Lipid peroxide and transition metals are required for the toxicity of oxidized low density lipoprotein to cultured endothelial cells

155

Btochtmlca et Btophvslca Acta 1096 (1991) 155-161 ~'~ 1991 Elsevier Soence Publishers B V 0925-4439/91/$03 50 ADONIS 0925443991000612

BBADIS 61019

Lipid peroxide and transition metals are required for the toxicity of oxidized low density lipoprotein to cultured endothelial cells Masafuml Kuzuya, Mlchataka Nmto, Chmkl FunakL Toshm Hayashl, Kanlchl Asal and Fumlo Kuzuya Department of Geriatrics, Nagoya Unlt,ersltV School of Medlune, Nagoya (Japan) (Recewed 7 June 1990) (Revised manuscript received 29 October 1990)

Key words Oxidized low density hpoprotem Atherosclerosls, Llptd peroxldahon, Cvtotoxloty Transltmn metal, (Bovine aortic endothehal cell)

The toxicity of oxidized low density lipoprotein (Ox-LDL) to cultured vascular endothelial cells was investigated The modification of low density lipoprotein (LDL) by copper led to the production of thiobarbituric acid-reacting substance (TBARS) and lipid hydroperoxlde (LPO). TBARS was distributed not only in lipoprotein, but also in the aqueous phase, whereas LPO was observed only in the hpoprotein particle During the incubation of LDL with copper, the copper bound to lipoprotein and formed a complex. The toxicity of products resulting from the oxidation of LDL to endothelial cells was recognized in Ox-LDL particles, not in the aqueous phase. Following dialysis of Ox-LDL against EDTA, copper which had bound to the Ox-LDL particle was released and the toxicity of Ox-LDL disappeared. The addition of copper to the dialyzed Ox-LDL restored the cytotoxicity. To a lesser extent this effect was also observed with the addition of iron. A study of the time-course of LDL oxidation showed that the toxicity of Ox-LDL depends upon the level of LPO, not upon the content of TBARS, the extent of negative charge or the protein adduct of aldehydes. These results demonstrate that transition metal is required for Ox-LDL toxicity and that the toxic moiety of the products resulting from LDL oxidation is LPO associated with the Ox-LDL particle.

Introduction Recent studies suggest that the oxadative modification of low density hpoprotem (LDL) may play an important role in the inltxation and progression of atherosclerosis [1,2] L D L can be modified by incubating it with endothehal cells [3], smooth muscle cells [4,5] or m o n o c y t e s / m a c r o p h a g e s [6,7] in the presence of trace amounts of transition metals The oxidative modification of L D L ts variously associated with lipid peroxadatlon [8], an increase in net negative charge [4], hydrolysis of phosphohpld [8] and fragmentation of apoproteln B [8,9] Recently It has become apparent

Abbreviations LDL, low density hpoprotem, Ox-LDL, oxidized LDL TBARS, tluobarbltunc acid-reacting substance, DME, Dulbecco's modified Eagle's medmm, MEM, Eagle's minimum essenual medmm, LDH, lactate dehydrogenase, LPO, lipid hydroperoxlde, MDA, malondlaldehyde Correspondence M Kuzuya, Department of Geriatrics, Nagoya University School of Medicine, Tsuruma-cho, Showa-ku Nagoya 466, Japan

that this biological modification is mediated via a free radical-induced peroxldation of L D L which can be mamlcked by such transition metals as copper or iron in the absence of cells [8,10] Oxidative modification of L D L increases its uptake by macrophages through the scavenger receptor [11,12] and stimulates monocyte chemotaxls [13,14] Oxidized L D L (Ox-LDL) also exhibits toxicity to various cells including cultured vascular endothehal cells [15-17] Although the endothelial damage caused by O x - L D L would be relevant to atherogenesls, the detaded mechanism of cell injury induced by O x - L D L is unknown Previous studies suggest that the toxan of O x - L D L resides In the lipid phase [15] and that the susceptibility to O x - L D L depends upon the phase of the cell cycle [18] and on the level of lntracellular glutathione [19] It is"not clear, however, how Ox-LDL induces injury to the target cell. and what toxac substances may be contained m O x - L D L This study focuses on these questions using a cultured bovine aortic endothelial cell as the target cell The evxdence presented indicates that the toxic moiety of Ox-LDL is the hpld hydroperoxlde (LPO) assocmted wtth the Ox-LDL particle, and that such a transmon

