to PC and triacylglycerol which have a higher proportion of 1 8 : 2 ( n - 6). T h e s e results indicate that dcsaturation products of 1 8 : 2 0 1 - 6) are preferentially incorporated into PI and PE. It is evident that h u m a n airway epithelial cells possess active A%, A"- and ,.lS-desaturases and perhaps A4-desaturases and thereby arc able to produce 2 0 : 4 ( n -6) and o t h e r fatty acids, Thus, the low level of 2 0 : 4 ( n - 6) in the cultured cells r e p o r t e d by Alpert et al. [12] should not be the result of a lack of A6-de saturase, but may be due to an insufficiency in substrate ( 1 8 : 2 ( n - 6 ) ) for A%desaturation. This is supported not only by the results of the assay of [ ~ C ] 1 8 : 2 1 n - 6 ) conversion, but also by analysis of fatty acid composition of m e m b r a n e phospholipid, which showed that 18 : 2(n - 6) s u p p l e m e n t a t i o n to culture m e d i u m increased both 18: 2(n - 6) and 20:41n -

273 6) to levels close to those f o u n d in fleshly isolated cells. It is notable that desaturation rate is age-related. Therefore, age could bc an important physiological factor affecting desaturase activity, thereby affecting

fatty acid composition of cell membrane. The presence of desaturase enzymes in the airway epithelium suggests that: (i) the cells have capability to produce a variety of polyunsaturated fatty acids which can bc esterified to p h o s p h o l i p i d s , thereby altering the composition of p h o s p h o l i p i d fatty acyl chains: and ~,21 pathologically induced alterations in both m e m b r a n e

fatty acid composition and eicosar, oid production in pulmonary diseases may, in part, be mediated by changes in desaturase activities, thereby resulting in a low level of a specific polyunsaturated fatty acid. Our results clearly show that 2 0 : 5 ( n - 3) (50 ~m) inhibited incorporation of radioactive linoleic aci