- Email: [email protected]
Activation status of T and NK cells in the endometrium throughout menstrual cycle and normal and abnormal early pregnancy
ELSEVIER
Activation Status of T and NK Cells in the Endometrium Throughout Menstrual Cycle and Normal and Abnormal Early Pregnancy Hong-Nerng Ho, Kuang-Han Chao, Chun-Kai Yu-Shih Yang, and Su-Cheng Huang
Chen,
ABSTRACT: Endometrial lymphocytes were studied at all stages throughout the menstrual cycle and early pregnancy by flow cytometry to examine different lymphocyte subpopulations and the expression of the T- and NK-cell activation markers. After pregnancy, CDS’CD3’ lymphocytes were decreased in the decidua. In both endometrium and decidua, more T cells expressed CD69, CD71, HLA-DR, and CD38 antigens than in peripheral blood. After pregnancy, CD7 1 ‘CD3’ lymphocytes were further increased. CD25’CDI’ lymphocytes decreased significantly in the endometrium and decidua of ectopic pregnancies, but not in the decidua of normal pregnancies. These findings indicate that T cells are regionally acti-
vated in the first trimester, and it may be the result of the stimulation by fetal antigens. NK cells were the most abundant cell type in the decidua, which expressed the phenotype CD16- CD56’, and CD57-CD56’. The proportion of activated decidual NK cells was increased in anembryonic pregnancies more than in normal pregnancies, although the total NK subpopulation was similar in both groups. This might result in increased NK cytotoxicity in anembryonic pregnancies. In conclusion, T cells are activated, but NK cytotoxicity is decreased in the decidua of early normal pregnancies. This might be important in the control of trophoblast growth and placental development. HuTnan Immunology 49, 130-136 (19961
ABBREVIATIONS FITC fluorescein isothiocyanate LGLs large granular lymphocytes natural killer NK
PBMCs PE
peripheral blood mononuclear phycoerythrin
cells
INTRODUCTION Since the decidua
is the maternal tissue in closest contact with the fetal tissue, it is likely that it contributes to the maintenance of pregnancy by forming a local network supporting cellular and humoral immunity. Immunohistochemical analysis of human decidua in the first trimester of pregnancy identified three principle leukocyte populations: macrophage, classic T lymphocytes, and an unusual lymphocyte population [l, 21, subsequently identified as expressing antigen characteristic of large granular lymphocytes (LGLs) 131. The most abundant cell type in human decidua has
From the Department of Obstetricsand Gynecology, College of Medicine and Hospital, National Taiwan University, Taipei, Taiwan. ROC. Address reprint requeststo: Dr. Hong-Nevng Ho, Department ofObstetrics and Gynecology,National Taiwan University Hospital, No. 7. Chung Shari South Road. 10016, Taipei, Taiwan, ROC. Received February 12. 1996: acceptedApril 23. 1996. Human Immunology 49, 130-136 (1996) 0 Amencan Society for Histocompatibilicy
and Immunogenetics,
1996
been identified with flow cytometry by the expression of its surface antigens, size, and granularity, as LGLs 141. These LGLs comprise 245% of all decidual cells and are mainly CD16_CD56’CD3141, which corresponds to a minor population of natural killer (NK) cells in peripheral blood. In contrast to decidual LGLs, CD16’CD56’CD3NK cells were the most abundant and the most efficient cytotoxic effecters in peripheral blood 151. In a previous study, we found the proportion of NK cells (defined as CD56’CD3or CDlb’CD3-) was significantly higher in the nonpregnant endometriurn than in the peripheral blood 161, and this difference was more apparent in the decidua during pregnancy 171. It was also noted that decidual NK cytotoxicity decreased in normal pregnancy, but not in anembryonic pregnancy or in recurrent spontaneous abortion 17}. Whether the increase in NK cytotoxicity in the cases of spontaneous abortion is due to the increase in the pro0198~8859/961$15,00 PI1 S1098-8859(96)00120-h
Activated T and NK in Endometrium
portion of CDlb’CD56’CD3NK cells needs to be clarified. Our previous report revealed that endometrial T lymphocytes were activated by increasing the expression of early and late T-cell activation markers (CD69 and HLADR) {bl. In 1992, Saito et al. [8} showed decidual T cells expressed CD69, HLA-DR, and CD25 antigen significantly, indicating that the T cells in the decidua were also regionally activated. NK activity could be regulated by lymphokines, and activated T cells had the ability to release lymphokines which could modulate the NK cytotoxicity. In this study, we explored whether the relatively higher levels of NK cell activity in anembryonic pregnancy is the result of activation of T or NK cells. The study was designed to investigate whether there was consistency between the activation of T and NK cells in the nonpregnant endometrium and decidua, and whether there was any difference in the level of activation in normal and abnormal pregnancies.
MATERIALS
AND
METHODS
Szlbjects. Sixty-five patients who received hysterectomies for benign reasons of nonendometrial pathology (mostly myoma uteri), 22 pregnant women who had elective abortions of normal pregnancies due to multiparity, 20 women who had anembryonic pregnancies, and 5 women with ectopic tubal pregnancy between 6 and 10 weeks of gestational age were enrolled in the study, with written informed consent. The pathologic results of the hysterectomy specimens showed no evidence of infection or inflammation in the 35 specimens in the follicular phase and the 30 specimens in the luteal phase. The luteal phase was further divided into early and late stages according to the last menstrual period, levels of progesterone, and the histologic dating. Anembryonic pregnancy was identified by sonography, which failed to show a fetal pole when the gestational sac was >25 mm. Specimens. Endometrial
131
and Decidua
tissues and venous blood obtained from 65 hysterectomy patients were collected on the day of hysterectomy. Decidual tissue and peripheral blood samples were taken from each pregnant woman at the time of abortion and laparoscopic surgery or laparotomy. Endometrial tissue was cut with a surgical knife immediately after hysterectomy, minced with a scalpel, suspended in RPMI-1640 medium, and pressed gently through a 380~pm and then a 45.7-pm sieve, as described by Chen et al. 161. The filtered solution was layered over a Ficoll-Hypaque gradient and centrifuged for 45 min at 400 x g. An enriched cell suspension was collected and then washed twice with RPMI-1640 medium. The recovered cells were checked for viability with trypan blue and counted. In the pregnant cases, to mini-
mize the contamination by blood, the decidual tissue was macroscopially separated from chorionic villi, washed twice with Hank’s balanced salt solution (HBSS: 1 g/liter D-glucose, 0.35 g/liter sodium bicarbonate, phenol red), cut into small pieces, washed twice again, and passed through a 1.9-mm mesh to remove the residual blood without enzymatic treatment. These samples were then filtered through a 45.7~pm stainless-steel mesh, and decidual mononuclear cells were isolated by FicollHypaque gradient sedimentation. Peripheral blood mononuclear cells were also isolated by Ficoll-Hypaque sedimentation. Immunophenotypic unalysis. The following
monoclonal antibodies (Becton Dickinson, San Jose, CA) were used: anti-CD45 FITC/CD14 PE (LeucoGATE), control yl/yza (rl FITCly,, PE), anti-CD3iCD19 (Leu-4 FITC/Leu-12 PE), anti-CD3/CDlb+CD56 (Leu-4 FITC/Leu-llc+l9 PE), anti-CD3/CD4 (Leu-4 FITCiLeu-3a PE), anti-CD3i CDS (Leu-4 FITCiLeu-2a PE), anti-CD3iCD69 (Leu-4 FITC/L23 PE), anti-CD3/CD25 (Leu-4 FITC/CD25 PE), anti-CD7 l/CD3 (CD7 1 FITC/Leu-4 PE), anti-CD3i HLA-DR (Leu-4 FITCIHLA-DR PE), and anti-CD3i CD38 (Leu-4 FITCILeu-17 PE). Incubation with monoclonal antibodies was done at 4°C for 30 min, and then the cells were washed twice in phosphate-buffered saline containing 2% fetal calf serum and 0.1% sodium azide. Immunofluorescence and dual-color flow cytometric analyses were performed using a FACScan cytofluorimeter (Becton Dickinson, San Jose, CA) with computer interfacing to Hewlett-Packard Consort 32 software for full-list-mode data storage, recovery and analysis. Dualcolor flow cytometry used an argon ion laser with 15 mW at 488-nm excitation. Triggering was set on the forward-scatter channel, and the threshold was adjusted to exclude debris. LeucoGATE was used to measure the proportion of lymphocytes in the sample being studied without any scatter gates. Then, the gare was set around the lymphocytes (CD45’CD14-) to exclude other cells from analysis. The Simultest control [mouse immunoglobulin (Ig)G, fluorescein isothiocyanate (FITC) + IgG,, phycoerythrin (PE)] was used for background control. For each experiment, 10,000 endometrial or intradecidual lymphocytes and 10,000 peripheral blood lymphocytes were evaluated. NK cytotoxicity by 5’Cr
release assay. Natural killer cell activity was measured using a chromium release assay (CRA) as described previously 19, 101. The K562 cells used as target cells were maintained in suspension culture in complete medium (RPM1 1640 medium supplemented with 10% fetal calf serum, 2 mM/ml of Lglutamine, 25 mM/ml of HEPES, 50 U/ml of penicillin, and 50 mg/ml streptomycin), and 2 x 10” K562 cells in
132
H.-N.
