Research Article

Protein Z gene variants and risk of Idiopathic Recurrent Pregnancy Loss in Saudi Arabian women

Received: September 10, 2017
Accepted: October 30, 2017
Published: November 29, 2017
Genet.Mol.Res. 16(4): gmr16039834
DOI: 10.4238/gmr16039834


Background: Protein Z (PZ) is a vitamin K-dependent plasma glycoprotein, it plays a key role in the physiologic inhibition of coagulation by acting as a cofactor in the inactivation of factor Xa. The relationship between PZ gene polymorphisms and pregnancy loss is controversial. To address this, we investigate the association between PZ (rs3024718, rs3024719, rs3024731, rs3024778 and rs3024772) single nucleotide polymorphisms (SNPs) and idiopathic recurrent pregnancy loss (IRPL) in in women from Tabuk region (North western region, Saudi Arabia). Methods: A case control study, including Sixty-three women with at least three unexplained recurrent pregnancy loss (RPL) were selected and matched with seventy-eight healthy and fertile women (controls). SNP genotyping was carried out by allelic discrimination using real time PCR. Results: The outcome was that the minor allele frequencies (MAF) are 0.23 vs 0.20, 0.24 vs 0.19, 0.02 vs 0.02, 0.26 vs 0.21, and 0.02 vs 0.01 respectively for rs3024718 A/G, rs024719 G/A, rs3024778 G/A, rs3024731 G/A, and rs3024772 A/G PZ polymorphisms. The genotype distribution was similar between women with RPL and control women (P>0.05) and none of the tested SNPs were associated with RPL under co-dominant, dominant, or recessive genetic models. The lack of association was also confirmed by haplotypes analysis showing an absence of RPL risk with the constructed haplotypes. Conclusion: It was concluded that the studied PZ polymorphisms are not associated with the risk of recurrent pregnancy loss in the studied population


Pregnancy loss (PL) affects up to 15% of reproducing couples (Li et al, 2002), 1% to 3% of those women have two or more consecutive pregnancy losses and are qualified as recurrent pregnancy loss (Allison and Schust, 2009). It may occur due to multiple reasons; include anatomical defects, chromosomal aberrations, endocrine factors, infections, and other immunological factors (El Hachem H et al, 2017). Approximately 50 % of recurrent pregnancy loss remain unexplained (Branch et al, 2010). The defective maternal hemostatic response leading to utero-placental thrombosis along with hypoxia has been hypothesized to subsequently lead to adverse pregnancy outcomes like placental abruption, intrauterine growth restriction or death and preeclampsia. This may include thrombosis in decidual vessels, impairment of trophoblast invasion, villitis and placental micro-thrombi (Rai et al, 2003). An abnormality of trophoblast invasion associated with thrombotic risk factors, such as thrombophilia were described (Khong et al, 1989). Furthermore, it was described that pregnancy and delivery are associated with a disruption of the hemostatic balance favouring a hypercoagulability state with increased level of clotting factor and decreased of fibrinolytic activity (D’Uva et al, 2008; Thornton and Douglas, 2010).

Protein Z is a vitamin K-dependent plasma glycoprotein synthesized by the liver. It plays a key role in the physiologic inhibition of coagulation by acting as a cofactor in the inactivation of factor Xa. Therefore, PZ promotes the assembly of thrombin under unknown circumstances, paradoxically enhance coagulation (Yin et al, 2000, Gorski MM, 2016). Therefore, it is not surprising that the decrease of protein Z level induce pro-coagulant state and is associated with thrombotic disorder. Several findings showing an association between low and high PZ concentrations and an increased risk of thrombotic diseases such as ischemic stroke (Staton et al, 2005, Vasse et al, 2001). However, until now, the association between the low levels of PZ and the risk of idiopathic RPL is the subject of debate and the results are controversial (Al-Shaikh et al, 2013). Based on the previous published data, several common and rare polymorphisms in the PZ gene were reported significantly to be associated with an altered plasma level of PZ (Lichy et al, 2004, Rice et al, 2001).