156 metal as copper xs required for the induction of LPOdependent Ox-LDL toxicity to the target cells Materials and Methods

Matermls Dulbecco's modified Eagle's medmm (DME) and phenol red free Eagle's minimum essentml medium (MEM) were purchased from NlSSUl Pharmaceutical (Tokyo, Japan) EDTA was obtained from Katayama Chemicals (Osaka, Japan) and 2-thlobarblturlc acid from Wako Chermcals (Osaka, Japan) Bovine calf serum was obtained from Cell Culture Laboratories (Cleveland, U S A ) Flasks and 16 mm 24-well plates were obtained from Cornlng Glassworks (Cornmg, U S A )

Cell ~ulture Endothelial cells were estabhshed m culture from the thoracic aorta of the fetal calf as previously described [20] They were maintained In D M E supplemented with 10% (v/v) calf serum, penlcdhn, streptomycin and amphotencln m a humidified atmosphere containing 5% CO z and 95% air Cells at passage 9-13 were used for all experiments

Preparatwn of oxMtzed low densay hpoprotem LDL (density 1 019 to 1 063) was isolated from normal human plasma by ultracentrlfugatlon as previously described [21] The LDL preparation was dialyzed at 4 ° C for 48 h against two changes of at least 100 vols of 0 15 M NaC1 (pH 7 4) For the preparation of Ox-LDL, LDL was diluted wtth sahne to the indicated concentration and incubated with 20 /~M CuSO 4 at 3 7 ° C for specified intervals In some experiments, ahquots were then dialyzed against at least 100 vols of 0 15 M NaC1 (pH 74) for 24 h at 4 ° C or dialyzed against 015 M NaCl with 0 1 mM EDTA (pH 7 4) for 24 h followed by further dialysis against 0 15 M NaC1 (pH 7 4) for 24 h at 4°C to exclude EDTA

Gel ftltratlon of oxidized LDL 1 ml of sample was diluted with 1 ml of sahne and loaded onto a column of Sephadex G-25M (PD-10, Pharmacia, Uppsala, Sweden) and eluted with sahne Fractions of 1 ml vol were collected and assayed for protein, TBARS, LPO, copper and cytotoxlclty The content of protein, TBARS or LPO was recovered more than 90% following gel filtration

ktt (Kyowa Medex, Tokyo, Japan) according to the methods of Ohlsht et al [22-24] Cumene hydroperoxide was used as a standard The fluorescence mtensity of Ox-LDL was measured on a spectrofluorometer (Hitachi F-3010) with excitation at 360 nm, emission at 430 nm and excitation at 400 nm, emission at 470 nm as protein adducts of reactwe aldehydes [2]

Assessment of injury to endothehal cells Cellular injury was assessed by measurmg the amount of lactate dehydrogenase (LDH) released from the cells into the medium Endothelial cells at confluence m 24-well plates (approx 5 3 105 cells/well) were rinsed with MEM twice and incubated in a total vol of 0 4 ml of MEM containing 30% (v/v) O x - L D L for 16 h at 3 7 ° C except where noted MEM used in the present study lacks transition metals including copper In some experiments, CuSO 4 in stock solution (100/~M in MEM) was added to the test medium at the concentrations indicated in the text Following incubation, L D H activity in the medmm was determined by spectrophotometrlc analysis of N A D H oxidation (Shimadzu spectrophotometer UV-160) [25] Each L D H activity was compared with that released from cells following the addition of Triton X-100 for 0 1% ( v / v ) final concentration (% total L D H release)

Other methods For the determination of polyunsaturated fatty acids in Ox-LDL, hplds were extracted from the samples according to Follch et al [26] Llplds were sapomfied with methanohc potassium hydroxide followed by methylatlon with 14% ( w / v ) B F : m e t h a n o l reagent [27] Fatty acid methyl esters were analyzed on capillary gas chromatography (Hewlett-Packard, 5890A) equipped with a flame ~onlzation detector, using a 25 m × 0 2 mm Carbowax 20 M column (Hewlett-Packard) Protein was assayed by the method of Lowry [28] using bovine albumin as the standard Agarose gel electrophoresls of Ox-LDL was carried out as described by Nobel [29] The concentration of copper was determined by atomic absorption spectrophotometry [30] using a Hitachi Model 180-60 Results are the means of duphcate deterlmnations unless otherwise noted Statistical significance was determined by Student's t test Probability level was P < 0 001 Results

