Brief Reviews |
From the Research Institute for Internal Medicine, Rikshospitalet, University of Oslo, Oslo, Norway.
Correspondence to Nils Olav Solum, Research Institute for Internal Medicine, Rikshospitalet, Pilestredet 32, 0027 Oslo, Norway. E-mail n.o.solum{at}rh.uio.no
| Abstract |
|---|
|
|
|---|
-granules. Theoretically, therefore,
microvesicles can be involved in both coagulation and inflammation.
Scott syndrome is probably caused by a defect in the activation of an
otherwise normal scramblase, resulting in a relatively severe bleeding
tendency. In Stormorken syndrome, the patients demonstrate a
spontaneous surface expression of aminophospholipids. Activated
platelets and the presence of procoagulant microvesicles have been
demonstrated in several clinical conditions, such as thrombotic and
idiopathic thrombocytopenia, disseminated intravascular coagulation,
and HIV-1 infection, and have been found to be associated with fibrin
in thrombosis. Procoagulant microvesicles may also be formed from other
cells as a result of apoptosis.
Key Words: microvesicles platelets coagulation hemostasis
| Introduction |
|---|
|
|
|---|
Tissue factor is a protein complexed with phospholipid (a cell membrane
or a membrane fragment). It triggers the coagulation cascade by
activating factor VII and forms a complex with the activated
coagulation factor. (By convention, the activated form of a
coagulation factor is written with lowercase a.) This complex directly
activates some factor X. It also activates factor IX,
which then binds, together with factor VIIIa, to negatively charged
phospholipid surfaces. (Some factor VIII is believed to be
activated by the first thrombin molecules formed.) These 2
activated coagulation factors form the so-called intrinsic
"tenase" complex on the surface of activated platelets.
Here factor IXa is acting as an enzyme with factor VIIIa as a cofactor.
This combination leads to the proteolytic activation of factor X. As
symbolized in the Figure
, factor Xa supported by factor Va,
which is bound to phospholipid, acts as the proteolytic enzyme of the
"prothrombinase" complex on the platelet surface. This converts
prothrombin into thrombin, the enzyme that subsequently transforms
fibrinogen into fibrin monomers. These polymerize into a clot that is
"stabilized" by a transamidase reaction catalyzed by factor XIIIa,
resulting in covalent linkages between adjacent fibrin fibrils. The
factor VIIatissue factorphospholipid complex probably is most
important as a trigger of coagulation, in view of the fact that it is
soon inactivated by the "tissue factor pathway
inhibitor" in complex with factor Xa. It has been
suggested that specific receptors contribute to the binding of the
activated coagulation factors to the activated
platelet.3 Thus, the effector cell protease receptor-1
is expressed on the surface of the activated platelet and
is supposed to act as a receptor for factor Xa.3 This is
in conjunction with factor Va and phospholipid and in the presence of
calcium ions. Traditionally, the contribution of platelets to the
coagulation process has been described as "platelet factor 3
activity."
|
| Coagulation Processes on the Platelet Surface |
|---|
|
|
|---|
-granules and is released or
translocated to the membrane during secretion as a partly proteolyzed
active molecule. It has been calculated that the platelets
contribute approximately one fifth of the total factor V of whole blood
(see Reference 44 ). Except for tiny amounts of factor Xa from
intracellular platelet stores,5 factor X and factors
VIII and IX are exclusively found in plasma. The catalytic subunit of
factor XIII, the a chain, is present in significant amounts in
platelets, surprisingly enough not in the
-granules but in the
cytosol. Factors IXa and Xa, which bind calcium ions with high affinity
because of their content of
-carboxyglutamic acid, demonstrate a
strongly enhanced binding on platelet activation. This is directly
related to the breakdown of the phospholipid asymmetry of the
platelet surface membrane. The inner leaflet of the
nonactivated platelet membrane contains most of the
aminophospholipids and almost all of the
phosphatidylserine, but a considerable amount of
this is found in the outer monolayer after platelet
activation.6 By use of isolated platelets and
different types of platelet activators in the presence
of defined concentrations of the other reactants, a direct quantitative
relation comparing the platelet prothrombinase activity and the
percentage of the total amount of
phosphatidylserine exposed on the platelet
surface has been shown.6 The negative charge of the
phosphatidylserine, hydrophobic interactions, and
the calcium-binding properties of the vitamin Kdependent coagulation
factors are essential for the catalytic activity at the
activated platelet surface. In the presence of calcium
ions, an anticoagulant protein from placenta, annexin V, binds to the
platelet surface with a markedly increased affinity after
platelet activation. Because it is available as a recombinant
protein, the fluorescence-labeled protein is much used as a
probe in flow cytometry to demonstrate the presence of exposed
phosphatidylserine on the surface of
activated platelets.
