The murine macrophage cell line J774A.1 (American Type Culture Collection, Rockville, MD) was grown in RPMI 1640 medium supplemented with 10% fetal calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin, 150 U/ml polymyxin B, and 50 μg/ml gentamycin in a humidified 5% carbon dioxide/95% air incubator at 37°C. All cell culture media and supplements were from Invitrogen (Paisley, UK). For cell culture experiments 48 well plates (Costar, Cambridge, MA) or culture slides (Beckton Dickinson Labware, Sunnyvale, CA) were used. Cells (0.5 × 106/800 μL) were allowed to adhere at 37°C for 2 hours. Thereafter, cells were washed with warm medium and the medium was replaced by Dulbecco Modified Eagles Medium (DMEM) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, and 50 μg/ml gentamycin.
Human blood platelet concentrates were kindly provided by the blood transfusion center of the University Hospital of Antwerp (Belgium). These platelet concentrates had been deleukocytized by filtration and contained only 1 to 2 white blood cells per 300,000 platelets. Platelets were washed in warm PBS and resuspended in serum free Dulbecco Modified Eagles Medium (DMEM) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, and 50 μg/ml gentamycin.
J774 macrophages, primed with recombinant mouse IFNγ (100 U/ml, Sigma), were incubated with washed human platelets (108 per 0.5 × 106 macrophages) 1 in the presence of an NSAID or its solvent. After an incubation period of 18 hours, the culture medium was removed from the macrophages, centrifuged, and stored at −20°C for nitrite determination. To study the specificity of NSAID treatment on the nitrite production evoked by platelet phagocytosis, experiments using lipopolysaccharide (LPS) from Salmonella typhosa as an alternative stimulus for iNOS induction were performed as well.
The following NSAIDs were examined (highest final solvent concentration is indicated): indomethacin (Federa, Brussels, Belgium, 0.5 mM TRIS pH 8.3), ibuprofen (Federa, Brussels, Belgium, 0.5% ethanol), acetylsalicylic acid (Federa, Brussels, Belgium, 0.5% ethanol), naproxen (Federa, Brussels, Belgium, 1% ethanol), meloxicam (Calbiochem, San Diego, CA, 0.2% DMSO), and sulindac sulfide (Sigma, St Louis, MO, 0.2% ethanol), the active metabolite of the prodrug sulindac. They were dissolved at the start of every experiment and the solutions were sterilized by filtration.
Nitric oxide synthase activity was assessed by measuring nitrite, a stable NO metabolite, using the Griess reaction. 7 Briefly, 200 μL culture medium was transferred to a 96-well microtiter plate and mixed with 40 μL of 6.5 M HCl and 37.5 mM sulphanilic acid (1:1). After 10 minutes, 20 μL 12.5 mM N-(1-naphtyl)-ethylene diamine dihydrochloride was added. After 30 minutes at 37°C, the absorbance was read at 540 nm using a Titertek Multiscan MCC/340 plate reader (Labsystems). Nitrite concentration was calculated from a standard curve of sodium nitrite in culture medium.
Oil Red O Stain
To examine whether platelet engulfment by macrophages was altered by NSAIDs, co-incubations of macrophages and platelets were stained with oil red O 8 after fixation of the cells with 4% formaldehyde (10 minutes). Nuclei were stained with hematoxylin.
Phagocytosis of platelets by macrophages was also evaluated by flow cytometry. J774 macrophages and washed human platelets were incubated for 45 minutes with DMEM supplemented with 5 μM of 5- and 6-([(4-chloromethyl) benzoyl] amino) tetramethylrhodamine (Cell Tracker Orange, Molecular probes) or 2.5 μM 5-chloromethylfluorescein diacetate (Cell Tracker Green, Molecular Probes), respectively. 9 Thereafter, the media were refreshed, and the macrophages were incubated with the platelets for 3.5 hours in the presence or absence of the highest concentrations of these NSAIDs that inhibited macrophage activation. This time point was based on previous experiments on the evaluation of platelet phagocytosis using confocal and electron microscopy. 1 Flow cytometry was performed using a FACSort (BD Biosciences, San Diego, CA). Forward and side scatters were set to include cells but to exclude debris. Green fluorescent-free platelets were monitored in the FL 1 channel, whereas orange fluorescent macrophages were measured in the FL 3 channel. Data were analyzed using CellQuestPro software (BD Biosciences). Phagocytosis (per cent) was calculated as the (number of cells containing cell tracker green and orange / total number of cells containing cell tracker orange) × 100.
