Clinical Science (1 979) 56,53-59
Fluorescent lipid-peroxidation products in synovial fluid
J . L U N E C A N D T. L. D O R M A N D Y Department of Chemical Pathology, Whittington Hospital, London
(Received 3 April 1978; accepted 14 August 1978)
Keele, Misra, Lehmeyer, Webb, Baehner & Rajagopalan, 1973; Klebanoff, 1975; Segal & Peters, 1976; Johnson & Lehmeyer, 1976) and as a consequence of tissue damage (Dormandy, 1978). However, established methods for detecting freeradical lipid-peroxidation products, in particular the thiobarbituric acid (TBA) reaction, have been of limited value in man. One reason may be that TBA-reactive compounds form non-reactive complexes (Schiff bases) with primary amino groups (Shin, Huggins & Carraway, 1972; Tappel, 1972; Bidlack & Tappel, 1973). These complexes can be identified only by their fluorescent spectral characteristics and other physical properties (Fletcher, Dillard & Tappel, 1973; Malshet & Tappel, 1973; Desai, Fletcher & Tappel, 1975; Trombly & Tappel, 1975). The present study was undertaken to establish whether these markers of free-radical activity might be demonstrable in synovial effusions.
Summary
1. Samples and extracts from synovial effusions were examined for the presence of fluorescent degradation products of free-radical oxidation (peroxidation). 2. Two classes of fluorescent compounds were identified and their fluorescence and physical properties are described. The fluorescence in the aqueous methanol phase of the extracts had the characteristics of conjugated Schiff bases. 3. Changes in synovial fluid and synovial fluid extracts in vitro induced by U.V. irradiation, prolonged incubation and the enzymic generation of oxidizing free radicals, with or without admixed peroxidizing arachidonic acid, and in the absence and presence of antioxidants, suggest that the fluorescent material is derived from peroxidized polyunsaturated lipids in damaged synovial tissue. Key words: fluorescence, free radical, lipidperoxidation products, synovial fluid.
Materials and methods
Abbreviation: TBA, thiobarbituric acid.
Synovialjluids Samples from 30 synovial fluids were examined. They were withdrawn from the knee-joints of patients with inflammatory or degenerative joint disease for diagnostic or therapeutic purposes. The patients were unselected, but fluids from patients with pyogenic arthritis and haemarthrosis were excluded, as were fluids from patients who had had previous intra-articular injections. All but five of the patients were on conventional analgesic-antiinflammatory therapy (dose ranges: ibuprofen, 50100 mg; salicylates, 2-6 g; indomethacin, 1-2 g;
Introduction
Much recent work suggests that free radicals may be generated from oxygen during inflammation both in phagocytosing cells (Babior, Kipness & Curnutte, 1973; Stossel, Mason & Smith, 1974; Shohet, Pitt, Baehner & Poplack, 1974; Salin & McCord, 1975: Yost & Fridovitch, 1974; Johnson, Correspondence: Dr T. L. Dormandy, Department of Chemical Pathology, Whittington Hospital, London N.19. 53
54
J. Lunec and T. L. Dormandy
methrazone, 0-6-1-2 g daily). Of these only salicylates have fluorescent characteristics somewhat similar to those of lipid-peroxidation products. For this reason fluids that gave a positive ferric chloride reaction for salicylates were also excluded. The samples were analysed within a few hours of collection or stored at -2OOC. Preliminary experiments showed that none of the characteristics studied changed significantly during storage. Chemicals Chloroform, methanol (spectroscopic grade), thiobarbituric acid (TBA) and butylated hydroxytoluene were obtained from British Drug Houses Ltd (Poole, Dorset, U.K.). Xanthine, xanthine oxidase, nitroblue tetrazolium and arachidonic acid (99% pure) were obtained from Sigma (London) Chemical Co., Kingston, Surrey, U.K. The xanthine oxidase was diluted to an activity of 0.25 i.u./ml with phosphate buffer [K,HPO,/KH,PO, (50 mmolh), pH 7.8, containing EDTA (sodium salt; 1 mmol/l)]. Arachidonic acid was sonicated (Rapidus Ultrasonics, Shipley, Yorks., U.K.) in phosphate-saline buffer [KH,PO,/K,HPO, (40 mmol/l)/NaCl (142 mmol/l), pH 7-41 to a final concentration of 3 mmol/l immediately before incubation. Butylated hydroxytoluene was added to the arachidonic acid emulsion before incubation with synovial fluid to a final concentration of 5 pmolll. Preparation of extracts Synovial fluid (500 pl) was added to 6 ml of chloroform/methanol (2 :1, v/v), which had previously been heated to 45OC for 10 min (Fletcher et al., 1973). Extracts were obtained by Vortex mixing for 2 min, centrifugation at 1000 g for 5 min, and removal of the protein precipitate. ‘Aqueous’ and ‘choroform’ phases were separated by the addition of 2 ml of deionized water to 5 ml of the chloroform/methanol extract and centrifugation for 5 min. Butylated hydroxytoluene was added to chloroform/methanol extracts stored at -2OOC to a final concentration of 2.5 mmol/l.
Fluorescence measurements All fluorescence spectral studies were performed on a Perkin-Elmer MPF-3L spectrofluorimeter (Perkin-Elmer, Beaconsfield, Herts., U.K.). The
instrument settings were: excitation and emission slits at 12 and 14 nm respectively; sensitivity settings from x 1 to x 30; wavelength calibration with quinine sulphate (10 mmol/l in 100 mmol/l of H,SO& fluorescent intensity calibration with a polymer standard (Perkin-Elmer block 5, compound 610, approximate concentration 5 x mol/l) arbitrarily defined as representing a fluorescent intensity of 100 units at x l sensitivity setting at its maximum emission of 475 nm when excited at 400 nm. All fluorescence measurements were performed at room temperature. Diene conjugation and absorbance at 270 nm Chloroform/methanol extracts of synovial fluids were analysed for changes in diene conjugation (absorbance at 234 nm) and absorbance at 270 nm as indicators of lipid peroxidation (Dodge & Phillips, 1966). All measurements were performed against a solvent blank in a Pye Unicam SP. 800 ultraviolet spectrophotometer (Pye Unicam, Cambridge, U.K.). Thiobarbituric acid reactivity The measurements were adapted from Dahle, Hill &. Holman (1962) and Stocks, Gutteridge, Sharp & Dormandy (1974). Synovial fluid (1 ml) was made to react with 1 ml of trichloroacetic acid (1 7 mol/l). The mixture was centrifuged at 2000 g for 10 min. Supernatant (1 ml) was heated to 100°C for 15 min with 0 - 5 ml of TBA solution [70 mmol of TBA/l in NaOH (50 mmol/l)]. The resulting mixture was cooled and centrifuged again. The absorbance of the supernatant was measured at 532 nm in a SP. 800 Pye Unicam spectrophotometer. Xanthinelxanthine oxidase effect The method was based on that described by Scudder, Stocks & Dormandy (1975). Portions of synovial fluid were diluted 1: 1 (v/v) with phosphate buffer, pH 7.8. Xanthine was added to 0.5 mmol/l final concentration. Xanthine oxidase, 200 pl, or deionized water, 200 pl, was added. The mixtures were incubated at 37OC for 10 min. In parallel series of mixtures mannitol was added, to a final concentration of 50 pmol/l. The generation of superoxide anion radicals was monitored by reduction with nitroblue tetrazolium (Scudder et al., 1975).
