Low Protein Z Associated with Venous Thrombophilia and with Ischemic Stroke

Normal blood coagulation provides rapid clot formation at places of interrupted vessel walls, surgical wounds, or traumatic tissue destruction. The biological symphony of coordinated procoagulant and anticoagulant activities guarantees fluidity of circulating blood through the intact arterial and venous vessels.[1] There are a number of established mechanisms by which the coagulant pathways are regulated by protease inhibitors, which allow intrinsic thrombin generation at a low physiological level (Fig. [1]). The tissue factor pathway inhibitor (TFPI) will block the extrinsic pathway of thrombin generation by forming an inactive complex with factor VIIa, tissue factor (TF), and factor Xa (Fig. [1]). Within the intact vessel wall, thrombin binds to thrombomodulin bound on the surface of endothelial cells by which thrombin converts from the most potent procoagulant factor into a potent anticoagulant factor in inhibiting itself by activating of protein C to activated protein C (APC) (Fig. [1]). APC in concert with its cofactor protein S (PS) inactivates factors VIIIa and Va (Fig. [1]).

TFPI deleted mice and protein C (PC) deleted mice are lethal due to microvascular thrombosis.[2] [3] Factor V Leiden (FVL) deleted mice express a strain-specific thrombotic phenotype; about 20% of FVL deleted mice die during the neonatal period with microvascular thrombosis.[4] Protein Z (PZ) deleted mice produce a grossly normal phenotype with no increased risk of thrombosis or bleeding at least in the absence of a thrombotic challenge.[4] Heterozygous FVL, TFPI, and PC are also asymptomatic in the unchallenged state.

PZ deleted mice markedly increase the mortality of FVL deleted mice to a nearly completely lethal condition.[4] The PZ heterozygous mice also affect the FVL homozygous mice and the survival of FVL heterozygous/PZ homozygous mice is significantly less than the FVL heterozygous/PZ heterozygous mice. Antifibrin(ogen) immunohistochemical analysis of the FVL deleted mice showed vascular thrombosis and hepatic fibrin deposition, the scope of which was related directly to the PZ genotype: PZ -/-> PZ +/-> PZ +/+. Thus the level of PZ affects the phenotypes of FVL mice. No such data are available for the influence of PZ deleted mice on FVIII deleted or FIX deleted mice whether the level of PZ modifies the bleeding diathesis. Broze et al. have crossed PZ-/-mice with factor XI-/- mice. The factor XI-/- mice do not have an obvious hemorrhagic phenotype. The combination of factor XI-/- and PZ-/- mice appeared normal and did not bleed abnormally in a tail vein bleeding time or following amputation of 2 cm of tail (George J. Broze, Jr, personal communication, April 8, 2002). Preliminary studies in humans[5] demonstrated that FVL patients with PZ levels below 1000 ug/l presented with significantly earlier age of thrombotic manifestation and that the frequency of thromboembolic events was significantly higher than it was in patients with PZ above 1000 ug/l. The question to be addressed is whether the low level of PZ is an additional cause and/or the consequence (consumed in the activated coagulation) of FVL thrombophilia.

PZ promotes the assembly of thrombin (IIa) with phospholipid vesicles in a Ca-dependent fashion, which might promote coagulation (procoagulant activity).[6] [7] Hogg and Stenflo showed in the bovine system that inactivated IIa bound to PZ with a reasonable Kd.[6] [7] Because of this finding it was predicted that PZ deficency may cause a bleeding tendency; however the binding of PZ to inactivated IIa was due to the fact that bovine PZ contains a carboxyterminal extension which IIa removes by proteolytic cleavage. The binding of inactivated IIa to bovine PZ appears to simple represent the binding of an enzyme to its substrate. When active IIa interacts with bovine PZ, it cleaves at this ``binding site'' and stable binding presurability would not occur. Human PZ does not bind inactivated IIa because it lacks the C-terminal extension present in bovine PZ (George J. Broze, Jr, personal communication, April 8, 2002). In 1995 Kemkes-Matthes and Matthes[8] reported that 21 of 36 patients with an idiopathic bleeding tendency had PZ levels below the lowest value of the healthy control group. Typical bleedings in patients with low PZ levels were positive Rumpel-Leede test, hematoma, and postoperative bleeding. The phenomenon of a positive Rumpel-Leede test may suggest whether PZ is physiologically involved in the interaction of thrombin and phospholipids surface of activated platelets (thrombin-induced platelet activation) at sites of endothelial cell damage. Gamba et al.[9] found normal PZ levels in 15 patients with idiopathic bleeding tendency. In a recent study of 200 healthy controls and 48 patients with a bleeding tendency, low PZ levels were not associated with bleeding.[10] Ten percent of normal blood donors have PZ<50% and do not bleed. This, plus the observations by Gambda et al.[9] and Ravi et al.[8], make PZ deficiency an unlikely cause of bleeding despite the data of Kemkes-Matthes and Matthes.[8]

In the presence of PZ, calcium, and phospholipids, PZ-dependent protease inhibitor causes fast inactivation of factor Xa demonstrating the anticoagulant potential of PZ.[11] Similar to the vitamin K-dependent procoagulant factors and anticoagulant factors, PZ is consumed during consumption coagulopathy. The PZ-dependent inactivation of FXa bound to phospholipids may precede or regulate the thrombin levels in vivo as well as the thrombin-induced factor VIII- and V-activation. It therefore may play a key role in the regulation (fine tuning) of the procoagulant-anticoagulant balance in the early initiation phase of blood coagulation, platelet, and vessel wall interaction at the level of the prothrombinase complex (Fig. [1], PI = phospholipid on the platelet surface).