The PZ gene is located on chromosome 13 in the region 13q34 and it is highly polymorphic (Bafunno et al, 2011). The most reported polymorphisms of the PZ gene are located at the major transcription site -103 A/G (rs3024718), -119 T/A (rs3024731) -475 G/A (rs3024719) (Almawi et al, 2013). A polymorphism in intron C 42G/A, intron F 79G/A (rs3024735) are also reported in the PZ gene (Almawi et al, 2013). A nucleotide substitution of G by C in exon II of the PZ gene, resulting in the replacement of Glu-30 with a Gln residue was identified (Santacroce et al, 2006). The functional aspects of those genetic variation have not thoroughly investigated but some of them are seeming to be linked with different level of PZ secretion (Almawi et al, 2013).

In view of its capacity to reduce plasma level of PZ, a possible relationship between PZ polymorphisms and the occurrence of thrombotic conditions and idiopathic RPL were suggested (Al-Shaikh et al, 2013, Santacroce et al, 2004). Different polymorphisms are shown to be associated with high risk of idiopathic RPL via an enhanced risk of severe placental insufficiency, soon after the connection of maternal and fetal circulations (Paidas et al 2005, Gris et al, 2002).

In the present study, we conduct a case-control study to investigate the contribution of selected variants of PZ promoter (rs3024718, rs3024719, and rs3024731) and exonic (rs3024778 and rs3024772) single nucleotide polymorphisms (SNPs) on increasing idiopathic RPL risk among Saudi women.

Material and Methods


The present study is a case-control retrospective study. Patients (n= 63) with at least three RPL were recruited from January 2015 to January 2017 at Gynecology and obstetrical department of King Khaled hospital in Tabuk, Saudi Arabia.

Patients are women reported with three or more RPL with unknown causes. All cases presented with miscarriage caused by anatomical abnormalities, previously known systemic disease, endocrine disorders including diabetes, previous venous or arterial thrombosis or a family history of thromboembolism are excluded from this study. Chromosomal abnormalities and Rh incompatibility were performed before inclusion in the study. Patients included in this study were confirmed to be negative for, any infections which may explain the pregnancy loss like, Toxoplasma gondii, rubella, cytomegalovirus, herpes simplex viruses, varicella zoster virus and human immunodeficiency virus. Transvaginal ultrasound was performed to confirm spontaneous abortion.


Controls subject include 78 fertile women with at least one successful pregnancy were recruited from January 2015 to January 2017 at Gynecology and obstetrical department of King Khaled hospital in Tabuk, Saudi Arabia. Control women were matched with patients according to several risk factors (smoking, and oral contraceptive use). Controls and patients present similar distribution according to age and body mass index (BMI). An informed consent was achieved from patients and controls subject before enrolling them in the study. The study that was conduct after approval from the research ethical committee at the university of Tabuk. Blood samples were collected from cases and controls in EDTA tubes and were processed for DNA extraction and real time PCR.

Protein Z genotyping

The common PZ rs3024718 (A-475G), rs3024731 (T-119A) and rs3024719 (G-103A), (rs3024778 (G208A) and rs3024772 (G844A) polymorphisms were selected for the study. The PZ genotyping was performed using the allelic discrimination method as described (Al-Shaikh, et al 2013). The analysis was performed using real-time polymerase chain reaction (PCR), according to the procedure and instruction of the manufacturers (Applied Biosystems).

Statistical analysis

Statistical analysis was performed on SPSS software. Categorical variables were expressed as percentages of total or as mean ± SD. To determine differences in means, the Student’s t test was used and Pearson χ2 or Fisher’s exact test was used to assess inter-group significance. Allele frequencies were calculated, Hardy–Weinberg equilibrium was tested for each polymorphism. All analyses were conducted under additive genetic effect, using SNPStats software Linkage disequilibrium analysis was performed using HaploView 4.1 ( HaploView), and haplotypes reconstruction was performed. Logistic regression analysis was performed to determine the odds ratios (OR) and 95 % confidence intervals (95% CI) associated with the RPL risk, taking the control women as the reference group.

