Characterization of hematogenous cellular constituents of the murine decidua: a surface marker study

Journal of Repr,,du(tu'e Immunology. 8 ( 1985 ) 213 - 234 I'lsevler

213

JRI 00381

Characterization of hematogenous cellular constituents of the murine decidua: a surface marker study Mary Kearns and Peeyush K.

kala *

Department of Anatomv. Ih,alth S('tem es ('enter. The L'ntt'ersttv of l~'r~h'rn O,ztart~,. London. Ontarm. Canada N(}A 5C1 (Accepted for p u b h c a t l o n 13 S e p t e m b e r 1985)

Decidual tissue, which includes typical (,qromal type) decidual cells as v, ell as infiltrating leukoc)tes. a p p e a r s to play a h~-'al i m m u n o r e g u l a t o r y role in the m a i n t e n a n c e of pregnancy m nature. The present study evaluated the ~,,,mtribution of n u m e r o u s leukocyte subset~ character=/ed on the basis of morpholog). c o m b i n e d with cell surface markers u~ the d e v e l o p m e n t of m u r m e decidua d u r i n g s y n g e n e w (CBA '~ x C B A 3' ) and allogeneic (CBA ? × C 5 7 B L / 6 , 3 ) pregnancy. C o l l a g e n a s e dispersed decidua ",.,.ere subjected to total and differential c o u n t s and cell surface labeling for a r a d i o a u t o g r a p h i c identificatmn of '.'anou~ markers: S-IgM on B cells, T h y - I on 1 cells, neltber m a r k e r on null lymphocyte,,, kvt- (1 or 2 or 1.2) a n t i g e n s on T cell subsets, Mac-I and [-A on macrophages, using ~'' i-labeled m o n o c l o n a l a n t i b o d i e s or a s a n d w i c h labeling with IZ~'l-protein A. The tolal cellularity of dectdua basalis s h o v e d a biphasJc rise m both pregnancies, with peaks on day I l and days 15 and 16, but the a l l o p r e g n a n t d e c l d u a shov, cd a bagher a c c u m u l a t i o n of all cell types i n d i c a t i n g that an allogeneJc c o n c e p t u s causes an a u g m e n t e d dcclduogenesis. The n u m b e r of dec)dual cells, the mosl frequent cell class, rose In a peak on day I1 followed b~, a decline possibly due to cell death. The n u m b e r of lymphocyte,,., the next frequent cell class, showed a parallel pattern initially, followed b). a sharp secondary rise on day l0 This rise m a ) be due u~ a w i t h d r a w a l of progesterone, an a n t i t n f l a m m a t o r v hormone. Null cells p r e d o m i n a t e d amongst decidual l y m p h o c y t e s (45-80%), as well as in the progestational e n d o m e t n u m (53%). i n d i c a t i n g a h o r m o n a l control of their accumulation. The frequency of B cells was low (I()-13c~l and T cells (25-459~) c o m p a r a b l e It, that in the blo<~, with Lyt-l only class being the most c o m m o n T cell subset AIIopregnant decidua al,,o s h o w e d a late rise m the total n u m b e r of Lyt-2 only cells '~,hich ma', have a suppressor I'unct~on. Macrophages, the next c o m m o n leukocyte class, all expressed Mac-1. Their n u m b e r rose to a plateau b,. day 12, but at a higher level in allopregnancy. I-A (needed for antigen p r e s e n t a t i o n ) v.as cxpre~,sed b', an increasing p r o p o m o n (5-60"~} of m a c r o p h a g e s with a d v a n c i n g ge'qatlon These findings provide a ba~q,. for further functional studies.

Key words: dec~dual cell~, leul~o¢'vte subset3 tn dectdua. !vmphocvte ¢uh~et.~ m de(tdua, h m p l m ~ t c ~urfa(e marker~. Thv-I antigen. L.vt antigen. I . A antigen. Mac. 1 antigen

• To whom all c o r r e s p o n d e n c e and reprint requests should be addresscd.

0 1 6 5 - 0 3 7 8 / 8 5 / $ 0 3 . 3 0 ' 1985 Elsevier Science Publishers B V. ( Biomedical Dl~ ision)

214 Introduction

l)ecidualization of the endometrium during pregnancy is a unique maternal response occurring in certain mammalian orders such as carnivores, insectivores, rodents and anthropoids (Finn and Porter, 1975), in which trophoblast cells of the conceptus breach the uterine epithelium to invade the endometrial stroma and vessels, thus exposing themselves to the maternal immune system, and not in others such as ungulates in which this epithelium remains es,sentially intact and thus shelters the trophoblast cells (Lala et al., 1985). This observation combined with the fact that a significant proportion of the trophoblast cells in the m o u s e ( C h a t t e r j t : e Hasrouni and Lala, 1982; Colavincenzo and Lala, 1984, 1985) and the human tSunderland et al., 1981: Montgomery and Lala, 1983) express Class 1 major histocompatibility (MHC) antigens remaining unsheltered from the maternal lymphocytes, suggest that decidua may play an important local role in protecting the allogeneic conceptus. Numerous functional roles have been attributed to the decidua: provision of nutrition to the embryo during the early post implantation period prior to the establishment of feto-placental circulatory exchange (Krehbiel. 1937). elaboration of prolactin (Riddick and Kusmik, 1976: Golander et al., 1979: DeZiegler and Gurpide, 1982), estrogen and progesterone (Siiteri and Seron-Ferre, 1981) and production of prostaglandins (Antebi el al., 1975: Williams and Downing. 1977). Cells of the decidua or their products are also capable of suppressing the afferent arm of numerous immune responses of lymphocytes in vitro: (1) T cell alloreactivity in the mouse (Kirkwood and Bell, 1981; Badet et al., 1983a.b: Slapsys and Clark. 1983: Lala et al.. 1983; Clark et al., 1984. 1985) and man ((Solandcr et al., 1981: I,ala et al., 1985: Parhar and Lala, 1985), (2) T cell response to polyclonal mitogens (Golander et al., 1981). (3) development of natural killer (NK) function in NK lineage cells of the dccidua (Scodras et al., 1985) and (4) B cell responsc tO DN P-polylysine (Globerson et al., 1976). l)ecidual tissuc, however, is heterogcneous in composition comprising typical stromal type deeidual cells (henceforth called decidual cells), hematogcnous leukocvtcs (lymphocytes, monocytc-macrophages and granulocytes), metrial gland cells (in the rodents, or their equivalents in other species), and other stromal elements, e.g. fibroblasts and blood vessels. Immunohistological study of placental bed for leukocyte surface markers in situ has provided valuable qualitative information on the leukocyte subsets in the human decidua (Bulmcr and Sunderland, 1983, 1984). The origin and possible functional significance of some of the cellular constituents of the dccidua in several species have been explored by several laboratories, however, the information on the relative contribution of the various cell populations and their subsets to the decidual development remains fragmentary. Recently Kreck et al. (1983) reported that the prcsence of a conceptus which is genetically different from the mother at the 1I-2 locus is associated with a higher accumulation of mononuclear leukocytes in the trophoblast giant cell circulation of the placenta, however, the migration potential of these cells into the decidua remains undetermined. The present study was an attempt to exanaine the kinetics of accumulation of various

215 cellular elements in the murine decidua during syngeneic and allogeneic pregnancy. as well as to characterize the leukocyte subsets in the decidua on the basis of several well recognized surface markers identified at the morphological level with a highly sensitive radioautographic method. Lymphocytes were examined for surface lgM (on B lymphocytes), Thy-I antigen (on T lymphocytes) and neither marker (on "'null'" lymphocytes). T lymphocyte subsets were further characterized on the basis of Lvt (Lyt-I and Lyt-2) markers, and macrophages on the basis of Mac-1 and I-A antigens. Since decidua basalis forms the placental bed, our attention was primarily devoted to this component of the murine decidua.

