108326 2395.2400

Human Reproduction Vol.19, No.10 pp. 2395–2400, 2004 Advance Access publication August 19, 2004 An increase in the absolute count of CD56dimCD161CD691NK cells in the peripheral blood is associated with a poorerIVF treatment and pregnancy outcome M.Y.Thum1,2,3,4, S.Bhaskaran2, H.I.Abdalla1, B.Ford2, N.Sumar2, H.Shehata3and A.S.Bansal2 1Lister Fertility Clinic, Lister Hospital, Chelsea Bridge Road, London SW1W 8RH, 2Immunology Department and 3Women HealthDepartment, Epsom and St Helier University Hospitals NHS Trust, Surrey, UK 4To whom correspondence should be addressed. E-mail: [email protected] BACKGROUND: Our aim was to evaluate the effect of the absolute count of the activation marker (CD69), IgG Fc receptor (CD16) and inhibitor marker (CD94) expression on peripheral blood natural killer (NK) cells on implan-tation and miscarriage rates after IVF treatment. METHODS: Prospective observational study of 138 randomlyselected women who underwent IVF treatment from December 2002 to September 2003. NK cells were identifiedas CD561 (dim 1 bright) and CD3 – by flow cytometry. The absolute counts of the CD691, CD161 and CD941expressing NK cells were recorded and their relation to IVF treatment outcome and miscarriage rate was ana- lysed. RESULTS: The mean (6 SD) absolute count of the CD56dimCD161CD691 NK cells for women who had asuccessful ongoing pregnancy was 0.61 3 106/l (6 0.31). For those women who failed to achieve a pregnancy, themean value of the absolute count of CD56dimCD161D691 NK cells was significantly (P 5 0.003) higher at1.66 3 106/l (6 0.52). The absolute count of CD56dimCD161CD941 and CD56dimCD161 NK cells did not show anystatistically significant differences between those women with successful and failed IVF treatment. Receiver opera-ting characteristic (ROC) curve analysis was performed to select a CD69 threshold for further statistical analysis.
The implantation rate (IR) was significantly lower (13.1%) and miscarriage rate (MR) was significantly higher(66.7%) for women with an absolute CD56dimCD161CD691 NK cell count of > 1.0 3 106/l compared to women with count below this value (IR 28.2% and MR 16.7%). Further analysis of the absolute count of CD56brightCD691and CD56brightCD941 NK cells did not show any significant difference between those women with successful andfailed IVF treatment. CONCLUSIONS: An increase in the absolute count of activated NK cells (CD56dimCD161CD691) in the peripheral blood is associated with a reduced rate of embryo implantation in IVF treatment. Fur-thermore, women with high CD56dimCD161CD691 peripheral blood NK cell absolute count, who are able toachieve pregnancy, have a significantly higher miscarriage rate.
Key words: activation markers/CD69/flow cytometry/IVF/natural killer cells A previous study by Beer et al. (1996) showed that an The natural killer (NK) cell is the most abundant immune elevated percentage of peripheral blood NK cells was associ- cell infiltrating the uterine implantation site (Moffett-King, ated with recurrent failed IVF treatment cycles. Later, Fukui 2002). It is the first line cellular immune defence mechanism et al. (1999) showed that increased peripheral blood NK cell and has a close contact with conceptus or placenta. NK cells cytotoxicity level was associated with an increased rate of comprise , 15% of all lymphocytes and are defined pheno- recurrent failed implantation after IVF treatment. More recent typically by expression of CD56 and lack of expression of studies have confirmed elevated CD69 expression on NK CD3 on the cell surface (Robertson and Ritz, 1990). The cells as being associated with recurrent miscarriage and infer- majority (, 90%) of human NK cells have low density tility of unknown aetiology (Ntrivalas et al., 2001). Finally, a expression of CD56 (CD56dim) and express high levels of recent small non-randomized study by Coulam and Roussev Fcg receptor IIIa (FcgRIIIa; CD16þ) whereas , 10% of NK (2003) revealed that infertile women undergoing IVF treat- cells are CD56brightCD16dim or CD56brightCD162 (Cooper ment also have a higher percentage of elevated CD69 et al., 2001). Uterine NK cells appear to be CD56bright and expression on NK cells as compared to multiparous women.
increase in number during the post-ovulatory luteal phase CD69 belongs to the C-lectin type superfamily and is a type II integral membrane protein consisting of a disuphide-linked Human Reproduction vol. 19 no. 10 q European Society of Human Reproduction and Embryology 2004; all rights reserved homodimer with two phosphorylated chains (Ziegler et al., soft catheter (Wallace) with transabdominal ultrasound guidance.
