Contribution of JAK2 and STAT3 Variants to the Genetic Susceptibility of Recurrent Spontaneous Miscarriage in a Tunisian Population

Abstract

Aims: Th1 and Th2 balance is crucial for maintenance of pregnancy, and intracellular JAK and STAT proteins significantly contribute to it. In view of evidence linking JAK2 and STAT3 variants with recurrent spontaneous miscarriage (RSM), here we investigated the association of JAK2 (rs2230724) and STAT3 (rs1053023 and rs1053004) to RSM susceptibility in Tunisians. Subjects and Methods: A retrospective case-control study. Subjects comprised 235 RSM cases and 235 control subjects. JAK2 and STAT3 were genotyped by the allelic discrimination method. Results: STAT3 rs1053023 and, to a lower extent, rs1053004 were significantly associated with RSM under the additive and dominant, but not recessive models. This remained significant after adjustment for the covariates age, smoking, and gravida. In contrast to STAT3 variants, JAK2 rs2230724 was not associated with RSM under any of the genetic models tested. Two-locus STAT3 (rs1053023/rs1053004) haplotype analysis revealed increased frequency of the C/G haplotype in patients with RSM. Multivariate regression analysis confirmed the association of C/G haplotype with RSM (p=0.001; odds ratio=2.01; 95% confidence interval=1.32-3.07), thus conferring RSM susceptibility nature. These differences remained significant after applying the Bonferroni correction for multiple testing (Pc=0.004). Conclusions: STAT3 rs1053023, more so than the STAT3 rs1053004 or JAK2 rs2230724 polymorphisms, is associated with RSM risk.

Introduction

Insofar as adequate cytokine balance is required for maintaining successful pregnancy, dysregulation of cytokine profile was reported to induce a number of pregnancy complications, including pre-eclampsia, intrauterine fetal death, and recurrent spontaneous miscarriage (RSM) (Wegmann et al., 1993; Raghupathy, 1997; Aluvihare et al., 2004, Koga and Mor, 2010; Manaster and Mandelboim, 2010). While several modifiable and nonmodifiable fetal and maternal RSM risk factors were identified, more than half of RSM cases remain idiopathic (Kwak-Kim et al., 2009; Choudhury and Knapp, 2001). An immunological basis underlying some RSM cases was suggested (Hill, 1995; Al-Khateeb et al., 2011), and was attributed to altered Th1/Th2 balance. This Th1-Th2 paradigm has been recently expanded to include Th17 and regulatory T (Treg) helper T-cell subsets, each with its unique cytokine secretion profile (Saito et al., 2010).

Antigen-specific T cell activation is a complex process, which converges at engagement of JAK kinases and the STAT family of intracellular proteins (Hauguel-de Mouzon and Guerre-Millo, 2006), of which the JAK2-STAT3 signaling pathway was reported to play an important role in trophoblast invasion and differentiation (Mulla et al., 2010), and is required for establishment and maintenance of pregnancy (Bao et al., 2006; Fitzgerald et al., 2008), as shown by the finding that dysregulated STAT3 expression and activation were associated with defective placentation (Tarrade et al., 2002; Catalano et al., 2005; Mulla et al., 2010). Since JAK2 and STAT3 expression are genetically determined, previous studies on the association of the JAK2 and STAT3 polymorphisms with RSM was explored, but with inconclusive findings (Mercier et al., 2007; Finan et al., 2010; Kaandorp et al., 2010). We evaluated the association of the JAK2 (rs2230724) and STAT3 (rs1053023 and rs1053004) gene variants with RSM in Tunisian subjects, and found that of these, STAT3 (rs1053023) is the only variant identified in association with RSM in our population. These three variants are single-nucleotide polymorphisms (SNPs). JAK2 rs2230724 is an exonic polymorphism, localized in exon 19. This variant was associated with susceptibility to acute leukemia and its subtypes, thus suggested to be a genetic determinant for acute leukemia and acute myeloid leukemia (Zhong et al., 2012). Both STAT3 variants rs1053023 and rs1053004 are localized in exon 24. STAT3 rs1053023 is associated with 28% risk reduction for B-cell non-Hodgkin's lymphoma (Butterbach et al., 2011), while STAT3 rs1053004 is significantly associated with idiopathic RSM (Finan et al., 2010).

