Uw

ARD

a

ARRA

KNNRCP

1

ptoibCCamr

(

0d

Journal of Reproductive Immunology 90 (2011) 105– 110

Contents lists available at ScienceDirect

Journal of Reproductive Immunology

j o ur nal homep age : w w w.elsev ier .com/ locate / j repr imm

terine and circulating natural killer cells and their roles in womenith recurrent pregnancy loss, implantation failure and preeclampsia

tsushi Fukui ∗, Ayano Funamizu, Megumi Yokota, Kenichi Yamada, Rika Nakamua,ie Fukuhara, Hidetaka Kimura, Hideki Mizunuma

epartment of Obstetrics and Gynecology, Hirosaki University Graduate School of Medicine, 5 Zaifu-cho, Hirosaki, Aomori 036-8562, Japan

r t i c l e i n f o

rticle history:eceived 25 December 2010eceived in revised form 29 March 2011ccepted 4 April 2011

eywords:K cellatural cytotoxicity receptorecurrent pregnancy lossytokine

a b s t r a c t

The regulation of uterine and circulating peripheral blood natural killer (NK) cells has beenassociated with reproductive conditions including recurrent pregnancy loss (RPL), implan-tation failure and preeclampsia. Natural cytotoxicity receptors (NCRs) are unique markersthat regulate NK cell cytotoxicity and cytokine production. The role of NCRs in reproduc-tive events has not yet been fully characterized. There is an NK1 (Type 1) shift in peripheralblood NK cells in non-pregnant women prone to RPL and implantation failure. The differentprofile of NCR expression in endometrial or aborted decidual NK cells suggests the pres-ence of abnormal regulation of NK cells in women with reproductive failure. Women with ahistory of RPL and preeclampsia carry immunological abnormalities of NCRs on peripheralblood NK cells during pregnancy. Evaluation of NKp46 on peripheral blood NK cells may

reeclampsia be applicable for the prediction of preeclampsia. The lower expression of NKp46+ NK cellsin women with preeclampsia may account for the higher production of NK1 cytokines –known as the NK1 shift – in pregnant women with preeclampsia. In this review, the expres-sion of NCRs in peripheral blood NK cells and endometrial or decidual NK cells is discussedin relation to reproductive failure.

. Introduction

Natural killer (NK) cells play an important role in humanregnancy and the systemic regulation of NK cells con-ributes to reproductive success. NK cells make up 5–10%f peripheral blood lymphocytes (PBL) and 70–90% of uter-ne lymphocytes are distinguished from other cell typesy the expression of NK-specific surface markers, i.e.,D56-positive cells. Human NK cell subsets exist as eitherD56bright or CD56dim NK cells. In addition, CD56bright cells

re mainly found in the decidua. In peripheral blood, theain population of NK cells consists of CD56dim cells. The

egulation of uterine and circulating peripheral blood NK

∗ Corresponding author. Tel.: +81 172 39 5107; fax: +81 172 37 6842.E-mail addresses: a.fukuipon@mac.com, atsushi@cc.hirosaki-u.ac.jp

A. Fukui).

165-0378/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.jri.2011.04.006

© 2011 Elsevier Ireland Ltd. All rights reserved.

cells may be associated with reproductive disorders suchas recurrent pregnancy loss (RPL), implantation failure andpreeclampsia.

Generally, NK cells can attack target cells without sen-sitization. The roles of NK cells are known to involvethe elimination of cancer cells or foreign substances. NKcells recognize the target cells using their activating andinhibitory receptors. In other words, NK cells can attack thecells that do not express major histocompatibility complex(MHC = HLA: human leukocyte antigen). There is no doubtthat there are abnormalities in the numbers or phenotypesof NK cells, with abnormal expression of NK cell surfacereceptors and cytokine production by NK cells in suchreproductive conditions. Therefore, it seems convincing

that the appropriate regulation of immune cells, especiallyNK cells, is important to reproductive success.