1993). It is a functional triggering molecule on activated NK Progesterone supplement for luteal support (Cyclogest; Shire Phar- cells and is one of the earliest cell surface activation markers maceuticals Ltd, UK), 400 mg once a day per vaginum or per rec- expressed (Yokoyama, 1999). It is capable of inducing cyto- tum, was commenced 1 day before embryo transfer and continued toxicity and stimulating cytokine production (Zingoni et al., until a pregnancy test was performed 2 weeks after embryo transfer.
2000). Besides mediating NK cell cytotoxicity, it also medi-ates other NK cell functions such as proliferation, tumour Flow cytometric NK activation and inhibition quantification assay necrosis factor (TNF-a) production and expression of other Peripheral blood was collected in heparinized tubes and analysed activation antigens (Borrego et al., 1999; Pisegna et al., within 24 h. Fifty millilitres of blood was placed in flow cytometric tubes (Becton Dickinson) and each incubated for 15 min at roomtemperature with mouse anti-human CD16 – fluorescein isothio- CD94 is an inhibitory marker of NK cell function. It is cyanate (FITC), anti-CD56 phycoerythrin (PE) (BD PharMingen), part of the killing inhibitory receptor (KIR) family which is a anti-CD3 PE Cy5 (Quest Biomedical), together with CD69 or CD94 sub-group of the C type lectin superfamily (Lopez-Botet APC (BD PharMingen) monoclonal antibodies (mAb). Isotypic con- et al., 1997). Borrego et al. (1999) demonstrated that NK cell trol mAb included mouse IgG1 FITC, IgG1 APC, IgG1 PE (BD cytotoxicity could be blocked by the CD94 inhibitory recep- PharMingen) and IgG1 PE– Cy5 (Quest Biomedical). In this lyse, no tor. Previous studies have shown that imbalances in CD69 wash procedure, 1 ml of Quicklysis lysing solution (Quest Biomedi- and CD94 expression on NK cell could result in infertility of cal) was added to each tube and incubated for a further 10 min at unknown aetiology or recurrent miscarriage (Ntrivalas et al., room temperature. Fifty millilitres of PerfectCount beads (Quest Biomedical) were then accurately pipetted to each tube and samples CD16 (also classified as FcgRIIIa) is one of the low affi- run with BD FACSCalibur flow cytometer. Cells negatively staining nity receptors for the Fc region of IgG. FcgRIIIa is an inte- for CD3, but positively for CD56, were selected and their CD69 andCD94 expression analysed using a Cell Quest software (BD) using a gral membrane protein expressed on NK cells, on a subset of T lymphocytes, and on a subpopulation of monocytes andmacrophages (Ravetch and Perussia, 1989). Previously Fukui et al. (1999) showed that an increased percentage of peri- pheral blood CD16þ NK cells was associated with failed All IVF data were collected in Medical System for IVF (Medical- Sys, UK) and analysed by Statistics Package for Social Sciences(SPSS, UK). Descriptive statistical analysis was performed initially The aim of this study was to document any association to examine the normal distribution of all continuous variances for between the absolute count of activation marker (CD69), IgG parametric statistical tests. The t-test was then used to compare the Fc receptor (CD16) and inhibitor marker (CD94) expressing mean value in two groups: pregnant and not pregnant. Receiver NK cells on the implantation and miscarriage rate after IVF operating characteristic (ROC) curve and area under curve (AUC) analysis were performed. The ROC curve represents the probabilityof true positive results (sensitivity) as a function of the probabilityof false positive results (1 - specificity). The AUC is a measure of the accuracy of a test. In order to perform the ROC curve calcu-lation, set threshold values for the state variable (CD69 absolute count) were selected. These are arbitrary values selected to analyse From December 2002 to July 2003, 138 patients undergoing IVF the association of treatment outcome. For each CD69 threshold, sep- treatment cycles were recruited into the study. Independent ethical arate curves were produced for treatment outcome and pregnancy approval was obtained from the Local Research Ethics Committee.