Subjects and Methods

RSM cases comprised 235 unrelated women (mean age 32.1±6.1), who were recruited from the outpatient OB/GYN clinics in for assessment of idiopathic RSM, as per the Royal College of Obstetricians and Gynecologists (RCOG) Guidelines (Table 1). The inclusion criteria included ≥3 consecutive pregnancy losses with no known risk factors, and with the same partner, which occurred during the first trimester of gestation. Exclusion criteria were Rh incompatibility, older age at first miscarriage (≥40 years), preclinical miscarriages and/or biochemical pregnancy, and pre-eclampsia (elevated systolic and diastolic blood pressure (BP) >145/95 mm Hg, or rise in systolic/diastolic BP>30/15 mmHg on at least two occasions). Additional exclusion criteria included systemic autoimmune disease, thyroid dysfunction, diabetes mellitus, anatomical disorders, infections (toxoplasmosis, human cytomegalovirus, rubella, HIV, Group B streptococci, Chlamydia trachomatis, hepatitis B and C, and bacterial vaginosis), and liver function abnormalities.

Table 1.

Characteristics of Study Participants

	Cases	Controls	p^a
Age at inclusion in study	32.1±6.1	32.8±5.8	0.012
Body-mass index (kg/m²)	25.2±3.9	25.6±4.1	0.322
Obesity (>30 kg/m²)	37 (13.1)^b	58 (15.3)	0.325
Smokers	9 (3.2)	19 (5.0)	0.002
Gravida	4.2±2.1	3.2±1.3	<0.001
Number of live births	0.8±0.9	3.2±1.0	<0.001
Abortions	3.5±1.2	0.0±0.0	NA

Tunisian subjects consisted of 235 cases and 235 control women.

Student's t-test for continuous variables; Pearson's chi-square tests for categorical variables.

Number of subjects (percent total).

NA, not applicable.

Controls comprised 235 multiparous ethnically matched women who had at least two live births, and no miscarriages, and consisted of university and hospital employees, or volunteers. All participants were Tunisian Arabs, and non-Arab Tunisians (Berber) were excluded. On all subjects, demographic details were recorded, and were verified from clinic records where available. The Research and Ethics Committee of the University of Monastir (Tunisia) approved the study protocol. Blood samples were taken from all participants in ethylenediaminetetraacetic acid-containing tubes by venipuncture for genomic DNA extraction, after signing a consent form agreeing to be included in the study.

Genotyping

JAK2 (rs2230724) and STAT3 (rs1053023 and rs1053004) polymorphisms were genotyped using the allelic (VIC- and FAM-labeled) discrimination method. TaqMan assays were ordered from Applied Biosystems (Table 2). The reaction was performed in a 6-μL volume on StepOne real-time PCR system, as recommended by the manufacturer (Applied Biosystems). Replicate-blinded quality control samples were included to assess reproducibility of the genotyping procedure; concordance was >99%.

Table 2.

Single-Nucleotide Polymorphisms Analyzed

Gene	SNP	Chromosome	Assay ID	HWE p^a
JAK2	rs2230724	9	C_22273141_10	1.00
STAT3	rs1053023	17	C___7530673_10	0.066
STAT3	rs1053004	17	C___1795285_1_	0.023

Location on chromosome based on dbSNP-build 125.

p determined using Haploview 4.2.

SNP, single-nucleotide polymorphism; HWE, Hardy-Weinberg equilibrium.

