Abstract
Protein Z (PZ) is a vitamin K-dependent factor characterized by its homology to other vitamin K-dependent factors (factors VII, IX, and X, protein C and protein S), but lacks any enzymatic activity. Instead, PZ acts as a cofactor for the inhibition of factor Xa through the serpin PZ-dependent protease inhibitor (ZPI). PZ deficiency is associated with a procoagulant state, highlighted by excessive FXa secretion and thrombin production, and is linked with several thrombotic disorders, including arterial vascular and venous thromboembolic diseases. A role for the PZ–ZPI complex in the regulation of physiological pregnancy has been demonstrated, highlighted by the progressive elevation in PZ levels in the first trimester of gestation, which then steadily decline toward delivery. An association between altered plasma PZ concentrations and adverse pregnancy outcomes (recurrent miscarriage, stillbirth, preeclampsia, intrauterine growth restriction, and placental abruption) has been reported. The mechanism by which PZ deficiency leads to adverse pregnancy outcomes is not clear, but it is multifactorial. It may be attributed to the anti-PZ IgG and IgM autoantibodies, which apparently act independently of classical antiphospholipid antibodies (lupus anticoagulant, anticardiolipin, and anti-β2-glycoprotein I antibodies). PZ deficiency has also been reported to be constitutional, and a number of variants in the PROZ (PZ) gene and SERPINA10 (ZPI) gene are linked with specific adverse pregnancy complications. This review summarizes the relationship between adverse pregnancy outcomes and acquired and constitutional PZ–ZPI deficiency, in order to understand whether or not PZ deficiency could be considered as a risk factor for poor pregnancy outcomes.
Introduction
The vitamin K-dependent anticoagulant plasma glycoprotein protein Z (PZ) was first isolated from bovine plasma in 1977 and from human plasma in 1984 and was shown to play an important role in the regulation of the coagulation cascade (Broze 2001, Vasse 2008). While in vitro studies have shown that bovine PZ could promote the assembly of thrombin with phospholipid surfaces, thereby enhancing coagulation, the human PZ form binds to thrombin poorly, with very little effect on the association of thrombin binding with phospholipids. More recent studies have shown that PZ forms a calcium ion-dependent complex with factor Xa on phospholipid surfaces, thereby serving as a cofactor for the inhibition of factor (F) Xa through a PZ-dependent protease inhibitor (ZPI; Huang et al. 2012).
Deficiency in PZ secretion and/or function is linked with a procoagulant state and several thrombotic disorders, including arterial and venous thrombosis. As the outcome of pregnancy is dictated to a large extent by the maintenance of adequate maternal and fetal blood circulation and as coagulation abnormalities are associated with adverse pregnancy outcomes, a role for the PZ–ZPI complex in the regulation of pregnancy has been suggested, and an association between altered plasma PZ levels and adverse pregnancy outcomes has been reported, often with apparently contradictory conclusions. This review summarizes the relationship between acquired and constitutional PZ–ZPI deficiency and adverse pregnancy outcomes, in particular, whether PZ deficiency could be a risk factor for poor pregnancy outcomes.
Biochemistry of PZ
PZ is a 62 kDa vitamin K-dependent single-chain glycoprotein, consisting of 360 amino acids containing a N-terminal γ-carboxyglutamic acid (Gla) domain necessary for its effective secretion (Souri et al. 2009), followed by two epidermal growth factor-like domains (light chain homolog), and a C-terminal pseudo-catalytic domain (heavy chain homolog) (Vasse 2008; Fig. 1). Although structurally related to other coagulation serine proteases (FVIIa, FIXa, FXa, and activated protein C), PZ lacks significant proteinase activity due to the presence of only Asp in its active center and lack of histidine and serine residues in the catalytic triad (replaced by Ala and Thr residues respectively).
Organization of the PZ (PROZ) and ZPI genes and protein products.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
Organization of the PZ (PROZ) and ZPI genes and protein products.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
Organization of the PZ (PROZ) and ZPI genes and protein products.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
PZ acts as a cofactor for the 72 kDa serpin ZPI, which rapidly (t½ <10 s) inhibits FXa and FXIa (Heeb et al. 2005, Huang et al. 2012), hence reducing thrombin generation (Koren-Michowitz et al. 2006, Vasse 2008, 2011; Fig. 2). Relative to PZ, ZPI is present in the plasma at higher levels (Han et al. 2000, Tabatabai et al. 2001), where it binds to all PZ at a 1:1 ratio (Han et al. 2000, Tabatabai et al. 2001), hence circulating as a stable PZ–ZPI complex, with virtually no free PZ being detected (Corral et al. 2007, Vasse 2008). Compared with other vitamin K-dependent factors, PZ exhibits 100-fold slower membrane binding and dissociation kinetics, due to the presence of an additional Gla residue at position 11 in the PROZ (PZ) protein (Vasse 2008, Souri et al. 2009).
ZPI–PZ complex. (A) Crystal structure of the ZPI (blue)–PZ (purple) complex (PDB ID: 3H5C). (B) ZPI–PZ complex interaction domains showing amino acids interacting at the interface. Labeled interacting amino acid residues (shown as stick representations) of ZPI are K68, D74, D238, K239, N261, D292, and T296, while those of PZ are H210, R212, E219, E244, H250, R298, R350 and Q357 and are colored by specific elements (carbon, green; nitrogen, blue; and oxygen, red). Images were prepared using PyMol.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
ZPI–PZ complex. (A) Crystal structure of the ZPI (blue)–PZ (purple) complex (PDB ID: 3H5C). (B) ZPI–PZ complex interaction domains showing amino acids interacting at the interface. Labeled interacting amino acid residues (shown as stick representations) of ZPI are K68, D74, D238, K239, N261, D292, and T296, while those of PZ are H210, R212, E219, E244, H250, R298, R350 and Q357 and are colored by specific elements (carbon, green; nitrogen, blue; and oxygen, red). Images were prepared using PyMol.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
ZPI–PZ complex. (A) Crystal structure of the ZPI (blue)–PZ (purple) complex (PDB ID: 3H5C). (B) ZPI–PZ complex interaction domains showing amino acids interacting at the interface. Labeled interacting amino acid residues (shown as stick representations) of ZPI are K68, D74, D238, K239, N261, D292, and T296, while those of PZ are H210, R212, E219, E244, H250, R298, R350 and Q357 and are colored by specific elements (carbon, green; nitrogen, blue; and oxygen, red). Images were prepared using PyMol.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
Human PROZ gene is located on chromosome 13q34, in proximity to F8 (FVII) and TSTA3 (FX) genes (Fig. 1). The PROZ gene spans 15 kb and is organized into nine exons, including an alternative exon (Souri et al. 2009). The liver is the main source of PZ, and plasma PZ levels are reduced in patients with chronic liver diseases (Kemkes-Matthes & Matthes 1995). Conflicting findings have been reported on the production of PZ by human endothelial cells (Kusanovic et al. 2007). In the liver, 30% of the synthesized PZ is converted to Gla inside the cells, before it is secreted into the plasma in a vitamin K- and Gla-30-dependent process (Vasse 2008, Souri et al. 2009). While PZ has a long half-life (2.5 days; Miletich & Broze 1987, Kusanovic et al. 2007), the half-life of the PZ–ZPI complex compared with that of either free ZPI or free PZ remains to be established (Kusanovic et al. 2007).
