SIEP publications are protected by copyright and all rights are reserved.  SIEP publications may not be reproduced in any form or by any means without written permission from the copyright owner. This includes the posting of electronic files on the Internet, transferring electronic files to other persons, distributing printed output, and photocopying. Requests for authorization to make photocopies should be directed to:   SIEP, 1697 Lark Lane, Cherry Hill, NJ 08003-3157, Tel: (856) 429 2699, Fax:  (856) 429 7414, Email: barnea@earlypregnancy.org

EARLY PREGNANCY: Biology and Medicine Editor-in-Chief: Eytan R. Barnea MD, FACOG

January 2003
Volume VI, Number 1
ISSN: 1537-6583
Pages: 248-271

R. Demir (1), U.A.Kayisli (1,2), C. Celik (1), E.T. Korgun (1), A.Y. Demir-Weusten (3)

(1)Department of Histology and Embryology, Faculty of Medicine, Akdeniz University, Antalya, Turkey; (2)Department of Obstetric and Gynecology, Reproductive Endocrinology Section, Medical School, Yale University, New Haven, CT; (3)Department of Obstetric and Gynecology, Research Institute Growth and Development (GROW), Maastricht University, Maastricht, The Netherlands

Key words: decidual granular leukocytes / early pregnancy / development / ultrastructure

Correspondence: Prof. Dr. Ramazan Demir, Department of Histology and Embryology, Faculty of Medicine, Akdeniz University, Campus, 07070 Anytalya-TURKEY, Tel and Fax: 0090-242-2274486, email: demir@med.akdeniz.edu.tr

Acknowledgement: We thank Hakan Er and Arife Demircan for their excellent technical assistance. The study was financially partly supported by the Research Fund of Akdeniz University, Antalya, Turkey, grant no:20.01.0102, and supported by the TEMGA in Medical Faculty as instrumentally.

Decidual granular leukocytes (DGLs) are the dominant cell populations in first trimester decidua, but the information about ultrastructural diversity of DGLs with immunohistochemical observations in early pregnancy is lacking. We hypothesised that human decidua could be one of the sites where differentiation, and maturation of DGLs occur. DGLs were examined fine structurally between 18-38 days post-conception (p.c.). Ultrastructurally, four main cell types of DGLs were distinguished; immature-DGLs, immature intermediate-DGLs, mature clear-DGLs and mature dark-DGLs. A cellular barrier formed by DGLs were observed, separated by a limiting membrane beneath the glandular epithelium. DGL fragments forming a compact structure around eroded capillary were seen. These ultrastructural diversities reflect the presence of DGLs at different stages of maturation, or with different levels of functional activity. The cellular barrier suggests a protective interaction between glandular epithelium and DGLs cyto-toxicity against trophoblastic invasion. DGL's are possibly natural effector cells, probably protect the materno–fetal unit from other effects causing disorders to placental development.

Introduction

The structures in pregnant human endometrium / decidua are the decidual glands, vessels and decidual stroma. The stroma structures are composed of main cell types: decidual cells, leukocytes including macrophages, T lymphocytes and large granular lymphocytes (LGLs), and undifferentiated mesenchymal cells; they are all loosely connected with each other. The uterine mucosa and decidual cells are regulated distinctly by steroid hormones and paracrine factors (Classen-Linke et al., 1998). The number of uterine NK (uNK) known as LGLs is changeable in the proliferative and in the mid-secretory phases endometrium (Bulmer et al., 1988) but increases and reaching a maximum in the late secretory phase of the cycle (Stewart-Akers et al., 1997; King et al., 1998), and they are particularly abundant around the time of implantation and during early pregnancy (King et al., 1998; Chao et al., 1999). These uNK cells have an unusual phenotype (King et al., 1999) and may control trophoblast migration (Chao et al., 1999), and indeed these cells seem to accompany the invading trophoblast (Beer et al., 1998).

In the human uterus, the defensive cells present at the feto–maternal interface are macrophages and uNK cells while the other bone marrow-derived cells are absent and thymus-derived cells are sparse (Loke and King, 1995). The presence of various activation markers has been demonstrated on CD4, CD8, CD56 positive cells and on T cells isolated from human decidual tissues (King et al., 1991; Saito et al., 1992). The increase in CD56+ cells in the luteal phase could be due to an influx of differentiated cells from blood or to proliferation of CD56+ cells in uterus from precursor cells (Loke and King, 1995). Proliferation has been shown to occur in vivo both by assessing mitoses of granulated cells (Pace et al., 1989) and more specifically by double immunostaining for CD45 or CD56 and the proliferation marker, Ki-67 (Tabibzadeh, 1990).

The endometrium contains a large number of macrophages (Hunt, 1994). Some stromal cells with oestrogen receptors respond to steroid hormones and they stimulate uterine growth factors attracting macrophages (Hunt and Robertson, 1996; Miller and Hunt, 1996). Distribution and density of macrophages patterns in cycling and in decidua changes due to the circulating estrogens and progesterone levels (Arck et al., 2000).

Recent studies have made evident that human endometrium expresses the cytokines such as stem cell factor (SCF), Interleukin 1 and 2 (IL-1, IL-2) and macrophage colony-stimulating factor (MCSF) regulating the proliferation and survival of hematopoietic cells (Kauma et al., 1996; Umekage et al., 1998; Kauma, 2000). These factors may be a source to trigger undifferentiated stromal mesenchymal cells of  human endometrium to differentiate distinct subtypes of decidual granular leukocytes (DGLs) (our unpublished data). Although the number of these cells increase at the implantation site, there is not enough study concerning their ultrastructural diversity and functions. Here, we used the term DGLs meaning of common term for all decidual granular leukocyte cells, because we couldn't apply some special antibodies belong to immunoelectronmicroscopical markers and phenotypic receptors for identification of the leukocyte cell sub-populations. But, we know that there was a striking diversity and clear definition between well-developed fine structure of decidual granular leukocyte and conventional leukocytes in the endometrium. These could be different with ultrastructural properties from conventional processes.