The main population of uterine NK cells consists ofCD16−/CD56bright cytokine-producing cells and that of

uctive I

106 A. Fukui et al. / Journal of Reprod

peripheral blood NK cells consists of CD16+/CD56dim cyto-toxic NK cells. It is reported that peripheral blood NKcell levels reflect changes in decidual NK cell levels (Parket al., 2010). We can easily gain access to the periph-eral blood; thus it is often used for diagnostic purposesto detect potential reproductive failure (Aoki et al., 1995;Ntrivalas et al., 2001). We have previously reported thatthe percentages of CD16+/CD56dim NK cells in peripheralblood and endometrium in IVF–ET cycles were signifi-cantly higher and that of CD16−/CD56bright NK cells wasmarkedly lower in reproductive failures compared withthe group of successful pregnancies (Fukui et al., 1999).Others have reported that examination of peripheral bloodNK cells would not be informative of events in the uterus(Saito, 2000; Moffett et al., 2004). It is evident that uterineNK cells are important for the achievement and mainte-nance of pregnancy. However, the function and usefulnessof peripheral blood NK cells are controversial.

In this review, expression of NCRs in peripheral bloodNK cells and endometrial or decidual NK cells wereassessed in relation to reproductive failure.

2. Natural cytotoxicity receptors

For the NK cell surface markers, natural cytotoxicityreceptors (NCRs) are unique markers that regulate NK cellcytotoxicity and cytokine production. NCRs, which includeNKp30, NKp44, and NKp46, are expressed exclusively onNK cells. The NKp46 receptor is a 46-kDa Type 1 membraneglycoprotein characterized by two C2-type Ig-like domainsin the extracellular portion and is associated with CD3�and Fc�RI� that become tyrosine-phosphorylated. TheNKp44 receptor is a 44-kDa Type 1 membrane glycopro-tein characterized by a single V-type Ig-like domain in theextracellular portion and is associated with KARAP/CAP12molecules. The NKp30 receptor is a 30-kDa Type 1 mem-brane glycoprotein characterized by a single V-type Ig-likedomain in the extracellular portion and associated withCD3� polypeptides (Biassoni et al., 2003). NCRs are themajor receptors involved in NK cytotoxicity and theirfunctions are the recognition and lysis of tumor cells byNK cells. The signal cascade of NCRs is as follows. First,cellular ligands, for which recognition by NCRs are notfully elucidated, bind to each NCR. Then, protein tyrosinekinases are activated. Protein tyrosine kinases phospho-rylate immunoreceptor tyrosine-based activation motifs(ITAM). Next, protein tyrosine kinases of zeta-chain asso-ciated protein of 70 kDa (Zap-70) or SYK are recruited andactivated. SLP76, p85-PI3K, LAT, PLC-g1, PLC-g2, and so onare then phosphorylated and activated. Finally, the intra-cellular release of calcium induces NK cell cytotoxicity andcytokine release (Fig. 1) (Biassoni et al., 2003).

NKp46 and NKp44, but not NKp30, recognize viralproteins such as the hemagglutinin of influenza, or thehemagglutinin-neuraminidase of the parainfluenza virus(Arnon et al., 2001; Mandelboim et al., 2001). The endoge-

nous cellular ligands recognized by NCRs are not known.Recently, NCR ligands were found to be expressed bymurine lymphoma and myeloma cell lines (Halftecket al., 2009) and in human primary nevi and melanomas

mmunology 90 (2011) 105– 110

(Cagnano et al., 2008; Lakshmikanth et al., 2009). TheNKp30 and NKp46 receptors are expressed on the surface ofactivated and non-activated NK cells. However, the NKp44receptor is expressed on the surface of activated NK cellsonly. In addition, the NKp30 and NKp46 receptors have afunction in the cytotoxic activity and cytokine productionof NK cells.

