Extraction of leucocytes from human decidua

Extraction of leucocytes from human decidua

Journal oflmmunological Methods, 104 (1987) 231-236 Elsevier 231 JIM 04534 Extraction of leucocytes from human decidua A comparison of dispersal te...

437KB Sizes 0 Downloads 79 Views

Journal oflmmunological Methods, 104 (1987) 231-236 Elsevier


JIM 04534

Extraction of leucocytes from human decidua A comparison of dispersal techniques Anne Ritson and Judith N. Bulmer Department of Pathology, Universityof Leeds, Leeds LS2 9JT, U.K. (Received 11 May 1987, accepted 23 June 1987)

Functional studies of human utero-placental tissues have been limited by poor characterisation of the morphology and antigenic phenotype of the cells under investigation. The present study documents the effect of dispersal methods on the viability and cellular composition of cell suspensions prepared from decidualised endometrium in early human pregnancy. First trimester decidua was subjected to both mechanical disaggregation and digestion with various enzyme combinations in an attempt to optimise the yield of infiltrating decidual leucocytes. Cell types were characterised with monoclonal antibodies using an alkaline phosphatase immunolabelling method. Mechanical disaggregation resulted in suspensions containing many large decidual cells and much cell debris but few leucocytic cells. Overall viability was low, although the viability of small leucocyte common antigen-bearing cells remained high. Enzymic digestion yielded cell suspensions rich in leucocytes with high viability and minimal contamination by other cell types. Collagenase produced a high yield of leucocytes with high viability and minimal disruption of surface antigens in contrast with pronase which caused extensive antigenic loss. The disaggregation method determines the yield of bone marrow derived cells in decidual cell suspensions and the extent of contamination by true decidual cells and epithelial cells. Key words: Decidua, human; Leukocyte, decidual; Leukocyte isolation


The survival of the fetus during normal pregnancy is an immunological paradox. The embryo inherits maternal and paternal antigens and fetally derived trophoblast cells in uteroplacental tissues are in intimate contact with maternal leucocytes throughout gestation (Pijnenborg et al., 1980). Immunohistological techniques have permitted characterisation of leucocyte common antigen-bearing (LCA ÷) cells within the decidualised endometrium lining the uterine cavity in pregnancy. Correspondence to: A. Ritson, Department of Pathology, University of Leeds, Leeds LS2 9JT, U.K.

Two main populations of LCA + cells have been described: decidual macrophages (leu M3 +, Dako-macrophage ÷, O K M I - , leu M I - , M a c l - ) (Bulmer and Johnson, 1984) are present in decidua throughout pregnancy, whereas lymphocytes (CD2 ÷, CD7 +, OKT10 +, CD3-, CD5-, CD4-, CD8-, CD25-) (Bulmer and Sunderland, 1984; Bulmer and Johnson, 1986; Bulmer et al., 1987) are detected mainly in first trimester decidua. The distribution of decidual lymphocytes mirrors that of endometrial granulocytes (Bulmer and Sunderland, 1983) which are morphologically similar to large granular lymphocytes (LGLs). However, decidual lymphocytes do not react with the NK cell markers leu-7 or leu-ll, although they label in-