Previous ultrastructural studies of decidua have been based on limited conventional electron microscopic findings of different cell populations (Wynn, 1974; Tekelioğlu-Uysal et al., 1975). Although these studies have described leukocyte populations in human decidua in the first trimester of pregnancy, but there have not enough information about the ultrastructural diversity and maturation aspects of the decidual leukocytes with immunohistochemical findings during very early pregnancy. Since many hematopoietic factors are described in human endometrium during early pregnancy we hypothesised that human decidua could be one of the sites where differentiation and maturation of DGLs occur. For this, the uniqueness of our samples is their particularly early gestations age. Information about the ultrastructure of decidual leukocytes early in gestation is lacking and this sample group is potentially important. Our aim is to show that decidual granulocytes show ultrastructural diversity due to the development and differentiation process in decidua during early pregnancy. In the present study, the entire range of human DGLs was examined fine structurally during early pregnancy (between 18-38 days post conception, p.c.).

Samples of human decidual tissue were obtained after legal termination of pregnancy by curettage for medical or psychosocial reasons, which were unlikely to affect decidual structure and function. None of the pregnancies were receiving hormone treatment; all of them were normal pregnant. Tissues were supplied from Department of Obstetrics and Gynecology, Medical Faculty, Akdeniz University, and Clinic of Obstetrics and Gynecology, Government Hospital, Antalya.

Tissue collection and storage
Samples of human decidual tissues from 21 women were studied. Early specimens (n = 8, aged 18 to 25 days post conception, p.c.; n = 6 aged 28 to 33 days p.c.) obtained by curettage or by hysterectomy, were classified as described in our previous study (Demir et al., 1989) and by studying embryonic developmental details following the Carnegie classification (O’Rahilly, 1973). Additional specimens (aged 35 to 38 days p.c. n = 7) were assessed by history and measurement of fetal weight and of length at the end of week 5 and the beginning of week 6 of pregnancy were obtained from clinically normal pregnancies interrupted by legal abortion or hysterectomy. The samples obtained between day 18 and 33 p.c. were examined for ultrastructural studies. In all cases, patient consent and local ethical committee approval were obtained.

Tissue preparation and evaluation of micrographs
Samples of human decidual tissue were fixed by immersion in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) at room temperature for 4 h, post-fixed in 1% phosphate-buffered osmium tetroxide for 2 h. The specimens were dehydrated in ascending concentrations of ethanol series and were embedded in Araldite-epoxy resin. Semi- and ultra-thin sections were taken by ultramicrotome. Semi-thin sections were stained with toluidine blue. Thin sections were double-stained with uranyl acetate and lead citrate and examined with a Zeiss EM10 and LEO906.

Measurement of fine components
To evaluate the components of the DGLs, fifteen transmission electronmicroscopy (TEM) micrographs standardized under a same magnification (2.500x original degree) with a Zeiss EM10 and LEO906, were chosen randomly between 36 micrographs, and all of the DGLs were countered. Certain cellular components were measured by a computer based image analyzer system (SAMBATECH, France, 1999), and described as; i) total numbers of DGLs and their granules, ii) the area of the cell and granule (µm²), iii) perimeter of the cell and granules (µm), iv) maximum and minimum diameter of cell and granules (µm) and v) shape of the cells and the granules as form factor. Following the quantification of all the fine structural parameters, electronmicrographs were divided into two group as early period and advanced period of early pregnancy (between days 18-25, and 26-38 of early pregnancy respectively), except the examination of the nuclei. The lobulation of nuclei were calculated totally for DGLs. The evaluation of the ultrastructural quantification was also made out for the nuclei of DGLs that were divided into three groups due to their lobulation appearing one, two, three and more lobes, and the means of them were calculated.

Statistical analysis
Statistical analyses for ultrastructural scores and data were normally distributed (tested by Kolmogorov-Smirnov test) throughout the pregnancy days. Analysis of variance (ANOVA) and the Tukey test were carried out for statistical analysis and pairwise multiple comparisons. Statistical calculations were performed using the Statistical Package for social sciences (SPSS, Chicago, IL, USA) and Sigmastat for Windows, version 2.0 (Jandel Scientific Corporation, San Rafael, CA, USA).

Ultrastructural observations
The ultrastructure of the DGLs varies considerably during the early part of pregnancy. Therefore, when examining the ultrastructure of the DGLs, the exact date in early pregnancy at which the large granulocytes were sampled should be specified. In most cases, DGLs outnumbered decidual cells in the implantation areas and these were clearly observed on comparative draw (Figs.1a, b) and TEM micrographs (Figs. 2, 3).

The cellular diversities were examined and counted quantitatively in two groups; early period between 18-24 days, and advanced period between 26-38 days of early pregnancy, and percentage for the each cell type was calculated (Fig. 4a). The majority of DGLs were observed ranged in size from 6.22 to 13.28 µm and had an irregular shape with eccentrically located nucleus. The means of the original sizes for DGLs and for their granules were calculated for five parameters between days 18 and 38 of early pregnancy (Figs. 5a-c). The cell surface usually formed numerous cytoplasmic thick extensions and thin processes. These processes were often established in areas of contact with other cells. The DGLs were never covered by a pericellular lamina (Figs. 2, 3). Some of the DGLs exhibited a characteristic elongated shape with a constriction of the cytoplasm and the nucleus, which is a feature of migrating cells (Figs. 3, 6). Small vacuoles of varying size were observed in the cytoplasm.