3. Peripheral blood NK cells and recurrentpregnancy loss and implantation failure

There is a shift toward Type 2 T cells in normal preg-nancy (Wegmann et al., 1993; Saito et al., 1999; Kwak-Kimet al., 2003). The Th1/Th2 concept has been extended bydemonstrating that natural killer (NK) cells can also showcomparable polarities in their cytokine secretion profiles(Carter and Dutton, 1995; Peritt et al., 1998). Cytokine pro-duction by both CD56bright NK cells and CD56dim NK cellsmay also shift away from Type 1 toward Type 2 duringpregnancy. Higuma-Myojo et al. (2005) first reported thecytokine profile of peripheral and decidual NK cells. Theyclearly described the NK1/NK2/NK3 (TGF-�-producing NKcells)/NKr1 (IL10-producing NK cells) paradigm duringpregnancy. However, there are no reports that demon-strate that IL4-, IL5-, and IL13-producing NK cells increaseduring pregnancy (Higuma-Myojo et al., 2005). On the con-trary, it is suggested that NK cell and lymphocyte responseare shifted away from a Type 1 immune response dur-ing pregnancy (Veenstra Van Nieuwenhoven et al., 2002).It was also reported that there were no changes in theTh1/Th2 or Tc1/Tc2 cell ratios. However, the NK1/NK2 cellratios were significantly decreased during normal preg-nancy (Borzychowski et al., 2005; Sargent et al., 2006)or NK1 cytokines such as TNF� and IFN� were signif-icantly higher in women with abortion (Higuma-Myojoet al., 2005; Thum et al., 2007; Saito et al., 2008). It seemsthat the maternal systemic immune response in pregnancy(Th1/Th2 shift) primarily involves NK cells (Sargent et al.,2007).

On the other hand, a propensity toward Th1 immuneresponses has been reported in women with recurrentpregnancy loss and implantation failure after IVF–ET sys-temically or locally. The presence of elevated Th1/Th2 cellratios, the high concentration of Th1 cytokines secreted byperipheral blood mononuclear cells (PBMCs), elevated NKcell cytotoxicity and levels, and the emergence of variousautoantibodies are the supporting evidence (Kwak-Kimet al., 2005).

In peripheral blood, we reported that the expression ofNKp46 cells in circulating NK cells is significantly lowerin women with recurrent pregnancy loss compared withnormal fertile controls (Fukui et al., 2006). This abnormalexpression of NKp46 on peripheral blood NK cells existswith non-pregnant status. NKp46 may function in bothcytotoxicity and cytokine production (Orange and Ballas,2006). The cytokine repertoire of peripheral NK cells ismainly Type 1 cytokines such as IFN� and TNF�, but NK

cells are capable of producing Type 2 cytokines such asIL4, IL5, and IL13 (Peritt et al., 1998; Loza and Perussia,2001). Our results, namely lower expression of NKp46 onperipheral blood CD56+ NK cells in women with RPL, may

A. Fukui et al. / Journal of Reproductive Immunology 90 (2011) 105– 110 107

oxicity

ectsftn(iwthllteatwoCCbNgitp

aaeeec

Fig. 1. Signal cascade of natural cytot

xplain NK cell abnormalities of cytokine production andytotoxicity in women with RPL. In fact, we reported thathe percentage of CD56bright/IFN�+/TNF�+ (NK1) cells wasignificantly higher in women with RPL and implantationailure compared with that of healthy controls. Moreover,he proportion of CD56bright/IL4+/IL10+ (NK2) cells was sig-ificantly lower in women with RPL than that of controlsFukui et al., 2008). We demonstrated that the NK1 shiftn peripheral blood NK cells was present in non-pregnant

omen with RPL and implantation failure. Then, we inves-igated whether women with RPL and implantation failureave aberrant correlation between NCRs and intracellu-

ar cytokine expression of NK cells. Recurrent pregnancyoss and implantation failure have a lack of, or nega-ive correlation between NCRs and intracellular cytokinexpression; namely, in normal fertile controls, the percent-ge of CD56bright/NKp30+ cells positively correlated withhat of CD56bright/IFN�+/TNF�+ cells. However, in womenith implantation failure, these did not correlate. More-

ver, in the implantation failure group, the percentage ofD56bright/NKp46+ cells negatively correlated with that ofD56bright/IFN�+/TNF�− cells. There were no correlationsetween any of the NCRs and Type 1 cytokine-expressingK cell levels in women with RPL. This observation sug-ests that excessive pro-inflammatory cytokine expressionn NK cells in RPL and implantation failure may be exertedhrough the NCRs or interruption of the signal transductionrocesses (Fukui et al., 2009).