Cyclooxygenase Inhibition and Determination of Prostaglandin E2
To document COX inhibition, we tested all NSAIDs on their ability to inhibit arachidonic acid-induced blood platelet aggregation. Blood samples (9 mL) were obtained from a cannulated carotid artery of pentobarbitone (30 mg/kg)-anesthetized New Zealand White rabbits and collected on trisodiumcitrate (1 mL; 3.8%). Platelet-rich plasma was prepared by centrifugation (10 minutes, 200 g). The samples were exposed to 330 μM arachidonic acid after 3 minutes incubation in the presence or absence of the NSAIDs in a Chronolog dual channel aggregometer (37°C, 900 rpm). 9 The maximum increase in light transmission after the addition of arachidonic acid was measured.
PGE2 was measured in the culture medium by radioimmunoassay. Dose interpolation was done with a four-parameter logistic function using the IBM-PC RIA data reduction package provided by M.L. Jaffe (Silver Spring, MD). 10
Surface-Enhanced Laser Desorption/Ionization Time-of-Flight (SELDI-TOF) Mass Spectrometry
Analysis of APP fragments in cell lysates of macrophages, platelets, and macrophages co-incubated with platelets for 3, 6, or 18 hours with or without indomethacin was performed by SELDI-TOF mass spectrometry using the ProteinChip system (PS20, protein chip array, Ciphergen Biosystems Inc., Fremont, CA). Cells were lysed for 30 minutes in 50 mM TRIS buffer containing 250 mM NaCl, 5 mM EDTA, 0.1% (vol/vol) Triton X 100, 1 mM dithiothreitol, 1 mM NaF, 10 mM β-glycerophosphate, 0.1 mM Na3VO4, and 0.1 mM paramethylsulphonylfluoride.
The chip arrays were incubated overnight at 4°C with the antibody (6E10 or IgG) to allow coupling. 6E10 is an antibody specific for the 17 N-terminal amino acids of β-amyloid peptides (672–688 of the full-length APP sequence). IgG antibody coating was used as a control to discriminate between specific and aspecific peptide binding to the protein chip. The samples were diluted and 20 μg total protein was spotted on the 6E10 or IgG antibody-coated chips. After 4 hours, non-bound peptides were removed by extensive washing with PBS and PBS containing 0.5% Triton X 100. The arrays were rinsed several times with a 1.0-mM HEPES buffer, and allowed to air dry. Thereafter, 1 μL of a 20% saturated solution of α-cyano-4-hydroxycinnamic acid in 0.5% (vol/vol) trifluoroacetic acid: 50% (vol/vol) acetonitrile was applied to each spot. Mass spectrometric analysis was performed by averaging a minimum of 100 laser shots in a Ciphergen SELDI Protein Biology system II, with a linear time-of-flight mass analyzer (mass accuracy: 0.1%). 11–14
Nitrite values were compared by using analysis of variance (ANOVA). A P value less than 0.05 was considered to be significant. All data are represented as mean ± SEM.
Effect of NSAIDs on Nitrite Production
Incubation of IFNγ-primed J774 macrophages with washed human platelets or LPS resulted in nitrite production by the macrophages. The extent of the nitrite production was comparable for both stimuli. Addition of indomethacin, ibuprofen, meloxicam, or sulindac sulfide to the culture medium concentration-dependently reduced nitrite production by macrophages evoked by platelet phagocytosis (Fig. 1, A–D), whereas acetylsalicylic acid or naproxen were without such an inhibitory effect (Fig. 1, E and F). As compared with indomethacin, sulindac sulfide, and meloxicam, the effect of ibuprofen was less pronounced. When macrophages were stimulated with LPS, nitrite production was not affected by any NSAID (Fig. 1, A–F).