Materials and Methods

Mit "e Inbred C B A / J and C 5 7 B L / 6 J mice (Jackson Laboratories. Bar Harbor, ME) were used at 12 to 14 wk of age. Syngeneic and allogcneic pregnancies were set up by mating C B A / J female micc respectively with C B A / J and C 5 7 B L / 6 J males. The day of appearance of a vaginal plug was taken as the day 0 of gestation. Animals were used at various gestational ages ranging between days 8 and 16. Preparation of single cell suspensions Spleen and thymus These organs were washed in ice-cold MEM containing 10~ newborn calf serum (10% NCS-MEM), minced on a wire screen (80 mesh per inch-" ) with fine scissors and gently passed through the screen with the same medium. One ml of NCS was placed underneath the cell suspension in a tube for 5 rain at 4°C in order to remove clumps which settle below the interface. Cells were washed and resuspended at a concentration of 20 x 10 ~ cells/ml of the medium for cell surface labeling. Blood In one set of experiments, leukocytes were recovered from cardiac blood by lysing the red cells in buffered 0.168 M NH4CI solution {for 10 min at 40(") followed by removal of the erythrocyte membranes by gentle spinning through a column of NCS. Cell pellets were resuspended in 10c~ NCS-MEM for cell surface labeling. Dectdual ttssue At gestational ages longer than day 8, embryos and placentae were removed from the uterus and well-defined decidua basalis underlying the placentae were completely dissected out and rinsed in ice-cold 10% NCS-MEM. On day 8 when the decidua basalis is still rudimentary, the whole uterus freed from embryos, membranes and placentae were collected for recovering cells from all decidual components including basalis, capsularis and lateralis. In one set of experiment.,,. non-pregnant progestational stage uterus was used to recover endometrial leukocytes. Tissues were. minced into fine pieces in Ca -~', M g " ' free phosphate buffered saline (pH 7.4) containing 0.02S EDTA and 0.3% collagenase (without the passage through a wire screen which disrupts decidual cells) and incubated in the same

216

medium for 30 min at 37°C in a shaking water bath to dissociate cells. (_'ells were then spun down and subjected to red blood cell lysis and clump removal as detailed earlier. They were then resuspended in 10% NCS-MEM for total nucleated cell counts, differential count and cell surface labeling, in one set of experiments the red cell lysis step was omitted so that the extent of Hood cell contamination in the decidual cell suspension could be evaluated. In this case, one uterine horn of the anesthesized mice was perfused via the uterine artery (Chatterjee-Hasrouni and Lala, 1982) with phosphate buffered saline (PBS, pH 7.4) to remove blood cells from the uterine circulation as much as possible to permit a comparison of the nucleated cell contents in the decidua of the perfused vs. nonperfused horns. Total and dtfferential count.~ of nucleated celA Total cell (inclusive of red cells) or nucleated cell counts were made in a Coulter counter (model ZBI, Coulter Electronics Inc., Healih, FL) and differential counts were performed under an oil emersion × 10el objective of a light microscope following staining of ethanol fixed smears or cytocentrifuge preparations with Giemsa. 4 ntihodtes Monospecific anti-Thv 1.2 [(AKR × J) anti-AKR × Cure] serum ( N I H catalogue member E Thy-l.2, cytotoxic titre 1/1000) was obtained through the courtesy of Dr. J.G. Ray of the National Institutes of Allergy and Infectious Diseases, Bethesda, MD. Affinity purified goat anti-mouse lgM was a gift of Dr. D.G. Osmond of McGill University. Antisera were spun at 160,000 × g, in an airdriven ultracentrifuge for 30 rain to remove protein aggregates, prior to their use for cell surface labeling. Monoc[onal antibodies (details provided in Table 1) against Thy-1. Lyt-I, Lyt-2 and I-A k (specificity la.17) antigens were purchased from Becton-Dickinson, Sunnyvale, CA, and against Mac-1 antigens from Hybritech, San Diego. CA.

I A BLI-. 1 Details of nlom~:lonal antih~lies employed ~

A ntibcv.lies

Source

I,,otype

Reference

Th~-I

[.vt- 1

serum a n d / o r a~*cate~,fluid of BAI.B/c mice culture supernatant

l.yt-2

culture supcrnatant

Mac-I

a,,cites serum

l.edhetter and Herzenherg { 1979) Ledbetter and lterzenherg (1979) l.edbener and Hcrzenherg (1979) Springer el al. (IgTg)

l-A~(la.17)

serum a n d / o r ascites fluid of tumor-bearing BAI.B/c mice

rat IgG2a rat Ig(;2a rat lg(52a rat lgG2h mouse IgG2h

()i et al. (197~1

" All ant=bodle~, v.ere received in pure form obtained hy ion exchange or afl'mJt~ chromatograph?,

217

Radimodmation of proteins A radioiodination of the following proteins with ~"~I was performed using the chloramine-T oxidation method of Greenwood et al. (1963): staphylococcal protein A (Pharmacia Fine Chemicals, Dorval, Quebec) and purified proteins of monoclonal Thy-1, Lyt-l, Lyt-2 and Mac-1 antibodies, and goat anti-mouse lgM. The final specific activity attained was 40-50 C i / g of protein in each case. Cell surface labeling Detection of surface Thy 1.2 antigen and lgM on small lymphocvtes m t,artous lymphotd organs and the decidua This was done as reported by Garnis and Lala (1978). Briefly, 2 x 10" cells in 0.1 ml vol. were incubated with 1/20 dilution of either monospecific anti-Thy-l.2 antibody or normal mouse serum (NMS) for 30 rain at 4°C. Following two washes through NCS gradients, the cells were finally incubated with ~251-goat antimouse IgM at a concentration of 2 ~ g / m l , washed two more times, and pellets were resuspended in a drop of FCS and smeared on gelatin coated slides for radioautography. With this procedure, the NMS-treated samples provided labeling of lgM" B cells. T h y - l . 2 - or T cells, in addition to B cells were labeled in the anti-Thy-l.2 treated sample, because most of the antibody activity was in the lgM fraction. Thus, the difference between the labeling index of this sample and the NMS-treated sample provided the incidence of Thy-l.2 ' cells. Lymphocytes remaining unlabeled in the anti-Thy-l.2 treated sample were considered as l g M - . Thv-I or null cells. In certain experiments cells were also directly labeled with radioiodinated monoclonal anti-Thy-1 antibody (0.4 tag/ml) to measure the incidence of T cells directly, as a check on the sandwich labeling protocol. Since the two measurements matched very closely, a direct labeling was abandoned in later experiments. It must also be pointed out that great care was taken to establish the morphological identity of the lymphocytes, to exclude the possibility of including any T h y - I " decidual cell (Bernard et al.. 1978; Kearns and Lala. 1985) in these measurements. Characterization of T cell subsets within the decidua and lymphoid organs Samples of 2 x 10 ~' cells in 0.I ml vol. were treated with radioiodinated anti-Lyt-1 (0.33 p,g/ml) or anti-Lyt-2 (0.9 ~ g / m l ) or a combination of both antibodies at these concentrations, for 1 h at 4°C. Cells were then washed three times through FCS and smeared on slides for radioautography. The incidence of various kyt antigen bearing subsets was computed as shown in Table 2. Detection of Mac-I and I-A antigens of monocyte--macrophages within the dectdua Cells were treated with radioiodinated anti-Mac-1 antibody (1 p.g/ml), followed by washing as described for Lyt antigens. To detect I-A antigens, a sandwich labeling with ~-'51-protein A was employed. Cells were treated with 1/20 dilution of anti-l-A ~ at 4°C for 30 min followed by two washes through NCS and then treated with ~2~l-protein A (1 p.g/ml) for 30 min followed by three washes prior to making smears.

