Multiple Species of Endothelial Microparticles (EMP) Released from Different Membrane Protein Clustering Domains Exhibit Distinctive Phenotypes and Activities.

INTRODUCTION: We have previously shown that EMP comprise multiple species of vesicles released from endothelial cells (EC) upon stimulation. However, the mechanism underlying EMP release is not clear, nor is their functional role. We postulated that EMP release is initiated by formation of discrete clusters of membrane proteins, each of which may release distinctive EMP characterized by the predominant protein in the cluster or raft. Therefore, each such subspecies may have distinctive activities in cell interaction or other function. In this study, we employed flow cytometry to investigate this postulated mechanism, and compared in vitro with in vivo findings.

METHODS: EMP were prepared by incubating renal endothelial cells (EC) with 10 ng/mL of TNF for 24hr. Two-color flow cytometry was used to analyze the phenotypic composition of the resulting EMP, the markers used including CD31, CD62E, CD51, CD54, annexinV (AnV), tissue factor (TF), and lectin Ulex europaeus I (Ulex). Fluorescence microscopy was used to study membrane protein movement and clustering.

RESULTS:(1) Phenotypic composition of EMP was evaluated in culture supernatants by flow cytometry, first by the number detected with each marker. Expressed in millions/mL, they were: by Ulex, 280; AnV, 52; CD54, 48; CD62E, 46; CD31, 34; TF, 36; and CD51, 8.(2) Two-color technique was used to establish the degree to which more than 1 marker (antigen) was present on the same EMP. It was found that only a small fraction (<5%) of CD54+ or CD62E+ EMP were also positive for CD31, and vice versa.(3) Cell interactions: Incubating the EMP mixture with neutrophils resulted in selective binding of CD54+ and CD62E+ EMP to the neutrophils and loss of 95% and 70% of free CD54+ and CD62E+ EMP, respectively, from the cell-free supernatants. EMP positive for the other markers showed little binding to leukocytes. These data confirm subspecies of EMP with little overlap of markers and differing affinity for leukocytes. (4) Fluorescence microscopy: Upon EC stimulation, a time-dependent movement of surface markers CD31 and CD54 resulted in their clustering to different locations prior to shedding of vesicles. Majority of vesicles were seen to shed from these clusters. This process may explain how EC can release multiple subspecies of EMP. (5a) In vivo: Levels of CD54+ EMP were always low or nearly undetectable in plasma from patients or normal controls. However, high levels of CD54+ EMP/leukocyte conjugates were found in several thrombotic and inflammatory disorders. This is consistent with in vitro findings. (5b) In vivo total MP: Study of plasma from 26 normal controls showed that MP measured by Ulex were about 3 to 4-fold higher than if measured by AnV. The majority of Ulex+ MP were negative for AnV.

SUMMARY:

• Our data support the hypothesis that upon activation or apoptosis, EC developed multiple membrane protein clusters as a prelude to EMP release.

• EMP species released from these membrane clusters exhibit distinctive phenotypes and activities such as leukocyte binding.

• AnV has been widely used a marker for total MP, but this will miss MP not expressing AnV.

We show that the lectin marker Ulex gives the highest counts of MP, in vitro and in vivo, suggesting that Ulex may be a better proxy than AnV for defining total MP.