outcome, making a total of 14 curves. x2 cross-tabulation test was Exclusion criteria: women with known immunological disease (anti- used to analyse the significance of differences in pregnancy rates, phospholipid antibodies, lupus anticoagulant, anticardiolipin anti- miscarriage rates and live birth rates between groups. Analysis of bodies), uterine abnormality (fibroid, uterine polyp, uterine septum), variance was then conducted to assess the duration and amount of fewer than two embryos available for transfer or endometrium thick- gonadotrophin required to achieve follicular maturity, estradiol ness , 8 mm before embryo transfer. All transvaginal ultrasound levels on hCG day, number of mature follicles, number of available scans were performed by a sonographer and exclusion of candidates embryos for transfer, number of oocytes collected and fertilization was performed without the knowledge of the NK cell blood test result. Blood samples were obtained on the day of vaginal oocytecollection prior to the procedure. Informed consent was provided byall subjects at recruitment.
Of the 138 women who underwent IVF, 12 were excluded from statistical analysis. Of these 12, four had failed fertiliza- Pituitary down-regulation was achieved with either nafarelin or tion, four had only one embryo available for transfer, one buserelin at mid-luteal phase. Ovarian stimulation was carried out had ovarian hyperstimulation syndrome and therefore did not with either recombinant FSH, hMG or urinary FSH. When follicles have embryo transfer, two had an endometrial thickness reached pre-ovulatory size (18 – 22 mm), 10 000 IU of hCG wasadministrated. Oocytes were aspirated using transvaginal ultrasound , 7.5 mm and one woman had poor quality embryos. None guidance 34 – 36 h after hCG administration. All embryos were of the women who participated in the study were excluded allowed to cleave and the best two or three embryos were selected due to abnormal uterine anatomy or known previous abnor- for transfer. Embryo transfer was performed on day 2 or 3 using a CD691 peripheral NK cells and IVF outcome CD69-expressing NK cells was significantly higher at Table I. Patients’ demographics and stimulation characteristic between 1.66 £ 106/l ^ 0.52 £ 106 (P ¼ 0.003). The absolute count of CD94 and CD16 expressing CD56dim NK cells and CD56bright NK cells showed no significant difference between those women with successful and failed IVF treatment.
For the ROC analysis, 14 curves were produced, one for treatment outcome (pregnancy rate) and one for pregnancy outcome (live birth rate) for each of the seven chosen CD69 absolute count threshold. Table III illustrates the test charac- teristics and AUC for each selected CD69 absolute count threshold. The AUC values were maximum at a CD69 absol- ute count of 1.0 £ 106/l for both treatment outcome (preg- nancy rate) and pregnancy outcome (live birth rate). This indicated that 1.0 £ 106/l is a better level for further statisti- cal analysis as compared to the other selected levels. Increas- ing the threshold level improved specificity and positive predictive value at the expense of sensitivity. Therefore the threshold value of 1.0 106/1 was selected for further statistical analysis without selecting a threshold level for the The study population was then divided into two groups based on the absolute count of CD69-expressing NK cells.
Group A women (n ¼ 87) had a count , 1.0 £ 10þ6/l while group B women (n ¼ 39) had a count . 1.0 £ 106/l.
bMean amount of gonadotrophin used for stimulation in IU (recombinant Table IV shows number of women in each group, their mean FSH, hMG or urinary FSH).
NA ¼ not applicable; NS ¼ difference not statistically significant age, causes of infertility, duration of infertility, basal FSH levels, mean number of previous IVF attempts and previousmiscarriages, stimulation characteristics and treatment out- Table I shows patients’ treatment outcome, mean age, come in both groups. There were no significant differences causes of infertility, duration of infertility, basal FSH levels, between group A and group B with regard to age, causes and mean number of previous IVF attempts, number of previous duration of infertility and basal FSH levels. The mean num- miscarriages and outcome of ovarian stimulation in the preg- ber of previous failed IVF attempts and the mean number of nant and non-pregnant groups. There were no significant previous miscarriages were significantly higher in group B as differences between the two groups with regard to any of the compared to group A. There was no significant difference between the two groups with regard to amount of gonado- Table II examines the relationship between the absolute trophin used for stimulation, estradiol levels on hCG day, count of CD69, CD94 and CD16 expressing CD56dim cells number of oocytes collected, fertilization rate, number of and the absolute count of CD69 and CD94 expressing available embryos for transfer or number of embryos trans- ferred. In group B, the implantation rate, pregnancy rate and ant versus non-pregnant. For women who had a success- live birth rate were significantly lower, and the miscarriage rate was significantly higher as compared to group A.