0.1 ml of complete medium were labeled by incubation with 100 pCi of ‘lCr. After 4 h incubation at 37°C the cells were washed three times with medium and resuspended to a concentration of 1 x 105/ml in complete medium. Aliquots (100 pl) of endometrial, decidual or peripheral blood lymphocytes at various concentrations were added in triplicate to loo-p1 aliquots of labeled target cells in U-bottomed, 96-well plates to give effector-totarget (E:T) ratios of 40: 1, 20: 1, 10: 1, and 5: 1. For each assay, maximum chromium release was determined following the addition of 100 pl 1% Triton X100, and spontaneous release was determined following the addition of 100 PI complete medium. After 4 h incubation at 37°C 100 pl of supernatant from each well was assayed in a y-counter. The percentage of specific chromium release was calculated as follows: Natural
Maximum
RESULTS In proliferative and early luteal phases, the B cells (CD1 9’CD3 -) were decreased in the endometrium compared to PBMC (Tables 1 and 2). The proportion of endometrial B cells increased after the late luteal phase, and the increase persisted in the early decidua tissue. However, the decidual B cells in patients with anembryonic pregnancies returned to the initial level observed in the endometrium of proliferative and early luteal phase. There was no difference in the proportion of T cells between peripheral blood mononuclear cells (PBMC) and endometrium in the proliferative and luteal phases, but there was an inverted CD&D8 ratio in the endometriurn. In early pregnancy, the T cells decreased in the decidua, mainly because of a reduced proportion of CDS’ lymphocytes, making the CD4CD8 ratio increase. The NK-cell population increased in the endometrium compared with PBMC; this increase occurred after normal and anembryonic pregnancies, but not after ectopic pregnancies. Nevertheless, there was no difference in NK-cell populations among the different phases of the menstrual cycle and between the normal and anembryonic pregnancies.
counts
counts - Spontaneous
x 100%.
counts
To compare NK cytotoxicity among different experiments and minimize the effect of different slopes of the lysis curve, cytotoxicity was expressed as lytic unit,,. One lytic unit,,, was designated as 20% specific lysis and was determined from the dose-response curve by a computer program Ill].
Activation
All results were expressed as the mean k standard deviation (SD). Statistical significance was tested using the paired and nonpaired Student t test. Ap value ~0.05 was considered significant. 1
Lymphocyte subpopulations and in normal, anembryonic, Proliferative phase
14.7 71.2 38.7 27.1 139.7 14.0 17.2 0.6 26.9 66.8 14.0
B cells T cells CD4 CD8 CD4iCD8 CD69iCD3 CD25iCD3 CD7 1/CD3 HLA-DRICD3 CD38iCD3 NK cells y Expressed as percentage
+ * + + f + f + + i i
of positwe
6.2’.’ 7.5 14.6’ 8.5b.‘ 74.9” 1o.4b 10.lb 0.5h.’ 11.1” 8.9h 8.3h
’p
< 0.05 compared
to lymphocytes
16.3 71.8 43.8 29.3 162.0 15.2 17.9 1.0 29.9 70.5 11.9
cells * SD on gated
’ p < 0.05 when PBMCs were compared
and T-cell activation in PBMCs in different and ectopic pregnancies* Late luteal phase (n = 14)
Early luteal phase (n = 16)
(n = 35)
with endometrial
of subjects
of T lymphocytes in endometrium and decidua.
There were increased CD69’CD3’, CD71’CD3’, HLADR’CD?‘, and CD38’CD3’ lymphocyte subpopulations in the endometrium and decidua compared with those in PBMC (Tables 1 and 2). The CD71’CD3’ lymphocyte subpopulation was further increased in the decidua than in the nonpregnant endometrium. On the contrary, the CD69’CD3’ and CD38’CD3’ lymphocytes in the decidua were further decreased compared with those in the nonpregnant endometrium. The pro-
Statistical analysis.
TABLE
* i * i + + if + k +
14.4 70.5 43.2 27.4 169.0 13.5 20.4 0.6 33.0 68.0 15.1
6.4”,’ 7.5 3.9h.‘ 6.3h 46.8”.‘ 8.5h 9.0h 1.7h 10.4’ 4.5b.’ 7.5b lymphocyte
with normal
i 7.3h + 6.4 i 6.0h‘ + 6.7h’ i 48.1’.‘ Yk7.4b + 8.2’ + 0.6’ i- 15.4” i- 5.1”.’ + 6.8’
population.
ot decidual pregnanca.
et al.
Proportion of various lymphocytes in endometrium and decidua.
killer cell lysis (%) =
Test counts - Spontaneous
Ho
lymphocytes
in Table 2.
Normal pregnancies (72 = 22) 11.1 71.4 36.6 33.3 113.0 16.0 17.2 0.3 26.6 58.7 15.5
f f + * + t + t + * i
4.0” 9.2h 8.9” 5.6” 35.2 6.5’ 5.0 0.3” 5.9’ 15.7b 7.5b
phases of menstrual
Anembryonic pregnancies (n = 20) 10.6 70.1 36.8 33.3 121.7 18.1 16.8 0.5 29.4 69.4 17.6
+ f f f 5 f + + + + +
5.8h 8.4h 8.2h 9.6” 51.4” 8.0b 6.6’ 1.1” 10.3’ 12.gh.‘ 8.2’
cycle
Ectopic pregnancies (n = 5) 9.8 76.9 44.1 29.3 156.0 6.9 16.9 1.2 27.8 63.9 9.1
i i i * * + * i * i +
3.2 2.9” 5.6”.’ 5.8 39.6” 2.0’.’ 6.1’ l.lb 7.4” 8.8h 2.7”.’
Activated
T and
TABLE
2
NK
in Endometrium
Lymphocyte endometrial anembryonic,
and
133
Decidua
subpopulations and the expression of activation markers of T and NK cells in lymphocytes in different phases of menstrual cycle and decidual lymphocytes in normal, and ectopic pregnancies”
Proliferative phase (n = 35)
Normal pregnancies (n = 22)
Late luteal phase (n = 14)
Early luteal phase (n = 16)
Anembryonic pregnancies (n = 20)
Ectopic pregnancies (n = 5)
B cells T cells CD4 CD8
2.8 67.2 17.5 44.8
f f + +
2.8’.’ 19.8 8.6 13.7’
2.3 65.5 17.2 48.4
+ e f f
1.6”,‘ 17.7’ 7.0 10.4
8.5 62.5 16.4 42.3
2 f + +
10.4 18.8 5.6 11.4
7.0 27.2 13.0 15.4
i f * i
7.0 14.0 5.9 8.4
3.1 27.0 12.9 16.5
f + f +
1.9 16.4 7.7 10.4
6.1 43.5 19.9 24.4
+ + + +
3.3 12.9 5.6 7.7
CD4/CD8 CD69/CD3 CD25/CD3 CD7 1/CD3 HLA-DRICD3 CD38/CD3 NK cells CDBICD3 CD69/CD3 CD25/CD3CD71/CD3HLA-DR/CD3CD38/CD3-
44.5 86.0 7.6 3.0 76.4 93.4 30.0 37.1 81.0 18.6 4.6 46.3 96.7
+ + + + + f f i + * i + +
7.6 16.2’ 5.9 2.1’ 10.0 3.5’ 20.2‘ 10.7’ 16.56.’ 11.9 3.7’ 18.5‘ 89.3’
41.8 84.0 6.5 3.1 79.9 89.8 32.1 37.2 81.0 15.0 3.2 39.9 93.4
f + + + + + * + + + + + f
7.7’ 18.1‘ 3.7’ 2.6 7.1 10.3’ 16.6 9.4 14.9h,‘ 12.1 4.7 16.1‘ 6.8
49.3 81.5 8.2 3.8 77.5 90.1 29.0 37.5 68.3 16.1 4.1 46.1 92.7
t * t -t i* t * + + + + i
17.8 17.2’ 5.0 2.3’ 14.3 2.8 19.5‘ 14.1‘ 18.8’ 8.1‘ 3.7 25.5’ 6.4
97.2 67.4 21.4 6.4 68.1 74.9 62.8 14.5 44.0 9.1 1.8 28.2 64.6
+ i + k + + + + f -t + + +
38.4 16.4 10.4 7.4 18.3 14.5 21.0 16.3 23.2 6.6 2.0 12.6 19.5
82.6 75.5 23.5 5.9 82.3 84.0 69.8 21.7 62.6 20.0 1.8 33.6 73.9
+ + k i + k + k ik + + +
24.0 16.2 9.8 5.2 15.8‘ 14.9 17.1 13.1 21.0 14.6 3.4 13.3 16.3
83.1 71.1 9.9 5.4 75.0 89.0 46.4 11.7 37.8 7.9 1.3 33.9 48.3
+ + 5 + i * + * + + k i +
9.3 14.1 2.5‘ 3.0 7.9 8.4 15.3‘ 2.5 18.9 3.8 0.7 6.2 15.1
“ Expressed as percentage of positive cells r SD on gated lymhocyte population 'p < 0.05
when lymphocytes of different phases were compared.
