Characterization of immune effector cells present in early murine decidua



93, 303-3 14 (1985)

Characterization of Immune Effector Cells Present in Early Murine Decidua’ P. GAMBEL,*,~B. A. CROY,~,~W. D. MOORE,?R. D. HUNZIKER,* T. G. WEGMANN,* AND J. ROSSANT*,~ *Department of Immunology and MRC Group on Immunoregulation, University of Alberta, Edmonton, Alberta, T6G 2H7, Canada, and tDepartment of Biological Sciences, Brock University, St. Catharines, Ontario, L2S 3A1, Canada Received February 5, 1985; accepted March 2, 1985 It has been suggested that murine decidual cells act as an important immunoregulatory population localized to the pregnant uterus. We have examined early murine decidua to determine if immune effector cells occur in the decidual environment in proximity to the conceptus. High levels of natural killer (NK) cell activity were found consistently in decidual cell suspensionswith peak activity occurring on Day 6.5 of gestation. NK activity declined as pregnancy proceeded and was not significant by Day 12.5 of gestation. Decidual cell suspensions did not appear to contain significant numbers of functional B or T effector cells. No antipaternal T-cell response could be demonstrated even in the decidua of immune mice. Lack of T-cell responseswas attributed to the absenceof T cells from decidua rather than to their inactivation because precursors of cytotoxic T lymphocytes (pCTL) could not be detected in decidual cell suspensions.Furthermore, the levels of pCTL detectable in spleen cell suspensionscould not be reduced by mixing spleen cells with 7.5day decidual cells. These results suggestthat B cells and T cells may not occur in early decidua while NK cells are present and regulated independently. 0 1985 Academic press, IIIC.

INTRODUCTION Uterine decidual cells play an important role in implantation and maintenance of mammalian pregnancy. They arise from stromal cells of the uterine endometrium which proliferate and differentiate under the influence of ovarian hormones in response to the implanting blastocyst (1, 2) or certain artificial stimuli (3). The ultimate origin of decidual cells remains disputed but it has been suggestedthat some decidual cells are bone marrow-derived (4-6). The biological functions of decidua may include nutrition for the embryo, secretion of pregnancy-associated hormones, protection of the mother from excessive trophoblast invasion, and protection of the embryo from maternal immune rejection (1, 2, 7-9). There is ’ This work was supported by grants from the Medical ResearchCouncil of Canada to J.R. and T.G.W. The Alberta Heritage Fund for Medical Research provided a postdoctoral fellowship to P.G., a graduate studentship to R.D.H., and a visiting scientist award to J.R. J.R. is also the recipient of an E. W. R. Steacie Fellowship from Natural Sciencesand Engineering Research Council, Canada. 2 Present address: Hospital Cochin, 27 rue du Faubourg Saint Jacques, 75674, Paris, Cedex 14, France. 3 To whom correspondence should be sent. 303 0008-8749185$3.00 Copyright 0 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.




direct and indirect evidence that decidual cells can regulate immune responses. Primary skin allografts survive longer in a pregnant or pseudopregnant uterus than in either a nonpregnant uterus or an orthotopic site (10). In presensitized hosts, however, the rapid rejection of intrauterine skin allografts cannot be prevented by decidua. Decidual tissues or their extracts have also been shown to suppress mixed leukocyte culture reactivity, in vitro generation of cytotoxic T cells (CTL) by adult splenocytes, and in vivo infiltration of sponge matrix allografts by CTL (1 l-l 5). The relevance of immune suppressor mechanisms to pregnancy immunoregulation in vivo is unclear because functional immune effector cells have never been demonstrated in the vicinity of a normal fetus. A number of potential immune effector cells have been identified in decidua, including FcR-positive cells, metrial gland granulocytes, and macrophages (2, 16) but the immunocompetency of decidual cells has not been thoroughly investigated. This study evaluates the functional activity of B, T, and NK cells in early murine decidua. MATERIALS


