During pregnancy, the maternal immune system is modified to enable survival of the semi-allogeneic fetus. It was originally proposed that type 2 cytokines (e.g. IL-4, IL-5, IL-6, IL-10) are prevalent, and type 1 cytokines (e.g. IFNγ, IL-2, IL-12, TNFα and TNFβ) are diminished, to prevent the activation of cell mediated immunity and inflammation (^{Wegmann et al., 1993}). However, the finding that pregnancy itself is an inflammatory state has led to a revision of this hypothesis and it is now apparent that both arms of the immune response are intensified, but with a stronger bias towards type 2 than type 1 responses in successful pregnancy (^{Redman and Sargent, 2003; Chaouat et al., 2007}). The causes behind this cytokine shift are not clearly defined, but it is known that the placenta produces type 2 cytokines such as IL-4 and IL-10, as well as progesterone (^{Sacks et al., 2001; Szekeres-Bartho et al., 2009}), which may influence maternal immunity (^{De Moraes-Pinto et al., 1997; Chaouat et al., 1999}).

Preeclampsia is the most common medical complication of pregnancy affecting 3–5% of all pregnancies. The maternal syndrome is characterised by new-onset hypertension and proteinuria which are caused by the excessive release of placenta derived factors into the blood. These factors stimulate an exaggerated maternal systemic inflammatory response of which endothelial dysfunction is a part (^{Redman and Sargent, 2009}). In preeclampsia, the shift away from type 1 responses either fails to occur or is reversed, with increased production of IL-18 and IFNγ relative to normal pregnancy (^{Germain et al., 2007}). This type 1 bias can be seen in T cells (^{Saito et al., 2007}) and more predominantly in NK and NKT-like (CD3+CD56+) cells (^{Borzychowski et al., 2005}). The latter cell type are of particular interest as recently a specific subset of invariant natural killer T cells (iNKT) has been defined, which link innate and adaptive immune responses (^{Matsuda et al., 2008}).

iNKT cells are a unique population that express an invariant T cell receptor (TCR), comprised of Vβ11 and Vα24 chains (^{Porcelli et al., 1993; Prussin and Foster, 1997}). They also share characteristics with NK cells, in that they are cytotoxic through the action of granzymes and perforins, and are potent producers of a wide variety of cytokines (^{Crowe et al., 2003; Van Kaer, 2004}). iNKT cells are important regulators of host defence in microbial and viral infections (^{Bendelac et al., 2007}), and are involved with the development or progression of a range of conditions, such as cancer, autoimmunity or allergy (^{Godfrey and Kronenberg, 2004}).

iNKT cells recognise and are activated by endogenous or exogenous lipid ligands presented by the major histocompatibility complex class I-like molecule, CD1d, on antigen presenting cells (^{Barral and Brenner, 2007}). They can be activated using a pharmacological agent; α-galactosylceramide (α-galcer) which is based on a glycosphingolipid isolated from the marine sponge Agelas mauritianus, and has antitumour activity in the mouse B16 melanoma model (^{Morita et al., 1995; Kawano et al., 1997}). Activation with α-galcer leads to the release of both type 1 and type 2 cytokines and the secondary activation of T cells, B cells and NK cells. The capacity of iNKT cells to produce cytokines can be defined by their differential expression of CD4 and CD8. CD4+ iNKT cells have a Th0 cytokine response in that they produce both type 1 and type 2 cytokines, CD8+ iNKT are primarily type 1 cytokine producers and double negative (CD4−/CD8−) iNKT produce type 1 cytokines but only low levels of IL-4 (^{Exley et al., 1997; Takahashi et al., 2000; Gumperz et al., 2002; Kim et al., 2002; Lee et al., 2002; Takahashi et al., 2002}).

Several observations suggest that iNKT cells may play a key role in pregnancy and preeclampsia. First, treatment of pregnant mice with a-galcer results in abortion due to iNKT IFN-γ release (^{Ito et al., 2000; Boyson et al., 2006}). Secondly, iNKT cells play a role in priming Th2 immune responses (^{Prussin and Foster, 1997; Singh et al., 1999}). Thirdly, iNKT cells regulate NK cell activation and proliferation (^{Carnaud et al., 1999; Eberl and Macdonald, 2000; Metelitsa et al., 2001}), cells that are known to be critical for the progression and success of pregnancy.

We have therefore investigated the absolute count, phenotype (CD4, CD8 and CD69 expression) and functional activity (IFNγ production) of iNKT cells in each trimester of pregnancy and in matched trios of non-pregnant, normal pregnant and preeclamptic women. Our aim was to determine whether there are changes in iNKT cells associated with pregnancy, and if so, how they might differ in preeclampsia.