Two phospholipid transportmediating proteins, or systems, are
postulated to maintain the steady-state phospholipid asymmetry of the
cell membrane, whereas a third one has been implicated in the fast
breakdown of the phospholipid asymmetry (scrambling) observed on
platelet activation. The first one, usually called the
aminophospholipid translocase and first described in
erythrocytes,7 is specific for aminophospholipids
(phosphatidylserine and phosphatidylethanolamine).
This protein is considered responsible for the transport of
aminophospholipids against a gradient from the outer to the inner
membrane leaflet in a process that is dependent on ATP and is inhibited
by calcium ions and sulfhydryl-blocking agents.8 A slow,
outwardly acting phospholipid-nonspecific "floppase," supposed to
counterbalance the aminophospholipid translocase, has also been
postulated.8 A bidirectional phospholipid-nonspecific
"scramblase" has been isolated from erythrocytes, cloned, and
sequenced.9 This protein is believed to be identical or
similar to the one responsible for the fast breakdown of the
phospholipid asymmetry during platelet activation.10
Human erythrocyte scramblase is a proline-rich type II membrane protein
with a short C-terminal external sequence, a single transmembrane
segment, and a long cytoplasmic extension.9 The
cytoplasmic region contains a calcium-binding segment as well as
several cysteinyl residues with at least one thioesterified fatty acid
acyl group.11 12 The biological activator is
believed to be calcium ions, the cytosolic level of which is
significantly increased on platelet activation (see Figure
).
Knowledge of the scramblase activity was to a large extent obtained by
use of platelets, erythrocytes, and lymphocytes from patients
suffering from Scott syndrome,13 14 which indicated that
the phospholipid transporters act similarly in these cells. The
syndrome is characterized first of all by a reduced ability to expose
aminophospholipids on the surface of the cells in response to an
increased cytosolic calcium ion concentration. (Clinical aspects are
discussed below.) Furthermore, scramblase activity has been studied by
use of "inside-out" vesicles from erythrocyte membranes containing
endogenous scramblase and also by use of liposomes with
detergent-extracted scramblase incorporated.10 15 The
picture that has developed from these studies is that of a phospholipid
transporter protein that is activated by calcium ions with the
activation related to a conformational alteration of the molecule.
Under experimental conditions, activation can be obtained by lowering
the pH to <6.0, even in the absence of calcium ions.15
The thioesterified fatty acid acyl group(s) in the cytoplasmic region
of the scramblase may also be involved in the
activation.11 It has been suggested that these
thioesterified fatty acid acyl groups may function by placing the
calcium-binding segment in the right position relative to the
membrane.12 In Scott syndrome, the
aminophospholipid-specific translocase appears to function
normally.16 The term scramblase has been used in the
present review even though the idea of scrambling in the sense of a
total randomization of the phospholipid distribution during
platelet activation has been challenged.17
| Microvesiculation |
|---|
|
|
|---|
-granule membrane, is also
present on the microvesicular surface.18 20 This is
explained by the hypothesis that both the phospholipid scrambling and
the secretion from the
-granules occur before the shedding of the
microvesicles. In line with this, it was observed that annexin V that
is present extracellularly during platelet activation allowed
the aminophospholipid surface to be formed but prevented the shedding
of microvesicles.18 This observation also has a bearing on
whether there is a direct causal relation between the phospholipid
scrambling and the formation of microvesicles. It suggests that the
surface expression of aminophospholipids in itself is not sufficient
for vesiculation even though it is a prerequisite for the process.