Mice. Random bred CD1 mice (Charles River, St. Constant, Quebec) were used in all experiments, unless otherwise indicated, to optimize the yield of decidual cells. C3H/HeNCrlBR, DBA/2NCrlBR, C57BL/6NCrlBR and BALB/cAnCrlBR mice were also purchased from Charles River. Female mice were selected for estrus and paired with males overnight. The presence of a vaginal plug the following morning indicated mating and that morning was counted as Day 0.5 of pregnancy. Pregnant females were killed at various days between 6.5 and 12.5 of gestation. Immunizations. C3H female mice were immunized by biweekly intraperitioneal (ip) injections of 2 X 10’ irradiated (1500 rads) spleen cells. Mice immunized against BALB/c received three injections, while mice immunized against DBA/2 received two injections. Preparation of single cell suspensions. Mice were killed by cervical dislocation and spleens and uteri were dissected and placed in Leibowitz medium (GIBCO, Grand Island, N.Y.). Spleens were pressed through 60-gauge stainless steel mesh and the resulting cell suspensions were washed, subjected to erythrocyte lysis with double distilled water, filtered through Nitex gauze (Tetko Inc., Elmsford, N.Y.), and resuspendedin RPM1 1640 (Flow Laboratories, Mississauga,Ont.) supplemented with 10% fetal bovine serum, 100 I.U./ml penicillin, 0.05 mg/ml streptomycin, 20 mM Hepes, and 5 X 10e5 A4 2-mercaptoethanol. Decidual tissue free from fetal contamination was prepared by dissecting decidual nodules out of the uterus and removing the entire conceptus by further dissection under a dissecting microscope. The decidual tissue was then minced with fine scissors in a solution of 1 mg/ml collagenase(Worthington Type IV, Flow Laboratories) and 25 pg/ml DNase (Sigma, St. Louis, MO.) in Leibowitz medium, using 1 ml of the solution per decidual capsule. The tissue suspension was incubated for 45 min in a shaking water bath at 37°C and the cells were then dispersed by vigorous pipetting and recovered following passagethrough Nitex. Erythrocytes were lysed and the decidual cells were washed three times, suspended in culture medium, and counted, using trypan blue dye exclusion to assessviability. In some experiments spleen cells were incubated 45 min in the collagenase-DNase solution, washed, enumerated, and used as control cells.








In vitro assays i. Mixed leukocyte cultures. Responder cells (2.5 X 105),from either spleen or decidua, were mixed with 2.5 X lo5 irradiated syngeneic or allogeneic spleen cells (1500 rads) in 0.2 ml supplemented RPM1 medium in U-bottom microtest plates and incubated at 37°C in 5% CO2 in a humidified incubator. After a 72-hr incubation, 0.5 mCi 3H-TdR (New England Nuclear, Boston, Mass.) was added for 16 hr. The cultures were then harvested (Titertek, Flow Laboratories) and assessed for isotope incorporation by scintillation counting. The mean of four to six replicate cultures was determined and used to calculate the stimulation index (SI) by dividing the mean cpm in the allogeneic mixtures by that in the syngeneic mixtures. Tube cultures were employed to sensitize cells for the 5’Cr-release assay. These cultures contained 3 X lo6 responder cells and 1 X lo6 irradiated splenocytes in 3 ml medium and were incubated for 108 hr. For limiting dilution analysis 50-500 responder cells from CD1 mice were cultured with 2.5 X lo5 irradiated DBA/2 splenocytes in medium containing 10% interleukin 2 (IL-2, Cell Products Inc., Buffalo, N.Y.) for 7 days. Control cultures contained irradiated DBA/2 spleen cells in medium containing 10% IL-2. Ninety-six replicates were tested in each experimental group. ii. “Cr-release assays. Decidual cells and spleen cells, either freshly prepared or after in vitro sensitization, were incubated in 0.2 ml medium with “Cr-labeled targets in V-bottom microtest plates for 6 hr. Targets included P8 15 (EI-2d), L 12IO (H-Zd), RI (H-2k) and YAC (H-2’) cells that had been labeled for 90 min with NaZ5’Cr04 (New England Nuclear; 1 mCi/ml). In most experiments 3 X lo6 targets were employed and effector:target ratios (E:T) ranged from 1OO:l to 5:l. After incubation, 0.1 ml of supernatant was removed from each culture and counted in a gamma detector for releasedisotope. Percentagespecific 51Crreleasewas calculated using the standard formula (17). In the limiting dilution experiments 5 X lo2 target cells were employed and isotope release from individual cultures was compared to the mean isotope release from 96 control cultures. Cultures were considered to be positive if the isotope release exceeded the control mean plus three standard deviations (18). iii. Mitogen stimulation. Cells (1 X 105) were cultured in U-bottom microtest plates with mitogen as follows: pokeweed mitogen (PWM; Sigma) loo-O.1 yg/ml; lipopolysaccharide (LPS; Difco, Detroit, Mich.), 500-0.25 pg/ml; phytohemagglutinin (PHA; Wellcome, Greenville, N.C.) dilutions of l/6-1/1 536; or concanavalin A (Con A; Difco) 96-1.5 pg/ml. Cultures were pulsed with 3H-TdR at 48 hr and harvested for scintillation counting after an additional 16-hr culture interval. Antiserum treatment. Up to 2 X lo7 cells from decidua or spleen were suspended in 1 ml of Leibowitz medium containing a dilution of 1:5000 anti-Thy 1.2 monoclonal antibody (New England Nuclear) or 1:25 anti-asialo GM1 (Wako Chemicals, Dallas, Tex.) and incubated at room temperature for 45 min. After washing, the cells were resuspended in 1 ml of 1:10 dilution of rabbit Low-Tox M complement (Cedarlane Laboratories, Hornby, Ont.) in Leibowitz medium and incubated at 37°C for 45 min. The cells were washed twice, resuspended in culture medium and counted. Colony-forming units-spleen (CFU-S) assay. C57BL/6, C3H, and DBA/2 female mice were given 950 rad gamma irradiation then injected intravenously with