218 I ABI.f: 2 ('omputatlon of the incidence of Lvt antigen-bearing subsets Antlb,~ly used (~) I.,,.t-I (b) Lyt-2 (c) LvI-I , Lvt-2 Inc,tlenceofl.',-I

(.'ells labeled l.vt-1. Lyt-1.2 Lyt-2. Lyt-1.2 l.vt-I. 1.,,i-2. Lvt-1.2 =(c)

(b)

b.-2 =(c)-(a) L,,-I.2 = ( a ) + ( b ) - ( c )

Radioautograph)" and et,aluation of radioautographs Slides were processed for radioautography using the method of Kopriwa and Leblond (1962). The threshold number of silver grains for positive labeling was determined from the grain count distribution on at least 200 cell-free areas of smcars equivalent to the size of the cell concerned, as described by Lala and Patt (1966). This was usually 6 grains for lymphocytes and 11 grains for monocytes/macrophages. To compute labeling indices, the grain count distribution on 300 500 lymphocytes and 100-200 monocytc/macrophages per samplc was determined.

ldenttfication of cells m srnears The morphology of mouse decidual cells was evaluated through a sequential examination of the decidual cell reaction from day 4 to day 14 of pregnancy using l p.m thick semithin Epon section for light microscopy and adjacent 100 nm thin serial sections f~,. electron microscopy. Typical, stromal type decidual cells were first unequivocably identified under an electron microscope based on their ultrastructural features as summarized earlier (Kearns and Lala, 1983, Lala et al., 1984). The same cells were're-examined in adjacent semithin sections under a light microscope to correlate their morphological appearances with those in smears. In both preparations they are identifiable as slightly vacuolated, uninucleate or occasionally binucleate cells of ovoid appearance. Their nuclei were eccentric in position, round or ovoid in shape, and had uniformly dispersed chromatin containing one or more nucleoli. In Giemsa-stained smears, these cells showed a basophilic cytoplasm and their sizes ranged usually between 12 and 30 /xm in diameter. They were easily distinguished from leukocytic elements, e.g. monocyte-macrophages, lymphocytes and granulocytes. Macrophages were highly vacuolated cells with copious cytoplasm having convoluted, or kidney shaped or occasionally rounded nuclei. Although displaying a similar size range as decidual cells, they showed a lighter staining cytoplasm and the ability to phagocytose latex particles, whereas decidual cells were nonphagocytic. Furthermore, decidual cells expressed surface Dec-I antigen, whereas macrophages or other leukocytes in the decidua did not (Kearns and Lala, 1985a; Kearns et al., 1985). Monocytes were smaller cells (10-14 p.m) with lightly staining basophilic cytoplasm and kidney shaped or clover leaf type nuclei. Small lymphocytes showed a round, purple staining eccentric nucleus and a thin rim of lightly staining blue cytoplasm, with diameters ranging between 6 and 8/~m. Polymorphonuclear leuko-

219

'51

n"aQ

Q

o.. L

4f

See ~ IP

• %.

•. o ,

Ii

.°,,"

• ,m, _--

a

~

"I oo



Fig. 1. Smear preparation of a collagenase dispersed single cell suspension of the murine decidua basalis on day 12 of gestation dlustrating various cell types within the decidual tissue (stained with Giemsa. x 250). D. decidual cells: L. lymphocyte:M. macrophage: MO. mon~x:yte. cytes displayed a segmented, doughnut shaped nucleus and lightly staining, slightly granulated cytoplasm with diameters ranging between 10 and 12 ~m. Other occasionally encountered cell types, which were too few to be quantitated in these preparations are metrial gland cells (ovoid cells, 25-50 p,m in diameter, containing numerous diastase resistant, Periodic Acid Schiff positive granules in pale cytoplasm) and nucleated erythroid cells. Figure 1 illustrates light microscope morphology of cells in smeared single cell suspension of decidua basalis.

Statistical analysis Standard error of the mean values were calculated and significance of the difference between mean values was evaluated with Student's t-test (Spiegel, 1961).

Results

Total cellularity of the dectdua The total number of nucleated cells in the collagenase dispersed preparations of the decidua basalis in syngeneic (CBA £ x CBA,3) and allogeneic (CBA ? x C57BL/6J,3 ) pregnancy showed a biphasic rise during mid- and late gestational intervals (Fig. 2). Peak cellularity occurred on day 11 and days 15 and 16 in both

22(I E • •

Syngeme,( ~,, 'ogen e,c 5 Y'~c~ene'c AIIOQeneic

} whole

dec,duO

T

t~ E~

L./T

ul

L

!

T 8:)

//

,:

t

## %,

u~

4,7



/I

I

\~1

1~,

/"

/

2 ¢.

2"

P

9

.0

DAY

"

12

3

4

15

16

OF GESTATION

Ftg. 2. Total cell count of the collagenase-dispersed decidua hasalts during the course of syngenclc and allogenelc pregnancy in the C B A / J mouse. Eight-day value~, represent total decidua related to a single conceptus rather than the dectdua haz>ali~. The values represent the mean of 3 determinatlom, excepting days 13 and 15 of allogene,c pregnancy which represent a single determination. Each determination includes 5-1U capsules (decldua basalis) from a single nlou~,e. Day 8 values represent the whole decldua. In this case. each determination includes materials pooled from 2 to 3 mice.

pregnancy types. The decidua basalis of allogeneically pregnant mice showed a higher size of cell population throughout pregnancy excepting days 15 and 16. The differences were significant (at P > 0.01 < 0.05) on days 9 and 12. N o significant difference in the cellularity of the whole decidua was observed on day 8 between the two pregnancy types. That the nucleated cell counts of the dispersed decidua werc not influenced by any significant contamination of blood leukocytes in the decidual circulation is evident from the data presented in Table 3. Firstly, in spite of a major reduction in the erythr(x:yte contamination in the dispersed decidua of perfused uterine horns, the nucleated cell counts did not show any obvious difference between the perfused vs. nonperfused decidua. Secondly, in the nonperfused decidua the erythrocyte: nucleated cell ratio was approximately 7 : 1 (day 8) or 5 : 1 (day 11), which can be translated into a meagre contamination of 1 circulating leukocyte per 200-300 nucleuted cells, assuming an approximate erythrocyte: leukocyte ratio of 1500 : 1 in the mouse blood. An analysis of the various cell classes within the decidua basalis (Fig. 3. Table 4) was based on the differential count of intact cells. Broken cells (which constituted the largest proportion at all intervals, as shown in Table 4) exhibited intact nuclei, the morphology of which was similar to those of fibroblast-like stromal cells seen in semithin sections. The nuclei were round, had a fine disperse chromatin and contained usually 2 - 3 distinct nucleoli. Thus it appears that stromal cells were most vulnerable to breakage in these preparations. Introduction of trypsin in the suspend-

221 TABLE 3 Extent o f e r y t h r o c y t e c o n t a m i n a t i o n in the d i s p e r s e d d e c t d u a tff n o n p e r f u s e d vs. p e r f u s e d h o r n s of a l l o p r c g n a n t C B A mice I )a'.' of gestatton

8 (thole decldua) 11 ( d e c i d u a hasalis)

Cell c o u n t ( × 10 ' ) per d e c t d u a Nonperfused horn"