0.61 £ 106/l ^ 0.31 £ 106. For those women who failed to achieve pregnancy, the mean value of the absolute count of CD69 is one of the earliest specific activation markersexpressed during large granulated lymphocyte activation,which includes the NK cell (Craston et al., 1997; Marzio Table II. Natural killer cell sub-population and CD69 expression between et al., 1999; Llera et al., 2001). Activated CD69þ NK cells will release cytokines which will further activate other NK cells and the cellular immune system (Marzio et al., 1999).
Previous studies have also shown that elevated CD69 expression on NK cells is associated with an increase in cyto- toxicity of NK cells towards target cells, which induces target cell lysis (Lanier et al., 1988; De Maria et al., 1994). Chao et al. (1999) have suggested that maternal NK cell CD69expression is involved in recognition of HLA-G and HLA-C Values are mean ^ SD absolute count ( £ 106/l) unless otherwise stated.
on the allogeneic embryonic and trophoblast cell surface. In NA ¼ not applicable; NS ¼ difference not statistically significant(P . 0.05).
theory, recognition of HLA-G and HLA-C expression is Table III. The test characteristics and area under the curve (AUC) for each selected CD69 absolute count threshold AUC ¼ area under the ROC (receiver operating characteristic) curve; PPV ¼ positive predictive value; NPV ¼ negativepredictive value thought to protect the embryo from destruction by NK cells significantly higher level of activated CD56dimCD16þCD69þ (Ellis et al., 1989; Kovats et al., 1990; King et al., 1997). Ho NK cells in the peripheral blood. This was evident despite et al. (1996) showed that NK cell cytotoxicity is decreased in the fact that there were no significant differences in patients’ a normal healthy pregnancy compared with an anembryonic demographic details, number of previous failed IVF attempts pregnancy. He suggested that activated NK cells, with CD69 or miscarriage, ovarian stimulation outcome, embryo quality expression on their cell surface, play an important role in the or number of embryos transferred. This appears to be the first control of trophoblast growth and placental development. In study to reveal elevated CD56dimCD16þCD69þ peripheral support of this, in vitro models show that activated CD69 blood NK cells in women who experience failed IVF treat- positive NK cells are capable of lysing trophoblasts (Helige ment. It has however been reported that women with inferti- lity needing IVF treatment and women who experience In this study, we explored the relationship between the recurrent spontaneous miscarriage have significantly higher IVF treatment outcome with the quantification of activation receptor (CD69) and inhibitory receptor (CD94) on peri- (Ntrivalas et al., 2001; Coulam et al., 2003). The mechanism pheral blood NK cells. Our results revealed that women who of implantation and the precise role of NK cells in implan- failed to achieve a pregnancy after IVF treatment have a tation are still not fully understood, but it can be speculated Table IV. Patients’ demographics, stimulation characteristic, treatment and pregnancy outcome in group A and B Duration of infertility (years) (mean ^ SD) No. of previous failed IVF attempts (mean ^ SD) No. of available embryos for transfer (mean ^ SD) a Anovulation and endometriosis.
b Mean amount of gonadotrophin used for stimulation (recombinant FSH, hMG or urinary FSH).
NA ¼ not applicable; NS ¼ difference not statistically significant (P . 0.05).
CD691 peripheral NK cells and IVF outcome that an excess of activated CD69þ NK cells might play a women with spontaneous recurrent miscarriages. However, negative role in successful implantation.
we cannot exclude the possibility that previous miscarriage We further evaluated CD56brightCD69þ cells, a sub- and failed IVF treatment may be the cause of the elevated population of NK cells, which are phenotypically similar to CD69þ NK cell count, nor that the selected threshold value uterine NK cells. The level of peripheral CD56brightCD69þ chosen may, by chance, divide the patients into two groups NK cells was not significantly different between the pregnant with very different clinical prognostic factors that determine and non-pregnant groups. In contrast, Kodama et al. (1998) the poorer outcome in the group with high CD69þ NK cell reported that increased numbers of CD56brightCD69þ NK count. This needs to be examined further in a new prospec- cells were found in the decidua in women with spontaneous tive study using logistic regression analysis adjusting for the miscarriage compared with normal pregnancies. This finding may suggest that although phenotypically similar, uterine In conclusion, our data suggest that an elevated level of CD56bright NK cells may not be related to peripheral blood CD56dimCD16þCD69þ peripheral blood NK cells is associ- CD56bright NK cells. Our study demonstrates that there is no ated with a reduced implantation rate of embryos in IVF association between the absolute count of CD56dimCD16þ treatment. Those women with an elevated peripheral blood CD94þ NK cells and CD56brightCD94þ NK cells with the CD56dimCD16þCD69þ NK cell count who achieve a preg- outcome of IVF treatment. However, further functional cyto- nancy after IVF manifest a significantly higher miscarriage rate. A high level of peripheral blood CD69þ NK cells may expression and cytotoxicity of NK cells. In contrast to a pre- play a negative role in IVF treatment outcome but this has to vious study by Fukui et al. (1999), the present study did not be explored in a new prospective study.
