Research Paper Mediators of Inflammation, 7, 327–333 (1998)

We separately studied the antioxidant properties of propofol (PPF), Diprivan (the commercial form of PPF) and intralipid (IL) (the vehicle solution of PPF in Diprivan) on active oxygen species produced by phorbol myristate acetate (10(-6) M)-stimulated human polymorphonuclear leukocytes (PMN: 5 x 10(5) cells/assay), human endothelial cells (5 x 10(5) cells/assay) or cell-free systems (NaOCl or H2O2/peroxidase systems), using luminol (10(-4) M)-enhanced chemiluminescence (CL). We also studied the protective effects of Diprivan on endothelial cells submitted to an oxidant stress induced by H2O2/MPO system: cytotoxicity was assessed by the release of preincorporated 51Cr. Propofol inhibited the CL produced by stimulated PMN in a dose dependent manner (until 5 x 10(-5) M, a clinically relevant concentration), while Diprivan and IL were not dose-dependent inhibitors. The CL produced by endothelial cells was dose-dependently inhibited by Diprivan and PPF, and weakly by IL (not dose-dependent). In cell free systems, dose-dependent inhibitions were obtained for the three products with a lower effect for IL. Diprivan efficaciously protected endothelial cells submitted to an oxidant stress, while IL was ineffective. By HPLC, we demonstrated that PPF was not incorporated into the cells. The drug thus acted by scavenging the active oxygen species released in the extracellular medium. IL acted in the same manner, but was a less powerful antioxidant.


Introduction
Active ox ygen spec ie s (AOS) are involved in tissue injury assoc iate d w ith many acute inflammatory proce sse s including sepsis, acute lung injury and se vere trauma. 1 ,2 The liposoluble anaesthe tic drug, propofol (2,6-diisoprop ylphenol) (PPF), shares a similar structure w ith phenolic antiox idants like the endogenous a -tocophe rol (vitamin E) w hich has be en show n to prote c t c ellular membranes against lipop erox idation processes induce d by AOS. 3 PPF has be en demonstrate d to inhibit the in v itro lipoperox idation of microsomes and mitochondria isolated from rat live r, at clinic ally re levant dose s, 4,5 and in a similar w ay as a -tocophe rol. 6 In addition, it has be en re ported that rats re ce iving PPF have he patic microsomes more re sis tant to lipoperox idation . 7 By its antiox idant effects, PPF stabilize s the rat liver mitochondrial me mbrane s ex pose d to an ox idative stre ss, inhibitin g the Ca 2+ induce d perme abilization of mitochondria. 8 PPF has also be en de monstrate d to prote ct erythroc yte me mbrane s against free radicals produc ed by the rmic decomposition of an azo-compound 9 and to inhibit hyaluronan dep olyme rization by AOS produce d by x anthine ox idase ac tivity. 1 0 PPF c ould also attenuate ischaemia re perfusion injury. 11 ,12 Finally, w e have show n that PPF c ould inhibit the pe rox idation of a lipidic e mulsion induce d by ferryl, ox oferryl or hydrox yl radicals, and that its inhibiting e ffec t w as as effic ie nt as that of a -toc op herol. 13 The effe cts of clinically re le vant dose s of Dipri-van®, the comme rc ial form of PPF, on AOS produce d by stimulate d polymorphonuclear ne utrophils (PMN), have be e n investigate d by as sessing differe nt ce ll functions such as deformability, che motac tism, re spiratory burst, phagocytosis and bacte ricidal ac tivity. These studies yie lded contradic tory re sults. Dipri-van® has be e n re ported to inhibit the cell mobility, 1 4 the PMN chemotac tically induce d polarization, 1 5,1 6 the PMN hydrogen pe rox ide (H 2 O 2 ) production, 1 7,1 8 as w e ll as the c ell c ap ac ity to phagocyte and kill bacte ria such as Sta p hy lo c o c c u s a u re u s and Es c he richia co li. 1 9 Che miluminesc enc e studie s pe rforme d on isolate d PMN stimulate d by N-formyl me thionyl phenylalanine (FMLP) or by zymosan re ported that Dip rivan® inhibite d the production of the AOS. 2 0 - 2 3 In c ontrast, other studies have show n that Dip rivan® did not modify the re spiratory burst and the p hagocytosis activity of PMN. 24,25 In fac t, intralipid (IL), the solve nt of PPF in Dip rivan®, has be e n demons trated to inhibit the mobility, the re spiratory burst and the phagocytosis capacity of PMN. 1 4,1 8 ,1 9 ,2 6 There fore , it may be important to asse ss sep arately the e ffec ts of Dipri-van®, IL and PPF.
The pre sent study w as designe d to inve stigate se parately the e ffec ts of PPF (dilute d in an ap propriate medium that did not interfere w ith the action of PPF), Diprivan® and IL on the stimulation of PMN, and on the production of H 2 O 2 by c ulture d e ndothelial cells, using luminol-e nhanc ed che miluminesc enc e (CL). 27 Ce ll-fre e syste ms w he re CL re sulted from sodium hypochlorite (NaOCl) or from the re action of H 2 O 2 w ith perox idases w e re also studied. In addition, the effe ct of Diprivan® on cultured human e ndothelial c ells submitted to an ox idant stress w as also studied, to ex amine the pote ntial intere st of this age nt in clinic al situations charac te rize d by an ex ce ssive PMN stimulation leading to an e ndothelium ox idativ e stre ss. 1 ,2 Using HPLC te chnique , w e quantified the inc orporation of the drug into the PMN or endothelial ce lls to establish the ex tra-or intra ce llular me chanism of action of PPF.