PZ function
The PZ–ZPI complex exerts its anticoagulant effect through the inactivation of phospholipid-bound FXa. While ZPI can inhibit FXa, its complexing with PZ accelerates the ZPI-mediated inhibition of FXa by 1000-fold (Al-Shanqeeti et al. 2005, Koren-Michowitz et al. 2006). Several mechanisms by which PZ acts as a cofactor in the modulation of the activity of ZPI, which include direct interaction of PZ with both FXa and ZPI at phospholipid surfaces (Dayer et al. 2012), forming the FXa–ZPI–PZ complex at the phospholipid surfaces (Dayer et al. 2012), have been postulated (Fig. 3). Specific interactions between the PZ Gla domain and the FXa Gla domain (Gla–Gla interaction) have been suggested to accelerate the inhibition rate (Huang et al. 2010, 2012, Dayer et al. 2012). PZ has also been suggested to induce structural changes in ZPI (Huang et al. 2012), whereby PZ aligns the inhibitory site of ZPI with the active site of FXa (Huang et al. 2010, Karimi et al. 2012; Fig. 3). This alters the secretion, localization, and clearance of ZPI (Broze 2001), hence facilitating the interaction between ZPI and FXa. Irrespective of the mechanism, the PZ–ZPI complex prevents thrombin generation in the early phases of coagulation, before the formation of the prothrombinase complex (Huang et al. 2010, 2012).
Two mechanisms proposed for the PZ-mediated inhibition of factor Xa by ZPI. (Left) Pre-formed circulating PZ–ZPI complex binds to factor Xa bound to phospholipid surfaces through the interaction of the Gla domain of PZ (in orange) and interacts with the Gla domain of factor Xa (in orange), (Right) ZPI is activated by glycosaminoglycans found on endothelial cell surface and as such may directly engage factor Xa and PZ.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
Two mechanisms proposed for the PZ-mediated inhibition of factor Xa by ZPI. (Left) Pre-formed circulating PZ–ZPI complex binds to factor Xa bound to phospholipid surfaces through the interaction of the Gla domain of PZ (in orange) and interacts with the Gla domain of factor Xa (in orange), (Right) ZPI is activated by glycosaminoglycans found on endothelial cell surface and as such may directly engage factor Xa and PZ.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
Two mechanisms proposed for the PZ-mediated inhibition of factor Xa by ZPI. (Left) Pre-formed circulating PZ–ZPI complex binds to factor Xa bound to phospholipid surfaces through the interaction of the Gla domain of PZ (in orange) and interacts with the Gla domain of factor Xa (in orange), (Right) ZPI is activated by glycosaminoglycans found on endothelial cell surface and as such may directly engage factor Xa and PZ.
Citation: REPRODUCTION 146, 2; 10.1530/REP-13-0072
The interaction between ZPI and FXa is a reversible event (Han et al. 2000, Heeb et al. 2005). In contrast to other serpin complexes, ZPI is proteolytically and rapidly cleaved at its C-terminus, which reduces its size from 72 to 68 kDa, resulting in an inactive ZPI with little or no functional activity (Broze 2001, Huang et al. 2010, 2012), and hence low free FXa levels for ZPI binding (Broze 2001). The resultant ZPI constitutes a poor substrate for the FXa–PZ complex (Broze 2001, Huang et al. 2012).
PZ synthesis and plasma levels
Mean plasma PZ levels range from 1.16 to 2.71 μg/ml (Miletich & Broze 1987) and are influenced by genetic and non-genetic factors, which include chronic liver diseases, age, gender, vitamin K levels, concurrent use of anticoagulants, and pregnancy (Al-Shanqeeti et al. 2005, Vasse 2008, Souri et al. 2009). Variations in PZ levels have been reported with regard to age, gender, and ethnic origin. Plasma PZ levels rapidly increase during the first few months of childhood and then slowly taper off, with adult levels being reached during puberty (Miletich & Broze 1987, Gowri et al. 2011), and higher PZ levels have been reported in men than in women (Miletich & Broze 1987). PZ levels are reduced in inflammation (Krawiec et al. 2011), and a negative correlation between PZ levels and plasma interleukin 1 (IL1) and IL6 levels has been reported (Undar et al. 1999, Vasse et al. 2002). Contradictory findings have also been reported, with both increased and decreased PZ levels being observed in obese individuals, which have been attributed to the altered expression of inflammatory cytokines associated with obesity (Ramsay et al. 2005), presence of anti-PZ antibodies (Pardos-Gea et al. 2008, Sater et al. 2011), ethnic background of the study subjects, and possibly increased excretion of PZ.