Synovialfluid lipids Results
Fluorescent complexes were demonstrable in 24 of the 30 synovial fluids examined. ,A composite fluorescence spectrum could be resolved into an aqueous phase and chloroform phase fluorescence by the addition of water. The bulk of the material showing the fluorescent characteristics of Schiff bases (excitation 355 nm; emission 450 nm) separated into the aqueous phase. On the scale defined by our fluorescence standard the range of intensities (emission peak heights) in this phase varied between 0 and 70 units. No attempt will be made in the present paper to relate these variations to clinical states or underlying pathological process, but a distinction will’ be made between ‘high-intensity’ fluids showning aqueous emission maxima between 50 and 70 units and ‘lowintensity’ fluids showing aqueous emission maxima between 0 and 20 units. The characteristic fluorescent spectra of the two phases of extracts from high-intensity fluids are shown in Fig. 1. Although some fluorescence of Schiff-base type was detectable in the chloroform
I
200
400
500
Wavelength (nm)
FIG. 1. Fluorescence spectra of the chloroform phase (--: excitation 320 nm;emission 395 nm) and aqueous phase (-: excitation 355 nm; emission 455 nm) of extracts of high-intensity synovial fluid. 5
phases, the main fluorescence in this phase had an emission peak in the U.V. range (395 nm) when excited at its excitation maximum (320 nm). This U.V. peak showed up as a ‘shoulder’ to the main peak (455 nm) in the aqueous phases when excited at the excitation maximum (355 nm). Both types of fluorescence spectra were demonstrable in corresponding extracts from peroxidizing arachidonic acid. Exposure of the chloroform/methanol extracts to U.V. irradiation led to an immediate increase in fluorescence. There was a ‘break’ in the rising curve after 3-4 min, which coincided with a marked decrease in the Stokes shift (from 355 nm/455 nm to 380 nm/425 nm) (Fig. 2 and Fig. 3). The effect of irradiation both before and after the break was significantly depressed by butylated hydroxytoluene (50 pmol/l). U.V. irradiation of the separated aqueous phases caused a linear increase in fluorescence. U.V. irradiation of the separated chloroform phases had no effect. U.V. irradiation of unextracted synovial fluids for up to 2 h did not increase the extractable fluorescence. The generation of fluorescent products and related changes in vifro were compared in (i) synovial fluid alone, (ii) synovial fluid to which peroxidizing arachidonic acid had been added, and (iii) peroxidizing arachidonic acid alone. In the chloroform/methanol extracts of peroxidizing arachidonic acid alone there was a marked increase in absorbance at both 234 and 270 nm. In extracts from the mixture of peroxidizing arachidonic acid and synovial fluid absorbance at 234 nm declined
0r 300
55
1
2
3
.4
5
6
7 46
9
1
0 0
Time of irradiation (min)
FIG. 2. U.v.-induced generation of fluorescence in the chloroform/methanol extracts of low-intensity synovial fluid (M), its inhibition by butylated hydroxytoluene (+-0) and the change in the Stokes shift (V-V).
J. Lunec and T. L. Dormandy
56
6oT-----50
2 3 4
1
0
5 6
7 8 9
loll
121314
I
Time of incubation (h)
200
300
400
500
600
FIG. 4. Fluorescence intensity changes in chloroform/methanol extracts (excitation 320 nm; emission 395 nm) from high-intensity synovial fluid (U---Q, high-intensity synovial fluid added to peroxidizing arachidonic acid (V---V) and peroxidizing arachidonic acid alone (o---o) during incubation at 37°C. Results are means f SEM of two experiments.
Wavelength (nm)
FIG. 3. Fluorescence spectral changes in chloroform/methanol extracts on exposure to U.V. irradiation. 0 time (-: excitation 360 nm; emission 440-460 nm); 6 min (---: excitation 375 nm; emission 445 nm); 8 min (---) (excitation 380 nm; emission 425 nm). slightly and absorbance at 270 nm showed no change. The fluorescence characteristics of the chloroform/methanol extracts of peroxidizing arachidonic acid and of the mixture of arachidonic acid and synovial fluid underwent similar changes: chloroform-extractable (u.v.) fluorescence declined (Fig. 4) and aqueous fluorescence increased (Fig. 5). The fluorescence characteristics of incubated synovial fluid alone did not change significantly. Table 1 shows the effect of butylated hydroxytoluene: this was most marked on the aqueous fluorescence of mixtures of synovial fluid and peroxidizing arachidonic acid after incubation. The characteristics of the fluorescence generated in the three systems are summarized in Table 2. No TBA-reactive material was present in any of the synovial fluids examined and none was generated either on exposure to U.V. light or on extended incubation. Fig. 5 shows the irregular and fluctuating increase in TBA-reactive material in incubated mixtures of synovial fluid and peroxidizing arachidonic acid. In peroxidizing arachidonic acid alone the increase was linear and correlated well with diene conjugation. Table 3 summarizes the effect of a free-radicalgenerating enzyme system on the endogenous lipids of synovial fluid. In contrast to U.V. ir-
8o
t
I
0
0
1
2
3 4 5 6
1
7 8 9 1011 1 2 1 3 1 4 1 5
Time of incubation (h)
FIG. 5. Changes in fluorescence in the aqueous phases (excitation 335 nm; emission 455 nm) of extracts from high-intensity synovial fluid (V---V), peroxidizing arachidonic acid added to high-intensity synovial fluid (+-a) and , peroxidizing arachidonic acid alone (0--0); L m corresponding changes in TBA reactivity in the mixture of peroxidizing arachidonic acid and high-intensity synovial fluid. Results are means f SEM of two experiments. radiation, the addition of such a system to unextracted fluid significantly increased extractable fluorescence. The effect was significantly inhibited by the hydroxyl radical scavenger mannitol (P < 0.001).