Whereas an increased surface expression of aminophospholipids can occur
without formation of microvesicles, microvesiculation has never been
reported to happen without an increased surface expression of
aminophospholipids. Protein tyrosine phosphatases are clearly involved
in microvesiculation,21 but the exact function of these
enzymes is unknown at present.
Physiologically, the surface expression of
phosphatidylserine is probably the most important
feature of the microvesicles, because this means that they are shed
with a procoagulant surface. In addition, the exposure of adhesion
molecules like P-selectin may be important in leukocyte interactions
and inflammatory reactions. | Calcium Ions in Aminophospholipid Scramblase Activity and Microvesiculation |
|---|
|
|
|---|
Whereas activation of calpain is not required for the expression of a procoagulant surface, such an activation may have an additive effect, in view of the fact that it has been shown that membrane-penetrating inhibitors of calpain can reduce the production of the procoagulant activities to some degree.22 However, whether the additive effect is related to shedding of procoagulant microvesicles is not known.
The free calcium ion concentration in the cytosol of resting
platelets is generally reported to be
10-7 mol/L, but strong activators
like the calcium ionophore A23187 can increase this >100-fold in the
presence of extracellular calcium ions. Whereas cytosolic
concentrations of
3 times the basal level were found to
activate the phospholipid scramblase but not
calpain,18 concentrations in the order of
10-5 mol/L induced both the scramblase and the
calpain activities.18 An important point is that these
activations are influenced by the extracellular calcium ion
concentration, indicating an influx of such ions during
activation.8 18 In experimental studies, calcium
ionophores and formation of the terminal complement complex C5b-9 allow
such influx because of hydrophobic permeability or passage through
membrane pores. Physiological agonists require
other mechanisms (see Figure
). It is interesting that in
addition to mobilization of calcium ions from intracellular stores, ADP
(and ATP) can activate a P2X1
purinoreceptor linked to the influx of calcium ions through a
nonselective cation channel in the platelet cytoplasmic
membrane.23 Because these 2 substances are released from
platelet "dense bodies," such an influx may take place
secondary to secretion. This may be why normal platelets increase
their cytosolic calcium ion concentration during the first 10 minutes
after an initial rise induced by a combination of collagen and
thrombin, whereas platelets from patients with dense granule
defects were not able to sustain the increased cytosolic calcium ion
concentration.24 Because the aminophospholipid translocase
is inhibited by calcium ions,8 a sustained calcium
concentration is probably necessary to maintain the scrambled surface
and the procoagulant activity (see Figure
).
| Clinical Aspects |
|---|
|
|
|---|
20% of the total amount of factor V in blood. Patients
severely deficient in plasma factor V are generally deficient in
platelet factor V as well. In a rare variant of the defect (factor
V Quebec), the platelet factor V activity was reported as <6% of
normal, whereas the plasma level was nearly normal.4 This
finding was associated with an abnormal bleeding tendency, pointing to
a role for platelet factor V in hemostasis. However, recent studies
have shown that other
-granular proteins are also proteolyzed in
this disorder,4 25 and the reason for the bleeding
tendency is unclear. The clinical condition associated to a defect in platelet procoagulant activity that is most studied and best understood is Scott syndrome.