Effect of NSAIDs on Platelet Aggregation and PGE2 Production
The production of PGE2 by macrophages stimulated with either platelets or LPS was significantly reduced in the presence of all NSAIDs (results not shown). Moreover, all NSAIDs inhibited arachidonic acid-induced platelet aggregation. After 3 minutes of exposure the percentage inhibition was: indomethacin (10, 50, 100 μM), 97% for all concentrations, ibuprofen (100, 200, 300 μM), 9%, 92%, and 96%, meloxicam (10, 30, 50 μM), 50%, 97%, and 97%, acetylsalicylic acid (50, 250, 500 μM), 13%, 96%, and 96% and naproxen (75, 150 μM), 12% and 97%. Sulindac sulfide was less effective as platelet inhibitor: 1 μM reduced the platelet aggregation with 3% and 10 μM with 18%.
Platelet Phagocytosis and Foam Cell Formation
Flow cytometry revealed that after 3.5 hours platelet phagocytosis was not altered in the presence of 100 μM indomethacin, 500 μM ibuprofen, 10 μM sulindac sulfide, and 50 μM meloxicam (Fig. 2).
Lipid accumulation in the macrophages as a result of platelet phagocytosis was not changed in the presence of 100 μM indomethacin, indicating that platelet phagocytosis was not altered by this NSAID (Fig. 2). Similarly, the other NSAIDs did not alter platelet phagocytosis (not shown).
Detection of APP-Fragments by SELDI-TOF Mass Spectrometry
The mass spectrum of the lysate of macrophages incubated for 3 hours with human blood platelets did not show peaks corresponding to Aβ1–40 or Aβ1–42 fragments (Fig. 3). However, in the lysate of macrophages incubated with blood platelets for 6 or 18 hours, two peaks at m/z 6277.5 and m/z 6541.5 were detected, which correspond to the [M + H]+ peaks of proteins with a molecular mass of 6276.5 Da and 6540.5 Da, respectively. The intensity of these peaks increased with an increasing incubation time of platelets and macrophages (Fig. 3), which paralleled nitrite production (0.49 ± 0.03 μM after 3 hours, 0.80 ± 0.07 μM after 6 hours and 6.8 ± 0.2 μM after 18 hours). For the 6276.5 Da protein, six different APP-derived peptides are theoretically possible within a mass range of ± 5 Da (Table 1). The peak at m/z 6541.5 was not detected in every experiment and will not be considered in the present study.
When macrophages were incubated with platelets in the presence of indomethacin for 18 hours, [M + H]+ peaks at m/z 6277.5 and 6541.5 disappeared completely. This was accompanied by peaks in the mass ranges m/z 5000–6000 and m/z 7000–8000 (Fig. 4). In all cases, a standard mixture gave the expected profile with [M + H]+ peaks at m/z 1953.5 (Aβ1–16, not shown), 4343.5 (Aβ1–40), and 4527.6 (Aβ1–42).
Recently, it was demonstrated that in neuronal cells a subset of NSAIDs including indomethacin, ibuprofen, and sulindac sulfide alter the γ-secretase cleaving site of APP. 6 As a result, the formation of Aβ1–42, the most toxic form of Aβ, was reduced in favor for the less toxic Aβ1–38. 6 Indeed, it has been suggested 6 that this interference with the γ-secretase cleaving site of APP may explain the benefit of some NSAIDs seen in epidemiological 15–17 and clinical studies 18–20 of Alzheimer's disease. Since specific inhibitors of γ-secretase are currently not available or cytotoxic (unpublished observations), we investigated a series of NSAIDs to test whether γ-secretase is involved in platelet-induced macrophage activation. These NSAIDs were merely selected as pharmacological tools to unravel the mechanisms of macrophage activation after platelet phagocytosis.