‘p < 0.05
compared to lymphocytes of subjects with normal pregnancies.
portion of CD25 ‘CD3’ lymphocytes decreased markedly in the endometrium compared with that in PBMC. However, this difference was not seen in the decidua of normal pregnancies, and the percentage of CD25’CD3’ lymphocytes was increased in anembryonic, but not ectopic pregnancies (Table 2).
the decidua were CD16-CD56’ (Table 3). The percentage of CD57-CD56’ lymphocytes was also greater than that of CD57’CD56’ lymphocytes in the decidua, making the CD57’CD56’/CD56’ ratio decrease. In the decidua, there are fewer B cells than NK cells (Table 2). Therefore, most of the CD3- lymphocytes could be considered NK cells. The CD3- lymphocytes with activation markers (CD69, CD25, CD71, HLA-DR, and CD38) were less expressed in the decidua than those in the endometrium. The CD69’CD3and CD25’CD3lymphocytes increased significantly in the decidua of
Subpopulations and activation of NK cells in the decidua and their cytotoxicity in normal and anembryonicpregnancies. The distribution of CD16’CD56’ lymphocytes was similar in PBMC and decidua, although most lymphocytes in TABLE
3
Subtypes of NK cells in PBMCs and decidua pregnancies and anembryonic pregnancies” Normal pregnancies (n = 22) PBMCs
NK CD16-CD56’ CD16+CD56’ CD16’/CD56CD16+CD56+/CD56+ CD57’CD56’ CD57-CD56’ CD57+CDSb’/CD56+
15.5 11.7 11.4 1.0 46.0 10.9 11.1 50.0
* + f f + f f f
7.5 5.6 8.4 0.9 26.0 5.3 4.9 14.6
Deciuda 62.8 59.6 8.7 0.7 15.4 6.4 62.4 11.6
+ f i t * k i +
21.0” 23.1” 4.7 0.6 11.5* 4.0” 22.8” 9.4b
of normal
Anembryonic pregnancies (n = 20) PBMCs 17.6 11.4 10.1 1.1 46.5 10.2 9.2 52.0
+ 8.2 + 6.1 Yk6.8 i 1.3 f 25.3 f 4.2 + 3.4 + 9.5
y Expressed as percentage of positive cells r SD on gated lymphocyte population. ’ p < 0.05 compared to peripheral blood lymphocytes.
Decidua 69.8 68.3 8.3 0.9 12.3 5.3 70.9 7.8
+ + + i i + f +
17.1/’ 18.Bh 4.3 1.2 9.6” 3.7h 17.4” 6.9’
134
H.-N. Ho et al.
anembryonic pregnancies compared to normal pregnancies. The NK cytotoxicity assay showed that decidual NK activity was significantly less than that of the peripheral blood activity in normal pregnancies (p = O.OOl), but significantly increased in anembryonic pregnancies (t, = 0.006) (Fig. 1).
DISCUSSION Our present study has shown that the nonpregnant endometrium is infiltrated by cells with phenotypes similar to those seen in the decidua, and there is no clearly discernible increase in their proportion from the proliferative to late luteal phase (which is the time the blastocyst would implant). It is impressive that there are almost no B cells in the uterus and there is a gradual increase in the percentage of B cells from the proliferative endometrium through secretory endometrium to decidua. However, the decrease of T cells and the increase of NK cells in the endometrium are noted only during the shifting from late luteal phase to early pregnancy. Nevertheless, the presence of these cells in the nonpregnant endometrium and the variation in their number in relation to the phases of the menstrual cycle suggest that their recruitment may be under hormonal control as well as a result of trophoblast stimulation. There are numerous bone marrow-derived cells in the human endometrium and early pregnancy decidua, which include tissue macrophages, NK cells and T cells. The functions of these endometrial and decidual leukoFIGURE 1 NK cytotoxicity (lytic unit,,) in the peripheral blood (0, Cl) and decidua (0, n ) of women with normal and anembryonic pregnancies. The means are indicated by a dash in each column. m
1
Normal Pregnancy (P=O.o013)
Anembryonic Pregnancy (P=O.W63)
cytes are not known. CD56 is expressed by lo-15% of peripheral blood lymphocytes, which are morphologically LGLs, mediating non-major histocompatibility complex-restricted cytotoxicity 1121. The largest population during the first trimester of pregnancy is CD1 6- or CD56-positive/CD3-negative NK cells. The proportion of NK cells in the endometrium further increased after normal and anembryonic, but not ectopic pregnancies. This might indicate that the further recruitment of NK cells in the early decidua is due to local stimulation by fetal antigens. The preparation of cells from human decidua has been reported to have NK activity which is confined to the subset of CD56-positive NK cells, which is also CD3 and CD16 negative [13}. This corresponds to a minor subset of PBMCs with only weak NK activity {S]. These findings are essentially consistent with our previous [6, 71 and present observations. This selective migration into the uterus does suggest some important role for these cells during early pregnancy. In peripheral blood, most CD56-positive cells are also CD16 positive and have high NK activity, which could explain the lower cytotoxic activity of decidual NK cells than that of PBMCs. The fraction of CD16’CD56’ lymphocytes is similar in both PBMC and decidua, whereas there is a marked increase of CD16-CD56’ lymphocytes in the decidua. This condition is also present in anembryonic pregnancies, although the suppression of decidual NK cytotoxicity disappears in the anembryonic pregnancies. Moreover, the CD16’CD56’iCD56’ ratio in the decidua is also similar in normal and anembryonic pregnancies. Therefore, the increase in NK cytotoxicity of anembryonic pregnancies could not be explained merely by the increase in CD16’CD56’ NK subpopulation. Compared with PBMCs, the ratio of CD4’CD3’ to CDS’CD3 + lymphocytes is inverted in the endometrium and decidua, although both the percentages of CD4’CD3’ and CDg’CD3’ lymphocytes decreased in the decidua. Whether the marked decrease of CDS’CD3’ lymphocytes in the decidua compared with that in the nonpregnant endometrium is necessary for maintaining the pregnancy needs further investigation. A significantly higher proportion of NK cells and an inverted CD4’CD3’/CDS’CD3’ ratio possibly reflect the change in local endometrial environment. After implantation of an embryo and during pregnancy, lymphocyte subpopulations alter with the stimulation by fetal antigens, and NK cells increase in number but decrease in cytotoxicity. Evidence of T-cell activation in human endometrium has been reported {I4-161. Our results confirm the reports that endometrial T lymphocytes are activated by increasing the expression of early and late T-cell activation markers. The expression of the CD69 antigen (an early activation marker for T and NK cells) has been
Activated T and NK in Endometrium
135
and Decidua
reported to appear shortly after stimulation by activation of the protein kinase C (PKC) pathway 117, 181. Zeigler et al. 1191 showed that the induction of CD69 mRNA in activated murine thymocytes and T cells is very rapid and usually precedes the interleukin (IL)-2cx receptor (CD25) expression. Our finding indicates that the T cells in the endometrium and decidua have already been stimulated via PKC. The HLA-DR antigen (a relatively late activation T-cell marker) was detected in about 30% of T cells in the peripheral blood lymphocytes of nonpregnant and pregnant women. In contrast, in the endometrium and decidua, most T cells showed HLA-DR antigen expression. Thus, we show here that endometrial and decidual T cells express an early activation antigen, CD69, and a late activation antigen, HLA-DR. Because the summation of the percentage of T lymphocytes expressing CD69 and HLA-DR greatly exceeded 100%; there may be a coexpression of early and late antigens on many T cells. Therefore, the endometrial and decidual T lymphocytes may be in a state of recent and persistent activation. In relation to this, it is interesting to note that the level of the CD25 antigen expression is downregulated in the nonpregnant endometrium, but its expression increases after pregnancy. The increased expression of T-cell-surface markers in the sequence CD25 and CD71 after pregnancy may be the result of the stimulation by fetal antigens, as previously described 120). In this study, CD25’CD3’ and CD7 1 ‘CD3 + subpopulations are increased in the decidual of normal pregnancies compared with those in nonpregnant endometrium. However, the CD25’CD3’ subpopulation in the decidua of ectopic pregnancies is less than that in normal pregnancies, and it return to a level similar to that in the nonpregnant endometrium. This observation further strengthens the suggestion that the increase in the decidual CD25’CD3’ subpopulation is due to local stimulation by fetal antigens. It has been reported that anti-CD8 treatment of pregnant mice reduces placental proliferation and phagocytosis, and that anti-CD4 antibody treatment affects placental proliferation 121). Therefore, it would appear that activated T cells may be involved in placental cell proliferation and function. Since trophoblast produces an IL-2-like molecule and expresses IL-2 mRNA at an early stage of pregnancy 122, 231, it is possible that this factor may activate T cells. However, it is also possible that other antigen-presenting cells in the decidua are involved in T-cell activation 1241. In the decidua, there are fewer B than NK cells (7.0 + 7.0% vs. 62.8 + 21.0%), and it might be reasonable to consider that most of the CD3- cells were NK cells. Therefore, the data in this study indicate that the NK cells in the decidua were activated (Table 2). However, the factors leading to activation of decidual NK cells in
viva and the function of such activated cells are still not known. It is possible that decidual NK cells might be expected to play a role in regulating placental growth and development through the production of cytokines or other factors, such as transforming growth factor B 125, 261 and thereby control the invasive growth of the trophoblast by direct cytotoxic activity or by regulating local maternal immune responses to trophoblasts 1271. It has been demonstrated that the NK activity of decidual NK cells against the standard NK target cell line K562 is low compared to that of peripheral blood NK cells 171. However, this difference was not present in anembryonic pregnancies (Fig. 1) and the proportion of NK-cell activation (CD69’CD3and CD25’CD3-) increased in these pregnancies. Whether the increase in this activated NK-cell subpopulation correlated with the augmentation of NK cytotoxicity needs further evaluation. In summary, our findings show that activated NK cells are present in the decidua. In addition, a significant proportion of decidual CD3’ cells also expresses CD25 and CD71, indicating that these cells are activated, although the mechanism of activation is unknown. Further studies are required to investigate whether and to what extent these cell types play a role in controlling trophoblast growth and placental development.