1.0-25 X 10’ syngeneic cells of bone marrow or decidual origin. Spleens were examined for hematopoietic colony formation on Day 12 or following death (19). RESULTS Are Natural Killer (NK) Cells Present in Decidua? To test for the presence of NK cells in early murine decidua we examined decidual cell suspensions, prepared from the decidua of CD1 mice on Days 6.5 to 12.5 of gestation, for their ability to lyse YAC target cells. Significant levels of YAC cell lysis were observed in cells from all stages of pregnancy between 6.5 and 9.5 days. The highest level of lytic activity was observed at Day 6.5, which is the earliest stage at which a reasonable amount of decidual tissue can be cleanly isolated. The activity declined steadily from that at Day 6.5 and by Day 12.5 no significant level of YAC killing was observed (Fig. 1). The total yield of decidual cells per implantation site also declined as pregnancy progressed (Table 1). Considerable variation in the activity of decidual cell preparations was observed in different experiments, which may reflect variation in the relative proportion of bone marrowderived cells in the final cell suspension. We have shown elsewhere (5) that collagenase digestion of decidua preferentially selects for bone marrow-derived cells and that the proportion of non-bone marrow-derived cells varies with the extent of the digestion. The lytic activity of spleen cells against YAC targets was not altered significantly by incubation of the splenocytes in collagenase-DNase, (data not shown) suggesting that the method of cell preparation did not alter the functional characteristics of the cell populations under investigation. For preliminary characterization of the lytic cells in decidua, 7.5-day decidua from CD1 mice were employed to obtain the highest possible yields of cells with significant NK activity. Cell suspensions were subjected to complement-mediated antibody lysis using anti-asialo GM1 or anti-Thy 1.2 (20). Treatment of decidual cells with anti-asialo GM1 and C’ greatly reduced the capacity of the cells to lyse YAC targets while treatment with anti-Thy 1.2 + c’ removed some but not all of the lytic activity (Table 2). Both treatments significantly reduced the capacity of splenocytes to lyse YAC targets. Treatment of decidual cells with both antisera consecutively was not possible, due to limitations in cell numbers.







II 5

III 12 5


FIG. 1. Time course of decidual NK cell activity during murine pregnancy.




Cell Yields from CD1 Decidual Capsules following Collagenase-DNase Treatment Day of pregnancy

Number of dissected decidual capsules

Mean cell yield per capsule (X 105)

6.5 7.5 8.5 9.5 10.5 12.5

32 195 103 77 51 34

2.5 3.0 2.0 1.9 2.0 0.7

Are Mixed Lymphocyte Reaction (MLR)-Responsive Cells Present in 7.5-Day Decidua? To assessthe capacity of decidual cells to recognize third party alloantigens, MLR cultures were established using well-washed 7.5day CD1 decidual cells as the responding cell population. Irradiated spleen cells from DBA/2 or CD1 served as the allogeneic and syngeneic stimulator cells, respectively. Control cultures consisted of mixtures of spleen cells, assayed within the same experiment. No significant response was observed in allogeneic cell mixtures containing decidual cells as responders in three of four experiments while significant responses were observed in the allogeneic spleen cell cultures in all of the experiments (Table 3). In the remaining experiment a very weak response by decidual cells was observed but the absolute values of the mean cpm were so low that the response may not be meaningful. Such a response was not observed in any subsequent experiments. TABLE 2 Characterization of Cell Surface Markers on Decidual NK Cells Mean % “Cr release f SEM from YAC targets at E:T Source of effector cells