Perfused horn =

Red cells

Nucleated cells

Red cells

Nucleated cells

28.32 53.23

4.12 10.71

994 22.44

423 10 68

" 5 c o n c e p t u s e s per h o r n o n e a c h o c c a s i o n ; d e c i d u a m e a c h h o r n were prowled heft~re count,, a n d the c o u n t s ~ e r e n o r m a h , , e d for a single d e c l d u a .

ing medium did not protect these cells, however, addition of DNAase lysed some of the nuclei of the broken cells, leading to a drop in the total nucleated cell count. These broken cells could also be excluded by further Ficoll-Paque separation. None of these procedures were adopted in the current study so that the total cellularity could be reliably determined. In spite of the possibility that all decidual cells may not have been released, or recovered intact after the dispersion procedure, typical decidual cells still formed the major cell class amongst the intact cells in both pregnancy types throughout gestation. Their number increased in both cases to a peak on day 11 followed by a decline. In allogeneic pregnancy the total number of decidual cells within the decidua 'basalis was higher than in syngeneic pregnancy. These results may suggest that the allogeneic placenta invokes an increased decidual cell development at the site of placentation. The general pattern of lymphocyte accumulation (Fig. 3) within the decidua basalis was similar for both pregnancy types, but the accumulation was more pronounced in allogeneic pregnancy. The first peak was seen on days I1 and 12 when the differences between the two pregnancy types were significant at P < 0.01. This was followed by a decline and a distinct second rise in late pregnancy in both cases. On day 16, lymphocyte numbers were comparable to that of the decidual cells. The frequency of monocyte-macrophages and granulocytes within the decidua basalis was initially very low in both forms of pregnancy (Table 4). Maximal accumulation of monocyte-macrophages (primarily macrophages and few monocytes) reached a plateau between days 11 and 16 (Fig. 3). They formed the second largest population of decidual leukocytes. Granulocyte population represented the smallest subset of leukocytes within the decidua. Their numbers reached a peak followed by a decline in late gestation.

Lymphocyte subsets withtn the decidua B, T and null lyrnphoc),tes Their incidence within the decidua is presented in Fig. 4. Null cells constituted the highest proportion of lymphocytes (45-80%) within the

222 60

SYNG,T.NE~C

ALLOGENEIC

aC

,o 6 5

g

/, f

,

~ ~ y~mpho

4

T'

}

.L'

2

,

/ ~0

~

/

Mono I m~ .

~

/} / -~.,

~'

\

, rn~ ,¢

! r

'

, "~" - - - -t-~-

L"

0.5 o.3

~"

,

O2

zo,o

No~ ~006

oo,

004

TI

jl I

'/'! / 1 '\i

i/

!/I

003

0.02

,I~/

00,

~C)

,2

,4

~6

8

,0

~z

~4

MS

DAYOF GESTATION F~g. 3. T o t a l c o u n t o f , . a n o u s c e l l t y p e s m the individual decJdua hasahs during the course of ,,}ngenetc and allogene~c pregnancy. The values repre,,ent the mean.,, o f 2 - 3 ( i n mo~,t ca.,,es 3) d e t e r m m a t m o n . , , excepting day 15 o f a l l o g e n e i c pregnancy which represent,,, a ~,mgle d e t e r m i n a t k m . E a c h determination i n c l u d e s 5 - 10 c a p s u l e s ( d e c i d u a ba~,alis) f r o m a .,,ingle mou.,,e. D a y 8 v a l u e s rcpre~,ent the whole d e c i d u a . I n this c a s e each determination ts based o n materials p(w)led f r o m 2 t o 3 m i c e . S v m b c , ls o n d a y 8 correspond to the respective cell type ~,ymb,,)ls o n other days.

decidua in both syngeneic and allogeneic pregnancy at various gestational ages, but this was equally high (53%) within the progestational endometrium of non-pregnant mice as given by the day 0 value. However. in the latter case (Table 5), this was 5% in thc spleen and 17% in the blood or the peritoneal space (a non-lymphoid tissue space), indicating that there was a preferential migration a n d / o r retention of null lymphocytes into the progestational endometrium as well as the decidua. Conversely, when compared to the blood or peritoneal space of non-pregnant mice, there was a paucity of B cell incidence in the endometrium prior to and during both types of

75.8

(4.9)

69.6

(7.8)

Broken cells

Day 9

(2.9)

84.2

(0.05)

0.05

(1.3)

1.6

(0.15)

12.9 (1.7) 1.4

Syn

Value oblamed from one determination. Numbers an parentheses indlca(e _+SE of the mean.

-

0.11

(0. I)

0.13

(0.1)

2.5

(2.2)

5.2

(2.9)

(1.0)

(3.7)

3.9

(2.6)

(1.4)

27

9.9 (2.2) 8.2

136 (2.1) 7.3

Allo

Metrial gland cells

Erylhyroid

Oranuh.~'yles

Mon~,~:yte- macrophage,,

l.ymphocyles

Decidual cells

(.ell type (%.)

Da> g

Syn

(;cslalional

age

(10.2)

66.1

-

(0.2)

0.2

(0.2)

0.2

(2.5)

29.4 (3.0) 3.2

AIIo

(3.2)

0.16 (0.1) 8(I.5

(0.2)

0.2

(0.45)

0.65

(0.09)

0.37

(3.0)

14.2 (4.0) 4.4

Svn

Day 10

(2,6)

0.1 (0.1) 57.4

(0.1)

0.l

(1.3)

3.0

(I 0)

27

(20)

29.1 41.9) 7.7

A[Io

(7.9)

0. I (0.1) 56.1

(0.16)

0.52

(0.15)

0.73

(0.27)

1.3

(0.4)

39.5 (8.3) 2.2

Svn

i)av 11

(2.3)

52.4

-

(3.0)

3.0

(0.5)

3.4

(02)

0.9

(5.2)

28.3 (5.2) 12.0

AIIo

(7.0)

73.6

(0.6)

0.8

(04)

0.9

(0.25)

1.7

(1,7)

17.0 (4.5) 6.0

Svn

L)av 12

(8.7)

73.1

(0.4)

0.4

(0.4)

1.0

(1.0)

2.5

(0.7)

15() (I.6) 8.0

AIIo

(2.6)

90.7

N ):,"

(5.6)

87.6

-

(0.1)

0.1

(0.2)

0.3

15

(3.3)

4.9 (20) 59

0 N

(0.~)

0 3"

i ) a s 16 Syn

(0.6)

2 6 "'

1 .g"

147:

AIIo

(0.05)

1.35

(0.15)

6.8 (2.1) U.75

Ssn

Da~, 15

D=fferenlial coun! Of the decLdual tissue of "~arlous ge.,,laUonal ages d u r i n g syngenelc and allogenelc pregnanc.s m (.'BA,,'J mice

TABLE 4

(2.4)

84. I

f~.7 (2.1) 6.7 (0.7) 2.4 (0.5) 036 (0.05) 0.1 (0.1)

AIIo

224

SYNGENEIC

e - - e

[3 " ce,ls

+-- -~ +~.

T - el+lit+ N u r l c411~$

l

7O 60

so ~

4o

[

-.l 1 1

~

iO I

u~

o

N ALLOGENEIC

t~

~

6o

~

so

~

4o

~

30 20 IO

8

DAY OF GESTATION

Fig. 4. The incidence of 1.1, T arid null lymphoc~,te ~,ubsets in the decidua at ',anous time points during ,~,ngenelc and allogeneic CBA,."J pregnancy. Zero da'," ~,alues represent l y m p h ~ k t e s contained ',x.ithin the progestatlonal nonpregnant uterus in diestrus. Vertical bars represent standard errors of the mean of 3 determinations. Each determination includes a pool of 2--3 mice. excepting the zero da', value which repre~,ents a p ~ l of five mice.

pregnancy (Fig. 4. Table 5). T cell incidence remained essentially unchanged. The trend in the total number of B, T and null lymphocytes in the decidua is shown in Fig. 5 for both types of pregnancy. While the number of all lymphocyte subsets increased during allogcneic pregnancy, only null lymphocyte numbers increased during syngeneic pregnancy. Furthermore, the number of corresponding subsets was always higher in allogeneic than in syngeneic pregnancy.