Among the studies on PZ levels, only four studies have addressed plasma ZPI levels, but with inconclusive findings. An association between reduced ZPI levels and peripheral arterial disease (Sofi et al. 2010) and venous thrombosis (Al-Shanqeeti et al. 2005) has been documented, while no difference in ZPI levels has been reported between normal controls and antiphospholipid antibody-positive cases. A recent report by Souri et al. (2012) has demonstrated higher plasma ZPI concentrations (measured by in-house ELISA) in pregnant women than in non-pregnant women, which was paralleled by increased FX levels, which were suggested to contribute to the maintenance of adequate placental circulation. Whereas plasma ZPI levels did not change in non-pregnant women with a history of recurrent miscarriage, plasma PZ levels were slightly reduced, suggesting a link between recurrent miscarriage and this paradoxical unaltered ZPI–mild PZ reduction, when compared with normal pregnancies (Souri et al. 2012).
PZ deficiency
PZ deficiency (<1 μg/ml) is linked with a procoagulant state, associated with excessive FXa secretion and thrombin production (Al-Shanqeeti et al. 2005, Koren-Michowitz et al. 2006). A recent meta-analysis has demonstrated that reduced PZ levels are linked with an increased risk of thrombotic events, such as arterial vascular and venous thromboembolic diseases (Sofi et al. 2010), and pregnancy complications. The latter include preeclampsia (PE; Erez et al. 2007, 2009), early fetal death (Vasse 2011), intrauterine growth restriction (Bretelle et al. 2005), and recurrent spontaneous miscarriage (Topalidou et al. 2009, AlShaikh et al. 2013).
Whether PZ deficiency constitutes an independent risk factor for thrombosis is inconclusive (Vasse et al. 2002, Al-Shanqeeti et al. 2005, Martinelli et al. 2005), since the association of PZ deficiency with an increased risk of thrombosis has been reported in the presence of other prothrombotic risk factors such as FV Leiden (Kemkes-Matthes & Matthes 1995, Martinelli et al. 2005), prothrombin G20210A mutation, and hyperhomocysteinemia (Martinelli et al. 2005). It has been further suggested that PZ deficiency does not constitute an independent risk factor for venous thromboembolism (VTE), but only increases the VTE risk with FV Leiden. Furthermore, a rare inherited coagulation disorder, hereditary combined vitamin K-dependent clotting factor deficiency (VKCFD), has been reported (Napolitano et al. 2010). This disorder results in a deficiency of the clotting factors FII, FVII, FIX, and FX and the coagulation inhibitors protein C, protein S, and PZ. VKCFD is linked with bleeding tendency with a variegate clinical picture and results from mutations of two enzymes of the vitamin K cycle: (type 1) defective γ-glutamyl carboxylase (Soute et al. 1992) or (type 2) functional deficiency in vitamin K 2,3-epoxide reductase complex (Oldenburg et al. 2000).
PZ in pregnancy
Pregnancy is associated with a state of hypercoagulation linked with excessive thrombin generation, which is crucial for controlling bleeding at delivery (Kist et al. 2008, Pabingert 2008). Growing evidence implicates coagulation abnormalities in adverse pregnancy outcomes, including recurrent and non-recurrent pregnancy losses (Pabingert 2008, Gris 2009), intrauterine growth retardation, placental abruption (Alfirevic et al. 2002), intrauterine fetal death, PE, and maternal or neonatal thrombosis (Michels & Tiu 2007). This has been evidenced by the reported increases in the levels of the clotting factors FVIII, FX, and Von Willebrand factor, along with those of FVII, which increase by 200% compared with pre-pregnancy levels, and those of fibrinogen, which gradually increase till they reach 1000% (Thornton & Douglas 2010). In addition, a decreased quantity of natural anticoagulants, such as protein S and protein C, and a reduction in the overall fibrinolytic activity accompany most pregnancy complications (Thornton & Douglas 2010). Both heritable and acquired thrombophilias have been implicated in pregnancy-associated hypercoagulation (Kist et al. 2008, Pabingert 2008), which include antithrombin III, protein C, and protein S deficiencies; altered activity of procoagulant factors, in particular, those precipitated by FV Leiden; and the prothrombin G20210A mutations. On the other hand, antiphospholipid antibodies (lupus anticoagulant and anticardiolipin antibodies) are responsible for the most common acquired thrombophilias linked with adverse pregnancy outcomes.
Progressively higher PZ levels have been observed with increasing gestational age in normal pregnancies, especially among obese pregnant women (Ramsay et al. 2005), which return to normal levels at around 6–12 weeks (Thornton & Douglas 2010), and have been attributed to pregnancy-associated imbalance of fibrinolytic and hemostatic mechanisms (Kusanovic et al. 2007). PZ levels correlate with gestational age, evidenced by the progressive increases in PZ levels (20%) during the three trimesters of pregnancy, which can be attributed to the compensatory mechanisms induced by increased FXa concentrations (Erez et al. 2007), and they decline thereafter by as much as 30%, to levels below those observed in the first trimester. The increase in PZ levels becomes attenuated in patients with abnormal pregnancy outcomes, including low-birth-weight babies or pregnancy-associated hypertension or diabetes (Gowri et al. 2011).
PZ in pregnancy complications
Insofar as PZ deficiency represents a procoagulant state and as pregnancy is linked with an aggravation of the procoagulant state that translates into adverse complications or poor outcomes, an association between altered plasma PZ concentrations and pregnancy complications has been reported. Reduced PZ levels (<1 μg/ml) have been linked with several pregnancy complications including fetal demise (Erez et al. 2009), intrauterine growth restriction (Bretelle et al. 2005), PE (Erez et al. 2009), small for gestational age (SGA) (Erez et al. 2009), HELLP syndrome (H, hemolysis; EL, elevated liver enzymes; LP, low platelet counts), which represents a severe form of PE and can be observed with or without preceding PE (Kaygusuz et al. 2011), and idiopathic recurrent miscarriage (Gris et al. 2002, AlShaikh et al. 2013).
An earlier French study has demonstrated a high prevalence of PZ deficiency in women with a first primary early (10–15 weeks of gestation) miscarriage, but not with recurrent embryonic loss (before 8 weeks of gestation), which is distinct from classical thrombophilia (Gris et al. 2002). This deficiency is unrelated to deficiencies of other vitamin K-dependent coagulation factors, including protein C and FVII, and persists despite vitamin supplementation (Gris et al. 2002). PZ deficiency has thus been proposed as a significant determinant of adverse pregnancy complication-associated thrombophilia. In light of changes in maternal–fetal circulation in the first trimester of pregnancy, the authors have suggested that PZ deficiency most probably favors a state of local thrombogenesis (Gris et al. 2002). PZ deficiency has been subsequently attributed to the presence of anti-PZ-specific IgG and IgM antibodies, the titers of which are inversely correlated with PZ concentrations in patients with recurrent fetal losses and with PZ deficiency (Gris et al. 2003). We later confirmed this in Bahraini women with idiopathic recurrent miscarriage (Sater et al. 2011).