0022-1759/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

232 tensely with NKH-1 (Ritson and Bulmer, 1987). The function of bone marrow derived cells in pregnancy decidua has been a source of interest. Antigen presenting ability has been detected in cells from murine decidua (Elcock and Searle, 1985) but the cell type responsible has not been well characterised and any in vivo role in pregnancy remains to be established. Small granulated lymphocytes in murine decidua in the first half of pregnancy have been shown to secrete a soluble factor which blocks lymphocyte responses to interleukin-2 (IL-2) (Clark et al., 1986). Suppressor activity has also been attributed to secretion of prostaglandin E 2 (PGE2) by decidual macrophages (Tawfik et al., 1986). Functional studies of human decidua have been more limited: the phenotypic characteristics of both decidual leucocyte populations provide limited clues as to their function. Decidual macrophages possess non-specific esterase and acid phosphatase activity (Bulmer and Johnson, 1984) which suggests a phagocytic role protecting the fetus by removal of immune, necrotic and protein debris. However, throughout pregnancy most decidual macrophages express HLA DR, DP and DQ antigens (Bulmer et al., unpublished results) which may suggest an antigen presenting function and there have been reports of antigen presenting capability in human decidua, although the cells responsible have not been characterised (Oksenberg et al., 1986). Culture supernatants from explants of early human pregnancy decidua have been shown to suppress lymphocyte responses to mitogens (Golander et al., 1981). Both 'large' and 'small' suppressor cells have been detected in human decidua (Daya et al., 1985): these cells have not been thoroughly characterised but it has been suggested that they may be analogous to granulated lymphoid suppressor cells in murine decidua. Most recent functional studies of mouse and human decidua have relied initially on the preparation of cell suspensions. Various enzymatic and mechanical disaggregation methods have been employed, although the possibility that they may yield different cell populations has received scant attention. Characterisation of morphology and antigenic phenotype of the various cells in functional studies has been a major problem which has limited correlation of results from different laboratories.

In the present study, several methods of dispersal have been investigated in order to determine the optimal technique for separation of leucocytic cells and compare populations obtained with enzymatic versus mechanical disaggregation.

Materials and methods

Preparation of tissues Decidua was obtained from elective termination of 12 normal pregnancies at 8-13 weeks gestation. Fragments of decidua were identified macroscopically by their greyish-white, solid appearance, washed free from blood in RPMI 1640 (Gibco, Paisley, U.K.) containing penicillin (100 I U / m l ) and streptomycin (100 /~g/ml), drained and weighed. Tissue was then divided into equal parts, each being subjected to one of the following treatments: (1) Teasing, where the tissue was gently teased apart with mounted needles. (2) Sieving, where the tissue was minced with opposing scalpels and pressed through a 100 /~m sieve. (3) Enzyme digestion using the five different solutions listed in Table I. The tissue was minced finely with opposing scalpels and incubated at room temperature with continuous agitation in the appropriate enzyme solution (2 vols. solution:l vol. tissue). After 30 rain the cell suspension was filtered through sterile gauze, washed by centrifugTABLE I ENZYME SOLUTIONS FOR TISSUE DISAGGREGATION Solution 1

Content 0.1% Dispase

Supplier BCL, Lewes,East Sussex


0.1% Type II collagenase 0.1% DNase

SigmaChemicalCo., Poole, Dorset


0.1% Type IV collagenase 0.1% DNase

Sigma ChemicalCo.


0.1% Pronase



0.1% Trypsin 0.1% DNase


233 ing at 400 x g, resuspending the cell pellet in fresh RPMI 1640, repeating the centrifugation and finally resuspending the cell pellet in complete medium (RPMI 1640, 10% heat-inactivated fetal calf serum (FCS), penicillin 100 I U / m l , streptomycin 100 # g / m l and glutamine 20 mM final concentration). Fresh enzyme solution was then added to the tissue fragments and the digest continued for a further 30 rain. It was hoped that minimising enzyme exposure would limit cell surface damage. The serial digest was continued until the tissue became gelatinous and difficult to filter. The cell suspension fractions were pooled after the washing step. The cells obtained with mechanical and enzymatic techniques were pelleted at 400 x g and incubated for 5 rain at 3 7 ° C in 0.84% ammonium chloride to lyse erythrocytes. Total cell count and viability were then determined with trypan blue. The cells were finally resuspended in complete medium and cultured overnight at 3 7 ° C in a gassed, humidified incubator, the aim of this step being to remove adherent cells from the suspension and enrich for leucocytes in the non-adherent fraction. Cell smears were then prepared by standard techniques immediately after red cell lysis and also after overnight culture. The smears were air dried overnight, fixed in acetone for 10 min at room temperature and stored at - 2 0 ° C wrapped in foil until use.