7.5-Day decidua 7.5Day decidua

Anti-asialo GM1 + C

19.4 f 1.1 6.8 f 0.5

13.4 iz 1.0 3.2 rt 2.0

8.5 k 1.5 3.0 + 1.6

0.4 +- 0.6 n.t.

7.5-Day pregnant spleen 7.5-Day pregnant spleen 7.5-Day prengant spleen

None C Anti-asialo GM1 + C’

13.1 f 2.2 19.9 + 1.4 4.5 f 1.5

8.7 f 1.2 11.8 f 2.0 2.4 + 0.7

2.7 * 0.3 7.6 + 0.7 0

2.5 2 0.3 3.2 + 0.1 n.t.

Nonpregnant spleen Nonpregnant spleen Nonpregnant spleen

None C Anti-asialo GM1 + C’

25.9 + 3.7 27.9 + 2.6 7.2 f 2.5

24.3 f 2.2 17.5 2 1.8 3.8 f 0.2

17.2 ?I 0.8 12.0 +- 0.8 0.6 I 1.4

5.0 + 0.8 7.1 rt 0.9 1.3 + 0.6

7.5-Day decidua 7.5-Day decidua

C Anti-Thy 1.2 + C

27.6 + 1.6 23.6 f 1.6

29.6 -c 0.5 23.3 f 0.6

23.0 +- 1.0 16.5 +- 3.0

15.5 f 1.5 10.3 k 1.9

7.5-Day pregnant spleen 7.5-Day pregnant spleen

C Anti-Thy 1.2 + C

26.9 f 0.5 16.9 f 1.8

18.4 +- 1.6 11.2 f 1.7

2.1 + 4.4 4.6 ?I 2.1

2.1 f 4.6 0


GAMBEL ET AL, TABLE 3 Examination of Decidual Cells as the Responder Cell Population in the MLR Response of 7.5-day decidual cells



I 2 3 4


X X x X


5 6


Allogeneic”” Syngeneic” (cpm X 103) (cpm X IO’) 0.13 -c 0.03 0.17 f 0.05 0.13 + 0.02 0.6 f 0.2

0.11 f 0. I2 f 0.12 + 0.2 +

0.01 0.03 0.02 0.1

0.5 f 0.09 5.2 f 0.9

0.5 f 0.1 4.4 + 0.9


Response of spleen cells Allogeneic” Syngeneic” (cpm X 103) (cpm X 103)

1.1 1.3 I.1 2.9”

8.5 f 8.9 f 7.1 k 40.7 f

1.0 1.2 0.6 1.4

1.0 1.2

35.6 + 8.6 7.3 * I.5

0.95 + 0.8 k 1.2 f 5.8 f


0.2 0.2 0.2 0.6

8.9* 11.1* 5.0* 6.9*

4.5 + 0.9 2.7 + 0.9

7.9* 2.6+

’ Values represent means of 6 replicate cultures + SEM. b In Experiments l-4 stimulator cells were irradiated third party DBA/2 splenocytes; in Experiments 5 and 6 stimulator cells were irradiated paternal BALB/c splenocytes. * P < 0.005.