Often the interaction between peripheral blood NK cellsnd trophoblasts is not well understood, and researchersre investigating this to define various roles of periph-

ral and decidual NK cells (Moffett et al., 2004; Bhallat al., 2006). NK cells preferentially kill targets with lowerxpression of major histocompatibility complex (MHC)lass I proteins, because fewer inhibitory receptors are

receptors (NCRs). See text for details.

engaging ligands. As a consequence, syncytiotrophoblastsare not free from peripheral blood NK cell cytotoxicity.Moreover, Moffett et al. (2004) reported that examinationof peripheral blood NK cells is not informative of what ishappening in the uterus. They stated that this is akin to esti-mating the number and activity of black cabs in TrafalgarSquare by analyzing red mini-cabs circulating on the M25”(Moffett et al., 2004). On the other hand, Park et al. (2010)reported that peripheral blood NK cells do reflect changesin decidual NK cells in women with recurrent miscarriages.Thus, this problem is still controversial.

To investigate this, we measured the correlationbetween NKp46+ NK cells in peripheral blood and in mid-secretory endometrium. There were significant positivecorrelations between the percentage of CD56+/NKp46+ andthe CD56bright/NKp46+ NK cells in peripheral blood andmidsecretory endometrium (Fukui, 2010a,b). Therefore, forthe expression of NKp46 on NK cells, it may be possibleto gain an understanding of the uterine NKp46+ NK cellsby evaluating peripheral blood NKp46+ NK cells. Addition-ally, in peripheral blood from pregnant women, we foundthe expression of NKp46 was significantly lower in womenwith a history of RPL compared with controls. Interestingly,this difference was observed at 12 and 20 weeks of gesta-tion, but disappeared after that. These data may mean thateven though their pregnancies are ongoing, at the time ofearly pregnancy, women with a history of RPL may experi-ence dysfunction of cytokine production by NK cells.

4. Natural cytotoxicity receptors and endometrium

The endometrial NK cell subpopulation before preg-nancy has not yet been fully characterized. It is reportedthat endometrial NK cells have a unique receptor reper-toire, or specifically, they are positive for NKp46, and

108 A. Fukui et al. / Journal of Reproductive Immunology 90 (2011) 105– 110

ral bloodlassified

Fig. 2. Representative dot plots of the expression of NCRs on (A) peripheLymphocytes are classified into CD56+ and CD56− cells. CD56+ cells are cendometrial CD56 with NCRs is evaluated.

negative or positive at low levels for NKp30 and NKp44,respectively, for both proliferative and secretory phaseendometrium (Manaster et al., 2008). However, our datashowed that endometrial NK cells are positive for notonly NKp46 but also NKp30 and NKp44 in secretory phaseendometrium (Fig. 2). We showed higher expression of

NKp30 and NKp44 on CD56dim NK cells in aborted deciduacompared with endometrium. Moreover, the expressionof NKp44 between CD56dim cells and CD56bright cells in

Fig. 3. Natural cytotoxicity receptors and NK cell-producing cytokines in repropregnancy loss, implantation failure and preeclampsia, there is a lower expresshigher production of TNF� and lower production of IFN�. Higher production of TNabnormal vessel remodeling. This is one of the causes of recurrent pregnancy loss

NK cells and (B) endometrial NK cells. A gate is set on the lymphocytes. into CD56bright and CD56dim cells. Co-expression of peripheral blood or

women with implantation failure was significantly dif-ferent; namely, the expression of NKp44 on CD56bright

cells was significantly up-regulated compared with con-trols. This may mean that there is high NK cell cytotoxicityin implantation failure (Fukui, 2010a,b). Further study isneeded to clarify how NCR in endometrial NK cells partici-

pate in reproduction.

It is reported that NKp46, NKp44, and NKp30 areexpressed on decidual NK cells (Zhang et al., 2008).

ductive failure. In women with reproductive failure such as recurrention of NKp46 on peripheral blood and endometrial NK cells. It leads toF� leads to an NK1 shift of NK cells and lower production of IFN� causes, implantation failure and preeclampsia.

uctive Im

TdtctTNtacdc2

5

ai2dpDd(te2rewpwpucprspenhcpIpo2otmtpPnTNTp

A. Fukui et al. / Journal of Reprod

hey found higher expression of NKp46 and NKp44 onecidual NK cells in women with spontaneous abor-ion. They concluded that enhanced decidual NK cellytotoxicity due to higher NKp46 and NKp44 might con-ribute to the susceptibility to spontaneous abortion.hey also showed the higher cytotoxicity for NCR-positiveK cells. However, they did not find any difference in

he expression of NKp30. We report here that therere significantly higher expressions of CD56dim/NKp44+

ells and CD56dim/NKp30+ cells in spontaneously abortedecidua compared with endometrium. These data indi-ates that aborted decidua has a higher cytotoxicity (Fukui,010a,b).