57
Synouialjluid lipids
TABLE1. Inhibition offluorescent-complexformation by butylated hydroxytoluene Emulsified arachidonic acid was added to low-intensity synovial fluid to a final concentration of 1.5 x mol/l (1). Butylated hydroxytoluene (BHT)was added to the mixture to a final concentration of 5 pnolll (2). The mixtures were incubated at 37OC for 75 h. The results are means f S.D. in three experiments. See the text for units. Before incubation
After incubation
Chloroform phase
Aqueous phase
Chloroformphase
Aqueous phase
1. Synovial fluid + peroxidizing
15.0 f 0.7
13.0 f 0.7
29.3 f 0.4
167.5 f 6.4
arachidonic acid 2. Synovial fluid + peroxidizing arachidonic acid + BHT
16.0 f 0.4
14.5 f 0.7
26.5 f 1.4
106.0 f 8.4
TABLE2. Characteristics of the fluorescent complexes in the aqueous phase of extracts from high-intensity synovial fluid before and after incubation with and without added arachidonic acid Phosphate-saline buffer (1) or emulsified arachidonic acid to a final concentration of 1-5 x mol/l (2) was added to low-intensity synovial fluid and incubated at 37OC for 75 h. The results are percentage changes as fluorescent intensities (means of three experiments) after: (a) addition of 10 pl of NaOH (4 mol/l) to 2 ml of extract; (b) re-neutralization with 10 4 of acetic acid (4 mol/l); (c)dialysis for 24 h against deionized water; (6)U.V. irradiation at 245 nm and 365 nm for 30 s. Ex., excitation; Em., emission. Fluorescence (nm) .
Before incubation
After incubation
I
Base
Acid
U.V.
Base
Ex.
Em.
(a)
(6)
(C)
ki)
(a)
Acid (6)
1. Synovial fluid
355
455
52
80
91
145
67
2. Synovial fluid +
355
455
60
75
80
160
360
460
120
130
14
129
peroxidizing arachidonic acid 3. Peroxidizing arachidonic acid
TABLE3. Effect of xanthinelxanthine oxidase system added to synovialfluid on extractable fluorescence Xanthme to a final concentration of 0.5 mmol/l (1) or xanthine to a final concentration of 0.5 mmolh and mannitol to a final concentration of 50 pmol/l (2)were added to 1 ml of high-intensity synovial fluid diluted with 1 ml of phosphate buffer. Xanthine oxidase (200 jd) or deionized water (200 pl) was added as indicated. The mixtures were incubated at 37OC for 10 min. The results are fluorescent intensities (mean f S.D. in six experiments) in the aqueous-phase extracts. (1) Synovial fluid (2) Synovial fluid + xanthine
(3) Synovial fluid + xanthine + mannitol
25.8 f 0.9 25.3 f 3.9 26.0 f 2.0
30.0 f 0.9 49.3 f 1.1 40.7 f 0.4
Discussion
For the purpose of the present study synovial effusions were regarded as products of a local inflammatory reaction whatever the destructive mechanism provoking the response. Although differences in composition, especially in protein and
Dialysis
Dialysis U.V.
(4
(a
87
93
142
55
83
86
153
140
170
17
145
trace-metal pattern, between fluids from ‘degenerative’ arthritis (or arthrosis) and ‘autoimmune’ or ‘inflammatory’ arthritis are well recognized (Scudder, White, McMurray & Dormandy, 1978) and reflect marked differences both in aetiology and natural history, both types of effusion have the basic characteristics of an inflammatory, i.e. reactive, exudate as distinct from a transudate (e.g. cardiac oedema fluid). Attempts to demonstrate lipid-peroxidation products in such material by means of the classical TBA reaction have been unsuccessful in the past, and the evidence pointing to the generation of oxygen-free radicals in synovial fluid has been indirect (McCord, 1974; PuigParellada & Planas, 1977). The findings reported in the present paper suggest that these products can be identified by their fluorescent spectral characteristics and that they are broadly divisible into two classes of compounds. A relatively non-polar class has a characteristic emission peak in the U.V. region (395 nm) when excited at 320 nm. The second, more polar class has the fluorescent characteristics
58
J. Lunec and T. L . Dormandy
of Schiff bases: an emission peak in the region of 460 nm when excited at 350 nm. Both types of fluorescence were demonstrable in the initial chloroform/methanol extracts of 80% of the unselected synovial effusions examined. Their presence and intensity has not yet been correlated with clinical findings. In extracts from fluids showing no initial fluorescence or only minimal fluorescence U.V. irradiation for 10 min generated a pattern indistinguishable from that shown by extracts of highintensity fluids. The changes induced by incubation at 37OC, with and without admixed peroxidizing arachidonic acid, by U.V. irradiation and by the enzymic generation of oxidizing free radicals point to a number of conclusions. They suggest that the fluorescent complexes in synovial fluid are largely derived from peroxidized polyunsaturated lipids. Lipid peroxides are known to undergo rapid secondary fragmentation (Dahle et al., 1962; Barber & Bernheim, 1967; Dormandy, 1969; Slater, 1972); both saturated and unsaturated aldehydes are formed (Gutteridge, Stocks & Dormandy, 1974), and some of these will readily form Schiff bases with primary amino groups (Shin et al., 1972; Fletcher et al., 1973; Malshet & Tappel, 1973; Bidlack & Tappel, 1973; Trombly & Tappel, 1975). Although the fluorescent spectral characteristics of peroxidizing arachidonic acid alone (which are not, of course, Schiff bases) are similar to the products of peroxidizing arachidonic acid/synovial fluid mixtures (Gutteridge, Lunec & Heys, 1978), the two groups of compounds possess contrasting physical properties. In particular the latter are non-dialysable and are susceptible to reversible quenching by alkali. The decrease in Stokes shift in irradiated chloroform/methanol extracts may reflect a process of polymerization (Undenfriend, 1969). These complexes may be formed largely from TBA-reactive peroxidation products, but they themselves no longer give the TBA reaction. Although the TBA reaction is therefore a sensitive and relatively specific measure of free-radical lipid oxidation in human tissue exposed in uitro to oxidant stress (Dormandy, 1971; Stocks, Offerman, Model1 & Dormandy, 1972; McMurray & Dormandy, 1974), it is less suited to the study of free-radical activity in uivo. It is possible that a number of secondary changes take place in free synovial fluid but both negative and positive evidence suggests that the initial free-radical oxidation products derive from inflamed or damaged synovial tissue. Indeed, formed synovial fluid has free-radical inhibitory
(antioxidant) properties (Scudder et al., 1978), probably a function of its caeruloplasmin content (Al-Timimi & Dormandy, 1978). To interpret the probable significance of these products it will clearly be necessary to establish their origin more precisely and to relate different patterns and concentrations to the associated or underlying pathological process (almost always modified in man by drug treatment). In view, however, of the extraordinarily high potency of many secondary products of peroxidized lipids in a wide range of biological systems (Schauenstein, 1967; Gutteridge, 1974; Gutteridge, Lamport & Dormandy, 1976; Barrowcliffe, Gutteridge & Dormandy, 1975; Williams & Peck, 1977) these free-radical reactions may be important regulators of the inflammatory process.
Acknowledgments
A project grant (G. 975/663/s) of the Medical Research Council is gratefully acknowledged. T. L. D. also wishes to acknowledge financial support by the Arthritis and Rheumatism Research Council.