Viable decidual cells were present in all MLR cultures from the day of their initiation until the day of harvest (Moore, unpublished data). Thus it appeared that cells capable of a proliferative responsefollowing recognition of third party alloantigens were absent or nonfunctional in decidua. Further experiments were conducted to determine if decidua might contain cells capable of an antipaternal immune response. C3H mice were mated to BALB/c males and decidual cell suspensions were prepared on Day 7.5 of gestation and tested for their ability to respond to BALB/c spleen cells in the MLR. In two replicate experiments no significant proliferative response was observed in the cultures containing decidual cells as responders. In both experiments, cultures containing splenocytes as responders gave the expected positive response (Table 3). Thus, 7.5-day decidua also lack cells capable of responding to paternal alloantigens. Are CTL Present in 7.5Day Decidua? The absence of cells capable of a response in the MLR does not preclude the possibility that functional CTL, possibly resulting from paternal antigen presentation at the time of mating, could be present in 7.5-day decidual cell suspensions. For these studies it was necessaryto use inbred strains of mice. Smaller litter sizes and increased frequency of nonproductive matings in these mice, compared with CD1 mice, limited the amount of tissue available in all experiments. In two experiments we tested freshly prepared C3H decidual cells, resulting from matings to BALB/c males, for their ability to lyse the paternal antigen-bearing targets P8 15 or L 1210. Control cultures in both experiments consisted of decidual cells mixed with “Crlabeled YAC cells and in one experiment sufficient cells were available to test for lytic activity against the maternal strain target RI. C3H decidual cells failed to lyse the paternal or maternal strain targets while C3H splenocytes, sensitized to BALB/c cells in vitro, lysed the H-2d targets (Table 4). In both experiments YAC cells were lysed by decidual cells and splenocytes, showing that the absenceof specific cytotoxic cells was not due to elimination of functional lytic cells from decidua by the cell




Examination of 7.5-Day Decidual Cells from Nonimmune and Immune Mice as Effecters of Antipatemal CML Mean % %

release f SEM at E:T ratios of

Target cells





P815 RI

0 0

0 0

0 0

0 0

P815 L1210 YAC

0 0 11.1 -c 1.2

3.5 + 2.7* 0 9.8 -c 0.9

0 0 2.5 f 0.5*

0 0 5.7 f 0.9


4.3 f 3.2* 0 20.9 f 0.4

4.1 f 1.9* 0 16.8 f 2.0

4.4 f 1.7* 0 9.7 f 0.6

0 0 5.3 f 0.5

Decidua, preimmune C3H X BALB/c mating

P815 L1210

2.7 Ifr l.l* 3.3 + 0.6*

0 0

0 0

0 0

MLR C3H anti-BALB/c

P815 L1210

51.4 f 1.5 40.3 + 4.5

50.9 f 1.0 26.5 + 1.9

31.3 f 1.4 15.1 + 2.6

4.5 + 0.8 0

Decidua, preimmune C3H X DBA/Z mating MLR, preimmune C3H anti DBA/2 MLR, nonimmune C3H anti DBA/Z




P815 P815

87.2 + 1.8 46.9 2 3.7

84.7 2 1.6 36.9 f 1.6

Source of effector cells Decidua; CD1 X CD1 mating Decidua; C3H X BALB/c mating Expt 1

Expt 2

4.8 + 2.6* 81.0 f 1.2 28.1 + 1.7

0 72.8 f 1.0 20.0 k 0.9

* Values not significantly different from 0; P > 0.005.

preparation procedures. These experiments suggestedthat functional CTL were not present in early murine decidua. Further, the absence of lysis of the NK insensitive tumor cells P8 15, L 1210, and RI by decidual cells capable of lysing YAC targets supported the conclusion that the lytic cells present in decidua were NK cells. To investigate further whether antipaternal CTL could ever be detected in decidua, we preimmunized C3H females with irradiated spleen cells from BALB/c or DBA/2 and then mated them to males of the immunizing strain. Decidual cells recovered 7.5 days after mating were tested for their ability to lyse P8 15 target cells (Table 4). Preimmunization did not result in the presence of detectable levels of functional CTL in the decidual cell suspensions. Inability to detect CTL was not due to failure of the assay, since P8 15 targets were lysed by splenocytes sensitized in vitro, or due to failure of the immunization procedures, as spleen cells from the immunized mice gave a secondary CTL response when exposed to the paternal antigens in vitro (Table 4). Thus functional CTL appear to be absent from early decidua in both immune and nonimmune mice. Are CTL Precursors (pCTL) Present in Decidua? Absence of functional T cells from decidua could result from the inactivation (suppression) of T cells or from their absence or active exclusion from the tissue. We attempted to distinguish between these possibilities by enumeration of pCTL (18) in decidua and determining whether mixing of decidual cells with spleen cells