Statistical analysis

Data were expressed as mean±SD (continuous variables), or as percentages of total (categorical variables), and intergroup significance was assessed by Student's t-test (continuous variables), and the χ² test (categorical variables). Allele frequencies were calculated by the gene-counting method; each SNP was tested for the Hardy-Weinberg equilibrium (HWE) using χ² goodness-of-fit test by Haploview 4.2 (www.broad.mit.edu/mpg/haploview/). Genotypic association of each SNP with RSM susceptibility was conducted under additive, dominant, and recessive models using SNPstats (http://bioinfo.iconcologia.net/SNPstats), taking homozygous wild-type genotype as the reference (odds ratio [OR]=1.00), after controlling for age, smoking, and body-mass index (BMI).

Results

Study subjects

The demographics and clinical characteristics of study subjects are shown in Table 1. BMI (25.2±3.9 kg/m² vs. 25.6±4.1 kg/m²; p=0.322) and the prevalence of obesity (>30 kg/m²) (p=0.325) were comparable between RSM cases and control women. In contrast, the prevalence of smokers (p=0.002), age (p=0.012), and gravida (p<0.001) was significantly different between RSM cases and control women. Accordingly, the latter were the covariates that were controlled for in subsequent analysis. In total, RSM cases reported a mean of 3.5 pregnancy losses, while control women reported an average of 3.2 live births/woman and no pregnancy losses.

JAK2 and STAT3 genotype distribution

The distribution of JAK2 rs2230724 (p=1.00), STAT3 rs1053023 (p=0.066), more so than rs1053004 (p=0.023) genotypes was in the HWE among study subjects (Table 2). Table 3 summarizes the minor allele frequencies (MAF) of the JAK2 and STAT3 variants in RSM cases and control subjects. MAF of both JAK2 rs2230724 (p=0.44) and STAT3 rs1053004 (p=0.10) were not significantly different between RSM cases and control women. In contrast, significantly higher MAF of STAT3 rs1053023 were seen in cases (p=0.006), and translated into increased RSM risk [OR (95%)=1.50 (1.13-1.97)].

Table 3.

Association of Single-Nucleotide Polymorphisms with Recurrent Spontaneous Miscarriage in Tunisian Arabs

	Cases MAF	Controls MAF	p^a	aOR (95% CI) ^b
JAK2 (rs2230724; A>G)	0.40	0.42	0.44	0.90 (0.70-1.16)
STAT3 (rs1053023; A>G)	0.36	0.28	0.006	1.50 (1.13-1.97)
STAT3 (rs1053004; C>T)	0.49	0.44	0.10	1.25 (0.97-1.62)

Study subjects comprised 235 RSM cases and 235 control women.

Crude p value.

aOR, adjusted OR, adjusted for gender and BMI.

Bold numbers indicate statistically significant values.

MAF, minor allele frequency; OR, odds ratio; CI, confidence interval; RSM, recurrent spontaneous miscarriage.

Table 4 summarizes the association analyses between JAK2 rs2230724 and STAT3 rs1053023 and rs1053004 variants, under additive, dominant, and recessive genetic models. STAT3 rs1053023 and, to a lower extent, rs1053004 showed a significant association with RSM under the additive and dominant, but not recessive models, which remained significant after adjustment for the covariates age, smoking, and gravida. In contrast to the STAT3 variants, JAK2 rs2230724 did not associate significantly with RSM under any of the genetic models tested.

Table 4.