Isolation of PMN
Human PMN w e re is olated from buffy coats of he althy donors (Blood Transfusion Ce ntre, Unive rsity Hospital of Liè ge) by de nsity gradie nt ce ntrifugation. Buffy coats w ere adde d w ith 1 volume of Polymorphp re p TM and ce ntrifuge d at room te mperature (500 3 g , 30 min). The supe rnatant w as diluted w ith 0.5 volume of 0.9% NaCl and ce ntrifuge d (1000 g 3 g , 20 min). The c olle cte d cells w e re w ashe d w ith a hypotonic solution (155 mM NH 4 Cl, 170 mM Tris-HCl, pH 7.4) to lyse re maining e rythrocyte s. After c entrifugation (800 3 g , 15 min), the c ollec te d c ells w e re w ashe d and adjuste d to 10 3 10 6 c ells/ml in NaCl 0.9% (stock solution). Viable c ells in this pre p aration, asse sse d by the ex clusion of trypan blue , w e re highe r than 98%.

Human endothelial cell culture
Endothelial ce lls w ere isolated from human umbilic al vein (HUVEC) by dispas e tre atment ac cording to Jaffe e t a l. 2 9 HUVEC w e re c ultured on 0.2% gelatine c oate d dishe s in M199 medium supp le mented w ith 10% heatinactivate d fetal c alf serum, 5% heat-inactivate d human serum, penic illin (100 U/ml), streptomycin (100 m g/ml), he parin (90 m g/ml) and ECGF (20 m g/ml). The c ells w e re used at passage 2.

Drug solutions
The comme rc ial form of PPF (Dip rivan® : PPF 10 mg/ ml) is formulated in intralipid (IL), a lipid ve hicle emulsion (10% soya be an emulsion, e gg phosphatides and glyce rol). Tests w e re performed separate ly w ith PPF, Diprivan® and IL. Diprivan® or e quivale nt volumes of IL w e re adde d to the re ac tion milie u and compared w ith re sults obtain ed in the abse nce of Dip rivan® and IL (c ontrol value taken as 100%). Be fore addition to the te st tube, pure PPF w as dissolve d in carbitol to obtain the appropriate conce ntration s. A volume of 5 m l of the drug solution w as alw ays adde d to a final volume of 0.5 ml of PBS. Control te sts w e re performe d w ith 5 m l of carbitol alone , and the re sulting CL value w as take n as 100% (control value ).