at limiting dilutions reduced the number of pCTL detectable in the splenocytes. Cultures of spleen cells but not decidual cells gave rise to significant numbers of CTL when cultured with irradiated allogeneic cells and IL-2 (Fig. 2). The estimated frequency of pCTL in spleen was l/3400 ( 18). Mixing of decidual cells with spleen cells did not reduce the number of pCTL detected in spleen; indeed, mixing raised the number of pCTL detectable in cultures having the lowest numbers of spleen cells. This suggeststhat the lack of cells in 7.5-day decidua that are capable of responding in the MLR and CTL assaysis more likely due to the absence of T cells from decidua than to T-cell suppression. Are Cells Responsive to B- and T-Cell Mitogens Present in Early Decidua? Cells derived from bone marrow are known to be present as a minority population in decidua by GPI, PGK, and H-2 marker studies of bone-marrow chimeric mice (4-6). To determine whether any of these cells respond to polyclonal activation, 7.5-day CD 1 decidual cell suspensions were tested for their ability to respond to B- and T-cell mitogens. The results of these studies are summarized in Fig. 3 and show that no stimulation of decidual cells occurred in the presence of Con A or PHA although control spleen cell cultures were activated. PWM, a B-cell activator, stimulated spleen cells but not decidual cells in four of four experiments. LPS stimulated spleen cells very strongly (SI > 50) and, when employed at very high doses( 125 pg/ml), produced weak but statistically significant stimulation of decidual cells in two of two experiments (SI = 3 and 6). To determine if this response to LPS was occurring in one subpopulation of decidual cells, 7.5-day CD1 decidual cell suspensions were separated into fractions of different cell sizes by velocity sedimentation (21) and each cell fraction was tested for its response to the high dose of LPS. Significant levels of stimulation were not observed in any fraction (data not shown). Thus it appeared that insignificant numbers of cells were present in decidua that were responsive to polyclonal B- and T-cell activators.



FIG. 2. Cytotoxic responses of CD1 spleen cells (O), 7Sday CD1 decidua (A), and mixtures of spleen cells plus equal numbers of CD1 decidual cells (0) after 7 days of culture with irradiated DBA/2 spleen cells in 10% IL-2.












FIG. 3. Responses of spleen cells (0) and decidual cells (0) to mitogens. Relative concentration indicated as 1 is 125 &ml for LPS, 100 rg/ml for PWM, i dilution for PHA, and 96 pg/ml for Con A.

Are CFU-S Present in 7.5-Day Decidua? Hematopoietic stem cells are progenitors of both the hematopoietic and lymphopoietic systems (22) and are cells of bone marrow origin. CFW-S were assayedto determine if hematopoietic stem cells were present in 7.5-day decidua. Bone marrow and decidua from inbred C57BL/6, C3H, or DBA/2 mice were prepared and injected iv into irradiated syngeneic recipients. As shown in Table 5, CFU-S were not detected in decidual cell suspensions from C3H or DBA/2 mice and the recipients died. Of the four C57BL/6 recipients receiving decidual cells three died and one survived. The spleen of the surviving mouse contained numerous very small foci of proliferating cells and resembled a spleen undergoing endogenous repopulation. In all of the experiments mice receiving syngeneic bone marrow survived to Day 12 and had detectable spleen colonies. DISCUSSION The purpose of these studies was to determine if early murine decidua contained any immunocompetent effector cells that could be potentially harmful to the fetus.


GAMBEL ET AL. TABLE 5 Assay for CFU-S in 7.5-Day Decidual Cell Suspensions


Strain C3H C57BL/6 C57BL/6 C57BL/6 DBA/2 C57BL/6

No. donor decidua 7 7 24 21

Cell No. infusedb (X 105) 17 1 2 3 25 20

Outcome Died day 9 Died day 5 Died day 5 Died day 6 Died day 11; no colonies Mottled spleen; tiny pinpoint foci of nroliferation’

a Irradiated recipients receiving BM survived in all experiments. b One recipient in each group. ’ Killed for assay day 12.