TABLE 5 Percentage of T. B and null I),mph~vcyte subsets m non-pregnant ('BA female mice in diestrus

Spleen BIc~d Peritoneal space t:teru~,

B

T

Null

70 39 58 9

25 44 25 3N

5 17 I7 53

225 8 r,C~,lIs

0--0

- -. - -

f Cet " Null

~'~IIS

C3

F ALLOSENE_.("

5 0 0 F SYNGENEIC c3 4001--

T

F

l

"

r..__

..

.!l

,'i

~ 2ool-

/

./

>" IOC

/ L

80P-



60

"

/

~ 5o ~ 4o 2- 2o

/

/

/

/

±

W

3 ,o

,L DAY OF G E S T A T I O N

Fig 5. Ah~,olute number of "I'. B and null I?,mphoeyte suhsets ,n the ('BA,.'J dccidua dunng ,,yngcneic and allogenelc pregnancy. Eight-day values represent the v, hole dectdua and other value.,, represent the decidua ha.,.alis. These values are deduced from Fig~,. 3 and 4.

Lyt subsets of T lvmphocvtes The incidence of Lyt lymphocyte subsets within the decidua is presented in Fig. 6. The combined incidence of all Lyt bearing lymphocytes was comparable to that of Thy-I " cells presented earlier in Fig. 4. Of the three subsets, Lyt-1 ~ cells formed the most predominant class amongst the decidual T lymphocytes in both forms of pregnancy. Changes within the individual subsets during the course of pregnancy were not significant at P = 0.05. A quantitation of the absolute number of these subsets (Fig. 7) revealed that the population size of each subset was larger in the decidua basalis (days 12 and 16) of allogeneic pregnancy than in its syngeneic counterpart. Between days 12 and 16, none of the changes in the three Lyt subsets were significant at P = 0.05, however. Lyt-2 only cells showed some 25% rise during allogeneic pregnancy. Mac-I and I.A antigenic marker on monoc)'te-macrophages Most or all (96-98%) of these cells in both pregnancy types labeled for Mac-I at all gestational intervals examined. I-A antigens (la.17) were examined on monocyte-macrophages only during syngeneic pregnancy on days 8, 12 and 14. The labeling ranged between 4.5 and 36.3% on day 8, 11.5 and 50% on days 12 and 30 and 61~, on day 14, indicating that only a subpopulation expressed I-A, the incidence of which increased with advancing gestational age. We had earlier shown that typical decidual cells are I - A , but a small subpopulation expresses Mac-I (Kearns and Lala, 1985). Figure 8 illustrates radioautographic preparations of cells of the decidua subjected to labeling for several surface markers.

SYNGF ~.~FtC

/

4O •/

z"'\\\/.

....

,

-e

I

....

\

/

role"

Lyt

t.yt - I

Lyl

I,~

\ •

l

T

JO t

cL

\



u

\\

"

....

~

II

1

t.

,6

L " ' ,F ".JF'IC

-I

.i~'--_

1"

ii5

I

,

__. ~A'~"

~

A __. ' ();

L,}",~T,'; r

L

i ~) ¢?ri

Fig. 6. The incidence of various l.~t subsets of lymphoc)tes in the dectdua at ,,anous tmlc points during syngenetc and allogeneic C B A / J pregnancy. Each value represents the mean of 3 determinations and ',erttcal bar,; represent standard errors. Each determination is based on a pool of 3 animals. O - - -0

Total L_y!

0--0

Lyt- L

A--&

Lyt-2

~''-C

Lyl " I,~'

ALLOGENEIC

SYNGENEIC ~oo e~

,

L~

LOC

-r ~

"~C ~C 5O

~

8c

n~

f ÷

T

T i

TT

T

D z_ ~ m

,'2

tO

,~,

DAY OF GESTATION

Fig 7 Absolute number of L',I lymphocyte subsets m the CBA,..'J decidua during s~ngeneJc and allogeneic pregnancy Da~. 8 ,..alues represent whole dectdua and other values, dectdua basalis. Vcrttcal bars are the standard errors of the mean These values are deduced from Figs. 3 and 6.

227 ..

,..°.

i

.(

,4.'

f

'

t !x.



~

v

m

Fig. 8. Illustration of cell types encountered in murine decidua hasalis. (a) and (b) are from a cell suspension of 12 day decidua of syngeneic pregnancy incubated with latex particles. (c)-(f) are from radioautographic preparations made for various cell surface markers. All magnifications are × 1890. excepting (e) which is × 2600. (a) A macrophage with intracytoplasmic latex particles. (b) A decidual cell showing no evidence of phagocytosts of latex particles. (c) Two lympht~ytes (one labeled, representmg a T or a B lymphocyte, the other unlabeled representing a null lymphocyte) from 12 day decidua of allogeneic pregnancy, exposed to anti-Thy-I antibody followed by I"~'l-anti-lgM. (d) Three labeled (Lyt bearing) lymphocytes from 12 day decidua of allogeneac pregnancy exposed to i.'5 I-labeled anti-Lyt-1 and anti-Lyt-2 antibodies. (e)-(g) are from 12 day decidua of syngene]c pregnancy showing (e) a Mac-I * macrophage. If) an I-A macrophage and (g) an I-A " macrophage.

228 l)i,lcussion

Present study was designed primarily to examine the kinetics of accumulation of various leukocyte subsets in the decidua during syngeneic and allogeneic pregnancy. The collagenase-dispersion procedure has been reported to recover all for most of tile marrow derived cellular elements in the decidua (Gambel et al., 1985). In spite of the possibility that all decidual cells may not have been recovered with this methodology, assuming a nonpreferential recovery during the two types of pregflancv, present results reveal an augmentation of decidual cellularity inclusive of typical decidual cclls as well as leukocytes in the presence of an allogeneic conccptus. The differential response, initially noted on day 9, may result from a recognition of paternal class 1 M H C alloantigens expressed on the trophoblast cells at this time (Chatterjee-Itasrouni and Lala, 1981). The peak cellularity of typical decidual cells noted on day 11 in both types of pregnancy is consistent with the histological picture of peak maturation of dccidua basalis seen at this timc (data not presented). The ~,ubsequcnt decline in cellularity possibly represents decidual cell death, also identifiable as pyknotic nuclei in histological sections. Dccidual cells arise by hypertrophy and proliferation of certain fibroblast like stromal cells. Whether these precursors represent a subset of or all endometrial .,,tromal cclls rcmains undetermined. Our earlier studies (Kearns and Lala, 1982) revealed that at least a subpopulation of decidual cells which can be recovered by a collagcnase dispersion of the deciduoma induced in pseudopregnant radiation bone marrow chimeras are ultimate dcscenders of the bone marrow. Recently two laboratorics (Fowlis and Ansell, 1985, Gambel ctal., 1985a) have reexamincd this issue using donor-specific iso-enzyme markers analysed on cell lysates of the deciduoma of pseudoprcgnancy or decidua of pregnancy. Although the iso-enzyme methods did not permit an examination of the phenotypic identity of the individual cells, they allowed an estimation of the overall cxtent of chimerism at the tissue level to reveal that ( 1 ) the degree of chimerism in the nondispersed decidua or deciduoma as well as in the fibrous matrix remaining after a collagenase dispersion was small or ncgligible in mice which showed variable degrees of chimerism in splenic white cells or bone marrow cells, (2) variable but appreciable bone marrow contribution was noted in the collagenase dispersed population. These findings indicate that (1) the bulk of the decidual stromal matrix left after collagenase dispersion does not have a bone marrow origin, (2) collagenase dispersion preferentially isolates marrow derived elements which may form a minor subpopulation within the decidua or deciduoma. Further ~tudies cmploying in situ labeling of cells for the donor phenotype should help to establish the identity of all marrow derived elements in the dccidua. Murinc decidual cells have been shown to be Thy-1 ", F O R , Mac-1 ", C , R , 1-A . l-J , and Lyt (Bernard et al., 1978: Lala et al., 1983: Kearns and Lala, 1985: Lala c t a l . , 1985) and to bear a unique lineage specific differentiation marker Dec-l, not shared by lymphomyeloid cells (Kearns and Lala, 1985) but present on decidual cells of other species (Kearns ct al., 1985). Abrogation of T cell alloreactivity by unfractionated decidua or its products has been reported by numerous laboratories (Kirkwood and Bell. 1981; Badet et al., 1983a. b). including