ACKNOWLEDGMENTS
The authors extend their sincere gratitude
to Professor. T. J.
Gill III, from the University of Pittsburgh, for critical discussion of the manuscript. This study was supported partially by grants from the National Science Council of the Republic of China (NSC84-2331-B002-287 and NSC85-2331-B002-113) and the Research Program of National Taiwan University Hospital (NTUH-84241-B24 and NTUH-84242-B16).
REFERENCES 1. Bulmer JN, Sunder-land CA: Immunohistological characterization of lymphoid cell population in the early human placental bed. Immunology 52:349, 1984. 2. Bulmer JN, Johnson PM: Immunohistological characterization of the decidual leucocytic infiltrate related to endometrial gland epithelium in early human pregnancy. Immunology 55:35, 1985. granulocytes in hu3. Ritson A, Bulmer JN: Endometrial man decidua react with a natural-killer (NK) cell marker, NKHl. Immunology 62:329, 1987. 4. Starkey PM, Sargent IL, Redman CWG: Cell population in human early pregnancy decidua: characterization and isolation of large granular lymphocytes by flow cytometry. Immunology 65:129, 1988. 5. Lanier LL, Le AM, Civin CI, Loken MR, Phillips JH: The relationship of CD16 (Leu-11) and Leu-19 (NKH-1) an-
136
H.-N.
tigen expression on human peripheral cytotoxic T lymphocytes. J Immunol
blood NK cells and 136:4480, 1986.
6. Chen CK, Huang SC, Chen CL, Yen MR, Hsu HC, Ho HN: Increased expressions of CD69 and HLA-DR but not CD25 or CD71 on endometrial T lymphocytes of nonpregnant women. Hum Immunol 42:227, 1995. 7. Chao KH, Yang YS, Ho HN, Chen SU, Chen HF, Dai HJ, Huang SC, Gill TJ III: Decidual natural killer cytotoxicity decreased in normal pregnancy but not in anembryonic pregnancy and recurrent spontaneous abortion. Am J Reprod Immunol 34:274, 1995. 8. Saito S, Nishikawa K, Morii T, Narita N, Enomoto M, Ichijo M: Expression of activation antigens CD69, HLADR, interleukin-2 receptor-alpha (IL-2Ra) and IL-2RB on T cells of human decidua at an early stage of pregnancy. Immunology 75:710, 1992. 9. Ho HN, Hultin LE, Mitsuyasu RT, Matud JL, Hausner MA, Bockstoce D, Chou CC, O’Rourke S, Taylor JMG, Giorgi JV: Circulating HIV-specific CD8’ cytotoxic T cells express CD38 and HLA-DR antigens. J Immunol 150:3070, 1993. 10. Ho HN, Chao KH, Chen HF, Wu MY, Yang YS, Lee TY: Peritoneal natural killer cytotoxicity and CD25’CD3’ lymphocyte subpopulation are decreased in women with stage III-IV endometriosis. Hum Reprod IO:267 1, 1995. 11. Pross HF, Baines MG, Rubin P, Shragge P, Patterson MS: Spontaneous human lymphocyte-mediated cytotoxicity against tumor target cells. IX. The quantitation of natural killer cell activity. J Clin Immunol 1:51, 1981. 12. Lanier LL, Le AM, Cwirla S, Federspiel N, Philips JH: Antigeneic, functional and molecular genetic studies of human natural killer cells and cytotoxic T lymphocytes not restricted by the major histocompatibility complex. Fed Proc 4512823, 1986. 13. Ferry BL, Starkey PM, Sargent IL, Watt GMO, Jackson M, Redman CWG: Cell populations in the human early pregnancy decidua: natural killer activity and response to interleukin-2 of CD56-positive large granular lymphocytes. Immunology 70:446, 1990. 14. Pace D, Morrison L, Bulmer JN: Proliferative activity in endometrial stromal granulocytes throughout menstrual cycle and early pregnancy. J Clin Path01 42:35, 1989. 15. Tabibzadeh S: Proliferative activity of lymphoid human endometrium throughout the menstrual Clin Endocrinol Metab 70~437, 1990.
cells in cycle. J
Ho et al.
16. Tabibzadeh S: Evidence of T-cell activation and potential cytokine activation in human endometrium. J Clin Endocrinol Metab 71:645, 1990. 17. Testi R, Phillips JH, Lanier LL: T cell activation 23 (CD69). J Immunol 143:1123, 1989. 18. Lanier LL, Burk DW, Rhodes L, Ding A, C, Phillips JH: Interleukin 2 activation cells rapidly induces the expression and of the Leu 23 activation antigen. J Exp 1988.
via Leu-
Evans E, Barney of natural killer phosphorylation Med 167:1572,
19. Zeigler SF, Levin SD, Johnson L, Copeland NG, Gilbert DJ, Jenkins NA, Baker E, Sutherland GR, Feldhaus AL, Rawdsdell F. The mouse CD69 gene, structure, expression and mapping to the NK gene complex. J Immunol 152:1228, 1994. 20. Ho HN, Chen HF, Chen SU, Chao KH, Yang YS, Huang SC, Lee TY, Gill TJ III. Gonadotropin releasing hormone (GnRH) agonist induced down-regulation of the CD3’CD25’ lymphocyte subpopulation in peripheral blood. Am J Reprod Immunol 33:243, 1995. 21. Athanassakis I, Chaouat G, Wegmann TG: The effects of anti-CD4 and anti-CD8 antibody treatment on placental growth and function in allogeneic and syngenic murine pregnancy. Cell Immunol 129:13, 1990. 22. Soubiran P, Zapitelli JP, Schaffar L: IL-2-like material is present in human placenta and amnion. J Reprod Immuno1 12:225, 1987. 23. Boehm KD, Kelley MF, Ilan J, Ilan J: The interleukin 2 gene is expressed in syncytiotrophoblast of human placenta. Proc Nat1 Acad Sci USA 86:656, 1989. 24. Oksenberg JR, Mor-Yosef S, Persity E, Schenker Y, Mozes E, Brantbar C: Antigen-presenting cells in human decidual tissue. Am J Reprod Immunol Microbial 11:82, 1986. 25. Dungy LJ, Siddigi TA, Khan S: Transforming growth factor-P1 expression during placental development. Am J Obstet Gynecol 165:853, 1991. 26. Saito S, Nishikawa K, Morii T, Enomoto M, Narita N, Motoyoshi K, Ichijo M: Cytokine production by CD16mCD56b”“h’ natural killer cells in the human early pregnancy decidua. Int Immunol 5:559, 1993. 27. King A, Loke YW. On the nature and function of human uterine granular lymphocytes. Immunol Today 12:432, 1991.