Effects of JAK2-STAT3 Genotypes on Recurrent Spontaneous Miscarriage Risk

		Genotype distribution		Additive model		Dominant model		Recessive model
		Controls	Patients	p^a	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)
JAK2 rs2230724	A/A	77 (33.0)^b	90 (38.3)	0.46	1.00	0.24	1.00	0.97	1.00
	G/A	114 (48.9)	103 (43.8)		0.77 (0.52-1.16)		0.80 (0.54-1.16)		0.99 (0.62-1.59)
	G/G	42 (18.0)	42 (17.9)		0.86 (0.51-1.45)	G/G vs. G/A+A/A		G/G+G/A vs. A/A
STAT3 rs1053023	A/A	125 (55.1)	101(43.0)	0.025	1.00	0.01	1.00	0.10	1.00
	A/G	79 (34.8)	98 (41.7)		1.54 (1.03-2.28)		1.63 (1.13-2.35)		1.60 (0.92-2.80)
	G/G	23 (10.1)	36 (15.3)		1.94 (1.08-3.84)	A/A vs. A/G+G/G		A/A+A/G vs. G/G
STAT3 rs1053004	T/T	83 (35.6)	61 (26.0)	0.07	1.00	0.023	1.00	0.70	1.00
	T/C	97 (41.6)	117 (49.8)		1.64 (1.07-2.51)		1.58 (1.06-2.35)		1.09 (0.71-1.67)
	C/C	53 (22.8)	57 (24.3)		1.46 (0.89-2.41)	T/T vs. T/C+C/C		T/T+T/C vs. C/C

p values adjusted for age, gender, and body-mass index.

Number of subjects (frequency).

Bold numbers indicate statistically significant values.

STAT3 haplotype analysis

Two-locus haplotypes were constructed based on the prevalence of individual SNPs, and linkage disequilibrium between them. Taking the common rs1053023/rs1053004 T/A haplotype as reference (OR=1.00), multivariate regression analysis confirmed the association of the C/G haplotype with RSM (p=0.001; OR=2.01; 95% confidence interval [CI]=1.32-3.07), thus conferring RSM susceptibility nature (Table 5). These differences remained significant for C/G (Pc=0.004) after applying the Bonferroni correction for multiple testing (Table 5).

Table 5.

Distribution of the STAT3 Haplotype

Haplotype ^a	Cases	Patients	p^b	Pc ^b	OR (95% CI)
T/A^c	0.375	0.339	-	-	1.00 (Reference)
C/A^c	0.349	0.302	0.730	0.995	1.06 (0.76-1.48)
T/G^c	0.189	0.169	0.460	0.915	1.16 (0.79-1.70)
C/G ^c	0.087	0.189	0.001	0.004	2.01 (1.32-3.07)

STAT3 haplotype (rs1053004 (T/C)/rs1053023 (A/G)) frequency determined by the maximum likelihood method.

Pearson's chi-square test.

Haplotype frequency.

Bold numbers indicate significant associations.

Discussion

Several lines of reasoning implicated altered JAK2/STAT3 activity with adverse pregnancy outcomes. The beneficial effect of Interleukin-11 for normal placentation in pregnancy involves upregulation of the α2-macroglobulin gene through the JAK2/STAT3 pathway (Bao et al., 2006), and JAK2 is a key factor for pregnancy-associated mammary epithelium proliferation and differentiation (Shillingford et al., 2002). In addition, STAT3 knockouts display embryonic lethality, at the gastrulation phase, further establishing a central role for STAT3 in successful pregnancy (Takeda et al., 1997), as evidenced by the findings that decidual cytokines utilize STAT3 as a common intracellular signaling protein, thereby allowing for effective invasion, proliferation, and differentiation of trophoblast cells, hence implantation (Catalano et al., 2005; Fitzgerald et al., 2008; Pu et al., 2012). Accordingly, dysregulation of STAT3 expression and activation during placental development are likely to provoke defective placentation, along with shallow invasion in pre-eclampsia or neoplasms such as human choriocarcinoma (Tarrade et al., 2002).

We analyzed the possible association between the JAK2 and STAT3 variants and RSM risk among Tunisian women. Among the acquired (modifiable) risk factors, RSM risk was linked with the prevalence of smoking, which was in apparent disagreement with a recent study on Chinese women reporting no such association (Zhang et al., 2010). This discrepancy may be related to the coffee consumption in our population compared to the Chinese cohort, which was reported to contribute to increased RSM risk in women smokers (Derbyshire and Abdula, 2008). Significant association was noted between RSM and STAT3 rs1053023, but not with JAK2 rs2230724 or STAT3 rs1053004, in apparent disagreement with the recent study of Finan, in which STAT3 rs1053004 more so than STAT3 rs1053023 was associated with RSM risk in Bahraini Arab women (Finan et al., 2010). This apparent contradiction is likely reconciled by differences in ethnic background of these distinct communities, as was earlier reported (Almawi et al., 2005).