The association of genotypes with RPL was conducted under codominant, dominant, and recessive models using SNPstats (, all P-values were two-tailed; P-value< 0.05 were considered statistically significant.


Study subjects. The clinical characteristics of study subjects are shown in table 1. Regarding the age and BMI, patients and controls were matched, and the statistical analysis show a non-significant difference between cases and controls (p=0.93 and p= 0.23). The mean age is 29.96 ± 5.31 and 30.08 ± 7.07 respectively for patients and controls. The analysis of data show also a non-significant difference between patients and controls in term of oral contraceptive consumption (P = 0.94). The mean number of miscarriage in patients is 3.05 ± 1.44, however the mean number of live birth in the control group is 3.12 ± 1.94 and it is significantly higher than observed in patients 0.2 ±1.06 (P < 0.001).

Parameters Cases Controls pa
Age at inclusion in study b 29,96 ± 5.31 30,08 ± 7.07 0.93
Body mass index b, kg/m2 27,16 ± 5.08 25,23 ±8.16 0.23
Obesity c, BMI >30 kg/m2 45.1% 40.5% 0.65
Oral contraceptive c 5.9% 5.6% 0.94
Number of live births b 0.2 ± 1.06 3.12 ± 1.94 <0.001
LA 0 0 NA
Abortions b 3.05 ± 1.44 0.0 ± 0.0 NA

Table 1. Clinical Characteristics of Study Participants

Association studies. The Genotypic distribution of the five polymorphisms studied shows no statistically significant evidence that the alleles frequencies deviated from Hardy-Weinberg equilibrium as shown in table 2. The prevalence of minor allele frequency (MAF) for the five polymorphisms studied were comparable between cases and controls 0.23 versus 0.20 (p=0.61), 0.24 versus 0.19 (P = 0.28), 0.02 versus 0.02 (P=0.84), 0.26 versus 0.21 (P = 0.61) and 0.02 versus 0.01 (P = 0.79) respectively for rs3024718, rs3024719, rs3024778, rs3024731 and rs3024772.

Name Localisation Allele Casesb MAF Controlsb MAF χ2 HWE pa aOR
(95% CI)
rs3024718 113813853 G 0.23 0.2 0.48 0.84 0.61 1.14
(0.57- 2.32)
rs3024719 113814225 A 0.24 0.19 0.09 0.75 0.28 1.53
(0.74 - 3.21)
rs3024778 113814465 A 0.02 0.02 0.06 0.81 0.84 0.82
(0.41 – 4.95)
rs3024731 113818815 A 0.26 0.21 0.15 0.70 0.61 1.23
(0.63 – 3.42)
rs3024772 113826100 A 0.02 0.01 0.05 0.81 0.79 1.6
(0.66 – 3.61)

Table 2. Allelic Distribution of PZ SNPs in IRPL Cases and Control Women

We find no significant association between the five PZ polymorphisms and the risk of RPL in different mode of transmission (recessive and dominant), the results are shown in table 3.