Activation Status of T and NK Cells in the Endometrium Throughout Menstrual Cycle and Normal and Abnormal Early Pregnancy Hong-Nerng Ho, Kuang-Han Chao, Chun-Kai Yu-Shih Yang, and Su-Cheng Huang
Chen,
ABSTRACT: Endometrial lymphocytes were studied at all stages throughout the menstrual cycle and early pregnancy by flow cytometry to examine different lymphocyte subpopulations and the expression of the T- and NK-cell activation markers. After pregnancy, CDS’CD3’ lymphocytes were decreased in the decidua. In both endometrium and decidua, more T cells expressed CD69, CD71, HLA-DR, and CD38 antigens than in peripheral blood. After pregnancy, CD7 1 ‘CD3’ lymphocytes were further increased. CD25’CDI’ lymphocytes decreased significantly in the endometrium and decidua of ectopic pregnancies, but not in the decidua of normal pregnancies. These findings indicate that T cells are regionally acti-
vated in the first trimester, and it may be the result of the stimulation by fetal antigens. NK cells were the most abundant cell type in the decidua, which expressed the phenotype CD16- CD56’, and CD57-CD56’. The proportion of activated decidual NK cells was increased in anembryonic pregnancies more than in normal pregnancies, although the total NK subpopulation was similar in both groups. This might result in increased NK cytotoxicity in anembryonic pregnancies. In conclusion, T cells are activated, but NK cytotoxicity is decreased in the decidua of early normal pregnancies. This might be important in the control of trophoblast growth and placental development. HuTnan Immunology 49, 130-136 (19961
ABBREVIATIONS FITC fluorescein isothiocyanate LGLs large granular lymphocytes natural killer NK
PBMCs PE
peripheral blood mononuclear phycoerythrin
cells
INTRODUCTION Since the decidua
is the maternal tissue in closest contact with the fetal tissue, it is likely that it contributes to the maintenance of pregnancy by forming a local network supporting cellular and humoral immunity. Immunohistochemical analysis of human decidua in the first trimester of pregnancy identified three principle leukocyte populations: macrophage, classic T lymphocytes, and an unusual lymphocyte population [l, 21, subsequently identified as expressing antigen characteristic of large granular lymphocytes (LGLs) 131. The most abundant cell type in human decidua has
From the Department of Obstetricsand Gynecology, College of Medicine and Hospital, National Taiwan University, Taipei, Taiwan. ROC. Address reprint requeststo: Dr. Hong-Nevng Ho, Department ofObstetrics and Gynecology,National Taiwan University Hospital, No. 7. Chung Shari South Road. 10016, Taipei, Taiwan, ROC. Received February 12. 1996: acceptedApril 23. 1996. Human Immunology 49, 130-136 (1996) 0 Amencan Society for Histocompatibilicy
and Immunogenetics,
1996
been identified with flow cytometry by the expression of its surface antigens, size, and granularity, as LGLs 141. These LGLs comprise 245% of all decidual cells and are mainly CD16_CD56’CD3141, which corresponds to a minor population of natural killer (NK) cells in peripheral blood. In contrast to decidual LGLs, CD16’CD56’CD3NK cells were the most abundant and the most efficient cytotoxic effecters in peripheral blood 151. In a previous study, we found the proportion of NK cells (defined as CD56’CD3or CDlb’CD3-) was significantly higher in the nonpregnant endometriurn than in the peripheral blood 161, and this difference was more apparent in the decidua during pregnancy 171. It was also noted that decidual NK cytotoxicity decreased in normal pregnancy, but not in anembryonic pregnancy or in recurrent spontaneous abortion 17}. Whether the increase in NK cytotoxicity in the cases of spontaneous abortion is due to the increase in the pro0198~8859/961$15,00 PI1 S1098-8859(96)00120-h
Activated T and NK in Endometrium
portion of CDlb’CD56’CD3NK cells needs to be clarified. Our previous report revealed that endometrial T lymphocytes were activated by increasing the expression of early and late T-cell activation markers (CD69 and HLADR) {bl. In 1992, Saito et al. [8} showed decidual T cells expressed CD69, HLA-DR, and CD25 antigen significantly, indicating that the T cells in the decidua were also regionally activated. NK activity could be regulated by lymphokines, and activated T cells had the ability to release lymphokines which could modulate the NK cytotoxicity. In this study, we explored whether the relatively higher levels of NK cell activity in anembryonic pregnancy is the result of activation of T or NK cells. The study was designed to investigate whether there was consistency between the activation of T and NK cells in the nonpregnant endometrium and decidua, and whether there was any difference in the level of activation in normal and abnormal pregnancies.
MATERIALS
AND
METHODS
Szlbjects. Sixty-five patients who received hysterectomies for benign reasons of nonendometrial pathology (mostly myoma uteri), 22 pregnant women who had elective abortions of normal pregnancies due to multiparity, 20 women who had anembryonic pregnancies, and 5 women with ectopic tubal pregnancy between 6 and 10 weeks of gestational age were enrolled in the study, with written informed consent. The pathologic results of the hysterectomy specimens showed no evidence of infection or inflammation in the 35 specimens in the follicular phase and the 30 specimens in the luteal phase. The luteal phase was further divided into early and late stages according to the last menstrual period, levels of progesterone, and the histologic dating. Anembryonic pregnancy was identified by sonography, which failed to show a fetal pole when the gestational sac was >25 mm. Specimens. Endometrial
131
and Decidua
tissues and venous blood obtained from 65 hysterectomy patients were collected on the day of hysterectomy. Decidual tissue and peripheral blood samples were taken from each pregnant woman at the time of abortion and laparoscopic surgery or laparotomy. Endometrial tissue was cut with a surgical knife immediately after hysterectomy, minced with a scalpel, suspended in RPMI-1640 medium, and pressed gently through a 380~pm and then a 45.7-pm sieve, as described by Chen et al. 161. The filtered solution was layered over a Ficoll-Hypaque gradient and centrifuged for 45 min at 400 x g. An enriched cell suspension was collected and then washed twice with RPMI-1640 medium. The recovered cells were checked for viability with trypan blue and counted. In the pregnant cases, to mini-
mize the contamination by blood, the decidual tissue was macroscopially separated from chorionic villi, washed twice with Hank’s balanced salt solution (HBSS: 1 g/liter D-glucose, 0.35 g/liter sodium bicarbonate, phenol red), cut into small pieces, washed twice again, and passed through a 1.9-mm mesh to remove the residual blood without enzymatic treatment. These samples were then filtered through a 45.7~pm stainless-steel mesh, and decidual mononuclear cells were isolated by FicollHypaque gradient sedimentation. Peripheral blood mononuclear cells were also isolated by Ficoll-Hypaque sedimentation. Immunophenotypic unalysis. The following
monoclonal antibodies (Becton Dickinson, San Jose, CA) were used: anti-CD45 FITC/CD14 PE (LeucoGATE), control yl/yza (rl FITCly,, PE), anti-CD3iCD19 (Leu-4 FITC/Leu-12 PE), anti-CD3/CDlb+CD56 (Leu-4 FITC/Leu-llc+l9 PE), anti-CD3/CD4 (Leu-4 FITCiLeu-3a PE), anti-CD3i CDS (Leu-4 FITCiLeu-2a PE), anti-CD3iCD69 (Leu-4 FITC/L23 PE), anti-CD3/CD25 (Leu-4 FITC/CD25 PE), anti-CD7 l/CD3 (CD7 1 FITC/Leu-4 PE), anti-CD3i HLA-DR (Leu-4 FITCIHLA-DR PE), and anti-CD3i CD38 (Leu-4 FITCILeu-17 PE). Incubation with monoclonal antibodies was done at 4°C for 30 min, and then the cells were washed twice in phosphate-buffered saline containing 2% fetal calf serum and 0.1% sodium azide. Immunofluorescence and dual-color flow cytometric analyses were performed using a FACScan cytofluorimeter (Becton Dickinson, San Jose, CA) with computer interfacing to Hewlett-Packard Consort 32 software for full-list-mode data storage, recovery and analysis. Dualcolor flow cytometry used an argon ion laser with 15 mW at 488-nm excitation. Triggering was set on the forward-scatter channel, and the threshold was adjusted to exclude debris. LeucoGATE was used to measure the proportion of lymphocytes in the sample being studied without any scatter gates. Then, the gare was set around the lymphocytes (CD45’CD14-) to exclude other cells from analysis. The Simultest control [mouse immunoglobulin (Ig)G, fluorescein isothiocyanate (FITC) + IgG,, phycoerythrin (PE)] was used for background control. For each experiment, 10,000 endometrial or intradecidual lymphocytes and 10,000 peripheral blood lymphocytes were evaluated. NK cytotoxicity by 5’Cr
release assay. Natural killer cell activity was measured using a chromium release assay (CRA) as described previously 19, 101. The K562 cells used as target cells were maintained in suspension culture in complete medium (RPM1 1640 medium supplemented with 10% fetal calf serum, 2 mM/ml of Lglutamine, 25 mM/ml of HEPES, 50 U/ml of penicillin, and 50 mg/ml streptomycin), and 2 x 10” K562 cells in
132
H.-N.