We analyzed the contribution of the JAK2 and STAT3 variants to RSM risk at the allele and genotype levels, the latter, including the additive, dominant, and recessive models. JAK2 rs2230724, STAT3 rs1053023 A>G, and rs1053004 C>T minor allele frequencies among control subjects were similar to frequencies established for Europeans/Caucasians and Asians, were higher that the rates seen for Asians (Japanese and Chinese), but were lower than frequencies established for African-Americans, or Sub-Saharan Africans. These findings confirm racial heterogeneity in the distribution of the JAK2 and STAT3 variants, and hence their disease association.

Since both STAT3 variants (rs1053004 and rs1053023) were in LD, 2-locus haplotypes were constructed based on the prevalence of the individual SNP and the LD between them. Of the three variants tested, slight deviation from the HWE was noted for rs1053004, which is not likely to be due to genotyping errors, but rather to population stratification, as was suggested elsewhere (Hosking et al., 2004; Vine and Curtis, 2009).

We found no significant association between JAK2 rs2230724, previously linked and altered cell proliferation and differentiation (Zhong et al., 2011), and RSM risk in both populations. Mercier reported that the JAK2 V617F mutation, which occurs in 0.20% of normal Southern European primigravidas 21-36 years of age, is linked with increased risk of pregnancy loss (Mercier et al., 2007), while Kaandorp reported that the V617F variant is not a clinically relevant etiologic factor in unexplained RSM (Kaandorp et al., 2010). Melillo et al. (2009) show that the fetal loss rate in our study was still high (24.6%). In particular, the rate of fetal loss due to spontaneous abortion in the first trimester was 15.6% in this study, while it ranged between 28.1% and 31.7% in other reports (Passamonti et al., 2007; Gangat et al., 2009). The risk of spontaneous abortion was 2.5 times higher in patients with ET than in the age-matched general population in the study of Melillo et al. (2009). Coupled to other studies demonstrating lack of association of the related variant V617F (rs77375493) with pregnancy complications (Dahabreh et al., 2009; Grandone et al., 2011), this questions the contribution of altered JAK2 activity in pregnancy complications.

PSTAT3 has been previously shown to be involved in fertility, implantation, and decidualization, hence the maintenance of pregnancy (Garcia et al., 2007). Altered STAT3 activity was linked with poor pregnancy outcome, evidenced by the reduced endometrial pSTAT3 activity in women with unexplained infertility (Dimitriadis et al., 2007). We focused on the 3′UTR region containing regulatory elements for STAT3 expression, and reasoned that altered STAT3 activity induced by specific STAT3 variants influences RSM susceptibility, thus contributing to pregnancy loss by setting in local inflammation at placental membranes (Dimitriadis et al., 2006), leading to accelerated miscarriage. At the haplotype level, the double-mutant C/G haplotype was associated with increased RSM risk (OR=2.01, 95% CI (1.32-3.07), after controlling for several covariates.

In conclusion, our study demonstrates that STAT3 rs1053023, but not STAT3 rs1053004 or JAK2 polymorphisms, is associated with an RSM risk. Our study has several strengths, namely the homogeneity of the population tested (only Tunisian Arabs from central Tunisia), and limiting the RSM cases to early miscarriages, thus minimizing the possibility of patient misclassification. Our study has some shortcomings, namely that we did not correlate the STAT3 genotypes with STAT3 activity, and to the selection of the SNPs, thus raising the possibility of the contribution of other variants in neighboring and distant genes in modifying RSM risk. A larger study is needed to confirm the current findings to reveal the mechanism by which altered STAT3 activity induced by specific genotypes influences overall RSM risk.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Al-Khateeb

, Sater

, Finan

et al. 2011. Analysis of interleukin-18 promoter polymorphisms and changes in interleukin-18 serum levels underscores the involvement of interleukin-18 in recurrent spontaneous miscarriage. Fertil Steril, 96:921-926.