References AL-TIMIMI,D.J. & DORMANDY, T.L. (1978) The inhibition of lipid autoxidation by human caeruloplasmin. Biochemical Journal, 168,283-288. BABIOR, B.M., KIPNES, R.S. & C U R N ~ EJ.T. , (1973) Biological defence mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. Journal of Clinical Investigation, 55 74 1-744. BARBER,A.A. & BERNHEIM, F. (1967) Lipid peroxidation: its measurement, occurrence and significance in animal tissue. Advances in Gerontological Research, 2,355. BARROWCLIFFE, T.W., GUITERIDGE, J.M.C. & DORMANDY, T.L. (1975) The effect of fatty-acid autoxidation products on blood coagulation. Thrombosis et Diathesis Haemorrhagica, 33,211-276.
BIDLACK,W.R. & TAPPEL,A.L. (1973) Fluorescent products of phospholipids during lipid peroxidation. Lipids, 8,203-207. DAHLE,L.K., HILL, E.G. & HOLMAN,R.T. (1962) The thiobarbituric acid reaction and the autoxidation of polyunsaturated fatty acid methyl esters. Archives of Biochemistry and Biophysics, 98,253-267. DESAI, I.D., FLETCHER,B.I. & TAPPEL,A.L. (1975) Fluorescent pigments from uterus of vitamin-E-deficient rats. Lipids, 10,307-309.
DODGE,J.T. & PHILLIPS,G.B. (1966) Autoxidation as a cause of altered lipid distribution in extracts from human red cells. Journal of Lipid Research, I , 387-394. DORMANDY, T.L. (1969) Biological rancidhication. Lancet, ii, 684-688.
DORMANDY, T.L. (1971) The autoxidation of red cells. British Journal of Haernatology, 20,457461. DORMANDY,T.L. (1978) Free-radical oxidation and antioxidants. Lancet, 1,647-650.
Synovialfluid lipids FLETCHER,B.L., DILLARD, C.J. & TAPPEL, A.L. (1973) Measurement of fluorescent lipid peroxidation products in biological systems and tissues. Analytical Biochemisfry, 32, 1-9. GUTTERIDGE, J.M.C., STOCKS,J. & DORMANDY, T.L. (1974) Thiobarbituric-acid-reacting substances derived from autoxidising linoleic and linolenic acids. Analytica Chimica Acta, 70,107-1 11. GUTITRIDGE, J.M.C. (1974) Consequences of lipid autoxidation in biological material. Ph.D.Thesis, London University. GUTTERIDGE, J.M.C., LAMPORT,P. & DORMANDY,T.L.(1976) The antibacterial effect of water-soluble compounds from autoxidising linolenic acid. Journal of Medical Microbiology, 4 105-1 10. GUTTERIDGE, J.M.C., LIJNEC,J., & HEYS, A. (1978) Fluorescence of peroxidised fatty acids. Analytical Lefters (In press). JOHNSON, R.B. & LEHMEYER, J.E. (1976) Elaboration of toxic oxygen products by neutrophils in a model of immune complex disease. Journal of Clinical Investigation, 51, 836841. JOHNSON, R.B., JR, KEELE,B.B., MISRA,H.O., LEHMEYER, J.E., K.V. (1975) WEBB,L.S., BAEHNER,R.L. & RAJAGOPALAN, The role of superoxide anion generation in phagocytic bactericidal activity. Journal of Clinical Investigation, 55, 1357. KLEBANOFF, S.J. (1975) Antimicrobial mechanisms in neutrophilic polymorphonuclear leukocytes. Seminars in Haemafology, 12,117-160. MCCORD, J.M. (1974) Free radicals and inflammation: Protection of synovial fluid by superoxide dismutase. Science, 185,529-530. MCMURRAY, W. & DORMANDY, T.L. (1974) Lipid autoxidation in human skeletal muscle. Clinica Chimica Acfa, 52, 105114. MALSHET, V.G. & TAPPEL, A.L. (1973) Fluorescent products of lipid peroxidation. Structural requirements of fluorescence ir, conjugated Schiff bases. Lipids, 8,194-198. PUIGPARELLADA, P. & PLANAS,J.M. (1977) Synovial fluid degradation induced by free radicals. In vifro action of several free radical scavengers and anti-inflammatory drugs. Biochemical Pharmacology, 27,535-537. SALIN, M.L. & MCCORD, J.M. (1975) Free radicals and inflammation. Journal of Clinical Investigation, 56, 13 191323. SCHAUENSTEIN, E. (1967) Autoxidation of polyunsaturated esters in water: chemical structure and biological activity of products. Journal of Lipid Research, 8,4 17.
59
SCUDDER,P., STOCKS,J. & DORMANDY,T.L. (1975) The relationship between erythrocyte superoxide dismutase and copper levels in normal subjects and in patients with rheumatoid arthritis. Clinica Chimica Acfa, 69,397-403. SCUDDER,P., WHITE, A.G., McMurray, W. & DORMANDY, T.L. (1978) Synovial fluid copper and related variables in rheumatoid and degenerative arthritis. Annals of fhe Rheumatic Diseases, 31,7 1-77. SEGAL, A.W. & PETERS,T.J, (1976) Characterisation of the enzyme defect in chronic granulomatous disease. Lancef, i, 1363-1365. SHIN,B.C.,HUGGINS, J.W. & CARRAWAY, K.L. (1972) Effects of pH concentration and aging on the malonaldehyde reactions with proteins. Lipids, 7,229-233. SHOHET,S.B., P m , J., BAEHNER,R.L. & POPLACK,D.G. (1974) Lipid peroxidation in the killing of phagocytised pneumococci. Infection and Immunity, 10,1321-1328. SLATER,T.F. (1972) Free-Radical Mechanisms in Tissue Injury. Pion, London. STOCKS,J., GUITERIDGE,J.M.C., SHARP,R.J. & DORMANDY, T.L. (1974) Assay using brain homogenate for measuring the antioxidant activity of biological fluids. Clinical Science and Molecular Medicine, 41,215-222. STOCKS,J., OFFERMAN,E.L., MODELL,C.B. & DORMANDY, T.L. (1972) The susceptibility to autoxidation of human red cell lipids in health and disease. British Journal of Haematology, 23,7 13-724. STOSSEL,T.P., MASON,R.J. & SMITH,A.L. (1974) Lipid peroxidation by human blood phagocytes. Journal of Clinical Investiga lions, 54,638-645. TAPPEL, A.L. (1972) Lipid peroxidation and fluorescent molecular damage to membranes. In: Pathological Aspects of Cell Membranes, pp. 145-172. Ed. Trump, B. F. & Arstila, A. Academic Press, New York. TROMBLY,R. & TAPPEL, A.L. (1975) Fractionation and analysis of fluorescent products of lipid peroxidation. Lipids, 10,441-450. UNDENFRIEND, S. (1969) Fluorescence Assay in Biology and Medicine, vol. 11. Academic Press, New York. WILLIAMS,T.J. & PECK, M.J. (1977) Role of prostaglandin mediated vasodilatation in inflammation. Nafure (London), 270,530-532. YOST,F.J., JR & FRIDOVITCH, I. (1974) Superoxide radicals and phagocytosis. Archives of Biochemisfry and Biophysics, 161, 395-401.