. Natural cytotoxicity receptors and preeclampsia

Pregnancy has been reported to be associated with shift away from Th1-type and bias toward Th2-typemmune responses (Wegmann et al., 1993; Kwak-Kim et al.,003, 2005). The Th1/Th2 concept has been extended byemonstrating that NK cells can also show comparableolarities in their cytokine secretion profiles (Carter andutton, 1995; Peritt et al., 1998). It is reported that pre-ominant Th1-type immunity is present in preeclampsiaSaito and Sakai, 2003). A later study demonstrated thathe production of the Th1-type cytokine, IFN�, is alsonhanced in NK cells (Germain et al., 2007; Saito et al.,010). Namely, the Type 1 shift of NK cells has beeneported in preeclampsia (Borzychowski et al., 2005). Inter-stingly, Borzychowski et al. (2005) reported that thereere no changes in the Th1/Th2 or Tc1/Tc2 cell ratios inreeclamptic women. However, the NK1/NK2 cell ratiosere significantly decreased in normal pregnancy com-ared with non-pregnant and preeclamptic women. Theysed pan Type 1 (IL18) and pan Type 2 (ST2L) lympho-yte function markers in flow cytometry to characterizeeripheral blood lymphocyte populations. Moreover, aber-ant NK cell activation, both locally in the decidua andystemically in the maternal blood may be the cause ofreeclampsia (Sargent et al., 2007). We have analyzed thexpression of NCRs on peripheral blood NK cells in preg-ant women with preeclampsia. Pregnant women whoad preeclampsia showed a significantly decreased per-entage in CD56bright/NKp46+ cells compared with that inregnant women without preeclampsia (Fukui, 2010a,b).

nterestingly, lower expression of CD56+/NKp46+ cells inregnant women was observed 3–4 months before thenset of preeclampsia. In other words, it was observed at0 weeks of gestation and continued until delivery. Onsetf preeclampsia in our study was at 34 ± 6 weeks of gesta-ion. Therefore, NKp46 has the potential to become a useful

arker for prediction of preeclampsia along with other fac-ors, such as soluble fms-like tyrosine kinase-1 (sFlt1) andlacental growth factor (PlGF) (Moore Simas et al., 2007;oon et al., 2009). CD56+/NKp46+ cells in pregnant womenegatively correlated with CD56bright/IFN�−/TNF�+ cells.

his means that evaluation of NKp46 on peripheral bloodK cells may be applicable for prediction of preeclampsia.he lower expression of NKp46+ NK cells in women withreeclampsia may account for the higher production of NK1

munology 90 (2011) 105– 110 109

cytokine that is known as the NK1 shift in pregnant womenwith preeclampsia.

6. Natural cytotoxicity receptor and IL22production

Recently, a new type of NKp46+ NK cell, which producesIL22, was reported (Crellin et al., 2010; Satoh-Takayamaet al., 2010; Veiga-Fernandes et al., 2010). These cells arecalled NCR22 cells (Satoh-Takayama et al., 2010) or NK-22cells (Cella et al., 2009). It is considered that IL22 productionby NCR22 cells might mediate mucosal immune defense,i.e., intestine and tonsils. These cells may become a keyfactor for reproduction, but this has not yet been charac-terized. Many interesting questions remain regarding thefunctions of NCR22 cells and the role of IL22 in reproductivefailure and various diseases.

7. Conclusion

Appropriate numbers and phenotypes of NK cells areimportant in pregnancy. There is abnormal expressionof NK cell surface antigens and dysregulation of NKcell cytotoxicity and cytokine production in women withreproductive failure (Fig. 3). The receptors and cytokines inperipheral blood and endometrial NK cells are critical forthe successful achievement and maintenance of pregnancy.Abnormal expression of these factors is associated not onlywith implantation failure and recurrent pregnancy loss, butalso the onset of preeclampsia.