Diversity of DGLs
The development processes and the diversity of DGLs were classified as indicated in Figs. 1a, b. During the very early period of pregnancy, between days 18 and 38 post conception (p.c.), the DGLs have not attained a fully differentiated state. Abundant DGLs were observed including immature and mature forms as well as intermediate type.

Subtypes of DGL population
DGLs showed diversification due to their maturation periods and different ultrastructural aspects. Three subtypes of DGLs were observed as follows: (i) immature DGLs (im-DGLs); (ii) mature DGLs (m-DGLs) that were fully differentiated with electron dense granules, and (iii) immature intermediate DGLs (imi-DGLs) between sub-populations (i) and (ii) (Figs.1a, 2). m-DGLs were observed in different two cell types as mature clear DGLs (mc-DGLs), and mature dark DGLs (md-DGLs) (Figs. 2, 7) Additionally, with some ultrastructural differences between every each subgroups were also seen. During statistical analysis, im-DGLs and imi-DGL cells were accepted as immature cell group. The granule number means of the md-DGLs and mc-DGL cells during these days were compared. The mean of the granule number for per md-DGLs was significantly higher than that of mc-DGLs in the early and advanced period (p<0.05) (Fig. 4b). On the other hand, the difference between the early period and advanced period for the granule number of md-DGL or of mc-DGL cells in each other was not statistically significant (p>0.05).

Precursors of DGLs
During implantation, many DGL precursors were recognized in the implantation areas (Figs. 2, 3). The first recognizable precursors of DGLs were regular spherical cells, which showed an immature lymphocyte-like shape (Figs. 2, 3). The cytoplasm was lightly stained with definitive organelles. Even at this early stage of maturation, they had some cytoplasmic pseudopodia-like extensions on their surface which was the same as observed in lymphocyte, as was their contour. The nucleus was oval-shaped and there were very abundant organelles: except a few small bodies, neither secretory granules, nor their multivesicular precursors were found. The nuclei of these precursors showed a striking difference while differentiating to the mature and intermediate forms of DGLs: a broad electron dense rim of darkly staining heterochromatin was observed on their periphery. The typical polymorph or polylobular and kidney-shaped nucleus of the mature cells were not observed in these cells population. Some of the precursor cells exhibited a characteristic elongated shape with a constriction of the cytoplasm and nucleus (Figs. 3). The imi-DGLs with clear cytoplasm including more organelles and very few small granules were seen near the immature DGLs (Figs. 2).

Mature DGLs
According to the staining of cytoplasmic contents and the granules two mature DGL types were recognized: mc-DGLs and md-DGLs. The first one had a clear cytoplasm with big and rare dense granules, and the second one had a dark cell cytoplasm with abundant number of granules in different sizes (Figs. 2, 6, 7). The nuclei of the clear cells, generally, showed one or two lobes, but in dark cells the nuclei generally showed more than two and three lobes in different shapes (Figs. 2, 6, 7). An increase in the number of mc-DGLs was observed while the number of md-DGLs tended to decrease as pregnancy progressed (Fig. 4a). These changes for these cells were not statistically significant (P>005). The characteristic irregular shape was associated with the cell surface modulated by cellular extensions, which varied in diameter. These cellular extensions appeared to be arranged parallel to the long axis of the cell and unusual microvilli-like or cytoplasmic pseudopodia were frequently observed for the both cell types (Figs. 2, 6, 7). Many pseudopodia-like cytoplasmic extensions with granules were observed on the cytoplasmic surface, the content was varying from 0.5 µm up to 2 µm in length and from 0.2 µm to 0.8 µm in width.

The nucleus was generally situated eccentrically. Because of the lobular shape of the nucleus, cells sometimes appeared having more than one nucleus: generally 1-4 nuclear lobes were seen per cell section (Figs. 3, 7). The total number of the DGLs counted in fifteen TEM micrographs for the comparison of the nuclear lobes was 184. The percentages of the nuclei with one lobe, two lobes or three and more lobes were %31.52, % 40.75 and %27.72 respectively (Fig. 8). The nuclei of mature DGLs showed a broad rim of heterochromatin, which were many times wider than that found in the precursors of all types. Dense chromatin was accumulated peripherally and occupied 2/3 part of the nucleus. Generally, young type of the dark cells showed one or two lobes, but full mature dark DGLs have three or four nuclear lobes with highly dense chromatin condensation. Full mature dark DGL was a typical neutrophil-like leukocyte. Mitotic stages could not be distinguished in mature DGLs. Sub-cellular organelles did not appear to be different from those in other cells. In addition to these cell populations, we have observed another cell type of DGL that are free cells or attached to endothelial cells in destroyed capillary areas. These cells are distinguishable from the other cells because of the darker cytoplasm including vacuoles and dense bodies, elongated cytoplasmic projections on the surface, and polymorph nuclei with dispersed broad rim of dense heterochromatin, and with distinctive nucleoli was observed only in capillary areas (Fig. 9). These cells gave macrophage-like cells impression. Around the areas that were close to the capillary, DGLs were surrounded by accumulation of cell fragments with or without secretory granules (Fig. 2); limiting destroyed capillary wall (Figs. 6, 9). Interestingly, many m-DGLs were arranged in side by side as a defensive cell line, and limited by a thick membrane from sub-glandular area, beneath the glandular epithelium (Fig. 10). This very interesting observation is declared first this  study. Regular stromal cells were also found between the basal lamina of glandular epithelium and the limiting membrane of DGLs line.