22~

the partial identification of suppressor activity in the 1.5 kI)a, 40-50 kDa and 1000' kDa fractions (Badet et al., 1983b) of culture supernatants. Rccentl~ functional studies with 90-97~ pure murine or human decidual cells have revealed that they can strongly suppress T cell alloreactivity (Lain et al.. 1983, 1985: Parhar and Lain. 1985), as well as the development of natural killer function in the NK lineage cells of the decidua (Scodras et al., 1985), in an MH('-nonrestricted manner by releasing prostaglandins, primarily P G E , . Thus an augmented deciduogenesis during allogeneic pregnancy may be of functional significance. Another cell class in the rodent decidua described as metrial gland ceils which display characteristic cytoplasmic granules (Peel et al.. 1979), cytoplasmic Ig (Mitchell et al., 1980) and surface Fc receptors for [gG (Craggs and Peel, 1983) has also been reported to be the product of bone marrow derived percursors (Peel et al., 1983). These cells were too few to be quantitated in the present preparations. Lymphocytes constituted the largest component of the leukocytes infiltrating decidua. The peak lymphocyte accumulation in the decidua on day 11 and its augmentation in allogeneic pregnancy observed m our study presents an interesting parallel to the histological findings of Krcek et al. (1983) on the kinetics of mononuclear leukocyte accumulation in the trophoblast giant cell circulation of the placenta. They noted a maximal accumulation on day 10 (which corresponds to day 11 according to ~ur terminology) and a higher accumulation in placentae which were genetically disparate from the mother at the H-2 locus, it is possible that these celb, are in transit for a migration into the adjacent decidua. The secondary rise in the number of lymphocytcs in the decidua during late pregnancy noted in the present study may be due to a decline in the level of progesterone, a hormone show n to have an anti-inflammatory function (Siiteri et al., 1977). Previous studies from this laboratory havc documented changes in the B, "F and null lymphocytes in the various maternal lymphoid organs during the course of syngeneic and allogeneic murine pregnancy (Chatteoee-Hasrouni ct al., 1980). The present analysis, extended to the decidua in similar mice revealed that the pattern of infiltration by these lymphocyte subsets characterized by a high null cell and low B cell level is predictably set at the progestational stage of the endomctrium before implantation and decidua formation. A further careful analysis of the surface markers on endometrial lymphocytes at different stages of the estrus cycle should reveal whether this pattern is influenced by ovarian hormones. For example, a cyclic leukocyte migration into the endometrium has been reported by Padykula and Tansey (1979). The paucity of B cell accumulation in the decidua is consistent with the findings of a paucity of LPS-responsive cells reported by Gambel et al. (1985b). The null lymphocyte subset, predominant in the decidua, was also found to increase in all lymphoid organs (bone marrow, peripheral blood, placcntal circulation, spleen and paraaortic lymph nodes draining thc uterus) of pregnant mice. more notably during the course of allogeneic than of syngcneic pregnancy (Chatterjee-Hasrouni et al., 1980). The peak incidence in the blood (on days 8 and 9) was 20,°/ in syngcneic and 40% in allogeneic pregnancy. Present studies on the decidua revealed even higher values (45-80%) suggesting a preferential influx a n d / o r retention of the null lymphocytes within the decidua. Furthermore, a higher total accumulation was

230 noted in the decidua of allogeneic than of syngeneic pregnancy. Decidual null lymphocytes may represent a functionally heterogeneous population. Two functional subsets have so far been recognized. One subset, having the sedimentation properties of small lymphocytes, a granular morphology and surface Fc receptors, is considered to exert a local immunoregulatory role (Slapsys and Clark, 1983; Clark et al., 1984) by releasing a 100 kDa suppressor factor capable of blocking the lymphocyte response to IL-2 (Clark et al., 1985). Another possible subset is the natural killer iNK) lineage lymphocytes, which recognize target structures on the murine trophoblast cells (Chatterjee-Hasrouni et al., 1984). Questions naturally arise, do NK lineage cells migrate to the decidua, and if so, do they remain functional in situ? Gambel et al. (1985b) reported readily detectable NK activity in 6.5 day decidua, but the activity gradually declined to background levels by day 11.5. It was further shown from our laboratory (Lala et al., 1985: Scodras et al., 1985) that 11-14 day decidua contained an abundance of NK lineage cells (null lymphocytes capable of binding YAC-1 lymphoma targets) but these cells were inactive. However, decidual NK activity could be restored in the presence of indomethacin, an inhibitor of prostaglandin synthesis, indicating a local prostaglandin mediated inactivation of the NK lineage cells. The suppressor cells have been identified as decidual cells and macrophages. Purified fraction of these cell types from the decidua were shown to suppress the development of the lytic function in NK lineage cells by a prostaglandin mediated mechanism. T lymphocytes, the most important participants in an allograft response, constitute functionally distinctive subsets recognizable by Lyt markers (Cantor and Boyse, 1975a,b). For example, Lyt-l,2 cells include precursor T cells which differentiate into Lyt-1 and Lyt-2 subsets. Lyt-1 cells include T helper cells both for antibody formation as well as alloreactivity (Cantor and Boyse, 1975a,b: Swain ct al.. 1979). in the latter case. they are believed to proliferate in response to Class 2 MHC antigens. Effector cells for DTH responses are also Lyt-1. This cell class is believed to play an important role in allograft rejection (Loveland and McKenzie, 1982). Finally, suppressor inducer ceils of Lyt-I phenotype have also been reported (Eardley et al., 1978). Lyt-2 cells include suppressor T cells, both for humoral and cellular immune responses, and cytotoxic T cells having specificity towards antigens presented on syngeneic cells or towards Class 1 MHC alloantigens (Cantor and Boyse, 1975a; Swain et al., 1979). Despite reported exceptions to these general rules. Lyt markers are considered to provide valuable indices of T cell function. A comparison of the changes in the various Lyt subsets within the maternal lymphoid organs (Chatterjee-Hasrouni, 1981; Lala et al., 1983) with the current findings on the decidua may be informative. Present findings of an absolute higher number of all three Lyt subsets in the decidua of allogeneic as compared to that of syngeneic pregnancy are very similar to those recorded for the paraaortic lymph nodes draining the uterus (Chatterjee-Hasrouni, 1981). Taken together, these results may be suggestive of a higher local recruitment of alloreactive T cells of all functional classes at the sites of antigen recognition. Lyt-1 + cells constituted the most predominant T cell class in both types of pregnancy, the incidences being very similar to those reported in the spleen (Chatterjee-Hasrouni, 1981~. Further studies including