Changes in iNKT cell numbers, phenotype and function are associated with progression of various diseases, for example in allergies such as asthma (^{Matangkasombut et al., 2009}), autoimmunities such as systemic lupus erythematosus (^{Gabriel et al., 2009}) and cancer (^{Molling et al., 2008}), often with associated cytokine imbalance. The ability of iNKT cells to drive and modulate immune responses makes them attractive candidates for being key regulators of immunity during pregnancy. Boyson et al. have previously shown that iNKT cells are present in first trimester (7–10 weeks) decidua at a frequency 10 times greater than seen in peripheral blood (^{Boyson et al., 2002}). It is possible that they interact with CD1d expressed on the invasive extravillous cytotrophoblast. Interestingly CD4+ve decidual iNKT cell clones stimulated with CD1d transfected C1R cells (a MHC class I negative EBV transformed B cell line) pulsed with α-galcer showed a strong bias to IFNγ (type 1) production compared to peripheral blood iNKT clones which were biased to IL-4 production (type 2).

In this study we focussed on iNKT cells in peripheral blood as our aim is to study their role in the maternal syndrome of preeclampsia which is a maternal systemic inflammatory disorder of the second half of pregnancy. We first investigated iNKT cells during each trimester of pregnancy, and found that while the number and type of iNKT cells do not change, the cells become activated, as demonstrated by increased CD69 expression on CD4−ve/CD8−ve and CD8+ iNKT cells during the third trimester. IFNγ production is reduced, which is consistent with previous observations showing a shift to a type 2 phenotype on NK and NKT-like cells (^{Borzychowski et al., 2005}) and a type 2 bias in cytokine production in pregnancy (^{Germain et al., 2007}). This data suggests that iNKT cells contribute to this cytokine bias. We were unable to measure the concomitant production of IL-4 due to ethical limitations on the volume of blood we were able to take for these studies, but from the results of Boyson et al. it might be expected that an increase in IL-4 production by iNKT cells would also be seen later in pregnancy. The increased activation of CD4 negative cells, typically type 1 cytokine producers, is paradoxical, and may represent a portion of cells that have been activated but have become anergic in response to chronic stimulation in vivo, a feature of activated human and mouse iNKT cells after long term exposure to α-galcer (^{Sullivan and Kronenberg, 2005}).

From the above results it might be expected that iNKT cells could play an important role in preeclampsia where increased IFNγ production by PBMC is seen compared to normal pregnancy (^{Germain et al., 2007}). However, no differences in iNKT cell number or phenotype were seen between women with preeclampsia and their matched normal pregnant and non-pregnant controls. A slight decrease in iNKT cell production of IFNγ was seen in normal pregnant compared to preeclamptic women, but this was not statistically significant. There has been one previous study of peripheral blood iNKT cells in preeclampsia (^{Miko et al., 2008}). In contrast to our study, it was found that the percentage of activated (CD69+ve) iNKT cells was significantly higher in preeclampsia compared to the matched normal pregnant and non-pregnant controls as was the percentage of iNKT cells producing IFNγ (as determined by intracellular cytokine staining). They did not investigate iNKT cells at other stages of gestation. The differences in results between the studies may be due to differences in the methodology. Miko et al. stained 50 μl whole blood for the iNKT TCR using the 6B11 antibody (^{Montoya et al., 2007}), and analysed at least 1000 iNKT cells, i.e. greater than 20,000 cells/ml, which is very high for iNKT cells. Our data are more consistent with other reports, for example ^{Molling et al. (2005)} who show an average of approximately 1000 iNKT/ml blood at 30 years age; our patients had a range of just 46 to 14,563 iNKT cells/ml blood. Miko et al. gated their lymphocytes on forward and side scatter only rather than using CD3 staining as in our study, which could have lead to the inclusion of false positive events in their analysis.

In summary, we show evidence for activation of iNKT cells and a decrease in IFNγ production as pregnancy progresses. However, no significant changes in CD69 expression or IFNγ production were seen between preeclamptic and normal pregnant women, suggesting that iNKT cells are not a major player in the pathogenesis of this condition.

This study was supported by a grant from the Wellcome Trust (ref GR079862MA) and by the Oxford Partnership Comprehensive Biomedical Research Centre with funding from the Department of Health's NIHR Biomedical Research Centres funding scheme. The views expressed in this publication are those of the authors and not necessarily those of the Department of Health. The authors would like to thank Carol Simms, Hazel Meacher, Nicola Higgins, Ali Chevassut and Linda Holden for their help in recruiting patients and collecting blood samples for this study.