Secretory granules
A striking feature of mature DGL was the presence of secretory granules with different sizes and diameters. The majority of these granules were round to oval and measured between 0.37 and 1.48 µm. Some granules were electron dense while others were translucent, making them appear empty. On sections throughout the DGL cells, an average of 60-80 granules were counted. On the average electron micrographs encompassing the cellular areas containing granules, more than a thousand granules have been counted. The means of granule numbers were compared and analysed statistically for early and advanced days of early pregnancy (Figs. 4b, 5c). Three different kinds of secretory granules were observed (Fig. 11). (i) The limiting membrane was electron dense and in direct contact with vesicles; (ii) moderately electron dense including some multivesicular structures; (iii) precursors of young secretory granules including multivesicular contents. All intermediate stages of development of the secretory granules were observed ranging from vesicles to secretory granules, which filled with electron dense or translucent content (Fig. 11).

Most secretory granules were membrane-bound but some non-membrane-bound granules were also observed in the same cell. During the development and maturation periods, secretory granules were surrounded by a single membrane and showed varying density condensation, which spreads homogeneously, according to the stage (Figs. 3, 11). Granules were dispersed in the cytoplasm and formed accumulations in some areas. Secretory granules were found side by side in the same cells. Some of them were situated close to the plasma membrane, and the number, the sizes, and the contents of granules were different due to the cell type characters; the granules of mc-DGLs were bigger, denser and but less in number than the granules of md-DGL cells (Figs. 2, 6, 7).

Fragmentation of DGLs and Secretory Granules Discharge
Secretory granules were also dispersed in cytoplasmic extensions, and these extensions were fragmented from the cell body to form an abundant accumulation like a compact structure between the DGLs cells and the capillary wall (Figs. 2, 3, 6, 7, 9). The representative sections through some DGLs, extensions containing the secretory granules were connected to each other by narrow bridges of cytoplasm. The secretory granules in these cytoplasmic extensions and the cell body were heterogeneous with respect to both their size and structure. A striking feature of mature DGLs was the presence of primary and secondary protrusions from the cytoplasmic extensions (Figs. 2, 9). These were only found when secretory granules were present. More fragmentation took place in these secondary cytoplasmic extensions. In this process, granules were not reduced in size through detachment of peripheral areas; the secretory granules were discharged from the cell by exocytosis and by fractionation of the cytoplasm (Figs. 2, 6, 7, 9). The fractionation was very interesting, and the mechanism was similar to that of apocrine secretion. Another mechanism for discharge appeared to be occurred by detachment of cellular extensions, which were frequently found in the areas remote from any granules. During normal implantation there seemed to be a strong discharge of secretory granules into the intercellular spaces. Of the abundant granules, most were found free rather than fused to each other in small fragments (Figs. 3, 6, 7). Following the discharge process, secretory granules became less electron dense and almost foamy appearance. Furthermore, we observed fibrillar type collagen replaced with the more homogenous substances in extracellular spaces as pregnancy progressed (Figs. 7).

Discussion

Human decidua is a leukocyte-rich tissue. Macrophages, LGL or NK cells, T cells and other DGLs are the most abundant leukocyte populations in decidua of first trimester. DGLs account for up to 70% of endometrial leukocytes (Stemmer, 2000), 10 – 15 % of all cells are leukocytes (Mincheve-Nilsoon et al., 1994). The maternal leukocytes present in human decidua during early pregnancy are composed of CD56+ uNK cells (70-80%) (Johnson et al., 1999), CD3+ T cells (10%), and CD14+ (10%) (Beer et al., 1998). The percentage of the distribution ratio of uterine leukocytes in our recent study (Özenci et al., 2001) was ~40% CD45+ leukocytes, and ~30% of these leukocyte was CD56+ cells and ~22% was CD14+ cells at weeks 4-5 of early pregnancy. The remaining part was probably T cells and other leukocytes. This difference may be due to the earlier stage of pregnancy observed in our recent study compared to previous studies (Bulmer et al., 1991; Beer et al., 1998) in which the whole first trimester was included, but not like our material that belonged to very early days of pregnancy. In our results, the density of CD56+ cells distribution were variable due to the decidual portion, and their number changed from 12% to 50% in total leukocyte population as pregnancy progressed (Özenci et al., 2001). There is not a real accordance for the percentages and their statistically significance of the decidual leukocytes between investigators were given in recently published reports. These findings suggest that while uNK-like cell population was increasing, the cell populations such as T cells and other leukocytes were decreasing in parallel to pregnancy age. Recently, a supporting data has reported by Vince and Johnson (2000) that T cell numbers decrease significantly in first trimester decidua, and they play an important role in the immunological maintenance of human pregnancy and they could be an important for implantation. So, the increase in numbers of uNK cells and macrophages at the implantation site strongly suggests that uNK cells are involved in the maintenance of normal pregnancy (Loke and King, 1995), and the aim that implantation is likely to involve an innate immune system becomes (Looke and King, 2000). Many immunological studies have been carried out on the decidua in recent years since the decidua was involved in the suppression of cellular and humoral interaction between maternal and embryonic side (Classen-Linke et al., 1998). It was suggested that the immunocompetent cells limited the deep invasion of trophoblastic cells into the decidua (Peel et al., 1982; Bulmer and Sunderland, 1983). Our recent results (Özenci et al. 2001) agree with this conclusion and the presence of abundant CD56+ cells throughout the human decidual stroma may be associated with the deep invasion of trophoblast. On the other hand, the ultrastructural findings of close localisation of DGLs to capillary wall and periglandula may be the main defensive cells against trophoblast over-penetration or against pathogenic agents originated from peripheral blood into decidua compared with the less numerous of other leukocytes. Ultrastructurally, the peri-glandular and vessel wall distribution of DGLs is remarkable in decidua, and the cells were found in an peri-epithelial location. These ultrastructural results are in accordance with our recent immunohistochemical observations, the DGLs can be visualised moving through the cell clusters, and limiting destroyed capillary wall by their dense fragments including loot of cytoplasmic remainders with or without secretory granules. These results suggest that these fragments could form a physical and immunological barrier against invasive trophoblasts.