Fluorescent lipid-peroxidation products in synovial fluid
J . L U N E C A N D T. L. D O R M A N D Y Department of Chemical Pathology, Whittington Hospital, London
(Received 3 April 1978; accepted 14 August 1978)
Keele, Misra, Lehmeyer, Webb, Baehner & Rajagopalan, 1973; Klebanoff, 1975; Segal & Peters, 1976; Johnson & Lehmeyer, 1976) and as a consequence of tissue damage (Dormandy, 1978). However, established methods for detecting freeradical lipid-peroxidation products, in particular the thiobarbituric acid (TBA) reaction, have been of limited value in man. One reason may be that TBA-reactive compounds form non-reactive complexes (Schiff bases) with primary amino groups (Shin, Huggins & Carraway, 1972; Tappel, 1972; Bidlack & Tappel, 1973). These complexes can be identified only by their fluorescent spectral characteristics and other physical properties (Fletcher, Dillard & Tappel, 1973; Malshet & Tappel, 1973; Desai, Fletcher & Tappel, 1975; Trombly & Tappel, 1975). The present study was undertaken to establish whether these markers of free-radical activity might be demonstrable in synovial effusions.
Summary
1. Samples and extracts from synovial effusions were examined for the presence of fluorescent degradation products of free-radical oxidation (peroxidation). 2. Two classes of fluorescent compounds were identified and their fluorescence and physical properties are described. The fluorescence in the aqueous methanol phase of the extracts had the characteristics of conjugated Schiff bases. 3. Changes in synovial fluid and synovial fluid extracts in vitro induced by U.V. irradiation, prolonged incubation and the enzymic generation of oxidizing free radicals, with or without admixed peroxidizing arachidonic acid, and in the absence and presence of antioxidants, suggest that the fluorescent material is derived from peroxidized polyunsaturated lipids in damaged synovial tissue. Key words: fluorescence, free radical, lipidperoxidation products, synovial fluid.
Materials and methods
Abbreviation: TBA, thiobarbituric acid.
Synovialjluids Samples from 30 synovial fluids were examined. They were withdrawn from the knee-joints of patients with inflammatory or degenerative joint disease for diagnostic or therapeutic purposes. The patients were unselected, but fluids from patients with pyogenic arthritis and haemarthrosis were excluded, as were fluids from patients who had had previous intra-articular injections. All but five of the patients were on conventional analgesic-antiinflammatory therapy (dose ranges: ibuprofen, 50100 mg; salicylates, 2-6 g; indomethacin, 1-2 g;
Introduction
Much recent work suggests that free radicals may be generated from oxygen during inflammation both in phagocytosing cells (Babior, Kipness & Curnutte, 1973; Stossel, Mason & Smith, 1974; Shohet, Pitt, Baehner & Poplack, 1974; Salin & McCord, 1975: Yost & Fridovitch, 1974; Johnson, Correspondence: Dr T. L. Dormandy, Department of Chemical Pathology, Whittington Hospital, London N.19. 53
54
J. Lunec and T. L. Dormandy
methrazone, 0-6-1-2 g daily). Of these only salicylates have fluorescent characteristics somewhat similar to those of lipid-peroxidation products. For this reason fluids that gave a positive ferric chloride reaction for salicylates were also excluded. The samples were analysed within a few hours of collection or stored at -2OOC. Preliminary experiments showed that none of the characteristics studied changed significantly during storage. Chemicals Chloroform, methanol (spectroscopic grade), thiobarbituric acid (TBA) and butylated hydroxytoluene were obtained from British Drug Houses Ltd (Poole, Dorset, U.K.). Xanthine, xanthine oxidase, nitroblue tetrazolium and arachidonic acid (99% pure) were obtained from Sigma (London) Chemical Co., Kingston, Surrey, U.K. The xanthine oxidase was diluted to an activity of 0.25 i.u./ml with phosphate buffer [K,HPO,/KH,PO, (50 mmolh), pH 7.8, containing EDTA (sodium salt; 1 mmol/l)]. Arachidonic acid was sonicated (Rapidus Ultrasonics, Shipley, Yorks., U.K.) in phosphate-saline buffer [KH,PO,/K,HPO, (40 mmol/l)/NaCl (142 mmol/l), pH 7-41 to a final concentration of 3 mmol/l immediately before incubation. Butylated hydroxytoluene was added to the arachidonic acid emulsion before incubation with synovial fluid to a final concentration of 5 pmolll. Preparation of extracts Synovial fluid (500 pl) was added to 6 ml of chloroform/methanol (2 :1, v/v), which had previously been heated to 45OC for 10 min (Fletcher et al., 1973). Extracts were obtained by Vortex mixing for 2 min, centrifugation at 1000 g for 5 min, and removal of the protein precipitate. ‘Aqueous’ and ‘choroform’ phases were separated by the addition of 2 ml of deionized water to 5 ml of the chloroform/methanol extract and centrifugation for 5 min. Butylated hydroxytoluene was added to chloroform/methanol extracts stored at -2OOC to a final concentration of 2.5 mmol/l.
Fluorescence measurements All fluorescence spectral studies were performed on a Perkin-Elmer MPF-3L spectrofluorimeter (Perkin-Elmer, Beaconsfield, Herts., U.K.). The
instrument settings were: excitation and emission slits at 12 and 14 nm respectively; sensitivity settings from x 1 to x 30; wavelength calibration with quinine sulphate (10 mmol/l in 100 mmol/l of H,SO& fluorescent intensity calibration with a polymer standard (Perkin-Elmer block 5, compound 610, approximate concentration 5 x mol/l) arbitrarily defined as representing a fluorescent intensity of 100 units at x l sensitivity setting at its maximum emission of 475 nm when excited at 400 nm. All fluorescence measurements were performed at room temperature. Diene conjugation and absorbance at 270 nm Chloroform/methanol extracts of synovial fluids were analysed for changes in diene conjugation (absorbance at 234 nm) and absorbance at 270 nm as indicators of lipid peroxidation (Dodge & Phillips, 1966). All measurements were performed against a solvent blank in a Pye Unicam SP. 800 ultraviolet spectrophotometer (Pye Unicam, Cambridge, U.K.). Thiobarbituric acid reactivity The measurements were adapted from Dahle, Hill &. Holman (1962) and Stocks, Gutteridge, Sharp & Dormandy (1974). Synovial fluid (1 ml) was made to react with 1 ml of trichloroacetic acid (1 7 mol/l). The mixture was centrifuged at 2000 g for 10 min. Supernatant (1 ml) was heated to 100°C for 15 min with 0 - 5 ml of TBA solution [70 mmol of TBA/l in NaOH (50 mmol/l)]. The resulting mixture was cooled and centrifuged again. The absorbance of the supernatant was measured at 532 nm in a SP. 800 Pye Unicam spectrophotometer. Xanthinelxanthine oxidase effect The method was based on that described by Scudder, Stocks & Dormandy (1975). Portions of synovial fluid were diluted 1: 1 (v/v) with phosphate buffer, pH 7.8. Xanthine was added to 0.5 mmol/l final concentration. Xanthine oxidase, 200 pl, or deionized water, 200 pl, was added. The mixtures were incubated at 37OC for 10 min. In parallel series of mixtures mannitol was added, to a final concentration of 50 pmol/l. The generation of superoxide anion radicals was monitored by reduction with nitroblue tetrazolium (Scudder et al., 1975).