| Scott Syndrome |
|---|
|
|
|---|
Studies involving the originally described patient16 27 28 showed that the immediate defect is a reduced ability to promote factor X and prothrombin activation that is due to a diminished surface exposure of phosphatidylserine and that this is associated with a reduced shedding of microvesicles. The defect is not associated with defects in aggregation, secretion, or granule storage function.26 29 The patients erythrocytes and lymphocytes also demonstrate the same basic defect. This is believed to reflect a deficiency in the function of the phospholipid scramblase described above, whereas the ATP- and sulfhydryl-dependent aminophospholipid translocase apparently functions normally.16 Calpain is also normally activated.19 A series of experiments indicate, however, that scramblase, as such, is a normal molecule in Scott syndrome and that the deficiency is related to a defect in the activation mechanism. Thus, the phospholipid scramblase activity of Scott erythrocyte membranes could be induced by lowering the pH to <6.0. This was also true after incorporation of the scramblase isolated from a detergent extract of Scott syndrome erythrocyte ghosts into liposomes.15 Furthermore, in this system, the phospholipid scramblase activity could be induced with the patients protein and with the protein obtained from normal erythrocytes whether activation was induced by addition of calcium ions at pH 7.4 or by acidification.15 However, inside-out vesicles from Scott erythrocytes containing the membrane-associated scramblase still demonstrated the deficiency as observed on treatment with calcium ions.15 This indicates that some additional component had been removed or modified in the detergent-extracted scramblase compared with that in the intact membrane. As already mentioned, one or more thioesterified fatty acid acyl groups in the scramblase have recently been implicated in the activation process.11
| Other Clinical Conditions With Aberrant Platelet Factor 3 Activity |
|---|
|
|
|---|
Recent studies have described a life-long bleeding disorder in 4 patients from 3 apparently unrelated families with a deficient platelet microvesiculation but a normal prothrombinase activity.31 Analyses of phospholipid distribution were not presented.
Stormorken syndrome, first described in 1985,32 has certain features that may justify its characterization as an "inverse Scott syndrome membrane anomaly." However, whereas the abnormality in Scott syndrome apparently is related to one basic anomaly, Stormorken syndrome is a multifaceted one in which the underlying cause of the abnormality is unknown. It refers to a family in which a grandmother, mother, and one son have been hospitalized several times for various ailments, including a bleeding tendency. Mother and son show essentially the same platelet deficiencies, the most prominent of which is a nearly full procoagulant activity even in the absence of stimulating agents.33 This phenomenon also corresponds to an increased binding of annexin V to their nonactivated platelets, as determined by flow cytometry. In accordance with this binding, microvesicles are present in nonactivated samples of the patients platelet-rich plasma at a higher level than that usually observed in patients with activated platelets.
In addition to the presence of platelets with a spontaneously expressed procoagulant surface and microvesicles, clot retraction is clearly reduced in these patients.33 All other coagulation and fibrinolytic activities tested were normal.33 Platelet aggregation with citrated platelet-rich plasma was in the lower normal range with all the usual agonists except collagen. This agonist gave a reduced platelet aggregation and secretion of ATP.33 However, these studies were complicated by a tendency to spontaneous platelet aggregation in whole blood,33 which may result in a loss of platelets and a selected platelet population after centrifugation. Normal amounts of glycoprotein IIb/IIIa complexes on the platelet surface but a reduced binding of the activation-dependent antibody PAC-1 to glycoprotein IIb/IIIa after platelet activation indicate that some of these aggregation receptors might be in a refractory state. Paradoxically, whereas the existence of the platelets in a "procoagulant" state would predict an increased thrombotic predisposition, the clinical picture of the syndrome is that of a moderate bleeding tendency. The clinical bleeding corresponds, however, to the situation observed under shear in a flow chamber. By use of purified human collagen type III as a trigger for thrombus formation, a clearly reduced thrombus volume was observed at shear rates of 650 s-1 and 2600 s-1, whereas platelet adhesion to the collagen surface was higher than normal.33 Finally, it should be added that the multifaceted aspects of the syndrome include asplenia, reduced platelet survival time, miosis, dyslexia, muscle fatigue, and ichthyosis.32 One may speculate whether the high number of microvesicles in plasma reflects the congenital lack of spleen. Surface exposure of phosphatidylserine is considered a signal for clearance for many cells. However, the platelet survival time is clearly reduced in these patients.33
Another clinical condition associated with bleeding in spite of an increased procoagulant platelet surface is Bernard Soulier syndrome.34 However, the giant size of the platelets and low platelet count complicate the interpretation of this observation. It is generally believed that the hemostatic defect in this syndrome is related to absence of the adhesion receptor complex glycoprotein Ib/V/IX.