231

an examination of L3T4 marker of the helper phenotype are needed to evaluate their functional potential. A late increase in the total number of l_yt-2 ' cells seen in the allopregnant decidua has also been noted within the spleen and paraaortic lymph nodes (Chatterjee-Hasrouni, 1981; Lala et al., 1983), which might include suppressor T cells. However, this may reflect a homeostatic response of the mother to the presence of the allogeneic conceptus, having no functional bearing on feto-placental survival. Monocyte-macrophages within the decidua, all expressing Mac-1 antigens, formed a significant population in mid to late gestational intervals particularly during allogeneic pregnancy. These may constitute two functional subsets: a suppressor class capable of blocking the activation of T lymphocytes (Hunt et al., 1984) or the development of natural killer function in the N K lineage cells (Scodras et al., 1985) and an antigen presenting class (Elcock and Searle, 1985) most likely represented by the I - A ' macrophages identified in the present study. In conclusion, it is reasonable to suggest that most or all cellular components of the decidua may play a major or minor local immunoregulatory role at the feto-maternal interface. Further functional studies with various cellular subsets in vitro, as well as in vivo are dependent on the success of obtaining highly pure populations with cell fractionation procedures.

Acknowledgements This study was supported by the Medical Research Council and the National Cancer Institute of Canada. We thank Sandy Slater for skilled technical help.

References Antebi, S.O., Bauminger, S., Zor, U. and Lindner, H R (1975) Prostaglandin svnthesas m decidual tissue of" the rat uterus. Prostaglandins 10. 991-999. Badet, M.T., Bell, S.C. and Billington, W.D. (1983a) Immunoregulatory activity of supernatants from short-term cultures of mouse decidual tissue, J. R e p r o d Fertil. 68. 351-358. Badet, M.T., Bell, S.C. and Billington, W.D. (1983b) Partial characterization of mlmunosuppresslve factors from short term cultures of murine decidual tissue. Ann. [mmunol. (Pans) 134(.'. 321- 329. Bernard, O., Scheid, M., RJpoche. M.A. and Bennett, I). (1978) Immunological studies of mouse decldual cells, l Membrane markers of decidual ceils Jn the days after implantation. J. Exp. Med. 148. 580-591. Bulmer, J N . and Sunderland. C.A. (1983) Bone-marrow, o n g m of endometrial granulocytes in the earb, h u m a n placental bed. J. Reprod. Immunol. 5, 383-387. Bulmer, J.N. and Sunderland. C.A. (1984) lmmunohistologtcal characteriz,atlon of lymphoid cell population in the early h u m a n placental bed. Immunology 52. 349- 356. Cantor, H. and Boyse, E.A. (1975a) Functional subclasses of T lymphocytes bearing different Ly antigens. 1. The generatton of functionally distinct T-cell subclasses is a differentiative process independenl of antigens. J. Exp. Med. 141, 1376-1389 Cantor. H. and Boyse. E A . (1975b) Functaonal subclasses of T lymph(~:ytes bearing different L.~, antigens, 11. Cooperation between subclasses of Lyt cells in the generation of kdler activity. J. l-xp, Med. 141, 1390-1399.

232 ('hatterjee-Hasrouni, S. (1991) Surface marker characterization of maternal T lymphocyte subsets during pregnancy. Proc. (.'an. Fed. Biol. Sty,.'. 24. 140. Chatterjce-Hasrounl, S. and l.ala, P K . (1981) M H C antigens on mouse trophoblast cell,,. PaucW, of la antigens despite the presence of H-2K and D. J. lmmunol. ]27. 2070 2073. Chatterjce-Hasrouni, S. and Lala, P.K. (1982) Localization of paternal H-2K ant=sen`` on murlne trophohlast cells in vivo. J. Exp. M e d 155, 1679-16149. (`'hatterjee-Ilasrouni, S.. Santer, V. and [,ala. P . K (1990) Characterlsatlon of maternal small [ymphc~'yte subsets durin 8 allogeneic pregnancy =n the mou~.e. (.'ell. lmmunol. 50, 290-3(14, ('hatterjee-Ha.``rouni. S.. Parhar. R. and Lala. P.K. (1984) l'valuation of the maternal natural killer cell population during the course of murme pregnancy. Cell. Immunol. I~4, 264-275. (.'lark. D.A.. Slapsys. R.M.. Croy. B.A. and Rossant. J. (19~4) Immunoregulation o f ho``t ~.er,,u`` graft responses m the uterus. Immunol. Today 5. 111-115. ('lark. D.A., Chaput, A., Walker. C. and Rosenthal. K.I,. (19~5) Act=re suppression of host v,,. graft reaction =n pregnant mice. VI. Soluble suppre~.~.or activity obtained from decidua of allopregnant mice blocks the response to ]L-2..I. Immunol. 134. 1659-]664. ('olavincenzo. V. and Lala, P.K. (1984) In vivo labeling pattern of cells of the routine placenta after a ',ingle injection of radio-labeled antipaternal H-2 antibody. Anat. Rec. 20H. 33A. ('olavincenzo. V. and Lala. P.K. (1985) Ultrastructural Ic~.'alizatlon of antipaternal 11-2 antibody binding m v~vo by cells of the murine placenta. Anal. Rec. 211.42A Craggs, R.I. and Peel, S. (1983) Immunological characterisation of surface receptor', on rat metrial gland cells. J. Reprt~l. lmmunol. 5. 27-37. DeZiegler, D. and Gurpide, E (1982) Production of prolactin by cultures of cells fr¢)m human decldua J Clin End(x:rinol. Metabl. 55. 511 515. Eardley. D.D.. Itugenberger. J., McVay-Bourdeau, l,..Shen. F.W.,Ger``hon, R.K. andCantor, H (1978) Immunoregulatory circuits among T cell sets I. l helper cells reduce other "l-cell ~,ets to exert feedback inhibition..I. Exp. Meal 147. 1106-1115 EIc(~k, J.M and Searle, R F . (1995) AntJgen-prcscntlngcapacity of mou``e decidua[ ti``sue and placenta. Am. J. Repro.d. Immunol. MacrohJol. 7.99-103, Finn. C A and Porter, D.G. (1975) In: The Uteru,,. pp. 74-95, Elek Science, Lmdon. Fou, li.``, D.J. and Ansell, J.D. (19951 Evidence that decidual ceils are not derived from bone marro'~. Transplantation 39. 445-446. (;ambel. P., Ro``sant. J.. Hun,qker. R D. and Wegmann, T.G. (19~5a1 ()rigm of decidual cell`` in murmc pregnancy and p``eudopregnancy. ] ransplantation 39. 443-445. (;ambel, P., Croy, B.A.. M~.~re, W.D.. Hun~,iker. R.D, Wegmann. T.G. and Rossant, J. (1985h) ('haracterlzallon of Lmmune effector cells present m early munne decldua. ('ell Immunol 93. 303-314 Garnis, S. and Lala. P.K (1978) Surface markers of .,.mall lymphoc'ytes appearing m the mouse Ehrhch ascites tumor, host spleen and blo~l. Immunology 34, 487-499, (ilobers~)n, A.. Bauminger, S., Abel. L. and Peleg, S. (19761 Decidual extracts suppress antibod'~ respon``e in vitro. Eur. J. Immuno]. 6. 120-122. Golander, A., Hurley, T., Barrett. J. and Handwerger, S. (1979) Synthe,.is of prolactin b~, human decidua m vitro. J. Endocrinol. 82. 263. Golander. A., Zakuth, V.. Shechter. Y. and Spirer, Z. ( 1991 ) Suppression of lymphocyte reacttvlty m vitro by a soluble factor secreted b>, explant`` of human decidua Eur. J. lmmunol. 11, g49-851. (;reenv..~)d. F.C., Hunter. W.M. and (Hover. J.S. (1963) The preparation of i~1 l-labeled human gr(va.th hormone of specific radioactivity. Biog.'hem. J. 89, 114 -123. Hunt, J.S., Manning. L.S. and W(~)d. G.W. (1984) Macrophages m murme uterus are immunosuppresslVe. ('ell. lmmunol. 85. 499-510 Kearns, M. and Lala. P.K. (1992) Bone marrow origin of dec=dual cell precursors m the pseudopregnant mouse uterus. J. Exp. Med. 155, 1537-1554. Kearns. M. and [.ala. P.K (19831 Life hastor~ of dec=dual cells. Am. J. Reprod. lmmunol. 3. 7g-g2. Kearns, M. and Lala. P.K. (1985) Rad~oautograph~c analysi`` of surface markers on decidual cells shared by cells of the lymphomyelo=d tissues. Am. J. Reprod lmmum)l. Microblol. 9, 39 47 Kearns. M. and Lala, P.K. (1985a) Mom~lonal antibodies defining marker(``) umque to the decidual cell lineage. Transplant, Prc~.'. 17. 911-915.