Chemiluminescence assays
Luminol enhanced chemilumine sce nce (CL) w as me asured in a Bio-Orbit 1251 Luminome te r. All the CL assays w e re performed in a final volume of 0.5 ml PBS. The CL value s w ere re c orded over time and computerized. Max imal light e mission (peak CL) and total emitte d light (area unde r the time curve) w ere cons idere d. Results w e re ex pre sse d as the pe rc entage inhibition of the c ontrol value .
PMN CL a s s a y CL w as me asured at 37°C in the pre senc e of 5 3 10 5 PMN (50 m l of the stock solution), luminol (10 -4 M) and appropriate c oncentrations of Diprivan®, IL or PPF dissolved in c arbitol. The re action w as started by addition of PMA (10 -6 M). The CL w as re corde d for 10 min. The re sults re pre se nt me ans ± SD of five indepe nde nt assays performe d w ith PMN isolate d from five diffe re nt buffy coats.

HUVEC CL a s s a y
Adhere nt HUVEC w ere w as hed three time s w ith PBS and gently scraped w ith a rubbe r polic eman into PBS. After c entrifugation, the ce lls w ere suspended in PBS (5 3 10 6 ce lls/ml). The as say w as pe rforme d at 37°C, in a final volume of 0.5 ml of PBS containin g 100 m l of the cell suspens ion (final numbe r of ce lls: 5 3 10 5 ), luminol (10 -4 M) and horseradish pe rox idase (HRP) (2.5 U/ml). The CL w as measure d for 20 min. The assays w ere re pe ated three times.

Na OCl CL a s s a y
The assay w as done in the pre senc e of luminol (5 3 10 -7 M). NaOCl (10 -5 M final c onc entration) w as injecte d and the pe ak value of CL w as immediately re c orded. Each assay w as done in triplicate and the ex pe riment w as re pe ated three times (n = 9).

Cytotoxicity assay on HUVEC
Cytotox icity w as assessed by me asuring the re lease of pre viously incorporated 5 1 Cr. Conflue nt HUVEC in six multiw e ll plates w e re labelled ove rnight by 10 m Ci/ml 51 Cr added in the c ulture me dium. HUVEC w ere w ashe d in HBSS to re move uninc orp orated 51 Cr, and then inc ubate d for 1 h in 1 ml of HBSS containin g MPO (5 m g), w ith or w ithout addition of Diprivan ® or IL. H 2 O 2 (10 -4 M) w as added to initiat e the enzymatic activity of MPO. After a further 2 h inc ubation at 37°C, the supernatants w ere c ollec te d and the cells w ere w ashe d three times w ith HBSS. Supe rnatant and w ashings w e re poole d and 5 1 Cr re le as e w as quantifie d by g c ounting. Ce lls w e re lysed in NaOH (1 N) and the intra ce llular 51 Cr w as counted. The pe rc entage of 5 1 Cr re lease w as c alculated for e ach te st c ondition. An index of c ytotox icity (IC) w as c alc ulated as de scribe d elsew he re . 28 The p rote ctive e ffec t of Diprivan® against the ox idativ e stress w as c alculated as follow s % prote ction = 100 3 1 1 -IC Dip rivan® IC stre ss 2 w here IC Dip r iv a n ® and IC s tr e s s w e re re spe ctively the IC obtain ed in the pre senc e or in the absenc e of Dip rivan®. The ex pe riment w as re p eated w ith HUVEC isolated from tw o differe nt donors (n = 9).

HPLC analysis of PPF incorporation into the cells
PMN or HUVEC (5 3 10 5 c ells) w ere incubate d w ith Dip rivan® or PPF (10 -4 M or 10 -3 M) for 1 h at 37°C in 1 ml of HBSS. Afte r incubation, the supe rnatant and the c ells w ere separate d and PPF w as ex tracte d by cyclohex ane afte r addition of thymol as inte rnal standard. 30 The solvent w as evaporate d to dryness at ambient te mperature under nitrogen. The re sidue w as dissolve d in HPLC mobile phase , acetonitrile /H 2 O/ trifluoroac etic ac id (600 :400 :1 by vol.), and submitted to HPLC analysis on RP-18 c olumn w ith UV dete ction. The ex pe riment w as re pe ated tw ice .