Synovialfluid lipids Results
Fluorescent complexes were demonstrable in 24 of the 30 synovial fluids examined. ,A composite fluorescence spectrum could be resolved into an aqueous phase and chloroform phase fluorescence by the addition of water. The bulk of the material showing the fluorescent characteristics of Schiff bases (excitation 355 nm; emission 450 nm) separated into the aqueous phase. On the scale defined by our fluorescence standard the range of intensities (emission peak heights) in this phase varied between 0 and 70 units. No attempt will be made in the present paper to relate these variations to clinical states or underlying pathological process, but a distinction will’ be made between ‘high-intensity’ fluids showning aqueous emission maxima between 50 and 70 units and ‘lowintensity’ fluids showing aqueous emission maxima between 0 and 20 units. The characteristic fluorescent spectra of the two phases of extracts from high-intensity fluids are shown in Fig. 1. Although some fluorescence of Schiff-base type was detectable in the chloroform
I
200
400
500
Wavelength (nm)
FIG. 1. Fluorescence spectra of the chloroform phase (--: excitation 320 nm;emission 395 nm) and aqueous phase (-: excitation 355 nm; emission 455 nm) of extracts of high-intensity synovial fluid. 5
phases, the main fluorescence in this phase had an emission peak in the U.V. range (395 nm) when excited at its excitation maximum (320 nm). This U.V. peak showed up as a ‘shoulder’ to the main peak (455 nm) in the aqueous phases when excited at the excitation maximum (355 nm). Both types of fluorescence spectra were demonstrable in corresponding extracts from peroxidizing arachidonic acid. Exposure of the chloroform/methanol extracts to U.V. irradiation led to an immediate increase in fluorescence. There was a ‘break’ in the rising curve after 3-4 min, which coincided with a marked decrease in the Stokes shift (from 355 nm/455 nm to 380 nm/425 nm) (Fig. 2 and Fig. 3). The effect of irradiation both before and after the break was significantly depressed by butylated hydroxytoluene (50 pmol/l). U.V. irradiation of the separated aqueous phases caused a linear increase in fluorescence. U.V. irradiation of the separated chloroform phases had no effect. U.V. irradiation of unextracted synovial fluids for up to 2 h did not increase the extractable fluorescence. The generation of fluorescent products and related changes in vifro were compared in (i) synovial fluid alone, (ii) synovial fluid to which peroxidizing arachidonic acid had been added, and (iii) peroxidizing arachidonic acid alone. In the chloroform/methanol extracts of peroxidizing arachidonic acid alone there was a marked increase in absorbance at both 234 and 270 nm. In extracts from the mixture of peroxidizing arachidonic acid and synovial fluid absorbance at 234 nm declined
0r 300
55
1
2
3
.4
5
6
7 46
9
1
0 0
Time of irradiation (min)
FIG. 2. U.v.-induced generation of fluorescence in the chloroform/methanol extracts of low-intensity synovial fluid (M), its inhibition by butylated hydroxytoluene (+-0) and the change in the Stokes shift (V-V).
J. Lunec and T. L. Dormandy
56
6oT-----50
2 3 4
1
0
5 6
7 8 9
loll
121314
I
Time of incubation (h)
200
300
400
500
600
FIG. 4. Fluorescence intensity changes in chloroform/methanol extracts (excitation 320 nm; emission 395 nm) from high-intensity synovial fluid (U---Q, high-intensity synovial fluid added to peroxidizing arachidonic acid (V---V) and peroxidizing arachidonic acid alone (o---o) during incubation at 37°C. Results are means f SEM of two experiments.
Wavelength (nm)
FIG. 3. Fluorescence spectral changes in chloroform/methanol extracts on exposure to U.V. irradiation. 0 time (-: excitation 360 nm; emission 440-460 nm); 6 min (---: excitation 375 nm; emission 445 nm); 8 min (---) (excitation 380 nm; emission 425 nm). slightly and absorbance at 270 nm showed no change. The fluorescence characteristics of the chloroform/methanol extracts of peroxidizing arachidonic acid and of the mixture of arachidonic acid and synovial fluid underwent similar changes: chloroform-extractable (u.v.) fluorescence declined (Fig. 4) and aqueous fluorescence increased (Fig. 5). The fluorescence characteristics of incubated synovial fluid alone did not change significantly. Table 1 shows the effect of butylated hydroxytoluene: this was most marked on the aqueous fluorescence of mixtures of synovial fluid and peroxidizing arachidonic acid after incubation. The characteristics of the fluorescence generated in the three systems are summarized in Table 2. No TBA-reactive material was present in any of the synovial fluids examined and none was generated either on exposure to U.V. light or on extended incubation. Fig. 5 shows the irregular and fluctuating increase in TBA-reactive material in incubated mixtures of synovial fluid and peroxidizing arachidonic acid. In peroxidizing arachidonic acid alone the increase was linear and correlated well with diene conjugation. Table 3 summarizes the effect of a free-radicalgenerating enzyme system on the endogenous lipids of synovial fluid. In contrast to U.V. ir-
8o
t
I
0
0
1
2
3 4 5 6
1
7 8 9 1011 1 2 1 3 1 4 1 5
Time of incubation (h)
FIG. 5. Changes in fluorescence in the aqueous phases (excitation 335 nm; emission 455 nm) of extracts from high-intensity synovial fluid (V---V), peroxidizing arachidonic acid added to high-intensity synovial fluid (+-a) and , peroxidizing arachidonic acid alone (0--0); L m corresponding changes in TBA reactivity in the mixture of peroxidizing arachidonic acid and high-intensity synovial fluid. Results are means f SEM of two experiments. radiation, the addition of such a system to unextracted fluid significantly increased extractable fluorescence. The effect was significantly inhibited by the hydroxyl radical scavenger mannitol (P < 0.001).