The present data confirm that platelets are phagocytosed by macrophages (flow cytometry), that the macrophages become activated (nitrite production) and transform to foam cells (oil red O). Foam cell formation can result from other mechanisms than platelet phagocytosis. Since we used serum-free medium, endocytosis of lipid droplets can be excluded. Engulfment of LDL or modified LDL bound to the platelets is possible but seems to be less important because we used washed human platelets. With flow cytometry, we could demonstrate that platelets are indeed phagocytosed by macrophages, confirming previous results. 1
Subsequently, our results clearly showed a reduced nitrite production when macrophages were incubated with platelets in the presence of indomethacin, ibuprofen, sulindac sulfide, and meloxicam as compared with the controls. Importantly, the concentrations of the NSAIDs used in the present study did not affect LPS-induced nitrite production, confirming previous reports. 21–23 The latter finding indicates that the inhibitory activity of NSAIDs on macrophage activation following platelet phagocytosis was not due to interference with the cell signaling pathways essential for iNOS induction or with the activity of iNOS. Indeed, Aeberhard et al 22 showed that only extremely high concentrations of NSAIDs, exceeding those of the present study, inhibit iNOS activity in LPS-stimulated macrophages. Moreover, the inhibitory activity of indomethacin, ibuprofen, sulindac sulfide, and meloxicam on macrophage activation evoked by platelet ingestion was not due to reduced platelet phagocytosis or processing. The different compounds did not inhibit the early platelet uptake as studied by flow cytometry, nor did they suppress the accumulation of platelet-derived lipid droplets inasmuch as oil red O staining of NSAID-treated macrophages did not differ from that of untreated macrophages. Furthermore, these experiments revealed no differences in platelet uptake between NSAID-treated and control macrophages.
Both PGE2 measurements in the supernatant and arachidonic acid-induced blood platelet aggregation assays showed that all NSAIDs inhibited COX, confirming their activity and excluding possible degradation of the drugs. However, inhibition of macrophage activation by the NSAIDs was not due to COX-inhibition. Except for ibuprofen, the concentrations of the NSAIDs required to affect the macrophage activation after platelet ingestion were higher than those necessary for COX inhibition. Secondly, sulindac sulfide, the least effective COX inhibitor, was even the most active inhibitor of macrophage activation following platelet phagocytosis. More importantly, although acetylsalicylic acid and naproxen inhibited COX, they did not alter nitrite production of macrophages induced by platelet phagocytosis. Indeed, the latter NSAIDs do not affect the γ-secretase cleaving site of APP. 6 Taken together, these findings suggest that the effect of some NSAIDs on nitrite formation could be due to interference with the γ-secretase cleaving site of APP, thereby reducing the formation of Aβ-like peptides.
Previously, Ogawa et al 24 reported that Aβ1–40 could stimulate nitrite production in IFNγ-primed J774 macrophages. They further showed that indomethacin and ibuprofen inhibited the Aβ-effect and proposed that the inhibitory effects of these NSAIDs on iNOS expression were mediated via peroxisome proliferator-activated receptor-γ (PPAR-γ). Unfortunately, we were unable to reproduce these experiments. We encountered a great batch-to-batch variability in the response to Aβ1–40 as well as Aβ1–42 and the nitrite production of IFNγ-primed macrophages after stimulation with either peptide was too low to study the effect of inhibitors. 1 Since Ogawa et al 24 did not investigate whether the inhibition by the NSAIDs was specific for Aβ1–40 by including an alternative stimulus, such as LPS, their experiments cannot be compared with the present study in which we used platelets as a stimulus. Furthermore, the active concentrations of indomethacin and ibuprofen were even higher than in the present study and could have interfered directly with the activity of iNOS, as reported by Aeberhard et al. 22 Since NSAIDs used in the present study did not interfere with the LPS-induced expression or activity of iNOS, it seems less likely that the concentrations used in the present study interfered with PPAR-γ nuclear receptors as proposed by Ogawa et al 24. However, it cannot be excluded that some of their inhibitory effects were due to interference with iNOS-inducing pathways distal from Aβ-production.