Assays for products of hptd peroxtdatton

Locahzatton of toxtc substance

The ttuobarblturlc acid-reacting substance (TBARS) in 100/~1 of the sample were measured fluorometrlcally as previously described [19] Malondialdehyde (MDA) formed from 1,1,3,3-tetramethoxypropane was used as a standard The hpld hydroperoxldes (LPO) m 100 /~1 of the sample were determined using the Determiner LPO

After incubating L D L with 20/~M CuSO 4 for 24 h, 1 ml of the sample was diluted with 1 ml saline and applied to the Sephadex G-25M column Fig 1B shows the elution patterns of protein, TBARS and LPO Two peaks of TBARS were found, the first was observed m fractions containing O x - L D L protein, and the other,

157 70

A m

50

I

Im

I

" ~ 30

10 0 10

B

~

8

)oo

parucles, an d that the p r e s e n c e of T B A R S in the aqueous phase is nontoxac to e n d o t h e l i a l cells T o evaluate the l o c a h z a u o n of copper, separate gel filtration was carried out to d e t e r m i n e the c o n c e n t r a t i o n of c o p p e r m each fraction Fig 2 shows the e l u u o n patterns of c o p p e r and p r o t e i n C o p p e r was f o u n d b o t h m h p o p r o t e m and in the a q u e o u s phase, suggesting that c o p p e r exists not only in ~ts free f o r m b u t also b o u n d to the O x - L D L p a r u c l e

Effect of dialysis on Ox-LDL toxlclt)'

6

i

'

~o,

,

,

~o~

~

E

o

0

2

3

4

5

6

7

Fract)on

8

9

10

11

12

13

A f t e r i n c u b a t i n g L D L with 20 / t M C u S O 4 for 24 h, the sample was dialyzed against 0 15 M N a C l c o n t a i n ing 0 1 m M E D T A at 4 ° C for 24 h followed by a further dialysis against 0 15 M NaC1 for 24 h to r e m o v e E D T A 1 ml of the s a m p l e was then d i l u t ed with 1 ml saline an d a p p h e d to the c o l u m n C o m p a r e d to F i g 1B, Fig 3C shows that the T B A R S p e a k n o t associated with the O x - L D L p r o t e i n d i s a p p e a r e d f o l l o w i n g dialysis, a nd that the peaks of T B A R S an d L P O associated with the O x - L D L p r o t ei n r e m a i n e d at the s a m e fractions as before dialysis M e a s u r e m e n t of the c o n c e n t r a t i o n of c o p p e r m each fraction d e m o n s t r a t e d an ab se nc e of

0

14

10 -

420

Number

Fig 1 Gel fdtrataon of oxidized LDL LDL (1 5 mg proteln/ml) was incubated with 20 p,M CuSO4 for 24 h at 37°C, then diluted with sahne (1 1, v/v) 2 ml of the sample was loaded onto Sephadex G-25M column and eluted with saline Fractions of 1 ml were collected and assayed for cytotoxacity (upper panel. A), protein (O), TBARS (zx) and hDd hydroperoxlde (LPO, o) (lower panel, B) To assess cytotoxaclty endothehal cells at confluence in 24-wells were incubated m MEM contalmng each fraction (40%, v/v) at 37 °C for 16 h LDH activity In the medium was then determined Results are expressed as descnbed in Materials and Methods Values in A represent the means of duphcate wells and in B represent the means of duphcate determmaUons that differed by no more than 5 and 9% respectwely

larger peak was seen in fractions not c o n t a l m n g protein, i n d i c a t i n g that T B A R S is not only present in h p o p r o tern, b u t also in the a q u e o u s phase T h e L P O an d p r o t e i n el u t i o n patterns coincided, no peak in L P O was o b s e r v e d that was n o t associated with O x - L D L protein, i n d i c a t i n g that in contrast to T B A R S , L P O was located in h p o p r o t e l n , n o t the a q u e o u s phase T o evaluate the cytotoXlClty of each fraction to cultured en d o t h el i al cells, the c o n f l u e n t cells in 24-well plates were ex p o s ed to the test m e d i u m consisting of M E M plus each fraction (40%, v / v ) for 16 h at 3 7 ° C F i g 1A shows that the extent of the toxicity of each fraction to the cells was c o r r e l a t e d with the c o n t e n t of p r o t e i n and L P O of each fraction T h e fractions cont a m i n g free T B A R S did n o t exhibit toxacity suggesting that the toxic substance is associated with the O x - L D L

I ~0

5 -

10

0..