233 Kearns, M., Parhar, R.S. and Lala, P.K. (19851 Decldual cell specific surface antigen(s) recognimd b', monoclonal antibc~lzes: tissue and species distribution J Immunol. 135, 1046-1052. Ktrkwt~'~.t.K.J. and Bell, S.C. (1981) Inhibitory activity of supernatant~, from murme decidual ,:ell cultures on the mixed lymphocyte reaction. J. Reprod. lmmunol. 3, 243-252. Kopriwa, B.M. and Leblond, C.P. (1962) Improvements in the coaling technique (:.f radioautography. J H i s t o c h e m C y t o c h e m 10. 269-284. Krcek, J.P.. Dickson, A.D. and Biddle, F.G. (19831 Mononuclear cell accumulations in the trophohlastlc grant cell circulation tn inbred stratus of mice. J. Anal. 136. 293-292. Krehbiel. R.H. (1937) Cylochemical studies of the decidua] reactlon in the rat during earl v pregnancv and m the prc~luction of deciduoma. Physiol Zc~:,l. I0. 212-235. Lala. P.K. and Pall, II.M. (1966) Cytokinetic analysis of tumor gro~.lh. Proc'. ,N]at]. Acad. Scl. U.SA. 56. 1753- 1942. Lala. P.K.. ('ha/terjee-Hasrouni. S.. Kearns. M., Montgomer.~. B. and Colavincenzo. V. (19831 Immunobiology of the feto-maternal interface. Immunol. Rev. 75. 87-II6. Lain, P.K., Kearns, M and £'c, lavincenzo. V. (1984) ('ells of the fclomalernal interface: their role in the maintenance of viviparous pregnancy. Am. J. Anal. 170. 5 0 1 - 5 1 7 Lain, P K . , Kearn.,.. M. and Parhar. R.S. (1985} Immunobiology of the decidual tissue: the maternal component of the fete, maternal interface. In lmmunoregulalJon and Fetal Survival (Gdl. 1. and Wegmann, T.G., eds.). Oxford University Press. New York (m press). [,edbetter, J.A. and Herzenberg (1979) Xenogeneic mom~lonal antibodies to mouse lymphoid differenlaalion antigen.,,, lmmunol. Rev. 47. 63 -90. Loveland, B.E. and McKenzic, I.F.C. ll9g2) Which T cells cause grMt rejection. "lransplanlation 33. 217-220. Mitchell. B.S.. Craggs, R. and Peel. S. (198(1) Localisation of immunoglohulin (IgG} within the rat metr~al gland. J. Reprod. Immunol. 2. 235-244. Montgomery. B. and Lala, P.K. (I 98310ntogeny of the M I l l " anlagens on human trophoblast cell.,, during the first trimester of pregnancy. J. Immunol. 13], 2348-2355 Oi. V.T.. Jones, P.O. G~xling. J.W.. Herzenberg. L.A and flerzenberg, L.A. (19781 Properties of monoclonal antibodies to mouse lg al]otypes. H-2 and la antigens. ('urr. Top. Microbiol. Immunol. 81, 115-129. Padykula. H.A. and Tansey. T.R. (19791 3"he occurrence of uterine stromal and intraepithelial monocvles and heterophils during normal late pregnancy in the rat. Anal. Rec. 193. 329-355. Parhar. RS. and Lain, P.K. (1985) Local immunosuppression of lymphocyte alloreactivily by human decidual cells. Anat. Rec. 211. 147A. Peel. S.. Stewart, I. and Bulmer, D. (1979) Metrial gland cells in deciduomata of pseudopregnancy. J. Anal. 129. 21-30. Peel. S.. Stewart. I. and Bulmer. D. (1983) Experimental evidence of the bone marrow origin of granulated metrtal gland cells of the mouse uterus. Cell Tissue Res. 233, 647-656. Riddick, D.H. and Kusmik. W.F. (1976) Decidua: a possible source of amniotic fluid prolactin. Am. J. Obstet. Gynecol. 127. 187-19~). Scodras, J.M.. Parhar. R.S. and Lala, P.K. (1985) Inactivation of natural killer lymphocytes m the murme dectdua by decidual cells and macrophages. Anat. Rec. 211. 171A. Siiteri. P.K. and Seron-Ferre, M. (1981) Some new thoughts on the feto-placental unit and parturition m primates. In: The Oregon Regional Primate Research Center Symposia on Primate Reproductive Biology. Vol. ] (Fetal Endocrinology). Academic Press. New York. Siiteri. P.K.. Febres. F.. Clemens, L.E. Chang. R.J.. Gondos, B. and Stltes. D. (1977) Progesterone and maintenance of pregnancy: Is progesterone nature's immunosuppressant? Ann. N Y. Acad. Set. 296. 384- 394. Slapsys, R.M. and Clark. D.A. (1983) Active suppression of host-versus graft reaction in pregnant mice. V Kinetics, specificity and m vivo activity of non-T suppressor cell.,, localized to the genital tract of mice during first pregnancy. Am. J. Reprod. immunol. 3, 65-71. Spiegel. M.R. (1961) In Theory and Problems of Statistics. pp. 69-70 and 188-191 (Schaum Publishing Co.. New York).

234 Springcr. "1., Galfre, G., Secher, D.S. and Milsteln, ('. (1978) Monoclonal xenogenelc antibc, dies to murmc cell surface antigens: identification of novel leukoc'yte differentiation antigens. Eur. J. lmmunol. 8. 539-551. Sunderland. ('.A.. Redman. C.W.G. and Stirrat, G.M. (1981) HLA A, B, C antigens are expressed ~,n nonvillous Irophoblast of the early h u m a n placenta. J. immunol. 127. 1614-1615. Swain. S.. Bakke. A . English. M. and Dutton. R.W. (1979) L.y phenotypes and M H ( recogmtlon: the allohelper that recognizes K or D i~, a mature Ly I. 2. 3 cell. J. lmmunol. 123. 2716- 2724. Wdliam.,,. K.l. and Downing. I (1977) Prostaglandin and thromboxane production by rat decidual mlcrosome.s. Prostaglandins 14. 813-817.