References

Aoki, K., et al., 1995. Preconceptional natural-killer-cell activity as a pre-dictor of miscarriage. Lancet 345, 1340–1342.

Arnon, T.I., et al., 2001. Recognition of viral hemagglutinins by NKp44 butnot by NKp30. Eur. J. Immunol. 31, 2680–2689.

Bhalla, A., et al., 2006. Comparison of the expression of human leukocyteantigen (HLA)-G and HLA-E in women with normal pregnancy andthose with recurrent miscarriage. Reproduction 131, 583–589.

Biassoni, R., et al., 2003. Human natural killer cell receptors: insights intotheir molecular function and structure. J. Cell. Mol. Med. 7, 376–387.

Borzychowski, A.M., et al., 2005. Changes in systemic type 1 and type 2immunity in normal pregnancy and pre-eclampsia may be mediatedby natural killer cells. Eur. J. Immunol. 35, 3054–3063.

Cagnano, E., et al., 2008. Expression of ligands to NKp46 in benign andmalignant melanocytes. J. Invest. Dermatol. 128, 972–979.

Carter, L.L., Dutton, R.W., 1995. Relative perforin- and fas-mediated lysisin T1 and T2 CD8 effector populations. J. Immunol. 155, 1028–1031.

Cella, M., et al., 2009. A human natural killer cell subset provides an innatesource of IL22 for mucosal immunity. Nature 457, 722–725.

Crellin, N.K., et al., 2010. Human NKp44+IL22+ cells and LTi-like cells con-stitute a stable RORC+ lineage distinct from conventional natural killercells. J. Exp. Med. 207, 281–290.

Fukui, A., 2010a. NK cells and its role in reproduction. Am. J. Reprod.Immunol. 64, 1.

Fukui, A., 2010b. Uterine and circulating natural killer cells and their rolesin women with recurrent pregnancy losses, implantation failures orpreeclampsia. J. Reprod. Immunol. 86, 14.

Fukui, A., et al., 1999. Natural killer cell subpopulations and cytotoxicityfor infertile patients undergoing in vitro fertilization. Am. J. Reprod.Immunol. 41, 413–422.

Fukui, A., et al., 2006. Expression of natural cytotoxicity receptors anda2V-ATPase on peripheral blood NK cell subsets in women with recur-

rent spontaneous abortions and implantation failures. Am. J. Reprod.Immunol. 56, 312–320.

Fukui, A., et al., 2008. Intracellular cytokine expression of peripheral bloodnatural killer cell subsets in women with recurrent spontaneous abor-tions and implantation failures. Fertil. Steril. 89, 157–165.

uctive I

110 A. Fukui et al. / Journal of Reprod

Fukui, A., et al., 2009. Correlation between natural cytotoxicity receptorsand intracellular cytokine expression of peripheral blood NK cells inwomen with recurrent pregnancy losses and implantation failures.Am. J. Reprod. Immunol. 62, 371–380.

Germain, S.J., et al., 2007. Systemic inflammatory priming in normal preg-nancy and preeclampsia: the role of circulating syncytiotrophoblastmicroparticles. J. Immunol. 178, 5949–5956.

Halfteck, G.G., et al., 2009. Enhanced in vivo growth of lymphoma tumorsin the absence of the NK-activating receptor NKp46/NCR1. J. Immunol.182, 2221–2230.

Higuma-Myojo, S., et al., 2005. Cytokine profile of natural killer cells inearly human pregnancy. Am. J. Reprod. Immunol. 54, 21–29.

Kwak-Kim, J.Y., et al., 2003. Increased T helper 1 cytokine responses bycirculating T cells are present in women with recurrent pregnancylosses and in infertile women with multiple implantation failures afterIVF. Hum. Reprod. 18, 767–773.

Kwak-Kim, J.Y., et al., 2005. T helper 1 and 2 immune responses in relation-ship to pregnancy, nonpregnancy, recurrent spontaneous abortionsand infertility of repeated implantation failures. Chem. Immunol.Allergy 88, 64–79.