Whether this anticoagulant activity of PZ plays a role on the phospholipid surface of activated platelets at places of damaged endothelial cells of arteries or microvasculature remains elusive. Subsequent thrombomoduline/thrombine-induced PC activation on intact endothelial cells is followed by the inactivation of FVa and this feedback mechanism is of physiological relevance to anticoagulate the thrombotic potential of circulating IIa (Fig. [1]).[1] In addition, antithrombin (AT) inhibits not only the circulating Xa and IIa but also IXa and XIa, which is enhanced several thousandfold by heparin as one of the most potent therapeutic anticoagulant agents (Fig. [1]).[1] Homozygous AT deficiency is not compatible with human life. Homozygous PC deficiency presents with congenital purpura fulminans, and homozygous FVL is associated with venous thrombophilia at adult age. In the coagulation concept presented above, dysfunctional and/or low levels of PZ only play a minor role as a risk factor for venous thrombophilia and may play a more prominent role as a risk factor for arterial thrombophilia. Further research on the physiological role of PZ as an anticoagulant factor in the regulation of blood coagulation is needed.

The results of a recent case control study indicate that inherited thrombophilia due to FVL or the prothrombin G20210A genotype are not associated with an increased risk of ischemic stroke in 100 young patients aged less than 45 years as compared to 238 control subjects.[12] In an elegant study, Vasse et al.[13] reported a PZ plasma deficiency in about 20% in 169 patients (as compared to 5% in 88 controls) with idiopathic ischemic stroke 3 months after the thrombotic event in the absence of vascular risk factors like dyslipidemia, hypertension, vascular malformation, or cardiac abnormalities. The mean value of PZ in 56 patients with a previous history of deep venous thrombosis did not significantly differ from the control group and three (5%) had PZ deficiency, a frequency similar to the healthy controls. A second measurement in 12 patients confirmed a persistent PZ deficiency in all 12 patients. A family study of three patients suggested an inherited deficiency. The authors conclude that PZ deficiency may be involved in thrombotic complications of atherosclerosis and reinforce the concept of a role of PZ as an anticoagulant factor in the regulation of blood coagulation, platelet, and vessel wall interaction. Low PZ plasma levels may become an independent risk factor for thrombotic complications in relatively early stages of idiopathic cerebral ischemic disease like transient ischemic attacks (TIAs) and minor stroke. A proper, well-designed, large, prospective population-based case control study in subgroups of patients with idiopathic TIAs and minor stroke is warranted.

REFERENCES

• 1 Michiels J J, Hamulyak K. Laboratory diagnosis of hereditary thrombophilia.  Semin Thromb Hemost . 1998;  24 309-320
• 2 Huang Z-F, Higuchi D, Lasky N, Broze G J. Tissue factor pathway inhibitor gene disruption produces intrauterine lethality in mice.  Blood . 1997;  90 944-951
• 3 Jalbert L R, Rosen E D, Moons L. Inactivation of the gene for anticoagulant protein C causes lethal perinatal consumptive coagulopathy in mice.  J Clin Invest . 1998;  102 1481-1488
• 4 Yin Z-F, Huang Z-F, Cui J. Prothrombotic phenotype of protein Z deficiency.  PNAS . 2000;  97 6734-6738
• 5 Kemkes-Matthes B, Matthes K J. Protein Z.  Semin Thromb Hemost . 2001;  27 551-556
• 6 Hogg P J, Stenflo J. Interaction of vitamin K-dependent protein Z with thrombin.  J Biol Biochem . 1991;  266 10953-10958
• 7 Hogg P J, Stenflo J. Interaction of human protein Z with thrombin.  Biochem Biophys Res Comm . 1991;  178 801-807
• 8 Kemkes-Matthes B, Matthes K J. Protein Z deficiency: a new cause of bleeding tendency.  Thromb Res . 1995;  79 49-55
• 9 Gambda G, Bertilino G, Montani N, Spedini P, Balduini C L. Bleeding tendency of unknown origin and protein Z levels.  Thromb Res . 1998;  90 291-295
• 10 Ravi S, Mauron T H, Lämmle B, Wuilleman W A. Protein Z in healthy human individuals and in patients with a bleeding tendency.  Brit J Haematol . 1998;  102 1219-1223
• 11 Han X, Fiehler R, Broze G J. Isolation of a protein Z-dependent plasma protease inhibitor.  Proc Natl Acad Sci USA . 1998;  95 9250-9255
• 12 Lopaciauk S, Bykowska K, Kwiecinsky H. Factor V Leiden, prothrombin G20210A variant, and methylenetetrahydrofolate reductase C677T genotype in young adults with ischemic stroke.  Clin Appl Thromb Hemost . 2001;  7 346-350
• 13 Vasse M, Guegan-Massardier, Borg J-Y, Woimant F, Soria C. Frequency of protein Z deficiency in patients with ischaemic stroke.  Lancet . 2001;  357 933-934