• ' 1

2

3

4

5

6

"-o.. 7

8

9

10 11 12 13 14

Fract)on N u m b e r

Fig 2 Elutlon pattern of copper Samples were dduted with sahne (1 1, v/v) and 2 ml of ahquot was apphed to Sephadex G-25M column and eluted with saline Fractions of 1 ml were collected and assayed for protein and copper The samples were prepared as follows, LDL (1 1 mg protem/ml) was mcubated with 20/~M CuSO4 for 24 h at 37°C (o) then dialyzed agamst either 0 15 M NaCl for 24 h (zx) or 0 15 M NaCl containing 0 1 mM EDTA for 24 h followed by further dialysis against 0 15 M NaCl for 24 h ([3) The elutlon pattern of protein (e) was obtaaned from the sample which was prepared by Incubation of LDL with 20 /tM CuSO4 for 24 h at 37°C Other samples demonstrated the same elutlon pattern of protein (results not shown) Values represent the means of duphcate deterrmnatlons that differed by no more than 10%

158 70

O x - L D L r e m a m e d to the same extent as that of the n o n d i a l y z e d sample (data not shown)

A 50

50 "1"

~ 30

Requtrement of copper ]or Ox-LDL tO~ClClt)

...a --

30

~

0 I-

lo

o

i 2 3 4 5 6 7 891011121314

0 , ~ - - C - - - ~ , I , I , I . ~ I 0

C

8

100

I

(5

¢

~05

4

,

50 ~

C o

ca.

T o further c o n f i r m that the c o p p e r is requtred for the toxicity of O x - L D L to en d o t h el i al cells, we c o m p a r e d the cytotoxlctty of O x - L D L b e f o r e an d after dialysis (Fig 4) Before dialysis O x - L D L was toxtc to the cells O x - L D L dtalyzed against NaC1 w t t h o u t E D T A was observed to have a toxtc effect to the same extent as O x - L D L wtthout dialysis H o w e v e r , no toxic effect was observed with O x - L D L which had been dialyzed against E D T A followed by further dlalysts against N a C l . at least over the periods of time studied (16 h) T h e a d d i t i o n of various c o n c e n t r a t i o n s of C u S O 4 to the test m e d i u m c o n t a m m g O x - L D L dialyzed a g a m s t E D T A led to a dose-related m c r e a s e in the release of L D H C u S O 4 at the c o n c e n t r a t i o n s used in this study, had no effect on cell vlablhty

Effect of duration of LDL oxidation on to:ttcltv

.

.

.

.

.

0 I

2

3

4

5

6

7

Fract)on

8

9

10

11

12

13

14

Number

F,g 3 Gel filtration of ox,dlzed L D L following dlalysls against EDTA L D L (1 5 mg protem/ml) was incubated wlth 20 p M CuSO4 for 24 h at 37°C and dlalyzed against 0 15 M NaC] containing 0 1 mM EDTA for 24 h followed by further dlalysls aga,nst 0 15 M NaCI for 24 h The sample was then dduted wlth saline and apphed to the column The fraction (I m]) eluted with sahne was assayed as described m Fig 1 In the assay for cytotoxlclty endothelial cells at

confluence in 24-wells were incubated m MEM containing each fracUon (40%, v/v) m the absence (upper panel, A) or presence (reset of upper panel, B) of 2 /~M CuSO4 at 37°C for 16 h Values in A and B represent the means of duplicate wells and m C represent the means of duphcate determinations that differed by no more than 5 and 9%, respecuvelv

T o e x a m i n e the effects of v ar y i n g the level of oxidation of L D L on the toxicity to e n d o t h e l i a l cells, L D L was i n c u b a t e d with 20 # M C u S O 4 at 37 ° C for various intervals ( 0 - 4 8 h) S am p l es were then dtalyzed against 0 1 m M E D T A for 24 h and further dialyzed against 0 15 M NaC1 for 24 h D u r m g m c u b a t l o n with C u S O 4, O x - L D L showed an increased e l e c t r o p h o r e t l c m o b i l i t y