In this study, sub-endothelial and intra-epithelial or peri-glandular localisation of the DGLs cell populations either in large cell aggregates or as individual cells suggests the local interactions of leukocytes with the epithelial or endothelial cells. In our ultrastructural observations, around the eroded capillary, formation of compact structure like a physical barrier by DGLs and their fragmentation could exhibit either protective mechanism for capillary wall or facilitation processes for the trophoblastic invasion. Furthermore, sub-glandular lines of the DGLs forming a set in the peri-glandular area could be related with the protection of the glandular epithelium since the conceptus is nourished by the uterine milk produced by the glandular epithelium (Demir and Akkoyunlu, 1999). The findings of the present study suggest that the glandular epithelium could be protected by a paracrine mechanism against trophoblastic destruction. In the case of the DGLs accumulation beneath the glandular epithelium, DGLs may be provoked by interleukin-8 expressed in human endometrial glandular cells (Arici et al., 1998), and monocyte chemotactic protein-1 (MCP-1) expressed in human endometrial stroma (Arici et al., 1999). Furthermore, this may be carried out by macrophage migrating inhibitory factor (MIF) localized in glandular epithelium as well as in decidual cells during first trimester (Arcuri et al., 2001).

Immunohistological studies have indicated that nearly 20% of leukocytes in both the endometrium and decidua are macrophages, identified by monoclonal antibodies to CD14+ (Loke and King, 1995). In our recent study (Özenci et al., 2001), 22% of the leukocytes were CD14+ cells in the fourth week of pregnancy, and these cells showed an increase like CD56+ cells as gestational age progressed. These results suggest that CD14+ cells that are located especially around vessel walls may have a role in regulation of fetomaternal blood flow, and may facilitate trophoblast invasion into the vessels. So the key to their presence in the decidua may be the distribution pattern in this immune-privileged tissue. Our ultrastructural observations of macrophage-like cells in capillary lumen support this functional process.

DGLs were heterogeneous with respect to the type of the cytoplasmic contents, ultrastructurally; mc-DGL, and md-DGL cells contained abundant cytoplasm granules with high electron dense and multivesicular bodies that are strong characteristics of cytotoxic cells (Koizumi et al., 1991). Recent studies have shown that human endometrium expresses the cytokines such as stem cell factor (SCF), Interleukin 1 and 2 (IL-1, IL-2) and macrophage colony-stimulating factor (MCSF) that regulate the proliferation and survival of hematopoietic cells (Kauma et al., 1996; Umekage et al., 1998; Kauma, 2000).

These factors could trigger undifferentiated stromal cells of human endometrium to differentiate to distinct subtypes of DGLs. Briefly, we have not been able to apply the specific antibody for immunoelectronmicroscobic identification, but we clearly described ultrastructural diversities during differentiation and maturation of the DGLs during early days of pregnancy. In the present study different types of DGLs were described according to their ultrastructural features between immature and mature cell types. This ultrastructural identification of progenitor, precursor and mature cells could reflect that DGLs could be developed from undifferentiated mesenchymal cells derived from the uterus stroma. From immature intermediate cell form to mature type were identifiable. Proliferation and differentiation has been shown to occur in vivo both by assessing mitoses of granulated cells (Pace et al., 1989, Tabibzadeh, 1990).

We have found that some of the DGLs have a reniform nucleus, short stubby cytoplasmic projections, and membrane-bound granules of varying size with or without a peripheral rim. Additionally, the present study described for the first time that; uterine DGLs undergo cellular diversification due to maturation periods in decidua during early pregnancy. Immature DGLs and mature types were observed as wall as immature intermediate cell type in same areas during early pregnancy. This is very important for cellular diversification and maturation processes. DGLs were statistically examined in two groups according to gestational days of early pregnancy (early and advanced periods), and when the percentage for the each cell type was compared, there was a significant increase for mc-DGL in advanced group. On the other hand, the mean of the granule number for per md-DGL was significantly higher than that of mc-DGL cells in the early and advanced period (* p<0.05). There is not a statistically significant difference between these periods for the cell dimensions or the granule dimensions (P> 0.05). The percentages of the nuclei with one lobe, two lobes or three and more lobes were 31.52%, 40.75% and 27.72% respectively.

With polymorph lobule heterocromatic nuclei are very similar to a neutrophy-like leukocyte, but with abundant electron dense granules they are different from a regular neutrophyl. The increase of nuclear lobulation (68.47% respectively) and abundant granules is indicating that these cells are in process of differentiation and transformation. Our results reveal that these cells are different ultrastructurally from typical LGL in peripheral blood (Kang et al., 1988) and they are not the same cell types that are activated as CD56+ stained cells, which demonstrated features of uterine granulated lymphocytes (King et al., 1991, Mincheve-Nilsoon et al., 1994). Our findings indicated that distinctive ultrastructural features of md-DGLs absolutely have different ultrastructural characters, they are similar to neutrophyl leukocyte with some features.

Free granules found in the extracellular spaces suggest that excretion of granules occur. Discharged secretory granules with cytoplasmic fragments were frequently observed in the intercellular spaces near the DGLs and other stromal cells of the decidua. They appear to be discharged from the DGLs by an apocrine mechanism. Another mechanism of discharge is likely to be the detachment of cellular extensions, which were frequently found in areas remote from any DGLs. Most discharged granules were fused to each other and were composed of electron lucent substances that accumulated in varying shapes outside the cells. These ultrastructural findings are interesting as they differ from the electron microscopic studies reported before (King et al., 1991; Mincheve- Nilsoon et al., 1994).

On the other hand, soluble collagen-like electron lucent substances were accumulated in intercellular spaces, and situated near to DGLs. The replacement of fibrils type collagen with the more homogenous substances in extracellular spaces could be resulted from the digestion of these fibrils by matrix metalloproteinases.