0.1 ml of complete medium were labeled by incubation with 100 pCi of ‘lCr. After 4 h incubation at 37°C the cells were washed three times with medium and resuspended to a concentration of 1 x 105/ml in complete medium. Aliquots (100 pl) of endometrial, decidual or peripheral blood lymphocytes at various concentrations were added in triplicate to loo-p1 aliquots of labeled target cells in U-bottomed, 96-well plates to give effector-totarget (E:T) ratios of 40: 1, 20: 1, 10: 1, and 5: 1. For each assay, maximum chromium release was determined following the addition of 100 pl 1% Triton X100, and spontaneous release was determined following the addition of 100 PI complete medium. After 4 h incubation at 37°C 100 pl of supernatant from each well was assayed in a y-counter. The percentage of specific chromium release was calculated as follows: Natural
Maximum
RESULTS In proliferative and early luteal phases, the B cells (CD1 9’CD3 -) were decreased in the endometrium compared to PBMC (Tables 1 and 2). The proportion of endometrial B cells increased after the late luteal phase, and the increase persisted in the early decidua tissue. However, the decidual B cells in patients with anembryonic pregnancies returned to the initial level observed in the endometrium of proliferative and early luteal phase. There was no difference in the proportion of T cells between peripheral blood mononuclear cells (PBMC) and endometrium in the proliferative and luteal phases, but there was an inverted CD&D8 ratio in the endometriurn. In early pregnancy, the T cells decreased in the decidua, mainly because of a reduced proportion of CDS’ lymphocytes, making the CD4CD8 ratio increase. The NK-cell population increased in the endometrium compared with PBMC; this increase occurred after normal and anembryonic pregnancies, but not after ectopic pregnancies. Nevertheless, there was no difference in NK-cell populations among the different phases of the menstrual cycle and between the normal and anembryonic pregnancies.
counts
counts - Spontaneous
x 100%.
counts
To compare NK cytotoxicity among different experiments and minimize the effect of different slopes of the lysis curve, cytotoxicity was expressed as lytic unit,,. One lytic unit,,, was designated as 20% specific lysis and was determined from the dose-response curve by a computer program Ill].
Activation
All results were expressed as the mean k standard deviation (SD). Statistical significance was tested using the paired and nonpaired Student t test. Ap value ~0.05 was considered significant. 1
Lymphocyte subpopulations and in normal, anembryonic, Proliferative phase
14.7 71.2 38.7 27.1 139.7 14.0 17.2 0.6 26.9 66.8 14.0
B cells T cells CD4 CD8 CD4iCD8 CD69iCD3 CD25iCD3 CD7 1/CD3 HLA-DRICD3 CD38iCD3 NK cells y Expressed as percentage
+ * + + f + f + + i i
of positwe
6.2’.’ 7.5 14.6’ 8.5b.‘ 74.9” 1o.4b 10.lb 0.5h.’ 11.1” 8.9h 8.3h
’p
< 0.05 compared
to lymphocytes
16.3 71.8 43.8 29.3 162.0 15.2 17.9 1.0 29.9 70.5 11.9
cells * SD on gated
’ p < 0.05 when PBMCs were compared
and T-cell activation in PBMCs in different and ectopic pregnancies* Late luteal phase (n = 14)
Early luteal phase (n = 16)
(n = 35)
with endometrial
of subjects
of T lymphocytes in endometrium and decidua.
There were increased CD69’CD3’, CD71’CD3’, HLADR’CD?‘, and CD38’CD3’ lymphocyte subpopulations in the endometrium and decidua compared with those in PBMC (Tables 1 and 2). The CD71’CD3’ lymphocyte subpopulation was further increased in the decidua than in the nonpregnant endometrium. On the contrary, the CD69’CD3’ and CD38’CD3’ lymphocytes in the decidua were further decreased compared with those in the nonpregnant endometrium. The pro-
Statistical analysis.
TABLE
* i * i + + if + k +
14.4 70.5 43.2 27.4 169.0 13.5 20.4 0.6 33.0 68.0 15.1
6.4”,’ 7.5 3.9h.‘ 6.3h 46.8”.‘ 8.5h 9.0h 1.7h 10.4’ 4.5b.’ 7.5b lymphocyte
with normal
i 7.3h + 6.4 i 6.0h‘ + 6.7h’ i 48.1’.‘ Yk7.4b + 8.2’ + 0.6’ i- 15.4” i- 5.1”.’ + 6.8’
population.
ot decidual pregnanca.
et al.
Proportion of various lymphocytes in endometrium and decidua.
killer cell lysis (%) =
Test counts - Spontaneous
Ho
lymphocytes
in Table 2.
Normal pregnancies (72 = 22) 11.1 71.4 36.6 33.3 113.0 16.0 17.2 0.3 26.6 58.7 15.5
f f + * + t + t + * i
4.0” 9.2h 8.9” 5.6” 35.2 6.5’ 5.0 0.3” 5.9’ 15.7b 7.5b
phases of menstrual
Anembryonic pregnancies (n = 20) 10.6 70.1 36.8 33.3 121.7 18.1 16.8 0.5 29.4 69.4 17.6
+ f f f 5 f + + + + +
5.8h 8.4h 8.2h 9.6” 51.4” 8.0b 6.6’ 1.1” 10.3’ 12.gh.‘ 8.2’
cycle
Ectopic pregnancies (n = 5) 9.8 76.9 44.1 29.3 156.0 6.9 16.9 1.2 27.8 63.9 9.1
i i i * * + * i * i +
3.2 2.9” 5.6”.’ 5.8 39.6” 2.0’.’ 6.1’ l.lb 7.4” 8.8h 2.7”.’
Activated
T and
TABLE
2
NK
in Endometrium
Lymphocyte endometrial anembryonic,
and
133
Decidua
subpopulations and the expression of activation markers of T and NK cells in lymphocytes in different phases of menstrual cycle and decidual lymphocytes in normal, and ectopic pregnancies”
Proliferative phase (n = 35)
Normal pregnancies (n = 22)
Late luteal phase (n = 14)
Early luteal phase (n = 16)
Anembryonic pregnancies (n = 20)
Ectopic pregnancies (n = 5)
B cells T cells CD4 CD8
2.8 67.2 17.5 44.8
f f + +
2.8’.’ 19.8 8.6 13.7’
2.3 65.5 17.2 48.4
+ e f f
1.6”,‘ 17.7’ 7.0 10.4
8.5 62.5 16.4 42.3
2 f + +
10.4 18.8 5.6 11.4
7.0 27.2 13.0 15.4
i f * i
7.0 14.0 5.9 8.4
3.1 27.0 12.9 16.5
f + f +
1.9 16.4 7.7 10.4
6.1 43.5 19.9 24.4
+ + + +
3.3 12.9 5.6 7.7
CD4/CD8 CD69/CD3 CD25/CD3 CD7 1/CD3 HLA-DRICD3 CD38/CD3 NK cells CDBICD3 CD69/CD3 CD25/CD3CD71/CD3HLA-DR/CD3CD38/CD3-
44.5 86.0 7.6 3.0 76.4 93.4 30.0 37.1 81.0 18.6 4.6 46.3 96.7
+ + + + + f f i + * i + +
7.6 16.2’ 5.9 2.1’ 10.0 3.5’ 20.2‘ 10.7’ 16.56.’ 11.9 3.7’ 18.5‘ 89.3’
41.8 84.0 6.5 3.1 79.9 89.8 32.1 37.2 81.0 15.0 3.2 39.9 93.4
f + + + + + * + + + + + f
7.7’ 18.1‘ 3.7’ 2.6 7.1 10.3’ 16.6 9.4 14.9h,‘ 12.1 4.7 16.1‘ 6.8
49.3 81.5 8.2 3.8 77.5 90.1 29.0 37.5 68.3 16.1 4.1 46.1 92.7
t * t -t i* t * + + + + i
17.8 17.2’ 5.0 2.3’ 14.3 2.8 19.5‘ 14.1‘ 18.8’ 8.1‘ 3.7 25.5’ 6.4
97.2 67.4 21.4 6.4 68.1 74.9 62.8 14.5 44.0 9.1 1.8 28.2 64.6
+ i + k + + + + f -t + + +
38.4 16.4 10.4 7.4 18.3 14.5 21.0 16.3 23.2 6.6 2.0 12.6 19.5
82.6 75.5 23.5 5.9 82.3 84.0 69.8 21.7 62.6 20.0 1.8 33.6 73.9
+ + k i + k + k ik + + +
24.0 16.2 9.8 5.2 15.8‘ 14.9 17.1 13.1 21.0 14.6 3.4 13.3 16.3
83.1 71.1 9.9 5.4 75.0 89.0 46.4 11.7 37.8 7.9 1.3 33.9 48.3
+ + 5 + i * + * + + k i +
9.3 14.1 2.5‘ 3.0 7.9 8.4 15.3‘ 2.5 18.9 3.8 0.7 6.2 15.1
“ Expressed as percentage of positive cells r SD on gated lymhocyte population 'p < 0.05
when lymphocytes of different phases were compared.