Statistical analysis
All the re sults w e re ex pre sse d in pe rc entage of control and pre sented as mean value s w ith the standard deviation. Statistic al analysis w as pe rforme d using the Student's t-te st. P< 0.05 w as considere d statistically signific ant. Dose -re sponse c urve s w ere fitted by re gression mode ls (Pearson's te st) using PPF conc entrations in logarithmic scale.

Effect of Diprivan®, IL and PPF on the CL of activated PMN
The effe cts of differe nt c oncentrations of IL and Dip rivan® on the CL of activated PMN are show n in Table 1. The perce ntage of inhibition w as variable, depending on the diffe re nt pre parations of PMN. How e ver in all case s, IL alone yie lded an inhibitory effect w hich did not differ significantly from the effect of Diprivan ®.
Pure PPF inhibited the CL of PMN in a dosedepende nt manner (Fig. 1), in the range of 5 3 10 -5 M (37.3 ± 6.3 perce ntage of inhibition) to 10 -3 M (93.5 An tio xida n t a ctivity o f pro po fo l, Dipriva n ® a n d in tra lipid Mediators of Inflammation · Vol 7 · 1998 329

Effects of Diprivan®, IL and PPF on the CL of HUVEC
IL and Diprivan® inhibited the HUVEC re sponse, Dip rivan® yelding a higher effe ct than IL (Fig. 2). For Dip rivan®, inhibition w as observed from 10 -6 M (25.8 ± 6.8 perce ntage of inhibition) to 10 -4 M (95.6 ± 1.1 perc entage of inhibition). A saturation e ffec t w as obtain ed w ith IL: inhibition of CL rapidly re ache d a plateau at 30.9 ± 0.2 perc entage of inhibition. Pure PPF inhibite d the CL of HUVEC in a dose-depe nde nt manner (Fig. 2) in the range of 2.5 3 10 -6 M to 10 -4 M (respe ctive ly 28.4 ± 5.8 and 96.0 ± 1.8 perce ntage of inhibition). For pure PPF and Diprivan® a line ar correlation w as found be tw ee n inhibition of CL and the drug conce ntration (respe ctive ly r 2 = 0.943 and 0.955). The PPF conc entrations re quired for CL inhibition of HUVEC w ere low er than for CL inhibition of PMN, but PMN produced more AOS than endothelial c ells: the CL of PMN w as 40 time s that of HUVEC.

Effects of Diprivan®, IL and PPF on the CL of NaOCl
The CL re sulting from NaOCl w as inhibited by Dip rivan® and IL, in a dose-depe nde nt manner (linear correlations w ith re spec tively r 2 = 0.953 and 0.940) (Fig. 3) but Diprivan® yielded the most inhibitin g effect. A signific ant inhibition (35.8 ± 1.0%) w as alre ady obtain ed w ith 3 3 10 -6 M of Diprivan®, w hile IL w as ine ffec tive . For pure PPF dissolved in c arbitol the inhibition of CL w as already significant for a PPF conce ntration of 5 3 10 -6 M (25.9 ± 6.7 perce ntage of inhibition) (Fig.  3). A linear c orre lation w as obtained be tw ee n inhibition of CL and the PPF (logarithmic scale) c onc entrations (r 2 = 0.914).

PPF incorporation into PMN and HUVEC
Small amounts of PPF w ere incorporate d into PMN. PMN incubated w ith 10 -3 M Dip rivan® or PPF in carbitol inc orporated re spec tively 2.5% and 5.1% of the drug; for 10 -4 M Diprivan® or PPF in carbitol, w e found re spe ctive ly 0% or 2.2% of inc orporation. No PPF inc orporation w as found into HUVEC.