I00-

,

*

,

~

-1-*

ro

~ 50I-.-

c o p p e r from all fractions following dialysis agamst E D T A (F~g 2) E v a l u a t i o n of cytotoxaclty showed that the toxicity of those fractions which had exhibited a cytotoxac effect before dialysis d i s a p p e a r e d c o m p l e t e l y after dialysis against E D T A (Fig 3A) T o e x a m i n e w h e t h e r the disapp e a r a n c e of this toxicity w o u l d be referable to the r e m o v a l of c o p p e r by dialysis, a low c o n c e n t r a t i o n of C u S O 4 (2 p M ) was a d d e d to the test m e d m m As shown in Fig 3B, the effect of the fractions c o n t a l m n g O x - L D L on the toxaclty to endothelial cells was restored to a p a t t e r n similar to that p r e c e d i n g dialysis A f t e r i n c u b a t i n g L D L for 24 h with 20 p M C u S O 4, dtalysls of O x - L D L against 0 15 M NaC1 w i t h o u t E D T A for 24 h d e m o n s t r a t e d that c o p p e r r e m a i n e d In the h p o p r o t e m particle, b u t d i s a p p e a r e d f r o m the a q u e o u s phase (Ftg 2) Th e toxicity of the fractions c o n t a i n i n g

0

Fig 4 Effects of d]alysls and copper on toxloty of Ox-LDL L D H

release was determmed in the test medium after incubating endothehal cells m 24-wells with test medium for 16 h The test medium consisted of MEM containing Ox-LDL (30%, v/v) prepared by one of the following methods LDL (1 6 mg protem/ml) was incubated with 20 /~M CuSO4 for 24 h at 37°C (without &alysls) then &alyzed against either 0 15 M NaCI for 24 h (dialysis against NaC1) or 0 15 M NaCI containing 0 1 mM EDTA for 24 h followed by further dialysis agmnst 0 15 M NaCI for 24 h Test medium containing varying concentrations of CuSO4 was also assayed Results are expressed m Materials and Methods Values represent the means + standard deviation of tnphcate wells *P < 0 001 vs Cu 0/~M

159 300

8 LPO,~ <

Ull

E

~;

200

E

4

< I-

2

v

Start

100 ~

pos=tton

(~ (A)

(B)

(C)

(D) (E)

o

F~B 5 Mobd~ty on agarose gel electrophores~s of variously oxidized LDL LDL (1 5 mg p r o t e m / m l ) was incubated w~th 20 ~M CuSO 4 at 3 7 ° C for 0 h (A), 8 h (B), 16 h (C), 24 h (D) and 48 h (E) Each sample was dialyzed against 0 15 M NaC1 conta~mng 0 1 mM EDTA for 24 h followed by further dmlys~s against 0 15 M NaC1 for 24 h

on agarose gel electrophoresls (Fig 5), indicating an mcreased net negative charge during lncubaUon The lonetlcs of the various products m Ox-LDL resulting from hpld peroxadatlon, and the cytotoxaoty of Ox-LDL are shown m Fig 6 The TBARS content of Ox-LDL mcreased with an lncreasmg length of mcubatton w~th CuSO 4 up to 48 h (Fig 6A), while the extent of LPO was dlphas~c LPO increased up to 16 h and then decreased at 24 and 48 h (Fig 6A) The fluorescence lntensttles of Ox-LDL measured with exotatlon at 360 nm, emission at 430 nm and excltatton at 400 nm, emtss~on at 470 nm were also monitored as the formauon of protein adducts of reactwe aldehydes The formauon of these fluorophors mcreased during periods of mcubatlon up to 48 h (Fig 6B) Agam m the absence of CuSO 4, there was no release of L D H from cultured cells with Ox-LDL at any level of oxtdatton In the presence of 2/~M CuSO~, the extent of the toxicity of O x - L D L was d~phaslc, there was an mcreased release of L D H up to 16 h, and with mcubatton periods longer than 24 h, the extent of L D H release decreased (Fig 6C), correlated closely w~th the level of LPO (F~g 6A) These results suggest that the toxic substance m Ox-LDL ts LPO, not TBARS or protein adducts of aldehydes Table I shows the amount of unsaturated fatty acids in Ox-LDL prepared by incubation w~th CuSO 4 for 0, 24 and 48 h Increasing the period of exposure to CuSO 4 led to a decrease m the content of polyunsaturated fatty a o d s m hpoprotem