More recent studies, including a meta-analysis (Sofi et al. 2010) and our case–control Bahraini study (AlShaikh et al. 2013), have confirmed the strong relationship between low PZ levels and adverse pregnancy complications. The meta-analysis of Sofi et al. (2010) involving 714 patients and 515 controls has demonstrated a strong association of PZ deficiency with pregnancy complications (OR (95% CI)=3.42 (2.51–4.66)). The study of AlShaikh et al. (2013) on 282 recurrent miscarriage cases and 281 control women has also demonstrated an almost fourfold increased risk of fetal loss with PZ deficiency, which is influenced by the specific PROZ genotypes (see below). PZ deficiency has also been observed in women with the HELLP syndrome; median PZ levels in patients with the HELLP syndrome and PZ levels correlate with platelet counts and changes in liver enzyme (LDH and AST) levels, thus prompting the speculation that this may be a consequence of a preceding liver dysfunction (Kemkes-Matthes & Matthes 1995).
The lack of a relationship between PZ deficiency and pregnancy complications has also been reported by smaller studies. A high prevalence of PZ deficiency is associated with PE, evidenced by lower maternal plasma PZ concentrations in PE women than in women with normal pregnancies (Paidas et al. 2005, Erez et al. 2007). This was in contrast to the findings of the Brettele study in which median plasma PZ concentrations were similar for PE patients and women with uncomplicated pregnancies (Bretelle et al. 2005). This has been attributed to the small sample size (50 non-pregnant and 34 healthy pregnant control women and 61 women with complicated pregnancies) and heterogeneity in patient presentation (PE, intrauterine growth restriction, and intrauterine fetal demise) and also to differences in ethnicity (Bretelle et al. 2005). The study of Grandone et al. (2004) has also reported that PZ deficiency is not linked with unexplained fetal loss, which is due to the small sample size, low plasma PZ cut-off values (1.43±0.76 μg/ml in healthy controls), and exclusion of women with known inherited (FV Leiden or FII G20210A mutations and protein C, protein S, or antithrombin deficiency) or acquired (antiphospholipid antibodies) thrombophilia. The above-mentioned studies involved lower numbers of subjects compared with the studies of Gris et al. (2002, 2003) and Sofi et al. (2010), and our study (Sater et al. 2011, AlShaikh et al. 2013), indicating study under-power. Interestingly, both the Bretelle et al. (2005) and Erez et al. (2007) studies have been reported as the main contributors to statistical heterogeneity for pregnancy complications in the meta-analysis of Sofi et al. (2010) (P for heterogeneity from 0.002 overall to 0.17 after their exclusions).
PZ autoantibodies and adverse pregnancy outcomes
The mechanism by which PZ deficiency leads to poor pregnancy outcomes is not clear and may be attributed to anti-PZ IgG and IgM autoantibodies (Gris et al. 2003, Sater et al. 2011) and the presence of functional mutations in the PROZ gene, in particular, in the G79A variant (Dossenbach-Glaninger et al. 2008, El-Hamid & El-Khayat 2011, AlShaikh et al. 2013). An earlier report by Gris involving 171 women with pathological pregnancies and 191 multiparous control women has demonstrated high levels of anti-PZ IgG and IgM antibodies, which are distinct from classical antiphospholipid/anticofactor antibodies, and a dose–effect relationship between anti-PZ antibody levels and poor pregnancy outcomes has been documented (Grandone et al. 2004). The association of high anti-PZ IgG and IgM autoantibody titers with poor pregnancy outcomes has been independently confirmed later in different populations (Paidas et al. 2005, Erez et al. 2009, Sater et al. 2011). Although the association between PZ levels and the presence of these autoantibodies has not been confirmed by all studies (Paidas et al. 2005), the combination of PZ deficiency and high anti-PZ autoantibody titers has been linked with an increased risk of pregnancy loss (Gris et al. 2003, Kusanovic et al. 2007).
Two small independent studies have yielded contradictory findings (Sailer et al. 2008, Erez et al. 2009). Though it did not reach statistical significance, the study of Sailer et al. (2008) has reported a trend to significance in the association of anti-PZ antibodies with adverse pregnancy outcomes. This is probably due to the low number of cases included, which resulted in adopting the 75th percentile of control subjects as the upper limit of the normal range for comparison, since the number of subjects whose antibody levels exceeded the 90th percentile of the controls was very low (Sailer et al. 2008). The study of Erez et al. (2009) involving 51 women has suggested that heightened anti-PZ antibody levels are not associated with fetal death, but rather with SGA, and that a high maternal anti-PZ IgM titer is linked with vascular placental lesions in PE patients but not in SGA neonate patients, thus prompting the conclusion that the pathological effects of anti-PZ antibodies are observed in select patients. In these studies, anti-PZ antibodies have been detected in varying titers in non-pregnant patients, thereby raising the speculation that anti-PZ antibodies constitute natural antibodies (Erez et al. 2009). Apart from the study of Gris et al. (2003), most of these studies did not address the correlation between plasma PZ levels and anti-PZ antibody titers and that pregnancy complications are observed only in patients with high titers of anti-PZ antibodies (Gris et al. 2003, Sater et al. 2011).
The mechanism by which anti-PZ autoantibodies contribute to adverse pregnancy outcomes remains to be established. Maternal plasma IgM anti-PZ autoantibody concentration >90th percentile has been associated with vascular placental lesions in PE patients, which results in abnormal placentation and pregnancy complications (Erez et al. 2009). Anti-PZ antibodies may also act by rapidly clearing PZ, either by enhancing immune complex formation associated with cellular or complement activation or/and by inducing the formation of inactive antibody-coated PZ molecules (Gris et al. 2003, Dorner et al. 2005). The latter mechanism is more plausible, as it has been shown to precipitate maternal hypercoagulation (Gris et al. 2003). Taken together, anti-PZ antibodies acting independently of classical antiphospholipid antibodies (lupus anticoagulant, anticardiolipin, and anti-β2-glycoprotein I antibodies) may predict the risk of pathologic pregnancies. This has been supported by the existent dose–effect relationship between anti-PZ antibody levels and adverse pregnancy outcomes (Gris et al. 2003, Sater et al. 2011).