The content of granules in uNK cells in rodent uterus has been reported (Parr et al., 1990) with undescription exactly. But, the contents of the granules in human DGLs have not yet been analyzed; granules from the two cell types (mc-DGLs and md-DGLs) were distinguishable. The content of granules is unknown but the possibility is that the granules contain the factors modulating the maternofetal immune system and trophoblastic invasion. Stemmer (2000) has suggested that all decidual leukocytes expressed activation markers and MHC class II. Against the trophoblastic invasion, expression of several proteins from the epithelial cells may be induced to prolong the structural and functional protection. Alternatively, some specific proteins in reactive T cells (Haregewoin et al., 1989) have been suggested to fulfill a surveillance function, and these reactive T cells could participate in the remodeling of decidua (Mincheve- Nilsoon et al., 1994).

The immunohistochemical findings (Özenci et al. 2001) correspond with ultrastructural observations for DGLs accumulation around the glandular epithelium, and around the eroded capillary. Thus, the DGLs most probably are the important sites for regulation of local immunologic response against embryonic substances derived from trofoblasts. During the early stages of pregnancy, the decidua interacts with extravillous trofoblasts (EVT) (Kaufmann and Castellucci, 1997) that are sometimes localized subepithelial to the decidual glands. On the other hand, normally, a strategic adaptation by EVT occurs, against destruction performed by DGLs cells. Of course, this adaptation must take place alongside invasion of the decidua, and a possibility is that EVT growth is regulated by HLA-G, and restricted by decidual lymphocytes (Mincheve-Nilsoon et al., 1994).

In summary, ultrastructurally, four main cell types of DGLs could be distinguished in this study; im-DGLs, imi-DGLs, mc-DGLs and md-DGLs. This ultrastructural identification of four cell types could reflect the presence of DGLs at different stages of maturation, or with different levels of functional activity. This cellular diversity suggested that DGLs could be derived from undifferentiated mesenchymal cells of uterus. These ultrastructural results could be helpful in elucidating the mechanisms of materno–fetal immune interaction during early pregnancy. These cells are situated in the vicinity of decidual glands, suggesting that glandular epithelium played a role in the activation processes. Human decidua from early pregnancy between days 18 and 38 p.c. of pregnancy contains numerous DGLs and their aggregates, cell clusters with CD45+, CD56+ and CD14+ cells (Özenci et al. 2001) active against the semi-allogenic embryo. Neutrophyl-like granulated decidual cells are possibly natural suppresser cells and probably protect the materno–fetal unit from other effects causing disorders to placental development.

For personal use. Only reproduce with permission from SIEP.

1. Arck PC, Hertwing K, Hagen E, Hildebrandt M, Kalpp BF. 2000. Pregnancy as a model of controlled invasion might be attributed to the ratio of CD3/CD8 to CD56. Am J of Reprod Immunol 44:1-8
   
   

2. Arcuri F, Ricci C, Letta F, Cintorino M, Tripodi SA, Cetin I, Garzia E, Schatz F, Klemi P, Santopietro R, Paulesu L. 2001. Macrophage migration inhibitory factor in the human endometrium: expression and localisation during the menstrual cycle and early pregnancy. Biol Reprod 64:1200-1205
   
   

3. Arici A, Seli E, Senturk LM, Gutierrez LS, Oral E, Taylor HS. 1998. Interleukin-8 in the human endometrium. J Clin Endocrinol Metab 83:1783-1787
   
   

4. Arici A, Senturk LM, Seli E, Bahtiyar MO, Kim G. 1999. Regulation of monocyte chemotactic protein-1 expression in human endometrial stromal cells by oestrogen and progesterone. Biol Reprod 61:85-90
   
   

5. Beer AE, Joanne YH, Kwak JY. 1998. Immunology of Normal Pregnancy. Immunol  Allergy Clin North America 18:253-254
   
   

6. Bulmer JN, Morrison L, Longfellow M. 1991. Granulated lymphocytes in human endometrium: histochemical and immunohistochemical studies. Human Reprod 6:791-798
   
   

7. Bulmer JN, Sunderland CA. 1983. Bone marrow origin of endometrial granulocytes in the early human placental bed. J Reprod Immunol 5:383-387
   
   

8. Bulmer JN, Lunny, D. P., and Hagin, S.V. 1988. Immunoregulatory characterisation of stromal leukocytes in non-pregnant human endometrium. Am J Reprod Immunol 17:83-90
   
   

9. Chao KH, Wu M Y, Chen CD. 1999. The expression of killer cell inhibitory receptors on natural killer cells and activation status of CD4+ and CD8+ T cells in the decidua of normal and abnormal early pregnancies. Human  Immunol 60:791-797
   
   

10. Classen-Linke I, Alfer J, Hey S, Krusche CA, Krusche M, Beir HM. 1998. Marker molecules of human endometrial differentiation can be hormonally regulated under in-vitro conditions as in-vivo. Human Reprod Update 4:539-549
    
    

11. Demir R, Akkoyunlu G. 1999. Ultrastructural features of the human decidual glandular epithelium during early pregnancy. Placenta 20:A21
    
    

12. Demir R, Kaufmann P, Castellucci M, Erbengi T, Katovski A. 1989. Fetal vasculogenesis and angiogenesis in human placental villi. Acta anat. 136:190-203
    
    

13. Haregewoin A, Soman GR, Holm C, Finberg RW. 1989. Human T cells respond to mycobacterial heat shock protein. Nature 340:309-313
    
    

14. Hunt JS. 1994. Immunologically relevant cells in the uterus. Biol Reprod 59:461
    
    