Subsequently, SELDI-TOF mass spectrometry was used to substantiate the γ-secretase involvement. This technique demonstrated that incubation of macrophages with platelets resulted in a time-dependent formation of a peptide with a molecular mass of 6276.5 Da, containing an Aβ-like domain. The properties and the importance of this peptide are still unclear. Only its molecular mass is known and the fact that it contains the epitope recognized by the 6E10 antibody. Six theoretically possible Aβ-like peptides with a molecular mass of 6276.5 ± 5 Da remained after excluding peptides without the epitope of the 6E10 antibody on the SELDI chip. Two of these candidates deserve further consideration. The first one, which has a mass of 6275.1 Da, is 58 amino acids long and contains the complete Aβ1–42 sequence (ie, amino acids 672–713 of full-length APP) at its C-terminal site, ending just at the γ-secretase cleaving site (Table 1). The difference between the theoretical mass and the measured mass was the lowest for this peptide as compared with the other five candidates. Another interesting candidate has a mass of 6273.3 Da, is 57 amino acids long and contains the complete Aβ1–42 sequence at its N-terminal site, starting with the β-secretase cleaving site (Table 1). The exact identity of the Aβ-like peptide with a mass of 6276.5 ± 5 Da formed during platelet phagocytosis by macrophages remains to be established. Also the importance of the presence of the complete Aβ1–42 sequence in this Aβ-like peptide remains to be determined, since Aβ1–42 itself had only a limited effect on the activation of both J774 macrophages, as discussed before, and microglia cells. 25 Nevertheless, the present study showed that addition of indomethacin to the co-incubations of platelets and macrophages inhibited the formation of this peptide, as well as nitrite production. This strongly suggests that the Aβ-like peptide with a mass 6276.5 ± 5 Da participates in macrophage activation following platelet phagocytosis.
In summary, our results demonstrate that NSAIDs known to alter the γ-secretase cleaving site of APP, as well as meloxicam, reduced macrophage activation following platelet phagocytosis, but not after LPS stimulation. Furthermore, NSAIDs like acetylsalicylic acid and naproxen, which do not affect γ-secretase cleavage were also without effect on macrophage activation after platelet ingestion. This indicates that the reduced macrophage activation is independent of COX-inhibition and points to an effect on γ-secretase. The inhibition of the formation of Aβ-related peptides by indomethacin, measured by SELDI-TOF, gives further support to the idea that NSAIDs act via interference with a γ-secretase-related pathway, although further research is necessary to identify the peptide(s). Finally, our findings indicate that the novel platelet-induced pathway of macrophage activation can be treated in a pharmacological way. This may open new perspectives for the development of specific drugs for intervention in the complex matter of plaque physiology. 5
The authors thank Prof. M. Claeys for critical comments on the manuscript. M. Fillet is a Senior Research Assistant at the National Fund for Scientific Research (FNRS, Belgium) and M.-P. Merville is a Research Associate (FNRS, Belgium). The use of SELDI-TOF mass spectrometry technology was made possible by cooperation with the Laboratory of Medicinal Chemistry (Prof. J. Gielen), University of Liège, Belgium. Human blood platelets were kindly provided by the Blood Transfusion Center, University Hospital of Antwerp, Antwerp, Belgium.
1. De Meyer GRY, De Cleen DMM, Cooper S, et al. Platelet phagocytosis and processing of beta-amyloid precursor protein as a mechanism of macrophage activation in atherosclerosis. Circ Res. 2002; 90:1197–1204.
2. Kockx MM, Cromheeke KM, Knaapen MWM, et al. Phagocytosis and macrophage activation associated with hemorrhagic microvessels in human atherosclerosis. Arterioscler Thromb Vasc Biol. 2003; 23:440–446.
3. Boyle JJ, Weissberg PL, Bennett MR. Human macrophage-induced vascular smooth muscle cell apoptosis requires NO enhancement of Fas/Fas-L interactions. Arterioscler Thromb Vasc Biol. 2002; 22:1624–1630.
4. Newby AC, Zaltsman AB. Fibrous cap formation or destruction—the critical importance of vascular smooth muscle cell proliferation, migration and matrix formation. Cardiovasc Res. 1999; 41:345–360.
5. Rajagopalan S, Meng XP, Ramasamy S, et al. Reactive oxygen species produced by macrophage-derived foam cells regulate the activity of vascular matrix metalloproteinases in vitro. Implications for atherosclerotic plaque stability. J Clin Invest. 1996; 98:2572–2579.