PROZ/SERPINA10 (ZPI) polymorphisms in adverse pregnancy outcomes
The wide variability in plasma PZ levels is attributed to the presence of genetic factors influencing PZ biosynthesis (Rice et al. 2001, Vasse et al. 2002), and common and rare gene variants have been reported in the PROZ locus (Rice et al. 2001, van Goor et al. 2008, Le Cam-Duchez et al. 2009, AlShaikh et al. 2013). The differential linkage disequilibrium (LD) pattern between PROZ variants controlling PZ production (Rice et al. 2001, van Goor et al. 2008, Le Cam-Duchez et al. 2009), and thus disease association (Rice et al. 2001, Nowak-Göttl et al. 2009, AlShaikh et al. 2013), indicates that PZ deficiency is, at least in part, constitutional (Miletich & Broze 1987). While the functional aspects of these variants have not been thoroughly assessed, some have been shown to influence plasma PZ levels, in particular G79A, A13G, and G42A; the lowest plasma PZ levels have been observed with the G/G genotype (A13G) and the A/A (G42A) and A/A (G79A) homozygous variant genotypes (Lichy et al. 2004).
The G79A PROZ gene variant (rs3024735) has been the most investigated and linked with several coagulation disorders, including stroke (van Goor et al. 2008), thromboembolism (Nowak-Göttl et al. 2009), coronary artery disease (Le Cam-Duchez et al. 2009), and poor pregnancy outcomes (Lichy et al. 2004, AlShaikh et al. 2013). Few studies have investigated the association of PROZ gene polymorphisms with poor pregnancy outcomes, with inconclusive findings. A small Egyptian study involving 40 women with recurrent miscarriage and 30 control women has reported on the higher prevalence of the G79A minor allele in controls than in cases, suggesting a protective role of the 79A allele in recurrent miscarriage (El-Hamid & El-Khayat 2011). Similarly, an Austrian study involving 49 cases and 48 control women has reported that the 79A allele, individually and in combination with other thrombophilic risk factors (factor V Leiden and increased factor VIII activity), is associated with lower PZ concentrations and a reduced risk of early (8–12 weeks of gestation) spontaneous fetal loss (Dossenbach-Glaninger et al. 2008).
In addition, two Greek studies involving small numbers of women with idiopathic fetal loss and controls have documented that while plasma PZ levels are significantly lower in the 79A allele carriers, the frequency of the 79A allele is similar between the cases and control women (Effraimidou et al. 2009, Topalidou et al. 2009). This suggests that low PZ levels, more than G79A, constitute a risk factor for adverse pregnancy outcomes. In contrast to these studies, we documented that the G79A minor allele was associated with an increased risk of adverse pregnancy outcomes in 287 Bahraini women with idiopathic miscarriage and 308 control women (AlShaikh et al. 2013), and both susceptible and protective PROZ haplotypes were identified (AlShaikh et al. 2013). These inconsistencies may be explained by differences in the genetic background of the studied populations, selection of cases and controls, PZ inter-individual variability, and small size of the cohorts in mainly retrospective studies.
In addition to the G79A variant, other PROZ variants have been shown to be associated with pregnancy-associated adverse effects. For example, the G-42A PROZ promoter variant has been associated with fetal losses (Grandone et al. 2008) and pulmonary embolism, but not with pregnancy-related deep venous thrombosis (Grandone et al. 2009). Furthermore, both rs3024719 (G103A) and rs3024731 (T119A) PROZ promoter variants have been associated with reduced PZ levels and an increased risk of fetal loss (AlShaikh et al. 2013). Based on the LD pattern between PROZ variants (van Goor et al. 2008, Le Cam-Duchez et al. 2009, AlShaikh et al. 2013), specific PROZ haplotypes have recently been shown to be associated with poor pregnancy outcomes (AlShaikh et al. 2013). Larger studies on different ethnic groups are needed to confirm the nature of the association of PROZ G79A and other variants with adverse pregnancy outcomes.
Few studies have investigated the link between SERPINA10 mutations and coagulation defects in humans, often with inconclusive results. ZPI deficiency resulting from mutations in the SERPINA10 gene contributes to thrombotic events (van De Water et al. 2004) and is associated with many coagulation disorders including venous thrombosis (Al-Shanqeeti et al. 2005) and atherosclerotic peripheral arterial disease (Sofi et al. 2010). Several mutations in the coding region of the SERPINA10 gene have been reported by van de Water, of which the nonsense mutations R67X and W303X create stop codons and thus lead to ZPI deficiency due to altered ZPI levels given their location within structurally important sites within the SERPINA10 gene (van De Water et al. 2004). R67X and W303X SERPINA10 variants have been associated with venous thromboembolism in New Zealander (van De Water et al. 2004) and Spanish (Corral et al. 2007) patients, but not in Italian patients (Razzari et al. 2006). Both SERPINA10 mutations were absent in Italian (Fabbro et al. 2007), Caucasian (Folsom et al. 2007), and different groups of Spanish populations (Gonzalez-Conejero et al. 2005), suggesting ethnic restriction in the distribution of these mutations (Gonzalez-Conejero et al. 2005, Fabbro et al. 2007). A lone study has reported a strong association of R67X (OR=2.66), and to a lesser extent W303X (OR=2.44), SERPINA10 variants and identified SERPINA10 haplotypes with early fetal loss, but not with embryonic miscarriages (AlShaikh et al. 2012). This extends the involvement of the genetic variants in the SERPINA10 and PROZ loci in determining the overall risk of adverse pregnancy outcomes.