Controls b Patients b Pa OR (95% CI)
Codominant A/A 49 (62.8%) 37 (58.7%) 0.74 1.00
A/G 27 (34.6%) 23 (36.5%) 1.13 (0.56-2.27)
G/G 2 (2.6%) 3 (4.8%) 1.99 (0.32-4.5)
Dominant A/A 49 (62.8%) 37 (58.7%) 0.62 1.00
A/G-G/G 29 (37.2%) 26 (41.3%) 1.19 (0.61-2.34)
Recessive A/A-A/G 76 (97.4%) 60 (95.2%) 0.48 1.00
G/G 2 (2.6%) 3 (4.8%) 1.9 (0.20-5.85)
Codominant G/G 50 (64.1%) 38 (60.3%) 0.64 1.00
G/A 25 (32%) 21 (33.3%) 1.11 (0.59-2.26)
A/A 3 (3.8%) 4 (6.3%) 1.75 (0.38-8.31)
Dominant G/G 50 (64.1%) 38 (60.3%) 0.46 1.00
G/A-A/A 28 (35.9%) 25 (39.7%) 1.56 (0.66-3.69)
Recessive G/G-G/A 75 (96.2%) 59 (93.7%) 0.87 1.00
A/A 3 (3.8%) 4 (6.3%) 1.69 (0.37-4.87)
Codominant T/T 50 (64.1%) 35 (55.6%) 0.59 1.00
T/A 23 (29.5%) 23(36.5%) 1.43 (0.69-2.94)
A/A 5 (6.4%) 5 (7.9%) 1.81 (0.37-4.74)
Dominant T/T 50 (64.1%) 35 (55.6%) 0. 3 1.00
T/A-A/A 28 (35.9%) 28 (44.4%) 1.43 (0.72-2.82)
Recessive T/T-T/A 73 (93.6%) 58 (92.1%) 0.73 1.00
A/A 5 (6.4%) 5(7.9%) 1.26 (0.37-4.48)
  G/G 77 (98.7%) 61 (96.8%)   1.00
G/A 1 (1.3%) 2 (3.2%) 0.44 -
  G/G 75 (96.2%) 61 (96.8%)   1.00
G/A 3 (3.8%) 2 (3.2%) 0.84 -

Table 3. Distribution of PZ Genotypes and risk of IRPL

Also, these differences remain non-significant after adjusting for age, BMI, and oral contraceptive. The individual SNPs prevalence of the five locus (rs3024718/ rs3024719/rs3024778/rs3024731/rs3024772) of PZ haplotypes were constructed and analyzed in table 4. Eight haplotypes were commonly detected in both cases and controls. However, GGTGA haplotype shows a high prevalence in cases with IRPL (0.02) compared to those in controls (0.005) however there was no significant difference between the two groups (P=0.38). The haplotypic analysis reveals a lack of association between the different haplotype and the risk of recurrent miscarriage, and was confirmed by the absence of any block of linkage disequilibrium LD.

Haplotype a Case c Controls c P b AOR (95% CI)
AGTGG 0.41 0.53 - 1 (REFERENCE)
AGAGG 0.156 0.22 0.59 1.78 (0.72 – 4.4)
AATGG 0.1 0.11 0.7 1.27 (0.31 – 7.39)
GGTGG 0.11 0.085 0.38 1.63 (0.55 – 4.85)
AAAGG 0.06 0.05 0.61 1.48 (0.34– 6.51)
GATGG 0.036 0.049 0.34 2.27 (0.43 – 6.65)
GGAGG 0.041 0.045 0.46 0.91 (0.58 – 5.23)
GGTGA 0.02 0.005 0.38 3.81 (0.78 – 4.71)