6. Weggen S, Eriksen JL, Das P, et al. A subset of NSAIDs lower amyloidogenic Abeta42 independently of cyclooxygenase activity. Nature. 2001; 414:212–216.
7. Schmidt HH, Nau H, Wittfoht W, et al. Arginine is a physiological precursor of endothelium-derived nitric oxide. Eur J Pharmacol. 1988; 154:213–216.
8. Luna LG. Manual of histologic staining methods of the armed forces institute of pathology. L.G.Luna, ed. Armed Forces Institute of Pathology: Washington, DC; 1968:140–142.
9. Baker GR, Sullam PM, Levin J. A simple, fluorescent method to internally label platelets suitable for physiological measurements. Am J Hematol. 1997; 56:17–25.
10. Zamora R, Bult H, Herman AG. The role of prostaglandin E2 and nitric oxide in cell death in J774 murine macrophages. Eur J Pharmacol. 1998; 349:307–315.
11. Merchant M, Weinberger SR. Recent advancements in surface-enhanced laser desorption/ionisation-time of flight-mass spectrometry. Electrophoresis. 2000; 21:1164–1167.
12. Austen BM, Frears ER, Davies H. The use of seldi proteinchip arrays to monitor production of Alzheimer's beta-amyloid in transfected cells. J Pept Sci. 2000; 6:459–469.
13. Vehmas AK, Borchelt DR, Price DL, et al. beta-Amyloid peptide vaccination results in marked changes in serum and brain Abeta levels in APPswe/PS1DeltaE9 mice, as detected by SELDI-TOF-based ProteinChip technology. DNA Cell Biol. 2001; 20:713–721.
14. Beher D, Wrigley JD, Owens AP, et al. Generation of C-terminally truncated amyloid-beta peptides is dependent on gamma-secretase activity. J Neurochem. 2002; 82:563–575.
15. Stewart WF, Kawas C, Corrada M, et al. Risk of Alzheimer's disease and duration of NSAID use. Neurology. 1997; 48:626–632.
16. McGeer PL, Schulzer M, McGeer EG. Arthritis and anti-inflammatory agents as possible protective factors for Alzheimer's disease. Neurology. 1996; 47:425–432.
17. Anthony JC, Breitner JCS, Zandi PP, et al. Reduced prevalence of AD in users of NSAID's and H2 receptor antagonists. Neurology. 2000; 54:2066–2071.
18. Rogers J, Kirby LC, Hempelman SR, et al. Clinical trial of indomethacin in Alzheimer's disease. Neurology. 1993; 43:1609–1611.
19. Aisen PS, Schmeidler J, Pasinetti GM. Randomized pilot study of nimesulide treatment in Alzheimer's disease. Neurology. 2002; 58:1050–1054.
20. Aisen PS, Schafer KA, Grundman M, et al. Effects of rofecoxib or naproxen vs placebo on Alzheimer disease progression: a randomized controlled trial. JAMA. 2003; 289:2819–2826.
21. Amin AR, Vyas P, Attur M, et al. The mode of action of aspirin-like drugs: effect on inducible nitric oxide synthase. Proc Natl Acad Sci U S A. 1995; 92:7926–7930.
22. Aeberhard EE, Henderson SA, Arabolos NS, et al. Nonsteroidal anti-inflammatory drugs inhibit expression of the inducible nitric oxide synthase gene. Biochem Biophys Res Commun. 1995; 208:1053–1059.
23. Salvemini D, Misko TP, Masferrer JL, et al. Nitric oxide activates cyclooxygenase enzymes. Proc Natl Acad Sci U S A. 1993; 90:7240–7244.
24. Ogawa O, Umegaki H, Sumi D, et al. Inhibition of inducible nitric oxide synthase gene expression by indomethacin or ibuprofen in beta-amyloid protein-stimulated J774 cells. Eur J Pharmacol. 2000; 408:137–141.
25. Meda L, Cassatella MA, Szendrei GI, et al. Activation of microglial cells by beta-amyloid protein and interferon-gamma. Nature. 1995; 374:647–650.