demonstrate any statistical difference in the absolute count ofCD56dimCD16þ NK cells and CD56brightCD16þ NK cellsbetween successful and failed IVF treatment. However, for statistical analysis these authors used percentage of CD16þ Avril T, Iochmann S, Brand D, Bardos P, Watier H and Thibault G (2003) NK cells instead of the absolute count, which may not be Human choriocarcinoma cell resistance to natural killer lysis due to defec- accurate, since the percentage can vary depending on the tive triggering of natural killer cells. Biol Reprod 69,627 – 633.
Beer AE, Kwak JYH and Ruiz JE (1996) Immunophenotypic profiles of per- composition of other lymphocytes in a sample.
ipheral blood lymphocytes in women with recurrent pregnancy losses and Further statistical analysis was carried out based on the infertile women with multiple failed in vitro fertilisation cycles. Am J absolute count of CD56dimCD16þCD69þ NK cells above and Borrego F, Robertson MJ, Ritz J, Pena J and Solana R (1999) CD69 is a below 1.0 £ 106/l in order to examine other variables. The stimulatory receptor for natural killer cell and its cytotoxic effect is value of 1.0 £ 106/l was selected for analysis after ROC test- blocked by CD94 inhibitory receptor. Immunology 97,159 – 165.
ing. This study is not intended to establish a threshold level Chao KH, Wu MY, Chen CD, Yang JH, Yang YS and Ho HN (1999) The for the CD69 NK cell testing procedure. Estimating a expression of killer cell inhibitory receptor on natural killer cells and acti- vation status of CD4þ and CD8þ T cell in the decidua of normal and threshold level and testing it in the same sample will create abnormal early pregnancy. Hum Immunol 60,791– 797.
bias. The selected threshold value should be tested and its Cooper MA, Fehniger TA and Caligiuri MA (2001) The biology of human value confirmed or rejected in a separate study with a new natural killer-cell subsets. Trends Immunol 22,633 – 640.
set of samples. The result of the ROC analysis revealed that Coulam C and Roussev R (2003) Correlation of NK cell activation and inhi- bition markers with NK cytotoxicity among women experiencing immuno- the sensitivity and negative predictive value were rather low, logical implantation failure after in vitro fertilization and embryo transfer.
perhaps because there are numerous factors affecting IVF outcome. For example, a woman with a low CD69 count Craston R, Koh M, McDermott A, Ray N, Prentice HG and Lowdell MW (1997) Temporal dynamics of CD69 expression on lymphoid cells.
may not achieve a pregnancy because her embryos are De Maria R, Cifone MG, Trotta R, Rippo MR, Festuccia C, Santoni A and Testi R (1994) Triggering of human monocyte activation through CD69, amember of the natural killer cell gene complex family of signal transdu- CD56dimCD16þD69þ NK cells . 1.0 £ 106/l, the implan- cing receptors. J Exp Med 180,1999 – 2004.
tation rate and the pregnancy rate were significantly lower Ellis SA, Strachan T, Palmer MS and McMichael AJ (1989) Complete and the miscarriage rate was significantly higher as compared nucleotide sequence of a unique HLA class I C locus product expressedon the human choriocarcinoma cell line BeWo. J Immunol 142, to women with CD56dimCD16þD69þ NK cells below this level. The combination of lower implantation rate and higher Emmer PM, Nelen W, Steegers E, Hendriks J, Veerhoek M and Joosten I miscarriage rate result in a significant lower live birth rate.
(2000) Peripheral natural killer cytotoxicity and CD56posCD16pos cells With this finding, we speculated that elevated peripheral acti- increase during early pregnancy in women with a history of recurrentspontaneous abortion. Hum Reprod 15,1163 – 1169.
vated NK cells may play a detrimental role in IVF treatment Fukui A, Fujii S, Yamaguchi E, Kimura H, Sato S and Saito Y (1999) Natu- outcome. The data also reveal that women with elevated peri- ral killer cell subpopulations and cytotoxicity for infertile patients under- pheral CD56dimCD16þD69þ NK cells had a significantly going in vitro fertilisation. Am J Reprod Immunol 41,413– 422.