We found that NK cells were present very early in decidual development (Day 6.5, the earliest time at which a significant number of pure decidual cells can be obtained) and that their lytic capacity declined as pregnancy progressed. By midpregnancy NK activity was undetectable. This decline in activity is coincident with the elevation of NK activity in spleen (23) and in yolk sac (24) and with the detection of local nonspecific uterine suppressor cells (15). Thus the loss of NK cell activity by midpregnancy may be accounted for by migration of NK cells from the decidua, regulation of NK activity by a pregnancy-associated suppressor cell, or differentiation of NK cells into suppressor cell populations. Nonspecific suppressive activity is, however, associated with the decidua at the time under investigation (Days 6.5-7.5) (11, 13) suggesting that NK cells in decidua may be refractory to some forms of immune suppression. NK-like activity was assessedby the ability of cells from decidua to lyse YAC target cells and their inability to lyse other tumor targets (Table 4). It seemslikely that the lytic activity can be attributed to a heterogeneous class of effector cells since not all the lytic cells in decidua carried the same surface antigens. Anti-asialo GM 1 and anti-Thy 1.2 treatments reduced NK activity to differing degrees but some activity remained after each treatment and it is possible that cells lacking both markers also contributed to the decidual NK activity. Subsets of NK cells have been described in lymphoid tissue (20, 25) and could also exist in decidua. One of the functions attributed to decidua is the prevention of extensive invasion of the fetal trophoblast into the maternal myometrium (1, 7). Early trophoblast is highly invasive in ectopic sites where there is no decidual response (26). NK cells could play a role in the limitation of trophoblast invasion by acting directly to lyse some of the trophoblast cells or indirectly to regulate the proliferation of trophoblast. This hypothesis predicts that deletion of NK activity from decidua would lead to destructive invasion of the myometrium by trophoblast. However, examination of pregnancies in NK-deficient mice has not supported this hypothesis (27). No immune effector cell populations other than NK cells were detected in early murine decidua. We obtained no evidence that decidua contained T cells capable








of recognizing or responding to alloantigens of paternal or third party genotypes. No functional CTL could be demonstrated in early decidua even in preimmune mice sensitized to paternal antigens. This latter experiment demonstrated that the mechanisms accounting for the absenceof functional CTL from the immediate area of the conceptus are very effective. To determine if lack of functional CTL in decidua reflected suppression of their lytic function within the decidua or the actual absence of this class of cells from decidual suspensions, we assayed for precursors of CTL within the decidua. pCTL could not be demonstrated in decidua. Carryover of a suppressor factor is an unlikely explanation for the absence of detection of pCTL since large volumes of medium were used in the digestion and washing procedures and since placental factors of this type have been reported to be somewhat unstable (28). It remains possible that cell-bound suppressor molecules or suppressor cells in the decidual suspensionscould have blocked the differentiation of pCTL. If this had been the case such suppressor cells would have been expected to block development of CTL from splenic pCTL. This effect was not observed and the opposite appeared to be true. At the lowest spleen cell numbers tested addition of decidual cells elevated the numbers of pCTL detected in spleen. From this experiment we concluded that the absence of functional CTL in early decidua was due to the absence of pCTL and CTL rather than due to blockade of pCTL differentiation to CTL. The ability of decidual cells and extracts to block the differentiation of CTL in mass cultures containing T cells (1 l-l 5, 28, 29) may therefore be a manifestation of control mechanisms that could play a secondary role in normal pregnancy and inactivate aberrent T-cell effecters within decidua. Alternatively they may represent a mechanism for active exclusion of T cells from decidua. The studies reported here were conducted with decidual tissues isolated during early pregnancy and it is also possible that later decidua (11, 15, 28, 29) contain different cell populations regulated by distinct mechanisms. Decidual cell suspensions, which are known to contain cells of bone marrow origin (4-6) did not contain cells responsive to polyclonal activation or CFU-S. The lack of a response to T-cell mitogens was consistent with the absence of T cells and their precursors demonstrated in other assays.Lack of a responseto B-cell mitogens is also probably due to the absence of B cells within decidua since previous immunochemical studies have suggested that mature B cells, identified by the presence of surface IgM, are not present in 7.5-day decidua (16). Decidual cells, infused at numbers comparable to those required for detection of CFU-S in bone marrow, did not sustain viability of irradiated recipients. Of three recipients receiving much larger doses of decidual cells, only one survived to the end of the experiment. The appearance of the spleen colonies in the single surviving recipient was atypical, suggestive of endogenous repopulation and we conclude that CFU-S are unlikely to be present in decidua. Cells derived from bone marrow are known to be present, at least as a minority population, within decidual cell suspensions. We have shown that hematopoietic stem cells are not included within this population. We have also demonstrated that functional B and T cells are not among the bone marrow-derived cell population present in decidua. NK cells are present in early decidua but their activity disappears by midpregnancy. Numerous investigators have demonstrated that decidual tissue has immunosuppressive activities (2, 1l-15, 28, 29). Our studies have addressedthe assumption that immunocompetent cells are present in decidua and therefore