Discussion Although the tOXlCtty of Ox-LDL to a variety of cells is well documented [15-19], the detaded mechamsm of cell injury and the identity of the toxac substance are poorly understood It has been demonstrated that the toxic motety of Ox-LDL resides m the hpld phase [15] and that the suscepubdlty to Ox-LDL depends upon the phase of cell cycle [18] The receptor-medmted uptake of Ox-LDL appears not to be reqmred for cell tnjury, since Ox-LDL is also toxac to the cells winch lack a scavenger receptor [15-17] We recently found

3oo

~

~i

,

,

j

B

~

,

~

~oo

200

EE -- Lu

100

Ec10o

~ E~

~ × ~

~E , ~ o

0

70

0

~

C

0

8

16

24

Oxtdtzed LDL 0ncubatlon t~rne for preparation

48

hours)

F~g 6 Effect of the duration of LDL oxidation on the various products and their toxicity LDL (1 5 mg p r o t e m / m l ) was incubated with 20 p.M CuSO4 at 37 ° C for the indicated intervals, then dmlyzed against 0 15 M NaC1 containing 0 1 mM EDTA for 24 h followed by further dlalys~s against 0 15 M NaC1 for 24 h TBARS, LPO, fluorescence intensity and cytotoxloty were deterrmned as described m Materials and Methods In the assay for cytotoxxclty, the test medmm supplemented with 2 pM CuSO4 was also measured Values m A and B represent the means of at least duphcate determinations wtuch differed by no more than 5% Values m C represent the means of duphcate wells which differed by no more than 5%

TABLE I

Polyunsaturated fatty actds m oxz&zed L D L LDL (1 5 mg p r o t e m / m l ) was incubated with 20 p.M CuSO 4 at 3 7 ° C for the indicated intervals Samples were then dialyzed against 0 15 M NaCI containing 0 1 mM EDTA followed by further dialysis against 0 15 M NaCI for 24 h Polyunsaturated fatty acids m Ox-LDL were deterrmned as described m Materials and Methods Values represent the means of duphcate determinations which differed by no more than 5% Fatty a o d

Incubation t ~ m e

(/t g / m l )

0h

24 h

48 h

18 20 20 20 20

104 170 94 4 21 8 487

52 104 36 2 44 80

24 60 12 7 20 56

3 3 4 5 6

160 that the susceptlbdlty of endothelial cells to Ox-LDL is dependent upon the mtracellular glutathlone level [19], and that the hpad peroxldatlon of the cellular membrane is involved in the cellular injury mfhcted by Ox-LDL [311 In this study we showed that the oxidative mo&ficataon of LDL mediated by copper led to the production of TBARS m both hpoprotem and m the aqueous phase, consistent w~th the findings of other investigators [8,32] LPO was present only m hpoprotem, not m the aqueous phase, indicating that TBARS m the aqueous phase as not derived from the breakdown of peroxides during the acid-heating stage of the TBA assay TBARS m the aqueous phase is likely to consist of free M D A released from LDL fatty acids durmg the peroxidatlon process, as suggested by Esterbauer et al [33] that most M D A formed during oxidation of LDL does not remain associated with LDL Our observation that the toxic substance resides m the Ox-LDL particle, not in TBARS m the aqueous phase, agrees with the landing of Morel et al [34] We found that copper can band to Ox-LDL and form a complex Copper ions may bind to amino acids in hpoprotem during incubation with LDL, since copper ions readdy ligate to the amino groups of protein [35] Copper was released from the copper-Ox-LDL complex during dmlysls against EDTA, suggesting that copper bands loosely to Ox-LDL and that the copper moiety can easdy be released from the complex by a chelating agent, such as EDTA It ~s possible that the formation of the copper-hpoprotem complex may contribute not only to a change m conformaUon of hpoprotem, but also lead to changes in its biological properties Although the free non-bound copper m the aqueous phase was removed during the dialysis of Ox-LDL against a &alysate w~thout EDTA, the copper associated w~th the Ox-LDL particle, and the toxicity of Ox-LDL, both remained In contrast the toxicity of Ox-LDL &sappeared by removing the copper associated with the Ox-LDL particle during &alysls against EDTA Furthermore, the ad&t~on of copper to the test medium containing Ox-LDL which had been dialyzed against EDTA restored th~s toxicity, indicating that copper is revolved m the toxicity of Ox-LDL Replacing copper with iron (FeC13, FeSO4) m the test medmm containing Ox-LDL dmlyzed by EDTA also led to L D H release, although the effect of iron was weaker than copper (data not shown) In the study of the time-course of L D L peroxidatlon, we observed that TBARS content, electrophoretlc mobdlty and protein adducts of aldehydes increased during the incubation of LDL with copper It is well known that various aldehydes are formed from polyunsaturated fatty acids during the oxidation of LDL [33] Of those aldehydes, 4-hydroxynonenal is a candidate for a toxic moiety in Ox-LDL, since this aldehyde has been re-