15. Hunt JS, Robertson SA. 1996. Uterine macrophages and environmental programming for pregnancy success. J Reprod Immunol 32:1
    
    

16. Johnson PM, Christmas SE, Vince GS. 1999. Immunological aspects of implantation and implantation failure. Human Reprod 14:26-36
    
    

17. Kang YH, Mitchell C, Watson LP. 1988. Ultrastructural and functional effects of lipopolysaccaride and interlukin-2 on human NK cells. Scanning Microscope 2:1567-1571
    
    

18. Kaufman P, Castellucci M. 1997. Extravillous trophoblast in the human placenta. A Review. Trophoblast Research 10:21-65
    
    

19. Kauma S, Huff T, Krystal G, Ryan J, Takacs P, Turner T. 1996. The expression of SCF and its receptor, c-kit in human endometrium and placental tissues during pregnancy. J Clin Endocrinol Metab. 81:1261-1266
    
    

20. Kauma SW. 2000. Cytokines in implantation. J Reprod Fertil Suppl 55:31-42
    
    

21. King A, Balendranp N, Woodings P, Carter NP, Loke YW. 1991. CD3- leukocytes present in human uterus during early placentation: phenotypic and morphologic characterisation of the CD56++ population. Dev Immunol 1:169-174
    
    

22. King A, Burrows T, Verma S, Hiby S, Loke YW. 1998. Human uterine lymphocytes. Hum Reprod 4:480-485
    
    

23. King A, Gardner L, Loke YW. 1999. Co-stimulation of human decidual natural killer cells by interleukin-2 and stromal cells. Human Reprod 14:656-663
    
    

24. Koizumi H, Lui CC, Zheng LM, Joag SV, Bayne NK, Holoshitz J, Young JDE. 1991. Expression of perforin and serine esterases by human T cells. J Exp Med 173:499-503
    
    

25. Loke YW, King A. 1995. Human Implantation Cell Biology and Immunology. London: Cambridge University Press
    
    

26. Loke YW, King A. 2000. Immunology of Implantation. Bailliere's Clinical Obstet and Gynaecol 5:827-837
    
    

27. Miller I, Hunt JS. 1996. Sex steroid hormones and macrophage function. Life  Sci 59
    
    

28. Mincheva-Nilsoon L, Baranov V, Yeung MMW, Hammarström S, Hammarström M-L. 1994. Immunomorphological studies of human decidua-associated lymphoid cells in normal early pregnancy. J Immunology 152:2020-2032
    
    

29. O'Rahilly R. 1973. Developmental stages in human embryos, part A. Publ. 631. Washington: Carnegie Institute
    
    

30. Özenci CC, Korgun ET, Demir R. 2001. Immunohistochemical detection of CD45+, CD56+, and CD14+ cells in human decidua during early pregnancy. Early Preg 5:164-175
    
    

31. Pace D, Morrison L, Bulmer JN. 1989. Proliferative activity in endometrial stromal granulocytes throughout the menstrual cycle and early pregnancy. J of Clin Pathol 42:35-39
    
    

32. Parr EL YL, Parr MB, Young DE,. 1990. Granulated metrial gland cells of pregnant mouse uterus are natural killer-like cells contain perforin and serine esterases. J Immonol 145:2365-2371
    
    

33. Peel S, Stewen I BD. 1982. The morphology of granulated metrial gland cells in lethally irradiated, bone marrow reconstituted mice. J Anat 135:849-856
    
    

34. Saito S, Nishikawa K, Morii T, Narita N, Enomoto M, Ichijo M. 1992. Expression activation atigenss CD69, HLA-DR, interlukin-2receptor alpha and IL-'Rbeta on T cells of human decidua at an early stage of pregnancy. Immunol 75:710-716
    
    

35. Stemmer SM. 2000. Current progress in early pregnancy investigation. Early Pregnancy 4:214-218
    
    

36. Stewart-Akers AM, Krasnow JS, Deloila JA. 1997. Decidual leukocyte populations in ectopic pregnancies. Fertil Steril 68:1103-1107
    
    

37. Tabibzadeh S. 1990. Proliferative activity of lymphoid cells in human endometrium throughout the menstrual cycle. J Reprod Immunol 18:124-130
    
    

38. Tekelioglu-Uysal M, Edwards RG, Kisinisci HA. 1975. Ultrastructural relationships between decidua, trophoblast and lymphocytes at the beginning of human pregnancy. J Reprod Fertil 42:431-438
    
    

39. Umekage H, Saito S, Morikawa H. 1998. Enhancement by SCF of interleukin-2 (IL-2)-induced DNA synthesis in human decidual CD16-CD56 bright NK cells mediated by increased expression of the IL-2 receptor alpha chain. J Reprod Immunol 40:1-24
    
    

40. Vince GS, Johnson PM. 2000. Leokocyte populations and cytokine regulation in  human uteroplacental tissues. Biochem Soc Trans 28:191-195
    
    

41. Wynn RM. 1974. Ultrastructural development of human decidua. Am J Obstet Gynecol 118:652-658
    

A panoramic view of the DGLs clusters are demonstrated according to their developmental processes at early pregnancy. Immature DGL (im-DGL) with immature intermediate DGL (imi-DGL) and mature DGL (m-DGL) types are seen (a). A hypothesis; the developmental stages of DGLs in decidua during early pregnancy; indifferent decidual granuloblast (in-DGB) derived from stem cell, may develop in different directions of decidual leukocytes. From the progenitor cell stage and following, proliferation, differentiation, transformation and maturation processes will be taking place respectively (b). Original draw.