necessitatethe presence of a variety of suppressor mechanisms. We have shown that only NK cells appear in normal decidua in early pregnancy. These cells may play a positive role in embryonic development rather than act as a destructive effector mechanism. Other potentially cytotoxic cells were not demonstrated in decidua and appear to be absent rather than immunosuppressed. Thus the importance in vivo of cells and factors described as promoting local intrauterine immunosuppression remains open to future interpretation and may involve nonimmunological regulatory processesduring pregnancy. ACKNOWLEDGMENTS We thank Julie Hood and Lily DeRusha for their excellent technical assistance.

REFERENCES 1. Finn, C. A., J. Reprod. Fertil. Suppl. 31, 105, 1982. 2. Bell, S. C., J. Reprod. Immunol. 5, 185, 1983. 3. Miller, B. G., and Emmens, C. W., J. Endocrinol. 43, 427, 1969. 4. Keams, M., and Lala, P. K., J. Exp. Med. 155, 1537, 1981. 5. Gambel, P., Hunziker, R. D., Rossant, J., and Wegmann, T. G., Transplantation, in press. 6. Fowlis, D., and Ansell, J. D., Transplantation, in press. 7. Kirby, D. R. S., In “The Early Conceptus, Normal and Abnormal” (W. W. Park, Ed.), pp. 68-74. Univ. of St. Andrews Press,Edinburgh, 1965. 8. Keams, M., and Lala, P. K., Amer. J. Reprod. Immunol. 3, 78, 1983. 9. Porter, D. G., Heap, R. B., and Flint, A. P. E., J. Reprod. Fertil. Suppl. 31, 113, 1982. 10. Beer, A. E., and Billingham, R. E., J. Reprod. Fertil. Suppl. 21, 59, 1974. Il. Kirkwood, K. J., and Bell, S. C.,J. Reprod. Immunol. 3, 243, 1981. 12. Lala, P. K., Chattejee-Hasrouni, S., Keams, M., Montgomery, B., and Colavincenzo, V., Zmmunol. Rev. 75, 87, 1983. 13. Croy, B. A., Rossant, J., Clark, D. A., and Wegmann, T. G., Transplantation 35, 627, 1983. 14. Slapsys, R. M., and Clark, D. A., J. Reprod. Immunol. 4, 355, 1982. 15. Slapsys, R. M., and Clark, D. A., Amer. .I. Reprod. Immunol. 3, 65, 1983. 16. Bernard, O., Scheid, M. P., Ripoche, M., and Bennett, D., J. Exp. Med. 148, 580, 1978. 17. Mishell, B. B., and Shiigi, S. M. (Eds.) “Selected Methods in Immunology,” p. 135. Freeman, San Francisco, 1980. 18. McDonald, H. R., Cerottini, J.-C., Ryser, J.-E., Maryanski, J. L., Taswell, C., Widmer, M. B., and Brunner, K. T., Immunol. Rev. 51, 93, 1980. 19. Till, J. E., and McCulloch, E. A., Radiat. Res. 14, 213, 1961. 20. Koo, G., Jacobson, J., Hammerling, G., and Hammerling, U., J. Immunol. 125, 1003, 1980. 21. Miller, R. G., and Phillips, R. A., J. Cell. F’hysiol. 73, 191, 1969. 22. Abramson, S., Miller, R. G., and Phillips, R. A., J. Exp. Med. 145, 1567, 1977. 23. Chattejee-Hasrouni, S., Parhar, R., and Lala, P. K., Cell. Immunol. 84, 264, 1984. 24. Dahl, C., Eur. J. Immunol. 13, 747, 1983. 25. Roder, J. C., Karre, K., and Kiessling, R., Prog. Allergy 28, 66, 1981. 26. Simmons, R. L., and Russell, P. S., Ann. N. Y. Acad. Sci. 99, 7 17, 1962. 27. Croy, B. A., Gambel, P., Rossant, J., and Wegman, T. G., Cell. Immunol. 93, 317, 1985. 28. Chaouat, G., Kolb, J.-P., and Wegmann, T. G., Immunol. Rev. 75, 31, 1983. 29. Clark, D. A., Slapsys, R. M., Croy, B. A., Kreck, J., and Rossant, J., Amer. J. Reprod. Immunol. 5, 78, 1984.