Conclusion
The role of the PZ–ZPI complex in normal pregnancies and pregnancy complications as a systemic or a local regulator remains unclear. By regulating FXa activity, the presence of the PZ–ZPI complex may provide a local defense mechanism against vascular injury accompanying poor placentation and fetal loss. As such, reduced PZ–ZPI activity stemming from specific anti-PZ autoantibodies and/or polymorphisms within the PROZ and SERPINA10 genes precipitates adverse pregnancy outcomes. PZ deficiency can also be acquired, and the role of contributing factors such as inflammation, obesity, smoking, hypertension, and autoimmunity in affecting plasma PZ and ZPI levels remains to be established. The apparently contradictory results reported for plasma PZ levels and the contribution of the PROZ and SERPINA10 polymorphisms can be explained by the limited number of individuals enrolled and the choice of the control groups. In conclusion, the exact role of the PZ–ZPI complex in the pathogenesis of poor pregnancy outcomes remains to be established, but cannot be dismissed, and future adequately powered studies that address the contribution of inherited and acquired risk factors are necessary.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the review reported.
Funding
This review did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Acknowledgement
The authors thank Ms Parambil Puthen V K Minimol for her assistance in preparing the manuscript.
References
Alfirevic Z, Roberts D & Martlew V 2002 How strong is the association between maternal thrombophilia and adverse pregnancy outcome? A systematic review. European Journal of Obstetrics, Gynecology, and Reproductive Biology 101 6–14. (doi:10.1016/S0301-2115(01)00496-1)
AlShaikh FS, Finan RR, Almawi AW, Mustafa FE & Almawi WY 2012 Association of the R67X and W303X non-sense polymorphisms in the protein Z-dependent protease inhibitor gene with idiopathic recurrent miscarriage. Molecular Human Reproduction 18 156–160. (doi:10.1093/molehr/gar069)
AlShaikh FS, Sater MS, Finan RR, Racoubian E, Abu-Hijleh TM, Mustafa FE & Almawi WY Protein Z variants associated with protein Z plasma levels and with risk of idiopathic recurrent miscarriage Reproductive Sciences 2013 doi:10.1177/1933719112473659)
Al-Shanqeeti A, Vlieg AVH, Berntorp E, Rosendaal FR & Broze GJ 2005 Protein Z and protein Z-dependent protease inhibitor. Thrombosis and Haemostasis 93 411–413. (doi:10.1160/TH04-11-0715)
Bretelle F, Arnoux D, Shojai R, D'Ercole C, Sampol J, Dignat F & Camoin-Jau L 2005 Protein Z in patients with pregnancy complications. American Journal of Obstetrics and Gynecology 19 1698–1702. (doi:10.1016/j.ajog.2005.04.006)
Broze GJ 2001 Protein Z-dependent regulation of coagulation. Determinants of levels and risk of venous thrombosis. Thrombosis and Haemostasis 86 8–13.
Corral J, Gonzalez-Conejero R, Hernandez-Espinosa D & Vicente V 2007 Protein Z/Z dependent protease inhibitor (PZ/ZPI) anticoagulant system and thrombosis. British Journal of Haematology 137 99–108. (doi:10.1111/j.1365-2141.2007.06548.x)
Dayer MR, Ghayour O & Dayer MS 2012 Mechanism of protein-Z-mediated inhibition of coagulation factor Xa by Z-protein-dependent inhibitor: a molecular dynamic approach. ISRN Hematology 2012 762728. (doi:10.5402/2012/762728)
Dorner T, Hoppe B, Salama A, Pruss A & Kiesewetter H 2005 Antibodies against protein Z and fetal loss: current perspective. Clinical and Experimental Medicine 5 50–54. (doi:10.1007/s10238-005-0065-4)
Dossenbach-Glaninger A, van Trotsenburg M, Helmer H, Oberkanins C & Hopmeier P 2008 Association of the protein Z intron F G79A gene polymorphism with recurrent pregnancy loss. Fertility and Sterility 90 1155–1160. (doi:10.1016/j.fertnstert.2007.07.1376)
Effraimidou S, Farmakiotis D & Topalidou M 2009 Protein Z levels and recurrent pregnancy loss. Fertility and Sterility 91 e27. (doi:10.1016/j.fertnstert.2008.12.100)
El-Hamid SA & El-Khayat W 2011 Relationship of the protein Z intron F G79A and IL6 C634G gene polymorphisms with the risk of recurrent pregnancy loss in Egyptian women. Journal of Investigative Medicine 59 655–660. (doi:10.231/JIM.0b013e31820c9c90)
Erez O, Hoppensteadt D, Romero R, Espinoza J, Goncalves L, Nien JK, Kusanovic JP, Fareed J, Gotsch F & Pineles B et al. 2007 Preeclampsia is associated with low concentration of protein Z. Journal of Maternal–Fetal & Neonatal Medicine 20 661–667. (doi:10.1080/14767050701495011)
Erez O, Romero R, Vaisbuch E, Mazaki-Tovi S, Kusanovic JP, Chaiworapongsa T, Than NG, Gotsch F, Kim CJ & Mittal P et al. 2009 Maternal anti-protein Z antibodies in pregnancies complicated by pre-eclampsia, SGA and fetal death. Journal of Maternal–Fetal & Neonatal Medicine 22 662–671. (doi:10.1080/14767050902801751)
Fabbro D, Barillari G & Damante G 2007 Mutations R67X and W303X of the protein Z-dependent protease inhibitor gene and venous thromboembolic disease: a case–control study in Italian subjects. Journal of Thrombosis and Thrombolysis 23 77–78. (doi:10.1007/s11239-006-9003-x)
Folsom AR, Cushman M, Rasmussen-Torvik LJ, Heckbert SR & Tsai MY 2007 Prospective study of polymorphisms of the protein Z-dependent protease inhibitor and risk of venous thromboembolism. Thrombosis and Haemostasis 97 493–494. (doi:10.1111/j.1538-7836.2007.02620.x)
