We read with great interest the review article, ‘Cellular microparticles: what are they bad or good for?’, by Freyssinet in the Journal [1]. It opens with a succinct review of the hypothesis that a necessary prelude to vesiculation and shedding of microparticles (MP) is migration of phosphatidylserine (PS) from the inner to the outer leaflet of the membrane bilayer, implying by the illustration that surface exposure of PS is a defining feature or signature of MP. Freyssinet supported the concept that the role of MP is not limited to procoagulant activities with thrombotic potential but extends to inflammation, and in addition they can behave as true diffusible vectors in transcelluar signaling. He discusses at some length evidence that MP can bind to and influence leukocyte activation. It should be added that our laboratory was the first to demonstrate clearly that platelet microparticles (PMP) bind to neutrophils, activating them and causing formation of grape-like clusters of PMP–neutrophil complexes [2]. In that paper we advanced the concept that PMP can play an important role as a messenger linking thrombosis and inflammation through interaction with neutrophils [2]. That work was not cited in the review.

He then concisely reviews some of the pertinent clinical publications, citing two of ours, namely, our original observation on PMP in patients with idiopathic thrombocytopenic purpura (ITP) in 1992 [3] and our comprehensive review of PMP in 1999 [4]. In our 1992 report, we demonstrated that PMP are hemostatically functional because ITP patients with high PMP do not bleed in spite of severe thrombocytopenia—certainly an example of ‘good’ MP—but we also found that those with unusually high MP suffered from ischemic small vessel diseases of the central nervous system, manifesting as recurrent transient ischemic attack, confirmed by magnetic resonance imaging [3], some patients progressing to advanced vascular dementia [3,5]; hence in this setting, they are certainly ‘bad’ MP. Our reports certainly fit into the dichotomy of MP into ‘good’ and ‘bad’ as Freyssinet proposed. However, good or bad may not be qualitative but quantitative in some clinical settings.

Among the most exciting recent developments in MP studies is the burgeoning field of endothelial microparticle (EMP) analysis, not covered in his review. We have recently reviewed this field [6]. An important issue arising in our EMP studies is the distinction between MP-bound and true soluble circulating markers of endothelial activation: we have shown that many of the markers assumed to be ‘soluble’ are in reality MP-bound, at least in part (as discussed in [6]). The pathophysiological mechanisms of release of true soluble and MP-bound species are quite different and the two forms may be functionally distinct as well (because true soluble forms generally lack transmembrane domains). It remains to be seen which form is the better marker of disease activity. Meanwhile, assays by ELISA methods measure the sum of both forms.

Finally, it is important to add our observation that not all MP are positive for PS. For example, the majority of EMP from activated endothelial cells were not positive for PS as judged by annexin V (AnV) binding, since many more EMP were counted by CD62E than by AnV. Even EMP from apoptotic cells, which are much richer in AnV binding, still gave only about half as many positives for AnV as for CD31 [6,7]. Therefore, assay methods which define total MP in terms of positivity for PS, as employed by others [8] and illustrated in the review [1], can give grossly misleading results; for example, will tend to underestimate MP arising from cellular activation as distinct from apoptosis. Thus the critical question of how best to measure MP deserves further investigation.

Our recent study of the mechanism of EMP generation indicated that capping precedes release of EMP from the membrane in a well-orchestrated manner, as opposed to random release from the membrane [9]. There may be many diverse routes governing release of MP, the floppase hypothesis being only one of them.

The review of Freyssinet is timely and informative, bringing to attention a number of interesting papers and raising many questions. We hope it will initiate momentum towards a forum to refine and standardize MP assay methods, and to further elucidate the true role(s) of MP in health and disease. Methods to identify ‘good’ and ‘bad’ MP would certainly be welcome, as that could substantially improve the early diagnosis of thrombotic and inflammatory disorders, which are by far the most common and devastating illnesses in the modern world.


  1. Top of page
  2. References