| Activated Platelets and Microvesicles in Various Clinical Conditions |
|---|
|
|
|---|
5%) of the
platelets (an exception is Stormorken syndrome, in which the
percentage of activated platelets is significantly higher).
Even so, it should be noted that microvesicles are able to bind soluble
fibrinogen, bind to immobilized fibrinogen, and coaggregate
with platelets.38 Microvesicles in plasma are normally observed and quantified by use of flow cytometry with the instrument gated with a platelet-specific probe so that only platelet-derived microvesicles are detected. One should keep in mind then that apoptosis in other cells also leads to breakdown of the phospholipid asymmetry, the formation of procoagulant surface blebs,39 and shedding of such membrane fragments, all of which go undetected when this approach is used. The extent and significance of nonplatelet microvesicles in the circulation in disease are not known at present.
It has recently been reported that platelets contain CD40L and express this on the surface after activation.40 This protein acts as a ligand for CD40 on B cells, monocytes, macrophages, and endothelial cells, triggering inflammatory reactions. We have confirmed the existence of this protein in platelets and found that CD40L from activated platelets is also present in a biologically active soluble form after secretion. In general, activation of endothelial cells and monocytes is known to induce synthesis and exposure of tissue factor. Because there seems to be some link between inflammation and coagulation, it should also be mentioned that platelets contain and release platelet-derived growth factor as well as chemokines, such as platelet factor 4, platelet basic protein/NAP-2, and RANTES.37
Received February 1, 1999; accepted April 21, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
K. Fischer, S. Voelkl, J. Berger, R. Andreesen, T. Pomorski, and A. Mackensen Antigen recognition induces phosphatidylserine exposure on the cell surface of human CD8+ T cells Blood, December 15, 2006; 108(13): 4094 - 4101. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Jayachandran, K. Karnicki, R. S. Miller, W. G. Owen, K. S. Korach, and V. M. Miller Platelet Characteristics Change With Aging: Role of Estrogen Receptor {beta} J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2005; 60(7): 815 - 819. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Ferenac, D. Polancec, M. Huzak, O. M. Pereira-Smith, and I. Rubelj Early-Senescing Human Skin Fibroblasts Do Not Demonstrate Accelerated Telomere Shortening J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2005; 60(7): 820 - 829. [Abstract] [Full Text] [PDF] |
||||
![]() |
I. del Conde, F. Nabi, R. Tonda, P. Thiagarajan, J. A. Lopez, and N. S. Kleiman Effect of P-Selectin on Phosphatidylserine Exposure and Surface-Dependent Thrombin Generation on Monocytes Arterioscler. Thromb. Vasc. Biol., May 1, 2005; 25(5): 1065 - 1070. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. OSTERUD and E. BJORKLID Role of Monocytes in Atherogenesis Physiol Rev, October 1, 2003; 83(4): 1069 - 1112. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. M. Monroe, M. Hoffman, and H. R. Roberts Platelets and Thrombin Generation Arterioscler. Thromb. Vasc. Biol., September 1, 2002; 22(9): 1381 - 1389. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. B. Brooks, J. L. Catalfamo, H. A. Brown, P. Ivanova, and J. Lovaglio A hereditary bleeding disorder of dogs caused by a lack of platelet procoagulant activity Blood, April 1, 2002; 99(7): 2434 - 2441. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. H. M. Smits, J. van der Linden, P. J. Blankestijn, and T. J. Rabelink Coagulation and haemodialysis access thrombosis Nephrol. Dial. Transplant., November 1, 2000; 15(11): 1755 - 1760. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1999 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |