Post on 21-Jul-2016
Changes in Endometrial Natural Killer Cell Expression of CD94,CD158a and CD158b are Associated with InfertilityEmma McGrath1,2,*, Elizabeth J. Ryan1,2,*, Lydia Lynch1, Lucy Golden-Mason1, Eoghan Mooney3, MaeveEogan4, Colm O’Herlihy4, Cliona O’Farrelly1,2
1Education & Research Centre, St. Vincent’s University Hospital Dublin, Dublin 4, Ireland;2Comparative Immunology Group, School of Biochemistry and Immunology, Trinity College Dublin, Dublin, Ireland;3Department of Pathology and Laboratory Medicine, National Maternity Hospital, Dublin 2, Ireland;4Department of Obstetrics and Gynaecology, University College Dublin, National Maternity Hospital, Dublin 2, Ireland
Introduction
Natural killer (NK) cells are the predominant lym-
phocyte population in the endometrium at the time
of implantation and during early pregnancy.1–3
These cells are believed to play a central role in
maintaining maternal tolerance of the semi-allogenic
foetus4 and elevated NK cell activity has been
implicated in pregnancy failure and infertility.5–7
The hypothesis that local endometrial NK cells acting
in an unregulated way could lead to infertility has
stimulated interest in the characterization of the
local endometrial NK cell population.
NK cells not only have a primary role in innate
immune responses against viruses and transformed
cells but also influence adaptive immunity. They are
Keywords
CD158a, CD158b, CD94, infertility, NK cell,
uterine
Correspondence
Cliona O’Farrelly, School of Biochemistry and
Immunology, Trinity College Dublin, Dublin 2,
Ireland.
E-mail: cliona.ofarrelly@tcd.ie
*These authors contributed equally to this
study.
Submitted March 5, 2008;
accepted January 16, 2009.
Citation
McGrath E, Ryan EJ, Lynch L, Golden-Mason L,
Mooney E, Eogan M, O’Herlihy C, O’Farrelly C.
Changes in endometrial natural killer cell
expression of CD94, CD158a and CD158b are
associated with infertility. Am J Reprod
Immunol 2009; 61: 265–276
doi:10.1111/j.1600-0897.2009.00688.x
Problem
Cycle-dependent fluctuations in natural killer (NK) cell populations in
endometrium and circulation may differ, contributing to unexplained
infertility.
Method of study
NK cell phenotypes were determined by flow cytometry in endometrial
biopsies and matched blood samples.
Results
While circulating and endometrial T cell populations remained constant
throughout the menstrual cycle in fertile and infertile women, circulat-
ing NK cells in infertile women increased during the secretory phase.
However, increased expression of CD94, CD158b (secretory phase),
and CD158a (proliferative phase) by endometrial NK cells from infertile
women was observed. These changes were not reflected in the
circulation.
Conclusion
In infertile women, changes in circulating NK cell percentages are found
exclusively during the secretory phase and not in endometrium; cycle-
related changes in NK receptor expression are observed only in infertile
endometrium. While having exciting implications for understanding NK
cell function in fertility, our data emphasize the difficulty in attaching
diagnostic or prognostic significance to NK cell analyses in individual
patients.
ORIGINAL ARTICLE
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
Journal compilation ª 2009 John Wiley & Sons A ⁄ S 265
cytolytic towards cells which lack expression of
major histocompatibility complex (MHC) class I mol-
ecules8 and are major producers of cytokines and
chemokines when activated.9–11 The functional plei-
otropy of NK cells reflects their phenotypic heteroge-
neity. NK cell subpopulations are identified by
differential expression of a range of NK receptors
(NKRs). These include CD56 which is an isoform of
the adhesion molecule NCAM (neural cell adhesion
molecule), the Fcc receptor CD16, c-type lectin
molecules, CD94 and CD161 and receptors with
immunoglobulin-like domains, the KIR (Killer cell
Immunoglobulin-like Receptor family of molecules)
family. Some correlation between phenotype and
function has been demonstrated, for example circu-
lating NK cells are primarily CD56dim CD16+ and are
potent cytotoxic cells, whereas decidual NK cells are
mostly CD56bright CD16) NK cells and are less cyto-
toxic than those of the periphery, but are believed to
play a role in tissue remodelling.12
NKRs control NK cell activity, by balancing signals
between activatory and inhibitory receptors. This bal-
ance is thought to be critical in preventing maternal
rejection of the fetus.13 Decidual NK cells express
NKRs belonging to both the KIR family (CD158a,
CD158b) and c-type lectin family (CD94, CD161)
and these molecules are thought to play an impor-
tant role in early pregnancy,14,15 although this fact
remains controversial.16 KIRs with long cytoplasmic
tails are inhibitory and contain an immuno-receptor
tyrosine-based inhibitory motif (ITIM) sequence.
KIRs with an ITIM sequence, including CD158a
(KIR2DL1), CD158b (KIR2DL2 ⁄ L3) and CD158e
(KIR3DL1), are inhibited from lysing target cells that
express MHC class I molecules that bind to that par-
ticular KIR.17 Depending on the NKG2 molecule with
which CD94 dimerizes, either an inhibitory or activa-
tory receptor is formed. NKG2A contains an ITIM in
its cytoplasmic region, which inhibits NK cell cyto-
toxicity.18 CD161 (NKR-P1A) is a c-type lectin recep-
tor expressed by NK cells that plays an important
role in the regulation of NK cytotoxicity.19 The role
of NKRs in maintaining a fine balance in the regula-
tion of NK cytotoxicity may be a key factor in suc-
cessful pregnancy. At the time of implantation,
CD56+ NK cells comprise 70–80% of the total endo-
metrial leukocyte population. Although the target
remains elusive, a correlation between NK cell cyto-
toxicity and infertility has been reported.20
CD56bright NK cells, the primary NK population in
endometrium, are less cytotoxic than CD56dim NK
cells and produce higher levels of cytokines.12 The
influence of cytokines on pregnancy maintenance
has been controversial with recent suggestions that a
balance favoring Th2 cytokines favors successful
pregnancy.21 Although there are conflicting data in
that IFN-c, a Th1 cytokine inhibited trophoblast
growth and induced apoptosis in vitro,22,23 this cyto-
kine was also shown to be necessary for foetal
implantation in mice.24
NK cell distribution and activation at the time of
fertilization and implantation are likely to play a
critical role in pregnancy outcome. However, unlike
decidual NK cells, NK cell populations from non-
pregnant endometrium have not been studied in
detail and, particularly, in relation to fertility. The
aim of this study was to examine endometrial lym-
phocyte subpopulations during the menstrual cycle
and to compare these cells in fertile women and
women with unexplained infertility.
Materials and methods
Subjects
Contemporaneous peripheral blood samples and
endometrial tissue biopsies (obtained by sharp curet-
tage) were collected from 28 pre-menopausal
women admitted for operative procedures under
general anaesthesia at the National Maternity Hospi-
tal, Dublin. The study was approved by the Hospi-
tal’s Ethics Committee and written informed consent
was obtained from all participants. Women in the
control group (n = 10, for the determination of T
and NK cell populations, n = 9 for the analysis of
NKRs) were previously parous and were attending
for a variety of procedures, mainly tubal ligation.
Women in the infertility group (n = 18 for determi-
nation of T and NK cell populations, four of whom
were omitted for analysis of NKRs) had normal ovu-
latory hormone profiles, normal pelvic ultrasound
and all semen analyses were reported as being
within normal limits. Women were recruited into
the study at the point of being admitted for tubal
patency assessment via laparoscopy and hydrotuba-
tion. Four patients had minimal endometriosis diag-
nosed and treated at laparoscopy. All but one of the
women had patent tubes confirmed at laparoscopy
and that patient subsequently had patent tubes con-
firmed at hysterosalpingogram, suggesting tubal
spasm at the time of laparoscopy. Any women with
factors that could interfere with normal endometrial
MCGRATH ET AL.
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
266 Journal compilation ª 2009 John Wiley & Sons A ⁄ S
physiology, such as endometrial hyperplasia, severe
endometriosis and women with recurrent miscar-
riages were excluded. None of the women were
undergoing IVF at the time of the study.
The stage of menstrual cycle was determined for
each subject by menstrual data (all women had reg-
ular menstrual cycles) and this was confirmed by
histological assessment.25
For analysis of peripheral blood lymphocyte popu-
lations, eight of the fertile endometrium samples (4
proliferative, 4 secretory) had a matched blood sam-
ple, while 15 of the infertile samples had a matched
blood sample. Of the infertile blood samples, 13
were used for NKR analysis (6 proliferative, 7 secre-
tory). Of the endometrial samples used for NKR
analysis, five fertile endometrium samples were in
the proliferative phase and four were in the secre-
tory phase, while six infertile endometrium samples
were in the proliferative phase and eight were in the
secretory phase.
Preparation of Single Cell Suspensions
Biopsies were collected in Hanks’ Balanced Salt
Solution (HBSS; GIBCO BRL, Paisley, UK) supple-
mented with antibiotics (100 U ⁄ mL penicillin,
100 mg ⁄ mL streptomycin; GIBCO BRL) and 5% fetal
bovine serum (FBS; GIBCO BRL), transported to the
laboratory, drained and weighed. Endometrial single
cell suspensions were prepared as previously
described.26 In brief, the tissue was washed thor-
oughly with HBSS to remove residual blood and was
then minced finely with opposing scalpels and
rotated for 20 min at 37�C in 5 mL of enzyme solu-
tion consisting of Roswell Park Memorial Institute-
1640 (RPMI-1640), 20 mm Hepes buffer (GIBCO
BRL), 1% FBS, 1% bovine serum albumin (BSA;
GIBCO BRL), 200 U ⁄ mL collagenase IV (Worthing-
ton Biochemicals, NJ, USA) and 35 U ⁄ mL DNase I
(Sigma, Poole, UK). In general, 5 mL of enzyme
solution were required per 1 g (or less) of tissue. The
suspension was then passed through 30 lm gauze to
remove tissue fragments and washed twice with
HBSS at 400 g for 10 min and resuspended in 1 mL
of complete RPMI-1640. Cell yields and viability
were assessed using ethydium bromide ⁄ acridine
orange staining. The cell suspension was adjusted to
1 · 106 cells ⁄ mL in RPMI-1640.
Peripheral blood mononuclear cells (PBMCs) were
prepared by standard density gradient centrifugation
over Lymphoprep (Nycomed, Roskilde, Denmark) at
400 · g for 25 min. The cells were then washed
twice with HBSS supplemented with Hepes buffer
solution (GIBCO BRL) and antibiotics. Cell pellets
were resuspended in 1 mL of RPMI-1640 medium
and cell yields were assessed as described before and
adjusted to 1 · 106 cells ⁄ mL in RPMI-1640 medium.
Cell Surface Staining of Cells for Flow Cytometric
Analysis
100 lL aliquots of both PBMCs and endometrial
mononuclear cells (1 · 105 cells) were labelled with
monoclonal antibodies (mAbs) directed against cell
surface markers classically associated with NK cells
(listed below). Optimized concentrations of mAbs
(0.3 lg ⁄ mL) were added to cells, which were then
incubated in the dark at 4�C for 30 min. Cells were
then washed twice with 1 mL PBS ⁄ 1% BSA ⁄ 0.05%
Sodium Azide and fixed in 1% paraformaldehyde
(Sigma Aldrich).
Flow Cytometric Analysis
For the detection of T (CD3+, CD56)) and NK
(CD3), CD56+) cells, anti-CD56 phycoerythrin (PE)
(clone MY31) and anti-CD3 peridin chlorophyll pro-
tein (PerCP) (clone SK7) (BD Biosciences, Erem-
bodegem, Belgium) were used. In addition, the
following mAbs were used to determine the pheno-
type of both pNK and endometrial NK cells: anti-
CD158a fluorescein isothiocyanate (FITC) (clone
HP-3E4), anti-CD158b FITC (clone CH-L), anti-
CD158e FITC (clone DX9), anti-CD161 FITC (clone
DX12), anti-CD94 FITC (clone HP-3D9), anti-CD16
FITC (clone NKP15); anti-NKG2A PE (Beckman
Coulter, Fullerton, CA, USA) was used in combina-
tion with CD56-FITC (clone NCAM16.2) (all, except
anti-NKG2A-PE, from BD Biosciences). Isotype
matched FITC, PE and PerCP conjugated controls
were used to exclude non-specific binding. Cells
were stored in the dark at 4�C and acquisition was
performed within 24 hr of staining. Cells were
analysed using a FACScan flow cytometer and Cell-
Quest� software (BD Biosciences). Lymphocytes
were gated by their density and granularity using
forward scatter and side scatter parameters. A total
of 10,000 events in the lymphocyte gate were
acquired for each sample. All samples included in
this study yielded sufficient cells for analysis. A dot
plot of side scatter against CD56 was used to confirm
that all CD56+ cells were in the lymphogate. The
ENDOMETRIAL NK CELLS AND UNEXPLAINED INFERTILITY
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
Journal compilation ª 2009 John Wiley & Sons A ⁄ S 267
proportions of T cells (CD3+CD56)) and NK cells
(CD3+CD56)) as a percentage of lymphocytes were
determined in each sample. In addition, the percent-
age of NK cells expressing each of the NKRs was
determined.
All results are expressed as median (range). Statis-
tical differences between groups were assessed using
the Mann–Whitney U-test (Statview for Windows,
Version 5.0.1; SAS Institute Inc., NC, USA). P < 0.05
was considered significant.
Results
Subjects
Twenty eight pre-menopausal women with an aver-
age age of 34 years (range 29–43) were recruited for
the study. The average age of the group of women
with a diagnosis of unexplained infertility was not
significantly different from that of the parous
women.
Circulating and Endometrial T and NK cell
Populations of Fertile and Infertile Women
We compared the percentage of T (CD3+CD56)) and
NK (CD3)CD56+) cells in the lymphocyte population
of the non-pregnant endometrium and peripheral
blood of fertile and infertile women. When the
phase of the menstrual cycle was ignored, there was
no significant difference in the percentage of these
populations of lymphocytes between infertile and
fertile women in either the endometrium (Fig. 1a)
or circulation (Fig. 1b). After gating on the lymphoid
population, the percentages of T and NK cells in the
endometrium from fertile women were 9.4% (range
2.0–54.4%) and 6.8% (range 2.5–12.8%) respec-
tively, while in infertile endometrial samples, the
percentage of T and NK cells was 6.2% (range
0.03–53.5%) and 4.7% (range 0.17–14.1%) of lym-
phocytes respectively (Fig. 1a). Likewise, there were
no significant differences in the percentages of T and
NK cell populations in peripheral blood from fertile
women 65.8% (range 58.5–76.0%) and 14.2%
(range 4.9–20.9%) of lymphocytes, respectively
when compared with infertile women; 64.3% (range
15–70.9%) and 13.5% (range 8.2–19.6%; Fig. 1b).
These data, when taken with the observation that
there was no difference in NK cell number from
fertile or infertile specimens (Table I), suggest no sig-
nificant change in the distribution of lymphocytes in
the periphery or endometrium in these infertile
women. However, minor trends may be masked by
the large inter-individual variation in percentages of
T and NK cells in the lymphocyte gate of both blood
and endometrial samples for both the fertile and
infertile groups of women (Fig. 1). Variation in NK
cell number between individual samples is shown in
Table I.
Cycle-related Fluctuations in T and NK cell
Populations
T cells (CD3+CD56)) in peripheral blood remained
constant throughout the menstrual cycle in both fer-
tile and infertile women (Fig. 2b). However, during
the secretory phase of the menstrual cycle, infertile
women had a significantly higher percentage of cir-
culating NK cells (16.9%, range 8.2–19.6% of lym-
phocytes) compared with fertile women (7.4%,
Endometrium
% ly
mph
ocyt
es
Blood
T cells T cells
NK cells NK cells
InfertileFertile
6050
40
3020
10
0
–10
1614121086420
–2
22201816141210864
8070
6050
40
30
20–10
(b)(a)
Fig. 1 T and natural killer (NK) cell numbers in (a) endometrium and
(b) blood of fertile and infertile women. (a) The percentages of T and
NK cells were determined in single cell suspensions isolated from
endometrial biopsies of both fertile (n = 10) and infertile (n = 18)
women by flow cytometry. (b) The percentages of T and NK cells were
determined in peripheral blood mononuclear cells of fertile (n = 8) and
infertile (n = 15) women. Results are expressed as a percentage of the
cells in the lymphocyte gate as determined by forward and side scat-
ter analysis.
MCGRATH ET AL.
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
268 Journal compilation ª 2009 John Wiley & Sons A ⁄ S
range 4.9–16.8% lymphocytes) in the same phase of
the menstrual cycle (P < 0.05) (Fig. 2b).
NK Cell Receptor Expression
While we found no difference in the percentages of
endometrial NK cells (CD3)CD56+) of fertile women
when compared with infertile women (Fig. 2a),
expressions of CD158a (P < 0.05), CD158b (P <
0.01) and CD94 (P < 0.05) were all significantly
lower on endometrial NK cells in fertile women
(3.8%, range 0.2–6.7%, 19.5%, range 10.9–36.2%
and 46.3%, range 14.6–79.1% of NK cells respec-
tively) compared with infertile (6.9%, range 0.6–
23.9%, 32.5%, range 20.3–56.7% and 73.3%, range
38.7–88.9% of NK cells respectively) (Fig. 3a). We
found no difference in the expression of CD158e,
CD16, CD161 or NKG2A on endometrial NK cells
isolated from fertile or infertile women (data not
shown). Expression of CD158a, CD158b, CD158e,
CD161, CD94, CD16 and NKG2A on peripheral
blood NK cells did not differ significantly between
fertile and infertile women irrespective of stage of
cycle (Fig. 3b; data not shown). These results indi-
cate that the NKRs, CD158a, CD158b and CD94 are
differentially expressed by endometrial NK cells of
infertile women when compared to fertile women.
Cycle-related Fluctuations in NK Cell Receptor
Expression
No significant fluctuations in expression of CD16,
CD161, CD158e and NKG2A on endometrial NK
cells throughout the menstrual cycle in either fertile
or infertile endometrium were observed (data not
shown). However, NK cells isolated from endome-
trial samples taken during the proliferative phase of
the menstrual cycle from infertile women showed
significantly (P < 0.05) higher expression of CD158a
Table I Absolute Numbers of NK Cells Obtained from
Endometrium or Matched Blood Samples
No. of NK cells ⁄ g tissue
Fertile
proliferative
aInfertile
proliferative
Fertile
secretory
cInfertile
secretory
69269 8570 4209 3196
3280 15972 12952 68643
48383 283082 14674 42188
64429 3092 52256 8136
5674
7939
No. of NK cells ⁄ ml blood
Fertile
proliferative
bInfertile
proliferative
Fertile
secretory
Infertile
secretory
7932 3668 3250 474
8542 7562 8278 3070
14894 8044 4482 4360
9618 3776 8154 6950
13808
11002
16904
Data from 4 ⁄ 6 endometrium samples in the proliferative phasea
and 4 ⁄ 6 matched blood samplesb from infertile women were
available.cData from 6 ⁄ 7 endometrium samples were available from infer-
tile women in the secretory phase.
*******
Endometrium(a) (b)
% ly
mph
ocyt
es
Blood
T cells
NK cells
Fertile, Proliferative Fertile, Secretory
Infertile, Proliferative Infertile, Secretory
Pro Sec
Pro
Pro Sec
Sec Pro Sec
60
50
40
30
20
10
0
25
20
15
10
5
0
80
70
60
50
40
30
20
10
0
16
14
12
10
8
6
4
2
0
Fig. 2 Cycle-related fluctuations in lymphocyte populations of (a)
endometrium and (b) peripheral blood. The percentages of T and natu-
ral killer cells in the (a) endometrial lymphocyte population and (b)
peripheral blood mononuclear cells of fertile women were determined
by flow cytometry. Results are expressed as a percentage of the cells
in the lymphocyte gate as determined by forward and side scatter
analysis. Individual results are displayed as circles and each bar repre-
sents the median of that group. *P < 0.05, by Mann–Whitney U-test.
ENDOMETRIAL NK CELLS AND UNEXPLAINED INFERTILITY
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
Journal compilation ª 2009 John Wiley & Sons A ⁄ S 269
(5.6%, range 0.8–10.9% of NK cells) compared with
that in fertile endometrium (2.1%, range 0.6–4.2%
of NK cells) (Fig. 4a). In addition, CD158b expres-
sion on NK cells was significantly higher in infertile
endometrium (37.0%, range 24.7–56.7% of NK
cells) in the secretory phase of the cycle compared
with that of fertile endometrium (21.4%, range
10.9–36.2% of NK cells) during the same phase of
the cycle (P < 0.05) (Fig. 4b). Expression of CD94
was significantly higher on NK cells in infertile
endometrium (72.2%, range 38.7–88.9% of NK
cells) in the secretory phase of the cycle compared
with that of fertile women (48.8%, range 29.7–
67.7% of NK cells, P < 0.05) (Fig. 4c).
The expression of CD158a (Fig. 4a) in the prolif-
erative phase and CD158b (Fig. 4b) and CD94
(Fig. 4c) in the secretory phase on CD3)CD56+ NK
cells were higher in infertile endometrium than in
fertile endometrium. These trends were also appar-
ent in the secretory phase regarding CD158a and in
the proliferative phase regarding CD158b and CD94.
However, these differences were not statistically sig-
nificant due to the large degree of variation between
individuals (Fig. 4).
We also assessed the expression of NKG2A on
CD3)CD56+ NK cells and found that there was no
significant difference in NKG2A expression between
fertile and infertile endometrium during the men-
strual cycle (data not shown).
Expression of the NKRs CD158a, CD158b,
CD158e, CD161, CD94, CD16 and NKG2A by
peripheral blood NK cells did not differ between fer-
tile and infertile women when the stage of cycle was
ignored (Fig. 3b; data not shown). Neither were
there any cycle-related changes in NKR expression
by peripheral blood NK cells (Fig. 5a-c and data not
shown).
CD56bright and CD56dim Populations
The relative proportions of CD56bright (�65%) and
CD56dim (�35%) NK cells in the endometrium dif-
fered from peripheral blood where �10% of NK cells
were CD56bright and �90% of NK cells CD56dim
(Figs 4, 5; data not shown). However, these relative
proportions did not change with the cycle nor was
there a significant difference between these cell pop-
ulations when comparing fertile and infertile endo-
metrium. The expression levels of CD158a, CD158b
and CD94 on the CD56bright and CD56dim endome-
trial NK cell populations in fertile and infertile
women in both the proliferative and secretory phase
of the menstrual cycle are summarized in Table II.
In infertile women, CD56dim endometrial NK cells
express significantly lower levels of CD158a
(P < 0.01), CD158b (P < 0.05) and CD94 (P < 0.05)
in the secretory phase compared with the expression
of these receptors on CD56dim endometrial cells in
the proliferative phase of the menstrual cycle. Fur-
thermore, we found that the expression of CD158b
was significantly higher on CD56bright endometrial
NK cells of infertile women in the proliferative phase
of the menstrual cycle compared to the CD56bright
cells of fertile women (P < 0.05, Table II). Whereas,
the expression of CD158a (P < 0.05), CD158b
(P < 0.05) and CD94 (P < 0.05) by CD56dim endo-
metrial NK cells of infertile women is significantly
lower than fertile controls, also in the proliferative
phase of the menstrual cycle. This analysis reveals
that the overall increase in NKRs observed in
InfertileFertile
% N
K C
ells
*
Blood
CD158a CD158b CD94
80
–100
20
40
60
100
Endometrium(a)
(b)
CD158a CD158b CD94
80
–100
20
40
60
100
*
**
Fig. 3 Expression of CD158a, CD158b and CD94 in (a) the endome-
trium and (b) the peripheral blood of fertile and infertile women, irre-
spective of stage of cycle. The percentage of NK cells (CD3)CD56+)
expressing CD158a, CD158b and CD94 was determined by multi-
parameter flow cytometry (a) in single cell suspensions isolated from
endometrial biopsies of fertile and infertile women and (b) peripheral
blood mononuclear cells were obtained at the same time as the
biopsy. Individual results are displayed as circles and each bar repre-
sents the median of that group. *P < 0.05 and **P < 0.01, respectively
by Mann–Whitney U-test.
MCGRATH ET AL.
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
270 Journal compilation ª 2009 John Wiley & Sons A ⁄ S
infertile endometrium is due specifically to the
increase in expression on the predominant
CD56bright population.
Discussion
To test the hypothesis that local endometrial NK
cells acting in an unregulated way could lead to
infertility, we compared NKR expression by the local
endometrial NK cell population with expression by
peripheral blood NK cells in fertile and infertile
women. Our results clearly demonstrate compart-
mentalized differences in NK cell phenotype in
women with unexplained infertility compared with
fertile women. Specifically, increased expression of
NKRs, CD94, CD158b (secretory phase) and CD158a
(proliferative phase) by endometrial NK cells from
infertile women was observed; however, these
changes were not reflected in the circulation.
While it has been reported that circulating NK cell
numbers and activity can be predictive of pregnancy
outcome,27,28 the usefulness of enumerating circulat-
ing NK cell numbers as a diagnostic tool in fertility
assessment remains controversial.29,30 NK cell num-
ber in the circulation is influenced by a number of
factors including age and stress and there is a wide
normal range. In addition, there is no evidence that
numbers of NK cells in the circulation reflect num-
bers of endometrial NK cells. Furthermore, the pre-
cise role of endometrial NK cells and the relative
contribution of cytokine secretion or cytotoxicity in
implantation and maintenance of a successful preg-
nancy remain elusive.30
We found that there was wide inter-individual
variation in the percentage of NK cells in the circula-
tion of both fertile and infertile women (Fig. 1b). A
statistically significant elevation in the proportion of
circulating NK cells was observed, but only during
the secretory phase of the menstrual cycle in infer-
tile women when compared with fertile women in
the same stage of the menstrual cycle (Fig. 2b).
However, the percentage of NK cells observed in all
CD
158A
CD
158B
CD
94
% N
K c
ells
Fertile Infertile
010203040506070
Proliferative
0
20
40
60
80
100
Proliferative
CD158b
CD94
Proliferative–2
0
1012
2468
010203040506070
Secretory
Secretory0
20
40
60
80
100
Secretory
CD158aCD158a; Proliferative phase
CD158b; SecretoryPhase
CD94; Secretory phase
2% of NK cells
6% of NK cells
21% of NK Cells
38% of NK Cells
48% of NK cells
74% of NK cells
0
10
20
30
40
50
–10
FS
C
SR1
CD56
E
FIT
C
Isotype control
Fertile Infertile
(a)
(b)
(c)
Fig. 4 Cycle-related fluctuations in CD158a,
CD158b and CD94 expression on endometrial
NK cells of fertile and infertile women. The
percentage of NK cells (CD3)CD56+) express-
ing (a) CD158a, (b) CD158b and (c) CD94 was
determined by flow-cytometry in single cell
suspensions isolated from endometrial biop-
sies of fertile (n = 9; 5 proliferative, 4 secre-
tory) and infertile (n = 14; 6 proliferative, 8
secretory) women. Results are expressed as a
percentage of total NK cells, with box-plots
representing the median and range. Represen-
tative dot-plots of each of the statistically sig-
nificant results are shown. *P < 0.05 by
Mann–Whitney U-Test.
ENDOMETRIAL NK CELLS AND UNEXPLAINED INFERTILITY
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
Journal compilation ª 2009 John Wiley & Sons A ⁄ S 271
% N
K c
ells
Fertile Infertile
Proliferative–10
0102030405060
–100
102030405060
Secretory
Proliferative–10
0102030405060
–100
102030405060
Secretory
30405060708090
100
Proliferative
CD158a
CD158b
CD94
30405060708090
100
Secretory
32% of NK cells
34% of NK cells
55% of NK cells
63% of NK cells
CD
158A
CD158a; Proliferative phase
CD
158B
CD158b; Secretory phase
CD
94
CD94; Secretory phase
CD56
SSC
FS
C
R1
FIT
C
PE
Isotype control
3% of NK cells
12% of NK cells
Fertile Infertile
(a)
(b)
(c)
Fig. 5 Expression of CD158a, CD158b and
CD94 on NK cells in peripheral blood of fertile
and infertile women is not cycle-dependant.
The percentage of NK cells (CD3)CD56+)
expressing (a) CD158a, (b) CD158b and (c)
CD94 was determined by multi-parameter flow
cytometry in peripheral blood mononuclear
cells isolated from blood from fertile (n = 8; 4
proliferative, 4 secretory) and infertile (n = 13;
6 proliferative, 7 secretory) women. Results
are expressed as a percentage of total NK
cells, with box-plots representing the median
and range of that group. Representative dot-
plots also illustrate that there was no differ-
ence between fertile and infertile samples.
Table II Expression Levels of CD158a, CD158b and CD94 on CD56bright and CD56dim Endometrial NK Cells of Fertile and Infertile Women in
the Proliferative or Secretory Phase of the Menstrual Cycle. Results Are Expressed as %, Range
CD56bright NK cells
Fertile proliferative aInfertile proliferative Fertile secretory Infertile secretory
CD158a 3.06%, 0.61–9.41 11.13%, 0.44–45.44 9.61%, 3.37–14.49 11.47%, 0.38–35.01
CD158b 25.48%, 12.56–44.06 44.93%, 33.14–79.21 35.5%, 41.9–40.68 46.08%, 4.8–58.73
CD94 83.1%, 72.95–96.02 94.75%, 85.85–97.58 78.15%, 49.05–87.79 87.81%, 75.47–95.54
CD56dim NK cells
Fertile proliferative bInfertile proliferative Fertile secretory cInfertile secretory
CD158a 40.84%, 5.373–58.54 22.32%, 5.83–57.36 13.26%, 7.36–20.23 6.85%, 0–37.61
CD158b 48.77%, 9.14–57.48 24.3%, 11.34–51.25 15.49%, 6.95–21.91 10.91%, 0.41–48.4
CD94 43.7%, 11.38–64.54 33.09%, 17.59–58.49 16.74%, 12.33–32.96 13.2%, 0.19–37.36
aCD56bright expression of CD158b differs significantly between fertile and infertile endometrial NK cells in the proliferative phase of the men-
strual cycle (P < 0.05).bCD56dim expression of CD158a (P < 0.05), CD158b (P < 0.05) and CD94 (P < 0.05) differs significantly between fertile and infertile endome-
trial NK cells in the proliferative phase of the menstrual cycle.cEndometrial CD56dim NK cells express lower levels of CD158a (P < 0.01), CD158b (P < 0.05) and CD94 (P < 0.05) in the secretory phase of
the menstrual cycle compared to the proliferative.
MCGRATH ET AL.
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
272 Journal compilation ª 2009 John Wiley & Sons A ⁄ S
patients was within the normal range (2–29%). In
addition, we found that the difference in percentage
did not translate into a difference in the absolute
number of NK cells in the circulation of fertile and
infertile women (Table I). Previously, we demon-
strated an increase in the number of endometrial NK
cells during the late secretory phase of the menstrual
cycle.31 We did not observe this increase in this
study, as the data were not categorized into early
and late secretory phase, due to the small number of
patient samples. Pooling of data is likely to contrib-
ute to the large inter-individual variations observed,
thus masking minor significant differences in this
cell population during the menstrual cycle. Further-
more, we detected no change in the absolute num-
ber of NK cells or the percentage of NK cells in the
endometrium of fertile women compared to infertile
women in the secretory phase of the menstrual
cycle. These results emphasize the difficulty of
attaching any diagnostic or prognostic significance to
NK cell analyses in individual patients.
Significant changes in NKR expression, however,
were observed on endometrial NK cells from infertile
women. We found that the expression of the NKRs,
CD158a, CD158b and CD94 on endometrial NK cells
of infertile women was consistently higher than that
on endometrial NK cells obtained from fertile
women regardless of stage of cycle (Fig. 3a). This dif-
ference was not observed in circulating NK cells iso-
lated at the same time (Fig. 3b). The majority of
previous studies investigating NKR expression in
women with infertility,32,33 recurrent spontaneous
abortions32 or failed in vitro fertilization5 examined
only circulating NK cells or decidual NK cells.13,14
Ntrivalas et al.32 reported a significant decrease in
the expression of CD158a and CD158b on peripheral
blood NK cells in women with implantation failures
compared to normal controls, a result not replicated
in this study (Fig. 3b). This may be because Ntrivalas
et al. studied women with recurrent spontaneous
abortions, who may have a further expansion of NK
cells than the women with unexplained infertility
who were recruited for this study.
Our data show that no correlation exists between
the expression of NKRs on circulating NK cells and
expression on local endometrial NK cell populations
(Fig. 3). Endometrial NK cells from non-pregnant
women have been less studied. It is possible that
these are separate NK cell populations of different
origins. Evidence is accumulating that there are mul-
tiple sites of NK cell differentiation in the healthy
adult.34 Previously, we demonstrated the presence of
cells with haematopoietic stem cell (HSC) phenotype
in human endometrium, a proportion of which
expressed CD56.35 We therefore suggested that NK
cell populations might differentiate in the uterus
from local haematopoietic progenitor cells.
NK cells represent an important class of lympho-
cyte in the uterine epithelium and are thought to
have a tissue remodelling role during the formation
of the placenta.3 Recently, it has been suggested that
decidual NK cells develop from existing endometrial
NK cells.12 The secretion of chemokines and angio-
genic growth factors by decidual NK cells has been
shown to regulate trophoblast invasion.36 Therefore,
higher expression levels of both CD158a and
CD158b, which exert an inhibitory influence on NK
cells, may affect fertility by suppression of local NK
cell activity (e.g. cytokine and growth factor secre-
tion) thereby affecting placental developmental pro-
cesses including trophoblast invasion and vascular
growth.
The heterodimer CD94 ⁄ NKG2A is a ligand for
HLA-E on foetal trophoblasts. As the interaction of
CD94 ⁄ NKG2A with HLA-E leads to inhibition of NK
cell activity, it is possible that changes in CD94
expression alter cytokine secretion, which may have
critical implications for placental development,
maternal uterine spiral arterial remodelling, antiviral
immune response as well as trophoblast survival.37
Additionally, HLA-G, which has been implicated in
the regulation of trophoblast invasion,38 has been
reported to interact with KIRs and CD94 ⁄ NKG2A;
however, there has been some controversy regarding
this relationship.39 Our data show an increase in
CD94 expression, as well as CD158a and CD158b
expression by NK cells isolated from infertile endo-
metrium, thus possibly altering NK cell ability to reg-
ulate trophoblast invasion.
The percentage of NK cells in the endometrium
apparently increases in the later stages of the secre-
tory phase,40 although this is not obvious from our
data as they were not separated into early and late
secretory phase (Fig. 2a). However, we have shown
that CD158b and CD94 expression on endometrial
NK cells isolated from infertile women during the
secretory phase was significantly higher than in par-
ous women (Fig. 4). In contrast, CD158a expression
on endometrial NK cells was significantly higher in
infertile women compared with parous women in
the proliferative phase of the menstrual cycle.
Although not significantly different due to the degree
ENDOMETRIAL NK CELLS AND UNEXPLAINED INFERTILITY
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
Journal compilation ª 2009 John Wiley & Sons A ⁄ S 273
of variation between individual samples, these trends
were also present in the secretory phase with respect
to CD158a and in the proliferative phase with respect
to CD158b and CD94, thus suggesting a tendency
towards greater expression of these NKRs in the
endometrium of infertile women (Fig. 4). Cycle-
dependent changes in NKR expression by NK cells
from matched peripheral blood samples were not
observed (Fig. 5), emphasizing the compartmentali-
zation of fertility-associated NK cell changes.
There did not appear to be any difference in per-
centage of CD56bright endometrial NK cells between
fertile endometrium and infertile endometrium. A
similar result was recently reported by Matteo et al,
2007.41 While we show an increase in NKR expres-
sion on CD56+ cells in the endometrium of infertile
women (Fig. 4), interestingly, CD56dim endometrial
NK cells expressed significantly reduced levels of
CD158a, CD158b, CD158e, CD16 and CD94 during
the secretory phase in infertile endometrium
(Table II). This was not observed in fertile endome-
trium. Although the CD56dim cell population of
endometrial NK cells is not generally studied in
detail, being the minority population, it is possible
that changes in this cell population during the secre-
tory phase of the menstrual cycle may impart a neg-
ative influence on maintaining fertility.
To conclude, this study shows that in women with
a diagnosis of unexplained infertility following stan-
dard investigation, fluctuations in peripheral blood
NK cell populations are cycle-dependent and are not
reflected in the local environment of the endome-
trium where NK cell expression of the NKRs; CD158a,
CD158b and CD94 is elevated. We observed wide
inter-individual variation in both NK cell number and
receptor expression. Furthermore, our sample size
was small, which may also contribute to masking
some potentially significant findings. A recent study
has elegantly shown that NK cells with differing
expression of KIR function differently in the context
of viral infection.42 However, further studies are
required both to confirm our findings and to establish
whether differential expression of these molecules is
sufficient to influence endometrial NK cell activity to
the extent that fertility is compromised.
Acknowledgments
We wish to thank the women who participated in
this research which was supported by a grant from
Enterprise Ireland.
References
1 Bulmer JN, Morrison L, Longfellow M, Ritson A, Pace
D: Granulated lymphocytes in human endometrium:
histochemical and immunohistochemical studies. Hum
Reprod 1991; 6:791–798.
2 Kodama T, Hara T, Okamoto E, Kusunoki Y, Ohama
K: Characteristic changes of large granular
lymphocytes that strongly express CD56 in
endometrium during the menstrual cycle and early
pregnancy. Hum Reprod 1998; 13:1036–1043.
3 Moffett-King A, Entrican G, Ellis S, Hutchinson J,
Bainbridge D: Natural killer cells and reproduction.
Trends Immunol 2002; 23:332–333.
4 Santoni A, Zingoni A, Cerboni C, Gismondi A:
Natural killer (NK) cells from killers to regulators:
distinct features between peripheral blood and
decidual NK cells. Am J Reprod Immunol 2007; 58:280–
288.
5 Thum MY, Bhaskaran S, Abdalla HI, Ford B, Sumar
N, Shehata H, Bansal AS: An increase in the absolute
count of CD56dimCD16+CD69+ NK cells in the
peripheral blood is associated with a poorer IVF
treatment and pregnancy outcome. Hum Reprod 2004;
19:2395–2400.
6 Vassiliadou N, Bulmer JN: Immunohistochemical
evidence for increased numbers of ‘classic’ CD57+
natural killer cells in the endometrium of women
suffering spontaneous early pregnancy loss. Hum
Reprod 1996; 11:1569–1574.
7 Kwak JY, Beer AE, Kim SH, Mantouvalos HP:
Immunopathology of the implantation site utilizing
monoclonal antibodies to natural killer cells in
women with recurrent pregnancy losses. Am J Reprod
Immunol 1999; 41:91–98.
8 Yokoyama WM, Kim S: How do natural killer cells
find self to achieve tolerance? Immunity 2006; 24:249–
257.
9 Farag SS, Caligiuri MA: Human natural killer cell
development and biology. Blood Rev 2006; 20:123–
137.
10 Colucci F, Caligiuri MA, Di Santo JP: What does it
take to make a natural killer? Nat Rev 2003; 3:413–
425.
11 Robertson MJ: Role of chemokines in the biology of
natural killer cells. J Leuk Biol 2002; 71:173–183.
12 Manaster I, Mandelboim O: The unique properties of
human NK cells in the uterine mucosa. Placenta 2008;
29:60–66.
13 Kusumi M, Yamashita T, Fujii T, Nagamatsu T,
Kozuma S, Taketani Y: Expression patterns of lectin-
like natural killer receptors, inhibitory CD94 ⁄ NKG2A,
and activating CD94 ⁄ NKG2C on decidual CD56bright
MCGRATH ET AL.
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
274 Journal compilation ª 2009 John Wiley & Sons A ⁄ S
natural killer cells differ from those on peripheral
CD56dim natural killer cells. J Reprod Immunol 2006;
70:33–42.
14 Eidukaite A, Siaurys A, Tamosiunas V: Differential
expression of KIR ⁄ NKAT2 and CD94 molecules on
decidual and peripheral blood CD56bright and
CD56dim natural killer cell subsets. Fertil Steril 2004;
81(Suppl. 1):863–868.
15 Yamada H, Shimada S, Kato EH, Morikawa M,
Iwabuchi K, Kishi R, Onoe K, Minakami H:
Decrease in a specific killer cell immunoglobulin-
like receptor on peripheral natural killer cells in
women with recurrent spontaneous abortion of
unexplained etiology. Am J Reprod Immunol 2004;
51:241–247.
16 Witt CS, Goodridge J, Gerbase-Delima MG, Daher S,
Christiansen FT: Maternal KIR repertoire is not
associated with recurrent spontaneous abortion. Hum
Reprod 2004; 19:2653–2657.
17 Moretta A, Moretta L: HLA class I specific inhibitory
receptors. Curr Opin Immunol 1997; 9:694–701.
18 Le Drean E, Vely F, Olcese L, Cambiaggi A, Guia S,
Krystal G, Gervois N, Moretta A, Jotereau F, Vivier E:
Inhibition of antigen-induced T cell response and
antibody-induced NK cell cytotoxicity by NKG2A:
association of NKG2A with SHP-1 and SHP-2 protein-
tyrosine phosphatases. Eur J Immunol 1998; 28:264–
276.
19 Lanier LL, Chang C, Phillips JH: Human NKR-P1A. A
disulfide-linked homodimer of the C-type lectin
superfamily expressed by a subset of NK and T
lymphocytes. J Immunol 1994; 153:2417–2428.
20 Matsubayashi H, Shida M, Kondo A, Suzuki T, Sugi T,
Izumi S, Hosaka T, Makino T: Preconception
peripheral natural killer cell activity as a predictor of
pregnancy outcome in patients with unexplained
infertility. Am J Reprod Immunol 2005; 53:126–131.
21 Chernyshov VP, Tumanova LE, Sudoma IA, Bannikov
VI: Th1 and Th2 in Human IVF Pregnancy with
Allogenic Fetus. Am J Reprod Immunol 2008; 59:352–
358.
22 Ho S, Winkler-Lowen B, Morrish DW, Dakour J, Li H,
Guilbert LJ: The role of Bcl-2 expression in EGF
inhibition of TNF-alpha ⁄ IFN-gamma-induced villous
trophoblast apoptosis. Placenta 1999; 20:423–430.
23 Yui J, Garcia-Lloret M, Wegmann TG, Guilbert LJ:
Cytotoxicity of tumour necrosis factor-alpha and
gamma-interferon against primary human placental
trophoblasts. Placenta 1994; 15:819–835.
24 Ashkar AA, Croy BA: Functions of uterine natural
killer cells are mediated by interferon gamma
production during murine pregnancy. Sem Immunol
2001; 13:235–241.
25 Noyes RW, Herbig AT, Rock J: Dating the endometrial
biopsy. Fertil Steril 1950;1:3–25.
26 Flynn L, Carton J, Byrne B, Kelehan P, O’Herlihy C,
O’Farrelly C: Optimisation of a technique for isolating
lymphocyte subsets from human endometrium.
Immunol Invest 1999; 28:235–246.
27 Beer AE, Kwak JY, Ruiz JE: Immunophenotypic
profiles of peripheral blood lymphocytes in women
with recurrent pregnancy losses and in infertile
women with multiple failed in vitro fertilization
cycles. Am J Reprod Immunol 1996; 35:376–382.
28 Coulam CB, Beaman KD: Reciprocal alteration in
circulating TJ6+ CD19+ and TJ6+ CD56+ leukocytes
in early pregnancy predicts success or miscarriage. Am
J Reprod Immunol 1995; 34:219–224.
29 Rai R, Sacks G, Trew G: Natural killer cells and
reproductive failure–theory, practice and prejudice.
Hum Reprod 2005; 20:1123–1126.
30 Moffett-King A: Natural killer cells and pregnancy.
Nat Rev 2002; 2:656–663.
31 Flynn L, Byrne B, Carton J, Kelehan P, O’Herlihy C,
O’Farrelly C: Menstrual cycle dependent fluctuations
in NK and T-lymphocyte subsets from non-pregnant
human endometrium. Am J Reprod Immunol 2000;
43:209–217.
32 Ntrivalas EI, Bowser CR, Kwak-Kim J, Beaman KD,
Gilman-Sachs A: Expression of killer
immunoglobulin-like receptors on peripheral blood
NK cell subsets of women with recurrent spontaneous
abortions or implantation failures. Am J Reprod
Immunol 2005; 53:215–221.
33 Michou VI, Kanavaros P, Athanassiou V, Chronis GB,
Stabamas S, Tsilivakos V: Fraction of the peripheral
blood concentration of CD56+ ⁄ CD16- ⁄ CD3- cells in
total natural killer cells as an indication of fertility
and infertility. Fertil Steril 2003; 80(Suppl. 2):691–697.
34 Huntington ND, Vosshenrich CA, Di Santo JP:
Developmental pathways that generate natural-killer-
cell diversity in mice and humans. Nat Rev 2007;
7:703–714.
35 Lynch L, Golden-Mason L, Eogan M, O’Herlihy C,
O’Farrelly C: Cells with haematopoietic stem cell
phenotype in adult human endometrium: relevance
to infertility? Hum Reprod 2007; 22:919–926.
36 Hanna J, Goldman-Wohl D, Hamani Y, Avraham I,
Greenfield C, Natanson-Yaron S, Prus D, Cohen-
Daniel L, Arnon TI, Manaster I, Gazit R, Yutkin V,
Benharroch D, Porgador A, Keshet E, Yagel S,
Mandelboim O: Decidual NK cells regulate key
developmental processes at the human fetal-maternal
interface. Nat Med 2006; 12:1065–1074.
37 Rabot M, Tabiasco J, Polgar B, Aguerre-Girr M,
Berrebi A, Bensussan A, Strbo N, Rukavina D, Le
ENDOMETRIAL NK CELLS AND UNEXPLAINED INFERTILITY
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
Journal compilation ª 2009 John Wiley & Sons A ⁄ S 275
Bouteiller P: HLA class I ⁄ NK cell receptor interaction
in early human decidua basalis: possible functional
consequences. Chem Immunol Allergy 2005; 89:72–83.
38 Roussev RG, Coulam CB: HLA-G and its role in
implantation (review). J Assist Reprod Gen 2007;
24:288–295.
39 Hunt JS, Petroff MG, Morales P, Sedlmayr P,
Geraghty DE, Ober C: HLA-G in reproduction: studies
on the maternal-fetal interface. Hum Immunol 2000;
61:1113–1117.
40 Mselle TF, Meadows SK, Eriksson M, Smith JM, Shen
L, Wira CR, Sentman CL: Unique characteristics of
NK cells throughout the human female reproductive
tract. Clin Immunol 2007; 124:69–76.
41 Matteo MG, Greco P, Rosenberg P, Mestice A, Baldini
D, Falagario T, Martino V, Santodirocco M, Massenzio
F, Castellana L, Specchia G, Liso A: Normal
percentage of CD56bright natural killer cells in young
patients with a history of repeated unexplained
implantation failure after in vitro fertilization cycles.
Fertil Steril 2007; 88:990–993.
42 Ahlenstiel G, Martin MP, Gao X, Carrington M,
Rehermann B: Distinct KIR ⁄ HLA compound
genotypes affect the kinetics of human antiviral
natural killer cell responses. J Clin Invest 2008;
118:1017–1026.
MCGRATH ET AL.
American Journal of Reproductive Immunology 61 (2009) 265–276 ª 2009 The Authors
276 Journal compilation ª 2009 John Wiley & Sons A ⁄ S