ported toxic to certain cells [36-39] A hpophihc aldehyde, 4-hydroxynonenal remains associated with the Ox-LDL particle, generating fluorophor (excitation 360 nm/emlssion 430 nm) in Ox-LDL protein [2,33] M D A also reacts with hpoprotem, leading to a fluorescence maximum at excitation 400 nm/emlsslon 470 nm [2] However, it is unlikely that such aldehydes are responsible for the toxicity of Ox-LDL to endothelial cells, since in our results, the time-dependent increase in fluorescence of Ox-LDL was not consistent with the observed dlphaslc response of toxicity of O x - L D L The incubation of LDL with copper induced an initial increase in the LPO content of O x - L D L then subsequently decreased This finding agrees with the recent report of Esterbauer et al [40] The decrease in LPO is explained by our observation that the long-term incubation of LDL with copper led to a depletion of polyunsaturated fatty acids resulting from the ultimate destruction of polyunsaturated fatty acids side-chains by hpld peroxidation [41] In the absence of copper no toxic effect was detected with Ox-LDL In contrast, in the presence of copper the toxic effect of Ox-LDL appeared and was strictly dependent on the level of LPO, not on the TBARS content, extent of negative charge or the fluorescence intensity of Ox-LDL These results suggest that the toxicity of Ox-LDL is not dependent upon the extent of the oxidation of LDL, and that the most hkely candidate for the toxic substance m Ox-LDL as LPO We demonstrated that LPO and such transition metals as copper or ~ron are reqmred for the reduction of Ox-LDL cytotoxicity Pure LPO is fairly stable at physiological temperature However, in the presence of transition metals including iron and copper. LPO as readily decomposed by a redox mechanism that generates such toxic products as free radicals and reactive aldehydes [42,43] Therefore, the requirement of transition metal for LPO-dependent cytotoxlclty amphes that the toxic products generated from LPO during its decomposition would contribute to the lethal injury of cultured endothehal cells If significant levels of hydroperoxides pre-exist an the Ox-LDL exposed to prooxidizing con&tlons, decomposition of these hydroperoxides would play a major role an the initiation of lipid peroxidatlon in the cell membrane leading to cell death

References 1 Steinberg, D , Parthasarathy, S, Carew, T E , Khoo, J ( and Wltztum, J L (1989)N Engl J Med 320, 915-924 2 Jurgens G , Hoff H F , Chlsolm G M and Esterbauer, H (1987) Chem Phys Llplds 45, 315-336 3 Henrtksen T Mahoney E M and Steinberg, D ( 1 9 8 1 ) P r o c Natl Acad Scl USA 78, 6499-6503 4 Hennksen, T , Mahoney, E M and Steinberg D (1983) Arteriosclerosis 3, 149-159 5 Hemecke, J W , Rosen, H and Chalt, A (1984)J C h n Invest 74 1890-1894