Discussion
The pre sent study w as designe d to investigate se parately the e ffec ts of pure PPF, IL and Diprivan® on the production of AOS by PMN and endothe lial ce lls, and on cultured human e ndothelial c ells submitted to an ox idant stre ss. First, the study emphas ized the dosedepende nt inhibiting e ffec t of PPF (dilute d in carbitol) on the luminol-e nhance d CL produc ed by stimulated ne utrophils and e ndothelial ce lls. In addition, it show ed the capacity of PPF and Diprivan® to inhibit the chemilumine sce nce of NaOCl and of the H 2 O 2 / perox idase syste ms. The se e ffec ts c ould be attribute d to the non-hypnotic prope rties of Diprivan®. Ac cording to other re sults of this study, IL, the emulsifie d vehicle solution of PPF in Diprivan®, yielde d a low er, but significant and dose -depende nt inhibitin g e ffec t on the CL of HUVEC and cell-fre e syste ms. The HPLC study show e d that PPF did not enter the c ells. The CL inhibition c ould thus be attributed to a scavenging action on AOS, and not to an inhibition of the AOS production by the c ells.
Considering the production of AOS by stimulate d PMN, the inhibiting effe cts of Diprivan® w ere variable de pending on c ell batches and dose-depe nde nt re lationship c ould not be cle arly e stablishe d. The inhibiting e ffec t of IL te ste d se parately did not significantly differ from that of Diprivan®. This inhibitory effect of IL is quite difficult to ex plain. It might be due to the physical prope rties of the An tio xida n t a ctivity o f pro po fo l, Dipriva n ® a n d in tra lipid  emulsion produc ing a ne phe lome tric e ffec t in the re action vessel during the CL measureme nt; how ever, this is unlikely as the volumes of IL use d in our assays w ere too small to obtain this ne phe lometric effect. A che mical re ac tion of IL w ith the AOS is more likely: IL contains mainly triglyce rides w ith saturated and unsaturated fatty acids. The unsaturated lipids could act as AOS scavengers. IL could also act dire ctly on ce ll me mbranes, and induce structure alterations leading to a de cre ase of the AOS re le as e in the ex trace llular medium w ith a de cre ase of the CL re sponse. Inhibitory effects of IL on the func tions (chemotax is, phagoc ytosis, bacte ricidal ac tivity, supe rox ide anion gene ration) of PMN w e re previously re ported and often attributed to changes of the struc ture of the c ell membranes. 14 ,15 ,1 8 Prop ofol alone (dilute d in a non interfe ring solvant) had dosedepende nt inhibiting e ffec t on the CL re sponse of PMN and endothe lial c ells, and w as active at clinically re le vant c oncentration s (range: 10 -5 to 10 -4 M).