57
Synouialjluid lipids
TABLE1. Inhibition offluorescent-complexformation by butylated hydroxytoluene Emulsified arachidonic acid was added to low-intensity synovial fluid to a final concentration of 1.5 x mol/l (1). Butylated hydroxytoluene (BHT)was added to the mixture to a final concentration of 5 pnolll (2). The mixtures were incubated at 37OC for 75 h. The results are means f S.D. in three experiments. See the text for units. Before incubation
After incubation
Chloroform phase
Aqueous phase
Chloroformphase
Aqueous phase
1. Synovial fluid + peroxidizing
15.0 f 0.7
13.0 f 0.7
29.3 f 0.4
167.5 f 6.4
arachidonic acid 2. Synovial fluid + peroxidizing arachidonic acid + BHT
16.0 f 0.4
14.5 f 0.7
26.5 f 1.4
106.0 f 8.4
TABLE2. Characteristics of the fluorescent complexes in the aqueous phase of extracts from high-intensity synovial fluid before and after incubation with and without added arachidonic acid Phosphate-saline buffer (1) or emulsified arachidonic acid to a final concentration of 1-5 x mol/l (2) was added to low-intensity synovial fluid and incubated at 37OC for 75 h. The results are percentage changes as fluorescent intensities (means of three experiments) after: (a) addition of 10 pl of NaOH (4 mol/l) to 2 ml of extract; (b) re-neutralization with 10 4 of acetic acid (4 mol/l); (c)dialysis for 24 h against deionized water; (6)U.V. irradiation at 245 nm and 365 nm for 30 s. Ex., excitation; Em., emission. Fluorescence (nm) .
Before incubation
After incubation
I
Base
Acid
U.V.
Base
Ex.
Em.
(a)
(6)
(C)
ki)
(a)
Acid (6)
1. Synovial fluid
355
455
52
80
91
145
67
2. Synovial fluid +
355
455
60
75
80
160
360
460
120
130
14
129
peroxidizing arachidonic acid 3. Peroxidizing arachidonic acid
TABLE3. Effect of xanthinelxanthine oxidase system added to synovialfluid on extractable fluorescence Xanthme to a final concentration of 0.5 mmol/l (1) or xanthine to a final concentration of 0.5 mmolh and mannitol to a final concentration of 50 pmol/l (2)were added to 1 ml of high-intensity synovial fluid diluted with 1 ml of phosphate buffer. Xanthine oxidase (200 jd) or deionized water (200 pl) was added as indicated. The mixtures were incubated at 37OC for 10 min. The results are fluorescent intensities (mean f S.D. in six experiments) in the aqueous-phase extracts. (1) Synovial fluid (2) Synovial fluid + xanthine
(3) Synovial fluid + xanthine + mannitol
25.8 f 0.9 25.3 f 3.9 26.0 f 2.0
30.0 f 0.9 49.3 f 1.1 40.7 f 0.4
Discussion
For the purpose of the present study synovial effusions were regarded as products of a local inflammatory reaction whatever the destructive mechanism provoking the response. Although differences in composition, especially in protein and
Dialysis
Dialysis U.V.
(4
(a
87
93
142
55
83
86
153
140
170
17
145
trace-metal pattern, between fluids from ‘degenerative’ arthritis (or arthrosis) and ‘autoimmune’ or ‘inflammatory’ arthritis are well recognized (Scudder, White, McMurray & Dormandy, 1978) and reflect marked differences both in aetiology and natural history, both types of effusion have the basic characteristics of an inflammatory, i.e. reactive, exudate as distinct from a transudate (e.g. cardiac oedema fluid). Attempts to demonstrate lipid-peroxidation products in such material by means of the classical TBA reaction have been unsuccessful in the past, and the evidence pointing to the generation of oxygen-free radicals in synovial fluid has been indirect (McCord, 1974; PuigParellada & Planas, 1977). The findings reported in the present paper suggest that these products can be identified by their fluorescent spectral characteristics and that they are broadly divisible into two classes of compounds. A relatively non-polar class has a characteristic emission peak in the U.V. region (395 nm) when excited at 320 nm. The second, more polar class has the fluorescent characteristics
58
J. Lunec and T. L . Dormandy
of Schiff bases: an emission peak in the region of 460 nm when excited at 350 nm. Both types of fluorescence were demonstrable in the initial chloroform/methanol extracts of 80% of the unselected synovial effusions examined. Their presence and intensity has not yet been correlated with clinical findings. In extracts from fluids showing no initial fluorescence or only minimal fluorescence U.V. irradiation for 10 min generated a pattern indistinguishable from that shown by extracts of highintensity fluids. The changes induced by incubation at 37OC, with and without admixed peroxidizing arachidonic acid, by U.V. irradiation and by the enzymic generation of oxidizing free radicals point to a number of conclusions. They suggest that the fluorescent complexes in synovial fluid are largely derived from peroxidized polyunsaturated lipids. Lipid peroxides are known to undergo rapid secondary fragmentation (Dahle et al., 1962; Barber & Bernheim, 1967; Dormandy, 1969; Slater, 1972); both saturated and unsaturated aldehydes are formed (Gutteridge, Stocks & Dormandy, 1974), and some of these will readily form Schiff bases with primary amino groups (Shin et al., 1972; Fletcher et al., 1973; Malshet & Tappel, 1973; Bidlack & Tappel, 1973; Trombly & Tappel, 1975). Although the fluorescent spectral characteristics of peroxidizing arachidonic acid alone (which are not, of course, Schiff bases) are similar to the products of peroxidizing arachidonic acid/synovial fluid mixtures (Gutteridge, Lunec & Heys, 1978), the two groups of compounds possess contrasting physical properties. In particular the latter are non-dialysable and are susceptible to reversible quenching by alkali. The decrease in Stokes shift in irradiated chloroform/methanol extracts may reflect a process of polymerization (Undenfriend, 1969). These complexes may be formed largely from TBA-reactive peroxidation products, but they themselves no longer give the TBA reaction. Although the TBA reaction is therefore a sensitive and relatively specific measure of free-radical lipid oxidation in human tissue exposed in uitro to oxidant stress (Dormandy, 1971; Stocks, Offerman, Model1 & Dormandy, 1972; McMurray & Dormandy, 1974), it is less suited to the study of free-radical activity in uivo. It is possible that a number of secondary changes take place in free synovial fluid but both negative and positive evidence suggests that the initial free-radical oxidation products derive from inflamed or damaged synovial tissue. Indeed, formed synovial fluid has free-radical inhibitory
(antioxidant) properties (Scudder et al., 1978), probably a function of its caeruloplasmin content (Al-Timimi & Dormandy, 1978). To interpret the probable significance of these products it will clearly be necessary to establish their origin more precisely and to relate different patterns and concentrations to the associated or underlying pathological process (almost always modified in man by drug treatment). In view, however, of the extraordinarily high potency of many secondary products of peroxidized lipids in a wide range of biological systems (Schauenstein, 1967; Gutteridge, 1974; Gutteridge, Lamport & Dormandy, 1976; Barrowcliffe, Gutteridge & Dormandy, 1975; Williams & Peck, 1977) these free-radical reactions may be important regulators of the inflammatory process.
Acknowledgments
A project grant (G. 975/663/s) of the Medical Research Council is gratefully acknowledged. T. L. D. also wishes to acknowledge financial support by the Arthritis and Rheumatism Research Council.