Almawi

, Keleshian

, Borgi

et al. 2005. Varied prevalence of factor V G1691A (Leiden) and prothrombin G20210A single nucleotide polymorphisms among Arabs. J Thrombosis Thrombolysis, 20:163-168.

Aluvihare

, Kallikourdis

, Betz

. 2004. Regulatory T cells mediate maternal tolerance to the fetus. Nat Immunol, 5:266-271.

Bao

, Devi

, Bowen-Shauver

et al. 2006. The role of interleukin-11 in pregnancy involves up-regulation of alpha2-macroglobulin gene through Janus kinase 2-signal transducer and activator of transcription 3 pathway in the decidua. Mol Endocrinol, 20:3240-3250.

Butterbach

, Beckmann

, de Sanjosé

et al. 2011. Association of JAK-STAT pathway related genes with lymphoma risk: results of a European case-control study (EpiLymph) Br J Haematol, 153:318-333.

Catalano

, Johnson

, Campbell

et al. 2005. Inhibition of Stat3 activation in the endometrium prevents implantation: a nonsteroidal approach to contraception. Proc Natl Acad Sci USA, 102:8585-8590.

Choudhury

, Knapp

. 2001. Human reproductive failure. I. Immunological factors. Hum Reprod Update, 7:113-134.

Dahabreh

, Jones

, Voulgarelis

et al. 2009. No evidence for increased prevalence of JAK2 V617F in women with a history of recurrent miscarriage. Br J Haematol, 144:802-803.

Derbyshire

, Abdula

. 2008. Habitual caffeine intake in women of childbearing age. J Hum Nutr Diet, 21:159-164.

10.

Dimitriadis

, Sharkey

, Tan

et al. 2007. Immunolocalisation of phosphorylated STAT3, interleukin 11 and leukaemia inhibitory factor in endometrium of women with unexplained infertility during the implantation window. Reprod Biol Endocrinol, 5:44.

11.

Dimitriadis

, Stoikos

, Tan

et al. 2006. Interleukin 11 signalling components signal transducer and activator of transcription 3 (STAT3) and suppressor of cytokine signaling 3 (SOCS3) regulate human endometrial stromal cell differentiation. Endocrinology, 147:3809-3817.

12.

Finan

, Mustafa

, Al-Zaman

et al. 2010. STAT3 polymorphisms linked with idiopathic recurrent miscarriages. Am J Reprod Immunol, 63:22-27.

13.

Fitzgerald

, Poehlmann

, Schleussner

et al. 2008. Trophoblast invasion: the role of intracellular cytokine signaling via signal transducer and activator of transcription 3 (STAT3) Hum Reprod Update, 14:335-344.

14.

Gangat

, Wolanskyj

, Schwager

et al. 2009. Predictors of pregnancy outcome in essential thrombocythemia: a single institution study of 63 pregnancies. European Journal of Haematology, 82:350-353.

15.

Garcia

, Tirado-Gonzalez

, Handjiski

et al. 2007. High expression of survivin and down-regulation of Stat-3 characterize the feto-maternal interface in failing murine pregnancies during the implantation period. Placenta, 28:650-657.

16.

Grandone

, Colaizzo

, Tiscia

et al. 2011. PRENACEL Study Group. Impact of common thrombophilias and JAK2 V617F on pregnancy outcomes in unselected Italian women. J Thromb Haemost, 9:496-501.

17.

Hauguel-de Mouzon

, Guerre-Millo

. 2006. The placenta cytokine network and inflammatory signals. Placenta, 27:794-798.

18.

Hill

. 1995. T-helper 1-type immunity to trophoblast: evidence for a new immunological mechanism for recurrent abortion in women. Hum Reprod, 10,suppl2:114-120.

19.

