Flow cytometry analysis of platelet cyclooxygenase-2 expression: induction of platelet cyclooxygenase-2 in patients undergoing coronary artery bypass grafting

Authors


Karsten Schrör, MD, Institut für Pharmakologie und Klinische Pharmakologie, Heinrich-Heine-Universität, Moorenstr. 5, D-40225 Düsseldorf, Germany. E-mail: kschroer@ uni-duesseldorf.de

Abstract

Summary. There are conflicting reports about the expression of cyclooxygenase (COX)-2 in human platelets. The present study describes a flow cytometric method for the measurement of platelet COX. Both COX-1 and COX-2 were shown to be expressed in platelets from patients undergoing a coronary artery bypass graft. There was a significant increase in COX-2 expression at day 5 as compared with pre-surgery values (mean fluorescence 12·31 ± 0·88 versus 9·15 ± 0·88; means ± SEM, n = 7, P < 0·05), whereas COX-1 levels did not change (13·45 ± 1·11 versus 12·38 ± 1·41; n = 7, P > 0·05).

Prostaglandin H synthase (PGHS) is a rate-limiting enzyme in the synthesis of prostanoids, for example thromboxane A2, from arachidonic acid. Inhibition of the cyclooxygenase (COX) activity of PGHS is the major target for the platelet-inhibitory actions of aspirin (Patrono, 1994). PGHS exists in two isoforms: a constitutive (designated as COX-1) and an inducible form (designated as COX-2) (Vane et al, 1998). Because circulating platelets lack a nucleus and are not capable of mRNA synthesis, it is generally assumed that thromboxane A2 synthesis is entirely dependent on COX-1 (Patrono et al, 2001).

After the original description of COX-2 mRNA expression in platelets by Matijevic-Aleksic and colleagues (Matijevic-Aleksic et al, 1995), we have reported that circulating platelets also express COX-2 protein (Weber et al, 1999; Hohlfeld et al, 2000). However, Patrignani (Patrignani et al, 1999) were not able to confirm these findings. The clarification of this issue is of clinical importance because COX-2 expression in platelets might be a potential factor in aspirin resistance. We have therefore established a flow cytometric method for the measurement of platelet COX-1 and COX-2 expression.

Materials and methods

Platelet-rich plasma was prepared from citrate (1:9, v/v)-anticoagulated blood as described previously (Weber et al, 2000). For flow cytometric analysis of platelet COX expression, a modification of a method, originally described for the measurement of VASP phosphorylation (Schwarz et al, 1999), was applied. Briefly, washed platelets (Weber & Schrör, 2001) were fixed with paraformaldehyde (1%) for 10 min, followed by the addition of Triton X-100 (0·3%) for another 10 min. Platelets were then pelleted by centrifugation (12 000 g for 15 s), washed in phosphate-buffered saline (PBS), pelleted again, and resuspended in PBS. Then, 1 µg of fluorescein isothiocyanate (FITC)-conjugated monoclonal anti-COX-1 (Cayman Chemical, Ann Arbor, MI, USA; Cat. no. 160111) or anti-COX-2 antibodies (Cayman; Cat. no. 160113) was incubated with 10 µl of platelet suspension for 15 min in the dark. Then, samples were diluted with isotoneR and analysed on the EPICS-XL cytometer (Beckman Coulter, Krefeld, Germany). The platelet population was identified on its forward and side scatter distribution, which was validated by positive staining with anti-CD42b antibodies (Dako, Hamburg, Germany). Detectors were set to logarithmic amplification and 30 000 platelets were analysed, using the System II software. Statistical ananlysis was performed using the Student's t-test (GraphPad INSTAT® 3·01, GraphPad Software, San Diego, CA, USA). P-values < 0·05 were considered significant. Western blotting was performed as described previously (Hermann et al, 2001). Membranes were probed with monoclonal anti-COX-1 (Cayman, Cat. no. 160110, 1:500) or anti-COX-2 antibodies (Cayman, Cat. no. 160112, 1:500), followed by incubation with peroxidase-conjugated secondary antibodies (1:3000). Bands were visualized by chemiluminescence (Roche Molecular Biochemicals, Mannheim, Germany).

Results and discussion

Possible contamination with extraplatelet sources of COX-2, such as leucocytes, is an issue of concern in COX-2 analysis (Patrignani et al, 1999). We have therefore established a flow cytometric method for the measurement of platelet COX expression, which excludes cells or particles outside the range of platelet size. The purity of the platelet population was validated by positive staining with anti-CD42b antibodies. To achieve high specificity, monoclonal anti-COX-2 antibodies directed against a peptide from the COX-2 amino acid sequence (amino acids 580–599), which is not present in COX-1, were used.

Using this technique, we found marked COX-2 fluorescence signals in permeabilized but not in non-permeabilized platelets (Fig 1A), indicating intracellular localization of platelet COX-2 immunoreactivity. The specificity of the reaction was verified using a peptide that was used as antigen for production of the anti-COX-2 antibody (Cayman, Cat. no. 360107, 10 µg/ml) and a peptide derived from the COX-1 sequence (Santa Cruz, Cat. no. sc-1752P, 10 µg/ml) respectively. The COX-2 peptide reduced the fluorescence signals to the levels of the isotypic control (Fig 1B). In contrast, the COX-1 peptide had only marginal effects (Fig 1C). Platelet COX-2 was detected by flow cytometry in 12 platelet samples obtained from healthy volunteers.

Figure 1.

Flow cytometric analysis of cyclooxygenase (COX)-2 in human platelets. COX-2-expression was measured in permeabilized (Triton X-100) and non-permeabilized platelets, respectively (A). The specificity of the reaction was verified using COX-2 (B) and COX-1 (C) blocking peptides. The fluorescence obtained with isotype-matched antibodies (isotypic control) is shown for comparison.

Next, the flow cytometric method was applied to measure the expression of platelet COX-2 in patients undergoing coronary artery bypass surgery. In accordance with previous findings by Hohlfeld and colleagues (Hohlfeld et al, 2000), a transient upregulation of COX-2 was found (Fig 2A). Interestingly, COX-1 expression did not change during the study period (Fig 2B). The flow cytometric data were validated by Western blotting, confirming the transient induction of COX-2 at unchanged COX-1 levels (Fig 2). The increase in COX-2 expression at day 5 as compared with pre-surgery values was statistically significant (mean fluorescence 12·31 ± 0·88 versus 9·15 ± 0·88; means ± SEM, n = 7, P < 0·05), whereas COX-1 levels did not change (mean fluorescence 13·45 ± 1·11 versus 12·38 ± 1·41; means ± SEM, n = 7, P > 0·05).

Figure 2.

Western blot and flow cytometric analysis of COX-2 (A) and COX-1 (B) in a patient undergoing a coronary artery bypass graft (CABG). Measurements were carried out before the operation (pre-OP), as well as 1, 5 and 10 d after the operation (post-OP). The flow cytometric data (mean fluorescence intensity, MnX) and histograms demonstrating COX expression before and 5 d after CABG are shown below the Western blots.

Taken together, using a flow cytometric method, the present study confirmed the expression of COX-2 in human platelets. In addition, this study demonstrated that the expression level of platelet COX-2 can be upregulated under certain clinical conditions. Further studies will address the question as to whether platelet COX-2 contributes to aspirin resistance, for example in patients undergoing coronary artery bypass grafting (Zimmermann et al, 2001).

Ancillary