Helige C, Hagendorfer G, Smolle J and Dohr G (2001) Uterine natural killer higher number of previous miscarriages and previous failed cells in a three-dimensional tissue culture model to study trophoblast inva- IVF treatments despite the fact that there were no statistically significant demographic differences between pregnant and Ho HN, Chao KH, Chen CK, Yang YS and Huang SC (1996) Activation sta- non-pregnant groups (Table I). This finding is in accord with tus of T, NK cells in the endometrium through menstrual cycle and normaland abnormal early pregnancy. Hum Immunol 49,130 – 136.
previous studies (Emmer et al., 2000; Ntrivalas et al., 2001) Moffett-King AM (2002) Natural killer cells and pregnancy. Nat Rev showing that levels of activated NK cells were raised in King A, Jokhi PP, Burrows TD, Gardner L, Sharkey AM and Loke YW Ntrivalas EI, Kwak-Kim JYH, Gilman-Sachs A, Chung-Bang H, Ng SC, (1996) Functions of human decidual NK cells. Am J Reprod Immunol 35, Beaman KD, Mantouvalos HP and Beer AE (2001) Status of peripheral blood natural killer cells in women with recurrent spontaneous abortions King A, Hiby SE, Vema S, Burrows TD, Gardner L, Sharkey and Loke YW and infertility of unknown aetiology. Hum Reprod 16,855– 861.
(1997) Uterine NK cells and trophoblast HLA class I molecules. Am J Pisegna S, Zingoni A, Pirozzi G, Cinque B, Cifone MG, Morrone S, Piccoli M, Frati L, Palmieri G and Santoni A (2002) Src-dependent Sky Kodama T, Hara T, Okamoto E, Kusunoki Y and Ohama Y (1998) Charac- activation controls CD69-mediated signaling and function on human NK teristic changes of large granular lymphocytes that strongly express CD56 in endometrium during the menstrual cycle and early pregnancy. Hum Ravetch JV and Perussia BV (1989) Alternative membrane forms of FcgRIII (CD16) on human natural killer cells and neutrophils cell type- Kovats S, Main EK, Librach C, Stubblebine M, Fisher SJ and De Mars R specific expression of two genes that differ in single nucleotide substi- (1990) A class I antigen, HLA-G, expressed in human trophoblast. Science Robertson MJ and Ritz J (1990) Biology and clinical relevance of human Lanier LL, Buck DW, Rhodes L, Ding A, Evans E, Barney C and Philip J natural killer cells. Blood 76,2421 – 2438.
(1988) Interleukin-2 activation of natural killer cells rapirly induces the Yokoyama WM (1999) Natural Killer Cells. In Paul WE (ed) Fundamental expression and phosphorylation of the Leu 23 activation antigen. J Exp Immunology, 4th edn. Lippincott-Raven, New York, pp. 575 – 603, Llera AS, Viedma F, Sanchez-Madrid F and Tormo J (2001) Crystal struc- Ziegler SF, Ramsdell F and Hjerrild KA (1993) Molecular characterisation ture of the C-type lectin-like domain from the human hematopoietic cell of the early activation antigen CD69: a type II membrane glycoproteinrelated to a family of natural killer cell activation antigens. Eur J Immunol receptor CD69. J Biol Chem 276,7312 – 7319.
Lopez-Botet M, Perez-Villar JJ, Carretero M, Rodriguez A, Melero I, Bellon Zingoni A, Palmieri G, Morrone S, Carretero M, Lopez-Botel M, Piccoli M, T, Llano M and Navarro F (1997) Structure and function of the CD94 Frati L and Santoni A (2000) CD69-triggered ERK activation and func- C-type lectin receptor complex involved in recognition of HLA class I tions are negatively regulated by CD94/NKG2-A inhibitory receptor. Eur J molecules. Immunol Rev 155,165 – 174.
Marzio R, Mauel J and Betz-Corradin S (1999) CD69 and regulation of the immune function Immunopharmacol. Immunotoxicol 21,565– 582.
Submitted on December 4, 2003; accepted on May 24, 2004

Source: http://www.ri-centre.co.uk/uploads/files/pdf/7_Hum_Reprod-2004-Thum-2395-4001.pdf