Mediators of Inflammation
The se ne w re sults c onfirme d our first observations of an in v itro antilipoperox idant effect of pure PPF 1 3 and w e re in agree ment w ith the data colle cte d from the lite rature , w hich demonstrate d the in vitro antiox idant and antiradical prop erties of PPF in ce llfree syste ms of e rythrocyte me mbrane, microsome and mitochondria membrane lipope rox idation. 4 -9 The e ffec ts of PPF on PMN functions like che motac tism, re spiratory burst and phagoc ytosis have be en ex te nsively investigated, 15 -1 9 but few data using the che miluminesc enc e te chnique are available . Four studie s 20 -23 re ported the inhibiting effe ct of Dipri-van® on the CL produc ed by N-formyl-me thionylleucyl-phenylalanine (FMLP) or zymosan-stimulate d PMN and on FMLP stimulate d PMN afte r priming w ith TNF-a . From these studies, it appe ared that IL w as less active than Diprivan®. 2 0 -22 Regarding the diffe re nt AOS producing syste ms in our study, the dosedepende nt inhibiting e ffec t of PPF dissolved in carbitol diffe re d from that observe d w ith Dip rivan®. In the majority of the ex p erimental c onditions, the effects of Diprivan® at low c onc entrations re fle cte d additiv e e ffec ts of IL and pure PPF. The perc entage of inhibition of Diprivan® at 10 and 30 m M PPF on the CL of NaOCl w as equivale nt to the sum of inhibition values measured w ith IL and PPF dissolve d in c arbitol. In the same w ay, considering the H 2 O 2 / HRP syste m, the CL inhibiting e ffec t of 10 m M Diprivan® w as the sum of the inhibition of IL and of PPF dissolve d in carbitol. In contrast, the CL assays p erformed on stimulated e ndothelial cells did not yie ld similar re sults. In these c onditions, the inhibitin g effe ct of inc re as ing c onc entrations of IL quickly re ache d a plateau. This plateau effec t re flects a lesser sc avenging activity of IL c ompared w ith PPF and a diffe re nc e in AOS production by HUVEC and PMN. Henc e, it is unlike ly that HOCl could be produc ed by HUVEC given the abse nce of MPO into the se c ells.
The ex perimental conditions using diffe re nt CL producing syste ms provide us w ith a partial understanding of the mechanism of action of pure PPF on the AOS. In fact, the CL of stimulate d PMN re flec ts the global production of AOS including supe rox ide anion (O 2 d ) and its de rivativ e hydrogen perox ide (H 2 O 2 ), hypochlorous acid (HOCl) produc ed by the e nzymatic activity of MPO, NO d , perox ynitrite (resulting from the re ac tion be tw e en O 2 d and NO d ) and other ex c ited ox yge n spec ie s such as singlet ox ygen and hydrox yl radic al. 31 PPF may ex e rt its inhibiting e ffec t on CL of PMN at differe nt le ve ls. It has alre ady be en show n to ne utralize hydrox yl radical. 13 From other studies, it appeared that PPF c ould affect the AOS produce d by the ac tivity of x anthin e ox idase , 10 and re store the re lax ing effe ct of aortic rings impaired by AOS. 3 2 Rec ently, PPF w as demons trated to be a sc avenge r of perox ynitrite. 33 ,3 4 The observations re porte d he re indicate that PPF also may ac t both on HOCl and H 2 O 2 , or dire ctly affect the ac tivity of perox idases. Its mechanisms of action on stimulate d endothelial c ells are still difficult to ac curately define; indee d, these c ells are able to produc e O 2 d and H 2 O 2 from intra ce llular x anthine ox idase, as w e ll as peroxynitrite from NO d .
Finally, the inhibiting property of PPF might be cons idere d as an unw anted side effe ct of the drug, sinc e it c ould re duce the phagocytosis activity of PMN. How ever in the ex p erimental conditions of the pre sent study, the PMN stimulation by PMA re sulte d in a degranulation process and the re lease of active enzymes and AOS outside the c ells, as demonstrate d by the measure ment of active granulocytic enzymes (MPO, elastase) in the re action milieu. Should the scave nging activity of PPF be direc te d only against the ex c ited ox ygen spe cie s re lease d outside the ce lls, PPF w ill be prote c tive against the ce llular and tissue destruc tion induced by an ox idative stress. Such a stress c ould be produc ed particularly by HOCl, a pote nt ox idant mole cule c ap able of re aching are as quite distant from its production site. By chromatographic analysis (HPLC), w e demonstrated that PPF does not enter the c ell (neither PMN nor endothelial ce lls) and there fore doe s not act intrac ellularly. More ove r, w e also de monstrate d that Diprivan® prote c ts endothe lial cells against an ox idativ e stre ss w hile intralipid alone w as w ithout effect, and w e re cently re porte d an inc re as ed antiox idant c apacity of plasma in patients anaesthe tize d w ith PPF. 3 5 The se re sults could suggest a pote ntial be ne fit of PPF in clinical situation s characte rize d by an ex ce ssive PMN activation in the absenc e of cytotox ic compounds re sulting from the re action of this age nt w ith AOS.
In conclusion, using the luminol-e nhanc ed chemiluminesc enc e te chnique , w e demonstrated that the inhibiting prope rty of PPF on the produc tion of AOS in diffe re nt ex perime ntal c onditions is also share d (but to a lesser degre e) by the lipidic e mulsion IL, w hich is the vehicle solution of PPF in Dip rivan®. Both the ac tive principle and the solve nt have dosedepende nt effec ts w hich are additive at low conc entrations. Diprivan ® w as also protectvie for e ndothelial ce lls submitte d to an ox idant stress. As PPF did not enter the cells, the drug w ould act by scave nging the active ox ygen spe cie s re le as ed in the ex tracellular me dium. IL w ould act in the same manne r, but w ith a low er antiox idant capac ity.