‘p < 0.05
compared to lymphocytes of subjects with normal pregnancies.
portion of CD25 ‘CD3’ lymphocytes decreased markedly in the endometrium compared with that in PBMC. However, this difference was not seen in the decidua of normal pregnancies, and the percentage of CD25’CD3’ lymphocytes was increased in anembryonic, but not ectopic pregnancies (Table 2).
the decidua were CD16-CD56’ (Table 3). The percentage of CD57-CD56’ lymphocytes was also greater than that of CD57’CD56’ lymphocytes in the decidua, making the CD57’CD56’/CD56’ ratio decrease. In the decidua, there are fewer B cells than NK cells (Table 2). Therefore, most of the CD3- lymphocytes could be considered NK cells. The CD3- lymphocytes with activation markers (CD69, CD25, CD71, HLA-DR, and CD38) were less expressed in the decidua than those in the endometrium. The CD69’CD3and CD25’CD3lymphocytes increased significantly in the decidua of
Subpopulations and activation of NK cells in the decidua and their cytotoxicity in normal and anembryonicpregnancies. The distribution of CD16’CD56’ lymphocytes was similar in PBMC and decidua, although most lymphocytes in TABLE
3
Subtypes of NK cells in PBMCs and decidua pregnancies and anembryonic pregnancies” Normal pregnancies (n = 22) PBMCs
NK CD16-CD56’ CD16+CD56’ CD16’/CD56CD16+CD56+/CD56+ CD57’CD56’ CD57-CD56’ CD57+CDSb’/CD56+
15.5 11.7 11.4 1.0 46.0 10.9 11.1 50.0
* + f f + f f f
7.5 5.6 8.4 0.9 26.0 5.3 4.9 14.6
Deciuda 62.8 59.6 8.7 0.7 15.4 6.4 62.4 11.6
+ f i t * k i +
21.0” 23.1” 4.7 0.6 11.5* 4.0” 22.8” 9.4b
of normal
Anembryonic pregnancies (n = 20) PBMCs 17.6 11.4 10.1 1.1 46.5 10.2 9.2 52.0
+ 8.2 + 6.1 Yk6.8 i 1.3 f 25.3 f 4.2 + 3.4 + 9.5
y Expressed as percentage of positive cells r SD on gated lymphocyte population. ’ p < 0.05 compared to peripheral blood lymphocytes.
Decidua 69.8 68.3 8.3 0.9 12.3 5.3 70.9 7.8
+ + + i i + f +
17.1/’ 18.Bh 4.3 1.2 9.6” 3.7h 17.4” 6.9’
134
H.-N. Ho et al.
anembryonic pregnancies compared to normal pregnancies. The NK cytotoxicity assay showed that decidual NK activity was significantly less than that of the peripheral blood activity in normal pregnancies (p = O.OOl), but significantly increased in anembryonic pregnancies (t, = 0.006) (Fig. 1).
DISCUSSION Our present study has shown that the nonpregnant endometrium is infiltrated by cells with phenotypes similar to those seen in the decidua, and there is no clearly discernible increase in their proportion from the proliferative to late luteal phase (which is the time the blastocyst would implant). It is impressive that there are almost no B cells in the uterus and there is a gradual increase in the percentage of B cells from the proliferative endometrium through secretory endometrium to decidua. However, the decrease of T cells and the increase of NK cells in the endometrium are noted only during the shifting from late luteal phase to early pregnancy. Nevertheless, the presence of these cells in the nonpregnant endometrium and the variation in their number in relation to the phases of the menstrual cycle suggest that their recruitment may be under hormonal control as well as a result of trophoblast stimulation. There are numerous bone marrow-derived cells in the human endometrium and early pregnancy decidua, which include tissue macrophages, NK cells and T cells. The functions of these endometrial and decidual leukoFIGURE 1 NK cytotoxicity (lytic unit,,) in the peripheral blood (0, Cl) and decidua (0, n ) of women with normal and anembryonic pregnancies. The means are indicated by a dash in each column. m
1
Normal Pregnancy (P=O.o013)
Anembryonic Pregnancy (P=O.W63)
cytes are not known. CD56 is expressed by lo-15% of peripheral blood lymphocytes, which are morphologically LGLs, mediating non-major histocompatibility complex-restricted cytotoxicity 1121. The largest population during the first trimester of pregnancy is CD1 6- or CD56-positive/CD3-negative NK cells. The proportion of NK cells in the endometrium further increased after normal and anembryonic, but not ectopic pregnancies. This might indicate that the further recruitment of NK cells in the early decidua is due to local stimulation by fetal antigens. The preparation of cells from human decidua has been reported to have NK activity which is confined to the subset of CD56-positive NK cells, which is also CD3 and CD16 negative [13}. This corresponds to a minor subset of PBMCs with only weak NK activity {S]. These findings are essentially consistent with our previous [6, 71 and present observations. This selective migration into the uterus does suggest some important role for these cells during early pregnancy. In peripheral blood, most CD56-positive cells are also CD16 positive and have high NK activity, which could explain the lower cytotoxic activity of decidual NK cells than that of PBMCs. The fraction of CD16’CD56’ lymphocytes is similar in both PBMC and decidua, whereas there is a marked increase of CD16-CD56’ lymphocytes in the decidua. This condition is also present in anembryonic pregnancies, although the suppression of decidual NK cytotoxicity disappears in the anembryonic pregnancies. Moreover, the CD16’CD56’iCD56’ ratio in the decidua is also similar in normal and anembryonic pregnancies. Therefore, the increase in NK cytotoxicity of anembryonic pregnancies could not be explained merely by the increase in CD16’CD56’ NK subpopulation. Compared with PBMCs, the ratio of CD4’CD3’ to CDS’CD3 + lymphocytes is inverted in the endometrium and decidua, although both the percentages of CD4’CD3’ and CDg’CD3’ lymphocytes decreased in the decidua. Whether the marked decrease of CDS’CD3’ lymphocytes in the decidua compared with that in the nonpregnant endometrium is necessary for maintaining the pregnancy needs further investigation. A significantly higher proportion of NK cells and an inverted CD4’CD3’/CDS’CD3’ ratio possibly reflect the change in local endometrial environment. After implantation of an embryo and during pregnancy, lymphocyte subpopulations alter with the stimulation by fetal antigens, and NK cells increase in number but decrease in cytotoxicity. Evidence of T-cell activation in human endometrium has been reported {I4-161. Our results confirm the reports that endometrial T lymphocytes are activated by increasing the expression of early and late T-cell activation markers. The expression of the CD69 antigen (an early activation marker for T and NK cells) has been
Activated T and NK in Endometrium
135
and Decidua
reported to appear shortly after stimulation by activation of the protein kinase C (PKC) pathway 117, 181. Zeigler et al. 1191 showed that the induction of CD69 mRNA in activated murine thymocytes and T cells is very rapid and usually precedes the interleukin (IL)-2cx receptor (CD25) expression. Our finding indicates that the T cells in the endometrium and decidua have already been stimulated via PKC. The HLA-DR antigen (a relatively late activation T-cell marker) was detected in about 30% of T cells in the peripheral blood lymphocytes of nonpregnant and pregnant women. In contrast, in the endometrium and decidua, most T cells showed HLA-DR antigen expression. Thus, we show here that endometrial and decidual T cells express an early activation antigen, CD69, and a late activation antigen, HLA-DR. Because the summation of the percentage of T lymphocytes expressing CD69 and HLA-DR greatly exceeded 100%; there may be a coexpression of early and late antigens on many T cells. Therefore, the endometrial and decidual T lymphocytes may be in a state of recent and persistent activation. In relation to this, it is interesting to note that the level of the CD25 antigen expression is downregulated in the nonpregnant endometrium, but its expression increases after pregnancy. The increased expression of T-cell-surface markers in the sequence CD25 and CD71 after pregnancy may be the result of the stimulation by fetal antigens, as previously described 120). In this study, CD25’CD3’ and CD7 1 ‘CD3 + subpopulations are increased in the decidual of normal pregnancies compared with those in nonpregnant endometrium. However, the CD25’CD3’ subpopulation in the decidua of ectopic pregnancies is less than that in normal pregnancies, and it return to a level similar to that in the nonpregnant endometrium. This observation further strengthens the suggestion that the increase in the decidual CD25’CD3’ subpopulation is due to local stimulation by fetal antigens. It has been reported that anti-CD8 treatment of pregnant mice reduces placental proliferation and phagocytosis, and that anti-CD4 antibody treatment affects placental proliferation 121). Therefore, it would appear that activated T cells may be involved in placental cell proliferation and function. Since trophoblast produces an IL-2-like molecule and expresses IL-2 mRNA at an early stage of pregnancy 122, 231, it is possible that this factor may activate T cells. However, it is also possible that other antigen-presenting cells in the decidua are involved in T-cell activation 1241. In the decidua, there are fewer B than NK cells (7.0 + 7.0% vs. 62.8 + 21.0%), and it might be reasonable to consider that most of the CD3- cells were NK cells. Therefore, the data in this study indicate that the NK cells in the decidua were activated (Table 2). However, the factors leading to activation of decidual NK cells in
viva and the function of such activated cells are still not known. It is possible that decidual NK cells might be expected to play a role in regulating placental growth and development through the production of cytokines or other factors, such as transforming growth factor B 125, 261 and thereby control the invasive growth of the trophoblast by direct cytotoxic activity or by regulating local maternal immune responses to trophoblasts 1271. It has been demonstrated that the NK activity of decidual NK cells against the standard NK target cell line K562 is low compared to that of peripheral blood NK cells 171. However, this difference was not present in anembryonic pregnancies (Fig. 1) and the proportion of NK-cell activation (CD69’CD3and CD25’CD3-) increased in these pregnancies. Whether the increase in this activated NK-cell subpopulation correlated with the augmentation of NK cytotoxicity needs further evaluation. In summary, our findings show that activated NK cells are present in the decidua. In addition, a significant proportion of decidual CD3’ cells also expresses CD25 and CD71, indicating that these cells are activated, although the mechanism of activation is unknown. Further studies are required to investigate whether and to what extent these cell types play a role in controlling trophoblast growth and placental development.