Monoclonal antibodies Four monoclonal antibodies (mAbs) were used to investigate the cell populations present in the

smears and all were titrated to determine the optimal dilution. Dilutions, specificities and sources are listed in Table II.

Immunocytochemistry Smears were labelled by an indirect alkaline phosphatase technique. The smears were rehydrated for 2 min in 0.05 M Tris-buffered 0.15 M saline pH 7.6 (TBS) and incubated for 1 h at room temperature in a humid chamber with primary mAbs appropriately diluted in TBS. The smears were then washed twice briefly in TBS and incubated for 45 min with an affinity-purified, alkaline phosphatase-conjugated rabbit anti-mouse immunoglobulin diluted 1/2000 in TBS (Northeast Biomedical Labs., Denham, U.K.). After two further brief TBS washes the reaction was developed with 0.05 M Tris pH 8.2 containing 1.0 m g / m l Fast Red TR Salt (Sigma Chemical Co., Poole, U.K.) and 0.2 m g / m l naphthol AS MX phosphate (Sigma Chemical Co.). Slides were lightly counterstained with Mayer's haematoxylin and mounted in glycerin jelly. Negative controls were performed for all decidual smears. Smears of peripheral blood lymphocytes, prepared by standard techniques, were used as positive controls. Unlabelled smears of decidual cell suspensions were stained with standard haematoxylin and eosin ( H & E ) and May Grtinwald Giemsa to permit assessment of morphology. The percentage of positive cells was determined by counting positive and negative cells over four fields and taking the mean of these figures. A minimum of 500 cells was counted per smear.


Specificity Leucocytecommon antigen (CD45)

Source Dakopatts, Denmark

Dilution 1/100 CS a


E rosette receptor (CD2)


1/50 CS


Tissue macrophages


1/100 CS

CAM 5.2

Low molecularweight cytokeratins(epithelialcells)

Becton Dickinson, CA

1/250 CS

a Culture supernatant.




Cell yield and viability Table III shows the cell yield per gram of decidua for each of the different disaggregation methods. Sieving gave the highest cell yield but this was mainly due to heavy contamination of the small decidual macrophages and lymphocytes by very large, round, vacuolated cells, presumably true decidual (decidualised stromal) cells. The viability, as assessed by trypan blue exclusion, of these large cells was very low while that of the smaller leucocytic cells remained high (86%). Teasing resulted in fewer large dead cells but they still formed a significant percentage of cells in suspension and the overall cell yield was low. The viability of small cells in the teased preparation was highest (93%). All enzymes gave comparable cell yield and viability (75-85%). In contrast with mechanically dispersed preparations these cell suspensions contained very few large decidual cells. H & E and Giemsa staining of smears showed that the smaller cells in the suspensions had the characteristic morphology of lymphocytes. All enzymatically prepared suspensions showed similar composition morphologically.

Dispersal method

Total cell yield per gram of tissue

Viability of leucocytes

% Dispase Collagenase II/DNase Collagenase IV/DNase Pronase Trypsin/DNase Sieving Teasing

7.9 x 106 14.6 x 106 15.0X 106 16.0 x 106 9.5 X 106 30.0 x 106 2.4 X 106

75 75 76 85 85 86 93

population, the relative proportions of T cells and macrophages remained similar except for tissues digested with pronase. Pronase-digested suspensions contained a higher proportion of cells labelling with Dako-macrophage (44%) and CAM 5.2 (21%) while fewer cells were LCA + (42%) and CD2 + (27%). Trypsin was also found to give variable immunocytochemical results in some tissues; three of the twelve specimens contained only small numbers of CD2 + cells (16, 24 and 26%). Trypsin digestion also resulted in a higher proportion of epithelial cells in the suspensions (20%).