Gonzalez-Conejero R, Perez-Ceballos E, Vicente V & Corral J 2005 Protein Z-dependent protease inhibitor W303X mutation in venous thrombosis. British Journal of Haematology 129 561–562. (doi:10.1111/j.1365-2141.2005.05486.x)
van Goor MP, Dippel DW, Jie KS, de Maat MP, Koudstaal PJ & Leebeek FW 2008 Low protein Z levels but not the protein Z gene G79A polymorphism are a risk factor for ischemic stroke. Thrombosis Research 123 213–218. (doi:10.1016/j.thromres.2008.02.006)
Gowri V, Mathew M, Gravell D, AlFalahi K, Zakwani I, Ganguly SS & Pathare AV 2011 Protein Z levels in pregnant Omani women: correlation with pregnancy outcome. Journal of Thrombosis and Thrombolysis 32 453–458. (doi:10.1007/s11239-011-0629-y)
Grandone E, Colaizzo D, Cappucci F, Cocomazzi N & Margaglione M 2004 Protein Z levels and unexplained fetal losses. Fertility and Sterility 82 982–983. (doi:10.1016/j.fertnstert.2004.03.049)
Grandone E, Colaizzo D, Cappucci F, D'Ambrosio RL, Vecchione G & Margaglione M 2008 An unreported mutation within protein Z gene is associated with very low protein levels in women with fetal loss. Fertility and Sterility 90 864–865. (doi:10.1016/j.fertnstert.2007.07.1318)
Grandone E, Favuzzi G, De Stefano V, Chinni E, Rossi E, Cappucci F & Margaglione M 2009 Protein Z g-42a variant and the risk of pregnancy-related venous thromboembolism in a cohort of Italian patients. Thrombosis Research 123 848–850. (doi:10.1016/j.thromres.2008.12.045)
Gris JC 2009 Thrombophilia and pregnancy loss: cause or association. Thrombosis Research 123 (Suppl 2) S105–S110. (doi:10.1016/S0049-3848(09)70022-2)
Gris JC, Quéré I, Dechaud H, Mercier E, Pinçon C, Hoffet M, Vasse M & Marès P 2002 High frequency of protein Z deficiency in patients with unexplained early fetal loss. Blood 99 2606–2608. (doi:10.1182/blood.V99.7.2606)
Gris JC, Amadio C, Mercier E, Lavigne-Lissalde G, Déchaud H, Hoffet M, Quéré I, Amiral J, Dauzat M & Marès P 2003 Anti-protein Z antibodies in women with pathologic pregnancies. Blood 101 4850–4852. (doi:10.1182/blood-2002-12-3802)
Han X, Fieher R & Broze GJ Jr 2000 Characterization of the protein Z-dependent protease inhibitor. Blood 96 3049–3055.
Heeb MJ, Cabral KM & Ruan L 2005 Down-regulation of factor IXa in the factor Xase complex by protein Z-dependent protease inhibitor. Journal of Biological Chemistry 280 33819–33825. (doi:10.1074/jbc.M506502200)
Huang X, Dementiev A, Olson ST & Gettins PG 2010 Basis for the specificity and activation of the serpin protein Z-dependent proteinase inhibitor (ZPI) as an inhibitor of membrane-associated factor Xa. Journal of Biological Chemistry 285 20399–20409. (doi:10.1074/jbc.M110.112748)
Huang X, Yan Y, Tu Y, Gatti J, Broze GJ Jr, Zhou A & Olson ST 2012 Structural basis for catalytic activation of protein Z-dependent protease inhibitor (ZPI) by protein Z. Blood 120 1726–1733. (doi:10.1182/blood-2012-03-419598)
Kaygusuz I, Firatli-Tuglular T, Toptas T, Ugurel V & Demir M 2011 Low levels of protein Z are associated with HELLP syndrome and its severity. Clinical and Applied Thrombosis/Hemostasis 17 214–219. (doi:10.1177/1076029609357738)
Kemkes-Matthes B & Matthes KJ 1995 Protein Z, a new haemostatic factor, in liver diseases. Haemostasis 25 312–316.
Kist WJ, Janssen NG, Kalk JJ, Hague WM, Dekker GA & de Vries JI 2008 Thrombophilias and adverse pregnancy outcome – a confounded problem. Thrombosis and Haemostasis 99 77–85. (doi:10.1160/TH07-05-0373)
Koren-Michowitz M, Levene NR, Volcheck Y, Jehoshua OG & Kornberg A 2006 Protein Z and its role in venous and arterial thrombosis. Israel Medical Association Journal 8 53–55.
Krawiec P, Gluszko P, Kwasny-Krochin B & Undas A 2011 Decreased protein Z levels in patients with rheumatoid arthritis: links with inflammation. Thrombosis and Haemostasis 106 548–550. (doi:10.1160/TH11-02-0134)
Kusanovic JP, Espinoza J, Romero R, Hoppensteadt D, Nien JK, Kim CJ, Erez O, Soto E, Fareed J & Edwin S et al. 2007 Plasma protein concentration in pregnant women with idiopathic intrauterine bleeding and in women with spontaneous preterm labor. Journal of Maternal–Fetal & Neonatal Medicine 20 453–463. (doi:10.1080/14767050701398272)
Le Cam-Duchez V, Barbay V, Bal Dit Sollier C, Drouet L, Coudert M, Soria C & Borg Y 2009 Haplotypic or genotypic combinations of three protein Z polymorphisms influence protein Z plasma level. Thrombosis and Haemostasis 101 212–214. (doi:10.1160/TH08-07-0478)
Lichy C, Kropp S, Dong-Si T, Genius J, Dolan T, Hampe T, Stoll F, Reuner K, Grond-Ginsbach C & Grau A 2004 A common polymorphism of the protein Z gene is associated with protein Z plasma levels and with risk of cerebral ischemia in the young. Stroke 35 40–45. (doi:10.1161/01.STR.0000106909.75418.E4)
Martinelli I, Razzari C, Biguzzi E, Bucciarelli P & Mannucci PM 2005 Low levels of protein Z and the risk of venous thromboembolism. Journal of Thrombosis and Haemostasis 3 2817–2819. (doi:10.1111/j.1538-7836.2005.01664.x)
Michels TC & Tiu AY 2007 Second trimester pregnancy loss. American Family Physician 76 1341–1346.
Miletich JP & Broze GJ Jr 1987 Human plasma protein Z antigen: range in normal subjects and effect of warfarin therapy. Blood 69 1580–1586.