ACKNOWLEDGMENTS
The authors extend their sincere gratitude
to Professor. T. J.
Gill III, from the University of Pittsburgh, for critical discussion of the manuscript. This study was supported partially by grants from the National Science Council of the Republic of China (NSC84-2331-B002-287 and NSC85-2331-B002-113) and the Research Program of National Taiwan University Hospital (NTUH-84241-B24 and NTUH-84242-B16).
REFERENCES 1. Bulmer JN, Sunder-land CA: Immunohistological characterization of lymphoid cell population in the early human placental bed. Immunology 52:349, 1984. 2. Bulmer JN, Johnson PM: Immunohistological characterization of the decidual leucocytic infiltrate related to endometrial gland epithelium in early human pregnancy. Immunology 55:35, 1985. granulocytes in hu3. Ritson A, Bulmer JN: Endometrial man decidua react with a natural-killer (NK) cell marker, NKHl. Immunology 62:329, 1987. 4. Starkey PM, Sargent IL, Redman CWG: Cell population in human early pregnancy decidua: characterization and isolation of large granular lymphocytes by flow cytometry. Immunology 65:129, 1988. 5. Lanier LL, Le AM, Civin CI, Loken MR, Phillips JH: The relationship of CD16 (Leu-11) and Leu-19 (NKH-1) an-
136
H.-N.
tigen expression on human peripheral cytotoxic T lymphocytes. J Immunol
blood NK cells and 136:4480, 1986.
6. Chen CK, Huang SC, Chen CL, Yen MR, Hsu HC, Ho HN: Increased expressions of CD69 and HLA-DR but not CD25 or CD71 on endometrial T lymphocytes of nonpregnant women. Hum Immunol 42:227, 1995. 7. Chao KH, Yang YS, Ho HN, Chen SU, Chen HF, Dai HJ, Huang SC, Gill TJ III: Decidual natural killer cytotoxicity decreased in normal pregnancy but not in anembryonic pregnancy and recurrent spontaneous abortion. Am J Reprod Immunol 34:274, 1995. 8. Saito S, Nishikawa K, Morii T, Narita N, Enomoto M, Ichijo M: Expression of activation antigens CD69, HLADR, interleukin-2 receptor-alpha (IL-2Ra) and IL-2RB on T cells of human decidua at an early stage of pregnancy. Immunology 75:710, 1992. 9. Ho HN, Hultin LE, Mitsuyasu RT, Matud JL, Hausner MA, Bockstoce D, Chou CC, O’Rourke S, Taylor JMG, Giorgi JV: Circulating HIV-specific CD8’ cytotoxic T cells express CD38 and HLA-DR antigens. J Immunol 150:3070, 1993. 10. Ho HN, Chao KH, Chen HF, Wu MY, Yang YS, Lee TY: Peritoneal natural killer cytotoxicity and CD25’CD3’ lymphocyte subpopulation are decreased in women with stage III-IV endometriosis. Hum Reprod IO:267 1, 1995. 11. Pross HF, Baines MG, Rubin P, Shragge P, Patterson MS: Spontaneous human lymphocyte-mediated cytotoxicity against tumor target cells. IX. The quantitation of natural killer cell activity. J Clin Immunol 1:51, 1981. 12. Lanier LL, Le AM, Cwirla S, Federspiel N, Philips JH: Antigeneic, functional and molecular genetic studies of human natural killer cells and cytotoxic T lymphocytes not restricted by the major histocompatibility complex. Fed Proc 4512823, 1986. 13. Ferry BL, Starkey PM, Sargent IL, Watt GMO, Jackson M, Redman CWG: Cell populations in the human early pregnancy decidua: natural killer activity and response to interleukin-2 of CD56-positive large granular lymphocytes. Immunology 70:446, 1990. 14. Pace D, Morrison L, Bulmer JN: Proliferative activity in endometrial stromal granulocytes throughout menstrual cycle and early pregnancy. J Clin Path01 42:35, 1989. 15. Tabibzadeh S: Proliferative activity of lymphoid human endometrium throughout the menstrual Clin Endocrinol Metab 70~437, 1990.
cells in cycle. J
Ho et al.
16. Tabibzadeh S: Evidence of T-cell activation and potential cytokine activation in human endometrium. J Clin Endocrinol Metab 71:645, 1990. 17. Testi R, Phillips JH, Lanier LL: T cell activation 23 (CD69). J Immunol 143:1123, 1989. 18. Lanier LL, Burk DW, Rhodes L, Ding A, C, Phillips JH: Interleukin 2 activation cells rapidly induces the expression and of the Leu 23 activation antigen. J Exp 1988.
via Leu-
Evans E, Barney of natural killer phosphorylation Med 167:1572,
19. Zeigler SF, Levin SD, Johnson L, Copeland NG, Gilbert DJ, Jenkins NA, Baker E, Sutherland GR, Feldhaus AL, Rawdsdell F. The mouse CD69 gene, structure, expression and mapping to the NK gene complex. J Immunol 152:1228, 1994. 20. Ho HN, Chen HF, Chen SU, Chao KH, Yang YS, Huang SC, Lee TY, Gill TJ III. Gonadotropin releasing hormone (GnRH) agonist induced down-regulation of the CD3’CD25’ lymphocyte subpopulation in peripheral blood. Am J Reprod Immunol 33:243, 1995. 21. Athanassakis I, Chaouat G, Wegmann TG: The effects of anti-CD4 and anti-CD8 antibody treatment on placental growth and function in allogeneic and syngenic murine pregnancy. Cell Immunol 129:13, 1990. 22. Soubiran P, Zapitelli JP, Schaffar L: IL-2-like material is present in human placenta and amnion. J Reprod Immuno1 12:225, 1987. 23. Boehm KD, Kelley MF, Ilan J, Ilan J: The interleukin 2 gene is expressed in syncytiotrophoblast of human placenta. Proc Nat1 Acad Sci USA 86:656, 1989. 24. Oksenberg JR, Mor-Yosef S, Persity E, Schenker Y, Mozes E, Brantbar C: Antigen-presenting cells in human decidual tissue. Am J Reprod Immunol Microbial 11:82, 1986. 25. Dungy LJ, Siddigi TA, Khan S: Transforming growth factor-P1 expression during placental development. Am J Obstet Gynecol 165:853, 1991. 26. Saito S, Nishikawa K, Morii T, Enomoto M, Narita N, Motoyoshi K, Ichijo M: Cytokine production by CD16mCD56b”“h’ natural killer cells in the human early pregnancy decidua. Int Immunol 5:559, 1993. 27. King A, Loke YW. On the nature and function of human uterine granular lymphocytes. Immunol Today 12:432, 1991.