Irnmunocytochernistry Table IV shows the percentage of cells in each suspension labelled with each of the mAbs listed in Table II. All the enzymatic methods gave high proportions of LCA+cells, while mechanically dispersed suspensions contained fewer leucocytes and a higher proportion of epithelial cells as detected by labelling with CAM 5.2. Within the LCA ÷


Bone marrow-derived cells present in the decidualised endometrium may play an important role in the survival of the fetus during pregnancy. There have been attempts to assign particular immunological functions to various cells within


Dispase Type II collagenase/DNase Type IV collagenase/DNase Pronase Trypsin/DNase Sieving Teasing

% Cells labelled DAKO-LC

DAKO T l l

DAKO macrophage

76 76 77 42 73 25 24

50 56 49 27 44 13 26

16 16 20 44 16 5 3

CAM 5.2 (epithelial cells) 5 11 9 21 20 54 52

235 human decidual tissues but documentation of the morphology and cell surface phenotype has been poor. The starting point for most of these investigations has been the preparation of a single cell suspension from decidual tissue. The present study shows that the dispersal method has a dramatic effect on the proportion of each cell type present in the resulting suspension. This possibility appears to have been largely ignored by other investigators studying the activities of either the 'true' decidualised stromal cells or the infiltrating bone marrow-derived cells in decidua. Lala et al. (1986) have reported the presence of lympho-myeloid markers on 'decidual' cells: however, by employing a collagenase digestion technique these authors have already preferentially selected a cell suspension rich in cells of bone marrow origin which in no way reflects the relative proportions of cell types within the intact tissue. Gambel et al. (1985) also reported disproportionately high numbers of bone marrow-derived cells in cell suspensions compared with intact tissue after collagenase digestion of murine decidua. Mild enzymatic digestion techniques selectively disaggregate cells least tightly bound and hence migratory leucocytes in decidua are likely to dissociate more easily (Gambel et al., 1985). Thus, mechanical dispersal techniques appear to be more appropriate for the study of stromaltype decidual cells or endometrial epithelial cells with only 25% LCA + cells present. This suspension is more likely to reflect the true proportion of cell types within the tissue. Similarly, for the study of LCA + cells, enzymatic digestion would appear to be the method of choice, the suspensions containing over 75% LCA + cells. However, enzymes may not preserve surface phenotype to the same extent: pronase had a profound effect on certain cell surface antigens. The low proportion of LCA + cells (42%) present in the pronase-digested suspension is likely to reflect surface antigen damage rather than altered composition since the morphology of smears stained with H&E and Giemsa was comparable with that of other enzymes. The increased number of macrophages in this preparation may indicate the inability of these cells to adhere to plastic after pronase treatment. Trypsin also produced inconsistent immunocytochemical

results predominantly affecting the CD2 antigen: levels of CD2 + cells as low as 16-26% were found in three of the 12 specimens. The variation in the findings for trypsin digestion could also reflect variation between different enzyme batches. Dispase and the collagenases all produced consistent immunocytochemical results but an increased cell yield was obtained with collagenase, which is an obvious advantage. Initial studies (results not shown) revealed that collagenase can also cause cell damage when used for prolonged periods and inconsistent immunolabelling was apparent after only 2.5 h of digestion. The use of the fractional digestion method minimised any surface damage and consequent antigenic loss and maintained the cells in apparently good condition. However, cell function may also be affected by enzyme digestion. Clark et al. (1986) reported loss of suppressor activity from suspensions prepared from murine decidua by digestion with certain types of collagenase and dispase compared with mechanical disaggregation. However, the precise details of technique were not included and it is possible that prolonged digestion could have caused cell damage. In summary, the proportions of various cell types obtained in decidual cell suspensions after different dispersal methods have been documented. It is essential to characterise these suspensions which form the starting point for subsequent purification and functional assays. Better documentation of cell types in decidual cell suspensions would allow correlation of results from different laboratories and hence produce a more coherent picture of decidual leucocytic function in normal human pregnancy.