161 6 Cathcart. M K . Morel. D W and Clusolm. G M (1985) J Leukocyte B,ol 38. 341-350 7 Parthasarathy. S. Prmtz. D J . Boyd. D . Joy. L and Steinberg D (1986) Arteriosclerosis 6. 505-510 8 Stembrecher. U P. Parthasarathy. S. Leake. D S. Wltztum. J L and Steinberg. D (1984) Proc Natl Acad Sc~ USA 81. 3883-3887 9 Fong. L G . Parthasarathy. S W~tztum J L and Steinberg. D (1987) J L,pld Res 28. 1466-1477 10 Hemecke. J W . Baker L. Rosen. H and Chalt. A (1986) J Chn Invest 77. 757-761 11 Sparrow. C P. Parthasarathy. S and Steinberg D (1989) J Blol Chem 264. 2599 2604 12 Aral. H. Klta. T. Yokode M . Narumlya. S and Kawal C (1989) Blochem Biophys Res Commun 159. 1375-1382 13 Qumn. M T . Parthasarathy S. Fong. L G and Steinberg. D (1987) Proc Natl Acad Scl USA 84. 2995-2998 14 Qumn. M T. Parthasarathy. S and Steinberg. D (1988) Proc Natl Acad Scl USA 85. 2805-2809 15 Hessler. J S. Morel. D W. Lewis. L J and Chtsolm. G M (1983) Arteriosclerosis 3. 215-222 16 Evensen. S A . Galdal. K S and Ndsen. E (1983) Atherosclerosls 49. 23-30 17 Morel. D W . Hessler. J R and Ch~solm. G M (1983) J Lipid Res 24 1070-1076 18 Kosug~. K . Morel. D W. D~Corleto. P E and Chlsolm. G M (1987) J Cell Physlol 130. 311-320 19 Kuzuya. M . Nalto. M Funakl. C Hayaslu. T. Asa,. K and Kuzuya F (1989)Biochem Biophys Res Commun 163. 14661472 20 Ryan. U S. Mortara. M and Whltaker. C (1980) Tissue and Cell 12. 619-635 21 Hatch. F T and Lees. R S (1968) Adv Lipid Res 6. 1-68 22 Ohlshl, N , Ohkawa, H , Muke, A Tatano, T and Yagl, K (1985) Blochem lnt 10, 205-211 23 Tatelchl, T, Yoshlrmne, N and Kuzuya, F (1987) Exp Geront 22, 103-111

24 Yagl. K . Kauchl. K . Salto. A . Kayahara. N . Tatano. T and Ohlshl. N (1986) Blochem Int 12. 367-371 25 Bergmeyer. H U . Bernt. E and Hess. B (1965) Methods of Enzymatic Analysis. pp 736-741 Academic Press. New York 26 Folch. J . Lees. M and Sloane-Stanley G H (1957)J BIol Chem 226. 497-509 27 Metcalfe. L D and Schmltz. A A (1961) Anal Chem 33. 363-364 28 Lowry O H . Rosebrough. N J Farr A L and Randall R J (1951) J Biol Chem 193. 265-275 29 Nobel. R P (1968) J Lipid Res 9. 693 700 30 Ichlda. T and Nobuoka. M (1969) Chn Chlm Acta 24. 299-303 31 Kuzuya. M Yamada. K Hayashl. T. Funakl. C. Na,to. M. Asal K and Kuzuya. F (1990)B,ochem Int 22. 567-573 32 Yokode. M . Kata T . Ktkawa. Y . Ogorochl. T . Narumtya. S and Kawal. C (1988)J Chn Invest 81. 720-729 33 Esterbauer H . Jurgens. G Quehenberger. O and Koller. E (1987) J Lipid Res 28. 495-509 34 Morel. D W. DICorleto P E and Ch~solm. G M (1984) Arteriosclerosis 4. 375-364 35 Lau S J and Sarkar. B (1981)B,ochem J 199. 649-656 36 Benedettl. A . Comport,. H and Esterbauer. H (1980) Blochlm B~ophys Acta 620 281-296 37 Krokan. H . Grafstrom. R C. Sundqwst. K . Esterbauer. H and Harris. C C (1985) Carcinogenesis 6. 1755-1759 38 Kaneko. T. Honda. S. Nakano. S and Matsuo. M (1987) Chem B~ol Interact 63 127-137 39 Kaneko. T. Kajl K and Matsuo. M (1988)Chem Blol Interact 67. 295-304 40 Esterbauer. H . Rotheneder. M D . Waeg. G . Stnegl. G and Jurgens. G (1990)Chem Res Toxacol 3. 77-92 41 Halhwell. B and Guttendge. J M C (1989) Free Radicals m B,ology and Medicine. pp 218-234. Clarendon Press. Oxford 42 Girottl. A W (1985) J Free Radical Blol Med 1.87-95 43 Halhwell. B and Guttendge J M C (1984)B.ochem J 219. 1-14