Lakshmikanth, T., et al., 2009. NCTs and DNAM-1 mediate NK cell recog-nition and lysis of human and mouse melanoma cell lines in vitro andin vivo. J. Clin. Invest. 119, 1251–1263.

Loza, M.J., Perussia, B., 2001. Final steps of natural killer cell maturation:a model for type 1-type 2 differentiation? Nat. Immunol. 2, 917–924.

Manaster, I., et al., 2008. Endometrial NK cells are special immature cellsthat await pregnancy. J. Immunol. 181, 1869–1876.

Mandelboim, O., et al., 2001. Recognition of haemagglutinins on virus-infected cells by NKp46 activates lysis by human NK cells. Nature 409,1055–1060.

Moffett, A., et al., 2004. Natural killer cells, miscarriage, and infertility. BMJ329, 1283–1285.

Moore Simas, T.A., et al., 2007. Angiogenic factors for the prediction ofpreeclampsia in high-risk women. Am. J. Obstet. Gynecol. 197 (244),e1–8.

Ntrivalas, E.I., et al., 2001. Status of peripheral blood natural killer cellsin women with recurrent spontaneous abortions and infertility ofunknown aetiology. Hum. Reprod. 16, 855–861.

Orange, J.S., Ballas, Z.K., 2006. Natural killer cells in human health anddisease. Clin. Immunol. 118, 1–10.

mmunology 90 (2011) 105– 110

Park, D.W., et al., 2010. Peripheral blood NK cells reflect changes in decid-ual NK cells in women with recurrent miscarriages. Am. J. Reprod.Immunol. 63, 173–180.

Peritt, D., et al., 1998. Differentiation of human NK cells into NK1 and NK2subsets. J. Immunol. 161, 5821–5824.

Poon, L.C., et al., 2009. First-trimester prediction of hypertensive disordersin pregnancy. Hypertension 53, 812–818.

Saito, S., 2000. Cytokine network at the feto-maternal interface. J. Reprod.Immunol. 47, 87–103.

Saito, S., Sakai, M., 2003. Th1/Th2 balance in preeclampsia. J. Reprod.Immunol. 59, 161–173.

Saito, S., et al., 1999. Distribution of Th1, Th2, and Th0 and the Th1/Th2cell ratios in human peripheral and endometrial T cells. Am. J. Reprod.Immunol. 42, 240–245.

Saito, S., et al., 2008. The balance between cytotoxic NK cells and regula-tory NK cells in human pregnancy. J. Reprod. Immunol. 77, 14–22.

Saito, S., et al., 2010. Th1/Th2/Th17 and regulatory T-cell paradigm inpregnancy. Am. J. Reprod. Immunol. 63, 601–610.

Sargent, I.L., et al., 2006. NK cells and human pregnancy – an inflammatoryview. Trends Immunol. 27, 399–404.

Sargent, I.L., et al., 2007. NK cells and pre-eclampsia. J. Reprod. Immunol.76, 40–44.

Satoh-Takayama, N., et al., 2010. IL7 and IL15 independently programthe differentiation of intestinal CD3-NKp46+ cell subsets from Id2-dependent precursors. J. Exp. Med. 207, 273–280.

Thum, M.Y., et al., 2007. The relationship of systemic TNFalpha andIFNgamma with IVF treatment outcome and peripheral blood NK cells.Am. J. Reprod. Immunol. 57, 210–217.

Veenstra Van Nieuwenhoven, A.L., et al., 2002. Cytokine production innatural killer cells and lymphocytes in pregnant women comparedwith women in the follicular phase of the ovarian cycle. Fertil. Steril.77, 1032–1037.

Veiga-Fernandes, H., et al., 2010. Natural killer receptors: the burden of aname. J. Exp. Med. 207, 269–272.

Wegmann, T.G., et al., 1993. Bidirectional cytokine interactions in the

maternal-fetal relationship: is successful pregnancy a Th2 phe-nomenon? Immunol. Today 14, 353–356.

Zhang, Y., et al., 2008. Expressions of natural cytotoxicity receptors andNKF2D on decidual natural killer cells in patients having spontaneousabortions. Fertil. Steril. 90, 1931–1937.