References AL-TIMIMI,D.J. & DORMANDY, T.L. (1978) The inhibition of lipid autoxidation by human caeruloplasmin. Biochemical Journal, 168,283-288. BABIOR, B.M., KIPNES, R.S. & C U R N ~ EJ.T. , (1973) Biological defence mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. Journal of Clinical Investigation, 55 74 1-744. BARBER,A.A. & BERNHEIM, F. (1967) Lipid peroxidation: its measurement, occurrence and significance in animal tissue. Advances in Gerontological Research, 2,355. BARROWCLIFFE, T.W., GUITERIDGE, J.M.C. & DORMANDY, T.L. (1975) The effect of fatty-acid autoxidation products on blood coagulation. Thrombosis et Diathesis Haemorrhagica, 33,211-276.
BIDLACK,W.R. & TAPPEL,A.L. (1973) Fluorescent products of phospholipids during lipid peroxidation. Lipids, 8,203-207. DAHLE,L.K., HILL, E.G. & HOLMAN,R.T. (1962) The thiobarbituric acid reaction and the autoxidation of polyunsaturated fatty acid methyl esters. Archives of Biochemistry and Biophysics, 98,253-267. DESAI, I.D., FLETCHER,B.I. & TAPPEL,A.L. (1975) Fluorescent pigments from uterus of vitamin-E-deficient rats. Lipids, 10,307-309.
DODGE,J.T. & PHILLIPS,G.B. (1966) Autoxidation as a cause of altered lipid distribution in extracts from human red cells. Journal of Lipid Research, I , 387-394. DORMANDY, T.L. (1969) Biological rancidhication. Lancet, ii, 684-688.
DORMANDY, T.L. (1971) The autoxidation of red cells. British Journal of Haernatology, 20,457461. DORMANDY,T.L. (1978) Free-radical oxidation and antioxidants. Lancet, 1,647-650.
Synovialfluid lipids FLETCHER,B.L., DILLARD, C.J. & TAPPEL, A.L. (1973) Measurement of fluorescent lipid peroxidation products in biological systems and tissues. Analytical Biochemisfry, 32, 1-9. GUTTERIDGE, J.M.C., STOCKS,J. & DORMANDY, T.L. (1974) Thiobarbituric-acid-reacting substances derived from autoxidising linoleic and linolenic acids. Analytica Chimica Acta, 70,107-1 11. GUTITRIDGE, J.M.C. (1974) Consequences of lipid autoxidation in biological material. Ph.D.Thesis, London University. GUTTERIDGE, J.M.C., LAMPORT,P. & DORMANDY,T.L.(1976) The antibacterial effect of water-soluble compounds from autoxidising linolenic acid. Journal of Medical Microbiology, 4 105-1 10. GUTTERIDGE, J.M.C., LIJNEC,J., & HEYS, A. (1978) Fluorescence of peroxidised fatty acids. Analytical Lefters (In press). JOHNSON, R.B. & LEHMEYER, J.E. (1976) Elaboration of toxic oxygen products by neutrophils in a model of immune complex disease. Journal of Clinical Investigation, 51, 836841. JOHNSON, R.B., JR, KEELE,B.B., MISRA,H.O., LEHMEYER, J.E., K.V. (1975) WEBB,L.S., BAEHNER,R.L. & RAJAGOPALAN, The role of superoxide anion generation in phagocytic bactericidal activity. Journal of Clinical Investigation, 55, 1357. KLEBANOFF, S.J. (1975) Antimicrobial mechanisms in neutrophilic polymorphonuclear leukocytes. Seminars in Haemafology, 12,117-160. MCCORD, J.M. (1974) Free radicals and inflammation: Protection of synovial fluid by superoxide dismutase. Science, 185,529-530. MCMURRAY, W. & DORMANDY, T.L. (1974) Lipid autoxidation in human skeletal muscle. Clinica Chimica Acfa, 52, 105114. MALSHET, V.G. & TAPPEL, A.L. (1973) Fluorescent products of lipid peroxidation. Structural requirements of fluorescence ir, conjugated Schiff bases. Lipids, 8,194-198. PUIGPARELLADA, P. & PLANAS,J.M. (1977) Synovial fluid degradation induced by free radicals. In vifro action of several free radical scavengers and anti-inflammatory drugs. Biochemical Pharmacology, 27,535-537. SALIN, M.L. & MCCORD, J.M. (1975) Free radicals and inflammation. Journal of Clinical Investigation, 56, 13 191323. SCHAUENSTEIN, E. (1967) Autoxidation of polyunsaturated esters in water: chemical structure and biological activity of products. Journal of Lipid Research, 8,4 17.
59
SCUDDER,P., STOCKS,J. & DORMANDY,T.L. (1975) The relationship between erythrocyte superoxide dismutase and copper levels in normal subjects and in patients with rheumatoid arthritis. Clinica Chimica Acfa, 69,397-403. SCUDDER,P., WHITE, A.G., McMurray, W. & DORMANDY, T.L. (1978) Synovial fluid copper and related variables in rheumatoid and degenerative arthritis. Annals of fhe Rheumatic Diseases, 31,7 1-77. SEGAL, A.W. & PETERS,T.J, (1976) Characterisation of the enzyme defect in chronic granulomatous disease. Lancef, i, 1363-1365. SHIN,B.C.,HUGGINS, J.W. & CARRAWAY, K.L. (1972) Effects of pH concentration and aging on the malonaldehyde reactions with proteins. Lipids, 7,229-233. SHOHET,S.B., P m , J., BAEHNER,R.L. & POPLACK,D.G. (1974) Lipid peroxidation in the killing of phagocytised pneumococci. Infection and Immunity, 10,1321-1328. SLATER,T.F. (1972) Free-Radical Mechanisms in Tissue Injury. Pion, London. STOCKS,J., GUITERIDGE,J.M.C., SHARP,R.J. & DORMANDY, T.L. (1974) Assay using brain homogenate for measuring the antioxidant activity of biological fluids. Clinical Science and Molecular Medicine, 41,215-222. STOCKS,J., OFFERMAN,E.L., MODELL,C.B. & DORMANDY, T.L. (1972) The susceptibility to autoxidation of human red cell lipids in health and disease. British Journal of Haematology, 23,7 13-724. STOSSEL,T.P., MASON,R.J. & SMITH,A.L. (1974) Lipid peroxidation by human blood phagocytes. Journal of Clinical Investiga lions, 54,638-645. TAPPEL, A.L. (1972) Lipid peroxidation and fluorescent molecular damage to membranes. In: Pathological Aspects of Cell Membranes, pp. 145-172. Ed. Trump, B. F. & Arstila, A. Academic Press, New York. TROMBLY,R. & TAPPEL, A.L. (1975) Fractionation and analysis of fluorescent products of lipid peroxidation. Lipids, 10,441-450. UNDENFRIEND, S. (1969) Fluorescence Assay in Biology and Medicine, vol. 11. Academic Press, New York. WILLIAMS,T.J. & PECK, M.J. (1977) Role of prostaglandin mediated vasodilatation in inflammation. Nafure (London), 270,530-532. YOST,F.J., JR & FRIDOVITCH, I. (1974) Superoxide radicals and phagocytosis. Archives of Biochemisfry and Biophysics, 161, 395-401.
