STRUCTURAL HETEROGENEITY OF PLATELET-ACTIVATING-FACTOR PRODUCED BY MURINE PREIMPLANTATION EMBRYOS

Platelet-activating factor (PAF) is 1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine with the alkyl moiety predominantly a mixture of saturated hexadecyl and octadecyl chains (C16:0 and C18:0 PAF, respectively). Previously, a PAF bioassay was compared with a radioimmunoassay for PAF (NEN Du Pont). Both assays were sensitive and quantitative, but the correlation between PAF measured by the bioassay compared to the radioimmunoassay was poor for murine embryo-derived PAF (r = 0.773, n = 88), while being completely adequate (r = 0.961) for a PAF standard which was an equimolar mixture of C16:0 and C18:0 PAF (C16:c1:0 PAF). This study compared a larger sample size of murine embryo-derived PAF (n = 154) and found that the poor correlation between the two assays persisted (r = 0.791). When dose-response curves were generated with C16:0, C16:0/C18:0 and C18:0 PAF (over a concentration range of 0.3 - 30 ng/ml), the concentrations which gave a 50% response were equivalent in the bioassay (i.e. 6 ng/ml), but differed in the radioimmunoassay (i.e. 1.5, 3 and 6 ng/ml, respectively). Following separation of murine embryo-derived PAF (from medium in which 30 two-cell embryos had been cultured for 24 h) into C16:0 and C18:0 PAF by reverse phase high performance liquid chromatography, 9/20 cultures produced 100% C16:0 PAF, 2/20 cultures produced 100% C18:0 PAF and the remaining 9/20 cultures produced varying proportions of both. Therefore, since embryos produced a heterogeneous mixture of C16:0 and C18:0 PAF and the radioimmunoassay has different sensitivities to these molecular species, these results explain the poor correlation between the bioassay and radioimmunoassay measurements of embryo-derived PAF, but also indicate that the radioimmunoassay can only be used if each of the separated molecular species are measured individually against appropriate standards.