Table 4. Distribution of 6-Locus PZ Haplotypes in IRPL Cases and Control Women


RPL is a multifactorial disorder, however its pathogenesis is still controversial. The present study investigates the genetic variation of protein Z gene on RPL risk in a group of women from Saudi Arabian population. Our finding shows no significant difference between the two groups in term of allelic frequencies and genotyping distribution for all analyzed polymorphisms. The MAF of the analyzed PZ SNPs except rs3024731 were comparable to those described in HapMap CEU sample. Several previously published studies have shown that the MAF of rs3024731 appear to be lower than what was found in the Caucasian population (Nawak-Gottl et al, 2009; Santacroce et al, 2004). In this study, we found that neither of the studied SNPs are associated with RPL in Saudi Arabian women which agrees with results reported by Al-Shaikh (Al-Shaikh et al, 2013). In contrast, the rs3024719 (G-103A) and rs3024731 (T-119A) polymorphisms are significantly higher in patients than in the control group and constitute an independent risk factor for recurrent miscarriage in Bahrain population (Al-Shaikh et al, 2013). Moreover, a Grecian cohort enrolled 51 cases with recurrent miscarriage, demonstrate no association between risk of pregnancy loss and PZ polymorphisms (Tapalidou et al, 2009). Our study is in agreement with European studies have demonstrate that PZ polymorphisms was not associated with PZ deficiency and risk of preeclampsia (LeCam-Duchez et al, 2005). In addition, several studies conclude that the risk of RPL was associated with lower level of protein Z but not with the rs3024735 (G79A) variant of PZ gene which was described protective against RPL (El-Hamid et al, 2011, Dossenbach-Glaninger et al, 2008). Differential Linkage disequilibrium LD between PZ function, PZ production and diseases were investigated. It was published that ATG haplotypes is associated with increase rate of PZ levels (Nawak-Gottl et al, 2009).

The effect of PZ deficiency in the process of pregnancy was described in the early stage when the placenta starts to replace the yolk sac to supply embryo by blood (Gris et al, 2003). The wild variability of PZ level was investigated in different studies involving Arab women’s and the results are controversial. An Omani study demonstrate that a PZ deficiency was observed in the first and second trimester of pregnancy, but all women had a normal pregnant outcome (Gowri et al, 2011). However, Bahranian women show that the low level of since the first trimester of pregnancy loss need to be explain by other variant of PZ and protein Z inhibitor (Al-shaikh et al, 2013). According the pathogenic mechanism of PZ deficiency, described by Padias and colleagues as thrombotic disorder, is different from the classical thrombophilia pathway (Paidas et al, 2005). Insofar the PZ act in thrombotic event by limiting the effect of increased generated factor Xa, the PZ form with protein Z protease inhibitor ZPI a complex PZ_ZPI. In addition, pregnancy loss was associated with high levels of antibodies anti-PZ (Dorner et al, 2005, Gris et al, 2003) without affecting the plasmatic levels of PZ. However, it is noted that low concentration of PZ is associated with patients with antiphospholipid antibodies (Forastiero et al, 2003). In view of the wide range of results observed in different populations the association between the common genetic variations within the PZ locus and the risk of recurrent pregnancy losses is controversial. Some results suggest that the effect of PZ is not related to the genetic variation but to PZ deficiency (Tapalidou et al, 2009) or/and to presence of antiphospholipid antibodies (Broze GJ Jr, 2001). It was also suggested that the PZ deficiency is often associated with lupus anticoagulant (Steffano et al, 2001, (Forastiero et al, 2003).


The lack of significant association between PZ polymorphisms and risk of recurrent pregnancy loss shown here suggest that PZ don’t present an individual risk factor for RPL but may be combined with other factors and require introducing more clinical parameters. The present work faced some limitation especially the small size of patients and controls included but it has the strength for the homogeneity in their ethnic background only saudi Arabia women were included and the combination of the five PZ polymorphisms located at promoter intron and exons region. Furthermore, it will be interesting to complete this finding by an assessment of PZ levels in womens with idiopathic recurrent pregnancy loss during the three trimesters of pregnancy. It is possible that among the large number of risk factors affecting the maternal hemostasis equilibrium, knowledge of the genetic variation of PZ combined with others risk factors may help to identify other pathways that would help in better understanding of the disease.


The current study was approved by the Research Ethics Committee at the University of Tabuk.


The author(s) declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.


We acknowledge the financial support by the Deanship of Scientific Research at the University of Tabuk, kingdom of Saudi Arabia, for the present study (grant S-1436-0114).

About the Authors

Corresponding Author

Walid Zammiti

Prince Fahd Ben Sultan Research Chair (PRFBC), Department of Medical Laboratory Technology, Faculty of Applied Medical Sciences, University of Tabuk, Kingdom of Saudi Arabia

[email protected]


Full PDF