Napolitano M, Mariani G & Lapecorella M 2010 Hereditary combined deficiency of the vitamin K-dependent clotting factors. Orphanet Journal of Rare Diseases 5 21. (doi:10.1186/1750-1172-5-21)
Nowak-Göttl U, Fröhlich B, Thedieck S, Huge A & Stoll M 2009 Association of the protein Z ATG haplotype with symptomatic nonvascular stroke or thromboembolism in white children: a family-based cohort study. Blood 113 2336–2341. (doi:10.1182/blood-2008-10-181461)
Oldenburg J, von Brederlow B, Fregin A, Rost S, Wolz W, Eberl W, Eber S, Lenz E, Schwaab R & Brackmann HH et al. 2000 Congenital deficiency of vitamin K dependent coagulation factors in two families presents as a genetic defect of the vitamin K–epoxide–reductase-complex. Thrombosis and Haemostasis 84 937–941.
Pabingert I 2008 Thrombophilia and its impact on pregnancy. Hämostaseologie 28 130–134.
Paidas MJ, Ku DH, Lee MJ, Manish S, Thurston A, Lockwood CJ & Arkel YS 2005 Protein Z, protein S levels are lower in patients with thrombophilia and subsequent pregnancy complications. Journal of Thrombosis and Haemostasis 3 497–501. (doi:10.1111/j.1538-7836.2005.01158.x)
Pardos-Gea J, Ordi-Ros J, Serrano S, Balada E, Nicolau I & Vilardell M 2008 Protein Z levels and anti-protein Z antibodies in patients with arterial and venous thrombosis. Thrombosis Research 121 727–734. (doi:10.1016/j.thromres.2007.07.009)
Ramsay JE, Stewart F, Friel H, Walker ID, Greer IA & McColl MDWest of Scotland PZ Study Group 2005 Protein Z in pregnancy: exaggerated rise in obese women. Journal of Thrombosis and Haemostasis 3 2584–2586. (doi:10.1111/j.1538-7836.2005.01607.x)
Razzari C, Martinelli I, Bucciarelli P, Viscardi Y & Biguzzi E 2006 Polymorphisms of the protein Z-dependent protease inhibitor (ZPI) gene and the risk of venous thromboembolism. Thrombosis and Haemostasis 95 909–910. (doi:10.1160/TH06-01-0060)
Rice GI, Futers TS & Grant PJ 2001 Identification of novel polymorphisms within the protein Z gene, haplotype distribution and linkage analysis. Thrombosis and Haemostasis 85 1123–1124.
Sailer T, Vormittag R, Koder S, Quehenberger P, Kaider A & Pabinger I 2008 Clinical significance of anti-protein Z antibodies in patients with lupus anticoagulant. Thrombosis Research 122 153–160. (doi:10.1016/j.thromres.2007.09.014)
Sater MS, Finan RR, Al-Hammad SA, Mohammed FA, Issa AA & Almawi WY 2011 High frequency of anti-protein Z IgM and IgG autoantibodies in women with idiopathic recurrent spontaneous miscarriage. American Journal of Reproductive Immunology 65 526–531. (doi:10.1111/j.1600-0897.2010.00925.x)
Sofi F, Cesari F, Abbate R, Gensini GF, Broze GJ Jr & Fedi S 2010 A meta-analysis of potential risks of low levels of protein Z for diseases related to vascular thrombosis. Thrombosis and Haemostasis 103 749–756. (doi:10.1160/TH09-09-0645)
Souri M, Iwata H, Zhang WG & Ichinose A 2009 Unique secretion mode of human protein Z: its GIa domain is responsible for inefficient, vitamin K-dependent and warfarin-sensitive secretion. Blood 113 3857–3864. (doi:10.1182/blood-2008-07-171884)
Souri M, Sugiura-Ogasawara M, Saito S, Kemkes-Matthes B, Meijers JC & Ichinose A 2012 Increase in the plasma levels of protein Z-dependent protease inhibitor in normal pregnancies but not in non-pregnant patients with unexplained recurrent miscarriage. Thrombosis and Haemostasis 107 507–512. (doi:10.1160/TH11-08-0591)
Soute BAM, Ulrich MM, Watson AD, Maddison JE, Ebberink RH & Vermeer C 1992 Congenital deficiency of all vitamin K-dependent blood coagulation factors due to a defective vitamin K-dependent carboxylase in Devon Rex cats. Thrombosis and Haemostasis 68 521–525.
Tabatabai A, Fiehler R & Broze GJ Jr 2001 Protein Z circulates in plasma in a complex with protein Z-dependent protease inhibitor. Thrombosis and Haemostasis 85 655–660.
Thornton P & Douglas J 2010 Coagulation in pregnancy. Best Practice & Research. Clinical Obstetrics & Gynaecology 24 339–352. (doi:10.1016/j.bpobgyn.2009.11.010)
Topalidou M, Effraimidou S, Farmakiotis D, Papadakis E, Papaioannou G, Korantzis I & Garipidou V 2009 Low protein Z levels, but not intron F G79A polymorphism, are associated with unexplained pregnancy loss. Thrombosis Research 124 24–27. (doi:10.1016/j.thromres.2008.09.017)
Undar L, Karadogan I & Ozturk F 1999 Plasma protein Z levels inversely correlate with plasma interleukin-6 levels in patients with acute leukemia and non-Hodgkin's lymphoma. Thrombosis Research 94 131–134. (doi:10.1016/S0049-3848(98)00210-2)
Vasse M 2008 Protein Z, a protein seeking a pathology. Thrombosis and Haemostasis 100 548–556. (doi:10.1160/TH08-01-0024)
Vasse M 2011 The protein Z/protein Z-dependent protease inhibitor complex. Systemic or local control of coagulation? Hämostaseologie 31 155–158. (doi:10.5482/ha-1161)
Vasse M, Denoyelle C, Legrand E, Vannier JP & Soria C 2002 Weak regulation of protein Z biosynthesis by inflammatory cytokines. Thrombosis and Haemostasis 87 350–351.
Van de Water N, Tan T, Ashton F, O'Grady A, Day T, Browett P, Ockelford P & Harper P 2004 Mutations within the protein Z-dependent protease inhibitor gene are associated with venous thromboembolic disease: a new form of thrombophilia. British Journal of Haematology 127 190–194. (doi:10.1111/j.1365-2141.2004.05189.x)