Acknowledgements We wish to thank the Medical and Operating Theatre staff of the General Infirmary at Leeds (Clarendon Wing) and St. James's University Hospital, Leeds, for their assistance in collecting specimens. This work was supported by a grant from the Yorkshire Regional Health Authority.


References Bulmer, J.N. and Jotmson, P.M. (1984) Macrophage populations in the human placenta and amniochorion. Clin. Exp. Immunol. 57, 393. Bulmer, J.N. and Johnson, P.M. (1986) The T-lymphocyte population in first trimester human decidua does not express the interleukin-2 receptor. Immunology 58, 685. Bulmer, J.N. and Sunderland, C.A. (1983) Bone marrow origin of endometrial granulocytes in the early human placental bed. J. Reprod. Immunol. 5, 383. Bulmer, J.N. and Sunderland, C.A. (1984) Immunohistological identification of lymphoid populations in the early human placental bed. Immunology 52, 349. Bulmer, J.N., Johnson, P.M. and Bulmer, D. (1987) Leucocyte populations in human decidua and endometrium. In: T.J. Gill and T.G. Wegmann (Eds.), Immunoregulation and Fetal Survival. Oxford University Press, Oxford, p. 111. Clark, D.A., Slapsys, R., Chaput, A., Walker, C., Brierley, J., Daya, S. and Rosenthal, K.L. (1986a) lmmunoregulatory molecules of trophoblast and decidual suppressor cell antigen at the maternofetal interface. Am. J. Reprod. Immunol. Microbiol. 10, 100. Clark, D.A., Brierley, J., Slapsys, R., Daya, S., Damji, N., Chaput, A. and Rosenthal, K. (1986b) Trophoblast-dependent and trophoblast-independent suppressor cells of maternal origin in murine and human decidua. In: D.A. Clark and B.A. Croy (Eds.), Reproductive Immunology 1986. Elsevier, Amsterdam, p. 219. Daya, S., Clark, D.A., Devfin, C. and Jarrell, J. (1985) Pre-

liminary characterisation of two types of suppressor cell in the human uterus. Fertil. Steril. 44, 778. Elcock, J.M. and Searle, R.F. (1985) Antigen presenting capacity of mouse decidual tissue and placenta. Am. J. Reprod. Immunol. Microbiol. 7, 99. Gambel, P., Rossant, J., Hunziker, R.D. and Wegrnann, T.G. (1985) Origin of decidual cells in murine pregnancy and pseudopregnancy. Transplantation 39, 443. Golander, G., Zakuth, V., Schechter, Y. and Spirer, Z. (1981) Suppression of lymphocyte reactivity in vitro by a soluble factor secreted by explants of human decidua. Eur. J. Immunol. 11,849. Lala, P.K., Parhar, R.S., Kearns, M., Johnson, S. and Scodras, J.M. (1986) Immunological aspects of the decidual response. In: D.A. Clark and B.A. Croy (Eds.), Reproductive Immunology 1986. Elsevier, Amsterdam, p. 190. Oksenberg, J.R., Mor-Josef, S., Persitz, E., Schenker, Y., Moses, E. and Brantbar, C. (1986) Antigen presenting cells in human decidual tissue. Am. J. Reprod. Immunol. Microbiol. 11, 82. Pijnenborg, R., Dixon, G., Robertson, W.B. and Brosens, I. (1980) Trophoblastic invasion of human decidua from 8-18 weeks of pregnancy. Placenta 1, 3. Ritson, A. and Bulmer, J.N. (1987) Endometrial granulocytes in human decidua react with a natural killer (NK) cell marker, NKH1. Immunology, in press. Tawfik, O.W., Hunt, J.S., Wood, G.W. (1986) A major soluble suppressor factor produced by uterine cells during murine pregnancy is prostaglandin E 2 (PGE2). J. Reprod. Immunol. Suppl. 126.