Hosking

, Lumsden

, Lewis

et al. 2004. Detection of genotyping errors by Hardy-Weinberg equilibrium testing. Eur J Hum Genet, 12:395-399.

20.

Kaandorp

, Lauw

, van der Schoot

et al. 2010. Prevalence of JAK2V617F mutation in women with unexplained recurrent miscarriage. J Thromb Haemost, 8:2837-2839.

21.

Koga

, Mor

. 2010. Toll-like receptors at the maternal-fetal interface in normal pregnancy and pregnancy disorders. Am J Reprod Immunol, 63:587-600.

22.

Kwak-Kim

, Yang

, Gilman-Sachs

. 2009. Recurrent pregnancy loss: a disease of inflammation and coagulation. J Obstet Gynaecol Res, 35:609-622.

23.

Manaster

, Mandelboim

. 2010. The unique properties of uterine NK cells. Am J Reprod Immunol, 63:434-444.

24.

Melillo

, Tieghi

, Candoni

et al. 2009. Outcome of 122 pregnancies in essential thrombocythemia patients: A report from the Italian registry. Am J Hematol, 84:636-640.

25.

Mercier

, Lissalde-Lavigne

, Gris

. 2007. JAK2 V617F mutation in unexplained loss of first pregnancy [Letter] N Engl J Med, 357:1984-1985.

26.

Mulla

, Myrtolli

, Brosens

et al. 2010. Antiphospholipid antibodies limit trophoblast migration by reducing IL-6 production and STAT3 activity. Am J Reprod Immunol, 63:339-348.

27.

Passamonti

, Randi

, Rumi

et al. 2007. Increased risk of pregnancy complications in patients with essential thrombocythemia carrying the JAK2 (617V>F) mutation. Blood, 110:485-489.

28.

, Duan

, Wang

et al. 2012. Thymic stromal lymphopoietin promotes the proliferation of human trophoblasts via phosphorylated STAT3-mediated c-Myc upregulation. Placenta, 33:387-391.

29.

Raghupathy

. 1997. Th1-type immunity is incompatible with successful pregnancy. Immunol Today, 18:478-482.

30.

Saito

, Nakashima

, Shima

et al. 2010. Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy. Am J Reprod Immunol, 63:601-610.

31.

Shillingford

, Miyoshi

, Robinson

et al. 2002. Jak2 is an essential tyrosine kinase involved in pregnancy-mediated development of mammary secretory epithelium. Mol Endocrinol, 16:563-570.

32.

Takeda

, Kurachi

, Yamamoto

et al. 1997. Alternative signaling mechanism of leukemia inhibitory factor responsiveness in a differentiating embryonic carcinoma cell. Endocrinology, 138:2689-2696.

33.

Tarrade

, Goffin

, Munaut

et al. 2002. Effect of matrigel on human extravillous trophoblasts differentiation: modulation of protease pattern gene expression. Biol Reprod, 67:1628-1637.

34.

Vine

, Curtis

. 2009. Markers typed in genome-wide analysis identifies regions showing deviation from Hardy-Weinberg equilibrium. BMC Res Notes, 2:29.

35.

Wegmann

, Lin

, Guilbert

et al. 1993. Bidirectional cytokine interactions in the maternal-fetal relationship: is successful pregnancy a TH2 phenomenon. Immunol Today, 14:353-356.

36.

Zhang

, Wei

, Niu

et al. 2010. Risk factors for unexplained recurrent spontaneous abortion in a population from southern China. Int J Gynaecol Obstet, 108:135-138.

37.

Zhong

, Wu

, Ma

et al. 2011. Association of Janus kinase 2 (JAK2) polymorphisms with acute leukemia susceptibility. Int J Lab Hematol.[Epub ahead of print] 10.1111/j.1751-553×.2011.01386.x.

38.

Zhong

, Wu

, Ma

et al. 2012. Association of Janus kinase 2 (JAK2) polymorphisms with acute leukemia susceptibility. Int J Lab Hematol., 34:248-253.