DGL cell types from the 18^th day of early pregnancy. Immature DGL (im-DGL) with light cytoplasm including definitive cell organelles, and euchromatic elliptical nucleus, mature dark DGLs (md-DGL) with electron lucent cytoplasm and abundant dense granules, and lobulated nucleus, mature clear DGLs (mc-DGL) with light cytoplasm including large a few electron dense granules and unlobular nucleus, and immature intermediate DGLs (imi-DGL) with rare granules are recognized. The accumulation of cellular fragments forming a compact barrier between the cell groups is seen (with arrow line, CF). Original magnification 12,500 x.

The imi-DGL exhibited a characteristic elongated shape with a construction of the cytoplasm and the nucleus, which is a feature of migrating cells (with broken lines). The migration direction is appearing between the DGL cell clusters and destroyed capillary (Cap). Secretory granules with different electron dense contents are seen in clear and dark m-DGLs from 22 day of pregnancy. Secretory granules present in cytoplasmic extensions of md-DGLs, which are connected to the cell by means of a narrow bridge of cytoplasm, are seen (with arrowheads). Homogenous substance accumulations are seen in intercellular spaces (with arrows). In the insert picture, high magnification of granules from a mc-DGL cell situated near the endothelium (E) is observing. Capillary area (Cap), red blood cells (BC). Original magnification 12.500x, ins. 40.000x.

184 DGLs evaluated on standardized magnification in 15 electronmicrographs, were classified and counted according to their ultrastructural features. Three different cell types, im-DGL (including imi-DGLs), md-DGL and mc-DGL were recognized. These cell types were examined in two groups (early and advanced periods), and percentage for the each cell type was calculated. There was a significant increase for mc-DGL in advanced group (a). The mean of the granule number for per md-DGL was significantly higher than that of mc-DGL cells in the early and advanced period (* p<0.05) (b).

a)

b)

Figure 5 a-c

A panoramic view of the measured dimensions of the cells . Descriptive schematisation of the cell and granule sizes that were used for quantification of area (Ar), perimeter (Pm), minimum diameter (Mnd), and maximum diameter (Mxd) (a). According to the parameters mentioned above (a), the quantification of the DGL dimensions and their granule sizes was counted. The evaluation of the cells (b) and granules (c) was carried out in early and advanced period. There is not a statistically significant difference between these periods for the parameters that evaluated for either of the cell dimensions (b) or the granule dimensions (c) (P> 0.05). Form factor (Ff) means the formation of the subjects and is regulated by the computerized analyser system, no difference significantly between two periods for this.

a) 

Fractionation of the DGLs from the 24^th day of early pregnancy. In most cases, the secretory granules have been discharged from the cell by means of an apocrine mechanism and a barrier has been formed with these cellular fractions. Through detachment and fragmentation of cytoplasmic pieces including secretory granules (with double arrows), accumulation occurs (with arrow line, CF) around the vascular area (Cap). It appears that they form a morphologic and immunologic barrier-like defence configuration. The md-DGL, mc-DGL, imi-DGL cells and also a moving cell configuration (with broken line) are seen. Free granules (with single arrows), and homogenous substance (with arrowheads) in intercellular spaces. Original magnification, 12.500x.

Typical mc-DGLs with characteristic features of light cytoplasm, eccentric placed unlobular nuclei and rare number of big dense secretory granules, and md-DGLs with characteristic features of eccentric nuclei, which showed a lobular condition with a broad rim of heterochromatin and many electron dense secretory granules in the cytoplasm of DGLs are seen. Cytoplasmic fragments of DGLs discharged from the cells, and accumulated by fusing to each other are filled with secretory granules (with arrowheads). An imi-DGL with a relation to other DGL cell types is seen. Interestingly, numerous dense substances (with double arrows) accumulated intercellular spaces, and cell fragments (with arrow lines, CF) are seen. Pregnancy day 26. Original magnification, 12.500x.

Nuclear lobe ratio of DGLs during early pregnancy. DGLs were classified in relation to nuclear shapes. The total number of the DGLs counted in fifteen TEM micrographs was 184. The percentages of the nuclei with one lobe, two lobes or three and more lobes were 31.52%, 40.75% and 27.72% respectively. These lobulation percent is seen in favour of multilobulatin. This is suggesting a sign for cellular maturation.

A cell type that has irregular shape appearing very dark cytoplasm including vacuoles, electron dense bodies, elongated cytoplasmic projections, and polymorph euchromatic nucleus, was observed in the destroyed capillary area (Cap). This cell type is giving an impression that is macrophage-like cell (ML). The destroyed capillary endothelium (E) has some cytoplasmic hills and foldings on the basal pole of cell lined by thin basement membrane (with arrowheads). Un-fragmented DGLs and destroyed capillary area were isolated by accumulation of the cellular fragments (with arrow line, CF) with or without secretory granules (with arrows). Stroma (ST). Pregnancy day 29. Original magnification, 12.500x.

Interestingly, this micrograph represents m-DGL cell line forming a cellular barrier separated by a limiting membrane (with double arrows) from sub-glandular area. These cells are arranged side by side as a defensive cellblock. A few regular decidual cells (DC) in stroma (ST) are observing between the basal lamina (with arrowheads) of glandular epithelium (GE) and the limiting membrane (with double arrows) of DGLs line. This finding suggest the glandular epithelium could be protected by itself against the DGLs toxicity, or against trophoblastic destruction. The insert picture shows the limiting membrane and the basal lamina. Pregnancy day 30. Original magnification, 12.500x, ins. 25.000x.

DGL cell surrounded by other decidual stromal cells (DC) associated with each others is seen. Various stages of secretory granule formation present in DGL cytoplasm from the 35^th day of early pregnancy. Three different kinds of secretory granules were seen. Limiting membrane was electron dense, and indirect contact with vesicles (1); moderately electron dense with some multivesicular structures (2); precursors of secretory granules including multivesicular contents (3). Nucleus (N), nucleolus (Nol). Original magnification, 12.500x.